Time-lagged induction of N₂O emission and its trade-off with NO emission from a nitrogen fertilized Andisol

Abstract

To understand the influence of basal application of N fertilizer on nitrification potential and N₂O and NO emissions, four soil samples were collected from an upland Andisol field just before (sample 1) and 4 (sample 2), 36 (sample 3) and 72 (sample 4) days after the basal application of N fertilizer during the Chinese cabbage growing season from 12 September to 30 November 2005. The potentials of N₂O production and nitrification of the soils were determined using a ¹⁵N tracer technique and the soils were incubated for 25 days at 25°C and 60% water-filled pore space (WFPS). The results revealed that as much as 84–97% N₂O and almost all NO were produced by nitrification. The ¹⁵N₂O emission peak occurred approximately 350 h after the beginning of incubation for samples 1 and 2, but just 48 h later in samples 3 and 4. Total ¹⁵N₂O emission during the 25-day incubation of samples 3 and 4 ranged from 190 to 198 µg N kg⁻¹ soil, which was significantly higher than the 99–108 µg N kg⁻¹ soil recorded in samples 1 and 2. Basal application of N fertilizer did not immediately increase the nitrification potential and the ratio of N₂O to N added, but did dramatically increase the nitrification potential and the ratio of N₂O to N added as (¹⁵NH₄)₂SO₄ 36–72 days after the basal N fertilizer was added. In contrast, NO emission was negatively correlated with nitrification potential and total N₂O emission. As a result, a trade-off relationship between total NO and N₂O emissions was identified. The results indicated that there was a time-lagged induction of the change of N turnover in the soil, which was possibly caused by slow population growth of the nitrifiers and/or a slow shift in the microbial community in the soil.

Key words:

• mineral N fertilizer
• nitric oxide
• nitrification
• nitrous oxide
• ¹⁵N tracer technique

INTRODUCTION

Atmospheric N₂O concentration has been increasing (Intergovernmental Panel on Climate Change 2001). Increases in the atmospheric concentration of N₂O contribute not only to global warming, constituting 6% of the anthropogenic greenhouse effect, but directly to the destruction of the stratospheric ozone through NO production (Intergovernmental Panel on Climate Change 2001). Nitric oxide participates in the regulation of the oxidant balance and is oxidized into NO₂ in the atmosphere, while redeposition of NO and NO₂ as acid rain contributes to acidification and eutrophication of ecosystems (Crutzen 1983; Nakahara et al. 2003; Vitousek et al. 1997). Agricultural soils are a major source of N₂O, contributing approximately 6.3 Tg N₂O-N per year or 35% of the global annual emission (Kroeze et al. 1999; Mosier et al. 1998), and are also considered to be an important source of NO, albeit not as important as fossil fuel combustion sources, accounting for approximately 5.5 Tg NO-N per year (Davidson and Kingerlee 1997). Nitrous oxide is produced in soils as a byproduct or intermediate product of nitrification and denitrification processes (Bouwman 1990; Minami 2005), as well as chemo-denitrification (van Cleemput 1998) and fungal transformations (Laughlin and Stevens 2002). The magnitude of N₂O and NO production and subsequent emission is determined by N supply, climatic factors and soil properties; the latter will determine the microbial reaction rate, soil aerobicity and gas diffusivity (Akiyama et al. 2006). Nitric oxide emissions are predominantly associated with nitrification and are favored by soil characteristics that provide good aeration, coarse texture and low soil water contents, whereas N₂O production and emission is stimulated by increased soil water content and compaction by denitrification (Cai et al. 1998; Granli and Bockman 1994).

In Japan, up to 50% of agricultural fields for consecutive upland crop cultivation are covered with volcanic ash soils or Andisols (Classification Committee of Cultivated Soils 1996). Volcanic ash soils tend to maintain their aerobic conditions because of their high porosity and, thus, as much as 87–92% of the N₂O produced is derived from autotrophic nitrification (Inubushi et al. 1996; Kusa et al. 2006). A number of previous field studies have shown that N₂O emissions from supplemental N fertilizer were higher than those from basal N fertilizer (Dobbie and Smith 2003; Jambert et al. 1997; Meng et al. 2005; Williams et al. 1999; Yan et al. 2001). For example, Xiong et al. (2006) did not observe any emission peaks of N₂O following the first application of N fertilizer, but did record peaks after the second to fifth applications of N fertilizer during five consecutive vegetable crops. Nishimura et al. (2005) also did not report any N₂O peaks after basal N fertilization in upland rice field, but did record peaks approximately 10 days after the application of supplemental N fertilizer. In contrast, Akiyama et al. (2000) found that the N₂O emission peak following basal N application was higher than that after supplemental fertilizer. To understand the dynamic change in the nitrification potential induced by basal N fertilization and the responses in microbial N₂O and NO emissions from agricultural soils to basal and supplemental application of N fertilizer, a laboratory incubation experiment was conducted using Andisol soil samples collected at different times after the basal application of N fertilizer during the Chinese cabbage growth period.

MATERIALS AND METHODS

Field experiment

The experiment was conducted at an experimental field in the National Institute for Agro-Environmental Sciences, Tsukuba, Japan (36°01′N, 140°07′E). The field includes six lysimeter plots, each with a 9 m² (3 m × 3 m) cross-sectional area and a 1.2 m depth. The soil is a light-colored Andisol (Hapludand; Classification Committee of Cultivated Soils 1996). The bulk density of the topsoil is 0.92 g cm⁻³. The soil pH in H₂O is 5.90. Total organic C and N contents in the topsoil are 31.3 g C kg⁻¹ soil and 2.60 g N kg⁻¹ soil, respectively (Akiyama et al. 2000). All plots were uniformly managed for sorghum cultivation for 2 years before experimentation began.

Prior to the sowing of Chinese cabbages on 12 September 2005, compound fertilizer, including urea (15 g N m⁻²), phosphate (15 g P₂O₅ m⁻²) and potassium (15 g K₂O m⁻²), together with fresh cattle manure (1 kg m⁻²) containing approximately 7 g N kg⁻¹ were used as basal fertilizer and spread uniformly on the soil and then incorporated into the top soil layer of 0–10 cm using a portable rotary tiller. No supplemental fertilizer was applied during the growing season. The cabbage was harvested on 30 November 2005.

Soil sampling

During the 80-day Chinese cabbage growing period, soil samples were collected for incubation four times on 12 September 2005, just before (hereinafter referred to as sample 1), 16 September, 4 days (sample 2), 18 October, 36 days (sample 3) and 23 November, 72 days (sample 4) after basal fertilization (Fig. 1). At each time, eight soil samples were taken from depths between 0 and 10 cm at different locations in each lysimeter plot and then all samples from the six lysimeter plots were mixed thoroughly to form a composite. The NH⁺ ₄-N and NO⁻ ₃-N concentrations in field samples 1, 2, 3 and 4 were 2.13 and 7.57, 245.75 and 52.35, 0.57 and 4.76, and 0.44 and 1.03 mg N kg⁻¹ soil, respectively. This indicates that most of the NO⁻ ₃ converted from the NH⁺ ₄ applied was either absorbed by Chinese cabbage and microbes or leached out of the top soil 36 days after the basal application of N fertilizer as a result of the light texture of the soil.

Incubation process

The soil samples collected from the field were rapidly transported to the laboratory for sieving (2 mm). Within

Figure 1  Experimental schedule.

24 h after sampling, each 15 g (on an oven-dried basis) fresh soil sample was weighed and distributed among 72 125-mL plastic bottles, and 50 g was added to nine 500-mL incubation bottles. All bottles were divided into three groups (i.e. 24 for each group). Pure water and (¹⁵NH₄)₂SO₄ solution or K¹⁵NO₃ (1.07 atom% ¹⁵N) solution were added to each bottle of group one and two, respectively, using a mini-pipette to adjust the soil to 60% water-filled pore space (WFPS) soil moisture and the concentration of NH⁺ ₄ or NO⁻ ₃ to 260 mg N kg⁻¹ (oven-dried weight). Pure water only was added to each bottle in group three to act as a control. All bottles were then covered with aluminum foil (punctured with needle holes to maintain an aerobic atmosphere in the bottles) and were incubated in the dark at 25°C. During the incubation, lost water was replaced every 3 days by mini-pipette. Three replicates of the 24 plastic bottles in each group were removed from the incubator at 12, 48, 108, 180, 276, 372, 492 and 612 h after the beginning of incubation to measure changes in soil NH⁺ ₄ and NO⁻ ₃ concentrations by incubating with 10% KCl extraction solution for 30 min. The extracted solutions were filtered and stored in a deep freezer (–18°C) until analysis. The NH⁺ ₄ and NO⁻ ₃ concentrations were measured using a colorimetric method and a TRAACS 2000 Continuous-Flow Analyzer (BRAN + LUEBBE, Hamburg, Germany). The ¹⁵N enrichment of NH⁺ ₄ and NO⁻ ₃ were determined using a diffusion methodology (Brookes et al. 1989) and analysis on a MAT 252 isotope ratio mass spectrometer (Thermo Electron, Bremen, Germany). Changes in soil (¹⁵NH₄)₂SO₄-derived NH⁺ ₄ concentration between successive sampling days were defined as the net nitrification rate between the sampling days.

Three incubation bottles from each group were used to measure gaseous N₂O and NO emissions at the above-mentioned times. For NO measurement, the bottles were allowed to aerate for approximately 30 min prior to being capped with a butyl rubber stopper. The first 5-mL sample was collected using a plastic syringe immediately after capping and analyzed immediately. Two more samples were collected and injected into a Chemiluminescence NO_x analyzer (Model 42C, Thermo Environmental Instruments, Franklin, MA, USA) at 3.5 min interval to determine NO concentration. After the third sampling, the bottle was re-opened and left until the air established equilibrium inside the bottle and then re-capped and incubated. Just after and 12 h after closure, 1 mL headspace gas was sampled using an airtight syringe and used for N₂O measurement. The N₂O concentrations were analyzed using a gas chromatograph equipped with an electron capture detector at 340°C; the main separating column was packed with Porapak Q and the carrier gas was 5% methane in argon (GC-ECD 8A, Shimadzu, Kyoto, Japan). Samples for ¹⁵N-N₂O measurement were stored in 50-mL gas-tight glass bottles and their isotopic enrichment was determined on a MAT 252 Mass Spectrometer with a PreCon system. The ¹⁵N-enriched gas fluxes were calculated from the atom% excess of samples, taking account of the atom% excess ¹⁵N of N applied.

Statistical analyses

Statistical analyses were conducted using SPSS. Statistically significant differences in total N₂O emission among soils were analyzed using anova and least significant difference (LSD) calculations at a 5% significance level (P = 0.05). Linear regression analysis was used to identify significant correlations between N₂O and NO emissions.

RESULTS

Nitrous oxide emission

Table 1 shows the net nitrification rate and cumulative N₂O emission from ¹⁵N-labeled fertilizer added to an

Table 1 Net nitrification rates and supplemental fertilizer derived N2O emission from the soils during the first 180 h and 25-day incubation after adding (15NH4)2SO4 or K15NO3 fertilizer and incubating at 25°C and 60% water-filled pore space

Fertilizer	Sample	Net nitrification rate			N2O emission			Ratio of N2O to N added (%)
		First 180 h (mg N kg^−1)	180–612 h (mg N kg^−1)	First 180  h/total (%)	First 180 h (µg N kg^−1)	Total (µg N kg^−1)	First 180  h/total (%)
(^15NH4)2SO4	1	91.3	169	35.1	30.7 ± 3.8 a	99.3 ± 12.6 a	31	0.038
	2	86.3	174	33.2	27.4 ± 2.2 a	108 ± 2 a	25	0.041
	3	210	49.6	80.9	179 ± 14 b	190 ± 18 b	94	0.072
	4	250	10.0	96.1	190 ± 6 b	198 ± 6.61 b	96	0.075
K^15NO3	1	–	–	–	2.06 ± 0.04 a	3.17 ± 0.84 a	65	0.001
	2	–	–	–	10.1 ± 0.6 b	17.8 ± 1.20 b	57	0.007
	3	–	–	–	4.64 ± 2.37 a	7.98 ± 5.70 ab	58	0.003
	4	–	–	–	1.80 ± 0.68 a	7.40 ± 1.12 ab	24	0.008
Mean ± standard deviation (n = 3). Values within the same column followed by different letters are significantly different at P < 0.05.

Figure 2  (a) Total N2O emission from the soils during the 25-day incubation with and without supplemental N addition. CK represents the treatment without supplemental N addition. (b) Supplemental (15NH4)2SO4-derived N2O emission from the soils during the 25-day incubation. Error bars denote the standard deviation (n = 3).

Andisol during the 25-day incubation period. There was no marked N₂O emission when ¹⁵ NO⁻ ₃-N was added to soils, even in soils that had been amended with fresh cattle manure of 1 kg m⁻² just 4 days before sampling of sample 2 (Table 1). In contrast, N₂O emissions occurred in pulses when ¹⁵NH⁺ ₄-N was added to all four samples (Fig. 2, Table 1). The ratio of ¹⁵N₂O emission from ¹⁵NO⁻ ₃-N added to ¹⁵NH⁺ ₄-N added was 0.03–0.04 in samples 1, 3 and 4, and 0.16 in sample 2. This result indicated that only 3–4% of cumulative N₂O was produced by denitrificiaton at 25°C and 60% WFPS; however, this proportion could be increased immediately after the application of chemical fertilizer and fresh cattle manure. The N₂O emission peaks occurred approximately 350 h after the beginning of incubation for samples 1 and 2, but only 48 h later in samples 3 and 4. Furthermore, the magnitude of the N₂O peaks in samples 3 and 4 was significantly larger than that in samples 1 and 2, despite that the NH⁺ ₄ concentration in sample 2 amounted to approximately 500 µg N g⁻¹ soil, which was nearly double that recorded in samples 1, 3 and 4 because of the application of basal N fertilizer

Table 2 Total N2O and NO emissions from the soils with and without supplemental N addition during the 25-day incubation

Fertilizer	Sample	Total N2O emission (µg N kg^−1 soil)				Total NO emission (µg N kg^−1 soil)				Ratio of NO/N2O (B/A)
		N added (A)	Control	From N added	Ratio of N2O to N added (%)	N added (B)	Control	From N added	Ratio of NO to N added (%)
(^15NH4)2SO4	1	105 ± 16.6 a	6.28 ± 0.73 a	98.7	0.038	8540 ± 268 c	0	8540	3.29	81
	2	226 ± 4.42 b	106 ± 13.0 b	120	0.046	8525 ± 668 c	6167 ± 411	2358	0.91	38
	3	214 ± 14.7 b	12.8 ± 11.1 a	201	0.077	1563 ± 198 a	0	1563	0.64	7
	4	222 ± 6.49 b	2.67 ± 0.22 a	219	0.084	3575 ± 101 b	0	3575	1.38	16
K^15NO3	1	7.18 ± 2.73 a	6.28 ± 0.73 a	0.90	0	0	0	0	0	−
	2	106 ± 8.0 b	106 ± 13.0 b	0	0	6407 ± 718	6167 ± 411	340	0.13	60
	3	6.58 ± 3.86 a	12.8 ± 11.1 a	−6.22	−0.002	0	0	0	0	−
	4	8.74 ± 5.13 a	2.67 ± 0.22 a	6.07	0.002	0	0	0	0	−
Mean ± standard deviation (n = 3). Values within the same column followed by different letters are significantly different at P < 0.05.

and (¹⁵NH₄)₂SO₄. This indicates that N₂O was slowly produced and emitted in samples 1 and 2, whereas it was eruptively released in samples 3 and 4.

Table 2 listed the total N₂O and NO emissions determined using a GC-ECD and a chemiluminescence NO_x analyzer, respectively, from the incubated soils with and without supplemental N addition. Total N₂O emission during the 25-day incubation at 60% WFPS and 25°C was significantly higher in samples 2, 3 and 4 than in sample 1. To exclude the likely influence of basal N fertilizer and mineralized N from soil organic matter on N₂O emission, the results for total N₂O emission (Table 2, Fig. 2a) were compared with those for ¹⁵N₂O emission (Table 1, Figs 2b,3). The pattern of cumulative ¹⁵N₂O emissions during incubation is illustrated in Fig. 3. Cumulative ¹⁵N₂O emissions during the 25-day incubation at 60% WFPS and 25°C in samples 3 and 4 were 190 and 198 µg N kg⁻¹ soil, respectively, which were significantly higher than the 99.3 µg N kg⁻¹ soil recorded in sample 1 and the 108 µg N kg⁻¹ soil recorded in sample 2 (Table 1). In addition, 25–31% of the cumulative ¹⁵N₂O emission was released in the first 180 h of incubation in samples 1 and 2, whereas as much as 94–96% was released for samples 3 and 4. In total, the emission factor of N₂O, the ratio of emitted ¹⁵N₂O-N to applied ¹⁵NH⁺ ₄-N, was 0.072–0.075% in samples 3 and 4, which was nearly twice the amount in samples 1 and 2 (0.038–0.041%).

Nitric oxide emission and the ratio of NO to N₂O

Similar to N₂O, NO emission was found only in NH⁺ ₄-amended soil and was under the detection limit in NO⁻ ₃-amended soil (Fig. 4). For samples 1 and 2, large NO emissions occurred throughout the experimental period until the end of incubation. In contrast, almost all NO was produced and emitted within the first 180 h of incubation in samples 3 and 4. Regression analysis showed that NO emission during the incubation period was significantly correlated with N₂O emission in all soils except sample 2 (Fig. 5).

Total NO emissions in samples 1 and 2 were 8540 and 8525 µg N kg⁻¹ soil, respectively, significantly higher than the emissions recorded in samples 3 and 4 (Table 2). However, NO emission from control sample 2, which was amended without (¹⁵NH₄)₂SO₄, amounted to 6167 µg N kg⁻¹ soil, resulting in the net NO emission from ¹⁵NH⁺ ₄ added to sample 2 to be only 2358 µg N kg⁻¹ soil. The large emission of NO from sample 2 may be attributed to the presence of a large amount of urea that was applied as basal fertilizer.

The ratios of NO to N₂O emitted in samples 1 and 2 were 81 and 38, respectively, and were significantly larger than those emitted in samples 3 and 4 (Table 2). While,

Figure 3  Change in cumulative (15NH4)2SO4-derived N2O emission from the soils during the 25-day incubation after adding supplemental (15NH4)2SO4. Error bars denote the standard deviation (n = 3).

Figure 4  Nitric oxide emission from the soils during the 25-day incubation after adding supplemental (15NH4)2SO4. Error bars denote the standard deviation (n = 3).

Figure 5  Correlation between NO and N2O emission from the soils during the 25-day incubation after adding (15NH4)2SO4.

Figure 6  Changes in (15NH4)2SO4-derived and concentrations in the soils during the 25-day incubation. Error bars denote the standard deviation (n = 3).

regression analysis for samples 1, 3 and 4 showed that total NO emission was negatively, albeit not significantly, correlated with total N₂O emission (R = –0.94, n = 3).

Variation in NH⁺ ₄ and NO⁻ ₃ concentration

After (¹⁵NH₄)₂SO₄ was added to soils, rapid decreases in ¹⁵NH₄-N and increases in ¹⁵NO₃-N concentrations occurred in samples 3 and 4 (Fig. 6) and as much as 81–96% of the ¹⁵NH⁺ ₄ added was nitrified to ¹⁵NO⁻ ₃ within the first 180 h of incubation (Table 1). In contrast, the magnitude of NH⁺ ₄ added decreased more slowly, but was completely consumed, during the 25-day incubation in samples 1 and 2, resulting in only 33–35% of the ¹⁵NH⁺ ₄ added being consumed in the first 180 h.

The changes in ¹⁵NO₃-N concentration during the 25-day incubation could be fitted to two types of models. Samples 1 and 2 followed a zero-order reaction model, whereas samples 3 and 4 followed a first-order reaction model (Table 3). However, changes in ¹⁵NO₃-N concentration in samples 3 and 4 also followed a zero-order model within the first 180 h of incubation, during this period nearly all of the ¹⁵NH₄-N was oxidized

Table 3 Reaction models and best-fit parameters obtained from the rates of nitrification after adding supplemental (15NH4)2SO4 to Andisols and incubating at 25°C and 60% water-filled pore space

Incubation period	Sample	Reaction equation	n	R ^2
25 days	1	NO^− 3−N = 0.456 t – 12.249	8	0.96**
	2	NO^− 3−N = 0.378 t – 7.904	8	0.98**
	3	NO^− 3−N = 69.369 Ln(t) – 175.42	8	0.93**
	4	NO^− 3−N = 67.945 Ln(t)  – 165.31	8	0.94**
First 180 h	3	NO^− 3−N = 1.144 t + 2.318	4	0.98*
	4	NO^− 3−N = 1.276 t + 1.325	4	0.99**
*P < 0.05;
**P < 0.01.

to ¹⁵NO⁻ ₃, revealing that a lack of NH⁺ ₄ in samples 3 and 4 reduced the increase in NO⁻ ₃ concentration. After 180 h the reaction model changed.

DISCUSSION

Effect of basal N fertilizer on nitrification potential and N₂O emission

Many studies examining the environmental factors affecting N₂O and NO emissions from cultivated soils have been undertaken. Soil temperature, moisture regime and fertilizer management are considered to be the major factors influencing emissions of nitrogen oxides (Davidson 1993; Granli and Bockman 1994). Studies have shown that a WFPS of 60% is the optimum moisture for nitrification (Granli and Bockman 1994). In the present study, we designed similar environmental conditions (60% WFPS moisture and 25°C) in the laboratory to compare soils sampled at different times after basal application of N fertilizer, to evaluate how the application of basal N fertilizer influences the nitrification potential, and N₂O and NO emissions from supplemental N fertilizer in an Andisol. In general, NO and N₂O are produced in soils by nitrification and denitrification synchronically; however, the emission ratio of NO/N₂O was relatively constant during nitrification (Davidson 1993). Inubushi et al. (1996) reported that 87–90% of N₂O produced was derived from autotrophic nitrification during the incubation of an Andisol at 60 and 80% water-holding capacity with NH⁺ ₄ added for 32 days at 30°C. Our results showed that N₂O emission in NO⁻ ₃-amended soil was equal to 3–16% of that in NH⁺ ₄-amended soil, even though fresh cattle manure was applied and no NO was released during the 25-day incubation at 60% WFPS and 25°C, indicating that N₂O and NO were predominantly produced by nitrification (Table 2).

Total N₂O emission during the 25-day incubation in samples 1 and 2 was similar, but only half of that in samples 3 and 4. For samples 3 and 4, as much as 94–96% of the cumulative ¹⁵N₂O emission from (¹⁵NH₄)₂SO₄ was released within 180 h of incubation; of course 81–96% of NH⁺ ₄ added was oxidized. In contrast, only 25–31% of the total N₂O emission occurred in the first 180 h in samples 1 and 2 when 33–35% of the NH⁺ ₄ added disappeared. This indicates that nitrifiers in samples 3 and 4 not only responded more rapidly to the NH⁺ ₄ added, but also converted NH⁺ ₄ into N₂O more efficiently. There was a significant temporal variation in nitrification rate constants, which sharply increased from 0.378–0.456 in samples 1 and 2 to 1.144–1.276 in samples 3 and 4 (Table 3), indicating that the application of basal N-amended fertilizer significantly changed the soil nitrification potential (Flowers and O’Callaghan 1983; Wheatley et al. 2001).

Starved ammonia oxidizers exhibit a lag phase of up to 150 h, which would be of considerable significance and probable disadvantage (Batchelor et al. 1997). As competition will be intense for the limited supply of N, organisms need to respond rapidly to gain a competitive advantage, while also investing the minimum energy. Ammonia oxidizers are poor competitors for NH⁺ ₄ compared to heterotrophic bacteria (van Niel et al. 1993). This is because ammonia oxidizers use CO₂ as a sole source of C and the energy for the CO₂ fixation originates from nitrification (Aleem et al. 1965). NH₃ and NO⁻ ₂ are not very effective energy sources and as much as 30 g NH₃ is needed for the build-up of 1 g dry mass of Nitrosomonas sp. (Schlegel 1992). Therefore, the active growth of ammonia oxidizers was very slow (Nishio and Fujimoto 1990; Schmidt 1982; Wheatley et al. 2001). The population of ammonia oxidizers did not change within 13 days, but increased from 8.9 × 10⁶ to 38 × 10⁶ cells g⁻¹ soil after 39 days of NH⁺ ₄ applied to a tomato field (Okano et al. 2004). The calculated generation time for ammonia oxidizers in the field is as long as 372.7 h (Okano et al. 2004). Furthermore, Mendum et al. (1999) and Avrahami et al. (2003) found that there was not a community shift in ammonia oxidizers during 4–6 weeks of incubation, but there was after 16 weeks of incubation, although the ammonia oxidizer population did not exhibit net growth. The application of N fertilizer could gradually stimulate the growth of Nitrosospira clusters 1 and 3 and Nitrosomonas cluster 7, which have high nitrification potentials with high ratios of N₂O to N added and become dominant in lieu of Nitrosospira clusters 2, 4 and 9 (Avrahami et al. 2003; Bruns et al. 1999; Kowalchuk et al. 2000; Webster et al. 2002).

Wheatley et al. (2001) found that there was an apparent maximum level of nitrification potential in a UK clay loam, which may be related to the biomass of nitrifiers, and N addition did not stimulate nitrifier growth in soils that had already attained the maximum nitrification potential. Okano et al. (2004) observed a similar phenomenon in which the population growth of ammonia oxidizers had ceased by the time most of the added NH⁺ ₄ had been depleted in the case of 1.5 mmol L⁻¹ (NH₄)₂SO₄ treatment. In contrast, growth reached a plateau at approximately 6.0 × 10⁷ cells g⁻¹ soil, although there was still a high NH⁺ ₄ concentration remaining in the case of 7.5 mmol L⁻¹ (NH₄)₂SO₄ treatment. Apparently, the stimulation effect of N fertilizer on population growth and nitrification activity of nitrifiers mainly depends on whether the soil attains the maximum nitrification potential. In the present study, it is unlikely that the tested soil reached the maximum nitrification potential before or just after basal application of N fertilizer, probably because of the size of the indigenous nitrifier population. Our study indicated that basal application of N fertilizer was unable to stimulate the population growth of nitrifiers immediately, which resulted in a relatively long response time for nitrification and decreased the ratio of N₂O to N added as basal fertilizer. However, the basal application of N fertilizer did reduce the response time and dramatically increase the ratio of N₂O to N added as supplemental fertilizer at 36 days or more after basal fertilization. This is likely to be because of the increased population of nitrifiers and/or a shift in the community of ammonia oxidizers from a low to high ratio of N₂O to N added.

Effect of basal fertilizer on NO emission

In the present study, NO emission was related significantly to N₂O emission during the incubation period in all soils except sample 2, indicating that N₂O and NO were simultaneously produced (Fig. 5). However, total NO emission during the 25-day incubation was negatively correlated with total N₂O emission (Table 1, R = −0.94, n = 3). In other words, increased nitrification potential in samples 3 and 4 enhanced N₂O emission, but reduced NO emission. This finding is strongly supported by the results of previous studies. Cheng et al. (2004) reported that the ratio of emitted N₂O relative to net nitrification rate in acid soils was lower than that in alkaline soils, but the ratio of emitted NO to net nitrification rate was significantly higher in acid soils. Hou et al. (2000) found that the same tested soil amended with urea released more N₂O but less NO at 70% than at 40% WFPS. Our finding further confirmed that there was a trade-off relationship between total NO and N₂O emissions and the ratio of NO/N₂O decreased with increased nitrification potential in the tested Andisol (Wolf and Russow 2000).

Yamulki et al. (1997) reported that NO, but not N₂O, was emitted from sterile acid soils, indicating that NO could also be produced by chemodenitrification. Gross NO production rate by chemodenitrification was not correlated with the amounts of NH⁺ ₄ and NO⁻ ₃ or with the rate of the overall nitrification, but was correlated with the concentration of HNO₂, which originated from autotrophic microbial oxidation of NH⁺ ₄ to NO⁻ ₂ (Venterea and Rolston 2000). The NO was produced by chemodenitrification through the aqueous decomposition of HNO₂ and reactions of HNO₂ with phenolic and other functional constituents of soil organic matter or with reduced metal cations (van Cleemput and Baert 1976; Nelson 1982; Stevenson 1994). Morrill and Dawson (1967) found that Nitrobacter populations could have experienced slightly greater lag effects than ammonia oxidizers, resulting in NO⁻ ₂ accumulation, which in turn provided the opportunity for abiotic HNO₂ and NO production prior to utilization by oxidative or reductive microbes. In the present study, although no NO⁻ ₂ was measured in soil, we could not exclude the contribution of chemodenitrification to NO production.

In contrast, extracellular NO is an oxidative intermediate in autotrophic oxidation of NH⁺ ₄ to NO⁻ ₂, although this has not been confirmed (Hooper 1982). In addition, NO is not necessarily a precursor for the formation of N₂O under nitrification (Wolf and Russow 2000). Therefore, it is possible that NO just “leaks” from the oxidization process of NH⁺ ₄. There is evidence that Nitrosomonas and Nitrobacter populations preferentially grow on the oppositely charged surfaces in soils, each tending to proliferate on the surfaces to which its primary substrates are absorbed (Underhill and Porsser 1987). The growth rate of nitrite oxidizers was faster than that of ammonia oxidizers after N addition (De Boer et al. 1989), which in turn accelerated the oxidation of NO⁻ ₂ and was also conducive to the oxidation of NH⁺ ₄. Microbial NO⁻ ₂ utilization occurring simultaneously with abiotic HNO₂ reduction could have competed for available NO⁻ ₂ and thereby attenuated abiotic NO production (Venterea and Rolston 2000). Furthermore, an increased oxidation rate of NH⁺ ₄ to NO⁻ ₂ and, in particular, of NO⁻ ₂ to NO⁻ ₃ induced by N addition possibly reduced “leaking” of NO from the atuotrophic oxidation of NH⁺ ₄. The third possibility is that the dominant ammonia oxidizers induced by N addition in samples 3 and 4 differed greatly from those in samples 1 and 2, and may have a high nitrification potential with a high ratio of N₂O produced to N added and a low ratio of NO produced to N added. Further study is needed to evaluate the contribution of chemodenitrification to NO production and to determine whether there is a community shift in ammonia oxidizers among these four soils.

CONCLUSION

An incubation experiment of a cultivated Andisol collected at different times before and after basal fertilization indicated that there was a time lagged induction in N₂O emissions, which was possibly caused by slow population growth of nitrifiers and/or a slow shift in the microbial community in the soil. The change in N turnover in the soil dramatically increased the ratio of N₂O to N added as supplemental fertilizer at 36 days or more after basal fertilization. In addition, slow oxidization of NH⁺ ₄ to NO⁻ ₂ and/or NO⁻ ₂ to NO⁻ ₃ in the soil during the lag-time resulted in relatively large NO emissions. In contrast, accelerated oxidation by increased nitrification potential increased N₂O emission, but reduced NO emission after the lag time, which in turn showed a trade-off relationship between total NO and N₂O emissions.

ACKNOWLEDGMENTS

This study was funded by the Japan Society for the Promotion of Science (JSPS) and partly by the Global Environment Research Fund S2_3a. We also thank the National Basic Research Program of China (2005CB121101) and the National Natural Science Foundation of China (40621001) for their financial support.