Management Strategy Impacts on Ammonia

Published online May 11, 2005

Management Strategy Impacts on Ammonia Volatilization from Swine Manure Diane M. Panetta, Wendy J. Powers,* and Jeffery C. Lorimor

Reproduced from Journal of Environmental Quality. Published by ASA, CSSA, and SSSA. All copyrights reserved.

ABSTRACT

of nitrogen to the atmosphere as ammonia (Kerr, 1995; Ferguson et al., 1998; Hobbs et al., 1998). Once excreted, however, management strategies must be used to reduce the proportion of excreted nitrogen that is volatilized, with special concern for the nitrogen volatilized as ammonia. Ultimately, a combination of dietary and postexcretion strategies will be needed to address air emission issues. When urine and feces are mixed in manure slurries, the urea, found primarily in urine, is hydrolyzed by microbial ureases in feces to form carbon dioxide and ammonium (Muck and Steenhuis, 1981). The ammonium may be released as gaseous ammonia under favorable temperature, moisture, and pH conditions (Van Horn et al., 1996). The dissociation of ammonia from ammonium is favored by alkaline conditions, suggesting that acidification of slurry may inhibit ammonia volatilization (Molloy and Tunney, 1983; Husted et al., 1991). Ammonia production is also temperature dependent, and favored by warm conditions (Power et al., 1994). Agitation frequency, or stirring of the slurry, may affect ammonia release by disturbance of the liquid–gas interface and by increasing uniformity throughout the slurry. Segregation of urine from feces at excretion should terminate the process at the start, by preventing urease from hydrolyzing the urinary urea (Mobley et al., 1995). A urease inhibitor like NBPT may also prevent this hydrolysis even when urine and feces are mixed (Varel et al., 1999). When urea is hydrolyzed, an agent that chemically converts or binds ammonium or ammonia in the slurry, as the plant extract of yucca is believed to do (Headon et al., 1991), would prevent ammonia from being converted to the gaseous form. Potential strategies to reduce ammonia volatilization include changing the physical, chemical, and microbiological factors that affect the formation of ammonium and ammonia and the release of ammonia from the manure solution. Our hypothesis was that these management strategies, evaluated in a laboratory-scale experiment, can be used to reduce the ammonia volatilized over time, the total amount of nitrogen lost, and the proportion of slurry nitrogen remaining in the form of ammonium, by intervention at different points in the process leading to ammonia volatilization. The objective was to determine how the pattern of nitrogen conversion and ammonia volatilization varies over time for various manure management strategies. To accomplish this, we studied the effects of (i) temperature, (ii) continuous stirring, (iii) addition of a nitrogen binder, (iv) addition of a urease inhibitor, (v) segregation of urine and feces, and (vi) pH on ammonia emission potential of swine manure.

Ammonia emitted from manure can have detrimental effects on health, environmental quality, and fertilizer value. The objective of this study was to measure the potential for reduction in ammonia volatilization from swine (Sus scrofa domestica) manure by temperature control, stirring, addition of nitrogen binder (Mohave yucca, Yucca schidigera Roezl ex Ortgies) or urease inhibitor [N-(n-butyl) thiophosphoric triamide (NBPT)], segregation of urine from feces, and pH modification. Swine manure [total solids (TS) ⫽ 7.6–11.2%, total Kjeldahl nitrogen (TKN) ⫽ 3.3–6.2 g/L, ammonium nitrogen (NH4ⴙ–N) ⫽ 1.0–3.3 g/L] was stored for 24, 48, 72, or 96 h in 2-L polyvinyl chloride vessels. The manure was analyzed to determine pre- and post-storage concentrations of TS and volatile solids (VS), TKN, and NH4ⴙ–N. The concentration of accumulated ammonia N in the vessel headspace (HSAN), post-storage, was measured using grab sample tubes. Headspace NH3 concentrations were reduced 99.3% by segregation of urine from feces (P ⬍ 0.0001). Stirring and NBPT (152 ␮L/L) increased HSAN concentration (119 and 140%, respectively). Headspace NH3 concentration increased by 2.7 mg/m3 for every 1ⴗC increase in temperature over 35ⴗC. Slurry NH4ⴙ–N concentrations were reduced by segregation (78.3%) and acidification to pH 5.3 (9.4%), and increased with stirring (4.8%) and increasing temperature (0.06 g/L per 1ⴗC increase in temperature over 35ⴗC). Temperature control, urine–feces segregation, and acidification of swine manure are strategies with potential to reduce or slow NH4ⴙ–N formation and NH3 volatilization.

E

missions of ammonia contribute to acid precipitation and eutrophication of surface waters (Lowe, 1995). Atmospheric ammonia forms fine particulate matter (PM2.5 or particles 2.5 ␮m in diameter or smaller; USEPA, 1997), an air pollutant regulated under the Clean Air Act. The volatilization of ammonia from swine manure is estimated to contribute more than 10% of all anthropogenic ammonia emissions in the United States (Battye et al., 1994). In addition to the environmental concerns, the volatilization of nitrogen reduces the fertilizer value of manure. More than $80 million of fertilizer value is lost in the United States, annually, because of nitrogen volatilization from manures (Mackie et al., 1998). For swine manure alone, the USEPA estimates that more than 400 000 Mg per year, or about 63% of the nitrogen, is volatilized under normal production conditions (USEPA, 2004). Dietary changes have been shown to affect the initial amount and form of nitrogen available for volatilization (Sutton et al., 1999). Numerous studies have demonstrated that reducing dietary crude protein results in reduced nitrogen excretion and, therefore, potential losses

D.M. Panetta and W.J. Powers, Department of Animal Science, and J.C. Lorimor, Department of Agricultural and Biosystems Engineering, Iowa State University, Ames, IA 50011. Received 12 Aug. 2004. Technical Reports. *Corresponding author ([email protected]). Published in J. Environ. Qual. 34:1119–1130 (2005). doi:10.2134/jeq2004.0313 © ASA, CSSA, SSSA 677 S. Segoe Rd., Madison, WI 53711 USA

Abbreviations: HSAN, headspace ammonia nitrogen; NBPT, N-(n-butyl) thiophosphoric triamide; TKN, total Kjeldahl nitrogen; TS, total solids; VS, volatile solids.

1119

1120

J. ENVIRON. QUAL., VOL. 34, MAY–JUNE 2005


MATERIALS AND METHODS Six trials were conducted using a direct measurement technique to evaluate the effects of (i) temperature variation, (ii) continuous stirring versus no stirring, (iii) addition of a nitrogen binder, (iv) addition of a urease inhibitor, (v) segregation of urine and feces, and (vi) pH adjustment on in vitro ammonia volatilization from swine manure.

Substrate Collection With the exception of the segregation trial, fresh manure was collected as needed from the flush gutters of solid concrete-floored pens housing grow-finish pigs fed a corn and soybean meal diet (11–12% crude protein). In addition to corn and soybean meal, the only remaining diet ingredients were minerals and vitamins. Buckets of manure were sealed and stored at ⫺20⬚C. Before each trial, the buckets to be used were thawed at 4⬚C. The manure from each bucket was handmixed and diluted with water in a 1:1 (v/v) ratio. For the segregation trial, urine and feces were collected directly from pigs on a 17.5% crude protein corn and soybean meal diet with minerals and vitamins. The excreta were separately frozen, thawed, and diluted 1:1 (v/v) with water to mimic the dilution that normally occurs in deep pit manure storage systems from pit charging and wasted drinking water. Following dilution, the substrate from each bucket was sampled for total solids (TS), volatile solids (VS), total Kjeldahl nitrogen (TKN), and ammonium nitrogen (NH4⫹–N) analyses. Total Kjeldahl nitrogen and ammonium nitrogen were analyzed on the wet samples to minimize nitrogen volatilization. All analyses were run in duplicate and followed procedures outlined by AOAC International (2002). Analysis of the initial substrate is provided as a reference point only (footnote of Tables 1–6), as the initial content was not considered a treatment and therefore not included in statistical analyses. The manure storage vessels were constructed of 10.2-cm polyvinyl chloride (PVC) pipe. Bottoms of the vessels were a PVC cap. Each vessel had an operating capacity of 2 L and a headspace of 280 cm3. Duplicate vessels were filled with 1 L of manure slurry for each combination of storage time (24, 48, 72, and 96 h) and treatment. Treatments and storage times were assigned randomly to vessels, after filling, and the bucket source of each vessel was recorded. The top of each vessel was closed with a rubber stopper and all vessels were left undisturbed at room temperature during their assigned storage periods. Because the vessels were stoppered, gases would accumulate in the headspace. However, because manure had been frozen before use, the microbial activity was depressed sufficiently that excess pressure did not build up in the headspace.

Treatments Temperature Effects Temperature effects on ammonia emission potential were evaluated by storing manure in the vessels, and maintaining them at 24, 35, and 42⬚C. Vessels were established, in duplicate, for each temperature ⫻ time (24, 48, 72, and 96 h) combination (n ⫽ 24). Room-temperature vessels (24⬚C) were stored on a laboratory bench top. The 35 and 42⬚C temperatures were maintained using one of two water baths, preheated to the prescribed temperature. Stirring Effects The effects of stirring on ammonia emission potential were evaluated by storing manure under continuous stirring or leav-

ing unstirred. Duplicate vessels were established for each stirring regime ⫻ time (24, 48, 72, and 96 h) combination (n ⫽ 16). Continuous stirring was achieved by placing the vessels, each containing a large stir bar, individually on stir plates (Isotemp Catalog #11-600-49SU; Fisher Scientific, Hampton, NH). The plates were set to continuously mix at 1200 rpm throughout the storage period. The unstirred vessels were left undisturbed on a laboratory benchtop at room temperature. Nitrogen Binder Effects Yucca extract as De-Odorase (Alltech, Nicholasville, KY) was added to the vessels at 0, 7.4, and 14.9 mg/L to evaluate the effects of a nitrogen-binding agent on ammonia emission potential. The binder was stirred into the substrate of each vessel before storage at room temperature. Each binder dose ⫻ time (24, 48, 72, and 96 h) combination was represented by duplicate vessels (n ⫽ 24). The 7.4 mg/L corresponded to the recommended dose (2 ounces per ton), while the 14.9 mg/L represented a double dose. The control group received no additive. Urease Inhibitor Effects The urease inhibitor NBPT (Agrotain International, St. Louis, MO) was evaluated as a manure amendment. Concentrations of the commercial product used were 0, 76, and 152 ␮L/L of manure with each volume of product containing 25% NBPT by weight. The 76 ␮L/L corresponded to the recommended dose for liquid manure, while the 152 ␮L/L represented a double dose. The control group received no additive. The product was pipetted and stirred into the substrate of each vessel before storage. Vessels were established, in duplicate, for each inhibitor dose ⫻ time (24, 48, 72, and 96 h) combination (n ⫽ 24), and stored at room temperature. Segregation Effects The impacts on ammonia emission potential when urine and feces are stored separately were evaluated by comparing dilute urine, dilute feces, and a mixture of the two as the substrates. Fresh feces and urine were collected from growfinish pigs into separate buckets and frozen. The buckets were thawed and diluted 1:1 (v/v) with water before filling the vessels to mimic the dilution that normally occurs in deep pit manure storage systems from pit charging and wasted drinking water. The mixture vessels contained urine and feces in an as-excreted ratio (Standard D384.1; American Society of Agricultural Engineers, 2003) with 466 mL of dilute urine and 534 mL of dilute feces per vessel. Duplicate vessels were established for each substrate ⫻ time (24, 48, 72, and 96 h) combination (n ⫽ 24), and stored at room temperature. pH Effects The effects of pH on ammonia emission potential were evaluated by storing manure in the vessels at an initial pH of 5.30, 6.59, and 8.85. Duplicate vessels represented each pH ⫻ time (24, 48, 72, and 96 h) combination (n ⫽ 24). The pH 6.59 corresponded to the original slurry pH, with no added acid or base. The pH 5.30 was obtained by addition of 17 to 18 mL of 4 M hydrochloric acid to each vessel. The pH 8.85 was obtained by addition of 8 mL of 5 M sodium hydroxide per liter of slurry. After addition of acid or base and before storage at room temperature, the starting pH of each vessel was measured using a combination pH electrode (Model 50200; Hach, Loveland, CO) and pH meter (Model 250; Denver Instrument, Denver, CO). The final pH of a sample from each vessel was measured after storage.

PANETTA ET AL.: AMMONIA VOLATILIZATION FROM SWINE MANURE


Measurements and Analyses At the end of each storage time, the designated vessels were opened, and headspace ammonia (HSAN) concentrations were measured using Dra¨ger sample tubes, which ranged from 0.25 to 700 ppm, and a Dra¨ger accuro bellows pump (Dra¨ger Safety, Pittsburgh, PA). The lower detection limit of the tubes used was initially 5 ppm. However, due to the low concentration of ammonia, tubes with a lower limit of detection of 0.25 ppm were adopted for the remaining trials. Room temperature was recorded daily during the temperature, N binder, and urease inhibitor trials, because of unusually warm laboratory conditions (see Calculations, below). The slurry in each vessel was mixed and sampled to analyze final TKN, NH4⫹–N, TS, and VS (and pH for the pH trial). Ammonium N and TKN analyses were begun on the wet manure immediately. Two 400-mL samples from each vessel were frozen until TS, VS, and pH could be analyzed. These samples were also used to re-run any duplicates that exceeded an intra-assay coefficient of variation of 5% for TKN and NH4⫹–N and 10% for TS and VS.

Calculations The mass of ammonia released into the headspace was calculated as: headspace concentration of ammonia in ppm multiplied by 1 ⫻ 10⫺6 and the molecular weight of ammonia, and divided by the molar volume of headspace gas (Lefcourt, 2001). The molar volume of headspace gas was determined using the Ideal Gas Law, where the volume in L/mol equals the gas constant (0.0821 L · atm/mol · K) multiplied by the absolute temperature and divided by the pressure. Measured ambient temperatures were used for the calculations when available, and 1 atm (0.101 MPa) pressure, although at the elevation of the laboratory (281 m above sea level), the atmospheric pressure was 0.967 atm (0.098 MPa) during the experiments (National Oceanic and Atmospheric Administration, 2003). The average temperature that occurred during the N binder trial was used in calculations for the subsequent stirring, segregation, and pH trials. Nitrogen loss was reported as the difference in mass of TKN in the slurry before and after storage. Ammonium N and TKN concentrations were used to calculate the percentage of TKN in the form of ammonium N before and after storage.

Data Analyses Statistical analyses were completed using the GLM procedure of SAS Version 8.2 (SAS Institute, 1990). Data were analyzed as a two-factor, completely randomized design. For the stirring, segregation, and pH trials, the model included treatment, hour of storage, and the treatment by hour interaction. For the N binder, urease inhibitor, and temperature trials, the same model was used to determine least squares means for the main effects. To evaluate the significance of treatment as a continuous effect for these experiments and generate a continuous plot, a second model was solved with storage time as a fixed effect and treatment as a continuous independent variable. Levels of significance are reported based on this model, which does not include the interaction due to consideration of treatment as a continuous effect. Contrast statements were used to determine storage time effects within a treatment and treatment effects within a storage time when the main effects of treatment and storage time, respectively, were significant. For those experiments where treatment was treated as a continuous effect, regression equations for each storage time were obtained using regression procedures and subgrouping

1121

by storage time. The slopes of the storage time regression equations were compared with each other in a pair-wise fashion using t tests of the treatment ⫻ hour parameter at each storage time. In all trials, simple correlation procedures were used to evaluate relationships between slurry NH4⫹–N data and HSAN data. Data were analyzed separately for each trial because of differences in manure sources and handling. Significance was declared at P ⱕ 0.05.

RESULTS AND DISCUSSION Temperature Effects The main effect of storage time on TS and VS concentrations was not significant (P ⬎ 0.05). As storage temperature increased from 24 to 35 or 42⬚C, TS and VS concentrations decreased (P ⬍ 0.05), with no difference between concentrations of either TS or VS in the slurries stored at 35 and 42⬚C (Table 1). Any loss of solids mass can be attributed to volatilization of nitrogen and carbon dioxide released during microbial respiration and urea hydrolysis. Storage temperature and time had no effect on TKN concentration (Table 1). However, slurry ammonium N concentrations averaged 1.08 ⫾ 0.07 g/L initially, and increased when storage vessels were heated above room temperature and were stored over time (Table 1). Using contrast statements it was found that after 24 h of storage, the 42⬚C vessels had greater NH4⫹–N concentrations than the vessels stored at 35 and 24⬚C (P ⬍ 0.05). At 96 h, the vessels stored at 35⬚C had greater NH4⫹–N concentrations than those at 24⬚C (Table 1). The accumulation of NH4⫹–N slowed following the first two days of storage. The increase in NH4⫹–N concentrations over time suggests that NH4⫹–N built up in the slurry because either the conversion of NH4⫹–N to volatile forms of nitrogen or the volatilization of nitrogen into the headspace occurred more slowly than NH4⫹–N production. During the first 24 h of storage, volatilization likely reflected the rapid conversion of urinary urea to ammonia. Because TKN concentration was unaffected by storage time or temperature, and NH4⫹–N concentration increased with storage time and temperature, the portion of TKN in the form of NH4⫹–N increased with temperature and time (Table 1). After storage at 24⬚C, the slurry NH4⫹–N, as a percentage of TKN, was 29 and 15% less than that of slurry stored at 42 and 35⬚C. The portion of NH4⫹–N increased 10.8 and 6.0% points the first two days of storage, and a total of 2.8% during the last 48 h. Plotting NH4⫹–N as a percentage of TKN over time illustrates the effect of temperature on NH4⫹–N formation over time. The TKN and ammonium data suggest that maintaining lower temperatures decreases the conversion of organic forms of manure nitrogen to ammonium nitrogen. Ammonia N concentration in the headspace was more than sevenfold greater at 42⬚C than at the lower temperatures despite no treatment effects on TKN concentration. This conflict is probably the result of method differences and sensitivity of those methods. The HSAN was measured with units of mg/m3. The regression analysis

1122


Table 1. Least squares means of temperature and storage time effects on swine manure composition† and headspace ammonia N concentration. Temperature


ⴗC 24

Time post-storage

TS

h

35

42

Temperature Storage time

%

24 48 72 96 24 48 72 96 24 48 72 96

10.10 10.36 10.17 10.20 9.10 8.40 7.65 8.68 9.68 8.79 8.16 8.96 0.75

24ⴗC 35ⴗC 42ⴗC 24 h 48 h 72 h 96 h

10.21b 8.46a 8.90a 9.63 9.18 8.66 9.28

SEM

VS

Treatment Storage time

0.0103 0.4520

TKN

NHⴙ4 –N

g/L 3.59 3.65 3.74 3.47 3.16 3.07 3.46 3.83 3.71 3.68 3.41 3.66 0.19 Main effect means§ 7.95b 3.61 6.47a 3.38 6.91a 3.62 7.58 3.49 7.14 3.46 6.67 3.54 7.06 3.66 Main effect significance levels 7.99 8.06 8.00 7.75 7.18 6.60 5.63 6.47 7.57 6.76 6.36 6.96 0.67

0.0201 0.4182

0.8461 0.7334

1.29 1.35 1.44 1.43 1.28 1.40 1.64 1.77 1.61 2.03 1.93 2.16 0.10 1.37a 1.52a 1.93b 1.39A 1.59B 1.67BC 1.78C (P )¶

⬍0.0001 0.0081

NHⴙ4 –N

Headspace NH3–N

% of TKN 35.8 36.9 38.4 41.1 40.4 45.5 47.3 46.1 43.4 55.3 56.7 58.9 1.2

mg/m3 1.9 2.4 2.7 2.9 1.7‡ 1.8 2.8 5.4 6.8 26 31 26 3.1

38.1a 44.8b 53.6b 39.9A 45.9B 47.4BC 48.7C

2.8a 2.9a‡ 22b 3.4A‡ 10B 12B 12B

⬍0.0001 0.0004

0.0004 0.4321

† Initial manure composition data were as follows: total solids (TS) content ⫽ 10.14%, volatile solids (VS) content ⫽ 8.11%, total Kjeldahl nitrogen (TKN) content ⫽ 3.70 g/L, ammonium nitrogen (NH4ⴙ–N) content ⫽ 1.08 g/L, ammonium nitrogen, percent of total Kjeldahl nitrogen (NH4ⴙ–N, % of TKN) ⫽ 29.1%. For each temperature–time combination, n ⫽ 2. ‡ One vessel was below the limit of detection (5 ppm). § Main effect means within a column having different letters differ significantly (P ⬍ 0.05). ¶ Main effects were determined with treatment as a continuous independent variable. Therefore the interaction of the main effects was not included in the statistical model.

yielded a single equation for the effect of temperature (in ⬚C) on HSAN: HSAN (mg/m3) ⫽ 1.01754 ⫻ temperature (⬚C) ⫺ 24.62994 The response of headspace ammonia N concentration to temperature may not be linear (R2 ⫽ 0.47). We were unable to test for quadratic or higher order effects because only three temperatures were evaluated. No storage time effects were observed for HSAN concentration (Table 1). Slurry ammonium N concentrations and ammonium N as a percentage of TKN were both positively correlated with headspace ammonia N concentrations (r ⫽ 0.85 and r ⫽ 0.85, respectively). Increased storage temperatures facilitated both ammonium N production and ammonia N volatilization. Ammonia N volatilization increased when the pool of readily volatilized slurry nitrogen (in the form of ammonium N) was increased. It also slowed as the slurry ammonium N accumulation slowed. It cannot be determined from correlation analyses if one of these events caused the other.

over time by an average of 0.125 ⫾ 0.065 g/d, pooled across treatments (Table 2). Continuous stirring and increased storage time increased the ammonium N as a percentage of TKN (Table 2). There was a significant treatment by hour interaction for ammonium N as a percentage of TKN (P ⫽ 0.0016). Left unstirred, a steady increase in manure ammonium N as a percentage of TKN over time was observed, while stirring yielded more erratic results over time. Headspace ammonia N concentration was increased 119% by stirring (Table 2), and was greater after 96 h of storage (7.8 mg/m3) than after 24 to 72 h (2.9, 3.1, and 3.4 mg/m3, pooled across stirring regimes; P ⫽ 0.0276; Table 2). The results suggest that stirring increased ammonium N formation from organic nitrogen. Stirring also increased volatilization of ammonia N, but not enough to deplete the ammonium N pool. The size of this pool (absolute and as a portion of the TKN) was positively correlated with the concentration of headspace ammonia N (r ⫽ 0.5, P ⫽ 0.0376 and r ⫽ 0.74, P ⫽ 0.0009, respectively).

Stirring Effects

Nitrogen Binder Effects

There were no differences in concentrations of TS, VS, and TKN between vessels attributed to stirring regime or storage time (Table 2). Stirring increased ammonium N concentrations, pooled across storage times, by 4.8% compared with not stirring (1.53 vs. 1.46 g/L; P ⫽ 0.0013; Table 2). Ammonium N concentration increased

The laboratory temperature was an average 22 ⫾ 0.7⬚C during the four days of the trial. Mean concentrations of TS, VS, and TKN (Table 3) were not affected by variation of the dose of yucca extract or the storage time. The changes in TKN concentration during storage did not follow a clear pattern over time, and may have been

1123


Table 2. Least squares means of stirring and storage time effects on swine manure composition† and headspace ammonia N concentration. Treatment

Time post-storage

TS

h


Stirred

Unstirred

Storage time

Treatment Storage time Treatment ⫻ storage time

%

24 48 72 96 24 48 72 96

9.68 10.72 10.19 10.88 11.03 10.31 10.76 10.38 0.34

8.57 8.48 8.49 8.66 8.83 8.63 8.52 8.68 0.09

stirred unstirred 24 h 48 h 72 h 96 h

10.37 10.62 10.36 10.51 10.47 10.63

8.55 8.67 8.70 8.55 8.50 8.67

SEM Stirring

VS

0.3176 0.8792 0.0760

TKN

NH4ⴙ–N

g/L 3.52 4.05 3.85 3.93 3.92 3.70 3.87 3.73 0.12 Main effect means‡ 3.84 3.81 3.72 3.87 3.86 3.83 Main effect significance levels

0.1064 0.1710 0.5488

0.7046 0.6045 0.0567

1.39 1.51 1.57 1.65 1.27 1.42 1.54 1.59 0.02 1.53b 1.46a 1.33A 1.47B 1.55C 1.62D (P )

0.0013 ⬍0.0001 0.3053

NH4ⴙ–N

Headspace NH3–N

% of TKN 39.5 37.3 40.9 42.0 32.4 38.6 39.7 42.8 0.7

mg/m3 5.5 2.7 5.6 9.7 0.33 3.4 1.3 5.9 1.4

39.9b 38.4a 35.9A 37.9B 40.3C 42.4D 0.0171 0.0001 0.0016

5.9b 2.7a 2.9A 3.1A 3.4A 7.8B 0.0158 0.0276 0.2594

† Initial manure composition data were as follows: total solids (TS) content ⫽ 11.22%, volatile solids (VS) content ⫽ 9.28%, total Kjeldahl nitrogen (TKN) content ⫽ 3.50 g/L, ammonium nitrogen (NH4ⴙ–N) content ⫽ 1.02 g/L, ammonium nitrogen, percent of total Kjeldahl nitrogen (NH4ⴙ–N, % of TKN) ⫽ 28.4%. For each stirring regime–time combination, n ⫽ 2. ‡ Main effect means within a column having different letters differ significantly (P ⬍ 0.05).

made apparent by a low standard error in this trial (0.03 g/L). Ammonium N concentration and ammonium N, as a percentage of TKN, increased over time (Table 3), indicating that ammonium N production occurred faster than nitrogen volatilization with no treatment differences. Headspace ammonia N concentration was significantly affected by hour (P ⬍ 0.0001), peaking at 48 h (1.7 mg/ m3 ; Table 3), with a significantly lower concentration at 72 h (1.0 mg/m3) suggesting ammonia N was converted to another form of nitrogen or removed from the headspace. This phenomenon occurred irrespective of yucca dose. While it is unclear what caused this decrease in ammonia N concentration, it is not explained by a room temperature difference, because temperatures at 48 and 72 h (22.0 and 22.8⬚C) resulted in only a 0.1 mol/L decrease in gas density at 72 h compared with 48 h. Vessels stored for at least 48 h had positive headspace ammonia N concentrations. Headspace ammonia N concentration was the only measurement that decreased with increased dose of yucca (Table 3). There is further evidence of this effect in the decreasing functions of the regressions at all storage times except 24 h (Fig. 1). There may have been a lag time of 24 to 48 h before enough ammonia volatilized to be able to observe dose effects of yucca. Figure 1a suggests that headspace ammonia N concentrations were approaching an equilibrium at each dose, as exhibited by the decreasing change in ammonia N concentration from one day to the next at all doses (Table 3). While storage time and the time ⫻ treatment interaction significantly improved the model (R2 ⫽ 0.80, P ⬍ 0.05), the slopes of the hourly regressions did not differ from each other, indicating that the dose effect was not dependent on time of storage. Therefore, we should expect decreases in ammonia N concentrations from each dose of yucca to be

proportionate across storage times, and total decreases to be greater after more than 24 h of storage. Headspace ammonia N concentrations were weakly correlated with ammonium N concentrations (r ⫽ 0.42, P ⬍ 0.05), and not with the percentage of TKN as ammonium N. Yucca may have been acting as a nitrogen binder, because it decreased ammonia volatilization without affecting slurry ammonium N concentrations. The results of this experiment do not distinguish in what form the nitrogen may have been bound or whether ammonia N volatilization was prevented by chemical binding or an effect on the physical characteristics of the slurry– headspace boundary.

Urease Inhibitor Effects The laboratory temperature averaged 25 ⫾ 0.3⬚C during the four days of the trial. Dose of NBPT and storage time had no effect on mean TS, VS, or TKN concentrations (Table 4), indicating no net TS or VS degradation or TKN loss. Ammonium N concentration and ammonium N as a percentage of TKN were significantly increased by increased storage time (Table 4), but not dose of NBPT. Ammonium N concentration was greater after 96 h (1.72 g/L) than the other storage times, which were not significantly different from each other. Ammonium N as a percentage of TKN was less at 24 h (31.6%) than after more time, with 48, 72, and 96 h values not different from each other (Table 4). The only significant dose effect of NBPT was on headspace ammonia N concentration (P ⫽ 0.0131). The addition of a double dose of NBPT (152 ␮L/L) increased headspace ammonia N concentration compared with the control and single dose, but the result of a single dose (76 ␮L/L) was not different from the control (Table 4).

1124


Table 3. Least squares means of N binder (yucca extract) and storage time effects on swine manure composition† and headspace ammonia N concentration. N binder


mg/L 0

7.4

14.9

Time post-storage h

Storage time


VS %

24 48 72 96 24 48 72 96 24 48 72 96

10.84 10.08 11.01 10.47 10.16 10.21 10.87 10.23 11.00 10.38 10.17 10.98 0.86

8.55 7.81 8.65 8.21 8.02 7.90 8.60 8.00 8.74 8.11 7.88 8.75 0.78

0 mg/L 7.4 mg/L 14.9 mg/L 24 h 48 h 72 h 96 h

10.60 10.37 10.63 10.67 10.23 10.68 10.56

8.31 8.13 8.37 8.44 7.94 8.38 8.32

SEM N binder

TS

0.9515 0.8555

TKN

NH4ⴙ–N

g/L 3.95 1.26 3.98 1.20 3.95 1.44 4.02 1.51 3.97 1.17 4.06 1.27 3.98 1.45 3.95 1.52 3.98 1.09 4.01 1.43 4.04 1.39 3.91 1.48 0.03 0.05 Main effect means‡ 3.98 1.35 3.99 1.35 3.99 1.35 3.97AB 1.17A 4.02B 1.30B 3.99AB 1.42C 3.96A 1.51C Main effect significance levels (P )§

0.8883 0.7913

0.7167 0.2648

0.9207 ⬍0.0001

NH4ⴙ–N

Headspace NH3–N

% of TKN 31.9 30.2 36.4 37.6 29.3 31.4 36.4 38.4 27.5 35.6 34.3 38.0 1.2

mg/m3 0.20 2.0 1.1 1.5 0.22 1.6 1.2 1.1 0.22 1.5 0.78 0.95 0.27

34.0 33.9 33.8 29.6A 32.4B 35.7C 38.0D

1.2 1.0 0.86 0.21A 1.7C 1.0B 1.2B

0.8800 ⬍0.0001

0.0513 ⬍0.0001

† Initial manure composition data were as follows: total solids (TS) content ⫽ 11.07%, volatile solids (VS) content ⫽ 8.86%, total Kjeldahl nitrogen (TKN) content ⫽ 3.82 g/L, ammonium nitrogen (NH4ⴙ–N) content ⫽ 1.14 g/L, ammonium nitrogen, percent of total Kjeldahl nitrogen (NH4ⴙ–N, % of TKN) ⫽ 29.9%. For each dose–time combination, n ⫽ 2. ‡ Main effect means within a column having different letters differ significantly (P ⬍ 0.05). § Main effects were determined with treatment as a continuous independent variable. Therefore the interaction of the main effects was not included in the statistical model.

This observation is in contrast to Varel et al. (1999) who observed that increases in ammonia concentrations in cattle feedlot manure were prevented for a five-week time period with the repeated application of NBPT. Differences between the studies probably stem from the frequency of application of the NBPT and the occurrence of the first application. In this study, NBPT was not added until after urine and feces had come in contact with each other and was added only once, at the onset of the storage period. In the current study, urea had the opportunity to volatilize as ammonia before manure collection. However, analyses of a subset of samples from the post-storage containers revealed that low concentrations of urea remained present in the manure samples (25–50 mg/L wet basis, compared with 90 mg/L for freshly mixed then frozen urine and feces and 6000 mg/L for urine only). Headspace ammonia N concentration increased with storage time (Table 4). The effect paralleled the pattern of ammonium N concentration, with vessels stored for 96 h having greater headspace ammonia N concentration than vessels stored for 24, 48, or 72 h. Slurry ammonium N concentration, but not ammonium N as a percentage of TKN, was positively correlated with headspace ammonia N concentration (r ⫽ 0.64, P ⬍ 0.05). The regression equations of ammonia N concentration against dose of NBPT show increasing functions at 24, 72, and 96 h and a lack of response to NBPT at 48 h (Fig. 1b). Except for vessels stored for 48 h, the slopes increase with storage time (3.20, 9.99, and 22.1 mg/m3 greater concentration of ammonia N for every additional microliter of the NBPT product added to a liter

of slurry for 24, 72, and 96 h lines, respectively; R2 ⫽ 0.76, P ⫽ 0.001). The significantly greater slope at 96 h compared with 24 and 48 h is evidence of a greater dose effect with increased storage time. The results of this experiment do not support the function of NBPT as a urease inhibitor, or give evidence of any effect of NBPT on urease activity. Rather, they suggest the facilitation of ammonia volatilization at doses exceeding the recommendation and no effect on any parameter at the recommended dose, relative to the control. The facilitated release of ammonia into the headspace was related to increased ammonium N concentrations in the slurry, although the variation in ammonium N concentration was independent of NBPT dose. Closer examination of the effect of NBPT when applied to freshly excreted manure is needed. However, with respect to swine manure storage, uniform application of a urease inhibitor on a regular basis may prove cumbersome, providing less than optimal results in practice.

Segregation Effects Urine samples were not analyzed for TS or VS concentrations (Table 5). There were no main effects of storage time on TKN concentration, pooled across substrates. However, urine, feces, and the mixture of urine and feces varied in TKN concentration (Table 5). Similarly, the concentration of ammonium N and ammonium N, as a percentage of TKN, also differed by substrate with substantially more ammonium N in the mixture than in either urine or feces, alone. This pattern was observed even ini-

1125



Fig. 1. Storage time and treatment effects on headspace ammonia N concentrations. Regression equations fit for (a) dose of N binder (yucca extract) (0–14.9 mg/L) P ⫽ 0.0513, hour P ⬍ 0.0001 and (b) dose of urease inhibitor N-(n-butyl) thiophosphoric triamide (NBPT) (0–152 ␮L/L) P ⫽ 0.0131, hour P ⫽ 0.0010.

1126


Table 4. Least squares means of urease inhibitor N-(n-butyl) thiophosphoric triamide (NBPT) and storage time effects on swine manure composition† and headspace ammonia N concentration. Urease inhibitor


␮L/L 0

76

152

Time post-storage h 24 48 72 96 24 48 72 96 24 48 72 96

Storage time


0 76 152 24 48 72 96

␮L/L ␮L/L ␮L/L h h h h

VS %

13.44 11.82 11.01 12.04 9.72 10.84 8.72 11.99 11.19 9.91 11.01 13.27 1.73

SEM Urease inhibitor

TS

12.08 10.32 11.35 11.45 10.86 10.25 12.43 0.5232 0.4046

TKN

NH4ⴙ–N

g/L 4.58 3.77 3.63 4.16 3.57 3.19 3.05 4.18 3.72 3.24 3.62 4.56 0.61 Main effect means‡ 9.68 4.04 8.31 3.50 9.14 3.79 9.19 3.96 8.86 3.40 8.26 3.43 9.87 4.30 Main effect significance levels

1.48 1.33 1.42 1.24A 1.30A 1.39A 1.72B (P )§

0.5392 0.4636

0.5827 0.0034

10.75 9.57 8.81 9.60 7.73 8.93 7.08 9.50 9.10 8.06 8.88 10.50 1.33

0.5176 0.1506

1.36 1.37 1.50 1.70 1.13 1.27 1.22 1.70 1.22 1.26 1.46 1.76 0.17

NH4ⴙ–N

Headspace NH3–N

% of TKN 29.7 36.7 42.7 40.9 32.0 40.6 40.0 40.7 33.2 39.3 40.7 38.6 2.4

mg/m3 0.23 0.94 1.1 1.9 1.0 0.97 1.5 2.7 0.72 0.99 2.7 5.2 0.66

37.5 38.3 37.9 31.6A 38.9B 41.1B 40.1B 0.7774 0.0001

1.0a 1.5a 2.4b 0.66A 0.97A 1.7A 3.2B 0.0131 0.0010

† Initial manure composition data were as follows: total solids (TS) content ⫽ 10.33%, volatile solids (VS) content ⫽ 8.21%, total Kjeldahl nitrogen (TKN) content ⫽ 3.69 g/L, ammonium nitrogen (NH4ⴙ–N) content ⫽ 1.15 g/L, ammonium nitrogen, percent of total Kjeldahl nitrogen (NH4ⴙ–N, % of TKN) ⫽ 31.5%. For each dose–time combination, n ⫽ 2. ‡ Main effect means within a column having different letters differ significantly (P ⬍ 0.05). § Main effects were determined with treatment as a continuous independent variable. Therefore the interaction of the main effects was not included in the statistical model.

tially (3.29, 0.81, and 0.73 g/L and 53.0, 16.1, and 9.3% of TKN for mixture, feces, and urine, respectively), with the mixture not intermediate as would have been expected. Upon mixture of the urine and feces and before analyses of the initial samples, there was significant formation of ammonium N in the slurry from non-ammonium TKN. The post-storage ammonium N concentration and ammonium N portion of TKN increased with increasing storage time (Table 5), with significant accumulation of ammonium N during each day of storage. The vessels containing the mixture steadily increased in ammonium N concentration between 24 and 96 h (Table 5). Vessels containing feces did not differ between 24 and 96 h of storage time (mean 0.92 g/L). For vessels containing urine, the only storage time effect on ammonium N concentration was 96 h (0.90 g/L) compared with vessels stored for 24 h (0.73 g/L; P ⫽ 0.0256). Urine had the greatest TKN concentration initially, but the formation of ammonium N from other TKN was slowed in the absence of feces. For ammonium N as a percentage of TKN, neither urine nor feces vessels differed between 24 and 96 h of storage time, but the vessels containing the mixture increased between 24 and 72 h (Table 5), and were not significantly greater at 96 h (70.2%) than at 72 h (67.9%, P ⫽ 0.0746). The greater ammonium N concentration of the mixture compared with that of feces and urine, separately, was reflected in the headspace ammonia N concentrations. The mixture volatilized at least 103 times more am-

monia N than the separate feces and urine (137 vs. 1.32 and 0.66 mg/m3; P ⬍ 0.0001; Table 5). At 24 h, the concentration of ammonia volatilized from feces was undetected (⬍0.25 ppm). For the mixture, all headspace ammonia N concentrations were positive, but not different according to storage time, suggesting that the ammonia was volatilized during the first 24 h, with no significant release in the remaining time periods. These findings are consistent with summarized findings of Sommer and Hutchings (2001). Despite high headspace ammonia N concentrations in the vessels containing the mixture, the movement of this nitrogen from the slurry to the headspace was insufficient to decrease final TKN slurry concentrations, possibly due to the fraction of TKN that was in the form of organic N or because the closed environment used in this study prevented a sufficient gradient in ammonia to cause significant loss of nitrogen to the headspace. Headspace ammonia N concentrations in this trial were highly correlated with slurry ammonium N concentrations and ammonium N as a percentage of TKN (r ⫽ 0.96, P ⬍ 0.0001 and r ⫽ 0.96, P ⬍ 0.0001, respectively). These very strong correlations are evidence of the effects mixing urine with feces has on both ammonium N production and the potential for ammonia N volatilization. Ammonium N production in manure largely requires the presence of both the substrate urea N from urine and the catalyst urease, which is present in feces. Production of ammonia N is driven by concentration of ammonium N and high pH, which is maintained by

1127


Table 5. Least squares means of segregation and storage time effects on swine manure composition† and headspace ammonia N concentration. Time post-storage

Substrate

TS‡

h


Urine†

Feces†

As-excreted mixture†

Substrate Storage time


%

24 48 72 96 24 48 72 96 24 48 72 96

– – – – 14.50 13.80 13.44 16.53 7.14 7.22 6.86 7.33 1.34

urine feces as-excreted mixture 24 h 48 h 72 h 96 h

– 14.57b 7.14a 10.82 10.51 10.15 11.93

SEM

VS‡

⬍0.0001 0.6018 0.7406

TKN

NH4ⴙ–N

g/L – 7.87 – 7.97 – 8.05 – 7.77 12.71 5.42 12.09 5.09 11.76 5.16 14.62 5.77 5.96 6.08 6.02 6.35 5.67 6.33 6.13 6.33 1.23 0.19 Main effect means¶ – 7.91c 12.79b 5.36a 5.95a 6.27b 9.34 6.46 9.05 6.47 8.71 6.51 10.38 6.62 Main effect significance levels (P ) ⬍0.0001 0.5886 0.7393

⬍0.0001 0.6890 0.2363

0.73 0.80 0.80 0.90 0.91 0.86 0.92 0.98 3.26 4.00 4.30 4.44 0.05

NH4ⴙ–N % of TKN 9.3 10.1 10.0 11.6 16.7 16.9 17.9 17.0 53.7 63.1 67.9 70.2 0.8

0.81a 0.92b 4.00c 1.63A 1.89B 2.01C 2.11D

⬍0.0001 ⬍0.0001 ⬍0.0001

Headspace NH3–N mg/m3 0.17 1.2 0.46 0.84 ⬍LOD§ 0.36 1.2 2.4 147 127 142 133 12

10.3a 17.1b 63.7c 26.6A 30.0B 31.9C 32.9C

0.66a 1.32a§ 137b 74§ 43 48 45

⬍0.0001 ⬍0.0001 ⬍0.0001

⬍0.0001 0.8738 0.9517

† Initial urine composition data were as follows: total Kjeldahl nitrogen (TKN) content ⫽ 7.97 g/L, ammonium nitrogen (NH4ⴙ–N) content ⫽ 0.73 g/L, ammonium nitrogen, percent of total Kjeldahl nitrogen (NH4ⴙ–N, % of TKN) ⫽ 9.3%. Initial feces composition data were as follows: total solids (TS) content ⫽ 13.18%, volatile solids (VS) content ⫽ 11.50%, total Kjeldahl nitrogen (TKN) content ⫽ 5.02 g/L, ammonium nitrogen (NH4ⴙ–N) content ⫽ 0.81 g/L, ammonium nitrogen, percent of total Kjeldahl nitrogen (NH4ⴙ–N, % of TKN) ⫽ 16.1%. Initial urine ⫹ feces mixture composition data were as follows: total solids (TS) content ⫽ 7.55%, volatile solids (VS) content ⫽ 6.34%, total Kjeldahl nitrogen (TKN) content ⫽ 6.21 g/L, ammonium nitrogen (NH4ⴙ–N) content ⫽ 3.29 g/L, ammonium nitrogen, percent of total Kjeldahl nitrogen (NH4ⴙ–N, % of TKN) ⫽ 53.0%. For each substrate–time combination, n ⫽ 2. ‡ Total solids and volatile solids contents of urine were not measured. § Two vessels were below the limit of detection (0.25 ppm). ¶ Main effect means within a column having different letters differ significantly (P ⬍ 0.05).

the buffering capacity of swine feces (Molloy and Tunney, 1983). Based on the observed values for urine and feces separately, calculated values for ammonium N concentration and headspace ammonia N concentration of the mixture are 0.87 g/L and 1.01 mg/m3, assuming manure is 53.4% feces and 46.7% urine (American Society of Agricultural Engineers, 2003). However, these calculated values represent only 22 and 0.7% of the observed values for the mixture because the calculated values, based on observations when urine and feces are segregated, do not account for the impact that mixing fecal urease with urinary urea has on liberation of urea to form ammonium N and then transform into ammonia. Manure management systems that completely prevent swine urine and feces from mixing have the potential to decrease ammonia volatilization by more than 99% (137 vs. 1.32 and 0.66 mg/m3; P ⬍ 0.0001; Table 5), in large part because of decreases in ammonium N formation. These findings are similar to experimental-scale systems that have compared their ammonia concentrations to published data from commercial systems. Kaspers et al. (2002) reported a 65% reduction in indoor ammonia concentrations compared with reported literature values. More recently, von Bernuth et al. (2004) measured ammonia concentrations no greater than 7.5 ppm using the Dra¨ger tubes. Neither of these studies had

control systems for comparison. However, both studies indicate that our results are similar to that observed in a full-scale system.

pH Effects Final manure pH varied according to the treatments prescribed (Table 6). Numerically, the final pH of the vessels were different from initial pH (Table 6). The pH changes may be attributed to the relative volatilization of ammonia, volatile fatty acids, carbon dioxide, and pH-adjusting additives, as well as buffering effects of the microbial population (Sommer and Husted, 1995; Husted et al., 1991). Total and volatile solids concentrations were not affected by treatment, and were not different between 24 and 96 h of storage (Table 6). The concentration of TKN changed with time stored (Table 6), but not pH. What caused TKN concentrations to increase between 24 and 72 h then decrease from 72 to 96 h is not evident. Ammonium N concentration and ammonium N as a percentage of TKN both increased over time and with increased pH (Table 6). A significant pH by time interaction effect (P ⬍ 0.0001 and P ⫽ 0.0013 for ammonium N concentration and ammonium N as a percentage of TKN, respectively) was observed. These results suggest that alkaline conditions facilitated the formation of am-

1128


Table 6. Least squares means of pH and storage time effects on swine manure composition† and headspace ammonia N concentration. pH

Time post-storage

End pH

h


5.30

6.59

8.85

pH Storage time


VS

TKN

5.41 5.33 5.32 5.27 6.23 6.27 6.25 6.25 6.99 7.01 7.05 6.97 0.02

5.30 6.59 8.85 24 h 48 h 72 h 96 h

5.33a 6.25b 7.00c 6.21B 6.20AB 6.20B 6.16A ⬍0.0001 0.0962 0.0632

g/L 9.05 3.10 7.98 3.11 8.18 3.24 7.58 3.16 7.36 2.96 8.35 3.18 7.63 3.31 7.40 3.19 8.89 3.08 7.42 3.20 7.69 3.20 7.35 3.13 0.52 0.05 Main effect means§ 9.22 8.20 3.15 8.68 7.69 3.16 9.00 7.84 3.15 9.54B 8.43 3.05A 8.97AB 7.92 3.17BC 8.89AB 7.83 3.25C 8.47A 7.44 3.16B Main effect significance levels (P )

NH4ⴙ–N

NH4ⴙ–N

1.16 1.15 1.17 1.19 1.13 1.28 1.33 1.37 1.26 1.42 1.46 1.51 0.01

% of TKN 37.3 36.8 36.0 37.5 38.2 40.1 40.0 42.9 41.0 44.2 45.6 48.4 0.6

%

24 48 72 96 24 48 72 96 24 48 72 96

SEM

TS 10.13 8.99 9.20 8.55 8.33 9.38 8.63 8.40 10.15 8.53 8.83 8.47 0.56

0.4247 0.1946 0.3977

0.3902 0.1948 0.4092

0.9760 0.0020 0.1927

1.16a 1.28b 1.41c 1.18A 1.28B 1.32C 1.36D ⬍0.0001 ⬍0.0001 ⬍0.0001

36.9a 40.3b 44.8c 38.8A 40.4B 40.5B 42.9C ⬍0.0001 ⬍0.0001 0.0013

Headspace NH3–N mg/m3 ⬍LOD‡ 0.26 0.39 0.19 0.45 1.0 0.92 1.0 2.7 10 8.4 8.5 1.2 0.28a‡ 0.84a 7.4b 1.6‡ 3.8 3.2 3.2 ⬍0.0001 0.0413 0.2015

† Initial manure composition data were as follows: total solids (TS) content ⫽ 9.73%, volatile solids (VS) content ⫽ 8.58%, total Kjeldahl nitrogen (TKN) content ⫽ 3.34 g/L, ammonium nitrogen (NH4ⴙ–N) content ⫽ 1.11 g/L, ammonium nitrogen, percent of total Kjeldahl nitrogen (NH4ⴙ–N, % of TKN) ⫽ 33.2%. For each pH–time combination, n ⫽ 2. ‡ Two vessels were below the limit of detection (0.25 ppm). § Main effect means within a column having different letters differ significantly (P ⬍ 0.05).

monium N from organic nitrogen, and that across pH, ammonium N built up over time. When manure was acidified, there was a greater ammonium N concentration at 96 h (1.19 g/L) than at 48 h (1.15 g/L; P ⫽ 0.0220). While manure that was unadjusted or alkalized showed steady increases in the percentage of TKN as ammonia N over time, no storage time differences (P ⬎ 0.05) were observed. Headspace ammonia N concentrations increased more than sevenfold when manure was alkalized (7.4 mg/m3) compared with when it was unadjusted or acidified (0.84, 0.28 mg/m3; Table 6). Ammonia volatilization continued long enough within the alkalized vessels to demonstrate an increase in headspace ammonia N concentration with storage time (Table 6). Headspace ammonia concentrations were undetected (⬍0.25 ppm) until 48 h in vessels containing acidified manure, and not different from zero in all acidified and unadjusted vessels. For the alkalized vessels, headspace ammonia N accumulation was positive at all storage times. Acidification did not reduce headspace ammonia N concentrations relative to the control (unadjusted; pH 6.59). However, acidification did maintain a reduced pH and prevented ammonium N concentration from increasing in the slurry during storage, which may contribute to significantly reduced ammonia volatilization potential when manure is stored for longer than 96 h. Positive correlations between headspace ammonia N concentrations and ammonium N concentrations, ammonium N as a percentage of TKN, and final pH (r ⫽ 0.79, P ⬍ 0.0001; r ⫽ 0.78, P ⬍ 0.0001; and r ⫽ 0.74, P ⬍ 0.0001, respectively) provide support.

General In none of the vessels was a large portion of the initial TKN (⬎0.004%) or ammonium N (⬎0.007%) detected in the post-storage headspace ammonia N. This may be because vessels remained closed until their storage time had expired. In all trials except the segregation trial, manure was collected after urine and feces had been allowed to mix, providing the opportunity for volatilization losses before experiment initiation. Even in the segregation trial, there was evidence that significant ammonium production had occurred in the time that elapsed between mixing the substrates and analyzing the bucket samples. Where TKN losses were observed, the corresponding headspace ammonia N accounted for a maximum of 0.8% of the nitrogen loss suggesting additional forms (unmeasured) of nitrogen loss. If this relationship holds for more conventional storage periods, it means that decreases in nitrogen content during slurry storage are probably the result of more than just ammonia volatilization; that is, other forms of nitrogen may be lost. In all trials, headspace ammonia N concentrations were positively correlated with slurry ammonium N concentrations. Headspace ammonia N concentrations were generally very sensitive to increases in slurry ammonium N concentrations. All significant treatment differences in ammonium N concentration corresponded with at least sevenfold greater increases in headspace ammonia N concentration, on a percentage change basis. This phenomenon may have resulted from additive effects of ammonium production on both the ammonium N concentration and pH of the slurry. Increases in both of


these factors facilitate the dissociation of ammonium and its subsequent volatilization as ammonia.


CONCLUSIONS Warmer temperatures increased TS and VS degradation, ammonium N formation, and ammonia volatilization. In a review of factors affecting ammonia emissions from field-applied manure, Sommer and Hutchings (2001) depict that as much as 50% of total nitrogen is volatilized as ammonia at a temperature of 30⬚C compared with 35% when the temperature is 25⬚C; a nonlinear response. In this study, the linear regression equation suggests that, using those temperatures, a five-degree temperature differential would have a greater impact, possibly due to the contained environment under which this study was conducted. However, the actual data support a nonlinear response such as that reported by Sommer and Hutchings (2001) that could not be evaluated using the three temperature points studied. Stirring facilitated both ammonium N production and ammonia N release. Yucca extract had no effect on ammonium N production, but decreased ammonia volatilization in proportion with the storage time and dose. Nitrogen binding may have been the mode of action, but further details on the way in which ammonia volatilization was prevented by yucca were not elucidated. N-(n-butyl) thiophosphoric triamide (NBPT) did not act as a urease inhibitor. Addition of NBPT had no effect on any parameter at the recommended dose, and increased ammonia volatilization at twice the recommended dose. The amount of ammonia released per dose of NBPT added increased with storage time. Segregation of urine from feces at excretion showed great potential as a strategy to prevent most of the initial formation of ammonium N and the subsequent ammonia N losses. Acidification decreased ammonium N concentration and ammonium N as a percentage of TKN. Alkalization increased headspace ammonia N concentrations. Once ammonium N is present in high concentrations, there are many factors that can affect its release as ammonia N. All of these factors need to be considered to prevent large losses of ammonia nitrogen. Therefore, prevention of ammonium N formation is a key intervention point for management strategies. Sommer and Hutchings (2001) report that approximately 50% of ammonia emissions occur within the first 24 h following land application, with emissions continuing for up to 10 d following application. Others have reported that emissions are greatest during the first 6 h following application (Chantigny et al., 2004). However, substantial emissions occur during storage of manure (Lorimor et al., 2000) suggesting that strategies to address emissions immediately after voiding will have benefit. Managing urine and feces separately may be a very promising strategy, depending on the feasibility of using the necessary facilities. For manure handled more conventionally, managers might consider keeping manure at a cool temperature, unstirred, and at a low pH to keep ammonium N concentrations low. The subsequent volatilization of any ammoniacal nitrogen that has formed may be partially prevented by addition of yucca to the slurry.

1129

The assumption that ammonia N volatilization is responsible for most of the nitrogen loss during slurry storage was probably not valid under the conditions of this experiment. It is not known what impact freezing the manure sources before the experiment had on microbial nitrogen processes. Further research using direct measurement of other volatilized nitrogen compounds is needed to determine the major forms of nitrogen lost from manure during storage. The design of this experiment only allowed for comparison of different levels of one strategy at a time. Further investigation is needed to compare the potential of different strategies to each other and to evaluate the effects of combinations of strategies. Furthermore, the use of more dynamic equipment that allows repeated measurements of a particular storage vessel over time is needed to validate the storage time effects observed with this static method. ACKNOWLEDGMENTS The authors thank the Iowa Pork Producers Association for financial support of this investigation, and Laura Flatow and Martha Jeffrey for assistance with laboratory analyses.

REFERENCES American Society of Agricultural Engineers. 2003. Standard D384.1: Manure production and characteristics. ASAE, St. Joseph, MI. AOAC International. 2002. Official methods of analysis. 17th ed. Rev. 1. AOAC Int., Gaithersburg, MD. Battye, R., W. Battye, C. Overcash, and S. Fudge. 1994. Development and selection of ammonia emission factors. Final report. USEPA Office of Res. and Development, Washington, DC. Chantigny, M.H., P. Rochette, D.A. Angers, D. Masse, and D. Cote. 2004. Ammonia volatilization and selected soil characteristics following application of anaerobically digested pig slurry. Soil Sci. Soc. Am. J. 68:306–312. Ferguson, N.S., R.S. Gates, J.L. Taraba, A.H. Cantor, A.J. Pescatore, M.L. Straw, M.J. Ford, and D.J. Burnham. 1998. The effect of dietary protein and phosphorus on ammonia concentration and litter composition in broilers. Poult. Sci. 77:1085–1093. Headon, D.R., K. Buggle, A. Nelson, and G. Killeen. 1991. Glycofractions of the yucca plant and their role in ammonia control. p. 95–108. In T.P. Lyons (ed.) Biotechnology in the feed industry. Alltech, Nicholasville, KY. Hobbs, P.J., T.H. Misselbrok, and B.F. Pain. 1998. Emission rates of odorous compounds from pig slurries. J. Sci. Food Agric. 77:341–348. Husted, S., L.S. Jensen, and S.S. Jorgensen. 1991. Reducing ammonia loss from cattle slurry by the use of acidifying additives: The role of the buffer system. J. Sci. Food Agric. 57:335–349. Kaspers, B., J. Koger, and T. van Kempen. 2002. Evaluation of a conveyor belt waste collection system for swine: Fecal drying efficiency and ammonia emission reductions. Dep. of Animal Sci., North Carolina State Univ., Raleigh. Kerr, B.J. 1995. Nutritional strategies for waste reduction management. p. 47. In New horizons in animal nutrition and health. Inst. of Nutrition, North Carolina State Univ., Raleigh. Lefcourt, A.M. 2001. Large environmental chamber: Ammonia recovery calibration. Appl. Eng. Agric. 17(5):683–689. Lowe, P.D. 1995. Social issues and animal wastes: A European perspective. p. 168–171. In Proc. Int. Livestock Odor Conf., Ames, IA. 16–18 Oct. 1995. Iowa State Univ., Ames. Mackie, R.I., P.G. Stroot, and V.H. Varel. 1998. Biochemical identification and biological origin of key odor components in livestock waste. J. Anim. Sci. 76:1331–1342. Mobley, H.L.T., M.D. Island, and R.P. Hausinger. 1995. Molecular biology of microbial ureases. Microbiol. Rev. 59:451–480. Molloy, S.P., and H. Tunney. 1983. A laboratory study of ammonia volatilisation from cattle and pig slurry. Ir. J. Agric. Res. 22:37–45.


1130


Muck, R.E., and T.H. Steenhuis. 1981. Nitrogen losses in free stall dairy barns. p. 406. In Livestock waste: A renewable resource. Proc. 4th Int. Symp. on Livestock Wastes, Amarillo, TX. 15–17 Apr. 1980. Am. Soc. Agric. Eng., St. Joseph, MI. Lorimor, J., W. Powers, and A. Sutton. 2000. Manure characteristics. MWPS-18. MidWest Plan Service, Ames, IA. National Oceanic and Atmospheric Administration. 2003. Past weather. Unedited local climatological data [Online]. Available at www.noaa. gov/pastweather.html (verified 13 Jan. 2005). NOAA, Asheville, NC. Power, J.F., B. Eghball, and J.A. Lory. 1994. Utilization of nutrients in beef cattle feedlot manure in the northern great plains. p. 161– 167. In Balancing animal production and the environment. Proc. Great Plains Animal Waste Conf. on Confined Animal Production and Water Quality, Denver. 19–21 Oct. 1994. GPAC Publ. 151. Great Plains Agric. Council, Fort Collins, CO. SAS Institute. 1990. SAS user’s guide: Statistics. Version 6a. SAS Inst., Cary, NC. Sommer, S.G., and S. Husted. 1995. The chemical buffer system in raw and digested animal slurry. J. Agric. Sci. 124:45–53. Sommer, S.G., and N.J. Hutchings. 2001. Ammonia emission from

field applied manure and its reduction—Invited paper. Eur. J. Agron. 15:1–15. Sutton, A.L., K.B. Kephart, M.W. Verstegen, T.T. Canh, and P.J. Hobbs. 1999. Potential for reduction of odorous compounds in swine manure through diet modification. J. Anim. Sci. 77:430–439. USEPA. 1997. National ambient air quality standards for particulate matter. Final rule. Fed. Regist. 62:38651. USEPA. 2004. National emission inventory—Ammonia emissions from animal husbandry operations. Draft report [Online]. Available at www.epa.gov/ttn/chief/ap42/ch09/related/nh3inventorydraft_jan2004. pdf (verified 13 Jan. 2005). USEPA, Washington, DC. Van Horn, H.H., G.L. Newton, and W.E. Kunkle. 1996. Ruminant nutrition from an environmental perspective: Factors affecting whole-farm nutrient balance. J. Anim. Sci. 74(12):3082–3102. Varel, V.H., J.A. Nienaber, and H.C. Freetly. 1999. Conservation of nitrogen in cattle feedlot waste with urease inhibitors. J. Anim. Sci. 77:1162–1168. Von Bernuth, R.D., J.D. Hill, E. Henderson, S. Godbout, D. Hamel, and F. Pouliot. 2004. Efficacy of a liquid/solid isolation system for swine manure. Paper 044131. In ASAE Annual Int. Meeting, Ottawa, ON, Canada. 1–4 Aug. 2004. Am. Soc. Agric. Eng., St. Joseph, MI.