The occurrence and distribution of selected herbicides, atrazine metabolites, and nitrate were determined for near-surface aquifers (within 50 feet of land surface) in the corn- and soybean-producing region of the midcontinental United States. The study region included all or parts of Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, and Wisconsin. Water samples were collected during the spring and summer of 1991 from 303 wells completed in near-surface unconsolidated and near-surface bedrock aquifers. At least one herbicide or atrazine metabolite was detected in 24 percent of 579 water samples analyzed for herbicides, based on a reporting limit of 0.05 microgram per liter. However, no herbicide concentration exceeded the U.S. Environmental Protection Agency's maximum contaminant levels or health advisory levels for drinking water. The most frequently detected herbicide compound was desethylatrazine, an atrazine metabolite (18.1 percent), followed by atrazine (17.4 percent); deisopropylatrazine, an atrazine metabolite (5.7 percent); prometon (5.0 percent); metolachlor (2.7 percent); alachlor (1.7 percent); simazine (1.0 percent); metribuzin (1.0 percent); and cyanazine (0.7 percent). The herbicides ametryn, prometryn, propazine, and terbutryn were not detected during this study. Nitrate concentrations equal to or greater than 3.0 milligrams per liter (excess nitrate) were detected in 29 percent of the 599 nitrate analyses, and ammonium concentrations equal to or greater than 0.01 milligram per liter were detected in 78 percent of the 584 ammonium analyses. Nitrate concentrations equal to or greater than the U.S. Environmental Protection Agency's maximum contaminant level for drinking water of 10 milligrams per liter were found in 6 percent of the samples.

The frequency of herbicide detection was, in part, affected by the analytical method's reporting limit. Results from this study show that the frequency of atrazine detection increases as the reporting limit decreases. Herbicide metabolite concentrations are critical to understanding the detection of herbicide residues. The frequency of detection of atrazine residue (atrazine + desethylatrazine + deisopropylatrazine) was 22.1 percent, which was more than the frequency of detection of atrazine alone (17.4 percent).

Prometon was detected more frequently than every other herbicide except atrazine. The prometon appears to be derived from areas of nonagricultural land use, such as golf courses and residential areas. Herbicides and excess nitrate were both rarely detected in the eastern part of the study region, even though this is an area of intense herbicide and nitrogen-fertilizer use.

Hydrogeologic factors, land use, agricultural practices, local features, and water chemistry were analyzed for possible relation to herbicide and excess-nitrate detections. Herbicides and excess nitrate were detected more frequently in near-surface unconsolidated aquifers than in nearsurface bedrock aquifers. The depth to the top of the aquifer was inversely related to the frequency of detection of herbicides and excess nitrate. The proximity of streams to sampled wells also affected the frequency of herbicide detection. Significant seasonal differences were determined for the frequency of herbicide detection, but not for the frequency of excess nitrate.