Increased precipitation variability causes boom-bust cycles in invasive annual grasses, while increasingly stressed perennials persist

Lauren M. Porensky and David L. Hoover, Rangeland Resources and Systems Research Unit, USDA-ARS, Fort Collins, CO, Troy W. Ocheltree, Forest and Rangeland Stewardship, Colorado State University, Fort Collins, CO

Background/Question/Methods:
Climate change is predicted to increase interannual precipitation variability in many regions of the world, including the semi-arid western edge of the North American Great Plains. In this region, heightened variability is predicted to be coupled with increased winter precipitation. The consequences of these shifts in precipitation variability and seasonality for plant communities and ecosystem function remain uncertain. Ecotones can be especially sensitive to shifts in abiotic conditions, since species in these regions are coexisting near the edges of their respective ranges. We investigated the separate and combined effects of shifting precipitation variability and seasonality on plants in a shrubland-grassland ecotone. We experimentally enhanced interannual precipitation variability by implementing precipitation reductions (-50% of ambient) or precipitation additions (+50%) in the first year of the experiment, and then reversing these treatments in the subsequent year. Precipitation variability treatments were crossed with added winter moisture treatments to simulate predicted shifts in precipitation seasonality, in which we increased winter moisture by 50% at the end of the dormant season. We predicted that increases in interannual variability and winter precipitation should favor deep-rooted shrubs at the expense of perennial grasses in this ecosystem due to differences in rooting profiles and soil moisture availability.
Results/Conclusions:
Over the first three years of the project, precipitation manipulation treatments had strong impacts on invasive annual grasses, but effects were lagged by one year. When compared to plots receiving ambient precipitation, these annuals were more than twice as abundant in plots that had been watered in both the winter and summer of the previous year, despite the fact that these same plots were receiving reduced precipitation during the current growing season. Leaf water potential measurements on perennial grasses and shrubs revealed that growing season droughts created more water stress in grasses than shrubs, as predicted. However, despite added stress, perennial grasses exposed to increased variability treatments continued to maintain their aboveground production, relative to ambient plots, during the first three years of the experiment. These results indicate that perennial grasses and shrubs in this grassland-shrubland ecosystem are resistant to highly variable weather. On the other hand, annual grasses may be highly sensitive to increased precipitation variability, with lagged effects. Our findings suggest that in semi-arid ecosystems where plants are already well-adapted to variable weather, a few years of enhanced variability may not have large effects on native perennial plants, but could have dramatic impacts on undesirable invasive species.