Predicting and preventing losses of imperiled fish species in an urbanizing environment
Wenger, Seth J.
MetadataShow full item record
Increasing urban land cover is a threat to many freshwater fish species. To effectively manage this threat we need to understand the nature of the stressors, develop relationships between stressors and measures of population viability, create a policy for limiting stressors and use predictive modeling to fine tune the policy and forecast the outcome of management. In this dissertation, I use the development of a Habitat Conservation Plan (HCP) for the imperiled fish species of the Etowah River basin, Georgia, USA, as a case study in applying this kind of conservation approach. I review the literature on urban effects on fishes, concluding that stormwater runoff from impervious surfaces is most likely to be the major threat to imperiled species of the Etowah. I then develop models relating the presence/absence of five species to effective impervious area (EIA) after accounting for historic land use, hydrogeomorphic variables and the confounding factors of incomplete detection and spatial autocorrelation. For a species (the Cherokee darter, Etheostoma etowahae) that shows a relationship between EIA and abundance but not presence, I propose an extension of existing methods to simultaneously model species presence, abundance and detection. I then explain the policy we have developed under the Etowah HCP for limiting the total volume of runoff from developed sites. This policy has the potential to limit EIA to sufficiently low levels to permit species persistence in the Etowah while accommodating future growth. Finally, I illustrate how predictive modeling can be used to guide application of this policy and forecast the outcomes, even with limited data, and propose this as a general approach to managing for imperiled species threatened by land use change.