BIG DATA IS FINALLY ENOUGH TO EVALUATE ECOSYSTEM VULNERABILITY IN THE ANTHROPOCENE (Invited Presentation)

The concept of resilience in the ecological literature has been developed and tested with diverse approaches and in diverse environments around the world. This concept has received such wide attention because it describes fundamental and valuable properties of complex systems: namely, the resistance to and recovery from perturbation. Now that extensive surface and subsurface data (e.g., hydrochemistry and ecosystem structure) are available, describing and predicting ecological resilience is also now possible at unprecedented spatiotemporal scales. Here, we attempt to combine observations and ideas from across the natural sciences to form a conceptual model that describes predictors and contingency of resistance and recovery. Specifically, we use emerging data science tools and large data sets (e.g., hydrologic, carbon, and nutrient data) from Critical Zone Observatories and Long-Term Ecological Research sites to describe a variety of relationships between resistance and recovery at the ecosystem scale. We show how these relationships are largely driven by emergent properties and collinearities from coevolution of ecosystems, climate, and human activity. We use this model to test the hypothesis that univariate estimates of ecosystem resilience will consistently fail to predict actual response to perturbation because of the multiple pressures associated with the Anthropocene. This multidimensional resilience hypothesis conceptualizes linkages among environmental pressures and interactions between ecosystem responses, while also giving Earth scientists and resource managers a path forward in the face of human pressures worldwide. We conclude that increased collaboration between data scientists and ecosystem scientists, all of whom work from regional to continental scales, could continue to trigger substantial advancements in the longstanding quest to link site-level mechanisms with large-scale patterns of ecological resilience.