Temperature, oxygen, and vegetation controlson decomposition in a James Bay peatlandMichael Philben1,2, James Holmquist3,4, Glen MacDonald3,4, Dandan Duan1,5, Karl Kaiser1,6,and Ronald Benner1

1Department of Biological Sciences and Marine Science Program, University of South Carolina, Columbia, South Carolina,USA, 2Present address Department of Earth Sciences, Memorial University of Newfoundland, St. John’s, Newfoundland andLabrador, Canada, 3Department of Ecology and Evolutionary Biology, University of California, Los Angeles, California, USA,4Institute of the Environment and Sustainability, University of California, Los Angeles, California, USA, 5Present address StateKey Laboratory of Organic Geochemistry, Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, Guangzhou,China, 6Present address Departments of Marine Sciences and Oceanography, Texas A&M University, Galveston, Texas, USA

Abstract The biochemical composition of a peat core from James Bay Lowland, Canada, was used toassess the extent of peat decomposition and diagenetic alteration. Our goal was to identify environmentalcontrols on peat decomposition, particularly its sensitivity to naturally occurring changes in temperature,oxygen exposure time, and vegetation. All three varied substantially during the last 7000 years, providing anatural experiment for evaluating their effects on decomposition. The bottom 50 cm of the core formedduring the Holocene Climatic Optimum (~7000–4000years B.P.), when mean annual air temperature was likely1–2°C warmer than present. A reconstruction of the water table level using testate amoebae indicatedoxygen exposure time was highest in the subsequent upper portion of the core between 150 and 225 cmdepth (from ~2560 to 4210 years B.P.) and the plant community shifted from mostly Sphagnum to vascularplant dominance. Several independent biochemical indices indicated that decomposition was greatestin this interval. Hydrolysable amino acid yields, hydroxyproline yields, and acid:aldehyde ratios of syringyllignin phenols were higher, while hydrolysable neutral sugar yields and carbon:nitrogen ratios were lowerin this zone of both vascular plant vegetation and elevated oxygen exposure time. Thus, peat formedduring the Holocene Climatic Optimum did not appear to be more extensively decomposed than peatformed during subsequent cooler periods. Comparison with a core from the West Siberian Lowland,Russia, indicates that oxygen exposure time and vegetation are both important controls on decomposition,while temperature appears to be of secondary importance. The low apparent sensitivity of decompositionto temperature is consistent with recent observations of a positive correlation between peat accumulationrates and mean annual temperature, suggesting that contemporary warming could enhance peatlandcarbon sequestration, although this could be offset by an increasing contribution of vascular plants tothe vegetation.

1. Introduction

Boreal peatlands are important sinks for atmospheric carbon dioxide and currently contain ~550 PgC dueto an imbalance between net primary production and organic matter decomposition [Yu et al., 2010].Organic matter stocks in peatlands are several times higher than in other ecosystems despite lower ratesof primary production, indicating that net accumulation is primarily due to limited decomposition[Schlesinger and Bernhardt, 2013; Thormann et al., 1999]. Decomposition is likely incomplete due to avariety of factors including low temperatures [Lafleur et al., 2005], restricted oxygen exposure [Belyeaand Clymo, 2001; McNicol and Silver, 2015; Philben et al., 2014b], and the poor litter quality of Sphagnum,the dominant moss in many peatland plant communities [Moore et al., 2007]. However, the relativeimportance of these factors remains poorly understood. The lack of a mechanistic understanding of peatdecomposition reduces our ability to predict the responses of peatland carbon reservoirs to futurechanges in temperature, growing season, and precipitation.

The duration of peat exposure to oxygen (oxygen exposure time) has been proposed as an important controlon the extent of peat decomposition [Philben et al., 2014b]. Decomposition is relatively rapid in the oxicsurface layer, and peat decomposition models indicate litter loses 80–90% of its initial mass before burial

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 729

PUBLICATIONSGlobal Biogeochemical Cycles

RESEARCH ARTICLE10.1002/2014GB004989

Key Points:• Biochemical indices were used todetermine decomposition in apeat core

• Decomposition was highest with highoxygen exposure time and vascularvegetation

• Higher temperatures during peatformation did not appear to affectdecomposition

Supporting Information:• Figure S1 and Table S1• Table S2• Table S3• Table S4

Correspondence to:M. Philben,mphilben@mun.ca

Citation:Philben, M., J. Holmquist, G. MacDonald,D. Duan, K. Kaiser, and R. Benner (2015),Temperature, oxygen, and vegetationcontrols on decomposition in a JamesBay peatland, Global Biogeochem. Cycles,29, 729–743, doi:10.1002/2014GB004989.

Received 22 SEP 2014Accepted 1 MAY 2015Accepted article online 6 MAY 2015Published online 1 JUN 2015

©2015. American Geophysical Union. AllRights Reserved.

beneath the water table [Clymo, 1992]. Oxygen is rapidly depleted below the water table by heterotrophicrespiration. The remineralization of organic carbon to carbon dioxide is much less thermodynamicallyfavorable in the absence of oxygen, and decomposition is slower under anaerobic conditions by an orderof magnitude or more [Belyea, 1996; Benner et al., 1984; Clymo, 1965]. Oxygen exposure time is mainly afunction of water table depth and peat accumulation rate, as drier conditions and lower accumulationrates result in slower burial beneath the oxic layer. However, increases in oxygen exposure time decreasethe accumulation rate by promoting decomposition. It is therefore an important feedback and regulatingmechanism for peat accumulation.

A recent study comparing the extent of peat decomposition along a latitudinal transect in the WestSiberian Lowland (WSL), Russia, found that differences in oxygen exposure time were more important indetermining the extent of decomposition than temperature or vegetation change [Philben et al., 2014b].Peat cores collected from the northern WSL were formed under cooler conditions, with mean annualtemperatures as much as 7°C lower than in the south. In contrast, oxygen exposure time was longer inthe north, as peat was buried beneath the water table less rapidly. The southern cores were generallySphagnum-dominated but contained recurring intervals of vascular plant dominance. There was nodifference in decomposition between Sphagnum and vascular plant layers of the southern cores,indicating that plant type was not the primary control on decomposition. The northern cores were moreextensively decomposed, indicating that oxygen exposure time was more important than temperaturein determining the extent of decomposition.

However, the variability of environmental conditions between the northern and southern WSL cores waslikely greater than those driven by in situ climatic change. Accumulation rates were as much as fivefoldhigher in the south, presumably resulting in similar differences in oxygen exposure time. Climate-drivenchanges in the water table depth will likely be significant but less dramatic. The temperature difference of7°C between the sites is similar to the upper extreme of the 3–7°C estimate for warming in the subarcticregion by 2100 [Intergovernmental Panel on Climate Change (IPCC), 2013]. It is possible that temperatureand vegetation could have a greater impact on peat decomposition over naturally occurring ranges ofenvironmental conditions.

The goal of this study is to test the effects of temperature, vegetation, and oxygen exposure time byexamining temporal variability in the extent of peat decomposition in a core from James Bay, Canada. Thecore represents over 7000 years of peat accumulation [Holmquist et al., 2014; Holmquist and MacDonald,2014], providing an opportunity to evaluate peatland responses to natural climate change. Elevatedsummer insolation associated with the Holocene thermal maximum (HTM) increased temperatures acrossnorthern Canada by an average of 1–2 above the twentieth century average between about 10,000 and5000 years B.P. [Kaufman et al., 2004]. One hundred twenty of 140 sites analyzed showed qualitativeevidence of increased temperature during this period, although the few records collected from thevicinity of James Bay Lowlands have been equivocal and suggest that the HTM was delayed to ~7000–4000 years. B.P. due to the persistence of the Laurentide Ice Sheet [Bunbury et al., 2012; Kaufman et al.,2004; O’Reilly et al., 2014]. In addition, the Medieval Climate Anomaly produced a temperature anomalyof about 2°C across northern Canada between 1000 and 700 years B.P., including the James Bay Lowland[Viau and Gajewski, 2009; Bunbury et al., 2012]. The ~2°C temperature anomalies experienced duringboth of these events are analogous to the projected warming over the next 20 years [IPCC, 2013]. Thisenables the estimation of the temperature sensitivity of peat decomposition in a natural system withoutexperimental manipulation.

The historical variability of oxygen exposure time can be estimated using the testate amoebae record. Eachspecies of amoebae inhabits a characteristic range of moisture conditions [Charman and Warner, 1992]. Thetestate amoebae community can therefore yield a reconstruction of the water table level at the site duringpeat formation [Booth, 2002; Woodland et al., 1998]. Calibration with modern analogues has resulted ina quantitative transfer function specific to eastern North America and a better understanding of theuncertainty associated with this approach [Booth, 2008]. Oxygen is rapidly depleted below the water table,making the water table reconstruction a useful proxy for past changes in oxygen exposure time.

We used the biochemical composition of peat as an integrative indicator of its extent of decomposition. Themajor classes of biological macromolecules are degraded at varying rates, and their carbon-normalized

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 730

yields (e.g., nmolmgC�1) can reveal the extent of organic matter decomposition. Hydrolysable neutralsugars are initially abundant in plant litter, primarily in cellulose and hemicellulose, but are among itsmost labile components and are removed rapidly [Opsahl and Benner, 1999]. Hydrolysable amino acids,in contrast, are enriched in microbes compared to plant litter, and their yields generally increase withdecomposition [Hobara et al., 2014; Tremblay and Benner, 2006]. Hydroxyproline, an amino acid found inall plants but not in microbes, is generally less reactive than bulk plant carbon, and its yield tends toincrease with decomposition [Philben and Benner, 2013]. The ratio of the acid and aldehyde monomersof vanillyl and syringyl phenols is also low in plant material but increases with decomposition [Hedgeset al., 1988; Opsahl and Benner, 1995]. Using multiple independent diagenetic indicators can increaseconfidence and improve understanding of the sensitivity of individual indicators. We hypothesize thatthese indicators will show more extensive decomposition in the sections of the core with longer oxygenexposure time.

2. Materials and Methods2.1. Study Site and Collection

This study uses a peat core selected from a network of eight cores collected in 2008 along a latitudinaltransect in the Ontario James Bay Lowland, Canada [Holmquist and MacDonald, 2014; Holmquist et al.,2014]. The JBL7 core (54°23′43″N, 89°31′20″W) was collected from an ombrotrophic, permafrost-freeSphagnum bog using a Russian auger. The core was covered in plastic wrap and aluminum foil and frozenupon return to the laboratory. Subsamples (1 cm) were collected using a bandsaw [Holmquist et al., 2014].An age-depth model for the core was constructed using 14C dating and the Bayesian “Bacon” model[Blaauw and Christen, 2011]. JBL7 had a basal age of 7675 calendar (cal) years B.P. and had an apparentlong-term carbon accumulation rate of 18.4 g Cm�2 yr�1 [Holmquist et al., 2014; Holmquist and MacDonald,2014]. Subsamples were dried in a Savant Speedvac vacuum centrifuge and ground in a Wiley mill forchemical analysis.

Samples of surface vegetation were collected from James Bay and were supplemented by samples fromToolik Lake, Alaska, and Codroy Valley, NL, Canada. These samples were analyzed to identify the chemicalsignatures of fresh plant inputs to the peat. Their chemical compositions are reported elsewhere [Philbenet al., 2014a, 2014b] and summarized herein (Table 1).

2.2. Chemical Analyses

Carbon and nitrogen contents were determined using a Costech ECS 4010 elemental analyzer aftercombustion at 980°C. The mean % deviation was ±0.7% for carbon and ±3% for nitrogen. Hydrolysableneutral sugars were analyzed according to Skoog and Benner [1997] with some modifications. Samples(~5mg dry weight) were placed in 2mL glass ampules and immersed in 12mol L�1 sulfuric acid for 2 hat room temperature. Ampules were then flame-sealed after the addition of 1.8mL deionized water and

Table 1. Elemental and Biochemical Composition of Peat-Forming Vegetation

Sample n wt %Na wt %Ca C:Nb (atm) Hypbc (nmol/mgC) THAAbd (%C) THNSae (%C)

SphagnumSphagnum fuscum 2 0.42 39.6 109.2 8.8 2.5 33.8Sphagnum magellanicum 2 0.29 42.1 170.0 7.2 1.8 37.9Sphagnum angustifolium 2 0.59 39.8 78.9 6.1 2.8 35.3

Vascular plantsCarex leaves 2 0.76 39.6 61.0 7.8 3.5 40.5Eriophorum leaves 2 1.62 44.8 32.0 10.1 7.7 38.1Ledum leaves 3 1.16 49.0 50.0 11.3 5.3 23.9Abies balsamea needle litter 3 1.07 50.1 55.0 8.3 5.2 22.4

aFrom Philben et al. [2014a].bFrom Philben et al. [2014b].cHydroxyproline.dTotal hydrolysable amino acids.eTotal hydrolysable neutral sugars.

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 731

immersed in a water bath at 100°C for 3 h. The hydrolysis was terminated in an ice bath. Ampuleswere cracked open, and deoxyribose was added as an internal standard. About 400 μL of hydrolysatewere neutralized with 2mL self-absorbed ion retardation resin (Biorad AG11 A8) [Kaiser and Benner,2000]. The sample was then deionized using a mixture of cation and anion exchange resins (BioradAG50 and AG2).

Neutral sugars were separated isocratically using 25mmol L�1 NaOH on a PA1 column in a Dionex 500 systemwith a pulsed amperiometric detector. Detector settings were based on Skoog and Benner [1997]. The sevenneutral sugars measured were fucose, rhamnose, arabinose, galactose, glucose, mannose, and xylose. Theanalytical precision of duplicate analyses performed on different days was ±2% for glucose and ±4% forother sugars.

Samples for hydrolysable amino acid analysis were added to glass ampules with 1mL of 6mol L�1

hydrochloric acid. The ampules were flame sealed and heated to 110°C for 20 h. The ampules wereopened after hydrolysis, and aliquots were dried under a stream of N2 gas. Samples were redissolved indeionized water, and norvaline was added as an internal standard.

Amino acids were purified by solid phase extraction and derivatized with propyl chloroformate using thecommercially available EZ:Faast kit for amino acid profiling (Phenomenex, USA). Derivatized sampleswere separated and quantified using gas chromatography (Phenomenex ZB-AAA column; 110–320°C at30°min�1) with flame ionization detection. The 16 quantified amino acids were alanine, glycine, valine,leucine, isoleucine, threonine, serine, proline, aspartic acid, hydroxyproline, glutamic acid, phenylalanine,lysine, histidine, tyrosine, and diaminopimelic acid.

Lignin phenols were analyzed by gas chromatography/mass spectrometry following CuO oxidation [Kaiserand Benner, 2012]. Samples (8–10mg) were mixed with 330mg CuO, 106mg of Fe(NH4)2(SO4)2 · 6H2O, and2.5mL of 2mol L�1 NaOH in Monel reaction vessels. Vessels were kept in argon-filled containers to excludemolecular oxygen. A mixture of transcinnamic acid and ethyl-vanillin was added as internal standards afteroxidation. Samples were purified by solid-phase extraction using Oasis hydrophilic lipophilic balancedcartridges and were dried under a stream of argon. Samples were then redissolved in 200μL dry pyridineand stored frozen until analysis.

Phenols were quantified using an Agilent 7890 gas chromatography system with an Agilent 5975C triple axismass detector with electron impact ionization. Samples were derivatized with N,O-bis(trimethylsilyl)trifluoroacetamide/trimethylchlorosilane reagent at 75°C for 15min prior to analysis. Mass spectrometersettings are described in detail by Kaiser and Benner [2012]. Analyzed phenols included p-hydroxy (P)phenols (p-hydroxybenzaldehyde, p-hydroxyacetophenone, and p-hydroxybenzoic acid); vanillyl (V)phenols (vanillin, acetovanillone, and vanillic acid); syringyl (S) phenols (syringaldehyde, acetosyringone,and syringic acid); cinnamyl (C) phenols (coumaric acid and ferulic acid); and 3,5-dihydroxy-benzoic acid.The analytical precision of duplicate analysis was ±4%.

2.3. Vegetation Reconstruction

The plant community at JBL7 was reconstructed using the carbohydrate composition and the lignin phenolyield of the peat. The carbohydrate approach exploited differences in carbohydrate composition betweenvascular plants and Sphagnum to calculate a carbohydrate Sphagnum index using the molar ratio of sugars:

CarbohydrateSphagnum index ¼ RhamnoseMannoseþ Xylose

This index was converted to a quantitative estimate of the Sphagnum contribution to total peat carbon usingthe equation:

fSphagnum ¼ 2:6998 SICarbð Þ3 � 4:5889 SICarbð Þ2 þ 4:3775 SICarbð Þ � 0:2161

where SICarb is the carbohydrate Sphagnum index and fSphagnum is the fractional contribution of Sphagnumto the peat carbon [Philben et al., 2014a]. The vascular plant contribution was calculated by difference. Theresults of the biochemical vegetation reconstruction were validated by comparison with the stratigraphy ofmajor macrofossil groups [Holmquist and MacDonald, 2014].

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 732

Lignin phenols were used as an independent estimate of the paleovegetation according to the equation:

fVascular ¼ VPIPhenol220

where f Vascular is the fraction of carbon from vascular plants, VPIPhenol is the phenol vascular plant index(nmol V + S phenols mgC�1), and 220 nmolmgC�1 is the average VPIPhenol in fresh vascular plants [Philbenet al., 2014a]. The Sphagnum contribution was calculated by difference.

2.4. Water Table Reconstruction

The water table was reconstructed by testate amoebae analysis using a method modified from Charman[2001] and Booth [2008]. Peat samples were boiled gently and spiked with a known number of exoticLycopodium spores to aid quantification. Testate amoebae were separated from the peat by filteringthrough a 250 um filter and reverse filtering through a 7 um filter. Samples were centrifuged and mixedwith Saffron dye and glycerol as a preservative. Tests were counted under a compound microscope with400X magnification and identified according to Charman [2001] as modified by Booth [2008]. At least 150tests were counted for samples with >6000 tests cm�3, and 50 tests were counted for other samples[Payne and Mitchell, 2009]. Water table depth was calculated, omitting the problematic taxa Difflugia prististype, and using weighted-average transfer function developed from 776 North American surface samples[Booth, 2008; Markel et al., 2010]. Stratigraphic plots, additional methodological data, and a list of taxarecorded are available in Holmquist [2013].

3. Results3.1. Chemical Composition of Peat-Forming Vegetation

Approximately 40% of the total carbon in fresh vegetation was characterized as hydrolysable neutral sugars,amino acids, and lignin phenols (±2% for Sphagnum and ±11% for vascular plants). Vascular plants hadslightly higher carbon (C) content (45.9 ± 4.8 standard deviation wt %C) and substantially higher nitrogen(N) content (1.15 ± 0.35wt %N) than Sphagnum mosses (40.5 ± 1.4 and 0.43 ± 0.15wt %, respectively;Table 1). Sphagnum therefore had a much higher C:N ratio than vascular plants (119 ± 43 versus 50 ± 13).The N content of all vegetation was strongly correlated with the amino acid yield (r2 = 0.99), so thepercentage of C as amino acids in vascular plants (5.43 ± 1.7%C as amino acids) was also approximatelydouble that of Sphagnum (2.37 ± 0.51%C as amino acids). Hydroxyproline yields were slightly higher onaverage in vascular plants than in Sphagnum (9.4 ± 1.6 versus 7.3 ± 1.4 nmol HypmgC�1). Neutral sugaryields were similar in both plant types but were more variable in vascular plants (31.2 ± 9.4%C as neutralsugars) than in Sphagnum (35.6 ± 2.1%C as neutral sugars).

3.2. Environmental History of JBL7

The paleovegetation reconstruction based on lignin phenols consistently indicated a greater contribution ofvascular plant material than the reconstruction based on carbohydrates. The carbohydrate method indicateda larger vascular plant contribution in only five of the 70 peat samples analyzed. Although the two methodsindicated differing magnitudes of vegetation change, they showed consistent patterns. From the surfaceto 150 cm depth, both methods indicated the plant community was >50% Sphagnum, except at 31 cmand 84–86 cm depths, where the phenol method indicated greater vascular plant inputs (Figure 1a). Bothmethods indicated a sustained shift to vascular plant dominance between 150 and 225 cm depth, wherethe phenol method indicated that 15 of the 16 samples were >50% vascular plant C. The carbohydratemethod indicated a smaller vascular plant contribution in this zone, but a distinct transition was stillapparent, as 13 of the 16 samples had >20% vascular plant C, compared to a baseline of about 0–10%above 150 cm. This period of vascular plant dominance lasted approximately 1650 years (4210–2560 calyears B.P.) according to the calibrated radiocarbon age model [Holmquist et al., 2014]. The plant communitywas more variable below 225 cm and appeared to be mostly Sphagnum except for vascular plant incursionsat 259, 303, 315, and 327 cm.

The testate amoebae-derived water table record indicated that conditions were relatively wet (water tableclose to the peat surface) above 150 cm depth (2560 cal years B.P.) (Figure 1b), although a trend towarddrier conditions over the last two centuries was also visible. The water table was generally lower below150 cm, indicating drier conditions during peat formation, including the driest conditions in the core at

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 733

172 cm depth (2740 cal years B.P.; 36 ± 8 cmwater table depth). The average water table depth below 150 cmwas 17 cm, compared to 13 cm above 150 cm. Like the vegetation record, the water table reconstructionindicated rapid change near 150 cm depth. However, the generally drier conditions persisted to the baseof the core, unlike the vascular plant vegetation which was mostly confined to the 150–225 cm interval.

3.3. Chemical Composition of JBL7 Peat

The C content of the peat varied between 39 and 51wt % C through most of the core. The deepest sampleshad lower C content (as little as 32%), likely due to dilution by the underlying mineral soil during the initialestablishment of the peatland on the mineral surface. The N content was much more variable than the Ccontent, with a nearly sevenfold difference between the highest and lowest N concentrations (2.0 and0.3wt %N, respectively; Figure 2). The N content was highest in the vascular plant-dominated zonebetween 150 and 225 cm depth. Samples at 183 and 215–227 cm were particularly enriched in N (>1.5wt%N). N content decreased beneath this zone and was <1.1 between 250 cm and the bottom of the core.The C:N ratio primarily reflected changes in N because the C content was relatively constant. The C:N

Figure 2. Elemental composition and C:N ratio of the peat. The green bar highlights the interval from 150 to 225 cm withvascular plant vegetation.

Figure 1. (a) Vegetation reconstruction using the lignin phenol and carbohydrate methods. The green fill shows the rangeof values obtained using the two methods. (b) Water table reconstruction based on the testate amoebae analysis. The linerepresents the mean estimate of water table depth while the shaded area shows the maximum and minimum estimates.The green bar highlights the interval from 150 to 225 cm with vascular plant vegetation. Best fit radiocarbon ages are fromHolmquist and MacDonald [2014].

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 734

ratio was highest in the upper 50 cm(>90) and generally declined withdepth. The lowest values were foundbetween 150 and 225 cm, with C:Nratios as low as 28 (Figure 2).

The analyzed biochemicals accountedfor 29 ± 6% of the peat C in the core.The percentage of C accounted forby neutral sugars (%C as NS) washighest in the upper 50 cm, wherethe yield in the peat was within therange of yields found in fresh plantmaterial (Figure 3). The sugar yielddeclined with depth but was lowestin the zone 150–225 cm (16 ± 4 %Cas NS). The yield was higher below250 cm (21 ± 4 %C as NS) but was stilllower than in fresh plant material.

Differences in botanical compositioncould influence the sugar yield inde-pendent of decomposition because

vascular plants had a lower average and wider range of neutral sugar yields (Table 1). An “expected” rangeof neutral sugar yields was calculated for each sample based on its reconstructed plant composition tocorrect for variable neutral sugar yields in vegetation. The expected sugar yield was calculated usingthe formula:

Expected yield ¼ %Vascular100

� �%CasNSð ÞVascular þ

%Sphagnum100

� �%CasNSð ÞSphagnum

where “%Vascular” and %Sphagnum” are the vascular plant and Sphagnum contents, respectively, based on anaverage of the carbohydrate and phenol reconstructions. (%C as NS)Sphagnum and (%C as NS)Vascular refer to theneutral sugar yields of fresh Sphagnum and vascular plant tissues, respectively. A range of expected yields wasgenerated by performing the calculation with the highest and lowest sugar yield found in each plant type.

The “sugar degradation index” was calculated as the ratio of the expected sugar yield to the observedyield. The index was calculated using the minimum and maximum expected yields to generate a range

of degradation indices for each sam-ple, and the average of these twovalues was reported. The sugardegradation index was <1.2 in theupper 50 cm of the core, indicatingthat the peat had similar neutral sugaryields compared to fresh vegetation(Figure 3). The index was highestbetween 150 and 225 cm (on average2.2± 0.5), reaching a maximum valueof 3.1 at 183 cm, the sample whichalso had the lowest C:N ratio. Theindex was lower below 225 cm butremained above 1. The sugar degra-dation index was strongly correlatedwith the N:C ratio (n=70, r2 = 0.71,p< 0.001; Figure 4), as samplesbetween 150 and 225 cm generallyhad higher N:C ratios andFigure 4. Correlation between the sugar degradation index and the N:C ratio.

Figure 3. Profile of the percentage carbon as neutral sugars and the sugardegradation index in the core. The gray shaded area on the left indicatesthe expected sugar yield based on the vegetation reconstruction and neutralsugar yields in fresh vegetation. The sugar degradation index is the ratio ofthe expected sugar yield to the observed yield. The green bar highlights theinterval from 150 to 225 cm with vascular plant vegetation.

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 735

degradation indices than samplesabove or below.

The hydrolysable amino acid yield(%C as amino acids) followed a trendsimilar to the N content and thesugar degradation index (Figure 5).The yield was lowest near the sur-face (1.9–2.3 in the upper 50 cm)and was highest between 150 and225 cm (3.9 ± 1.0%C as amino acids).Like the other indicators, the aminoacid yield was highest at 183 and219 cm, with values of 5.4 and5.8%C as amino acids, respectively.However, the amino acid yield wasalso relatively high at 64 cm (5.2% Cas amino acids) unlike the N contentor the sugar degradation index.Overall, the amino acid yield wasstrongly correlated with the totalnitrogen content (n= 70, r2 = 0.82,p< 0.001; Figure S1 in the support-

ing information). The slope of the best fit line (0.34) indicated that on average, amino acids accountedfor 34% of the total nitrogen and the strong correlation indicates that this ratio was relatively constantthroughout the core.

Yields of hydroxyproline (Hyp) generally followed the patterns of the other indicators, with somedifferences. Yields in the upper 50 cm were low and within the range of fresh vegetation, while yieldswere highest between 150 and 225 cm (Figure 5). The highest yield in the core was 49 nmol HypmgC�1,more than double the highest yield found in fresh plant material. However, the pattern of peaks in theHyp yield was somewhat different than peaks in the other indicators. The local maximum at 183 cmcooccurred with maxima in the sugar degradation index and nitrogen content, but the peak in Hyp yieldat 155 cm was not reflected in the other indicators. The peak at 64 cm was also found in the N contentbut not the sugar degradation index. In addition, Hyp yields below 225 cm were generally in the samerange as fresh plant material and similar to peat near the surface, while the N content and sugar

degradation index were elevatedcompared to the surface.

The C:N ratio, sugar degradation index,and hydroxyproline yield indicatedthe extent of bulk peat decomposi-tion, whereas acid:aldehyde (Ad/Al)ratios of lignin phenols were usedto assess the degree of oxidation ofvascular plant detritus. The Ad/Alratios of both vanillyl and syringylphenols were low (<0.4) throughoutthe core (Figure 6). The ratio ofvanillyl phenols did not match thepattern of the other degradationindices. The ratio was higher andmore variable near the surface, withthe highest ratio found at 38 cm,but the total yield of vanillyl phenolswas low in this interval due to the

Figure 6. Profile of acid:aldehyde (Ad:Al) ratios of (left) vanillyl and (right)syringyl phenols in the core. The green bar highlights the interval from150 to 225 cm with vascular plant vegetation.

Figure 5. Percentage of (a) carbon as total hydrolysable amino acids and(b) hydroxyproline yields with depth in the core. The green bar highlightsthe interval from 150 to 225 cm with vascular plant vegetation, and thegray bar in Figure 5b indicates the range of Hyp yields in fresh plantmaterial (Table 1).

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 736

Sphagnum-dominated plant community, so the ratio was likely sensitive to minor variations in plantcomposition. Unlike the other indicators, the Ad/Al ratio of vanillyl phenols was not elevated between150 and 225 cm. The Ad/Al ratio of syringyl phenols was more similar to the patterns of other indices.The syringyl phenol Ad/Al ratio was elevated in the 150–225 cm zone and was higher than values aboveor below, with the exception of a single sample at 31 cm.

4. Discussion4.1. Environmental History and Oxygen Exposure Time

The lignin phenol method for paleovegetation reconstruction consistently indicated a greater contribution ofvascular plants to the peat than the carbohydrate method. The differences were relatively minor nearthe peat surface but were greatest in the highly decomposed interval between 150 and 225 cm, where thephenol method often indicated 100% of the C was from vascular plants compared to <50% using thecarbohydrate method. The lignin phenol method overestimated the vascular plant abundance in thisinterval, as both Sphagnum and vascular plant macrofossils were present [Holmquist and MacDonald,2014]. This supports the assessment of Philben et al. [2014a] that the lignin phenol method likelyoverestimates the vascular plant contribution in extensively decomposed peats because lignin is moreresistant to biodegradation than bulk plant C, especially under anaerobic conditions [Benner et al., 1984].The carbohydrate method, in contrast, likely overestimates the Sphagnum contribution due to relativelyhigher microbial production of rhamnose compared to other sugars and differences in reactivity betweenSphagnum and vascular plant-derived sugars [Philben et al., 2014a]. However, it is also possible thatdifferences between the methods are due to biases in the vegetation indices given the variability of ligninphenol yields among vascular plants and the relatively small sample set.

The position of the water table is a key factor in determining the duration of peat exposure to molecularoxygen, i.e., oxygen exposure time, because oxygen is rapidly depleted below the water table. The averagewater table level was lower in the bottom portion of the core (>150 cm depth) than in the upper portion,so oxygen exposure time was correspondingly longer. Based on the water table reconstruction and theBacon age-depth model, peat between 50 and 150 cm was buried below the water table after 280 yearson average, compared to 480 years below 150 cm, indicating an approximately 70% increase in oxygenexposure time. In addition, the higher vascular plant input between 150 and 225 cm would likely increaseoxygen time in this interval and underlying peat. Wetland plant roots release oxygen into the otherwiseanoxic rhizosphere [Armstrong and Beckett, 1987]. The magnitude of rhizosphere oxidation in peatlandsvaries by species but is sufficient to have a substantial impact on methane oxidation rates [Fritz et al.,2011; Watson et al., 1997]. This effect could further increase the difference in oxygen exposure timebetween the lower (>150 cm) and upper portions of the core.

The water table was also relatively low in the upper 50 cm of the core. Themost recently deposited peat formedwith a water table depth of 28±8 cm, comparable to or drier than most of the samples below 150 cm. Peat inthis interval was not more extensively decomposed despite the drier conditions. The surface layer includes thecurrent acrotelm, where peat is still experiencing relatively rapid decomposition. The apparent accumulationrate was higher in this interval than from 150 to 225 cm (1.8 versus 0.5mm/yr), which could indicate lowoxygen exposure time despite drier conditions. However, incomplete decomposition near the surface caninflate these estimates [Frolking et al., 2014; Tolonen and Turunen, 1996]. The 150–225 cm interval likelyexperienced the longest oxygen exposure time of peat in the current catotelm, although peat currently nearthe surface could ultimately experience similar exposure times.

4.2. Peat Decomposition

Several independent indicators showed the peat between 150 and 225 cm experienced the highest degreeof decomposition. The C:N ratio, which was lowest in the 150–225 cm interval, is often used as a diageneticindicator in peats based on the assumption of conservative N recycling. While organic C is respired to CO2

during decomposition, N is retained by microbes and mostly remains within the system [Kuhry and Vitt,1996; Malmer and Holm, 1984; Malmer and Wallen, 2004]. Although the C:N ratio can be influenced bychanges in atmospheric N deposition and vegetation change [Bragazza et al., 2006; Gerdol et al., 2007], it isstrongly correlated with independent indicators of decomposition, including the syringyl phenol acid:aldehyde ratios and hydroxyproline yields used in this study [Philben et al., 2014b; Zaccone et al., 2008].

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 737

Hydrolysable amino acids provide a molecular perspective of the C:N ratio, as they accounted for ~46% oftotal nitrogen in the peat. The amino acid yield was strongly correlated with the N:C ratio, showing thatthis fraction was relatively constant throughout the core. Even in the extensively decomposed 150–225 cminterval, which had a nitrogen content up to 4 times higher than the rest of the core, the fraction ofnitrogen as amino acids varied between 32 and 51%, similar to peat in surrounding layers. Amino acids arepreferentially removed during decomposition of marine organic matter [Amon et al., 2001; Cowie andHedges, 1992, 1994] and during the formation of soil organic matter [Friedel and Scheller, 2002; Schmidtet al., 2000], resulting in a decline in %N as amino acids, but this is not observed in peatlands [this study;Philben et al., 2014b]. This supports the hypothesis that nitrogen is conservatively recycled in peatlands[Malmer and Holm, 1984], as substantial losses would likely result in the preferential removal of aminoacids and a lower % nitrogen as amino acids in highly decomposed peats. Instead, it appears thatmicrobes continuously regenerate amino acids, maintaining them as a relatively constant fraction oftotal nitrogen.

Hydrolysable neutral sugars are the most abundant class of biochemicals in peat-forming plants, accountingfor up to 41% of the peat C (Table 1). Hydrolysable neutral sugars are removed more rapidly than bulk C indecaying plant detritus, so lower sugar yields are indicative of more extensive decomposition [Opsahl andBenner, 1999]. This can be complicated by vegetation change, but even accounting for the difference inneutral sugar yields between Sphagnum and vascular plants by using the sugar degradation index, the150–225 cm interval was more decomposed than the surrounding peat. The strong correlation betweenthe sugar degradation index and the N:C ratio increases confidence in its use as a diagenetic indicator.

A number of other diagenetic indicators have been proposed based on the composition of neutral sugars,such as the mole % of glucose [Hedges et al., 1994; Opsahl and Benner, 1999] or the ratio of plant-derivedsugars (arabinose and xylose) to microbial-derived sugars (rhamnose, fucose, galactose, and mannose)[Glaser et al., 2000; Guggenberger et al., 1994]. The mole % glucose profile is consistent with the otherindicators of decomposition, as it is generally high (55–70%) in the peat but is considerably lower (<45%)at 183, 219–227, and 243 cm, zones of extensive decomposition according to the other indicators. Unlikethe other indicators, however, only peat from these five depths was significantly different from the rest ofthe core and the mole % glucose was not typically lower in the 150–225 cm interval. Compositional indicesrelated to plant versus microbial-derived sugars did not correlate with the sugar yield or other indicatorsof decomposition used in this study. This is likely due to the substantial differences in sugar compositionbetween vascular plants and Sphagnum. Sphagnum is relatively rich in rhamnose and fucose, whereasvascular plants are enriched in arabinose and xylose [Comont et al., 2006; Jia et al., 2008]. Vegetationchange can therefore cause dramatic changes in the mole % of these sugars. This makes the neutral sugarcomposition very useful for vegetation reconstruction but complicates its utility in assessing decompositionin peatlands.

Hydroxyproline (Hyp) is of similar or greater resistance to decomposition than bulk plant C, and its yieldcommonly increases with decomposition even though it is not synthesized by bacteria or fungi [Philbenand Benner, 2013]. Yields above the range found in vascular plants can therefore be interpreted as anotherindependent indicator of extensive decomposition. Hydroxyproline yields, like the C:N ratio and the sugardegradation index, indicate the most highly degraded peats between 150 and 225 cm and in particularsamples at 183 and 219 cm. The yields found in these samples (up to 49 nmol HypmgC�1) are similar tothose found in highly decomposed peat in West Siberian Lowland [Philben et al., 2014b].

In contrast to the other indicators, the acid:aldehyde (Ad/Al) ratios of lignin phenols do not clearly indicateextensive decomposition in the 150–225 cm zone. The Ad/Al ratio of syringyl phenols was elevated in thisinterval, but there was no apparent change in the Ad/Al ratio of the vanillyl phenols compared to thebaseline of surrounding peat. In addition, both Ad/Al ratios were relatively low (<0.4 for V phenols, <0.3for S phenols) compared with those in highly decomposed peat (0.6–1.5 [Philben et al., 2014b; Williams andYavitt, 2003]) and mineral soils (>1.5 [Feng and Simpson, 2007; Otto and Simpson, 2006]).

The apparent disagreement between the Ad/Al ratios and the other independent indicators of decomposition(C:N ratios, amino acid yields, Hyp yields, and sugar degradation indices) is consistent with previous studiesshowing a poor agreement between Ad/Al ratios and other indicators [Disnar et al., 2008; Schellekenset al., 2015]. This is likely because these indicators have widely varying sensitivities and are therefore

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 738

useful over different time scales. Decomposition models indicate that peat entering the catotelm has lostapproximately 80–90% of its initial mass [Clymo, 1992], so the ideal diagenetic indicator would be mostsensitive over this range. In a 4 year decomposition study of three types of vascular plant litter, theneutral sugar yield was most sensitive in the early and middle stages of decomposition and declined byan average of 30% after 70% mass loss [Opsahl and Benner, 1999]. Hydroxyproline yields doubled after80% mass loss and increased another 40% between 80 and 90% mass loss [Philben and Benner, 2013].Lignin phenol Ad/Al ratios, in contrast, increased little up to 90% mass loss but increased exponentially withsubsequent mass loss [Opsahl and Benner, 1995]. This suggests that neutral sugar and hydroxyproline yieldsare more sensitive to differences in decomposition between peat samples, while substantial changes to theAd/Al ratios only occur with more extensive decomposition. Differences in sensitivity among the variousindicators emphasize the utility of using multiple indicators in interpreting the extent of decomposition.

4.3. Controls on Decomposition

The pattern of decomposition in the JBL7 core, with extensively decomposed peat between 150 and 225 cmand less decomposed peat above and below, provides a unique opportunity to examine the controls on peatdecomposition. First, it illustrates that the extent of decomposition is not always correlated with peat age asassumed in classic static models [Clymo, 1984; Frolking et al., 2001]. The vegetation succession at JBL7 differsfrom many peat sequences in that it shows a transition from Sphagnum-dominated after initiation, tovascular plant dominated, and then a shift back to Sphagnum dominance. This pattern is consistent withthe plant macrofossil record of the core [Holmquist and MacDonald, 2014]. Most peatlands begin asminerotrophic fens inhabited mainly by sedges and become increasingly ombrotrophic, allowing Sphagnumto outcompete vascular plants [van Breemen, 1995], although succession from bog to fen has also beenobserved [Van Bellen et al., 2013]. The unusual pattern in the present core provides an opportunity toevaluate the effects of vegetation and associated environmental conditions on decomposition. In a commonpeat profile with a Sphagnum layer underlain by a vascular plant-dominated interval, both the age of thepeat and the contribution of vascular plants increase with depth, making it difficult to separate the effects ofpeat age and litter quality on decomposition.

The JBL7 core contains a layer of mostly vascular plant remains in the middle of the core, avoiding thecovariance of age and vegetation type found in most peat profiles. Peat near the base of the core wasthousands of years older than peat between 150 and 225 cm, yet the latter was more extensivelydecomposed according to several independent indicators. This demonstrates that differences in litterquality and environmental conditions in the acrotelm can be more significant for the extent of peatdecomposition than lengthy exposure to decomposition under anaerobic conditions in the catotelm.This is remarkable because peat typically spends only tens or hundreds of years in the acrotelm compared tothousands of years in the catotelm, but it is consistent with theoretical modeling of peat decomposition[Frolking et al., 2001].

The results also indicate that temperature was likely not the primary driver of decomposition in this core. TheHolocene thermal maximum (HTM) lasted from approximately 7000–4000 years B.P. at sites near James Bay,with temperatures up to 1–2°C warmer than present [Kaufman et al., 2004]. This event approximatelycorresponds to the interval between 217 and 318 cm according to the Bacon age-depth model [Holmquistet al., 2014], so the bottom ~100 cm of the core likely formed under warmer conditions than the rest ofthe core. However, the extent of decomposition in this layer was similar to the 100–150 cm interval, wherethe plant community was also Sphagnum-dominated, and lower than in the 150–225 cm interval.

Samples at 86 and 88 cm were formed during the Medieval Climate Anomaly (1000–700 years B.P.), withtemperature about 2°C above the twentieth century average [Viau and Gajewski, 2009; Bunbury et al.,2012]. Like those formed during the Holocene thermal maximum, there is no evidence that the warmertemperatures enhanced decomposition, as the sugar degradation index, amino acid yield, andhydroxyproline yield remained low during this interval. The sample at 84 cm had an elevated sugardegradation index but was formed after the Medieval Climate Anomaly according the Bacon model(639 years B.P.). The extent of peat decomposition was low during both the Holocene thermal maximumand the Medieval Climate Anomaly despite higher temperatures, indicating that temperature is notthe primary control of decomposition. The low apparent temperature sensitivity of decomposition isconsistent with results from the West Siberian Lowland (WSL), where cores from the northern WSL were

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 739

more decomposed than those fromthe southern WSL despite experien-cing mean annual temperatures asmuch as 7°C cooler than those in thesouth [Philben et al., 2014b].

The zone of elevated decomposition inJBL7 coincided with a sustained transi-tion to vascular plant vegetation andthe interval with longest oxygen expo-sure time, complicating efforts toresolve the relative importance ofthe two factors. In a different core(SIB06) collected from the West SiberianLowland, several peat layers are domi-nated by vascular plant remains butappear less extensively decomposedthroughout (Figure 7) [Philben et al.,2014a, 2014b]. This cannot be explainedby differences in the vascular plantcommunities between regions, as the

lignin phenol and neutral sugar compositions indicated that a variety of plant types including sedgesand gymnosperm needles were present in both cores.

To reconcile these apparently contradictory results, we used a simplified version of the Peat DecompositionModel [Frolking et al., 2001] to simulate differences in the decomposition trajectory between vascular plantsand Sphagnum at different levels of oxygen exposure. The model uses litter bag decomposition rates of therespective plant types as initial decay constants, which decline with time due to decreasing litter qualityaccording to the equation:

m tð Þ ¼ m0

1þ k0t

where m0 and m(t) are the mass remaining at time 0 and time t and k0 is the initial decay constant ofSphagnum or vascular plant litter (see Table S1 in the supporting information for model parameters).The decay constant declines sharply after the designated oxygen exposure time, simulating low rates ofanaerobic decomposition below the water table. We used this approach to model 500 years ofdecomposition for cohorts of Sphagnum and vascular plants with 10 years, 30 years, and unlimitedoxygen exposure times. The modeled mass loss was compared to the expected range of sensitivities forthe biochemical indicators of decomposition. This varies by indicator, with the sugar degradation index

Figure 8. Mass loss from Sphagnum and vascular plants during 500 years of decomposition with unlimited, 10 year or30 year oxygen exposure time. The model uses a simplified version of the Peat Decomposition Model [Frolking et al.,2001] with decay rates declining with time and decreasing 100-fold after the oxygen exposure time. The shaded areacorresponds to the approximate range of sensitivity of the biochemical indicators (70–100% mass loss).

Figure 7. Relationship between the botanical composition and extent ofdecomposition in JBL7 (this study) and SIB06, a core from West SiberianLowland, Russia (data from Philben et al. [2014a]). The plant communitywas calculated in both cores as the mean of results from the neutral sugarand lignin phenol methods (section 2.3).

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 740

sensitive at>70%mass loss and Ad/Al ratios likely requiring>90%mass loss to show extensive decomposition(section 3.2).

The model indicated that with unlimited oxygen exposure time, Sphagnum and vascular plant decompositionconverge in the long term and both would appear extensively decomposed (Figure 8). This scenario isequivalent to long-term litter bag decomposition studies conducted in well-drained soils, in which theplant species and litter quality influence decomposition rates during the initial stages of decay but donot affect the ultimate extent of mass loss [Prescott, 2010]. With a limited oxygen exposure time, theextent of decomposition of vascular plants and Sphagnum diverges during early diagenesis and thedifferences are preserved due to slow anaerobic decomposition in the catotelm. Vascular plant remainswould appear extensively decomposed, while Sphagnum mass loss is not sufficient to be detected bythe indicators. This scenario is representative of the bottom portion of JBL7, where oxygen exposuretime was relatively high, but only the vascular plant-dominated layers appeared to be extensivelydecomposed. In contrast, a short oxygen exposure time would generate differences in decompositionbetween Sphagnum and vascular plants, but with a lower overall extent of decomposition for both planttypes. Although vascular plants are more decomposed, they remain near the lower limit of the sensitivity ofthe degradation indices, and neither type is interpreted as being extensively decomposed. This scenario isequivalent to SIB06, where neither intervals with Sphagnum or vascular plant dominance show evidence ofextensive decomposition due to limited oxygen exposure time.

These patterns illustrate the impact of both oxygen exposure time and vegetation on the extent ofdecomposition observed in the cores. Vascular plant-dominated peat is generally more extensivelydecomposed than Sphagnum-dominated peat, demonstrated by the close relationship betweenvegetation and the extent of decomposition in JBL7. The extent of decomposition also appears to behighly sensitive to small changes in oxygen exposure time, as shown by the differences betweenvascular plant-dominated intervals of JBL7 and SIB06. In addition, the convergence of Sphagnum andvascular plant decomposition with time in the absence of oxygen limitation indicates that oxygenexposure time is the ultimate control on decomposition in peatlands.

Our finding that peat decomposition is primarily controlled by oxygen exposure time and plant type ratherthan temperature has important implications for future changes in peatland C balance. It suggests thatwarmer temperatures in the arctic will have little direct impact on peatland decomposition, althoughan indirect effect is likely due to thaw in permafrost peatlands. Permafrost thaw exposes previouslyinaccessible peat to decomposition, but this appears to be offset by increased plant growth [Robinson andMoore, 2000; Turetsky et al., 2007]. In contrast to decomposition, Sphagnum net primary production isstrongly correlated with photosynthetically available radiation integrated over the growing season (PAR0),which is increasing due to climate change [Loisel et al., 2012]. This is consistent with observations that netC accumulation over the last 2000 years is correlated with mean annual temperature in the West SiberianLowland [Beilman et al., 2009] and in the James Bay Lowland [Holmquist et al., 2014]. In addition, globalpeat C accumulation rates declined during the transition from the Medieval Climate Anomaly to the LittleIce Age, likely due to decreasing temperature [Charman et al., 2013]. Within the climate window of peatformation, net primary production is more sensitive to temperature than decomposition, suggesting thatnet C accumulation in peatlands could increase with future warming in the subarctic. This could be offsetby the replacement of Sphagnum with faster decomposing vascular plants due to warming or increasingatmospheric N deposition in polluted peatlands [Berendse et al., 2001; Limpens et al., 2011].

ReferencesAmon, R. M. W., H. P. Fitznar, and R. Benner (2001), Linkages among the bioreactivity, chemical composition, and diagenetic state of marine

dissolved organic matter, Limnol. Oceanogr., 46(2), 287–297.Armstrong, W., and P. Beckett (1987), Internal aeration and development of stela anoxia in submerged roots, New Phytol., 105(2),

221–245.Beilman, D. W., G. M. MacDonald, L. C. Smith, and P. J. Reimer (2009), Carbon accumulation in peatlands of West Siberia over the last 2000

years, Global Biogeochem. Cycles, 23, GB1012, doi:10.1029/2007GB003112.Belyea, L. R. (1996), Separating the effects of litter quality and microenvironment on decomposition rates in a patterned peatland, Oikos, 77,

529–539.Belyea, L. R., and R. Clymo (2001), Feedback control of the rate of peat formation, Proc. R. Soc. London, Ser. B, 268(1473), 1315–1321.Benner, R., A. E. Maccubbin, and R. E. Hodson (1984), Anaerobic biodegradation of the lignin and polysaccharide components of

lignocellulose and synthetic lignin by sediment microflora, Appl. Environ. Microbiol., 47(5), 998–1004.

AcknowledgmentsData supporting Figures 1–6 are providedin Tables S1–S3 in the supportinginformation. Radiocarbon ages shownin Figure 1 can be found in Holmquistand MacDonald [2014]. This work wassupported by the National ScienceFoundation (grants NSF-0843417,NSF-0843685, and NSF-0628598). Wethank S. Frolking and an anonymousreviewer for their comments andsuggestions that improved the manu-script. We also thank S. Ziegler andF. Murrin for providing vegetation andfungi from Newfoundland, K. Koba andS. Hobara for vegetation from ToolikLake, and F. Chapelle for access to aCostech CHN elemental analyzer.

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 741

Berendse, F., et al. (2001), Raised atmospheric CO2 levels and increased N deposition cause shifts in plant species composition andproduction in Sphagnum bogs, Global Change Biol., 7(5), 591–598, doi:10.1046/j.1365-2486.2001.00433.x.

Blaauw, M., and J. A. Christen (2011), Flexible paleoclimate age-depth models using an autoregressive gamma process, Bayesian Anal., 6,457–474, doi:10.1214/ba/1339616472.

Booth, R. K. (2002), Testate amoebae as paleoindicators of surface-moisture changes on Michigan peatlands: Modern ecology and hydrologicalcalibration, J. Paleolimnol., 28(3), 329–348.

Booth, R. K. (2008), Testate amoebae as proxies for mean annual water-table depth in Sphagnum-dominated peatlands of North America,J. Quat. Sci., 23(1), 43–57.

Bragazza, L., et al. (2006), Atmospheric nitrogen deposition promotes carbon loss from peat bogs, Proc. Natl. Acad. Sci. U.S.A., 103(51),19,386–19,389, doi:10.1073/pnas.0606629104.

Bunbury, J., S. A. Finkelstein, and J. Bollmann (2012), Holocene hydro-climatic change and effects on carbon accumulation inferred from apeat bog in the Attawapiskat River watershed, Hudson Bay Lowlands, Canada, Quat. Res., 78(2), 275–284.

Charman, D. (2001), Biostratigraphic and palaeoenvironmental applications of testate amoebae, Quat. Sci. Rev., 20(16), 1753–1764.Charman, D., and B. Warner (1992), Relationship between testate amoebae (Protozoa: Rhizopoda) and microenvironmental parameters on a

forested peatland in northeastern Ontario, Can. J. Zool., 70(12), 2474–2482.Clymo, R. (1965), Experiments on breakdown of Sphagnum in two bogs, J. Ecol., 53, 747–758.Clymo, R. (1992), Models of peat growth, Suo, 43(4–5), 127–136.Clymo, R. S. (1984), The limits to peat bog growth, Philos. Trans. R. Soc. London, Ser. B, 303(1117), 605–654.Comont, L., F. Laggoun-Defarge, and J. R. Disnar (2006), Evolution of organic matter indicators in response to major environmental

changes: The case of a formerly cut-over peat bog (Le Russey, Jura Mountains, France), Org. Geochem., 37(12), 1736–1751, doi:10.1016/j.orggeochem.2006.08.005.

Cowie, G. L., and J. I. Hedges (1992), Sources and reactivities of amino acids in a coastal marine environment, Limnol. Oceanogr., 37(4),703–724.

Cowie, G. L., and J. I. Hedges (1994), Biochemical indicators of diagenetic alteration in natural organic matter mixtures, Nature, 369(6478),304–307, doi:10.1038/369304a0.

Disnar, J.-R., J. Jacob, M. Morched-Issa, N. Lottier, and F. Arnaud (2008), Assessment of peat quality by molecular and bulk geochemicalanalysis: Application to the Holocene record of the Chautagne marsh (Haute Savoie, France), Chem. Geol., 254(1–2), 101–112, doi:10.1016/j.chemgeo.2008.06.004.

Feng, X., and M. J. Simpson (2007), The distribution and degradation of biomarkers in Alberta grassland soil profiles, Org. Geochem., 38(9),1558–1570.

Friedel, J. K., and E. Scheller (2002), Composition of hydrolysable amino acids in soil organic matter and soil microbial biomass, Soil Biol.Biochem., 34(3), 315–325, doi:10.1016/s0038-0717(01)00185-7.

Fritz, C., V. A. Pancotto, J. Elzenga, E. J. Visser, A. P. Grootjans, A. Pol, R. Iturraspe, J. G. Roelofs, and A. J. Smolders (2011), Zero methaneemission bogs: Extreme rhizosphere oxygenation by cushion plants in Patagonia, New Phytol., 190(2), 398–408.

Frolking, S., N. T. Roulet, T. R. Moore, P. J. Richard, M. Lavoie, and S. D. Muller (2001), Modeling northern peatland decomposition and peataccumulation, Ecosystems, 4(5), 479–498.

Frolking, S., J. Talbot, and Z. M. Subin (2014), Exploring the relationship between peatland net carbon balance and apparent carbonaccumulation rate at century to millennial time scales, The Holocene, 24(9), 1167–1173, doi:10.1177/0959683614538078.

Gerdol, R., A. Petraglia, L. Bragazza, P. Iacumin, and L. Brancaleoni (2007), Nitrogen deposition interacts with climate in affecting productionand decomposition rates in Sphagnum mosses, Global Change Biol., 13(8), 1810–1821, doi:10.1111/j.1365-2486.2007.01380.x.

Glaser, B., M.-B. Turrion, D. Solomon, A. Ni, and W. Zech (2000), Soil organic matter quantity and quality in mountain soils of the Alay Range,Kyrgyzia, affected by land use change, Biol. Fertil. Soils, 31(5), 407–413.

Guggenberger, G., B. T. Christensen, and W. Zech (1994), Land-use effects on the composition of organic matter in particle-size separates ofsoil: I. Lignin and carbohydrate signature, Eur. J. Soil Sci., 45(4), 449–458.

Hedges, J. I., R. A. Blanchette, K. Weliky, and A. H. Devol (1988), Effects of fungal degradation on the CuO oxidation products of lignin:A controlled laboratory study, Geochim. Cosmochim. Acta, 52(11), 2717–2726.

Hedges, J. I., G. L. Cowie, J. E. Richey, P. D. Quay, R. Benner, M. Strom, and B. R. Forsberg (1994), Origins and processing of organic matter in theAmazon River as indicatated by carbohydrates and amino acids, Limnol. Oceanogr., 39(4), 743–761.

Hobara, S., T. Osono, D. Hirose, K. Noro, M. Hirota, and R. Benner (2014), The roles of microorganisms in litter decomposition and soilformation, Biogeochemistry, 118(1–3), 471–486.

Holmquist, J. R. (2013), Holocene peatland carbon accumulation, ecology, and hydrology in the Canadian James Bay Lowlands PhD dissertation,Univ. of California, Los Angeles, Los Angeles, Calif.

Holmquist, J. R., and G. M. MacDonald (2014), Peatland succession and long-term apparent carbon accumulation in central and northernOntario, Canada, Holocene, 24(9), 1075–1089.

Holmquist, J. R., G. M. MacDonald, and A. Gallego-Sala (2014), Peatland initiation, carbon accumulation, and 2 ka depth in the James BayLowland and adjacent regions, Arct. Antarct. Alp. Res., 46(1), 19–39.

Intergovernmental Panel on Climate Change (IPCC) (2013), Climate Change 2013: The Physical Science Basis. Contribution of Working Group I tothe Fifth Assessment Report of the Intergovernmental Panel on Climate Change, edited by T. F. Stocker et al., 1535 pp., Cambridge Univ. Press,Cambridge, U. K., and New York, doi:10.1017/CBO9781107415324.

Jia, G. D., J. A. J. Dungait, E. M. Bingham, M. Valiranta, A. Korhola, and R. P. Evershed (2008), Neutral monosaccharides as biomarker proxies forbog-forming plants for application to palaeovegetation reconstruction in ombrotrophic peat deposits, Org. Geochem., 39(12), 1790–1799,doi:10.1016/j.orggeochem.2008.07.002.

Kaiser, K., and R. Benner (2000), Determination of amino sugars in environmental samples with high salt content by high performance anionexchange chromatography and pulsed amperometric detection, Anal. Chem., 72(11), 2566–2572, doi:10.1021/ac991407t.

Kaiser, K., and R. Benner (2012), Characterization of lignin by gas chromatography and mass spectrometry using a simplified CuO oxidationmethod, Anal. Chem., 84(1), 459–464, doi:10.1021/ac202004r.

Kaufman, D., T. Ager, N. Anderson, P. Anderson, J. Andrews, P. Bartlein, L. Brubaker, L. Coats, L. C. Cwynar, and M. Duvall (2004), Holocenethermal maximum in the western Arctic (0–180 W), Quat. Sci. Rev., 23(5), 529–560.

Kuhry, P., and D. H. Vitt (1996), Fossil carbon/nitrogen ratios as a measure of peat decomposition, Ecology, 77(1), 271–275, doi:10.2307/2265676.

Lafleur, P. M., T. R. Moore, N. T. Roulet, and S. Frolking (2005), Ecosystem respiration in a cool temperate bog depends on peat temperaturebut not water table, Ecosystems, 8(6), 619–629, doi:10.1007/s10021-003-0131-2.

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 742

Limpens, J., et al. (2011), Climatic modifiers of the response to nitrogen deposition in peat-forming Sphagnummosses: A meta-analysis, NewPhytol., 191(2), 496–507, doi:10.1111/j.1469-8137.2011.03680.x.

Loisel, J., A. V. Gallego-Sala, and Z. Yu (2012), Global-scale pattern of peatland Sphagnum growth driven by photosynthetically activeradiation and growing season length, Biogeosciences, 9(7), 2737–2746, doi:10.5194/bg-9-2737-2012.

Malmer, N., and E. Holm (1984), Variation in the C/N-quotient of peat in relation to decomposition rate and age determination with 210 Pb,Oikos, 43, 171–182.

Malmer, N., and B. Wallen (2004), Input rates, decay losses and accumulation rates of carbon in bogs during the last millennium: Internalprocesses and environmental changes, Holocene, 14(1), 111–117, doi:10.1191/0959683604h1693rp.

Markel, E. R., R. K. Booth, and Y. Qin (2010), Testate amoebae and δ13C of Sphagnum as surface-moisture proxies in Alaskan peatlands,

Holocene, 20, 463–475.McNicol, G., and W. L. Silver (2015), Non-linear response of carbon dioxide and methane emissions to oxygen availability in a drained

histosol, Biogeochemistry, 123(1–2), 299–306, doi:10.1007/s10533-015-0075-6.Moore, T. R., J. L. Bubier, and L. Bledzki (2007), Litter decomposition in temperate peatland ecosystems: The effect of substrate and site,

Ecosystems, 10(6), 949–963, doi:10.1007/s10021-007-9064-5.Opsahl, S., and R. Benner (1995), Early diagenesis of vascular plant tissues: Lignin and cutin decomposition and biogeochemical implications,

Geochim. Cosmochim. Acta, 59(23), 4889–4904, doi:10.1016/0016-7037(95)00348-7.Opsahl, S., and R. Benner (1999), Characterization of carbohydrates during early diagenesis of five vascular plant tissues,Org. Geochem., 30(1),

83–94, doi:10.1016/s0146-6380(98)00195-8.O’Reilly, B. C., S. A. Finkelstein, and J. Bunbury (2014), Pollen-derived paleovegetation reconstruction and long-term carbon accumulation at

a fen site in the Attawapiskat River Watershed, Hudson Bay Lowlands, Canada, Arct. Antarct. Alp. Res., 46(1), 6–18.Otto, A., andM. J. Simpson (2006), Evaluation of CuO oxidation parameters for determining the source and stage of lignin degradation in soil,

Biogeochemistry, 80(2), 121–142.Payne, R. J., and E. A. Mitchell (2009), How many is enough? Determining optimal count totals for ecological and palaeoecological studies of

testate amoebae, J. Paleolimnol., 42(4), 483–495.Philben, M., and R. Benner (2013), Reactivity of hydroxyproline-rich glycoproteins and their potential as biochemical tracers of plant-derived

nitrogen, Org. Geochem., 57, 11–22, doi:10.1016/j.orggeochem.2013.01.003.Philben, M., K. Kaiser, and R. Benner (2014a), Biochemical evidence for minimal vegetation change in peatlands of the West Siberian Lowland

during the medieval climate anomaly and little ice age, J. Geophys. Res. Biogeosci., 119, 808–825, doi:10.1002/2013JG002396.Philben, M., K. Kaiser, and R. Benner (2014b), Does oxygen exposure time control the extent of organic matter decomposition in peatlands?,

J. Geophys. Res. Biogeosci., 119, 897–909, doi:10.1002/2013JG002573.Prescott, C. E. (2010), Litter decomposition: What controls it and how can we alter it to sequester more carbon in forest soils?,

Biogeochemistry, 101(1–3), 133–149, doi:10.1007/s10533-010-9439-0.Robinson, S. D., and T. R. Moore (2000), The influence of permafrost and fire upon carbon accumulation in high boreal peatlands, Northwest

Territories, Canada, Arct. Antarct. Alp. Res., 32(2), 155–166, doi:10.2307/1552447.Schellekens, J., P. Buurman, T. W. Kuyper, G. D. Abbott, X. Pontevedra-Pombal, and A. Martínez-Cortizas (2015), Influence of source vegetation

and redox conditions on lignin-based decomposition proxies in graminoid-dominated ombrotrophic peat (Penido Vello, NW Spain),Geoderma, 237–238, 270–282, doi:10.1016/j.geoderma.2014.09.012.

Schlesinger, W. H., and E. S. Bernhardt (2013), Biogeochemistry: An Analysis of Global Change, p. 443, Academic Press, San Diego, Calif.Schmidt, M. W. I., H. Knicker, and I. Kögel-Knabner (2000), Organic matter accumulating in Aeh and Bh horizons of a Podzol—Chemical

characterization in primary organo-mineral associations, Org. Geochem., 31(7–8), 727–734, doi:10.1016/S0146-6380(00)00045-0.Skoog, A., and R. Benner (1997), Aldoses in various size fractions of marine organic matter: Implications for carbon cycling, Limnol. Oceanogr.,

42(8), 1803–1813.Thormann, M., A. Szumigalski, and S. Bayley (1999), Aboveground peat and carbon accumulation potentials along a bog-fen-marsh wetland

gradient in southern boreal Alberta, Canada, Wetlands, 19(2), 305–317, doi:10.1007/bf03161761.Tolonen, K., and J. Turunen (1996), Accumulation rates of carbon in mires in Finland and implications for climate change, The Holocene, 6(2),

171–178, doi:10.1177/095968369600600204.Tremblay, L., and R. Benner (2006), Microbial contributions to N-immobilization and organic matter preservation in decaying plant detritus,

Geochim. Cosmochim. Acta, 70(1), 133–146, doi:10.1016/j.gca.2005.08.024.Turetsky, M. R., R. K. Wieder, D. H. Vitt, R. J. Evans, and K. D. Scott (2007), The disappearance of relict permafrost in boreal North America:

Effects on peatland carbon storage and fluxes, Global Change Biol., 13(9), 1922–1934, doi:10.1111/j.1365-2486.2007.01381.x.Van Bellen, S., M. Garneau, A. A. Ali, A. Lamarre, É. C. Robert, G. Magnan, H. Asnong, and S. Pratte (2013), Poor fen succession over

ombrotrophic peat related to late Holocene increased surface wetness in subarctic Quebec, Canada, J. Quat. Sci., 28(8), 748–760,doi:10.1002/jqs.2670.

van Breemen, N. (1995), How Sphagnum bogs down other plants, Trends Ecol. Evol., 10(7), 270–275.Viau, A. E., and K. Gajewski (2009), Reconstructing Millennial-Scale, Regional Paleoclimates of Boreal Canada during the Holocene, J. Clim.,

22(2), 316–330, doi:10.1175/2008JCLI2342.1.Watson, A., K. D. Stephen, D. B. Nedwell, and J. R. Arah (1997), Oxidation of methane in peat: Kinetics of CH4 and O2 removal and the role of

plant roots, Soil Biol. Biochem., 29(8), 1257–1267.Williams, C. J., and J. B. Yavitt (2003), Botanical composition of peat and degree of peat decomposition in three temperate peatlands,

Ecoscience, 10(1), 85–95.Woodland, W. A., D. J. Charman, and P. C. Sims (1998), Quantitative estimates of water tables and soil moisture in Holocene peatlands from

testate amoebae, Holocene, 8(3), 261–273.Yu, Z. C., J. Loisel, D. P. Brosseau, D. W. Beilman, and S. J. Hunt (2010), Global peatland dynamics since the Last Glacial Maximum, Geophys. Res.

Lett., 37, L13402, doi:10.1029/2010GL043584.Zaccone, C., D. Said-Pullicino, G. Gigliotti, and T. Miano (2008), Diagenetic trends in the phenolic constituents of Sphagnum-dominated peat

and its corresponding humic acid fraction, Org. Geochem., 39(7), 830–838.

Global Biogeochemical Cycles 10.1002/2014GB004989

PHILBEN ET AL. CONTROLS ON PEAT DECOMPOSITION 743