GAP analysis of fish in the hydrologic unit 12090205 of central Texas



Journal Title

Journal ISSN

Volume Title


Texas Tech University


GAP analysis has been successfully used to predict the biodiversity and develop conservation priorities on land on a broad scale since 1987. Its application to aquatic ecosystems started in New York in 1995 on a watershed scale and in Missouri in 1997 on a statewide scale. No complete standard method is currently in use. This project attempted to apply GAP analysis to the eight-digit hydrologic unit 12090205 of the Colorado River basin in central Texas to identify and prioritize opportunities for conserving fish biodiversity in the riverine ecosystems, and to demonstrate the feasibility of applying tiie GAP analysis approach to the aquatic ecosystem in Texas.

The regular GAP analysis procedures were followed in the study. Sampled fish data were compiled and put into a customized Microsoft Access® relational database. The watershed-wide "known" distribution maps were produced by geographically linking each sample to the National Hydrologic Dataset (NHD) which allows the graphical display of sampling locations, and spatially cross-referencing these data in a tabular format. The riverine ecosystems were classified into Valley Segment Types using ArcGIS® (Environmental Systems Research Institute, Inc). The eight variables for the classification were Size, Size Discrepancy, Gradient, Valley Wall Interaction Points, FIow, Rocktype, Floodplain and Land Use/Land Cover. Additional auxiliary attributes, Lowland, Soil pH, Pool, Lake, Backwater, Gravelpits, Mouth and Sewage, were also assigned to each segment. Habitat affinities were compiled and extracted from a number of available sources. New habitat-affinity information was also generated from the sampling database in conjunction with the valley-segment datalayer. Habitat-affinity models were then created using Structured Query Language (SQL) and were used to predict species occurrence on valley segments for each species known to occur in the study area. Logistic-regression models also were developed and employed to predict of species occurrence for the purposes of references and comparison. The variables Water Quality, Land Use/Land Cover, Road/Rail Road and Dam were used to classify the valley segments to develop an index of environmental quality for fish. The environmental quality index, the predicted fish biodiversity and the number of fish species of special concern (i.e., endemic to Texas, endangered/threatened and with a state rank from S2 to S4) were combined to develop tiie conservation priority ranks for each segment. The segments with a prior conservation rank were calibrated with the current conservation status and feasibility.

The results show that I) the segments with high environmental quality are in Barton Creek, in the upper portion of Onion Creek and in the Balcones Canyonlands; 2) predicted fish diversity is expected to be highest in midsized streams; and 3) three groups of segments were identified as candidate conservation sites. They are i) the lower portion of Barton Creek, ii) the group of segments including three segments of Colorado River, Coldwater Creek, and three adjacent segments, and iii) two lower segments of Onion Creek and two segments of Colorado River right downstream of the City of Austin. The segments with high conservation priority after calibration are in the second group, and they are the gap for conservation; 4) the GAP analysis approach used in this study was shown to be a feasible method for developing conservation priorities in central Texas.



Central, Endangered ecosystems -- Texas, Central -- Remote-sensing, Fishes -- Ecology -- Texas, Central -- Remote-sensing, Biotic communities -- Texas, Central -- Colorado River, Fish communities -- Texas, Stream conservation -- Texas -- Colorado River, Fish populations -- Estimates -- Texas -- Colorado