Pesticides in Near‐Surface Aquifers: An Assessment Using Highly Sensitive Analytical Methods and Tritium

Abstract

In 1992, the U.S. Geological Survey (USGS) determined the distribution of pesticides in near‐surface aquifers of the midwestern USA to be much more widespread than originally determined during a 1991 USGS study. The frequency of pesticide detection increased from 28.4% during the 1991 study to 59.0% during the 1992 study. This increase in pesticide detection was primarily the result of a more sensitive analytical method that used reporting limits as much as 20 times lower than previously available and a threefold increase in the number of pesticide metabolites analyzed. No pesticide concentrations exceeded the U.S. Environmental Protection Agency's (USEPAs) maximum contaminant levels or health advisory levels for drinking water. However, five of the six most frequently detected compounds during 1992 were pesticide metabolites that currently do not have drinking water standards determined. The frequent presence of pesticide metabolites for this study documents the importance of obtaining information on these compounds to understand the fate and transport of pesticides in the hydrologic system. It appears that the 56 parent compounds analyzed follow similar pathways through the hydrologic system as atrazine. When atrazine was detected by routine or sensitive analytical methods, there was an increased likelihood of detecting additional parent compounds. As expected, the frequency of pesticide detection was highly dependent on the analytical reporting limit. The number of atrazine detections more than doubled as the reporting limit decreased from 0.10 to 0.01 µg/L. The 1992 data provided no indication that the frequency of pesticide detection would level off as improved analytical methods provide concentrations below 0.003 µg/L. A relation was determined between groundwater age and the frequency of pesticide detection, with 15.8% of the samples composed of pre‐1953 water and 70.3% of the samples of post‐1953 water having a detection of at least one pesticide or metabolite. Pre‐1953 water is less likely to contain pesticides because it tends to predate the use of pesticides to increase crop production in the Midwest. Pre‐1953 water was more likely to occur in the near‐surface bedrock aquifers (50.0%) than in the near‐surface unconsolidated aquifers (9.1%) sampled.