Responses to climate change in hot desert ecosystems: Connecting local to global scales

Background/Question/Methods

It is increasingly recognized that global connectivity among the land, atmosphere, and oceans is driving dynamics at finer scales, and pushing thresholds of change in ecological systems. The consequences of this connectivity in resources, propagules, and information to the interplay between drivers and responses across scales can result in dynamics that are not easily predicted based on local drivers. Because both drivers and responses are expected to change under future climate scenarios, understanding and prediction require an explicit examination of these cross-scale interactions. Three major classes of connectivity events link local ecological dynamics with broader scale drivers. These events differ in initial driver characteristics (intensity, spatial extent, duration), rate of change in response through space, and the relative importance of driver characteristics versus ecological properties and land cover through time. As a result, the impacts of drivers on ecological systems can be synchronized, attenuated or amplified through time and space. Here, I examine how these three classes can interact during the course of climatic events to result in surprising ecological responses, and a potential state change reversal in the largest desert of North America. 

Results/Conclusions

Unusually large amounts of summer rainfall in 2008 in the American Southwest were related to a hurricane that began as a thunderstorm in West Africa, and propagated with increasing intensity across the Atlantic Ocean. Interactions with the land surface of North America attenuated hurricane intensity spatially, but it was still sufficient to result in an extreme, infrequent rain event throughout the Southwest.  Ecologically, this rain pulse created a synchronous, broad-scale recruitment of perennial grasses in degraded shrublands that had undergone desertification over the past century. Successive years of average and below-average rainfall amplified the growth and spatial propagation of grasses via plant-soil feedbacks at the scale of individual plants. Continued interactions across scales are expected to push the system towards a new state codominated by grasses and shrubs. Because similar wet years in the 1990s did not generate grass recovery, ecological properties can constrain the potential of this system to respond to extreme climatic events. These complex cross-scale dynamics are not unique to this system, and suggest that effective predictions of future dynamics will require a consideration of changing drivers and landuse across scales that include both contiguous and non-adjacent locations.