Evaluation of agricultural best-management practices in the Conestoga River headwaters, Pennsylvania: Characterization of surface-runoff and ground-water quantity and quality in a small carbonate basin near Churchtown, Pennsylvania, prior to terracing and implementation of nutrient management: Water-quality study of the Conestoga River headwaters, Pennsylvania

Abstract

The U.S. Geological Survey, in cooperation with the Pennsylvania Department of Environmental Protection1 , conducted a study as part of the U.S. Department of Agriculture's Rural Clean Water Program to determine the effects of agricultural best-management practices on surface-water and ground-water quality in the Conestoga River headwaters basin. This report describes Field-Site 1 and characterizes the surface-runoff and ground-water quantity and quality at the site from January 1983 through September 1984, before the implementation of terracing and nutrient-management best-management practices. The 22.1-acre site, part of two dairy farms, was cropland used primarily for the production of corn and alfalfa, and is underlain by carbonate rock. During the 21-month study period, 91.2 inches of precipitation fell, of which 66 percent occurred during the 1984 water year. Of the 169 storms of 0.10 inch or larger, 97 produced measurable runoff. The average annual application of nutrients to the 14.4 acres of cornfields was 410 pounds per acre of nitrogen and 110 pounds per acre of phosphorus. About three times more nutrients were applied during 1984 water year than during the 1983 water year. The approximate soil-nitrate concentration as nitrogen for the study period was 8.2 milligrams per kilogram in the top 8 inches of soil. Runoff for the study period totalled 714,000 cubic feet, or 9.8 percent of total precipitation. Eighty-eight percent of the runoff occurred during the 1984 water year. Regression analyses indicate that total runoff was controlled primarily by total precipitation amounts during storms and by antecedent moisture conditions; mean event and maximum instantaneous discharges are controlled by precipitation intensity and by antecedent moisture conditions. Regression analyses also suggest that crop cover on the corn acreage affected runoff amounts and rates. Mean concentrations in runoff ranged from 24 to 73,000 milligrams per liter for suspended sediment, 0.45 to 24 milligrams per liter for total phosphorus, and 1.5 to 55 milligrams per liter for total nitrogen. Of the total nitrogen in instantaneous runoff samples, a median of 72 percent was total organic nitrogen. Results of multiple regression analyses suggest that mean nitrogen concentrations in runoff (1) increased with increased surface-nitrogen applications made prior to runoff, and (2) were diluted with increased precipitation duration. Runoff from storms on frozen ground produced the highest loads of nitrogen in runoff. Over the 21-month study period, an estimated 258 tons of suspended sediment, 314 pounds of nitrogen, and 176 pounds of phosphorus were transported with runoff from the site. Of this, 88 percent of the nitrogen and phosphorus loads and 94 percent of the suspended-sediment load were discharged during the wet 1984 water year. The loads for the study period represent 2.5 percent of the total nitrogen and 5.5 percent of the total phosphorus applied to the site as manure and commercial fertilizer. The ground-water basin at the site is slightly larger than the surface-water basin. Ground-water levels, except at one well, responded quickly to recharge; water levels peaked several hours to a day after precipitation. Surface-applied materials moved rapidly to the water table through macropores and slowly by transport through micropores in the unsaturated zone. Median concentrations of dissolved nitrate ranged from 9.2 to 13 milligrams per liter as nitrogen for ground-water samples collected monthly from five wells and a spring. Dissolved nitrate comprised 93 percent (median percentage) of the total nitrogen in ground-water samples. The effect of recharge on ground-water nitrate concentrations was affected by variations in nutrient availability at the land surface and in the unsaturated zone. A lag time of 1 to 3 months was observed between the time that nitrogen was applied to the land surface and local maximums in nitrate concentrations were detected in ground water unaffected by recharge events. Approximately 3,187,000 cubic feet of ground water and an associated 2,200 pounds of nitrate-nitrogen discharged from the site during the study period. For the study period, 42 percent of the precipitation recharged to ground water, 10 percent became runoff, and 48 percent evapotranspired. Inputs of nitrogen to the study area were estimated to be 93 percent from manure, 5 percent from commercial fertilizer, and 2 percent from precipitation. Nitrogen outputs from the system were estimated to be 38 percent to crop uptake, 39 percent to volatilization, 20 percent to ground-water discharge, and 3 percent to surface runoff.