Nitrous oxide production by nitrification and

Soil Biology & Biochemistry 36 (2004) 687–699 www.elsevier.com/locate/soilbio

Nitrous oxide production by nitrification and denitrification in soil aggregates as affected by O2 concentration K. Khalila, B. Marya,*, P. Renaultb b

a INRA, Unite´ d’Agronomie, rue Fernand Christ, 02007 Laon F-02007 Cedex, France INRA, Unite´ Climat, Sol et Environnement, Site Agroparc, 84914 Avignon Cedex 9, France

Received 14 August 2003; received in revised form 10 December 2003; accepted 13 January 2004

Abstract Nitrous oxide emitted by soils can be produced either by denitrification in anoxic conditions or by nitrification in presence of O2. The relative importance of the two processes, particularly under varied partial pressures of O2, is not always known. This paper focuses on the influence of O2 concentration on N2O production by nitrification and denitrification in an arable Orthic Luvisol. Soil aggregates (2– 3 mm size), water unsaturated, received 116 mg N kg21 as ammonium sulphate labelled with 15N and were incubated during 14 days at different O2 partial pressures: 0, 0.35, 0.76, 1.5, 4.3 and 20.4 kPa. A 15N tracing technique was used to quantify nitrification and denitrification rates. 15 N2O and 15N2 were measured. Oxygen pressure appeared to strongly influence both nitrification and denitrification rates and also N2O emissions. Nitrification rates were reduced by a factor of 6 – 9 when O2 decreased from 20.4 to 0.35 kPa. They were highly correlated with O2 consumption rates. Denitrification mainly occurred in complete anoxic conditions. The proportion of N2O emitted by denitrification was estimated by two independent methods: one based on 15N tracing using isotope composition of NH4, NO3 and N2O, the other based on the measurement of the 15N2O:15N2 ratio. The two methods gave close results. The highest N2O emissions were obtained under complete anoxic conditions and were due to denitrification. However, N2O emissions almost as important were obtained at day 14 with 1.5 kPa O2 pressure, and they were due to nitrification. Nitrification was the main source of N2O at O2 concentrations greater than 0.35 kPa. The amounts of N2O-N emitted by nitrification were linearly related to the amounts of N nitrified, but the slope of the regression was highly dependent on O2 concentration: it varied from 0.16 to 1.48% when O2 concentration was reduced from 20.4 to 0.76 kPa. Emissions of N2O by nitrification may then be quite significant if nitrification occurs at a reduced O2 concentration. q 2004 Elsevier Ltd. All rights reserved. Keywords: Denitrification; Nitrification; Nitrous oxide; Oxygen; 15N

1. Introduction Nitrous oxide is involved in the global greenhouse effect (Smith, 1990; IPCC, 1996). Its emission from soils results mainly from biological denitrification and nitrification (Groffman, 1991; Heńault and Germon, 1995; Conrad, 1996). A better knowledge of the contribution of each process should help to predict and mitigate N2O emissions by cultivated soils. Nitrification and denitrification in soil often occur in close vicinity so that a substantial part of the NO2 3 formed by nitrification in an oxic zone can diffuse towards an anaerobic zone where it can be denitrified into N2 * Corresponding author. Tel.: þ 33-323236483; fax: þ 33-323793615. E-mail address: [email protected] (B. Mary). 0038-0717/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.soilbio.2004.01.004

(Nielsen et al., 1996). The simultaneous occurrence of nitrification and denitrification in soil associated with N fertilisation has been suggested recently (Zanner and Bloom, 1995; Nielsen et al., 1996; Abbasi and Adams, 1998). Bremner and Blackmer (1981) observed nitrous oxide emissions in ‘well-aerated soils’; the emissions were not correlated with NO2 3 but were significantly correlated with NHþ concentrations. Parton et al. (1996) found that 4 N2O fluxes through nitrification could be proportional to soil N turnover and that only high levels of soil NH4 (. 3 mg N kg21soil) affected N2O emissions. The sources of N2O can be identified by using selective inhibitors, sterilisation or by adding substrates. The disadvantage of nitrification inhibitors is that the prevention of NO2 3 production may affect the rate of denitrification. Sterilisation can be used to separate abiotic from biotic

688

K. Khalil et al. / Soil Biology & Biochemistry 36 (2004) 687–699

2 sources. Adding NHþ 4 or NO3 as substrates cannot provide definitive identification of the sources of N2O unless the substrates are 15N labelled. In this case, the relative importance of nitrification and denitrification can be assessed by measuring and comparing the isotope enrich2 ments of the N2O, NHþ 4 and NO3 pools (Stevens et al., 1997). The availability of O2 in soil is one of the main factors regulating nitrification, denitrification and the release of N2O. Oxygen pressure is the main factor controlling denitrification through the activity and synthesis of denitrifying enzymes in soil (Tiedje, 1988). Denitrification has been considered for long as a strictly anaerobic process, but it is now well established that it can also occur in apparently aerobic environments because many soil denitrifying micro-organisms are able to produce N2O over a wide range of oxygen pressures. Conversely, nitrification is a strictly aerobic process since the NHþ 4 oxidation enzyme of nitrifying organisms requires O2 for activation (Wood, 1986). The effect of O2 on nitrification and associated N2O emissions has been studied more in microbial cultures than in soils. Bollmann and Conrad (1998) showed that N2O emitted by nitrification in soils was important at partial pressures higher than 0.1 –0.5 kPa O2. Goreau et al. (1980), using sediment slurries, found that production of N2O by nitrification reached a maximum at 0.2 kPa O2 pressure. The mechanism of N2O production by nitrification is not completely elucidated (Blackmer et al., 1980; Stevens and Laughlin, 1998). Three main hypotheses have been proposed:

1. A constant proportion of NHþ 4 can be converted to N2O during nitrification, resulting from various reactions of intermediates. Conrad (1996) proposed that N2O production could be the consequence of intermediate (HNO) formation during oxidation of NH2OH to NO2 2 . HNO oxidation could also lead to the formation of another unknown compound, which would then be oxidised to NO2 2 . This hypothesis was retained in several models (Linn and Doran, 1984; Davidson, 1993; Parton et al., 1996). 2. The use of NO2 2 as an alternative electron acceptor during NHþ oxidation for growth of nitrifiers when O2 4 pressure is not high enough to supply them with all the required O2 (Ritchie and Nicholas, 1972; Goreau et al., 1980; Poth and Focht, 1985). The effect of O2 pressure has been shown experimentally either directly by varying O2 partial pressure (Bollmann and Conrad, 1998) or indirectly by varying soil moisture (Zanner and Bloom, 1995) and it is accounted for in some models (Grant, 1995); 2 3. The partial oxidation of NHþ 4 into NO2 in aerobic 2 conditions, followed by NO2 diffusion to anaerobic (or microaerobic) regions and its subsequent reduction into N2O by denitrification.

The aim of this work was to quantify nitrification and denitrification and their specific contribution to N2O emissions in response to the O2 concentration in soil atmosphere. The study was then conducted on small soil aggregates, water unsaturated, using 15N isotope technique to quantify nitrification and denitrification rates and N2 production.

2. Materials and methods 2.1. Soil sampling and conservation until measurements Experiments were performed on an Orthic Luvisol (FAO classification) sampled at Mons-en-Chausseé in Northern France (498800 N; 38600 E). The soil was cultivated with maize in 2000. Its properties were as follows: clay, 194 g kg21; silt, 706 g kg21; sand, 68 g kg21; pHwater, 8.20; total CaCO3, 32 g kg21; organic C, 8.52 g kg21; total N, 1.00 g kg21. Soil clods were sampled in the ploughed layer (10 –30 cm depth) after digging a trench, on September 2000. Two sets of clods were separated: clods D; with a massive structure and no visible porosity (resulting from compaction due to traffic) and clods G; with a fragmentary structure and visible porosity (Richard et al., 1999). The clods were gently broken down immediately after sampling and then calibrated at field moisture condition (0.184 g water g21 dry soil): we kept clods between 2.5 and 3 cm size. In order to reduce microbial activity during storage, the clods were air-dried during 3 days to obtain a residual moisture close to 0.10 g g21 soil (corresponding to a water suction of 2 2 MPa). At the beginning of the present experiment (June 2002), the clods were sieved to obtain aggregates between 2 and 3.15 mm. The soil then contained 0.07 g water g21 soil; its NO2 3 content was 10.7 mg N kg21 and its NHþ 4 content was 0.1 mg N kg21. The aggregates were rewetted by spraying deionised water to obtain a residual moisture close to 0.18 g g21, and were pre-incubated at 20 8C in airtight jars during 7 days. 2.2. Experiment 1: effect of NHþ 4 concentration The objective was to assess the influence of NHþ 4 concentration on nitrification rate. Twenty five grams fw (fresh weight) of soil aggregates were placed in 125 ml plasma flasks, and various amounts of NHþ 4 were added as (NH4)2SO4 solution to obtain 0, 80, 116 or 170 mg NHþ 4N kg21 soil. The addition of this solution rose the soil water content to 0.19 g g21. Twenty one flasks were closed and incubated at 20 8C for 14 days: six flasks were used for gas measurements (three flasks with soil and three flasks without soil used as blanks), three flasks were used for final soil pH measurement and 12 flasks for mineral nitrogen measurements. N2O and mineral N were analysed at day 2, 4, 7, 10 and 14 (three replicates at each date). Initial soil pH


and mineral N were also measured at day 0 on additional samples. At each date, the atmosphere of flasks was renewed by opening the flasks for a few minutes. The same procedure and measurements were applied for each NH þ 4 concentration. 2.3. Experiment 2: effect of O2 pressure Its objective was to assess the influence of O 2 concentration on nitrification, denitrification and N2O production, at a given soil NHþ 4 concentration, chosen as 116 mg N kg21 soil according to results of experiment 1. Twenty-five grams f.w. of soil aggregates were put into 125 ml flasks. A 15N labelled (NH4)2SO4 solution (50 atom% enrichment) was added to each soil sample to obtain 21 concentration and a residual the 116 mg NHþ 4 -N kg moisture of 0.21 g water g21 soil. The flasks were closed and placed under 6 different O2 concentrations (nominally 0, 0.5, 1, 2, 4.8 and 21 kPa O2), O2 being mixed with pure N2. Each flask received 3 successive cycles of 3 min vacuum and 3 min filling with a given mixture of O2/N2 at atmospheric pressure. Eighteen flasks were incubated at 20 8C for 14 days. The procedure was the same than in experiment 1, except that at each measurement date the atmosphere of flask was renewed with the corresponding O2/N2 gas mixture by the procedure described above. The same measurements were performed as in experiment 1. Additional measurements consisted in O2 concentration and 15 N atom% excess in atmosphere gas (N2, N2O) and mineral 2 N in soil (NHþ 4 and NO3 ). Oxygen consumption occurred during each measurement interval, so that the mean O2 pressure in each treatment was in fact: 0, 0.35, 0.76, 1.50, 4.30 and 20.4 kPa. These values will be used as effective O2 concentrations in the following.

2.4. Gas measurements For both experiments, at each date, 1 ml of gas was sampled from the flask with a syringe and injected into an automatic CN analyser (Carlo Erba, ANA 1500, Milan, Italy) for measuring O2 concentration. The analyser was adapted to O2 measurement by replacing the previous Porapak QS column by a molecular sieve column (60 –80 mesh, 1.8 m, 50 8C) which separates O2 and N2 gas and reducing the dead volume in the tube trapping CO2 and H2O gas. The CN analyser was coupled to an integrator (Shimadzu, C-R6A, Chromatopac), the specific thermal conductivity of O2 and N2 was accounted for in the calibration procedure using the six O2/N2 standard gas previously mentioned. The analysis of 15N2 was performed on a 0.2 ml gas aliquot sampled with a syringe, also injected in a modified C – N analyser. The gas sample successively passed through a water absorbent, a CO2 trap, a N2O cryogenic trap and an

689

oven filled with reduced copper to remove O2, before entering an isotope ratio mass spectrometer (Fisons, Isochrom-EA, Manchester, England). The analysis of N2O and 15N2O was conducted on a 250 ml gas sample taken in dual-ended flasks previously vacuumed. The analysis was made with the mass spectrometer after pre-concentration of N2O using the ‘trace gas’ system (Micromass, Manchester, England). 2.5. Soil mineral N and pH measurements The mineral N content of clods was extracted with a KCl 1 M solution (soil:solution ratio ¼ 1:5). Measurements were performed with a TRAACS 2000 analyser (Bran and Luebbe, Germany) using the methods proposed by 2 Kamphake et al. (1967) for NO2 3 and NO2 analysis and 15 þ Krom (1980) for NH4 analysis. The N-NHþ 4 and 2 15 N-(NO2 3 þ NO2 ) were separated successively by microdiffusion and collected on a glass fiber disc (6 mm diameter) impregnated with 10 ml of 1 M H2SO4 solution (Brooks et al., 1989). Each disc was then placed in a tin capsule and analysed with the C – N analyser-mass spectrometer equipment. The pH measurements were performed after shaking the soil with ultra pure water (1:2 massic ratio) during 10 min. The mixture was left for 2 min and the pH was recorded every minute during 5 min using a pH-meter and calomel/glass electrodes. Measurements were done on 10 replicates at the beginning of incubation and on three replicates at the end of incubation in each treatment.

2.6. Nitrogen rates calculations N rates were calculated using measurements of mineral -N and -15N and FLUAZ model (Mary et al., 1998). This model combines a numerical method for solving the differential system given by the N and 15N mass equations and a non linear fitting program for optimising the N rates parameters by minimising the difference between observed and simulated N and 15N data (amounts and isotopic excess of 2 NHþ 4 and NO 3 ). Nitrification rate could have been estimated from NO2 3 experimental accumulation, but this estimate would not be as accurate as using the model. The model was used in experiment 2 to calculate mineralisation ðmÞ; nitrification ðnÞ and denitrification ðdÞ for the different O2 concentrations. We assumed that immobilisation was negligible, since no readily decomposable organic carbon was added. NH3 volatilisation was also neglected, on the basis of previous measurements (results not shown). These assumptions were confirmed by the measured 15N balance 15 15 15 NO2 NO2 N2O þ 15N2) that did not (15NHþ 4 þ 2 þ 3 þ differ significantly from 100% in any of the treatments. We also assumed that the mineralisation rate was the same in all treatments at a given time. This assumption

690


allowed to calculate more precisely nitrification and denitrification rates. Nitrification rates were supposed to follow first order kinetics, whereas mineralisation and denitrification followed zero order kinetics during each time interval. FLUAZ model does not directly account for NO2 2 2 compartment. It considers the sum (NO2 2 þ NO3 ) and therefore calculates the ammonium oxidation rate (rate ni ), 2 i.e. oxidation into NO2 2 . The oxidation rate into NO3 (rate na ) was calculated afterwards using measurements of NO2 2 pool ðNi Þ; as follows: na ¼ ni 2

DNi Dt

ð1Þ

The nitrification rates were then compared with O2 consumption rates. Regarding the stoichiometry of the two

stages of nitrification relative to O2 consumption, we calculated the following nitrification rate:

n¼

3ni þ na 4

ð2Þ

which should be linked to the O2 consumption rate by nitrification by a 1:2 mol mol21 ratio.

3. Results 3.1. Effect of NHþ 4 concentration Fig. 1 shows the evolution of soil mineral N pools during the 14 days incubation at atmospheric O 2 concentration, measured in experiment 1. Ammonium and nitrate concentrations remained almost stable in the control (i.e. without addition of NHþ 4 ), whereas they decreased and increased versus time, respectively, for the three NHþ 4 treatments (addition of 80, 116 and 170 mg N kg21, Fig. 1a and b). Nitrite concentration increased then decreased with time in all treatments except in the control where it was negligible. The higher the initial 2 concentration of NHþ 4 , the higher was the peak of NO2 . 21 The maximal value was 22.7 mg N kg for the addition of 170 mg N kg21 (Fig. 1c). The net nitrification kinetics were calculated using the changes in mineral N contents (Table 1). They appear to þ depend on the initial NHþ 4 concentrations. Without NH4 addition, the net nitrification rate was low, in average 0.27 mg N kg21 d21: it was equal to net mineralisation rate. 2 The highest net NHþ 4 and NO2 oxidation rates were þ observed with the highest NH4 concentration (170 mg N kg21): they reached 19.1 and 23.7 mg N kg21 d21, respectively. The initial net rates of NHþ 4 oxidation were greater than the initial rates of NO2 2 oxidation. Then the net NO2 2 oxidation rates ðna Þ increased more with time than the NHþ 4 oxidation rates ðni Þ: Table 1 2 Net oxidation rates (mg N kg21 d21) of NHþ 4 ðni Þ and NO2 ðna Þ in soil incubated at atmospheric O2 concentration, for different time intervals and initial ammonium concentrations (0, 80, 116 and 170 mg N kg21). Note: na was calculated on the basis of NO2 3 accumulation and ni on the basis of 2 NHþ 4 disappearance and NO2 variation Period (days)

Fig. 1. Variation of mineral N (mg N kg21) with time in soil with various amounts of NHþ 4 added at time 0. A0: control (no addition), A80: 80 mg N kg21, A116: 116 mg N kg21; A170: 170 mg N kg21. (a) NHþ 4 2 concentration; (b) NO2 3 concentration; (c) NO2 concentration. Vertical bars are the standard errors.

0 –2 2 –4 4 –7 7 –10 10 –14

A0

A80

A116

A170

ni

na

ni

na

ni

na

ni

na

0.0 0.6 0.2 0.5 0.1

0.0 0.6 0.2 0.5 0.1

15.1 12.2 8.0 0.0 0.1

10.6 10.5 12.2 0.0 0.1

9.4 15.3 12.3 8.5 1.2

5.9 11.2 13.7 12.1 1.2

17.8 14.8 19.1 18.1 0.5

12.4 8.8 19.8 23.7 1.4


691

2 was the disappearance of NHþ 4 and the increase of NO3 , and 2 the greater was the NO2 accumulation. The maximal NO2 2 concentration found in soil at atmospheric O2 pressure was 15.3 mg N kg21 (Fig. 2c). The nitrification and denitrification rates, calculated using measurements of mineral N and 15N (Table 2) and FLUAZ model, are given in Table 3. Significant denitrification rates (i.e. greater than 0.2 mg N kg21 d21 corresponding to the accuracy of FLUAZ calculations) were only observed in the 0 and 0.35 kPa O2 pressure treatments. Nitrification 2 rates (both NHþ 4 and NO2 oxidation rates) were markedly reduced when O2 concentration decreased: they decreased by a factor of 4 – 10 when O2 pressure fell from 20.4 to 0.35 kPa. The nitrification rate at 0.35 kPa O2 remained very low and about constant throughout the incubation, at 1.6 mg N kg21 d21. The nitrification rates ðna Þ increased versus time in all other treatments, except for the highest two O2 concentrations: the nitrification rates reached a peak at day 7 and 10 for the 4.3 and 20.4 kPa O2 pressures respectively, and then decreased probably because of the exhaustion of the exchangeable NHþ 4 . The second step of nitrification (NO2 2 oxidation) in these two treatments was slightly slower at the beginning but then increased faster than the first step (NHþ 4 oxidation), as shown previously (experiment 1). For the lowest O2 pressures, 0.35 and 0.76 kPa, the two steps proceeded at the same rate, so that there was no NO2 2 accumulation. The reduction in O2 concentration then affected primarily the NHþ 4 oxidation step.

Fig. 2. Variation of mineral N (mg N kg21) with time in soil incubated at 2 different partial pressures of O2. (a) NHþ 4 concentration; (b) NO3 2 concentration; (c) NO2 concentration. Vertical bars are the standard errors.

3.2. Effect of O2 concentration 3.2.1. Mineral N and 15N Fig. 2 shows the evolution of soil mineral N pools during the 14 days of incubation at various O2 concentrations, measured in experiment 2 at a fixed initial NHþ 4 concentration (116 mg N kg 21). Ammonium concentration decreased and NO2 3 concentration increased in all treatments, except at 0% O2 concentration: in this case, NHþ 4 concentration remained more or less constant and NO2 3 concentration fell from 10.7 to 0.1 mg N kg21 within 7 days. NO2 2 concentration increased then decreased with time, except for the lowest O2 concentrations (0 and 0.35 kPa); in these treatments, nitrite concentration remained negligible during the whole incubation period. The mineral N content responded markedly to O 2 concentration. The higher the O2 concentration, the faster

3.2.2. N2O, 15N2O and 15N2 gases The cumulative N2O emissions by the soil submitted to different O2 concentrations are presented in Fig. 3a. The kinetics of nitrous oxide emissions was highly dependent on O2 concentration in the flask atmosphere. The anaerobic situation (0% O2) resulted in the highest production of N2O. The emission took place rapidly, since the amount of N2O produced at day 2 was 0.73 mg N kg21, corresponding to a rate of 30 nmol N2O kg21 s21. It levelled off after day 7, when NO2 3 concentration was very small. At day 14, the amount of N2O evolved was 1.16 mg N kg21. The emission rate also decreased with time with the 0.35 kPa O2 pressure but the intensity was much smaller than in the anaerobic treatment. The kinetics of emission were very different in the treatments with 0.76 and 1.5 kPa O2 pressure: the N2O emission rate increased with time. The amounts of N2O produced at day 14 decreased in the following order of treatments: 0 kPa O2 < 1.4 kPa O 2 . 0.76 kPa O2 . 4.3 kPa O2 . 20.4 kPa O2 < 0.35 kPa O2. This order seems quite logical except for the last two treatments, the similarity of which is a coincidence since N2O was produced mainly by nitrification at 20.4 kPa O2 and by denitrification at 0.35 kPa O2. The cumulative N2O production levelled off towards the end of the incubation, simultaneously to the nitrification process. The corresponding measurements of the ratio 15 N2O:15N2 are given in Fig. 3b. Knowing that N2 emitted

4.07 (3.49) 24.56 (4.84) 31.91 (3.13) 8.39 (5.67) 2.57 (0.71) 9.94 (0.01) 28.91 (4.29) 37.62 (4.46) 37.61 (0.31) 38.03 (0.24) 11.64 (2.92) 30.98 (2.02) 32.92 (4.56) 19.55 (6.52) 11.77 (2.50) 9.97 (0.40) 24.34 (1.43) 32.34 (1.53) 36.84 (0.47) 37.82 (0.24) 39.73 (2.72) 38.43 (1.80) 38.60 (1.22) 38.25 (2.82) 18.63 (1.85) 20.70 (2.12) 37.30 (6.83) 26.08 (2.41) 29.53 (5.14) 35.87 (1.06) 8.01 (0.78) 20.83 (0.66) 28.35 (1.37) 32.95 (1.17) 35.76 (0.15) 40.65 (1.97) 38.65 (2.07) 38.12 (2.29) 40.19 (1.09) 39.94 (0.26) 17.58 (2.02) 27.64 (0.40) 24.80 (0.55) 31.56 (1.36) 33.25 (1.05) 7.08 (3.24) 17.21 (1.76) 23.55 (0.25) 28.49 (0.15) 33.69 (0.11) 41.25 (1.26) 41.66 (0.57) 40.28 (1.60) 40.15 (1.26) 40.80 (0.38) 4.07 (3.43) 10.14 (3.80) 3.70 (2.10) 2.00 (1.45) 26.18 (4.75) 4.85 (0.50) 12.66 (0.44) 18.04 (0.30) 23.20 (0.60) 27.05 (0.69) 40.69 (2.07) 39.82 (2.86) 38.81 (3.27) 38.76 (2.27) 40.09 (1.37) 0.31 (0.05) 0.87 (0.42) 0.01 (0.01) 0.18 (0.04) 0.03 (0.04) 0.00 (0.01) 0.00 (0.01) 0.01 (0.01) 0.01 (0.01) 0.01 (0.01) 10–14

7–10

4–7

2–4

41.24 (1.50) 41.43 (0.69) 38.45 (1.02) 38.51 (0.72) 39.65 (0.61) 0–2

NO2 3 Period (days)

41.22 (1.29) 40.75 (0.84) 39.66 (1.03) 26.75 (0.26) 7.06 (0.01)

NO2 3 NHþ 4 N2O NO2 3 NHþ 4 NHþ 4 NO2 3 NHþ 4 NHþ 4 NHþ 4

N2O

0.35 kPa O2 0 kPa O2

NO2 3

N2 O

0.76 kPa O2

N2O

1.5 kPa O2

NO2 3

N2O

4.3 kPa O2

20.4 kPa O2

N2O

K. Khalil et al. / Soil Biology & Biochemistry 36 (2004) 687–699 Table 2 15 2 þ N atom% excess of NHþ 4 , NO3 and N2O (mean and standard errors within brackets) measured in soil or gas samples at different incubation dates and different partial pressures of O2. The atom% excess of NH4 and NO2 pools are the average of the values measured at the beginning and the end of each time interval 3

692

by the soil is exclusively derived from N2O, we can assume that both gases have the same isotopic composition. Therefore, the ratio N2O:N2 is equal to the measured ratio 15 N2O:15N2. This ratio was small, between 0.06 and 0.13, in the treatments 0 and 0.35 kPa O2. At day 14, the ratio was equal to 0.11 in the anaerobic treatment, so that the amount of N2 evolved was estimated at 11.1 mg N kg21. The ratio was much higher in all other treatments, with the maximum for a 1.5 kPa O2 pressure. These results suggest that N2O production by denitrification (up to N2 production) was the dominant process at 0 and 0.35 kPa O2 pressure, whereas N2O production mainly derived from nitrification in the other treatments since little N2 was produced. 3.3. Proportion of N2O derived from nitrification and denitrification The use of 15N tracing provides another means of determining the origin of N2O emissions. Table 2 shows the mean isotope compositions of N2O in atmosphere, NHþ 4 and 15 NO2 N 3 in soil during the five time intervals studied. The þ atom% excess of NH4 decreased slowly with time except for the highest O2 concentrations for which it fell rapidly at the end of the incubation due to the simultaneous consumption of 15NHþ 4 ions by nitrification and production of unlabelled NHþ 4 by ammonification of native organic N. Simultaneously, the 15N atom% excess of NO2 3 increased due to nitrification of the labelled NHþ 4 , except in the 0% O2 treatment. In this case, the NO2 3 remained unlabelled during the whole incubation period, which confirms that there was no nitrification. The isotopic excess of N2O was also close to 0 in this treatment, indicating that the N2O produced came from the unlabelled NO2 3 initially present in soil. In the 0.35 kPa O2 treatment, the isotopic composition of þ N2O was much closer to that of NO2 3 than to that of NH4 , confirming that denitrification was the dominating process. However, it is noticeable in this treatment and in others (particularly at 4.3 and 20.4 kPa O2) that the 15N atom% excess of N2O can be lower than the 15N atom% excess of 2 both NHþ 4 and NO3 . Such a result indicates that at least one of the two pools is not uniformly labelled. This is attributed to an incomplete diffusion of the added 15NHþ 4 within soil aggregates before being nitrified due to its adsorption on solid phase. The non uniformity was much more important þ for NO2 3 than for NH4 , since the soil contained much more 2 þ NO3 than NH4 (10.7 versus 0.1 mg N kg21). Two independent methods were used to calculate the proportion of N2O from nitrification or denitrification. The first method relies on the measurements of 15N2. It assumes that N2 emitted by the soil has the same composition than N2O, and that the ratio N2O:N2 due to denitrification is constant. We estimated its value in the treatment 0% O2 at 0.12 (see Fig. 7a). The second method relies on measure2 ments of 15N isotopic excess of N2O, NHþ 4 and NO3 . It assumes that N2O is derived from two pools, one being labelled (with the composition of NHþ 4 ), the other pool


693

Table 3 21 21 2 Nitrification rates (NHþ d ), calculated using measured N and 15N pools 4 oxidation rate, ni and NO2 oxidation rate, na ) and denitrification rate ðdÞ (mg kg and FLUAZ model in soil samples incubated at various partial pressures of O2 Period (days)

0–2 2–4 4–7 7–10 10–14

0 kPa O2

0.35 kPa O2

0.76 kPa O2

1.5 kPa O2

4.3 kPa O2

20.4 kPa O2

ni

na

d

ni

na

d

ni

na

d

ni

na

d

ni

na

d

ni

na

d

0.0 0.0 0.0 0.0 0.0

0.0 0.0 0.0 0.0 0.0

3.2 1.4 0.5 0.0 0.0

1.7 1.7 1.5 1.6 1.3

1.7 1.7 1.5 1.6 1.2

0.2 0.0 0.0 0.4 0.0

2.8 3.2 3.5 3.3 4.2

2.7 3.1 3.5 3.2 4.2

0.0 0.0 0.0 0.0 0.0

4.1 6.1 6.2 6.5 8.2

3.6 5.5 6.0 7.0 8.4

0.0 0.0 0.0 0.0 0.0

7.3 9.8 11.0 13.3 4.2

5.5 7.0 11.8 15.4 4.3

0.0 0.0 0.0 0.0 0.0

10.2 16.0 13.8 9.2 0.6

6.7 11.9 15.2 12.8 0.7

0.0 0.0 0.0 0.0 0.0

being unlabelled. It calculates the N2O isotopic abundance versus the abundance of each pool and their respective contribution to N2O production. The comparison between this calculated value and the measured 15N atom% excess of N2O enables to estimate the proportion of N2O emitted from the labelled pool, i.e. by nitrification. The detail of both methods is given in Appendix. Although method 1 was based on more questionable assumptions than method 2, the two methods gave close estimates of the proportion of N2O emitted from nitrification, since the regression equation between the two estimates was y ¼ 0:99x and the coefficient of determination was r 2 ¼ 0:85 ðn ¼ 30Þ: Thus we considered the average of the two estimates. The amounts of N2O emitted by nitrification and denitrification were calculated in each treatment with this method (Fig. 4). As expected, denitrification was the unique process responsible for N2O production in anoxic conditions (0% O2) and it stopped towards the end of the incubation due to the disapperance of NO2 3 . At 0.35 kPa O2, denitrification remained the major process producing N2O, except at the end of experiment when emissions by nitrification increased. For samples with 0.76 and 1.5 kPa O2, the proportion of N2O by nitrification was more important than denitrification at all measurement intervals. Maximum N2O production by nitrification was obtained for the 1.5 kPa O2 pressure. The emissions by nitrification decreased when O2 concentration increased from 4.3 to 20.4 kPa O 2, particularly after day 7, corresponding to the end of the nitrification process. However, emissions by denitrification took place during the first 2 days, even at the highest O2 concentrations (4.3 and 20.4 kPa).

and 20.4 kPa O2): r 2 ¼ 0:94; n ¼ 25 (Fig. 6). The slope of the regression, 2.02 ^ 0.12 mol O2 consumed mol21 N nitrified, is not significantly different from 2 mol O2 mol21 N, which is the theoretical value for O2 consumption due to nitrification (calculated according to Eq. (2)). The measured O2 consumption is the sum of O2 consumption by nitrification and respiration. The intercept of the regression line was 0.22 ^ 0.06 mmol 2 kg21 d21. This value represents the O2 consumption in absence of nitrification, due to respiration. We can assume that the respiration in treatments varying in nitrification rate was the same. The consumption of oxygen was then mainly due to nitrification, even in the 0.35 kPa O2 treatment with the lowest nitrification rate.

3.4. Relationship between O2 consumption and nitrification rate The kinetics of cumulative nitrification and O2 consumption were very similar (Fig. 5), both processes being favoured by the increase in O2 pressure during the whole incubation period. Indeed we found a very good correlation between O2 consumption rates and nitrification rates in the five treatments with oxygen (0.35, 0.76, 1.5, 4.3

Fig. 3. Kinetics of N2O and N2 emissions (mg N kg21) from soil incubated at different partial pressures of O2. (a) Cumulative N2O; (b) cumulative N2O: N2 ratio (equal to the 15N2O: 15N2 ratio).

694


Fig. 4. N2O emission rate (mg N kg21 d21) coming from nitrification and denitrification at different time intervals in soil incubated at different partial pressures of O2.

3.5. Relationship between N2O emissions and nitrification/denitrification rates The amounts of N2O produced were compared to the amounts of N nitrified or denitrified, previously calculated with FLUAZ (Table 3). In the 0 and 0.35 kPa O2 treatments, the amounts of N2O emitted by denitrification were highly correlated with N denitrified: the coefficient of determination being equal to 0.99 and the slope, which represents the ratio N2O:(N2O þ N2), equal to 0.11 (Fig. 7a). In the 0.76, 1.5, 4.3 and 20.4 kPa O2 treatments, the amounts of N2O emitted by nitrification were highly correlated with N nitrified (Fig. 7b). We obtained specific regression lines for each O2 concentration, which slopes, intercepts and correlation coefficients are given at Table 4. Results show that the intercept was not significantly different from 0 when we consider the N2O emitted by nitrification alone. The slope markedly increased when O2 availability decreased: the proportion of nitrified N evolved as N2O varied from 0.16 to 1.48% when O2 pressure fell from 20.4 to 0.76 kPa. If we consider the total N2O emission, the proportion varied from 0.24 to 1.80%. The total N2O emissions measured during each time interval were also compared with the soil NO2 2 concentrations at the end of each interval. The relationship was highly dependent on O2 concentration (Fig. 8). Linear relationships were found at 20.4 and 4.3 kPa O2 pressure, but not at lower O2 pressures. No significant correlation was found when we considered NO2 2 concentrations present in

soil at the beginning of each time interval. This indicates that NO2 2 was not a single determinant of N2O emissions by nitrification. 3.6. pH evolution The mean pH value at the beginning of incubation was 8.04 ^ 0.03. At day 14, the pH had increased to 8.20 ^ 0.01 in soil incubated at 0% O2, but had decreased at other O2 concentrations. It was 8.03 ^ 0.01, 7.93 ^ 0.01, 7.78 ^ 0.03, 7.74 ^ 0.01 and 7.81 ^ 0.01 at 0.76, 1.5, 4.3 and 20.4 kPa O2 pressure, respectively. The pH decrease was greatest for the highest O2 pressure, except for the 20.4 kPa treatment. The aerobic conditions favoured nitrification and consequently pH decrease, whereas at 0% O2, pH had increased because there was only denitrification. In the 20.4 kPa O2 treatment, the nitrification was over at day 14 and the higher pH is attributed to a pH reequilibration in soil (due to the carbonates equilibrium).

4. Discussion 4.1. Validity of the estimates of N rates One important condition required for a successful application of isotope methods to quantify N transformations is that the added inorganic 15N becomes homogeneously mixed with the native, unlabelled inorganic N in soil (Davidson et al., 1991). The comparison of the isotope


Fig. 5. Kinetics of nitrification and O2 consumption in soil incubated at different partial pressures of O2. (a) Cumulative nitrification (mg N kg21) calculated with FLUAZ model; (b) cumulative O2 consumption (mmol O2 kg21) measured. Nitrification rate is defined as ðð3ni þ na Þ=4Þ; where ni 2 is the NHþ 4 oxidation rate and na is the NO2 oxidation rate (see text). Vertical bars are the standard errors. 2 composition of NHþ 4 , NO3 and N2O (Table 3) suggests that this condition was not fulfilled in our experiment. This may be attributed to a slow diffusion of the added 15NHþ 4 within soil aggregates due to its high adsorption capacity. The 15 NO2 3 produced by nitrification could have been unevenly distributed soon after its production. Stevens et al. (1997), using a similar 15N application procedure, found that the NO2 3 pool was rather uniform, enough to perform

695

Fig. 7. Relationships between cumulative N2O emissions and cumulative N denitrified (a) or nitrified (b) in soil incubated at different O2 pressures. þ Nitrification is defined as NO2 2 production, i.e. cumulative NH4 oxidation rates. Regression equations are given at Table 4.

calculations of the contribution of nitrification and denitrification to N2O emission, using an isotope dilution equation. The difference with our experiment is that these authors used urea instead of NHþ 4 . Urea can diffuse rapidly within soil aggregates before being hydrolysed into NHþ 4, whereas NHþ 4 cannot. The lack of uniformity in our experiment has different consequences: (a) it should not affect the validity of Table 4 Linear regression parameters of cumulative N2O production (mg N kg21) versus cumulative nitrification (mg N kg21). Total N2O is the measured N2O emission whereas N2O ‘by nitrification’ is calculated using coefficient a (see Appendix). Nitrification is defined as NO2 2 production, i.e. cumulative NHþ 4 oxidation rates O2 pressure (kPa)

Fig. 6. Relationship between O2 consumption rates and nitrification rates (mmol kg21 d21) measured for the different time intervals and different O2 treatments. Nitrification rate is defined as ðð3ni þ na Þ=4Þ; where ni is the 2 NHþ 4 oxidation rate and na is the NO2 oxidation rate (see text).

0.76 1.5 4.3 20.4

Total N2O

N2O by nitrification

Slope (%)

Intercept (mg kg21)

r2

Slope (%)

Intercept (mg kg21)

r2

1.80 1.20 0.43 0.24

-0.02 0.01 0.07 0.01

0.994 0.999 0.951 0.979

1.48 1.09 0.42 0.16

0.00 0.00 0.00 0.00

0.983 0.999 0.892 0.915

696


Fig. 8. Relationship between total N2O emissions rates (mg N kg21 d21) 21 and NO2 2 concentrations (mg N kg ) measured in soil at the end of each time interval, for different partial pressures of O2.

nitrification estimates (using FLUAZ), because nitrification must have occurred in the soil zone where all the added þ 2 15 NHþ 4 was present. In this zone, the NH4 and NO3 can be considered as uniformly labelled. Davidson et al. (1991) have shown that an heterogeneous distribution of tracer in soil may have little effect on the rate estimates by tracing technique. The validity of our estimates of nitrification rates is also confirmed by the linear relationship with the O2 consumption rate, the slope of which is equal to the theoretical value (Fig. 6). (b) It affected the calculation of the proportion of N2O derived from nitrification, since the application of the isotope dilution equation was not possible. We proposed another method of calculation which accounts for two origins of N2O, one being derived 2 from labelled NHþ 4 and the other from unlabelled NO3 . The first origin is attributed to nitrification and the second one to denitrification. Although this assumption may be discussed, this calculation provided results which were in good agreement with a second independent method, based on the measured ratio 15N2O:15N2. This agreement supports our conclusions. Furthermore, the ratio N2O:(N2O þ N2) for denitrification, equal to 0.11 (Fig. 7a), is close to the estimates previously made with 15N2 measurements (0.10) and with the C2H2 inhibition method (0.13; Khalil et al., 2004). 4.2. Denitrification in well aerated aggregates Both above discussed methods have indicated that denitrification occurred even at atmospheric O2 pressure. This result is surprising since the moisture content was chosen (at 0.21 g g21 soil) to obtain unsaturated aggregates, considering the shrinking-swelling curve previously established on that soil (Sillon, 1999). The explanation may come from the technique of adding solution to soil aggregates using a pipette which does not deliver the same volume of solution to each aggregate; consequently, some aggregates may have been saturated after this addition. Anoxic conditions occur if the radius of a saturated aggregate is greater than a critical value (Renault and Stengel, 1994).

The critical radius, calculated using O2 consumption rate measured and O2 concentration in the atmosphere as proposed by these authors, varied from 1.0 mm in the treatment 0.35 kPa O2 to 3.5 mm in the treatment 20.4 kPa O2. Therefore, anoxic denitrification was possible in some of the 3 mm size aggregates at 4.3 and 20 kPa O2. A second possible explanation is the occurrence of aerobic denitrification, although the mechanism of this process, biological or chemical, remains unclear. Aerobic biological denitrification activity can be due to denitrifying bacteria such as Pseudomonas, Aeromonas, Moraxella, Paracoccus, Microvirgula and Arthrobacter, which are able to use nitrate as terminal electron acceptor in the presence of oxygen (Robertson and Kuenen, 1984; Robertson et al., 1989). It could also result from autotrophic nitrifying bacteria like Nitrosomonas which use both oxygen and nitrite as terminal electron acceptor (Robertson et al., 1989; Schmidt and Bock, 1997). Laughlin and Stevens (2002) showed that fungi could be responsible for N2O production by denitrification under aerobic conditions in temperate grassland soils. Garrido et al. (2002) concluded that aerobic denitrification was the main process for N2O production in Redoxic Luvisol in atmospheric conditions. Chemical N2O production, already observed in acid soils (Nelson and Bremner, 1970), is very unlikely in our calcareous soil. 4.3. Effect of O2 availability on nitrification and N2O production Our results show that O2 availability greatly influences the nitrification rates, which were reduced by a factor of 6– 9 when O2 pressure decreased from 20.4 to 0.35 kPa. Our results are close to those obtained by Goreau et al. (1980) for pure cultures of Nitrosomonas europaea, in which the nitrification rate decreased by a seven fold factor when O2 was reduced from 20 to 0.5 kPa. Very few results have been reported in soils about the direct O2 effect. Our results also allow to estimate the Michaelis constant of NHþ 4 oxidation relative to O2 pressure, at the beginning of the experiment when NHþ 4 concentrations were much higher than the Michaelis constant for NHþ 4 which is about 1 mol m23 (Laanbroek and Gerards, 1993), corresponding to 3 mg N kg21. The nitrification rates versus O2 concentration were fitted to Michaelis-Menten kinetics to obtain Vmax and Km at three time intervals (0 –2, 2 –4 and 4– 7 days). We found Km ¼ 2:5 ^ 0:6 kPa O2, corresponding to 1.1 ^ 0.3 mol O2 m23 air. This value is much greater than those reported by Laanbroek and Gerards (1993) for N. europaea grown in continuous cultures: 1.3 – 15 mmol m23. Such a difference is surprising and cannot be easily explained. The Vmax value, calculated with our results, increased with time from 0.8 to 1.3 mmol kg21 d21 at 20 8C, due to the growth of nitrifiers in soil. Oxygen concentration also exerted a marked effect on N2O production. We found that the yield of N2O emission by nitrification, i.e. the amount of N2O-N emitted per unit of


NH4-N oxidised, increased rapidly from 0.16 to 1.48% when O2 pressure fell from 20.4 to 0.76 kPa. The maximum yield was obtained at 0.76 kPa but we had not enough accuracy to make calculations at 0.35 kPa O2 pressure. Anderson et al. (1993) found that the optimum O2 pressure for N2O production in pure culture was 0.3 kPa O2 for N. europaea and 2 – 4 kPa for the heterotrophic nitrifier Alcaligenes faecalis. They obtained a yield of 1% at 5 kPa O2 (in our study 0.42% at 4.3 kPa). Goreau et al. (1980) reported that N2O production by N. europaea in pure culture increased by 3 –4 times when O2 pressure fell from 20 to 1 kPa. They measured higher yields than ours: 0.3, 0.9, 3 and 8% at 20, 5, 1 and 0.5 kPa O2 pressure, respectively. However, they also measured yields in isolates of Nitrosomonas, Nitrosolobus and Nitrosospira from soils which were similar to ours at 21 kPa O2: 0.47, 0.09 and 0.11%, respectively. Bollmann and Conrad (1998), working on soils, found that the maximum N2O emission by nitrification occurred at 0.5 kPa O2. Using the results published by Stevens et al. (1997), we could calculate a yield similar to ours at atmospheric pressure: 0.18 ^ 0.01%, independent of the water content (40, 50 and 60% WHC). A large variation in yield has been reported at 20 kPa O2: 0.03 –1% (Garrido et al., 2002), 0.5– 2% (Bolle, 1986), 0.09 –0.28% (Breitenbeck et al., 1980) and 0.02% (Tortoso and Hutchinson, 1990). The origin of N2O emitted by nitrification is still on debate. Ritchie and Nicholas (1972) suggested that NHþ 4 oxidisers reduced NO2 2 to N2O to minimize intracellular accumulation of NO2 2 which is toxic. Remde and Conrad (1990) showed that N2O could derive from nitrite produced inside the cells. Poth and Focht (1985) confirmed this hypothesis for N. europaea cultivated in pure culture at various O2 pressures and called it denitrification of nitrite, the nitrite being the terminal electron acceptor. They indicated that O2 was required in NHþ 4 oxidising bacteria only for the first oxidation step: oxidation of NHþ 4 into hydroxylamine by mono-oxigenase.

5. Conclusions In this study, we have evaluated a method to characterise the effect of O2 pressure on N2O emissions associated with the nitrification process in a Luvisolic soil. The method combines the use of small unsaturated aggregates and the use of 15N tracer to calculate precisely nitrification rates and distinguish between denitrification and nitrification. Results show that O2 pressure exerts a progressive and strong effect on both the nitrification rates and the proportion of N2O emitted per unit of N nitrified. These results are consistent with those obtained in pure cultures of autotrophic nitrifiers. The high O2 requirement for nitrification suggests that part of the nitrification process in arable soils could occur in microsites at an O2 pressure lower than the atmospheric one. Consequently this would slow down the nitrification rate

697

and increase the proportion of nitrified N emitted as N2O. This hypothesis deserves to be evaluated in mechanistic models simulating nitrification and denitrification processes as a function of soil structure and biological activity.

Acknowledgements This work was supported by the French programs PNSE and GESSOL, the Region Picardie and INRA. We thank F. Barrois and G. Alavoine for technical assistance and O. Delfosse for improving gas equipment and conducting the isotope analyses.

Appendix A Let a be the proportion of N2O derived from nitrification and ð1 2 aÞ the proportion derived from denitrification. Two methods can be used to calculate this proportion:

A.1 Method 1: use of the ratio

15

N2O:15N2

Three hypotheses are made: (H1) the ratio N2O:N2 due to denitrification process remains constant with time and treatments; (H2) N2O and N2 emitted by denitrification have the same origin and isotope composition, so that the ratio N2O:N2 is equal to the ratio 15N2O:15N2; (H3) no N2 is produced during nitrification. Let Q be the total amount of 15N2O produced (mg N kg21) Let D be the total amount of 15N2 produced (mg N kg21) Let QD be the amount of 15N2O produced by denitrification (mg N kg21) Let RD be the ratio 15N 2O: 15N 2 associated to denitrification Let R be the overall ratio 15N2O:15N2 measured RD ¼

QD ðhypothesis H3Þ D

ðA1:1Þ

RD ¼

Qð1 2 aÞ ðhypothesis H2Þ D

ðA1:2Þ

R¼

Q D

RD ¼ Rð1 2 aÞ ðhypothesis H1Þ

ðA1:3Þ

ðA1:4Þ

698


Since:

Table 5 14

Population Proportion Atom% N isotope abundance frequency P1 P2 P1 þ P2

a 12a 1

A1 A2 A

a1 a2

15

N isotope frequency

Isotope ratio

b1 ¼ 1 2 a1 R1 b2 ¼ 1 2 a2 R2 R

so that:

15

N2O

2b1 a1

ðA2:1Þ

Isotope ratio of population 2: 2b2 a2

ðA2:2Þ

Mass 44 of the mixture: M44 ¼ a21 a þ a22 ð1 2 aÞ

ðA2:3Þ

Mass 45 of the mixture: M45 ¼ 2a1 b1 a þ 2a2 b2 ð1 2 aÞ

ðA2:4Þ

Isotope ratio of the mixture: R¼

M45 2a1 b1 a þ 2a2 b2 ð1 2 aÞ ¼ M44 a21 a þ a22 ð1 2 aÞ

ðA2:5Þ

Eqs. (A2.1) and (A2.2) yield: a1 ¼

2 2 þ R1

ðA2:6Þ

a2 ¼

2 2 þ R2

ðA2:7Þ

b1 ¼

R1 2 þ R1

ðA2:8Þ

b2 ¼

R2 2 þ R2

ðA2:9Þ

Replacing a1 ; b1 ; a2 and b2 in Eq. (A2.5) gives: R¼

aR1 ð2 þ R2 Þ2 þ ð1 2 aÞR2 ð2 þ R1 Þ2 að2 þ R2 Þ2 þ ð1 2 aÞð2 þ R1 Þ2

A¼

a¼

This method assumes that the N2O produced has two origins, each one having a specific isotopic composition. N2O signal is measured at mass 44 (14N14N16O) and mass 45 (14N15N16O). (Table 5) Isotope ratio of population 1:

R2 ¼

ðA2:11Þ

it comes:

ðA1:5Þ

A.2

R1 ¼

R 2þR

aA1 ð1 2 A1 Þ þ ð1 2 aÞA2 ð1 2 A2 Þ að1 2 A1 Þ þ ð1 2 aÞð1 2 A2 Þ

ðA2:12Þ

so that:

R a¼12 D R

Method 2: use of the isotope composition of

A¼

ðA2:10Þ

ð1 2 A2 ÞðA2 2 AÞ ð1 2 A2 ÞðA2 2 AÞ 2 ð1 2 A1 ÞðA1 2 AÞ

ðA2:13Þ

References Abbasi, M.K., Adams, W.A., 1998. Loss of nitrogen in compacted grassland soil by simultaneous nitrification and denitrification. Plant and Soil 200, 265 –277. Anderson, I.C., Poth, M., Homstead, J., Burdige, D., 1993. A comparison of NO and N2O production by the autotrophic nitrifier Nitrosomonas europaea and the heterotrophic nitrifier Alcaligenes faecalis. Applied and Environmental Microbiology, 3525– 3533. Blackmer, A.M., Bremner, J.M., Schmidt, E.L., 1980. Production of nitrous oxide by ammonia oxidizing chemoautotrophic microorganisms in soil. Applied and Environmental Microbiology 40, 1060–1066. Bolle, H.J., Seiler, W., Bolin, B., 1986. Other greenhouse gases and aerosols. In: Bolin, B., Jager, J., Warrick, R.A., Doos, B.R. (Eds.), The Greenhouse Effect, Climatic Change and Ecosystems, Wiley, New York, pp. 157–203. Bollmann, A., Conrad, R., 1998. Influence of O2 availability on NO and N2O release by nitrification and denitrification in soils. Global Change Biology 4, 387–396. Breitenbeck, G.A., Blackmer, A.M., Bremner, J.M., 1980. Effects of different nitrogen fertilizers on emission of nitrous oxide from soil. Geophysical Research Letters 7, 85 –88. Bremner, J.M., Blackmer, A.M., 1981. Terrestrial nitrification as a source of atmospheric nitrous oxide. In: Delwiche, C.C., (Ed.), Denitrification, Nitrification and Atmospheric Nitrous Oxide, Wiley, New York, pp. 151 –170. Brooks, P.D., Stark, J.M., McInteer, B.B., Preston, T., 1989. Diffusion method to prepare soil extracts for automated nitrogen-15 analysis. Soil Science Society of America Journal 53, 1707–1711. Conrad, R., 1996. Soil microorganisms as controllers of atmospheric trace gases (H2, CO, CH4, OCS, N2O and NO). Microbiology Review 60, 609 –640. Davidson, E.A., Hart, S.C., Shanks, C.A., Firestone, M.K., 1991. Measuring gross nitrogen mineralization, immobilization and nitrification by 15N isotopic pool dilution in intact soil cores. Journal of Soil Science 42, 335–349. Davidson, E.A., 1993. Soil water content and the ratio of nitrous oxide to nitric oxide emitted from soil. In: Uremland, R.S., (Ed.), Biogeochemistry of Global Change, Chapman & Hall, New York, pp. 369–386. Garrido, F., Heńault, C., Gaillard, H., Pe´rez, S., Germon, J.C., 2002. N2O and NO emissions by agricultural soils with low hydraulic potentials. Soil Biology & Biochemistry 34, 559–575. Goreau, T.J., Kaplan, W.A., Wofsy, S.C., McElroy, M.B., Valois, F.W., Watson, S.W., 1980. Production of NO2 2 and N2O by nitrifying bacteria at reduced concentrations of oxygen. Applied and Environmental Microbiology 40, 526 –532. Grant, R.F., 1995. Mathematical modelling of nitrous oxide evolution during nitrification. Soil Biology & Biochemistry 27, 1117–1125. Groffman, P.M., 1991. Ecology of nitrification and denitrification in soil evaluated at scales relevant to atmospheric chemistry. In: Rogers, J.E., Whitman, W.B. (Eds.), Microbial Production and Consumption of

K. Khalil et al. / Soil Biology & Biochemistry 36 (2004) 687–699 Greenhouse Gases: Methane, Nitrogen Oxides and Halomethanes, American Society for Microbiology, Washington, DC, pp. 201–217. Heńault, C., Germon, J.C., 1995. Quantification de la deńitrification et des e´missions de protoxyde d’azote (N2O) par les sols. Agronomie 15, 321–355. IPCC, 1996. Contribution of working group I to the second assessment report of the intergovernmental panel on climate change. In: Houghton, J.T., Meira Filho, L.G., Callander, B.A., Harris, N., Kattenberg, A., Maskell, K. (Eds.), Climate change 1995: the science of climate change, Cambridge University Press, Cambridge. Kamphake, L.J., Hannah, S.A., Cohen, J.M., 1967. Automated analysis for nitrate by hydrazine reduction. Water Research 1, 205–216. Khalil, K., Renault, P., Mary, B., Gue´rin, N., 2004. Modelling denitrification including the dynamics of denitrifiers and their progressing ability to reduce nitrous oxide: comparison with batch experiments. European Journal of Soil Science in press. Krom, M.D., 1980. Spectrometric determination of ammonia: a study of a modified Berthelot reaction using salicylate and dichloroisocyanurate. The Analyst 105, 305 –316. Laanbroek, H.J., Gerards, S., 1993. Competition for limiting amounts of oxygen between Nitrosomonas europaea and Nitrobacter winogradskyi grown in mixed continuous cultures. Archives of Microbiology 159, 453–459. Laughlin, R.J., Stevens, R.J., 2002. Evidence for fungal dominance of denitrification and codenitrification in a grassland soil. Soil Science Society of America Journal 66, 1540–1548. Linn, D.M., Doran, J.W., 1984. Effect of water-filled pore space on carbon dioxide and nitrous oxide production in tilled and non-tilled soils. Soil Science Society of America Journal 48, 1267– 1272. Mary, B., Recous, S., Robin, D., 1998. A model for calculating nitrogen fluxes in soil using 15N tracing. Soil Biology & Biochemistry 30, 1963–1979. Nelson, D.W., Bremner, J.M., 1970. Gaseous products of nitrite decomposition in soils. Soil Biology & Biochemistry 2, 203 –215. Nielsen, T.H., Nielsen, L.P., Revsbech, N.P., 1996. Nitrification and coupled nitrification-denitrification associated with a soil-manure interface. Soil Science Society of America Journal 60, 1829–1840. Parton, W.J., Mosier, A.R., Ojima, D.S., Valentine, D.W., Schimel, D.S., Weier, K., Kulmala, A.E., 1996. Generalized model for N2 and N2O production from nitrification and denitrification. Global Biogeochemical Cycles 10, 401–412. Poth, M., Focht, D.D., 1985. 15N kinetic analysis of N2O production by Nitrosomonas europaea: an examination of nitrifier denitrification. Applied and Environmental Microbiology 49, 1134–1141.

699

Remde, A., Conrad, R., 1990. Production of nitric oxide in Nitrosomonas europaea by reduction of nitrite. Archives of Microbiology 154, 187– 191. Renault, P., Stengel, P., 1994. Modeling oxygen diffusion in aggregated soils. I. Anaerobiosis inside the aggregates. Soil Science Society of America Journal 58, 1017–1023. Richard, G., Boizard, H., Roger-Estrade, J., Boiffin, J., Gue´rif, J., 1999. Study of soil compaction due to traffic in field conditions: pore space and morphological analysis. Soil and Tillage Research 51, 151– 160. Ritchie, G.A.F., Nicholas, D.J.D., 1972. Identification of the sources of nitrous oxide produced by oxidative and reductive processes in Nitrosomonas europaea. Biochemistry Journal 126, 1181–1191. Robertson, L.A., Cornelisse, R., Vos, P.D., Hadioetomo, R., Kuenen, J.G., 1989. Aerobic denitrification in various heterotrophic nitrifiers. Antonie Van Leeuwenhoek Journal of Microbiology 56, 289– 299. Robertson, L.A., Kuenen, J.G., 1984. Aerobic denitrification, a controversy revived. Archives of Microbiology 139, 351 –354. Schmidt, I., Bock, E., 1997. Anaerobic oxidation with nitrogen dioxide by Nitrosomonas eutropha. Archives of Microbiology 167, 106–111. Sillon, J.F., 1999. Etude expe´rimentale et mode´lisation des effets de la structure de la couche laboureé sur le re´gime hydrique du sol nu en phase de desse`chement. PhD Thesis. INA-PG, Paris, 222 pp. Smith, K.A., 1990. Greenhouse gas fluxes between land surfaces and the atmosphere. Progress in Physical Geography 14, 349–372. Stevens, R.J., Laughlin, R.J., Burns, L.C., Arah, J.R.M., Hood, R.C., 1997. Measuring the contributions of nitrification and denitrification to the flux of nitrous oxide from soil. Soil Biology & Biochemistry 29, 139– 151. Stevens, R.J., Laughlin, R.J., 1998. Measurement of nitrous oxide and dinitrogen emissions from agricultural soils. Nutrient Cycling in Agroecosystems 52, 131–139. Tiedje, J.M., 1988. Ecology of denitrification and dissimilatory nitrate reduction to ammonium. In: Zehnder, A.J.B., (Ed.), Biology of anaerobic microorganisms, Wiley, New York, pp. 179– 244. Tortoso, A.C., Hutchinson, G.L., 1990. Contributions of autotrophic and heterotrophic nitrifiers to soil NO and N2O emissions. Applied and Environmental Microbiology 56, 1799– 1805. Wood, P.M., 1986. Nitrification as a bacterial energy source. In: Prosser, J.I., (Ed.), Nitrification, IRL Press, Oxford, pp. 39–62. Zanner, C.W., Bloom, P.R., 1995. Mineralization, nitrification and denitrification in histosols of northern Minnesota. Soil Science Society of America Journal 59, 1505–1511.