dissolved organic carbon in alaskan boreal forest: sources ......dissolved organic carbon in alaskan...

Ecosystems (2007) DOI: 10.1007/s10021·007·9101·4

Dissolved Organic Carbon inAlaskan Boreal Forest: Sources,Chemical Characteristics, and

Biodegradability

Kimberly P. Wickland,1,* Jason C. Neff,2 and George R. Aiken1

1us Geological Survey, 3215 Marine St., Boulder, Colorado 80303,USA; 2Department of Geological Sciences, University of Colorado,Boulder, Colorado 80303,USA

ABSTRACT

The fate of terrestrially-derived dissolved organiccarbon (DOC) is important to carbon (C) cycling inboth terrestrial and aquatic environments, and re-cent evidence suggests that climate warming isinfluencing DOC dynamics in northern ecosystems.To understand what determines the fate of terres-trial DOC, it is essential to quantify the chemicalnature and potential biodegradability of this DOC.We examined DOC chemical characteristics andbiodegradability collected from soil pore waters anddominant vegetation species in four boreal blackspruce forest sites in Alaska spanning a range ofhydrologic regimes and permafrost extents (WellDrained, Moderately Well Drained, Poorly Drained,and Thermokarst Wetlands). DOC chemistry wascharacterized using fractionation, UV-Vis absor-bance, and fluorescence measurements. Potentialbiodegradability was assessed by incubating thesamples and measuring CO2 production over1month. Soil pore water DOC from all sites wasdominated by hydrophobic acids and was highlyaromatic, whereas the chemical composition ofvegetation leachate DOC varied significantly withspecies. There was no seasonal variability in soil

pore water DOC chemical characteristics or biode-gradability; however, DOC collected from thePoorly Drained site was significantly less biode-gradable than DOC from the other three sites (6%loss vs. 13-15% loss). The biodegradability of veg-etation-derived DOC ranged from 10 to 90% loss,and was strongly correlated with hydrophilic DOCcontent. Vegetation such as Sphagnum moss andfeathermosses yielded DOC that was quicklymetabolized and respired. In contrast, the DOCleached from vegetation such as black spruce wasmoderately recalcitrant. Changes in DOC chemicalcharacteristics that occurred during microbialmetabolism of DOC were quantified using frac-tionation and fluorescence. The chemical charac-teristics and biodegradability of DOC in soil porewaters were most similar to the moderately re-calcitrant vegetation leachates, and to the micro-bially altered DOC from all vegetation leachates.

Key words: dissolved organic carbon; decompo-sition; fluorescence; boreal forest; Alaska; blackspruce; thermokarst.

Received 16 May 2007; accepted 21 September 2007*Corrcsponding author: e-mail: [email protected]

INTRODUCTION

Dissolved organic carbon (DOC) cycling in north-ern terrestrial ecosystems, especially in areas af-fected by permafrost, is of particular interest in light

esippText BoxThis file was created by scanning the printed publication. Text errors identified by the software have been corrected: however some errors may remain.

K. P. Wickland and others

of changing climate at northern latitudes. The largeamounts of organic carbon (C) in biomass and soils,combined with relatively slow decomposition rates,provides a potentially large DOC source, whilepermafrost prevents deep percolation into soils soDOC can be efficiently transported to surface wa-ters. Changes in temperature and permafrostdepths are likely impacting DOC production, pro-cessing, and transport in northern terrestrial eco-systems (Frey and Smith 2005; Striegl and others2005). Warmer temperatures may be stimulatingDOC production and/or consumption rates byinfluencing microbial activity, and thawing per-mafrost is likely altering hydrologic flow paths,leading to increased residence time of DOC in soils.To understand the mechanisms that underlie thispotentially important C flux in high-latitude eco-systems, it is essential to: (1) know the chemicalcharacteristics of the terrestrially-derived DOC, (2)quantify the potential biodegradability of the DOC,and (3) understand environmental controls onDOC chemistry and biodegradability.

DOC is a complex mixture of low and highmolecular weight compounds that originates fromvegetation, litter, soil leachates. plant root exu-dates, and microbial enzymes and biomass (Thur-man 1985; Guggenberger and Zech 1994). DOCpools and fluxes in terrestrial ecosystems may besmall compared to other terrestrial C pools andfluxes (Neff and Asner 2001; Moore 2003), butthese pools and fluxes have important roles in bothterrestrial and aquatic C cycles. DOC is a substratefor microbial activity (McArthur and others 1985;Amon and Benner 1996), it can be a readily sta-bilized form of C through sorption to mineral soils(McDowell and Wood Neff and Asner 2001),and it serves as a link between terrestrial andaquatic systems (Dalva and Moore 1991; Schiff andothers 1998; Cole and others 2007). Microbialmetabolism of DOC is affected by environmentalconditions such as temperature (Christ and David 1996)

and oxygen availability (Bastviken and oth-ers 2004), and by the chemical structure of DOCmolecules (Qualls and Haines 1992). The fluxes of terrestrially-derived DOC within and between soilsand aquatic systems are highly dependent on thesource strength and the amount of water movingthrough soils to surface waters (Hope and others

1994; Aitkenhead and McDowell as well ason the amount of microbial metabolism and sorp-tion the DOC undergoes during transport (Kaw -higashi and others 2004).

To increase our understanding of terrestrial DOCcycling in northern ecosystems, we characterizedthe chemical nature and the potential biodegrad-

ability of DOC collected from terrestrial sources ininterior Alaska. We focused on black spruce [ (Piceamariana Mill.) B.S.P.] forest, which constitutes amajor ecosystem type in the North American borealforest (Van Cleve and Dyrness 1988; Hall andothers 1997), We chose black spruce systems be-cause they are most commonly found on perma-frost soils, but also exist in areas withoutpermafrost; they have large amounts of soil organicC (SOC) and, thus, are a potentially large DOCsource; and they are found in areas with soils thatrange from poorly to well drained. Their commonassociation with permafrost makes many poorlydrained black spruce forests particularly vulnerableto effects of climate warming, as permafrost thawcan cause changes in drainage conditions (Osterk-amp and others 2000; Camill and others 2001;Wickland and others 2006) and increase theamount of SOC in the active layer as it deepens.Current and future climate warming may increasesoil biological activity and decomposition (Shaverand others 1992; Oechel and others 1993; Gouldenand others 1998), and potentially increase DOCrelease from peatlands (Frey and Smith 2005),depending on hydrologic connectivity with surfacewaters (Striegl and others 2005).

Groundcover vegetation species anel tree pro-ductivity in black spruce systems are influenced bydrainage conditions and permafrost. In welldrained systems, trees are more productive andlight penetration through the canopy is limited,creating ideal conditions for feathermosses and li-chens to dominate the understory vegetation (Bis-bee and others 2001 ). In poorly drained systemsthe trees are slow growing and sparse, and Sphaq-num mosses thrive in the high light conditionswhen moisture is not limiting (Bisbee and others

2001). Moderately drained systems have interme- diate tree productivity and the understory is dominated by feathermosses. To represent the current range of conditions in which black spruce systems exist, we chose three sites in central Alaska that included a well drained forest without permafrost, a moderately drained forest having permafrost, and a poorly drained forest having permafrost. Within the poorly drained site there are several thermokarst collapse wetlands that have formed due to localized permafrost thaw.

We collected DOC from soil pore waters several times annually during multiple years to capture a seasonal and site variability of DOC chemical characteristics and biodegradability. We also leached DOC from dominant vegetation for chemical characterization and biodegradability. Our objec- tives were to: (1) compare the chemical charac-

Dissolved organic carbon in Alaskan boreal forest

teristics of DOC from black spruce forest sites hav-ing differences in permafrost, drainage, and vege-ration cover, (2) quantify the potentialbiodegradability of DOC from a range of blackspruce forest soils and vegetation, and (3) examinethe influence of DOC chemical characteristics onpotential biodegradability.

SITE DESCRIPTION

The study sites are located in central Alaska (Fig-ure 1) in regions underlain by discontinuous per-mafrost. The sites are all populated by black spruce,but vary in drainage, permafrost depth, organiclayer thickness, stand density, and dominantgroundcover vegetation (Table 1). Two sites arelocated in the Donnelly Flats area near DeltaJunction, AK (63°53'N, 145°44'W) and two sitesare in the Tanana River floodplain near Fairbanks,AK (64°42'n 148°19'W). Soils in the Donnelly flatsarea are mainly derived from Donnelly moraineand wind-blown loess (O'Neill 2000), and areunderlain by deposits of sand and gravel. A "WellDrained" site and a "Moderately Well Drained" siteare located here in mature black spruce forest (80-100 years-old) (Marries and others 2001, 2004). Inearly spring, water collects in the surface soils ontop of the seasonal ice layer. As the seasonal icelayer thaws, water drains into deeper soils by earlyJune at the Well Drained site, whereas the presenceof permafrost prevents water from completelydraining at the Moderately Well Drained site. The

"Poorly Drained" site near Fairbanks is a matureblack spruce stand (up to 130 years-old). Perma-frost and underlying alluvial material prevent ver-tical movement of water. and the water table iswithin 35-40 cm of the surface during the entiresnow-free season. Isolated thermokarst featureshave formed within the Poorly Drained site wherepermafrost has thawed ("Thermokarst Wetlands"site). In these areas the ground surface is 0.5-1 mlower than the surrounding forest there is standingwater in many places, and there are standing deadblack spruce and tamarack trees. No permafrost ispresent to a depth of at least 2.2 m in the Ther-mokarst Wetlands (see Wickland and othersfor detailed site description).

METHODS

Soil Pore Water Collection and AnalysesSoil pore water samples were collected withintransects ranging from 40 to 100 m in length (onetransect at Poorly Drained, two transects each atModerately Well Drained and Well Drained). Fiveto ten sampling locations were distributed amongeach transect spaced 10-20 m apart. In addition,there were five sampling locations distributedwithin the Thermokarst Wetlands site. The depth ofavailable water varied with site and season, as thewater table was most often located immediatelyabove the ice layer. In addition, local spatial het-erogeneity of the water table at each site resulted invarying sample collection locations within the


transects during the study. A stainless steel probe(3 mm i.d.) slotted at the bottom 2 cm was insertedinto the soil and water was drawn through theprobe using a peristaltic pump attached to the endof the probe. Samples ranging in volume from 1 to2 L were filtered in the field through GelmanAquaPrep 600TM capsule filters (0.45 um) intopre-baked amber glass bottles with Teflon-linedcaps (the first 200 mL of pore water were used torinse the filters and discarded). The samples werekept on ice during transport to the lab and refrig-erated until analysis for DOC concentration, UVabsorbance, fluorescence, and DOC fractionation.

Soil pore water samples were collected duringthe following time periods:

Well Drained = July 2003, April and May 2004,and May 2005, ranging from 0.5 to 17 cm depth;

Moderately Well Drained = September 2002, May-October 2003, May-September 2004, and May2005, ranging from 0.5 to 34 cm depth;

Poorly Drained = July-September 2002 and 2003,April-September 2004, and May 2005, rangingfrom 2 to 43 cm depth;

Thermokarst Wetlands = July-September 2002and 2003, and March-September 2004, rangingfrom 6 to 80 cm depth.

Vegetation LeachatesWe collected samples of representative vegetationfrom each site for leaching, including P. mariananeedles (brown needles still on trees) and barkand twigs from the Well Drained and Poorly

Drained sites; Hylocomium splendens (Hedw.) B.S.G.(stair-step moss) and Pleurorium schreberii (Brid.)Mitt. (red-stemmed Ieathermoss) from the Mod-erately Well Drained site (live, clipped; hereafterreferred to as "mixed feathermoss"); Sphagnumanqustifolium (Russow) C. Jens (live, clipped) andEriophorum anqustifolium Honck. (Jive, clipped)from the Thermokarst Wetlands; and Betula natiaL. leaves (senesced leaves still on branches) fromthe Poorly Drained site. We also collected Betulapapyrifera Marsh. leaves (senesced, recently drop-ped) from a stand next to the Poorly Drained site,as this tree species is commonly found near blackspruce forest. Soluble organic carbon was obtainedby leaching air-dried vegetation. Ten to fifty gramsof vegetation (dry weight) was combined with 5-9 L of 0.001 N NaHCO3 solution in 9 L clear pyrexjugs covered with foil. The NaHCO3, was used tobuffer pH and to mimic ionic strength of naturalsystems. The samples were aerated continuouslyusing fish tank pumps and sintered glass tubes toprevent anoxia. The vegetation samples wereleached for 7-14 days, during which the solutionswere periodically sampled for DOC concentration,UV absorbance, fluorescence, and DOC fraction-ation. Replicate Sphagnum and feathermoss-mixvegetation samples were subjected to 3-monthlong leaching periods to examine changes inleachate DOC chemical characteristics with lime.The volumes removed during sampling (50-500 mL) were replaced with an equal volume ofdilute NaHCO3 solution. Samples were immedi-ately filtered through 0.45-um Supor syringe fil-


ters or 0.45-um capsule filters depending onsample volume.

DOC AnalysesPore water samples were analyzed for DOC con-centration within 2 weeks of collection, and vege-tation leachate samples were analyzed within oneday of collection using an O.I. Analytical Model700 TOC Analyzer via the platinum catalyzed per-sulfate wet oxidation method (Aiken and others1992). The instrument was calibrated using aminimum of live standards spanning the range ofsample concentrations, and standards and sampleswere run in duplicate (instrument std. dev. ± 0.2mg C L-1). Samples were diluted with deionized(DI) water by weight prior to DOC analysis to fallwithin the optimum range of the TOC analyzer

UV-Vis Absorption AnalysesSamples were analyzed for UV-Vis absorption usinga Hewlett-Packard Model 8453 photo-diode arrayspectrophotometer (^ = 200-800 nm) and a l=cmpath-length quartz cell. DI water was used as aninstrument blank. Samples were diluted by weightwith DI to be within the range of the instrument.Results are reported for absorption at ^ = 254 nm.the wavelength associated with aromatic com-pounds (Chin and others 1994). The standarddeviation for a UV measurement at 254 nm is±O.002 AU. Specific UV absorbance (SUVA) of DOCgives an "average" molar absorptivity for all themolecules contributing to the DOC in a sample, andit has been used as a measure of DOC aromaticity(Chin and others 1994; Weishaar and others 2003).SUVA was determined by dividing UV-Vis absor-bance at ^ = 254 nm by DOC concentration, whereeach variable was measured at the same dilution

DOM FluorescenceTo further characterize the optically active portionof the dissolved organic matter (DOM), we mea-sured 3-dimensional fluorescence of a subset of soilpore waters and vegetation leachates using a Jobin-Yvon Horiba Fluoromax-3TM fluorometer. These3-D fluorescence intensities are referred to asExcitation-Emission Matrices (EEMs). Beforeanalysis, an aliquot of each sample was allowed to warm to room temperature and diluted, whennecessary, to remain within the range of thedetector. UV-Vis was measured after dilution, as

above. EEMs were collected over an excitationrange of 240-450 nm every 5 nm. and an emissionrange of 300-600 nm every 2 nm. A series of cor-rections were made to the EEMs to ensure thatthey were comparable among samples. DI waterEEMs, which served as blanks, were collected dailyand subtracted from each sample EEM. The blank-subtracted EEMs were then corrected for instru-ment biases using instrument-specific excitationand emission corrections files provided by themanufacturer. The EEMs were then Raman-nor-malized using the area under the Raman scatterpeak (350 nm excitation wavelength) obtainedfrom the corresponding DI water blank. We thencorrected EEMs for inner filter effects using theUV-Vis absorbance spectra. This correction ac-counts for the absorption of excitation and emis-sion light by the sample (Ohno 2002). The resultingcorrected EEMs were plotted using MatLab with 30contour lines, after normalizing fluorescenceintensities to DOC concentration. The location ofmaximum fluorescence intensity (Fmax) for eachEEM was determined on regions unaffected byfirst- and second-order Rayleigh scattering (Ingleand Crouch l988).

Two indices based on fluorescence spectra, thefluorescence index (FI) and the humification in-dex (HIX), were determined from the EEMs. Thefluorescence index is calculated as the ratio of theintensities at excitation (ex) and emission (em)wavelengths ex370/em450 and ex370/em500, andhas been used to distinguish between microbially-derived (FI = 1.7-2.0) and terrestrially-derived(FI < 1.4) aquatic fulvic acids (McKnight andothers 2001). The humification index is based onthe suggestion that as decomposition, or humifi-cation, of fluorescing molecules proceeds, theiremission spectra will shift towards longer wave-lengths due to lower molecular H:C ratios (Zsol-nay and others 1999). Therefore the sum offluorescence intensities at long wavelengths canbe compared to the intensities at shorter wave-lengths to quantify the relative degree of humifi-cation. We calculated the humification index(HIX) from the DOC normalized fluorescencevalues as:

s the sum of the fluorescence

intensity at emission wavlengths

2002). The HTX values range from 0

to 1, with higher values indicating an increasing degree of humification (Ohno 2002).


Parallel factor analysis (PARAFAC), a statisticalmodeling technique, was applied to the correctedEEMS to identify fluorescing components accord-ing to their unique excitation and emission patterns(Stedmon and others 2003; Cory and McKnight2005). We quantified the relative contribution ofthirteen different fluorescing components previ-ously identified by Cory and McKnight (2005).Seven of these components (Q1-Q3, SQ1-SQ3 andHQ) are identified as quinone-like molecules,which vary in redox state and degree of conjuga-tion. Tyrosine and tryptophan are identified asprotein-like fluorphores, and the remaining fourcomponents are not associated with any particularmolecule class (Cory and McKnight

DOC FractionationWe used a resin-based method of DOC fraction-ation as a further means to characterize DOCchemistry. Soil pore water and vegetation leachatesamples were chromatographically separated intofive different fractions (hydrophobic acids, hydro-phobic neutrals, transphilic neutrals, hydrophilicorganic matter, and transphilic acids) using Am-berlite XAD-8 and XAD-4 resins (Aiken and others1992). The resins preferentially sorb different clas-ses of organic acids based on aqueous solubility ofthe solute, chemical composition of the resin, resinsurface area, and resin pore size. The amount oforganic matter within each fraction was calculatedusing the DOC concentration and the sample massof each fraction, and are presented as percentagesof total DOC. UV-Vis absorption was run onhydrophobic acids (HPOA), hydrophilic organicmatter (HP1), and transphilic acid (TPIA) fractions.Fluorescence was run on HPOA and HPI fractionsfrom soil pore water collected in May 2005. Thefractions were brought to neutral pH prior tofluorescence analysis using NaOH. A select numberof samples were fractionated in duplicate, and theaverage values are presented. The standard devia-tion for the mass percentages of the fractionationanalysis was ±2%.

DOC IncubationsPotential DOC biodegradation was determined byincubating samples in sealed serum bottles at 22°Cand measuring CO2 production in the headspaceover 28-31 days. Vegetation leachates were sub-sampled after 1 week of leaching for incubation.All samples were diluted to approximately 10 mgC L-I with DI water to prevent excessive microbialgrowth, and 50 mL aliquots were dispensed into100 mL amber glass serum bottles (12 replicate

bottles per sample). An inoculant was prepared bymixing soil from each site with DI water, filtering aportion of the solution through a 1.6 um glass fiberfilter, and sequentially diluting the filtrate for a10-3 serial dilution. One mL of dilute soil solutionwas added to each serum bottle. The DOC contentof the dilute soil solution was below detectionlimits. One-eighth of a pre-baked (400°C for 4 h)glass fiber filter was added to each bottle to providea surface for microbial establishment, and the bot-tles were sealed using butyl rubber stoppers.Within 2 h of sealing the serum bottles, we ana-lyzed CO2 in the equilibrated headspace of theserum bottles by withdrawing 0.5-mL aliquots ofheadspace (four reps per bottle) and injecting theminto a nitrogen carrier stream passing through aLicor 6252 infrared CO2 analyzer. The mean of thefour injections was used to calculate headspace CO2concentration. Calibration curves were createdusing a minimum of three standards. On the firstand last days of the incubations, three replicatebottles per sample were acidified with 2 mL 42.5%,H3PO4 and the headspace was analyzed for dis-solved inorganic carbon (DIC) on the Licor CO2analyzer to ensure that all carbonate species wereaccounted for. The remaining replicate serum bot-tles were analyzed for headspace CO2 concentra-tion on Days 1, 3, 5,7, 14,21, and 28. The volumeof headspace removed for analysis was replaced byan equal volume of N2 after each sampling. Be-tween analyses the bottles were placed upside-down on a shaker set at 100 rpm and covered withfoil. On days 14 and 28 three replicate bottles persample Were opened and analyzed for DOC con-centration, SUVA, and fluorescence after head-space analysis. We calculated dissolved CO2concentrations (headspace + aqueous) from the labanalyses and known CO2 equilibrium constants(Plummer and Busenberg 1982) adjusted forambient temperature and pressure (Striegl andothers 2001).

Total CO2 production over the incubation periodwas quantified as the difference between initial andfinal DIC concentrations, as this accounts for anychanges in pH during the incubation, which wouldalter the ratio of dissolved versus gaseous CO2, Thechange in DIC was assumed to originate completelyfrom respiration of DOC, thus equaling DOC con-sumption. Biodegradation is expressed as % DOCmineralized (mg C-DIC produced/initial mg c-DOC).

The decomposition constants for each DOCsample were determined from the time series ofcumulative DIC production, which we calculatedby multiplying the CO2 concentrations by the ratio


of total change in DIC: total change in CO2,assuming a constant pH (Kawahigashi and others2004). We filled a single exponential model (Eq. 2)and a double exponential model (Eq. 3) to theincubation results assuming one or two distinctDOC pools (Kalbitz and others 2003):

and "late" (September-October) prior to statisticalanalyses. This allowed us to test for site by seasoneffects, and to compare the Well Drained site,where soil pore water was present only during theearly season, with the other sites. Two-factor AN-OVA without replication was used to test PARA-FAC results for significant differences betweensamples (p < 0.05).

RESULTS

Soil Pore Water DOC Concentrations andChemical CharacteristicsSoil pore water DOC concentrations and fractionswere not significantly different between sites orseasons, so we present the mean values here (Ta-ble 2). Although spatial and temporal variabilitywere insignificant, general trends among the siteswere evident. DOC concentrations tended to begreatest at the wettest sites (Poorly Drained andThermokarst Wetlands) and lowest at the WellDrained site. The SUVA values were high at all sites,indicating high aromatic content. The dominantDOC fraction at all sites was HPOA, accounting formore than 50% of the total DOC. The next largestfractions were HPI and TPIA, whereas the neutralfraction (hydrophobic plus transphilic neutrals) wasconsistently the smallest fraction at all sites.

We measured fluorescence on whole water DOCand on the HPOA and HPI fractions from onesample date (May 2005) for Well Drained, Moder-ately Well Drained, and Poorly Drained. The fluo-rescence signatures of samples from each site werevery similar, and we show one example in Figure 2.(Moderately Well Drained, whole water DOC). Thewhole waters and DOC fractions all displayed

Statistical AnalysesWe used repeated-measures ANOYA followed byUnique N HSD post-hoc analyses (p < 0.05; Statis-tica 7.0) where appropriate to test for significantdifferences in soil pore water DOC chemical char-acreristics and DOC incubation results betweensites (p values are reported when significant). Wepartitioned the samples into seasons, designated"early" (March-May), "middle" (June-August),

where t = time (days), kI = decomposition rateconstant of "stable" DOC (slowly mineralizableDOC pool, day-1, a = the portion of the total DOCpool that is stable (%), k2 = decomposition rateconstant of "labile" DOC (rapidly mineralizableDOC pool, day -1). The curves were fitted using aleast-squares regression (Levenberg-Marquardtmethod) in Staristica 7.0. We fitted both models toall sample incubations and determined whichmodel best described each sample.


maximum fluorescence intensities (Fmax) in a re-gion that is associated with fulvic acid fluorophores(ex < 250/em440-504; Stedmon and Markager

2005) (Table 3). PARAFAC analyses reveal similartrends in the dominant fluorescing components ofthe DOC in each sample (Table 3). The dominantfluorophore in the whole water DOC and HPOAfraction samples was Component 4, identified as ahydroquinone (HQ) by Cory and McKnight (2005,whereas Component 2, a terrestrially-derived qui-none (Q2; Cory and McKnight 2005), was thegreatest contributor to the HPI fractions. There isno statistical difference in the relative abundance ofthe fluorphores for the whole water DOC or theDOC fractions between sites.

The FI values for the soil pore water DOC sam-ples range from 1.14 to 1.51 (Table 3), and areconsistently lower for the whole water samples andHPOA fractions than the HPI fractions. The HIXvalues of the samples from the different sites arerelatively similar (Table 3) and suggest that overallthe DOC is highly humified. The HPI fractions fromeach site have consistently lower HIX values thanthe HPOA fractions, although differences were notstatistically significant, possibly due to the smallsample number.

Vegetation Leachate DOC Yields andChemical CharacteristicsIn contrast to the soil pore waters, the vegetationleachates had large ranges in DOC yields (mg C

tics (Table 4). Betula papyrijera leaves yielded thehighest amount of DOC, whereas the lowest DOCyields were from the mixed feathermoss and P.mariana bark and twigs. In addition to differencesin the magnitude of DOC yields, there were dif-ferences in the timing of maximum DOC yields(Table 4 Figure 3A). P. mariana and E. angustifoli-um litter leached increasing amounts of DOCthroughout the leaching period, whereas DOCconcentrations of the two moss and B. papyriferaleachates actually decreased after about 24 h (Fig-ure 3A). SUVA values varied widely (Table 4). butall the vegetation leachates were less aromatic thansoil pore water DOC (Table 2). There was a largerange in the relative amounts of DOC fractionsamong the leachates (Table 4. Figure 3B), althoughall the leachates had higher % HPI and lower %HPOA than the soil pore water DOC.

Fluorescent properties of leachate DOC varied inintensity and in dominant regions (Table 4, Fig-ures 4, 5). The P. mariana needles and B. papyriferaleaves leachates had the greatest Fmax, whereas theSphagnum leachate had the lowest. The P. marianabark and twigs and the E. anqustifolium leachareshad Fmax values in the same region as the soil porewaters Fmax values. All of the other leachates hadFmax values at ex280/em310-350, which is a regioncharacterized by the protein-like fluorophorestryptophan and tyrosine (Coble 1996; Baker 2002;Jaffe and others 2004).

The relative proportions of the different DOCfractions were dynamic during the course ofleaching, particularly the HPOA and HPI frac-tions. The ratio of HPOA:HPI increased with timeto varying degrees in all the leachates (Fig-ure3B), but for different reasons depending onthe vegetation. Shifts in HPOA:HPI were attrib-utable to net increases in HPOA in P. marianavegetation leachates. and to net decreases in HPIin the moss leachate, and the B. papyriferaleachate. The increases in HPOA:HPI in the E.anqustifolium and B. nana leachates were due toroughly equal net increases in HPOA and netdecreases in HPI over time.

During the longer-term leaching of Sphagnumand mixed feathermoss, there were distinct chan-ges in DOC fluorescence intensities and regions(Figure 5). Quantifiable changes in fluorescingcomponents, as identified by PARAFAC, and the FIand HIX values over time in the leachates indicateprominent shifts in the DOC chemical characteris-tics (Table 5). After 24 h of leaching Compouent 8,which corresponds to tryptophan (Cory andMcKnight 2005) was the most abundant fluoro-

9CI1V1

S


phore in both leachates. After 3 months, the rela-tive contribution of tryptophan decreased by 83-97 %. The other amino acid-like fluorophore.Component 13 or tyrosine, also decreased in rela-tive abundance in both leachates with time. Theinitial rates of disappearance of tryptophan andtyrosine (24-194 h) were greater in the Sphaqnumleachate than the feathermoss leachate. Compo-nent 4 (hydroquinone) was the dominant fluor-phore at 194 h and 3 months. The componentsthat increased the most in relative contribution tofluorescence were Components 7 and 9 (semiqui-nones) in both leachates. and Component 6(unidentified) in the feathermoss and Component12 (a quinone) in the Sphagnum leachates. Con-current with these changes, the FI and HIX valuesincreased over time (Table 5).

DOC Incubations of Soil Pore Water andVegetation LeachatesIn contrast to the soil pore water DOC chemicalcharacteristics, DOC biodegradability exhibitednoticeable trends both among sites and, in somecases, with season. The DOC collected from thePoorly Drained site was the least biodegradable,followed by the Well Drained site and Thermok-arst Wetlands, whereas DOC from the ModeratelyWell Drained site was the most biodegradablewhen considering % mineralized DOC (Table 6).We present the means of the incubation resultshere. DOC decomposition was best described by asingle pool model for all soil pore waters. Thedecomposition constants (kI) of the soil porewaters followed a similar pattern as the % min-eralized, with Poorly Drained DOC having theslowest kl, followed by Well Drained, Thermok-arst Wetlands, and Moderately Well DrainedDOC. The half-life of DOC is almost twice as longfor the Poorly Drained site than for the WellDrained and Thermokarst Wetlands sites, andmore than 2.5 times that for the Moderately WellDrained site. When we compared samples basedon season, the half-life of Poorly Drained-DOCwas significantly longer than Moderately WellDrained-DOC during the middle and late seasons(p < 0.01 and p < 0.05, respectively). In addition,the half-life of Poorly Drained-DOC was signifi-cantly longer than Thermokarst Wetlands-DOCduring the middle season (p < 0.001). Seasonalvariation in DOC biodegradability was significantonly at the Poorly Drained site, where DOC half-life was approximately twice as long during themiddle season as during the early and late sea-sons (p < 0.05).

best described using a two pool model. SphagnumDOC had the fastest decomposition constants forthe labile and stable DOC pools, followed closelyby mixed feathermoss DOC for the stable pool.The P. mariana leachates had the slowestdecomposition constants, and were similar tothose calculated for soil pore water DOC.

Potential DOC biodegradability of all soil porewaters and vegetation leachates combined waspositively associated with initial % HPI content(Figure 6, r2 = 0.82).

The vegetation leachates DOC had a widerange in potential biodegradability, ranging fromabout 11%) mineralization for P. mariana leach-ates to about 90% mineralization for mossleachates (Table 6). When calculating DOC bio-degradation, we included any net decreases inDOC concentration that we measured during thefirst week of leaching where appropriate (seeFigure 3A). DOC decomposition was best de-scribed by a single pool model for all leachatesexcept Sphaqnum and B. papyrifera, which were


Potential biodegradability was negatively associ- water DOC biodegradability alone was not signifi-ated with % HPOA, but to a lesser degree cantly correlated with % HPI, % HPOA, or any(r

2 = 0.50; not shown). The variability in soil pore other measured DOC chemistry parameter.


generally slow to degrade, whereas DOC leachedfrom representative vegetation contained poten-tially large amounts of hydrophilic compounds andranged from slow to fast biodegradability. Soil porewater DOC chemistry was not reflective of vege-tation leachate DOC chemistry probably becausethe highly biodegradable compounds were quicklyconsumed and respired as CO2 and/or CH4, leavingthe more recalcitrant compounds to accumulate inthe soil solution (van Hees and others 2005). .Forestsoils characterized by poor drainage and shallowpermafrost contained the least labile pore waterDOC, suggesting that the large DOC pools in thesesystems have undergone significant microbial pro-cessing, and that relatively recalcitrant DOC accu-mulates as a result of slow hydrologic transport outof the system.

Soil Pore Water DOC ChemicalCharacteristics and BiodegradabilitySoil pore water DOC concentrations were generallyhighest in the wettest sites (Thermokarst Wet-lands > Poorly Drained > Moderately WellDrained > Well Drained), which is likely a result of

DISCUSSION

DOC collected from black spruce forest soils rang-ing in drainage and permafrost regime was domi-nated by hydrophobic compounds and was


differences in organic soil layer thickness, wateravailability, redox potential, and water residencetimes. Organic layer thickness increases by almosttenfold from the Well Drained to the ThermokarstWetlands sites, which has two implications for DOCconcentrations. First, there is a much larger po-tential DOC source in the sites that have more soilorganic C (Hope and others 1997). Second, siteswith thicker O horizons have comparatively deepmineral soil horizons, so pore water DOC must betransported further to sorb to mineral soils. Shallowpermafrost further isolates DOC from mineral soils,particularly at the Poorly Drained site, by blockingvertical transport. Differences inwater availabilityamong the sites may affect DOC concentrations byinfluencing DOC production (Christ and David1996), and by increasing the potential for OCcompound dissolution. Free soil pore water ispresent at varying depths during the entire snow-free season at the Thermokarst Wetlands andPoorly Drained sites, and is often present at theModerately Well Drained site. However, free soilpore water is rarely present at the Well Drained siteexcept during spring snowmelt before the seasonalice layer disappears. Finally, poor drainage condi-tions likely result in longer residence times of wa-ter, and thus of DOC, at the Thermokarst Wetlandsand Poorly Drained sites. Hongve ( 1999) suggestedthat high DOC concentrations found in poorlydrained peat are due primarily to long waterretention times.

Soil pore water DOC collected from all the siteswas highly aromatic and dominated by the HPOAfraction (Table 2'), which is consistent with otherstudies of soil pore waters (Guggenberger and Zechl994); Cronan and Aiken 1985; McKnight andothers 1985); Qualls and Haines 1991; Micbaelsonand others 1998; Smolander and Kitunen 2002).Fluorescence analyses further highlight the simi-larity in DOC chemical characteristics among thesites (Table 3). Fluorescence analyses of the DOCfractions indicate that there may be differences insources and degree of humification of the differentfractions (McKnight and others 2001; Ohno 2002).In particular, the FI and HIX values suggest that insoil pore waters the HPOA fraction is strongly de-rived from terrestrial plant sources and hasundergone extensive microbial processing, whereasthe HPI fraction is a mixture of terrestrially- andmicrobially-derived compounds and is relativelyless humified. Comparison of the HIX values of theHPI fractions from the three sites suggest that thisfraction is more degraded in Poorly Drained-DOC.

The range in decomposition rate constants (kl) ofsoil pore water DOC was similar to values mea-

sured for DOC from other terrestrial ecosystems(Kalbitz and others 2003; McDowell and others

2006), The relatively low potential biodegradabilityof soil pore water DOC is consistent with the gen-erally low hydrophilic content, as that is regardedas the most biodegradable fraction (Qualls andHaines 1992; Jandl and Sollins 1997; Qualls 2005).Among the sites, Poorly Drained-DOC was the leastbiodegradable, and had a significantly longer half-life than Moderately Well Drained- and Thermok-arst Wetlands-DOC. The differences in biodegrad-ability are likely due to a combination of DOCresidence time (that is, age of DOC), extent of priormicrobial processing, and of source material. Veg-etation may account for much of the difference inDOC biodegradability between the Poorly Drainedand Thermokarst Wetlands sites (Table 1). Theprimary vegetation at the Thermokarst Wetlandssite is Sphagnum moss, a source of bighly labile DOCas shown in this study. The Poorly Drained site hascomparatively less Sphagnum moss, and it has moreblack spruce trees, which are a source of more re-calcitrant DOC (this study). Low oxygen availabil-ity at the Thermokarst Wetland site may also serveto slow in situ decomposition (Bastviken and others

2004), resulting in the "preservation" of more la-bile DOC compounds. The Well Drained site hasthe highest density of black spruce trees and theleast moss cover (Table 1), and thus potentially theleast labile DOC sources. However, the smallerDOC pool at the Well Drained site is transient as thesoils drain quickly after spring thaw, so there maynot be the opportunity to accumulate more re-calcitrant DOC in soils at this site.

Vegetation Leachate DOC ChemicalCharacteristics and BiodegradabilitySoluble OC yields from various boreal forest vege-tation species ranged from less than 1 to 7%, of drylitter mass, although these yields are conservativeas some DOC was likely lost to respiration, espe-cially from the highly biodegradable DOC leach-ates. These values are similar in magnitude to DOCyields measured from a wide range of vegetationfrom very different ecosystems, including subalpineand tropical plant species (Cleveland and others2004). The relatively narrow range in DOC yieldsthat we measured from the various vegetationtypes is in striking contrast to the very broad rangein potential biodegradability of this DOC, whichspanned from approximately 10 to 90%, DOC lossover one month. The variability in potential DOCbiodegradation was highly correlated with the rel-ative amounts of hydrophilic DOC in the vegeta-


lion leachates (Figure 6). Qualls (2005) more nar-rowly defined hydrophilic neutrals as the most la-bile DOC fraction, and hydrophilic acids as the leastlabile. We did not separate the hydrophilic fractioninto neutrals and acids (the fraction we call HPIincludes both), and this may explain why someleachates. such as P. mariana needles, had a largeHPI fraction but a low potential biodegradability(33%) DOC as HPI at the start of the incubation,II % total DOC mineralized). Presumably, the HPIfraction from P. mariana needles was predomi-nantly composed of HPI-acids. This may be the casefor the soil pore water HPI DOC as well.

The high potential biodegradability of DOC lea-ched from the mosses is opposite of what we ex-pected, based on low decomposition rates reportedfor both Sphagnum moss and Hylocomium splendens(Hobbie 1996; Aerts and others 1999). SlowSphagnum decomposition rates have been attrib-uted to low concentrations of nutrients, high con-centrations of decay-resistant compounds, and therelease of decay inhibitors (Painter 1991; Johnsonand Damman 1993; Verhoeven and Toth 1995;Aerts and others 1999). However, our resultsdemonstrate that DOC leached from some Sphag-num and feather mosses is highly biodegradable,and results from other studies provide supportingevidence of this. Moore and Dalva (2001) mea-sured moderately high rates of CO2 productionduring incubation of fresh Sphagnum fuscum andSphaqnum maqellanicum moss under saturated con-ditions. It is possible that DOC leached from theSphagnum fueled this respiration. Carlton and Read(1991) found that Pleurozium schreberi leachatecontained sufficient nutrients to support growth ofmycorrhizal fungi in pure culture. Therefore, somemosses contain very labile soluble C compounds,while at the same lime have recalcitrant structuralcomponents. Whether this is the case for most mossspecies needs to be investigated further.

Transformation of DOC Chemistry byMicrobial MetabolismThe changes in the HPOA:HPI ratios of the vege-tation leachates (Figure 3B) and the moss leachatefluorescence (Figure 5) demonstrate that as DOCundergoes microbial processing certain compoundsare selectively removed while other compoundsremain, increasing in relative abundance. Specifi-cally, the labile HPI fraction was metabolized anddecreased in absolute abundance whereas theHPOA fraction became relatively enriched. In termsof tile fluorescent components, the protein-likecomponents were preferentially metabolized while

certain quinone- and semiquinone-like compo-nents increased in relative abundance. Over time,the distinct chemical signatures of the fresh vege-tation leachates became more similar to each other,and to the chemical characteristics of DOC in thesoil pore waters.

The transformation of DOC by microbial metab-olism may in part explain the absence of significantsite-dependent and seasonal variability in soil porewater DOC chemical characteristics, and explainthe distinct differences between vegetation leach-ates and soil pore water DOC properties (Figure 7).For example, mosses may contribute significantamounts of DOC to the forest floor (Wilson andCoxson l999; Moore 2003), but the chemical sig-nature of fresh moss-derived DOC was absent fromsoil pore waters because much of this DOC is highlylabile and disappears on the order of days. Al-though much of the moss-derived DOC is miner-alized quickly, the microbially altered DOCremaining in the leachate resembles the chemicalcharacter of soil pore water DOC. The chemicalcharacter of the soil pore water DOC also resemblesP. mariana bark and twigs leachate DOC, althoughthis type of biomass is relatively less abundant inthese sites and has less direct contact with soils thanthe mosses. Other sources of DOC, including plantroot exudates and microbial enzymes and biomass(Thurman 1985; Guggenberger and Zech 1994),likely contribute to soil pore water DOC chemicalcharacteristics as well.

CONCLUSIONS

Boreal black spruce forest soils generally containedrelatively recalcitrant DOC in their pore waters, al-though this was not due to the lack of labile DOCsource material. Sphagnum mosses and feathermos-ses, which often form continuous groundcover onthe forest floor, leached DOC that is characterized byhigh hydrophilic DOC content and high potentialbiodegradability. Most of the moss-derived DOC canbe respired on the order of days, leaving behind morerecalcitrant DOC that has not been consumed, andDOC from other vegetation, to accumulate in soilpore waters. The chemical nature of DOC from dif-ferent vegetation species, rather than differences inthe rate of DOC supply, dictates the fate of plant-derived DOC in soils. Drainage condition, whichvaries with permafrost depth, appears to be animportant influence on both the size of the DOC pooland of the extent of accumulation of more re-calcitrant DOC compounds. The biodegradation ofplant and litter-derived DOC may be a significantsource of heterotrophic respiration in surface soils of


ACKNOWLEDGMENTS

We thank R. Cory and D. McKnight for theirassistance in running and interpreting fluorescenceand PARAFAC analyses. The following individualsassisted with sample collection and laboratoryanalyses: K. Butler, J. Jeppson, T. Sachs, and K.Cawley. Valuable comments on earlier versions ofthis manuscript were provided by N. Mladenov, G.Nee. and two anonymous reviewers. We thank S.Grandy for insightful discussions of carbon chem-istry and microbial metabolism, and R. Striegl forassistance with understanding the role of carbonateequilibrium in the DOC incubations. Any use oftrade, firm, or product names is for descriptivepurposes only and does not imply endorsement bythe US Government.

REFERENCES

Aerts R, Verhoeven .ITA, Whigman DF. 1999. Plaru-mediau-dcontrols on nutrient cycling in temperate Icns and bogs.Ecology 80(7):2170-81.

Aiken GR, McKnight DM, Thorn KA, Thurman EM. 1992. Iso-lation of hydrophilic organic acids from water using nonionirmacroporous resins. Org Geochem 18(4):567-73.

Airkenhead JA, McDowell WHo 2000. Soil C:N ratio as a pre-dictor of annual riverine DOC Ilux at local and global scale-s.Global Biogeochem Cycles 14:127-38.

Amon RMW, Benner R. 1996. Photochemical arid microbialconsumption of dissolved organic carbon and dissolved ox y-gen in the Amazon River system. Gcorhirn Cosmochim lIeta60(10): 1783-92.

Baker A. 2002. Spectophotomctric discrimination of river dis-solved organic matter. Hydml Process 16:3203-13.

Bastviken D, Persson L Odham G, Tranvik L. 2004. Degradationof dissolved organic matter in oxic and anoxic lake water.Limnol Oceanogr 49(1):109-16.

Bisbee KE, Gower ST, Norman JM, Nordheim EV. 2001. Envi-ronmenral controls on ground cover specks composition andproductivity in a boreal black spruce forest. Oecologia129:261-70.

Camill P, Lynch JAo Clark JS, Adams JB, Jordan B. 2001.Changes in biomass, aboveground net primary production.and peat accumulation following pcrmalrost thaw in Iheboreal peatlands of Manitoba, Canada. Ewsyslems 4:4(,1-78.

Carlton TJ, Read DJ. 1991. Eciomvcorrhizas and nutrienttransfer in conifer leather moss ecosvstems. Can.1 Bot 69:778-85.

Chasar LS, Chanton JP, Glaser PH, Siegel DI, Rivers .JS. 2000.Radiocarbon and stable carbon isotopic evidence for t rans-port and transformation of dissolved organic carbon. dis-solved inorganic carbon, and methane in a northernMinnesota pearland. Global lliogenchem Cycles 14(4): 1095-l08.

Chen W, Westerhoff P, Leenhccr .IA. Booksh K. 2003. Fluores-cence excitation-emission matrix regional integration to

black spruce forests and Sphagnum-dominated wet-lands, as has been suggested for other terrestrialecosystems (Jandl and Sollins 1997; Chasar andothers 2000; van Hees and others 2005).

Based on the results of this study, we suggest thatthere are at least two mechanisms through whichclimate change at northern latitudes may be influ-encing the chemical characteristics and cycling ofterrestrially-derived DOC: shifts in vegetation dis-tributions and changes in water residence times.Shifting vegetation distributions at local scales, suchas those resulting from thermokarst wetland for-mation, and regional scales, such as increasing shrubabundance in the arctic tundra (Sturm and others2005), may be changing DOC chemical characteris-tics through the introduction of new types of plant-derived DOC. It may be difficult, however, to detectthese chemical changes in situ due to changes in theDOC chemical character that occur during microbialmetabolism. Further studies of plant-derived DOCleachate chemical characteristics and biodegrad-ability will be instrumental in determining the ef-fects of vegetation distribution shifts. Changes inwater residence times across the landscape as a resultof permafrost thaw may be having significant impacton the chemical nature and biodegradability of DOCtransported to aquatic systems. Increased waterresidence times allows greater opportunity formicrobial metabolism of DOC in soils, whereas de-creased residence times likely results in the transportof more labile DOC to surface waters.


quantify spectra for dissolved organic matter. Environ SciTechnol 37:5701-10.

Chin YP, Aiken GR, O'Loughlin E. 1994. Molecular weight.polydispersuity. and spectroscopic properties of aquatic humicsubstances. Environ Sci Technol 2R:1853-8.

Christ MJ, David MH. 1996. Temperature and moisture effectson the production of dissolved organic carbon in a spodosol.Soil Bioi Biochem 28:1191-9.

Cleveland CC, Neff .IC, Townsend AR, Hood E. 2004. Cornposi-tion, dynamics, and fate of leached dissolved organic matter interrestrial ecosystems: results from a decomposition experi-ment. Ecnsystems 7:275-8).

Coble P. 1996. Characterization of marine and terrestrial DOMin seawater using excitarion-cmission matrix spectroscopy.Mar Chem 51 :325-46.

Cole JJ, Prairie YT, Caraco NF. McDowell WH, Tranvik U, StrieglRG, Duarte CM, Kortelaincn p, Downing JA, Middelburg JJ,Melack .J. 2007. Plumbing the global carbon cycle: integratinginland waters into the terrestrial carbon budget, Ecosystemsdoi:l0.1007 ls10021-006·9013-8.

Cory RM, McKnight DM. 2005. Fluorescence spectroscopy re-veals ubiquitous presence of oxidized and reduced quinonesin dissolved organic mailer. Environ Sci Technol 39:8142-9.

Cronan CS, Aiken GR. 1985. Chemistry and transport of solublehumic substances in forested watersheds of the AdirondackPark, New York. Gcochim Cosmochim Acta 49: 1697-705.

Dillva M, Moore TR. 1991. Sources and sinks of dissolved organiccarbon in a Iorcsrcd swamp catchment. Biogeochemistry15:1-19.

Frey KE, Smith LC. 2005. Amplified carbon release from vastWest Siberian peatlands by 2100, Geophysical Research Let-ters 32 L09401, doi: Io.1029/2D04GL022025.

Goulden ML. Wolsey SC, Harden JW, Trumbore SE, Crill PM,Gower ST, Fries T, Daube BC, Fan S·M, Sutton DJ, Bazzaz A,Mungn .TW. 1998. Sensitivity of boreal forest carbon balanceto soil thaw. Science 279:214--7.

Guggcnhergcr G, Zcch W. 1994. Dissolved organic carbon inforest floor leachates: simple degradation products or humicsubstances? Sci Total Environ 152:37-47.

Hall FG, Knapp DE, Huemrnrich KF. 1997. Physically basedclassification and satellite mapping of biophysical character-istics in the southern boreal forest. J Geophys Res D Atmos102:29567-80.

Hobbie SE. 1996,. Temperature and plant species control over litterdecomposition in Alaskan tundra. Ecol Monogr 66:503-22.

Hongve D. 1999. Production of dissolved organic carbon in for-ested catchments. J Hvdrol 224:91-9.

Hope D, Billet MF, Cresscr MS. 1994. A review of the export ofcarbon in river water: Ilu xes and processes. Environ Pollur84:301-24.

Hope D, Billet MF, Cresser MS. 1997. The export of organiccarbon in two river systems in NE Scotland . .I Hydrol 193:61-82.

Ingle .JD, Crouch SR. 1988. Spectrochemical analysis. UpperSaddle River (N.I): Prentice-Hail, p 608.

Jaffe R, Boyer .lN, Lu X, Maie N. Yang C, SCullyNM, Mock S.2004. Source characterization of dissolved organic matter in asub: ropical mangrove-dominated estuary by fluorescenceanalysis. Mar Chern R4:195-210.

Jandl R, Sullins P. 1997. water-extractable soil carbon in rela-tion to the helowground carbon cycle. Biol Fertil Soils 25: 196-201.

Johnson LC, Damman AWH. 1993. Decay and irs regulation inSphagnum peatlands. Adv Bryol 5:249-96.

Kalhitz K, Schmerwitz .I, Schwesig D, Matzner E. 2003. Bio-degradation of soil-derived dissolved organic matter as relatedto its properties. Geoderma 113:273-91

Kawahigashi M, Kaiser K. Kalbitz K, Rodionov A, GuggenbergerG. 2004. Dissolved organic matter in small streams along agradient from discontinuous to continuous permafrost. GlobalChange Bioi 10: 1576-86.

Manies KL, Harden JW, Yoshikawa K, Randerson J. 2001. Theeffect of soil drainage on tire and carbon cycling in centralAlaska. In: Galloway J, Ed. US Geological Survey ProfessionalPaper 1678, Studies by the US Geological Survey in Alaska. pp145-52.

Manies KL, Harden JW, Silva SR, Briggs PH, Schmid BM. 2004.Soil data from Picea mariana stands near Delta Junction, AK ofdifferent ages and soil drainage type. US Geological SurveyOpen File Report 2004--1271. 19p.

McArthur JV, Marzolf GR, Urban JE. 1985. Response of bacteriaisolated from a pristine prairie stream to concentration andsource of soluble organic carbon. Appl Environ Microbiol49( 1):238-41.

McDowell WH, Wood T. 1984. Podzolization: soil processescontrol dissolved organic carbon concentrations in streamwater. Soil Sci 137:23-32.

McDowell WH, Zsolnay A. Aitkenhead-Peterson JA, GregorichEG, Jones DL, Jodemann D, Kalbitz K, Marschner B, SchwesigD. 2006. A comparison of methods to determine biodegrad-able dissolved organic carbon from different terrestrial sour-ces. Soil Bioi Biochem 38:1933-42.

McKnight D, Thurman EA, Wershaw RL. 1985. Biogeochemistryof aquatic humic substances in Thoreau's Bog, Concord,Massachusets. Ecology 66: 13 39-52.

McKnight DM, Boyer EW, Westerhoff PI


Painter TG. 1991. Lindow Man, Tollund Man, and other peat-bog bodies: the preservative and antimicrobial action ofsphagnan. a reactive glycuronoglycan with tanning andsequestering properties. Carhohydr Polym 15:123-42.

Plummer LN, Busenberg E. 1982. The solubility of calcite, ara-gonite and vaterite in COrH20 solutions between 0 and90°C, and an evaluation of the aqueous model for CaCO,-COrH10. Geochim Cosmochim Acta 44:1011-40.

Qualls RG. 2005. Biodegradability of fractions of dissolved or-ganic carbon leached from decomposing leaf litter. Environ SciTechnol 39:1616-22.

Qualls RG, Haines BL. 1991. Geochemistry of dissolved organicnutrients in water percolating through a forest ecosystem. SoilSci Soc Am J 55: 1112-1123.

Qualls RG, Haines BL. 1992. Biodegradability of dissolved or-ganic matter in forest rhroughtall. soil solution, and streamwater. Soil Sci Soc Am J 56:578-86.

Schiff S, Aravena R, Mewhinnev E, Elgood R, Warner B, DillonP, Trurnbore SE. 199R. Precambrian shield wetlands: hydro-logic control of the sources and export of dissolved organicmatter. Clim Change 40: 167-RR.

Shaver GR, Billings WI), Chapin FS III. Giblin AE, NadelhofterKJ, Oechel W C, Rastetter EB. 1992. Global change and thecarbon balance of arctic eCOSYSTems. Bioscience 42:433-41.

Smolander A, Kitunen V. 2002. Soil microbial activities andcharacteristics of dissolved organic C and N in relation to treespecies. Soil Bioi Biochem 34:651-60.

Stedmon CA, Markager S, Bro R. 2003. Tracing dissolved organicmailer in aquatic. environments using a new approach tofluorescence spectroscopy. Mar Chem 82:239-54.

Stedmon CA, Markager S. 2005. Resolving the variability indissolved organic matter fluorescence in a temperate estuaryand its catchment using PARAFAC analysis. Limnol Oceanogr50(2):686-97.

Striegl RG, Kortelainen P, Chanton JP, Wickland KP, Bugna GC,Rantakari M. 2001. Carbon dioxide partial pressure and 13Ccontent of north Temperate and boreal lakes at spring ice melt.Limnol Oceanogr 46(4):941-5.

Striegl RG, Aiken GR, Dornblaser MM, Raymond PA, WicklandKP. 2005. A decrease in discharge-normalized DOC export bythe Yukon River during slimmer through autumn. Geophy,Res Lett 32, L21413, doi: 10.1029/2005GU124413.

Sturm M, Srhimel J, Michaelson G, Welker .JM, Oherbauer SF.Liston GE, Fahnestock J, Romanovsky V. 2005. Winter bio-logical processes could help convert arctic t undra to shrub-land. Bioscience 55(1):17-26.

Thurman EM. 1985. Organic geochemistry of natural waters.Dordrecht: Martinus Nijhoff/Dr W Junk Puhlishers. p 497.

Van Cleve K, Dyrness CT. 1983 Introduction and overview of amultidisciplinary research project: the structure and functionof a black spruce (Piceamariana) forest in relation to other fire-affected taiga ecosystems. Can J For Res 13:695-702.

van Bees PAW, Jones DL Finlay R, Godbold DL, Lundstrom US.2005. The carbon we do not see-the impart of low molecularweight compounds on carbon dynamics and respiration inforest soils: a review. Soil Biol Biochem 37: 1-13.

Verhoeven JTA, Toth E. 1995. Decompostion of Cares andSphaqnum litter in fens: effects of litter quality and inhibi-tion by living tissue homogenates. Soil Biol Biochem27(3):271-5.

Weishaar JL, Aiken GR, Bergamaschi BA, Fram MS, Fujii RMopper K. 2003. Evaluation of specific ultraviolet absor-bance as an indicator of the chemical composition andreactivity of dissolved organic carbon. Environ Sci Technol37:4702-8.W

Wickland KP, Striegl RG, Neff JC, Sachs T. 2006. Ellens ofpermafrost melting on CO2 and CH4, exchange of a poorlydrained black spruce lowland. J Geoph vs Res G Biogeosci 111,G02011, doi:JO.1029/2005JG000099.

Wilson JA, Coxson DS. 1999. Carbon flux in a subalpine sprucefir forest: pulse release from Hylocomium splcndcns leath er-moss mats. Can J Bot 77:564-9.

Zsolnay A, Baigar E, Jimenez M, Steillweg 13, Sacromandi F.1999. Differentiating with lIuorescence spectroscopy thesources of dissolved organic matter in soils subjected to drying.Chemosphere 38:45-50.

dissolved organic carbon in alaskan boreal forest: sources ......dissolved organic carbon in alaskan...

Documents