Nitrous Oxide Loss from Poultry Manure-Amended Soil after Rain

Land application of poultry wastes in Kentucky will increase as the broiler industry grows. If poultry manure stimulates N2O loss from soil it will reduce the fertilizer N value of this waste. In contrast, stimulated NjO loss in grass filter strips receiving the runoff from manured fields could help reduce contamination of surface water by NOf. Our objectives were to determine (i) if poultry manure stimulated N2O loss in soil after rainfall and (ii) if there was an edge-of-field effect on N2O loss in grass filters intercepting runoff from amended soil. Soil covers were used to measure N2O loss from a well-drained, poultry manure-amended, silt loam soil immediately after simulated rainfall and were also used to measure N2O loss from grass filters intercepting their surface runoff. Nitrous oxide loss from manureamended soil was greater than from unamended controls and ranged from 5 to 13 mg N2O-N m~ 2 h". The maximum N2O loss was equivalent to 3.2 kg N2O-N ha"' d~'. Nitrous oxide loss from grass niters intercepting runoff ranged from 0.1 to 1.4 mg NiO-N m~ h" and was significantly greater than portions of the grass filters that did not intercept runoff. Nitrous oxide loss from poultry manureamended soils was greater than N2O loss typically measured from waste-amended agricultural soils. However, it only represented up to 0.7% of the total N in the applied manure. B PRODUCTION IN KENTUCKY has dramatically increased in recent years and by 1997 annual production is expected to exceed 250 million birds. Because approximately 1500 kg of waste is generated per 1000 birds in a 10-wk growing cycle, the waste generated by this industry's growth, both manure and litter, will be considerable (Edwards and Daniel, 1992). Poultry wastes are typically land-applied in Kentucky, as in much of the southeastern USA, since this practice is cheap and convenient relative to processing methods such as composting. Crops can recover some of the N from the waste, which reduces the inorganic fertilizer N required to sustain high yields. However, compared with other animal wastes, relatively little is known about Department of Agronomy, N-122 Agricultural Science Building-North, Univ. of Kentucky, Lexington, KY 40546-0091. Contribution of the Kentucky Agricultural Exp. Stn. Manuscript no. 94-3-11. Received 19 Oct. 1994. *Corresponding author ([email protected]). Published in J. Environ. Qual. 24:1091-1096 (1995). the environmental effects of poultry waste disposal (Edwards and Daniel, 1992). One environmental consequence of poultry waste disposal may be stimulation of nitrous oxide (N2O) loss from soil. Nitrous oxide contributes to global warming and ozone layer destruction (Davidson, 1991). However, in terms of N management, a more immediate interest is that it represents a largely irreversible loss of N from soil. Experiments employing poultry waste as a soil amendment have been conspicuously absent in studies of N2O loss from agricultural soil (Eichner, 1990). More importantly, few of these field studies measured N2O loss during periods when it was likely to be at its greatest— immediately after rain (Firestone and Tiedje, 1979). Coyne et al. (1994) monitored N2O loss after simulated rainfall on a well-drained silt loam soil amended with stored poultry manure and measured rates that averaged 0.18 kg N2O-N ha" 1 d". However, up to 3 wk elapsed between the date of manure application and rainfall simulation. Furthermore, it was not obvious from this study whether poultry manure stimulated N2O loss from soil relative to unamended soil. Evidence that recently incorporated fresh poultry manure stimulates N2O loss from tilled soils immediately after rainfall still needs to be examined. Although N2O emission represents lost N from a nutrient management perspective, it would be beneficial in terms of reducing NOf contamination of surface water and groundwater by runoff from manured fields. Grass filters are used as a best management practice to intercept surface runoff before it reaches waterways. Groffman et al. (1991) noted that grass filters had greater denitrification potential than control soils, although they did not demonstrate this effect by measuring N gas loss in field conditions. They suggested that adding readily available C, such as manure, might enhance denitrification in filter strips and remove some of the NOf that infiltrated grass filters before it reached groundwater (Groffman et al., 1991). Nitrous oxide is an intermediate in denitrification and other microbial pathways that reduce NOa"; conseAbbreviations: BCD, electron capture detector; MPN, most probable number; COD, chemical oxygen demand. 1092 J. ENVIRON. QUAL., VOL. 24, NOVEMBER-DECEMBER 1995 quently, grass filters that trap runoff from manureamended soil may have elevated N20 loss and reflect increased denitrification activity. Coyne et al. (1994) observed that N20 loss from grass filters receiving runoff from poultry manure-amended soil was less than in the manure-amended soil itself. However, the edge-of-field effect was not examined by comparing N20 loss from grass filters receiving runoff to suitable controls. Nitrous oxide loss may be stimulated in grass filters abutting manured plots and receiving their runoff compared with grass filters not similarly affected. If so, it represents a route by which N can be removed from surface runoff before it contributes to nonpointsource pollution of adjoining waterways or infiltrates to groundwater. In this article we report on the stimulation of N20 loss from tilled soil amended with fresh, undercage layer manure immediately after a runoff-producing rain. We also report on the edge-of-field effect of runoff from manure-amended soil on N20 lOSS in adjoining grass filters. Our objectives were to quantify NzO loss immediately after rainfall in both cases and relate these measurements to the potential denitrification activity in both the manureamended soil and grass filter strips. MATERIALS AND METHODS Site. We did the study at the University of Kentucky Agricultural Experiment Station in L~xington during June and July 1993 on plots that were previously used by Coyne t al. (1994) for rainfall simulation studies. The soil was a well-drained, Maury silt loam soil (fine, mixed, mesic Typic Paleudalf) with an average natural slope of 9 % and soil permeability ranging from 5 to 15 cm h-~ (Blevins et al., 1990). Six individual tilled plots 4.6 m wide by 18.2 m long (Plots 1 and 2) or 13.7 long (Plots 3, 4, and 5) were prepared (Fig. 1). An additional 18.2 m long plot was prepared but rendered unusable by storm damage during this study. A grass filter strip either 9.0 or 4.5 m in length abutted the downslope dge of each tilled strip as shown in Fig. 1. These grass filters were a mixed sod composed of tall fescue (Festuca arundinacea L.) and Kentucky bluegrass (Poa pratensis L.). Site Treatment. Moore et al. (1983) previously described the rain simulator used in our study. Due to its size (five individual units, hooked in tandem, each with dimensions of 4.6 m by 6.1 m), the time necessary to prepare each plot for rain simulation, and the length of rain simulation on each plot, it was impractical to attempt more than one simulation per week. We minimized this temporal difference by strictly maintaining the same timing of manure application and rain simulation. Undercage poultry manure from a laying house was collected by 0900 h and uniformly spread over a tillage strip at 16.0 Mg ha-~ (wet wt.). The manure was incorporated into the tillage strip to a depth of 15 cm with a chisel plow as the only tillage practice. Tillage was in the direction of the slope and was completed by 1200 h. A border surrounding the manure-amended soil was tilled at the same time. We left the tillage strips uncovered after manure application unless rain was forecast. In that event, we covered the strips with black plastic tarps. We did not cover the grass filters. Meteorological conditions during the study are shown in Fig. 2. The average nutrient composition of the manure was 2.8 % total N, 2.9% total P, and 1.8% total K on a wet-weight basis. Average moisture content was 34%. The rate used was equivalent to an application of 448 kg N ha-~, about three o0 o