River regulation and water use are among the foremost problems faced by aquatic resource managers in North America. Broader-scale influences, such as urbanization and climate change, also can affect water quantity, alter the seasonality of streamflows, and directly influence animal population dynamic processes, such as dispersal and reproduction. Thus, the development of effective ecological flow policies requires consideration of temporal and spatial dynamics of stream-dwelling biota. To evaluate the influence of hydrologic alteration on fishes, we developed a dynamic, spatially-explicit, landscape-scale model to predict the changes in the structure and persistence of fish communities in the upper Flint River basin, Georgia. We simulated the response of fish communities to varying levels of water development using historic streamflow data and used the simulated results to develop relationships between the degree of hydrologic alteration and changes in fish community structure. The nature of these relationships, however, was influenced by model assumptions regarding fish population dynamics. We intend our approach to be used in an adaptive framework, within which assumptions are tested and model components iteratively improved to increase our understanding of physical and ecological processes, thereby improving water resource management.