Abstract

Little existing information was available describing pesticide occurrence in ground water of Wyoming, so the U.S. Geological Survey, in cooperation with the Wyoming Department of Agriculture and the Wyoming Department of Environmental Quality on behalf of the Wyoming Ground-water and Pesticides Strategy Committee, collected ground-water samples twice (during late summer/early fall and spring) from 296 wells during 1995–2006 to characterize pesticide occurrence. Sampling focused on the State’s ground water that was mapped as the most vulnerable to pesticide contamination because of either inherent hydrogeologic sensitivity (for example, shallow water table or highly permeable aquifer materials) or a combination of sensitivity and associated land use.

Because of variations in reporting limits among different compounds and for the same compound during this study, pesticide detections were recensored to two different assessment levels to facilitate qualitative and quantitative examination of pesticide detection frequencies—a common assessment level (CAL) of 0.07 microgram per liter and an assessment level that differed by compound, referred to herein as a compound-specific assessment level (CSAL). Because of severe data censoring (fewer than 50 percent of the data are greater than laboratory reporting limits), categorical statistical methods were used exclusively for quantitative comparisons of pesticide detection frequencies between seasons and among various natural and anthropogenic (human-related) characteristics.

One or more pesticides were detected at concentrations greater than the CAL in water from about 23 percent of wells sampled in the fall and from about 22 percent of wells sampled in the spring. Mixtures of two or more pesticides occurred at concentrations greater than the CAL in about 9 percent of wells sampled in the fall and in about 10 percent of wells sampled in the spring. At least 74 percent of pesticides detected were classified as herbicides. Considering only detections using the CAL, triazine pesticides were detected much more frequently than all other pesticide classes, and the number of different pesticides classified as triazines was the largest of all classes.

More pesticides were detected at concentrations greater than the CSALs in water from wells sampled in the fall (28 different pesticides) than in the spring (21 different pesticides). Many pesticides were detected infrequently as nearly one-half of pesticides detected in the fall and spring at concentrations greater than the CSALs were detected only in one well. Using the CSALs for pesticides analyzed for in 11 or more wells, only five pesticides (atrazine, prometon, tebuthiuron, picloram, and 3,4-dichloroaniline, listed in order of decreasing detection frequency) were each detected in water from more than 5 percent of sampled wells. Atrazine was the pesticide detected most frequently at concentrations greater than the CSAL.

Concentrations of detected pesticides generally were small (less than 1 microgram per liter), although many infrequent detections at larger concentrations were noted. All detected pesticide concentrations were smaller than U.S. Environmental Protection Agency (USEPA) drinking-water standards or applicable health advisories. Most concentrations were at least an order of magnitude smaller; however, many pesticides did not have standards or advisories.

The largest percentage of pesticide detections and the largest number of different pesticides detected were in samples from wells located in the Bighorn Basin and High Plains/ Casper Arch geographic areas of north-central and southeastern Wyoming. Prometon was the only pesticide detected in all eight geographic areas of the State.

Pesticides were detected much more frequently in samples from wells located in predominantly urban areas than in samples from wells located in predominantly agricultural or mixed areas. Pesticides were detected distinctly less often in samples from wells located in predominantly rangeland/undeveloped areas. The frequency of pesticide detection in samples from wells located in either predominantly agricultural or mixed land-use areas was intermediate to those samples from urban and rangeland/undeveloped areas. Using the CAL, the proportion of wells with at least one pesticide detected was significantly different among the four land-use categories for both the fall and the spring.

Pesticide detections in ground water were examined in relation to hydrogeology (aquifer type, water-level depth, well depth, and well type). The percentage of wells with at least one pesticide detected was larger for wells completed in unconsolidated-deposit aquifers than for wells completed in bedrock aquifers. Using the CAL, the proportion of wells with at least one pesticide detected was significantly different between unconsolidated and bedrock aquifers sampled in the fall but not in the spring. Also using the CAL, the proportion of wells with at least one pesticide detected was significantly different among different categories of well depth and well type. The proportion of samples with pesticides detected at concentrations greater than their respective CSALs (detection frequency) was significantly different among water-leveldepth categories for one pesticide—deethylatrazine (spring but not fall samples). In most cases, pesticide detection frequencies decreased as well depth increased, but no clear increase or decrease in pesticide detection frequencies was noted as water-level-depth increased. The proportion of samples with pesticides detected at concentrations greater than their respective CSALs was significantly different among different categories of well depth and well type for three pesticides (prometon, tebuthiuron, and bromacil, listed in order of decreasing detection frequency) detected in both the fall and spring, and for one pesticide detected (diuron) in the spring.

Pesticide detections in ground water were examined in relation to selected soil properties (organic matter content, soil permeability, and soil hydrologic index) from the State Soil Geographic (STATSGO) database mapped within a 500-meter (1,640-foot) radius surrounding each sampled well. The proportion of samples with pesticides detected at concentrations greater than their respective CSALs was significantly different among organic-matter-content categories for four pesticides (atrazine, tebuthiuron, flumetsulam, and fipronil sulfide, listed in order of decreasing detection frequency) detected in the fall and in one pesticide (tebuthiuron) detected in the spring. The proportion of samples with pesticides detected at concentrations greater than their respective CSALs was significantly different among soil-permeability categories for three pesticides (deethylatrazine, tebuthiuron, and aldicarb sulfone, listed in order of decreasing detection frequency) detected in the fall and for two pesticides (atrazine and tebuthiuron) detected in the spring. The proportion of samples with pesticides detected at concentrations greater than their respective CSALs was significantly different among soil-hydrologic-index categories (derived from several soil properties) for two pesticides (atrazine and deethylatrazine) detected in both fall and spring. For most pesticides, no clear trend of increasing or decreasing frequency of detection was noted as organic matter content and soil permeability increased, and there was not a clear trend among the soil-hydrologic-index categories.

One water-quality constituent (nitrate) and three different water-quality characteristics (specific conductance, pH, and dissolved oxygen) were examined in relation to the occurrence of pesticides in ground water. Using the CAL, the proportion of wells with at least one pesticide detected was significantly different between the nitrate concentration categories of less than and greater than or equal to 1.1 milligrams per liter for the fall and the spring. The proportion of samples detected at concentrations greater than their respective CSALs was much larger for the “elevated” nitrate concentration category than for the other nitrate concentration category for almost every pesticide tested, and the difference was significant between the two nitrate concentration categories for three pesticides (atrazine, prometon, and deethylatrazine, listed in order of decreasing detection frequency) in both the fall and spring, one pesticide (tebuthiuron) in the fall, and one pesticide (bromacil) in the spring.

The proportion of samples with pesticides detected at concentrations greater than their respective CSALs (detection frequency) was much larger for the saline-water category than for the freshwater category (on the basis of specific conductance) for almost every pesticide tested. The proportion of samples with pesticides detected at concentrations greater than their respective CSALs was significantly different between the two specific conductance categories for one pesticide (tebuthiuron) in both the fall and spring and for one pesticide (prometon) only in the fall.

Using the CAL, the proportion of wells with at least one pesticide detected was significantly different among the three pH categories—acidic, circumneutral, and alkaline—for the fall and spring. The proportion of samples with pesticides detected at concentrations greater than their respective CSALs was significantly different among the three pH categories for three pesticides (prometon, tebuthiuron, and bromacil, listed in order of decreasing detection frequency) detected in both the fall and spring and for one pesticide (diuron) in the spring. For many pesticides, no clear trend of increasing or decreasing frequency detection was noted as pH increased.

The proportion of samples with pesticides detected at concentrations greater than their respective CSALs was significantly different between the two dissolved oxygen categories (representative of anoxic and oxic conditions) for one pesticide (tebuthiuron) in the fall.

Posted March 12, 2009