Herbicides and their transformation products in source-water aquifers tapped by public-supply wells in Illinois, 2001-02

During 2001-02, ground-water samples were collected from 117 public-supply wells distributed throughout Illinois to evaluate the occurrence of herbicides and their transformation products in the State?s source-water aquifers. Wells were selected using a stratified-random method to ensure representation of the major types of source-water aquifers in the State. Samples were analyzed for 18 herbicides and 18 transformation products, including 3 triazine and 14 chloroacetanilide products. Herbicide compounds (field-applied parent herbicides and their transformation products) were detected in 34 percent of samples. A subset of samples was collected unfiltered to determine if analytical results for herbicides in unfiltered samples are similar to those in paired filtered samples and, thus, can be considered equally representative of herbicide concentrations in ground water supplied to the public. The study by the U.S. Geological Survey was done in cooperation with the Illinois Environmental Protection Agency. Parent herbicides were detected in only 4 percent of all samples. The six most frequently detected herbicide compounds (from 5 to 28 percent of samples) were chloroacetanilide transformation products. The frequent occurrence of transformation products and their higher concentrations relative to those of most parent herbicides confirm the importance of obtaining information on transformation products to understand the mobility and fate of herbicides in ground-water systems. No sample concentrations determined during this study exceeded current (2003) Federal or State drinking-water standards; however, standards are established for only seven parent herbicides. Factors related to the occurrence of herbicide compounds in the State?s source-water aquifers include unconsolidated and unconfined conditions, various hydrogeologic characteristics and well-construction aspects at shallow depths, and proximity to streams. Generally, the closer an aquifer (or well location) is to a recharge area and (or) the stronger the hydraulic connection between an aquifer and a recharge area, the younger the ground water and the more vulnerable the aquifer will be to contamination by herbicide compounds. The weak relation between current (2001) statewide application rates of herbicides and current (2001-02) occurrence of herbicide compounds in source-water aquifers indicates that additional factors must be considered when relating herbicide-application rates to occurrence. These factors include historical application rates and the mobility and persistence of the various herbicide compounds in ground-water systems. Frequency of detection and concentrations of herbicides compounds in the State?s source-water aquifers are indicated to be highest during the spring, when crops are planted and herbicides primarily are applied. Excess nitrate (concentrations of nitrate, as nitrogen, higher than 3 milligrams per liter) in ground water strongly indicates the co-occurrence of herbicide compounds. However, nitrate concentrations are not a reliable indicator of herbicide-compound concentrations. The inverse relation found between current use of land for corn and soybean production and current occurrence of herbicide compounds in underlying aquifers indicates that various factors, along with current agricultural land use, contribute to herbicide occurrence. These factors include, among others, land-use history, ground-water age, ground-water-flow patterns, geology, soil microbiology, and chemistry and persistence of the herbicide compounds. Detection of agriculture-specific herbicide compounds in 71 percent of samples from urban areas with no current or recent agricultural land use near the sampled wells indicates that recharge to certain high-capacity supply wells may originate at considerable distances (up to about 10 miles) from the wells. Essentially no difference was found between the analytical results for herbicides in paired unfiltered and filtered samples,