j ourna l homepage wwwe lsev ie r com locate geoderma
Conifer litter identity regulates anaerobic microbial activity in wetlandsoils via variation in leaf litter chemical composition
Joseph B Yavitt a Christopher J Williams ba Department of Natural Resources Cornell University Ithaca NY 14853 USAb Department of Earth and Environment Franklin amp Marshall College Lancaster PA 17603 USA
httpdxdoiorg101016jgeoderma2014120230016-7061copy 2015 Elsevier BV All rights reserved
a b s t r a c t
a r t i c l e i n f o
Article historyReceived 4 August 2014Received in revised form 23 December 2014Accepted 27 December 2014Available online xxxx
KeywordsAnaerobic respirationConiferHemicelluloseMethanogenesisPeat soilPectinPlant species identity
Plant litter is a crucial source of energy and nutrients for soilmicroorganisms and thus plant species identitymayhelp account for variation in rates of soil microbial activity via species dependent differences in litter chemistryThis may be particularly important in forested wetlands where the composition of conifer tree species can varyWeexplored the relationship between conifer litter identity and anaerobicmicrobial activity by adding fresh nee-dle litter fromnine conifer species to soils from two contrasting forestedwetlands and quantifying rates of anaer-obic respiration (CO2 production) and methanogenesis (CH4 production) We characterized litter chemistry asthe amounts of structural polysaccharides and polymers (cellulose hemicellulose pectin lignin) and examinedchanges in these fractions during incubation Litter from Pinus species supported the largest rates of CH4 produc-tion whereas litter from Picea species supported the lowest rates Litter from two deciduous conifers(Metasequoia Larix) supported intermediate rates of CH4 production Tree species identity had less impact onrates of anaerobic CO2 production Rates of CH4 production correlated with hemicellulose and with covalentlybonded pectin whereas rates of anaerobic CO2 production showed a positive correlation with initial acid-detergent lignin concentration and with pectin Gaining a better understanding of how hemicellulose and pectinin decomposing plant litter promote anaerobic microbial activity especially specialized methanogenesis in soilprovides insight that links plant litter identity and leaf traits to anaerobic microbial activity in the underlying soil
copy 2015 Elsevier BV All rights reserved
1 Introduction
Woody vegetation can be a dominant feature in some peatland eco-systems (Rydin and Jeglum 2013) and interest in the functional role oftrees in peatlands has increased in concert with the recent expansion ofPinus Picea and Larix onto cool temperate peatlands in North Americaand in Europe (Berg et al 2009 Heijmans et al 2013 Pellerin andLavoie 2003) Because trees increase the capture of atmospheric carbonthey can in turn promote increased rates of heterotrophicmicrobial ac-tivity in soil via an added amount of litter to the soil (Jackson et al2002) In forested peatlands leaf litter is the largest source of fresh car-bon added to the soil (Coles and Yavitt 2004) and thus understandinglinkages between tree leaf litter and soilmicrobial activity is necessary ifwe are to fully explain carbon dynamics in peatlands This is particularlygermane to the production of methane (CH4) and carbon dioxide (CO2)derived ultimately from leaf litter decay and thus consideration of treespecies identity might help predict emission of these atmospheric gas-ses in response to current and future changes in tree species composi-tion on peatlands
illiamsfandmedu
This follows the growing body of research that links plant speciesaboveground to soil microorganisms belowground (Zak et al 2003)The rationale is that plant leaves show a suite of chemical characteristicsthat might confer distinctiveness to soil microorganisms For exampleplant leaves can be viewed along a spectrum from long-lived evergreenleaves that are thick with low nutrient content and high amounts ofstructural compounds to seasonally deciduous leaves that are thinnerwith higher nutrient content and fewer structural compounds (Wrightet al 2004) This spectrum has been shown to influence the rate of leafdecomposition eg slower rates for evergreen leaves and faster ratesfor deciduous leaves (Cornwell et al 2008) Since microbial activity gen-erally corresponds to the rate of litter decomposition variation in litterchemistry should provide a crucial control on microbial activity Further-more while the leaf trait spectrum is familiar for angiosperms the ques-tion whether it extends to conifers from evergreen Picea and Pinus todeciduous Larix and Metasequoia in peatlands is unclear Also undersome circumstances deciduous conifers may produce more leaf litterthan evergreen conifers (Middleton and McKee 2004 Williams et al2003)
Most studies of litter decomposition are prolonged taking the firstmeasurement of mass loss only after several months to one year andfollowing the residue for three or more years As a result we knowmuch less aboutmicrobial colonization of fresh litter at the very earliest
142 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
stage of decomposition (Koide et al 2005 Moorhead and Sinsabaugh2006) The situation is especially acute in wetland soils where commonwisdom suggests slow rates of litter decay in particular with little oxy-gen and via anaerobic metabolism (Yavitt et al 1997)
Here we propose that the mechanistic link to soil microbial activityoccurs in the suite of chemical compounds that make up plant cellwalls Plant cell walls are combinations of three dominant structuralpolysaccharides (cellulose hemicellulose and pectin) the lignin poly-mer and various proteins and a variety of lowmolecular weight solublecompounds (Berg and McClaugherty 2003) A general model is thatcellulosemicrofibrils are embedded in amatrix of hemicellulose pectinand proteins (Keegstra 2010) In secondary cell walls lignin surroundsthe cellulose and is bound to the hemicellulose (Cosgrove 2005) There-fore the traditional litter decay model posits that initial decompositionof leaf litter consumes the soluble fraction and non-protected cellulose(Berg and Staaf 1980) leaving other components to accumulate in theresidue However recent research has shown that initial decompositioncan include the hemicellulose (Delaney et al 1996) pectin (Jenkins andSuberkropp 1995) and lignin (Klotzbuumlcher et al 2011) Moreover dif-ferent components of plant cell walls are known to vary in how accessi-ble and digestible they are to different microbial decomposers (Dinget al 2012) Thus variation in the relative amounts of these plant cellwall components among different species of conifers may play a rolein regulating the early stages of litter decay and the supply of substratesto the soil microbial community (Koide et al 2005)
We examined how tree species identitymight ultimately regulatemi-crobial activity in soil We used a common garden approach and addedneedle litter fromnine conifer species to soils from two forestedwetlandswith contrasting soil types We focused specifically on anaerobic produc-tion of CO2 and CH4 (methanogenesis) We expected a priori that CO2
production would exhibit relatively weak relationships with leaf litteridentity The reason being that CO2 comes from a plethora of decarboxyl-ation reactions fermentations and anaerobic respirations such that thenet effect on CO2 would be muted However for methanogenesis weexpected greater fidelity with leaf litter identity given that methanogensuse only a limited number of substrates derived specifically from fermen-tation and that fermentation specifically would depend on the suite offermentable material per plant species (Drake et al 2009 Reith et al2002) This follows the finding from Bremer et al (2007) that plantspecies identity influenced themicroorganisms that carried out soil deni-trification Therefore specific predictions were for greater rates of micro-bial activity with deciduous rather than long-lived foliage and thatvariation in the plant cell wall composition of litter among specieswould in part explain variation in rates of anaerobic microbial activity
2 Materials and methods
Soils were collected from two sites in central New York State Aver-age annual temperature for the region is 65 degC and average annual pre-cipitation is 1170 mm June and July typically have the highest meanmonthly precipitation
One site Labrador Hollow (local name hereafter peat soil) is a for-ested peatland located 8 km southeast of Tully New York (LabradorHollow Unique Area 42781932degN 76041628degW) The site has aclosed-canopy forest composed of Acer rubrum L (red maple) and sev-eral conifer species Pinus strobus L (eastern white pine) Tsugacanadensis (L) Carriegravere (eastern hemlock) and Larix laricina (Du Roi)K Koch (tamarack or larch) Vaccinium corymbosum L (highbush blue-berry) Toxicodendron radicans (L) Kuntze (poison ivy) and a mixtureof pteridophytes are common in the understory Sphagnum mossescover about 75 of the ground and include the following species Sphag-num squarrosum Crome Sphagnum girgensohnii Russow Sphagnumfimbriatum Wilson Sphagnum henryense Warnst and Sphagnumrussowii Warnst Beneath Sphagnum is a 0 to 10 cm thick layer ofdense roots overlying about 8 m of mesic Typic Medisaprists peat soil
Below that glacial till and sand cover Devonian age shale and limestone(Fisher et al 1970) Further details are given in Coles and Yavitt (2004)
The second site Sapsucker Woods (hereafter swamp soil) is a for-ested wetland in Ithaca New York (42477291degN 76451523degW) locat-ed 47 km to the southeast of the first site Sapsucker is a closed canopyforest dominated byA rubrum Alnus incana (L)Moench ssp rugosa (DuRoi) RT Clausen (Grey alder) and T canadensis Trees occur on elevatedhummocks surrounded by water-filled depressions The soil (Alden se-ries) is a mesic Mollic Endoaquept very poorly drained in depressionsand low areas It is approximately 1 m deep with a fragipan soil horizonthat prevents drainage andmaintains a highwater table throughout theyear Soil for the incubations was collected from the depressionsbetween tree hummocks
Needle litter was collected from trees growing within close proxim-ity (ca 1 km) of each other on the same soil and exposed to identicalclimate in the Cornell Plantations Natural Area We collected needlesfrom three deciduous conifers L laricina Metasequoia glyptostroboidesHu and Cheng (Dawn redwood) and Taxodium distichum (L) Rich(Bald cypress) and from six evergreen conifers (Pinus banksianaLamb (Jack pine) Pinus rigida Mill (Pitch pine) P strobus (Easternwhite pine) Picea abies (L) Karst (Norway spruce) Picea glauca(Moench) Voss (White spruce) and Piceamariana (Mill) Britton Sternsand Poggenb (Black spruce)) We collected needles at the end of thegrowing season in October by shaking branches gently indicating awell-formed abscission zone and collecting the needles that landed ina plastic bag placed beneath the branches Needles were air dried atlowhumidity Sub-sampleswere oven dried at 105 degC to ascertainmois-ture content Other portions were sampled for chemical analyses (de-tailed below) We constructed litterbags containing a total of 1 g dry-weight-equivalent litter using a 15-cm2 piece of fine mesh material(0675 mmmesh size) gathered and sealed with plastic thread
We quantified plant litter effects on rates of soil CH4 production andanaerobic CO2 production using in vitro incubation studies Five repli-cate soil samples were collected randomly combined and thoroughlymixed by hand before 30-g portions of soil were placed into individualincubation jars (N = 3 per leaf litter type) on the date of collectionWe added 20 mL of de-ionized water before we sealed the jar with alid that had a gas-tight septum to allow sampling gas in the headspaceof the sealed jar The jar headspace was made anoxic by evacuating forseveral minutes using a vacuum pump and refilling with O2-free N2The evacuation was repeated three times Jars with soil were incubatedfor 15 days at 20 degC and concentrations of CO2 and CH4 in the headspacewere measured every third day in order to establish the in vitro basalrate of microbial CO2 and CH4 production The pre-incubation periodalso reduced the amount of inherent substrate for soil microorganisms(see Wang et al 2003) thereby increasing the impact of the addedleaf litter on their activity
Sampleswere analyzed bygas chromatographyusing aflame ioniza-tion detector for CH4 and a thermal conductivity detector for CO2 Thegas chromatograph has a 275 m by 318 mm column of Poropak Q 80100 mesh (Waters Chromatography Milford MA) maintained at 50 degCto separate the gasses The flow rate of the He carrier gas was30 mL minminus1 The injector temperature was 110 degC
Following the pre-incubation period we opened the jars and addeda litterbag to the surface of the soil We tapped the bag to insure contactwith soil before re-sealing the jar and establishing anoxic conditionsand incubating at 20 degC as described above We measured CH4 andCO2 in the headspace every three to four days for 24 days At the endof the incubation period the litterbag was removed and the residue re-maining was analyzed for its biochemical content
We determined the composition of hemicellulose cellulose and lig-nin using the detergent fiber technique (Van Soest 1994) First wecombined and thoroughly mixed material per species before beingground through a Wiley Mill Duplicate subsamples (05 g) per specieswere subjected to sequential neutral detergent fiber (Van Soest et al1991) acid detergent fiber and acid detergent lignin digestions
143JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
(Method 97318 AOAC 2012) using an ANKOM fiber analyzer (ANKOMTechnology Macedon New York USA) The neutral detergent fiber di-gestion dissolves soluble cell wall components resulting in a residuecomposed of hemicellulose cellulose and lignin The acid detergentfiber digestion dissolves hemicellulose (hereafter referred to as deter-gent hemicellulose) leaving a residue of cellulose and lignin The gravi-metric difference between the neutral detergent fiber and the aciddetergent fiber residue provides an estimate of detergent hemicelluloseconcentration Treatment of the acid detergent fiber residue with 72ww sulfuric acid dissolves cellulose leaving a residue of acid-detergent lignin Samples were ash-corrected after the final acid diges-tion to account for any contamination of mineral soil in the acid-detergent lignin fraction
We also used a non-detergent sequential extraction procedure(McLeod et al 2007) to characterize the composition of the plant cellwall of litter before and again after incubation Briefly litter tissue wasextracted as seven fractions with (1) 10 formic acid to remove looselybound polysaccharides with (2) a phosphate buffer to dissolve and re-move loosely boundwater soluble pectinwith (3) cyclohexanediaminetetra acetic acid (CDTA) to remove calcium which releases bound pec-tin in the middle lamellae with (4) urea to cleave hydrogen bondsand separate pectin bonded covalently to cellulose as well as cell wallproteins with (5) sodium carbonate to remove any remaining pectinin the primary cell wall with (6) sodium hydroxide to dissolve
Fig 1 Mean production rates of CH4 (top panel) and CO2 (bottom panel) for two wetland soi24 days following the addition of different types of conifer litter Control soils were incubated
hemicelluloses (hereafter referred to as extractable hemicellulose)and finally with (7) formic acid to remove all remaining non-structural sugars
We performed analysis of variance (ANOVA) for CH4 production andfor CO2 production We examined the effect of soil type conifer (decid-uous pine spruce) and species Species was nested within coniferPearson correlation coefficients were calculated to evaluate the rela-tionship between CH4 and CO2 production rates and litter chemistry
3 Results
Rates of CH4 production in control soil without added litter were b-1 nmol gminus1 dayminus1 (Fig 1) Likewise control soil without added litter(Fig 1) had rates of CO2 production b 1 μmol gminus1 dayminus1 It is very likelythat slow rates of CH4 production and CO2 production in the controlsoils without added litter resulted from the pre-incubation periodbeing long enough to deplete substrates for resident microbial popula-tions Therefore we assume that production rates for CH4 and CO2 insoils with added litter are indicative of conifer litter identity and thesubstrates supplied by each species that supported anaerobic microbialactivity
Over the course of the 24-day incubation rates of CH4 productionwere much greater in soils with added litter (Fig 1) although rateswith added Taxodium litter were significantly less than rates with
ls (Labrador Hollow peat soil Sapsucker Woods swamp soil) incubated anaerobically forwithout litter addition Error bars are +minus one standard error of the mean
Table 2Plant litter type and pre-incubation concentrations of extractable cell wall fractions (car-bohydrates) expressed in g kgminus1 dry tissue
a Fractions are soluble carbohydrates (Fraction1)water soluble pectin (Fraction2) pectinin themiddle lamella (Fraction 3) pectin bound covalently to cellulose (Fraction 4) cell wallpectin (Fraction 5) extractable hemicellulose (Fraction 6) and residual non-cellulose carbo-hydrates (Fraction 7)
144 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
other litters The impact of litter addition on CH4was significantly largerfor the peat soil from Labrador Hollow (mean= 124 nmol gminus1 dayminus1)than for the swamp soil from Sapsucker Woods (49 nmol gminus1 dayminus1)(F1 54 = 2543 P = 00023) The significant effect of conifer type(F6 54 = 491 P = 00004) was evident for both soils Litter from Pinussupported the largest rates of CH4 production 187 nmol gminus1 dayminus1
when added to peat soil and 63 nmol gminus1 dayminus1 when added toswamp soil Picea litter supported the lowest CH4 production rates78 nmol gminus1 dayminus1 with peat soil 39 nmol gminus1 dayminus1 with swamp soilHence the deciduous conifers (Metasequoia Larix Taxodium) hadintermediate average rates of CH4 production 104 nmol gminus1 dayminus1
with peat soil 39 nmol gminus1 dayminus1 with swamp soil However theseaverage rates increase to 156 nmol gminus1 dayminus1 with peat soil and70 nmol gminus1 dayminus1 with swamp soil excluding Taxodium
Rates of anaerobic CO2 production were greater with litter addi-tions as opposed to control soils (Fig 1) Rates of anaerobic CO2 pro-duction were significantly greater with litter added to the peat soil(32 μmol gminus1 dayminus1) than the swamp soil (13 μmol gminus1 dayminus1)(F1 54 = 11580 P b 00001) The significant effect of conifer type(F6 54 = 263 P = 00259) was evident for both soils although dif-ferences among conifer types were larger for the peat soil Whenadded to the peat soil litter from the deciduous conifers supportedlarger rates of CO2 production (mean = 36 μmol gminus1 dayminus1) thanlitter from Pinus (mean = 33 μmol gminus1 dayminus1) or Picea (mean =31 μmol gminus1 dayminus1) Taxodium litter supported the largest ratesfor both soil types
The deciduous conifer litter had much less detergent hemicelluloseand cellulose than that in the evergreen conifers (Table 1) Thereforethe deciduous conifers had a much larger proportion of pectins andstructural proteins On average Pinus litter contained about 10 morecellulose and 14 more acid-detergent lignin than Picea litter
The sumof extractable cell wall components ranged from34 to 61of litter mass (Table 2) The largest fractions were loosely bound poly-saccharides (Fraction 1) pectin covalently bonded to cellulose (Fraction4) and extractable hemicellulose (Fraction 6) There was some varia-tion in values among members of a group of conifers For instanceTaxodium hadmore loosely bound polysaccharides and pectin covalent-ly bonded to cellulose but less extractable hemicellulose compared toMetasequoia and Larix P banksiana contained less extractable hemicel-lulose compared to the other two species of Pinus Among the sprucesP glauca was distinctive with more loosely bound polysaccharideswhereas P abies had less pectin covalently bonded to cellulose
There were considerable changes in the different fractions of cell wallcomponents during incubation (Fig 2) Loosely bound polysaccharides(Fraction 1) and pectin covalently bonded to cellulose (Fraction4) showed the largest declines The large increase inwater-soluble pectin(Fraction 2) was particularly noticeable In four of the litters (TaxodiumP banksiana P glauca and P abies) there was a large increase in the ex-tractable hemicellulose fraction (Fraction 6) during incubation Therealsowas a soil type effect on the changes in extractable carbohydrate frac-tions For example quantities of loosely bound polysaccharides (Fraction
Table 1Plant litter type and pre-incubation concentrations of detergentfiber fractions (g kgminus1 ashfree dry weight)
Litter Type Hemicellulose Cellulose Lignin Residuala
a Residual includes soluble cell-wall components including β-glucans and pectins as wellas soluble simple sugars amino acids water-soluble phenolics and insoluble protein
1) and pectin covalently bonded to cellulose exhibited greater declines inthe peat soil than in the swamp soil whereas the observed increase inwater soluble pectin (Fraction 2) also was greater in peat than in theswamp soil
We compared rates of soil CH4 production and CO2 production withinitial litter concentrations of detergent hemicellulose cellulose andacid-detergent lignin for all litter types (Table 3) We also comparedchanges in the extractable cell wall fractions (1 to 7) that occurred inall litter types during the incubations with soil CH4 production andCO2 production (Table 3) Therewas a relatively strong positive correla-tion (r ge 063 P b 01) between rates of CH4 production in both soiltypes and the initial detergent hemicellulose content of the litter Thenegative correlation (r=minus071 P b 005) between rates of CH4 produc-tion in the peat soil and pectin covalently bonded to cellulose meansthat the post-incubation increase in pectin covalently bonded to cellu-lose reduced CH4 production CO2 production in both soils was positive-ly correlated with the initial acid-detergent lignin content (rge 072 P b
005) Post incubation water-soluble pectin was positively correlated(r = 067 P b 005) with CO2 in the swamp soil
4 Discussion
Our comparative study examined several relationships between leaflitter fromnine different conifer taxa and anaerobicmicrobial activity intwo wetland soils Although our analysis is limited to correlation thedata provide new insights into plantndashsoil relationships For exampleseasonally deciduous leaves are thought to decay faster than evergreenleaves according to the Leaf Economics Spectrum (Cornwell et al2008) This is generally attributed to the seasonally deciduous leaveshaving low leaf mass per area and lower CN ratios compared to long-lived evergreen leaves However our finding of no significant differencein CO2 production for deciduous versus evergreen conifers suggests thatamore complex linkage exists between leaf litter traits and concomitantmicrobial trace gasproductionOur chemical analyses indicate 1) pectinand hemicelluloses vary among leaf litter types and their concentra-tions can change substantially in the early stages of leaf litter decompo-sition and 2) changes in pectin and hemicelluloses during leaf litterdecomposition might be fueling anaerobic microbial activity Thereforethe specific chemical attributes of the litter influence the anaerobic mi-crobial response to litter inputs
41 General soil microbial response to litter addition
Our data agree with findings of Reith et al (2002) that fresh Picealitter could promote anaerobic metabolism by soil microorganismsOverall the rates of anaerobic CO2 production in this study were typicalof rates found in other wetland soils (Yavitt et al 1997) Methane pro-duction and anaerobic CO2 productionwere greater in the peat soil than
Fig 2 Post-incubation versus pre-incubation comparison of sequentially extracted cell wall fractions (carbohydrates) of nine conifer litter types following a 24-day incubation in peat soilor swamp soil Degree of color value indicates themagnitude of changewhereas color indicates increases (green) or decreases (red) in a fraction following incubation The absolute changein concentrations of each fraction is listed within cells as grams carbohydrate per kilogram dry tissue (For interpretation of the references to color in this figure legend the reader is re-ferred to the web version of this article)
145JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
in the swamp soil This supports the general notion that rates of litterdecay are slowed by anaerobic metabolism in peat soils compared toperhaps greater aerobic metabolism in swamp soils (Moore et al2005) It is also likely that peat soils harbored a larger more active
Table 3Pearson correlation matrix for pre-incubation litter detergent fiber (cellulose hemicellu-lose lignin) concentrations and post incubation litter extractable cell wall fractions (1ndash7) versus trace gas production for each soil type
P b 001 001 b P b 005 005 b P b 01a Fractions are detergent fiber (cellulose hemicellulose lignin) and sequentially extracted
soluble carbohydrates (Fraction 1) water soluble pectin (Fraction 2) pectin in the middlelamella (Fraction 3) pectin bound covalently to cellulose (Fraction 4) cell wall pectin(Fraction 5) extractable hemicellulose (Fraction 6) and residual non-cellulose carbohy-drates (Fraction 7)
community of microorganisms capable of anaerobic metabolism Thewater-saturated peat provides a superior microbial habitat for anaero-bic activity and methanogenesis On the other hand the swamp soil isvariably saturated it ranges from being waterlogged during springsnowmelt to having a thin unsaturated surface layer during summerdry periods We sampled in the fall when a thin unsaturated layer wasapparent at the soil surface Thus our measured rates of anaerobic CO2
production in the swamp soil are consistent with the notion that facul-tative anaerobes can constitute about 10 of the aerobic microbial pop-ulation even in well-drained soil (Tiedje et al 1984) Our study alsoconfirms that the seasonally saturated conditions provide a tolerablehabitat for the persistence of methanogenic microorganisms whichare adapted to tolerate variable redox conditions (DeAngelis et al2010)
42 Leaf litter traits
In the context of the leaf economic spectrum it is logical to assumethat litter fromdeciduous coniferswould support greater rates ofmicro-bial activity than litters from evergreen Pinus or Picea This assumptionis based in part on the notion that seasonally deciduous leaves havebetter litter quality for decomposition However the prediction wasnot upheld and Pinus litter supported the greatest rates of CH4 produc-tion The microbial response to litter addition in our study reveals theuse of a simple binomial leaf trait characterization (evergreen vs decid-uous) as a predictor of soil microbial response to be lacking This findingcorroborates those of Moore et al (2005) that deciduous Larix needles
146 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
placed on the surface of peat soils decomposed slower than Pinus andPicea needles in sites in Canada
Why does the species identity of litter matter For one Taxodiumproduced themost unusual patternwith very slow rates of CH4 produc-tion but the greatest rate of CO2 production Other studies have foundthat Taxodium needles will decompose rapidly in wetlands especiallywhen exposed to alternating oxic and anoxic conditions (Battle andGolladay 2001) and thus the large CO2 production is not surprisingAlthoughwetlands dominated by Taxodium are known sources of atmo-spheric CH4 (Vann andMegonigal 2003) it is likely that fresh Taxodiumneedles have biochemical traits that slow the rate of CH4 production butnot CO2 production at the earliest stage of anaerobic decomposition Forexample Taxodium is known to produce taxoquinone and several otherditerpenoids that may have an inhibitory effect on microbes (Cervanteset al 2000 Zaghloul et al 2008) Perhaps it is not until such com-pounds degrade that Taxodium litter can be turned into methanogenicsubstrate
We also found that soilmicrobial response varied at the generic levelwith Pinus generally promoting greater CH4 than Picea litter Our findingthat P banksiana (Jack pine) litter supported CH4 production is particu-larly interesting because this species has been encroaching on peatlandsinNorth America recently (Pellerin and Lavoie 2003) Thus P banksianaencroachment may promote enhanced soil CH4 production and poten-tial emissions of atmospheric CH4 Our speculation that encroachmentof woody vegetation onto otherwise open peatlands may result inenhanced CH4 production awaits experimental confirmation Admitted-ly other environmental factors that regulate CH4 productionmay trumpthe influence of litter identity Nevertheless our results and a scarcity ofinformation in the literature suggest further study may be warranted
43 Leaf litter carbohydrate chemistry
Our analyses of conifer leaf litter quality in terms of different deter-gent fiber fractions and cell wall components in fresh litter were de-signed to determine how constituents other than cellulose and ligninmight affect microbial colonization and activity Most previous researchhas focused on cellulose because cellulose is often assumed to be themost abundant carbohydrate fraction by mass (Berg and Staaf 1980)In contrast the roles of hemicellulose and pectin in microbial respira-tion have not been studied extensively However pectin and hemicellu-lose act as the glue that holds cellulose fibers and lignin together in theplant cell wall (Cosgrove 2001 2005) Thus microbial decomposersattack compounds sequentially pectin and hemicellulose before cellu-lose then lignin Hemicellulose is a complex polymer with differentsugars and ratios of the different sugars can vary as a function of plantspecies and site fertility (Strakovaacute et al 2010) For instance Berg andMcClaugherty (2003) reported that arabinose and galactose in thehemicellulose fraction of Pinus needle litter decomposed immediatelyafter litterfall in pine forests whereas loss of mannose and xylose oc-curred after one year
Although detergent fiber analysis of plant litter has been applied inmany ecological studies of litter quality (ie results in Table 1) the se-quential extraction of cell wall components (ie results in Table 2) hasbeen used mainly in plant physiology studies (Fry 1989) mdash althoughseeMcLeod et al (2007)We quantified hemicellulose by bothmethodsand this gave conflicting results the sequential extraction method gavemuch larger values for the deciduous conifers whereas values weresimilar using both methods for the Pinus and Picea Indeed Theanderand Westerlund (1993) proposed alternatives to detergent hemicellu-lose arguing that the procedure is non-specific and might includesome pectin Jung and Lamb (2004) found that the detergent fibermethod overestimated the amount of hemicellulose in alfalfa leavesMuch more comparative work using both methods is warranted fornon-forage plants given that sequential extraction of cell wall compo-nents has the potential to providemore information about litter qualityin particular about pectin
Overall Fraction 2 (weakly bound water soluble pectin) was largerin leaf litter post-incubation than in fresh litter (Fig 2) The largeincrease in Fraction 2 across all litter types during the incubation wasunexpected as it could not be production of pectin per se Rather it like-ly represents the liberation of weakly bound sugars following incuba-tion that are then extracted in this fraction Messenger et al (2012)also found that amounts of extractable sugars increased during decom-position which they attributed to decomposition of cellulose proteinsand low-molecular-weight metabolites that are sugar-rich The anaero-bic metabolism in our study suggests another mechanism the sugarsmight be related to fermentation Low-molecular-weight organicacids derived from fermentation when in excess supply can lead togluconeogenesis (essentially reverse glycolysis) and the production ofcarbohydrates (Dijkstra et al 2011) The increase in extractable hemi-cellulose content of some of the litters (Fraction 6) also was surprisingPresumably this represents sugars that coalesce into a sodiumhydroxide-extractable fraction and is not decomposer production ofhemicellulose per se (Hoch 2007)
Fraction 4 (pectin covalently bound to cellulose) in generaldecreased in leaf litters during incubation This reflects the fracturingof plant cell walls during the earliest stage of decomposition Unfortu-nately the mechanism is not clear from the data we obtained Presum-ably fungal hyphae are involved ie filaments penetrate the plant cellwall and degrade it physically as well as chemically Clearly morework is needed to determine the fate of the liberated pectin ie is itmade soluble or metabolized to CO2 andor CH4
44 Changes in carbohydrate fractions versus microbial activity
Inmost cases positive correlations between components in leaf litterand microbial activity are interpreted as that component fueling activi-ty Likewise a negative relationship is viewed as either a recalcitrantcomponent or a component that is inhibitory tomicrobial activity How-ever our data are more complex as in some cases the negative correla-tion (low quantities of a carbohydrate fraction versus a rapid rate ofmicrobial activity) could be explained as rapid microbial metabolismcausing the loss (see below)
For Fraction 4 the straightforward positive correlationwith CH4 pro-duction suggests that slow fermentation of cell wall pectin provides aslow release of methanogenic substrate Previous studies demonstratethat the aerobic decay of pectin is linked to CH4 production (McLeodet al 2008) and thus our data extend this relationship to anaerobicme-tabolism Likewise pectin fermentation is known to fuel methanogensin anaerobic lake sediments (Schink and Zeikus 1982)
In contrast hemicellulose showed a positive correlation with CH4
production based on the detergent analysis but a negative correlationwith extractable Fraction 6 This is clearly troubling however in the se-quential extraction the earlier urea-extraction (Fraction 4) also removesstructural proteins (Fry 1989) so the negative correlation may reflectthe influence of protein and enzymes on CH4 production The sugar con-tent of the urea-extractable fraction in terms of arabinose galactosemannose xylose and others in each species is not known (cf Renardet al 1991) yet thismay also influence rates of CH4 production It is pos-sible that variation in the composition of monomers liberated fromhemicellulose may determine the supply of fermentation products tomethanogens
Rates of anaerobic CO2 production had positive relationships withacid-detergent lignin and with loosely bound water-soluble pectinBecause only a subset of microorganisms can degrade lignin completely(Osono 2007 Talbot and Treseder 2011) there is usually a negative re-lationship (Aerts 1997) especially in long-termdecomposition studiesFurthermore most of the research has been done with aerobic notanaerobic decay In contrast our correlation was with the initial acid-detergent lignin composition and we do not know how acid-detergent lignin quantity and quality changed during the incubationIt is also possible that the CO2 was not derived from lignin as such but
147JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
rather from hemicelluloses that occur intimately with acid-detergentlignin (Webster et al 2005) In contrast in soil incubated without O2pectin is a very good substrate for fermentation (Chin et al 1998) Itis likely that the relationship between pectin and anaerobic CO2 produc-tion represents the CO2 released from hydrolysis and fermentation aspectin is altered In other words pectin might be the first compoundthat microbial decomposers encounter and thus it modulates microbialactivity at the onset of decomposition asmeasured in the present study
5 Conclusions
About 6Mg haminus1 of leaves and needles fall to the ground each year inforestedwetlands (Reich et al 1997) andunder some circumstances de-ciduous conifers may produce much higher amounts (Middleton andMcKee 2004Williams et al 2003) This litter is a crucial source of energyfor soilmicroorganisms (Coles andYavitt 2004)Most studies of litter de-composition are prolonged making the first measurement of mass lossafter a few months and following the residue for one or more yearsThus we know much less about microbial colonization of fresh litter atthe very earliest stage of decomposition (Moorhead and Sinsabaugh2006) The situation is especially acute in forested wetlands where com-mon wisdom suggests slow rates of litter decay We suggest that finescale details of the decay process have been missed Hemicelluloses andpectin are probably the first molecular compounds to be released fromplant cell walls as they intertwine with lignin and cellulose and preventspecialized enzymes such as cellulases and lignases from working As aconsequence a moderate pulse of CH4 production should occur as freshlitter decays Moreover the encroachment of some types of conifersinto northern peatlands which reflects a shift in the dominant wetlandplant functional type may promote a new linkage between abovegroundcarbon fixation and belowground carbon mineralization
Acknowledgments
We thankTracy Bartella for assistance in thefield and laboratoryWethank three anonymous reviewers for constructive comments on themanuscript
References
Aerts R 1997 Climate leaf litter chemistry and leaf litter decomposition in terrestrialecosystems a triangular relationship Oikos 439ndash449
AOAC 2012 Official Methods of Analysis 19th ed Association of Official Analytical Chem-ists Gaithersburg MD
Battle JM Golladay SW 2001 Hydroperiod influence on breakdown of leaf litter incypress-gum wetlands Am Midl Nat 146 128ndash145
Berg B McClaugherty C 2003 Plant Litter SpringerBerg B Staaf H 1980 Decomposition rate and chemical changes of Scots pine needle lit-
ter II Influence of chemical composition Ecol Bull 373ndash390Berg EE Hillman KM Dial R DeRuwe A 2009 Recent woody invasion of wetlands on
the Kenai Peninsula Lowlands south-central Alaska a major regime shift after18 000 years of wet Sphagnumndashsedge peat recruitment Can J For Res 392033ndash2046 httpdxdoiorg101139X09-121
Bremer C Braker G Matthies D Reuter A Engels C Conrad R 2007 Impact of plantfunctional group plant species and sampling time on the composition of nirK-typedenitrifier communities in soil Appl Environ Microbiol 73 6876ndash6884
Cervantes FJ Velde S Lettinga G Field JA 2000 Competition between methanogenesisand quinone respiration for ecologically important substrates in anaerobic consortiaFEMS Microbiol Ecol 34 161ndash171
Chin K-J Rainey FA Janssen PH Conrad R 1998 Methanogenic degradation of poly-saccharides and the characterization of polysaccharolytic clostridia from anoxic ricefield soil Syst Appl Microbiol 21 185ndash200
Coles JRP Yavitt JB 2004 Linking belowground carbon allocation to anaerobic CH4
and CO2 production in a forested peatland New York State Geomicrobiol J 21445ndash455 httpdxdoiorg10108001490450490505419
Cornwell WK Cornelissen JHC Amatangelo K Dorrepaal E Eviner VT Godoy OHobbie SE Hoorens B Kurokawa H Peacuterez-Harguindeguy N Quested HMSantiago LS Wardle DA Wright IJ Aerts R Allison SD Van Bodegom PBrovkin V Chatain A Callaghan TV Diacuteaz S Garnier E Gurvich DE Kazakou EKlein JA Read J Reich PB Soudzilovskaia NA Vaieretti MV Westoby M 2008Plant species traits are the predominant control on litter decomposition rates within bi-omes worldwide Ecol Lett 11 1065ndash1071 httpdxdoiorg101111j1461-0248200801219x
Cosgrove DJ 2001 Wall structure and wall loosening A look backwards and forwardsPlant Physiol 125 131ndash134
Cosgrove DJ 2005 Growth of the plant cell wall Nat Rev Mol Cell Biol 6 850ndash861httpdxdoiorg101038nrm1746
DeAngelis KM Silver WL Thompson AW Firestone MK 2010 Microbial communi-ties acclimate to recurring changes in soil redox potential status Environ Microbiol12 3137ndash3149
Delaney MT Fernandez IJ Simmons JA Briggs RD 1996 Red Maple and White Pinelitter quality initial changes with decomposition Technical Bulletin No 162 Techni-cal Bulletin University of Maine Orono ME
Dijkstra P Blankinship JC Selmants PC Hart SC Koch GW Schwartz E HungateBA 2011 Probing carbon flux patterns through soil microbial metabolic networksusing parallel position-specific tracer labeling Soil Biol Biochem 43 126ndash132
Ding SY Liu YS Zeng Y Himmel ME Baker JO Bayer EA 2012 How does plantcell wall nanoscale architecture correlate with enzymatic digestibility Science 3381055ndash1060
Drake HL Horn MA Wuumlst PK 2009 Intermediary ecosystem metabolism as a maindriver of methanogenesis in acidic wetland soil Environ Microbiol Rep 1 307ndash318
Fisher D Isachsen Y Rickard L 1970 Geologic map of New York State consisting of 5sheets Niagara Finger Lakes Hudson-Mohawk Adirondack and Lower Hudson Mapand Chart Series No 15
Fry S 1989 Analysis of cross-links in the growing cell walls of higher plants Plant FibersSpringer pp 12ndash36
Heijmans MM Knaap YA Holmgren M Limpens J 2013 Persistent versus transienttree encroachment of temperate peat bogs effects of climate warming and droughtevents Glob Chang Biol 19 2240ndash2250
Hoch G 2007 Cell wall hemicelluloses as mobile carbon stores in non-reproductiveplant tissues Funct Ecol 21 823ndash834
Jackson RB Banner JL Jobbaacutegy EG PockmanWTWall DH 2002 Ecosystem carbonloss with woody plant invasion of grasslands Nature 418 623ndash626 httpdxdoiorg101038nature00910
Jenkins CC Suberkropp K 1995 The influence of water chemistry on the enzymaticdegradation of leaves in streams Freshw Biol 33 245ndash253 httpdxdoiorg101111j1365-24271995tb01165x
Jung HJG Lamb JFS 2004 Prediction of cell wall polysaccharide and lignin concentra-tions of alfalfa stems from detergent fiber analysis Biomass Bioenergy 27 365ndash373
Klotzbuumlcher T Kaiser K Guggenberger G Gatzek C Kalbitz K 2011 A new conceptu-al model for the fate of lignin in decomposing plant litter Ecology 92 1052ndash1062
Koide K Osono T Takeda H 2005 Fungal succession and decomposition of Camellia ja-ponica leaf litter Ecol Res 20 599ndash609
McLeod AR Newsham KK Fry SC 2007 Elevated UV-B radiation modifies the ex-tractability of carbohydrates from leaf litter of Quercus robur Soil Biol Biochem 39116ndash126 httpdxdoiorg101016jsoilbio200606019
McLeod AR Fry SC Loake GJ Messenger DJ Reay DS Smith KA Yun B-W 2008Ultraviolet radiation drives methane emissions from terrestrial plant pectins NewPhytol 180 124ndash132 httpdxdoiorg101111j1469-8137200802571x
Messenger DJ Fry SC Yamulki S McLeod AR 2012 Effects of UV-B filtration on thechemistry and decomposition of Fraxinus excelsior leaves Soil Biol Biochem 47133ndash141 httpdxdoiorg101016jsoilbio201112010
Middleton BA McKee KL 2004 Use of a latitudinal gradient in bald cypress (Taxodiumdistichum) production to examine physiological controls of biotic boundaries and po-tential responses to environmental change Glob Ecol Biogeogr 13 247ndash258 httpdxdoiorg101111j1466-822X200400088x
Moore TR Trofymow JA Siltanen M Prescott C Group CW 2005 Patterns of decom-position and carbon nitrogen and phosphorus dynamics of litter in upland forest andpeatland sites in central Canada Can J For Res 35 133ndash142 httpdxdoiorg101139x04-149
Moorhead DL Sinsabaugh RL 2006 A theoretical model of litter decay and microbialinteraction Ecol Monogr 76 151ndash174 httpdxdoiorg1018900012-9615(2006)076[0151ATMOLD]20CO2
Osono T 2007 Ecology of ligninolytic fungi associated with leaf litter decompositionEcol Res 22 955ndash974 httpdxdoiorg101007s11284-007-0390-z
Pellerin S Lavoie C 2003 Recent expansion of jack pine in peatlands of southeasternQueacutebec a paleoecological study Ecoscience 10 247ndash257
Reich PB Grigal DF Aber JD Gower ST 1997 Nitrogen mineralization and produc-tivity in 50 hardwood and conifer stands on diverse soils Ecology 78 335ndash347httpdxdoiorg1018900012-9658(1997)078[0335NMAPIH]20CO2
Reith F Drake HL Kuumlsel K 2002 Anaerobic activities of bacteria and fungi in moder-ately acidic conifer and deciduous leaf litter FEMS Microbiol Ecol 41 27ndash35httpdxdoiorg101111j1574-69412002tb00963x
Renard CMGC Voragen AGJ Thibault J-F Pilnik W 1991 Studies on appleprotopectin IV apple xyloglucans and influence of pectin extraction treatments ontheir solubility Carbohydr Polym 15 387ndash403 httpdxdoiorg1010160144-8617(91)90089-U
Rydin H Jeglum JK 2013 The Biology of Peatlands 2e Oxford University PressSchink B Zeikus JG 1982 Microbial ecology of pectin decomposition in anoxic lake sedi-
ments Microbiology 128 393ndash404 httpdxdoiorg10109900221287-128-2-393Strakovaacute P Anttila J Spetz P Kitunen V Tapanila T Laiho R 2010 Litter quality
and its response to water level drawdown in boreal peatlands at plant speciesand community level Plant Soil 335 501ndash520 httpdxdoiorg101007s11104-010-0447-6
Talbot JM Treseder KK 2011 Interactions among lignin cellulose and nitrogen drive lit-ter chemistryndashdecay relationships Ecology 93 345ndash354 httpdxdoiorg10189011-08431
148 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
Theander O Westerlund E 1993 Quantitative analysis of cell wall components In JungHG Buxton DR Hatfield RD Ralph J (Eds) Forage Cell Wall Structure and Di-gestibility Am Soc Agron Madison WI pp 83ndash104
Tiedje JM Sexstone AJ Parkin TB Revsbech NP 1984 Anaerobic processes in soilPlant Soil 76 197ndash212 httpdxdoiorg101007BF02205580
Van Soest PJ 1994 Nutritional Ecology of the Ruminant Cornell University Press IthacaNY
Van Soest PJ Robertson JB Lewis BA 1991 Methods for dietary fiber neutral deter-gent fiber and nonstarch polysaccharides in relation to animal nutrition J DairySci 74 3583ndash3597
Vann CD Megonigal JP 2003 Elevated CO2 and water depth regulation of methaneemissions comparison of woody and non-woody wetland plant species Biogeo-chemistry 63 117ndash134 httpdxdoiorg101023A1023397032331
Wang WJ Dalal RC Moody PW Smith CJ 2003 Relationships of soil respiration tomicrobial biomass substrate availability and clay content Soil Biol Biochem 35273ndash284
Webster EA Halpin C Chudek JA Tilston EL Hopkins DW 2005 Decomposition insoil of soluble insoluble and lignin-rich fractions of plant material from tobacco withgenetic modifications to lignin biosynthesis Soil Biol Biochem 37 751ndash760 httpdxdoiorg101016jsoilbio200409012
Williams CJ LePage BA Vann DR Tange T Ikeda H Ando M Kusakabe T TsuzukiH Sweda T 2003 Structure allometry and biomass of plantation Metasequoia
glyptostroboides in Japan For Ecol Manag 180 287ndash301 httpdxdoiorg101016S0378-1127(02)00567-4
Wright IJ Reich PB Westoby M Ackerly DD Baruch Z Bongers F Cavender-Bares JChapin T Cornelissen JHC Diemer M Flexas J Garnier E Groom PK Gulias JHikosaka K Lamont BB Lee T Lee W Lusk C Midgley JJ Navas M-LNiinemets U Oleksyn J Osada N Poorter H Poot P Prior L Pyankov VIRoumet C Thomas SC Tjoelker MG Veneklaas EJ Villar R 2004 The worldwideleaf economics spectrum Nature 428 821ndash827 httpdxdoiorg101038nature02403
Yavitt JB Williams CJ Wieder RK 1997 Production of methane and carbon dioxide inpeatland ecosystems across North America effects of temperature aeration and or-ganic chemistry of peat Geomicrobiol J 14 299ndash316 httpdxdoiorg10108001490459709378054
Zaghloul AM Gohar AA Naiem ZA-AM Abdel Bar F 2008 Taxodione a DNA-binding compound from Taxodium distichum L (Rich) Z Naturforsch C J Biosci63 355
Zak DR Holmes WE White DC Peacock AD Tilman D 2003 Plant diversity soilmicrobial communities and ecosystem function are there any links Ecology 842042ndash2050 httpdxdoiorg10189002-0433
142 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
stage of decomposition (Koide et al 2005 Moorhead and Sinsabaugh2006) The situation is especially acute in wetland soils where commonwisdom suggests slow rates of litter decay in particular with little oxy-gen and via anaerobic metabolism (Yavitt et al 1997)
Here we propose that the mechanistic link to soil microbial activityoccurs in the suite of chemical compounds that make up plant cellwalls Plant cell walls are combinations of three dominant structuralpolysaccharides (cellulose hemicellulose and pectin) the lignin poly-mer and various proteins and a variety of lowmolecular weight solublecompounds (Berg and McClaugherty 2003) A general model is thatcellulosemicrofibrils are embedded in amatrix of hemicellulose pectinand proteins (Keegstra 2010) In secondary cell walls lignin surroundsthe cellulose and is bound to the hemicellulose (Cosgrove 2005) There-fore the traditional litter decay model posits that initial decompositionof leaf litter consumes the soluble fraction and non-protected cellulose(Berg and Staaf 1980) leaving other components to accumulate in theresidue However recent research has shown that initial decompositioncan include the hemicellulose (Delaney et al 1996) pectin (Jenkins andSuberkropp 1995) and lignin (Klotzbuumlcher et al 2011) Moreover dif-ferent components of plant cell walls are known to vary in how accessi-ble and digestible they are to different microbial decomposers (Dinget al 2012) Thus variation in the relative amounts of these plant cellwall components among different species of conifers may play a rolein regulating the early stages of litter decay and the supply of substratesto the soil microbial community (Koide et al 2005)
We examined how tree species identitymight ultimately regulatemi-crobial activity in soil We used a common garden approach and addedneedle litter fromnine conifer species to soils from two forestedwetlandswith contrasting soil types We focused specifically on anaerobic produc-tion of CO2 and CH4 (methanogenesis) We expected a priori that CO2
production would exhibit relatively weak relationships with leaf litteridentity The reason being that CO2 comes from a plethora of decarboxyl-ation reactions fermentations and anaerobic respirations such that thenet effect on CO2 would be muted However for methanogenesis weexpected greater fidelity with leaf litter identity given that methanogensuse only a limited number of substrates derived specifically from fermen-tation and that fermentation specifically would depend on the suite offermentable material per plant species (Drake et al 2009 Reith et al2002) This follows the finding from Bremer et al (2007) that plantspecies identity influenced themicroorganisms that carried out soil deni-trification Therefore specific predictions were for greater rates of micro-bial activity with deciduous rather than long-lived foliage and thatvariation in the plant cell wall composition of litter among specieswould in part explain variation in rates of anaerobic microbial activity
2 Materials and methods
Soils were collected from two sites in central New York State Aver-age annual temperature for the region is 65 degC and average annual pre-cipitation is 1170 mm June and July typically have the highest meanmonthly precipitation
One site Labrador Hollow (local name hereafter peat soil) is a for-ested peatland located 8 km southeast of Tully New York (LabradorHollow Unique Area 42781932degN 76041628degW) The site has aclosed-canopy forest composed of Acer rubrum L (red maple) and sev-eral conifer species Pinus strobus L (eastern white pine) Tsugacanadensis (L) Carriegravere (eastern hemlock) and Larix laricina (Du Roi)K Koch (tamarack or larch) Vaccinium corymbosum L (highbush blue-berry) Toxicodendron radicans (L) Kuntze (poison ivy) and a mixtureof pteridophytes are common in the understory Sphagnum mossescover about 75 of the ground and include the following species Sphag-num squarrosum Crome Sphagnum girgensohnii Russow Sphagnumfimbriatum Wilson Sphagnum henryense Warnst and Sphagnumrussowii Warnst Beneath Sphagnum is a 0 to 10 cm thick layer ofdense roots overlying about 8 m of mesic Typic Medisaprists peat soil
Below that glacial till and sand cover Devonian age shale and limestone(Fisher et al 1970) Further details are given in Coles and Yavitt (2004)
The second site Sapsucker Woods (hereafter swamp soil) is a for-ested wetland in Ithaca New York (42477291degN 76451523degW) locat-ed 47 km to the southeast of the first site Sapsucker is a closed canopyforest dominated byA rubrum Alnus incana (L)Moench ssp rugosa (DuRoi) RT Clausen (Grey alder) and T canadensis Trees occur on elevatedhummocks surrounded by water-filled depressions The soil (Alden se-ries) is a mesic Mollic Endoaquept very poorly drained in depressionsand low areas It is approximately 1 m deep with a fragipan soil horizonthat prevents drainage andmaintains a highwater table throughout theyear Soil for the incubations was collected from the depressionsbetween tree hummocks
Needle litter was collected from trees growing within close proxim-ity (ca 1 km) of each other on the same soil and exposed to identicalclimate in the Cornell Plantations Natural Area We collected needlesfrom three deciduous conifers L laricina Metasequoia glyptostroboidesHu and Cheng (Dawn redwood) and Taxodium distichum (L) Rich(Bald cypress) and from six evergreen conifers (Pinus banksianaLamb (Jack pine) Pinus rigida Mill (Pitch pine) P strobus (Easternwhite pine) Picea abies (L) Karst (Norway spruce) Picea glauca(Moench) Voss (White spruce) and Piceamariana (Mill) Britton Sternsand Poggenb (Black spruce)) We collected needles at the end of thegrowing season in October by shaking branches gently indicating awell-formed abscission zone and collecting the needles that landed ina plastic bag placed beneath the branches Needles were air dried atlowhumidity Sub-sampleswere oven dried at 105 degC to ascertainmois-ture content Other portions were sampled for chemical analyses (de-tailed below) We constructed litterbags containing a total of 1 g dry-weight-equivalent litter using a 15-cm2 piece of fine mesh material(0675 mmmesh size) gathered and sealed with plastic thread
We quantified plant litter effects on rates of soil CH4 production andanaerobic CO2 production using in vitro incubation studies Five repli-cate soil samples were collected randomly combined and thoroughlymixed by hand before 30-g portions of soil were placed into individualincubation jars (N = 3 per leaf litter type) on the date of collectionWe added 20 mL of de-ionized water before we sealed the jar with alid that had a gas-tight septum to allow sampling gas in the headspaceof the sealed jar The jar headspace was made anoxic by evacuating forseveral minutes using a vacuum pump and refilling with O2-free N2The evacuation was repeated three times Jars with soil were incubatedfor 15 days at 20 degC and concentrations of CO2 and CH4 in the headspacewere measured every third day in order to establish the in vitro basalrate of microbial CO2 and CH4 production The pre-incubation periodalso reduced the amount of inherent substrate for soil microorganisms(see Wang et al 2003) thereby increasing the impact of the addedleaf litter on their activity
Sampleswere analyzed bygas chromatographyusing aflame ioniza-tion detector for CH4 and a thermal conductivity detector for CO2 Thegas chromatograph has a 275 m by 318 mm column of Poropak Q 80100 mesh (Waters Chromatography Milford MA) maintained at 50 degCto separate the gasses The flow rate of the He carrier gas was30 mL minminus1 The injector temperature was 110 degC
Following the pre-incubation period we opened the jars and addeda litterbag to the surface of the soil We tapped the bag to insure contactwith soil before re-sealing the jar and establishing anoxic conditionsand incubating at 20 degC as described above We measured CH4 andCO2 in the headspace every three to four days for 24 days At the endof the incubation period the litterbag was removed and the residue re-maining was analyzed for its biochemical content
We determined the composition of hemicellulose cellulose and lig-nin using the detergent fiber technique (Van Soest 1994) First wecombined and thoroughly mixed material per species before beingground through a Wiley Mill Duplicate subsamples (05 g) per specieswere subjected to sequential neutral detergent fiber (Van Soest et al1991) acid detergent fiber and acid detergent lignin digestions
143JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
(Method 97318 AOAC 2012) using an ANKOM fiber analyzer (ANKOMTechnology Macedon New York USA) The neutral detergent fiber di-gestion dissolves soluble cell wall components resulting in a residuecomposed of hemicellulose cellulose and lignin The acid detergentfiber digestion dissolves hemicellulose (hereafter referred to as deter-gent hemicellulose) leaving a residue of cellulose and lignin The gravi-metric difference between the neutral detergent fiber and the aciddetergent fiber residue provides an estimate of detergent hemicelluloseconcentration Treatment of the acid detergent fiber residue with 72ww sulfuric acid dissolves cellulose leaving a residue of acid-detergent lignin Samples were ash-corrected after the final acid diges-tion to account for any contamination of mineral soil in the acid-detergent lignin fraction
We also used a non-detergent sequential extraction procedure(McLeod et al 2007) to characterize the composition of the plant cellwall of litter before and again after incubation Briefly litter tissue wasextracted as seven fractions with (1) 10 formic acid to remove looselybound polysaccharides with (2) a phosphate buffer to dissolve and re-move loosely boundwater soluble pectinwith (3) cyclohexanediaminetetra acetic acid (CDTA) to remove calcium which releases bound pec-tin in the middle lamellae with (4) urea to cleave hydrogen bondsand separate pectin bonded covalently to cellulose as well as cell wallproteins with (5) sodium carbonate to remove any remaining pectinin the primary cell wall with (6) sodium hydroxide to dissolve
Fig 1 Mean production rates of CH4 (top panel) and CO2 (bottom panel) for two wetland soi24 days following the addition of different types of conifer litter Control soils were incubated
hemicelluloses (hereafter referred to as extractable hemicellulose)and finally with (7) formic acid to remove all remaining non-structural sugars
We performed analysis of variance (ANOVA) for CH4 production andfor CO2 production We examined the effect of soil type conifer (decid-uous pine spruce) and species Species was nested within coniferPearson correlation coefficients were calculated to evaluate the rela-tionship between CH4 and CO2 production rates and litter chemistry
3 Results
Rates of CH4 production in control soil without added litter were b-1 nmol gminus1 dayminus1 (Fig 1) Likewise control soil without added litter(Fig 1) had rates of CO2 production b 1 μmol gminus1 dayminus1 It is very likelythat slow rates of CH4 production and CO2 production in the controlsoils without added litter resulted from the pre-incubation periodbeing long enough to deplete substrates for resident microbial popula-tions Therefore we assume that production rates for CH4 and CO2 insoils with added litter are indicative of conifer litter identity and thesubstrates supplied by each species that supported anaerobic microbialactivity
Over the course of the 24-day incubation rates of CH4 productionwere much greater in soils with added litter (Fig 1) although rateswith added Taxodium litter were significantly less than rates with
ls (Labrador Hollow peat soil Sapsucker Woods swamp soil) incubated anaerobically forwithout litter addition Error bars are +minus one standard error of the mean
Table 2Plant litter type and pre-incubation concentrations of extractable cell wall fractions (car-bohydrates) expressed in g kgminus1 dry tissue
a Fractions are soluble carbohydrates (Fraction1)water soluble pectin (Fraction2) pectinin themiddle lamella (Fraction 3) pectin bound covalently to cellulose (Fraction 4) cell wallpectin (Fraction 5) extractable hemicellulose (Fraction 6) and residual non-cellulose carbo-hydrates (Fraction 7)
144 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
other litters The impact of litter addition on CH4was significantly largerfor the peat soil from Labrador Hollow (mean= 124 nmol gminus1 dayminus1)than for the swamp soil from Sapsucker Woods (49 nmol gminus1 dayminus1)(F1 54 = 2543 P = 00023) The significant effect of conifer type(F6 54 = 491 P = 00004) was evident for both soils Litter from Pinussupported the largest rates of CH4 production 187 nmol gminus1 dayminus1
when added to peat soil and 63 nmol gminus1 dayminus1 when added toswamp soil Picea litter supported the lowest CH4 production rates78 nmol gminus1 dayminus1 with peat soil 39 nmol gminus1 dayminus1 with swamp soilHence the deciduous conifers (Metasequoia Larix Taxodium) hadintermediate average rates of CH4 production 104 nmol gminus1 dayminus1
with peat soil 39 nmol gminus1 dayminus1 with swamp soil However theseaverage rates increase to 156 nmol gminus1 dayminus1 with peat soil and70 nmol gminus1 dayminus1 with swamp soil excluding Taxodium
Rates of anaerobic CO2 production were greater with litter addi-tions as opposed to control soils (Fig 1) Rates of anaerobic CO2 pro-duction were significantly greater with litter added to the peat soil(32 μmol gminus1 dayminus1) than the swamp soil (13 μmol gminus1 dayminus1)(F1 54 = 11580 P b 00001) The significant effect of conifer type(F6 54 = 263 P = 00259) was evident for both soils although dif-ferences among conifer types were larger for the peat soil Whenadded to the peat soil litter from the deciduous conifers supportedlarger rates of CO2 production (mean = 36 μmol gminus1 dayminus1) thanlitter from Pinus (mean = 33 μmol gminus1 dayminus1) or Picea (mean =31 μmol gminus1 dayminus1) Taxodium litter supported the largest ratesfor both soil types
The deciduous conifer litter had much less detergent hemicelluloseand cellulose than that in the evergreen conifers (Table 1) Thereforethe deciduous conifers had a much larger proportion of pectins andstructural proteins On average Pinus litter contained about 10 morecellulose and 14 more acid-detergent lignin than Picea litter
The sumof extractable cell wall components ranged from34 to 61of litter mass (Table 2) The largest fractions were loosely bound poly-saccharides (Fraction 1) pectin covalently bonded to cellulose (Fraction4) and extractable hemicellulose (Fraction 6) There was some varia-tion in values among members of a group of conifers For instanceTaxodium hadmore loosely bound polysaccharides and pectin covalent-ly bonded to cellulose but less extractable hemicellulose compared toMetasequoia and Larix P banksiana contained less extractable hemicel-lulose compared to the other two species of Pinus Among the sprucesP glauca was distinctive with more loosely bound polysaccharideswhereas P abies had less pectin covalently bonded to cellulose
There were considerable changes in the different fractions of cell wallcomponents during incubation (Fig 2) Loosely bound polysaccharides(Fraction 1) and pectin covalently bonded to cellulose (Fraction4) showed the largest declines The large increase inwater-soluble pectin(Fraction 2) was particularly noticeable In four of the litters (TaxodiumP banksiana P glauca and P abies) there was a large increase in the ex-tractable hemicellulose fraction (Fraction 6) during incubation Therealsowas a soil type effect on the changes in extractable carbohydrate frac-tions For example quantities of loosely bound polysaccharides (Fraction
Table 1Plant litter type and pre-incubation concentrations of detergentfiber fractions (g kgminus1 ashfree dry weight)
Litter Type Hemicellulose Cellulose Lignin Residuala
a Residual includes soluble cell-wall components including β-glucans and pectins as wellas soluble simple sugars amino acids water-soluble phenolics and insoluble protein
1) and pectin covalently bonded to cellulose exhibited greater declines inthe peat soil than in the swamp soil whereas the observed increase inwater soluble pectin (Fraction 2) also was greater in peat than in theswamp soil
We compared rates of soil CH4 production and CO2 production withinitial litter concentrations of detergent hemicellulose cellulose andacid-detergent lignin for all litter types (Table 3) We also comparedchanges in the extractable cell wall fractions (1 to 7) that occurred inall litter types during the incubations with soil CH4 production andCO2 production (Table 3) Therewas a relatively strong positive correla-tion (r ge 063 P b 01) between rates of CH4 production in both soiltypes and the initial detergent hemicellulose content of the litter Thenegative correlation (r=minus071 P b 005) between rates of CH4 produc-tion in the peat soil and pectin covalently bonded to cellulose meansthat the post-incubation increase in pectin covalently bonded to cellu-lose reduced CH4 production CO2 production in both soils was positive-ly correlated with the initial acid-detergent lignin content (rge 072 P b
005) Post incubation water-soluble pectin was positively correlated(r = 067 P b 005) with CO2 in the swamp soil
4 Discussion
Our comparative study examined several relationships between leaflitter fromnine different conifer taxa and anaerobicmicrobial activity intwo wetland soils Although our analysis is limited to correlation thedata provide new insights into plantndashsoil relationships For exampleseasonally deciduous leaves are thought to decay faster than evergreenleaves according to the Leaf Economics Spectrum (Cornwell et al2008) This is generally attributed to the seasonally deciduous leaveshaving low leaf mass per area and lower CN ratios compared to long-lived evergreen leaves However our finding of no significant differencein CO2 production for deciduous versus evergreen conifers suggests thatamore complex linkage exists between leaf litter traits and concomitantmicrobial trace gasproductionOur chemical analyses indicate 1) pectinand hemicelluloses vary among leaf litter types and their concentra-tions can change substantially in the early stages of leaf litter decompo-sition and 2) changes in pectin and hemicelluloses during leaf litterdecomposition might be fueling anaerobic microbial activity Thereforethe specific chemical attributes of the litter influence the anaerobic mi-crobial response to litter inputs
41 General soil microbial response to litter addition
Our data agree with findings of Reith et al (2002) that fresh Picealitter could promote anaerobic metabolism by soil microorganismsOverall the rates of anaerobic CO2 production in this study were typicalof rates found in other wetland soils (Yavitt et al 1997) Methane pro-duction and anaerobic CO2 productionwere greater in the peat soil than
Fig 2 Post-incubation versus pre-incubation comparison of sequentially extracted cell wall fractions (carbohydrates) of nine conifer litter types following a 24-day incubation in peat soilor swamp soil Degree of color value indicates themagnitude of changewhereas color indicates increases (green) or decreases (red) in a fraction following incubation The absolute changein concentrations of each fraction is listed within cells as grams carbohydrate per kilogram dry tissue (For interpretation of the references to color in this figure legend the reader is re-ferred to the web version of this article)
145JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
in the swamp soil This supports the general notion that rates of litterdecay are slowed by anaerobic metabolism in peat soils compared toperhaps greater aerobic metabolism in swamp soils (Moore et al2005) It is also likely that peat soils harbored a larger more active
Table 3Pearson correlation matrix for pre-incubation litter detergent fiber (cellulose hemicellu-lose lignin) concentrations and post incubation litter extractable cell wall fractions (1ndash7) versus trace gas production for each soil type
P b 001 001 b P b 005 005 b P b 01a Fractions are detergent fiber (cellulose hemicellulose lignin) and sequentially extracted
soluble carbohydrates (Fraction 1) water soluble pectin (Fraction 2) pectin in the middlelamella (Fraction 3) pectin bound covalently to cellulose (Fraction 4) cell wall pectin(Fraction 5) extractable hemicellulose (Fraction 6) and residual non-cellulose carbohy-drates (Fraction 7)
community of microorganisms capable of anaerobic metabolism Thewater-saturated peat provides a superior microbial habitat for anaero-bic activity and methanogenesis On the other hand the swamp soil isvariably saturated it ranges from being waterlogged during springsnowmelt to having a thin unsaturated surface layer during summerdry periods We sampled in the fall when a thin unsaturated layer wasapparent at the soil surface Thus our measured rates of anaerobic CO2
production in the swamp soil are consistent with the notion that facul-tative anaerobes can constitute about 10 of the aerobic microbial pop-ulation even in well-drained soil (Tiedje et al 1984) Our study alsoconfirms that the seasonally saturated conditions provide a tolerablehabitat for the persistence of methanogenic microorganisms whichare adapted to tolerate variable redox conditions (DeAngelis et al2010)
42 Leaf litter traits
In the context of the leaf economic spectrum it is logical to assumethat litter fromdeciduous coniferswould support greater rates ofmicro-bial activity than litters from evergreen Pinus or Picea This assumptionis based in part on the notion that seasonally deciduous leaves havebetter litter quality for decomposition However the prediction wasnot upheld and Pinus litter supported the greatest rates of CH4 produc-tion The microbial response to litter addition in our study reveals theuse of a simple binomial leaf trait characterization (evergreen vs decid-uous) as a predictor of soil microbial response to be lacking This findingcorroborates those of Moore et al (2005) that deciduous Larix needles
146 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
placed on the surface of peat soils decomposed slower than Pinus andPicea needles in sites in Canada
Why does the species identity of litter matter For one Taxodiumproduced themost unusual patternwith very slow rates of CH4 produc-tion but the greatest rate of CO2 production Other studies have foundthat Taxodium needles will decompose rapidly in wetlands especiallywhen exposed to alternating oxic and anoxic conditions (Battle andGolladay 2001) and thus the large CO2 production is not surprisingAlthoughwetlands dominated by Taxodium are known sources of atmo-spheric CH4 (Vann andMegonigal 2003) it is likely that fresh Taxodiumneedles have biochemical traits that slow the rate of CH4 production butnot CO2 production at the earliest stage of anaerobic decomposition Forexample Taxodium is known to produce taxoquinone and several otherditerpenoids that may have an inhibitory effect on microbes (Cervanteset al 2000 Zaghloul et al 2008) Perhaps it is not until such com-pounds degrade that Taxodium litter can be turned into methanogenicsubstrate
We also found that soilmicrobial response varied at the generic levelwith Pinus generally promoting greater CH4 than Picea litter Our findingthat P banksiana (Jack pine) litter supported CH4 production is particu-larly interesting because this species has been encroaching on peatlandsinNorth America recently (Pellerin and Lavoie 2003) Thus P banksianaencroachment may promote enhanced soil CH4 production and poten-tial emissions of atmospheric CH4 Our speculation that encroachmentof woody vegetation onto otherwise open peatlands may result inenhanced CH4 production awaits experimental confirmation Admitted-ly other environmental factors that regulate CH4 productionmay trumpthe influence of litter identity Nevertheless our results and a scarcity ofinformation in the literature suggest further study may be warranted
43 Leaf litter carbohydrate chemistry
Our analyses of conifer leaf litter quality in terms of different deter-gent fiber fractions and cell wall components in fresh litter were de-signed to determine how constituents other than cellulose and ligninmight affect microbial colonization and activity Most previous researchhas focused on cellulose because cellulose is often assumed to be themost abundant carbohydrate fraction by mass (Berg and Staaf 1980)In contrast the roles of hemicellulose and pectin in microbial respira-tion have not been studied extensively However pectin and hemicellu-lose act as the glue that holds cellulose fibers and lignin together in theplant cell wall (Cosgrove 2001 2005) Thus microbial decomposersattack compounds sequentially pectin and hemicellulose before cellu-lose then lignin Hemicellulose is a complex polymer with differentsugars and ratios of the different sugars can vary as a function of plantspecies and site fertility (Strakovaacute et al 2010) For instance Berg andMcClaugherty (2003) reported that arabinose and galactose in thehemicellulose fraction of Pinus needle litter decomposed immediatelyafter litterfall in pine forests whereas loss of mannose and xylose oc-curred after one year
Although detergent fiber analysis of plant litter has been applied inmany ecological studies of litter quality (ie results in Table 1) the se-quential extraction of cell wall components (ie results in Table 2) hasbeen used mainly in plant physiology studies (Fry 1989) mdash althoughseeMcLeod et al (2007)We quantified hemicellulose by bothmethodsand this gave conflicting results the sequential extraction method gavemuch larger values for the deciduous conifers whereas values weresimilar using both methods for the Pinus and Picea Indeed Theanderand Westerlund (1993) proposed alternatives to detergent hemicellu-lose arguing that the procedure is non-specific and might includesome pectin Jung and Lamb (2004) found that the detergent fibermethod overestimated the amount of hemicellulose in alfalfa leavesMuch more comparative work using both methods is warranted fornon-forage plants given that sequential extraction of cell wall compo-nents has the potential to providemore information about litter qualityin particular about pectin
Overall Fraction 2 (weakly bound water soluble pectin) was largerin leaf litter post-incubation than in fresh litter (Fig 2) The largeincrease in Fraction 2 across all litter types during the incubation wasunexpected as it could not be production of pectin per se Rather it like-ly represents the liberation of weakly bound sugars following incuba-tion that are then extracted in this fraction Messenger et al (2012)also found that amounts of extractable sugars increased during decom-position which they attributed to decomposition of cellulose proteinsand low-molecular-weight metabolites that are sugar-rich The anaero-bic metabolism in our study suggests another mechanism the sugarsmight be related to fermentation Low-molecular-weight organicacids derived from fermentation when in excess supply can lead togluconeogenesis (essentially reverse glycolysis) and the production ofcarbohydrates (Dijkstra et al 2011) The increase in extractable hemi-cellulose content of some of the litters (Fraction 6) also was surprisingPresumably this represents sugars that coalesce into a sodiumhydroxide-extractable fraction and is not decomposer production ofhemicellulose per se (Hoch 2007)
Fraction 4 (pectin covalently bound to cellulose) in generaldecreased in leaf litters during incubation This reflects the fracturingof plant cell walls during the earliest stage of decomposition Unfortu-nately the mechanism is not clear from the data we obtained Presum-ably fungal hyphae are involved ie filaments penetrate the plant cellwall and degrade it physically as well as chemically Clearly morework is needed to determine the fate of the liberated pectin ie is itmade soluble or metabolized to CO2 andor CH4
44 Changes in carbohydrate fractions versus microbial activity
Inmost cases positive correlations between components in leaf litterand microbial activity are interpreted as that component fueling activi-ty Likewise a negative relationship is viewed as either a recalcitrantcomponent or a component that is inhibitory tomicrobial activity How-ever our data are more complex as in some cases the negative correla-tion (low quantities of a carbohydrate fraction versus a rapid rate ofmicrobial activity) could be explained as rapid microbial metabolismcausing the loss (see below)
For Fraction 4 the straightforward positive correlationwith CH4 pro-duction suggests that slow fermentation of cell wall pectin provides aslow release of methanogenic substrate Previous studies demonstratethat the aerobic decay of pectin is linked to CH4 production (McLeodet al 2008) and thus our data extend this relationship to anaerobicme-tabolism Likewise pectin fermentation is known to fuel methanogensin anaerobic lake sediments (Schink and Zeikus 1982)
In contrast hemicellulose showed a positive correlation with CH4
production based on the detergent analysis but a negative correlationwith extractable Fraction 6 This is clearly troubling however in the se-quential extraction the earlier urea-extraction (Fraction 4) also removesstructural proteins (Fry 1989) so the negative correlation may reflectthe influence of protein and enzymes on CH4 production The sugar con-tent of the urea-extractable fraction in terms of arabinose galactosemannose xylose and others in each species is not known (cf Renardet al 1991) yet thismay also influence rates of CH4 production It is pos-sible that variation in the composition of monomers liberated fromhemicellulose may determine the supply of fermentation products tomethanogens
Rates of anaerobic CO2 production had positive relationships withacid-detergent lignin and with loosely bound water-soluble pectinBecause only a subset of microorganisms can degrade lignin completely(Osono 2007 Talbot and Treseder 2011) there is usually a negative re-lationship (Aerts 1997) especially in long-termdecomposition studiesFurthermore most of the research has been done with aerobic notanaerobic decay In contrast our correlation was with the initial acid-detergent lignin composition and we do not know how acid-detergent lignin quantity and quality changed during the incubationIt is also possible that the CO2 was not derived from lignin as such but
147JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
rather from hemicelluloses that occur intimately with acid-detergentlignin (Webster et al 2005) In contrast in soil incubated without O2pectin is a very good substrate for fermentation (Chin et al 1998) Itis likely that the relationship between pectin and anaerobic CO2 produc-tion represents the CO2 released from hydrolysis and fermentation aspectin is altered In other words pectin might be the first compoundthat microbial decomposers encounter and thus it modulates microbialactivity at the onset of decomposition asmeasured in the present study
5 Conclusions
About 6Mg haminus1 of leaves and needles fall to the ground each year inforestedwetlands (Reich et al 1997) andunder some circumstances de-ciduous conifers may produce much higher amounts (Middleton andMcKee 2004Williams et al 2003) This litter is a crucial source of energyfor soilmicroorganisms (Coles andYavitt 2004)Most studies of litter de-composition are prolonged making the first measurement of mass lossafter a few months and following the residue for one or more yearsThus we know much less about microbial colonization of fresh litter atthe very earliest stage of decomposition (Moorhead and Sinsabaugh2006) The situation is especially acute in forested wetlands where com-mon wisdom suggests slow rates of litter decay We suggest that finescale details of the decay process have been missed Hemicelluloses andpectin are probably the first molecular compounds to be released fromplant cell walls as they intertwine with lignin and cellulose and preventspecialized enzymes such as cellulases and lignases from working As aconsequence a moderate pulse of CH4 production should occur as freshlitter decays Moreover the encroachment of some types of conifersinto northern peatlands which reflects a shift in the dominant wetlandplant functional type may promote a new linkage between abovegroundcarbon fixation and belowground carbon mineralization
Acknowledgments
We thankTracy Bartella for assistance in thefield and laboratoryWethank three anonymous reviewers for constructive comments on themanuscript
References
Aerts R 1997 Climate leaf litter chemistry and leaf litter decomposition in terrestrialecosystems a triangular relationship Oikos 439ndash449
AOAC 2012 Official Methods of Analysis 19th ed Association of Official Analytical Chem-ists Gaithersburg MD
Battle JM Golladay SW 2001 Hydroperiod influence on breakdown of leaf litter incypress-gum wetlands Am Midl Nat 146 128ndash145
Berg B McClaugherty C 2003 Plant Litter SpringerBerg B Staaf H 1980 Decomposition rate and chemical changes of Scots pine needle lit-
ter II Influence of chemical composition Ecol Bull 373ndash390Berg EE Hillman KM Dial R DeRuwe A 2009 Recent woody invasion of wetlands on
the Kenai Peninsula Lowlands south-central Alaska a major regime shift after18 000 years of wet Sphagnumndashsedge peat recruitment Can J For Res 392033ndash2046 httpdxdoiorg101139X09-121
Bremer C Braker G Matthies D Reuter A Engels C Conrad R 2007 Impact of plantfunctional group plant species and sampling time on the composition of nirK-typedenitrifier communities in soil Appl Environ Microbiol 73 6876ndash6884
Cervantes FJ Velde S Lettinga G Field JA 2000 Competition between methanogenesisand quinone respiration for ecologically important substrates in anaerobic consortiaFEMS Microbiol Ecol 34 161ndash171
Chin K-J Rainey FA Janssen PH Conrad R 1998 Methanogenic degradation of poly-saccharides and the characterization of polysaccharolytic clostridia from anoxic ricefield soil Syst Appl Microbiol 21 185ndash200
Coles JRP Yavitt JB 2004 Linking belowground carbon allocation to anaerobic CH4
and CO2 production in a forested peatland New York State Geomicrobiol J 21445ndash455 httpdxdoiorg10108001490450490505419
Cornwell WK Cornelissen JHC Amatangelo K Dorrepaal E Eviner VT Godoy OHobbie SE Hoorens B Kurokawa H Peacuterez-Harguindeguy N Quested HMSantiago LS Wardle DA Wright IJ Aerts R Allison SD Van Bodegom PBrovkin V Chatain A Callaghan TV Diacuteaz S Garnier E Gurvich DE Kazakou EKlein JA Read J Reich PB Soudzilovskaia NA Vaieretti MV Westoby M 2008Plant species traits are the predominant control on litter decomposition rates within bi-omes worldwide Ecol Lett 11 1065ndash1071 httpdxdoiorg101111j1461-0248200801219x
Cosgrove DJ 2001 Wall structure and wall loosening A look backwards and forwardsPlant Physiol 125 131ndash134
Cosgrove DJ 2005 Growth of the plant cell wall Nat Rev Mol Cell Biol 6 850ndash861httpdxdoiorg101038nrm1746
DeAngelis KM Silver WL Thompson AW Firestone MK 2010 Microbial communi-ties acclimate to recurring changes in soil redox potential status Environ Microbiol12 3137ndash3149
Delaney MT Fernandez IJ Simmons JA Briggs RD 1996 Red Maple and White Pinelitter quality initial changes with decomposition Technical Bulletin No 162 Techni-cal Bulletin University of Maine Orono ME
Dijkstra P Blankinship JC Selmants PC Hart SC Koch GW Schwartz E HungateBA 2011 Probing carbon flux patterns through soil microbial metabolic networksusing parallel position-specific tracer labeling Soil Biol Biochem 43 126ndash132
Ding SY Liu YS Zeng Y Himmel ME Baker JO Bayer EA 2012 How does plantcell wall nanoscale architecture correlate with enzymatic digestibility Science 3381055ndash1060
Drake HL Horn MA Wuumlst PK 2009 Intermediary ecosystem metabolism as a maindriver of methanogenesis in acidic wetland soil Environ Microbiol Rep 1 307ndash318
Fisher D Isachsen Y Rickard L 1970 Geologic map of New York State consisting of 5sheets Niagara Finger Lakes Hudson-Mohawk Adirondack and Lower Hudson Mapand Chart Series No 15
Fry S 1989 Analysis of cross-links in the growing cell walls of higher plants Plant FibersSpringer pp 12ndash36
Heijmans MM Knaap YA Holmgren M Limpens J 2013 Persistent versus transienttree encroachment of temperate peat bogs effects of climate warming and droughtevents Glob Chang Biol 19 2240ndash2250
Hoch G 2007 Cell wall hemicelluloses as mobile carbon stores in non-reproductiveplant tissues Funct Ecol 21 823ndash834
Jackson RB Banner JL Jobbaacutegy EG PockmanWTWall DH 2002 Ecosystem carbonloss with woody plant invasion of grasslands Nature 418 623ndash626 httpdxdoiorg101038nature00910
Jenkins CC Suberkropp K 1995 The influence of water chemistry on the enzymaticdegradation of leaves in streams Freshw Biol 33 245ndash253 httpdxdoiorg101111j1365-24271995tb01165x
Jung HJG Lamb JFS 2004 Prediction of cell wall polysaccharide and lignin concentra-tions of alfalfa stems from detergent fiber analysis Biomass Bioenergy 27 365ndash373
Klotzbuumlcher T Kaiser K Guggenberger G Gatzek C Kalbitz K 2011 A new conceptu-al model for the fate of lignin in decomposing plant litter Ecology 92 1052ndash1062
Koide K Osono T Takeda H 2005 Fungal succession and decomposition of Camellia ja-ponica leaf litter Ecol Res 20 599ndash609
McLeod AR Newsham KK Fry SC 2007 Elevated UV-B radiation modifies the ex-tractability of carbohydrates from leaf litter of Quercus robur Soil Biol Biochem 39116ndash126 httpdxdoiorg101016jsoilbio200606019
McLeod AR Fry SC Loake GJ Messenger DJ Reay DS Smith KA Yun B-W 2008Ultraviolet radiation drives methane emissions from terrestrial plant pectins NewPhytol 180 124ndash132 httpdxdoiorg101111j1469-8137200802571x
Messenger DJ Fry SC Yamulki S McLeod AR 2012 Effects of UV-B filtration on thechemistry and decomposition of Fraxinus excelsior leaves Soil Biol Biochem 47133ndash141 httpdxdoiorg101016jsoilbio201112010
Middleton BA McKee KL 2004 Use of a latitudinal gradient in bald cypress (Taxodiumdistichum) production to examine physiological controls of biotic boundaries and po-tential responses to environmental change Glob Ecol Biogeogr 13 247ndash258 httpdxdoiorg101111j1466-822X200400088x
Moore TR Trofymow JA Siltanen M Prescott C Group CW 2005 Patterns of decom-position and carbon nitrogen and phosphorus dynamics of litter in upland forest andpeatland sites in central Canada Can J For Res 35 133ndash142 httpdxdoiorg101139x04-149
Moorhead DL Sinsabaugh RL 2006 A theoretical model of litter decay and microbialinteraction Ecol Monogr 76 151ndash174 httpdxdoiorg1018900012-9615(2006)076[0151ATMOLD]20CO2
Osono T 2007 Ecology of ligninolytic fungi associated with leaf litter decompositionEcol Res 22 955ndash974 httpdxdoiorg101007s11284-007-0390-z
Pellerin S Lavoie C 2003 Recent expansion of jack pine in peatlands of southeasternQueacutebec a paleoecological study Ecoscience 10 247ndash257
Reich PB Grigal DF Aber JD Gower ST 1997 Nitrogen mineralization and produc-tivity in 50 hardwood and conifer stands on diverse soils Ecology 78 335ndash347httpdxdoiorg1018900012-9658(1997)078[0335NMAPIH]20CO2
Reith F Drake HL Kuumlsel K 2002 Anaerobic activities of bacteria and fungi in moder-ately acidic conifer and deciduous leaf litter FEMS Microbiol Ecol 41 27ndash35httpdxdoiorg101111j1574-69412002tb00963x
Renard CMGC Voragen AGJ Thibault J-F Pilnik W 1991 Studies on appleprotopectin IV apple xyloglucans and influence of pectin extraction treatments ontheir solubility Carbohydr Polym 15 387ndash403 httpdxdoiorg1010160144-8617(91)90089-U
Rydin H Jeglum JK 2013 The Biology of Peatlands 2e Oxford University PressSchink B Zeikus JG 1982 Microbial ecology of pectin decomposition in anoxic lake sedi-
ments Microbiology 128 393ndash404 httpdxdoiorg10109900221287-128-2-393Strakovaacute P Anttila J Spetz P Kitunen V Tapanila T Laiho R 2010 Litter quality
and its response to water level drawdown in boreal peatlands at plant speciesand community level Plant Soil 335 501ndash520 httpdxdoiorg101007s11104-010-0447-6
Talbot JM Treseder KK 2011 Interactions among lignin cellulose and nitrogen drive lit-ter chemistryndashdecay relationships Ecology 93 345ndash354 httpdxdoiorg10189011-08431
148 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
Theander O Westerlund E 1993 Quantitative analysis of cell wall components In JungHG Buxton DR Hatfield RD Ralph J (Eds) Forage Cell Wall Structure and Di-gestibility Am Soc Agron Madison WI pp 83ndash104
Tiedje JM Sexstone AJ Parkin TB Revsbech NP 1984 Anaerobic processes in soilPlant Soil 76 197ndash212 httpdxdoiorg101007BF02205580
Van Soest PJ 1994 Nutritional Ecology of the Ruminant Cornell University Press IthacaNY
Van Soest PJ Robertson JB Lewis BA 1991 Methods for dietary fiber neutral deter-gent fiber and nonstarch polysaccharides in relation to animal nutrition J DairySci 74 3583ndash3597
Vann CD Megonigal JP 2003 Elevated CO2 and water depth regulation of methaneemissions comparison of woody and non-woody wetland plant species Biogeo-chemistry 63 117ndash134 httpdxdoiorg101023A1023397032331
Wang WJ Dalal RC Moody PW Smith CJ 2003 Relationships of soil respiration tomicrobial biomass substrate availability and clay content Soil Biol Biochem 35273ndash284
Webster EA Halpin C Chudek JA Tilston EL Hopkins DW 2005 Decomposition insoil of soluble insoluble and lignin-rich fractions of plant material from tobacco withgenetic modifications to lignin biosynthesis Soil Biol Biochem 37 751ndash760 httpdxdoiorg101016jsoilbio200409012
Williams CJ LePage BA Vann DR Tange T Ikeda H Ando M Kusakabe T TsuzukiH Sweda T 2003 Structure allometry and biomass of plantation Metasequoia
glyptostroboides in Japan For Ecol Manag 180 287ndash301 httpdxdoiorg101016S0378-1127(02)00567-4
Wright IJ Reich PB Westoby M Ackerly DD Baruch Z Bongers F Cavender-Bares JChapin T Cornelissen JHC Diemer M Flexas J Garnier E Groom PK Gulias JHikosaka K Lamont BB Lee T Lee W Lusk C Midgley JJ Navas M-LNiinemets U Oleksyn J Osada N Poorter H Poot P Prior L Pyankov VIRoumet C Thomas SC Tjoelker MG Veneklaas EJ Villar R 2004 The worldwideleaf economics spectrum Nature 428 821ndash827 httpdxdoiorg101038nature02403
Yavitt JB Williams CJ Wieder RK 1997 Production of methane and carbon dioxide inpeatland ecosystems across North America effects of temperature aeration and or-ganic chemistry of peat Geomicrobiol J 14 299ndash316 httpdxdoiorg10108001490459709378054
Zaghloul AM Gohar AA Naiem ZA-AM Abdel Bar F 2008 Taxodione a DNA-binding compound from Taxodium distichum L (Rich) Z Naturforsch C J Biosci63 355
Zak DR Holmes WE White DC Peacock AD Tilman D 2003 Plant diversity soilmicrobial communities and ecosystem function are there any links Ecology 842042ndash2050 httpdxdoiorg10189002-0433
143JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
(Method 97318 AOAC 2012) using an ANKOM fiber analyzer (ANKOMTechnology Macedon New York USA) The neutral detergent fiber di-gestion dissolves soluble cell wall components resulting in a residuecomposed of hemicellulose cellulose and lignin The acid detergentfiber digestion dissolves hemicellulose (hereafter referred to as deter-gent hemicellulose) leaving a residue of cellulose and lignin The gravi-metric difference between the neutral detergent fiber and the aciddetergent fiber residue provides an estimate of detergent hemicelluloseconcentration Treatment of the acid detergent fiber residue with 72ww sulfuric acid dissolves cellulose leaving a residue of acid-detergent lignin Samples were ash-corrected after the final acid diges-tion to account for any contamination of mineral soil in the acid-detergent lignin fraction
We also used a non-detergent sequential extraction procedure(McLeod et al 2007) to characterize the composition of the plant cellwall of litter before and again after incubation Briefly litter tissue wasextracted as seven fractions with (1) 10 formic acid to remove looselybound polysaccharides with (2) a phosphate buffer to dissolve and re-move loosely boundwater soluble pectinwith (3) cyclohexanediaminetetra acetic acid (CDTA) to remove calcium which releases bound pec-tin in the middle lamellae with (4) urea to cleave hydrogen bondsand separate pectin bonded covalently to cellulose as well as cell wallproteins with (5) sodium carbonate to remove any remaining pectinin the primary cell wall with (6) sodium hydroxide to dissolve
Fig 1 Mean production rates of CH4 (top panel) and CO2 (bottom panel) for two wetland soi24 days following the addition of different types of conifer litter Control soils were incubated
hemicelluloses (hereafter referred to as extractable hemicellulose)and finally with (7) formic acid to remove all remaining non-structural sugars
We performed analysis of variance (ANOVA) for CH4 production andfor CO2 production We examined the effect of soil type conifer (decid-uous pine spruce) and species Species was nested within coniferPearson correlation coefficients were calculated to evaluate the rela-tionship between CH4 and CO2 production rates and litter chemistry
3 Results
Rates of CH4 production in control soil without added litter were b-1 nmol gminus1 dayminus1 (Fig 1) Likewise control soil without added litter(Fig 1) had rates of CO2 production b 1 μmol gminus1 dayminus1 It is very likelythat slow rates of CH4 production and CO2 production in the controlsoils without added litter resulted from the pre-incubation periodbeing long enough to deplete substrates for resident microbial popula-tions Therefore we assume that production rates for CH4 and CO2 insoils with added litter are indicative of conifer litter identity and thesubstrates supplied by each species that supported anaerobic microbialactivity
Over the course of the 24-day incubation rates of CH4 productionwere much greater in soils with added litter (Fig 1) although rateswith added Taxodium litter were significantly less than rates with
ls (Labrador Hollow peat soil Sapsucker Woods swamp soil) incubated anaerobically forwithout litter addition Error bars are +minus one standard error of the mean
Table 2Plant litter type and pre-incubation concentrations of extractable cell wall fractions (car-bohydrates) expressed in g kgminus1 dry tissue
a Fractions are soluble carbohydrates (Fraction1)water soluble pectin (Fraction2) pectinin themiddle lamella (Fraction 3) pectin bound covalently to cellulose (Fraction 4) cell wallpectin (Fraction 5) extractable hemicellulose (Fraction 6) and residual non-cellulose carbo-hydrates (Fraction 7)
144 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
other litters The impact of litter addition on CH4was significantly largerfor the peat soil from Labrador Hollow (mean= 124 nmol gminus1 dayminus1)than for the swamp soil from Sapsucker Woods (49 nmol gminus1 dayminus1)(F1 54 = 2543 P = 00023) The significant effect of conifer type(F6 54 = 491 P = 00004) was evident for both soils Litter from Pinussupported the largest rates of CH4 production 187 nmol gminus1 dayminus1
when added to peat soil and 63 nmol gminus1 dayminus1 when added toswamp soil Picea litter supported the lowest CH4 production rates78 nmol gminus1 dayminus1 with peat soil 39 nmol gminus1 dayminus1 with swamp soilHence the deciduous conifers (Metasequoia Larix Taxodium) hadintermediate average rates of CH4 production 104 nmol gminus1 dayminus1
with peat soil 39 nmol gminus1 dayminus1 with swamp soil However theseaverage rates increase to 156 nmol gminus1 dayminus1 with peat soil and70 nmol gminus1 dayminus1 with swamp soil excluding Taxodium
Rates of anaerobic CO2 production were greater with litter addi-tions as opposed to control soils (Fig 1) Rates of anaerobic CO2 pro-duction were significantly greater with litter added to the peat soil(32 μmol gminus1 dayminus1) than the swamp soil (13 μmol gminus1 dayminus1)(F1 54 = 11580 P b 00001) The significant effect of conifer type(F6 54 = 263 P = 00259) was evident for both soils although dif-ferences among conifer types were larger for the peat soil Whenadded to the peat soil litter from the deciduous conifers supportedlarger rates of CO2 production (mean = 36 μmol gminus1 dayminus1) thanlitter from Pinus (mean = 33 μmol gminus1 dayminus1) or Picea (mean =31 μmol gminus1 dayminus1) Taxodium litter supported the largest ratesfor both soil types
The deciduous conifer litter had much less detergent hemicelluloseand cellulose than that in the evergreen conifers (Table 1) Thereforethe deciduous conifers had a much larger proportion of pectins andstructural proteins On average Pinus litter contained about 10 morecellulose and 14 more acid-detergent lignin than Picea litter
The sumof extractable cell wall components ranged from34 to 61of litter mass (Table 2) The largest fractions were loosely bound poly-saccharides (Fraction 1) pectin covalently bonded to cellulose (Fraction4) and extractable hemicellulose (Fraction 6) There was some varia-tion in values among members of a group of conifers For instanceTaxodium hadmore loosely bound polysaccharides and pectin covalent-ly bonded to cellulose but less extractable hemicellulose compared toMetasequoia and Larix P banksiana contained less extractable hemicel-lulose compared to the other two species of Pinus Among the sprucesP glauca was distinctive with more loosely bound polysaccharideswhereas P abies had less pectin covalently bonded to cellulose
There were considerable changes in the different fractions of cell wallcomponents during incubation (Fig 2) Loosely bound polysaccharides(Fraction 1) and pectin covalently bonded to cellulose (Fraction4) showed the largest declines The large increase inwater-soluble pectin(Fraction 2) was particularly noticeable In four of the litters (TaxodiumP banksiana P glauca and P abies) there was a large increase in the ex-tractable hemicellulose fraction (Fraction 6) during incubation Therealsowas a soil type effect on the changes in extractable carbohydrate frac-tions For example quantities of loosely bound polysaccharides (Fraction
Table 1Plant litter type and pre-incubation concentrations of detergentfiber fractions (g kgminus1 ashfree dry weight)
Litter Type Hemicellulose Cellulose Lignin Residuala
a Residual includes soluble cell-wall components including β-glucans and pectins as wellas soluble simple sugars amino acids water-soluble phenolics and insoluble protein
1) and pectin covalently bonded to cellulose exhibited greater declines inthe peat soil than in the swamp soil whereas the observed increase inwater soluble pectin (Fraction 2) also was greater in peat than in theswamp soil
We compared rates of soil CH4 production and CO2 production withinitial litter concentrations of detergent hemicellulose cellulose andacid-detergent lignin for all litter types (Table 3) We also comparedchanges in the extractable cell wall fractions (1 to 7) that occurred inall litter types during the incubations with soil CH4 production andCO2 production (Table 3) Therewas a relatively strong positive correla-tion (r ge 063 P b 01) between rates of CH4 production in both soiltypes and the initial detergent hemicellulose content of the litter Thenegative correlation (r=minus071 P b 005) between rates of CH4 produc-tion in the peat soil and pectin covalently bonded to cellulose meansthat the post-incubation increase in pectin covalently bonded to cellu-lose reduced CH4 production CO2 production in both soils was positive-ly correlated with the initial acid-detergent lignin content (rge 072 P b
005) Post incubation water-soluble pectin was positively correlated(r = 067 P b 005) with CO2 in the swamp soil
4 Discussion
Our comparative study examined several relationships between leaflitter fromnine different conifer taxa and anaerobicmicrobial activity intwo wetland soils Although our analysis is limited to correlation thedata provide new insights into plantndashsoil relationships For exampleseasonally deciduous leaves are thought to decay faster than evergreenleaves according to the Leaf Economics Spectrum (Cornwell et al2008) This is generally attributed to the seasonally deciduous leaveshaving low leaf mass per area and lower CN ratios compared to long-lived evergreen leaves However our finding of no significant differencein CO2 production for deciduous versus evergreen conifers suggests thatamore complex linkage exists between leaf litter traits and concomitantmicrobial trace gasproductionOur chemical analyses indicate 1) pectinand hemicelluloses vary among leaf litter types and their concentra-tions can change substantially in the early stages of leaf litter decompo-sition and 2) changes in pectin and hemicelluloses during leaf litterdecomposition might be fueling anaerobic microbial activity Thereforethe specific chemical attributes of the litter influence the anaerobic mi-crobial response to litter inputs
41 General soil microbial response to litter addition
Our data agree with findings of Reith et al (2002) that fresh Picealitter could promote anaerobic metabolism by soil microorganismsOverall the rates of anaerobic CO2 production in this study were typicalof rates found in other wetland soils (Yavitt et al 1997) Methane pro-duction and anaerobic CO2 productionwere greater in the peat soil than
Fig 2 Post-incubation versus pre-incubation comparison of sequentially extracted cell wall fractions (carbohydrates) of nine conifer litter types following a 24-day incubation in peat soilor swamp soil Degree of color value indicates themagnitude of changewhereas color indicates increases (green) or decreases (red) in a fraction following incubation The absolute changein concentrations of each fraction is listed within cells as grams carbohydrate per kilogram dry tissue (For interpretation of the references to color in this figure legend the reader is re-ferred to the web version of this article)
145JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
in the swamp soil This supports the general notion that rates of litterdecay are slowed by anaerobic metabolism in peat soils compared toperhaps greater aerobic metabolism in swamp soils (Moore et al2005) It is also likely that peat soils harbored a larger more active
Table 3Pearson correlation matrix for pre-incubation litter detergent fiber (cellulose hemicellu-lose lignin) concentrations and post incubation litter extractable cell wall fractions (1ndash7) versus trace gas production for each soil type
P b 001 001 b P b 005 005 b P b 01a Fractions are detergent fiber (cellulose hemicellulose lignin) and sequentially extracted
soluble carbohydrates (Fraction 1) water soluble pectin (Fraction 2) pectin in the middlelamella (Fraction 3) pectin bound covalently to cellulose (Fraction 4) cell wall pectin(Fraction 5) extractable hemicellulose (Fraction 6) and residual non-cellulose carbohy-drates (Fraction 7)
community of microorganisms capable of anaerobic metabolism Thewater-saturated peat provides a superior microbial habitat for anaero-bic activity and methanogenesis On the other hand the swamp soil isvariably saturated it ranges from being waterlogged during springsnowmelt to having a thin unsaturated surface layer during summerdry periods We sampled in the fall when a thin unsaturated layer wasapparent at the soil surface Thus our measured rates of anaerobic CO2
production in the swamp soil are consistent with the notion that facul-tative anaerobes can constitute about 10 of the aerobic microbial pop-ulation even in well-drained soil (Tiedje et al 1984) Our study alsoconfirms that the seasonally saturated conditions provide a tolerablehabitat for the persistence of methanogenic microorganisms whichare adapted to tolerate variable redox conditions (DeAngelis et al2010)
42 Leaf litter traits
In the context of the leaf economic spectrum it is logical to assumethat litter fromdeciduous coniferswould support greater rates ofmicro-bial activity than litters from evergreen Pinus or Picea This assumptionis based in part on the notion that seasonally deciduous leaves havebetter litter quality for decomposition However the prediction wasnot upheld and Pinus litter supported the greatest rates of CH4 produc-tion The microbial response to litter addition in our study reveals theuse of a simple binomial leaf trait characterization (evergreen vs decid-uous) as a predictor of soil microbial response to be lacking This findingcorroborates those of Moore et al (2005) that deciduous Larix needles
146 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
placed on the surface of peat soils decomposed slower than Pinus andPicea needles in sites in Canada
Why does the species identity of litter matter For one Taxodiumproduced themost unusual patternwith very slow rates of CH4 produc-tion but the greatest rate of CO2 production Other studies have foundthat Taxodium needles will decompose rapidly in wetlands especiallywhen exposed to alternating oxic and anoxic conditions (Battle andGolladay 2001) and thus the large CO2 production is not surprisingAlthoughwetlands dominated by Taxodium are known sources of atmo-spheric CH4 (Vann andMegonigal 2003) it is likely that fresh Taxodiumneedles have biochemical traits that slow the rate of CH4 production butnot CO2 production at the earliest stage of anaerobic decomposition Forexample Taxodium is known to produce taxoquinone and several otherditerpenoids that may have an inhibitory effect on microbes (Cervanteset al 2000 Zaghloul et al 2008) Perhaps it is not until such com-pounds degrade that Taxodium litter can be turned into methanogenicsubstrate
We also found that soilmicrobial response varied at the generic levelwith Pinus generally promoting greater CH4 than Picea litter Our findingthat P banksiana (Jack pine) litter supported CH4 production is particu-larly interesting because this species has been encroaching on peatlandsinNorth America recently (Pellerin and Lavoie 2003) Thus P banksianaencroachment may promote enhanced soil CH4 production and poten-tial emissions of atmospheric CH4 Our speculation that encroachmentof woody vegetation onto otherwise open peatlands may result inenhanced CH4 production awaits experimental confirmation Admitted-ly other environmental factors that regulate CH4 productionmay trumpthe influence of litter identity Nevertheless our results and a scarcity ofinformation in the literature suggest further study may be warranted
43 Leaf litter carbohydrate chemistry
Our analyses of conifer leaf litter quality in terms of different deter-gent fiber fractions and cell wall components in fresh litter were de-signed to determine how constituents other than cellulose and ligninmight affect microbial colonization and activity Most previous researchhas focused on cellulose because cellulose is often assumed to be themost abundant carbohydrate fraction by mass (Berg and Staaf 1980)In contrast the roles of hemicellulose and pectin in microbial respira-tion have not been studied extensively However pectin and hemicellu-lose act as the glue that holds cellulose fibers and lignin together in theplant cell wall (Cosgrove 2001 2005) Thus microbial decomposersattack compounds sequentially pectin and hemicellulose before cellu-lose then lignin Hemicellulose is a complex polymer with differentsugars and ratios of the different sugars can vary as a function of plantspecies and site fertility (Strakovaacute et al 2010) For instance Berg andMcClaugherty (2003) reported that arabinose and galactose in thehemicellulose fraction of Pinus needle litter decomposed immediatelyafter litterfall in pine forests whereas loss of mannose and xylose oc-curred after one year
Although detergent fiber analysis of plant litter has been applied inmany ecological studies of litter quality (ie results in Table 1) the se-quential extraction of cell wall components (ie results in Table 2) hasbeen used mainly in plant physiology studies (Fry 1989) mdash althoughseeMcLeod et al (2007)We quantified hemicellulose by bothmethodsand this gave conflicting results the sequential extraction method gavemuch larger values for the deciduous conifers whereas values weresimilar using both methods for the Pinus and Picea Indeed Theanderand Westerlund (1993) proposed alternatives to detergent hemicellu-lose arguing that the procedure is non-specific and might includesome pectin Jung and Lamb (2004) found that the detergent fibermethod overestimated the amount of hemicellulose in alfalfa leavesMuch more comparative work using both methods is warranted fornon-forage plants given that sequential extraction of cell wall compo-nents has the potential to providemore information about litter qualityin particular about pectin
Overall Fraction 2 (weakly bound water soluble pectin) was largerin leaf litter post-incubation than in fresh litter (Fig 2) The largeincrease in Fraction 2 across all litter types during the incubation wasunexpected as it could not be production of pectin per se Rather it like-ly represents the liberation of weakly bound sugars following incuba-tion that are then extracted in this fraction Messenger et al (2012)also found that amounts of extractable sugars increased during decom-position which they attributed to decomposition of cellulose proteinsand low-molecular-weight metabolites that are sugar-rich The anaero-bic metabolism in our study suggests another mechanism the sugarsmight be related to fermentation Low-molecular-weight organicacids derived from fermentation when in excess supply can lead togluconeogenesis (essentially reverse glycolysis) and the production ofcarbohydrates (Dijkstra et al 2011) The increase in extractable hemi-cellulose content of some of the litters (Fraction 6) also was surprisingPresumably this represents sugars that coalesce into a sodiumhydroxide-extractable fraction and is not decomposer production ofhemicellulose per se (Hoch 2007)
Fraction 4 (pectin covalently bound to cellulose) in generaldecreased in leaf litters during incubation This reflects the fracturingof plant cell walls during the earliest stage of decomposition Unfortu-nately the mechanism is not clear from the data we obtained Presum-ably fungal hyphae are involved ie filaments penetrate the plant cellwall and degrade it physically as well as chemically Clearly morework is needed to determine the fate of the liberated pectin ie is itmade soluble or metabolized to CO2 andor CH4
44 Changes in carbohydrate fractions versus microbial activity
Inmost cases positive correlations between components in leaf litterand microbial activity are interpreted as that component fueling activi-ty Likewise a negative relationship is viewed as either a recalcitrantcomponent or a component that is inhibitory tomicrobial activity How-ever our data are more complex as in some cases the negative correla-tion (low quantities of a carbohydrate fraction versus a rapid rate ofmicrobial activity) could be explained as rapid microbial metabolismcausing the loss (see below)
For Fraction 4 the straightforward positive correlationwith CH4 pro-duction suggests that slow fermentation of cell wall pectin provides aslow release of methanogenic substrate Previous studies demonstratethat the aerobic decay of pectin is linked to CH4 production (McLeodet al 2008) and thus our data extend this relationship to anaerobicme-tabolism Likewise pectin fermentation is known to fuel methanogensin anaerobic lake sediments (Schink and Zeikus 1982)
In contrast hemicellulose showed a positive correlation with CH4
production based on the detergent analysis but a negative correlationwith extractable Fraction 6 This is clearly troubling however in the se-quential extraction the earlier urea-extraction (Fraction 4) also removesstructural proteins (Fry 1989) so the negative correlation may reflectthe influence of protein and enzymes on CH4 production The sugar con-tent of the urea-extractable fraction in terms of arabinose galactosemannose xylose and others in each species is not known (cf Renardet al 1991) yet thismay also influence rates of CH4 production It is pos-sible that variation in the composition of monomers liberated fromhemicellulose may determine the supply of fermentation products tomethanogens
Rates of anaerobic CO2 production had positive relationships withacid-detergent lignin and with loosely bound water-soluble pectinBecause only a subset of microorganisms can degrade lignin completely(Osono 2007 Talbot and Treseder 2011) there is usually a negative re-lationship (Aerts 1997) especially in long-termdecomposition studiesFurthermore most of the research has been done with aerobic notanaerobic decay In contrast our correlation was with the initial acid-detergent lignin composition and we do not know how acid-detergent lignin quantity and quality changed during the incubationIt is also possible that the CO2 was not derived from lignin as such but
147JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
rather from hemicelluloses that occur intimately with acid-detergentlignin (Webster et al 2005) In contrast in soil incubated without O2pectin is a very good substrate for fermentation (Chin et al 1998) Itis likely that the relationship between pectin and anaerobic CO2 produc-tion represents the CO2 released from hydrolysis and fermentation aspectin is altered In other words pectin might be the first compoundthat microbial decomposers encounter and thus it modulates microbialactivity at the onset of decomposition asmeasured in the present study
5 Conclusions
About 6Mg haminus1 of leaves and needles fall to the ground each year inforestedwetlands (Reich et al 1997) andunder some circumstances de-ciduous conifers may produce much higher amounts (Middleton andMcKee 2004Williams et al 2003) This litter is a crucial source of energyfor soilmicroorganisms (Coles andYavitt 2004)Most studies of litter de-composition are prolonged making the first measurement of mass lossafter a few months and following the residue for one or more yearsThus we know much less about microbial colonization of fresh litter atthe very earliest stage of decomposition (Moorhead and Sinsabaugh2006) The situation is especially acute in forested wetlands where com-mon wisdom suggests slow rates of litter decay We suggest that finescale details of the decay process have been missed Hemicelluloses andpectin are probably the first molecular compounds to be released fromplant cell walls as they intertwine with lignin and cellulose and preventspecialized enzymes such as cellulases and lignases from working As aconsequence a moderate pulse of CH4 production should occur as freshlitter decays Moreover the encroachment of some types of conifersinto northern peatlands which reflects a shift in the dominant wetlandplant functional type may promote a new linkage between abovegroundcarbon fixation and belowground carbon mineralization
Acknowledgments
We thankTracy Bartella for assistance in thefield and laboratoryWethank three anonymous reviewers for constructive comments on themanuscript
References
Aerts R 1997 Climate leaf litter chemistry and leaf litter decomposition in terrestrialecosystems a triangular relationship Oikos 439ndash449
AOAC 2012 Official Methods of Analysis 19th ed Association of Official Analytical Chem-ists Gaithersburg MD
Battle JM Golladay SW 2001 Hydroperiod influence on breakdown of leaf litter incypress-gum wetlands Am Midl Nat 146 128ndash145
Berg B McClaugherty C 2003 Plant Litter SpringerBerg B Staaf H 1980 Decomposition rate and chemical changes of Scots pine needle lit-
ter II Influence of chemical composition Ecol Bull 373ndash390Berg EE Hillman KM Dial R DeRuwe A 2009 Recent woody invasion of wetlands on
the Kenai Peninsula Lowlands south-central Alaska a major regime shift after18 000 years of wet Sphagnumndashsedge peat recruitment Can J For Res 392033ndash2046 httpdxdoiorg101139X09-121
Bremer C Braker G Matthies D Reuter A Engels C Conrad R 2007 Impact of plantfunctional group plant species and sampling time on the composition of nirK-typedenitrifier communities in soil Appl Environ Microbiol 73 6876ndash6884
Cervantes FJ Velde S Lettinga G Field JA 2000 Competition between methanogenesisand quinone respiration for ecologically important substrates in anaerobic consortiaFEMS Microbiol Ecol 34 161ndash171
Chin K-J Rainey FA Janssen PH Conrad R 1998 Methanogenic degradation of poly-saccharides and the characterization of polysaccharolytic clostridia from anoxic ricefield soil Syst Appl Microbiol 21 185ndash200
Coles JRP Yavitt JB 2004 Linking belowground carbon allocation to anaerobic CH4
and CO2 production in a forested peatland New York State Geomicrobiol J 21445ndash455 httpdxdoiorg10108001490450490505419
Cornwell WK Cornelissen JHC Amatangelo K Dorrepaal E Eviner VT Godoy OHobbie SE Hoorens B Kurokawa H Peacuterez-Harguindeguy N Quested HMSantiago LS Wardle DA Wright IJ Aerts R Allison SD Van Bodegom PBrovkin V Chatain A Callaghan TV Diacuteaz S Garnier E Gurvich DE Kazakou EKlein JA Read J Reich PB Soudzilovskaia NA Vaieretti MV Westoby M 2008Plant species traits are the predominant control on litter decomposition rates within bi-omes worldwide Ecol Lett 11 1065ndash1071 httpdxdoiorg101111j1461-0248200801219x
Cosgrove DJ 2001 Wall structure and wall loosening A look backwards and forwardsPlant Physiol 125 131ndash134
Cosgrove DJ 2005 Growth of the plant cell wall Nat Rev Mol Cell Biol 6 850ndash861httpdxdoiorg101038nrm1746
DeAngelis KM Silver WL Thompson AW Firestone MK 2010 Microbial communi-ties acclimate to recurring changes in soil redox potential status Environ Microbiol12 3137ndash3149
Delaney MT Fernandez IJ Simmons JA Briggs RD 1996 Red Maple and White Pinelitter quality initial changes with decomposition Technical Bulletin No 162 Techni-cal Bulletin University of Maine Orono ME
Dijkstra P Blankinship JC Selmants PC Hart SC Koch GW Schwartz E HungateBA 2011 Probing carbon flux patterns through soil microbial metabolic networksusing parallel position-specific tracer labeling Soil Biol Biochem 43 126ndash132
Ding SY Liu YS Zeng Y Himmel ME Baker JO Bayer EA 2012 How does plantcell wall nanoscale architecture correlate with enzymatic digestibility Science 3381055ndash1060
Drake HL Horn MA Wuumlst PK 2009 Intermediary ecosystem metabolism as a maindriver of methanogenesis in acidic wetland soil Environ Microbiol Rep 1 307ndash318
Fisher D Isachsen Y Rickard L 1970 Geologic map of New York State consisting of 5sheets Niagara Finger Lakes Hudson-Mohawk Adirondack and Lower Hudson Mapand Chart Series No 15
Fry S 1989 Analysis of cross-links in the growing cell walls of higher plants Plant FibersSpringer pp 12ndash36
Heijmans MM Knaap YA Holmgren M Limpens J 2013 Persistent versus transienttree encroachment of temperate peat bogs effects of climate warming and droughtevents Glob Chang Biol 19 2240ndash2250
Hoch G 2007 Cell wall hemicelluloses as mobile carbon stores in non-reproductiveplant tissues Funct Ecol 21 823ndash834
Jackson RB Banner JL Jobbaacutegy EG PockmanWTWall DH 2002 Ecosystem carbonloss with woody plant invasion of grasslands Nature 418 623ndash626 httpdxdoiorg101038nature00910
Jenkins CC Suberkropp K 1995 The influence of water chemistry on the enzymaticdegradation of leaves in streams Freshw Biol 33 245ndash253 httpdxdoiorg101111j1365-24271995tb01165x
Jung HJG Lamb JFS 2004 Prediction of cell wall polysaccharide and lignin concentra-tions of alfalfa stems from detergent fiber analysis Biomass Bioenergy 27 365ndash373
Klotzbuumlcher T Kaiser K Guggenberger G Gatzek C Kalbitz K 2011 A new conceptu-al model for the fate of lignin in decomposing plant litter Ecology 92 1052ndash1062
Koide K Osono T Takeda H 2005 Fungal succession and decomposition of Camellia ja-ponica leaf litter Ecol Res 20 599ndash609
McLeod AR Newsham KK Fry SC 2007 Elevated UV-B radiation modifies the ex-tractability of carbohydrates from leaf litter of Quercus robur Soil Biol Biochem 39116ndash126 httpdxdoiorg101016jsoilbio200606019
McLeod AR Fry SC Loake GJ Messenger DJ Reay DS Smith KA Yun B-W 2008Ultraviolet radiation drives methane emissions from terrestrial plant pectins NewPhytol 180 124ndash132 httpdxdoiorg101111j1469-8137200802571x
Messenger DJ Fry SC Yamulki S McLeod AR 2012 Effects of UV-B filtration on thechemistry and decomposition of Fraxinus excelsior leaves Soil Biol Biochem 47133ndash141 httpdxdoiorg101016jsoilbio201112010
Middleton BA McKee KL 2004 Use of a latitudinal gradient in bald cypress (Taxodiumdistichum) production to examine physiological controls of biotic boundaries and po-tential responses to environmental change Glob Ecol Biogeogr 13 247ndash258 httpdxdoiorg101111j1466-822X200400088x
Moore TR Trofymow JA Siltanen M Prescott C Group CW 2005 Patterns of decom-position and carbon nitrogen and phosphorus dynamics of litter in upland forest andpeatland sites in central Canada Can J For Res 35 133ndash142 httpdxdoiorg101139x04-149
Moorhead DL Sinsabaugh RL 2006 A theoretical model of litter decay and microbialinteraction Ecol Monogr 76 151ndash174 httpdxdoiorg1018900012-9615(2006)076[0151ATMOLD]20CO2
Osono T 2007 Ecology of ligninolytic fungi associated with leaf litter decompositionEcol Res 22 955ndash974 httpdxdoiorg101007s11284-007-0390-z
Pellerin S Lavoie C 2003 Recent expansion of jack pine in peatlands of southeasternQueacutebec a paleoecological study Ecoscience 10 247ndash257
Reich PB Grigal DF Aber JD Gower ST 1997 Nitrogen mineralization and produc-tivity in 50 hardwood and conifer stands on diverse soils Ecology 78 335ndash347httpdxdoiorg1018900012-9658(1997)078[0335NMAPIH]20CO2
Reith F Drake HL Kuumlsel K 2002 Anaerobic activities of bacteria and fungi in moder-ately acidic conifer and deciduous leaf litter FEMS Microbiol Ecol 41 27ndash35httpdxdoiorg101111j1574-69412002tb00963x
Renard CMGC Voragen AGJ Thibault J-F Pilnik W 1991 Studies on appleprotopectin IV apple xyloglucans and influence of pectin extraction treatments ontheir solubility Carbohydr Polym 15 387ndash403 httpdxdoiorg1010160144-8617(91)90089-U
Rydin H Jeglum JK 2013 The Biology of Peatlands 2e Oxford University PressSchink B Zeikus JG 1982 Microbial ecology of pectin decomposition in anoxic lake sedi-
ments Microbiology 128 393ndash404 httpdxdoiorg10109900221287-128-2-393Strakovaacute P Anttila J Spetz P Kitunen V Tapanila T Laiho R 2010 Litter quality
and its response to water level drawdown in boreal peatlands at plant speciesand community level Plant Soil 335 501ndash520 httpdxdoiorg101007s11104-010-0447-6
Talbot JM Treseder KK 2011 Interactions among lignin cellulose and nitrogen drive lit-ter chemistryndashdecay relationships Ecology 93 345ndash354 httpdxdoiorg10189011-08431
148 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
Theander O Westerlund E 1993 Quantitative analysis of cell wall components In JungHG Buxton DR Hatfield RD Ralph J (Eds) Forage Cell Wall Structure and Di-gestibility Am Soc Agron Madison WI pp 83ndash104
Tiedje JM Sexstone AJ Parkin TB Revsbech NP 1984 Anaerobic processes in soilPlant Soil 76 197ndash212 httpdxdoiorg101007BF02205580
Van Soest PJ 1994 Nutritional Ecology of the Ruminant Cornell University Press IthacaNY
Van Soest PJ Robertson JB Lewis BA 1991 Methods for dietary fiber neutral deter-gent fiber and nonstarch polysaccharides in relation to animal nutrition J DairySci 74 3583ndash3597
Vann CD Megonigal JP 2003 Elevated CO2 and water depth regulation of methaneemissions comparison of woody and non-woody wetland plant species Biogeo-chemistry 63 117ndash134 httpdxdoiorg101023A1023397032331
Wang WJ Dalal RC Moody PW Smith CJ 2003 Relationships of soil respiration tomicrobial biomass substrate availability and clay content Soil Biol Biochem 35273ndash284
Webster EA Halpin C Chudek JA Tilston EL Hopkins DW 2005 Decomposition insoil of soluble insoluble and lignin-rich fractions of plant material from tobacco withgenetic modifications to lignin biosynthesis Soil Biol Biochem 37 751ndash760 httpdxdoiorg101016jsoilbio200409012
Williams CJ LePage BA Vann DR Tange T Ikeda H Ando M Kusakabe T TsuzukiH Sweda T 2003 Structure allometry and biomass of plantation Metasequoia
glyptostroboides in Japan For Ecol Manag 180 287ndash301 httpdxdoiorg101016S0378-1127(02)00567-4
Wright IJ Reich PB Westoby M Ackerly DD Baruch Z Bongers F Cavender-Bares JChapin T Cornelissen JHC Diemer M Flexas J Garnier E Groom PK Gulias JHikosaka K Lamont BB Lee T Lee W Lusk C Midgley JJ Navas M-LNiinemets U Oleksyn J Osada N Poorter H Poot P Prior L Pyankov VIRoumet C Thomas SC Tjoelker MG Veneklaas EJ Villar R 2004 The worldwideleaf economics spectrum Nature 428 821ndash827 httpdxdoiorg101038nature02403
Yavitt JB Williams CJ Wieder RK 1997 Production of methane and carbon dioxide inpeatland ecosystems across North America effects of temperature aeration and or-ganic chemistry of peat Geomicrobiol J 14 299ndash316 httpdxdoiorg10108001490459709378054
Zaghloul AM Gohar AA Naiem ZA-AM Abdel Bar F 2008 Taxodione a DNA-binding compound from Taxodium distichum L (Rich) Z Naturforsch C J Biosci63 355
Zak DR Holmes WE White DC Peacock AD Tilman D 2003 Plant diversity soilmicrobial communities and ecosystem function are there any links Ecology 842042ndash2050 httpdxdoiorg10189002-0433
Table 2Plant litter type and pre-incubation concentrations of extractable cell wall fractions (car-bohydrates) expressed in g kgminus1 dry tissue
a Fractions are soluble carbohydrates (Fraction1)water soluble pectin (Fraction2) pectinin themiddle lamella (Fraction 3) pectin bound covalently to cellulose (Fraction 4) cell wallpectin (Fraction 5) extractable hemicellulose (Fraction 6) and residual non-cellulose carbo-hydrates (Fraction 7)
144 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
other litters The impact of litter addition on CH4was significantly largerfor the peat soil from Labrador Hollow (mean= 124 nmol gminus1 dayminus1)than for the swamp soil from Sapsucker Woods (49 nmol gminus1 dayminus1)(F1 54 = 2543 P = 00023) The significant effect of conifer type(F6 54 = 491 P = 00004) was evident for both soils Litter from Pinussupported the largest rates of CH4 production 187 nmol gminus1 dayminus1
when added to peat soil and 63 nmol gminus1 dayminus1 when added toswamp soil Picea litter supported the lowest CH4 production rates78 nmol gminus1 dayminus1 with peat soil 39 nmol gminus1 dayminus1 with swamp soilHence the deciduous conifers (Metasequoia Larix Taxodium) hadintermediate average rates of CH4 production 104 nmol gminus1 dayminus1
with peat soil 39 nmol gminus1 dayminus1 with swamp soil However theseaverage rates increase to 156 nmol gminus1 dayminus1 with peat soil and70 nmol gminus1 dayminus1 with swamp soil excluding Taxodium
Rates of anaerobic CO2 production were greater with litter addi-tions as opposed to control soils (Fig 1) Rates of anaerobic CO2 pro-duction were significantly greater with litter added to the peat soil(32 μmol gminus1 dayminus1) than the swamp soil (13 μmol gminus1 dayminus1)(F1 54 = 11580 P b 00001) The significant effect of conifer type(F6 54 = 263 P = 00259) was evident for both soils although dif-ferences among conifer types were larger for the peat soil Whenadded to the peat soil litter from the deciduous conifers supportedlarger rates of CO2 production (mean = 36 μmol gminus1 dayminus1) thanlitter from Pinus (mean = 33 μmol gminus1 dayminus1) or Picea (mean =31 μmol gminus1 dayminus1) Taxodium litter supported the largest ratesfor both soil types
The deciduous conifer litter had much less detergent hemicelluloseand cellulose than that in the evergreen conifers (Table 1) Thereforethe deciduous conifers had a much larger proportion of pectins andstructural proteins On average Pinus litter contained about 10 morecellulose and 14 more acid-detergent lignin than Picea litter
The sumof extractable cell wall components ranged from34 to 61of litter mass (Table 2) The largest fractions were loosely bound poly-saccharides (Fraction 1) pectin covalently bonded to cellulose (Fraction4) and extractable hemicellulose (Fraction 6) There was some varia-tion in values among members of a group of conifers For instanceTaxodium hadmore loosely bound polysaccharides and pectin covalent-ly bonded to cellulose but less extractable hemicellulose compared toMetasequoia and Larix P banksiana contained less extractable hemicel-lulose compared to the other two species of Pinus Among the sprucesP glauca was distinctive with more loosely bound polysaccharideswhereas P abies had less pectin covalently bonded to cellulose
There were considerable changes in the different fractions of cell wallcomponents during incubation (Fig 2) Loosely bound polysaccharides(Fraction 1) and pectin covalently bonded to cellulose (Fraction4) showed the largest declines The large increase inwater-soluble pectin(Fraction 2) was particularly noticeable In four of the litters (TaxodiumP banksiana P glauca and P abies) there was a large increase in the ex-tractable hemicellulose fraction (Fraction 6) during incubation Therealsowas a soil type effect on the changes in extractable carbohydrate frac-tions For example quantities of loosely bound polysaccharides (Fraction
Table 1Plant litter type and pre-incubation concentrations of detergentfiber fractions (g kgminus1 ashfree dry weight)
Litter Type Hemicellulose Cellulose Lignin Residuala
a Residual includes soluble cell-wall components including β-glucans and pectins as wellas soluble simple sugars amino acids water-soluble phenolics and insoluble protein
1) and pectin covalently bonded to cellulose exhibited greater declines inthe peat soil than in the swamp soil whereas the observed increase inwater soluble pectin (Fraction 2) also was greater in peat than in theswamp soil
We compared rates of soil CH4 production and CO2 production withinitial litter concentrations of detergent hemicellulose cellulose andacid-detergent lignin for all litter types (Table 3) We also comparedchanges in the extractable cell wall fractions (1 to 7) that occurred inall litter types during the incubations with soil CH4 production andCO2 production (Table 3) Therewas a relatively strong positive correla-tion (r ge 063 P b 01) between rates of CH4 production in both soiltypes and the initial detergent hemicellulose content of the litter Thenegative correlation (r=minus071 P b 005) between rates of CH4 produc-tion in the peat soil and pectin covalently bonded to cellulose meansthat the post-incubation increase in pectin covalently bonded to cellu-lose reduced CH4 production CO2 production in both soils was positive-ly correlated with the initial acid-detergent lignin content (rge 072 P b
005) Post incubation water-soluble pectin was positively correlated(r = 067 P b 005) with CO2 in the swamp soil
4 Discussion
Our comparative study examined several relationships between leaflitter fromnine different conifer taxa and anaerobicmicrobial activity intwo wetland soils Although our analysis is limited to correlation thedata provide new insights into plantndashsoil relationships For exampleseasonally deciduous leaves are thought to decay faster than evergreenleaves according to the Leaf Economics Spectrum (Cornwell et al2008) This is generally attributed to the seasonally deciduous leaveshaving low leaf mass per area and lower CN ratios compared to long-lived evergreen leaves However our finding of no significant differencein CO2 production for deciduous versus evergreen conifers suggests thatamore complex linkage exists between leaf litter traits and concomitantmicrobial trace gasproductionOur chemical analyses indicate 1) pectinand hemicelluloses vary among leaf litter types and their concentra-tions can change substantially in the early stages of leaf litter decompo-sition and 2) changes in pectin and hemicelluloses during leaf litterdecomposition might be fueling anaerobic microbial activity Thereforethe specific chemical attributes of the litter influence the anaerobic mi-crobial response to litter inputs
41 General soil microbial response to litter addition
Our data agree with findings of Reith et al (2002) that fresh Picealitter could promote anaerobic metabolism by soil microorganismsOverall the rates of anaerobic CO2 production in this study were typicalof rates found in other wetland soils (Yavitt et al 1997) Methane pro-duction and anaerobic CO2 productionwere greater in the peat soil than
Fig 2 Post-incubation versus pre-incubation comparison of sequentially extracted cell wall fractions (carbohydrates) of nine conifer litter types following a 24-day incubation in peat soilor swamp soil Degree of color value indicates themagnitude of changewhereas color indicates increases (green) or decreases (red) in a fraction following incubation The absolute changein concentrations of each fraction is listed within cells as grams carbohydrate per kilogram dry tissue (For interpretation of the references to color in this figure legend the reader is re-ferred to the web version of this article)
145JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
in the swamp soil This supports the general notion that rates of litterdecay are slowed by anaerobic metabolism in peat soils compared toperhaps greater aerobic metabolism in swamp soils (Moore et al2005) It is also likely that peat soils harbored a larger more active
Table 3Pearson correlation matrix for pre-incubation litter detergent fiber (cellulose hemicellu-lose lignin) concentrations and post incubation litter extractable cell wall fractions (1ndash7) versus trace gas production for each soil type
P b 001 001 b P b 005 005 b P b 01a Fractions are detergent fiber (cellulose hemicellulose lignin) and sequentially extracted
soluble carbohydrates (Fraction 1) water soluble pectin (Fraction 2) pectin in the middlelamella (Fraction 3) pectin bound covalently to cellulose (Fraction 4) cell wall pectin(Fraction 5) extractable hemicellulose (Fraction 6) and residual non-cellulose carbohy-drates (Fraction 7)
community of microorganisms capable of anaerobic metabolism Thewater-saturated peat provides a superior microbial habitat for anaero-bic activity and methanogenesis On the other hand the swamp soil isvariably saturated it ranges from being waterlogged during springsnowmelt to having a thin unsaturated surface layer during summerdry periods We sampled in the fall when a thin unsaturated layer wasapparent at the soil surface Thus our measured rates of anaerobic CO2
production in the swamp soil are consistent with the notion that facul-tative anaerobes can constitute about 10 of the aerobic microbial pop-ulation even in well-drained soil (Tiedje et al 1984) Our study alsoconfirms that the seasonally saturated conditions provide a tolerablehabitat for the persistence of methanogenic microorganisms whichare adapted to tolerate variable redox conditions (DeAngelis et al2010)
42 Leaf litter traits
In the context of the leaf economic spectrum it is logical to assumethat litter fromdeciduous coniferswould support greater rates ofmicro-bial activity than litters from evergreen Pinus or Picea This assumptionis based in part on the notion that seasonally deciduous leaves havebetter litter quality for decomposition However the prediction wasnot upheld and Pinus litter supported the greatest rates of CH4 produc-tion The microbial response to litter addition in our study reveals theuse of a simple binomial leaf trait characterization (evergreen vs decid-uous) as a predictor of soil microbial response to be lacking This findingcorroborates those of Moore et al (2005) that deciduous Larix needles
146 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
placed on the surface of peat soils decomposed slower than Pinus andPicea needles in sites in Canada
Why does the species identity of litter matter For one Taxodiumproduced themost unusual patternwith very slow rates of CH4 produc-tion but the greatest rate of CO2 production Other studies have foundthat Taxodium needles will decompose rapidly in wetlands especiallywhen exposed to alternating oxic and anoxic conditions (Battle andGolladay 2001) and thus the large CO2 production is not surprisingAlthoughwetlands dominated by Taxodium are known sources of atmo-spheric CH4 (Vann andMegonigal 2003) it is likely that fresh Taxodiumneedles have biochemical traits that slow the rate of CH4 production butnot CO2 production at the earliest stage of anaerobic decomposition Forexample Taxodium is known to produce taxoquinone and several otherditerpenoids that may have an inhibitory effect on microbes (Cervanteset al 2000 Zaghloul et al 2008) Perhaps it is not until such com-pounds degrade that Taxodium litter can be turned into methanogenicsubstrate
We also found that soilmicrobial response varied at the generic levelwith Pinus generally promoting greater CH4 than Picea litter Our findingthat P banksiana (Jack pine) litter supported CH4 production is particu-larly interesting because this species has been encroaching on peatlandsinNorth America recently (Pellerin and Lavoie 2003) Thus P banksianaencroachment may promote enhanced soil CH4 production and poten-tial emissions of atmospheric CH4 Our speculation that encroachmentof woody vegetation onto otherwise open peatlands may result inenhanced CH4 production awaits experimental confirmation Admitted-ly other environmental factors that regulate CH4 productionmay trumpthe influence of litter identity Nevertheless our results and a scarcity ofinformation in the literature suggest further study may be warranted
43 Leaf litter carbohydrate chemistry
Our analyses of conifer leaf litter quality in terms of different deter-gent fiber fractions and cell wall components in fresh litter were de-signed to determine how constituents other than cellulose and ligninmight affect microbial colonization and activity Most previous researchhas focused on cellulose because cellulose is often assumed to be themost abundant carbohydrate fraction by mass (Berg and Staaf 1980)In contrast the roles of hemicellulose and pectin in microbial respira-tion have not been studied extensively However pectin and hemicellu-lose act as the glue that holds cellulose fibers and lignin together in theplant cell wall (Cosgrove 2001 2005) Thus microbial decomposersattack compounds sequentially pectin and hemicellulose before cellu-lose then lignin Hemicellulose is a complex polymer with differentsugars and ratios of the different sugars can vary as a function of plantspecies and site fertility (Strakovaacute et al 2010) For instance Berg andMcClaugherty (2003) reported that arabinose and galactose in thehemicellulose fraction of Pinus needle litter decomposed immediatelyafter litterfall in pine forests whereas loss of mannose and xylose oc-curred after one year
Although detergent fiber analysis of plant litter has been applied inmany ecological studies of litter quality (ie results in Table 1) the se-quential extraction of cell wall components (ie results in Table 2) hasbeen used mainly in plant physiology studies (Fry 1989) mdash althoughseeMcLeod et al (2007)We quantified hemicellulose by bothmethodsand this gave conflicting results the sequential extraction method gavemuch larger values for the deciduous conifers whereas values weresimilar using both methods for the Pinus and Picea Indeed Theanderand Westerlund (1993) proposed alternatives to detergent hemicellu-lose arguing that the procedure is non-specific and might includesome pectin Jung and Lamb (2004) found that the detergent fibermethod overestimated the amount of hemicellulose in alfalfa leavesMuch more comparative work using both methods is warranted fornon-forage plants given that sequential extraction of cell wall compo-nents has the potential to providemore information about litter qualityin particular about pectin
Overall Fraction 2 (weakly bound water soluble pectin) was largerin leaf litter post-incubation than in fresh litter (Fig 2) The largeincrease in Fraction 2 across all litter types during the incubation wasunexpected as it could not be production of pectin per se Rather it like-ly represents the liberation of weakly bound sugars following incuba-tion that are then extracted in this fraction Messenger et al (2012)also found that amounts of extractable sugars increased during decom-position which they attributed to decomposition of cellulose proteinsand low-molecular-weight metabolites that are sugar-rich The anaero-bic metabolism in our study suggests another mechanism the sugarsmight be related to fermentation Low-molecular-weight organicacids derived from fermentation when in excess supply can lead togluconeogenesis (essentially reverse glycolysis) and the production ofcarbohydrates (Dijkstra et al 2011) The increase in extractable hemi-cellulose content of some of the litters (Fraction 6) also was surprisingPresumably this represents sugars that coalesce into a sodiumhydroxide-extractable fraction and is not decomposer production ofhemicellulose per se (Hoch 2007)
Fraction 4 (pectin covalently bound to cellulose) in generaldecreased in leaf litters during incubation This reflects the fracturingof plant cell walls during the earliest stage of decomposition Unfortu-nately the mechanism is not clear from the data we obtained Presum-ably fungal hyphae are involved ie filaments penetrate the plant cellwall and degrade it physically as well as chemically Clearly morework is needed to determine the fate of the liberated pectin ie is itmade soluble or metabolized to CO2 andor CH4
44 Changes in carbohydrate fractions versus microbial activity
Inmost cases positive correlations between components in leaf litterand microbial activity are interpreted as that component fueling activi-ty Likewise a negative relationship is viewed as either a recalcitrantcomponent or a component that is inhibitory tomicrobial activity How-ever our data are more complex as in some cases the negative correla-tion (low quantities of a carbohydrate fraction versus a rapid rate ofmicrobial activity) could be explained as rapid microbial metabolismcausing the loss (see below)
For Fraction 4 the straightforward positive correlationwith CH4 pro-duction suggests that slow fermentation of cell wall pectin provides aslow release of methanogenic substrate Previous studies demonstratethat the aerobic decay of pectin is linked to CH4 production (McLeodet al 2008) and thus our data extend this relationship to anaerobicme-tabolism Likewise pectin fermentation is known to fuel methanogensin anaerobic lake sediments (Schink and Zeikus 1982)
In contrast hemicellulose showed a positive correlation with CH4
production based on the detergent analysis but a negative correlationwith extractable Fraction 6 This is clearly troubling however in the se-quential extraction the earlier urea-extraction (Fraction 4) also removesstructural proteins (Fry 1989) so the negative correlation may reflectthe influence of protein and enzymes on CH4 production The sugar con-tent of the urea-extractable fraction in terms of arabinose galactosemannose xylose and others in each species is not known (cf Renardet al 1991) yet thismay also influence rates of CH4 production It is pos-sible that variation in the composition of monomers liberated fromhemicellulose may determine the supply of fermentation products tomethanogens
Rates of anaerobic CO2 production had positive relationships withacid-detergent lignin and with loosely bound water-soluble pectinBecause only a subset of microorganisms can degrade lignin completely(Osono 2007 Talbot and Treseder 2011) there is usually a negative re-lationship (Aerts 1997) especially in long-termdecomposition studiesFurthermore most of the research has been done with aerobic notanaerobic decay In contrast our correlation was with the initial acid-detergent lignin composition and we do not know how acid-detergent lignin quantity and quality changed during the incubationIt is also possible that the CO2 was not derived from lignin as such but
147JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
rather from hemicelluloses that occur intimately with acid-detergentlignin (Webster et al 2005) In contrast in soil incubated without O2pectin is a very good substrate for fermentation (Chin et al 1998) Itis likely that the relationship between pectin and anaerobic CO2 produc-tion represents the CO2 released from hydrolysis and fermentation aspectin is altered In other words pectin might be the first compoundthat microbial decomposers encounter and thus it modulates microbialactivity at the onset of decomposition asmeasured in the present study
5 Conclusions
About 6Mg haminus1 of leaves and needles fall to the ground each year inforestedwetlands (Reich et al 1997) andunder some circumstances de-ciduous conifers may produce much higher amounts (Middleton andMcKee 2004Williams et al 2003) This litter is a crucial source of energyfor soilmicroorganisms (Coles andYavitt 2004)Most studies of litter de-composition are prolonged making the first measurement of mass lossafter a few months and following the residue for one or more yearsThus we know much less about microbial colonization of fresh litter atthe very earliest stage of decomposition (Moorhead and Sinsabaugh2006) The situation is especially acute in forested wetlands where com-mon wisdom suggests slow rates of litter decay We suggest that finescale details of the decay process have been missed Hemicelluloses andpectin are probably the first molecular compounds to be released fromplant cell walls as they intertwine with lignin and cellulose and preventspecialized enzymes such as cellulases and lignases from working As aconsequence a moderate pulse of CH4 production should occur as freshlitter decays Moreover the encroachment of some types of conifersinto northern peatlands which reflects a shift in the dominant wetlandplant functional type may promote a new linkage between abovegroundcarbon fixation and belowground carbon mineralization
Acknowledgments
We thankTracy Bartella for assistance in thefield and laboratoryWethank three anonymous reviewers for constructive comments on themanuscript
References
Aerts R 1997 Climate leaf litter chemistry and leaf litter decomposition in terrestrialecosystems a triangular relationship Oikos 439ndash449
AOAC 2012 Official Methods of Analysis 19th ed Association of Official Analytical Chem-ists Gaithersburg MD
Battle JM Golladay SW 2001 Hydroperiod influence on breakdown of leaf litter incypress-gum wetlands Am Midl Nat 146 128ndash145
Berg B McClaugherty C 2003 Plant Litter SpringerBerg B Staaf H 1980 Decomposition rate and chemical changes of Scots pine needle lit-
ter II Influence of chemical composition Ecol Bull 373ndash390Berg EE Hillman KM Dial R DeRuwe A 2009 Recent woody invasion of wetlands on
the Kenai Peninsula Lowlands south-central Alaska a major regime shift after18 000 years of wet Sphagnumndashsedge peat recruitment Can J For Res 392033ndash2046 httpdxdoiorg101139X09-121
Bremer C Braker G Matthies D Reuter A Engels C Conrad R 2007 Impact of plantfunctional group plant species and sampling time on the composition of nirK-typedenitrifier communities in soil Appl Environ Microbiol 73 6876ndash6884
Cervantes FJ Velde S Lettinga G Field JA 2000 Competition between methanogenesisand quinone respiration for ecologically important substrates in anaerobic consortiaFEMS Microbiol Ecol 34 161ndash171
Chin K-J Rainey FA Janssen PH Conrad R 1998 Methanogenic degradation of poly-saccharides and the characterization of polysaccharolytic clostridia from anoxic ricefield soil Syst Appl Microbiol 21 185ndash200
Coles JRP Yavitt JB 2004 Linking belowground carbon allocation to anaerobic CH4
and CO2 production in a forested peatland New York State Geomicrobiol J 21445ndash455 httpdxdoiorg10108001490450490505419
Cornwell WK Cornelissen JHC Amatangelo K Dorrepaal E Eviner VT Godoy OHobbie SE Hoorens B Kurokawa H Peacuterez-Harguindeguy N Quested HMSantiago LS Wardle DA Wright IJ Aerts R Allison SD Van Bodegom PBrovkin V Chatain A Callaghan TV Diacuteaz S Garnier E Gurvich DE Kazakou EKlein JA Read J Reich PB Soudzilovskaia NA Vaieretti MV Westoby M 2008Plant species traits are the predominant control on litter decomposition rates within bi-omes worldwide Ecol Lett 11 1065ndash1071 httpdxdoiorg101111j1461-0248200801219x
Cosgrove DJ 2001 Wall structure and wall loosening A look backwards and forwardsPlant Physiol 125 131ndash134
Cosgrove DJ 2005 Growth of the plant cell wall Nat Rev Mol Cell Biol 6 850ndash861httpdxdoiorg101038nrm1746
DeAngelis KM Silver WL Thompson AW Firestone MK 2010 Microbial communi-ties acclimate to recurring changes in soil redox potential status Environ Microbiol12 3137ndash3149
Delaney MT Fernandez IJ Simmons JA Briggs RD 1996 Red Maple and White Pinelitter quality initial changes with decomposition Technical Bulletin No 162 Techni-cal Bulletin University of Maine Orono ME
Dijkstra P Blankinship JC Selmants PC Hart SC Koch GW Schwartz E HungateBA 2011 Probing carbon flux patterns through soil microbial metabolic networksusing parallel position-specific tracer labeling Soil Biol Biochem 43 126ndash132
Ding SY Liu YS Zeng Y Himmel ME Baker JO Bayer EA 2012 How does plantcell wall nanoscale architecture correlate with enzymatic digestibility Science 3381055ndash1060
Drake HL Horn MA Wuumlst PK 2009 Intermediary ecosystem metabolism as a maindriver of methanogenesis in acidic wetland soil Environ Microbiol Rep 1 307ndash318
Fisher D Isachsen Y Rickard L 1970 Geologic map of New York State consisting of 5sheets Niagara Finger Lakes Hudson-Mohawk Adirondack and Lower Hudson Mapand Chart Series No 15
Fry S 1989 Analysis of cross-links in the growing cell walls of higher plants Plant FibersSpringer pp 12ndash36
Heijmans MM Knaap YA Holmgren M Limpens J 2013 Persistent versus transienttree encroachment of temperate peat bogs effects of climate warming and droughtevents Glob Chang Biol 19 2240ndash2250
Hoch G 2007 Cell wall hemicelluloses as mobile carbon stores in non-reproductiveplant tissues Funct Ecol 21 823ndash834
Jackson RB Banner JL Jobbaacutegy EG PockmanWTWall DH 2002 Ecosystem carbonloss with woody plant invasion of grasslands Nature 418 623ndash626 httpdxdoiorg101038nature00910
Jenkins CC Suberkropp K 1995 The influence of water chemistry on the enzymaticdegradation of leaves in streams Freshw Biol 33 245ndash253 httpdxdoiorg101111j1365-24271995tb01165x
Jung HJG Lamb JFS 2004 Prediction of cell wall polysaccharide and lignin concentra-tions of alfalfa stems from detergent fiber analysis Biomass Bioenergy 27 365ndash373
Klotzbuumlcher T Kaiser K Guggenberger G Gatzek C Kalbitz K 2011 A new conceptu-al model for the fate of lignin in decomposing plant litter Ecology 92 1052ndash1062
Koide K Osono T Takeda H 2005 Fungal succession and decomposition of Camellia ja-ponica leaf litter Ecol Res 20 599ndash609
McLeod AR Newsham KK Fry SC 2007 Elevated UV-B radiation modifies the ex-tractability of carbohydrates from leaf litter of Quercus robur Soil Biol Biochem 39116ndash126 httpdxdoiorg101016jsoilbio200606019
McLeod AR Fry SC Loake GJ Messenger DJ Reay DS Smith KA Yun B-W 2008Ultraviolet radiation drives methane emissions from terrestrial plant pectins NewPhytol 180 124ndash132 httpdxdoiorg101111j1469-8137200802571x
Messenger DJ Fry SC Yamulki S McLeod AR 2012 Effects of UV-B filtration on thechemistry and decomposition of Fraxinus excelsior leaves Soil Biol Biochem 47133ndash141 httpdxdoiorg101016jsoilbio201112010
Middleton BA McKee KL 2004 Use of a latitudinal gradient in bald cypress (Taxodiumdistichum) production to examine physiological controls of biotic boundaries and po-tential responses to environmental change Glob Ecol Biogeogr 13 247ndash258 httpdxdoiorg101111j1466-822X200400088x
Moore TR Trofymow JA Siltanen M Prescott C Group CW 2005 Patterns of decom-position and carbon nitrogen and phosphorus dynamics of litter in upland forest andpeatland sites in central Canada Can J For Res 35 133ndash142 httpdxdoiorg101139x04-149
Moorhead DL Sinsabaugh RL 2006 A theoretical model of litter decay and microbialinteraction Ecol Monogr 76 151ndash174 httpdxdoiorg1018900012-9615(2006)076[0151ATMOLD]20CO2
Osono T 2007 Ecology of ligninolytic fungi associated with leaf litter decompositionEcol Res 22 955ndash974 httpdxdoiorg101007s11284-007-0390-z
Pellerin S Lavoie C 2003 Recent expansion of jack pine in peatlands of southeasternQueacutebec a paleoecological study Ecoscience 10 247ndash257
Reich PB Grigal DF Aber JD Gower ST 1997 Nitrogen mineralization and produc-tivity in 50 hardwood and conifer stands on diverse soils Ecology 78 335ndash347httpdxdoiorg1018900012-9658(1997)078[0335NMAPIH]20CO2
Reith F Drake HL Kuumlsel K 2002 Anaerobic activities of bacteria and fungi in moder-ately acidic conifer and deciduous leaf litter FEMS Microbiol Ecol 41 27ndash35httpdxdoiorg101111j1574-69412002tb00963x
Renard CMGC Voragen AGJ Thibault J-F Pilnik W 1991 Studies on appleprotopectin IV apple xyloglucans and influence of pectin extraction treatments ontheir solubility Carbohydr Polym 15 387ndash403 httpdxdoiorg1010160144-8617(91)90089-U
Rydin H Jeglum JK 2013 The Biology of Peatlands 2e Oxford University PressSchink B Zeikus JG 1982 Microbial ecology of pectin decomposition in anoxic lake sedi-
ments Microbiology 128 393ndash404 httpdxdoiorg10109900221287-128-2-393Strakovaacute P Anttila J Spetz P Kitunen V Tapanila T Laiho R 2010 Litter quality
and its response to water level drawdown in boreal peatlands at plant speciesand community level Plant Soil 335 501ndash520 httpdxdoiorg101007s11104-010-0447-6
Talbot JM Treseder KK 2011 Interactions among lignin cellulose and nitrogen drive lit-ter chemistryndashdecay relationships Ecology 93 345ndash354 httpdxdoiorg10189011-08431
148 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
Theander O Westerlund E 1993 Quantitative analysis of cell wall components In JungHG Buxton DR Hatfield RD Ralph J (Eds) Forage Cell Wall Structure and Di-gestibility Am Soc Agron Madison WI pp 83ndash104
Tiedje JM Sexstone AJ Parkin TB Revsbech NP 1984 Anaerobic processes in soilPlant Soil 76 197ndash212 httpdxdoiorg101007BF02205580
Van Soest PJ 1994 Nutritional Ecology of the Ruminant Cornell University Press IthacaNY
Van Soest PJ Robertson JB Lewis BA 1991 Methods for dietary fiber neutral deter-gent fiber and nonstarch polysaccharides in relation to animal nutrition J DairySci 74 3583ndash3597
Vann CD Megonigal JP 2003 Elevated CO2 and water depth regulation of methaneemissions comparison of woody and non-woody wetland plant species Biogeo-chemistry 63 117ndash134 httpdxdoiorg101023A1023397032331
Wang WJ Dalal RC Moody PW Smith CJ 2003 Relationships of soil respiration tomicrobial biomass substrate availability and clay content Soil Biol Biochem 35273ndash284
Webster EA Halpin C Chudek JA Tilston EL Hopkins DW 2005 Decomposition insoil of soluble insoluble and lignin-rich fractions of plant material from tobacco withgenetic modifications to lignin biosynthesis Soil Biol Biochem 37 751ndash760 httpdxdoiorg101016jsoilbio200409012
Williams CJ LePage BA Vann DR Tange T Ikeda H Ando M Kusakabe T TsuzukiH Sweda T 2003 Structure allometry and biomass of plantation Metasequoia
glyptostroboides in Japan For Ecol Manag 180 287ndash301 httpdxdoiorg101016S0378-1127(02)00567-4
Wright IJ Reich PB Westoby M Ackerly DD Baruch Z Bongers F Cavender-Bares JChapin T Cornelissen JHC Diemer M Flexas J Garnier E Groom PK Gulias JHikosaka K Lamont BB Lee T Lee W Lusk C Midgley JJ Navas M-LNiinemets U Oleksyn J Osada N Poorter H Poot P Prior L Pyankov VIRoumet C Thomas SC Tjoelker MG Veneklaas EJ Villar R 2004 The worldwideleaf economics spectrum Nature 428 821ndash827 httpdxdoiorg101038nature02403
Yavitt JB Williams CJ Wieder RK 1997 Production of methane and carbon dioxide inpeatland ecosystems across North America effects of temperature aeration and or-ganic chemistry of peat Geomicrobiol J 14 299ndash316 httpdxdoiorg10108001490459709378054
Zaghloul AM Gohar AA Naiem ZA-AM Abdel Bar F 2008 Taxodione a DNA-binding compound from Taxodium distichum L (Rich) Z Naturforsch C J Biosci63 355
Zak DR Holmes WE White DC Peacock AD Tilman D 2003 Plant diversity soilmicrobial communities and ecosystem function are there any links Ecology 842042ndash2050 httpdxdoiorg10189002-0433
Fig 2 Post-incubation versus pre-incubation comparison of sequentially extracted cell wall fractions (carbohydrates) of nine conifer litter types following a 24-day incubation in peat soilor swamp soil Degree of color value indicates themagnitude of changewhereas color indicates increases (green) or decreases (red) in a fraction following incubation The absolute changein concentrations of each fraction is listed within cells as grams carbohydrate per kilogram dry tissue (For interpretation of the references to color in this figure legend the reader is re-ferred to the web version of this article)
145JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
in the swamp soil This supports the general notion that rates of litterdecay are slowed by anaerobic metabolism in peat soils compared toperhaps greater aerobic metabolism in swamp soils (Moore et al2005) It is also likely that peat soils harbored a larger more active
Table 3Pearson correlation matrix for pre-incubation litter detergent fiber (cellulose hemicellu-lose lignin) concentrations and post incubation litter extractable cell wall fractions (1ndash7) versus trace gas production for each soil type
P b 001 001 b P b 005 005 b P b 01a Fractions are detergent fiber (cellulose hemicellulose lignin) and sequentially extracted
soluble carbohydrates (Fraction 1) water soluble pectin (Fraction 2) pectin in the middlelamella (Fraction 3) pectin bound covalently to cellulose (Fraction 4) cell wall pectin(Fraction 5) extractable hemicellulose (Fraction 6) and residual non-cellulose carbohy-drates (Fraction 7)
community of microorganisms capable of anaerobic metabolism Thewater-saturated peat provides a superior microbial habitat for anaero-bic activity and methanogenesis On the other hand the swamp soil isvariably saturated it ranges from being waterlogged during springsnowmelt to having a thin unsaturated surface layer during summerdry periods We sampled in the fall when a thin unsaturated layer wasapparent at the soil surface Thus our measured rates of anaerobic CO2
production in the swamp soil are consistent with the notion that facul-tative anaerobes can constitute about 10 of the aerobic microbial pop-ulation even in well-drained soil (Tiedje et al 1984) Our study alsoconfirms that the seasonally saturated conditions provide a tolerablehabitat for the persistence of methanogenic microorganisms whichare adapted to tolerate variable redox conditions (DeAngelis et al2010)
42 Leaf litter traits
In the context of the leaf economic spectrum it is logical to assumethat litter fromdeciduous coniferswould support greater rates ofmicro-bial activity than litters from evergreen Pinus or Picea This assumptionis based in part on the notion that seasonally deciduous leaves havebetter litter quality for decomposition However the prediction wasnot upheld and Pinus litter supported the greatest rates of CH4 produc-tion The microbial response to litter addition in our study reveals theuse of a simple binomial leaf trait characterization (evergreen vs decid-uous) as a predictor of soil microbial response to be lacking This findingcorroborates those of Moore et al (2005) that deciduous Larix needles
146 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
placed on the surface of peat soils decomposed slower than Pinus andPicea needles in sites in Canada
Why does the species identity of litter matter For one Taxodiumproduced themost unusual patternwith very slow rates of CH4 produc-tion but the greatest rate of CO2 production Other studies have foundthat Taxodium needles will decompose rapidly in wetlands especiallywhen exposed to alternating oxic and anoxic conditions (Battle andGolladay 2001) and thus the large CO2 production is not surprisingAlthoughwetlands dominated by Taxodium are known sources of atmo-spheric CH4 (Vann andMegonigal 2003) it is likely that fresh Taxodiumneedles have biochemical traits that slow the rate of CH4 production butnot CO2 production at the earliest stage of anaerobic decomposition Forexample Taxodium is known to produce taxoquinone and several otherditerpenoids that may have an inhibitory effect on microbes (Cervanteset al 2000 Zaghloul et al 2008) Perhaps it is not until such com-pounds degrade that Taxodium litter can be turned into methanogenicsubstrate
We also found that soilmicrobial response varied at the generic levelwith Pinus generally promoting greater CH4 than Picea litter Our findingthat P banksiana (Jack pine) litter supported CH4 production is particu-larly interesting because this species has been encroaching on peatlandsinNorth America recently (Pellerin and Lavoie 2003) Thus P banksianaencroachment may promote enhanced soil CH4 production and poten-tial emissions of atmospheric CH4 Our speculation that encroachmentof woody vegetation onto otherwise open peatlands may result inenhanced CH4 production awaits experimental confirmation Admitted-ly other environmental factors that regulate CH4 productionmay trumpthe influence of litter identity Nevertheless our results and a scarcity ofinformation in the literature suggest further study may be warranted
43 Leaf litter carbohydrate chemistry
Our analyses of conifer leaf litter quality in terms of different deter-gent fiber fractions and cell wall components in fresh litter were de-signed to determine how constituents other than cellulose and ligninmight affect microbial colonization and activity Most previous researchhas focused on cellulose because cellulose is often assumed to be themost abundant carbohydrate fraction by mass (Berg and Staaf 1980)In contrast the roles of hemicellulose and pectin in microbial respira-tion have not been studied extensively However pectin and hemicellu-lose act as the glue that holds cellulose fibers and lignin together in theplant cell wall (Cosgrove 2001 2005) Thus microbial decomposersattack compounds sequentially pectin and hemicellulose before cellu-lose then lignin Hemicellulose is a complex polymer with differentsugars and ratios of the different sugars can vary as a function of plantspecies and site fertility (Strakovaacute et al 2010) For instance Berg andMcClaugherty (2003) reported that arabinose and galactose in thehemicellulose fraction of Pinus needle litter decomposed immediatelyafter litterfall in pine forests whereas loss of mannose and xylose oc-curred after one year
Although detergent fiber analysis of plant litter has been applied inmany ecological studies of litter quality (ie results in Table 1) the se-quential extraction of cell wall components (ie results in Table 2) hasbeen used mainly in plant physiology studies (Fry 1989) mdash althoughseeMcLeod et al (2007)We quantified hemicellulose by bothmethodsand this gave conflicting results the sequential extraction method gavemuch larger values for the deciduous conifers whereas values weresimilar using both methods for the Pinus and Picea Indeed Theanderand Westerlund (1993) proposed alternatives to detergent hemicellu-lose arguing that the procedure is non-specific and might includesome pectin Jung and Lamb (2004) found that the detergent fibermethod overestimated the amount of hemicellulose in alfalfa leavesMuch more comparative work using both methods is warranted fornon-forage plants given that sequential extraction of cell wall compo-nents has the potential to providemore information about litter qualityin particular about pectin
Overall Fraction 2 (weakly bound water soluble pectin) was largerin leaf litter post-incubation than in fresh litter (Fig 2) The largeincrease in Fraction 2 across all litter types during the incubation wasunexpected as it could not be production of pectin per se Rather it like-ly represents the liberation of weakly bound sugars following incuba-tion that are then extracted in this fraction Messenger et al (2012)also found that amounts of extractable sugars increased during decom-position which they attributed to decomposition of cellulose proteinsand low-molecular-weight metabolites that are sugar-rich The anaero-bic metabolism in our study suggests another mechanism the sugarsmight be related to fermentation Low-molecular-weight organicacids derived from fermentation when in excess supply can lead togluconeogenesis (essentially reverse glycolysis) and the production ofcarbohydrates (Dijkstra et al 2011) The increase in extractable hemi-cellulose content of some of the litters (Fraction 6) also was surprisingPresumably this represents sugars that coalesce into a sodiumhydroxide-extractable fraction and is not decomposer production ofhemicellulose per se (Hoch 2007)
Fraction 4 (pectin covalently bound to cellulose) in generaldecreased in leaf litters during incubation This reflects the fracturingof plant cell walls during the earliest stage of decomposition Unfortu-nately the mechanism is not clear from the data we obtained Presum-ably fungal hyphae are involved ie filaments penetrate the plant cellwall and degrade it physically as well as chemically Clearly morework is needed to determine the fate of the liberated pectin ie is itmade soluble or metabolized to CO2 andor CH4
44 Changes in carbohydrate fractions versus microbial activity
Inmost cases positive correlations between components in leaf litterand microbial activity are interpreted as that component fueling activi-ty Likewise a negative relationship is viewed as either a recalcitrantcomponent or a component that is inhibitory tomicrobial activity How-ever our data are more complex as in some cases the negative correla-tion (low quantities of a carbohydrate fraction versus a rapid rate ofmicrobial activity) could be explained as rapid microbial metabolismcausing the loss (see below)
For Fraction 4 the straightforward positive correlationwith CH4 pro-duction suggests that slow fermentation of cell wall pectin provides aslow release of methanogenic substrate Previous studies demonstratethat the aerobic decay of pectin is linked to CH4 production (McLeodet al 2008) and thus our data extend this relationship to anaerobicme-tabolism Likewise pectin fermentation is known to fuel methanogensin anaerobic lake sediments (Schink and Zeikus 1982)
In contrast hemicellulose showed a positive correlation with CH4
production based on the detergent analysis but a negative correlationwith extractable Fraction 6 This is clearly troubling however in the se-quential extraction the earlier urea-extraction (Fraction 4) also removesstructural proteins (Fry 1989) so the negative correlation may reflectthe influence of protein and enzymes on CH4 production The sugar con-tent of the urea-extractable fraction in terms of arabinose galactosemannose xylose and others in each species is not known (cf Renardet al 1991) yet thismay also influence rates of CH4 production It is pos-sible that variation in the composition of monomers liberated fromhemicellulose may determine the supply of fermentation products tomethanogens
Rates of anaerobic CO2 production had positive relationships withacid-detergent lignin and with loosely bound water-soluble pectinBecause only a subset of microorganisms can degrade lignin completely(Osono 2007 Talbot and Treseder 2011) there is usually a negative re-lationship (Aerts 1997) especially in long-termdecomposition studiesFurthermore most of the research has been done with aerobic notanaerobic decay In contrast our correlation was with the initial acid-detergent lignin composition and we do not know how acid-detergent lignin quantity and quality changed during the incubationIt is also possible that the CO2 was not derived from lignin as such but
147JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
rather from hemicelluloses that occur intimately with acid-detergentlignin (Webster et al 2005) In contrast in soil incubated without O2pectin is a very good substrate for fermentation (Chin et al 1998) Itis likely that the relationship between pectin and anaerobic CO2 produc-tion represents the CO2 released from hydrolysis and fermentation aspectin is altered In other words pectin might be the first compoundthat microbial decomposers encounter and thus it modulates microbialactivity at the onset of decomposition asmeasured in the present study
5 Conclusions
About 6Mg haminus1 of leaves and needles fall to the ground each year inforestedwetlands (Reich et al 1997) andunder some circumstances de-ciduous conifers may produce much higher amounts (Middleton andMcKee 2004Williams et al 2003) This litter is a crucial source of energyfor soilmicroorganisms (Coles andYavitt 2004)Most studies of litter de-composition are prolonged making the first measurement of mass lossafter a few months and following the residue for one or more yearsThus we know much less about microbial colonization of fresh litter atthe very earliest stage of decomposition (Moorhead and Sinsabaugh2006) The situation is especially acute in forested wetlands where com-mon wisdom suggests slow rates of litter decay We suggest that finescale details of the decay process have been missed Hemicelluloses andpectin are probably the first molecular compounds to be released fromplant cell walls as they intertwine with lignin and cellulose and preventspecialized enzymes such as cellulases and lignases from working As aconsequence a moderate pulse of CH4 production should occur as freshlitter decays Moreover the encroachment of some types of conifersinto northern peatlands which reflects a shift in the dominant wetlandplant functional type may promote a new linkage between abovegroundcarbon fixation and belowground carbon mineralization
Acknowledgments
We thankTracy Bartella for assistance in thefield and laboratoryWethank three anonymous reviewers for constructive comments on themanuscript
References
Aerts R 1997 Climate leaf litter chemistry and leaf litter decomposition in terrestrialecosystems a triangular relationship Oikos 439ndash449
AOAC 2012 Official Methods of Analysis 19th ed Association of Official Analytical Chem-ists Gaithersburg MD
Battle JM Golladay SW 2001 Hydroperiod influence on breakdown of leaf litter incypress-gum wetlands Am Midl Nat 146 128ndash145
Berg B McClaugherty C 2003 Plant Litter SpringerBerg B Staaf H 1980 Decomposition rate and chemical changes of Scots pine needle lit-
ter II Influence of chemical composition Ecol Bull 373ndash390Berg EE Hillman KM Dial R DeRuwe A 2009 Recent woody invasion of wetlands on
the Kenai Peninsula Lowlands south-central Alaska a major regime shift after18 000 years of wet Sphagnumndashsedge peat recruitment Can J For Res 392033ndash2046 httpdxdoiorg101139X09-121
Bremer C Braker G Matthies D Reuter A Engels C Conrad R 2007 Impact of plantfunctional group plant species and sampling time on the composition of nirK-typedenitrifier communities in soil Appl Environ Microbiol 73 6876ndash6884
Cervantes FJ Velde S Lettinga G Field JA 2000 Competition between methanogenesisand quinone respiration for ecologically important substrates in anaerobic consortiaFEMS Microbiol Ecol 34 161ndash171
Chin K-J Rainey FA Janssen PH Conrad R 1998 Methanogenic degradation of poly-saccharides and the characterization of polysaccharolytic clostridia from anoxic ricefield soil Syst Appl Microbiol 21 185ndash200
Coles JRP Yavitt JB 2004 Linking belowground carbon allocation to anaerobic CH4
and CO2 production in a forested peatland New York State Geomicrobiol J 21445ndash455 httpdxdoiorg10108001490450490505419
Cornwell WK Cornelissen JHC Amatangelo K Dorrepaal E Eviner VT Godoy OHobbie SE Hoorens B Kurokawa H Peacuterez-Harguindeguy N Quested HMSantiago LS Wardle DA Wright IJ Aerts R Allison SD Van Bodegom PBrovkin V Chatain A Callaghan TV Diacuteaz S Garnier E Gurvich DE Kazakou EKlein JA Read J Reich PB Soudzilovskaia NA Vaieretti MV Westoby M 2008Plant species traits are the predominant control on litter decomposition rates within bi-omes worldwide Ecol Lett 11 1065ndash1071 httpdxdoiorg101111j1461-0248200801219x
Cosgrove DJ 2001 Wall structure and wall loosening A look backwards and forwardsPlant Physiol 125 131ndash134
Cosgrove DJ 2005 Growth of the plant cell wall Nat Rev Mol Cell Biol 6 850ndash861httpdxdoiorg101038nrm1746
DeAngelis KM Silver WL Thompson AW Firestone MK 2010 Microbial communi-ties acclimate to recurring changes in soil redox potential status Environ Microbiol12 3137ndash3149
Delaney MT Fernandez IJ Simmons JA Briggs RD 1996 Red Maple and White Pinelitter quality initial changes with decomposition Technical Bulletin No 162 Techni-cal Bulletin University of Maine Orono ME
Dijkstra P Blankinship JC Selmants PC Hart SC Koch GW Schwartz E HungateBA 2011 Probing carbon flux patterns through soil microbial metabolic networksusing parallel position-specific tracer labeling Soil Biol Biochem 43 126ndash132
Ding SY Liu YS Zeng Y Himmel ME Baker JO Bayer EA 2012 How does plantcell wall nanoscale architecture correlate with enzymatic digestibility Science 3381055ndash1060
Drake HL Horn MA Wuumlst PK 2009 Intermediary ecosystem metabolism as a maindriver of methanogenesis in acidic wetland soil Environ Microbiol Rep 1 307ndash318
Fisher D Isachsen Y Rickard L 1970 Geologic map of New York State consisting of 5sheets Niagara Finger Lakes Hudson-Mohawk Adirondack and Lower Hudson Mapand Chart Series No 15
Fry S 1989 Analysis of cross-links in the growing cell walls of higher plants Plant FibersSpringer pp 12ndash36
Heijmans MM Knaap YA Holmgren M Limpens J 2013 Persistent versus transienttree encroachment of temperate peat bogs effects of climate warming and droughtevents Glob Chang Biol 19 2240ndash2250
Hoch G 2007 Cell wall hemicelluloses as mobile carbon stores in non-reproductiveplant tissues Funct Ecol 21 823ndash834
Jackson RB Banner JL Jobbaacutegy EG PockmanWTWall DH 2002 Ecosystem carbonloss with woody plant invasion of grasslands Nature 418 623ndash626 httpdxdoiorg101038nature00910
Jenkins CC Suberkropp K 1995 The influence of water chemistry on the enzymaticdegradation of leaves in streams Freshw Biol 33 245ndash253 httpdxdoiorg101111j1365-24271995tb01165x
Jung HJG Lamb JFS 2004 Prediction of cell wall polysaccharide and lignin concentra-tions of alfalfa stems from detergent fiber analysis Biomass Bioenergy 27 365ndash373
Klotzbuumlcher T Kaiser K Guggenberger G Gatzek C Kalbitz K 2011 A new conceptu-al model for the fate of lignin in decomposing plant litter Ecology 92 1052ndash1062
Koide K Osono T Takeda H 2005 Fungal succession and decomposition of Camellia ja-ponica leaf litter Ecol Res 20 599ndash609
McLeod AR Newsham KK Fry SC 2007 Elevated UV-B radiation modifies the ex-tractability of carbohydrates from leaf litter of Quercus robur Soil Biol Biochem 39116ndash126 httpdxdoiorg101016jsoilbio200606019
McLeod AR Fry SC Loake GJ Messenger DJ Reay DS Smith KA Yun B-W 2008Ultraviolet radiation drives methane emissions from terrestrial plant pectins NewPhytol 180 124ndash132 httpdxdoiorg101111j1469-8137200802571x
Messenger DJ Fry SC Yamulki S McLeod AR 2012 Effects of UV-B filtration on thechemistry and decomposition of Fraxinus excelsior leaves Soil Biol Biochem 47133ndash141 httpdxdoiorg101016jsoilbio201112010
Middleton BA McKee KL 2004 Use of a latitudinal gradient in bald cypress (Taxodiumdistichum) production to examine physiological controls of biotic boundaries and po-tential responses to environmental change Glob Ecol Biogeogr 13 247ndash258 httpdxdoiorg101111j1466-822X200400088x
Moore TR Trofymow JA Siltanen M Prescott C Group CW 2005 Patterns of decom-position and carbon nitrogen and phosphorus dynamics of litter in upland forest andpeatland sites in central Canada Can J For Res 35 133ndash142 httpdxdoiorg101139x04-149
Moorhead DL Sinsabaugh RL 2006 A theoretical model of litter decay and microbialinteraction Ecol Monogr 76 151ndash174 httpdxdoiorg1018900012-9615(2006)076[0151ATMOLD]20CO2
Osono T 2007 Ecology of ligninolytic fungi associated with leaf litter decompositionEcol Res 22 955ndash974 httpdxdoiorg101007s11284-007-0390-z
Pellerin S Lavoie C 2003 Recent expansion of jack pine in peatlands of southeasternQueacutebec a paleoecological study Ecoscience 10 247ndash257
Reich PB Grigal DF Aber JD Gower ST 1997 Nitrogen mineralization and produc-tivity in 50 hardwood and conifer stands on diverse soils Ecology 78 335ndash347httpdxdoiorg1018900012-9658(1997)078[0335NMAPIH]20CO2
Reith F Drake HL Kuumlsel K 2002 Anaerobic activities of bacteria and fungi in moder-ately acidic conifer and deciduous leaf litter FEMS Microbiol Ecol 41 27ndash35httpdxdoiorg101111j1574-69412002tb00963x
Renard CMGC Voragen AGJ Thibault J-F Pilnik W 1991 Studies on appleprotopectin IV apple xyloglucans and influence of pectin extraction treatments ontheir solubility Carbohydr Polym 15 387ndash403 httpdxdoiorg1010160144-8617(91)90089-U
Rydin H Jeglum JK 2013 The Biology of Peatlands 2e Oxford University PressSchink B Zeikus JG 1982 Microbial ecology of pectin decomposition in anoxic lake sedi-
ments Microbiology 128 393ndash404 httpdxdoiorg10109900221287-128-2-393Strakovaacute P Anttila J Spetz P Kitunen V Tapanila T Laiho R 2010 Litter quality
and its response to water level drawdown in boreal peatlands at plant speciesand community level Plant Soil 335 501ndash520 httpdxdoiorg101007s11104-010-0447-6
Talbot JM Treseder KK 2011 Interactions among lignin cellulose and nitrogen drive lit-ter chemistryndashdecay relationships Ecology 93 345ndash354 httpdxdoiorg10189011-08431
148 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
Theander O Westerlund E 1993 Quantitative analysis of cell wall components In JungHG Buxton DR Hatfield RD Ralph J (Eds) Forage Cell Wall Structure and Di-gestibility Am Soc Agron Madison WI pp 83ndash104
Tiedje JM Sexstone AJ Parkin TB Revsbech NP 1984 Anaerobic processes in soilPlant Soil 76 197ndash212 httpdxdoiorg101007BF02205580
Van Soest PJ 1994 Nutritional Ecology of the Ruminant Cornell University Press IthacaNY
Van Soest PJ Robertson JB Lewis BA 1991 Methods for dietary fiber neutral deter-gent fiber and nonstarch polysaccharides in relation to animal nutrition J DairySci 74 3583ndash3597
Vann CD Megonigal JP 2003 Elevated CO2 and water depth regulation of methaneemissions comparison of woody and non-woody wetland plant species Biogeo-chemistry 63 117ndash134 httpdxdoiorg101023A1023397032331
Wang WJ Dalal RC Moody PW Smith CJ 2003 Relationships of soil respiration tomicrobial biomass substrate availability and clay content Soil Biol Biochem 35273ndash284
Webster EA Halpin C Chudek JA Tilston EL Hopkins DW 2005 Decomposition insoil of soluble insoluble and lignin-rich fractions of plant material from tobacco withgenetic modifications to lignin biosynthesis Soil Biol Biochem 37 751ndash760 httpdxdoiorg101016jsoilbio200409012
Williams CJ LePage BA Vann DR Tange T Ikeda H Ando M Kusakabe T TsuzukiH Sweda T 2003 Structure allometry and biomass of plantation Metasequoia
glyptostroboides in Japan For Ecol Manag 180 287ndash301 httpdxdoiorg101016S0378-1127(02)00567-4
Wright IJ Reich PB Westoby M Ackerly DD Baruch Z Bongers F Cavender-Bares JChapin T Cornelissen JHC Diemer M Flexas J Garnier E Groom PK Gulias JHikosaka K Lamont BB Lee T Lee W Lusk C Midgley JJ Navas M-LNiinemets U Oleksyn J Osada N Poorter H Poot P Prior L Pyankov VIRoumet C Thomas SC Tjoelker MG Veneklaas EJ Villar R 2004 The worldwideleaf economics spectrum Nature 428 821ndash827 httpdxdoiorg101038nature02403
Yavitt JB Williams CJ Wieder RK 1997 Production of methane and carbon dioxide inpeatland ecosystems across North America effects of temperature aeration and or-ganic chemistry of peat Geomicrobiol J 14 299ndash316 httpdxdoiorg10108001490459709378054
Zaghloul AM Gohar AA Naiem ZA-AM Abdel Bar F 2008 Taxodione a DNA-binding compound from Taxodium distichum L (Rich) Z Naturforsch C J Biosci63 355
Zak DR Holmes WE White DC Peacock AD Tilman D 2003 Plant diversity soilmicrobial communities and ecosystem function are there any links Ecology 842042ndash2050 httpdxdoiorg10189002-0433
146 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
placed on the surface of peat soils decomposed slower than Pinus andPicea needles in sites in Canada
Why does the species identity of litter matter For one Taxodiumproduced themost unusual patternwith very slow rates of CH4 produc-tion but the greatest rate of CO2 production Other studies have foundthat Taxodium needles will decompose rapidly in wetlands especiallywhen exposed to alternating oxic and anoxic conditions (Battle andGolladay 2001) and thus the large CO2 production is not surprisingAlthoughwetlands dominated by Taxodium are known sources of atmo-spheric CH4 (Vann andMegonigal 2003) it is likely that fresh Taxodiumneedles have biochemical traits that slow the rate of CH4 production butnot CO2 production at the earliest stage of anaerobic decomposition Forexample Taxodium is known to produce taxoquinone and several otherditerpenoids that may have an inhibitory effect on microbes (Cervanteset al 2000 Zaghloul et al 2008) Perhaps it is not until such com-pounds degrade that Taxodium litter can be turned into methanogenicsubstrate
We also found that soilmicrobial response varied at the generic levelwith Pinus generally promoting greater CH4 than Picea litter Our findingthat P banksiana (Jack pine) litter supported CH4 production is particu-larly interesting because this species has been encroaching on peatlandsinNorth America recently (Pellerin and Lavoie 2003) Thus P banksianaencroachment may promote enhanced soil CH4 production and poten-tial emissions of atmospheric CH4 Our speculation that encroachmentof woody vegetation onto otherwise open peatlands may result inenhanced CH4 production awaits experimental confirmation Admitted-ly other environmental factors that regulate CH4 productionmay trumpthe influence of litter identity Nevertheless our results and a scarcity ofinformation in the literature suggest further study may be warranted
43 Leaf litter carbohydrate chemistry
Our analyses of conifer leaf litter quality in terms of different deter-gent fiber fractions and cell wall components in fresh litter were de-signed to determine how constituents other than cellulose and ligninmight affect microbial colonization and activity Most previous researchhas focused on cellulose because cellulose is often assumed to be themost abundant carbohydrate fraction by mass (Berg and Staaf 1980)In contrast the roles of hemicellulose and pectin in microbial respira-tion have not been studied extensively However pectin and hemicellu-lose act as the glue that holds cellulose fibers and lignin together in theplant cell wall (Cosgrove 2001 2005) Thus microbial decomposersattack compounds sequentially pectin and hemicellulose before cellu-lose then lignin Hemicellulose is a complex polymer with differentsugars and ratios of the different sugars can vary as a function of plantspecies and site fertility (Strakovaacute et al 2010) For instance Berg andMcClaugherty (2003) reported that arabinose and galactose in thehemicellulose fraction of Pinus needle litter decomposed immediatelyafter litterfall in pine forests whereas loss of mannose and xylose oc-curred after one year
Although detergent fiber analysis of plant litter has been applied inmany ecological studies of litter quality (ie results in Table 1) the se-quential extraction of cell wall components (ie results in Table 2) hasbeen used mainly in plant physiology studies (Fry 1989) mdash althoughseeMcLeod et al (2007)We quantified hemicellulose by bothmethodsand this gave conflicting results the sequential extraction method gavemuch larger values for the deciduous conifers whereas values weresimilar using both methods for the Pinus and Picea Indeed Theanderand Westerlund (1993) proposed alternatives to detergent hemicellu-lose arguing that the procedure is non-specific and might includesome pectin Jung and Lamb (2004) found that the detergent fibermethod overestimated the amount of hemicellulose in alfalfa leavesMuch more comparative work using both methods is warranted fornon-forage plants given that sequential extraction of cell wall compo-nents has the potential to providemore information about litter qualityin particular about pectin
Overall Fraction 2 (weakly bound water soluble pectin) was largerin leaf litter post-incubation than in fresh litter (Fig 2) The largeincrease in Fraction 2 across all litter types during the incubation wasunexpected as it could not be production of pectin per se Rather it like-ly represents the liberation of weakly bound sugars following incuba-tion that are then extracted in this fraction Messenger et al (2012)also found that amounts of extractable sugars increased during decom-position which they attributed to decomposition of cellulose proteinsand low-molecular-weight metabolites that are sugar-rich The anaero-bic metabolism in our study suggests another mechanism the sugarsmight be related to fermentation Low-molecular-weight organicacids derived from fermentation when in excess supply can lead togluconeogenesis (essentially reverse glycolysis) and the production ofcarbohydrates (Dijkstra et al 2011) The increase in extractable hemi-cellulose content of some of the litters (Fraction 6) also was surprisingPresumably this represents sugars that coalesce into a sodiumhydroxide-extractable fraction and is not decomposer production ofhemicellulose per se (Hoch 2007)
Fraction 4 (pectin covalently bound to cellulose) in generaldecreased in leaf litters during incubation This reflects the fracturingof plant cell walls during the earliest stage of decomposition Unfortu-nately the mechanism is not clear from the data we obtained Presum-ably fungal hyphae are involved ie filaments penetrate the plant cellwall and degrade it physically as well as chemically Clearly morework is needed to determine the fate of the liberated pectin ie is itmade soluble or metabolized to CO2 andor CH4
44 Changes in carbohydrate fractions versus microbial activity
Inmost cases positive correlations between components in leaf litterand microbial activity are interpreted as that component fueling activi-ty Likewise a negative relationship is viewed as either a recalcitrantcomponent or a component that is inhibitory tomicrobial activity How-ever our data are more complex as in some cases the negative correla-tion (low quantities of a carbohydrate fraction versus a rapid rate ofmicrobial activity) could be explained as rapid microbial metabolismcausing the loss (see below)
For Fraction 4 the straightforward positive correlationwith CH4 pro-duction suggests that slow fermentation of cell wall pectin provides aslow release of methanogenic substrate Previous studies demonstratethat the aerobic decay of pectin is linked to CH4 production (McLeodet al 2008) and thus our data extend this relationship to anaerobicme-tabolism Likewise pectin fermentation is known to fuel methanogensin anaerobic lake sediments (Schink and Zeikus 1982)
In contrast hemicellulose showed a positive correlation with CH4
production based on the detergent analysis but a negative correlationwith extractable Fraction 6 This is clearly troubling however in the se-quential extraction the earlier urea-extraction (Fraction 4) also removesstructural proteins (Fry 1989) so the negative correlation may reflectthe influence of protein and enzymes on CH4 production The sugar con-tent of the urea-extractable fraction in terms of arabinose galactosemannose xylose and others in each species is not known (cf Renardet al 1991) yet thismay also influence rates of CH4 production It is pos-sible that variation in the composition of monomers liberated fromhemicellulose may determine the supply of fermentation products tomethanogens
Rates of anaerobic CO2 production had positive relationships withacid-detergent lignin and with loosely bound water-soluble pectinBecause only a subset of microorganisms can degrade lignin completely(Osono 2007 Talbot and Treseder 2011) there is usually a negative re-lationship (Aerts 1997) especially in long-termdecomposition studiesFurthermore most of the research has been done with aerobic notanaerobic decay In contrast our correlation was with the initial acid-detergent lignin composition and we do not know how acid-detergent lignin quantity and quality changed during the incubationIt is also possible that the CO2 was not derived from lignin as such but
147JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
rather from hemicelluloses that occur intimately with acid-detergentlignin (Webster et al 2005) In contrast in soil incubated without O2pectin is a very good substrate for fermentation (Chin et al 1998) Itis likely that the relationship between pectin and anaerobic CO2 produc-tion represents the CO2 released from hydrolysis and fermentation aspectin is altered In other words pectin might be the first compoundthat microbial decomposers encounter and thus it modulates microbialactivity at the onset of decomposition asmeasured in the present study
5 Conclusions
About 6Mg haminus1 of leaves and needles fall to the ground each year inforestedwetlands (Reich et al 1997) andunder some circumstances de-ciduous conifers may produce much higher amounts (Middleton andMcKee 2004Williams et al 2003) This litter is a crucial source of energyfor soilmicroorganisms (Coles andYavitt 2004)Most studies of litter de-composition are prolonged making the first measurement of mass lossafter a few months and following the residue for one or more yearsThus we know much less about microbial colonization of fresh litter atthe very earliest stage of decomposition (Moorhead and Sinsabaugh2006) The situation is especially acute in forested wetlands where com-mon wisdom suggests slow rates of litter decay We suggest that finescale details of the decay process have been missed Hemicelluloses andpectin are probably the first molecular compounds to be released fromplant cell walls as they intertwine with lignin and cellulose and preventspecialized enzymes such as cellulases and lignases from working As aconsequence a moderate pulse of CH4 production should occur as freshlitter decays Moreover the encroachment of some types of conifersinto northern peatlands which reflects a shift in the dominant wetlandplant functional type may promote a new linkage between abovegroundcarbon fixation and belowground carbon mineralization
Acknowledgments
We thankTracy Bartella for assistance in thefield and laboratoryWethank three anonymous reviewers for constructive comments on themanuscript
References
Aerts R 1997 Climate leaf litter chemistry and leaf litter decomposition in terrestrialecosystems a triangular relationship Oikos 439ndash449
AOAC 2012 Official Methods of Analysis 19th ed Association of Official Analytical Chem-ists Gaithersburg MD
Battle JM Golladay SW 2001 Hydroperiod influence on breakdown of leaf litter incypress-gum wetlands Am Midl Nat 146 128ndash145
Berg B McClaugherty C 2003 Plant Litter SpringerBerg B Staaf H 1980 Decomposition rate and chemical changes of Scots pine needle lit-
ter II Influence of chemical composition Ecol Bull 373ndash390Berg EE Hillman KM Dial R DeRuwe A 2009 Recent woody invasion of wetlands on
the Kenai Peninsula Lowlands south-central Alaska a major regime shift after18 000 years of wet Sphagnumndashsedge peat recruitment Can J For Res 392033ndash2046 httpdxdoiorg101139X09-121
Bremer C Braker G Matthies D Reuter A Engels C Conrad R 2007 Impact of plantfunctional group plant species and sampling time on the composition of nirK-typedenitrifier communities in soil Appl Environ Microbiol 73 6876ndash6884
Cervantes FJ Velde S Lettinga G Field JA 2000 Competition between methanogenesisand quinone respiration for ecologically important substrates in anaerobic consortiaFEMS Microbiol Ecol 34 161ndash171
Chin K-J Rainey FA Janssen PH Conrad R 1998 Methanogenic degradation of poly-saccharides and the characterization of polysaccharolytic clostridia from anoxic ricefield soil Syst Appl Microbiol 21 185ndash200
Coles JRP Yavitt JB 2004 Linking belowground carbon allocation to anaerobic CH4
and CO2 production in a forested peatland New York State Geomicrobiol J 21445ndash455 httpdxdoiorg10108001490450490505419
Cornwell WK Cornelissen JHC Amatangelo K Dorrepaal E Eviner VT Godoy OHobbie SE Hoorens B Kurokawa H Peacuterez-Harguindeguy N Quested HMSantiago LS Wardle DA Wright IJ Aerts R Allison SD Van Bodegom PBrovkin V Chatain A Callaghan TV Diacuteaz S Garnier E Gurvich DE Kazakou EKlein JA Read J Reich PB Soudzilovskaia NA Vaieretti MV Westoby M 2008Plant species traits are the predominant control on litter decomposition rates within bi-omes worldwide Ecol Lett 11 1065ndash1071 httpdxdoiorg101111j1461-0248200801219x
Cosgrove DJ 2001 Wall structure and wall loosening A look backwards and forwardsPlant Physiol 125 131ndash134
Cosgrove DJ 2005 Growth of the plant cell wall Nat Rev Mol Cell Biol 6 850ndash861httpdxdoiorg101038nrm1746
DeAngelis KM Silver WL Thompson AW Firestone MK 2010 Microbial communi-ties acclimate to recurring changes in soil redox potential status Environ Microbiol12 3137ndash3149
Delaney MT Fernandez IJ Simmons JA Briggs RD 1996 Red Maple and White Pinelitter quality initial changes with decomposition Technical Bulletin No 162 Techni-cal Bulletin University of Maine Orono ME
Dijkstra P Blankinship JC Selmants PC Hart SC Koch GW Schwartz E HungateBA 2011 Probing carbon flux patterns through soil microbial metabolic networksusing parallel position-specific tracer labeling Soil Biol Biochem 43 126ndash132
Ding SY Liu YS Zeng Y Himmel ME Baker JO Bayer EA 2012 How does plantcell wall nanoscale architecture correlate with enzymatic digestibility Science 3381055ndash1060
Drake HL Horn MA Wuumlst PK 2009 Intermediary ecosystem metabolism as a maindriver of methanogenesis in acidic wetland soil Environ Microbiol Rep 1 307ndash318
Fisher D Isachsen Y Rickard L 1970 Geologic map of New York State consisting of 5sheets Niagara Finger Lakes Hudson-Mohawk Adirondack and Lower Hudson Mapand Chart Series No 15
Fry S 1989 Analysis of cross-links in the growing cell walls of higher plants Plant FibersSpringer pp 12ndash36
Heijmans MM Knaap YA Holmgren M Limpens J 2013 Persistent versus transienttree encroachment of temperate peat bogs effects of climate warming and droughtevents Glob Chang Biol 19 2240ndash2250
Hoch G 2007 Cell wall hemicelluloses as mobile carbon stores in non-reproductiveplant tissues Funct Ecol 21 823ndash834
Jackson RB Banner JL Jobbaacutegy EG PockmanWTWall DH 2002 Ecosystem carbonloss with woody plant invasion of grasslands Nature 418 623ndash626 httpdxdoiorg101038nature00910
Jenkins CC Suberkropp K 1995 The influence of water chemistry on the enzymaticdegradation of leaves in streams Freshw Biol 33 245ndash253 httpdxdoiorg101111j1365-24271995tb01165x
Jung HJG Lamb JFS 2004 Prediction of cell wall polysaccharide and lignin concentra-tions of alfalfa stems from detergent fiber analysis Biomass Bioenergy 27 365ndash373
Klotzbuumlcher T Kaiser K Guggenberger G Gatzek C Kalbitz K 2011 A new conceptu-al model for the fate of lignin in decomposing plant litter Ecology 92 1052ndash1062
Koide K Osono T Takeda H 2005 Fungal succession and decomposition of Camellia ja-ponica leaf litter Ecol Res 20 599ndash609
McLeod AR Newsham KK Fry SC 2007 Elevated UV-B radiation modifies the ex-tractability of carbohydrates from leaf litter of Quercus robur Soil Biol Biochem 39116ndash126 httpdxdoiorg101016jsoilbio200606019
McLeod AR Fry SC Loake GJ Messenger DJ Reay DS Smith KA Yun B-W 2008Ultraviolet radiation drives methane emissions from terrestrial plant pectins NewPhytol 180 124ndash132 httpdxdoiorg101111j1469-8137200802571x
Messenger DJ Fry SC Yamulki S McLeod AR 2012 Effects of UV-B filtration on thechemistry and decomposition of Fraxinus excelsior leaves Soil Biol Biochem 47133ndash141 httpdxdoiorg101016jsoilbio201112010
Middleton BA McKee KL 2004 Use of a latitudinal gradient in bald cypress (Taxodiumdistichum) production to examine physiological controls of biotic boundaries and po-tential responses to environmental change Glob Ecol Biogeogr 13 247ndash258 httpdxdoiorg101111j1466-822X200400088x
Moore TR Trofymow JA Siltanen M Prescott C Group CW 2005 Patterns of decom-position and carbon nitrogen and phosphorus dynamics of litter in upland forest andpeatland sites in central Canada Can J For Res 35 133ndash142 httpdxdoiorg101139x04-149
Moorhead DL Sinsabaugh RL 2006 A theoretical model of litter decay and microbialinteraction Ecol Monogr 76 151ndash174 httpdxdoiorg1018900012-9615(2006)076[0151ATMOLD]20CO2
Osono T 2007 Ecology of ligninolytic fungi associated with leaf litter decompositionEcol Res 22 955ndash974 httpdxdoiorg101007s11284-007-0390-z
Pellerin S Lavoie C 2003 Recent expansion of jack pine in peatlands of southeasternQueacutebec a paleoecological study Ecoscience 10 247ndash257
Reich PB Grigal DF Aber JD Gower ST 1997 Nitrogen mineralization and produc-tivity in 50 hardwood and conifer stands on diverse soils Ecology 78 335ndash347httpdxdoiorg1018900012-9658(1997)078[0335NMAPIH]20CO2
Reith F Drake HL Kuumlsel K 2002 Anaerobic activities of bacteria and fungi in moder-ately acidic conifer and deciduous leaf litter FEMS Microbiol Ecol 41 27ndash35httpdxdoiorg101111j1574-69412002tb00963x
Renard CMGC Voragen AGJ Thibault J-F Pilnik W 1991 Studies on appleprotopectin IV apple xyloglucans and influence of pectin extraction treatments ontheir solubility Carbohydr Polym 15 387ndash403 httpdxdoiorg1010160144-8617(91)90089-U
Rydin H Jeglum JK 2013 The Biology of Peatlands 2e Oxford University PressSchink B Zeikus JG 1982 Microbial ecology of pectin decomposition in anoxic lake sedi-
ments Microbiology 128 393ndash404 httpdxdoiorg10109900221287-128-2-393Strakovaacute P Anttila J Spetz P Kitunen V Tapanila T Laiho R 2010 Litter quality
and its response to water level drawdown in boreal peatlands at plant speciesand community level Plant Soil 335 501ndash520 httpdxdoiorg101007s11104-010-0447-6
Talbot JM Treseder KK 2011 Interactions among lignin cellulose and nitrogen drive lit-ter chemistryndashdecay relationships Ecology 93 345ndash354 httpdxdoiorg10189011-08431
148 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
Theander O Westerlund E 1993 Quantitative analysis of cell wall components In JungHG Buxton DR Hatfield RD Ralph J (Eds) Forage Cell Wall Structure and Di-gestibility Am Soc Agron Madison WI pp 83ndash104
Tiedje JM Sexstone AJ Parkin TB Revsbech NP 1984 Anaerobic processes in soilPlant Soil 76 197ndash212 httpdxdoiorg101007BF02205580
Van Soest PJ 1994 Nutritional Ecology of the Ruminant Cornell University Press IthacaNY
Van Soest PJ Robertson JB Lewis BA 1991 Methods for dietary fiber neutral deter-gent fiber and nonstarch polysaccharides in relation to animal nutrition J DairySci 74 3583ndash3597
Vann CD Megonigal JP 2003 Elevated CO2 and water depth regulation of methaneemissions comparison of woody and non-woody wetland plant species Biogeo-chemistry 63 117ndash134 httpdxdoiorg101023A1023397032331
Wang WJ Dalal RC Moody PW Smith CJ 2003 Relationships of soil respiration tomicrobial biomass substrate availability and clay content Soil Biol Biochem 35273ndash284
Webster EA Halpin C Chudek JA Tilston EL Hopkins DW 2005 Decomposition insoil of soluble insoluble and lignin-rich fractions of plant material from tobacco withgenetic modifications to lignin biosynthesis Soil Biol Biochem 37 751ndash760 httpdxdoiorg101016jsoilbio200409012
Williams CJ LePage BA Vann DR Tange T Ikeda H Ando M Kusakabe T TsuzukiH Sweda T 2003 Structure allometry and biomass of plantation Metasequoia
glyptostroboides in Japan For Ecol Manag 180 287ndash301 httpdxdoiorg101016S0378-1127(02)00567-4
Wright IJ Reich PB Westoby M Ackerly DD Baruch Z Bongers F Cavender-Bares JChapin T Cornelissen JHC Diemer M Flexas J Garnier E Groom PK Gulias JHikosaka K Lamont BB Lee T Lee W Lusk C Midgley JJ Navas M-LNiinemets U Oleksyn J Osada N Poorter H Poot P Prior L Pyankov VIRoumet C Thomas SC Tjoelker MG Veneklaas EJ Villar R 2004 The worldwideleaf economics spectrum Nature 428 821ndash827 httpdxdoiorg101038nature02403
Yavitt JB Williams CJ Wieder RK 1997 Production of methane and carbon dioxide inpeatland ecosystems across North America effects of temperature aeration and or-ganic chemistry of peat Geomicrobiol J 14 299ndash316 httpdxdoiorg10108001490459709378054
Zaghloul AM Gohar AA Naiem ZA-AM Abdel Bar F 2008 Taxodione a DNA-binding compound from Taxodium distichum L (Rich) Z Naturforsch C J Biosci63 355
Zak DR Holmes WE White DC Peacock AD Tilman D 2003 Plant diversity soilmicrobial communities and ecosystem function are there any links Ecology 842042ndash2050 httpdxdoiorg10189002-0433
147JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
rather from hemicelluloses that occur intimately with acid-detergentlignin (Webster et al 2005) In contrast in soil incubated without O2pectin is a very good substrate for fermentation (Chin et al 1998) Itis likely that the relationship between pectin and anaerobic CO2 produc-tion represents the CO2 released from hydrolysis and fermentation aspectin is altered In other words pectin might be the first compoundthat microbial decomposers encounter and thus it modulates microbialactivity at the onset of decomposition asmeasured in the present study
5 Conclusions
About 6Mg haminus1 of leaves and needles fall to the ground each year inforestedwetlands (Reich et al 1997) andunder some circumstances de-ciduous conifers may produce much higher amounts (Middleton andMcKee 2004Williams et al 2003) This litter is a crucial source of energyfor soilmicroorganisms (Coles andYavitt 2004)Most studies of litter de-composition are prolonged making the first measurement of mass lossafter a few months and following the residue for one or more yearsThus we know much less about microbial colonization of fresh litter atthe very earliest stage of decomposition (Moorhead and Sinsabaugh2006) The situation is especially acute in forested wetlands where com-mon wisdom suggests slow rates of litter decay We suggest that finescale details of the decay process have been missed Hemicelluloses andpectin are probably the first molecular compounds to be released fromplant cell walls as they intertwine with lignin and cellulose and preventspecialized enzymes such as cellulases and lignases from working As aconsequence a moderate pulse of CH4 production should occur as freshlitter decays Moreover the encroachment of some types of conifersinto northern peatlands which reflects a shift in the dominant wetlandplant functional type may promote a new linkage between abovegroundcarbon fixation and belowground carbon mineralization
Acknowledgments
We thankTracy Bartella for assistance in thefield and laboratoryWethank three anonymous reviewers for constructive comments on themanuscript
References
Aerts R 1997 Climate leaf litter chemistry and leaf litter decomposition in terrestrialecosystems a triangular relationship Oikos 439ndash449
AOAC 2012 Official Methods of Analysis 19th ed Association of Official Analytical Chem-ists Gaithersburg MD
Battle JM Golladay SW 2001 Hydroperiod influence on breakdown of leaf litter incypress-gum wetlands Am Midl Nat 146 128ndash145
Berg B McClaugherty C 2003 Plant Litter SpringerBerg B Staaf H 1980 Decomposition rate and chemical changes of Scots pine needle lit-
ter II Influence of chemical composition Ecol Bull 373ndash390Berg EE Hillman KM Dial R DeRuwe A 2009 Recent woody invasion of wetlands on
the Kenai Peninsula Lowlands south-central Alaska a major regime shift after18 000 years of wet Sphagnumndashsedge peat recruitment Can J For Res 392033ndash2046 httpdxdoiorg101139X09-121
Bremer C Braker G Matthies D Reuter A Engels C Conrad R 2007 Impact of plantfunctional group plant species and sampling time on the composition of nirK-typedenitrifier communities in soil Appl Environ Microbiol 73 6876ndash6884
Cervantes FJ Velde S Lettinga G Field JA 2000 Competition between methanogenesisand quinone respiration for ecologically important substrates in anaerobic consortiaFEMS Microbiol Ecol 34 161ndash171
Chin K-J Rainey FA Janssen PH Conrad R 1998 Methanogenic degradation of poly-saccharides and the characterization of polysaccharolytic clostridia from anoxic ricefield soil Syst Appl Microbiol 21 185ndash200
Coles JRP Yavitt JB 2004 Linking belowground carbon allocation to anaerobic CH4
and CO2 production in a forested peatland New York State Geomicrobiol J 21445ndash455 httpdxdoiorg10108001490450490505419
Cornwell WK Cornelissen JHC Amatangelo K Dorrepaal E Eviner VT Godoy OHobbie SE Hoorens B Kurokawa H Peacuterez-Harguindeguy N Quested HMSantiago LS Wardle DA Wright IJ Aerts R Allison SD Van Bodegom PBrovkin V Chatain A Callaghan TV Diacuteaz S Garnier E Gurvich DE Kazakou EKlein JA Read J Reich PB Soudzilovskaia NA Vaieretti MV Westoby M 2008Plant species traits are the predominant control on litter decomposition rates within bi-omes worldwide Ecol Lett 11 1065ndash1071 httpdxdoiorg101111j1461-0248200801219x
Cosgrove DJ 2001 Wall structure and wall loosening A look backwards and forwardsPlant Physiol 125 131ndash134
Cosgrove DJ 2005 Growth of the plant cell wall Nat Rev Mol Cell Biol 6 850ndash861httpdxdoiorg101038nrm1746
DeAngelis KM Silver WL Thompson AW Firestone MK 2010 Microbial communi-ties acclimate to recurring changes in soil redox potential status Environ Microbiol12 3137ndash3149
Delaney MT Fernandez IJ Simmons JA Briggs RD 1996 Red Maple and White Pinelitter quality initial changes with decomposition Technical Bulletin No 162 Techni-cal Bulletin University of Maine Orono ME
Dijkstra P Blankinship JC Selmants PC Hart SC Koch GW Schwartz E HungateBA 2011 Probing carbon flux patterns through soil microbial metabolic networksusing parallel position-specific tracer labeling Soil Biol Biochem 43 126ndash132
Ding SY Liu YS Zeng Y Himmel ME Baker JO Bayer EA 2012 How does plantcell wall nanoscale architecture correlate with enzymatic digestibility Science 3381055ndash1060
Drake HL Horn MA Wuumlst PK 2009 Intermediary ecosystem metabolism as a maindriver of methanogenesis in acidic wetland soil Environ Microbiol Rep 1 307ndash318
Fisher D Isachsen Y Rickard L 1970 Geologic map of New York State consisting of 5sheets Niagara Finger Lakes Hudson-Mohawk Adirondack and Lower Hudson Mapand Chart Series No 15
Fry S 1989 Analysis of cross-links in the growing cell walls of higher plants Plant FibersSpringer pp 12ndash36
Heijmans MM Knaap YA Holmgren M Limpens J 2013 Persistent versus transienttree encroachment of temperate peat bogs effects of climate warming and droughtevents Glob Chang Biol 19 2240ndash2250
Hoch G 2007 Cell wall hemicelluloses as mobile carbon stores in non-reproductiveplant tissues Funct Ecol 21 823ndash834
Jackson RB Banner JL Jobbaacutegy EG PockmanWTWall DH 2002 Ecosystem carbonloss with woody plant invasion of grasslands Nature 418 623ndash626 httpdxdoiorg101038nature00910
Jenkins CC Suberkropp K 1995 The influence of water chemistry on the enzymaticdegradation of leaves in streams Freshw Biol 33 245ndash253 httpdxdoiorg101111j1365-24271995tb01165x
Jung HJG Lamb JFS 2004 Prediction of cell wall polysaccharide and lignin concentra-tions of alfalfa stems from detergent fiber analysis Biomass Bioenergy 27 365ndash373
Klotzbuumlcher T Kaiser K Guggenberger G Gatzek C Kalbitz K 2011 A new conceptu-al model for the fate of lignin in decomposing plant litter Ecology 92 1052ndash1062
Koide K Osono T Takeda H 2005 Fungal succession and decomposition of Camellia ja-ponica leaf litter Ecol Res 20 599ndash609
McLeod AR Newsham KK Fry SC 2007 Elevated UV-B radiation modifies the ex-tractability of carbohydrates from leaf litter of Quercus robur Soil Biol Biochem 39116ndash126 httpdxdoiorg101016jsoilbio200606019
McLeod AR Fry SC Loake GJ Messenger DJ Reay DS Smith KA Yun B-W 2008Ultraviolet radiation drives methane emissions from terrestrial plant pectins NewPhytol 180 124ndash132 httpdxdoiorg101111j1469-8137200802571x
Messenger DJ Fry SC Yamulki S McLeod AR 2012 Effects of UV-B filtration on thechemistry and decomposition of Fraxinus excelsior leaves Soil Biol Biochem 47133ndash141 httpdxdoiorg101016jsoilbio201112010
Middleton BA McKee KL 2004 Use of a latitudinal gradient in bald cypress (Taxodiumdistichum) production to examine physiological controls of biotic boundaries and po-tential responses to environmental change Glob Ecol Biogeogr 13 247ndash258 httpdxdoiorg101111j1466-822X200400088x
Moore TR Trofymow JA Siltanen M Prescott C Group CW 2005 Patterns of decom-position and carbon nitrogen and phosphorus dynamics of litter in upland forest andpeatland sites in central Canada Can J For Res 35 133ndash142 httpdxdoiorg101139x04-149
Moorhead DL Sinsabaugh RL 2006 A theoretical model of litter decay and microbialinteraction Ecol Monogr 76 151ndash174 httpdxdoiorg1018900012-9615(2006)076[0151ATMOLD]20CO2
Osono T 2007 Ecology of ligninolytic fungi associated with leaf litter decompositionEcol Res 22 955ndash974 httpdxdoiorg101007s11284-007-0390-z
Pellerin S Lavoie C 2003 Recent expansion of jack pine in peatlands of southeasternQueacutebec a paleoecological study Ecoscience 10 247ndash257
Reich PB Grigal DF Aber JD Gower ST 1997 Nitrogen mineralization and produc-tivity in 50 hardwood and conifer stands on diverse soils Ecology 78 335ndash347httpdxdoiorg1018900012-9658(1997)078[0335NMAPIH]20CO2
Reith F Drake HL Kuumlsel K 2002 Anaerobic activities of bacteria and fungi in moder-ately acidic conifer and deciduous leaf litter FEMS Microbiol Ecol 41 27ndash35httpdxdoiorg101111j1574-69412002tb00963x
Renard CMGC Voragen AGJ Thibault J-F Pilnik W 1991 Studies on appleprotopectin IV apple xyloglucans and influence of pectin extraction treatments ontheir solubility Carbohydr Polym 15 387ndash403 httpdxdoiorg1010160144-8617(91)90089-U
Rydin H Jeglum JK 2013 The Biology of Peatlands 2e Oxford University PressSchink B Zeikus JG 1982 Microbial ecology of pectin decomposition in anoxic lake sedi-
ments Microbiology 128 393ndash404 httpdxdoiorg10109900221287-128-2-393Strakovaacute P Anttila J Spetz P Kitunen V Tapanila T Laiho R 2010 Litter quality
and its response to water level drawdown in boreal peatlands at plant speciesand community level Plant Soil 335 501ndash520 httpdxdoiorg101007s11104-010-0447-6
Talbot JM Treseder KK 2011 Interactions among lignin cellulose and nitrogen drive lit-ter chemistryndashdecay relationships Ecology 93 345ndash354 httpdxdoiorg10189011-08431
148 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
Theander O Westerlund E 1993 Quantitative analysis of cell wall components In JungHG Buxton DR Hatfield RD Ralph J (Eds) Forage Cell Wall Structure and Di-gestibility Am Soc Agron Madison WI pp 83ndash104
Tiedje JM Sexstone AJ Parkin TB Revsbech NP 1984 Anaerobic processes in soilPlant Soil 76 197ndash212 httpdxdoiorg101007BF02205580
Van Soest PJ 1994 Nutritional Ecology of the Ruminant Cornell University Press IthacaNY
Van Soest PJ Robertson JB Lewis BA 1991 Methods for dietary fiber neutral deter-gent fiber and nonstarch polysaccharides in relation to animal nutrition J DairySci 74 3583ndash3597
Vann CD Megonigal JP 2003 Elevated CO2 and water depth regulation of methaneemissions comparison of woody and non-woody wetland plant species Biogeo-chemistry 63 117ndash134 httpdxdoiorg101023A1023397032331
Wang WJ Dalal RC Moody PW Smith CJ 2003 Relationships of soil respiration tomicrobial biomass substrate availability and clay content Soil Biol Biochem 35273ndash284
Webster EA Halpin C Chudek JA Tilston EL Hopkins DW 2005 Decomposition insoil of soluble insoluble and lignin-rich fractions of plant material from tobacco withgenetic modifications to lignin biosynthesis Soil Biol Biochem 37 751ndash760 httpdxdoiorg101016jsoilbio200409012
Williams CJ LePage BA Vann DR Tange T Ikeda H Ando M Kusakabe T TsuzukiH Sweda T 2003 Structure allometry and biomass of plantation Metasequoia
glyptostroboides in Japan For Ecol Manag 180 287ndash301 httpdxdoiorg101016S0378-1127(02)00567-4
Wright IJ Reich PB Westoby M Ackerly DD Baruch Z Bongers F Cavender-Bares JChapin T Cornelissen JHC Diemer M Flexas J Garnier E Groom PK Gulias JHikosaka K Lamont BB Lee T Lee W Lusk C Midgley JJ Navas M-LNiinemets U Oleksyn J Osada N Poorter H Poot P Prior L Pyankov VIRoumet C Thomas SC Tjoelker MG Veneklaas EJ Villar R 2004 The worldwideleaf economics spectrum Nature 428 821ndash827 httpdxdoiorg101038nature02403
Yavitt JB Williams CJ Wieder RK 1997 Production of methane and carbon dioxide inpeatland ecosystems across North America effects of temperature aeration and or-ganic chemistry of peat Geomicrobiol J 14 299ndash316 httpdxdoiorg10108001490459709378054
Zaghloul AM Gohar AA Naiem ZA-AM Abdel Bar F 2008 Taxodione a DNA-binding compound from Taxodium distichum L (Rich) Z Naturforsch C J Biosci63 355
Zak DR Holmes WE White DC Peacock AD Tilman D 2003 Plant diversity soilmicrobial communities and ecosystem function are there any links Ecology 842042ndash2050 httpdxdoiorg10189002-0433
148 JB Yavitt CJ Williams Geoderma 243ndash244 (2015) 141ndash148
Theander O Westerlund E 1993 Quantitative analysis of cell wall components In JungHG Buxton DR Hatfield RD Ralph J (Eds) Forage Cell Wall Structure and Di-gestibility Am Soc Agron Madison WI pp 83ndash104
Tiedje JM Sexstone AJ Parkin TB Revsbech NP 1984 Anaerobic processes in soilPlant Soil 76 197ndash212 httpdxdoiorg101007BF02205580
Van Soest PJ 1994 Nutritional Ecology of the Ruminant Cornell University Press IthacaNY
Van Soest PJ Robertson JB Lewis BA 1991 Methods for dietary fiber neutral deter-gent fiber and nonstarch polysaccharides in relation to animal nutrition J DairySci 74 3583ndash3597
Vann CD Megonigal JP 2003 Elevated CO2 and water depth regulation of methaneemissions comparison of woody and non-woody wetland plant species Biogeo-chemistry 63 117ndash134 httpdxdoiorg101023A1023397032331
Wang WJ Dalal RC Moody PW Smith CJ 2003 Relationships of soil respiration tomicrobial biomass substrate availability and clay content Soil Biol Biochem 35273ndash284
Webster EA Halpin C Chudek JA Tilston EL Hopkins DW 2005 Decomposition insoil of soluble insoluble and lignin-rich fractions of plant material from tobacco withgenetic modifications to lignin biosynthesis Soil Biol Biochem 37 751ndash760 httpdxdoiorg101016jsoilbio200409012
Williams CJ LePage BA Vann DR Tange T Ikeda H Ando M Kusakabe T TsuzukiH Sweda T 2003 Structure allometry and biomass of plantation Metasequoia
glyptostroboides in Japan For Ecol Manag 180 287ndash301 httpdxdoiorg101016S0378-1127(02)00567-4
Wright IJ Reich PB Westoby M Ackerly DD Baruch Z Bongers F Cavender-Bares JChapin T Cornelissen JHC Diemer M Flexas J Garnier E Groom PK Gulias JHikosaka K Lamont BB Lee T Lee W Lusk C Midgley JJ Navas M-LNiinemets U Oleksyn J Osada N Poorter H Poot P Prior L Pyankov VIRoumet C Thomas SC Tjoelker MG Veneklaas EJ Villar R 2004 The worldwideleaf economics spectrum Nature 428 821ndash827 httpdxdoiorg101038nature02403
Yavitt JB Williams CJ Wieder RK 1997 Production of methane and carbon dioxide inpeatland ecosystems across North America effects of temperature aeration and or-ganic chemistry of peat Geomicrobiol J 14 299ndash316 httpdxdoiorg10108001490459709378054
Zaghloul AM Gohar AA Naiem ZA-AM Abdel Bar F 2008 Taxodione a DNA-binding compound from Taxodium distichum L (Rich) Z Naturforsch C J Biosci63 355
Zak DR Holmes WE White DC Peacock AD Tilman D 2003 Plant diversity soilmicrobial communities and ecosystem function are there any links Ecology 842042ndash2050 httpdxdoiorg10189002-0433
