RELATIONSHIPS BETWEEN WATERSHED STRESSORS AND SEDIMENT CONTAMINATION IN CHESAPEAKE BAY ESTUARIES

Three methods for assessing the relationships between estuarine sedime nt contaminant levels and watershed stressors for 25 Chesapeake Bay su b-estuaries were compared. A geographic information system (GIS) was u sed to delineate watersheds for each sub-estuary and analyze land use pattern (area and location of developed, herbaceous and forested land) and point source pollution (annual outflow and contaminant loading) u sing three landscape analysis methods: (1) a watershed approach using the watershed of the estuary containing the sampling station, (2) a 'p artial watershed' approach using the area of the watershed within a 10 km radius of the sampling station and (3) a 'weighted partial watersh ed' approach where stressors within the partial watershed were weighte d by the inverse of their linear distance from the sampling station. N ine sediment metals, 16 sediment organics and seven metals loading var iables were each reduced to one principal component for statistical an alyses. Relationships between the first principal components for sedim ent metals and organics concentrations and watershed stressor variable s were analyzed using rank correlation and stepwise multiple regressio n techniques. For both metals and organics, the watershed method yield ed R(2) values considerably lower than the partial and weighted partia l watershed analysis methods. Regression models using stressor data ge nerated by the weighted partial watershed landscape analysis method ex plained 76% and 47% of the variation in the first principal component for sediment metals and organics concentrations, respectively. Results suggest that the area of developed land located in the watershed with in 10 km of the sediment sampling station is a major contributing fact or in the sediment concentrations of both metals and organics.