at SciVerse ScienceDirect

Soil Biology & Biochemistry 54 (2012) 36e47

Contents lists available

Soil Biology & Biochemistry

journal homepage: www.elsevier .com/locate/soi lbio

pH controls over anaerobic carbon mineralization, the efficiency of methaneproduction, and methanogenic pathways in peatlands across anombrotrophiceminerotrophic gradient

Rongzhong Ye a, Qusheng Jin b, Brendan Bohannan a, Jason K. Keller c, Steven A. McAllister a,Scott D. Bridghama,*

aUniversity of Oregon, Institute of Ecology and Evolution, 5289 University of Oregon, Eugene, OR 97403, USAbUniversity of Oregon, Department of Geological Sciences, Eugene, OR 97403, USAcChapman University, School of Earth and Environmental Sciences, Orange, CA 92866, USA

a r t i c l e i n f o

Article history:Received 14 December 2011Received in revised form15 May 2012Accepted 16 May 2012Available online 5 June 2012

Keywords:pHAnaerobic carbon mineralizationAcetate poolingEfficiency of methane productionMethane pathwaysOmbrotrophiceminerotrophic gradientPeatlands

* Corresponding author. Tel.: þ1 541 346 1466.E-mail addresses: rzye2000@yahoo.com (R. Y

(S.D. Bridgham).

0038-0717/$ e see front matter � 2012 Elsevier Ltd.doi:10.1016/j.soilbio.2012.05.015

a b s t r a c t

Methane (CH4) production varies greatly among different types of peatlands along an ombrotrophiceminerotrophic hydrogeomorphic gradient. pH is thought to be a dominant control over observeddifferences in CH4 production across sites, and previous pH manipulation experiments have verified theinhibitory effect of low pH on CH4 production. In this experiment, we asked (i) if the major effect of lowpH is direct inhibition of one or both pathways of methanogenesis and/or inhibition of ‘upstream’

fermentation that provides substrates for methanogens, and (ii) to what extent is pH sufficient to explaindifferences in CH4 production relative to other factors that co-vary across the gradient. To address thesequestions, we adjusted the pH of peat slurries from 6 peatlands to 4 levels (3.5, 4.5, 5.5, and 6.5) thatreflected their range of native pH, maintained these pH levels over a 43-day anaerobic laboratoryincubation, and measured a suite of responses within the anaerobic carbon cycle. Higher pH causeda significant increase in CO2 production in all sites. Regardless of site, time, and pH level, the reduction ofinorganic electron acceptors contributed to <12% of total CO2 production. Higher pH caused acetatepooling by Day 7, but this effect was greater in the more ombrotrophic sites and lasted throughout theincubation, whereas acetate was almost completely consumed as a substrate for acetoclastic methano-genesis by Day 43 in the minerotrophic sites. Higher pH also enhanced CH4 production and this processwas up to 436% more sensitive to changes in pH than CO2 production. However, across all sites and pHlevels, CH4 production accounted for <25% of the total gaseous C production. Fermentation appeared tobe the main pathway for anaerobic C mineralization. Our results indicate that low pH inhibits CH4

production through direct inhibition of both methanogenesis pathways and indirectly through its effectson fermentation, but the direct effects are stronger. The inability of acetoclastic methanogenesis to fullycompensate for acetate pooling in ombrotrophic peats at higher pH suggests that CH4 production isinhibited by some factor(s) in addition to pH in these sites. We examine a variety of other potentialinhibitory mechanisms and postulate that humic substances may provide an important inhibitory effectover CH4 production in ombrotrophic peatlands.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Peatlands cover less than 3% of the earth’s land surface, yet theycontain 513 Pg soil carbon (C) (Maltby and Immirzi, 1993), or about22% of the world’s soil C pool to 3m depth (Jobbágy and Jackson,

e), bridgham@uoregon.edu

All rights reserved.

2000). Wetlands are responsible for between 20 and 40% ofglobal methane (CH4) emissions, with an important but poorlydefined contribution from northern peatlands (Denman et al.,2007). The formation and maintenance of peatlands requiresa water table close to the surface (Belyea and Baird, 2006), and thusa substantial portion of the soil profile normally undergoes anaer-obic mineralization, which produces carbon dioxide (CO2) and CH4as end products. Given that CH4 has approximately 25 times theglobal warming potential of CO2 (Forster et al., 2007), it is essentialto understand the fundamental controls over the efficiency and

R. Ye et al. / Soil Biology & Biochemistry 54 (2012) 36e47 37

pathways of CH4 production during anaerobic C mineralization inpeatlands.

During anaerobic C mineralization, organic polymers areinitially hydrolyzed into monomers, followed by the fermentationof monomers to short-chain fatty acids (including acetate), simplealcohols, dihydrogen (H2), and CO2. These fermentation productsare subsequently oxidized coupled with the reduction of severalinorganic terminal electron acceptors (TEAs), such as nitrate (NO3

�),manganese (Mn(IV)), ferric iron (Fe(III)), and sulfate (SO4

2�)(Megonigal et al., 2004). Humic substances have recently beenidentified as important organic TEAs in natural environments, andmay be particularly important in peatland soils (Lovley et al., 1996;Heitmann et al., 2007; Keller et al., 2009; Lipson et al., 2010). Ingeneral, microbes will preferentially utilize these TEAs, producingCO2 as a byproduct, before CH4 production becomes important,resulting in a high ratio of CO2:CH4 produced during anaerobicrespiration. After these more favorable TEAs have been depleted,methanogenesis dominates as the terminal step of anaerobic Cmineralization. In freshwater wetlands, CH4 production mostlyoccurs via acetate disproportionation (acetoclastic methano-genesis) and CO2 reduction by H2 (hydrogenotrophic methano-genesis). Acetoclastic methanogenesis generally accounts fortwo-thirds of CH4 production resulting in an overall 1:1 ratio ofCO2:CH4 production under methanogenic conditions (Conrad,1999).

Different types of peatlands within a landscape are typicallydefined by a hydrogeomorphic gradient ranging fromprecipitation-fed (ombrotrophic) bogs to groundwater-fed (min-erotrophic) fens. The source of incoming water and associatednutrients and cations explains a range of ecosystem characteristics(Vitt et al., 1995; Bridgham et al., 1996, 1998). Ombrotrophic bogsare typically highly acidic (pH � 4). Their acidity is primarilygenerated from organic acids and secondarily from sequestration ofcations in peat, and their pH reflects the buffering characteristics oforganic acids because by definition all mineral inputs are onlyatmospheric (Clymo, 1987; Urban et al., 1987). In contrast, miner-otrophic fens receive inputs of water and minerals from ground-water and/or overland runoff, which provide substantially moremineral-derived alkalinity and higher pHs (Bridgham et al., 1999).With increasing pH, fens can be further subdivided into acidic fens,intermediate fens, and rich fens (Sjörs, 1950). In addition to soil pH,base cation concentrations, soil nitrogen and phosphorus avail-ability, vegetation community structure, and soil carbon qualityvary dramatically along with this gradient.

Given that inorganic TEAs are often at very low concentrationsin peatlands (Vile et al., 2003a; Keller and Bridgham, 2007), it mightbe expected that methanogenesis dominates anaerobic C miner-alization in these systems (resulting in a 1:1 ratio of CO2:CH4). Incontrast to this expectation, observed CO2:CH4 ratios in peats aretypically much higher and can vary by several orders of magnitudeamong different types of peatlands, suggesting distinctive path-ways and controls of anaerobic C mineralization (Bridgham et al.,1998; Segers, 1998; Vile et al., 2003b; Yavitt and Seidman-Zager,2006; Keller and Bridgham, 2007; Galand et al., 2010). Thepredominance of CO2 production in peatlands has been attributedto either humic substances acting as TEAs (Yavitt and Seidman-Zager, 2006; Keller and Bridgham, 2007) or the buildup offermentation products (Vile et al., 2003a; Galand et al., 2010).However, a number of other factors that affect anaerobic C cyclingco-vary along the ombrotrophiceminerotrophic gradient and couldinfluence the production rates of CO2 and CH4 and the efficiency ofCH4 production (i.e., the CO2:CH4 ratio), including C quality,nutrient status, and soil pH (Segers, 1998; Blodau, 2002). Further-more, methanogenic pathways and archaeal community composi-tion differ dramatically along this gradient (Horn et al., 2003;

Dettling et al., 2007; Keller and Bridgham, 2007; Kotsyurbenkoet al., 2007). Thus, the relative importance of various factorscontrolling observed variations in anaerobic C mineralization indifferent peatland types remains elusive.

Of the many potential controlling factors, pH is known to posefundamental physiological restrictions on soil microbial commu-nities and therefore impact C mineralization (Goodwin and Zeikus,1987a). Low rates of CH4 production and high CO2:CH4 ratios arecommonly observed in anaerobic incubations of ombrotrophicpeatland soils (Bridgham et al., 1998; Avery et al., 2002; Galandet al., 2005; Dettling et al., 2007; Keller and Bridgham, 2007).Manipulations of soil pH demonstrate that excess acidity contrib-utes to the low CH4 production in these peatlands (Williams andCrawford, 1984; Dunfield et al., 1993; Kotsyurbenko et al., 2004).

There are a number of potential mechanism by which pHcould regulate anaerobic carbon mineralization and CH4 produc-tion, but the importance of these mechanisms across theombrotrophiceminerotrophic gradient has not been well studied.Valentine et al. (1994) suggested that low CH4 production in moreacidic conditions was a result of substrate limitation caused by pHsuppressions of fermentation. Addition of fermentative products,such as acetate and CO2/H2, has been shown to stimulate CH4productions in a bog (Bräuer et al., 2004) and an intermediate fen(Wüst et al., 2009). However, other studies have found an accu-mulation of acetate in ombrotrophic peatlands and suggested thatlow pH causes a disconnect between acetogenesis and acetoclasticmethanogenesis (Shannon and White, 1996; Duddleston et al.,2002; Keller and Bridgham, 2007). Hydrogenotrophic methano-genesis often dominates in ombrotrophic peatlands whereasacetoclastic methanogenesis dominates in more minerotrophicconditions (Duddleston et al., 2002; Galand et al., 2005; Dettlinget al., 2007). Again, manipulative pH experiments suggest thatacidity is a primary control over the relative importance of thetwo methanogenesis pathways (Kotsyurbenko et al., 2007). Yavittand Seidman-Zager (2006) postulated that soil pH was themain factor that determines the dominance of acetoclastic versushydrogenotrophic methanogenesis in peatlands.

Thus, while it is well established that low pH has a negativeeffect on CH4 production, its effects co-varywithmany other factorsalong the ombrotrophiceminerotrophic gradient in peatlands andthe specific mechanisms for the pH inhibition are unclear.Accordingly, we asked the following questions in this study: Doombrotrophic peatlands have low CH4 production rates solelybecause of their low pHs or are other factors also important? Is theinhibitory effect of low pH due to its direct effect on methanogensor due to an “upstream” effect on fermentation and productionof H2 and acetate? Does low pH inhibit one methanogenesispathwaymore than another? In the present study, wemanipulatedsoil pH and analyzed anaerobic C mineralization in six peatlandswith different native pHs (ranging from 3.7 to 6.0) along anombrotrophiceminerotrophic gradient. To our knowledge, no otherstudy has performed amanipulative pH experiment in such a broadrange of peatlands. Our objectives were to determine to whatextent that differences in soil pH among these peatlands control (i)acetate accumulation, (ii) CH4 and CO2 production rates, (iii) theefficiency of CH4 production vs. CO2 production, and (iv) thedominant methanogenic pathway.

2. Materials and methods

2.1. Site description

We sampled six peatlands in the Upper Peninsula of Michigan,USA that spanned the ombrotrophiceminerotrophic gradient basedupon dominant vegetation and soil pH. Five of these peatlands are

R. Ye et al. / Soil Biology & Biochemistry 54 (2012) 36e4738

located on the property of the University of Notre Dame Environ-mental Research Center and the sixth site is approximately 100 kmeast in Crystal Falls, Michigan, USA. Based upon our extensiveprevious research in the peatlands of this region and a survey ofmore than 20 potential peatlands in 2008 (unpublished data), siteswere chosen to be representative of the region and to incorporatedifferences in CH4 production and CO2:CH4 ratios.

Bog 1 (N46�60600, W88�1602500) is an ombrotrophic bog with welldeveloped hummocks and minimal hollows and lawn area. Vege-tation is dominated by >90% cover of Sphagnum spp. mosses withstunted (<1-m height) ericaceous shrubs such as leatherleaf(Chamaedaphne calyculata (L.) Moench), small cranberry (Vacciniumoxycoccos L.), and bog Labrador tea (Rhododendron groenlandicumOeder), and scattered low-stature black spruce (Picea mariana(Mill.) Britton, Sterns & Poggen). Peat depth isw380 cm, and the pHof the surface peat was 3.7 at the time of sampling. The water tableaveraged �27 cm during the growing season (all water-tabledepths are reported from hollow surfaces). Bog 2 (N46�1305700,W89�340700) has similar vegetation to bog 1 except that it has muchless hummock development. It has a peat depth of w490 cm, anda pH of 4.1. It had an average water-table depth of �16 cm. Theacidic fen (N46�1204800, W89�300200) has a Sphagnum spp. lawnwithlittle cover from other species in the area of sampling. Peat depth isw500 cm, pH was 4.1, and average water-table depth was �10 cm.The intermediate fen (N46�1501700, W89�3202000) also has >90%cover of Sphagnum spp., the same ericaceous shrubs as the bogsites, the deciduous shrubs speckled alder (Alnus incana (L.)Moench, ssp. rugosa (Du Roi) Clausen) and purple chokeberry(Photinia floribunda (Lindl.) Robertson & Phipps), and a variety ofgraminoid and forb species. Peat depth isw340 cm, soil pHwas 4.5,and the average water-table depth was �17 cm. The rich fen site(N46�1302700, W89�2905300) is dominated by upright sedge (Carexstricta Lam.) tussocks with leatherleaf also present on the tussocks.It has w6.4 m of peat, a pH of 5.9, and consistently about 30 cm ofstanding water. The cedar swamp site (N46�1404100, W89�3203800)has an overstory of northern white cedar (Thuja occidentalis L.) andspeckled alder and a mixed understory of mosses, forbs, andunvegetated ground. Peat depth is w290 cm, pH was 6.0, and thewater-table depth averaged �6 cm. Three of the sites (bog 2, theacidic fen, and the rich fen) bordered lakes, reflecting a commonsituation of peat in-filling of lakes in this region, but observationsand differences in chemistry between the lakes and porewatersuggest no significant exchange of surface or groundwater betweenthe lakes and bog 2 and the acidic fen. In contrast, the rich fenreceives continuous surface water flow from the adjacent lake,giving it a minerotrophic status.

2.2. Sample preparation

Four soil cores were randomly collected in hollows from eachsite with PVC tubes (10-cm diameter) to a depth of 15 cm below thewater table in August 2009. The samples were taken at a relativelydry period, so the water-table level was�34 cm at bog 1,�23 cm atbog 2, �15 cm at the acidic fen, �23 cm at the intermediatefen, þ35 cm at the rich fen, and �16 at the cedar swamp. Uponextraction, cores were intermediately capped after filling the coreswith porewater to prevent oxidation of the soil. Soil cores andwater samples were transported on ice to our laboratory at theUniversity of Oregon and frozen at �20 �C until use.

In the laboratory, peat cores were processed in a glove box filledwith 98% N2 and 2% H2 (Coy Laboratory Products Inc., Grass Lake,Michigan, USA). Each core was homogenized following the removalof large roots, woody material, and green vegetation. Moisturecontent was measured on a subsample that was dried at 60 �C for 3days. Rubbed fiber content and the von Post indexwere determined

on a subsample according to Parent and Caron (2006) and Clymo(1983), respectively, to estimate the degree of decomposition ofthe peats. Lower rubbed fiber content generally indicates a greaterdegree of decomposition and hence a lower carbon quality,although the plant material from which the peat is derived is alsoimportant (Bridgham et al., 2001). In contrast, a low humificationvalue of the von Post index suggests a lower degree of decompo-sition. Four additional subsamples were prepared from each corefor the pH manipulation experiment, with individual cores servingas the replicate unit. Approximately 120 g of field-moist peat wastransferred to a 440mLMason jar, slurriedwith 240mL of degassedand deionizedwater, and adjusted to a pH of 3.5, 4.5, 5.5, or 6.5 witheither 10 N HCl or 10 N NaOH. We adjusted the pH daily on the first7 days of incubation and once every 2 or 3 days thereafter in theglove box since pH changes in most of the peat slurries were lessthan 0.2 pH units after 7 days. After pH adjustment, peat slurrieswere capped and bubbled vigorously with oxygen- free N2 gas for10e15 min, followed by incubation at 17 �C in the dark, which wasthe average field temperature when the soil cores were extracted.

2.3. Analyses of porewater chemistry

On the 2nd, 7th, 15th, and 43rd day of incubation, 20 mL ofwater was collected from each slurry sample in the glove box, fol-lowed by additions of equal amounts of degassed and deionizedwater. The dilution effect of these additions was corrected for whencalculating the concentrations of dissolved constituents at differentsampling events. Water samples were centrifuged at 5000 rpm for5 min and filtered with Whatman GF/F glass-fiber filter paper,followed by quantification of reduced iron (Fe (II)) as described byGibbs (1976). An aliquot of the water sample was frozen at �20 �Cfor additional chemical analyses. Concentrations of SO4

2�, NO3�, and

nitrite (NO2�) were determined with a Dionex DX500 ion chroma-

tography system equipped with an Ionpac AS11 column, an AESsuppressor, and an ED50 electrochemical detector (Dionex Corpo-ration, Bannockburn, Illinois, USA). The detection limits for SO4

2�,NO3

�, and NO2� were 4, 16, and 4 mM, respectively. Acetate concen-

trations were measured with a Dionex DX500 ion chromatographysystem equipped with a HC-75 (Hþ) column (Hamilton CompanyUSA, Reno, Nevada, USA) and a Dionex AD20 absorbance detector,with a detection limit of 11 mM.

2.4. Analyses of methanogenic pathways and CO2, CH4, and acetateproduction

Immediately after collection of water samples and pH adjust-ment, approximately 10 g of slurried peat was removed from eachMason jar and transferred into a 40 mL I-CHEM vial (Thermo FisherScientific Inc., Rockwood, Tennessee, USA) and subsequentlybubbled vigorously with oxygen-free N2 gas for 5e10 min. Toestimate potential rates of hydrogenotrophic methanogenesis,0.1 mL of 3.5 mCi mL�1 NaH14CO3 was added to each of the slurries,followed by gently shaking and incubation at 17 �C in the dark. Afterincubation for 2 days, slurries were shaken to release trapped gasbubbles. Headspace gases were then analyzed for CO2 and CH4 bygas chromatography using a flame ionization detector equippedwith a methanizer (SRI Instruments, Torrance, California, USA).14CO2 and 14CH4 were determined simultaneously with an in-lineradioactive gas detector (LabLogic Systems Inc., Brandon, Florida,USA). Total CO2, CH4, 14CO2, and 14CH4 production were calculatedfrom both gas and liquid phases, adjusting for solubility, tempera-ture, and pH (Stumm and Morgan, 1995). Rates of respiration andmethanogenesis were calculated from total CO2 and CH4 produc-tion during the 2-day incubation and expressed per gram drymass peat. Methane production efficiency was defined as CH4

Table 1Selected properties of native peat samples across an ombrotrophiceminerotrophic gradient (mean � 1 standard error, n � 4).

Site pH von Post index Rubbed fiber volume content (%) Fe(II) (mM) SO42� (mM) Acetate (mM)

Bog 1 3.7 � 0.0 H3 38 � 10 23.3 � 9.3 18.7 � 6.8 BDBog 2 4.1 � 0.0 H3 38 � 10 3.9 � 1.9 14.9 � 1.8 20.4 � 7.1Acidic fen 4.1 � 0.0 H3/H4 35 � 9 1.5 � 0.5 20.8 � 9.6 61.4 � 29.7Intermediate fen 4.5 � 0.0 H4/H5 13 � 4 0.9 � 0.4 25.3 � 8.3 60.3 � 9.0Rich fen 5.9 � 0.0 H5 15 � 5 1.2 � 0.5 6.0 � 3.6 94.9 � 18.7Cedar swamp 6.0 � 0.1 H5 11 � 3 0.36 � 0.2 12.2 � 6.0 75.6 � 21.2

Note: BD, below detection limit; both NO3� and NO2

� were not detected.

R. Ye et al. / Soil Biology & Biochemistry 54 (2012) 36e47 39

production/(CO2 production þ CH4 production) and expressed asa percentage.

Rates of hydrogenotrophic methanogenesis (CH4,hyd) werecalculated as described by Keller and Bridgham (2007) with slightmodifications: CH4,hyd ¼ a[SCO2]a/Atg, where a is the recoveredactivity of 14CH4, SCO2 is the available CO2 pool in the liquid andgaseous phases, a (¼1.12) is the 14C:12C isotope fractionation factorfor hydrogenotrophic methanogenesis, A is the activity of availableH14CO3

� in the liquid and gaseous phases, t is the incubation time,and g is the dry mass of peat in the slurry. The size of the CO2 pool inthe peat slurry was minimal initially because of the flushing of theheadspace and reached its maximum at the end of the incubation asa result of anaerobic respiration, so using SCO2 at the end of theincubation (as done in Keller and Bridgham, 2007) overestimatesCH4,hyd. We therefore assumed that the increase of CO2 productionduring the 2-day incubationwas linear and used themid-point valueof SCO2 to estimate CH4,hyd. Similarly, we assumed that theconsumption of H14CO3

� was linear, and A was calculated from theactivity of H14CO3

� at the mid-point of the incubation. We haveextensively compared both analytical approaches in a number ofexperiments and found the “traditional” approach used by Keller andBridgham (2007) and others (e.g., Avery et al.,1999) can substantiallyoverestimate hydrogenotrophic methanogenesis, even to the extentthat it is sometimes greater than total methanogenesis. In thecurrent experiment, hydrogenotrophic methanogenesis rates werealways lower using the current method, but generally acetoclasticmethanogenesis dominated with both analytical approaches (seebelow) and provided similar statistical results.

Rates of acetoclastic methanogenesis were calculated by sub-tracting hydrogenotrophic production from total CH4 production.In preliminary experiments, we have found that acetoclasticmethanogenesis rates are underestimated with 14C-acetate, asreported by Avery et al. (1999). Net acetate production/consump-tion was calculated as the increase/decrease in acetate concentra-tions in each slurry between two consecutive time points.

2.5. Statistical analyses

Results were analyzed using restricted maximum likelihood inthe MIXED procedure of SAS 9.1 (SAS Institute) (Littell et al., 2006).

Table 2Effect of pH, peat type, and incubation time on selected variables of peat samples.

Source DF F-values

Fe(II) Acetate CO2 CH4

pH 3 529.8** 13.8** 99.5** 134.8*Peat 5 27.5** 77.4** 60.5** 96.5*Time 3 21.6** 182.4** 117.0** 26.4*pH*Peat 15 14.9** 4.0** 2.1** 2.9*pH*Time 9 4.9** 4.4** 1.9 3.1*Peat*Time 15 2.8** 32.7** 1.7* 4.1*pH*Peat*Time 45 1.1 3.1** 1.2 0.7

Note: *, p < 0.05; **, p < 0.01.

The fixed effects were pH, peat type (site), time, and their inter-actions with time as a repeated variable. Significant differencesbetween individual pH levels, sites, and time intervals wereanalyzed with Tukey’s test at a ¼ 0.05. All data were tested fornormality and log-transformed if the transform resulted in signif-icant improvements in overall distribution.

3. Results

3.1. Physicochemical analyses

Sites had a wide range of in situ pH (3.7e6.0) that represents therange typically found along the ombrotrophiceminerotrophicgradient in this region (Table 1). Initial rubbed fiber volume wasmuch higher in the ombrotrophic peats than in the minerotrophicpeats (Table 1), indicating the more decomposed nature of thelatter (i.e. less availability of labile C) and the woody origin of thesapric peat from the cedar swamp. The rubbed fiber values wererelatively low, especially in the ombrotrophic sites, because of thelow water-table at the time that the samples were taken and,hence, the great depth at which the samples were taken. The vonPost humification score and porewater acetate concentrationsgenerally increased along the ombrotrophiceminerotrophicgradient (Table 1). Nitrite and NO3

� were not detected in any ofthe water samples during the course of experiment. Initial SO4

2�

concentrations were �25 mM but were below detection in allsamples on the 2nd day of incubation and thereafter (Table 1).

Fe(II) concentrations increased as pH decreased, with a largeincrease in concentration during the first week at pH 3.5 and 4.5(Table 2, Fig. 1). At pH 3.5 and 4.5, samples from the more miner-otrophic sites (intermediate fen, rich fen, and cedar swamp)consistently had higher Fe(II) concentrations than those from othersites (Fig. 1).

Acetate concentrations were also significantly influenced by pHadjustment, with complicated interactions among pH, site, andincubation time (Table 2). On Day 2, pH had little effect on acetateconcentrations in any of the samples (Fig. 2a). Acetate concentra-tions increased dramatically between Day 2 and Day 7 in mostsamples within each pH level, especially those from the moreombrotrophic sites (Fig. 2a, b). Acetate concentrations were

CH4/(CO2 þ CH4) Acetoclastic CH4 Hydrogenotrophic CH4

* 41.1** 96.8** 91.4*** 100.9** 66.5** 62.4*** 112.8** 31.4** 7.32*** 2.4** 3.5** 1.85** 1.5 1.7 3.2*** 3.8** 2.6** 4.9**

0.8 0.8 1.7

Fig. 2. Acetate concentrations in peat samples across an ombrotrophiceminerotrophicgradient (from left to right) adjusted to various pH levels. Bars indicate mean � 1standard error. a, Day 2; b, Day 7, c, Day 15; d, Day 43.Fig. 1. Fe(II) concentrations in peat samples across an ombrotrophiceminerotrophic

gradient (from left to right) adjusted to various pH levels. Bars indicate mean � 1standard error. a, Day 2; b, Day 7, c, Day 15; d, Day 43.

R. Ye et al. / Soil Biology & Biochemistry 54 (2012) 36e4740

typically greater at higher pH after Day 2 in peat from the threemore ombrotrophic sites (bog 1, bog 2, and acidic fen) and inintermediate fen peat on Days 7 and 15, but pH had little effect onacetate concentrations in peat from the two most minerotrophicsites (rich fen and cedar swamp), which remained low throughoutthe experiment. Acetate concentrations also decreased graduallyfrom the Day 7 to Day 43 at pHs > 4.5 in rich fen and cedar swamppeat and from Day 15 to Day 43 in intermediate fen peat(Fig. 2bed).

3.2. CO2 production

There was a positive effect of increasing pH on CO2 productionrates across the peatland gradient, although that effect varied

somewhat by peat type (Table 2). By Day 2 an increase in pH from3.5 to 6.5 increased the rates of CO2 production by 607% in bog 1,332% in bog 2, 175% in the rich fen, and 208% in the cedar swamp(Fig. 3a). These pH effects were maintained throughout the exper-iment, although increases in peat from individual sites were notalways significant (Fig. 3). In general, CO2 production ratesdecreased from Day 2 to Day 7 and remained relatively stablethereafter.

3.3. CH4 production

Methane production rate also increased as pH increased, butthis effect varied greatly with site and time (Table 2, Fig. 4). On Day2, the CH4 production rate was up to 12 times smaller at a pH of 3.5compared to higher pHs in peat from the intermediate fen and richfen, with no pH effect in peat from the other sites (Fig. 4a). On Days

Fig. 4. Rates of total methanogenesis in response to pH adjustment in peat samplesacross an ombrotrophiceminerotrophic gradient (from left to right). Bars indicatemean � 1 standard error. a, Day 2; b, Day 7, c, Day 15; d, Day 43.

Fig. 3. Rates of anaerobic CO2 production in response to pH adjustment in peatsamples across an ombrotrophiceminerotrophic gradient (from left to right). Barsindicate mean � 1 standard error. a, Day 2; b, Day 7, c, Day 15; d, Day 43.

R. Ye et al. / Soil Biology & Biochemistry 54 (2012) 36e47 41

7, 15, and 43, lower pH typically had a strong inhibitory effect onCH4 production rates in all sites, with CH4 production rates oftenseveral orders of magnitude greater at a pH of 6.5 compared to a pHof 3.5 (Fig. 4bed). At pH> 3.5, the rates of CH4 production generallyincreased fromDay 2 to Day 15 (Fig. 4a, b). One of the bog sites (bog2) was an outlier with very small CH4 production rate overall andno significant pH effect.

Methane production efficiency (CH4/(CO2 þ CH4)) was signifi-cantly affected by pH, which was dependent on site but not time ofincubation (Table 2). Averaged across time, increasing pH from 3.5to 6.5 enhanced CH4 production efficiency by 159%e334% (Fig. 5).Bog 2 was again an outlier with very low CH4 production efficiencyand no effect of pH.Within each pH level, the rich fen generally hadthe highest CH4 production efficiency, followed by the intermediatefen, cedar swamp, acidic fen, bog 1, and bog 2 (Fig. 5). Interestingly,CH4 production efficiency was consistently highest at pH 5.5

regardless of the site, although it was not always statisticallydifferent from pH 4.5 and 6.5 (Fig. 5).

On Day 2, rates of acetoclastic CH4 production were 3e12times smaller at pH 3.5 than at higher pHs in the three mostminerotrophic peats (intermediate fen, rich fen, and cedarswamp), but there was no effect in peat from the other sites(Fig. 6a). Acetoclastic methanogenesis was inhibited by low pH inpeat from all sites on the last three sampling dates, except that peatfrom bog 2 had low overall rates and no significant pH effect untilDay 43. On Day 15, the rates of acetoclastic methanogenesis in peatfrom the intermediate fen and rich fen were much higher than thethree most ombrotrophic peats (bog 1, bog 2, and acidic fen), whichwas also observed on Day 43 (Fig. 6c, d). Rates of acetoclasticmethanogenesis in most samples increased gradually fromDay 2 toDay 15, especially in peat from the intermediate fen, rich fen, andcedar swamp at pH > 4.5 (Fig. 6aec). However, the rates decreasedafter Day 15 except in peat from the intermediate fen.

Low pH generally inhibited hydrogenotrophic methanogenesisin peat from all sites throughout the incubation (Fig. 7). The pHeffect was not significant in acidic fen peat on Day 2 and in peatfrom bog 2 and the intermediate fen on Day 7 (Fig. 7a, b). Therewere large changes in rates through time but this varied by site.

Fig. 6. Rates of acetoclastic methanogenesis in response to pH adjustment in peatsamples across an ombrotrophiceminerotrophic gradient (from left to right). Barsindicate mean � 1 standard error. a, Day 2; b, Day 7, c, Day 15; d, Day 43.

Fig. 5. pH effects on the efficiency of CH4 production relative to total gaseous C(CO2 þ CH4) production in peat samples across an ombrotrophiceminerotrophicgradient (from left to right). Bars indicate mean � 1 standard error.

R. Ye et al. / Soil Biology & Biochemistry 54 (2012) 36e4742

The contribution of acetoclastic methanogenesis to total CH4production in peat from bog 1 generally decreased as pH increasedfrom 3.5 to 6.5, though the difference was not always significant(Fig. 8). Similar pH effects were also observed in bog 2 on Day 43and in the acidic fen on Day 15, but not in the three most miner-otrophic sites (intermediate fen, rich fen, and cedar swamp) (Fig. 8).However, regardless of pH, acetoclastic methanogenesis dominatedthe total CH4 production (>54%) in all the peats, especially after Day2 (>66%).

4. Discussion

In the present study, we addressed the questions of whetherthe differences in pH in peatlands across an ombrotrophiceminerotrophic gradient are sufficient to explain the large varia-tion in the efficiency of CH4 production and what are the under-lying mechanisms of this pH effect in the context of the entireanaerobic C cycle. As with any manipulative biogeochemical labo-ratory study, extrapolation of rates to in situ conditions should bedone with extreme caution, and even more so in an experimentwhere something as fundamental as pH is changed with its myriadeffects on abiotic and biotic processes. However, like many othermanipulative studies, our experiment provides insights to under-standing mechanisms that cannot be achieved with a comparativefield approach. Overall our results indicate that, as we expected,low pH is an important control over anaerobic C mineralization andCH4 production in peatlands, but other factors are of at least equalimportance in explaining low rates of CH4 production and low CH4production efficiency in ombrotrophic peatlands. Based upon thecurrent experiment and ancillary experiments, we examine otherfactors likely to inhibit CH4 production.

4.1. CO2 production

Our results suggest strong pH limitation on anaerobic Cmineralization as increasing pH from 3.5 to 6.5 greatly stimulatedrates of CO2 production in all sites across the peatland gradient,but particularly in bog 1 and bog 2 on Day 2 (Fig. 3). The two bogshad the lowest native pH along with high rubbed fiber contentsand low humification scores (i.e. high C quality) (Table 1). Whenpotential pH limitation in these bog soils was removed, more Csubstrates weremade available tomicrobial decomposers through

increased fermentation, as evidenced by greater CO2 and acetateproduction.

We did not quantify denitrification or sulfate reduction rates,but we assume that they had minimal contribution to anaerobic Cmineralization given that neither NO3

� nor NO2� were detected in

our samples and SO42� concentrations were below the detection

limit after 24 h. Low rates of these processes are common in manynorthern peatlands (Blodau et al., 2002; Duddleston et al., 2002;Vile et al., 2003a), although high rates of sulfate reduction havebeen observed in some peatlands during short-term incubations(Wieder et al., 1990; Vile et al., 2003b).

Iron reduction is also generally assumed to be unimportant innorthern peatlands due to low availability of Fe(III) (Blodauet al., 2002). Although we did not measure the concentrations ofFe(III) directly in the present study, our results clearly indicated

Fig. 8. pH effects on the percentage of acetoclastic CH4 production relative to total CH4

production in peat samples across an ombrotrophiceminerotrophic gradient (from leftto right). Bars indicate mean � 1 standard error. a, Day 2; b, Day 7, c, Day 15; d, Day 43.

Fig. 7. Rates of hydrogenotrophic methanogenesis in response to pH adjustment inpeat samples across an ombrotrophiceminerotrophic gradient (from left to right). Barsindicate mean � 1 standard error. a, Day 2; b, Day 7, c, Day 15; d, Day 43.

R. Ye et al. / Soil Biology & Biochemistry 54 (2012) 36e47 43

substantial accumulation of Fe(II) at pH � 4.5, especially inmore minerotrophic peats (Fig. 1). More minerotrophic peatlandsgenerally have much higher peat mineral content (Bridgham et al.,2001; Gorham and Janssens, 2005), but more ombrotrophic peat-lands have higher porewater iron concentrations because iron ismore soluble at low pH (Vitt et al., 1995). We demonstrated theinteraction of these factors in this study, with maximal porewaterFe2þ concentrations in minerotrophic peats at low pH. Using theincrease in Fe(II) concentration over time and the ratio of 4 molFe(III) reduced to 1 mol CO2 mineralized to estimate the carbonequivalents of Fe reduction (Megonigal et al., 2004), our resultsdemonstrate that Fe reduction contributed <12% of total CO2

production at pH 3.5 and<3% at pH� 4.5. In a separate experimentwe measured Fe reduction rates by the difference in oxalate-extractable Fe(II) under in situ conditions using peat from thesame sites and sampling event as the current experiment. Wefound that Fe reduction accounted for <8% of anaerobic CO2production and was much lower in most samples (data not shown).Since Mn(IV) concentrations in peat soils are much less than Fe(III)

(Gorham and Janssens, 2005), we assumed that Mn reduction wasnegligible in the present study.

Acetoclastic methanogenesis produces CO2, whereas hydro-genotrophic methanogenesis consumes CO2. In the current exper-iment there was a net production of CO2 by these two processesbecause acetoclastic methanogenesis was the dominant pathway(see below). We used the ratio of 1 mol acetate oxidized to 1 molCO2 generated during acetoclastic methanogenesis and the ratio of1 mol CO2 reduced to 1 mol of CH4 produced during hydro-genotrophic methanogenesis to calculate the CO2 balance duringthe CH4 production (Conrad, 1999). The net production of CO2during methanogenesis explained only 1e25% of the CO2production.

In sum,methanogenesis and the reduction of inorganic TEAs canonly explain a very small proportion of CO2 production, especiallyat higher pH where Fe reduction was unimportant and respirationrates were highest. Large amounts of CO2 production that cannot be

R. Ye et al. / Soil Biology & Biochemistry 54 (2012) 36e4744

explained by the consumption of inorganic TEAs has been repeat-edly demonstrated in peatlands (Duddleston et al., 2002; Vile et al.,2003a and b; Yavitt and Seidman-Zager, 2006; Keller andBridgham, 2007). Galand et al. (2010) found that anaerobic CO2production in a minerotrophic fen occurred primarily via aceto-genesis, whereas it was primarily through an unknown pathway inan oligotrophic fen and a bog. Keller and Bridgham (2007) exam-ined anaerobic C pathways seasonally in a bog, intermediate fen,and rich fen in northern Michigan, where 29e85% of the CO2production was through an unknown pathway, with a somewhathigher percentage in the rich fen. In the current study, we foundthat an unexplained pathway contributed 60e99% of CO2 produc-tion, with lower percentages in the three most minerotrophic sites(intermediate fen, rich fen, and cedar swamp) at pH � 4.5.

Some have suggested that the unexplained CO2 production inpeatlands is due to humic substances acting as organic electronacceptors (Segers, 1998; Yavitt and Seidman-Zager, 2006;Heitmann et al., 2007; Keller and Bridgham, 2007). Humicsubstances may act indirectly as electron shuttles for Fe and sulfatereduction or act directly as electron acceptors (Heitmann et al.,2007), but since Fe and sulfate reduction rates were likelyminimal in this experiment, the direct effect as an electronacceptor is most pertinent for this experiment. Reduction of humicsubstances is more thermodynamically favorable than methano-genesis (Cervantes et al., 2000) and both concentrations of dis-solved and solid humic substances are typically very high inpeatlands (Thurman, 1985). Studies have also demonstrated clearinhibitory effects of humic substances on methanogenesis eitherby substrate competition or by direct toxicity to methanogens(Cervantes et al., 2000; Keller et al., 2009; Minderlein and Blodau,2010). Humic substances are important electron acceptors andinhibit methanogenesis in peats from our study sites (unpublisheddata). Nonetheless, the peats utilized in the current experimentwere collected and processed under anaerobic conditions and thehumic substances may have already been completely reduced. It istherefore doubtful that the effect of humic substances is substantialenough to explain the remaining CO2 production and low CH4production observed in the current experiment.

Many fermentative processes produce CO2 (Reddy and DeLaune,2008), and it has been suggested that the unexplained CO2production may be explained by the buildup of fermentative lowmolecular weight alcohols and acids (Duddleston et al., 2002; Vileet al., 2003a and b; Galand et al., 2010). In the present study, weobserved significant acetate pooling along with increasing CO2production (Figs. 2 and 3). It is likely that fermentation is the ulti-mate explanation for the unexplained CO2 production in thecurrent experiment.

The consistently higher CO2 and acetate production with anincrease in pH in all sites (Figs. 2 and 3) indicates a fundamental pHlimitation of fermenters and rates of fermentation (Goodwin andZeikus, 1987b; Valentine et al., 1994; Bergman et al., 1999).Increased pH also greatly increases the solubility of organic matterin wetland soils (Clark et al., 2005; Grybos et al., 2009), which mayincrease the availability of labile C to microbes and hence enhanceCO2 production.

4.2. Acetate dynamics

Porewater acetate concentrations vary greatly over time inpeatlands (Shannon and White, 1996; Keller and Bridgham, 2007),reflecting the temporal dynamics of production and consumptionmechanisms. These concentrations were initially low in all samples,and particularly so in the bog peats (Table 1). This is potentiallybecause the samples were collected in mid-summer when thewater table was relatively far from the surface, providing an

extended aerobic zone that would promote the aerobic degradationof acetate, while also reducing the anaerobic zone for acetateproduction and accumulation.

It is generally thought that under anaerobic condition acetate ismostly derived from incomplete oxidation of organic C in hetero-trophic respiration with TEAs and fermentation reactions (Reddyand DeLaune, 2008). As discussed previously, heterotrophic respi-ration with TEAs likely played a minor role in C mineralization inthe present study and hence was not a significant source of acetateproduction. Acetate can also be produced through the process ofhomoacetogenesis (i.e., acetate formation from CO2 and H2), butthermodynamics suggests that hydrogenotrophic methanogensshould outcompete homoacetogens for their common substrate, H2(Zinder and Anguish, 1992), except possibly at low temperaturewhen H2 is sufficient (Hoehler et al., 1999; Kotsyurbenko et al.,2001). In the present study, we did not measure H2 concentra-tions, but we did determine rates of homoacetogenesis in samplesnear their native pHs by quantifying the incorporation of 14CO3

� intoacetate (unpublished data). In contrast to accepted theory, homo-acetogenesis was an important (if not dominant) pathway ofacetate production in these peats at least their native pHs. Thus,fermentation reactions were likely the dominant source of acetatein our experiment.

Acetate rarely accumulates in most anaerobic environmentsbecause of its high turnover rates (King et al., 1983; Kotsyurbenkoet al., 2004). However, accumulation of acetate has been docu-mented repeatedly in acidic bogs (Shannon andWhite, 1996; Averyet al., 1999; Duddleston et al., 2002; Keller and Bridgham, 2007;Hines et al., 2008). Much greater acetate pooling is often observedin ombrotrophic peatlands than in minerotrophic peatlands (Kellerand Bridgham, 2007; Hines et al., 2008), although Galand et al.(2010) observed the opposite trend. These studies have ascribedthis phenomenon to a variety of factors that co-vary along theombrotrophiceminerotrophic gradient, including plant commu-nity composition, microbial community composition, and the lownutrient and low pH characteristics of ombrotrophic peats.

Acetate accumulated to very high concentrations throughoutthe incubation in the two bog peats and the acidic fen peat, espe-cially at higher pHs (Fig. 2), indicating much greater acetateproduction relative to consumption by acetoclastic methanogenesisand other potential pathways, such as TEA reduction. The higherrates of acetate production in the more ombrotrophic sites athigher pH likely resulted from the less decomposed nature of theombrotrophic peats (i.e., more labile C) and hence greater rates offermentation when the native pH limitation of these peats wasameliorated. In the intermediate fen peat, there was similar acetateaccumulation through Day 15, but there was substantial acetateconsumption by Day 43 at pHs 5.5 and 6.5, likely because of highrates of acetoclastic methanogenesis at these pHs during thisperiod (Fig. 6). The rich fen and cedar swamp peats had a muchlower degree of acetate pooling, and almost all of the acetate wasconsumed at pH > 3.5 by the end of the incubation. These miner-otrophic peat types likely had lower rates of acetate productionthan the more ombrotrophic peats because their rates of aceto-clastic methanogenesis were not substantially greater, althoughother pathways of acetate consumption cannot be completely dis-counted. Low rates of acetate production in the rich fen and cedarswamp samples are likely attributable to their more decomposednature of the peats (i.e., less labile C for fermentation) inferred fromtheir higher humification scores and lower rubbed fiber contents(Table 1).

We demonstrate that the acetate pooling sometimes observedin ombrotrophic peats is not due solely to their naturally low pH,because we found even greater acetate accumulation in these sitesat higher pH. We surmise that acetate pooling in these sites is due

R. Ye et al. / Soil Biology & Biochemistry 54 (2012) 36e47 45

to their low rates of acetoclastic methanogenesis, reflecting theirlow native pH and other potential inhibitory factors, includinghumic-like substance derived from Sphagnum spp. mosses, asdiscussed below.

4.3. Methanogenesis

Methanogens exist in peat soils across a wide range of pH(Dunfield et al., 1993; Horn et al., 2003; Kotsyurbenko et al., 2004,2007). Dunfield et al. (1993) found that optimal pH values formethanogenesis were about 2 pH units higher than the native pH inacidic peats and only 0e1 unit higher in near neutral peats. Our pHmanipulation experiment revealed that raising a native pH of 3.7(bog 1) by 2e3 pH units resulted in up to 697% increase in CH4production, while reducing a native pH of 6.0 (cedar swamp) by2e3 pH units suppressed CH4 production potentials by up to 2105%(Fig. 4), indicating strong pH limitation on methanogenesis acrossthe ombrotrophiceminerotrophic gradient, which agrees withother manipulative studies (Williams and Crawford, 1984; Dunfieldet al., 1993; Valentine et al., 1994; Bräuer et al., 2004). Studies ofpure cultures of methanogens also demonstrate pH effects (Sizovaet al., 2003; Bräuer et al., 2006).

Our results demonstrate that increasing pH increased the effi-ciency of CH4 production across the gradient, with a maximumpercentage consistently at a pH of 5.5 (Fig. 5). Increasing pHincreased fermentative activity and generally the availability ofacetate (and presumably H2), and this increase in methanogenicsubstrates is likely partially responsible for the increased efficiencyof CH4 production. Regardless of the pH, CH4 production contrib-uted <25% of total gaseous C production (CO2 þ CH4) (Fig. 5), incontrast to the optimal ratio of efficiency of 50% in a purelymethanogenic system (Zinder, 1993). A lower CH4 efficiency inombrotrophic versus minerotrophic peats at their native pHshas been repeatedly demonstrated in previous studies (e.g.,Bridgham et al., 1998; Yavitt and Seidman-Zager, 2006; Keller andBridgham, 2007). Even after experimentally increasing the pH inombrotrophic peats, CH4 efficiency was generally much lower thanin more minerotrophic peats with the same pH despite relativelyminor differences in CO2 production and increased acetate accu-mulation. This suggests that while pH is important, differences inCH4 production across the ombrotrophiceminerotrophic gradientare controlled by other factors in addition to soil pH.

Peat derived from different plant sources has very different Cquality and CH4 production rates (Yavitt et al., 1987; Bridgham et al.,1998), but the pooling of acetate and the dominance of acetoclasticmethanogenesis in the ombrotrophic sites strongly suggest that thelow CH4 efficiency in these sites cannot be explained predomi-nantly by C quality. However, the role of acetate as a substratefor acetoclastic methanogenesis is complicated by its role as anorganic acid at low pH (pK ¼ 4.7) and its consequent inhibitoryeffect on acetoclastic methanogens (Russell, 1991; Bridgham andRichardson, 1992). Bräuer et al. (2004) suggested that acetate canbe utilized by acetoclastic methanogens when concentrations areless than 5e10 mM. In the present study, none of the peat sampleshad acetate concentrations >2 mM, suggesting acetate protonationwas unlikely to cause significant inhibitory impacts.

A number of additional physiochemical variables vary alongthe ombrotrophiceminerotrophic gradient and could play a rolein regulating CH4 dynamics in peatlands. For example, methano-genesis may be limited in ombrotrophic peats by trace metallimitation (Basiliko and Yavitt, 2001), but we have found no suchlimitation in peat from our sites (unpublished data). Nutrientavailability also differs across this gradient, but previous worksuggests that nutrient limitation does not fully explain differences inanaerobic carbon cycling in peatlands (e.g., Keller et al., 2005, 2006).

An alternative explanation for low CH4 production potential inmore ombrotrophic peats is that Sphagnum mosses have highconcentrations of humic-like substances (Williams et al., 1998) thatcan be quite inhibitory to microbes (Aerts et al., 1999; Minderleinand Blodau, 2010). It is possible that methanogens are particu-larly sensitive to these compounds. Hines et al. (2008) found lowCH4 efficiency in peatlands dominated by Sphagnum spp. mossesand that CH4 efficiency increased with an increasing proportion ofvascular plant cover. Our results are consistent with the hypothesisthat increasing Sphagnum spp. cover decreases CH4 efficiency: plantcover was dominated by Sphagnum spp. in the two bogs and acidicfen, was somewhat less dominant in the intermediate fen, wasabsent in the rich fen because of the consistently high water table,and was a subdominant ground layer in the cedar swamp. Ina subsequent experiment we added the humic-substance analog,anthraquinone-2,6-disulfonate (AQDS), to peat from the bog 1 andthe rich fen, and it was highly inhibitory ofmethanogenesis in bog 1but acted as an organic TEA in the rich fen (unpublished data).These findings together are suggestive that methanogenesis inombrotrophic peatlands is strongly inhibited by humic-likesubstances that may be derived from Sphagnum spp. mosses.

Our experiment demonstrates that low pH limits methano-genesis rates and CH4 efficiency, but it is also clear that other factorsare important in limiting methanogen activity in all peatlands,and particularly in ombrotrophic peatlands. Despite substantialresearch to date on this topic, the mechanisms limiting methano-genesis in peatlands remain elusive; however the possibility thatorganic substances directly inhibit methanogenesis deservesconsiderably more attention.

4.4. Methanogenic pathways

In freshwater habitats, CH4 is theoretically produced fromacetate and H2/CO2 at a ratio of 2:1 (Conrad, 1999). While CH4production in most natural anaerobic habitats follows this ratiorelatively closely, peatlands are often an exception. More minero-trophic surface peats typically are dominated by acetoclasticmethanogenesis, whereas more ombrotrophic peats, deep peat,and far northern peats are often dominated by hydrogenotrophicmethanogenesis (Chasar et al., 2000; Keller and Bridgham, 2007;Galand et al., 2010). It has been suggested that low pH is the mainfactor that determines the dominance of hydrogenotrophic meth-anogenesis in ombrotrophic peats (Horn et al., 2003; Yavitt andSeidman-Zager, 2006; Kotsyurbenko et al., 2007).

Our results clearly demonstrate that higher pH stimulated therates of both hydrogenotrophic and acetoclastic CH4 production insamples across the peatland gradient (Figs. 6 and 7); however pHdid not have substantial effects on the relative importance of thetwo pathways (Fig. 8). Acetoclastic methanogenesis generallyaccounted for greater than the expected 66% of CH4 production inall sites after Day 2 and hydrogenotrophic methanogenesis neverdominated, even in the two bog sites at the lower pH. These resultsare in contrast to those of Kotsyurbenko et al. (2007), who founda shift from acetoclastic to hydrogenotrophic methanogenesis ina peat with a moderately acidic native pH (4.8) as pH decreasedfrom 6.0 to 3.8 at temperatures �15 �C. They also isolated anacidophilic hydrogenotrophic methanogen of the genus Meth-anobacterium that was postulated to be primarily responsible forthe hydrogenotrophic CH4 production at low pH in their system.Methanobacterium was the dominant methanogen genus in all ofour sites in a May 2010 sampling, but bog 1 also had the highestproportion ofMethanosarcina, a genus that is putatively acetoclastic(unpublished data).

It is possible that in our experiment bubbling the slurries at thebeginning of the CH4 pathway measurements reduced CO2 and H2

R. Ye et al. / Soil Biology & Biochemistry 54 (2012) 36e4746

levels enough to limit hydrogenotrophic methanogenesis. But wehave found in parallel experiments that CO2 quickly accumulates inthe headspace and would only be potentially limiting for a briefperiod, whereas H2 concentrations are always low even afterextended incubations indicating rapid coupling of production andconsumption. These H2 dynamics are typical of other peatlandstudies (Horn et al., 2003; Yavitt and Seidman-Zager, 2006).

As discussed above, in an ancillary experiment we foundsubstantial rates of homoacetogenesis in peats near their nativepHs (unpublished data). Homoacetogenesis may affect the relativedominance of methanogenic pathways, because acetate is thesubstrate for acetoclastic methanogenesis and homoacetogenspotentially compete for H2 with hydrogenotrophic methanogens.Thus, the low contribution of hydrogenotrophic methanogenesis toCH4 productionmay have been at least partially due to high rates ofhomoacetogenesis.

It has also been suggested that hydrogenotrophic methano-genesis is more dominant under situations of low C quality(Hornibrook et al., 1997). Our results contradict this hypothesis ashydrogenotrophic methanogenesis was greatest in the bog sites athigher pH (Fig. 8), where we also observed the highest CO2 produc-tion and acetate accumulation (Figs. 2 and 3). Overall, heterotrophicactivity (CO2 production þ acetate production) was uncorrelatedwith % acetoclastic methanogenesis (r ¼ �0.03, p ¼ 0.61).

Overall, our results suggest that pH is not a major factor in therelative importance of the two methanogenesis pathways in peat-lands. Given the contradictory results of our study and others in theliterature, we suggest that further experiments are necessary toexamine the interaction of pH, C availability, homoacetogenesis,and temperature on the relative contribution of the CH4 pathwaysin a broad range of peat types.

5. Conclusions

pH adjustment significantly altered many facets of anaerobic Cmineralization in six peat typeswith a broad range of native pH, butnot always in expected ways. Both CO2 and acetate productionweregreater at higher pH in all sites in the early stage of experiment,demonstrating a general pH control over fermentation in peatlandsacross the ombrotrophiceminerotrophic gradient. Acetoclastic CH4production accounted for <25% and the reduction of various TEAscontributed to <12% of total gaseous C production, indicatingthat fermentation was the major pathway for anaerobic C miner-alization across the different peat types and pHs. Increased pHpromoted significant acetate accumulation in ombrotrophicpeats, mainly because of enhanced acetogenesis, whereas acetateproduction and consumption were tightly coupled in minero-trophic peats. Our results suggest that the acetate accumulationoften observed in ombrotrophic peatlands is not mainly driven bytheir low pH.

More C flow was directed to methanogenesis as pH increased toa pH of 5.5. Methane production via the hydrogenotrophic pathwaywas not greater at lower pHs, suggesting that pH alone is not thereason that this pathway is often so important in ombrotrophicpeatlands under field conditions. Stimulation of both hydro-genotrophic and acetoclastic methanogenesis at higher pH mayhave been a direct pH effect or due to an increase in substrateavailability, but the large acetate pooling in the ombrotrophic peatsat higher pH indicates that acetoclastic methanogens were not ableto adjust to their environmental conditions as rapidly as acetogenicbacteria over the extended incubation period. Similar disequilib-rium between acetate production and consumption may explainthe temporary acetate pooling often observed in situ. The relativelylow CH4 efficiency in the more ombrotrophic peats across pHssuggests that pH alone does not explain the low CH4 production in

these types of peatlands. Overall, this study suggests that pH isa primary factor controlling fermentative activity in peatlands. pHis also an important control over rates of CH4 production, but otherfactors are likely of equal or greater importance in explaining thelow CH4 production typically observed in ombrotrophic peatlands.

Acknowledgments

We thank N.M. Eisenhut and A. Harvey for their help during theexperiment, James Ziemer and Yvonne Ziemer for access to theirprivate field site, and the University of Notre Dame EnvironmentalResearch Center for access to field sites and laboratory facilities.This work was supported by NSF grant DEB-0816575. A. Harveywas supported by the University of Oregon Summer Programfor Undergraduate Research. Comments from two anonymousreviewers greatly improved this manuscript.

References

Aerts, R., Verhoeven, J.T.A., Whigham, D.F., 1999. Plant-mediated controls onnutrient cycling in temperate fens and bogs. Ecology 80, 2170e2181.

Avery Jr., G.B., Shannon, R.D., White, J.R., Martens, C.S., Alperin, M.J., 2002. Controlson methane production in a tidal freshwater estuary and a peatland: methaneproduction via acetate fermentation and CO2 production. Biogeochemistry 62,19e37.

Avery Jr., G.B., Shannon, R.D., White, J.R., Martens, C.S., Alperin, M.J., 1999. Effect ofseasonal changes in the pathways of methanogenesis on the d13C values ofpore-water methane in a Michigan peatland. Global Biogeochemical Cycles 13,475e484.

Basiliko, N., Yavitt, J.B., 2001. Influence of Ni, Co, Fe, and Na additions on methaneproduction in Sphagnum-dominated northern American peatlands. Biogeo-chemistry 52, 133e153.

Bergman, I., Lundberg, P., Nilsson, M., 1999. Microbial carbon mineralization in anacid surface peat: effects of environmental factors in laboratory incubations.Soil Biology and Biochemistry 31, 1867e1877.

Belyea, L.R., Baird, A.J., 2006. Beyond “The limits to peat bog growth”: cross-scalefeedback in peatland development. Ecological Monographs 76, 299e322.

Blodau, C., 2002. Carbon cycling in peatlands e a review of processes and controls.Environmental Reviews 10, 111e134.

Blodau, C., Roehm, C.L., Moore, T.R., 2002. Iron, sulfur, and dissolved carbondynamics in a northern peatland. Archiv Fur Hydrobiologie 154, 561e583.

Bräuer, S.L., Cadillo-Quiroz, H., Yashiro, E., Yavitt, J.B., Zinder, S.H., 2006. Isolation ofa novel acidiphilic methanogen from an acidic peat bog. Nature 442, 192e194.

Bräuer, S.L., Yavitt, J.B., Zinder, S.H., 2004. Methanogenesis in McLean bog, an acidicpeat bog in upstate New York: stimulation by H2/CO2 in the presence ofrifampicin, or by low concentrations of acetate. Geomicrobiology Journal 21,433e443.

Bridgham, S.D., Ping, C.L., Richardson, J.L., Updegraff, K., 2001. Soils of northernpeatlands: histosols and gelisols. In: Richardson, J.L., Vepraskas, M.J. (Eds.),Wetland Soils: Genesis, Hydrology, Landscapes, and Classification. CRC press,Boca Raton, FL, pp. 343e370.

Bridgham, S.D., Pastor, J., Janssens, J., Chapin, C., Malterer, T., 1996. Multiple limitinggradients in peatlands: a call for a new paradigm. Wetlands 16, 45e65.

Bridgham, S.D., Pastor, J., Updegraff, K., Malterer, T.J., Johnson, K., Harth, C., Chen, J.,1999. Ecosystem control over temperature and energy flux in northernpeatlands. Ecological Applications 9, 1345e1358.

Bridgham, S.D., Richardson, C.J., 1992. Mechanisms controlling soil respiration (CO2and CH4) in southern peatlands. Soil Biology & Biochemistry 24, 1089e1099.

Bridgham, S.D., Updegraff, K., Pastor, J., 1998. Carbon, nitrogen, and phosphorusmineralization in northern wetlands. Ecology 79, 1545e1561.

Cervantes, F.J., Velde, van der, S., Lettinga, G., Field, J.A., 2000. Competition betweenmethanogenesis and quinine respiration for ecologically important substratesin anaerobic consortia. FEMS Microbiology Ecology 34, 161e171.

Chasar, L.S., Chanton, J.P., Glaser, P.H., Siegel, I., 2000. Methane concentration andstable isotope distribution as evidence of rhizospheric processes: comparison ofa fen and bog in glacial Lake Agassiz peatland complex. Analysis of Botany 86,655e663.

Clark, J.M., Chapman, P.J., Adamson, J.K., Lane, S.N., 2005. Influence of drought-induced acidification on the mobility of dissolved organic carbon in peatsoils. Global Change Biology 11, 791e809.

Clymo, R.S., 1983. Peat. In: Gore, A.J.P. (Ed.), Ecosystems of the World 4A: Mires:Swamp, Bog, Fen, and Moor. Elsevier, Amsterdam.

Clymo, R.S., 1987. Interactions of Sphagnumwith water and air. In: Hutchinson, T.C.,Meema, K.M. (Eds.), Effects of Atmospheric Pollutants on Forests, Wetlands andAgricultural Ecosystems. Springer, Berlin, pp. 513e529.

Conrad, R., 1999. Contribution of hydrogen to methane production and controlof hydrogen concentrations in methanogenic soils and sediments. FEMSMicrobiology Ecology 28, 193e202.

R. Ye et al. / Soil Biology & Biochemistry 54 (2012) 36e47 47

Denman, K.L., Brasseur, G., Chidthaisong, A., Ciais, P., Cox, P.M., Dickinson, R.E.,Hauglustaine, D., Heinze, C., Holland, E., Jacob, D., Lohmann, U.,Ramachandran, S., da Silva Dias, P.L., Wofsy, S.C., Zhang, X., 2007. Couplingsbetween changes in the climate system and biogeochemistry. In: Solomon, S.,Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L.(Eds.), Climate Change 2007: The Physical Science Basis. Contribution ofWorking Group I to the Fourth Assessment Report of the IntergovernmentalPanel on Climate Change. Cambridge University Press, Cambridge, UK,pp. 499e587.

Dettling, M.D., Yavitt, J.B., Cadillo-Quiroz, H., Christine, S., Zinder, S.H., 2007. Soil-methanogen interactions in two peatlands (Bog, fen) in central New York State.Geomicrobiology Journal 24, 247e259.

Duddleston, K.N., Kinney, M.A., Kiene, R.P., Hines, M.E., 2002. Anaerobic microbialbiogeochemistry in a northern bog: acetate as a dominant metabolic endproduct. Global Biogeochemical Cycles 16, 1063. doi:10.1029/2001GB001402.

Dunfield, P., Knowles, R., Dumont, R., Moore, T.R., 1993. Methane production andconsumption in temperate and subarctic peat soils: response to temperatureand pH. Soil Biology Biochemistry 25, 321e326.

Forster, P., Ramaswamy, P., Artaxo, P., Berntsen, T., Betts, R., Fahey, D.W., Haywood, J.,Lean, J., Lowe, D.C., Myhre, G., Nganga, J., Prinn, R., Raga, G., Schulz, M.,Dorland, R.V., 2007. Changes in atmospheric constituents and in radioactiveforcing. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B.,Tignor, M., Miller, H.L. (Eds.), Climate Change 2007: The Physical Science Basis.Contribution of Working Group I to the Fourth Assessment Report of theIntergovernmental Panel on Climate Change. Cambridge University Press,Cambridge, UK, pp. 129e234.

Galand, P.E., Yrjälä, K., Conrad, R., 2010. Stable carbon isotope fractionation duringmethanogenesis in three boreal peatland ecosystems. Biogeosciences 7,3893e3900.

Galand, P.E., Fritze, H., Conrad, R., Yrjälä, K., 2005. Pathways for methanogenesis anddiversity of methanogenic archaea in three boreal peatland ecosystems. Appliedand Environmental Microbiology 71, 2195e2198.

Gibbs, C.R., 1976. Characterization and application of ferrozine iron reagent asa ferrous iron indicator. Analytical Chemistry 48, 1197e1201.

Goodwin, S., Zeikus, J.G., 1987a. Ecophysiological adaptations of anaerobic bacteriato low pH: analysis of anaerobic digestion in acidic bog sediments. Applied andEnvironmental Microbiology 53, 57e64.

Goodwin, S., Zeikus, J.G., 1987b. Physiological adaptations of anaerobic bacteria tolow pH: metabolic control of proton motive force in Sarcina ventriculi. Journal ofBacteriology 169, 2150e2157.

Gorham, E., Janssens, J.A., 2005. The distribution and accumulation of chemicalelements in five peat cores form the mid-continent to the eastern coast of NorthAmerica. Wetlands 25, 259e278.

Grybos, M., Davranche, M., Gruau, G., Petitjean, P., Pédrot, M., 2009. Increasing pHdrives organic matter solubilization from wetland soils under reducing condi-tions. Geoderma 154, 13e19.

Heitmann, T., Goldhammer, T., Beer, J., Blodau, C., 2007. Electron transfer of dis-solved organic matter and its potential significance for anaerobic respiration ina northern bog. Global Change Biology 13, 1771e1785.

Hines, M.E., Duddleston, K.N., Rooney-Varga, J.N., Fields, D., Chanton, J.P., 2008.Uncoupling of acetate degradation from methane formation in Alaskanwetlands: connections to vegetation distribution. Global Biogeochemical Cycles22, GB2017. doi:10.1029/2006GB002903.

Hoehler, T.M., Albert, D.B., Alperin, M.J., Martens, C.S., 1999. Acetogenesis from CO2in an anoxic marine sediment. Limnology and Oceanography 44, 662e667.

Horn, M.A., Matthies, C., Küsel, K., Schramm, A., Drake, H.L., 2003. Hydro-genotrophic methanogenesis by moderately acid-tolerant methanogens ofa methane-emitting acidic peat. Applied and Environmental Microbiology 69,74e83.

Hornibrook, E.R.C., Longstaffe, F.J., William, S.F., 1997. Spatial distribution ofmicrobial methane production pathways in temperate zone wetland soils:stable carbon and hydrogen isotope evidence. Geochimica et CosmochimicaActa 61, 745e753.

Jobbágy, E.G., Jackson, R.B., 2000. The vertical distribution of soil organic carbon andits relation to climate and vegetation. Ecological Applications 10, 423e436.

Keller, J.K., Bauers, A.K., Bridgham, S.D., Kellogg, L.E., Iversen, C.M., 2006. Nutrientcontrol of microbial carbon cycling along an ombrotrophiceminerotrophicpeatland gradient. Journal of Geophysical Research Biogeosciences 111, G03006.doi:10.1029/2005JG000152.

Keller, J.K., Bridgham, S.D., 2007. Pathways of anaerobic carbon cycling across anombrotrophiceminerotrophic peatland gradient. Limnology and Oceanography52, 96e107.

Keller, J.K., Bridgham, S.D., Chapin, C.T., Iversen, C.M., 2005. Limited effects of long-term fertilization on carbon mineralization in a Minnesota fen. Soil Biology &Biochemistry 37, 1197e1204.

Keller, J.K., Weisenhorn, P.B., Megonigal, J.P., 2009. Humic acids as electron accep-tors in wetland decomposition. Soil Biology and Biochemistry 41, 1518e1522.

King, G.M., Klug, M.J., Lovley, D.R., 1983. Metabolism of acetate, methanol, andmethylated amines in intertidal sediments of Lowes Cove, Maine. Applied andEnvironmental Microbiology 45, 1848e1853.

Kotsyurbenko, O.R., Chin, K.J., Glagolev, M.V., Stubner, S., Simankova, M.V.,Nozhevnikova, A.N., Conrad, R., 2004. Acetoclastic and hydrogenotrophicmethane production and methanogenic populations in an acidic West-Siberianpeat bog. Environmental Microbiology 6, 1159e1173.

Kotsyurbenko, O.R., Glagolev, M.V., Nozhevnikova, A.N., Conrad, R., 2001. Compe-tition between homoacetogenic bacteria and methanogenic archaeal forhydrogen at low temperature. FEMS Microbiology Ecology 38, 153e159.

Kotsyurbenko, O.R., Friedrich, M.W., Simankova, M.V., Nozhevnikova, A.N.,Golyshin, P.N., Timmis, K.N., Conrad, R., 2007. Shift from acetoclastic toH2-dependent methanogenesis in a West Siberian peat bog at low pH valuesand isolation of an acidophilic Methanobacterium strain. Applied and Environ-mental Microbiology 73, 2344e2348.

Lipson, D.A., Jha, M., Raab, T.K., Oechel, W.C., 2010. Reduction of iron (III) and humicsubstances plays a major role in anaerobic respiration in an Arctic peat soil.Journal of Geophysical Research Biogeosciences 115, G00I06. doi:10.1029/2009JG001147.

Littell, R.C., Milliken, G.A., Stroup, W.W., Wolfinger, R.D., Schabenberger, O., 2006.SAS for Mixed Models. SAS Institute Inc., Cary, NC.

Lovley, D.R., Coates, J.D., Blunt-Harris, E.L., Phillips, E.J.P., Woodward, J.C., 1996.Humic substances as electron acceptors for microbial respiration. Nature 382,445e448.

Maltby, E., Immirzi, P., 1993. Carbon dynamics in peatlands and other wetland soils,regional and global perspectives. Chemosphere 27, 999e1023.

Megonigal, J.P., Hines, M.E., Visscher, P.T., 2004. Anaerobic metabolism: linkages totrace gases and aerobic processes. In: Schlesinger, W.H. (Ed.), Biogeochemistry.Elsevier-Pergamon, Oxford, UK, pp. 317e424.

Minderlein, S., Blodau, C., 2010. Humic-rich peat extracts inhibit sulfate reduction,methanogenesis, and anaerobic respiration but not acetogenesis in peat soils ofa temperate bog. Soil Biology and Biochemistry 42, 2078e2086.

Parent, L.E., Caron, J., 2006. Physical properties of organic soils and growing media:particle size and degree of decomposition. In: Carter, M.R., Gregorich, E.G.(Eds.), Soil Sampling and Methods of Analysis. Taylor & Francis Group, Florida,pp. 871e883.

Reddy, K.R., DeLaune, R.D., 2008. Biogeochemistry of Wetlands: Science andApplication. Taylor and Francis group LLC, Boca Raton, FL.

Russell, J.B., 1991. Intracellular pH of acid-tolerant ruminal bacteria. ApplyEnvironmental Microbiology 57, 3383e3384.

Segers, R., 1998. Methane production and methane consumption: a review ofprocesses underlying wetland methane fluxes. Biogeochemistry 41, 23e51.

Shannon, R.D., White, J.R., 1996. The effects of spatial and temporal variations inacetate and sulfate on methane cycling in two Michigan peatlands. Limnologyand Oceanography 41, 435e443.

Sizova, M.V., Panikov, N.S., Tourova, T.P., Flanagan, W., 2003. Isolation and charac-terization of oligotrophic acido-tolerant methanogenic consortia froma Sphagnum peat bog. FEMS Microbiology Ecology 45, 301e315.

Sjörs, H., 1950. On the relation between vegetation and electrolytes in northSwedish mire water. Oikos 2, 241e257.

Stumm, W., Morgan, J.J., 1995. Aquatic Chemistry: Chemical Equilibria and Rates inNatural Waters. Wiley, New York.

Thurman, E.M., 1985. Organic Geochemistry of Natural Waters. Kluwer AcademicPublishers, Dordrecht, The Netherlands.

Urban, N.R., Eisenreich, S.J., Gorham, E., 1987. Proton cycling in bogs: geographicvariation in northeastern North America. In: Hutchinson, T.C., Meema, K.M.(Eds.), Effects of Atmospheric Pollutants on Forests, Wetlands, and AgriculturalEcosystems. Springer-Verlag, Berlin, pp. 577e598.

Valentine, D.W., Holland, E.A., Schimel, D.S., 1994. Ecosystem and physiologicalcontrols over methane production in northern wetlands. Journal of GeophysicalResearch 99, 1563e1571.

Vile, M.A., Bridgham, S.D., Wieder, R.K., 2003a. Response of anaerobic carbonmineralization rates to sulfate amendments in a boreal peatland. EcologicalApplications 13, 720e734.

Vile, M.A., Bridgham, S.D., Wieder, R.K., Novák, M., 2003b. Atmospheric sulfurdeposition alters pathways of gaseous carbon production in peatlands. GlobalBiogeochemical Cycles 17, 1058e1064.

Vitt, D.H., Bayley, S.E., Jin, T.L., 1995. Seasonal variation in water chemistry overa bog-rich fen gradient in continental western Canada. Canadian Journal ofFisheries and Aquatic Sciences 52, 587e606.

Wieder, R.K., Yavitt, J.B., Lang, G.E., 1990. Methane production and sulfate reductionin two Appalachian peatlands. Biogeochemistry 10, 81e104.

Williams, C.J., Yavitt, J.B., Cleavitt, N.L., 1998. Cupric oxide oxidation products ofnorthern peat and peat-forming plants. Canadian Journal of Botany 76, 51e62.

Williams, R.T., Crawford, R.L., 1984. Methane production in Minnesota peatlands.Applied and Environmental Microbiology 47, 1266e1271.

Wüst, P.K., Horn, M.A., Drake, H.L., 2009. Trophic links between fermenters andmethanogens in a moderately acidic fen soil. Environmental Microbiology 11,1395e1409.

Yavitt, J.B., Seidman-Zager, M., 2006. Methanogenic conditions in northern peatsoils. Geomicrobiology Journal 23, 119e127.

Yavitt, J.B., Lang, G.E., Wieder, R.K., 1987. Control of carbon mineralization to CH4and CO2 in anaerobic, Sphagnum-derived peat from Big Run bog, West Virginia.Biogeochemistry 4, 141e157.

Zinder, S.H., 1993. Physiological ecology of methanogens. In: Ferry, J.G. (Ed.),Methanogenesis: Ecology, Physiology, Biochemistry and Genetics. Chapman &Hall, NY, pp. 128e206.

Zinder, S.H., Anguish, T., 1992. Carbon monoxide, hydrogen, and formate metabo-lism during methanogenesis from acetate by thermophilic cultures of meth-anosarcina and methanothrix strains. Applied and Environmental Microbiology58, 3323e3329.