Herbicides in the Pecatonica and Yahara Rivers in Southwestern Wisconsin, May 1996-July 1996

¹ U.S. Geological Survey
² Wisconsin Department of Agriculture, Trade, and Consumer Protection

Note: This fact sheet is also available in Acrobat (.pdf) format: Download the PDF file (350 Kbytes)

Introduction

Figure 1. Location of sampling sites for herbicides.

Table 1. Land use characteristics for the two herbicide sampling watersheds
(land use data from DATCP; Ag, agriculture; Other includes forests, lakes, wetlands,
and urban areas; APA, Atrazine Prohibition Area)
--------------------------------------------------------------------------------------
                     USGS Station    Drainage Area
Station name           Number        (square miles)   Percent land use or land cover
--------------------------------------------------------------------------------------
                                                        Ag        Other        APA
                                                      --------------------------------
Pecatonica River     05434500           1,034           96         4            4
  at Martintown

Yahara River         05430175             517           75        25           94
  near Fulton
--------------------------------------------------------------------------------------

Table 2. Estimate pounds of pesticide active ingredient applied during the
spring of 1996 (application data from DATCP)
--------------------------------------------------------------------------------------
                     Estimated pounds of pesticide active ingredient applied in
   Watershed                            the spring of 1996
--------------------------------------------------------------------------------------
                   Acetochlor    Alachlor    Atrazine    Metolachlor    Cyanazine
                  --------------------------------------------------------------------
Pecatonica R.       42,000        32,000      85,000      106,000          88,000

Yahara River        26,000        20,000       3,200       66,000          56,000
--------------------------------------------------------------------------------------

Samples were collected at each site weekly from May 15 to July 15, 1996, without regard to streamflow conditions at the sampling site. When the instantaneous discharge of the stream increased by at least 50 ft3/s (cubic feet per second) more than the mean daily discharge of the previous day, additional daily storm runoff samples were collected until the streamflow decreased.

Samples were collected with a US D-77 water-quality sampler using the Equal-Width-Increment (EWI) method (Edwards and Glysson, 1988). The sample was transferred to a teflon-lined sample splitter and then split into two pre-cleaned, one-liter brown glass bottles. The samples were analyzed for commonly used herbicides by the DATCP water-quality laboratory. The sampler and churn splitter were cleaned by methods outlined in Shelton (1994). For quality assurance (QA), samples of deionized (MilliRO and MilliQ) water were processed through the sampler and splitting equipment and then analyzed for herbicides by the DATCP laboratory. These QA samples, collected and processed near the beginning and at the end of the sampling periods, contained no detectable concentrations of the herbicides analyzed in this study.

Results

Figure 2. Herbicide concentration, minimum detection concentration, number of samples above the minimum detection concentration, and number of samples collected at the Pecatonica and Yahara Rivers.

Data for the Yahara River show that atrazine and acetochlor were the most frequently detected herbicides. Fifty-seven percent of the samples had concentrations of these herbicides above the laboratory analytical detection limit (fig. 2). Cyanazine and metolachlor were the next most detected herbicides; 43 percent of the samples were above the laboratory analytical detection limit. Alachlor and alachlor ESA were detected in 29 percent and 14 percent, respectively, of 14 samples collected.

The maximum herbicide concentration for the Pecatonica River was 8.07 µg/L (micrograms per liter) for cyanazine on June 17, 1996 (fig. 2 and fig. 3). The next highest concentration detected was for atrazine (with metabolites)-5.8 µg/L on May 29 (fig. 2 and fig. 3). The median concentrations of herbicides in the Pecatonica River samples were 0.90 µg/L atrazine, 0.86 µg/L metolachlor, 0.61 µg/L cyanazine, and 0.27 µg/L acetochlor. Median concentrations are reported here only for herbicides for which less than 50 percent of the sample concentrations were below the detection limit.

Figure 3. Discharge hydrographs, herbicide concentration, and period of herbicide concentration for the Pecatonica and Yahara Rivers from May 15 to July 15, 1996. (Laboratory analytical detection limit for acetochlor is 0.10 µg/L, for atrazine and alachlor is 0.15 µ'g/L; for metolachlor is 0.25 µg/L; for cyanazine is 0.50 µg/L and for alachlor ESA is 1.0 µg/L.)

In comparison, the median concentration of atrazine for three predominantly agricultural sites in eastern Wisconsin-Duck Creek, North Branch Milwaukee River, and the Milwaukee River-ranged from approximately 0.04 µg/L to 0.30 µg/L (Sullivan and Richards, 1996). Median cyanazine and metolachlor concentrations at the Pecatonica River were also higher than at these three sites.

At the Yahara River, atrazine (with metabolites) had the highest concentration-9.93 µg/L on June 19 (fig. 2 and fig. 3). Cyanazine had the next highest concentration-7.43 µg/L on June 17. The maximum atrazine concentration detected in samples from the Yahara River was greater than that detected in eastern Wisconsin streams, but less than that for streams monitored in Indiana and Nebraska (Sullivan and Richards, 1996). The median concentrations of herbicides in the Yahara River samples were 0.18 µg/L for atrazine, and 0.17 µg/L for acetochlor. The median atrazine concentrations for the Yahara River samples were within the range of median atrazine concentrations for the three streams in eastern Wisconsin mentioned above.

Herbicide concentrations in the Pecatonica and Yahara Rivers differed from those in other streams in Wisconsin and streams in other midwestern states. The maximum cyanazine concentration for the Pecatonica and the Yahara Rivers was greater than that detected in streams in eastern Wisconsin, Indiana, and Nebraska. The median atrazine concentrations for both the Pecatonica and Yahara Rivers were less than the median concentrations for watersheds in Indiana and Nebraska that drain extensive corn and soybean agricultural areas. Watersheds in Minnesota and Missouri, where grazing and small grain farming are the major agricultural activities, had lower median atrazine concentrations than either the Yahara or Pecatonica Rivers. The median cyanazine concentration for the Pecatonica River was greater than that detected in streams in Indiana, Nebraska, Minnesota, and Missouri. The Pecatonica had lower median metolachlor concentrations however, compared to a Nebraska watershed (Sullivan and Richards, 1996).

The concentration of herbicides that are transported to a stream is a function of the physiochemical properties, such as water solubility, soil half-life, and soil sorption coefficient of the herbicides; timing and rate of the herbicide application; and overland runoff from rainfall and farming practices (Goolsby and Pereira, 1995). Peak herbicide concentration in streams is dependent upon the timing of the herbicide application and when rainfall occurs in amounts great enough to generate overland runoff. Concentrations of all herbicides were less than the laboratory analytical detection limit in late June and continued to stay below the analytical detection limit until data collection ended in July. The streamflow and the concentrations of six of the herbicides detected in samples collected at the Pecatonica and the Yahara Rivers and the approximate period of herbicide application are shown in figure 3.

At the Pecatonica River, the peak concentration of atrazine was detected in late May to early June, whereas the peak cyanazine and metolachlor concentrations occurred during a major runoff period in the middle of June (fig. 3). The monthly mean streamflow for the Pecatonica River at Martintown during May was 22 percent greater than the mean monthly flow for the period of record (1940-95) (Holmstrom and others, 1997). During June, streamflow was 122 percent greater and during July, it was 127 percent greater than mean monthly flow for the period of record.

At the Yahara River, atrazine, cyanazine, and metolachlor concentrations peaked in the middle of June (fig. 3). The monthly mean streamflow at the Yahara River during May was 20 percent greater than the mean monthly flow for the period of record (1977-95); during June it was 224 percent greater and during July it was 99 percent greater (Holmstrom and others, 1997).

The load for 5 of the 6 herbicides studied was determined for both rivers. Herbicide load is the mass transport of the herbicide by the river and is a function of water discharge and concentration of the herbicide. The integration method was used to determine the load at each site (Porterfield, 1972). For samples that contained herbicide concentrations below the laboratory analytical detection limit, the concentrations were set to one-half the analytical detection limit for the purpose of load calculation. The daily loads were summed for the period May 15 to July 15 (fig. 4). The calculated herbicide loads at the Pecatonica River ranged from 47.2 lbs (pounds) of alachlor to 484 lbs for atrazine with metabolites. For the Yahara River, loads ranged from 36.1 lbs of alachlor to 289 lbs for atrazine with metabolites (fig. 4). The unit area loads (load divided by drainage area) were greater in the Yahara River watershed than in the Pecatonica River watershed for all the herbicides except for metolachlor (fig. 4). The unit area load for the Pecatonica River ranged from 0.05 lbs/mi2 (pounds per square mile) for alachlor to 0.47 lbs/mi2 for atrazine with metabolites. For the Yahara River, unit area loads ranged from 0.07 lbs/mi2 for alachlor to 0.55 lbs/mi for atrazine with metabolites (fig. 4). For the most part, the unit area loads of herbicides in the Pecatonica and Yahara Rivers were consistent with the application rates of herbicides for each watershed except for atrazine (table 2, fig. 4). Alachlor and acetochlor were the least-used herbicides and had the lowest unit area loads for both watersheds (fig. 4). In the Yahara River watershed, metolachlor was applied at rates higher than cyanazine but the unit area load for metolachlor was less than that for cyanazine (fig. 4). In the Pecatonica River watershed, metolachlor was the most heavily applied herbicide and had the second highest unit area load (fig. 4). In both watersheds, atrazine had the highest unit area load as compared to the other herbicides but was the least-applied herbicide in the Yahara River watershed and the third-most herbicide applied in the Pecatonica River watershed (table 2, fig. 4).

Ninety-four percent of the Yahara River watershed is in the atrazine prohibition area (fig. 1). Because much less atrazine is applied in this watershed compared to the Pecatonica River watershed (table 2), the unit area loads would be expected to be less. It is important to note, however, that the area just upstream from the sampling site is not in the atrazine prohibition area (fig. 1). Thus the high concentrations of atrazine detected in June-9.93 µg/L and 6.25 µg/L-may be a result of herbicides applied near the sampling site, and may not be representative of the whole watershed.

The median atrazine concentration for the Yahara River samples was less than that for the Pecatonica River samples. The higher unit area load for the Yahara River may be a result of the extreme streamflow in June (224 percent greater than the long-term mean monthly flow for June). Seventy-nine percent of the atrazine load for the study period was transported in 5 days, or 8 percent of the days monitored at the Yahara River.

Conclusions