Abstract

This report presents the results of a study by the U.S. Geological Survey, in cooperation with the Kentucky Department of Agriculture, on nutrients, select pesticides, and suspended sediment in the karst terrane of the Sinking Creek Basin.

Streamflow, nutrient, select pesticide, and suspended-sediment data were collected at seven sampling stations from 2004 through 2006. Concentrations of nitrite plus nitrate ranged from 0.21 to 4.9 milligrams per liter (mg/L) at the seven stations. The median concentration of nitrite plus nitrate for all stations sampled was 1.6 mg/L. Total phosphorus concentrations were greater than 0.1 mg/L, the U.S. Environmental Protection Agency’s recommended maximum concentration, in 45 percent of the samples. Concentrations of orthophosphates ranged from less than 0.006 to 0.46 mg/L. Concentrations of nutrients generally were larger during spring and summer months, corresponding to periods of increased fertilizer application on agricultural lands. Concentrations of suspended sediment ranged from 1.0 to 1,490 mg/L at the seven stations. Of the 47 pesticides analyzed, 14 were detected above the adjusted method reporting level of 0.01 micrograms per liter (μg/L). Although these pesticides were detected in water-quality samples, they generally were found at less than part-per-billion concentrations. Atrazine was the only pesticide detected at concentrations greater than U.S. Environmental Protection Agency drinking water standard of 3 μg/L, and the maximum detected concentration was 24.6 μg/L.

Loads and yields of nutrients, selected pesticides, and suspended sediment were estimated at two mainstream stations on Sinking Creek, a headwater station (Sinking Creek at Rosetta) and a station at the basin outlet (Sinking Creek near Lodiburg). Mean daily streamflow data were available for the estimation of loads and yields from a stream gage at the basin outlet station; however, only periodic instantaneous flow measurements were available for the headwaters station; mean daily flows at the headwater station were, therefore, estimated using a mathematical record-extension technique known as the Maintenance of Variance-Extension, type 1 (MOVE.1). The estimation of mean daily streamflows introduced a large amount of uncertainty into the loads and yields estimates at the headwater station.

Total estimated loads of select (five most commonly detected) pesticides from the Sinking Creek Basin were about 0.01 to 1.2 percent of the estimated application, indicating pesticides possibly are retained within the watershed. Mean annual loads [(in/lb)/yr] for nutrients and suspended sediment were estimated at the two Sinking Creek mainstem sampling stations. The relation between estimated and measured instantaneous loads of nitrite plus nitrate at the Sinking Creek near Lodiburg station indicate a reasonably tight distribution over the range of loads. The model for loads of nitrite plus nitrate at the Sinking Creek at Rosetta station indicates small loads were overestimated and underestimated. Relations between estimated and measured loads of total phosphorus and orthophosphate at both Sinking Creek mainstem stations showed similar patterns to the loads of nitrite plus nitrate at each respective station. The estimated mean annual load of suspended sediment is about 14 times larger at the Sinking Creek near Lodiburg station than at the Sinking Creek near Rosetta station.

Estimated yields of nutrients and suspended sediment increased from the headwater to downstream monitoring stations on Sinking Creek. This finding suggests that sources of nutrients and suspended sediment are not evenly distributed throughout the karst terrane of the Sinking Creek Basin. Yields of select pesticides generally were similar from the headwater to downstream monitoring stations. However, the estimated yield of atrazine was about five times higher at the downstream station on Sinking Creek than at the headwater station on Sinking Creek. A predominantly cultivated agricultural land area of the karst drainage basin drains into Sinking Creek just downstream of the headwater station. Because the daily mean streamflow was estimated at the headwater monitoring station, the error in the estimated nutrient, select pesticide, and suspended-sediment loads and yields are subject to considerable and unknown biases and imprecision (greater standard error of predictions than reported). Additional streamflow and water-quality data are needed to improve the reliability of the load estimates and the errors associated with them at the upstream and downstream stations on Sinking Creek.

First posted August 25, 2010