Book Description

In addition to causing an increase in mean temperatures, climate change is also altering precipitation regimes across the globe. General circulation models project both latitude-dependent changes in precipitation mean and increases in precipitation variability. These changes in water availability will impact terrestrial primary productivity, the fixation of carbon dioxide into organic matter by plants. In my thesis, I addressed the following three questions: 1.) What will be the relative effect of changes in the mean and standard deviation of annual precipitation on mean annual primary production? 2.) Which ecosystems will be the most sensitive to changes in precipitation? 3.) Will increases in production variability be disproportionately greater than increases in precipitation variability? I gathered 58 time series of annual precipitation and aboveground net primary production (ANPP) from long-term ecological study sites across the globe. I quantified the sensitivity of ANPP at each site to changes in precipitation mean and variance. My results indicated that mean ANPP is about 40 times more sensitive to changes in precipitation mean than to changes in precipitation variance. I showed that semi-arid ecosystems such as shortgrass steppe in Colorado or typical steppe in Inner Mongolia may be the most sensitive to changes in precipitation mean. At these sites and several others, a 1% change in mean precipitation may result in a change in ANPP that is greater than 1%. To address how increases in interannual precipitation variability will impact the variability of ANPP, I perturbed the variability of observed precipitation time series and evaluated the impact of this perturbation on predicted ANPP variability. I found that different assumptions about the precipitation-ANPP relationship had different implications for how increases in precipitation variability will impact ANPP variability. Increases in ANPP variability were always directly proportional to increases in precipitation variability when ANPP was modeled as a simple linear or a lagged function of precipitation. However, when ANPP was modeled as a nonlinear, saturating function of precipitation, increases in ANPP variability were disproportionately low compared to increases in precipitation variability during wet years but disproportionately high during dry years. My thesis addresses an existing research gap regarding the long-term impact of increases in interannual precipitation variability on key ecosystem functioning. I showed that increases in precipitation variability will have negligible impacts on ANPP mean and have disproportionately large impacts on ANPP variability only when ANPP is a concave down, nonlinear function of precipitation. My work also demonstrates the importance of the precipitation-ANPP relationship in determining the magnitude of impacts to ANPP caused by changes in precipitation. Finally, my thesis highlights the potential for considerable changes in ANPP variability due to increases in precipitation variability.