Pesticides in groundwater in the Anacostia River and Rock Creek watersheds in Washington, D.C., 2005 and 2008

The U.S. Geological Survey (USGS), in cooperation with the District Department of the Environment, conducted a groundwater-quality investigation to (a) determine the presence, concentrations, and distribution of selected pesticides in groundwater, and (b) assess the presence of pesticides in groundwater in relation to selected landscape, hydrogeologic, and groundwater-quality characteristics in the shallow groundwater underlying the Anacostia River and Rock Creek watersheds in Washington, D.C. With one exception, well depths were 100 feet or less below land surface. The USGS obtained or compiled ancillary data and information on land use (2001), subsurface sediments, and groundwater samples from 17 wells in the lower Anacostia River watershed from September through December 2005, and from 14 wells in the lower Anacostia River and lower Rock Creek watersheds from August through September 2008.

Twenty-seven pesticide compounds, reflecting at least 19 different types of pesticides, were detected in the groundwater samples obtained in 2005 and 2008. No fungicides were detected. In relation to the pesticides detected, degradate compounds were as or more likely to be detected than applied (parent) compounds.

The detected pesticides chiefly reflected herbicides commonly used in urban settings for non-specific weed control or insecticides used for nonspecific haustellate insects (insects with specialized mouthparts for sucking liquid) or termite-specific control. Detected pesticides included a combination of pesticides currently (2008) in use, banned or under highly restricted use, and some that had replaced the banned or restricted-use pesticides. The presence of banned and restricted-use pesticides illustrates their continued persistence and resistance to complete degradation in the environment. The presence of the replacement pesticides indicates the susceptibility of the surficial aquifer to contamination irrespective of the changes in the pesticides used.

A preliminary review of the data collected in 2005 and 2008 indicated that differences in the surficial geology, land use (as a surrogate for pesticide use), and above-average precipitation for most of 2004 through 2008, as well as differences in the number and performance of USGS laboratory methods used, could have led to more pesticides detected in groundwater samples collected in 2008 than in groundwater samples collected in 2005. Thus, although data from both years of collection were used for interpretive analysis, emphasis was placed on the analysis of the data obtained in 2008.

The presence of pesticides in shallow groundwater (less than approximately 100 ft (feet), or 30 m (meters), below land surface) indicated at least the upper surficial aquifer in Washington, D.C. was susceptible to contamination. One or more herbicides or insecticides were detected in groundwater samples collected from 50 percent of the shallow wells sampled in 2005, and from 62 percent of the shallow wells sampled in 2008.

Differences among types of pesticides in shallow groundwater were apparent. The most frequently detected class of herbicides was the s-triazine compounds—atrazine, simazine, or prometon, or the atrazine-degradate compounds—2-chloro-4-ethylamino-6-amino-s-triazine (desethylatrazine or CIAT) and 2-chloro-4-isopropylamino-6-amino-s-triazine (hydroxyatrazine or OIET). The next most frequently detected classes of herbicides were the chloroacetanilides, including metolachlor and acetochlor, and the ureic herbicides, including diuron (and degradate, 3,4-dichloroaniline), fluometuron, metsulfuron methyl, sulfameturon, bromacil, and tebuthiuron.

Insecticides also were detected, but less frequently than herbicides, with one or more insecticides present in groundwater samples from 38 percent of shallow wells sampled in 2008. Detected insecticides included parent or degradate compounds commonly used for either nonspecific or haustellate (sucking) insects, including chlorpyrifos and dichlorodiphenyldichloroethane (p,p’-DDD; a degradate of dichlorodiphenyltrichloroethane, DDT), and for termite control, including dieldrin, chlordane, heptachlor epoxide, (a degradate of heptachlor), fipronil, and the sulfone and sulfide degradates of fipronil.

The concentrations of individual pesticides in shallow groundwater in both years were low. Maximum concentrations were no greater than a few tenths of a microgram per liter (μg/L); typical concentrations often were less than 0.1 μg/L. Multiple pesticides, however, commonly were present in groundwater. For example, in 2008, approximately 88 percent (7 of 8) of the wells that yielded a sample with at least one detectable pesticide contained five or more pesticides. The highest number of detections occurred in a groundwater sample from well WE Ca 32, which is located in a highly developed urban area; this sample contained 15 different pesticide residues.

In relation to human and aquatic health, no pesticide concentration in either 2005 or 2008 exceeded Federal drinking-water standards. Groundwater samples from a few sites, however, contained levels of chiefly banned or restricted-use pesticides that exceeded other human-health and (or) aquatic-health guidelines. For example, concentrations of dieldrin in 2008 groundwater samples from three wells—WE Ca 32 (0.028 μg/L), WE Ba 11 (0.016 μg/L), and WW Ac 8 (0.014 μg/L)—fell within the range of concern for 2004 Federally approved non-regulatory USGS Health-Based Assessment benchmarks (0.002 to 0.2 μg/L), and exceeded earlier (1999) Federal criteria for drinking water (0.000052 μg/L). Other individual compounds whose concentrations exceeded 1999 Federal guidelines for samples from one or more of these three sites, or another site, included p,p’-DDD, dichlorodiphenyldichloroethylene (p,p’-DDE; another degradate of DDT), chlordane, and heptachlor epoxide. Pesticide concentrations in groundwater also were compared to three aquatic-health guidelines for freshwater (United States, Great Lakes, or Canada). One or more of these guidelines were exceeded in groundwater samples obtained in 2005 or 2008 for one or more of the compounds chlordane, dieldrin, heptachlor epoxide, p,p’-DDE, p,p’-DDD, and chlorpyrifos.

The spatial distribution of pesticides in the shallow groundwater appeared to be related, in part, to land use, a surrogate for pesticide use. Although most of the wells sampled in this study are in parklands or other relatively open and accessible space, multiple pesticides most often were detected in 2008 groundwater samples collected from wells where a considerable percentage (in excess of 60 percent) of the land within a 500-m radius is developed space (residential, commercial, or other urban infrastructure). Insecticides were detected in wells surrounded by at least 50 percent, and most commonly by more than 80 percent, development. Well WE Ca 32, the site associated with the highest number of pesticide residues in groundwater (8 herbicides and 7 insecticides), is in a small residential park, where 99 percent of the surrounding land is well-maintained residential and commercial development.

The vertical distribution of detected pesticides in shallow groundwater appeared to be related, in part, to depth below land surface, surficial-bedrock type, and differences in the chemistry of shallow groundwater. Pesticides were detected at relatively shallow depths in wells that may not have fully penetrated the shallow aquifer. For wells in which at least one pesticide was detected, the median depth below land surface to the top of the well screen was 5.8 m, and the maximum depth was 8.5 m.

Among the types of surficial materials in which wells were completed—alluvium, terrace deposits, or Potomac Formation sub- or outcrops in the Coastal Plain Province, and saprolite or fractured bedrock (Laurel and Sykesville Formations) underlying saprolite in the Piedmont Province—no pesticides were detected in groundwater associated with wells completed in the alluvium or fractured bedrock. Detections occurred in some but not all wells completed in the other surficial materials. Overall, the pattern in occurrence appeared related to the local permeability of these sediments and groundwater chemistry. Groundwater with multiple pesticide detections tended to occur in permeable sediments (absent any appreciable overlying clay, silt, or clay-silt layers), in conjunction with other common urban contaminants (elevated chloride in excess of tens to hundreds of milligrams per liter (mg/L), and oxic, rather than reduced, groundwater as evidenced by elevated (in excess of 5 mg/L) concentrations of nitrate).

The results of this investigation were compared to results from two other similar and recent studies on pesticide occurrence in the shallow aquifer. These included a study in the nearby Maryland and Delaware Coastal Plain Physiographic Province and one in the Maryland and Virginia Piedmont Physiographic Province. Results from these studies were similar to the current study in relation to (a) the types, frequencies, concentrations, and mixtures of pesticides detected; (b) compounds that exceeded human-and aquatic-health criteria; and (c) the occurrence and distribution of pesticides within the surficial aquifer in relation to depths, sediment types, and groundwater chemistries.