Grasslands in a changing climate: Demographic responses to natural and experimental variability in precipitation in perennial grasslands confronted with climate change

Background/Question/Methods

Ecosystems across the globe are experiencing dramatic effects of climate change, including shifts towards more extreme conditions, in terms of both temperature and water availability. For many systems, including drylands of the southwestern US, such changes intensify temporal variability in water availability, leading to increased frequency of extreme events, from drought to deluge. Understanding the implications of this change for community composition and population dynamics requires investigation of the demographic responses of local species to water availability. Further, it is critical to quantify responses to conditions outside of historical regimes and through time, which requires multi-year experiments and observation. Here we use a rare combination of a multi-year rainfall manipulation experiment and demographic monitoring in five grassland communities across an elevation gradient in northern Arizona to ask: 1) how does water availability influence vital rates and population growth rates for dominant perennial grasses across an elevation gradient? 2) Which species or sites are more vulnerable to reduced water availability? 3) What are the implications for population persistence and community composition? Further, we explore whether functional and physiological traits provided additional insight into predicting population responses.

Results/Conclusions

Experimental treatments, which excluded or supplemented natural rainfall, occurred against a backdrop of variable conditions, which included extreme drought bookended by heavy rainfall in the first and last years of the study period. Ambient drought reduced growth and increased the rate of shrinkage of plants to smaller sizes. Further, establishment of new individuals was low across the entire study period. These patterns were enhanced under experimental rainfall exclusion (drought treatments), but attenuated with rainfall addition, though these patterns depended on species and site. Accordingly, results indicate low population growth rates over the study period, with many values below 1, indicating population declines for many species and sites, particularly in response to natural and experimental drought. Growth rates generally increased with rainfall supplementation, though in many cases this was not sufficient to recover population growth rates. Future analyses will explore how functional and physiological traits mediate demographic responses. Together, our results indicate that natural conditions are already bringing high variability and extremes, from drought to deluge, and indicate that demographic responses to further reduction in water availability from drought is unlikely to be counterbalanced by extreme rainfall events.