Published online 20 November 2006
Published in Vadose Zone J 5:1236-1245 (2006)
DOI: 10.2136/vzj2006.0073
© 2006 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
ORIGINAL RESEARCH
Nitrous Oxide and Ammonia Emissions from Urine-Treated Soils
Texture Effect
Olga Singurindy,
Brian K. Richards*,
Marina Molodovskaya and
Tammo S. Steenhuis
Dep. of Biological and Environmental Engineering, Riley-Robb Hall, Cornell Univ., Ithaca, NY 14853
* Corresponding author (bkr2{at}cornell.edu)
Received 16 May 2006.
 |
ABSTRACT
|
|---|
Urine-treated soils have been implicated as an important source of gaseous N losses to the atmosphere. The goal of our research was to quantify NH3 and N2O emissions from urine under different simplified and controlled soil conditions and to relate these results to urinary-N transformation processes in soil. We studied the influence of soil texture (coarse vs. fine sand), moisture distribution with depth, air-filled pore space, and rate of air movement, which affected both soil drying processes and emission rates. Laboratory experiments were performed with synthetic urine in both aerobic and anaerobic conditions. Texture was the most important factor controlling NH3 volatilization and N2O emission factors in urine-treated sands. Generally, the finer the sand texture, the higher the input of denitrification to the total N2O emissions; however, the air-filled pore space threshold, below which denitrification became dominant, was greater in coarse sand. In addition, the evaporation rate of the urine water component was found to be an important parameter in total NH3 and N2O emissions. Ammonia volatilization occurred during the first day of the treatments, with volatilization rates closely related to evaporation rates in both sand textures. Finer sand texture caused reductions in the urine evaporation rate and therefore in NH3 volatilization rates. Urine evaporation increased air-filled pore space, thereby improving aeration conditions in the sand that contribute to nitrification dominance of N2O production. Moreover, evaporation of urine, enriched with dissolved N2O, increased total N2O emission. Results from these simplified experiments need confirmation under field soil conditions.
|
INTRODUCTION
|
|---|
GLOBAL agricultural N inputs to the atmosphere now exceed those from natural sources. Agriculture contributes up to 70% of N2O (Janzen et al., 1998) and approximately 90% of NH3 emissions (Boyer et al., 2002). In dairy farming, the main source of gaseous N losses is the spreading of animal waste. More than half of the excreted N from the animals occurs in urine. Urine deposition is an especially important source of N in grazed pastures where N turnover in urine patches is intense and the potential for environmental losses is significant (Petersen et al., 2004).
Urea, the main compound of urine, rapidly hydrolyzes to NH4+ when brought in contact with soil, resulting in subsequent NH3 volatilization (Haynes and Williams, 1992). Volatilized NH3 is a chemically active gas and readily combines with HNO3 and H2SO4 in cloud droplets, which prolongs their existence in the atmosphere and therefore influences the geographic distribution of acidic depositions. A large fraction of emitted NH3 is redeposited close to the source, while the remainder reacts with acidic gases to form fine particles that are transported across long distances. In contrast to NH3, N2O is a greenhouse gas that contributes to global warming (Duxbury, 1994). This gas is also oxidized in the stratosphere to form nitrogen oxides (NOx), which participate in photochemical processes that destroy ozone (Crutzen, 1981). Nitrous oxide is produced in soil by microbiological nitrification and denitrification processes. Due to the different effects of N2O and NH3, it is difficult to compare their relative environmental impacts, but emissions of both have to be mitigated.
Urine-N is rapidly available to plants, in contrast to fecal N that mineralizes over a number of years (Deenen and Mindelkoop, 1992). Urinary transfer in soil is complex, characterized by interactions between urine input and soil and environmental factors. A number of field studies examined gaseous NH3 emission from cattle urine-affected soils, finding a variation in extent caused by a number of soil properties such as cation exchange capacity, the initial soil pH, and nitrification activity. In general, the proportion of total urinary N volatilized as NH3 has ranged from about 4 to 27% in the UK (Lockyer and Whitehead, 1990), from about 12 to 36% in New Zealand (e.g., Sherlock and Goh, 1984), and from about 4 to 18% in the Netherlands (Vertregt and Rutgers, 1987). Whitehead and Raistrick (1993) found a range of 6.8 to 41.3% of the total urinary N as NH3 in a laboratory study, which was inversely correlated with the soil cation exchange capacity.
The Intergovernmental Panel of Climate Change (2000) identified a default urine-N2O emission factor of 2%, the emission factor being the fraction of N converted to N2O (Velthof et al., 2003). Van Groenigen et al. (2005) reported that the median N2O emission factor from 22 laboratory and field studies was 1.3%. It was agreed that the most important factors controlling the two N2O producing processes (nitrification and denitrification) are soil mineral N (NH4+ and NO3–) concentrations, partial O2 pressure and related water-filled pore space, and, in the case of denitrification, C to fuel the heterotrophic processes (e.g., Clough et al., 2003). Hynes and Knowles (1980) demonstrated that both nitrification and denitrification may occur simultaneously in separate microsites on opposite sides of aerobic–anaerobic interfaces. Abbasi and Adams (1998) and Abbasi et al. (1997) observed concurrent nitrification–denitrification processes in the soils with high moisture content, where N loss depends on the maintenance of the aerobic surface. In this case, nitrification was followed by downward diffusion of NO3– and subsequent denitrification in the anaerobic zone.
It is difficult to predict N2O and NH3 emission patterns because of the complexity of the interrelationships among factors and processes (Müller et al., 1997). The majority of experimental studies investigating urine-treated soils measured the gaseous N emission near the soil surface, with little focus on the processes that produce N2O and NH3 within the soil profile. Soil texture affects aeration conditions (e.g., Dziejowski et al., 1997) and soil drying processes (e.g., Coussot, 2000), causing heterogeneity in soil conditions, which can potentially have a strong influence on both N2O and NH3 emissions from urine-treated soils. Several studies addressed N2O and NH3 emission patterns relative to soil mitigation strategies (e.g., Anger et al., 2003; Van Groenigen et al., 2005). Ball et al. (1999) observed the importance of compaction, which may result in higher anaerobicity through lowering gas diffusivity and increasing the water-filled pore space.
Few studies have been directed toward understanding the basic processes that govern N2O and NH3 emissions from urine. This information is essential to mitigating gaseous-N emission via management practices. The main goal of our research was to quantify NH3 and N2O emissions from urine under different soil conditions and to relate these results to urinary-N transformation processes in soil. Specifically, we studied the influence of soil texture, moisture distribution with depth, air-filled soil pore space, and rate of air movement, which affected both soil drying processes and emission rates.
|
MATERIALS AND METHODS
|
|---|
Two types of experiments were performed at a constant temperature of 25°C, each using two sand types and synthetic urine mixtures, as summarized in Table 1. Experiments consisted of (i) no-flow experiments where the sample units were sealed (static headspace) and (ii) experiments where air was passed along the sample surface (flow-through).
Sand was used as a surrogate for soil to avoid confounding effects of N loss from organic matter that would be present in field soil. Two sand textures were used: coarse (1-mm diameter) and fine (0.25 mm), with a total porosity of 
36% in each case.
The synthetic urine used was composed of urea (at either 7.1 or 11.5 g/L to simulate two levels of dietary N intake), glycine (2.9 g/L), KHCO3 (13.8 g/L), KCl (2.5 g/L), KBr (4.2 g/L), and K2SO4 (1.4 g/L), following de Klein et al. (2003). The concentration of 11.5 g/L of urea is representative of real cow urine patches of 592 kg N/ha (10 L/m2). Natural cow urine (not sterile, as it contained some fecal matter and bedding) was collected from a local dairy farm in New York State, filtered, and added to the synthetic urine at a rate of 0.5% to initiate microbial activity in the system. The sample units were sealed in 14-cm-diam. plastic desiccators (Bel-Art F42010, Bel-Art Products, Pequannock, NJ) packed at sand depths of 1.0 or 5.0 cm.
Static Headspace Experiments
No-flow conditions were chosen for N2O flux measurements so that air inside the chamber was not diluted by external air, thereby giving greater sensitivity for detecting small fluxes. Three subexperiments were conducted.
Subexperiment 1 was focused on the study of the influence of urine-filled pore space on N2O emission in two sand texture systems and two different sand depths. Synthetic urine with 11.5 g/L urea was added to each sand sample to bring the urine-filled pore space to 10, 15, 30, 50, 60, 70, 85, and 90%, and was mixed carefully with sand in separate pans. Then the wetted sand samples were packed into the sample units by layers to prevent formation of air pockets. After sealing, samples were incubated at constant temperature. Incubations were performed under aerobic conditions. Samples of headspace air (6 mL) were taken twice a day. After each sampling, the sample units were opened shortly to room atmosphere and then sealed again. Every second day the sand samples were weighed to monitor the moisture loss by evaporation. The total treatment duration of each sample was 30 d.
In the second subexperiment, N2O emissions from sand samples treated with urine containing different urea concentrations (7.1 and 11.5 g/L) were measured. The experiment consisted of four treatments, which included both sand textures packed to 5-cm depth, mixed with the two urine types added to bring the urine-filled pore space to 15, 45, and 80%. The rest of the procedure and sample analysis were similar to the Subexperiment 1.
In the third subexperiment, the input of nitrification and denitrification processes to the total N2O emission from the sand surface was studied. Sands were mixed with urine (11.5 g/L of urea) to 10, 15, 30, 50, 60, 70, and 85% of urine-filled pore space, and then were packed to 5-cm depth. There was a lack of NO3– for the denitrification process at the beginning of the experiment. To identify the initial stage of incubation when NO3– began to accumulate in the sand–urine system, a series of preliminary tests was performed involving aerobic incubations, which were analyzed for NO3– concentration every 6 h for 4 d. Results demonstrated that NO3– reached its maximum after 2 d (length of the initial stage). Sand samples for the subexperiment were thus incubated aerobically for 2, 5.5, 9, 12.5, 16, 19.5, 23, and 26.5 d. After the aerobic incubation interval, the air inside the desiccators was exchanged with pure N2 and allowed an additional 4 d of anaerobic incubation.
The incubation procedure and N2O analysis were similar to Subexperiment 1. The sum of N2O emitted during the anaerobic part of each of the eight incubation intervals demonstrated the total amount of emitted N2O during denitrification. The difference in N2O emission values between these results and the results obtained from controlled samples, which were incubated under aerobic conditions for 30 d, represented the input of nitrification to N2O production. The same treatment was performed for each of the above-mentioned samples with different urine-filled pore spaces. In addition, sand samples were analyzed for total, organic, and inorganic C contents to check the influence of anaerobic conditions on microbial activity.
Flow-Through Chamber Experiments
The advantage of the flow-through chamber used for Subexperiments 4 and 5 was that it better reproduced field conditions. Subexperiment 4 permitted observation of NH3 volatilization as influenced by horizontal air flux at different urine-filled pore spaces. A schematic diagram of the experimental apparatus is presented in Fig. 1
. Both sand types were mixed with urine (11.5 g/L of urea) to yield 15, 30, and 60% urine-filled pore space. Sample units similar to those for the closed chamber experiments were used. To prevent gas leaks, silicon glue was used to seal the top and bottom of the sample unit parts, which were then clamped between two wood plates. Input and output ports were placed close to the sample surface on opposite sites of the unit lid, directing air into the headspace. The experiments were performed with constant air flow rates using an air pump (TopFin 50) at flow rates of 250 to 2000 mL/min. The air outflow was routed through a plastic cylinder containing 250 mL of 0.025 M H2SO4, where emitted NH3 was absorbed. The total duration of the experiment was 10 d.
Subexperiment 5 studied the influence of horizontal flux and sand drying on N2O-forming processes. Nitrous oxide forming processes were linked to the changes of urine distribution with depth caused by evaporation. Both sand types were mixed with urine (11.5 g/L of urea) for 80% of urine-filled pore space. An experimental apparatus similar to the NH3 volatilization experiments was used (Fig. 1). The only difference was that the bottom part of the sample unit was separated into four equal sectors by vertical plastic strips attached to the bottom, and then the unit was packed with sand to the depth of 5 cm. Twelve holes were drilled on four sides of the sample unit (three holes on each side) to take air samples from the depths of 2, 3.5, and 5 cm in each sector. The holes were sealed with rubber septa and 2-mL air samples were taken with syringes through these holes. Air was injected at a flow rate of 1250 mL/min for 24 h and then was reduced to 125 mL/min; air samples (6 mL) were taken from tubing near input and output ports with syringes every 4 h during the first 2 d and then every 8 h for 14 d. The air and urine samples were analyzed for N2O concentrations. The urine samples were evaporated and the N2O concentration was measured in the gas phase (Stevenson, 1987). Total duration of the experiment was 16 d. Flow through the sample unit was stopped every second day to take the air and urine samples from the headspace and from the different sand depths through the side holes of one sector. The sample unit was then opened and the sand sector where the air samples were taken was evacuated and replaced with pure sand. Afterward the sample unit was closed and the experiment was continued. The evacuated sand sample was sectioned and analyzed for gravimetric moisture content, NO3–, NH4+, and oxidation–reduction potential.
Analysis
Nitrous oxide concentrations were measured by a Varian 3700 gas chromatograph (Varian, Inc., Palo Alto, CA) with a Ni63 electron capture detector. The salicylic acid photometric method (American Public Health Association, 1985) was used to measure NO3– concentration in the solution separated from the sand. Total, organic, and inorganic C contents were determined via persulfate and H3PO4 oxidation using a Model 1010 total organic carbon analyzer (O-I-Analytical, College Station, TX). Ammonia absorbed in 0.025 M H2SO4 solution concentration was determined by spectrophotometric phenate method with a Spectronic 501 (Spectronic Analytical Instruments, Garforth, UK) (American Public Health Association, 1985). Air relative humidity was measured gravimetrically using anhydrous Ca2SO4 (Drierite, Xenai, OH). Solution analysis for NO3– and NH4+ were performed according to the American Public Health Association (1985). Oxidation–reduction potential (redox) was measured by redox monitor (American Marine Inc., Ridgefield, CT). The pH measurements were performed using pH meter Model AR50 (Fisher Scientific, Hampton, NH). To correct redox to a common pH, the normalization factor of –59 mV per unit pH was used (Bohn et al., 1979).
Statistical Analysis
The error in measured concentration values within each experiment was in all cases less than 0.5%. All experiments were performed and reported in triplicate. Mean values and standard deviations of experimental data were calculated using Microsoft Excel. Standard error bars are included on the graphs to show the variation of the data but in some cases are not apparent because the variation was less than the size of the plotting symbol. A probability level of 5% (P = 0.05) was used to test the statistical significance of texture effects on cumulative NH3 volatilization and N2O emission during different treatments.
|
RESULTS AND DISCUSSION
|
|---|
As outlined above, there were different time scales for the NH3 volatilization and N2O emission processes, therefore our results are presented thematically rather than chronically. We first examine the evolution of NH3 volatilization as influenced by changes in sand and urine properties, and then consider N2O emissions.
Ammonia Volatilization (Subexperiment 4)
In general, the proportion of total urinary N volatilized as NH3 ranged from about 7 to 27% depending on the type of treatment. Similar ranges were reported in previous studies for NH3 volatilization from urine-treated soils (Whitehead and Raistrick, 1993). As expected, the most NH3 was volatilized during the first day of the treatments.
The pattern of NH3 volatilization was divided into two stages (Fig. 2
), initial and advanced. The initial stage of the experiment was defined as a time interval from the beginning of the experiment during which 85% of the NH3 was volatilized. The NH3 volatilization rate during the initial stage was closely related to the urine evaporation rate (essentially, water component of urine) in both sand types (Fig. 2). The initial stage was shorter for the coarse sand because of the greater intensity of the drying processes and varied between 5 and 7 h and 10 to 12 h in the coarse and fine sands, respectively. In both sand types, the length of the initial stage was limited by the initiation of microbial activity caused by nitrification. This process resulted in decreasing NH4+ concentration in the urine (Fraser et al., 1994). The fact that urine evaporation limited NH3 volatilization made sand texture the most important factor that influenced this process (Coussot, 2000). The main tendency was that the finer sand texture caused a reduction in the urine evaporation rate and therefore in the NH3 volatilization rate at a given air flow rate and urine-filled pore space.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 2. Cumulative amount of N-NH3 and urine losses with time from fine and coarse sand, 2000 mL/min flow rate, 60% of urine-filled pore space, and 75% air relative humidity. Arrows show the initial stage of the experiment when 85% of the NH3 was volatilized.
|
|
In contrast, during the advanced stage of the experiment, the urine evaporation and NH3 volatilization rates followed different patterns, with the urine evaporation rate being much greater in both sand types. The evaporation rate decreased monotonically during the experiments and then became constant. The NH3 volatilization rate was greater in the fine sand than in the coarse sand, in which it approached zero after 20 to 25 h of incubation. The factor limiting NH3 volatilization rates at the advanced stage of the experiment was transformation of NH4+ by nitrification processes, which depended on urine distribution with depth and therefore on sand aeration conditions, as discussed below.
Figure 3
demonstrates the relationship between air flow rate and the rate of initial NH3 loss. Generally, the NH3 volatilization rate increased with increasing flow rate until reaching a plateau in both sands. Ammonia volatilization rates appeared to be directly proportional to air flow at exchange rates below 1250 mL/min (five headspace volumes per minute) for coarse sand and 1000 mL/min (four volumes per minute) for fine sand. In coarse sand, the initial volatilization rates were greater and the plateau was reached at greater air flow rates. The results demonstrate that the maximum initial rates of NH3 volatilization (2.4%/h loss) and the largest total amount of NH3 emitted during the initial stage (15% loss) were found in the treatments having flow rates >1250 mL/min in coarse sand, which is equivalent to wind speed of 0.5 m/min. The number of volumes per minute when the NH3 emission rate became constant (Fig. 3) was lower than in studies using urea and (NH4)2SO4-treated soils, where plateau was reached at about 15 exchange volumes per minute (e.g., Akiyama et al., 2004; Kissel et al., 1977). This lack of correspondence can be caused by the differences in air relative humidity (RH) circulated along the soil surface in the various studies (Stevenson, 1982). Our results demonstrate that the RH of air flowing along the sand sample had a marked effect on the evaporation process and therefore on the NH3 volatilization rate. When air with a RH of 13% was circulated over the sample, volatilization and evaporation rates were greater than in the cases presented in Fig. 2, where the RH was 75%. In addition, when the RH was 13%, the air flow rate when the NH3 emission rate became constant was greater than in experiments where the RH was 75%. The strong influence of RH on the NH3 volatilization rate makes the soil (sand)–urine system different from soil–manure and soil–fertilizer systems where NH3 loss rates are largely independent of air RH (e.g., Chao and Kroontje, 1964).
View larger version (9K):
[in this window]
[in a new window]
|
Fig. 3. Relationship between the initial NH3 volatilization rate and the air flow rate for two sand types, 60% urine-filled pore space, and 75% air relative humidity.
|
|
The sensitivity of the NH3 volatilization rate to pH changes in urine-treated soils is well known (e.g., Clough et al., 2003). In our specific experimental system, the effect of pH shifts on NH3 volatilization process can be regarded as negligible relative to the other factors, because pH values were stable during the initial stages, and only ranged from 6.6 to 7.1 at the advanced stage of the experiment. The changes in pH might be limited due to the presence of bicarbonates in the system that can potentially buffer fluctuations (Stevenson, 1982).
Nitrous Oxide Emission
No-Flow Experiments (Subexperiments 1, 2, and 3)
Figure 4
demonstrates the air-filled pore space (complementary with urine-filled pore space) distribution with depth in two sand types with 10, 60, and 90% of initial urine-filled pore space. It should be noted that both sands have the same porosity, therefore the same bulk urine content. Nearly identical profiles were observed in the 10 and 90% cases. When the initial urine-filled pore space was 60%, the air-filled pore space variation with depth became significant, with a difference between two sand textures. In the fine sand, the air-filled pore space decreased monotonically with depth, while the coarse sand was close to saturation at the bottom and the top was drier than that of the fine sand. Sand texture has been shown to be one of the most important factors controlling urine distribution in the profile. In addition, the analysis of similar experiments indicated that the maximum differences between two sand types occurred at urine-filled pore space levels between 30 and 80%.
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 4. Distribution of air-filled pore space (complementary with urine-filled pore space) with depth for fine and coarse sands at 10, 60, and 90% of initial urine-filled pore space (5 h after filling).
|
|
Figure 5
demonstrates the total amount of N2O emitted during the experiment from the sand surface as a function of air-filled pore space. There were significant differences in the N2O emission factor between coarse and fine sands in 5-cm-deep samples (Fig. 5a). The range of 20 to 70% air-filled pore space (corresponding to 80–30% initial urine-filled pore space) demonstrates the maximum differences between the two sands. Note that this was the same range where the maximum differences in air-filled pore space distribution with depth were found (see Fig. 4). The maximum N2O emission factor was observed in fine sand, reaching 12.3% at 40% air-filled pore space. In coarse sand, the maximum emission factor of 7.4% was observed at 50% air-filled pore space. The values obtained in this series of experiments were greater than the median 1.3% emission factor from field measurements presented by van Groenigen et al. (2005). The reason for such a difference is that no-flow conditions reduced the NH4+ loss from the system, which normally would have been substantial at the initial stages of incubation (see above).
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 5. A generalized relationship between air-filled pore space and N2O fluxes in fine and coarse sand and two sample thicknesses: (a) 5 cm, and (b) 1 cm; hatched area demonstrates the input of denitrification.
|
|
It is interesting that there was little difference in the relationship between air-filled pore space and N2O fluxes for the two sands when 1-cm-deep profiles were used under the same experimental conditions (Fig. 5b), probably due to the sand depth being insufficient to cause differences in urine distribution and therefore O2 availability. In both cases, N2O emission was observed between 30 and 80% of air-filled pore space and reached a maxima at 60%. The general shape of the curves corresponds to the shape of the relationship between N2O emission and water-filled porosity presented by Davidson et al. (2000) for different soil types (without urine treatment).
In addition, Fig. 5a presents the contributions of nitrification and denitrification to the total N2O emission factor in the 5-cm sand experiments. The hatched area is the input of denitrification to total N2O production in both sands. The general tendency in coarse sand was that nitrification was the dominant process at almost all air-filled porosities where N2O emission was observed. The input of denitrification was observed at air-filled pore space <55%. When anaerobic conditions were more pronounced (air-filled pore space <30%), denitrification input disappeared, probably because of the formation of N2, which is the primary final product of the denitrification process (although direct confirmation of N2 production requires isotopic analysis, Davidson et al., 2000). In contrast to the coarse sand, the N2O production from nitrification was lower in the fine sand. Above 50% of air-filled pore space, nitrification was almost the only source, while the denitrification contribution increased rapidly with decreasing air-filled pore space down to 15%. These results demonstrated, however, that the air-filled pore space below which denitrification became a significant component of the total N2O production ("critical" value) was greater in the coarse sand. The opposite tendency was reported by Pilot and Patrick (1972), who found that the finer the soil texture, the greater critical air-filled pore space it caused. This lack of correspondence was probably due to the bottom of the coarse sand being close to saturation (Fig. 4), which can potentially cause the formation of anaerobic conditions.
The shape of the curves and boundary between nitrification and denitrification contributions shown in Fig. 5 are a function of the depth of the sand sample, as well as sand type, temperature, and urine chemical composition used in any given experiment. Moreover, the method for estimation of nitrification and denitrification inputs to the total N2O emission is based on the number of approximations (see above). Nevertheless, our results clearly indicate the role of texture in the N2O emission factor in urine-treated soils.
Through-Flow Experiments (Subexperiment 5)
Figure 6
presents the time evolution of air-filled pore space with depth when an initial horizontal air flux of 1250 mL/min was applied. The chosen air-flow rate simulates a constant wind speed of 0.5 m/min so that this variable would not limit NH3 volatilization as determined from Fig. 3. In the coarse sand, the urine evaporation process caused an increasing difference in air-filled pore space between the top and bottom of the profile. Most urine evaporated from the top 3-cm layer where the air-filled pore space increased from 40 to 90%, while the bottom of the sample was more saturated, with air-filled pore space increasing from 10 to 40%. Generally, in coarse sand, the air-filled pore space distribution with depth reflected a pronounced aerobic–anaerobic transition zone, which moved deeper with time as evaporation progressed. In contrast, in the fine sand, the air-filled pore space decreased with depth monotonically and urine evaporation simultaneously affected all sand depths, probably because of the greater influence of capillary rise. The total loss of urine by evaporation was 18% greater in the coarse sand. Urine was the only source of N and most of the N2O formation processes occurred in urine, therefore this difference was significant for the urine–sand system.
View larger version (38K):
[in this window]
[in a new window]
|
Fig. 6. Air-filled pore space and N2O emission factor changes with depth and time: (a) fine-textured sand, (b) coarse-textured sand; 1250 mL/min initial air flow rate. Underlined numbers are N2O emission factors in gaseous form. Numbers in parentheses represent N2O dissolved in urine.
|
|
In addition, Fig. 6 presents the amount of emitted and dissolved N2O at the different sand depths. The sum of two numbers indicates the total N2O production. In general, in both sand types, the most gaseous emission of N2O was found within the 40 to 70% range of air-filled pore space. This range corresponds very well to Fig. 5b. The thickness of this horizon (40–70% air-filled pore space) was greater in the fine sand, therefore the production of N2O was more intensive in this sand type. In the fine sand (Fig. 6a), the thickness of this horizon increased from 1 to 3 cm from Days 2 to 8 of the experiment, and then decreased to 1 cm after 16 d. In contrast, in the coarse sand (Fig. 6b), its width was 1 cm throughout the experiment. In both sands, the N2O production horizons moved deeper with time because of the sand drying process.
Nitrous oxide is reasonably soluble in water (56.7 mL N2O/100 mL water at 25°C, Gabel and Schultz, 1973), therefore gaseous emission measurements didn't indicate the total N2O production in sand. Regarding this, in both sands the following tendencies were found (Fig. 6): (i) at air-filled pore space of 30%, the initial formation of N2O in dissolved form was found, (ii) at an air-filled pore space of 40%, the content of dissolved N2O in most cases was higher than the gaseous form, and (iii) at an air-filled pore space of 50%, the amount of dissolved N2O was small, probably because of its fast release to the gaseous form. In addition, the accumulation of N2O in solution was observed in the fine sand at different depths. At a depth of 2.5 cm, it increased up to 0.08% during the first 4 d, at the depth of 3.5 cm it increased from 0.05 to 0.1% during 4 to 8 d, and at the depth of 4.5 cm during 8 to 12 d, it increased from 0.03 to 0.06%. These results demonstrate that the amount of dissolved N2O was not negligible under the experimental conditions.
The rates of N2O production within the sand profile were reflected in the patterns of N2O flux from the sand surface. Figure 7
shows the evolution of N2O emission from the sand surface during the experiment presented in Fig. 6. In the coarse sand, where most N2O was formed along the aerobic–anaerobic interface, there was little change in total N2O emission from the sand surface during the experiment. In contrast, in the fine sand, the N2O emission was significantly greater during Days 6 to 10, reaching a maximum of 0.65% after 7 d. An examination of Fig. 6a indicates that during the same time interval, the maximum width of the N2O production region occurred, and emission was intensified by urine evaporation enriched with dissolved N2O. The total N2O emission factors obtained during this series of the experiments were 2.5 and 1.4% for the fine and coarse sands, respectively. These values are lower then that obtained in the no-flow conditions (above) and close to the mean N2O emission factor (1.3%) obtained under field conditions (Van Groenigen et al., 2005).
As was mentioned above, evaporation of urine increased the air-filled pore space in both sands, which improved the aeration conditions within the sand matrix and therefore reduced the contribution of denitrification to total N2O production. The measurements of NO3– and NH4+ in the urine samples were used as indications of the N2O production processes. In both sands, when the air-filled pore space was 30 to 40% in the deep horizons (>3.5 cm), NO3– formation was found during the first 2 to 3 d of the experiment in both sand types. This NO3– accumulation was accompanied by a reduction of dissolved NH4+ in urine that indicated the existence of the nitrification process. This nitrification process was limited by the amount of O2 initially present in the sand matrix or dissolved in urine. The intensity of this process was greater in the fine sand. After that, some NO3– still appeared in the lower horizons, but the concentration of NH4+ was constant, which indicated the possibility of downward diffusion of NO3– from the upper sand horizons. This NO3– can be a potential source for denitrification. When the air-filled pore space was <20%, no N2O formation was found, probably because anaerobic conditions favored formation of N2. In contrast, when the air-filled pore space was >70% as a result of sand drying, the concentration of NH4+ in solution was close to zero and no NO3– was detected; the amount of N2O found in the gaseous form in the pore space was probably the result of upward diffusion from the deeper horizons.
Oxidation–Reduction Potential and Nitrous Oxide Production (Subexperiments 1 and 5)
As observed above, sand texture had a strong influence on O2 supply and therefore sand redox status, changing the balance of nitrification and denitrification inputs to total N2O formation. Figure 8
presents the summary of the relationship between sand redox and N2O production in both sand textures. The experimental results indicate two ranges of redox, –50 to 50 and 250 to 400 mV, when the majority of N2O production was observed. Within the range of –50 to 50 mV, an anaerobic environment provided denitrification-dominant conditions. In contrast, good aeration conditions within the range of 250 to 400 mV caused nitrification to be the dominant process. Almost no generation of N2O was found when the redox was below –50 mV, probably due to the stronger reduction of N2O to N2 (Clough et al., 1998). Kewei and Patrick (2004) also reported significant N2O production in different types of rice (Oryza sativa L.) soils by nitrification when redox ranged between 200 and 400 mV. The field study of Kralova (1992) confirmed that in sandy pasture soils, the N2/N2O ratio was 1:1 at redox 0 mV and N2O production was reduced to zero at redox between 50 and 150 mV. Within the range between 50 and 250 mV, the lack of gaseous N emission by N2O formation can be compensated by formation of NO gas, which is predominantly formed at these redox values (Fuerhacker et al., 2001).
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 8. Relationship between sand oxidation–reduction potential and N2O emission factor in static headspace and flow-through experiments.
|
|
The comparison between N2O emission factors in two redox ranges indicates that during the flow-through experiments, most N2O was formed in nitrification-dominated regions (250–400 mV), with a maximum of 0.8% at 310 mV. During the same experiment, the N2O emission factors in denitrification-dominated regions (–50 to 50 mV) were much lower, ranging between 0.05 and 0.13%. In contrast, in constant headspace experiments where the evaporating urine losses were negligible, N2O formation was intensive in both nitrification- and denitrification-dominated regions. This corresponds with results presented in Fig. 5a, where the ratio between nitrification and denitrification inputs was 1:1 under the same experimental conditions (fine sand, 40% air-filled pore space). The difference in the intensity of N2O production in denitrification-dominated regions between two experimental types was observed mostly because the sand drying process increased air-filled pore space (Fig. 6), thus minimizing the formation of anaerobic conditions.
|
CONCLUSIONS
|
|---|
The results from these controlled and intentionally simplified systems showed that texture was the most important soil factor controlling NH3 volatilization and N2O emission factors in urine-treated sands. The two sand texture types had different air-filled porosity distributions with depth, the critical parameter for O2 supply that in turn controls redox status. The effect of air-filled pore space was primarily one of influencing the interplay between the nitrification and denitrification processes in urine-treated sand, with subsequent changes in N2O emissions. Generally, the finer the sand texture, the greater the input of denitrification to the total N2O production; however, the air-filled porosity threshold below which denitrification dominated was greater in coarse sand.
Sand texture also controlled urine evaporation (i.e., the water component of urine) rates from urine-treated sands, which played an essential role because urine was the only source of N and most of the N2O formation processes occurred in urine. The majority of NH3 volatilization occurred during the first day of the tests, with the volatilization rate closely related to the urine evaporation rate in both sand types. Finer sand texture caused reductions in the urine evaporation rate and therefore in NH3 volatilization rates at the initial stage of the experiments.
The urine evaporation rate was found to be an important parameter in total N2O emissions as well. First, urine evaporation increased the air-filled porosity, thereby improving the aeration conditions in the sand that contributed to nitrification dominance of N2O production. The effect of increased air-filled pore space could also be enhanced by faster release of accumulated N2O at the deeper layers, because diffusion in water is 104 lower than in air (Stevenson, 1982). Second, evaporation of urine, enriched with dissolved N2O, increased total N2O emission.
The interplay among these factors and processes contributes to the intensity of gaseous NH3 and N2O emissions from urine-treated fields, where wind is an integral part of the environmental conditions. Results from these simplified experiments require confirmation under more complex experimental conditions (more complex soils with additional N sources present) and ultimately under field soil conditions.
|
ACKNOWLEDGMENTS
|
|---|
We thank Dr. S.K. Giri for assistance and useful discussion. This material is based on work supported by the National Research Initiative Air Quality Program of the Cooperative State Research, Education, and Extension Service, U.S. Department of Agriculture, under Agreement no. 123527, and Vaadia-BARD Postdoctoral Fellowship Award no. FI-357-2004 from BARD, the U.S.–Israel Binational Agricultural Research and Development Fund.
|
REFERENCES
|
|---|
- Abbasi, M.K., and W.A. Adams. 1998. Loss of nitrogen in compacted grassland soil by simultaneous nitrification and denitrification. Plant Soil 200:265–277.[CrossRef]
- Abbasi, M.K., Z. Shah, and W.A. Adams. 1997. Concurrent nitrification and denitrification in compacted grassland soil. p. 47–54. In S.C. Jarvis and B.F. Pain (ed.) Gaseous nitrogen emission from grasslands. CAB Int., Wallingford, UK.
- Akiyama, H., I.P. Mc Taggart, B.C. Ball, and A. Scott. 2004. N2O, NO, and NH3 emissions from soil after the application of organic fertilizers, urea and water. Water Air Soil Pollut. 156:113–129.
- American Public Health Association. 1985. Standard methods for examination of water and wastewater. APHA, Washington, DC.
- Anger, M., C. Hoffman, and W. Kühbauch. 2003. Nitrous oxide emissions from artificial urine patches applied to different N-fertilized swards and estimated annual N2O emissions for differently fertilized pastures in an upland location in Germany. Soil Use Manage. 19:104–111.[CrossRef]
- Ball, B.C., J.P. Parker, and A. Scott. 1999. Soil and residue management effects on cropping conditions and nitrous oxide fluxes under controlled traffic in Scotland: 2. Nitrous oxide, soil N status and weather. Soil Tillage Res. 52:191–201.
- Bohn, L.H., B.L. McNeal, and G.A. O'Connor. 1979. Soil chemistry. John Wiley & Sons, New York.
- Boyer, E.W., C.L. Goodale, N.A. Jaworsk, and R.W. Howarth. 2002. Anthropogenic nitrogen sources and relationships to riverine nitrogen export in the northeastern USA. Biogeochemistry 57:137–169.[CrossRef]
- Chao, T.T., and W. Kroontje. 1964. Relationship between ammonia volatilization, ammonia concentration, and water evaporation. Soil Sci. Soc. Am. Proc. 28:393–395.
- Clough, T.J., S.F.D. Ledgard, M.S. Sprosen, and M.J. Kear. 1998. Fate of 15N labeled urine on four soil types. Plant Soil 199:195–203.[CrossRef]
- Clough, T.J., R.R. Sherlock, and F.M. Kelliher. 2003. Can liming mitigate N2O fluxes from a urine-amended soil? Austr. J. Soil. Res 41:439–457.[CrossRef]
- Coussot, P. 2000. Scaling approach of the convective drying of a porous medium. Eur. Phys. J. B 1:557–566.
- Crutzen, P.J. 1981. Atmospheric chemical processes of the oxides of nitrogen, including nitrous oxide. p. 17–44. In C.C. Delwiche (ed.) Denitrification, nitrification and nitrous oxide. John Wiley & Sons, New York.
- Davidson, E.A., M. Keller, H.E. Erickson, L.V. Verchot, and E. Veldkamp. 2000. Testing a conceptual model of soil emissions of nitrous and nitric oxides. BioScience 50:667–680.[CrossRef][ISI]
- Deenen, P.J.A.G., and N. Mindelkoop. 1992. Effects of cattle dung and urine on nitrogen uptake and yield of perennial ryegrass. Neth. J. Agric. Sci. 40:469–482.
- De Klein, C.A.M., L. Barton, R.R. Sherlock, Z. Li, and R.P. Littlejohn. 2003. Estimating a nitrous oxide emission factor for animal urine from some New Zeland pastoral soils. Aust. J. Soil Res. 41:381–399.[CrossRef]
- Duxbury, J.M. 1994. The significance of agricultural sources of greenhouse gases. Fert. Res. 38:151–163.
- Dziejowski, J.E., A. Rimmer, and T.S. Steenhuis. 1997. Preferential movement of oxygen in soils? Soil Sci. Soc. Am. J. 61:1607–1610.[Abstract/Free Full Text]
- Fraser, P.M., K.C. Cameron, and R.R. Sherlock. 1994. Lysimeter study of the fate of nitrogen in animal urine returns to irrigated pasture. Eur. J. Soil Sci. 45:439–447.[CrossRef]
- Fuerhacker, M., H. Bauer, R. Ellinger, U. Sree, H. Schmid, F. Zibuschka, and H. Puxbaum. 2001. Relationship between release of nitric oxide and CO2 and their dependence on oxidation reduction potential in wastewater treatment. Chemosphere 44:1213–1221.[Medline]
- Gabel, R.A., and B. Schultz. 1973. Solubility of nitrous oxide in water, 20–80°C. Anesthesiology 38:75–81.[CrossRef][ISI][Medline]
- Haynes, R.J., and P.H. Williams. 1992. Changes in soil solution composition and pH in urine affected areas of pasture. J. Soil Sci. 43:323–334.[CrossRef]
- Hynes, R.K., and R. Knowles. 1980. Denitrification, nitrogen-fixation and nitrification in continuous-flow laboratory soil columns. Can. J. Soil Sci. 60:335–363.
- Intergovernmental Panel on Climate Change. 2000. Good practice guidance and uncertainty management in national greenhouse gas inventories. OECD/ODCE, Paris, France.
- Janzen, H.H., R.L. Desjardins, J.M.R. Asselin, and B. Grace. 1998. The health of our air — toward sustainable agriculture in Canada. Publ. 1981/E. Research Branch, Agric. and Agri-Food Canada, Ottawa, ON.
- Kewei, Y., and W.H. Patrick, Jr. 2004. Redox window with minimum global warming potential contribution from rice soils. Soil Sci. Soc. Am. J. 68:2086–2091.[Abstract/Free Full Text]
- Kissel, D.E., H.L. Brewer, and G.F. Arkin. 1977. Design and test of a field sampler for ammonia volatilization. Soil Sci. Soc. Am. J. 41:1133–1138.[Abstract/Free Full Text]
- Kralova, M. 1992. Oxidation–reduction potentials in soil. Rostl. Vyroba 38(1): 39–46.
- Lockyer, D.R., and D.C. Whitehead. 1990. Volatilization of ammonia from cattle urine applied to grassland. Soil Biol. Biochem. 22:1137–1142.[CrossRef]
- Müller, C., R.R. Sherlock, and P.H. Williams. 1997. Mechanistic model for nitrous oxide emission via nitrification and denitrification. Biol. Fertil. Soils 24:231–238.[CrossRef]
- Petersen, S.O., P. Roslev, and R. Bol. 2004. Dynamics of a pasture soil microbial community after deposition of cattle urine amended with [13C]urea. Appl. Environ. Microb. 6363–6369.
- Pilot, L., and W.H. Patrick. 1972. Nitrate reduction in soils: Effect of soil-moisture tension. Soil Sci. 114(4):312–316.
- Sherlock, R.R. and K.M. Goh. 1984. Dynamics of ammonia volatilization from simulated urine patches and aqueous urea applied to pasture.1. Field experiments. Fertil. Res. 5:181–195.
- Stevenson, F.J. (ed.) 1987. Nitrogen in agricultural soils. Agron. Monogr. 22. ASA, CSSA, and SSSA, Madison, WI.
- Van Groenigen, J.W., P.J. Kuikman, W.J.M. deGroot, and G.L. Velthof. 2005. Nitrous oxide emission from urine-treated soil as influenced by urine composition and soil physical conditions. Soil Biol. Biochem. 37:463–473.[CrossRef]
- Velthof, G.L., P.J. Kuikman, and O. Oenema. 2003. Nitrous oxide emission from animal manures applied to soil under controlled conditions. Biol. Fertil. Soils 37:221–230.
- Vertregt, N., and B. Rutgers. 1987. Ammonia volatilization from urine patches in grassland. p. 85–91. In V.C. Nielsen et al. (ed.) Volatile emissions from lifestock farming and sewage operations. Elsevier, London.
- Whitehead, D.C., and N. Raistrick. 1993. The volatilization of ammonia from cattle urine applied to soils as influenced by soil properties. Plant Soil 148:43–51.[CrossRef]
This article has been cited by other articles:
|
|
|
|
 
A. J. Franzluebbers and J. A. Stuedemann
Early Response of Soil Organic Fractions to Tillage and Integrated Crop-Livestock Production
Soil Sci. Soc. Am. J.,
May 1, 2008;
72(3):
613 - 625.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
