Understanding how species coevolve within complex ecological communities is one of the greatest technical challenges in evolutionary biology. Mathematical models of coevolution with even modest amounts of biological reality usually defy transparent, general analyses of coevolution. This project will develop innovative mathematical tools for analyzing coevolution in complex communities, including both evolutionary and population dynamics. The two main analytic approaches will make use of different mathematical approximations available when the strength of selection is relatively weak or strong. The scope and accuracy of these new methods will be verified using computer simulations. The new mathematical tools will also be used to address two important questions in community coevolution. First, does the number of potential hosts influence transitions between generalist and specialist parasites? Second, does coevolution lead to different ecological network structure in mutualistic vs. antagonistic communities? In each case, theoretical predictions will be tested in the laboratories of empirical collaborators. Ongoing studies of host use in bacteriophage will evaluate predictions for specialist-generalist transitions in parasites; data from well-studied plant-insect interactions will test predictions for the coevolution of community network structure.
The proposed work will contribute to the continued integration of analytic methods into the biological sciences. Results will be widely disseminated through publication, scientific presentations, and software packages developed during the project. Training will introduce at least three post-doctoral researchers, four graduate students, and five undergraduates to cutting-edge analytical and computational techniques. The research collaboration also reinforces existing interactions between biology and mathematics departments in geographically proximate Washington State University and the University of Idaho. In addition, the project supports a unique international collaboration, and strengthens emerging regional ties between mathematical biologists at institutions in Washington, Idaho, and British Columbia.
Most species exist within complex ecological communities where they interact with many other species that are mutualists, competitors, prey, hosts, predators, or parasites. Predicting how evolution influences the structure, function, and dynamics of these complex ecological communities has been a long-standing challenge in ecology. What makes prediction so difficult in these cases is that, at a minimum, evolutionary changes within each species of the community must be described and tracked simultaneously. Although this can be feasible for small communities (e.g., two or three species), it becomes a formidable mathematical and computational challenge as the number of interacting species increases. This project developed new mathematical and computational tools that allow evolution among many interacting species to be modeled and quantified in a way that reveals its role in shaping ecological communities. In addition to developing a flexible set of tools that allows evolution to be studied in a mathematically rigorous way within complex communities, work on this project addressed several specific topics of applied biological importance. To highlight one example led by Dr. Emily Jones, a postdoctoral researcher supported by this grant, we investigated a long-standing prediction---dating back to Darwin---that biological invasions are more likely to be initiated by species that are more distant evolutionary relatives to those species currently within the native community. Using the mathematical and computational tools developed in the course of this project, we formalized and analyzed these earlier verbal predictions. Our results show that, although evolutionary relationships among species do have power to predict whether a species will invade a native community, whether a distantly related species is more likely or less likely to invade depends on the ecological details of interactions among the original resident species and any potential invading species. This work will help conservation biologists and wildlife managers to develop more effective threat assessments posed by potentially invasive species. Another important outcome of this project was the training of post-doctoral researchers, graduate students, and undergraduate students to formulate mathematical models and design computer simulations that allow us to better understand biological processes and logically project their consequences. Among those individuals trained on this grant, the postdoctoral researchers moved onto academic research positions, the graduate students took post-doctoral positions and jobs at information technology companies (GFI Informatique, Reasoning Mind Inc.), one undergraduate student went on to medical school, and two other undergraduate participants went on to graduate school.
