Methanogenic populations in the Sonora Margin sediments.
The methane isotopic ratio measured previously in these samples suggested that methane in the Sonora Margin shallow sediments was mainly of biogenic origin (
41). Furthermore, our results show that at least 91% of the biogenic methane in surface (0 to 7 cmbsf) sediments was produced by methylotrophic methanogenesis, suggesting that among the tested substrates, methylated amines were the main methane precursors in these sediments. Occurrence of methylotrophic methanogenesis throughout the sediments was supported by detection of 16S rRNA sequences related to
M. burtonii and
M. alaskense in enrichment cultures amended with trimethylamine. These methylotrophic methanogens that can generate methane by disproportionation of methylated amines appear to be widespread in cold seep environments (
14,
42). However, these environments might harbor only low abundances of
Methanoccocoidetes lineages, as related sequences were rarely directly detected without previous methanogenic enrichments (
43–45) or specific functional-gene amplifications (
46–49). Enrichment steps are generally required for the detection and identification of
Methanococcoides lineages in cold seep sediments (
13,
14,
42,
50,
51). The presence and activity of these methanogens in these sulfate-rich sediments (22 to 5 mM sulfate) (
23), as observed previously in other marine sediments (
14,
47,
50,
52,
53), were probably a consequence of utilization of noncompetitive methanogenic substrates, such as methylamines (
17,
19,
54). Methylated amines were presumably available in the surface sediments of WM14 and EWM14, as marine invertebrates, observed in high densities over these sediments, can accumulate large amounts of osmolytes (e.g., betaine and trimethylamine
N-oxide) and choline (widespread in cell membranes) in their tissues that can be subsequently released in the sediments and degraded to smaller methylated amines (e.g., TMA,
N,
N-dimethylglycine, and
N,
N-dimethylethanolamine) (
Fig. 4) (
55). For example, TMA concentrations in marine sediments were previously shown to be related to the abundance of benthic invertebrates (
56). Furthermore, degradation of choline and betaine to TMA has been reported for the deltaproteobacterial lineages
Desulfovibrio (
57),
Desulfobacterium (
58), and
Desulfuromonas (
59), detected previously by a 16S rRNA survey in the Sonora Margin sediments (
60). However, it has recently been demonstrated that
Methanococcoides species can also directly utilize choline and betaine to produce methane and therefore bypass the need for the bacterial-degradation step (
Fig. 4) (
19,
61). Hence, the use of invertebrate-derived substrates might explain the widespread occurrence of
Methanococcoides in organic-rich marine environments, such as cold seeps (
14,
42,
44,
45), tidal flats (
47,
53,
62), whale fall (
63), and mangrove sediments (
64), usually colonized by benthic invertebrates. These results also support studies showing cooccurrence of sulfate reduction and methylotrophic methanogenesis in marine sediments (
17,
65).
In contrast to methylotrophic methanogenesis, hydrogenotrophic methanogenesis rates were below those measured previously in seep and nonseep marine sediments (<0.4 to 30 nmol cm
−3 day
−1 [
28]) but were similar to hydrogenotrophic methanogenesis rates measured in the Amsterdam and Mercator mud volcanoes (
42,
66). Although methylotrophic methanogenesis dominated in surface sediments, the proportion of hydrogenotrophic methanogenesis increased with depth, representing up to 50% of the methane production at the bottom of the EWM14 core. In these organic-rich sediments, hydrogen could be produced by fermentation of organic matter by heterotrophic bacteria (
49), such as members of the phylum
Firmicutes (
Fig. 4), previously detected in significant proportions in these environmental samples (
24,
60). The presence of active hydrogenotrophic methanogenesis in these sediments was also supported by the growth of methanogens in enrichment cultures amended with H
2-CO
2. All the 16S rRNA sequences detected in these enrichments were affiliated with the genus
Methanogenium (order
Methanomicrobiales) and were detected previously using qPCR in the original environmental samples (
23). The characterized
Methanogenium strains are psychrophilic to thermophilic methanogens (0 to 62°C), mainly isolated from marine sediments, and can use formate or H
2-CO
2 as a substrate. Three distinct lineages of
Methanogenium were identified (groups 1, 2, and 3 [
Fig. 3]) in these enrichment cultures of Sonora Margin cold seep sediments. Sequences affiliated with
Methanogenium group 1 were detected from all enrichments amended with H
2-CO
2 and formed a distinct phylogenetic group that might represent a new lineage. A second group (
Methanogenium group 2), closely related to
M. cariaci (98% sequence similarity) strains previously identified in other cold seep sediments (
13,
44,
67), was detected only in enrichments from sediments underlying the white mat amended with acetate and H
2-CO
2. A third group of sequences (
Methanogenium group 3) distantly related to
M. marinum (93% sequence similarity) was detected in only two enrichment cultures amended with acetate and H
2-CO
2 from EWM14 sediments (6 to 10 cmbsf) and could also represent a new genus within the order
Methanomicrobiales. Similarly, different putative H
2-CO
2-utilizing
Methanomicrobiales lineages related to
Methanocorpusculum,
Methanoculleus, and
Methanomicrobium were also detected previously in the neighboring hot hydrothermal sediments of the Guaymas Basin (
49).
Methanomicrobiales were the only hydrogenotrophic methanogens detected in these shallow sediments, suggesting that members of this order could be responsible for most of the hydrogenotrophic methanogenesis in Sonora Margin sediments.
Acetate has been previously proposed as a significant substrate for methanogenesis in the hydrothermal sediments of the Guaymas Basin (
49). However, rates of acetate methanogenesis in these cold seep sediments were very low (1/20 of H
2-CO
2 methanogenesis), as they were below the typical rates measured in these environments (
28). Moreover, no methanogens were enriched with acetate as a sole carbon and energy source, although aceticlastic methanogens related to
Methanosarcina baltica were previously detected in these sediments using different enrichment conditions (incubation temperature, 25°C) (
20). Putative mesophilic aceticlastic methanogens were not detected, as opposed to the hydrothermal sediments of the basin (
49), suggesting that aceticlastic methanogens in the Sonora Margin were at low abundance and therefore difficult to enrich. Aceticlastic methanogens in these sediments could also be outcompeted for acetate by sulfate-reducing communities detected previously (
60) and associated with high sulfate concentrations (
68).
Have we caught them all?
In this study of shallow sediments of the Sonora Margin using culture-based approaches, four different methanogenic lineages were identified, whereas only one was detected from the same environmental samples, using culture-independent methods. This suggests that enrichment cultures can lower the detection limits of methanogens in these environments (
14,
42). Moreover, detection of lineages affiliated with
Methanosarcinales and
Methanomicrobiales is consistent with previous culture-independent surveys of archaeal communities associated with the Sonora Margin shallow sediments (
23,
24). Contrary to results from qPCR and pyrosequencing studies, the sizes of the amplicons in this culture-dependent study allowed phylogenetic identification and characterization of the methanogen community. However, members of the order
Methanococcales were previously quantified in abundance similar to that of the
Methanomicrobiales (
21). Mesophilic species of the
Methanococcales are known to be extremely sensitive to osmotic changes (
26) and have also been detected in low proportion in the hydrothermal sediments of the Guaymas Basin (
49). Hence, the lack of
Methanococcales lineages in our enrichment cultures might be due to the sample depressurization during the core recovery or to unsuitable culture conditions (e.g., temperature and time of incubation). Thus, despite the identification of several methanogen lineages, all the lineages might not have been detected.
Several studies showed that Sonora Margin sediments harbor high concentrations of ANME lineages (1, 2, and 3) distributed throughout the upper 20 cm of sediments (
23,
24,
60). Commonly proposed as methane oxidizers, some ANME lineages might also produce methane (
68–70) and be physiologically versatile (
23,
68). Despite their abundance in the environmental samples, ANME aggregates disappeared rapidly in the cultures, and no ANME sequences were detected from these methane-producing enrichments. This might suggest that ANME were not methane producers under our culture conditions. However, we could not exclude the possibility that ANME lineages could use alternative methanogenic substrates, such as methanol, as recently proposed (
68).
Together, these results indicated that the high methane concentrations measured in the Sonora Margin cold seeps are partially produced in the shallow sediments by active methanogens dominated by methylotrophic
Methanococoidetes, whereas the proportion of CO
2-reducing methanogens related to
Methanogenium increased with sediment depth (
Fig. 4). Aceticlastic methanogens represented a minority of the methanogen community. However, the methanogenic contributions of other shallow, uncultured microorganisms and ANME lineages using different substrates, as well as deeply buried microorganisms, remain to be explored.