PARENT SESSION
Thursday, August 10, 1:30-5:00 pm
COS 100 - Conservation ecology II
L-12, Lobby Level, Cook Convention Center
Presiders: N Nicholas

Predicting persistence of imperiled fish species in an urbanizing landscape.

Wenger, Seth^*,1, Freeman, Mary², Freeman, Byron^{1, 3}, ¹ Institute of Ecology, Athens, GA, USA² U.S. Geological Survey, Athens, GA, USA³ Georgia Museum of Natural History, Athens, GA, USA

ABSTRACT- Management decisions for imperiled species often must be made on the basis of limited information. Nevertheless, these decisions should be rational, repeatable, and based on the best available scientific information. We developed a statistical predictive modeling approach to guide policies for protection of nine imperiled stream fish species as part of a Habitat Conservation Plan in the Etowah River Basin, Georgia, USA. We developed logistic regression models accounting for incomplete detectability and spatial autocorrelation based on data from 357 collection records. We then applied these models across the ranges of the target fish species to predict occurrence under various development buildout scenarios in the Etowah basin. A key predictor, effective impervious area (EIA), varied in each scenario. Working with local governments, we introduced an innovative stormwater management program to limit the volumes of runoff from new development and maintain pre-development hydrologic function in critical watersheds, thus limiting EIA to levels predicted to maintain high occurrence probabilities of target species. Some development nodes with higher runoff levels were permitted; we worked cooperatively with governments to model the impact of proposed nodes and limit their size and location appropriately. Although uncertainties in the predictions are large, the method provides a rational basis for decisions and a framework that can be continually improved through an adaptive management process.

Key words: conservation, urban, fish