Origin and fate of acetate in an acidic fen

Abstract

Acetate is a central intermediate in the anaerobic degradation of organic matter, and the resolution of its metabolism necessitates integrated strategies. This study aims to (1) estimate the contribution of acetogenesis to acetate formation in an acidic fen (pH ~ 4.9), (2) assess the genetic potential for acetogenesis targeting the fhs gene encoding formyltetrahydrofolate synthetase (FTHFS) and (3) unravel the in situ turnover of acetate using stable carbon isotope pore-water analysis. H₂/CO₂-supplemented peat microcosms yielded ¹³C-depleted acetate (−37.2‰ vs. VPDB (Vienna Peedee belemnite standard) compared with −14.2‰ vs. VPDB in an unamended control), indicating the potential for H₂-dependent acetogenesis. Molecular analysis revealed a high diversity and depth-dependent distribution of fhs phylotypes with the highest number of operational taxonomic units in 0–20 cm depth, but only few and distant relationships to known acetogens. In pore waters, acetate concentrations (0–170 μM) and δ¹³C-values varied widely (−17.4‰ to −3.4‰ vs. VPDB) and did not indicate acetogenesis, but pointed to a predominance of sinks, which preferentially consumed ¹²C-acetate, like acetoclastic methanogenesis. However, depth profiles of methane and δ¹³C_CH4 revealed a temporarily and spatially restricted role of this acetate sink and suggest other processes like sulfate and iron reduction played an important role in acetate turnover.

acetogenesis, isotope analysis, fhs, peat, methanogenesis

Introduction

Peatlands cover only ~3% of the Earth's surface (Rydin & Jeglum, 2006) but have accumulated 30% of global soil carbon (Gorham, 1991). Consequently, they play a key role in global carbon cycling. The sequestration of carbon is facilitated in water-saturated acidic peat soils, where the degradation of organic matter is strongly hampered because of the lack of oxygen. Under anoxic conditions, plant-derived organic matters, for example cellulose and its early degradation products, are first subjected to hydrolysis by extracellular enzymes (McInerney & Bryant, 1981). Monomers formed in peatlands such as glucose and xylose (Kotsyurbenko, 2005) are then degraded to short-chain organic acids (e.g. acetate, propionate), alcohols, CO₂, and H₂ by primary and secondary fermenters (Wüst et al., 2009), followed by the transformation into CO₂ and CH₄, when nitrate, sulfate, oxidized manganese, and iron become depleted (Paul et al., 2006).

Acetate in particular has received attention because of its central role in ‘intermediary ecosystem metabolism’ (Drake et al., 2009). It is not only produced via fermentation but also via acetogenesis, that is, the reductive synthesis of acetate from CO₂ via the acetyl-CoA pathway (Drake et al., 2006). Acetogenic microorganisms synthesizing acetate can couple the fermentation of monomeric compounds, like sugars, with the reduction of CO₂. Acetogens compete also with methanogens for C1 substrates like formate, H₂–CO₂, and methanol. Hence, their metabolic versatility enables them to contribute on different levels to the overall anaerobic carbon degradation. Acetogens cannot be targeted using 16S rRNA-based approaches, because they do not form phylogenetically coherent groups. Instead, the gene fhs that codes for the enzyme 10-formyl tetrahydrofolate synthetase (FTHFS) of the acetyl-CoA pathway can be analyzed as functional marker for acetogenic microbial communities. The FTHFS primer set designed by Leaphart & Lovell (2001) has been applied to animal gut systems that have a high carbon throughput (Ottesen & Leadbetter, 2010), but also to marine environments with low organic C contents (Lever et al., 2010) and a variety of freshwater environments (Leaphart et al., 2003; Xu et al., 2009; Hunger et al., 2011). Unfortunately, this molecular approach provides only limited insight into the genetic potential for acetogenesis because the primer set also targets the sequences of purinolytic clostridia, sulfate reducers, methanogens, and other organisms that possess the FTHFS enzyme (Drake et al., 2006).

Similar to acetate formation, consumption of acetate is performed by a broad spectrum of microorganisms including acetoclastic methanogens (Hedderich & Whitman, 2006), but also nitrate, Fe- and sulfate-reducing prokaryotes (Rabus et al., 2006). The consistently low acetate concentrations in surface freshwater and marine habitats (Wellsbury et al., 2002) reflect a steady state between production and consumption. The stable carbon isotopic composition of pore-water acetate provides an approach to elucidate the different metabolic processes involved in acetate turnover and their relative rates in the field (Blair et al., 1987; Gelwicks et al., 1989; Blair & Carter, 1992; Heuer et al., 2009). The isotopic composition of acetate has been determined in various marine (Blair et al., 1987; Heuer et al., 2006, 2009) and freshwater sediments (Sugimoto & Wada, 1993; Gelwicks et al., 1994; Conrad et al., 2002; Krüger et al., 2002; Fey et al., 2004; Heuer et al., 2010). When fermentation is the main source of acetate, δ¹³C values of acetate (δ¹³C_ace) closely reflect the carbon isotopic composition of total organic carbon (TOC). In contrast, acetogenesis is associated with a distinct kinetic isotope effect and preferential utilization of the ¹²C-isotope from the dissolved inorganic carbon (DIC) pool (Gelwicks et al., 1989). In sediments, acetate resulting from acetogenesis often shows a distinct ¹³C-depletion relative to TOC (δ¹³C_ace < δ¹³C_TOC) (Heuer et al., 2006, 2009, 2010). In contrast, the kinetic isotope effect associated with acetoclastic methanogenesis results in a ¹³C-enriched acetate pool. Though the preferential consumption of isotopically light acetate leads to ¹³C-depleted CH₄, δ¹³C-values of CH₄ from acetoclastic methanogenesis (−65‰ to −50‰ vs. VPDB (Vienna Peedee belemnite standard) are high compared with those of CH₄ originating from CO₂ reduction (−110‰ to −60‰ vs. VPDB) (Whiticar, 1999). The stable carbon isotopic composition of acetate is a promising indicator for deciphering metabolic processes in natural environments, but it has so far not been used to elucidate sources and sinks of acetate in CH₄-emitting peatlands, for which the concept of intermediary ecosystem metabolism was established (Drake et al., 2009).

In the acidic fen Schlöppnerbrunnen (Germany), extracellular enzyme activity and the trophic links between moderately acid tolerant fermenters and methanogens have been intensively studied (Hamberger et al., 2008; Reiche et al., 2009; Wüst et al., 2009). Acetate appears to be an important intermediate of the carbon cycle in this fen with a broad concentration range from 0 to 2 mM (Küsel et al., 2008). However, the different sources of acetate in this peatland ecosystem have not yet been elucidated, and it is still unclear whether acetate acts as the main precursor for methanogenesis. Both acetoclastic as well as hydrogenotrophic methanogenesis have been reported (Knorr et al., 2008; Reiche et al., 2008), and archaeal 16S rRNA genes indicative of both obligate acetoclastic methanogens (Methanosaetaceae) and H₂/CO₂- and formate-consuming methanogens (Methanomicrobiaceae) have been detected (Wüst et al., 2009), leaving the role of acetate in these turnover processes unclear. Moreover, the previous research revealed the presence of active Fe- and sulfate-reducing communities, especially in surface-near fen soil layers (Loy et al., 2004; Schmalenberger et al., 2007; Küsel et al., 2008) that could play a key role in acetate consumption. On the basis of these earlier findings, the main objectives of the present study are to (1) estimate the contribution of acetogenesis to acetate formation in the peat soil, (2) investigate the acetogenic microbial communities present in different peat layers, and (3) unravel the turnover of acetate along a 40-cm deep pore-water profile using stable carbon isotope analysis of acetate and other carbon compounds.

Materials and methods

Field site and sampling methods

Samples were obtained from the acidic fen Schlöppnerbrunnen (fen area: 0.8 ha, pH 4.7 in the pore water) located in the northern Fichtelgebirge in east-central Germany (50°7′54′'N, 11°52′51′'E, 700 m above sea level). The fen has been intensively studied since 2001, and more detailed geochemical characteristics of the pore water and the solid phase are given elsewhere (Küsel et al., 2008; Reiche et al., 2008, 2010).

Peat soil samples for microcosms and DNA extraction were obtained in February 2009 and April 2010, respectively, from location M (Reiche et al., 2010). Spatial heterogeneity was accounted for by taking three replicate cores (0–40 cm depth) located in close proximity. Peat cores (3 or 8 cm inner diameter) were divided into four segments of each 10 cm length. Triplicate core material of each of the four depth zones (0–10; 10–20; 20–30; 30–40 cm) was pooled and mixed in either sterile 50-mL plastic tubes or bags. Subsamples for DNA extraction were immediately frozen in liquid N₂ and stored at −80 °C until DNA extraction was performed. Peat water for the microcosm studies was collected in February 2009 in plastic bottles from drainage ditches running in approximately 20 m distance from site M. Samples for microcosm experiments were transported to the laboratory at 4 °C and immediately processed or stored at 4 °C for a few days until use.

Pore-water samples were obtained with a self-made dialysis sampler from the upper 40 cm of peat soil in February 2009 and June 2010. The dialysis chamber was composited of a stainless steel frame holding a plastic piece that consisted of replicate chambers covering 40 cm depth with a resolution of 1 cm. The 5-mL cells were covered with a semi-permeable polyether sulfone membrane. At the sampling site, the peat was always water saturated ensuring the diffusion of pore-water compounds into the chambers of the dialysis sampler during 2 weeks of exposure. During sampling in February 2009, the area was covered by a 40-cm snow layer but the peat soil was never frozen below 2 cm. Pore-water samples were preserved in the following way: (1) for organic metabolites, 0.5-mL subsamples were filtered (0.2 μm, nylon filter) into 2-mL glass vials with natural rubber/TEF septum vial caps, analyzed or stored at −20 °C; (2) for CO₂ and CH₄ analysis, each 0.5-mL aliquot of unfiltered pore water was acidified with 20 μL 4 M HCl in closed 2-mL glass vials and stored upside down at 4 °C to prevent gas outflow until measurement; (3) for DIC, ~ 2-mL unfiltered subsamples were stored without headspace in 2-mL vials frozen at −20 °C.

Microcosm incubation experiment

To determine the potential acetate-forming processes in the peat soil, 3 g (fresh weight) peat soil from the uppermost (0–10 cm) and lowermost (30–40 cm) depth horizons were added to 50 mL anoxic (N₂), sterile (autoclaved, 30 min, 120 °C) medium in sterile 150-mL incubation flasks (Mueller and Krempel, Buelach, Switzerland) under a continuous flow of sterile Ar. Flasks were closed with sterile butyl rubber stoppers and screw caps, and were incubated in the dark at 15 °C without agitation with an initial overpressure of 150 mbar Ar. The medium consisted of peat drainage water supplemented with 0.9 mL L⁻¹ B-Vitamin solution (Drake, 1994), 9.0 mL L⁻¹ trace element solution TES-SL9 (Tschech & Pfennig, 1984), and 4 mL L⁻¹ yeast extract (w/v 3%, final concentration 0.012%). As electron donors, either glucose sterilized with a 0.2-μm filter to reach a final concentration of 2 mM or a gas phase of sterile H₂/CO₂ (80 : 20, v/v) were added to each of the three replicate bottles. The experiment included controls without amendment of H₂/CO₂ or glucose. For gas analysis, the headspace was sampled with sterile Ar–flushed syringes after shaking the bottles thoroughly to remove gas bubbles from the peat soil. For the quantification of short-chain fatty acids, sugars, and alcohols, 1-mL subsamples of the liquid phase were analyzed throughout the incubation period of 54 days. After 34 days of incubation, additional liquid phase samples were taken for stable carbon isotope analysis of acetate.

Analysis of fhs gene sequences in soil microbial communities

DNA was extracted from soil samples obtained in April 2010 of the following depths: 0–10 cm, 10–20 cm, and 30–40 cm using the RNA PowerSoil^® Total RNA Isolation kit in combination with the DNA Elution Accessory kit (MO BIO Laboratories, Carlsbad, CA) according to the manufacturer's instructions. The fhs gene encoding formyltetrahydrofolate synthetase (FTHFS, EC 6.3.4.3) was amplified from the samples of each depth using Taq Pol (JenaBioscience, Jena, Germany) and the primer set FTHFS-F: (5′-TTYACWGGHGAYTTCCATGC-3′)/FTHFS-R (5′-GTATTGDGTYTTRGCCATACA-3′) (Lovell & Leaphart, 2005), following the cycling conditions given therein. PCR products of the expected size of 1.1 kb were purified using the NucleoSpin^® Extract II PCR clean-up kit (Macherey-Nagel GmbH & Co. KG, Düren, Germany). Cloning of the purified PCR products was performed using the pGEM^®-T Easy Vector System (Promega Corporation, Madison, WI) with chemically competent Escherichia coli JM 109 (Promega Corporation). Inserts were sequenced by Macrogen (South Korea) yielding partial reads with a maximum length of 900 bp. A chimera check was performed using the online tool ‘Bellerophon’ (Huber et al., 2004). An fhs gene database was created in ARB (Ludwig et al., 2004) using fhs sequences from other studies (Lovell & Leaphart, 2005; Henderson et al., 2010). In addition, the BLAST search tool (http://blast.ncbi.nlm.nih.gov/Blast.cgi,) was employed in February 2011 to retrieve fhs sequences from cultured and uncultured microorganisms that were most closely related to the fhs sequences obtained in this study. A basic alignment of deduced FTHFS amino acid sequences of 25 known acetogens and sulfate reducers was constructed using the ClustalW aligner incorporated into ARB, and deduced FTHFS amino acid sequences obtained in this study as well as sequences obtained from Genbank were aligned against this initial alignment. A phylogenetic tree was generated as a consensus tree on the basis of trees constructed using neighbor-joining, parsimony and maximum likelihood methods with 5000, 5000, 100 bootstraps, respectively. The tree calculations were based on amino acid positions 198–423 according to Lovell et al. (1990). Operational taxonomic units (OTUs) were defined on amino acid level based on a 97% sequence identity cutoff using DOTUR (Schloss & Handelsman, 2005), and the coverage of the clone libraries was calculated according to Singleton et al. (2001). For the analysis of phylogenetic affiliation, sequences were assigned to clusters as revealed by the structure of the phylogenetic tree rather than OTUs. Here, each cluster included sequences that shared a sequence identity of 76% or more. In addition, we calculated acetogen similarity (HS) scores (Henderson et al., 2010) to gain further information whether the deduced FTHFS amino acid sequences obtained in this study were likely to originate from true acetogens, which would be indicated by HS score values of 90% or higher. Because only partial sequence reads were obtained, the HS score calculated here was normalized to the number of amino acid positions available for HS score calculations in our sequence data [usually 31, 32 or 37 compared with a maximum of 40 positions originally suggested by Henderson et al. (2010)]. The fhs gene sequences obtained in this study have been submitted to Genbank under the accession numbers JN652667–JN652786.

Biogeochemical analysis

Quantitative biogeochemical analysis

Preparation of samples and measurement of concentrations of short-chain fatty acids, sugars, and alcohols in pore water or fluids from the peat soil incubation experiments via high performance liquid chromatography (HPLC), combined with a refractive index and UV detector, were performed as described by Reiche et al. (2008). Concentrations of gases (CO₂, H₂, CH₄) in the headspace of microcosms and of CH₄ in the pore-water samples were quantified with a Hewlett Packard Co. 5890 series II gas chromatograph combined with a thermal conductivity (CO₂, H₂) or a flame ionization detector (CH₄) according to Küsel & Drake (1995). Prior to gas quantification in the microcosms, gas pressure was measured with a TensioCheck TC 1066 (Tensio-Technik, Geisenheim, Germany) needle manometer. Negative pressure within bottles resulting from, for example, H₂ consumption was compensated for with sterile Ar.

Redox potential or pH were measured using a WTW pH meter (pH 330, Weilheim, Germany) combined with an InLab Redox micro-electrode or an InLab 423 combination pH micro-electrode (Mettler Toledo, Giessen, Germany), respectively. Prior to the determination of Fe(II) and total Fe (Fe_total)), 0.5 mL pore water were incubated in either 1.5 or 9.5 mL of 0.5 M HCl for 1 h at room temperature to dissolve the total amount of Fe. Afterward, the HCl-extractable Fe(II) was measured spectrophotometrically (Uvikon 931, Kontron Instruments) at 512 nm following the phenanthroline method after Tamura et al. (1974). Fe(III) was calculated as the difference between Fe_total and Fe(II).

Stable carbon isotope analysis

The carbon isotopic composition of acetate was analyzed by isotope-ratio-monitoring liquid chromatography/mass spectrometry (irm-LC/MS) according to Heuer et al. (2009) using a combination of ThermoFinnigan Surveyor HPLC ThermoFinnigan LC IsoLink interface and ThermoFinnigan Delta Plus XP mass spectrometer (Krummen et al., 2004; Heuer et al., 2006). The carbon isotopic composition of CH₄ was analyzed by isotope-ratio-monitoring gas chromatography/mass spectrometry (irm-GC/MS) using a Thermo Finnigan Trace GC, connected to a ThermoFinnigan DELTA Plus XP mass spectrometer via GC combustion III interface and routine open split as reported previously (Heuer et al., 2010). For carbon isotope analysis of acetate and methane, the detection limits were 15 and 20 μM, respectively, and the precision was better than 0.4‰ (1σ). The carbon isotopic composition of DIC was analyzed using a gas bench coupled to a Finnigan MAT 252 mass spectrometer as reported previously (Heuer et al., 2010) with a precision better than 0.1‰ (1σ). Carbon isotope analysis of TOC was performed on a dual-inlet mass spectrometer (Finnigan MAT Delta E) as described previously (Heuer et al., 2010), using freeze-dried samples that were ground and homogenized in a mortar, decalcified by addition of 6 N HCl and subsequently dried overnight at 60 °C. The precision of the analysis was 0.1‰ (1σ). All isotope ratios are given in δ¹³C notation (per mil, ‰) relative to the Vienna Peedee Belemnite Standard (VPDB), with δ¹³C = [(R_sample − R_VPDB)/R_VPDB] × 10³, with R = ¹³C/¹²C and R_VPDB = 0.0112372 ± 2.9 × 10⁻⁶.

Statistical analysis

The IBM spss Statistics Version 20.0.0 Software package was used to perform linear regression analysis.

Results

Potential acetate-forming processes in peat microcosms

When peat soil (0–10 cm) was supplemented with 2.3 mM glucose, the substrate concentration declined rapidly within the first 3 days (Fig. 1a). This rapid consumption of glucose resulted in the formation of mainly H₂ (6 mM), acetate (1.9 mM) (Fig. 1a), ethanol, lactate, propionate, and butyrate (< 2.8, < 0.4, < 0.3, and < 1.0 mM, respectively) (data not shown). After 3 days, H₂ concentrations decreased while acetate concentrations increased, reaching a plateau concentration of 2.5 ± 0.6 mM after 11 days, which was sustained during the rest of the incubation time. After 34 days of incubation, acetate had a δ¹³C-value of −14.9‰ and was slightly depleted in ¹³C relative to the added substrate glucose that had a δ¹³C-value of −11.0‰. The TOC derived from peat (n = 4, 0–40 cm) had a δ¹³C value of −26.6 ± 0.4‰. The formation of CH₄ was detectable after a lag phase of 19–34 days with a final CH₄ concentration of 0.1 ± 0.1 mM after 50 days. Similar substrate consumption and product formation patterns were observed in the glucose treatment with peat soil from 30 to 40 cm depth with the exception that no CH₄ was detected (Fig. 1d).

Effects of supplemented glucose (a, d) and H₂/CO₂ (b, e) on peat soil obtained from 0 to 10 cm and 30 to 40 cm depth (February, 2009) compared to an unamended control (c, f). Treatments were performed in triplicates and averages ± standard deviations are given. Samples for isotopic analysis of acetate were obtained after 34 days.

In H₂/CO₂ supplemented peat soil microcosms derived from 0 to 10 cm depth, complete consumption of H₂ was observed after 50 days of incubation (Fig. 1b), and only acetate was formed as a reaction product. Acetate concentrations reached a plateau at 1.9 mM (±0.3 mM) after 20 days. CH₄ formation started after a lag phase of 14 days with a final CH₄ concentration of 0.9 ± 0.5 mM after 50 days (Fig. 1b). After 34 days, acetate had a δ¹³C-value of −37.2‰ and was depleted in ¹³C relative to the peat TOC with a δ¹³C value of −26.6 ± 0.4‰. In the H₂/CO₂ treatments with peat soil from 30 to 40 cm depth (Fig. 1d–f), initial H₂ consumption rates between days 4 and 19 approximated 0.54 μmol mL⁻¹ day⁻¹ similar to the rate observed from 0 to 10 cm depth (0.56 μmol mL⁻¹ day⁻¹). Although consumption of H₂ was incomplete during the incubation time, the concentration of acetate was slightly higher (3.2 mM) than in the peat samples from 0 to 10 cm depth (Fig. 1b); the δ¹³C-value of acetate was −37.2‰. Again, CH₄ formation could not be detected during the incubation period.

Unamended peat soil (0–10 cm) reached lower concentrations of acetate and H₂ (< 0.9 mM), lactate, propionate, and butyrate than the glucose and H₂/CO₂-supplemented microcosms (Fig. 1c). CH₄ formation started after 19 days, and a concentration of 1.4 ± 0.2 mM was reached after 50 days in the unamended controls. After 34 days, acetate had a δ¹³C-value of −14.2‰ and was enriched in ¹³C relative to TOC (−26.6 ± 0.4‰).

Pore-water biogeochemistry

In February 2009, acetate was present in the peat pore water in high concentrations of up to 170 μM (Fig. 2a). A peak at 9 cm depth was followed by a zone of lower concentrations between 12 and 26 cm (mean 40 μM). In the deeper peat zones starting at ~30 cm depth, acetate reached again the concentrations similar to those in the surface-near soil layers. The carbon isotopic composition of acetate varied over the depth profile with δ¹³C-values ranging from −17.4‰ to −3.4‰ and averaged around −12.3 ± 3.6‰ (Fig. 2e). The strongest ¹³C-enrichment in the acetate pool was found at 14–25 cm depth and correlated well with low acetate concentrations (< 51 μM) (Fig. 2a). Throughout the depth profile, acetate was enriched in ¹³C compared with bulk TOC (−26.6‰). Apart from acetate, the following organic metabolites were also present at the time of sampling with highest concentrations primarily below 20 cm depth: ethanol (< 270 μM), butyrate (< 190 μM), propionate (< 80 μM) and formate (< 30 μM) (Fig. 2d). The pore water was acidic with a mean pH of 4.9, and the concentrations of dissolved CO₂ were on average only 0.7 mM with single peaks of up to 3 mM in the upper peat layers and with a slight increase in deeper soil layers (R² = 0.355, T = −3.535, P < 0.001). Similarly, the concentrations of pore-water CH₄ (Fig. 2b) were low (depth-integrated mean: 8 μM). The concentrations of both CO₂ and CH₄ in the headspace vials used for sampling were below the detection limit for stable carbon isotope analysis. HCl-extractable Fe²⁺ varied only slightly over the whole profile (average, 165 μM) (data not shown). Concentrations of HCl-extractable Fe³⁺ were even lower (< 94 μM) and could not be detected below 15 cm (data not shown).

Pore-water profiles of concentrations of acetate (a, f), CH₄ (b, g), CO₂ (c, h), other volatile fatty acids and alcohols (organics) (d, i) and δ¹³C values for TOC, acetate, DIC, and CH₄ (% vs. VPDB) (e, j) in February 2009 (a-e) and June 2010 (f–j).

In June 2010, acetate concentrations were highest (165 μM) at 12 cm depth (Fig. 2f). Below, they decreased gradually with depth down to ~ 30 μM (R² = 0.614, T = −6.553, P < 0.001). At all depths, acetate was enriched in ¹³C compared with TOC (−26.6‰). δ¹³C values of acetate ranged from −22.0‰ to −13.8‰ (depth-integrated mean: −17.2 ± 1.8‰, n = 24) (Fig. 2j). In the upper 20 cm, the acetate pool showed an increase in ¹³C with depth (R² = 0.833, T = 7.809, P < 0.001), but below 23 cm δ¹³C-acetate was more or less constant at −17.6 ± 0.7‰ (n = 9) (R² = 0.122, T = 1.116, P < 0.293). The second quantitatively important volatile fatty acid present in the profile was propionate with concentrations up to 140 μM, following a similar depth-dependent pattern as acetate (Fig. 2i). Other organic metabolites were present at lower concentrations including lactate (< 50 μM) and formate (< 10 μM). With a mean pH of 5.6, the pore water was less acidic than in February 2009. CO₂ concentrations were more than twice as high as in February 2009 with a depth-integrated mean of 1.5 mM and a peak in 7–15 cm depth (mean of this depth interval: 2.1 mM), followed by a decrease down to 1.3 mM in 23 cm depth (R² = 0.649, T = −4.511, P = 0.001) and being more or less stable below 20 cm (R² = 0.112, T = −1.548, P = 0.138). Maximum Fe²⁺ concentrations (1.3 mM) were more than four times higher in June 2010 compared with February 2009 with a clear peak at 5–7 cm depth and lower concentrations of 250 μM below 24 cm (data not shown). Relatively low concentrations of Fe³⁺ were found within the more aerated upper 10 cm of peat soil (< 184 μM) (data not shown). In sharp contrast to February 2009, concentrations of pore-water CH₄ were high with a maximum of up to 272 μM at 9–13 cm and a depth-integrated mean of 139 μM (Fig. 2g). The maximum at around 10 cm depth coincided with high concentrations of acetate, CO₂ (Fig. 2h), and also Fe²⁺ (300–1400 μM). Below this depth, the concentration of acetate, CO₂, CH₄, and Fe²⁺ declined. δ¹³C values of CH₄ declined steadily between 5 cm depth (−50.3‰) and 26 cm depth (−59.9‰) (R² = 0.550, T = −5.070, P < 0.001). Below, the elevated δ¹³C values were observed in two depth intervals before the carbon isotopic composition of CH₄ reached a steady mean δ¹³C-value of −61.7 ± 0.7‰ (n = 10) in the 30- to 40-cm-depth interval (R² = 0.288, T = −2.009, P = 0.072) (Fig. 2j). While δ¹³ values of CH₄ decreased with depth, δ¹³C values of DIC increased from −15.1‰ at 2 cm to −9.2‰ at 10 cm depth (R² = 0.873, T = 5.238, P < 0.01). Below 32 cm, DIC concentrations were too low for stable carbon isotope analysis.

Phylogenetic affiliation of the detected fhs gene sequences

A total of 135 sequence reads were obtained from the three different depth zones (0–10 cm; 10–20 cm; 30–40 cm), of which 120 were identified as fhs gene sequences by nucleotide-based BLAST search with sequence similarities to fhs gene sequences obtained in other studies ranging from 69% to 92%. Highest similarity was found to nucleotide sequences related to Gordonibacter pamelaeae (92%). The fhs gene sequences obtained in this study could be assigned to eight different classes and 15 families within the domain bacteria (Supporting Information, Table S1). They affiliated with the families Hyphomicrobiaceae (20% of total number of sequences), Rhizobiaceae (16%), Acetobacteraceae (13%), Rhodobacteraceae (10%), Methylobacteriaceae (8%), Phyllobacteriaceae (3%), Sphingomonadaceae (3%), and Xanthobacteraceae (2%), all belonging to the α-Proteobacteria, as well as Coriobacteriaceae (Actinobacteria, 14%), Desulfovibrionaceae (δ-Proteobacteria, 3%), Methylococcaceae (γ-Proteobacteria, 2%), Acidobacteriaceae (Acidobacteria, 1%), Clostridiaceae (Clostridia, 1%), Verrucomicrobiaceae (Verrucomicrobiae, 1%) and one unclassified family (6%, β-Proteobacteria, unclassified Burkholderiales,Methylibium sp.) (Table S1).

Representatives of most of these families occurred in all three depth zones investigated; however, affiliation with δ-Proteobacteria,γ-Proteobacteria, and Verrucomicrobiae was restricted to the upper peat soil (0–10 cm) while fhs sequences related to Acidobacteria and Clostridia were observed only in the deepest soil layer below 30 cm. Using a 97% sequence identity cutoff based on deduced FTHFS amino acid sequences, a total of 43 OTUs were observed with the highest number of detected OTUs (25) in 0–10 cm depth, followed by depth 10–20 and 30–40 cm (Table 1). The majority of the detected OTUs were restricted to one of these depth zones with the highest number of unique depth-specific OTUs occurring in zone 0–10 cm (Table 1).

Results of fhs clone-library analysis for the three different depth layers investigated

Depth	# Sequences	# Detected OTUs	Coverage%	OTUs unique to this depth
0–10 cm	46	25	63.0	17
10–20 cm	48	20	72.9	13
30–40 cm	26	10	69.2	7

OTU assignment was based on deduced FTHFS amino acid sequences using a 97% sequence identity cutoff.

Phylogenetic trees based on 226 amino acid positions were calculated using neighbor-joining, parsimony, and maximum likelihood methods, which yielded similar tree topologies. The sequences obtained in this study formed 11 clearly distinguishable clusters with high bootstrap support, while six additional phylotypes were only represented by single sequences (Fig. 3). The clusters included sequences that shared a sequence identity of 76% or more on amino acid level. The largest fraction of sequences fell into two clusters related to sequences obtained from the marine subsurface or to Granulibacter bethesdensis (‘fen soil cluster 2 and 3’) and into one more distinct cluster (‘fen soil cluster 4’), each cluster comprising sequences from at least two of the three investigated depth zones. The observed sequence identity cutoff for cluster formation agreed well with the cutoff of 76.4% recently suggested by Hunger et al. (2011) for species definition of acetogens based on FTHFS protein sequences. A cluster-based analysis of depth-dependent changes in FTHFS phylotype diversity partly confirmed the results obtained with the 97% cutoff for OTU formation: Phylotypes in the depth zones 0–10 cm were affiliated with 11 clusters while phylotypes in depth zone 30–40 cm represented only six clusters. In depth zone 10–20 cm, the number of represented clusters was similarly low (7); however, the number of different phylotypes based on the 97% sequence identity cutoff (20 OTUs) was comparable to depth zone 0–10 cm (25).

Phylogenetic tree of deduced FTHFS amino acid sequences (226 amino acid positions) constructed as a consensus tree of neighbor-joining (5000 bootstraps), parsimony (5000 bootstraps), and maximum likelihood analysis (100 bootstraps). Partial FTHFS sequences obtained in this study (bold and framed) and FTHFS sequences of known acetogens are shown (bold and italic). Bootstrap values of more than 70% obtained for both the maximum likelihood and the parsimony tree are given at the nodes. Further information given along with the sequences obtained in this study: within brackets: number of sequences (n), acetogen similarity score (HS), behind brackets: sampling depth (I: 0–10 cm, II: 10–20 cm, IV: 30–40 cm). The Methylobacterium cluster consists of the accession numbers: AY279316, CP001298, CP000908, FP103042, and CP000943. Fen soil cluster 1 consists of Actinobacteria, α-, β-, and δ-Proteobacteria. The bar indicates 10% sequence divergence.

Discussion

Potential contribution of acetogenesis to acetate formation

H₂-dependent growth under chemolithotrophic conditions is considered to be a hallmark of acetogens. Nonetheless, this process is thermodynamically difficult because of the thermodynamic constraints of the entry-level redox reactions for CO₂ in the acetyl-CoA pathway as the standard redox potential of the CO₂/CO half-cell (−520 mV) is approximately 100 mV more negative than that of the 2H⁺ + 2e⁻/H₂ half-cell (−420 mV) (Drake et al., 2006). The H₂ threshold of Acetobacterium woodii using CO₂ as terminal electron acceptor is 23 μM that equals 520 parts per million (ppm), which is higher than those for H₂-utilizing sulfate reducers or methanogens (Cord-Ruwisch et al., 1988). In microcosms supplemented with high concentrations of H₂/CO₂, immediate H₂ consumption led to a formation of acetate but not of CH₄, demonstrating that acetogens and not methanogens can respond rapidly to high H₂ concentrations. Assuming that H₂ was exclusively consumed by acetogenesis, the yield of acetogenic activity was calculated for time intervals of continuous substrate depletion and simultaneously increasing acetate concentrations. Four mol H₂ can theoretically lead to the production of 1 mol acetate via acetogenesis (Kerby & Zeikus, 1983), which means that from 10.7 mol H₂ consumed, theoretically 2.7 mol of acetate could be formed. However, only 37% of this value were detected independent of the peat depth suggesting additional more competitive sinks for H₂ like Fe(III)- or sulfate reduction. The reduction of Fe(III) is very likely, because the upper 0–10 cm of this fen soil are especially enriched in Fe(hydr)oxides (Küsel et al., 2008) that are available for microbial reduction (Reiche et al., 2008). Small amounts of sulfate might have been also available in our microcosms, because sulfate concentrations peak regularly during winter, reaching the concentrations of up to 140 μM in the upper 0−10 cm declining with depth (Küsel et al., 2008). Nonetheless, based on the concentration differences expected, the solid Fe(III) pool should be the more important electron sink in this fen.

Acetogenesis by A. woodii is associated with a carbon isotope effect of ε_ace/DIC −60.0‰ (Gelwicks et al., 1989), with ε_ace/DIC defined as (α_ace/DIC − 1) where α_ace/DIC equals (¹³C_ace − ¹²C_ace)/(¹³C_DIC − ¹²C_DIC). Thus, H₂-dependent acetogenesis from DIC (−14.1‰) can be expected to yield acetate with a δ¹³C value of −69.9‰. In our microcosm experiment amended with H₂, the δ¹³C value of acetate was −37.2‰ and the clear ¹³C-depletion relative to TOC (−26.6‰) points to active acetogenesis. We estimated that acetogenesis contributed 24–30% of the pore-water acetate pool based on a two-source isotope mixing model (e.g. Heuer et al., 2009) and the assumptions that (1) the δ¹³C value of acetate from acetogenesis is −69.9‰, (2) fermentation of peat organic matter (−26.6 ± 0.4‰) goes along with little isotope fractionation (< 3‰) and yields acetate with δ¹³C-values ranging from −27.0‰ to −23.2‰, and (3) the isotopic composition of residual acetate in the microcosm experiment remains unaffected by processes consuming acetate, in particular, because there was no indication for acetoclastic methanogenesis.

Assuming homoacetogenic glucose utilization in the glucose-amended peat microcosms, 6.9 mmol acetate could theoretically be produced according to the 1 : 3 glucose-to-acetate ratio (Fontaine et al., 1942; Barker & Kamen, 1945). However, glucose consumption led to the production of H₂ and ethanol together with short-chain organic acids including acetate, which indicated that the observed glucose consumption primarily resulted from the activity of fermenting microorganisms (Gottschalk, 1986). In addition, acetate produced in the glucose-amended microcosms was only slightly depleted in ¹³C when compared with glucose (−14.9‰ vs. −11.0‰), again pointing to fermentation as the primary pathway of acetate formation. More specifically, we would expect homoacetogenesis to yield a 1 : 2 mixture of acetate from acetogenesis and fermentation with a δ¹³C value of −30.5‰ to −29.0‰ (based on the two-source isotope mixing model and assumptions outlined above and on the carbon isotopic composition of glucose), but the actually observed δ¹³C-acetate of −14.9‰ suggests that the contribution of acetogenesis was only 7–10%. Nevertheless, the main function of the acetyl-CoA pathway during homoacetogenic conversion of sugars is only the recycling of reduced electron carriers coupled to the reduction of CO₂ to synthesize the third mole of acetate, as the eight reducing equivalents are collectively generated during glycolysis and the oxidation of pyruvate (Drake et al., 2006).

The presence of primary fermenters in this fen was demonstrated by 16S rRNA gene-based analyses of the microbial communities in the peat soil, which revealed the sequences belonging to Hafnia alvei, a Enterobacteraceae fermenting both glucose and xylose (Wüst et al., 2009). In addition, fermenters belonging to Clostridiaceae,Acidobacteria (Wüst et al., 2009), and Actinomycetales (Hamberger et al., 2008) were identified.

The accumulation of acetate in an acidic environment like this fen could be toxic, because undissociated acetic acid can permeate the cellular membrane at an extracellular pH_out of 4.7 leading to a decoupling of the membranous proton motive force (Luli & Strohl, 1990). However, other alkalinity-generating anaerobic processes can counteract this effect by an increase of pH.

In our microcosm experiment, acetogens might be involved also in the flow of reducing equivalents derived from glucose via the subsequent consumption of H₂ released by the primary fermenters. Indeed, acetogens can even grow commensally with fermenters on sugars via the interspecies transfer of H₂ formate and lactate (Göβner et al., 1999). Although not statistically confirmed, H₂-dependent acetogenesis would explain the observed second small increase of acetate concentrations detected during the incubation as well as the slight ¹³C-depletion of acetate relative to glucose. Our microcosm results point to a prevailing role of acetogens in H₂ utilization and a restricted contribution of acetogens to fermentable sugar consumption during the anaerobic degradation of organic matter in peatlands. However, the high H₂ concentrations supplied or even obtained during amended glucose metabolism might not reflect field conditions. Indeed, concentrations of H₂ in the peat pore water are low (usually below 1 nM) in plant mesocosms from the same study site (Knorr et al., 2008). Under these low H₂ concentrations, acetogens should be outcompeted by sulfate reducers and methanogens, suggesting that our experimental design led to an overestimation of the in situ contribution of acetogenesis to acetate formation. However, the capacity of an acetogen to compete for H₂ can increase significantly when alternative electron acceptors other than CO₂ are utilized such as aromatic acrylates (e.g. caffeate or ferulate) (Drake et al., 2006). The H₂ threshold of A. woodii is 520 p.p.m. when CO₂ is utilized as the terminal electron acceptor but decreases to 3 p.p.m. when caffeate is used (Cord-Ruwisch et al., 1988). Similar aromatic compounds should be present in humic-rich peatlands. In addition, microenvironments with elevated H₂ concentrations within the peat soil could provide habitats with more favorable conditions for H₂-dependent acetogens, and our experiments demonstrated that acetogens respond more rapidly to elevated H₂ concentrations than methanogens. In addition, low temperatures, which prevail at this site, are known to favor acetogenesis from H₂/CO₂ over methanogenesis (Conrad et al., 1989).

Presence and potential metabolic function of fhs genes

Analysis of the potential acetogenic community in the peat soil using FTHFS as a molecular marker revealed the presence of a phylogenetically diverse microbial community that harbored the fhs gene. FTHFS catalyzes the formation of 10-formyltetrahydrofolate resulting in acetyl-CoA. However, the fhs gene is also present in nonacetogens, because metabolic reactions that bear close biochemical resemblance to the acetyl-CoA pathway of acetogens are utilized by nonacetogenic bacteria (e.g. sulfate-reducing bacteria) and members of the domain Archaea (e.g. methanogens) for either the assimilation of CO₂ (i.e. carbon) into biomass or the oxidation of acetate (Drake et al., 2006). Thus, this molecular marker does not always allow unambiguous conclusions regarding the phylogenetic affiliation of the corresponding microorganisms or their metabolic functions. Highest fhs gene diversity in combination with the highest number of unique depth-specific phylotypes was observed in the surface-near soil layers (Table 1). This high diversity was observed both on the microdiversity level, when using a 97% sequence identity cutoff for OTU assignment, as well as on the cluster level using a 76% identity cutoff. These findings suggest that the small-scale heterogeneity of microhabitats in the peat soil may support a high number of different species with different metabolic functions but also a high number of different ecotypes with adaptations to differences in oxygen availability or to the spectrum of available carbon sources. Field observations have shown that the upper soil layers are subject to fluctuations in aeration caused by drying or snow melt events (Küsel et al., 2008; Reiche et al., 2009). In addition, hydrogeological modeling suggested dynamic horizontal water movements in the fen, which, together with the micro-topography of the fen, may result in variable geochemical conditions on both temporal and spatial scales (Frei et al., 2010). Plant roots, especially those of sedges and other typical fen species, might locally provide oxygen and carbon substrates in their rhizospheres, supporting the presence of a more diverse microbial community of aerobic, microaerophilic, and anaerobic microorganisms compared with peat soil layers not affected by plant roots. Despite being classically referred to as strict anaerobes, acetogens have been isolated from plant roots and other habitats subjected to O₂ fluctuations (Peters & Conrad, 1995; Wagner et al., 1996; Kuhner et al., 1997; Göβner et al., 1999; Küsel et al., 2001). Acetogens are able to cope with oxidative stress because of their capacity to reductively remove O₂, peroxide, and superoxide (Küsel et al., 2001; Karnholz et al., 2002). Because many enzymes of the acetyl-CoA pathway are extremely sensitive to O₂, certain acetogens can metabolically bypass the need to use the acetyl-CoA pathway by shifting the flow of reductant toward catabolic processes that are less sensitive to O₂. Clostridium glycolicum RD-1, an aerotolerant acetogen isolated from sea grass roots, uses acetaldehyde, pyruvate, and protons as terminal electron acceptors when conditions become increasingly oxic, and ethanol, lactate, and H₂ become the reduced end products (Küsel et al., 2001). Indeed, the ability of certain acetogens to cope with oxidative stress appears to be maximized when sugars or other fermentable substrates are available (Drake et al., 2006).

The generally low similarity of the fhs gene sequences obtained in this study to those of cultured microorganisms points to the presence of novel species involved in carbon cycling in acidic peat soils. The majority of sequences obtained in this study were not affiliated with the cluster A of fhs sequences belonging to ‘true’ acetogens as described by Lovell & Leaphart (2005). This finding was further supported by the fact that most of the FTHFS protein sequences obtained in this study had HS score values below 60%, suggesting that they belong to nonacetogenic microorganisms (Henderson et al., 2010). In case of a low HS score, FTHFS is assumed to play a catalytic role in acetate-oxidizing sulfate-reducing bacteria that utilize the acetyl-CoA pathway in reverse (Lee & Zinder, 1988; Schauder et al., 1989). However, our results of sequence analysis do not necessarily exclude acetogenic functions of the corresponding organisms as the Lovell cluster A of FHTFS sequences was recently found to be inconsistent (Ottesen & Leadbetter, 2010). In fact, fhs sequences of some acetogens may be closely related to fhs sequences of nonacetogens as found for Clostridium sp. CA6 (Henderson et al., 2010) but also vice versa for Treponema azotonutricium ZAS-9 (Salmassi & Leadbetter, 2003). Additionally, discoveries of taxonomically and physiologically hitherto unknown acetogens are not rare as illustrated by the recent isolations of novel acetogenic bacteria like Alkalibaculum bacchi (Allen et al., 2010) and Moorella perchloratireducens (Balk et al., 2008) and of even acetogenic archaea (Lessner et al., 2006; Henstra et al., 2007).

Although only HS scores > 90% were assumed to be indicative of true acetogens (Henderson et al., 2010), our relatively high HS scores for an Eggerthella-related sequence (clone MI_64; 74%) and a sequence related to Clostridium cylindrosprorum (73%) might also point to an acetogenic function of the corresponding organisms given a possible underestimation of the HS score, as not all defined amino acid positions could be included in the analysis. Moreover, the interpretation of HS scoring is limited by the fact that FTHFS protein sequences are only available for some of the ~ 100 acetogenic species that are currently known, and fhs genes of yet uncultured acetogenic bacteria and archaea still need to be integrated into the databases that form the basis for HS scoring.

Full coverage of the acetogenic community with the current primer set cannot be expected and might provide an explanation why only few acetogens were detected. The FTHFS primer pair (Leaphart & Lovell, 2001) has been applied in various environments (Pester & Brune, 2006; Matsui et al., 2010; Ottesen & Leadbetter, 2010). However, because also a broad range of nonacetogenic FTHFS sequences are targeted, its ability to analyze the acetogenic community at a given site is limited. Most promising seems the approach to combine the analysis of the fhs gene with the analysis of another gene encoding an enzyme of the acetyl-CoA pathway, the acetyl-CoA synthetase, as recently suggested by Gagen et al. (2010). Interestingly, the fhs gene diversity observed in this study was higher than that reported from a stable isotope probing (SIP) experiment where peat material from the same fen site was incubated with ¹³C-labeled formate (Hunger et al., 2011). Here, differences in the fhs-bearing microbial communities between these two studies might be due to community shifts during the 38-day laboratory incubation in the SIP study; to spatial heterogeneities in the distribution of fhs gene phylotypes across the fen site or the fact that SIP emphasized the active community while DNA-derived clone libraries include gene homologs from inactive, fhs-bearing microorganisms.

Biogeochemical evidence for in situ turnover of acetate

Pore-water concentrations of acetate were in the range of the depth-integrated average of 82 μM of previous investigations at the site (Küsel et al., 2008). In principle, the isotopic composition of pore-water acetate is controlled by (1) the isotopic composition of its precursors, (2) the isotopic fractionations associated with its synthesis and consumption, and (3) the relative rates of all processes that influence its pool size (Gelwicks et al., 1989). Acetate resulting solely from fermentative processes is expected to have a similar carbon isotopic composition as the substrate. For example, under low H₂ partial pressure fermentation of saccharides by Clostridium papyrosolvens has been shown to result in the production of acetate that is slightly enriched in ¹³C (< 3‰) compared with the saccharides (Penning & Conrad, 2006). In the Schlöppnerbrunnen fen, where the TOC of peat has an average δ¹³C value of −26.6 ± 0.4‰, we expect fermentation to yield acetate with δ¹³C values ranging from −27.0‰ to −23.2‰. In contrast, with δ¹³C-DIC ranging from −15.1‰ to −9.2‰ and a kinetic isotope effect of ε_ace/DIC −60.0‰ (Gelwicks et al., 1989), H₂-dependent acetogenesis can be expected to produce acetate with δ¹³C values between −70.8‰ and −65.3‰. Thus, with varying contributions of fermentation and acetogenesis, acetate production is expected to result in δ¹³C values ranging from −70.8‰ to −23.2‰. However, in this study, we observed δ¹³C values of acetate averaging around −12.3‰ in February 2009 and −17.2‰ in June 2010. Isotopically heaviest acetate coincided with an almost depleted pore-water acetate pool or over depth declining acetate concentrations (Fig. 2a and f). The ¹³C-enrichment of pore-water acetate relative to TOC is likely to result from isotope effects associated with acetate consumption and points to a strong impact of acetate sinks on the pore-water acetate pool (Fig. 4). In particular, the ¹³C-enrichment suggests that the concentration minimum in the pore-water profile of acetate is related to strong consumption rather than low production, underlining the central metabolic role of acetate in this fen. Acetate released from the roots of vascular plants such as Carex spp. and Eriophorum spp. (Strom et al., 2003) would have no impact on δ¹³C_ace as its isotopic composition is comparable with δ¹³C_TOC (Boutton, 1991).

Theoretical model of sources and sinks for acetate including the involved processes (italic) and microbial guilds (dark boxes). δ¹³C values are given in brackets below important carbon compounds. Proposed carbon isotopic fractionations are presented in circles. Based in part on McInerney & Bryant, 1981. References: ^aGelwicks et al., 1989; ^bPenning & Conrad, 2006; ^cHeuer et al., 2010; ^dGoevert & Conrad, 2008; ^eConrad 2005.

Earlier peat microcosm studies using methyl fluoride as selective inhibitor to determine acetoclastic methanogenesis suggested that the proportion of acetoclastic methanogenesis varied between 42% and 90% in surface-near peat layers of this site (Reiche et al., 2008). Here, δ¹³C values of pore-water CH₄ approximated −50.3‰ at 5 cm depth in contrast to a steady mean δ¹³C-value of −61.7‰ below 30 cm. δ¹³C_CH4 declines and δ¹³C_DIC increases in the upper 10 cm in June 2010 as δ¹³C_CH4 decreases while δ¹³C_DIC is stable with depth. This pattern implicates that acetoclastic methanogenesis was predominant in shallow surface layers and hydrogenotrophic methanogenesis possibly was the main precursor for CH₄ in older, less reactive and deeper peat layers similar to observations made by Hornibrook et al. (1997, 2000). Acetate is often reported as the main source of CH₄ especially in upper younger fen soil (Popp et al., 1999; Chasar et al., 2000), where easily available organic matter is introduced by growing plants and oxic/anoxic cycling leads to higher metabolic dynamics. The presence of Carex-dominated plant communities might also favor acetoclastic methanogenesis, as the presence of these plants is often related to methanogenic microbial activity using this pathway (Rooney-Varga et al., 2007). Consumption of CH₄ is indicated by the low CH₄ concentrations together with high δ¹³C values of CH₄ in the upper 10 cm of the peat soil (Fig. 2g and k). Carbon isotopic fractionation during aerobic CH₄ consumption varies between 5‰ and 30‰ (Whiticar, 1999), and methanotrophic bacteria could intercept CH₄ originating from deeper soil layers near the peat surface (Nedwell & Watson, 1995). Even anaerobic oxidation of methane can occur simultaneously with methanogenesis in humic acid-rich minerotrophic fens (Smemo & Yavitt, 2007). Pore-water CO₂ concentrations at our study site were lower than found for other peatlands (Hornibrook et al., 1997, 2000) because of the low pH at our site. Where highest CO₂ concentrations coincided with lowest δ¹³C values of DIC and high CH₄ concentrations (10 cm depth, Fig. 2g, h and j), hydrogenotrophic methanogenesis might occur. Calculations of Gibbs free energies in a previous study showed that in peat mesocosms, hydrogenotrophic methanogenesis was a feasible process close to the water level with a maximum of CH₄ production (Knorr et al., 2008).

Other potential sinks for acetate in this fen are Fe(III) and sulfate reduction that have been shown to even co-occur with methanogenesis (Loy et al., 2004; Paul et al., 2006; Küsel et al., 2008). Although sulfate in the pore water was almost depleted (concentration < 30 μM) in June 2010, activity of sulfate reducers such as Syntrophobacter,Desulfobacteraceae, and Desulfomonile might have prevented acetate accumulation especially during winter as suggested in earlier studies (Loy et al., 2004). Depending on the metabolic pathway used for acetate oxidation, namely the acetyl-CoA pathway or the tricarboxylic acid cycle, sulfate reducers show an acetate isotopic fractionation ε_{Ace- substrate} of −19‰ or −2‰, respectively (Goevert & Conrad, 2008). In particular, the former pathway could explain the observed ¹³C-enrichment of the pore-water acetate pool. Maximum Fe(II) concentrations (1300 μM) coincided with concentration minima of acetate, suggesting the participation of acetate-utilizing Fe(III) reducers in the isotopic fractionation of acetate observed in 2010 and possibly also in 2009 when no CH₄ could be detected. Earlier incubation experiments demonstrated that rates of Fe(II) formation are stable during the year, and methanogenesis begins mostly when Fe(II) formation reaches a plateau (Reiche et al., 2008). Another sink for acetate could be its syntrophic oxidation. Although this process is thermodynamically more favorable at high temperature, some lines of evidence exist that this process can also take place at lower temperatures (Hattori, 2008). However, syntrophic oxidation of acetate might only lead to low isotopic fractionation and is therefore hard to detect (Conrad & Klose, 2011). These collective results suggest that multiple sinks for acetate are potentially present at this acidic fen where environmental factors such as water table, O₂ input, and plant coverage additionally alter the ongoing processes.

Conclusion

Our results have clearly shown that acidic peat soil harbors acetogenic potential, especially when supplemented with H₂/CO₂. However, both microcosm experiments and δ¹³C analysis of inorganic and organic compounds in the peat pore water indicated that acetate is mainly derived from fermentative processes rather than from acetogenesis under field conditions. In the field, isotopic fractionation further points to strong acetate sinks, underlining the central role of acetate in ‘intermediary ecosystem metabolism’. Analysis of the potential acetogenic microbial community in the fen soil using FTHFS as a molecular marker revealed a diverse microbial community harboring the fhs gene especially in the upper 10 cm of the peat profile. However, only few sequences were distantly related to known acetogens. To gain unambiguous insight into the genetic potential for acetogenesis in this fen, more research is needed focusing on the link between the phylogeny of this functional marker and the metabolism of the corresponding microorganisms.

Acknowledgements

This work is part of the research group FOR 562 ‘Dynamics of soil processes under extreme meteorological boundary conditions’ kindly supported by the Deutsche Forschungsgemeinschaft DFG. Stable carbon isotope analysis and the contribution of V.B.H. were supported by the DFG-Research Center/Cluster of Excellence ‘The Ocean in the Earth System’ (MARUM) and Deutsche Forschungsgemeinschaft project grant Hi 616/9-2.

Supporting Information

Additional Supporting Information may be found in the online version of this article:

Table S1. Class- and family-level identities of fhs gene clones from fen Schlöppnerbrunnen.

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

Author notes

Alfons Stams