Improving inference from biological assessments
Carroll, Gwendolyn Denise
MetadataShow full item record
Since the establishment of the US Clean Water Act in 1972, biological assessments using fish and macroinvertebrates have been employed to detect stream perturbations and are required to assess water quality in locations across the globe. Despite the widespread use of such techniques, many questions remain unanswered concerning relationships between biological factors and physical habitat. As Georgia continues to grow, it is important to identify how that growth will impact aquatic communities and water quality. A better understanding of linkages between ecological drivers and aquatic communities is useful for predictive modeling of watershed conditions and the development of management strategies, providing suggestions to help maintain or improve the health of a watershed. This study investigates relationships between aquatic communities and 1) urbanization and 2) turbidity, SSC, and percent fines, as well as examining interrelationships between fish and macroinvertebrate assemblages. Our results indicate biotic community metrics were highly sensitive to land cover, specifically the fractions of urban land, imperviousness, and forest. Nearly all fair, good, and excellent fish IBI scores occurred in basins with more than 50 percent forest and less than 15 percent urban area and less than 4 percent impervious surface, indicating that Georgia Piedmont streams may be more sensitive to urbanization effects than streams in other parts of the country. Overall, sites with less urbanization had greater litter standing crops during December; however, higher rates of retention occurred in more urbanized areas. We infer that urban streams balance litter export with additional horizontal inputs from storm drains that act to increase the litter source area. Macroinvertebrate shredder taxa richness was negatively affected by watershed landuse, but shredder abundance and percent composition were not. Shredder abundance and composition were not correlated to litter availability. Neither baseflow suspended sediment concentrations, baseflow turbidity, nor bed particle size distributions were significantly related to watershed imperviousness or other land use metrics. However, these metrics added significantly to explanatory models of important biological metrics. Macroinvertebrate and fish biotic indices do not provide the same information with regard to water quality, although fish index of well being scores are more highly correlated than IBI scores.