|Title||Sensitivity of grassland ecosystemsacross the Great Plains to present and future variability|
|Year of Publication||2008|
|Number of Pages||1 -161|
|University||Colorado State University|
|City||Fort Collins, CO|
|Thesis Type||Ph.D. Thesis|
Patterns and controls of aboveground net primary productivity (ANPP) have been of long-standing interest to ecologists because ANPP integrates key aspects of ecosystem structure and function through time. In many terrestrial biomes, water availability is a primary constraint to ANPP, and it is an ecosystem driver that will be affected by future climate change. To understand the sensitivity of temperate grasslands to inter- and intra-annual variability in precipitation, I analyzed long-term ANPP data, conducted a multi-site experimental manipulation in which the number of growing season rainfall events was varied, and simulated the effects of altered rainfall regimes using a terrestrial ecosystem model (DAYCENT). I conducted this research within the Great Plains of North America—a region characterized by a strong west-east precipitation-productivity gradient and three distinct grassland types—the semi-arid shortgrass, the mixed-grass prairie, and the mesic tallgrass prairie. My results demonstrate that temperate grasslands are indeed sensitive to both inter- and intra-variability in precipitation, but the ANPP response is contingent upon ecosystem structure and typical soil water levels. Additionally, both management strategies and topographic location may interact with precipitation to enhance or diminish coherence in the ANPP response. At the dry end of the gradient (semi-arid steppe), fewer, but larger rain events led to increased periods of above-average soil water content, reduced plant water stress and increased ANPP. The opposite response was observed at the mesic end of the gradient (tallgrass prairie), where longer dry intervals between large events led to extended periods of below-average soil water content, increased plant water stress, and reduced ANPP. Mixed grass prairie was intermediate along the gradient, characterized by the greatest plant species richness, and the most sensitive to within-season variability in rainfall. Comparison of these experimental data to model simulations revealed key differences in soil water dynamics and ANPP patterns, suggesting that more experimental data is needed to parameterize biological and physical processes that drive model simulations. In conclusion, these results highlight the difficulties in extending inference from single site experiments to whole ecosystems or biomes and demonstrate the complexity inherent in predicting how terrestrial ecosystems will respond to novel climate conditions.