Date of Graduation


Document Type


Degree Type



Davis College of Agriculture, Natural Resources and Design


Wildlife and Fisheries Resources

Committee Chair

J. Todd Petty.


Chapter one is a direct empirical evaluation of metacommunity theory using stream fishes in a heavily mined and least impaired region of the Monongahela River basin. Metacommunity ecology, a new multi-scale paradigm, has provided important insights into mechanisms controlling regional patterns of assemblage structure. Metacommunity ecology offers four competing hypothetical models---neutral, patch dynamic, species sorting, and mass effect---each differing slightly in its underlying mechanistic framework.;Our results overwhelmingly supported the mass effect metacommunity model, which suggests that regional patterns of diversity are maintained by local habitat heterogeneity and dispersal among local assemblages. However, our findings also indicated that natural communities are structured simultaneously by a range of mechanisms, falling within a three dimensional continuum according to the relative importance of environmental heterogeneity, dispersal limitation, and competitive asymmetry. These findings emphasized that knowledge of habitat connectivity and spatial processes is crucial to the management of stream fish biodiversity in this system.;The second chapter details an experimental study in which we used field-based density perturbations to address questions concerning the importance of local and regional processes in determining local population dynamics. The relative importance of local controls (e.g., survival and recruitment) versus regional controls (e.g., dispersal) of stream fish population dynamics remains largely unknown. The overriding objective of this research was to quantify and characterize the extent to which the local population dynamics of two Appalachian, small stream fishes---brook trout (Salvelinus fontinalis) and mottled sculpin (Cottus bairdi)---were influenced by internal (i.e., population density, survival, and recruitment) versus external (i.e., immigration) processes as well as the extent to which those dynamics differed given the local (e.g. source or sink habitat) and regional context (i.e., proximity to a source versus sink habitat).;Our results indicated that these small streams that are on the periphery of mottled sculpin distributions and within the core of the brook trout geographic distribution in this region may serve as functional sources for brook trout and functional sinks for mottled sculpin. We also demonstrated that, although the local context likely leads to predictable patterns in local processes, the regional context also modifies local dynamics. However, model inference provided little support for density dependence in estimates of apparent survival for either species. Nevertheless, there was some support for a density effect on the probability of transitioning among age classes both brook trout and mottled sculpin. Our study also resulted in some interesting, but unexpected, findings. For example, there was a nearly linear decreasing temporal trend in the immigration rate for both species among removal sites over the course of the study. This trend was coupled with a lagged increase in survival rates. Together, the findings suggested that the immediate recovery of the experimentally perturbed reaches was due an influx of predominantly mobile fish and that the subsequent return to a more locally driven dynamic over time was due to delayed accumulation of a more sedentary population segment.;The third chapter was a demonstration of the deflation of stream fish assemblages at the watershed scale as a result of accumulating local impacts. Our results confirmed that there are measurable biological consequences to cumulative mining impacts within the regional stream network. We also observed substantial variability in species specific responses to local and regional conditions. Finally, we were able to generate explicit estimates of the relative biological costs associated with local and regional impairment on species abundance and likelihood of occurrence as well as assemblage composition and integrity. However, we did not observe threshold responses of fish metacommunities. Instead, watershed scale losses accumulated continuously across a range of watershed conditions. These findings indicated that effective management of stream fishes must consider the watershed context and should seek to affect local and regional conditions through the restoration and protection of interconnected drainage networks. (Abstract shortened by UMI.).