In peatlands, hydrogentrophic methanogenesis is the predominant pathway of CH₄ formation, especially in deeper layers, accounting for 50 to 100% of total CH₄ production (12, 18, 28, 40). However, acetoclastic methanogenesis has also been found to play an important role in acidic bogs (1, 2, 21). Relatively little is known about the archaeal communities inhabiting peatlands. Recent studies of different wetlands revealed the presence of methanogens belonging to the Methanomicrobiaceae, Methanobacteriaceae, Methanococcaceae, Methanosarcinaceae, and Methanosaetaceae as well as new archaeal lineages within the Euryarchaeota (3, 7, 8, 13, 17, 34, 36, 37). However, the role of the methanogenic populations in CH₄ production under different in situ conditions is unknown.

Attempts to isolate acidophilic or acidotolerant methanogens in pure culture have failed until very recently, although acid-tolerant enrichment cultures have been reported (6, 11, 15, 34, 41). It was only after we finished our study that Bräuer and coworkers reported the successful isolation of a moderately acidophilic methanogen belonging to the Methanomicrobiales order (5).

The aim of the present study was to investigate how high acidity and low temperature can affect the functioning of the methanogenic community, its structure, and, hence, methane production in a peat bog, as well as to obtain a pure culture of an acidophilic methanogen. We used the same bog samples as in our previous study (21).

We obtained peat samples from Bakchar Bog, which is located in West Siberia (57°N, 83°E). The main unforested part of the bog is covered with continuous Sphagnum moss and patches of vascular plants (Carex, Menyanthes, and Equisetum spp.). The detailed location of the bog and structure of the plant community have been described earlier (21, 29). The samples were taken in July 1999 at a depth of 30 to 50 cm below the water table from the site covered with Equisetum. The peat pH values were in a range of 3.5 to 5.5, with pH 4.8 at the sampling site. The bog surface temperature varied during the summer season from 5 to 35°C but never exceeded 15°C at the depth of sampling.

The samples were transferred into sterile plastic bottles, transported to the laboratory, and then stored at 4°C for up to 3 months before analysis. The peat samples were handled anoxically under a N₂ atmosphere as described before (21). An aliquot (15 ml) of the samples was placed into a sterile serum bottle (60 ml), gassed with N₂, closed with a sterile black butyl rubber stopper, and then incubated in darkness without shaking. The samples were adjusted to different pH values by the addition of either 1 M HCl or 1 M NaOH. The total incubation time depended on temperature and pH and varied from 20 to 360 days at 25°C (pH 4.8) and 4°C (pH 3.8), respectively. The production rates of CH₄ were determined as described previously (21). Experiments were done in duplicate or triplicate. The δ¹³C of CH₄ and CO₂ was determined at the end of the incubations as described before (21).

Radioactive tracer experiments were done similarly to those described above but with Hungate tubes (20 ml). The tubes were filled with 10-ml peat samples, evacuated, gassed with N₂, and closed with rubber stoppers. The tubes were incubated at different temperatures, and the concentrations of intermediates were monitored until they reached steady state (21). The experiments were initiated by injecting 0.5 ml of carrier-free ¹⁴C-labeled Na-[2-¹⁴C]acetate (1 × 10⁶ dpm, 53 mCi mmol⁻¹; Amersham) or NaH¹⁴CO₃ (1 × 10⁶ dpm, 54 mCi mmol⁻¹) into each tube. Radioactive and nonradioactive CH₄ and CO₂ were analyzed in samples taken from the headspace. Acetate turnover rates, the respiratory index [¹⁴CO₂/(¹⁴CO₂ + ¹⁴CH₄)] for degradation of [2-¹⁴C]acetate, the fraction of acetate-dependent CH₄ production, and the fraction of CH₄ produced from H₂ and CO₂ were determined as described before (21). All experiments were done in triplicate.

The sampling for DNA extraction and terminal restriction fragment length polymorphism (T-RFLP) analysis of archaeal 16S rRNA genes was carried out using duplicate peat samples taken at the end of incubations (using all except those incubated at 4°C), and the samples were stored at −20°C. The procedure was described in detail before (21).

Methanogenic archaea were enriched in a low-ionic-strength liquid medium containing (in milligrams per liter of distilled water) NH₄Cl (16.5), MgCl₂·6H₂O (25), CaCl₂ (8.4), KCl (16.4), KH₂PO₄ (16.8), NaHCO₃ (100), Na-acetate (30), yeast extract (20), and resazurin (10), as an indicator of anoxic conditions. The medium was supplemented with 0.5% (vol/vol) of a vitamin solution (42), trace element stock solutions (21), and H₂ plus CO₂ (80:20, vol:vol) as substrate. The medium was buffered with morpholinoethanesulfonic acid (10 mM) and reduced with titanium citrate (0.5 mM). The initial pH of the medium was adjusted to 4.5 using 0.1 N HCl. Peat samples from different research sites were used as inocula. As soon as visual turbidity developed, CH₄ production was measured and an aliquot was transferred to fresh medium of the same composition. The same medium was used to assay the pH and temperature dependence of the cultures.

The procedure used for DNA extraction from methanogenic cultures has been described previously (16). Primers Ar12f (25) and Ar1542r (19) were used to amplify the nearly complete 16S rRNA gene of the isolated methanogens. Primer combinations ME1/ME2 and MCRf/MCRr were used to amplify mcrA fragments. The PCR mixtures and the other thermal cycling parameters were as previously described (24, 33) but with 32 cycles and an annealing temperature of 50°C for the amplification of the 16S rRNA gene. PCR products were purified with a QIAquick PCR purification kit (QIAGEN, Hilden, Germany). Sequencing of PCR products was performed using a BigDye Terminator cycle sequencing kit on an ABI 377A DNA sequencer (Applied Biosystems, Foster City, CA). Sequences were assembled using the Lasergene software package (DNAStar, Madison, WI). Phylogenetic analysis of the sequence data was made using the ARB software package and its databases (23) as previously described (24). The 16S rRNA gene and mcrA sequences obtained in this study have been designated in the GenBank database under the accession numbers DQ677518 and DQ677519, respectively.

Rates of CH₄ production, acetate turnover, and fractions of the relative contribution of acetoclastic and hydrogenotrophic methanogenesis were measured in peat samples incubated at different pH values and temperatures (Table 1). Rates of CH₄ production decreased with decreasing pH values and temperatures. Under most incubation conditions, acetoclastic and hydrogenotrophic methanogenesis contributed to total CH₄ production at ratios of about 65 and 35%, respectively, as expected from our previous study of the same peat bog (21), the results of which are included in Table 1. However, hydrogenotrophic methanogenesis became the dominant path of CH₄ formation (>92%) when the peat samples were incubated at pH 3.8 and at low temperatures (4 or 15°C) (Table 1). Note that these are realistic in situ conditions. The relative increase of the role of H₂-dependent methanogenesis was confirmed by the δ¹³C of produced CH₄, which became increasingly depleted in ¹³C with decreasing pH values (Fig. 1). The ratio of δ¹³CO₂ to δ¹³CH₄ gives an approximation of the apparent fractionation between CO₂ and CH₄, which can be calculated as follows: (δ¹³CO₂ +1,000)/(δ¹³CH₄ +1,000). The ratio of δ¹³CO₂ to δ¹³CH₄ increased from 1.053 at pH 6 to up to 1.072 at pH 3.8, confirming a relative increase of hydrogenotrophic methanogenesis at low pH values (18, 39).

TABLE 1. Acetate turnover and methanogenesis in peat samples incubated at different temperatures and pHsa

FIG. 1. δ13CH4 values measured in peat samples (Equisetum site) adjusted to different pH values and incubated at 4°C.

The archaeal community structure in the samples incubated at different temperatures and pH values was determined by T-RFLP analysis. Most of the incubations exhibited the same T-RFLP patterns that had been reported in our previous study (21), i.e., the predominance of members of the methanogenic families Methanosarcinaceae (T-RF, 185 bp) and Methanomicrobiaceae (T-RF, 392 bp). Incubation at 15°C versus 25°C affected the T-RFLP patterns only a little, but pH values showed a marked effect (Fig. 2). Notably, at the lowest pH, 3.8, at 15°C, we observed the appearance of a substantial T-RF of 91 bp which represents members of Methanobacteriaceae. Also, at pH 6.0 and 15°C, a novel T-RF (257 bp) appeared which, however, has not yet been identified.

FIG. 2. T-RFLP patterns of archaeal small-subunit rRNA gene fragments amplified from DNA extracts obtained from peat samples (Equisetum site) incubated at 15°C and different pH values. The x axis shows the lengths (bp) of the T-RFs. The y axis shows the (more ...)

The stimulation of hydrogenotrophic methanogenesis at low pH values prompted us to attempt the isolation of acidophilic methanogens. Three methanogenic strains, MB2, MB3, and MB4, were enriched in a low-ionic-strength medium supplemented with H₂ plus CO₂ at pH 4.5 and subsequently isolated as pure cultures after repeated transfers. All three strains were from different samples collected at different peat bog research sites covered with Carex, Equisetum, and Menyanthes species (21). The organisms have similar morphologies and are long rods, growing both as single cells and in associations (Fig. 3). The cells are often irregularly twisted. All strains had identical 16S rRNA gene sequences and thus belong to the same species. Phylogenetic analysis revealed that the new strains represent a new species within the genus Methanobacterium (Fig. 4); the closest relative was Methanobacterium congolense with 97% sequence identity. The McrA amino acid sequence of strain MB4 was most similar to that of Methanothermobacter thermautotrophicus X07794 as the closest cultivated relative (90% sequence similarity). The isolated strains grew only on H₂-CO₂ or formate as the energy source when a mineral medium that contained tiny amounts of yeast extract (0.002%) and acetate (0.003%) as additional carbon sources was used. The growth rate was very low (on the order of days), as the cultures were fully grown after about 2 weeks. Most notably, the cultures grew at low pH values and low temperatures and produced CH₄ over a range of pH values (3.8 to 6.0) and temperatures (5 to 30°C), with maximal rates at pH 5.0 to 5.5 and 25 to 30°C.

FIG. 3. Phase-contrast microscopy of a methanogenic archaeon, strain MB4, isolated from Bakchar Bog.

FIG. 4. Phylogenetic positions of strain MB4 and two further new strains of acidophilic methanogens. The 16S rRNA gene tree was constructed from a distance matrix using neighbor-joining analysis. The scale bar indicates 0.1 base pair changes per nucleotide sequence (more ...)

Our study shows that pH is an important factor influencing not only the rate of methanogenesis but also the CH₄ production pathway and the methanogenic archaeal community. A shift from acetoclastic to H₂-dependent methanogenesis occurred between pH 4.7 and 3.8 (Table 1). The predominance of hydrogenotrophic methanogenesis has also been observed in other acidic peatlands (10, 22, 27) and is possibly explained by the existence of acetic acid in its free form at a pH of <4.7. Acetic acid can pass through the cell membrane and act as a decoupler of the proton motive force. However, this is probably not the only explanation, since acetoclastic methanogenesis has occasionally been observed in acidic peatlands (1, 2, 21). Furthermore, hydrogenotrophic methanogens and other microorganisms are also exposed to the acetic acid present in their environment and thus must have mechanisms to compensate for decoupling by acetic acid.

A strong signal of a T-RF of 91 bp, characteristic for Methanobacteriaceae, was detected in samples at pH 3.8 and 15°C, indicating that the relative proportion of this methanogenic group increased under these conditions. Importantly, all species of the Methanobacteriaceae family grow with H₂ plus CO₂ as the substrate to produce CH₄. Hence, we conclude that Methanobacteriaceae became prevalent at low pH values, thus explaining the predominance of H₂-dependent CH₄ production observed in the isotopic experiments.

This conclusion is in agreement with the successful enrichment and isolation of a Methanobacterium sp. capable of growth with H₂ and CO₂ as a sole source of carbon and energy. The unique feature of this organism is its ability to grow at a pH as low as 3.8. No known methanogenic archaeon has such a low minimal pH for growth (5, 14, 31). The isolation of virtually the same Methanobacterium sp. from different field sites indicates its widespread distribution and possible ecological importance in Bakchar Bog. The phylogenetic affiliation of the isolated methanogen to a new species within the genus Methanobacterium indicates that members of this methanogenic genus may have a large capacity for adaptation to extreme pH conditions. It should be noted that the genus Methanobacterium already contains two alkaliphilic and one moderate acidophilic species (20, 30, 43), thus collectively having the widest growth range concerning pH (from 3.8 to 9.9). However, there are also reports on enrichments (6, 34) of acidiphilic archaea belonging to novel archaeal clusters. One of these was recently isolated (5). Like our isolate, this archaeon is also a hydrogenotrophic methanogen but belongs to a novel cluster within the order Methanomicrobiales.

Our study expands our knowledge about the mechanisms of CH₄ formation and the diversity of methanogenic archaea in acidic peat bogs.

We thank K.-J. Chin for help in T-RFLP analysis and Bianca Pommerenke, Peter Claus, and Melanie Klose for skillful technical assistance.

This work was financially supported by the DFG T86 grant “Methanogenesis at low pH” and a fellowship provided by the Max Planck Society.