Local environmental factors (e.g., soil chemistry, light) interact with the characteristics of plants to influence the abundance of species in a community. Dispersal can also influence species composition, but may be much more stochastic in its effect if it is a relatively uncommon event. The relative importance of environmental factors and dispersal, and the way that their roles may change at different spatial scales, is not well known. Plant communities on islands in Lake Michigan will be used as a model system to test the relative importance of environmental factors versus dispersal in determining plant community composition at a variety of spatial scales. Individuals of three clonal 'phytometer' species will be planted into the native shoreline environment and monitored for two growing seasons. The degree to which plant performance varies with distance will indicate the plant?s eye view of habitat heterogeneity in the environment. Species-area relationships will be analyzed for breaks in the slope, an indication of changes in plant-relevant (instead of human-defined) habitat factors. To complement field studies, a simulated plant community with known degrees of habitat sorting and dispersal limitation will be sampled using a variety of designs, and analyzed using several methods currently debated in the literature. The project will establish methods for similar approaches in other ecological communities to facilitate comparison and generalization of this important ecological question.
The broader impacts include research experiences for about ten students at an undergraduate institution. The scientific results will be available via electronic appendices and posted on-line for comparative analyses, as well as educational uses. A plant community simulation program will be developed and made freely available to ecologists for use in teaching and research.
