Temperature Response of Aerobic Methane-Oxidizing Bacteria in Lake Sediments from King George Island, Maritime Antarctica

Abstract

Methane (CH₄) is a major greenhouse gas that contributes substantially to global warming. The release of biogenic CH₄ into the atmosphere is a critical factor in global climate change and can be enhanced by increasing temperature. Over the past 50 years, Maritime Antarctica has been among the most rapidly warming regions of the planet. Methane-oxidizing bacteria (MOB) can oxidize a substantial fraction of the CH₄ produced by methanogenic archaea before it reaches the atmosphere. However, a major knowledge gap in the global CH₄ cycle in the Southern Ocean and Antarctica concerns its biological consumption by MOB, which act as an important biological sink. Although temperature is known to strongly influence CH₄ oxidation rates, its effects on the structure of MOB communities and their associated phospholipid fatty acid (PLFA) profiles in polar environments remain poorly understood. In this study, the effect of temperature on the structure of the active community of aerobic MOB in the sediment of a lake on Fildes Peninsula in Maritime Antarctica was investigated using stable isotope probing of phospholipid fatty acids (PLFA-SIP) and 16S rRNA gene amplicon sequencing. Differential abundance analysis of microcosms incubated at 5 and 20 °C for 20 and 40 days showed Methylobacter and Crenothrix as the main MOB at both temperatures, while PLFA C_16:1ω7c and C_16:1ω5c, biomarkers of gammaproteobacterial MOB, increased their concentration. The rise in temperature from 5 to 20 °C decreased the diversity of the MOB community, suggesting certain vulnerability due to lack of redundancy of function. This study provides new insights into the impact of temperature on the structure of MOB and the total bacterial community in a polar lake system.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00248-026-02717-2.

Introduction

Methane (CH₄) is a major greenhouse gas that contributes substantially to global warming. The release of biogenic CH₄ into the atmosphere is a critical factor in global climate change and can be enhanced by increasing temperature [1]. Over the past 50 years, Maritime Antarctica has been among the most rapidly warming regions on the planet [2, 3].

Methane production by methanogenic archaea has been reported in Antarctic ecosystems, particularly in lake sediments and soils across Maritime Antarctica [4–8]. However, a major knowledge gap in the global CH₄ cycle in the Southern Ocean and Antarctica concerns its biological consumption by methane-oxidizing bacteria (MOB), which act as an important biological sink [9]. In many ecosystems, including wetlands and aquatic sediments, MOB can oxidize a substantial fraction of the CH₄ produced by methanogenic archaea before it reaches the atmosphere. This process can occur under both oxic and anoxic conditions, the latter involving alternative electron acceptors such as sulfate, nitrite, nitrate, iron, or manganese. The activity of MOB is modulated by environmental factors including CH₄ and oxygen availability, the presence of alternative electron acceptors, temperature, pH, and substrate quality, all of which influence their growth rates and oxidation efficiency [10–12]. Anaerobic oxidation of CH₄ is performed by methanotrophic archaea or bacteria, whereas aerobic oxidation is primarily carried out by bacteria belonging to the phyla Pseudomonadota and Verrucomicrobiota. Recently, a candidate aerobic methanotrophic species has also been described within the phylum Actinomycetota [13]. Within Pseudomonadota, gammaproteobacterial MOB are mainly represented by the families Methylococcaceae and Crenotrichaceae and are classified as type I MOB. This classification distinguishes gammaproteobacterial MOB (type I) from alphaproteobacterial MOB (type II) based on physiological and biochemical characteristics, including phospholipid fatty acid (PLFA) profiles [10, 11]. Although traditionally considered obligately aerobic, several methanotrophic bacteria, particularly members of the class Gammaproteobacteria, have been shown to be active under apparently anoxic conditions in lake sediments [14–18].

Stable isotope probing of phospholipid fatty acids (PLFA-SIP) is a valuable approach for characterizing biologically active MOB communities in diverse environments. This method relies on the detection of isotopically labeled PLFA [19] and is notable for its high sensitivity, making it a powerful tool for studying MOB activity and diversity [20]. However, due to the limited availability of reference PLFA profiles for MOB, particularly from cold environments such as Antarctica, it remains challenging to assign specific PLFA signatures to individual taxa. This limitation can be mitigated by combining PLFA-SIP with 16S rRNA gene amplicon sequencing, which provides complementary information on community structure and taxonomic composition under varying environmental conditions [18, 21]. Some species of MOB possess distinctive PLFA profiles that allow differentiation between major methanotrophic groups. In particular, type I MOB are characterized by fatty acids (FA) such as C_16:1ω6c, C_16:1ω5c,t and C_18:1ω9c, whereas type II MOB typically contain C_16:1ω7t, C_18:1ω9t, C_18:1ω8c, C_18:2ω7c,12c and C_18:2ω6c,12c. These PLFA signatures are therefore useful markers for distinguishing between these groups [20]. Within the genus Methylobacter, the predominant FA is C_16:1ω7c, accompanied by lower proportions of C_14:0, C_16:1ω6c, C_16:1ω5c and C_16:0. The absence of C_16:1ω8c and the relatively low abundance of C_14:0 and C_16:0 further distinguish Methylobacter from other type I MOB [22].

Rising temperature accelerate CH₄ emissions, creating a positive feedback loop that further exacerbates global warming [23]. Although temperature is known to strongly influence CH₄ oxidation rates, its effects on the structure of MOB communities and their associated PLFA profiles in polar environments remain poorly understood [24–26]. In Maritime Antarctica, research on the response of MOB to temperature change is particularly limited due to the region’s extreme and highly variable environmental conditions. Consequently, further investigations are needed to assess how climate-driven warming may alter the structure and functioning of these microbial communities and their role in the CH₄ cycle. Our previous studies have shown that increasing temperatures significantly enhance microbial metabolic rates, with implications for CH₄ oxidation efficiency [27]. We also reported that aerobic MOB affiliated with Methylobacter clade 2 (type I, family Methylococcaceae) dominate many Antarctic lake sediments. Despite these advances, the response of biologically active MOB communities to temperature increases in Antarctic lake sediments remains insufficiently resolved, particularly with respect to changes in diversity, community structure, and lipid biomarker composition. We hypothesized that increasing temperature would (i) alter the structure of the active MOB community, leading to reduced diversity and the dominance of a limited number of taxa, and (ii) induce changes in MOB-associated phospholipid fatty acid (PLFA) profiles reflecting physiological and/or compositional shifts. To test these hypotheses, we investigated the effect of temperature on the biologically active MOB community in a lake sediment from the Fildes Peninsula (Maritime Antarctica) using a combination of PLFA-SIP and high-throughput 16S rRNA gene amplicon sequencing.

Materials and Methods

Study Site and Sample Collection

The Fildes Peninsula is located in the southwest of King George Island (62°08’-62°14’S and 59°02’-58°51’W), in the South Shetland archipelago, Maritime Antarctica. The sample was collected on February 2020 at 3.0 m depth from Lake C-2020 (62°13′16’’S, 58°57′26’’W) (Fig. 1). Surface sediments (upper 10 cm) were collected by scuba divers in sterile polyethylene containers and refrigerated (4 °C) until processed at the laboratory. The field campaign was supported by the Instituto Antártico Uruguayo (IAU) in the Artigas Antarctic Scientific Base (BCAA) in King George Island. See Table S1 for the physicochemical characteristics of this sample.

Fig. 1.

Fildes Peninsula map (a) showing the Antarctic stations of different countries (yellow circles) and the studied lake C-2020 (red symbol) and illustrative photograph of the sampling site (b). The map was modified from PANGAEA (Data Publisher for Earth and Environmental Science, Bremen, Germany. hdl:10013/epic.44322)

Physicochemical Analyses

Temperature and dissolved oxygen at the bottom water column near the sediment were measured in situ with a HOBO Dissolved Oxygen Logger U26-001 (Onset, USA). pH, and conductivity of the porewater of the sediment samples were measured with an Oakton waterproof pH/CON 450 Meter (Oakton Instruments). Inorganic anions and organic acids of low molecular weight were determined by ion chromatography (IC), Briefly 1 g of the sediment sample was sonicated (3 × 1 min cycles), diluted in 10 mL of deionized water, filtered (22 µm GFF filter), the supernatant was analyzed in a Metrohm 861 Advanced compact ion chromatograph (Metrohm AG, Herisau, Switzerland) with a Metrosep A sup 7–250 column and 3.6 mM sodium carbonate (NaCO3) as eluent. The content of total nitrogen (TN %) and total organic carbon (TOC %) was measured with an elemental analyzer (HT Flash, Thermo Fisher Scientific, Waltham Massachusetts, USA) as described previously [28].

PLFA-SIP Assay

Five grams of sediments were diluted at 50% (w/v) with 5 mL of lake water in 120 mL amber glass serum vials under aerobic conditions and sealed with butyl rubber stoppers (110 mL of head space). Subsequently, 1% (v/v) of a mixture 2:98 (v/v) of ¹³CH₄ (99% 13C, Sigma-Aldrich, cat. no. 490229-1L) and ¹²CH₄ (99,995% Linde Uruguay), respectively, were added in the head space. The experiment involved triplicate microcosms at two temperatures, 5 and 20 °C, and two incubation periods, 20 and 40 days (Fig. S1, Table S2). Similarly, blank vials were prepared under the same conditions, but autoclaved (30 min, 121 °C) before CH₄ addition to account for non-biological CH₄ oxidation. These vials showed no CH₄ consumption.

We selected these temperatures for the PLFA-SIP assay based on our previous experiment [27]. We chose 5 °C because it was the temperature to which these microorganisms were normally exposed, and 20 °C was a temperature at which the oxidation potential of CH₄ increased. At 20 days, a set of vials was refed, air injected and incubated another 20 days. Simultaneously, as controls, another set of 3 vials was prepared similarly but with 1% (v/v) ¹²CH₄ as substrate (no ¹³CH₄ added). Methane consumption was quantified in the gas phase by periodical sampling 0.5 mL and subsequent analysis on a GC-2014 gas chromatograph (Shimadzu Scientific) equipped with a Porapak-Q column (length 6.0 ft, 1/8 in. OD, 2.1 mm ID), column temperature 55 °C, detector temperature 140 °C, nitrogen as carrier gas (30 mL min⁻¹) and a flame ionization detector.

Fatty Acids Extraction and Analysis

Extraction of free FAs was performed according to the combination of methods described elsewhere [28, 29]. Lyophilized sediments (~ 5 g) were spiked with an internal standard (myristic acid-D27) and then extracted with a mixture of dichloromethane and methanol (DCM:/MeOH, 3:1, v/v) by ultrasound sonication (3 × 10 min cycles). The total lipid extract of each sample was concentrated to ~ 0.5 mL using rotary evaporation and then digested overnight at room temperature in a mixture of methanolic potassium (6% w/w) then separated into neutral and acidic fractions. The neutral lipid fraction was obtained by extracting the methanolic potassium mixture with 30 mL (three times) of n-hexane (Hx), rotavaporating, and recovering it with ~ 1 mL of Hx:DCM (9:1, v/v). The acidic lipid fraction was obtained by adding HCl (until a pH of 2) to the remaining methanolic potassium mixture and extracting it with 30 mL of Hx (three times), then it was concentrated using rotary evaporation and collected with DCM. The acid fraction was analyzed by gas chromatography-mass spectrometry (GC–MS), by injection after derivatization with BF3 in MeOH to form fatty acid methyl esters (FAME).

GC–MS analysis was performed using a 8860 GC system coupled to a 5977B MSD with a single axis detector (Agilent Technologies) operating with electron ionization at 70 eV and scanning from m/z 50 to 650. 1 µl of analytes were injected and then separated on an HP-5MS column (30 m × 0.25 mm i.d. × 0.25 µm film thickness) with He as a carrier gas at a constant flow of 1.1 mL.min⁻¹. For analyzing the acidic fraction, the oven temperature was programmed from 70 ºC to 130 ºC at 20 ºC.min⁻¹ and then to 300 °C at 10 °C.min⁻¹ (held 15 min). The injector temperature was set at 290 ºC, the transfer line was at 300 °C, and the MS source was at 240 °C. Compounds identification was based on the comparison of mass spectra with reference materials, and their quantification on the use of external calibration curves of FAME (C₈ to C₂₄). All chemicals and standards were supplied by Sigma Aldrich (San Luis, Missouri, USA). The recovery of the internal standards averaged 81 ± 11%.

Compound-Specific Isotope Analysis of Fatty Acids

The carbon isotopic composition of individual FAs in the sediment samples was determined by coupling a gas chromatograph-mass spectrometer (Trace GC 1310 ultra/ISQ QD MS) with an isotope-ratio mass spectrometry system (MAT 253 IRMS, Thermo Fisher Scientific). For the GC analysis, the oven temperature program was set to gradually increase from 70 ºC to 130 ºC at 20 ºC.min⁻¹, and to 300 ºC at 10 ºC.min⁻¹ (held for 15 min). The parameters for IRMS analysis were as follows: electron ionization at 100 eV, Faraday cup collectors m/z 44, 45, and 46, and temperature of the CuO/NiO combustion interface at 1000 ºC. The samples were injected in a PTV injector in splitless mode, with an inlet temperature of 50 ºC, and then a ramped temperature of 2.5 ºC.sec⁻¹ until 320 ºC (held 2.5 min), using helium as carrier gas at a constant flow of 1.1 mL.min⁻¹. The carbon isotopic ratio of the individual FAs separated by GC was measured using carbon dioxide spikes of known isotopic composition, introduced directly into the MS source, three times at the beginning and end of every run. A reference mixture (Indiana University, United States) of known isotopic compositions of FAs (F8-3) was run every three samples to check the accuracy of the isotopic ratio determined by the GCMS-IRMS. For the alkanoic acids, the δ¹³C data were calculated from the FAME values, correcting them for the one carbon atom added in the methanolysis [30]. The stable-carbon isotopic composition is expressed as the δ¹³C ratio, in per mile notation (‰), and relative to the Pee Dee Beelemite (PDB) standard.

DNA Extraction, Amplification, Sequencing, and Taxonomic Assignment

DNA extraction for each replicate was carried out from 0.25 g homogenized samples with DNeasy PowerSoil kit (Qiagen, USA). Amplification, sequencing, and quality were performed at Novogene Corporation Inc. (https://en.novogene.com/, California, USA), using the Illumina MiSeq sequencing platform, with the following primers: 341F-CCTAYGGGRBGCASCAG and 806R- GGACTACNNGGGTATCTAAT, targeting the variable region V3-V4 of the 16S rRNA gene for Bacteria. For analysis of the 16S rRNA gene sequences, the DADA2 pipeline [31] was employed to obtain the amplicon sequence variants (ASVs) as described in a previous work [32]. Briefly, raw reads were quality-filtered using DADA2 (R package). Filtering was performed with filterAndTrim, discarding reads containing ambiguous bases (maxN = 0) and applying an expected-error threshold of 2 for forward and reverse reads (maxEE = c(2,2)). Reads were truncated at 240 bp for forward reads and 220 bp for reverse reads (truncLen = 240, 220), and truncation was triggered at the first base with a quality score ≤ 2 (truncQ = 2). PhiX reads were removed during filtering (rm.phix = TRUE). No 5′ trimming was applied (trimLeft = 0). The forward and reverse reads acquired were merged, and chimeras were removed. Finally, taxonomic assignments of 16S rRNA ASVs were performed based on the SILVA database (v138.1) [33].

Data Analysis

Sequence analysis was conducted in R using the Phyloseq package [34] on the dataset obtained through taxonomic assignment and the visualization and graphic design of the obtained data were executed using the ggplot2 package [35]. Sequences affiliated with mitochondria, chloroplasts, Archaea, and Eukarya were excluded. Subsequently, read counts were transformed into relative abundance. For MOB, ASVs that were defined based on their taxonomic assignment using the SILVA database are selected from the OTU table for further analysis. The alpha diversity of the total bacterial and MOB community was described by the number of ASVs observed, and the Simpson and Shannon indices. The Wilcoxon rank sum test was used to statistically compare the different conditions tested. Beta diversity was visualized using multidimensional scaling (MDS) generated with the Bray–Curtis distance metric for both communities. To identify significant differences between the groups (non-incubated samples, samples incubated at 5 °C for 20 and 40 days, and samples incubated at 20 °C for 20 and 40 days), PERMANOVA was utilized. Rarefaction curves were generated using the Vegan package [36] to assess sequencing depth.

Differential microbial abundance was determined by comparing each test condition: temperature (5 and 20 °C) and time (20 and 40 days) of incubation versus time zero (not incubated samples), using DESeq2 [37]. An alpha threshold of 0.01 was selected to determine statistical significance for each taxon. Initially, the analysis was performed at the genus level without considering ASVs that did not reach that taxonomic level. In further analysis, when a more specific taxonomic classification was not possible, ASVs were reported at the lowest taxonomic rank reliably assigned by the classifier by 'best_hit'.

To improve the taxonomic assignment of the MOB ASVs of interest, the similarity of each of the 32 ASVs assigned to MOB and the 2 additional ASVs that only reached the class level were compared with the closest type strain using the 16S-based tool ID from EzBioCloud [38]. To establish the phylogenetic relationships of the selected ASVs, a phylogenetic tree using the neighbor-joining method [39] was constructed with MEGA 7.0 software [40], utilizing Kimura's two-parameter model [41] to calculate the evolutionary distance. 16S rRNA gene sequences from characterized MOB were also incorporated. A bootstrap analysis with 1000 replications was conducted to assess the reliability of the branches [42]. For the phylogenetic tree 426 bp were used.

Results

Isotope Analysis of PLFA Biomarkers

Total FAs were quantified in the range from 192.3 (time zero, not incubated) to 587.5 µg∙g⁻¹dw (40 days) (Table 1). Saturated FAs ranged from 155.7 to 557.9 µg∙g⁻¹dw, and unsaturated ones from 9.1 to 14.0 µg∙g⁻¹dw. An increase in total FA concentrations, including C₁₆ and C₁₈ fatty acids, was observed over the incubation period at both 5 and 20 °C. The presence of C_16:1ω7c and C_16:1ω5c (known biomarkers of MOB) stands out, with concentrations between 2.3–3.6 and 0.9–1.4 µg·g⁻¹dw, respectively.

Table 1.

General composition of fatty acids from de PLFA-SIP assay and potential biomarkers of the methane-oxidizing bacteria (MOB) community relative to dry weight (µg·g⁻¹dw) from microcosm samples of the Antarctic freshwater lake sediment C-2020 (different temperatures and incubation periods). Values shown in bold indicate fatty acids considered characteristic or of taxonomic relevance for methanotrophic microorganisms

	Time zero (not incubated)	5 °C—20 days	5 °C—40 days	20 °C—20 days	20 °C—40 days
Total FAs	192.3	311.5 ± 39.9	329.2 ± 73.8	462.2 ± 83.8	587.5 ± 289.1
Saturated	155.7	282.4 ± 47.8	301.8 ± 77.3	435.6 ± 76.6	557.9 ± 290.4
Unsaturated	14.0	10.3 ± 4.8	9.1 ± 2.9	12.1 ± 3.4	12.1 ± 4.2
Branched	20.3	16.9 ± 5.2	15.8 ± 2.7	12.6 ± 3.3	14.7 ± 3.5
C14-C18	69.4	95.6 ± 4.4	88.2 ± 12.0	98.6 ± 18.1	105.2 ± 14.3
C16	32.5	52.5 ± 3.2	46.6 ± 8.8	57.2 ± 8.9	59.4 ± 12.3
C16:1ω7c	2.3	3.1 ± 2.3	2.6 ± 1.5	3.6 ± 1.5	2.9 ± 1.3
C16:1ω5c	1.4	1.3 ± 0.3	1.3 ± 0.2	0.9 ± 0.3	1.2 ± 0.2
C18	17.1	24.6 ± 1.3	23.9 ± 3.1	26.6 ± 7.1	30.1 ± 6.0

The specific isotopic composition of FAs varied widely comparing ¹²CH₄ and ¹³CH₄ microcosm samples (Fig. 2). In general, the variation of δ¹³C was higher with FAs C_16:1ω7c (-25.0 to 26.5‰), C_16:1ω5c (-29.6 to 33.2‰), and C_18:0 3-OH (not detected to 51.5‰). At 5 °C the highest differences were observed in the ¹³CH₄ microcosm samples which reached values for the FAs C_16:1ω7c, C_16:1ω5c, and C_18:0 3-OH of 9.4, 15.1, and 16.2‰, respectively at 20 days, and increased to 26.5, 33.2 and 51.5‰, respectively at 40 days. At 20 °C and 20 days, the FAs C_16:1ω7c, C_16:1ω5c and C_18:0 3-OH reached values of -10.3 and -10.8‰ and 0 (not detected), respectively, and increased to 7.4, 18.5 and 2.0‰, respectively at 40 days.

Fig. 2.

Compound-specific isotopic composition (δ¹³C) of fatty acids in the ¹³CH₄ PLFA-SIP microcosm assays. The particularly divergent δ¹³C values at both incubation temperatures, 5 and 20 °C (A and B, respectively), are highlighted with grey ellipses where the δ¹³C enriched in fatty acids C_16:1ω7c, C_16:1ω5c, and C_18:0 3-OH are identified in incubated samples with ¹³CH₄ at time zero 0 (T₀, not incubated), and after 20 and 40 days

Furthermore, at 40 days and 20 °C a slight increase in the FA C_{18:1 ω7c} was verified, reaching a value of -12.4‰, compared to 5 °C and the control (-18.1‰ and -21.2, respectively). Incubations with ¹²CH₄ as substrate at 5 and 20 °C showed δ¹³C values close to time zero (T₀, not incubated), between -29.6 and -11.3‰ in incubation at 5 °C and between -26.1 and -15.6‰ in incubation at 20 °C. No considerable variations in δ¹³C were observed in control samples fed with only ¹²CH₄, which guaranteed low or no contamination with natural ¹³CH₄, and serves as a reference for the identification of the enriched fractions.

Structure of the MOB Community with Increasing Temperature

For the 16S rRNA gene, a total of 727,955 sequences grouped in 1,088 ASVs were obtained, of which 69,074 sequences were grouped in 32 ASVs associated with MOB (Table S3). Rarefaction curves showed that the plateau phase was reached for all the lake sediments, with asymptotic trends, indicating that the sequencing depth was sufficient to identify the minimum number of reads obtained per sample (Fig. S2). Alpha diversity indices were estimated to assess bacterial community diversity. Taxonomic richness was evaluated as the number of observed ASVs, while diversity and evenness were assessed using the Shannon and Simpson indices, respectively. Differences in Shannon and Simpson indices among microcosm groups—defined as non-incubated controls and microcosms incubated at 5 °C and 20 °C for 20 and 40 days—were statistically significant, whereas no significant differences were observed in taxonomic richness (Fig. 3A, Table S4). Regarding Shannon diversity, the greatest change was observed in microcosms incubated at 20 °C for 40 days, with median values of 3.96 and a mean of 3.97 ± 0.16, compared to the non-incubated control, which showed median and mean values of 5.19 and 5.07 ± 0.22, respectively. Similarly, regarding Simpson's uniformity, incubations at 20 °C for 40 days showed a decrease in the uniformity of the distribution of ASVs, with median values of 0.91 and an average of 0.91 ± 0.01, where for both incubation times, a significant decrease was identified compared to the not incubated control, which presented values of 0.98 and 0.98 ± 0.01, respectively.

Fig. 3.

Comparison of alpha and beta diversity of the total bacterial community and the methane-oxidizing community from microcosm samples of the Antarctic freshwater lake sediment C-2020 (different temperatures and incubation periods, each condition included 3 vials with ¹²CH₄ and 3 vials with ¹³CH₄ and ¹²CH₄). Boxplots of the alpha diversity indices calculated using Observed Richness, Shannon, and Simpson, were compared in the total bacterial community (A) and the MOB community (B), under different experimental conditions. Different letters indicate significant differences for each sample at different temperatures (Wilcoxon rank sum test, p < 0.05). Additionally, multidimensional scaling (MDS) was carried out using the Bray–Curtis distance for both, total bacterial community (C) and MOB community (D) under different experimental conditions

Regarding the MOB community (Fig. 3B, Table S4), no significant changes were detected in the observed richness between the different microcosms. However, Shannon diversity index showed a significant decrease at 40 days.

The Shannon index values at 5 °C (median 1.60, average of 1.59 ± 0.08), and at 20 °C (median 1.64, average 1.62 ± 0.32) were significantly lower compared to the not incubated control (median of 2.06, average 2.10 ± 0.07, Wilcoxon test, p < 0.05). Similarly, it was observed how the uniformity measured by the Simpson index showed a decrease at 40 days in the uniform distribution of ASVs at 5 °C (median 0.65, average of 0.65 ± 0.04), and at 20 °C (median 0.72, average 0.69 ± 0.13), respectively.

For the total bacterial community, the MDS obtained showed that each type of microcosms was grouped (Fig. 3C), and the differences between groups were assessed for significance using permutation analysis of variance (PERMANOVA).

The analysis revealed that the differences between the centroids for each condition tested were significant (p < 0.05), except for the comparison between not incubated (time zero) and microcosms incubated at 5 °C for 20 days. When considering only the MOB community (Fig. 3D), grouping were again observed for each type of microcosm, indicating significant differences in MOB communities in all tested conditions (p < 0.05).

The taxonomic assignment revealed 28 phyla, of which 20 had a relative abundance greater than or equal to 0.1% (Fig. S3). On average, all samples showed high abundance of Pseudomonadota (formerly Proteobacteria) (12.6–34.6%), Bacillota (formerly Firmicutes) (18.2–37.5%), Bacteroidota (18.0–27.4%), Actinobacteriota (6.3–16.1%), and lower abundance of Chloroflexota (formerly Chloroflexi), Campylobacterota, Acidobacteriota, Verrucomicrobiota, Nitrospirota, and Caldisericota. The main variations in average abundance compared to not incubated microcosms are observed in Pseudomonadota and Bacillota. In the not incubated control, the relative abundance of 12.6% and 37.5%, respectively, were identified. However, under the different conditions tested, the abundance of Pseudomonadota increased, reaching a maximum of 34.6% for microcosms at 5 °C for 40 days, while Bacillota decreased its abundance to a minimum of 18.2% also at 5 °C for 40 days. Subsequently, after incubating the samples for 40 days at 20 °C, a considerable decrease in the abundance of the phylum Campylobacterota was observed, from 11.1% to 0.8%, as well as an increase in the abundance of Actinobacteriota, from 6.3% to 16.1%. Finally, it can be seen how in the not incubated controls all the identified phyla are represented, while, depending on the incubation conditions, the disappearance of certain groups can be seen, at least for the depth of the sequencing reached. For example, Babelota (formerly Dependientae) decreases its abundance to 0% in incubations at 5 °C, while Sumerlaeota, Edwardsiibacteriota (formerly Edwardsbacteria), and Synergistota do so in incubations at 20 °C.

Regarding the MOB diversity in the not incubated microcosms (Fig. 4), on average, a relative abundance of the MOB of 3.0E-02 ± 1.1E-02% was observed of the total bacterial community. It included ASVs that only reached the level of the family Methylomonadaceae (2.5E-04 ± 1.9E-04%), and the genera Crenothrix (9.1E-03 ± 2.5E-03%) and Methylobacter (2.0E-02 ± 0.8E-02%). In the latter, some ASVs reached the rank of species, with M. tundripaludum being the last taxonomic rank reached (7.9E-04 ± 3.7E-04%). At 5 °C, an increase in the relative abundance of MOB can be observed, where for 20 and 40 days they reached an average relative abundance of 1.0E-01 ± 0.2E-01 and 2.0E-01 ± 0.3E-01%, respectively. At 40 days, the abundance of Methylomonadaceae, Crenothrix, Methylobacter, and M. tundripaludum reached values of 1.1E-03 ± 0.5E-03, 5.7E-02 ± 1.6E-02, 1.5E-01 ± 0.3E-01, and 1.1E-03 ± 2.7E-03%, respectively. At 20 °C, MOB abundance showed a non-significant increasing trend compared to the non-incubated control, reaching values of 5.0E-02 ± 2.0E-02 and 7.0E-02 ± 3.0E-02 after 20 and 40 days, respectively. The abundance of Methylomonadaceae, Crenothrix, Methylobacter, and M. tundripaludum at 40 days of incubation reached values of 2.4E-04 ± 0.6E-04, 3.8E-02 ± 2.5E-02, 3.4E-02 ± 1.4E-02, and 2.2E-03 ± 4.3E-03%, respectively.

Fig. 4.

Diversity of methane-oxidizing bacteria from microcosm samples of the Antarctic freshwater lake sediment C-2020 (different temperatures and incubation periods). Relative abundances for each condition are shown as a proportion of the total bacterial community, based on 16S rRNA gene sequence analysis. Taxonomic assignments are reported according to the best SILVA database hit and, for methane-oxidizing bacteria, at the lowest taxonomic rank reliably assigned

Differential abundance analyses revealed significant changes in certain genera (Fig. 5, Table S5). In these cases, one or more ASVs associated with a specific genus experienced variations in their abundance as a function of incubation time and temperature. At 5 °C for 20 days, an increase in the abundance of ASVs belonging to 6 different genera and a decrease in one ASV associated with Crenothrix were observed. At 40 days of incubation, a decrease in ASV abundance was noted in 4 genera and an increase in 8, with Crenothrix and Methylobacter showing the most significant changes in abundance. The greatest differences in composition were evident in incubation at 20 °C. At 20 days, 9 genera increased in abundance and 7 decreased (one associated with Crenothrix). After 40 days, it was observed that 10 genera decreased and 17 increased in abundance. It is relevant to note that the number of genera composed of multiple ASVs experienced both increasing and decreasing changes at 40 days of incubation at 20 °C.

Fig. 5.

Differential abundance analysis with DESeq2 at genus level from microcosm samples of the Antarctic freshwater lake sediment C-2020 (different temperatures and incubation periods). Bacterial genera that were statistically different (p < 0.01) between time zero (T₀, not incubated) and the different conditions tested: at 5 °C for 20 and 40 days (A and B, respectively), and at 20 °C for 20 and 40 days (C and D, respectively) are shown. Bacterial groups that were significantly enriched in the tested condition showed a positive log2FoldChange (> 0), while those that were significantly depleted showed a negative one (< 0)

These results indicate that the greatest differences in the composition of microbial diversity occur at 40 days. It was observed that regardless of the incubation temperature, members of MOB genera such as Methylobacter and Crenothrix increased their abundance. Furthermore, it was detected that abrupt changes in the incubation temperature from 5 °C to 20 °C, produced significant changes in the total bacterial community structure. In the analysis of differential abundance to the last taxonomic rank reached (Table S5), for microcosms at 5 and 20 °C, an increase in members of the class Gammaproteobacteria and Alphaproteobacteria was observed. Both classes included MOB. However, in a subsequent analysis, only 2 ASVs from Gammaproteobacteria were found to be closely related to MOB. The other ASVs considered were not associated with known alphaproteobacterial MOB.

The non-methanotrophic bacterial groups whose relative abundance decreased were the genera Clostridium, Magnetospirillum, Sulfuricurvum, Sulfurimonas, Cryptanaeobacter, Nitrospira, Phaselicystis, Caldisericum, and the Blvii28 wastewater-sludge group. Regarding the increase of non-methanotrophic bacterial groups ASVs assigned to 'Ca. Planktophila', 'Ca. Paracaedibacter', Leptolinea, Ferruginibacter, and the BSV13 group were identified at both temperatures. Only the genus 'Ca. Solibacter' significantly increased its abundance at 5 °C, while members of Polymorphobacter, Phenylobacterium, Ferriphaselus, Paludibaculum, Rhodobacterium, Paenibacillus, Longivirga, Sediminibacterium, Pseudolabrys, Sulfuriferula, and the PMMR1 group only increased at 20 °C.

Phylogeny and Taxonomic Affiliation of Identified MOB ASVs

To delve deeper into the taxonomic affiliation of the 32 MOB ASVs identified, we also used the EzBioCloud 16S-based ID tool to identify the sequences (Table S6). As can be seen, all the ASVs that were initially assigned to Crenothrix had the closest representative within the genus Methylobacter, except ASV546 and ASV1045, which had the closest similarity to Methylosoma difficile LC 2^T (94.63%) and Methylocorpusculum oleiharenae XLMV4^T (92.99%), respectively. Other ASVs that only reached the taxonomic rank of family, such as ASV643, ASV655, ASV689, ASV918, and ASV1062, which were assigned to Methylomonadaceae (now correctly and validly known as Methylococcaceae), also showed close similarity to members of the genus Methylobacter. From the analysis of differential abundance extended to the level of the last taxonomic rank reached, ASV281 and ASV420 showed an increase in their abundance and their closest similarity were Methylobacter psychrophilus Z-0021^T (93.69% and 91.59%, respectively). However, they had previously been assigned to the taxonomic rank of the class Gammaproteobacteria.

In summary, all those ASVs that showed the closest similarity to members of the Methylobacter genus showed the closest similarity to Methylobacter tundripaludum SV96^T or Methylobacter psychrophilus Z-0021^T according to EzBioCloud. ASVs that showed changes in abundance as an effect of temperature were taken into account for the construction of a phylogenetic tree. ASV5, ASV9, ASV10, ASV17, ASV44, ASV281, ASV342, ASV420, ASV546, and ASV914 were selected, along with reference strains of known MOB. The phylogenetic reconstruction (Fig. 6) shows that ASV5, ASV9, ASV10, ASV44, ASV281, ASV420, and ASV914 grouped within Methylobacter clade 2, together with Methylobacter oryzae KRF1^T, ‘Ca. Methylobacter titanis’ D1-2020 MAG004^Ts, Methylobacter svalbardensis LS7-T4A^T, M. tundripaludum SV96^T, and M. psychrophilus Z-0021^T. Within this large group, ASV281, ASV420, and ASV914 formed a small clade, which is far from the aforementioned strains. In contrast, ASV17 and ASV342 are located in a separate clade, and ASV546 forms a clade together with Crenothrix polyspora 5 and M. difficile LC 2^T. None of the ASVs studied grouped within the Methylobacter clade 1.

Fig. 6.

Neighbor-joining tree based on partial 16S rRNA gene sequences showing the taxonomic relation of the selected methane-oxidizing bacteria ASVs, which exhibited significant differences in the analysis of differential abundance. Bootstrap values (> 50%) based on 1000 re-samplings are listed at the nodes. Methylocystis parvus OBBP^T was used as an outgroup. Bar, 0.02 substitutions per nucleotide position

Discussion

As expected, the total bacterial community was markedly affected in both alpha and beta diversity only at 20 °C for 40 days. In contrast, the MOB community showed more variable alpha diversity across temperatures and incubation periods; however, the analysis of beta diversity of MOB community revealed that microcosms incubated at 20 °C for 40 days clustered separately from the remaining samples. This suggests that the strong disturbance imposed by the temperature shift on the overall bacterial community may indirectly influence specific functional groups, such as MOB [43, 44]. In line with this interpretation, Avila-Nuñez et al. (2024) [45] highlighted the importance of interactions between MOB and other microbial groups in shaping their activity.

The FAs C_16:1ω7c, C_16:1ω5c, and C_18:0 3-OH were predominantly labeled with ¹³C. The first two FAs are associated with specific microorganisms, particularly for C_16:1ω5c type I MOB, since C_16:1ω7c are present in type I and type II MOB. It is know that bacteria modify the composition of certain FAs in their cell membrane to adapt to different environmental conditions, including temperature [46, 47], especially MOB [48, 49]. At lower temperatures, they typically increase the proportion of unsaturated FAs to maintain membrane fluidity and prevent solidification [50]. This may explain the higher concentration of these FAs at 5 ºC than at 20 °C. As mention above, the presence of these FAs in conjunction with the 16S rRNA gene amplicon sequencing results suggests that Methylobacter could be the main active MOB in this sediment.

It should also be noted that the observed changes in unsaturated FAs may be due to a multi-causal effect, including changes in abundance and membrane restructuring of the MOB. An increase in the FA C_18:1ω7c compared to the not incubated controls was also observed at 20 °C. This may be due, in the case of cross-feeding, to an increase in the relative abundance of the ASVs of certain taxa identified in the samples that have this FA in their cell membranes such as Polymorphobacter, Phenylobacterium, Ferriphaselus, Paludibaculum, Rhodobacter, Longivirga, Pseudolabrys, and Sulfuriferula [51–58]. However, it can also be observed in Methylobacter [59], so it could be have been due to a turnover in membrane structuring between 5 and 20 °C.

The fatty acid C_18:0 3-OH has been associated mainly with type II MOB [60–62] and their source has been identified as the lipid A component of lipopolysaccharides. The presence of C_18:0 3-OH suggests that active type II MOB were present in the samples. However, their absence in the results of 16S rRNA gene amplicon sequencing may be due to biases in the primers used to recover these microorganisms, or there may indeed be an absence of this group in the sediments, and the type I MOB found in this Antarctic lake had this FA in their cell membranes. A similar situation was observed in Methylocystis heyeri, a type II MOB that contains C_18:1ω8c, which was previously considered to be a signature of FA of type I MOB [63].

In summary, the FAs C_16:1ω7c and C_16:1ω5c, distinctive markers of type I MOB were identified in our samples at 5 and 20 ºC as belonging to the active MOB. Regarding C_18:0 3-OH, its detection could not be clearly assigned to a specific type of MOB. These results were different from those obtained by Knoblauch et al. (2008) [64] with Siberian soils, in which raising the temperature from 0 to 22 °C produced a shift in the MOB community with increased concentration of FA C_18:1 but decreased concentration of FA C_16:1.

A priori, differences in relative abundances of MOB between 5 and 20 °C could be expected. However, when considering the increase in foldchange of MOB ASVs in the differential abundance analysis, no significant differences were evident between temperatures. Out of the 32 ASVs assigned to known MOB, only 10 of them (including 2 that were originally associated with Gammaproteobacteria), showed a significant change in their abundance at 5 and 20 °C. This result indicates that within the MOB community in these sediments, only a fraction was really active in the experimental conditions, capable of exhibiting a significant change in their abundance. Within these ASVs, not all showed an increase under the different conditions tested, indicating a different structuring of the MOB community at different temperatures, similar to what was described in northern polar lakes [26]. The ASV44 increased significantly only at 20 °C after 40 days, indicating it may represent a mesophilic MOB [65]. The presence of mesophilic MOB that can survive and thrive in prolonged cold conditions has been previously demonstrated [66]. ASV9, ASV10, ASV420, and ASV17 exhibited the same trend at 20 °C for 40 days. Nevertheless, these four ASVs also increased at 5 °C indicating these MOB were psychrophilic or psychrotrophic. ASV10 and ASV420 increased at 5 °C only at 40 days, which may be attributed to MOB with lower generation times compared to others in the community. In addition, ASV9 and ASV17 increased their abundance also at 5 °C at both incubation times. In a previous study conducted in the same geographic area [67], lake water temperature was identified as a variable that significantly affects microbial community structure. ASV281, ASV914, and ASV342 showed a significant increase at 5 °C, and only ASV5 showed a significant increase in all conditions tested. Finally, of those 10 ASVs identified as potential MOB (ASV5, ASV9, ASV10, ASV17, ASV44, ASV281, ASV342, ASV420, ASV546, and ASV914) 7 were found to be located phylogenetically close to clade 2 of the Methylobacter genus, characterized by being active members in sediments and adapting to cold environments [68, 69], and better competitors against other MOB under reduced oxygen tensions [70]. The ubiquity of clade 2 may represent its critical ecological role in mediating CH₄ fluxes in freshwater sediments and soils worldwide. This clade had also been proposed as the main responsible for CH₄ oxidation in Maritime Antarctica, but it remained to determine which active microorganisms fulfilled that role [27]. In particular, the closest described MOB for these 7 ASVs and for ASV17 and ASV342 (outside the clade) were M. tundripaludum and M. psychrophilus, MOB associated with cold environments [71, 72]. ASV546 was located further away, and phylogenetically close to M. difficile and C. polyspora, both microorganisms associated with MOB in lake sediments [73, 74]. In turn, it was the only MOB that decreased significantly at 5 and 20 °C at 20 days of incubation. This may be due to the conditions tested, such as substrate concentration, since M. difficile (the closest one) grows at low oxygen concentrations [74] or by competitive exclusion by the other MOB in the community. However, the decrease in the abundance of this ASV indicates that it is potentially active in the community but was not favored under the conditions tested. It should be noted that many of these ASVs were assigned to Crenothrix using the SILVA database. Moreover, a more detailed analysis using the EzBioCloud 16S rRNA gene database, in combination with phylogenetic reconstruction, could allow a more precise taxonomic assignment of the obtained ASVs originally classified using the SILVA database.

As was mentioned above, a greater MOB diversity represented by more ASVs was observed at 5 °C than at 20 °C. This suggests that although the function in the environment was maintained at both temperatures, the structure of the community at 20 °C is less diverse, making the ecosystem more vulnerable to changes due to a lack of functional redundancy [75]. The decrease in the abundance of those non-methanotrophic microorganisms was caused by the atmosphere, the substrate used in the experiments, and the temperature of incubation. Notably, no significant decreases in the abundance of these groups were observed at 5 °C for 20 days, a short incubation time at a temperature close to that found in situ in the lake. This incubation time allowed for better visualization of the effects of temperature, but did not provide clear values of the increase in the abundance of MOB, as well as their FAs in the PLFA-SIP assay. It was also observed that the majority of the genera increased their relative abundance at 20 °C, indicating that temperature played a significant role in the observed differences between conditions.

Conclusions

Considering together the results obtained from the analysis of FAs profiles from the PLFA-SIP assay and the 16S rRNA gene amplicon sequencing, it was confirmed that aerobic methanotrophy at 5 and 20 °C was mainly due to members of Methylobacter clade 2, among other potentially new taxa of MOB. The increase in temperature from 5 to 20 °C decreased the diversity of the MOB community, suggesting certain vulnerability due to lack of redundancy of function. It was observed that ASVs showing a significant decrease or increase in abundance predominantly correspond to genera with few described species lacking psychrophilic or psychrotrophic representatives.

From this study, new perspectives emerge to expand the knowledge of methanotrophy in Maritime Antarctica, such as the isolation of species of MOB not yet described and the study of MOB and the rest of the bacterial community as a complex assemblage that deserves attention for the ecological role they play in this system. Our results provide interesting new data on methanotrophy and the effect of temperature in Maritime Antarctica, contributing to the knowledge of information about this particular area.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

DMR and RJM designed the study and experimental details. DMR conducted the experiments, completed the analysis, and wrote the initial draft. RJM, DC, and LSG contributed to the manuscript and editing. All authors contributed to the article and approved the submitted version.

Funding

Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. This work was partially supported by ANII (Agencia Nacional de Investigación e Innovación), Proyecto de Iniciación a la Investigación 2021 N° 22320200200176UD—CSIC (Comisión Sectorial de Investigación Científica, Udelar). DMR was supported by a scholarship from CAP-Universidad de la República and ANII (POS_NAC_2021_1_169845), Uruguay. DMR and RJM are members of the National Research System (SNI-ANII); RJM is a member of PEDECIBA. DC and LSG were funded by the Spanish Ministry of Science and Innovation/State Agency of Research MCIN/AEI/ https://doi.org/10.13039/501100011033 and “ERDF A way of making Europe” through the grants PID2022-140180OB-C21 (DC), RYC2018-023943-I and PID2021-126746NB-I00 (LSG).

Data availability

The datasets generated for this study can be found in the NCBI, BioProject ID PRJNA1081369.

Declarations

Competing interest

The authors declare no competing interests.