Samples of water, bed material, fish, and soil were collected in four small drainage basins in Pennsylvania in 1969-71 and analyzed to determine the concentrations of chlorinated-hydrocarbon insecticides. Water samples only were also analyzed for phenoxy-acid herbicides. Each basin studied represents a predominant land-use classification—forested, general farming, residential, and orchard farming. All water and fish samples showed pesticide concentrations less than the U.S. Public Health Service's (1969) recommended maximum permissible concentration. However, no fish were found in the orchard area stream at the time collection was attempted. DOT or one of its metabolites was the most frequently occurring insecticide and was detected in all media sampled except the forestedarea soil. The highest observed combined concentration of DOT and its metabolites in storm-runoff samples was 11.4 micrograms per litre in a sample collected from the residential area stream, but the median was higher (0.12 microgram per litre) in the orchard area than in the residential area (0.02 microgram per litre). A sample of the top 0.5 inch (13 millimetres) of orchard soil contained 40,000 micrograms per kilogram of DOT and its metabolites, even though DOT had not been used in the orchards for several years prior to this study. Maximum concentrations detected in other orchard media are 330 micrograms per kilogram in bed material and 3.45 micrograms per litre in storm runoff. Dieldrin was the second most frequently occurring insecticide. Other insecticides detected were chlordane, heptachlor epoxide, lindane, and a trace of aldrin in one fish sample. At least one of the following herbicides—2,4-D, silvex, or 2,4,5-T—was detected in each stream.
