Characterization and control of agricultural nonpoint source pollution.

Abstract

Saline Valley Rural Clean Water Project was one of twenty national projects developed to evaluate methods for controlling agricultural nonpoint source pollution. The project found a volunteer-based land treatment approach inadequate for meeting participation and phosphorus reduction goals. Low participation, high variability in pollutant loading, and a limited monitoring design prevented the project from documenting relationships between applied best management practices and water quality improvements. An over-estimate of the severity of nonpoint phosphorus loading from the watershed further limited the project's chances for success, although monitoring data revealed a 56 percent decrease in total phosphorus loading occurred as a result of point-source controls. Temporal and spatial variability in pollutant export were quantified to evaluate water quality monitoring results. Cumulative loading distributions indicated that on average, 76, 56, and 51 percent of annual loads for solids, total-P, and soluble-P occurred within 28 days. Daily monitoring of storms revealed that the project's weekly sampling scheme resulted in extreme loading errors. Loading adjustments, based on a daily precipitation record, indicated annual mean loads were under estimated by 81, 66, 53, and 54 percent for solids, total-P, soluble-P, and nitrate. Loading patterns were also highly variable among stations. Results suggest that monitoring within sub-basins would improve chances of detecting water quality changes resulting from land treatment application. Nutrient-enrichment experiments were performed to examine transformations in pollutant fluxes during transport. Mass balance calculations of phosphorus and ammonia fluxes indicated that within a 2.5 km reach, 48 percent of the pulsed load was removed, 28 percent dispersed, and 24 percent remained within the initial pulse. Removal rates ranged from 17 to 46 gm hr$\sp{-1}$ for phosphorus and from 4 to 80 gm hr$\sp{-1}$ for ammonia. Removal rates varied seasonally and in response to changes in the stream's substrate following stormflow. Nutrient cycling during transport significantly affected pollutant fluxes and may obscure changes in export from the watershed.