Ammonia Volatilization Losses during Sprinkler Irrigation of Animal Manure

Recent work at Clemson University has provided 32 new data sets of ammonia volatilization losses during sprinkler irrigation of liquid manure. These 32 observations were pooled with 23 additional data sets that were available in the literature. The combined data includes losses from traveling gun, center pivot, and impact sprinkler irrigation of dairy, swine, and beef manure. Manure type did not affect volatilization losses. Total ammoniacal nitrogen (TAN = NH3-N + NH4-N) concentration differences between samples collected from the irrigated wastewater and samples collected in containers placed on the ground were measured as an estimate of ammonia volatilization loss. The TAN concentrations of the ground collected samples were not statistically different from TAN concentrations of the irrigated waste. In addition, it was determined that evaporation and drift were not major factors in the quantification of TAN losses. Therefore, volatilization loss from manure during the irrigation event was not found to be significant. INTRODUCTION Sprinkler irrigation of animal manure and wastewater onto crop, forage, and pasture land to recycle plant nutrients is a common practice in many regions of the United States. The total ammoniacal nitrogen (TAN = NH4-N + NH3-N) in liquid animal manure can account for 28% to 85% of the total Kjeldahl nitrogen (Chastain, et al., 2001; Montes, 2002) depending on the moisture content and animal species. A portion of the TAN can potentially be lost as a part of the land application process as ammonia volatilization. Ammonia volatilization loss during irrigation of manure and wastewater is an important issue due to the fact that regulatory agencies in the United States, Canada, and Europe either have prohibited the use of irrigation as a land application method or are considering the prohibition of this land application technique in order to reduce ammonia emissions from agriculture. Volatilization losses can potentially occur during collection, transfer, storage, treatment, and land application. The majority of the volatilization losses are associated with storage, treatment, and land application of manure (MWPS, 1985; Chastain et al., 2001; and Montes, 2002). The ammonia-N losses associated with land application can occur during the application process, or over a 1 to 4 day period following application (Meisinger, and Jokela, 2000; Montes, 2002). The available data indicates that the majority of the volatilization losses associated with land application occur following the application event (Meisinger, and Jokela, 2000; Montes, 2002). Many extension publications (e.g. MWPS, 1985; Dougherty et al., 1998) consider the ammonia loss due to irrigation to be greater than for land applied slurries and solid manure. However, extensive review of the literature (Chastain et al., 2001; Montes, 2002) and recent work on the ammonia losses following irrigation of lagoon effluent (Montes and Chastain, 2003) indicate that the losses following irrigation are no greater than for other land application 2005 Southern Conservation Tillage Systems Conference Clemson University 148 Poster Proceedings methods. The volatilization losses following irrigation of dilute liquid manures, such as lagoon supernatant, are much lower than for other land application scenarios (≈ 2% of TAN applied). The mass of TAN lost is a function of the solids content of the manure, application depth, and the amount of manure intercepted by plant foliage or residue (Chastain et al., 2001; Montes, 2002). The percentage of the TAN in the ammonia form is strongly dependent on pH. Most manure has a pH in the range of 7.0 to 8.0. About 8% to 10% of the TAN is in the ammonia form for most liquid manures (Jayaweera and Mikkelsen, 1990; Zhang, 1992; American Petroleum Institute, 1981; 1995; Ruxton, 1995; Cumby et al., 1995; Denmead et al., 1982). Therefore, only a small fraction of the TAN has the potential to be lost during the irrigation process. Several studies have reported ammonia volatilization losses of 10 to 25% during irrigation of liquid swine manure (Sharpe and Harper, 1997; Westermann et al., 1995; Safley et al., 1992). Safley et al. (1992) attributed the majority of the irrigation losses to the influence of evaporation and drift. Earlier work by Welsh (1973), concluded that volatilization losses during the irrigation of dairy slurry, liquid swine manure, and effluent from an oxidation ditch were insignificant. Recent work at Clemson University (Montes and Chastain, 2000; Montes, 2002) supported the observations by Welsh. Only three studies (Montes, 2002; Safley et al., 1992; and Welsh, 1973) had the quantification of ammonia losses during irrigation as a primary objective and the conclusions are mixed with regards to the importance of ammonia volatilization losses during the irrigation event. Only one of these studies (Montes, 2002) included rigorous statistical and error analyses. The objectives of this paper are to: • perform a pooled statistical analysis of the available data related to ammonia volatilization losses during irrigation of animal manure, and • perform a critical analysis of the impact of evaporation and drift on volatilization losses during the irrigation event. METHODS A summary of the available data on ammonia volatilization losses during irrigation of animal manure is presented in Table 1. Ammonia volatilization losses were calculated from the data reported by the authors based on the difference in TAN concentration before and after irrigation. These losses ranged from -33% to 26%. The mean ammonia loss ranged from -2.5% to 13% with an overall mean of 4.0% of the TAN applied. Negative ammonia loss values imply that NH3 was gained during the irrigation process. While this is obviously impossible, it indicates a significant amount of uncertainty in the quantification of ammonia losses. The factors that have been proposed to affect the magnitude of ammonia loss during irrigation include: air temperature, relative humidity, irrigation pressure, drop diameter, spray velocity, TAN content of the irrigated manure, and pH (Pote et al., 1980; Denmead et al., 1982; Brunke et al., 1988; Sharpe and Harper, 1997). These factors have been suggested as the cause of the variability in measuring ammonia volatilization losses. However, most of the authors did not perform any type of error analysis on their data collection procedures. 2005 Southern Conservation Tillage Systems Conference Clemson University 149 Poster Proceedings Table 1. Summary of available data on volatilization losses during sprinkler irrigation of manure. Description Irrigated TAN (mg/L) Irrigated (TS %) pH Ammonia Loss (%) n Reference Big Gun: Dairy, Beef, Swine 187 to 850 0.3 to 8.4 7.4 to 7.9 -2.5 (-12.4 to 9.8) 5 Welsh (1973) Center Pivot: Swine 299 to 327 0.14 to 0.17 7.4 to 7.5 4.9 (-2.1 to 18.4) 12 Safley et al. (1992) Big Gun: Swine 214 to 510 0.11 to 0.37 7.1 to 7.7 2.9 (0.5 to 9.4) 6 Safley et al. (1992) Big Gun: Swine 242 1 NR 2 NR 5.7 (-5.0 to 24) 3 Westermann et al. (1995) Solid Set: Swine 53 1 NR NR 13 NR Sharpe and Harper (1997) Solid Set: Swine 109 to 1183 0.05 to 0.57 7.6 to 8.6 0.3 (-33 to 26) 32 Montes (2002) 1 Not given directly, estimated from application data given in reference. 2 NR = not reported In the investigation by Welsh (1973), samples were taken from the manure storage structure before irrigation and from ground collected samples following the irrigation event. The difference in TAN concentration was used to estimate NH3 loss due to the irrigation process. The study, conducted in Minnesota, included four different manure types with very different characteristics as is reflected by the large range in total solids and TAN concentration shown in Table 1. The average ammonia loss was -2.5% and was not significantly different from zero. Safley et al. (1992) studied ammonia losses during sprinkler irrigation of swine lagoon effluent using center pivot and traveling gun irrigation equipment in North Carolina. Ammonia losses were estimated by calculating the difference in TAN concentration between samples taken from the lagoon and samples taken from liquid caught on the ground during irrigation. The TAN concentration difference between irrigated and ground collected samples in the data presented by Safley et al. (1992) ranged from -2.1% to 18.4% with a mean of 3.9%. The studies by Westermann et al. (1995), and Sharpe and Harper (1997) did not include all of the data required to be included in the present study. The TAN concentrations in the irrigated manure were estimated from nutrient application rate information provided in the publications. Consequently, these data were not included in the pooled statistical analysis. Montes (2002) collected similar ammonia volatilization data for sprinkler irrigation from two swine lagoons in South Carolina. Montes collected irrigated lagoon water samples from a sampling port in the irrigation pipe on the discharge side of the irrigation pump. The ground collected samples were the composite of samples collected in 8 locations within the irrigated plots. The data from the studies by Welsh (1973), Safley et al. (1992), and Montes (2002) were pooled into common statistical analyses. The quantities that were included were: TS, TAN, TKN (total Kjeldahl nitrogen), and pH. The change in TS between the irrigated and ground collected samples was included to provide a measure of evaporation losses. Both TAN and TKN were 2005 Southern Conservation Tillage Systems Conference Clemson University 150 Poster Proceedings included since a significant reduction in TAN during irrigation would also result in a reduction in TKN. Data on pH were included since the fraction of TAN that is in the ammonia form depends on manure pH. ANALYSIS AND RESULTS Pooled linear regression analyses were performed for the irrigated and ground collected concentrations of TS, TAN, and TKN. The least-squares best fit for each constituent was represented by the following equation form: