How biological communities respond to environmental change is determined both by the tolerances of individual species and by interactions among species, such as competition, predation, and mutualism. However, little progress has been made in understanding how biotic interactions influence species distributions or in incorporating their effects into models of species assemblages at regional to continental scales. This project will develop and test new methods that consider species interactions by accounting for patterns of species co-occurrence to predict how species and biological communities respond to changes in climate. Observed changes in the distributions of plants and mammals in eastern North America during the last 21,000 years will be combined with independent simulations of past climates to examine how the strength and direction of species interactions vary across broad regions and in response to changes in climate. These data and analyses will also be used to test whether ecological models that consider species co-occurrences have a better predictive ability than existing approaches that model species individually and to identify possible limits to our ability to predict how biological communities may respond to future changes in climate. Results will help address the grand challenge of understanding how changes in climate alter natural systems and their associated ecosystem services.
Beyond providing important insights into how species interactions influence the response of biological communities to changes in climate and our ability to predict these responses, this project will also provide cross-disciplinary training of undergraduate and graduate students and early-career scientists in global change ecology, paleoecology, and statistical modeling. Minority and first-generation undergraduate students will be recruited through campus programs at Frostburg State University and the University of California-Merced. Public outreach to rural communities will be conducted through public lectures and engagements at open house events in an underserved region. In support of open, reproducible science, all publications, datasets, and computer code will be made publicly available.
