Occurrence of selected trace elements and organic compounds and their relation to land use in the Willamette River Basin, Oregon, 1992-94

Between 1992 and 1994, the U.S.Geological Survey conducted a study of trace elements and organic compounds in the Willamette River Basin, Oregon, as part of the Willamette River Basin Water Quality Study. Low-level analyses were performed for trace elements, volatile organic compounds, organochlorine compounds, and pesticides. Overall, 94 water samples were collected from 40 sites, during predominantly high-flow conditions, representing urban, agricultural, mixed, and forested land uses. Although most observed concentrations were relatively low, some exceedances of water-quality criteria for acute and chronic toxicity and for the protection of human health were observed. Concentrations of chromium, copper, lead, and zinc in unfiltered water were well correlated with concentrations of suspended sediment. The highest trace-element concentrations generally were found at urban sites that receive a large portion of their runoff from industrial areas, particularly at high suspended-sediment concentrations. In contrast, concentrations of trace elements in some urban streams draining primarily residential areas appeared to approach a maximum as sediment concentrations increased. Whether this difference was due to a difference in the nature of the suspended sediments or to different concentrations in the aqueous phases from the two site types was not addressed. Eight organochlorine compounds were detected at 14 sites. Lindane, dieldrin, and DDT or its metabolites were each detected in about 30 percent of the samples, predominantly in samples collected from agricultural and urban areas. Polychlorinated biphenyl (PCB) compounds were detected in samples from two urban sites. For samples in which DDT and its metabolites were examined for partitioning, the largest proportion of the mass of DDT and its metabolites was associated with suspended sediment. In contrast, dieldrin and lindane were almost completely (greater than 99%) associated with the dissolved phase. Sixty-one of the 94 pesticides analyzed in filtered water were documented to have been used in the basin in 1987; 43 of these were detected at least once during 1992-94. An additional five were detected that were not documented in the 1987 estimates. Although a comparison between the frequency of detected pesticides and 1987 estimates of pesticide usage in the basin showed generally little correlation, some patterns of detections did appear to reflect land use in the basin. Of the 25 most frequently detected pesticides, 3 were found primarily at urban sites, 6 were found primarily at agricultural sites, and 7 were found at all types of sites except forested. The four most commonly detected pesticides in the basin, observed at all except forested site types, were atrazine, metolachlor, simazine, and diuron. A greater variety of compounds was detected at sites in the northern portion of the basin than in the southern portion of the basin probably because the northern portion has more diverse agricultural practices and a larger urban component. Possible reasons for the lack of agreement between pesticide detections and pesticide usage estimates include (1) uncertainty in the usage estimates due to spatial and temporal variability or due to changes in agricultural practices since the 1987 estimates were compiled, (2) chemical or biological transformations in the compounds after application, (3) variable hydrologic conditions among sites at the time of sampling, or (4) the ability of laboratory analytical procedures to detect low concentrations of some analytes. Results from repeated samplings at two sites during sequential storms in the fall of 1994 indicated that concentrations and loads of several constituents, including suspended sediment, suspended organic carbon, DDT, metolachlor, and atrazine were highest during peak flows of the first or second significant storms of the fall. Samplings during subsequent storms indicated that instantaneous concentrations and loads were generally reduced; however, data were not sufficient to compare overall transport during sequential storms.