Desert Fishes Council, Death Valley, CA, November 2000.

Identifying areas of conservation priority for native fishes in the Gila River Basin using GIS.

P.J. Unmack & W.L. Minckley

The southwestern United States provides a stark example of a depauperate and unique native fish fauna decimated by environmental depredation, introduced non-native species, and other factors. The fauna has declined to the point where most species are biologically imperiled. Due to the restricted availability of surface water and concern for the continued existence of the fauna, the region has been intensively sampled and substantial records exist from which to build a comprehensive database of fish occurrences. As reported at last year’s meeting, individual records for the Gila River Basin have been compiled from museum records, primary, and "gray" literature. Data were geographically referenced using GIS. Several components were incorporated into determining conservation priorities: present day richness for native species, threatened and endangered species richness, percentage and absolute decline in native fishes, non-native species richness, etc. Our results consisted of a table containing many factors (as outlined above) summarized by watersheds and stream reaches. These data were best suited for graphical display. There are many ways conservation priorities can be determined; we incorporated two factors, the ratio of threatened species richness to total species richness, and intactness (whether any species had been lost or not). These were divided into seven categories, with the number of watersheds for each shown in parentheses: no fish records (117), not intact and no threatened species (88), intact and no threatened species (25), not intact and 15-35% threatened species (15), intact and 15-35% threatened species (17), not intact and >40% threatened species (11), and intact and >40% threatened species (6). Results and data from our analyses are available on the Internet at http://fish.la.asu.edu/gilafish. Several problems arose. For example, how should a stream that once had seven native species but now has only four natives and many introduced species rank relative to one that has only ever had two native fishes and today has no exotics? Distribution of early collection records was biased by access limitations rather than covering representative areas. In recent years, few voucher specimens have been retained, and specimens of introduced species were rarely kept, so that now most recent records tend to be in "gray" literature which cannot be confirmed, checked for misidentifications, or re-examined when taxonomic changes occur. Another problem is the lack of negative data, it is not possible to know if species weren't collected because they were not present, or if they were not adequately sampled. GIS provided an excellent means by which this type of data can be analyzed. Assimilation of data into a GIS allows ease of display and high query capability. More importantly, GIS allows spatial analysis of relations among biotic and abiotic factors, including fish occurrence relative to permanency of water, geologic structure, stream gradients, altitude, land use, species associations, etc.