As part of the U.S. Geological Survey National Water-Quality Assessment Program, 32 monitoring wells were installed near cropland in parts of northern Alabama and Middle Tennessee to characterize the effect of row-crop agriculture on shallow ground-water quality. The wells were completed in regolith overlying carbonate bedrock. These geologic units are part of the Mississippian carbonate aquifer, a source of drinking water for domestic and municipal supply in the area. The majority of these wells were sampled in the spring of 2000 for inorganic constituents, nutrients, pesticides, and selected pesticide degradates. Land use and soil characteristics were delineated for a 1,640-foot radius buffer area around each well to relate water quality to environmental factors. A strong association among soil characteristics, land use, and hydrogeology limited the analysis of the effect of these factors on nitrate and pesticide occurrence.
Nitrate and pesticide concentrations generally were low, and no samples exceeded established drinking-water maximum contaminant levels. The maximum concentration of nitrate was about 8 milligrams per liter as nitrogen, and the median concentration was 1 milligram per liter. Nitrate concentrations were strongly correlated to dissolved-oxygen concentrations, and ratios of chloride to nitrate indicate nitrate concentrations were affected by denitrification in about a third of the samples. A pesticide or pesticide degradate was detected at concentrations greater than 0.01 microgram per liter in 91 percent of the samples. Pesticides with the highest use typically were detected most frequently and at the highest concentrations; however, glyphosate had the highest estimated use but was not detected in any samples. Fluometuron and atrazine, two high-use pesticides, were detected in 83 and 70 percent, respectively, of the samples from wells where the pesticide was applied in the buffer area. Maximum concentrations of fluometuron and atrazine were 2.13 and 1.83 micrograms per liter, respectively. Detection rates of pesticide degradates were similar to parent pesticides, and concentrations of degradates generally were comparable to or greater than the parent pesticide. Pesticide detections were correlated to dissolved-oxygen concentrations, suggesting that pesticides are most likely to be detected at high concentrations where ground-water residence time is short and the rate of recharge is fast.
Nitrate and pesticide data collected in this study were compared to data collected from similar agricultural land-use studies conducted by the National Water-Quality Assessment Program throughout the Nation. Nitrate concentrations generally were lower in this study than in samples from other agricultural areas; however, pesticides were detected more frequently in samples from wells in this study. For example, atrazine and its degradate, deethylatrazine, were detected in 62 and 47 percent, respectively, of water samples in this study but were detected in about 25 percent of the 851 wells sampled for agricultural land-use studies nationwide. In national study areas where atrazine use is greater than in the lower Tennessee River Basin, atrazine was detected in 30 percent of the water samples. Pesticides used on cotton were detected much more frequently in this study, but many of the study areas nationwide have smaller amounts of cotton acreage than the lower Tennessee River Basin.
Similarities in nitrate concentrations and the pesticides detected frequently in this agricultural land-use study and a network of drinking-water wells in the same area completed in bedrock in the Mississippian carbonate aquifer (sampled in a previous study) indicate the aquifer is susceptible to contamination from nonpoint sources. Nitrate concentrations were not statistically different for the two well networks and were correlated to total pesticide concentrations in both networks. Although detection frequencies and maximum concentrations