1 Introduction

There is growing interest and concern about greenhouse gas emissions from natural lakes and constructed reservoirs. These are one of the significant natural sources of methane, a greenhouse gas, with an estimated worldwide release of 12–70 Tg year−1 (Barros et al. 2011; Deemer et al. 2016), accounting for 6–16% of annual natural methane emissions (Goldman et al. 2016). Methane is produced in the anoxic sediments of lake/reservoir as a result of the microbial decomposition of organic matter. In addition, lake/reservoir sediments sequester a greater amount of extra labile organic materials than ocean sediments (Quadra et al. 2020). As a result, sediments in lakes/reservoirs are frequently regarded as ‘hot spots’ of methane production (Bastviken et al. 2004; He et al. 2018). While quantitative estimates of the terminal methane emission flux from lake/reservoir has been established through direct measurements and large-scale modelling (Maeck et al. 2013; Rosentreter et al. 2021), the transformation and mechanism of these events in sediments receive insufficient attention (Knittel and Boetius 2009).

Microbial activities regulate both the synthesis and consumption of methane in addition to methane diffusion and advection in sediments. It has been estimated that 50–95% (median of 90%) of the methane produced in freshwater lakes is oxidised prior to reaching the atmosphere (Bastviken et al. 2008). Methanotrophs oxidise methane biologically via aerobic and/or anaerobic processes. In lakes, the primary site of methane oxidation is expected to be in the oxic layer of sediments, or the layer of overlying water (Bastviken et al. 2008). However, recent study suggests that anaerobic oxidation of methane (AOM) (Wells et al. 2020) can occur in non-marine sediments via denitrification (Ettwig et al. 2010), sulfate reduction (Beal et al. 2009; Norði et al. 2013), and iron reduction via direct coupling or re-oxidation of sulfide (He et al. 2018).

Diverse biological niches for methane generation and consumption have been established in the complex sedimentary environment with multiple redox ions (Lloyd et al. 2011; Bodelier et al. 2013; Xiao et al. 2017; He et al. 2019). Denitrification, iron reduction, sulfate reduction, and methanogenic processes are listed in thermodynamic order based on the classical redox reaction sequence (Melton et al. 2014; Hansel et al. 2015). In the natural environment, however, certain reactions may overlap or reverse, and the boundaries of the reaction zone may even cross (Sela-Adler et al. 2017). Metal-AOM is found to exist directly below the oxic/anoxic boundary in iron-rich lake (Crowe et al. 2011) and below the zone of methanogenesis where nitrate and sulfate were entirely eliminated (Sivan et al. 2011). On the other hand, metal-AOM and sulfate-AOM were detected concurrently in the surface sediments of Lake Ørn (Norði and Thamdrup 2014). Additionally, anthropogenic eutrophication (Egger et al. 2015) and fast sediment deposition (Riedinger et al. 2014) could trigger the up-shift of sulphate-methane transition zone (SMTZ) in costal sediments. We hypothesise that, given the complex redox environment in reservoir sediments, methane production and oxidation may coexist in both freshwater as well as coastal sediments (Sivan et al. 2011; Xiao et al. 2017; Maltby et al. 2018), and that the geochemical gradient along a sediment profile may result in AOM niche differentiation depending on the electron acceptors.

We investigated a sediment core collected from Hongfeng Reservoir to explore the niche pattern of methane biological transformation. The reservoir is a lake-type reservoir with high concentration of nutrients and pollutants in its sediments, which have accumulated since the reservoir was constructed in 1958. In particular, we studied methane production and metabolism in freshwater sediments from the evidence of microbial colonisation combined with geochemical characteristics, focusing on the function of AOM driven by multi-electron acceptors on methane release from freshwater sediments. To achieve these objectives, the niche partitioning of methane production and oxidation were explored along the sediment core, through geochemical characteristics and microbial communities of samples in situ and lab incubation.

2 Materials and methods

2.1 Site description and sample collection

Hongfeng Reservoir is a eutrophic lake-type reservoir located in Qingzhen, Guizhou Province, China (106°19′–106°28′E, 26°26′–26°35′N; Fig. S1). The reservoir has a storage capacity of 6.01 × 108 m3, and the average and maximum water depths are 10.5 m and 45 m, respectively. Surface water samples and one sediment core (diameter of 65 mm) were collected from the central area of the reservoir (106.4151° E, 26.478533° N) with a 10-m water depth in January 2017 (Fig. S1). The core was ~38 cm in length, and reached the soil layer corresponding to the period before the construction of the dam. The surface sediments (~1 cm) were dark brown (Fig. S1) and may have contained iron oxides, but the deeper sediments were black. The sediment core was sliced in situ into 1 cm (0–20 cm) or 2 cm (20–38 cm) sub-samples. Porewater was extracted simultaneously using a Rhizon sampler (Rhizosphere Research Products, The Netherlands) connected to vacuum sampling bottles, and was then stored at 4 °C. Some of the solid samples were frozen at −20 °C for DNA extraction and sequencing, and some were kept at 4 °C for later incubation. Aliquots of sediments were dried in an oven at 105 °C for 12 h and then filtered through a 200-mesh strainer. Some aliquots were powdered manually in an agate mortar for subsequent geochemical analysis.

2.2 Geochemical analytical methods

2.2.1 Sediment sample analyses

A subsample was decalcified with 10% HCl before being washed twice with deionised water and dried at 50 °C for the detection of total organic carbon (Donis et al. 2017) determination using a CHN-O rapid elemental analyser (Heaeus, Germany).

2.2.2 Determination of components and isotopic compositions of porewater

The concentrations of anions (SO42− and NO2−) in the porewater were determined using an ion chromatographer (IC; ICS-1100, Thermo, CA, USA) equipped with a column of AG 19 (4 mm × 250 mm). The concentrations of formic acid, acetic acid, and propionic acid in porewater were analysed using an IC equipped with a column of Ion Pac AS11-HC (4 mm × 250 mm). Dissolved iron was determined by inductively coupled plasma mass-spectrometry (ICP–MS; Thermo iCAP Q, CA, USA) after samples were first diluted with 1% HNO3.

The concentrations and stable carbon isotopic (δ13C) compositions of CH4 and CO2 from the porewater were determined by gas chromatography (GC; GC-C/TC III) isotope-ratio mass spectrometry (IRMS; Delta V Advantage IRMS) and trace ultra GC (Thermo Finnigan). The chromatographic column was a HP-PLOT Q (30 m × 0.32 mm × 20.00 μm; J & W), and the injection temperature was 120 °C at a speed of 1.5 mL min−1. The temperature of the burner was 960 °C, and that of the reducing furnace was 600 °C. The precision of the δ13CCH4 and δ13CCO2 measurements was ± 0.2‰. The concentration and carbon isotopic compositions of dissolved inorganic carbon (DIC) were determined using GasBench-IRMS (Delta V Advantage, USA).

2.3 Laboratory incubation

To measure the methane production potential, fresh sediment (equivalent to a 5.0 g dry weight) was placed into a 125 mL serum bottle using a cap-cut sterile syringe. The serum bottles with sediments were vacuumed and gas charged with N2 for three cycles to achieve anaerobic conditions. The bottles were immediately sealed with thick butyl rubber stoppers and aluminium caps and incubated in a 25 °C incubator. In addition, sodium acetate (3.0 mmol kg−1) was added as a substrate for methanogens. For the AOM potential measurement, the sediment incubation experiment was conducted as above, and was supplemented with methane at an initial concentration of approximately 8.3 mL L−1 after being sealed and N2 replacement in the headspace. The headspace gas (1 mL) was collected by a micro-syringe for daily methane measurements. To maintain the ordinary pressure, 1 mL of pure N2 was injected back into the bottle. The CH4 concentration was determined by GC (GC-900, Shanghai Kechuang Chromatography Instrument Co. Shanghai, China) equipped with a hydrogen flame ionisation detector (FID), and expressed as per kilogram of dry weight (gdw) of sediment (Wassmann et al. 1998). All of the incubation experiments for methane production/oxidation potential were conducted in triplicate and, to the extent practicable, external contamination was avoided.

2.4 Microbial community

The DNA of sediment samples was extracted using a FastDNA® Spin Kit for Soil according to the manufacturer’s instructions. The quality and concentration of the extracted DNA was assessed via spectroscopic analysis (NanoDrop Technologies).

2.4.1 Illumina sequencing and analysis

The microbial communities of the collected sediments were analysed by Illumina MiSeq sequencing. Microbial sequencing was performed using the MiSeq Illumina platform at Major Biotechnology Company (Shanghai, China) according to the methods of Caporaso et al. (2012). Briefly, the V3–V4 and V4 regions of 16S bacterial and archaeal ribosomal DNA (rDNA) were amplified, respectively. The custom degenerate primer pairs of barcode-338F (5′-ACTCCTACGGGAGGCAGC-3′)/907R (5′-CCGTCAATTCMTTTRAGTTT-3′) (570 bp) and barcode-524F10extF (5′-TGYCAGCCGCCGCGGTAA-3′)/Arch958RmodR (5′-YCCGGCGTTGAVTCCAATT-3′) (434 bp) were used to generate an amplicon to construct libraries for bacteria and archaea, respectively. After sequencing, the quality of the raw data was checked (FastQC v0.11.8)) and filtered (PRINSEQ), and sequences less than 400 bp were eliminated from the resultant data. In total, 561,817 and 2,067,765 unique sequences were ultimately obtained for bacteria and archaea, respectively.

The sequence data were analysed using QIIME (version 1.17) (Deemer et al. 2016), and the sequences with a similarity of > 97% were clustered using Usearch (version 7.0, http://drive5.com/uparse/). This resulted in 6132 and 1885 operational taxonomic units (OTUs) for bacteria and archaea, respectively. The taxonomic assignment was performed using the Ribosomal Database Project (RDP) classifier (Goldman et al. 2016) and a training set extracted from the Silva108 database (Mendonça et al. 2012). Based on the results of the OTU clustering and annotation analysis, further data analysis was performed using the Mage’s I-Sanger platform (http://www.i-sanger.com/), which integrates various R language packages for microbial community analysis. All sequences have been submitted to the Sequence Read Archive under the BioProject accession numbers SAMN09011655 to SAMN09011709.

2.4.2 Quantitative PCR

Copy numbers of the functional genes (mcrA) were determined by real-time polymerase chain reaction (PCR) using an iCycleriQ 5 thermocycler (Bio-Rad, CA, USA). The PCR primers were MLF (5′GGTGGTGTMGGATTCACACARTAYGCWACAGC3′) and MLR (TTCATTGCRTAGTTWGGRTAGTT). To optimise the real-time PCR reaction system, some DNA extracts were diluted 100-fold or tenfold and used as a template. Serial plasmid dilutions of the respective functional genes (2.3 × 103 to 2.3 × 108 per action, r2 = 0.99) were employed as standards. All of the template DNA and standard samples were conducted in triplicate. The 20 μL reaction mixtures included 1 μL of template DNA, 12.0 μL of SYBR Premix Ex Taq (Takara BioInc, Shiga, Japan), and 500 nM of each primer. All PCR runs began with an initial denaturation at 95 °C for 5 min, 94 °C for 1 min, 55 °C for 1 min, and 72 °C for 30 s. This was followed by a melting curve analysis from 65 to 98 °C at 0.2 °C per reading with a 6-s hold time. Fluorescence was read during each cycle at 83 °C.

3 Results

3.1 Carbon content and carbon isotopic signature

The δ13CCH4 value (Vienna Pee Dee Belemnite, VPDB) decreased from −50.0‰ at the sediment surface to less than −68.1‰ at a depth of 11 cm, and then remained relatively constant between −68.1 and −70.2‰ below 12 cm (Fig. 1A). The values of δ13CCO2 (VPDB) and δ13CDIC (VPDB) varied proportionally and inversely with respect to the δ13CCH4 values. The δ13CCO2 and δ13CDIC values were nearly constant for the first 3 cm (−16‰ and −8‰, respectively), and subsequently increased below a depth of 12 cm to relatively constant values of between −1.15 and −2.68‰ for δ13CCO2, and between 2.70 and 6.40‰ for δ13CDIC.

Fig. 1
figure 1

Profiles of methane and carbon isotopic compositions of methane, CO2, and DIC (VPDB) (A), nitrite (B), sulfate (C), iron (D), and TOC (E) in the sediment porewater in the Hongfeng Reservoir. DIC and TOC stand for dissolved inorganic carbon and total organic carbon. The squares in B, C, and D outline the inferred AOM zones discussed in the zonation of AOM

Full size image

The TOC content (Fig. 1E) fluctuated between from 1.3 and 4 wt%, with a peak of 3.5–4.0 wt% at a depth of 10–17 cm. The TOC content decreased with two steps above and below the peak. The TOC content in the 3–8 cm layer was comparable to that in the 19–22 cm layer (2.2–2.3 wt%). The TOC content in the first 2 cm and 23–24 cm layer was 1.9 wt% and 2.2 wt%, respectively, and was relatively constant down the core at ~1.5 wt%.

The concentration profiles of methane, nitrate, iron, and sulfate are shown in Fig. 1B–E. The methane concentration increased from the top surface (0.054 mM) to the 10–11 cm layer (1.5 mM), and then declined quickly to 0.040 mM at a depth of 13 cm, it subsequently increased to 0.94 mM at a depth of 17 cm, and fluctuated between 0.025 and 0.58 mM below 18 cm (Fig. 1B). The nitrite concentration decreased from 10 μM at the surface to 5 μM at a depth of 5 cm, and then increased to 40 μM at a depth of 14 cm. It subsequently reduced to 16–23 μM at a depth of 16–24 cm before remaining relatively constant down the remaining core (Fig. 1B). The sulfate concentration decreased from 0.6 mM at the surface to less than 0.1 mM at 10–12 cm, which was followed by a high value at ~13 cm and then relatively constant values of 0.04–0.1 mM down the remaining core (Fig. 1C). The iron concentration declined from 44 μM at the surface to 24 μM at a depth of 4 cm, and was then relatively constant at ~10 μM down the rest of the core (Fig. 1D).

3.2 Incubation results for methanogens and anaerobic methanotrophs

The incubation results for the methane production potential both without and with methanogenic substrate (i.e. sodium acetate), and the consumption potential are shown in Fig. 2A–C. The methane production rate was comparatively high (0.03–3.94 µg CH4 g−1 sediment d−1) in sediments at a depth of 0–13 cm, and decreased along the core with the exception of that at ~20 cm (0.19 µg CH4 g−1 sediment d−1) (Fig. 2A). Moreover, the methane production rate increased dramatically in the treatment that included sodium acetate, especially at a depth of > 10 cm (Fig. 2B).

Fig. 2
figure 2

The methane production rates along the sediment profile incubated without substrate (A) and with acetate (B), and the anaerobic methane oxidation rates along the sediment profile (C) in the Hongfeng Reservoir

Full size image

Because all of the sediments showed a high potential of AOM during the first 2 months of incubation, the following discussion based on this period (Fig. 2C). An extraordinarily high value (6.47 µg CH4 g−1 sediment d−1) was measured at a depth of ~18 cm, followed by a fast decline. Another active AOM zone was observed at a depth of 5–10 cm (1.37–2.14 µg CH4 g−1 sediment d−1).

3.3 Potential methanogen species and functional species involved in AOM

Functional molecular genetic markers that are specific to methanogens were quantified by real-time qPCR (Fig. 3A). The mcrA gene copy numbers were relatively constant throughout the core, although there were two peaks at 12 cm (6.72 × 107 copies g−1 sediment) and 22 cm (4.92 × 107 copies g−1 sediment).

Fig. 3
figure 3

The abundance of mcrA gene (A) and the relative abundance (B) of dominate methanogens at order level detected along the sediment profile in the Hongfeng Reservoir

Full size image

The archaea community in the sediments was dominated by Euryarchaeota and Bathyarchaeota, which collectively accounted for 76.0–85.1% of the total community (Fig. S2B). The main methanogens in the sediments were Methanosarcinales (3.09–29.18%), Methanomicrobiales (0.55–14.23%), and Methanobacteriales (0.69–8.23%) at the order level (Fig. 3B; Fig. S2). Interestingly, Methanobacteriales was mainly distributed at a depth of 6–18 cm (1.17–8.23%), whereas its relative abundance in the surface layer (1–5 cm) was < 0.78% and disappeared below 18 cm (< 0.01%). The abundance of Methanomicrobiales in the sediments was slightly higher above a depth of 15 cm (3.86–14.23%) than of that below 16 cm (0.55–4.79%).

The Illumina sequencing results showed diverse and abundant functional groups involved in NO3−/NO2−, SO42−, and metal-dependent anaerobic methane oxidation (Fig. 4; Fig. S3). The abundance of typical nitrite-dependent anaerobic methane oxidation bacteria (Candidatus Methylomirabilis oxyfera) was significantly higher below 18 cm (40.2–48.7%) than at shallower depths (0–17 cm, 31.7–38.2%) (Fig. 4B). Nitrate-dependent AOM was performed by ANME-2d, and which was mostly prevalent in the upper layers (0–19 cm, 1.6–6.5%), and that the relative abundance varied in a narrow range of 0.60–0.98% below a depth of 20 cm (Fig. 4B). The abundance of ANME-2d in the sediments was higher above a depth of 18 cm (0.640–2.23%) in comparison to below 18 cm (0.23%; Fig. 4B). For the archaeal anaerobic methanotrophs in the sediment, the groups of ANME-1a/b, ANME-3 were observed along the sediment profile (Fig. 4A). The relative abundance of ANME-1a/b was higher than ANME-3 above a depth of 18 cm, but this tendency was totally reversed below the depth of 20 cm. The relative abundance of ANME-1a varied in a narrow range of 0–0.081%, with the exception of a high value of 0.193% at 2 cm (Fig. 4A). The ANME-1b was increased with the depth in the surface layers (0–3 cm), and peaked at 9–18 cm (0.096–0.324%). The abundance of ANME-3 was similarly increased in the upper 3 cm (up to 0.147%) in the sediments, which was higher in the depth of 13–28 cm (0.076–0.162%) in comparison to other layers (< 0.06%; Fig. 4A).

Fig. 4
figure 4

The relative abundance of ANMEs (A), M.oxyfera and ANME-2d (B) along the sediment profile in the Hongfeng Reservoir

Full size image

4 Discussion

4.1 Intersecting of methane production and consumption

The opposite trends of the porewater profiles of the δ13CCH4 value versus the δ13CCO2 and δ13CDIC (Fig. 1A) indicated a typical methane cycle zonation, with a clear zone of methane oxidation dominated the upper sediments (< 10 cm), and methane production prevailed in the lower sediments (> 10 cm). When plotted against the logarithmic methane concentrations (Fig. S4), and after eliminating the residual heavy methane, a linear trend of the δ13CCH4 value suggests that methane was oxidised in the sediments, whereby the corresponding δ13CDIC were more positive (−8.6 to > 3‰ VPDB). The aerobic consumption of methane obviously occurred in the very top surface (< 1 cm) (Fig. S1) because oxygen can penetrate (< 8 mm) to this depth, even in an oligotrophic lake (Melton et al. 2014). The sediments became anaerobic while oxygen decreased rapidly with increasing sediment depth, which led to a shift in methane consumption to anaerobic oxidation. As Lloyd et al. (2011), Norði et al. (2013), and Riedinger et al. (2014) reported, the symmetrical concentration profiles of methane versus sulfate, iron, and nitrate/nitrite (Fig. 1B–D) suggest that methane oxidation may have been coupled with sulfate (< 10 cm), iron (< 5 cm) and nitrite separately (12–24 cm). There may be the impact of organic matter (humic substance) especially on the surface sediment, but we were unable to completely rule out it in this study.

The lab incubation data demonstrated the high potential of both anaerobic methane production and oxidation along the entire depth of the sediment core (Fig. 5) (Heuer et al. 2010; Norði et al. 2013; Norði and Thamdrup 2014; Mach et al. 2015; Rissanen et al. 2021a, b; Yang et al. 2021). Interestingly, the methane production rate was obviously higher at a depth of 0–13 cm than that at ≥ 14 cm (Fig. 2A), and only when the extra substrate was added did the production rates increase dramatically (> 10 cm) to exceed those in the surface sediment (Fig. 2B). In accordance with the geochemical data, an extremely high activity of methane oxidation was identified at ~18 cm (Fig. 2C). Another active AOM zone was observed at 5–10 cm, which agreed with the changes in the electron acceptors (Fig. 1B−D). Therefore, methane production and consumption multiplied along the sediment core, and the consumption potential was much higher than production with or without additional substrate (Su et al. 2019).

Fig. 5
figure 5

Schematic representation of methane production and oxidation along the sediment core profile in the Hongfeng Reservoir. The relative microbial activity of functional groups, involved in the methanogenesis and methanotrophs, is indicated by color. The AOM process is driven by anaerobic methane oxidising bacteria (ANME), sulfate reducing bacteria (SRB), and M. oxyfera

Full size image

To sum up, the coexistence methane production and oxidation was proven by the two distinct production zones that were parallel to two evident AOM zones (Figs. 2 and 5), rather than the stereotypical pattern of oxidation/production along the redox gradient.

4.2 Zonation of methane production and limitation

4.2.1 Abundant methanogens along the sediments core

Abundance functional gene of methanogenic (methyl-coenzyme M reductase, mcrA gene) existed along the sediment core profile (Fig. 3A). Within the top 10 cm, higher abundance of the mcrA gene corresponded to the more vigorous potential of methane production.

The copy numbers of the mcrA gene (Fig. 3A) matched the methane production potential with substrate addition (Fig. 2B), which indicated that methanogens were stimulated by sufficient substrate and became active. This pattern was obvious at a depth of 10–17 cm, where the high TOC content and the anaerobic environment facilitated anaerobic digestion. The metabolic production might have also furthered promoted the growth of methanogens (e.g. via a hydrogenotrophic pathway) (Chen et al. 2021). Therefore, the activity of methanogenesis was constrained by the bioavailable substrate rather than by the abundance of methanogens.

4.2.2 Niche partitioning of various nutrient types of methanogens along the sediment core

The vertical distribution of various nutrient types of methanogens was revealed along the sediment core profile (Fig. 3B). Correspondingly, Methanomicrobiales and Methanobacteriales were mainly distributed within the upper ~20 cm. These hydrogenotrophic methanogens are able to produce methane using compounds containing H2, CO2, formic acid, alcohol, and propanol as energy and carbon sources (Euler et al. 2020). Therefore, hydrogenotrophic might have been an active methanogenic pathway in the upper layers. The relative abundance of Methanobacteriales peaked at 10–17 cm, and then disappeared below 18 cm (< 0.01%), which demonstrates the critical role of Methanobacteriales in such a fermentation layer.

The dominant methanogens were Methanosarcinales, which mainly consist of Methanothrix at the genus level, and are typical aceticlastic methanogens. When supplemented with sodium acetate, Methanothrix recovered the activity of methane production in the deeper layers. Methanothrix is an obligate aceticlastic methanogenic archaeon that can adapt to low concentrations of acetic acid of 7–70 μmol L−1 (Westermann et al. 1989) due to acetyl-CoA synthetase with a high affinity to acetic acid (Jetten et al. 1990). In general, there is a large amount of nutrients and organic matter sequestrated in the sediments of reservoirs (e.g. sodium acetate) (Quadra et al. 2020). Hence, aceticlastic methanogens dominate and play a critical role in methane production in deep sediments (Scholten and Stams 2000).

4.3 Zonation of methane oxidation and metabolism

A complex oxidation pattern was discovered in the sediment cores. With the exception of the top surface layer (< 1 cm) that may have been oxidised by oxygen, methane consumption in the deeper layers was coupled with SO42−-AOM, metal-AOM, and NO3−/NO2−-AOM (Figs. 1, 2, and 4; Fig. S3). Mayr et al. (2020) recently found such a niche partitioning of the taxa of different methane oxidisers in four lake sediments, which effectively mediated methane oxidation along with the oxygen–methane counter gradient. The functional groups involved in these processes include the consortia of anaerobic methanotrophic archaea, particular sulfate reducing bacteria (Desulfobulbus) (Bhattarai et al. 2019), ferric iron reducing bacteria (Geobacter) (Gao et al. 2017) and NC10 (Lee et al. 2018).

4.3.1 SO4 2–AOM and metal–AOM in the subsurface layer

The profiles of oxidants and functional groups indicate that sulfate was the primary oxidant for AOM in the surface layer (< 10 cm) and at ~13 cm, and subsequently drove the apparent iron oxides to mediate AOM (< 5 cm), which is similar to the report of He et al. (2018). The relative abundance of ANME and Desulfobulbus in the upper layer confirmed the interdependence of sulfate reducing bacteria and AOM. The reduced sulfur may have then transferred electrons to iron oxides, which could have been further driven by microbial iron oxidation performed by the identified iron-oxidising bacteria of Ferritrophicum and Crenothrix (0–17 cm). The relative high abundance of ANME-2d was demonstrated to potentially transfer the elements by extracellular electron transfer (Oni and Friedrich 2017) and interspecific electron transfer (Cai et al. 2018). Melton et al. (2014) demonstrated that Fe oxidisers could overcome the competition pressure to survive in lake sediments resulting in a high abundance of poorly crystalline iron. Norði et al. (2013) observed that ANME-2d drived AOM in iron-rich freshwater lake sediments where sulfate and Fe(III) coexisted. Unfortunately, although Cai et al. (2018) reported the enrichment and characterisation of a novel archaeon Candidatus “Methanoperedens ferrireducens” which couples anaerobic oxidation of methane to Fe(III) reduction, the responsible microorganisms for metal-AOM are still difficult to define. The potential of iron-AOM is uncertain in our study due to the preferential sulfate reduction and undetected iron minerals.

4.3.2 denitrification–AOM in the middle layer

Both of the incubation activity (Fig. 2C) and carbon isotope composition (Fig. 1A) of the sediments indicated the presence of an AOM zone in the intermediate layer of the sediments (18–20 cm). Correspondingly, abundant of nitrate- (Can. Methanoperedens and ANME-2d) and nitrite-dependent (NC10, M. oxyfera) methane anaerobic-oxidation microorganisms (Kurth et al. 2019) were concurrence at the depth of 18–20 cm. In a strictly anoxic environment, the coexistence of CH4, NO2−/NO3− drove the AOM (Lee et al. 2018). ANME-2d archaea can independently complete the nitrate-driven anaerobic oxidation reaction of methane without the participation of other microorganisms, which can reduce nitrate to nitrite while oxidising methane (Haroon et al. 2013). Furthermore, ANME-2d archaea has been found in many reactors to be coexist with M. oxyfera bacteria (Hu et al. 2015; Lomakina et al. 2019). Recently, Nie et al. (2021) found that ANME-2d is syntrophic with NC10 via microbial metabolites exchange within consortia for simultaneous nitrate- and sulfate-dependent AOM. In collaboration with sulfate-reducing bacteria (SRB) and iron reducing bacteria, active ANME-2d would provide a connection between the carbon, nitrogen, iron and sulfur cycles occurring in freshwater environments. Therefore, the anaerobic zone of freshwater sediments is an ideal habitat for the N-DAMO reaction as reported in lake ecosystems (Deutzmann and Schink 2011; Mayr et al. 2020).

5 Conclusion

An interaction intersecting zonation of methane production and oxidation has been revealed in the freshwater sediments based on the porewater concentration of methane, carbon isotopic composition, the incubation activity, and the abundance of functional genes. That is, two distinct AOM zones were concurrent with two production zones. The availability of substrate controls the production ability more than abundance and diversity of methanogens, simultaneously, and the abundance and activity of functional microorganisms of AOM are critical for quantifying the aquatic methane efflux from such environments. This study sheds new light on the metabolism and mechanism behind the biogeochemical cycles for carbon, sulfur, nitrogen, and metals in reservoir sediments. Furthermore, the critical role of AOM in mitigating methane release from the reservoir has been confirmed, which can help to better remedy the widespread methane emissions from freshwater sediments.