Introduction

Methane (CH4) is a key contributor to global warming. Rice fields are a major source of CH4, accounting for approximately 5–19 % of global CH4 emissions (IPCC 2007). CH4 is produced by methanogenic archaea as a final metabolic product of the anaerobic degradation of organic matters (Liesack et al. 2000). All known methanogens belong to tone of seven orders within the phylum Euryarchaeota: Methanomicrobiales, Methanobacteriales, Methanosarcinales, Methanocellales, Methanococcales, Methanopyrales (Sakai et al. 2009), and the recently discovered Methanomassiliicoccales (Borrel et al. 2014). Methyl coenzyme-M reductase (MCR), the key enzyme of methanogenesis (Ermler et al. 1997), is a specific marker for methanogens. This enzyme catalyzes the reduction of methyl-coenzyme M, which leads to the release of methane (Ellermann et al. 1988). mcrA gene is a functional marker gene encoding the MCR. This gene is highly conserved in all methanogens (Lehmacher and Klenk 1994; Springer et al. 1995), which renders it suitable for application in molecular ecology studies, and has been widely used for investigation of methanogenic communities in rice paddy soils (Ma et al. 2012; Lu et al. 2015).

Returning rice straw to rice field soil is a common practice to maintain the soil fertility in Asian agriculture (Tirol-Padre et al. 2005). However, incorporation of straw also increases CH4 production and emissions (Denier Van Der Gon and Neue 1995; Zou et al. 2005). Several incubation studies have shown a highly dynamic structure of methanogenic archaeal communities during straw decomposition, with members responding differently to organic residue type, incubation temperature, and duration of incubation (Conrad and Klose 2006; Peng et al. 2008; Conrad et al. 2012; Lu et al. 2015). It is thought that the concentrations of acetate and H2 are the key factors that control methanogenic community composition during the degradation of rice straw (Conrad 2007).

The intensive use of inorganic nitrogen (N) fertilizer in rice fields has been dramatically increased to meet the increasing food demands of a continuingly growing population (Zhu and Chen 2002). In rice field soil, most N is applied in inorganic form, such as urea. While applied for utilization by plants, the oxidation of ammonia during nitrification in oxic surface soils results in the production of nitrate which may have consequences for methane production, as nitrate in flooded rice fields has been widely observed to have an inhibitory effect on CH4 production (Klüber and Conrad 1998a, 1998b; Lu et al. 2000; Yuan and Lu 2009). Two main mechanisms have been reported to inhibit CH4 production from the presence of nitrate. Firstly, denitrification intermediates such as nitrite (NO2 −), nitric oxide (NO), and nitrous oxide (N2O) from nitrate reduction can inhibit methanogenic microorganisms (Clarens et al. 1998; Roy and Conrad 1999). Secondly, CH4 production can be reduced when nitrate reducers outcompete methanogens for common substrates. During the phase of reduction of NO3 −, NO2 −, and N2O, the partial pressure of H2 can decrease below concentrations that are required to support CH4 production (Klüber and Conrad 1998b; Roy and Conrad 1999). The production of organic matter decomposition also serves as an important electron source for iron(III) and sulfate (SO4 2−) reduction in anaerobic respiration processes, and CH4 production is suppressed by the reduction of these electron acceptors (Yao et al. 1999). It has been shown that nitrate amendment resulted in suppression of CH4 production in anoxic incubation of excised rice roots (Scheid et al. 2003). However, there are few studies examining the effects of combined rice straw and nitrate amendments on specific methanogenic populations.

Therefore, to determine which populations were influenced under different amendments, rice field soil was anaerobically incubated at a moderate temperature (30 °C) in the presence of rice straw alone, nitrate alone, and both straw and nitrate. CH4 production, soil physicochemical properties, and the DNA and complementary DNA (cDNA)-based methanogenic communities were analyzed with incubation time. The specific aims were to investigate the influence of rice straw and/or nitrate addition on (i) CH4 production and (ii) the structure and dynamics of the active methanogenic communities.

Material and methods

Preparation and incubation of soil slurries

Rice field soil samples were taken from the plow layer (0–20 cm) of rice fields (28° 12′ 22.2″ N, 116° 56′ 02.2″ E) in Jiangxi province of China. The soil type was silt loam and pH was 6.42. Other characteristics and pre-treatment of soil were described in a previous study (Bao et al. 2014). Soil sample was air-dried and passed through a 2-mm sieve. Rice straw were dried for 48 h at 60 °C and cut into <1-mm-long pieces. Glass bottles (100 ml) were filled with 15 g of dry soil, 0.15 g of dry straw and 21.5 ml of autoclaved and degassed water and shaken manually to homogenize the soil slurries. Then, the bottles were closed with sterilized butyl stoppers and crimp seals and flushed with nitrogen gas three times. One milliliter of KNO3 solution (200 mg kg−1 N dry soil) was added to generate the final concentration of about 10 mM nitrate in soil slurries (final ratio of water to soil was 1.5:1). All bottles were incubated at 30 °C in the dark for 40 days without shaking. Four treatments were established, i.e., control (no straw or nitrate) (Ctrl), rice straw alone (S), nitrate alone (N), and rice straw and nitrate added together (S + N). For chemical and microbial analysis at each time point, three replicate bottles were established per treatment, and one set of replicates (n = 3) was used to track the partial pressures of CH4, CO2, H2, and N2O that were produced during the incubation. Over the first 15 days of incubation, measurements were carried out on each day, then measurements were taken every 2 to 7 days.

Chemical analyses

CH4 and CO2 were analyzed using a gas chromatograph (Shanghai Precision and Scientific Instrument, China) equipped with a flame ionization detector, a catalytic methanizer for CO2 (Chrompack, nickat replacement reator) and a 80-cm-long Propack QS 50/100 mesh column operated at 50 °C, with N2 used as the carrier gas. A reductive gas detector (RGD2) (Trace Analytical, Menlo Park, CA, USA) was used to determinate low H2 concentrations (<10 Pa) with a molecular sieve column and synthetic air (80 % N2 and 20 % O2) as the carrier gas. Gas samples were detected with a thermal conductivity detector with a molecular sieve column and N2 as the carrier gas when H2 partial pressure was >10 Pa. N2O was analyzed using an Agilent 7980A gas chromatograph (Agilent Technologies) equipped with an ECD. Concentrations of Fe(II), NO3 −, NO2 −, SO4 2−, and acetate in soil solutions were determined as described previously (Bao et al. 2014).

Nucleic acid extraction and purification

Soil samples were taken for nucleic acid extraction on days 0, 5, 11, and 23 according to the patterns of CH4 production in all treatments. Total DNA and RNA were co-extracted and purified using a protocol described previously (Noll et al. 2005; Ma et al. 2012). For the analysis of DNA, no further purification steps were carried out and nucleic acids were stored at −20 °C. For the analysis of rRNA, DNA was removed by DNase treatment before purification using the RNeasy® Mini Kit (Qiagen) as described previously (Xu et al. 2012). The quality of DNA and rRNA extracts was quantified by UV-vis spectrophotometer (ND-1000, NanoDrop, USA) and checked by 1 % agarose gel electrophoresis.

cDNA synthesis

rRNA was transcribed to cDNA using the PrimeScript® first-strand cDNA Synthesis Kit (Takara) according to Xu et al. (2012). To verify the absence of DNA, a control reaction was performed with water instead of reverse transcriptase. cDNA was stored at −20 °C until subsequent analysis.

Terminal restriction fragment length polymorphism analysis

Terminal restriction fragment length polymorphism (T-RFLP) analyses were used to determine the composition of methanogenic communities. mcrA genes and transcripts were amplified by PCR using the same primers used for cloning with the forward primer additionally labeled with 6-carboxyfluorescein. Purified PCR products of mcrA genes and transcripts were digested at 37 °C for 3 h with Sau96I (Fermentas) (Lueders et al. 2001). Digestion products were size separated using a 3730xl Genetic Analyzer (Applied Biosystems). The relative abundance of each terminal restriction fragment (T-RF) was calculated as described previously (Xiao et al. 2013).

Quantitative analysis of mcrA genes and transcripts

The abundance of mcrA genes and their transcripts was determined using the mlas/mcrA-rev primer pair (Steinberg and Regan 2009). Reactions were performed on an iCycler iQ™ Thermocyler (Bio-Rad, USA) using a previously described method (Xu et al. 2012; Ma et al. 2012). For generating standard curves, 2.49 × 102 to 2.49 × 108 copies per reaction were used.

PCR, cloning, sequencing, and phylogenetic analysis

PCR reactions followed the protocols as described previously (Lueders et al. 2001). The primers were MCRf/MCRr. DNA and cDNA were used as PCR templates for mcrA gene and transcript analyses, respectively. According to the results from T-RFLP, the community structures of methanogens on day 11 showed great differences among almost all treatments compared with other sampling time points based on both mcrA gene and its transcript analysis. Thus, clone libraries were constructed from mcrA gene and transcript amplicons derived from N and S + N treatments sampled on day 11 to allow assigning T-RFs to specific sequence groups using a pGEM®-T Easy Vector kit (Promega) according to the manufacturer’s instruction.

Fifty-three and 110 randomly selected positive clones were sequenced from mcrA gene and transcript, respectively. The clones with correct insert size were sequenced in SinoGenoMax Co., Ltd. (Beijing, China). Clones with >95 % sequences similarity were considered to be the same operational taxonomic unit (OTU). We used the representative sequence from each OTU for phylogenetic analysis. The mcrA sequences were aligned with the related sequences using the Clustal X program. Neighbor-joining phylogenetic trees were constructed from the aligned sequences with MEGA 4.1 (Tamura et al. 2007) by using p-distance. Bootstrapping (1000 replicate reconstructions) was used to estimate the reliability of phylogenetic reconstructions.

The nucleotide sequences of mcrA gene and mcrA transcript fragments determined in this study were submitted in GenBank database under the following accession numbers: mcrA gene: KU052990–KU053010 and mcrA transcript: KU052963–KU052989.

Statistical analysis

Redundancy analysis (RDA) of T-RFLP profiles was performed to summarize the variations in methanogenic communities that could potentially be explained by the measured variables (treatments, sampling time, and all physicochemical parameters) using CANOCO 4.5 software (Microcomputer Power, Ithaca, NY, USA). RDA was also performed to analyze the variations of CH4 production and all variables (variations of methanogens (including the relative abundance of T-RFs, the copy numbers of mcrA genes or transcripts), treatments, sampling time, and all physicochemical parameters), with significance determined by Monte Carlo permutation test (P < 0.05). RDA was chosen according to the results of detrended correspondence analysis (DCA).

Results

Methane production and soil chemistry

Straw addition increased CH4 production. CH4 production in the S treatment was higher than in the S + N treatment during incubation. CH4 production started with longer lag phases in both N and Ctrl treatments, and CH4 production in the N treatment was slightly lower than in the Ctrl during incubation. The CH4 partial pressures at the end of the incubation reached 2.54, 17.44, 1.05, and 11.52 kPa in the Ctrl, S, N, and S + N treatments, respectively (Fig. 1a).

Fig. 1
figure 1

The time course of CH4 (a), CO2 (b), H2 (c), acetate (d), NO3 − (e), NO2 − (f), N2O (g), Fe(II) (h), and SO4 2− (i) in anaerobic incubation with treatments of Ctrl, S, N, and S + N, respectively. Data are mean ± SE (n = 3). Ctrl means control, S means rice straw, N means nitrate, and S + N means straw in combination with nitrate

Full size image

CO2 production occurred from the beginning of the incubation, and straw addition increased CO2 production. There were no differences of CO2 production not only between the S and S + N treatments (Fig. 1b), but also between the N and Ctrl treatments. H2 production increased in the S treatment with two distinct peaks in H2 concentrations. A lower concentration of H2 with a lag phase was observed in the S + N treatment. The obvious accumulation of H2 in the S and S + N treatments occurred within about 10 days. No apparent H2 production was found in the N treatment. H2 accumulated greatly in the Ctrl treatment within 1 week of incubation (Fig. 1c). Straw addition significantly increased acetate production. Two peaks of acetate accumulation were observed in the S + N treatment, with obvious consumption taking longer with the S treatment. The obvious consumption of acetate in the N treatment was shown within 3 days of incubation. No acetate accumulated in the Ctrl treatment (Fig. 1d).

NO3 − increased after 1 day of incubation but decreased immediately in both N and S + N treatments. NO3 − reduced within 3 days of incubation in the S + N treatment, but another high peak of NO3 − appeared on the fifth day in the N treatment, and the reduction of them was finished within 6 days of incubation. NO2 − production in the S + N treatment was greater than in the N treatment. Reduction of NO2 − was finished within 3 and 4 days of incubation in the S + N and N treatments, respectively. Greater production and longer reduction period of N2O were observed in the N treatment compared with the S + N treatment. Reduction of N2O took about 3 and 14 days in the S + N and N treatments, respectively (Fig. 1e–g, respectively).

Fe(III) reduction was greatly inhibited in the N treatment; the reduction of them nearly took 40 days after incubation. The patterns of Fe(III) reduction in other treatments were similar (Fig. 1h). SO4 2− obviously accumulated after 3 days of incubation in all treatments except the S treatment. A very sharp peak of SO4 2− was observed in the S + N treatment but decreased quickly within 4 days. The reduction of SO4 2− in the N treatment took a longer time than that in the Ctrl treatment (Fig. 1i).

Quantification of mcrA genes and transcripts

Straw addition increased the abundance of both mcrA genes and their transcripts. The abundance of mcrA genes and transcripts was higher in the S treatment than in the S + N treatment, with values reaching 1.44 × 1011 genes and 1.15 × 1013 transcripts g−1 dry soil in the S treatments, respectively. In S + N treatment, the abundance of mcrA genes and transcripts was 1.22 × 1011 genes and 8.97 × 1012 transcripts g−1 dry soil, respectively. The mcrA gene and transcript abundances remained relatively constant during the entire incubation period in the Ctrl and N treatments. The abundance of mcrA genes in the Ctrl and N treatments ranged from 1.54 × 109 to 2.07 × 1010 copies g−1 dry soil, and the abundance of mcrA transcripts in these two treatments ranged from 1.27 × 1010 to 7.18 × 1011 copies g−1 dry soil (Fig. 2a, b). The patterns of variations in mcrA genes and transcripts were significantly (P < 0.01) correlated with CH4 production. The abundance of mcrA genes and transcripts explained 72 % (r 2 = 0.72) and 96 % (r 2 = 0.96) of the observed variations in CH4 production, respectively.

Fig. 2
figure 2

The copy numbers of mcrA genes (a) and mcrA transcripts (b) (mean ± SE, n = 3) in all treatments over sampling time based on real-time PCR

Full size image

Dynamics of methanogenic archaeal community structure

T-RFLP analyses of the mcrA genes and transcripts were performed on different samples taken on days 0, 5, 11, and 23 to determine the population dynamics of methanogenic communities. In total, seven and eight characteristic terminal restriction fragments (T-RFs) were identified as major peaks in T-RFLP profiles of the mcrA genes (Fig. 3) and their transcripts (Fig. 4), respectively. The phylogenetic trees showed that methanogenic community mainly comprised Methanosarcinaceae, Methanocellaceae, Methanosaetaceae, and Methanobacteriaceae. Clone sequences were also used to relate T-RFs to the respective methanogens, i.e., four T-RFs (390, 422 , 489 , and 268 bp) to Methanosarcinaceae, two T-RFs (234 and 402 bp) to Methanocellaceae, and T-RFs of 502 and 416 bp to Methanobacteriaceae and Methanosaetaceae, respectively (Table S1, Figs. S1 and S2).

Fig. 3
figure 3

Community structures of methanogenic community based on T-RFLP analysis targeting mcrA genes in anaerobic incubation with treatments of Ctrl, S, N, and S + N and sampling time in rice field soil. Only the relative abundances (mean ± SE, n = 3) of major terminal restriction fragments (T-RFs) (relative abundance >1 %) are shown. Msr Methanosarcinaceae, Msa Methanosaetaceae, Mcl Methanocellaceae, Mb Methanobacteriaceae

Full size image
Fig. 4
figure 4

Community structures of methanogenic community based on T-RFLP analysis targeting mcrA transcripts in anaerobic incubation with treatments of Ctrl, S, N, and S + N and sampling time in rice field soil. Only the relative abundances (mean ± SE, n = 3) of major terminal restriction fragments (T-RFs) (relative abundance >1 %) as digested by Sau96I enzyme from PCR products are shown. Msr Methanosarcinaceae, Msa Methanosaetaceae, Mcl Methanocellaceae, Mb Methanobacteriaceae

Full size image

The T-RFLP patterns derived from mcrA genes revealed that the dynamics of methanogenic community were similar between the S and S + N treatments and between the Ctrl and N treatments. Methanosarcinaceae were predominant in all treatments. The relative abundance of Methanocellaceae increased in the S and S + N treatments on day 23 but decreased in the S treatment on day 5. Methanobacteriaceae decreased in the Ctrl treatment from day 0 to day 23, but only decreased on day 23 in the other three treatments. Methanosaetaceae increased in the N treatment but decreased in treatments with straw on both day 11 and day 23 (Fig. 3).

mcrA cDNA-based T-RFLP fingerprints were assumed to represent the community of active methanogens. cDNA-based T-RFLP profiles were distinct from those derived from genes. The dynamics of methanogenic community were different between treatments and showed strong variation across sampling time. Methanosarcinaceae increased in treatments without N on day 5 and day 11 but decreased on day 23. In the N and S + N treatments, Methanosarcinaceae decreased on day 5 but increased on day 11 and day 23. The relative abundances of Methanocellaceae, Methanobacteriaceae, and Methanosaetaceae showed great fluctuation in all treatments and over sampling time. Methanocellaceae increased in the N treatment on day 5. Methanobacteriaceae showed increases in the S and S + N treatments on day 11 and day 23. Methanosaetaceae showed increases in the N and S + N treatments on day 5 but decreased on days 11 and 23. Methanosaetaceae also decreased in the S treatment after 5 days of the incubation (Fig. 4).

Redundancy analysis (RDA) of T-RFLP profiles was performed to summarize the percentage of total variation in methanogenic communities that could potentially be explained by the measured variables (treatments, sampling time, and all physicochemical parameters) (Table 1). At the gene level, treatments (Ctrl, S, N, and S + N), sampling time (days 0, 5, 11 and 23), and physicochemical parameters (CO2, H2, acetate, NO3 −, NO2 −, N2O, SO4 2−, and Fe(II)) as explanatory variables accounted for 81 % of the total variation. The factors with a significant effect were sampling time and SO4 2− and CO2 concentrations, which contributed to 30, 21, and 12 % of the total variation of methanogenic community, respectively. Based on the mcrA transcript level, treatments, sampling time, and physicochemical parameters as explanatory variables accounted for 77 % of the total variation. Sampling time and Fe(II) concentration as significant effect factors contributed to 25 and 18 % of the total variation of methanogenic community, respectively.

Table 1 Redundancy analysis (RDA) of methanogenic composition (based on DNA and cDNA level) observed in all the treatments (Ctrl, S, N, S + N) throughout all sampling times (days 0, 5, 11, 23)
Full size table

Effects of variables on CH4 production

Table 2 shows the effects of all measured variables (treatments, sampling time, all physicochemical parameters, the relative abundance of T-RFs based on DNA and cDNA level, and the copy numbers of mcrA gene and their transcript on CH4 production by RDA (Table 2)). All factors together explained 89 % of the total variance of CH4 production. Significant factors included CO2 (75 %), SO4 2− (4 %), acetate (3 %), the relative abundance of T-RFs of 422 bp (Methanosarcinaceae, based on DNA level, 2 %), and mcrA gene abundance (1 %).

Table 2 Redundancy analysis (RDA) for the dependent variables on CH4 production
Full size table

Discussion

Physicochemical parameters

As reported in previous studies (Peng et al. 2008; Lu et al. 2015), our study showed that straw addition increased CH4 production. The addition of nitrate with straw reduced CH4 production compared with the treatment with straw alone. The amounts of CH4 production were similar between Ctrl and the treatment with nitrate alone. The differences of CH4 production between treatments most likely related to the production of substrates, such as H2 and acetate (Conrad 2007). A recent study found that the peak of acetate concentration coincided with the beginning of increased methane production (Wegner and Liesack 2015). H2 and acetate in microcosms with both straw and nitrate addition were lower than those in microcosms with straw alone, which was possibly due to competition for methanogenic substrates by denitrifying bacteria or the toxic effects of denitrification intermediates (NO2 −, NO, and N2O) on fermentative bacteria which were involved in acetate or H2 production (Meyer 1981) or possible toxic effect of products of nitrate reduction (NO, NO2 −, and N2O) on methanogens (Chidthaisong and Conrad 2000). Thus, the inhibitory effect of denitrification intermediates on methanogens was a possible contributing factor for lower CH4 production in treatment with both straw and nitrate. Such toxic effects have been reported in anoxic soil or pure culture studies (Klüber and Conrad 1998a; Roy and Conrad 1999). For example, the suppression of methanogenic acetate utilization by methanogens has prolonged although nitrate reduction was completely finished in methanogenic soil incubation (Klüber and Conrad 1998b). The microbial reduction of accumulated SO4 2− that was released from clays and hydrous aluminum oxides during anoxic incubation in treatments with nitrate addition may decrease electron donors (H2 and acetate) available to the methanogens (Yao et al. 1999) and thus decreased CH4 production. The deficiency of elector donors in treatments with nitrate alone possibly resulted in low reduction rates of Fe(III), so further influenced CH4 production.

Responses of the methanogenic communities to treatments

Straw addition resulted in increases in biomass and activity of methanogens. These growth and activity were most likely supported by substrates (H2 and acetate) derived from the decomposition of organic matter (Conrad and Klose 2006; Conrad et al. 2012). Nitrate showed inhibitory effect on both mcrA gene and transcript abundances in the presence of straw. Ctrl and treatment with nitrate-alone addition showed similar patterns of mcrA gene and transcript abundances, with low abundances of mcrA genes and transcripts during the entire incubation period. These results could be related to differences in the rates of substrate production between treatments and indicated that energy substrates were the limited factors to microbial growth and activity. While both mcrA gene and transcript abundances were positively correlated with methane production, transcript abundance was more significant, as observed in other studies (Freitag and Prosser 2009; Ma et al. 2012; Xu et al. 2012), indicating that the variations of mcrA transcript abundance were more powerful in predicting CH4 production.

We analyzed the succession and dynamics of the methanogenic community with the purpose of understanding the effects of straw and nitrate addition. The methanogenic community, as determined by gene-based T-RFLP analysis, changed in response to straw addition but did not alter in nitrate-alone treatment. However, examining the transcriptionally active community revealed a more dynamic and complicated response to straw and/or nitrate addition. This probably resulted from that the transcription level analysis is much more sensitive than the gene level, as activity of cells is directly associated with turnover and synthesis of rRNA (Molin and Givskov 1999).

Based on analysis of mcrA genes, Methanosarcinaceae were found to be the predominant archaea of the methanogenic community, and straw addition (alone) stimulated their growth in the early stage of incubation, as observed previously (Peng et al. 2008; Conrad et al. 2012). Methanosarcinaceae are known as fast-growing and substrate-versatile methanogens and not only can use high concentrations (0.2–1.2 mM) of acetate, but also can utilize H2-CO2 or methanol (Jetten et al. 1990; Conrad 2007). In straw-amended treatments, acetate and H2 accumulated to high concentrations resulting from anaerobic straw degradation in the same incubation stage. Nitrate inhibited the Methanosarcinaceae during the early stages of incubation, particularly on day 5 at the transcript level, presumably due to the toxicity of denitrification intermediates directly on Methanosarcinaceae or indirectly by inhibiting fermentative bacteria (Fig. 1c, 1d), and subsequent substrate production. Similar observations have been reported before (Scheid et al. 2003; Meyer 1981), where the growth of methanogens was inhibited even after nitrate was completely reduced, indicating the inhibitory effect of nitrate-derived intermediates on methanogen growth.

Methanosaeta spp. have been described as slow-growing methanogens, utilizing acetate over a lower range of concentrations (7–70 μM) (Jetten et al. 1992). Methanosaetaceae are usually dominant when acetate concentrations decrease to a low level (Peng et al. 2008). In this study, the relative abundance of Methanosaetaceae in gene level obviously increased in the Ctrl and nitrate-alone treatment on days 5 and 11. A clear shift from Methanosarcinaceae to Methanosaetaceae was observed on day 5 in treatments with nitrate by analysis of transcripts. An expected change of acetate concentrations was observed in these treatments, with accumulation in the early stages of incubation with a subsequent decrease; this was strongly in accordance with earlier reports (Jetten et al. 1992; Peng et al. 2008). Moreover, higher levels of nitrate (NO3 −), nitrous oxide (N2O), and sulfate (SO4 2−) in the same period in these treatments probably affected Methanosarcinaceae and Methanosaetaceae. These results further demonstrated that analysis of mcrA transcripts was more sensitive and powerful than analysis of mcrA genes.

Methanocellaceae are known as RC-I methanogens and play a key role in CH4 production in paddy soil (Lu and Conrad 2005). Hydrogenotrophic Methanocellaceae were favored in nitrate treatments in early incubation stage (on day 5) in both gene and transcript levels. Wu et al. (2009) have observed the increase of hydrogenotrophic methanogens on the rice roots with the increasing application of N fertilizer in paddy soil. The increase of Methanocellaceae (based on both gene and transcript levels) may be stimulated due to application of N fertilizer in this study. Additionally, SO4 2− nearly kept at high level in nitrate-alone treatment during the whole incubation period, which probably stimulated the increase of Methanocellaceae. Lyu and Lu (2015) have reported that Methanocellales (as Methanocella arvoryzae and Methanocella paludicola) were most likely to use sulfate as a sulfur source and assimilate sulfate into biomass. High activity of Methanocellaceae (based on transcript level) was also observed in later incubation stage (on day 23) in all treatments. This shift could be explained by the change of H2 concentration in the headspace. Low H2 concentrations were favored Methanocellaceae (Lu and Conrad 2005; Peng et al. 2008). Previous studies found that Methanocellaceae predominated methanogenic activity at high temperatures (e.g., 45 °C) (Peng et al. 2008; Lu et al. 2015). Methanocellaceae increased at 30 °C in our study, indicating that they can also play an important role in CH4 production at moderate temperatures.

Methanobacteriaceae using H2 plus CO2 have been isolated from rice field soil (Joulian et al. 1998). In this study, Methanobacteriaceae were favored in the treatments with straw application, which could be explained by variation of H2 concentration (Lu and Conrad 2005; Peng et al. 2008), indicating that Methanobacteriaceae contributed to CH4 production in straw-applied rice field soil.

In conclusion, nitrate amendment suppressed the straw-induced increase in CH4 production in laboratory incubations. Methanogenic community structures, as determined by mcrA gene analysis, changed in response to straw addition, but at mcrA transcript level, methanogenic community structures showed more sensitive, revealing differential response to a combined addition of nitrate and straw, as well as temporal fluctuations. These results suggest that nitrate application could alleviate CH4 production during anaerobic degradation of rice straw in paddy soils by inhibiting the production of substrates such as H2 and acetate and thus might influence CH4 emission from rice fields and even its contribution to global warming. Investigating the CH4 production and the dynamics of the methanogenic archaeal community upon rice straw and/or nitrate addition in paddy soils could provide insights into C and N biogeochemical cycles in rice fields and even further important suggestions for regulating CH4 emission from rice field soils via agricultural management practices.