WATER QUALITY FROM THE BLUEGRASS WATERSHED, AUDUBON
L.S. Seigley, G.R. Hallberg, and G.A. Miller
Iowa Department of Natural Resources, Geological Survey Bureau,
Technical Information Series 26, 1993, 30 p.
From 1987 to 1992, surface water and groundwater quality was monitored in Bluegrass watershed, a 1,024-acre watershed in north-central Audubon County. Groundwater from shallow monitoring wells, active private wells, inactive private wells, and tile lines, and surface water from Bluegrass Creek were sampled for bacteria, nitrate, and pesticides to determine water quality in and around the watershed. Sites were sampled monthly. Rainfall was collected after rainfall events and analyzed for pesticides. The majority of the rural population in western Iowa use shallow (20 to 40 feet deep), large-diameter (three to four feet) "seepage" wells located in the surficial deposits. The groundwater yielded by these wells is susceptible to contamination from modern land-surface activities.
Total coliform and fecal coliform bacteria samples were collected from the actively used private wells. Ninety-two percent of the samples had total coliform bacteria present. The total coliform bacteria results showed no seasonal trends. Thirty-one percent of the samples from the active wells were positive for fecal coliform bacteria.
Results of the nitrate monitoring showed several trends related to landuse, climate, geological materials, and landscape position. Nitrate-N results from the monitoring wells were related to topographic position of the well and the geologic materials in which the wells were screened. Landuse around the wells had minimal effect on nitrate-N concentrations. Wells situated in the uplands in loess (generally well-drained soils) had the highest nitrate-N concentrations, generally greater than 20 mg/L. Wells in Roberts Creek Member alluvium had the lowest nitrate-N concentrations, generally less than 0.2 mg/L. The low nitrate-N concentrations are likely the result of denitrification.
Nitrate-N concentrations from the active private wells declined through time, from 6.7 mg/L in 1987 to 2.7 mg/L in 1992. Annual median concentrations for the individual wells varied, ranging from less than 0.2 mg/L to 28.4 mg/L. The variability in nitrate-N concentrations among the wells is related to landscape position, the geologic materials in which the wells are situated, and land management around the wells. Well W1 showed the most significant decline in nitrate-N concentrations of all monitored sites. Median nitrate-N concentrations at well W1 declined from 14.9 mg/L in 1987 to 0.9 mg/L in 1990, 1991, and 1992. In 1987, the land surrounding W1 was taken out of row-crop production and placed in the Conservation Reserve Program. The dramatic decline in nitrate-N concentrations continued after the drought years of 1988 and 1989, and suggests that landuse caused this trend.
The inactive private wells were not used on a regular basis, and results from these wells could not be used to analyze trends in nitrate-N concentrations to landuse changes.
The tile lines had the highest annual median nitrate-N concentrations, varying from 13.1 mg/L to 15.6 mg/L. Annual concentrations for tile lines as a category declined from 1987 to 1989, increased in 1990, and declined from 1990 to 1992. Except for major changes in landuse, it was difficult to relate changes in nitrate-N concentrations from individual tile lines to changes in landuse practices. Nitrate-N concentrations from Bluegrass Creek showed trends similar to the tile lines, although annual median concentrations were lower than concentrations for the tile lines. For Bluegrass Creek, nitrate-N concentrations declined from 1987 to 1989, increased in 1990, and declined from 1990 to 1992.
Data from the pesticide monitoring showed few if any trends related to landuse, climate, geological materials, and landscape position. Unlike the nitrate data, the monitoring wells showed no trends between the pesticide data and the geologic materials in which the wells were situated nor the surrounding landuse. Herbicides were still detected in settings where denitrification removed the nitrate. Five herbicides were detected in the monitoring wells (alachlor, atrazine, cyanazine, metolachlor, and metribuzin) at concentrations ranging from 0.10 to 0.43 µg/L. Six herbicides were detected in the active private wells (alachlor, atrazine, cyanazine, metolachlor, pendimethalin, and trifluralin) at concentrations from 0.10 to 24.00 µg/L. In addition to those herbicides detected in the active private wells, metribuzin was also detected in the inactive wells. Concentrations of detected pesticides from the inactive wells ranged from 0.10 to 28.00 µg/L. Alachlor, atrazine, cyanazine, metolachlor, pendimethalin, and trifluralin were detected in the tile lines at concentrations from 0.10 to 7.40 µg/L. Five herbicides (alachlor, atrazine, cyanazine, metolachlor, and pendimethalin) were detected in Bluegrass Creek at concentrations from 0.10 to 4.60 µg/L. Precipitation in Bluegrass watershed was monitored for pesticides from May 1989 through September 1992. Eight herbicides (alachlor, atrazine, cyanazine, eptam, metolachlor, metribuzin, pendimethalin, and propachlor) were detected at concentrations from 0.10 to 28.00 µg/L. The majority of detections were less than 1.00 µg/L. The herbicides detected in rainfall showed a relationship to pesticide use in the state. Atrazine, alachlor, and cyanazine are three of the most commonly used herbicides and were the most commonly detected in Bluegrass precipitation. Insecticides are less frequently used, and no insecticides were detected in the rainfall. According to farm survey information from Bluegrass watershed, all of the herbicides detected in rainfall, except propachlor, were used in the watershed.