1 Introduction

Rice cultivation has been identified as a leading source of anthropogenic nitrous oxide (N2O) emission, since it provides suitable conditions for the microbial N2O-forming processes nitrification and denitrification after, sometimes excessive, inputs of chemical or organic fertilizers (Jacobson 2005; Shcherbak et al. 2014). As a greenhouse gas, N2O is 265 times more potent than carbon dioxide (CO2) over a time horizon of 100 years, and it is also responsible for stratospheric ozone depletion (Li et al. 2011). Agricultural N2O emissions account for more than 60% of global anthropogenic N2O emissions (Pachauri et al. 2014). With a projected increase in global rice consumption of around 45% by 2050 to meet the expected needs of the growing population for food (Rejesus et al. 2012), increasing rice production without increasing the release of N2O emissions, therefore, presents a major and urgent scientific and societal challenge (Fan et al. 2019; Chen et al. 2019).

Various management strategies have been developed to mitigate N2O emissions from rice cultivation, including water management modification (Ly et al. 2013), organic material application (Shin et al. 1996; Knoblauch et al. 2011) and fertilizer additives. Over the past few years, amendment of croplands with biochar, produced from pyrolysis of crop straw, seems to be increasing in popularity owing to its wide availability and extensive advantageous characteristics, including improving nutrient retention and reducing indirect gaseous emissions (Steinbeiss et al. 2009; Sarkhot et al. 2012). Although some meta-analyses have indicated that biochar application could reduce N2O emissions from soils (Cayuela et al. 2014), results are inconsistent among studies (e.g. Fan et al. 2017; He et al. 2018). For instance, some studies report a notable increase in N2O emissions following biochar application to soils (Wells and Baggs 2014), while other studies have found that the addition of biochar decreased N2O emissions (Yanai et al. 2007; Wang et al. 2013; Felber et al. 2014), and the reduction is suggested to be due to enhanced soil aeration and pH, reduced availability of inorganic N, and release of polycyclic aromatic hydrocarbons (PAHs). Previous studies have investigated the influence of biochar addition on nitrifier or denitrifier populations, but the results are still being debated. Thus, the underlying possible mechanisms for the impacts of biochar on N2O emissions require further investigation, both theoretically and experimentally (Spokas et al. 2012; Yu et al. 2013).

One of the most popular agricultural nitrification inhibitors, 3,4-dimethylpyrazole phosphate (DMPP), has been proven to be useful and readily applicable for decreasing N2O emissions directly, by suppressing the growth of ammonia-oxidizing bacteria (AOB) as well as possibly ammonia-oxidizing archaea (AOA), or indirectly by decreasing denitrifier activity and growth through reducing the presence of nitrate (NO3−) (Shi et al. 2017). Chen et al. (2019) conducted an incubation study and suggested that the combination of biochar and DMPP can effectively reduce N2O emissions and the dynamic changes of AOB and nosZ-N2O reduction caused by the applied biochar and/or DMPP played an important role in controlling emissions. However, to date, only limited experimental evidence exists describing the integrated functions of biochar and DMPP on N2O emissions, and the relevant mechanisms have not been verified.

Therefore, a year-round observation of N2O emissions from paddy soil was conducted in Northeast China. Specifically, this work (1) examined and evaluated the effects of application of biochar and/or DMPP on N2O emissions from paddy soil and (2) systematically explored the soil microbial functional genes in the processes of soil nitrification and denitrification by regulating the effects of biochar and/or DMPP on N2O emissions. Such information is expected to contribute to progressing the development of appropriate management practices to mitigate the climatic impacts from paddy fields.

2 Materials and methods

2.1 Experimental site

The field experiment was performed at Fangzheng Rice Science and Technology Experimental Station (45° 85′ N, 128° 82′ E) in Heilongjiang province, which has a long history of agriculture in northeastern China. The mean annual air temperature and precipitation are approximately 4.5 °C and 530 mm. The soil is classified as a Udoll (World Reference Base). The rice was irrigated with a layer of water 3 to 5 cm deep from 1 week before planting to 1 month before harvesting, which took place around 7th October.

2.2 Experimental design and field management

Five treatments were performed in triplicate in a randomized design, with an area of 7 m × 5 m (35 m2) for each plot: CK (no fertilizer), synthetic fertilizers (NPK only), NPK with biochar, NPK with DMPP, and the combination of NPK, biochar, and DMPP. Fertilizers applied were urea (46% N), triple superphosphate (19% P), and KCl (50% K). Before transplanting, mixed fertilizer (150 kg N ha−1; 75 kg P2O ha−1; and 75 kg K2O ha−1) was applied as a basal fertilizer. The inhibitor DMPP was applied at a rate of 2% of the applied urea nitrogen and the biochar was applied at 24 t ha−1. They were homogenously combined with the fertilizer and then applied into the soil with the soil depth ranging from 0 to 20 cm. Fertilization was carried out 1 day before rice transplanting. Rice seedlings (cv. Meifeng 9) were transplanted into each plot evenly with 3–5 rice seeds per hole at spacings of 15 cm × 30 cm (10 days after wetting of the soil). Field management was consistent with local practices. The top 20 cm of the soil had the following properties: pH of 6.11, soil organic carbon (SOC) of 25.8 g kg−1, total nitrogen (TN) of 2.04 g kg−1, soil available N (AN) of 24.81 mg kg−1, available P (AP) of 23.06 mg kg−1, and available K (AK) of 113.02 mg kg−1.

2.3 Biochar production and characterization

The biochar was produced through pyrolysis of rice straw at 600 °C utilizing a slow-pyrolysis process. The C, H, and N contents of the biochar were measured using a CN analyzer (Leco TruSpec CN, USA). The pH of the biochar was measured at a ratio of 1:2.5 (w/v) biochar-water mixture with a combination electrode. The biochar used in this study had the following properties: pH of 9.40, total nitrogen (TN) of 8.06 g kg−1, and total carbon (TC) of 567.0 g kg−1.

2.4 Gas sampling and auxiliary measurements

A static chamber was utilized to evaluate the in situ fluxes of N2O from the paddy field (Wu et al. 2019). Static chambers were constructed of transparent plexiglass and consisted of a chamber base (10 cm in height) and a chamber body (30 cm in diameter and 50 cm in height). The base was inserted into the soil and the body was attached to the base at the time of sampling. A small fan was fixed inside the top of the chamber to homogenize the headspace gases. Gas samples were collected every 2–3 days during the first 2 months, then sampling frequency was reduced to once per week for approximately 16 weeks. During midseason aeration (50 days after planting, 20th June), and the following re-flooding period, gas samples were collected daily. All the measurements were performed at a fixed time between 09:30 and 11:00 am every sampling day. Three samples were taken from the headspace of each chamber, at 0, 45 and 90 min after closing the chamber, using 50 ml syringes. The samples were transferred into tinfoil gas-collecting bags (300 ml) and taken to the laboratory for analysis of N2O concentration using a gas chromatograph (Agilent 7890B, Delaware, USA). Soil temperature was measured continuously in the chamber with a temperature probe (10 cm soil depth), while the air temperature inside the chamber was measured using a thermocouple 20 cm from the top.

The hourly N2O fluxes were obtained using the following formula:

$$ \mathrm{F}=\uprho \times \mathrm{h}\times \triangle \mathrm{c}/\triangle \mathrm{t}\times 273/\left(273+\mathrm{T}\right)\times 60 $$

In which, F represents N2O flux (μg m−2 h−1); ρ is the density of N2O gas under standard conditions (1.25 kg m−3); △c/△t is the change in N2O concentration in the headspace gas (10−9 V V−1 min−1); T is the temperature in the chamber enclosure (°C); and h is the height of the chamber (cm).

A linear interpolation method was employed to calculate the cumulative N2O emission from each treatment over the trial period. Precipitation and air temperature data were obtained from measuring equipment at the experiment site.

2.5 Soil sampling and measurement

Soil samples were collected from each plot (0–20 cm depths) at five stages of rice plant growth (seedling, tillering, elongation, heading, and maturity). Three individual samples were taken from three random points within the plot and then mixed thoroughly. The concentrations of NH4+-N and NO3−-N were measured by sieving fresh soil (< 2 mm) and extracting with 2 M KCl with a KCl-to-soil ratio of 10:1. The MgO-Devarda’s alloy distillation method was used to analyze the sample extracts. Soil total N was measured by Kjeldahl digestion while available K was extracted with ammonium acetate and measured using flame photometry. Soil available phosphorus (Olsen P) was measured calorimetrically after extraction with 0.5 M NaHCO3. Soil organic C was determined using the Walkley-Black wet oxidation method (Allen et al. 1987). Soil available N was analyzed through quantification of alkali-hydrolysable N in a Conway diffusion unit with Devarda’s alloy in the outer chamber and boric acid-indicator solution in the inner chamber (Shen et al. 2004).

2.6 Soil microbial analysis

The extraction of total DNA contents was carried out using an Omega DNA extraction kit for soil (E.Z.N.A. Soil DNA Kit; Omega Bio-Tek Inc., GA, USA) following the instructions of the manufacturer. The AOB amoA, AOA amoA, nirK, and nirS genes in the soil samples were measured by quantitative qPCR (Roche Light Cycler® 96, Switzerland) with the primers shown in Table 1. 10 ml of 2 × SYBR green mix (Takara, Dalian, China), 0.4 ml of each primer (10 mmol), 2 μl of template DNA, and ddH2O, formed the total volume of 20 ml. A standard curve was applied for the verification of the nitrifier genes, and the denitrifier functional genes were obtained using 10-fold serial dilutions of known copy amounts of the plasmid DNA. For these genes, the PCR amplification efficiency was between 91 and 110%, and R2 values were greater than 0.98. All qPCR was done by performing melt-curve analysis and agarose gel electrophoresis to amplify the specific products.

Table 1 Real-time PCR primers and conditions used for amplification of nitrifier and denitrifier genes
Full size table

2.7 Statistical analysis

A one-way analysis of variance (ANOVA) was performed on all data. Duncan post hoc tests were used to assess the statistical significance of the biochar and DMPP effects on soil properties, N2O emission, nitrifiers, and denitrifiers using the SPSS software package for Windows (Version 16.0, SPSS Inc., Chicago, IL, USA). The Tukey multiple-comparison test was performed to further evaluate the significance (p < 0.05) of the detected effects within any treatments.

Structural equation modeling (SEM) was used to test how relationships among soil properties, microbial abundance and N2O production respond to the different treatments. The analysis was conducted using AMOS 17.0 (Amos, Development Corporation, Meadville, PA, USA). The hypothetical relationships among the variables in the models were constructed based on results of correlation analyses (Table 2). Prior to the SEM analyses, the distributions of all involved variables were examined for normality. Several tests were used to assess model fit, i.e., the X2 test, comparative fit index (CFI), adjusted goodness-of-fit index (AGFI), and root mean square error of approximation (RMSEM).

Table 2 Correlation coefficients of soil physiochemical properties with N2O emission fluxes and abundance of the ammonia-oxidizing archaea gene (amoA-AOA), ammonia-oxidizing bacteria gene (amoA-AOB), nitrite reductase (nirS and nirK) gene, and N2O reductase (nosZ) gene
Full size table

3 Results

3.1 Climatic variables and N2O emissions

The average daily soil temperature ranged from 15 to 25 °C (10 cm depth), and the precipitation was 410 mm during the experimental period (Fig. 1). In general, N2O emissions were primarily induced by fertilization, reaching a peak on the first day after fertilization, then decreasing sharply and returning to background levels 80 days after rice transplanting. In addition, substantial N2O emissions were detected at the mid-season drainage stage. Throughout most of the growing season, N2O fluxes from the DMPP and biochar treatments had similar patterns to the fertilizer-only treatment (Fig. 2). Among the fertilized treatments, daily mean N2O fluxes followed the order: NPK > NPK + biochar > NPK + DMPP + biochar > NPK + DMPP. The N2O emissions were always lower in the DMPP and/or biochar treatment, leading to 39.6%, 18.9%, and 32.1% lower cumulative net N2O emissions for the NPK + DMPP, NPK + DMPP + biochar and NPK + biochar treatments, respectively, compared with fertilizer-only treatment (Fig. 2).

Fig. 1
figure 1

Daily rainfall, soil (20 cm), and air temperature during field experiments in 2018

Full size image
Fig. 2
figure 2

Effects of different treatments on N2O fluxes and cumulative N2O emissions in paddy fields. Error bars represent standard error of the mean (n = 3). Different letters indicate significant differences (p < 0.05) among treatments. Arrows indicate fertilizer application. F-D-F-M, flooding–mid-season drainage–reflooding–moist irrigation

Full size image

3.2 Nitrifier and denitrifier gene abundance

In general, the trends in the variation of AOA amoA and AOB amoA gene abundances were different during the different rice growing stages (Fig. 3). The AOA amoA abundance increased gradually and reached a peak at the elongation stage, while AOB amoA abundance peaked at the seedling stage and then decreased gradually. The detected abundance of AOB amoA, ranging from 4.7 × 108 to 11.3 × 108, was greater than that of AOA amoA (3.1 × 107 to 9.37 × 107), and both were much higher under the fertilized treatments compared with the control. Compared to NPK, the abundance of AOA was enhanced by biochar and/or DMPP during all growth stages, especially in the DMPP treatment at the elongation stage (p < 0.05). However, compared with the NPK treatment, the application of DMPP decreased AOB amoA abundance by 18–27%. The applied biochar increased AOB amoA abundance by 7–15% over all the rice growth stages and the applied biochar with DMPP decreased the abundance of AOB, but the reduction was lower than that of the DMPP-only treatment.

Fig. 3
figure 3

Effects of different treatments on copy numbers of ammonia-oxidizing bacteria (AOB) amoA gene (a), ammonia oxidizing archaea (AOA) amoA gene (b), nirS gene (c), nirK gene (d), and nosZ gene (e) in paddy soil at different stages of rice growth. Error bars represent standard error of the mean (n = 3). Columns with the same letter are not significantly different (p ≥ 0.05) (Note different y-axis unit magnitudes for (a) and (b))

Full size image

Overall, the nitrite reductase genes (nirK and nirS) in the soil showed similar trends over time, increasing initially after fertilizer application and then decreasing over the rice growth period. The abundances of nirK were greater than those of nirS. Compared with the control, the application of fertilizers significantly increased nirK and nirS growth. The application of biochar and/or DMPP decreased the abundance of nirS at the seedling, tillering, and maturity stages (p < 0.05). The abundance of nirK was reduced by DMPP and the combination of DMPP and biochar during the whole growth period but was increased by biochar only at the elongation and heading stages (p < 0.05).

As shown in Fig. 3, the abundance of the N2O reductase gene (nosZ) increased gradually after rice transplanting (seedling stage), and the peak appeared at the elongation stage for DMPP treatments. Compared with the control, the abundance of the nosZ gene increased slightly in the NPK treatment, but the differences were not significant at most growth stages (except the maturity stage). However, the application of biochar and/or DMPP increased the abundance of nosZ genes significantly during the whole growth period compared with the fertilizer-only treatment (p < 0.05), with the highest values occurring in the biochar and DMPP treatment.

3.3 Relationships between N2O flux and soil properties or microbial dynamics

The correlations between N2O emission, soil properties, and N-cycling genes, were calculated to identify the key factors controlling N2O emission in the paddy soil. The N2O emission flux was positively correlated with soil ammonium content, soil nitrate content, soil pH, and total N content (p < 0.01). Among the N2O emission-related genes, N2O flux exhibited a significant positively correlation with AOA (r = 0.378, p < 0.01), but a negative correlation with AOB (r = − 0.384, p < 0.01). Among the denitrifiers, N2O flux was negatively correlated with nosZ (r = − 0.339, p < 0.01). Furthermore, based on the SEM analyses, our data also showed that NH4+-N, NO3--N, pH, AOA, AOB, nirS, and nosZ, rather than other soil properties, exerts a direct and dominant effect on N2O, explaining 79% of variation for N2O emissions (Fig. 4).

Fig. 4
figure 4

The structural equation model (SEM) showing the hypothesized relationships between soil properties, microbial abundance, and N2O production. The arrow thickness indicates the strength of the standardized path coefficients. The solid line arrows represent positive path coefficients, and the dashed lines indicate negative path coefficients. Numbers on the arrow indicate significant standardized path coefficients (*p < 0.05; **p < 0.01), proportional to the arrow width. R2 indicates the variance of endogenous variable explained by the model. Goodness-of-fit statistics are shown underneath the modeling frames. NH4+, ammonium; NO3–, nitrate; CFI, comparative fit index; AGFI, adjusted goodness-of-fit index; RMSEM, root mean square error of approximation

Full size image

4 Discussion

4.1 N2O emissions from paddy soil

Consistent with previous studies (Maljanen et al. 2003; Meng et al. 2005; Baggs et al. 2010), N2O fluxes in all the treatments increased on the first day after fertilization and then declined quickly. The relatively high N availability from fertilizer application and the increase in soil water content from irrigation could have directly led to the enhancement of nitrification and/or denitrification and the increase in N2O emission (Gregorutti and Caviglia 2017). This is also supported by the fact that N2O flux showed positive correlations with NH4+-N and NO3−-N in the paddy soils (Table 2; Fig. 4). The abundances of AOA and AOB were greatly enhanced by 5.5–14.2% and 29.9–37.6%, respectively, after the application of fertilizer compared to the control treatment (Fig. 3), which confirmed that fertilizer can promote the activity of ammonia-oxidizing microbes by providing substrate for them (Wu et al. 2011), and thus consequently stimulating N2O emission (Fig. 3). Although a high N rate was applied, it is likely that most of the applied N could be taken up by the plants before being denitrified into N2O. This could explain why N2O emission was observed only after the fertilization event and lasted for only a few days. Notably, over the whole growing season, the N2O emissions were low during the flooding period, and an emission peak appeared at the initial period of the drainage stage, quickly declining and returning to the background by the start of the midseason aeration. These data suggest that soil water status played an important role in N2O losses. Part of the reason for this peak in N2O fluxes may be that the drainage process made the soil system aerobic, thus promoting nitrification and N2O production (Li et al. 2009). In addition, N2O accumulated in the deep soil could have been released along soil cracks during this period (Cai and Laughlin 2001).

Some studies have shown that denitrification of NO3−-N in anaerobic soil could be the reason for N2O losses after irrigation (Aulakh and Singh 1997). However, no N2O flux peak was found during the re-flooding period in this work. Similarly, Li et al. (2009) and Ma et al. (2009) found negligible N2O fluxes and NO3−-N in the soil during the flooding period and suggested that this could probably be attributed to the pressure of standing water preventing N2O from being released into the atmosphere and the dissolved N2O then being fully denitrified to N2 within the soil (Granli and Bøckman 1994). In the present study, the cumulative rates of N2O emission were between 0.33 and 0.53 kg N ha−1 and were within the range of 0.33 to 4.42, reported for rice fields around the world (Akiyama et al. 2005).

4.2 Effects of biochar and/or DMPP on N2O emissions

In the present study, N2O emissions were always lower with biochar than without, indicating that the addition of biochar effectively suppressed N2O losses. Previous work has also found that biochar addition could significantly increase the yield of rice and reduce N2O emissions from paddy soils (Zhang et al. 2012; Cayuela et al. 2014). In general, biochar easily combines with minerals in soil to form organic-inorganic complexes, which are difficult for soil microorganisms to utilize, and biochar can also increase soil pH, improve soil aeration, and reduce soil bulk density, inhibiting denitrification (Zhang et al. 2017). However, some researchers reported that biochar had no significant effect on N2O flux and cumulative emission during the long-term submergence of rice during the growing season. This has been ascribed to a range of factors, such as the biochar source, application rate and application method, soil type and water management, which could have affected the inhibitive efficiency of the biochar (Li et al. 2015; Lin et al. 2017). Results from the current study also highlighted that the addition of DMPP could significantly reduce N2O emissions, by 39.6% in the paddy field. It has been reported previously that nitrification inhibitors can fundamentally affect the nitrogen transformation processes in the soil, reducing the supply of substrate for nitrification and denitrification and consequently decreasing the emission of N2O (Benckiser et al. 2013). In addition, the application of biochar and DMPP showed synergistic effects on N2O emission, producing a 16% further reduction relative to the biochar-only treatment, indicating that applying the biochar with DMPP could improve the benefit of biochar in decreasing N2O losses (Fig. 2). Similarly, He et al. (2018) have also confirmed that the addition of biochar with inhibitors could more effectively reduce N2O emissions compared to biochar only. It is interesting to note that the inhibition efficiency of biochar combined with DMPP on N2O emissions was lower than that of the DMPP-only treatment (Fig. 2), which indicated that biochar may reduce the mitigating effect of DMPP on N2O emissions in the paddy field. This is in agreement with previous short-term laboratory incubation studies which have shown that the sorption of DMPP by biochar reduced its effectiveness for nitrification inhibition (Chen et al. 2019).

4.3 Underlying microbial mechanisms of biochar and/or DMPP decreasing N2O emissions

It has been suggested that ammonia oxidizers play a vital role in the biochemical cycling of nutrients and N2O emissions from flooded anaerobic soil (Paranychianakis et al. 2013; Zheng et al. 2016). Both AOA amoA and AOB amoA genes were highly abundant, with AOB numerically dominant over AOA, in this study, which is similar to findings from studies carried out in other paddy soils (Cai and Laughlin 2001). The addition of biochar significantly increased both bacterial and archaeal nitrifiers (Fig. 3). Levicnik-Hofferle et al. (2012) suggested that the abundance of AOA could be enhanced by ammonia produced from biochar or organic matter mineralization. Lin et al. (2017) studied the effects of wheat straw-derived biochar on N2O emissions and showed that adding biochar to paddy soils increased the abundance of ammonia oxidizers. However, the increase in the abundance of AOA and AOB caused by biochar addition is contrary to its potential for reducing N2O emission by nitrification inhibition. In view of the higher microbial activity in the biochar treatment (Fig. 3), the potential mechanisms for the mitigation effect may be attributable to its aeration regulation and promotion of complete denitrification (Harter et al. 2014; Padhye 2017).

In the current study, AOB and AOA showed opposite responses to DMPP addition, with decreased AOB abundance and increased AOA abundance, and N2O emissions from those soils were relatively low. These results further confirm that DMPP inhibits nitrification by reducing AOB abundance, which is in line with previous observations (Kleineidam et al. 2011; Benckiser et al. 2013; Shi et al. 2017; Fan et al. 2019). The contrasting responses of AOB and AOA to DMPP may be related to pH-associated niche differentiation, their differing enzyme systems, and the mixotrophic growth of AOA (Jia and Conard 2009). It has been reported that the bacteria might be more sensitive to different inhibitors than archaea because of their essential metabolic and cellular differences (Shen et al. 2013). It is interesting to note that DMPP could inhibit AOB and N2O emission, but biochar addition reduced this effect to some extent in the present study (Figs. 1 and 3). According to Hink et al. (2018), in the process of soil aerobic ammoxidation, the amount of N2O produced by AOA is lower than that by AOB. Therefore, it could be inferred that the DMPP and DMPP plus biochar treatments may have transferred N2O generation from an AOB-dominant process to the low N2O-yielding AOA equivalent (Fig. 2), which is evidenced by the positive relationship between N2O emissions and AOA abundance, but the negative relationship between N2O and AOB in this study (r = − 0.384, p < 0.05).

It has been shown that under anaerobic conditions, denitrification is the predominant source of N2O production. As with the inhibition of AOB, the application of biochar and/or DMPP had a significant impact on nirK and nirS-type denitrifiers in our study, suggesting that the two additives also affected denitrification to influence N2O emissions (Cayuela et al. 2014; Fan et al. 2019). Di et al. (2009) reported that nitrification inhibitor could reduce nirK gene abundance by affecting AOB populations, which also bear the nirK gene. For the nirS and the nirK denitrifiers, their growth was reduced by DMPP, but nirK was increased by biochar at the elongation and heading stages. Thus, as reported by Hallin et al. (2018), the different responses of these denitrifiers to biochar and/or DMPP treatment may be linked to their niche differentiation.

In the current study, it was found that the application of biochar or DMPP enhanced the expression of the nosZ gene, and the increase in nosZ was also negatively correlated with N2O emissions (r = − 0.339, p < 0.01). Therefore, it appears that biochar and DMPP have an effect on denitrification, inducing the transient expression of nosZ, and thus stimulating the complete reduction of N2O to N2. Consistent with this result, Hatch et al. (2005) observed that, compared with soil without DMPP treatment, N2O production decreased and N2 increased during anaerobic soil incubation with DMPP. These findings support the hypothesis of Anderson et al. (2011) who suggest that the reduction of N2O emission resulting from biochar application may be due to enhancement of the growth and activity of microorganisms capable of complete denitrification. As such, stimulation of nosZ gene production appears to be the most important factor in reducing N2O emissions (Conthe et al. 2018). Moreover, the DMPP with biochar treatment had a greater effect on the abundance of nosZ than the biochar or DMPP-only treatments (Fig. 3), indicating that the combined use of DMPP and biochar may promote full denitrification to N2, and further reduce N2O emissions.

5 Conclusion

The data obtained from this study showed that in rice paddy fields, application of fertilizer significantly increases N2O emissions, mainly during the midseason aeration period, which could be due to the relatively high N availability and dramatic changes in soil water content. N2O emissions were significantly reduced by DMPP and/or biochar amendment. The combination of biochar with DMPP reduced N2O emission to a greater extent than biochar alone but not as much as DMPP alone, and this was likely to have been due to the sorption of DMPP by biochar when applied together. To some extent, the different impacts of biochar and DMPP on N2O emission were related to the different response patterns of N-cycling genes. Overall, application of DMPP to paddy fields is an effective strategy to mitigate N2O emissions by regulating the abundance of ammonia oxidizers and N2O reducers.