Global Biogeochemical Cycles

Anaerobic microbial biogeochemistry in a northern bog: Acetate as a dominant metabolic end product

Authors


Abstract

[1] Field measurements and incubation techniques were used to determine the dynamics of acetate formation, iron reduction, and methanogenesis in surficial peat of an Alaskan bog. Acetate concentrations were ∼100 μM early in the season and decreased to ∼20 μM in July when the water table decreased. Acetate levels increased rapidly to ∼1000 μM when the water table rose to the surface in August. Acetate production in anaerobic slurries occurred at rates of 2.8–420 nmol carbon mL−1 day−1, which was 7–120 times more rapid than CH4 production. Experiments utilizing 14C-acetate confirmed that methanogenesis was not acetoclastic although acetate was converted very slowly to CO2. Peat incubated anaerobically for 4.5 months at 24°C never produced methane from acetate, suggesting that anaerobic acetate accumulation would have occurred all season if the water table had remained high. CO2 production was the most rapid process measured in laboratory incubations (up to 750 nmol mL−1 day−1) and appeared to be due primarily to fermentation. Acetate was the primary organic terminal product of anaerobic decomposition in the bog, and acetate was ultimately oxidized to CO2 via aerobic respiration and to a much lesser extent anaerobically by Fe reduction.

1. Introduction

[2] Wetlands occupy a large portion of the global terrestrial landscape of high latitude environments [Matthews and Fung, 1987; Ping and Moore, 1992]. These environments represent a large carbon reservoir that is a potentially important source of CO2 if warming occurs [Olson et al., 1985]. Studies have shown that tundra ecosystems are changing from sinks to sources of CO2 [Oechel et al., 1993] as a result of increasing global temperatures. Northern wetlands are also sources of CH4, comprising up to one third of the global wetland source [Fung et al., 1991]. Consequently, climatic changes will affect the role of northern wetlands in the exchange of trace gases.

[3] Northern soils are often waterlogged, and under these conditions decomposition of organic C below the very surface layer occurs mainly via anaerobic metabolism. Since these systems are often poor in S, N, and oxidized Fe, the terminal step in decomposition is usually methanogenesis. Biogenic CH4 in nature is primarily derived from acetate fermentation (acetoclastic methanogenesis) and from H2 oxidation coupled with CO2 reduction (H2/CO2) [Oremland, 1988]. Acetate is considered the most important carbon intermediate in anaerobic systems, where it rarely accumulates, but rather turns over rapidly due to uptake by bacteria at the end of the microbial food web [King et al., 1983; Lovley and Klug, 1983]. However, acetate has been shown to accumulate transiently in some freshwater and marine sediments due to temporal separation of acetate producing and consuming processes [Avery et al., 1999; Hoehler et al., 1999; Sansone and Martens, 1982; Shannon and White, 1996]. In most anaerobic sediments, approximately two thirds of CH4 produced is the result of acetoclastic methanogenesis [Jones and Simon, 1985; Lovley et al., 1982; Wu et al., 1997]. However, some environments, such as bogs and poor fens, may favor methanogenesis from H2/CO2 [Chasar et al., 2000; Conrad, 1999; Lansdown et al., 1992; Popp et al., 1999]. In addition, although methanogenesis is thermodynamically favored over homoacetogenesis (CO2 + 2H2 → CH3COOH), some habitats support an active acetogenic bacterial population in lieu of a methanogenic one, i.e., some flooded soils [Kusel and Drake, 1994], and animal intestines [Brauman et al., 1992; Degraeve et al., 1994]. The cause of this dominance is unknown.

[4] The present study was undertaken to investigate the anaerobic microbial biogeochemistry of a boreal bog system. In particular, we combined field and laboratory measurements to determine the role of acetate in anaerobic carbon decomposition, and its relative importance with regard to methanogenesis and Fe(III) reduction in bog peat. In this system, acetate was a relatively stable end product of anaerobic organic decomposition, and its accumulation far exceeded carbon flow to methane.

2. Methods

2.1. Site Description

[5] Samples were collected during the growing season of 1999 in Turnagain Bog, which is a 200-ha ombrotrophic convex peatland located in Anchorage, AK, (60°10′N 149°11′W). Summer and winter temperatures in the area vary from 8° to 19°C and from –16° to 6.4°C, respectively, with a mean annual temperature of 2°C. The study area was a Sphagnum moss-Myrica gale (Sweet gale) region in the center of the bog containing plants typical of ombrotrophic bogs. The surface layers (2–4 cm) of the bog contain actively growing Sphagnum moss, whereas the layers below contain brown, decaying peat. The in situ pH is 4.66–5.1. A detailed description of the bog is given by Hogan and Tande [1983]. A boardwalk was constructed in July 1998 to provide for minimal disturbance throughout the study area. The bog remained frozen through mid-May, 1999. Within 5 days of spring thaw, temperature sensors, a water level sensor, and pore water samplers were placed in the bog and a barrier put in place to discourage moose traffic.

2.2. In Situ Measurements

[6] Water level in the bog was determined by using a water level sensor and data logger system (WL300, 0–91.4 cm range, Global Water, Gold River, California) inserted into a well made of slotted PVC pipe (152.4 L × 5.08 cm I.D.). Water level measurements were recorded hourly and are reported as daily averages. The surface of the peat did not change appreciably with respect to the well during the study; therefore no correction for peat movement was made when calculating the water level relative to the peat surface. Temperature was determined at depths of 2, 4, 6, 10, 16, and 26 cm by using temperature sensors and data loggers (Onset Computer Corp., Pocasset, Massachusetts). A sensor was also placed in the shade underneath the boardwalk to record ambient air temperature. Temperature was recorded hourly and reported as daily averages. Precipitation data were obtained near by from the Anchorage International Airport station of NOAA's National Climate Data Center.

2.3. Pore Water Analysis

[7] Pore water was collected weekly from Teflon "sippers" that consisted of porous filter units designed to filter eluent for ion chromatography (Dionex Corp., Sunnyvale, California). These units were connected to 1/8 inch Teflon tubing and placed within the peat at depths of 2, 4, 6, 10, 16, 20, and 26 cm. The terminal porous portion (4 cm in length) was wrapped in Teflon tape so that water was collected only through a 1-cm window at the tip. Pore water for ion analysis was collected by using a syringe attached to the tubing. After rinsing the tubing and syringe with ∼10 mL of pore water, 1-2 mL of water was collected and filtered through a 0.2-μm syringe filter (13 mm, IC Acrodisc, Pall Gelman Sciences, Ann Arbor, Michigan) into clean distilled H2O rinsed plastic scintillation vials. Vials were held on ice for transport to the lab and analyzed for acetate, SO42− and NO3 concentrations within 5 hours of collection. Water for Fe(II) analysis (1 mL) was immediately filtered through a 0.45-μm nylon filter (13 mm, Acrodisc) into microcentrifuge tubes containing 10 μl of concentrated nitric acid.

[8] Acetate was measured by injecting 25-μL samples into a Dionex 1200 ion chromatograph with an AS14 4-mm ion exclusion column with suppression and with a sodium carbonate eluent. Acetate standards were prepared fresh in the laboratory from sodium acetate. Sulfate and NO3 were measured in the same samples and concentrations determined by comparison to dilutions of Dionex 5-anion standard. Reduced Fe was measured colorimetrically by using Ferrozine [Stookey, 1970].

2.4. Laboratory Incubation Experiments

[9] To examine anaerobic microbial activity within the bog, peat was collected periodically (usually 1–2 week intervals) throughout the season in a manner similar to that described by Kiene and Hines [1995]. Samples were collected from Sphagnum depressions. Green surficial Sphagnum moss was discarded (upper ∼2–4 cm) and a sample of brown peat (∼5 × 5 x 20 cm) was removed with a knife and placed in a 1-L glass jar. The jar was filled to the top with bog water so that no air pockets were present, and sealed with airtight lids. Jars were placed on ice for transport to the laboratory. Generally, samples were used immediately to create slurries. Occasionally, samples were stored at 4°C for 1–2 weeks prior to use.

[10] Slurries were prepared by draining excess water from the peat, sorting to remove woody plant roots and green Sphagnum, and placing 100–300 g (wet weight) into a blender. Bog water was added and the mixture homogenized so that a ratio of ∼1:3 wet weight peat:total volume slurry was achieved. Slurries prepared in this manner contained 0.04–0.08 g dried peat mL−1 slurry. Aliquots (50 mL) of slurry were transferred to 72-mL vials, and the vials were sealed with Teflon-lined butyl rubber septa, sealed with aluminum crimps, and flushed with N2 for 5–10 min. Slurries were preincubated in the dark for 12–18 hours. Following preincubation, vials were flushed with N2 again prior to analysis. All incubations were conducted in the dark at in situ temperatures (determined by averaging the 6–26 cm depth temperatures in the bog on the day peat was collected). Vials were incubated in the inverted position to minimize the possibility of oxygen entry through the stoppers. An equal volume of N2 was injected into the vials upon removal of samples for CH4, CO2, acetate and Fe(II) analysis. Replicate vials were amended with 10 mM 2-bromo-ethanesulfonic acid (BES, Sigma-Aldrich, St. Louis, Missouri), a known inhibitor of methanogenesis, by adding 1 mL of a stock solution of BES (500 mM, pH 5.0). Additions of BES had no impact on the pH of the peat slurries, which averaged 5.35 ± 0.2.

[11] Headspace samples for CH4 analyses (1 mL) were removed by a 1.0-mL plastic syringe. Samples were injected into a Shimadzu GC-9A gas chromatograph equipped with a 6 port valve and sample loop, a flame ionization detector, and a stainless steel column (25°C) packed with Poropak Q. Helium was used as carrier gas at a flow rate of 35 mL min−1. Total CH4 data were corrected for partitioning between gas and liquid phases and are reported as total μmol CH4 L−1 slurry. Carbon dioxide was measured by removing headspace samples (0.5 to 1.0 mL) from vials by a gas-tight Hamilton microsyringe with a side bore needle and injecting into a Shimadzu GC-14A gas chromatograph equipped with a thermal conductivity detector (injector/detector temperature = 90°C; TCD = 80°C) and a Teflon column (1/8 inch, 30°C) containing Poropak Q. Helium was used as carrier gas at a flow rate of 20 mL min−1. Total CO2 was calculated in the manner described for CH4 analysis.

[12] Slurry samples (1 mL) for acetate analysis were collected from vials by a plastic 1.0-mL syringe. Samples were centrifuged (12,000 × g) for 5 min to pellet peat, and the supernatant was filtered through a 0.2-μm IC syringe filter and frozen until analysis within 1 week of collection. Samples were thawed on ice and analyzed for acetate and other ions as described above. Slurry samples (0.5–1.0 mL) were also removed for Fe(II) analysis. Microcentrifuge tubes (0.6 mL volume) were totally filled and centrifuged to removed peat as described for acetate. The supernatant was filtered through a 0.45-μm syringe filter, fixed in nitric acid, and analyzed for Fe(II) as described above. For all analysis, triplicate vials were sampled at 24- to 48-hour intervals for 5–7 days and rates calculated from linear regression analyses.

[13] Occasionally, following the incubation of slurries for 5–7 days, vials were opened and contents pooled into a beaker in which the headspace was flushed continuously with N2. While gently mixing, aliquots of slurry (3–4 mL) were transferred to vials (13-mL capacity) that were continuously flushed with N2. After aliquots were transferred, vials were sealed with Teflon-lined gray butyl rubber septa and aluminum crimps and the headspace flushed with N2 for 10 min. The vials were incubated for 12–18 hours at in situ temperature, after which the headspace was flushed with N2 again and 2-14C-acetate (∼0.1 μCi) added. Duplicate or triplicate vials were sacrificed periodically (usually 8- to 24-hour intervals) for 24–48 hours by adding 1 mL of 6N HCl to the vials and shaking overnight. Headspace samples were then analyzed on a GC equipped with a gas proportional counter (RAGA, Raytest Inc., Germany) for 14CH4 and 14CO2.

[14] A depth profile was conducted in which a peat sample, 5 × 5 × 26 cm, with the green Sphagnum layer intact, was collected and returned to the laboratory. The peat was carefully cut into 6-cm sections representing depths of 0–6, 6–12, 12–18, and 18–24 cm. The 0- to 6-cm-depth section contained green Sphagnum. Each depth section was slurried as described above and incubated anaerobically at 25°C. Slurry samples were removed daily and analyzed for acetate.

[15] To examine the impact of long term incubation of peat at warmer temperature, the peat was collected and incubated intact with bog water under anaerobic conditions in a 1-L jar at 24°C for 4.5 months. Samples of bog water were removed with a 1.0-mL syringe and analyzed for acetate. The peat was then slurried and transferred to vials as described above and incubated anaerobically at 24°C. Samples of headspace and slurry were periodically removed for CH4 and acetate determinations.

3. Results

3.1. Temperature, Water Level, and Precipitation

[16] The bog water level remained near the surface of the peat throughout early June and then decreased 35 cm by mid-July (Figure 1a). Slight increases in water level were detected after brief periods of precipitation, yet overall the water level steadily decreased. Water level increased again rapidly in early to mid-August, and the peat remained saturated throughout the remainder of the study. Air and surficial peat (2–10 cm) temperatures were high (∼16.5° and 13.5°C, respectively) immediately following thaw. Temperatures decreased in late May and then increased ∼0.5°C day−1 during early June (Figure 1b). The average temperature of the upper 10 cm of peat was approximately 15°C from June to mid-August and decreased steadily at a rate of 0.2°C day−1 from late August to the end of the study.

Figure 1.

Time series plots of (a) water level (thick solid line) and precipitation (thin solid line), and (b) average temperature in peat (2–10 cm, thick line; 16–26 cm, shaded line) and air (thin hair) at Turnagain Bog.

3.2. Pore Water Analysis

[17] During the period when the water level decreased below 15 cm, we were unable to collect pore water from the sipper situated at a depth of 2 cm. However, we were able to collect pore water from sippers situated at 4 cm depth and deeper. Water collected above the water table probably represented water held by Sphagnum via capillary action. The concentrations of SO42− and NO3 remained below 6 μM for the majority of the study at all depths (data not shown). Nitrate increased to greater than 50 μM in the surface layer of Sphagnum (2 cm) on 8 and 28 June, and again in peat at depths of 6 and 16 cm on 5 July and at 16 cm on 30 August. Sulfate increased to greater than 70 μM at 2 cm on 17 May and again at a depth of 6 cm on 5 July. These increases may have been due to deposition from air during precipitation, or regeneration of NO3 and SO42− under aerobic conditions when the water table was low. Despite these infrequent increases in concentration, SO42− and NO3 were not present in sufficient quantities to support significant SO42− and NO3 reducing activity within the bog without recycling. We did not observe any decreases in SO42− in slurry incubations, and NO3 was not detected in slurries.

[18] Pore water acetate concentrations were highest when the peat was water saturated (Figure 2). Concentrations decreased from >150 to <20 μM as the water table dropped through mid-July, presumably due to aerobic oxidation of acetate in oxic peat. The acetate-producing activity was localized to the surficial layers of the peat (upper ∼20 cm), and concentrations increased at a rate of ∼20 μM day−1 in the upper 10 cm following the water level increase in early August. Maximum acetate concentrations occurred at or above 16 cm depth when the peat was saturated and reached levels near 1.0 mM. No acetate accumulation was apparent below a depth of 16 cm. Acetate concentrations decreased in September as peat temperatures decreased.

Figure 2.

Contour plot of acetate concentration in pore water at Turnagain Bog. Sippers were installed at depths of 2, 4, 6, 10, 16, 20, and 26 cm and samples collected through a 1-cm window at the tip. From mid-June to mid-August we were unable to collect water at 2 cm depth.

[19] Dissolved Fe concentrations were affected greatly by even slight changes in precipitation and water level (Figure 3). Increases in water level in early June and July resulted in increases in Fe concentrations as high as 470 and 300 μM, respectively. Dissolved Fe concentrations decreased to below 50 μM when the water level was lowest due to the oxidation of Fe(II). Following the increase in water level in mid August, Fe accumulated to as high as 600 μM in the surficial layers at a net rate of ∼15 μM day−1. The concentration of Fe at 16–26 cm also increased late in the season, but to a much lesser extent than in surficial layers.

Figure 3.

Same as Figure 2 but for reduced Fe concentration in pore water at Turnagain Bog.

3.3. Laboratory Incubation Studies

[20] Acetate and CH4 were produced simultaneously and increased linearly in laboratory incubations of slurried peat collected when the water table was no more than 15 cm below the surface of the peat (Figure 4). Methane was measured in all vials, including those amended with BES, and methanogenesis was inhibited in the presence of BES (data not shown). Rates of acetate production in the presence of BES were the same as in the absence of BES, indicating that acetate was not used methanogenically. Acetate carbon was produced at rates 7–120 times higher than CH4, indicating that acetate was a dominant metabolic end product in Turnagain Bog peat (Table 1). Carbon dioxide was produced in both the presence and absence of BES. However, CO2 production rates in the presence of BES were 10–30% less than those measured in its absence, indicating that BES had some effect on terminal metabolism.

Figure 4.

Methane (solid circles) and acetate (open circles) production during incubations of slurried peat collected from Turnagain Bog. Error bars represent ±1 standard deviation of the mean of three replicate vials. Absence of error bars indicates that the error is smaller than the symbol.

Table 1. Rate of CO2, Acetate, and CH4 Production in Incubations of Slurried Peat From Turnagain Bog, Anchorage, Alaska
DateT, °CRate,a μmol L−1 day−1Ratio of CO2:Acetate: CH4,C Basis
CO2AcetateCH4
  • a

    Rates were determined using linear regression.

  • b

    Values in parentheses are r2 of the linear regression.

  • c

    Not determined.

17 May8374.1 (1.0)b115.6 (0.97)9.0 (0.98)41 : 26 : 1
8 June12NDc28.2 (0.93)5.6 (0.99)ND : 10 : 1
21 June12604.2 (1.0)211.2 (1.0)3.6 (1.0)168 : 119 : 1
23 Aug.14740.9 (0.99)39.6 (0.98)2.5 (0.97)296 : 32 : 1
27 Sept.10211.8 (0.88)1.4 (0.89)0.4 (0.96)530 : 7 : 1

[21] When the water table was below 15 cm (21 June to 23 August), there was usually a 1- to 3-day lag period before acetate and CH4 accumulated (data not shown). The peat samples were likely oxic or suboxic and the production of CH4 and acetate were inhibited due to re-establishment of anaerobic conditions and the utilization of alternate electron acceptors, most likely Fe, early in the incubation period. Dissolved Fe was measured in incubations of peat collected in August and September and was found to accumulate at rates of 58.6 and 31.2 μM day-1, respectively. Maximum rates of dissolved Fe accumulation occurred during the first 2–4 days of incubation, reaching concentrations of 250–475 μM, after which rates decreased or ceased. Initial concentrations of Fe in slurries were similar to the average concentration measured in pore water on the day the peat sample was collected.

[22] Slurry incubations amended with 2-14C-acetate demonstrated that CH4 production was not acetoclastic (Table 2). At no time during the study was 14C-acetate metabolized to 14CH4, regardless of whether peat was collected when the water level was decreasing (28 June), low (1 August), or high (27 September). Even when pore water acetate levels were decreasing in September, samples did not degrade acetate to CH4. Acetate was degraded very slowly to CO2 during incubations, with <1.2% utilized during the 12- to 24-hour incubation period. On the basis of turnover rate data for 14C-acetate to 14CO2 and concentrations of acetate in slurries, the production of CO2 from acetate accounted for <10% of the total CO2 produced during incubations.

Table 2. Turnover of 2-[14C]Acetate in Slurried Peat From Turnagain Bog, Anchorage, Alaska
DateT, °CRate of Turnover,a %/hourTurnover to, %
14CO214CH4
  • a

    Values in parentheses are the range of duplicate samples.

28 June120.07 (0.017)1000
1 August150.04 (0.02)1000
27 Sept.100.02 (0.004)1000

[23] Intact peat samples incubated anaerobically at 24°C for 4.5 months accumulated acetate to near 10 mM in pore water. High gas pressure, presumably due to accumulation of CH4 and CO2, was evident after the extended incubation, although gas concentrations were not measured. When that peat was slurried and incubated anaerobically in vials, acetate and CH4 accumulated at rates of 200 and 22 μM day−1, respectively (Figure 5).

Figure 5.

Methane (solid circles) and acetate (open circles) production in slurried peat following long-term incubation. Whole peat from Turnagain Bog was incubated anaerobically at 24°C for 4.5 months prior to creating slurries. Error bars represent ±1 standard deviation of the mean of three replicate vials. Absence of error bars indicates that the error is smaller than the symbol.

3.4. Depth Profile of Acetate Production

[24] Substantial net acetate accumulation in anaerobically incubated peat (25°C) was mainly associated with the upper layers of the peat, which contain the live Sphagnum plants (Figure 6). This result suggests that labile precursors available in these upper layers might fuel the majority of acetate production. The accumulation rates measured in slurried peat were similar to those in whole peat (data not shown), suggesting that slurrying per se did not excessively stimulate acetate accumulation. Slurries of peat from 15–20 cm which did not accumulate acetate when unammended did so if Sphagnum from the 0- to 5-cm layer was included in the slurry (data not shown), further suggesting that degradation of the relatively fresh Sphagnum material was responsible for acetate production.

Figure 6.

Rate of acetate accumulation in slurried peat with depth. Whole peat samples were separated to represent depths of 0–6, 6–12, 12–18, and 18–24 cm. Error bars represent ±1 standard deviation of the mean of three replicate vials. Absence of error bars indicates that the error is smaller than the symbol.

4. Discussion

4.1. Acetate as an End Product of Anaerobic Metabolism

[25] Acetate was an important terminal product during anaerobic decomposition in Turnagain Bog, surpassing both methanogenesis and Fe reduction. Acetate is generally considered to be an important intermediate in anaerobic carbon metabolism [Jones and Simon, 1985; King et al., 1983; Lovley, 1982; Lovley and Phillips, 1986]. However, in Turnagain Bog acetate accumulation was a terminal step in anaerobic decomposition (Figures 2 and 4; Table 1). Acetate levels decreased twice during the season, once when the water table dropped, and again at the end of the season when peat temperature was decreasing. The former was likely due to aerobic oxidation of acetate under oxic conditions, since O2 respiration in these peats rapidly consumes the acetate produced in situ [Hines et al., 2001]. The decrease in acetate concentrations in September was likely a result of decreases in acetate production at lower temperature, the diffusion of acetate away from the its source region and subsequent oxidation by O2, and the use of acetate anaerobically by Fe-reducing bacteria. The loss of acetate in September was not due to the initiation of acetoclastic methanogenesis, since we did not observe an increase in acetate formation in the presence of BES, nor did we detect 14CH4 formation from 2-14C-acetate.

[26] It is conceivable that during slurry preparation acetoclastic methanogens were selectively inhibited relative to H2/CO2 utilizers, explaining the lack of acetate use in slurries in the presence of CH4 formation. However, acetate continued to accumulate during the 4.5-month anaerobic incubation (Figure 5) indicating that the lack of acetate use was not due to an ephemeral change in population structure. The rapid increase in acetate in situ in August also supported the incubation results, since the former pore water data were not derived from peats removed from the bog or manipulated in any way. In addition, the rate of acetate accumulation in the field (Figure 2) in August was similar to that in slurries collected in August (Table 1).

[27] The accumulation of acetate carbon in Turnagain Bog peat was 7–120 times more rapid than CH4 accumulation (Table 1), which demonstrates the quantitative significance of acetate as a terminal product of anaerobic metabolism. Rates of CH4 production were low in these peats. However, if the acetate that accumulated was converted quantitatively to CH4, rates of methanogenesis in Turnagain Bog would have been similar to those reported for other wetlands in which much of the CH4 is derived from acetate [Avery et al., 1999; Goodwin and Zeikus, 1987].

[28] Shannon and White [1994, 1996] and Avery et al. [1999], working in a Michigan peatland, reported a transient accumulation of acetate during spring, to over 1.0 mM in some instances, that preceded active acetoclastic methanogenesis. In their system, acetate concentrations in summer were <20 μM, and methanogenesis from acetate accounted for more than half of the CH4 produced during that period. During 1999, we never observed a decrease in acetate in Turnagain bog attributable to methanogenesis, and concentrations of acetate in August approached 1.0 mM. However, the aeration of peat during the water table drop in July could have created a situation similar to what might occur during the spring. The methanogenic population in surficial peat probably decreased in size due to O2 stress. The regrowth of the population as the water table increased again could have led to a sequence in which acetoclastic methanogens did not become established as quickly as those utilizing H2/CO2. The temperature decreased before the acetoclastic methanogens had sufficient time to repopulate, so they never became important contributors to methane formation. When Turnagain Bog peat is compared to the Michigan peatland studied by Shannon and White [1994, 1996], it is also possible that other unknown differences in the chemical and/or physical conditions accounted for the marked differences in the behavior of acetate.

[29] The fact that peat incubated at 24°C for 4.5 months never converted acetate to CH4 suggested that acetate accumulation is a rather permanent feature of this system. During 1998, the water table in Turnagain Bog never fell below the surface. Although we did not conduct a detailed seasonal study that year, we did measure acetate production rates in slurries on several occasions throughout the summer. Acetate accumulated similarly to what is noted here for 1999 and acetate was not converted to CH4 in 1998 even in August and September. Therefore, it appears that peat in Turnagain Bog accumulates acetate throughout the active season. It has been suggested that the transient increase in acetate in temperate bogs is the result of different temperature optima for acetoclastic and autotrophic methanogenic populations [Svensson, 1984] or different growth rates exhibited by acetoclastic methanogens and CO2 reducers [Vogels et al., 1988]. As a consequence, the phenomenon of transient acetate accumulation would be dependent on temperature and the occurrence of anaerobic conditions for a long enough period [Shannon and White, 1996]. Turnagain Bog peats remained about 5°C cooler than the sites studied by Shannon and White [1994]. Hence, the acetate accumulation in Turnagain Bog could be conceived as equivalent to an extended spring in more southern wetlands. It is also possible that an acetoclastic methanogenic population has never matured in Turnagain Bog and a 4.5-month incubation at 24°C was insufficient for one to appear. Whatever the cause, acetate was an important end product of metabolism and remained as the dominant organic terminal product of organic matter decomposition throughout most of the season.

4.2. Acetate and CO2 Production

[30] The origin of acetate in incubation experiments was not clear. Acetate could have originated from fermentation, acetogenesis, or SO42− reduction. Acetate is a product of incomplete oxidation processes by SO42− reducing bacteria [Odom and Singleton, 1993], and this process appears to occur in the wetlands studied by Shannon and White [1996]. However, SO42− reduction is not a source of acetate in Turnagain Bog peat, since SO42− concentrations were very low and did not decrease noticeably in slurry incubations. Autotrophic methanogens generally out-compete acetogens for H2 in anaerobic habitats [Zinder and Anguish, 1992]. However, acetogenic bacteria are able to out-compete methanogens in some habitats, including ones at low temperature. Hence, it is possible that acetate in bog peat was derived from homoacetogenesis.

[31] Carbon dioxide production was the most rapid process measured in incubation vessels (Table 1). Bridgham et al. [1998] reported that CO2:CH4 ratios in incubated bog samples were much higher than in samples of minerotrophic wetlands. The only significant respiration process capable of producing CO2 in these anaerobic peats was Fe reduction. Although Fe reduction was prevalent, Fe reduction rates were not rapid enough to account for more than 15% of the CO2 produced (see below). Sulfate and NO3 respiration were extremely slow or not detected in slurries, and without acetoclastic methanogenesis, CH4 originated almost exclusively from the consumption of CO2. Since no CO2 was produced from acetoclastic methanogenesis, the bulk of the CO2 produced must have been derived from fermentation reactions.

4.3. Implications for Methanogenesis

[32] Methane in Turnagain Bog was derived almost exclusively from H2/CO2. Acetate and H2/CO2 are considered as the primary precursors of CH4 in most anaerobic habitats. However, acetate was not converted to CH4 in Turnagain Bog peat. Methane can also be derived from one carbon (C1) compounds such as methylated nitrogen and sulfur compounds and methanol, although these precursors are insignificant in most environments. We demonstrated in samples from Turnagain Bog and other similar wetlands that C1 compounds were not utilized by methanogenic bacteria under most circumstances [Hines et al., 2001; Kiene and Hines, 1995]. Therefore, most if not all CH4 originated from H2/CO2. Studies utilizing measurements of the δ13C of CH4 showed that oligotrophic wetlands can produce CH4 that is derived almost completely from H2/CO2 [Lansdown et al., 1992], and it was suggested that oligotrophic conditions favor methanogenesis by this pathway [Chasar et al., 2000; Popp et al., 1999]. In addition, evidence exists that low temperature favors the use of H2/CO2 over the use of acetate by methanogens [Nozhevnikova et al., 1994]. The fact that bogs in warmer climates do support this process suggests that climate warming may cause northern wetlands to produce more CH4 from acetate in the future. However, such a switch did not occur in our 4.5-month incubation at 24°C (Figure 5).

4.4. Iron Reduction

[33] Iron reduction was prevalent in Turnagain Bog. Ombrotrophic wetlands would not normally be expected to support active Fe reduction, since Fe inputs are usually insignificant. However, the Anchorage area occasionally receives elemental input as volcanic ash; the most recent occurring in 1992 during the eruption of Mount Spurr located 80 km west. The ash contained about 5% Fe [Swanson et al., 1995] and is the most likely source of Fe for Fe reduction in the bog. Pore water Fe concentrations were sensitive to water table height and precipitation events (Figure 3). Dissolution of Fe occurred soon after precipitation and prior to significant water table increases, suggesting that Fe(III)-reducing bacteria were active without an appreciable lag. The changing water table in the bog probably caused both the reduction and oxidation of Fe to occur, which would have enhanced Fe reduction since freshly precipitated Fe is more easily reduced by bacteria [Lovley and Phillips, 1986]. It appeared that only a fraction of the Fe present was easily reducible, since the addition of freshly prepared Fe oxide lead to an immediate increase in Fe reduction (data not shown).

[34] Despite the occurrence of Fe reduction in the bog, it was not rapid enough to drastically affect the ability of the bog to accumulate acetate (Figure 2). The ability of Fe-reducing bacteria to consume acetate has been demonstrated [Lovley et al., 1992]. However, 8 moles of Fe are required to oxidize 1 mole of acetate and the increases of Fe(II) in bog pore waters was only capable of consuming about 10% of the acetate that was produced.

5. Summary and Conclusions

[35] Acetate was the dominant organic end product of anaerobic decomposition in Turnagain Bog; surpassing CH4 production by as much as 120 fold. Although net acetate accumulation in late summer may have been a transient process, the fact that peat held anaerobically for 4.5 months did not display acetoclastic methanogenesis suggested that acetate accumulation is a season-long phenomenon in this bog. Iron reduction was prevalent in bog peat, but was only capable of utilizing a small portion of the acetate that was formed. The ultimate fate of acetate produced anaerobically was its aerobic degradation to CO2 once it diffused into either oxidizing surficial regions or near oxidizing roots of vascular plants. Methanogenesis is a minor process in the degradation of organic matter in the bog, and it appears that fermentation may be the primary terminal degradation process occurring.

Acknowledgments

[36] We thank the Ted Stevens Anchorage International Airport for the use of their land. This work was supported by NSF grants 9630053, 9909861, 0095034 (M. E. H.), and DEB 9632421 (R. P. K.). M. A. K. was supported by the NSF Research Experience for Undergraduates (R. E. U.) Program.

Ancillary