M. H. Johnson

IMPACT OF SWINEWASTE APPLICATION ON GROUND AND STREAM WATER QUALITY IN AN EASTERN COASTAL PLAIN WATERSHED

Nonpoint source pollution from agriculture has been a major concern, particularly where intensive agricultural operations exist near environmentally sensitive waters. To address these concerns, a water quality project was initiated in Duplin County, North Carolina, in the 2044-ha Herrings Marsh Run watershed. A swine farm within this monitored watershed expanded its operation from 3,300 to more than 14,000 animals. Groundwater nitrate-N increased significantly in three of the seven wells located adjacent to the spray field and in the adjoining riparian zone. Stream nitrate-N concentrations have increased after the expansion of the swine operation in the colder months, but concentrations have remained approximately the same during the warmer months. Stream ammonia-N mean concentrations after expansion have increased as well as the frequency and magnitude of ammonia-N concentration spikes. Ortho-phosphate concentrations in the stream water have been relatively consistent over the study period. The riparian zone is reducing the impact of spray field groundwater nitrate concentrations and ammonia loadings in an adjacent stream.

AMMONIA VOLATILIZATION FROM CONSTRUCTED WETLANDS THAT TREAT SWINE WASTEWATER

Increasingly, large–scale animal production occurs in confinement where large per–unit–area quantities of waste are generated. With the increased scale of production, new environment–friendly technologies are needed to deal with the waste. Constructed wetlands are considered an alternative treatment, but it is not known if volatilization of free ammonia (NH3) governs nitrogen removal in these systems. The objective of this research was to quantify the NH3 volatilization from constructed wetlands that treat swine wastewater. In May and July of 2000, a specially designed enclosure was used to measure NH3 volatilization from constructed wetlands receiving swine wastewater. Laboratory and field calibration tests indicated that the enclosure was effective at measuring NH3 volatilization. Wetland tests indicated that NH3 volatilization was occurring. From average hourly rates, it was estimated that 7% to 16% of the nitrogen load to the wetlands was removed through NH3 volatilization. Although NH3 losses should not be ignored, results indicated that NH3 volatilization was not responsible for removing the majority of nitrogen from the swine wastewater.

Flow-proportional, time-composited, and grab sample estimation of nitrogen export from an eastern coastal plain watershed

The balance between resources expended and information obtained is an integral aspect of water quality investigations. As part of a Water Quality Demonstration Project in the eastern Coastal Plain, we monitored stream water quality at the watershed outlet. Four methods of assessing stream water quality were compared. These methods were time-composite sampling with continuous flow measurements (TC), flow-proportional sampling with independent measurement of flow (FP), grab sampling with instantaneous flow measurements (IG), and grab sampling for quality assurance/quality control checks using daily USGS flow measurements (UG). Flow measurements using the TC and IG methods were highly correlated (r 2 = 0.97). Because of more intensive measurements during high flow, the FP method sampled greater flow rates during the sampling period. For all four methods, nitrate-N, ammonia-N and total Kjeldahl nitrogen (TKN) concentrations were not correlated to stream flow. Because of the significantly greater flow sampled, the FP method predicted significantly greater mass loading rates for both nitrate-N, ammonia-N, and TKN. Grab sampling (IG and UG) and the TC methods were not significantly different for the entire study period; however, a few monthly differences were significant. These results suggest that an appropriate sampling method should adequately weight sampling of both storm and base flows.

DISSOLVED PHOSPHORUS TRANSPORT DURING STORM AND BASE FLOW CONDITIONS FROM AN AGRICULTURALLY INTENSIVE SOUTHEASTERN COASTAL PLAIN WATERSHED

The high density of animal production in southeastern Coastal Plain watersheds has caused some soils to contain excess amounts of plant–available soil phosphorus (P). Runoff, erosion, and leaching can transport P to surface water systems and out of these watersheds. High P concentrations in downstream aquatic ecosystems can increase the risk of eutrophication. Our objectives were to determine stream dissolved phosphorus (DP) mass loads transported under storm and base flow conditions and to examine relationships between precipitation, stream flow, and DP concentrations and export loads from an agriculturally intensive Coastal Plain watershed. This watershed was separated into four subwatersheds, and stream flows at their outlets were separated into base and storm flow conditions. Over the 2–year study period, stream base flow accounted for the majority of total stream flow at all outlets (58% to 73%). Average stream total DP mass loads at the watershed outlet in 1994 and 1995 were 234 and 477 mg DP ha–1 d–1, and higher DP mass loads (57% to 71% of the cumulative total) were exported during base flow conditions. In 1995, a series of intense storm events over two months caused a large DP pulse (approximately 63% of the stream’s yearly annual DP mass load) to exit the watershed. Regression analysis showed a linear relationship (P < 0.001) between log10 instantaneous stream flow and log10 DP export. Our results showed that more DP was exported during stream base flow conditions. However, intensive summer storms can greatly accelerate stream DP export from this agriculturally intensive Coastal Plain watershed.

Measurement of greenhouse gas emissions from agricultural sites using open-path optical remote sensing method.

Improved characterization of distributed emission sources of greenhouse gases such as methane from concentrated animal feeding operations require more accurate methods. One promising method is recently used by the USEPA. It employs a vertical radial plume mapping (VRPM) algorithm using optical remote sensing techniques. We evaluated this method to estimate emission rates from simulated distributed methane sources. A scanning open-path tunable diode laser was used to collect path-integrated concentrations (PICs) along different optical paths on a vertical plane downwind of controlled methane releases. Each cycle consists of 3 ground-level PICs and 2 above ground PICs. Three- to 10-cycle moving averages were used to reconstruct mass equivalent concentration plum maps on the vertical plane. The VRPM algorithm estimated emission rates of methane along with meteorological and PIC data collected concomitantly under different atmospheric stability conditions. The derived emission rates compared well with actual released rates irrespective of atmospheric stability conditions. The maximum error was 22 percent when 3-cycle moving average PICs were used; however, it decreased to 11% when 10-cycle moving average PICs were used. Our validation results suggest that this new VRPM method may be used for improved estimations of greenhouse gas emission from a variety of agricultural sources.

Forage Subsurface Drip Irrigation Using Treated Swine Wastewater

The rapid expansion of animal production in the eastern U.S. in the 1990s resulted in large quantities of concentrated animal waste that must be utilized or disposed of in an efficient and environmentally friendly manner. To address these environmental concerns for wastewater utilization, we installed a subsurface drip irrigation system to apply treated swine wastewater effluent for bermudagrass hay production. The overall study objective was to determine the feasibility of using subsurface drip irrigation (SDI) for treated wastewater effluent applications. The specific objectives for the SDI system were to compare bermudagrass hay production using (1) commercial and wastewater effluent for nutrients, (2) two SDI lateral spacings (0.6 and 1.2 m) installed at 0.3 m below the surface, and (3) two irrigation application rates based on calculated evapotranspiration (ETc) requirements (75% or 100%). The two-year study measured hay yields, hay biomass, soil nutrients, and soil water nutrients. The SDI system was successfully operated for two years applying effluent and commercial fertilizer to supply the nutrient requirements of the bermudagrass hay. Bermudagrass hay production for 2004 and 2005 ranged from 5.65 to 14 Mg ha-1. Results from the SDI system indicated no significant differences between the SDI lateral spacings or irrigation application rates. Treatments using wastewater effluent had significantly higher hay yields and significantly higher nutrient biomass removal rates than the commercial fertilizer treatments. Nitrate-N observed in soil water lysimeters increased with depth, indicating the potential for leaching without proper management. Soil nitrogen and carbon were not significantly different for any of the treatments but did vary slightly over the life of the project.

NITRATE-N DISTRIBUTION AND TRENDS IN SHALLOW GROUNDWATER ON AN EASTERN COASTAL PLAINSWATERSHED

Nonpoint source pollution from agriculture has been a major concern, particularly where intensive agricultural operations exist near environmentally sensitive waters. To address these nonpoint source pollution concerns, a Water Quality Demonstration Project (WQDP) was initiated on the Herrings Marsh Run (HMR) watershed in Duplin County, North Carolina. The WQDP was implemented to determine water quality benefits from voluntary adoption of improved management practices. In the WQDP, 84 groundwater monitoring well sites were established on 21 farms selected to represent the major farming practices on the watershed. On the HMR watershed, nitrate-N contamination of groundwater was not a wide spread problem. Seventy-four percent of the groundwater monitoring sites had nitrate-N less than the drinking water standard of 10 mg/L. Mean nitrate-N concentrations were below 10 mg/L on 16 of the 21 farms. Of the four farms with nitrate-N exceeding 10 mg/L, one farm had mean nitrate-N that exceeded 20 mg/L. This farm had an undersized and overloaded swine wastewater spray field. After the spray field was expanded and application rates were reduced, groundwater nitrate-N concentrations declined; but they continued to exceed 20 mg/L. Other farms with swine waste spray fields had mean groundwater nitrate-N concentrations <20 mg/L throughout the study period. Groundwater nitrate-N concentrations under row crops were <10 mg/L on all but two farms. Three of the four farms with nitrate-N concentrations exceeding 10 mg/L were in a subwatershed of the HMR that had the highest concentration of animal waste application and excess nitrogen applied. Of the 21 farms, three farms had a significant increasing trend in groundwater nitrate-N while four farms had a significant decreasing trend. The overloaded swine wastewater spray field had a significant decreasing nitrate-N trend. Most farms with concentrations less than 10 mg/L had no detectable trend in nitrate-N concentration during the study. These findings indicate that nitrate-N contamination of groundwater is not a widespread problem on the HMR watershed even though it is intensively farmed.

GLEAMS simulation of groundwater nitrate-N from row crop and swine wastewater spray fields in the eastern coastal plain

Nonpoint source pollution of surface and groundwater resulting from agricultural management practices is a major water quality problem. This problem was assessed on a demonstration watershed in the Cape Fear River Basin of North Carolina, during a five-year study. Groundwater was monitored in a row crop field (corn/wheat/soybean) and a swine waste spray field (Coastal bermuda grass). Groundwater nitrate-N concentrations averaged 6.5 mg/L in the row crop field. Nitrate-N concentrations in groundwater at the swine waste spray field exceeded 80 mg/L. Nitrate-N concentrations were simulated in both fields with the GLEAMS model. The GLEAMS model simulated groundwater nitrate-N concentrations with mean residuals (simulated-observed) ±1.3 mg/L and ±19 mg/L, respectively, for the row crop and the swine waste spray field. Groundwater nitrate-N concentrations have been reduced in the spray field by using improved management practices and the GLEAMS model simulated this nitrate-N concentration reduction. These simulation results show that the GLEAMS model can be used to predict nitrate-N loading of groundwater of these agricultural management systems.

Optimal Sensor Locations for the Backward Lagrangian Stochastic Technique in Measuring Lagoon Gas Emission

This study evaluated the impact of gas concentration and wind sensor locations on the accuracy of measuring gas emission rates from a lagoon environment using the backward Lagrangian stochastic (bLS) inverse-dispersion technique. Path-integrated concentrations (PICs) and three-dimensional (3D) wind vector data were collected at different locations within the lagoon landscape. A floating 45 m × 45 m perforated pipe network on an irrigation pond was used as a synthetic distributed emission source for the controlled release of methane. A total of 961 15-min datasets were collected under different atmospheric stability conditions over a 2-yr period. The PIC location had a significant impact on the accuracy of the bLS technique. The location of the 3D sonic anemometer was generally not a factor for the measured accuracies with the PIC positioned on the downwind berm. The PICs across the middle of the pond consistently produced the lowest accuracy with any of the 3D anemometer locations (<69% accuracy). The PICs located on the downwind berm consistently yielded the best bLS accuracy regardless of whether the 3D sonic anemometer was located on the upwind, side, or downwind berm (accuracies ranged from 79 to 108%). The accuracies of the emission measurements with the berm PIC-berm 3D setting were statistically similar to that found in a more ideal homogeneous grass field. Considering the practical difficulties of setting up equipment and the accuracies associated with various sensor locations, we recommend that wind and concentration sensors be located on the downwind berm.

Oxygen transfer in marsh-pond-marsh constructed wetlands treating swine wastewater

Oxygen transfer efficiencies of various components of the marsh-pond-marsh (M-P-M) and marsh-floating bed-marsh (M-FB-M) wetlands treating swine wastewater were determined by performing oxygen mass balance around the wetlands. Biological oxygen demand (BOD) and total nitrogen (TN) loading and escaping rates from each wetland were used to calculate carbonaceous and nitrogenous oxygen demands. Ammonia emissions were measured using a wind tunnel. Oxygen transfer efficiencies of the aerated ponds were estimated by conducting the ASCE standard oxygen transfer test in a tank using the same aeration device. Covering pond water surface with the floating bed slightly decreased oxygen transfer efficiency. The diffused membrane aeration (26.7 kg O2 ha-1 d-1) of M-P-M was surprisingly not as effective as plant aeration in the marsh (38.9 to 42.0 kg O2 ha-1 d-1). This unusually low oxygen transfer efficiency of the diffused aeration was attributed to its low submergence depth of 0.8 m compared to typical depth of 4.5 m. The wetlands consisting entirely of marsh removed similar amounts of C and N without investing additional equipment and energy costs of aerating ponds in the middle of wetlands.