Environ. Sci. Technol.

2001, 35, 2347-2351

5.5 Tg N2O-N/year, about 30-35% of the total emissions

Nitrous Oxide Emissions from (5); however, this estimation does not include N2O production
Aerated Composting of Organic from waste management systems. Although the treatment
of organic wastes has been frequently reported to generate
Waste N2O (6, 7), the total emission from these processes remains
poorly quantified.
Nitrous oxide is a byproduct of nitrification and deni-
Y A O W U H E , * ,† Y U H E I I N A M O R I , †
trification, which are the main mechanisms for nitrogen
MOTOYUKI MIZUOCHI,† HAINAN KONG,†
removal in waste treatment processes; therefore, it is not
NORIO IWAMI,† AND TIEHENG SUN‡
surprising to detect high N2O emission from a variety of waste
National Institute for Environmental Studies, Onogawa 16-2,
treatment facilities. Zheng et al. (8) reported that in anoxic
Tsukuba, Ibaraki 3050053, Japan, and Institute of Applied
treatment of wastewater 0-19% of nitrogen was removed in
Ecology, Chinese Academy of Sciences, 72 Wenhua Road,
form of N2O. In activated sludge systems, Mizuochi et al. (9)
Shenyang, 110015, P.R. China
and Hanaki (10) observed a production of 3 and 24.1 mg of
N2O-N, respectively, from the treatment of 1 m3 wastewater.
Apparent N2O emissions were also detected from treatment
The composting of high organic content wastes has been of solid waste. Tsujimoto et al. (7) reported emissions of 7.8
and 40.2 g‚day-1 from an active and a closed landfill site,
shown to produce nitrous oxide (N2O). This study was
respectively. Availability of oxygen is a determining factor in
initiated to investigate the mechanisms of N2O emissions N2O production. Nitrification under strict aerobic condition
from aerated composting and to determine the optimal and denitrification under strict anaerobic condition result
operational conditions that minimize N2O emissions. The in negligible N2O production. N2O is mainly formed at
results of our experiment in laboratory-scale composters moderate O2 concentrations (6, 7).
showed that more than 95% of N2O was produced during the Up to now, most solid organic waste has been landfilled
later period of composting when readily available carbon or incinerated; however, more and more concern has arisen
sources had been depleted. Significant increases in in recent years because of leachate problems and air pollution
N2O emission after nitrite (NO2-) addition, and good NO2- - associated with these systems. In particular, significant
N2O correlation, indicates that N2O was transformed emissions of greenhouse gases were detected from these
processes (7,11-13). In view of this, aerobic composting has
from NO2-. Extremely high N2O generation was observed
been suggested as a more acceptable alternative to landfill
after NO2- addition in the presence and absence of composted disposal since the predominant aerobic environment in-
cattle manure. This suggests an identical mechanism for volved can mitigate greenhouse gas emission by reducing
N2O production in both treatments. However, the addition of methane generation (14). Nevertheless, the potential of N2O
composted cattle manure resulted in an earlier initiation production from aerobic composting is unknown. In a
of the main N2O generation period. Intermittent feeding of practical scale sludge composting study, Czepiel et al. (6)
fresh food waste postponed the main N2O generation found that N2O flux was a function of the compost age, pile
period, and reduced the mass-based N2O emissions by depth, temperature, and water filled pore space. Aeration
20%. was reported to increase N2O production.
The present paper describes the results of a comprehen-
sive study conducted in laboratory-scale reactors to quantify
N2O generation from the aerated composting of food waste.
Introduction The effects of several factors were investigated to identify
Interest in the sources of atmospheric N2O has recently been the main parameters controlling N2O emission. The resulting
stimulated by the recognition that this trace gas exerts emission rate data were then used to optimize the operational
significant influence on the chemistry of the stratosphere conditions for aerated composting to minimize N2O emission.
and on the earth’s thermal balance. Reaction of N2O with
singlet atomic oxygen represents the dominant source of Experimental Section
nitric oxide in the stratosphere which plays a catalytic role Waste and Amendments (Table 1). To keep the uniformity
in ozone destruction (1, 2). The high efficiency of N2O in of food waste composition, synthetic food waste was used
absorbing infrared radiation (IR) also makes it an important instead of practical pre- and postcustomer garbage. It was
greenhouse gas (3). Though atmospheric N2O accounts for fortified according to a Standard Composition for Food Waste
only 6% of the greenhouse effect, the rapid increase of its used by Ministry of Construction, Japan, which contains
mixing ratio in the atmosphere (currently at a rate of 0.25- chicken bone (8.16%, by weight), fish (10.20%), apple
0.31%/year) (4), has drawn attention to the need for research (10.20%), banana peel (10.20%), grape peel (10.20%), cabbage
on all sources and sinks of this gas. (18.38%), carrot (18.38%), rice (10.20%), and tea residue
In the global inventory of N2O, natural sources such as (4.08%). Chicken bone, fish, and rice were boiled. All materials
undisturbed soil, photolytic processes in the ocean, and were reduced in size by a disposer (Emerson Electric Co.).
atmospheric formation are estimated to account for about The food mixture was stored in the dark in a freezer prior to
60% of total N2O emissions (4, 5). Significant anthropogenic use. A commercial sawdust-based product of National Ltd.
sources including agricultural and industrial activities release (Japan), Biochip, was used as the bulking agent. In some
reactors composted cattle manure (CCM) purchased from
* To whom correspondence should be addressed. Current ad- Joriku Ranch (Ibaraki, Japan) was added, in simulation of
dress: School of Biological Sciences, The Flinders University of South
the compost addition process in some small-scale composters
Australia, GPO Box 2100, Adelaide 5001, Australia. E-mail: li.li@
flinders.edu.au. in Japan, to investigate its effect on N2O generation.
† National Institute for Environmental Studies. Composting Unit. Cylinder-shaped composters (18 L)
‡ Institute of Applied Ecology. were used in the study. As shown in Figure 1, waste and
10.1021/es0011616 CCC: $20.00  2001 American Chemical Society VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 2347
Published on Web 04/19/2001
TABLE 1. Amount of Food Waste and Amendments (fresh weight) in Different Experiments
food composted
waste cattle manure sawdust aeration rate
experiments (kg) (kg) (kg) replicate (L‚min-1‚kg-1a)
reproducibility test 2.0 0.0 1.5 2 0.70
2.0 1.0 1.5 2 0.70
single-charged composters 2.0 0.0 1.5 2 0.70
2.0 1.0 1.5 2 0.70
intermittent-charged composters 2.0 + 4.0 + 6.0b 1.0 1.5 2 0.70c
effect of aeration rates 2.0 1.0 1.5 2 0.14
2.0 1.0 1.5 2 0.56
effect of NO2- addition 2.0 0.0 1.5 2 0.70
2.0 1.0 1.5 2 0.70
a Liters per minute per kilogram of initial dry weight of food waste. b On days 5 and 14, 4.0 and 6.0 kg of fresh food waste were added into

the composters. c After readdition of food waste, aeration rate was increased correspondingly.

eye at an accuracy of (0.1%, (0.1%, and (0.1 ppm for O2,

CO2, and NH3, respectively.
Waste Mixture Sampling and Analysis. Immediately
after air sampling, the waste mixture was sampled from at
least five different points in each composter and mixed
thoroughly. Subsamples were then removed to determine
pH and gravimetric water content. For analysis of NH4+-N,
NO3--N, NO2--N, and TN (total N) concentration, 2.00 g of
waste mixture was extracted by 20 mL of 2 M KCl solution
for 30 min. After being centrifuged at 4000 rev‚min-1 for 10
min, the supernatant was filtered through glass fiber filter
FIGURE 1. Schematic of laboratory composting apparatus.
(0.45 µm, Whatman), analyzed by an automated colorimetric
method on a TRAACS-800 instrument.
amendments were thoroughly mixed in the composter by To determine concentration of dissolved organic carbon
slow stirring (0.8 rev‚min-1) of the motor. Four mixing arms (DOC) in waste, another 2.00 g subsample was extracted by
were fixed to the central shaft and at the ends of two adjacent 20 mL of water (Milli-Q SP TOC, Millipore) for 30 min,
arms a connection plate was installed with only a narrow centrifuged at 4000 rev‚min-1 for 10 min, and filtered by
space left in the middle for the thermocouple. Such a design glass fiber filter (0.45 µm, Whatman). The supernatant was
enabled the mixing paddles to move across the composter stored at 4 °C and analyzed within 2 days on a TOC analyzer
to guarantee a uniform mixing of waste and amendments. (Shimadzu TOC-5000, Japan).
Aeration was achieved by continuously compressing air into
the composter through a false floor. To mitigate effects of Results
ambient temperature fluctuation on composting, the com- Reproducibility of Composter Performance. The reproduc-
poster was kept in a thermostatic chamber throughout the ibility of the performance of the composters was assessed by
testing period. Moreover, two electrical heaters were used, the analysis of oxygen and temperature profiles from two
one to heat the aeration flow, the other to warm a 0.5 L‚min-1 preliminary experiments, one with CCM the other without.
airflow passed through the space between composter and In both experiments water content in the waste mixture was
the wall of chamber to establish a thermostatic environment. kept at around 65%. In the presence of CCM, both duplicates
A trough was installed beneath the false floor to collect showed a 44-45 °C temperature peak at day 3 and a stabilizing
leachate which was then recycled into the composter. A temperature profile (30.6-32.3 °C) after day 10, whereas in
thermocouple buried into the waste mixture automatically the absence of CCM, temperature increased to 45-47 °C at
measured temperature. day 5 and stabilized after day 11. The average difference
Gas Sampling and Analysis. Air samples were collected between the replicate temperature profiles in two experi-
at 12-h intervals within the first week. After day 8, sampling ments was less than 2 °C throughout the composting period.
frequency was gradually reduced to a rate of once in 3 days. Oxygen percentage in the offgases first decreased to 13 and
A battery-operated pump (Sibata MP-2N, Japan) drew the 15% in treatments with and without CCM, respectively, and
headspace samples into aluminum bags. Samples were then returned to 19.1-19.5% after day 11. The average
immediately transported to the lab, and allowed to equilibrate discrepancy between oxygen levels in offgases from com-
to room temperature and analyzed by a Shimadzu GC-14A posters with identical initial conditions was less than 0.5%.
gas chromatograph (Shimadzu Co., Japan) within 4 h. Analysis of the temperature and oxygen profiles showed that
Samples and standards were cleaned-up across two glass- the composting system was able to provide reproducible test
made columns packed with magnesium perchlorate (Nacalai conditions for further experiments.
Tesque, Inc., Japan) and ASCARITE II (Thomas Scientific), Nitrous Oxide Emission. Significant N2O production was
respectively, to remove moisture and CO2 as they were detected in all experimental runs. Figure 2 presents two
injected into the gas chromatograph injection loop. Analysis typical N2O emission profiles versus composting time
of N2O ((0.02 µL‚L-1) was accomplished using an electron obtained in treatments with and without CCM, respectively.
capture detector after constituent separation by a Poropak A N2O emission peak was observed immediately after the
Q column (2m × 3 mm i.d. stainless steel). The carrier gas beginning of composting, and the peak value in treatments
was a 95% Ar-5% CH4 mixture at a flow rate of 40 mL‚min-1. with and without CCM was 3.32 and 2.73 µL‚L-1, respectively.
Temperatures of detector and oven were 340 and 80 °C, Charging of CCM increased the average peak value of N2O
respectively. O2, CO2, and NH3 concentrations in the exhaust emission by 0.59 µL‚L-1. After 2 days, N2O concentration in
air were measured by gas detector tubes (Gastec Co., Japan) the exhaust gas rapidly decreased to 0.53 µL‚L-1, slightly
packed with materials which change colors after reaction higher than the atmospheric background level of 0.45 µL‚L-1
with O2, CO2, and NH3. The measurement was conducted by (ranging from 0.41 to 0.47 µL‚L-1 during the composting

2348 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001
 FIGURE 4. Ammonia volatilization from composters aerated at
FIGURE 2. Profiles of N2O concentration in the exhaust gas versus
different rates.
composting time.

FIGURE 3. Profiles of N2O concentration in treatments with CCM

at various aeration rates. FIGURE 5. Changes of DOC in waste mixture and N2O concentration
period). From days 2 to 12, N2O production was low with an in the exhaust gas with composting time, the inset shows a plot
of N2O concentration in offgas against available carbon content (n
average emission rate of 3.38 µL‚h-1 from 1 kg food waste
) 111).
(dry weight). After day 12, N2O production from CCM-
amended treatments increased steadily and entered the
second (main) production period. A peak of 202.67 µL‚L-1 CH4 is 10 times lower than that of N2O (9). Therefore, among
was detected on day 36, then N2O generation decreased slowly the three aeration rates employed in the study, 0.56
to around 0.6 µL‚L-1 after day 100. In treatments without L‚min-1‚kg-1 appeared to be the most suitable aeration rate
CCM, no significant change occurred in N2O production until if both N2O and CH4 emissions were considered.
day 55. Though a peak was detected at day 110, the average Effect of Intermittent Feeding of Fresh Waste. To assess
peak value was much lower compared with that in CCM- the effect of carbon availability on N2O generation, DOC in
applied composters. Nitrous oxide generated during the main the waste mixture was determined in parallel with gas
production period accounted for more than 95% of total analysis. In Figure 5, DOC in waste mixture increased sharply
emission throughout the experimental processes in trials with to 33.28 mg‚g-1 on day 3, showing a higher DOC production
and without CCM. than removal, then decreased to 3.45 mg‚g-1 on day 10. During
Effect of Aeration Rate. The effect of aeration rate on this period, N2O emission was relatively low, ranging from
N2O production was studied by aerating composters with 0.45 to 3.32 µL‚L-1, most of N2O was produced after day 12
identical water contents and waste/amendment ratios at (Figures 2 and 5). A plot of N2O concentration against DOC
different aeration rate. As the aeration rate was reduced from content in the waste mixture (inset of Figure 5) showed no
0.70 to 0.56 and 0.14 L‚min-1‚kg-1, N2O emission decreased statistically significant correlation. However, high N2O pro-
significantly (Figure 3). The second N2O emission peak, which duction occurred while DOC content was low, indicating
was observed at 0.70 L‚min-1‚kg-1 (Figure 2, treatments with that N2O was mostly released after the depletion of available
CCM), was reduced by more than 90% at 0.56 L‚min-1‚kg-1 carbon source. The result implies that intermittent feeding
and did not appear at 0.14 L‚min-1‚kg-1. However, low of fresh waste in the presence of enough bulking agent and
aeration prolonged ammonification period and resulted in amendments, as it is widely used in commercially available
high ammonia volatilization (Figure 4), which would reduce composters which require sawdust addition once a month,
the nutrient value of the composted product and cause more may influence N2O production.
difficulties with odor control. Moreover, as the aeration rate To investigate the effect of intermittent feeding of waste
was reduced methane production increased. The mass-based on N2O generation, additional 4 and 6 kg fresh food waste
methane production was 1.72, 31.23, and 1328.94 mL‚kg-1 were added at days 5 and 14, respectively, into two CCM-
(dry weight), respectively, at 0.70, 0.56, and 0.14 L‚min-1‚kg-1. amended treatments with an initial loading rate of 2 kg. Fresh
High methane production at 0.14 L‚min-1‚kg-1 offset its waste was charged immediately after the pH in the waste
benefit in N2O emission control though IR absorbance of mixture increased to 9.0. Similar DOC accumulation-

VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 2349
 FIGURE 7. Correlation between N2O concentration in offgas and
NO2- content in waste mixture.

FIGURE 6. Effect of intermittent feeding of food waste on N2O

production

consumption patterns were observed after addition of

different amount of waste on days 0, 5, and 14 (DOC in waste
mixture reached a maximum concentration of 31.88, 61.94,
and 84.32 mg‚g-1 3 days after addition and returned to 6.18,
6.54, and 6.39 mg‚g-1 5 days after addition, respectively).
The result suggested that sawdust used in the experiment
was enough for removing organic carbon. A small N2O
emission peak was detected immediately after each addition
(Figure 6). Consequently, three emission peaks were obtained
within 15 days in intermittent-charged composters, while in
single-charged ones only one peak appeared during the same
period. Nitrous oxide in the exhaust gas from single-charged
composters began to increase after day 12 and reached a
FIGURE 8. Effect of nitrite addition on N2O production in composters
maximum concentration of 202.67 µL‚L-1 on day 36. In
in the presence (solid circles) and in the absence of CCM (open
intermittent-charged composters, however, a significant triangles).
increase in N2O production occurred on day 47 and reached
a maximum concentration of 953.23 µL‚L-1 on day 70. The
occurrence of the main production period was postponed in Discussion
intermittent-feeding treatments. The mass-based N2O emis- Our results demonstrate the potential for significant N2O
sion was 1.19 and 0.95L‚kg-1 (dry waste) for single-charged production from aerated composting of food waste. By
and intermittent-charged treatments, respectively, indicating microbial degradation, organic nitrogen was transformed
a 20% reduction in N2O production in intermittent-feeding into ammonia, leading to accumulation of ammonia in the
treatments. waste mixture and high ammonia volatilization (Figure 4).
Nitrous Oxide-Nitrite Correlation. Ammonia, nitrite, Significant N2O emission started concurrently with ammonia
nitrate, and TN concentration in waste mixture were oxidation, which was evidenced by the sequential occurrence
monitored simultaneously with N2O, and the data was then of ammonia, nitrite, and nitrate accumulation. More than
utilized to formulate their relationships with N2O concentra- 95% of N2O was released during the later period of composting
tion in the exhaust gas. Significant increases in ammonia, while available carbon had been used up (Figure 5). The
nitrite, and nitrate concentrations in waste mixture occurred time at which the N2O emission peak appeared depended
on days 8-10, 28-30, and 60-90, respectively, depending on the waste application pattern and amendment. Intermit-
on the aeration rate and waste application pattern. The result tent feeding of fresh waste postponed its occurrence, whereas
suggested that nitrification was the predominant nitrogen amendment of CCM resulted in a peak earlier in the
transformation process in our system. A good linear cor- composting cycle. The result implies that large amount of
relation between nitrite and N2O concentration was found N2O can be released from old compost piles in which
in all composting trials, independent of feeding patterns of ammonia oxidation has started.
waste and amendments (Figure 7). For ammonia, nitrate, Previous studies on aerobic wastewater treatment pro-
and TN, however, no statistically significant correlation was cesses such as activated sludge and systems with high
obtained. To provide further evidence for the dependence nitrification activity showed that high NO2- level was
of N2O production on nitrite in the aerated composting conducive to N2O production (8, 15-18). Similarly, linear
system, NaNO2 solution was added into composters to NO2--N2O correlation was observed in our study, implying
increase the theoretical nitrite concentration (calculated on the existence of a nitrogen transformation pathway from
the base of dry waste mixture) in waste mixture by 0.5 mg‚g-1. nitrite to N2O in our system. Significant N2O emission after
Six hours after addition of nitrite, N2O concentrations as high nitrite addition further certified that nitrite was one of main
as 8838.83 and 1978.96 µL‚L-1 were detected in the offgases parent chemicals of N2O in the aerated composting process.
of the composters with and without CCM, respectively (Figure The dependence of N2O production on nitrite was found
8) and then sharply decreased to normal level (0.53-0.60 both in the presence and absence of CCM, suggesting an
µL‚L-1) within 3 days. identical mechanism for N2O production in both treatments.

2350 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001
 In contrast to the results obtained previously in manure Acknowledgments
land application (19-21) and composting systems (22), no The research was financially supported by Environmental
significant correlation was found between N2O production Agency, Japan. We are grateful to Dr. Toshihiro Sankai and
and ammonia, nitrate and total nitrogen. Aeration enabled Ms. Keiko Kuto for their generous provision of equipment
the system to be predominantly an oxic environment, and for food waste preparation, and Ms. Hideko Arai and Ms.
reduction of nitrate was poor though anoxic/anaerobic Midori Toyama for their help on N, P analysis. Also, we wish
microsites might exist in large waste particles. As a result of to express our appreciation for the assistance of Dr. Nicholas
this, NO3- concentration as high as 3.53 mg‚g-1 detected in McClure and Mr. Brett Roman in manuscript revisions.
CCM-amended treatments at the end of composting did not
cause apparent change in N2O generation. Poor NH4+-N2O Literature Cited
correlation was possibly a consequence of the extensive
(1) Crutzen, P. J. J. Geophys. Res. 1971, 76, 7311-7327.
period needed for ammonia oxidation. (2) Crutzen, P. J. Q. J. Royal Meteorol. Soc. 1970, 96, 320-325.
High N2O production occurred after the readily available (3) Wang, W. C.; Yung, Y. L.; Lacis, A. A.; Mo, T.; Hansen, J. E. Science
carbon source was depleted. In intermittent-feeding com- 1976, 194, 685-690.
posters, occurrence of the main N2O-production peaks was (4) Khalil, M. A. K.; Rasmussen, R. A. J. Geophys. Res. 1992, 97,
postponed from days 36 to 70. It’s unlikely the increased 14651-14660.
(5) De Soete, G. in Proceedings of the 7th International Workshop
waste amount, which was expected to prolong DOC removal
on Nitrous Oxide Emissions; Cologne, Germany, April 21-23,
process, postponed N2O generation. With enough bulking 1997; pp IX-XIII.
agent, applying large amount of food waste did not prolong (6) Czepiel, P.; Douglas, E.; Harriss, R.; Crill, P. Environ, Sci, Technol.
DOC removal process, as evidenced by the similar DOC 1996, 30 (8), 2519-2525.
accumulation-consumption patterns observed after read- (7) Tsujimoto, Y.; Masuda, J.; Fukuya, J.; Ito, H. J. Air Waste Manage.
dition of different amount of waste in our study. Multiple Assoc. 1994, 44, 1313-1314.
(8) Zheng, H.; Hanaki, K.; Matsuo, T. Wat. Sci. Technol. 1994, 30
feedings of fresh waste supported a continuous DOC (6), 134-141.
production-consumption process and inhibited ammonia (9) Mizuochi, M.; Inamori, Y.; Sato, K. Japanese J. Water Treat. Biol.
oxidation, and hence N2O production was postponed. 1993 (Suppl.), 86-87.
In single-charged composters, a two-peak N2O emission (10) Hanaki, K.; Wantawin, C.; Gujer, W. Water Res. 1990, 24, 289-
curve was obtained after 5 months. The first peak observed 296.
(11) Boltze, U.; de Freitas, M. H. Waste Manage. Res. 1997, 15, 463-
immediately after feeding of fresh waste came from the
476.
release of N2O stored in food waste. Within 6 h, 200 g of food (12) Kjeldsen, P.; Fischer, E. V. Waste Manage. Res. 1995, 13, 467-
waste stored in refrigerator produced 0.75 µL of N2O according 484.
to our investigation. N2O was then sealed in the clump after (13) Liu, A. H.; Kunz, C. O. Environ. Sci. Technol. 1981, 15 (4), 436-
being frozen and released after thawing. 440.
The first N2O emission peak in treatments with CCM was (14) Toms, P.; Leskiw, J.; Hettiaratchi, P. in Presentation at the 88th
Annual Meeting & Exhibition; San Antonio, TX, June 18-23,
higher than that without CCM, as a result of N2O generation 1995; pp 18-23.
from the mature compost which was a known N2O generator- (15) Schulthess, R. V.; Kuhni, M.; Gujer, W. Water Res. 1995, 29,
(22-24). There is no direct evidence that CCM addition 215-226.
imported new N2O generation mechanisms based on the (16) Schulthess, R. V.; Wild, D.; Gujer, W. Water Sci. Technol. 1994,
presently available data, but this study clearly showed that 30, 123-132.
(17) Wild, D.; Schulthess, R. V.; Gujer, W. Water Sci. Technol. 1994,
addition of CCM resulted in an earlier occurrence of N2O.
30, 113-122.
In all experiments, a high N2O emission peak was found (18) Wild, D.; Schulthess, R. V.; Gujer, W. Water Sci. Technol. 1995,
at the end of the composting process (Figures 2 and 3). 31, 45-54.
Considering the good NO2--N2O correlation during this (19) Cates, R. L., Jr.; Keeney, D. R. J. Environ. Qual. 1987, 16 (4),
period and the immediate increase in N2O production after 443-447.
NO2- addition, N2O might be produced from NO2-. Usually (20) Boustany, R. G.; Crozier, C. R.; Rybczyk, J. M.; Twilley, R. R.
Wetlands Ecol. Manage. 1997, 4, 273-283.
this transformation step occurs under anoxic or anaerobic (21) Burford, J. R.; Bremner, J. M. Soil Biol. Biochem. 1975, 7, 389-
conditions (25, 26); however, aerobic denitrification has been 394.
frequently documented in recent years (27, 28). With the (22) Mahimairaja, S.; Bolan, N. S.; Hedley, M. J. Soil Biol. Biochem.
high aeration rates employed in the study, our composting 1995, 27 (9), 1223-1225.
system was kept predominantly aerobic throughout the (23) Haus, R.; Käding, H.; Leipnitz, W.; Bautzer, W. SPIE 1995, 2506,
748-755.
experiment; however, anoxic or anaerobic microsites might
(24) Kuroda, K.; Osada, T.; Yonaga, M.; Kanematu, A.; Nitta, T.; Mouri,
still exist inside the waste particles similarly to the coexistence S.; Kojima, T. Bioresour. Technol. 1996, 56, 265-271.
of aerobic and anaerobic microenvironment in aerobic soils (25) Flessa, H.; Beese, F. Soil Sci. Soc. Am. J. 1995, 59 (4), 1044-1051.
(29, 30), denitrification conducted by denitrifiers in these (26) Bremner, J. M. Nutr. Cycling Agroecosyst. 1997, 49 (1-3), 7-16.
microsites might contribute to NO2- reduction. If this is the (27) Robertson, L. A.; Kuenen, J. G. in Microbial Control of Pollution;
case, N2O production should be proportional to the available Fry, J. C., Gadd, G. M., Herbert, R. A., Jones, C. W., and Watson-
Craik, I. A.; Cambridge University Press: Cambridge, 1992.
carbon as shown previously (21, 31) because denitrifiers are (28) Zhao, H. W.; Mavinic, D. S.; Oldham, W. K.; Koch, F. A. Water
mostly heterotrophs. However, most of the N2O was produced Res. 1999, 33 (4), 961-970.
after the depletion of available organic carbon in our study, (29) Goodroad, L. L.; Keeney, D. R. Soil Biol. Biochem. 1984, 16 (1),
a result inconsistent with the active metabolism of denitrifiers. 39-43.
Considering the sequential appearance of NH4+, NO2-, and (30) Parkin, T. B. Soil Sci. Soc. Am. J. 1987, 51, 1194-1199.
NO3- peaks, which proved the existence of nitrifiers, nitrifier (31) Cabrera, M.; Chiang, S.; Merka, W.; Pancorbo, O.; Thompson,
S. Plant Soil 1994, 163, 189-196.
denitrification appeared to be the predominant mechanism
for N2O production in our system. ES0011616

VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 2351