Document: EDW-3-88-3

Sediment modeling and fish assemblages in the upper Little Tennessee River basin: Assessing the legacy.

GARDINER, E.P.* ¹, J.L.MEYER ¹, W.O.MCLARNEY ^2,3 and P.V.BOLSTAD ⁴

University of Georgia, Athens, GA 30502 USA ¹
Tennessee Valley Authority ²
Western North Carolina Alliance ³
University of Minnesota ⁴

Abstract:
Sediment is a common pollutant with long-term impacts on streams of the southeastern United States. We implemented the revised universal soil loss equation (RUSLE) to estimate erosion in the Upper Little Tennessee River watershed (1153 sq. km) in Macon County, North Carolina during three years for which imagery was available: 1950, 1970, and 1992. We simulated sediment transport and delivery to stream channels to derive predicted total sediment loads on each date to all streams in the study area for each year. The model demonstrated that sediment loads to streams were generally lower in 1992 than in either of the two previous years. We used the simulated sediment load data to predict the total abundance of (1) all fish, (2) sediment-intolerant, and (3) sediment-tolerant taxa present in 48 streams sampled between 1992 and 1996. We predicted that all fish and sediment-intolerant fish would be negatively correlated with simulated sediment load summed across all three years as well as for 1992 alone. We predicted no relationship between sediment load and the total abundance of sediment tolerant taxa. The latter prediction was true for all regression models analyzed. The first set of hypotheses was also supported. There appeared to be a response above a sediment load threshold, so we examined the twelve sites with the highest loads in 1992. Predicted sediment load for 1992 explained over 80% of the variation in the abundance of sediment-intolerant taxa. The sum of predicted loadings for all three years explained 78% of the variation in intolerant abundance and 52% of the variation in all taxa. We have demonstrated the importance and utility of distributed sediment modeling for exploring explicit linkages between land cover and fish assemblage structure. This approach is particularly important when using historical fish collections for which ancillary data, such as geomorphological measurements, are not available.

Keywords: RUSLE, sediment modeling, stream habitat, fish, geomorphology, history, land cover, land use

This abstract is being presented at: 11:30 AM in session:
Oral Session #70: Aquatic Ecology.