INVESTIGATION OF TEMPORAL AND SPATIAL DIFFERENCES IN FISH COMMUNITIES OF THE HIWASSEE RIVER WATERSHED

Human-induced disturbances, such as addition of infrastructure, deforestation, flow modification, and pollution can negatively impact fish biodiversity within watersheds. Thus, changes in fish communities and assemblages can serve as a bioindicator of water quality deterioration and habitat degradation. In addition, community comparisons of subwatersheds within basins can be indicative of historical impacts. We investigated whether changes have occurred within a 15-year period in the Hiwassee River Watershed (HRW) temporally as well as whether communities differ spatially in subwatersheds within the basin. Using fish community data provided by the North Carolina Department of Environmental Quality, we analyzed eight different subwatersheds in the HRW across three sampling periods (2004, 2009, and 2014) using community metrics such as richness, evenness, Shannon\u27s diversity index (SDI) and ecological distance. Significant differences among means between years and subwatersheds were detected using analysis of variance with a posthoc Tukey\u27s test and Kruskal-Wallis, respectively, using BiodiversityR in R 3.5.1. Similarity among sites was determined using principle components analysis and constructing a dendrogram from an ecological distance matrix. Fish communities were not significantly different (P = 0.64) temporally. However, there were significant differences among two of the eight subwatersheds, Little Brasstown Creek and Shooting Creek, in mean evenness (P = 0.005) and SDI (P \u3c 0.001). Based upon available water quality data and land use measurements, differences among communities may be related to poor water quality resulting from activities such as unsustainable methods of agriculture and livestock operation, deforestation, flow modification and septic system leakage. Thus, human-induced disturbances such as these can reduce overall fish biodiversity in the HRW if stream best management practices are not observed