Use of Real-Time Water-Quality Monitoring and Multiple-Regression Analysis for Estimation of Atrazine Loads in Streamflow, South-Central Kansas

Abstract

Atrazine, one of the triazine herbicides, is of particular concern in the United States because of potential effects on water quality as it relates to total maximum daily loads (TMDLs) and water-quality standards in streams. Reliable estimates of atrazine loads (mass transport) are important because, by combining streamflow and concentration, they provide a better indication of how much atrazine is being transported by a stream than by looking at concentrations alone. For this study, results of real-time monitoring of selected water-quality constituents will be used to help estimate long-term atrazine loads.

To optomize sample collection during runoff, automated samplers and a set of probes for specific conductance, pH, temperature, dissolved oxygen, and turbidity were installed at a U.S. Geological Survey stream-gaging station on the Little Arkansas River near Sedgwick, south-central Kansas. The use of automated equipment for sample collection and enzyme-linked immunosorbent assay (ELISA) for herbicide analysis allowed for the low cost and efficient collection of about 1,200 samples from 1995 through 1998. The large number of samples were needed because of the high degree of temporal variability. Analysis turnaround time for the ELISA procedure was about 2 days. The results of this procedure were reported as atrazine because gas chromatoraphy/mass spectrometry (GC/MS) confirmation of selected analytical results indicated that at least 80 percent of the triazine-herbicide concentration determined by ELISA was atrazine.

Although ELISA is a cost-effective and efficient method of detecting atrazine, it would be useful to estimate atrazine concentrations and loads by using other constituents that are monitored continuously in-stream such as discharge, specific conductance, and dissolved oxygen, which are related to atrazine because of their association with runoff events, time, temperature, and/or biological components. The development of a nonlinear least-square, multiple-regression equation would allow the determination of a direct or indirect relation between atrazine and the other constituents and would enable the estimation of concentrations and loads.

Additional information about Real-Time Water-Quality in Kansas can be found at: http://ks.water.usgs.gov/studies/qw/

Christensen, V.G., Jian, Xiaodong, and Ziegler, A.C., 1999, Use of real-time water-quality monitoring and multiple-regression analysis for estimation of atrazine loads in streamflow [abst.], in Abstracts of the Seventh Symposium on Chemistry and Fate of Modern Pesticides, September 14-16, 1999, Lawrence, Kansas: International Association of Environmental Analytical Chemistry, p. 89.