USGS Fact Sheet 096-00
July 2000

Download this fact sheet as a PDF file (520 KB)

Download a free copy of Acrobat Reader

Prepared in cooperation with the
CITY OF WICHITA, KANSAS

Occurrence of Pesticides in Streams of the Cheney Reservoir Watershed, South-Central Kansas, 1997-99

-Chad R. Milligan and Larry M. Pope

Red Rock Creek near Pretty Prairie, Kansas, 1998.
(Photograph by Chad Milligan, USGS.)

Table of Contents

Background
Methods of Study
Streamflow
Occurrence of Pesticides
Conclusions and Implications
References

The quality of water in the Cheney Reservoir watershed in south-central Kansas is important to about 300,000 people in the Wichita area that rely on the reservoir as a source of drinking water. About 52 percent of the watershed is used for the production of grain crops, which generally rely on the use of pesticides for efficient production. Some pesticides subsequently may be transported in runoff water to Cheney Reservoir. The potential exposure to pesticides is of concern. The purpose of this fact sheet is to describe the occurrence of pesticides in streams within the Cheney Reservoir watershed. Pesticide concentrations are compared with Federal water-quality criteria for pesticides such that public and private organizations can evaluate the potential risks from pesticide exposure.

Background

Methods of Study

A network of six streamflow-gaging/water-quality sampling sites was established in the Cheney Reservoir watershed in the fall of 1996 (fig. 1). Three of the sites were located on the North Fork Ninnescah River (two upstream from Cheney Reservoir and one at the reservoir outflow) and three on associated tributary streams. These sites were used to define the water-quality characteristics of the North Fork Ninnescah River and selected tributary streams.

Streamflow samples were collected manually at the sites according to methods presented in Horowitz and others (1994) during both low-flow (base-flow) and high-flow (storm-runoff) conditions. Water-quality samples were collected to represent the variability in seasonal and hydrologic conditions recorded at each sampling site. Samples for determination of pesticide concentrations were analyzed at the USGS National Water-Quality Laboratory in Denver, Colorado, using gas chromatography/mass spectrometry according to methods presented in Zaugg and others (1995).

Stream-water elevations (stage) were recorded continuously at the six sampling sites using pressure transducers and data-collection platforms. Daily mean streamflow (cubic feet per second) was calculated using stage-to-stage/streamflow relations according to methods presented in Kennedy (1983). Streamflow data for 1997-99 are presented in annual data reports such as Putnam and others (1999) or are on file at the USGS office in Lawrence, Kansas.

Streamflow

Annual mean streamflow in the Cheney Reservoir watershed varied substantially among 1997, 1998, and 1999, and also in comparison to long-term mean annual streamflow (fig. 2). The annual mean streamflow for sampling site 4 (fig. 1) was larger during 1998 (162 ft³/s, cubic feet per second) than either 1997 or 1999 and was 9 percent larger than the long-term mean annual (1966-99) streamflow. Because sampling site 4 was relocated in 1996 to its current location (fig. 1) from a site about 4 miles downstream, annual mean streamflows for 1997-99 were adjusted to approxi-mate annual mean streamflows at that previous location. This adjustment was necessaryto make comparison to long-term (1966-99) annual mean stream-flow.

Annual mean discharge from Cheney Reservoir (sampling site 6, fig. 1) was largest in 1999 (180 ft³/s) and was 31 percent larger than the long-term mean annual (1965-99) streamflow. Much of the relatively large 1999 annual mean discharge from Cheney Reservoir was the result of lowering the reservoir level for dam maintenance. No long-term streamflow data exist for sampling sites 1, 2, 3, and 5 (fig. 1) because these sites were established during this study.

Occurrence of Pesticides

For the purposes of this study, 26 herbicides, 1 herbicide metabolite (degradation product), and 19 insecticides were analyzed to determine the occurrence of pesticides in the Cheney Reservoir watershed in streamflow samples collected from 1997-99. Of the 26 herbicides and 1 metabolite analyzed, 16 were detected in water from sampling sites in the watershed (table 1). The average number of pesticides detected per water sample did not vary substantially among the six sampling sites (four at sampling site 3 to five at sampling site 5, fig. 1). Two of the herbicides, atrazine and metolachlor, were detected in nearly every sample analyzed. Atrazine was detected in 162 of 163 samples (99 percent), and metolachlor was detected in 156 of 163 samples (96 percent). The metabolite of atrazine, deethylatrazine, was detected in 157 of 163 samples (96 percent). Three other herbicides with a large number of detections were alachlor, detected in 100 of 163 samples (61 percent); tebuthiuron, detected in 43 of 154 samples (28 percent); and simazine, detected in 35 of 163 samples (21 percent). Eleven herbicides that were not detected in stream-flow samples from the Cheney Reservoir watershed are listed in table 2. Analytical method reporting limits (MRL's) ranged from 0.001 to 0.004 Īg/L (micrograms per liter) for those herbicides.

Of the herbicides detected in streamflow samples during this study, only alachlor and atrazine had concentrations greater than the U.S. Environmental Protection Agency (1995) established Maximum Contaminant Level (MCL) or Health Advisory Level (HAL). An MCL is the maximum permissible level (on an annual basis) of a contaminant in water that is deliveredto any user of a public water system. An HAL is defined as the maximum concentration of a chemical in drinking water that is not expected to cause adverse noncarcinogenic (noncancer-causing) effects over a lifetime of exposure, with a built-in margin of safety (U.S. Environmental Protection Agency, 1995). Alachlor was detected in 11 stream-flow samples at concentrations greater than the 2.0-µg/L MCL, with a maximum concentration of 28.6-Īg/L. Atrazine was detected in 15 streamflow samples at concentrations greater than the 3.0-µg/L MCL, with a maximum concentration of 135 µg/L. Although deethylatrazine was detected in almost every streamflow sample, no designated MCL for this metabolite has been established. The largest concentrations of alachlor and atrazine occurred in samples of runoff during the late spring and summer and probably are related to the application of these herbicides followed by runoff-producing rainfall.

Median concentrations for all but three herbicides (alachlor, atrazine, and metolachlor) and the metabolite deethylatrazine detected in streamflow samples from the Cheney Reservoir watershed were less than analytical method reporting limits (MRL's) (table 1). The three herbicides with median concentrations larger than their respective MRL's, however, were substantially less than their MCL's.

Median concentrations for alachlor, atrazine, and metolachlor varied substantially among the six sampling sites in the Cheney Reservoir watershed (fig. 3). The largest median concentrations for these three herbicides were from the Red Rock Creek subwatershed (sampling site 5, fig. 1) and were at least 24, 6.6, and 20 times larger (respectively) than median concentrations in water samples from any other sampling site.

Nineteen insecticides were analyzed for this study. Of those, six were detected in streamflow samples (table 1). The most frequently detected insecticide, carbofuran, was detected in 15 of 154 samples (10 percent). The other five insecticides were detected at frequencies ranging from 0.6 to 3.9 percent. Maximum concentra-tions ranged from 0.008 µg/L for azinphos (methyl-) to 1.16 µg/L for carbofuran. These maximum concentrations are substantially less than established MCL or HAL values. Insecticides that were not detected in streamflow samples are listed in table 2. MRL's ranged from 0.001 to 0.017 µg/L for these insecticides.

Conclusions and Implications

References