Introduction

Coastal marsh ecosystems are believed to be net sinks for carbon sequestration because they accumulate C at a rate of up to 100 times higher than that in terrestrial forest ecosystems [1,2,3]. However, coastal marsh ecosystems are also important sources of greenhouse gases because, compared to other ecosystems, they are characterized by a wide range of anaerobic and biogeochemical conditions along both spatial and temporal gradients. These environmental gradients, as well as the high soil sulfate concentration, may change the dynamics of methane (CH4) and carbon dioxide (CO2) emissions [4]. Because such greenhouse gas emissions are mainly from soil microbial activities, understanding the composition and diversity of active microbial communities in coastal ecosystems, particularly those related to CH4 dynamics, is very important because CH4 is considered nearly 25 times more efficient than CO2 in trapping heat [5].

CH4 emission varies across salinity regimes in tidal wetlands because the sulfate present in salt water affects the competition between sulfate reduction and methanogenesis [6]. Moreover, in strongly tide-influenced wetlands, 52–90% of the CH4 produced in situ is reoxidized to CO2 by methanotrophs before it reaches the atmosphere [7,8,9].

Canonical aerobic methanotrophs include three types. Type I includes the family Methylococcaceae belonging to the γ-Proteobacteria, type II includes the families Methylocystaceae and Beijerinckiaceae belonging to the α-Proteobacteria [10], and type X belongs to the γ-Proteobacteria [11,12,13]. Most methanotrophs, with the exception of Methylocella and Methyloferula, contain the pmoA gene encoding the particulate methane monooxygenase, which catalyzes the first step of methane oxidation [14, 15]. Thus, pmoA has been used as a molecular marker to evaluate the diversity and habitat preferences of cultured and uncultured aerobic methanotrophic bacteria [16].

Methanotrophic activity in coastal ecosystems such as freshwater tidal wetlands and brackish wetland sediments has been discussed in previous reports [8, 17], whereas there have been only few studies on the mangrove ecosystem. In tropical and subtropical areas, mangrove forests are one of the dominant ecosystems in coastal regions, occupying more than 15.2 million hectares globally [3, 18, 19]. Because of the diverse salinity ranges due to different tidal frequency, the CH4 emission was between 0.003 and 7.7 mg C m−2 h−1 [20,21,22]. Gomes et al. [23] have found some methanotrophs, such as Methylomonas, in the rhizosphere of mangrove roots after a 75-day incubation period of mangrove seedlings in mesocosms. However, the active methanotrophs inhabiting mangrove forest soils and their activities are still unclear because the abundance of methanotrophs is relatively low in coastal ecosystems and they are difficult to isolate [12].

DNA-based stable isotope probing (DNA-SIP) with a 13CH4 label has been demonstrated to be a successful tool for the identification of active methanotrophs in rice paddy, grassland soils [24,25,26], geothermal springs, and lake sediments [27,28,29]. In combination with isopycnic centrifugation, it was possible to extract the pmoA gene to quantify the diversity of the active methanotrophs incorporating 13C in the targeted soils [30].

This study was designed to quantify and evaluate the methanotrophs in natural tidal mangrove forests in northern Taiwan, using the DNA-SIP technique, and by quantitative polymerase chain reaction (qPCR) and high-throughput sequencing of the pmoA and 16S ribosomal RNA (rRNA) genes. This study aims to understand how methanotrophic communities vary in mangrove forest soils that experience different frequency of tides and how efficiently they oxidize CH4.

Materials and Methods

Sampling Site

This study was carried out in the Tamsui estuary, Taipei, Taiwan (121° 26′ E, 25° 9′ N), a large portion of which (around 100 ha) is covered with Kandelia obovata (Fig. 1). The average atmospheric temperature in the Tamsui estuary was between 14.7–31.4 °C with an average of 22.4 °C [31]. Tall K. obovata trees, about 2–3.5 m high, are the dominant plants in most of the area (designated as tidal mangrove), where tidal cycles occur every day. At a higher elevation (1–1.5 m above the sea level), located next to the tall mangrove forest, soil is only inundated during monthly high tides, and dwarfed K. obovata plants of about 0.5–1.5 m high are the dominant plant community (designated as dwarf mangrove) [32]. The tidal mangrove soil has an average salinity of 3.65 psu, sulfate (SO4 2−) concentration of 269.6 μg S g−1 soil, and soil oxidation–reduction (redox) potential of 180 mV, and the dwarf mangrove soil has a salinity of 2.88 psu, SO4 2− concentration of 610.5 μg S g−1 soil, and soil redox potential of 370 mV [33], which makes the mangrove forests oligohaline (0.5–5.0 psu) systems. The annual mean ammonium and nitrate concentrations were respectively 1.2 and 0.44 mg N L−1 [34], and the water temperature was 14.7–30.5 °C [31] in the main channel of the Tamsui River.

Fig. 1
figure 1

Map of the study area and relative locations of the tidal and dwarf mangrove forests

Full size image

A total of 36 soil core samples (18 from each site) were collected in January 2016 from both dwarf and tidal mangrove soils during low tides on a random walk, using polyvinyl chloride (PVC) tubes (1.5 cm in diameter and 30 cm long). Any visible litter on the soil surface was removed prior to collection. The collected samples were brought to the laboratory and immediately pushed out from the bottom of the PVC tubes, using rubber stoppers. Then, the top 1 cm of the soil samples was sliced off from the bulk soil, using a knife pretreated with 70% ethanol. The sliced-off 1-cm soil samples from the 18 cores from the same site were well mixed in a plastic zipper bag as a composite sample and stored at 4 °C at their initial moisture content until experiments were conducted.

Sub-samples were taken and oven-dried at 105 °C for 72 h to determine the soil moisture content. As supplemental information, total organic C and total N were measured by the combustion method using a Fisons NA1500 elemental analyzer (ThermoQuest Italia, Milan, Italy).

Soil Incubation

Six replicates of samples were re-distributed from the composite sample from each site and incubated within a week after the soil cores were collected from the mangroves. For each replicate, 10 g of soil was weighed from the mixed soil samples, placed into a 146-mL Erlenmeyer flask (125-mL flask with 21 mL of extra headspace above the flask scale), and sealed with a double-sided rubber stopper. Then, the six replicates from each site were separated into two sets. The headspace for one set of samples was replaced with 1.4 mL of pure (99.9%) 12CH4, and that for the other set of samples was replaced with 1.4 mL of pure (99 atom % 13C) 13CH4 (Sigma–Aldrich, St. Louis, MO, USA) to make the headspace contain approximately 1.2% CH4 for all samples.

A 1-mL gas sample was extracted from the headspace of each flask with a syringe to determine the initial CH4 concentration, and 1 mL of nitrogen gas was injected back into the flask to maintain the constant pressure in the headspace. The extracted gas samples were immediately analyzed using a gas chromatograph with a thermal conductivity detector (GC-TCD, 9790II, Fuli Instruments, Zhejiang, China) with five standard CH4 concentrations: 0; 5000; 7500; 10,000; and 15,000 ppmv.

The flasks were incubated in a dark incubator (LM-570RD, Yihder Technology Co., Taiwan) at 20 °C for 15 days, which was comparable to but slightly lower than the field temperature conditions to better capture the change in CH4 concentrations through time. Only the set of samples with injected 12CH4 was periodically tested (every week) during the incubation for the remaining CH4 concentration in the headspace to measure the CH4 oxidation rate while preventing dilution of the 13CH4 concentration in the headspace of the isotope-labeled samples. The incubation was completed when no 12CH4 was detected in the headspace of the flasks.

DNA Extraction and SIP Gradient Fractionation

Total DNA was extracted from 0.8 g of the incubated soil samples, using the PowerSoil® DNA isolation kit (MO BIO Laboratories, Inc., Carlsbad, CA, USA) according to the manufacturer’s instructions. The extracted DNA was stored at − 20 °C until further analysis.

The extracted DNA (6 μg) from each sample was partitioned into 15 density gradient fractions by the density gradient centrifugation and fractionation method described by Lu and Jia [35]. The abundance of the methanotrophic pmoA gene in fractions 2–14 from all 13CH4-labeled samples was then tested by DNA amplification using qPCR [36] with the A189F/mb661r primer pair. The cycling conditions were as follows: 5 min at 95 °C, followed by 40 cycles of 95 °C for 10 s, 55 °C for 30 s, 72 °C for 30 s, and 80 °C for 5 s with acquisition after both last cycling steps. Then, 13C-incubated methanotroph species were identified in fractions with the highest pmoA gene abundance using the Illumina MiSeq platform. The reads were processed and quality-controlled using QIIME [37]. Then, the trimmed sequences were classified using mothur [38] with instructions and reference database provided by Dumont et al. [39]. The abundance of identified methanotrophic sequences in the heavy fractions of each replicate sample at the same site was then pooled for calculation of the relative abundance.

Three replicates of fresh soil total DNA samples at each site were extracted and analyzed for pmoA gene abundance using the same qPCR technique. The sample with the highest pmoA gene abundance was identified for the methanotroph species using the protocol described above.

High-Throughput Sequencing of 16S rRNA Gene and Data Analysis

To access the bacterial community structure in different gradient fractions, the 16S rRNA gene was analyzed in fractions 2–14 of the 12C- and 13C-incubated soil samples at both mangrove sites. Because of the numerous amounts of DNA samples that needed to be processed, only one of the triplicate samples was analyzed for the 16S rRNA genes.

The V4 region of the 16S rRNA gene was amplified with primers 515F and 806R [40]. Both the forward and reverse primers included barcoded sequences and adapters. PCR was performed at 94 °C for 3 min, followed by 25 cycles of 94 °C for 30 s, 55 °C for 30 s, and 72 °C for 30 s and a final extension at 72 °C for 6 min using the KAPA HiFi HotStart ReadyMix Kit (KAPA Biosystems, MA, USA). The 16S rRNA genes were sequenced using the Illumina MiSeq platform. Sequence analysis was performed using the Ribosomal Database Project (RDP) pipeline (http://pyro.cme.msu.edu). The taxonomic information was analyzed using the naïve Bayesian rRNA classifier [41] in RDP.

The bacterial community structure in the fresh soil was also analyzed using the V4 region of the 16S rRNA gene for comparison. Similar to the analysis of the pmoA gene in fresh mangrove soils, three replicates of total DNA samples from fresh soil at each site were extracted, and the V4 region of the 16S rRNA gene in the sample with the highest pmoA gene abundance was amplified and sequenced using the Illumina MiSeq platform.

Statistical Analysis

Two-way factorial analysis was performed to determine whether the abundance of the pmoA gene was spatially or temporally different. Student’s paired t test was applied to further test the means of dependent variables when the two-way factorial analysis revealed a difference. Because the abundance of the pmoA gene was non-linearly distributed, log transformation of the raw data was performed prior to the statistical analysis.

All statistical analyses were performed using JMP 11.0 (SAS Inc., Cary, NC, USA). P < 0.05 was considered significant for all tests.

Results

The moisture contents in the top layer (0–1 cm) of the dwarf and tidal mangrove soils were 23.5 and 30%, respectively. The total organic C content was higher in the tidal mangrove soils (35.4 mg C g−1 soil) than in the dwarf mangrove soils (8.4 mg C g−1 soil). Similarly, the total soil N content was also higher in the tidal mangrove forest soils (2.4 mg N g−1 soil), whereas it was negligible in the dwarf mangrove forest soils (0.1 mg N g−1 soil).

The results showed that the replicates of both dwarf and tidal mangrove soils had similar initial CH4 concentrations (12,100 ppmv). The CH4 concentration quickly decreased in the dwarf mangrove samples, and CH4 was completely depleted from these samples on day 8, whereas it was not depleted from the tidal mangrove samples until day 15. On average, the dwarf mangrove and tidal mangrove soils utilized the headspace CH4 at a rate of 15.13 and 9.63 μg g−1 day−1, respectively (Fig. 2).

Fig. 2
figure 2

Potential CH4 oxidation rates in incubated dwarf and tidal mangrove soils

Full size image

qPCR analysis of the fresh and incubated soil samples showed that the pmoA gene abundance significantly increased after incubation (P = 0.003), and the abundance of the pmoA gene was higher in tidal mangrove soils (fresh: 5.7 × 107 copies g−1 soil; incubated: 2.4 × 108 copies g−1 soil) than in dwarf mangrove soils (un-incubated: 1.7 × 107 copies g−1 soil; incubated: 8.5 × 107 copies g−1 soil) (P = 0.02). However, no interaction was found between the sites and the treatments (i.e., fresh and incubated soils).

qPCR analysis of the DNA extracted and fractioned from the incubated soil samples showed that the pmoA gene was present in both dwarf and tidal mangrove forest soils. About 60–100% of the pmoA gene was successfully separated by ultracentrifugation and distributed in heavy fractions (layers 6–9) of the 13C-labeled dwarf and tidal mangrove soil samples. On the contrary, 70–100% of the pmoA gene was in light fractions (layers 11–13) of the non-isotope-labeled (incubated with 12CH4) dwarf and tidal mangrove soil samples (Fig. 3).

Fig. 3
figure 3

Relative abundance of the pmoA gene and 16S rRNA gene in fractionated DNA samples of dwarf and tidal mangrove soils. Solid symbols represent the 13CH4-labeled gene. Open symbols represent the 12CH4-labeled gene. Only one representative result is presented from the 3 replicate samples. The ratio of the maximum quantities of pmoA genes was calculated from the copies of the pmoA genes in one fraction layer divided by the number of pmoA copies in the layer with the highest quantity of pmoA genes × 100%

Full size image

For the pmoA analysis using the MiSeq platform, 22,460 (dwarf: 10,098 sequences; tidal: 12,362 sequences) and 564,497 (dwarf: 260,888 sequences; tidal: 303,609 sequences) pmoA gene sequences were retrieved, respectively, from the fresh total soil DNA and 13C-labeled DNA peak layer from both mangrove sites, and 30 and 26 OTUs were respectively identified at the level of 93% identity [42] in the fresh soil samples and the isotope-labeled DNA peak layer soil samples.

In the fresh mangrove forest samples, the majority of the sequences (11,717 sequences) belonged to type Ib methanotrophs, 7490 sequences of the pmoA gene belonged to type Ia methanotrophs, and 2241 sequences of the pmoA gene belonged to type II methanotrophs (Fig. 4). Methanotrophs affiliated with the cluster deep-sea-5 appeared to be the most dominant methanotroph genus among all methanotrophic genes in the mangrove forests and were more dominant in the tidal mangrove soils than in the dwarf mangrove soils (64.4 and 29.8%). In contrast, more Methylobacter and Methylocystis were found in the dwarf mangrove soils than in the tidal mangrove soils.

Fig. 4
figure 4

Proportion of bacterial phylotypes affiliated with methanotrophs in the fresh and incubated 13C-DNA peak fraction using pmoA gene sequences and averaged from the three replicate samples at one site

Full size image

In the incubated samples, the majority of the sequences (476,320 sequences) belonged to type Ia methanotrophs, 48,029 sequences of the pmoA gene belonged to type Ib methanotrophs, and only 34 sequences of the pmoA gene belonged to type II methanotrophs. Methylosarcina was the most dominant methanotrophs in both mangrove forests and comprised approximately 50% of the relative abundance among all methanotrophs. Methylobacter appeared to have higher abundance in the dwarf mangrove soils, whereas Methylomonas had higher abundance in the tidal mangrove soils.

The total microbial composition analyzed from the 16S rRNA gene sequence using MiSeq sequencing of the fresh soil samples (dwarf: 248,775 sequences; tidal: 300,400 sequences) is shown in the supplementary file (Table S1 and Fig. S2). Among the methanotrophs identified from 16S rRNA, a large portion (dwarf: 58.3%; tidal: 63.5%) of unclassified methanotrophs was present in both dwarf and tidal mangrove soils by 16S rRNA MiSeq sequencing (Fig. 5). Among the identified methanotrophs, both a type II methanotroph, i.e., Methylocystis, and type Ia methanotrophs, i.e., Methylobacterium and Methylobacter, were found in the dwarf mangrove soils. However, the type Ia Methylosarcina, Methylobacter, and Methylomonas were mostly found in the tidal mangrove soils.

Fig. 5
figure 5

Proportion of bacterial phylotypes affiliated with methanotrophs in the fresh and incubated 13C-DNA peak fraction using 16S rRNA gene sequences

Full size image

Results of 16S rRNA gene analysis from the isotope-labeled DNA peak layers (dwarf fraction 7: 207,308 sequences; tidal fraction 6: 223,021 sequences) also identified a large portion of unclassified methanotrophs in both dwarf and tidal mangrove soils, and their sequences were most similar to those of Methylomarinum commonly found in marine sediments (Fig. 6) [43]. In contrast, approximately 30% of the methanotrophs was composed of Methylosarcina and Methylomonas in the tidal mangrove soils.

Fig. 6
figure 6

Phylogeny of pmoA and 16S rRNA lineages detected in dwarf (DF) and tidal (TL) mangrove soils. The neighbor-joining tree was constructed by using the MEGA 7.0 software package. Representatives of the OTUs detected in the two genes are shown in bold italic font. The pmoA sequences of type II methane-oxidizing bacteria were used as the outgroup. The scale bar represents 10% sequence divergence for the pmoA gene and 2% sequence divergence for the 16S rRNA gene. The values following the OTU numbers are the percentage of reads for the pmoA genes and the 16S rRNA gene in peak fractions from dwarf (DF) and tidal (TL) mangrove soil samples

Full size image

Discussion

Activity of Methanotrophs and Feasibility of Using 13CH4 DNA-SIP

Methanotrophs present in the dwarf mangrove soils showed a stronger CH4 oxidation potential than those present in the tidal mangrove soils. This may be due to the difference in the frequency of tides between the two sampling locations, resulting in different redox potentials (dwarf: 370 mV; tidal: 180 mV) of the soils [33]. Since the methanotrophs identified in the present study were aerobic, it is reasonable to assume that the CH4 oxidation potential is higher in the dwarf site than in the tidal site. The absolute abundance of methanotrophs was higher at tidal mangrove than at dwarf mangrove based on the results of the qPCR for the pmoA genes in both the fresh and incubated soil samples. However, the methanotrophic activity was higher in the dwarf mangrove soils since CH4 was depleted within 8 days in the incubated dwarf mangrove soil samples. The reason for the contradiction between the abundance of methanotrophs and their activity is unclear. However, the diffusion rates of CH4 and O2 are both more than 10,000 times higher in air [44] than in water [45].

Although the soil moisture from the collected samples was similar at the two sites, possibly because they were collected during the low tide, the different soil redox potentials implied that the soil retained a higher water content for a longer period of time in the tidal mangrove than in dwarf mangrove. Therefore, less CH4 and O2 could be available for methanotrophs in the tidal mangrove soils. Moreover, the soil total organic C and N contents were higher at the tidal mangrove than at the dwarf mangrove. Because the soil organic matter has a high water-holding capacity [46, 47], it may also help to keep water in the soil for a long period and prevent gas diffusion. The physiochemical difference between the two mangrove forest soils may further affect the methanotrophic composition in situ, resulting in the significantly different CH4 oxidizing activities observed at the two sites. This may provide a possible explanation for the low methanotrophic activity in the tidal mangrove soils.

When comparing the abundance of pmoA genes in the fresh mangrove forest soils to that in other ecosystems, the number of pmoA copies was similar to that of other coastal ecosystems [48] but was slightly lower than that of other freshwater wetlands in East Asia [49, 50]. This may be due to the presence of SO4 2− in coastal soils that reduced the CH4 efflux [4] and indirectly inhibited the growth of methanotrophs [12].

The patterns of abundance of the 13C- and 12C-labeled pmoA gene in different gravity fractions were similar to those detected in previous studies that used the same technique [30], indicating that active methanotrophic genes were successfully labeled with 13CH4. Therefore, the methanotrophs identified by this technique could be regarded as bacteria that participated in CH4 oxidation during the 15-day soil incubation period.

Methanotrophs Identified from the pmoA gene

Henckel et al. [51] indicated that type I methanotrophs are more common at a relatively low soil water content because they prefer a high-oxygen environment compared to type II methanotrophs. In the present study, we collected the mangrove soil samples from the top layer (0–1 cm) during low tides when the soil should be mostly aerobic. This may have resulted in mostly type I methanotrophs in the fresh soils at both sites. Although the two mangrove sites experienced different tidal frequencies, type I methanotrophs were dominant in both sites. A possible reason may be that the soil redox at both sites was still in a relatively oxidative condition [52, 53], which provided a favorable environment for the growth of type I methanotrophs at both sites even though the compositions were different at the two sites. However, as previously mentioned, the different compositions of type I methanotrophs at the two sites still affected their overall CH4-oxidizing potential.

Research has also shown that type I methanotrophs are usually found in soils with a limited CH4 supply because they grow better than type II methanotrophs in a low-CH4 environment [12]. On the contrary, type II methanotrophs such as Methylocystis are usually found in high-CH4 systems [54]. Coastal ecosystems typically release less CH4 than freshwater ecosystems because of the presence of SO4 2− in soil [4, 6, 55,56,57]. In mangrove forests, CH4 emission is 0.003–7.7 mg C m−2 h−1 [20,21,22], which is lower than that in other freshwater ecosystems such as rice paddies (4.3–16.4 mg C m−2 h−1) [58, 59] and wet ponds (0.75–40.5 mg C m−2 h−1) [60]. This may be another reason why mostly type I methanotrophs were predominant in the fresh mangrove soils.

In addition, previous research showed that the type I methanotrophic population was stimulated in incubated soil using the same DNA-SIP technique [61,62,63], possibly because the communities could rapidly react to the added CH4 concentration [64, 65]. It is therefore possible that type I methanotrophs were activated through the nature of the experimental design even though type II methanotrophs inhabited the soils.

In the fresh dwarf mangrove soil, nearly 20% of the methanotrophs were categorized as type II Methylocystis. Several studies have shown that type II methanotrophs can perform N2 fixation to provide a nutrient source [12, 66]. As the soil at the dwarf mangrove contained low soil TN, this may favor the growth of some type II methanotrophic bacteria. In addition, methanotrophs belonging to deep-sea-5 cluster were more abundant in the fresh tidal mangrove soils than in the dwarf mangrove soils. Although the exact reason for the higher abundance of the deep-sea-5 cluster was not clear, it may involve the more frequent tides at the tidal site that somehow provided environmental conditions more similar to those of marine sediments, as the deep-sea-5 cluster was mostly found in the marine ecosystems [16].

Among the methanotrophs identified using the pmoA gene after incubation, type Ia methanotrophs were the most dominant CH4-oxidizing bacteria in the dwarf and tidal mangrove soils. In particular, Methylosarcina expanded its population to nearly 50% of the relative abundance and became the predominant methanotroph in both mangrove soils after incubation. Research has shown that Methylosarcina can inhabit a wide pH (4.0–9.0) and temperature (4–37 °C) range [67]. The exact reason for the fast growth of Methylosarcina is unclear, but it could be due to our selection of experimental conditions suitable for its growth.

Research has shown that Methylobacter that has high adaptation to various salinity conditions is commonly found in freshwater to brackish water environments. Some species in this genus can even grow in water containing more than 3% NaCl [68]. Although the salinity in the dwarf and tidal mangrove soils was similar at the time when the soil samples were collected [33], coastal soils were found to have a higher salinity at higher elevations in summer because of less tidal inundation and high evapotranspiration during this time, which leads to the accumulation of salt in the surface soil [4]. Therefore, the higher relative abundance of Methylobacter in the dwarf mangrove soil than in the tidal mangrove soil may imply that it tolerates high salinity levels, especially in summer when salinity of soil-pore water could be potentially high.

Previous research has indicated that most species belonging to Methylomonas grow at pH 6.5–7.0 [68, 69], whereas the optimal pH for Methylobacter is pH 7.0 [68]. Although the optimal pH for the two methanotrophs is not widely different, the soil pH values in the dwarf mangrove soils (pH = 6.34) were slightly lower than those in the tidal mangrove soils (pH = 7.55) [33]. Therefore, this pH difference might have contributed to the larger proportion of the Methylomonas community in the tidal mangrove soils (fresh: 1.8%; incubated: 18.7%) compared to the dwarf soils (fresh: 0.0%; incubated: 3.5%).

The different compositions of methanotrophic communities in the fresh dwarf and tidal mangrove soils suggest the influence of the tide frequencies on the methanotrophic composition in mangrove soils.

Methanotrophs Detected in Mangrove Soils by 16S rRNA Gene Sequencing

The results of the 16S rRNA gene sequencing showed that the majority of methanotrophs in both dwarf and tidal mangrove soils were unclassified (OTUs 5–7) and similar to Methylomarinum, which is commonly found in marine sediments. However, after aligning the pmoA primers and the sequences of the methanotrophs identified in the present study, we found that all the identified methanotrophs contained the same conserved genes, and their respective sequence regions basically matched the sequences of the pmoA primers used in this study. This result showed that the methanotrophs of OTUs 5–7 were different from Methylomarinum species because we could not find the same OTUs (OTUs 5–7) after extracting the pmoA genes from the mangrove forest soils. Therefore, the exact composition of the OTU 5–7 methanotrophs is still unknown, possibly because of the short fragments of the 16S rRNA gene that were sequenced. However, the results clearly indicate that methanotrophs living in estuarine mangroves may be widely different from those inhabiting freshwater systems. Other than the unclassified methanotrophs, the composition of the rest of the methanotrophs identified using the 16S rRNA gene was similar to that determined using the pmoA genes at both sites. This implies that both methods can be successfully applied in tidal mangrove forests.

In addition, as previously mentioned, the dwarf mangrove soils showed high activity of methanotrophic bacteria but low abundance of pmoA genes. The methanotrophic communities in dwarf and tidal mangrove soils identified from the pmoA genes showed similar composition. Additionally, the unclassified methanotroph in dwarf mangrove soils largely increased in relative abundance during the incubation. This implies that the high methane-oxidizing potential of dwarf mangrove soils may be mainly attributed to this methanotroph, which has yet to be classified.

Several studies have found that anaerobic oxidation of CH4 may be an important pathway in coastal freshwater and brackish wetlands because microbes use sulfate as an alternative electron acceptor when oxygen is depleted [17, 70, 71]. Since we used aerobic incubation to label aerobic methanotrophs in the present study, we may consider incubation of subsurface mangrove soils under anaerobic conditions in further studies to evaluate the importance of anaerobic oxidation of CH4 in tidal mangroves as the present study may only identify aerobic methanotrophs in the mangrove forest soils.

Conclusions

Our study indicated that shallow mangrove soils mostly contained active type I methanotrophs. The methanotrophic activity was higher in dwarf mangrove soils, which are only inundated during monthly high tides and have a higher soil redox potential than that of tidal mangrove soils. The compositions of methanotrophs in both sites were similar when using the 13CH4 DNA-SIP and pmoA genes, with methanotrophs belonging to the deep-sea-5 cluster and Methylosarcina being the most abundant methanotrophs in the fresh dwarf and tidal mangrove soils, whereas Methylosarcina appeared to be the most active methanotroph after the incubation.

The results of the high-throughput sequencing analysis of the 16S rRNA gene indicated that a considerable number of methanotrophs were unclassified, implying that methanotrophic communities in estuarine mangroves may differ from those in freshwater ecosystems. Moreover, the increment of the unclassified methanotroph in the dwarf mangrove soils may be critical for methane oxidation in the mangrove forest. Overall, this study showed that the 13CH4 DNA-SIP technique enabled successful identification of active methanotrophic communities and also demonstrated that environmental fluctuations in mangrove ecosystems could result in differences in methanotrophic activity.