Freshwater ecosystems are experiencing considerable degradation as increasing human demands affect both water quality and quantity. The challenge of maintaining the ecological integrity of rivers requires tools for evaluating the effects of these changes. Metacommunity theory seeks to explain species’ distributions at multiple scales and can serve as a valuable framework for addressing trends in riverine biodiversity. Combining several elements of this framework, we assume that patterns of abundance in the linked local communities of a river network emerge from interactions between the configuration and accessibility of habitat, as determined by flow regime and fluvial geomorphology. We have developed a simple, generalized simulation model to test predictions concerning the potentially complex interactions between flow-mediated disturbance and disruption of dispersal pathways. The model represents local communities as nodes in a spatial graph and records changes in their composition as species experience mortality, reproduce, disperse and establish.
Results/Conclusions

Given our exemplary networks and regional species pool sizes, we present results that indicate how increasing network path lengths, removing edges to simulate lost dispersal opportunities, and altering the sequence of mortality-inducing flows can have countervailing impacts on between-community similarity over time. In particular, we find that decreased connectivity on directed graphs produces more divergent communities relative to undirected networks, but that increasing a parameter for density-independent control of establishment success can reduce differences between local communities. Finally, we discuss the implications of our theoretical results for the management of real rivers currently facing biotic homogenization related to flow modification and habitat loss.