The role of terrestrially derived organic carbon in thecoastal ocean: A changing paradigm and the priming effectThomas S. Bianchi1

Department of Oceanography, Texas A&M University, College Station, TX 77843

Edited by Paul G. Falkowski, Rutgers, The State University of New Jersey, New Brunswick, NJ, and approved November 2, 2011 (received for reviewAugust 10, 2011)

One of the major conundrums in oceanography for the past 20 y has been that, although the total flux of dissolved organic carbon (OC; DOC)discharged annually to the global ocean can account for the turnover time of all oceanic DOC (ca. 4,000–6,000 y), chemical biomarker andstable isotopic data indicate that there is very little terrestrially derived OC (TerrOC) in the global ocean. Similarly, it has been estimatedthat only 30% of the TerrOC buried in marine sediments is of terrestrial origin in muddy deltaic regions with high sedimentation rates. Ifvascular plant material—assumed to be highly resistant to decay—makes up much of the DOC and particulate OC of riverine OC (along withsoil OC), why do we not see more TerrOC in coastal and oceanic waters and sediments? An explanation for this “missing” TerrOC in the oceanis critical in our understanding of the global carbon cycle. Here, I consider the origin of vascular plants, the major component of TerrOC,and how their appearance affected the overall cycling of OC on land. I also examine the role vascular plant material plays in soil OC, inlandaquatic ecosystems, and the ocean, and how our understanding of TerrOC and “priming” processes in these natural systems has gainedconsiderable interests in the terrestrial literature, but has largely been ignored in the aquatic sciences. Finally, I close by postulating thatpriming is in fact an important process that needs to be incorporated into global carbon models in the context of climate change.

decomposition | global carbon cycling | terrestrial organic matter

Understanding the alteration ofmaterials flowing from riversto the ocean has been an in-creasing area of research over

the past few decade(s), and presentlyglobal community programs such as theInternational Geosphere Biosphere Pro-gram and its major project, Land OceanInteraction in the Coastal Zone, areleading the research in this field (1). Nat-ural organic matter (OM), one of themost important components of riverinematerial, is the largest reactive reservoir ofreduced carbon on Earth, with soil con-taining 1,600 Pg C, sediments 1,000 Pg C,and the ocean 685 Pg C as dissolved OM(DOM), comparable to the global atmo-spheric CO2 reservoir (2–6) (Fig. 1). Ter-restrially derived organic carbon (OC;TerrOC) is a heterogeneous mix of recentvascular plant detritus, associated soil OC(SOC), older fossil (i.e., petrogenic) OCfrom carbonate rock erosion, and blackcarbon (e.g., soil organic charcoals andanthropogenic soots) (refs. 2, 7 and refs.therein). Approximately 90% of the OC(15,000,000 Pg C) on Earth resides inshales and other sedimentary rocks (7).Vascular plant material is believed tospend little time sequestered in in-termediate reservoirs such as soils, fresh-water sediments, and river deltas andcontributes minimally to the old 14C agesoften observed on continental shelves.This results in part from the large contri-bution of petrogenic relative to vascularplant material found in marine sediments(8). All the aforementioned TerrOCmaterials are important in regulating theglobal carbon and oxygen cycles.It is well known that terrestrial and

aquatic biogeochemistry have developed

independently of each other over theyears (9). In terrestrial ecosystem studies,more studies are now focusing on primingeffect experiments that examine the addi-tion of labile compounds to soils that re-sults in enhanced release of soil-derivedcarbon and nitrogen—compared withthose without additions. The general termto describe this is the “priming effect,” firstintroduced by Bingemann et al. (10) andlater reviewed by Kuzyakov et al. (11). Theactual process of priming was first de-scribed in agricultural literature by usingthe decomposition of green manure oflegume plants in soils (12). However, itwas not until the middle 1940s and 1950sthat the priming effect was more formallyrecognized with renewed interest (13, 14).Although terrestrial studies continue toincorporate priming as an important pro-cess in soil carbon cycling, aquatic studiesare seriously lagging behind, especiallyrelative to the dramatic increase in thenumber of priming studies in the pastdecade in the terrestrial literature (15).

Evolution of Land Plants and TheirImpact on Chemical Ecology of NaturalEcosystemsVascular plants represent the largestcomponent of living biomass on Earth (570Pg C). The oldest fossil proven to bea vascular land plant is Cooksonia, fromthe late Silurian Period (ca. 420 Ma) (16).Much of the paleobotanical work in the1970s suggested that the timeframe for theinitial diversification of land plants wasbetween the late Silurian and early Devo-nian (410 Ma). Later work has suggesteda time frame that begins in the Ordovician(ca. 480 Ma) and lasts over a period ofmore than 100 Myr (24). Lignin, a complex

heterogeneous phenylpropanoid polymerthat occurs throughout the cell walls withinthe xylem tissues of vascular plants, evolvedin land plants from their aquatic ancestorsapproximately 475 Ma (17). Lignin pro-vided structural integrity for the entire plantas well as resistance to cell collapse undertension associated with water transport. Al-though lignin-like compounds have beenidentified in green algae (18), the presenceof “true” lignin was only recently discoveredin the walls of certain intertidal red algae(19). Nevertheless, the overwhelminglydominant source of vascular plant detritusand the associated moieties of lignin, inboth dissolved and particulate OC of natu-ral ecosystems, are terrestrial biomes.In terrestrial ecosystems, most de-

composition studies have traditionallyconsidered lignin to be the most re-calcitrant component in the degradation ofTerrOC. In fact, lignin is one of the mostabundant natural organic compound onearth, second only to cellulose. Thesecompounds have had a significant impacton the biogeochemistry of both terrestrialand aquatic ecosystems, particularly inthe context of decomposer evolution.Therefore, how has a molecule like lignin(in both woody and nonwoody plant tis-sues) affected the overall kinetics of OCcycling on land since its appearance in themid-Paleozoic (ca. 350 Ma), and what im-pact has this had on aquatic ecosystems?The fossil record of wood decay revealsfossils that have white-pocket rot fungi and

Author contributions: T.S.B. wrote the paper.

The author declares no conflict of interest.

This article is a PNAS Direct Submission.1E-mail: tbianchi@tamu.edu.

www.pnas.org/cgi/doi/10.1073/pnas.1017982108 PNAS | December 6, 2011 | vol. 108 | no. 49 | 19473–19481

PERSPECTIV

ED

ownl

oade

d by

gue

st o

n Ju

ly 2

, 202

0 D

ownl

oade

d by

gue

st o

n Ju

ly 2

, 202

0 D

ownl

oade

d by

gue

st o

n Ju

ly 2

, 202

0

spindle-shaped zones of delignified woodthat occur as early as the Triassic (ca. 220Ma), which, very interestingly, suggests thatthis decay process has not changed formillions of years.

Enzyme Dynamics and TerrOC DecayFungal Enzymes. Fungi are the primary de-composers of woody and nonwoody vas-cular plant materials in terrestrial eco-system. The decomposition pathway forsome Basidiomycotina fungi (i.e., brown-rot fungi) involves an initial decay of woodthat begins with an extracellular non-enyzmatic attack by generating hydroxylradicals (·OH) via the Fenton reaction(20). In a different pathway of decay, otherBasidiomycotina (i.e., white-rot fungi) si-multaneously decompose cellulose andlignin using free radical attack that canbreak a variety of bonds in the large phe-nylpropanoid heteropolymer (21).Wetland systems, which are important

transitional zones between terrestrial andaquatic ecosystems, have marine fungi thatare important in decomposing vascularplant materials (22) (Fig. 1). Many ofthe dominant marine fungi [e.g., eumy-cotic, mitosporic, and chytrids (39)]have enzymes capable of decomposingwetland litter. The primary fungal speciesin decomposing salt marsh cordgrass(Spartina alterniflora) is Phaeosphaeriaspartinicola, whereas that in mangrovesis Halophytophthora spp. (23).

Bacterial Enzymes. Bacteria that decomposeTerrOC are generally divided into twocategories: the primary degraders (con-

sidered erosional bacteria) and initiatedigestion of woody and nonwoody plantmaterial, which allows for the secondarydegraders (considered scavenging bacte-ria), to consume metabolites from the pri-mary degraders. Sulfate reducing bacteriaDesulfovibrio spp. are an example ofscavenging bacteria when considering theirrole in wood decay (24). Some rate con-trolling effects on decay are differentspecies of wood, water temperature,oxygen availability, and salinity. The bac-teria commonly associated with wooddecay are cellulolytic aerobes such asCytophaga and Cellvibrio and anaerobicspecies such as Clostridium, Bacillus,Pseudomonas, Anthrobacter, Flavobac-terium, and Spirillum. Under anaerobicconditions, lignin is thought to bewell preserved. However, methanogenicdegradation of lignin derivatives hasbeen reported (25). In fact, lignin com-ponents of softwood and hardwood havebeen shown to partially degrade to gas-eous end products under anaerobic condi-tions (26).New molecular techniques that use the

16S rRNA or its encoding gene (rDNA)can be used as molecular biomarkers forthe presence of wood-decaying bacteria(28). Identifying the putative genes in genedatabases can provide important in-formation on the source of the enzymeand its potential functions (28); this ap-proach is key in our understanding ofglobal carbon cycling. Because not allbacteria found in wood are actually wood-decaying species, in general, the rRNAdiversity is typically less than 20 taxa,

particularly in samples from marine sites(27). It has been shown that the enzymesin sediments have a wider range of sub-strates than their water column analogues(28). Therefore, understanding the genesresponsible for the production of enzymescapable of lignin decay is key in our un-derstanding of global carbon cycling.

Enzyme Strategies and Limitations. The re-liance of microbes on extracellular en-zymes as an energy acquisition strategyraises a number of questions in ecosystemsciences. When do microbes produce ex-tracellular enzymes? Moreover, at whatcosts do extracellular enzymes play a cen-tral role in the heterotrophic utilization ofOC by microbial communities? Particulatesubstrates usually need to be degraded toa size of approximately 600 Da before theycan actively be transported across cell wallsby porins (29). Hydrolysis by extracellularenzymes is commonly considered a rate-limiting step in the remineralization ofOC (28, 30). Extracellular enzymes areeither released freely into solution ormay remain tethered to the cell (30).Some enzymes are constitutive (pro-duced regardless of cellular substrate con-ditions) and others are produced only inresponse to specific environmental cues(28). Enzymes can also be blocked or in-activated, rendering them ineffective inthe process of carbon remineralization(28) (Fig. 2). Substrates may also slow theprocess of carbon remineralization if theyare adsorbed to particles (e.g., clays) orcomplexed by macromolecules (e.g., hu-mic substances) (28, 31). Not all substrateshave to be hydrolyzed to monomers beforebeing transported into the cell, so simplylooking at the uptake of monomers as in-dicators of carbon remineralization maybe somewhat restrictive (28, 32). Addi-tionally, exo-acting extracellular enzymescleave units from the terminal end ofa polymer chain, where endo-acting en-zymes are specially targeted to certainsubstrates (32), further complicating thestudy of monomer uptake.As extracellular enzymes function out-

side the cell, the reaction products maydiffuse away from the enzyme-secretingmicrobe (33, 34). Therefore, microbes maybenefit by intercepting the products with-out secreting their own enzymes. Thisstrategy has been referred to as “cheating”and should arise whenever there are mul-tiple organisms that can benefit from theresource investment of a single organism(33). Thus, enzyme diffusion must be ex-tremely slow for producers to competewell in natural environments; such slowdiffusion could occur if enzymes were toencounter more tortuous diffusion path-ways in soils, sediments, or particles be-cause of pore structures (35). Most naturalterrestrial soil systems contain a large

Vascular Plants (570)t<1 yr

Soil Organic Litter (1600)Soil Mineral Horizon

t-10-10,000 yr

Sedimentary Rocks (15,000,000)t=infinite

Floodplain "reactors"t<1 yr - infinite

Ocean DIC (38,000)t=2,000 yr

Ocean DOC (685)t=1,000-6,000 yr

Ocean POC (3)t=0-10 yr

Recent Sediments (150)t=0-10,000 yr

net marine primary production (50)

OC deposition (0.15)

eolian transport (0.1)

river transport (0.5)

burial (0.1)

net terrestrial primary production (60)

Kerogen weathering/fossil fuel utilization (5)

physical weathering (0.1) humification (50)

flux of 'pre-aged' TeRR OC

2

uplift

6.3 ~92

~9058

deco

mpo

sitio

n

resp

iratio

n

59

photosynthesis120

Fig. 1. The global carbon cycle (units are in PgC or Pg C·y−1). Sources of inventories and fluxes are refs.2–8, 48, 59, 61; diagram modified from Drenzek (114).

19474 | www.pnas.org/cgi/doi/10.1073/pnas.1017982108 Bianchi

Dow

nloa

ded

by g

uest

on

July

2, 2

020

degree of heterogeneity: organism sub-strates, microbes, and mineral particlesform a 3D matrix of aggregates and porespaces of different sizes. Aggregates havebeen shown to be “hot spots” for microbialand enzyme activities (e.g., refs. 35, 36);this increase in enzyme activity has beenattributed to quorum sensing (system ofstimulus and response correlated withpopulation density) (37).

Other Metazoan Effects on Wood and LigninDecay. Perhaps the most well known wood-decomposing organism inmarine systems isthe shipworm (Teredo navalis) (38), whichresults in extensive biomechanical erosionof the woody material. Biomechanicaldisruption of soils can certainly affect thedistribution of oxygen, surface litter, andenzyme pathways. The scales of habitatutilization in soils depend mainly on thesize of the organisms: a few microns forbacteria, less than 100 μm for fungi, be-tween 100 μm and 2 mm for microar-thropods, and between 2 and 20 mm formacroarthropods (39). For example,macrodetritovores such as earthworms areimportant in processing SOC and wood(e.g., termites, bark beetles), there aremacrodetritivores in coastal wetlands(e.g., crabs and amphipods) that arecapable of digesting large amounts ofvascular plant litter (40). TerrOC re-sidues that have been fragmented andmoistened during macrofaunal gut trans-fer are more easily digested by microflora.For example, macrofauna like earthwormsand termites are large enough to developsymbiotic relationships with the microflorain their guts (41), but also allow for moreeffective breakdown of any remainingTerrOC in their feces. Fresh litter is oftenmixed down into the soil profiles by mac-

rofaunal organisms such as earthworms,thus injecting less degraded TerrOC atdeeper soil horizons. In fact, a quote byCharles Darwin further captures the im-portance of bioturbation in soils: “. . .inmany parts of England a weight of morethan 10,516 kg of dry earth annually passesthrough the bodies of earthworms and isbrought to the surface on an acre ofland. . .” (42).

Storage and Cycling of TerrOC in SoilsSoils, as the largest active OC pool onearth, contain an estimated 1,600 Pg ofOC (43) (Fig. 1). One estimate of theSOC inventory suggests that 250 Pg C isbound within cellulose, 175 Pg in lignin,150 Pg in hemicellulose and other polysac-charides, and 5 to 10 Pg in protein, lipids,and cutin, in addition to a large fractionthat remains unrecognizable (44). Meanresidence time of SOC increases withdepth from 10 to 10,000 y in the soil min-eral horizon (Fig. 1), dramatically in-creasing at greater than 0.2 m (45).There are three general mechanisms

associated with stabilizing TerrOC in soils:(i) inherent recalcitrance of specificorganic molecules that resist decay bymicrobes and enzymes; (ii) chemicalstabilization caused by interactions be-tween organic molecules, like surfacecondensation (sorption), which resultsin decreased availability of organic sub-strates; and (iii) physical protection oforganic substrates against decomposersfrom occlusion of substrates within ag-gregates (46). Sorption processes are animportant mechanism for transforminglabile DOC defined in the literature as OClower than 0.2, 0.45, or 0.7 μm in size) intostable SOC (47). The process of leachingDOC from forest floor litter with sub-sequent transport and sorption to lowermineral phase horizons has been in-creasingly recognized as an importantmechanism for SOC formation and stabi-lization. TerrDOC collected within themineral soil is chemically more similar tosoil OC than to fresh litter as a result ofsorption and exchange as water percolatesthrough soils (47) (Fig. 2).

Transport of TerrOC from Soils toGlobal OceanFrom Soils to Inland Waters. Approximately1.9 Pg C·y−1 is delivered to inland watersfrom the landscape (48). The distributionof molecules of different charge, weight,and size controls the reactivity of Terr-DOC in soils and its release to aquaticenvironments. Differences in the compo-sition of TerrDOC as it passes through theupper mineral soils are largely as a resultof microbial breakdown. Recent work hasshown that eroded soil particles can re-lease TerrOC and N when entering riversand oceans and that as much as 50% of

this OC is available for aquatic metabo-lism (49). TerrDOC released in summerfrom soils during high temperature hasbeen shown to be mainly derived fromoxidative breakdown of lignin, whereasTerrDOC released in winter and spring isdominated by carbohydrates; the reasonsfor this are not yet fully understood (50).In crop-growing regions, the major fac-

tors controlling TerrOC export are landuse and drainage associated with agricul-tural practices. Recent metadata analysisfurther supports that event-based fluxes ofTerrDOC in streams and rivers are veryimportant in forested watersheds (51).More specifically, the majority (70%) ofevent-based TerrDOC flux occurs duringthe rising hydrograph and during largeevents. TerrOC quantity, source, anddegradation state are dramatically alteredunder flooding conditions (52).

TerrOC Cycling in Inland Waters. Twice theamount of C is delivered to inland watersthan is delivered to the ocean (1.9 Pg C·y−1

vs. 0.9 Pg C·y−1), suggesting that TerrOCin these systems is more actively consumedthan previously thought. Heterotrophy isa condition in which respiration (R) ex-ceeds production (P; i.e., P/R < 1), andcan be inferred for many aquatic systemsbased on high partial pressure of CO2(53). Recent estimates of CO2 efflux fromstreams and rivers suggests that TerrOC isnot as recalcitrant as previously thought(54). Net heterotrophy of global fluvialrespiration from headwaters throughestuaries is approximately 2.1 Pg C·y−1,representing a global net heterotrophy of0.6 Pg C·y−1 (55). This high annual turn-over contradicts the paradigm that Ter-rOC cycles over longer timescales (48).Respiration and net ecosystem productionincreases from headwaters through riversto estuaries as the respiration of OC de-clines relative to gross primary production(55). This supports the River ContinuumConcept (RCC) (56) (discussed later) withfurther extension into estuaries, wheremetabolic performance is highest inheadwaters where most of the microbialbiomass and metabolic processes are as-sociated with stream-bed surface, whichallows for excellent exchange of nutrientsand oxygen and removal of wastes.The longitudinal change from hetero-

trophic headwaters, to more autotrophicmidorder streams and small rivers, andfinally to heterotrophic large rivers isa fundamental property of the RCC (56).There is a gradual succession in the mi-crobial community composition alongthis longitudinal gradient from headwa-ters to large rivers and even further outto estuaries (57). TerrDOC and terrestri-ally derived particulate OC (TerrPOC)produced within small channels are con-sidered to be recalcitrant and thus difficult

CO

Biomass

2

DOC

hydrolysis

high molecular weight organic matter

endo-acting extracelluar enzymes

exo-acting extracellular enzymes

mono-end oligomers

Fig. 2. Hydrolysis of high molecular weight OMto mono/oligomers (blue) by a variety of endo-acting (gold and red) and exo-acting (green, or-ange) extracellular enzymes. The identity of or-ganisms producing enzymes, structural specificityof enzymes, size spectrum of hydrolysis products,and identity of organisms consuming substratesare not specified in this diagram. This highlightssome of the major unknown factors in the earlysteps of carbon transformation in many aquaticsystems. Modified from Arnosti (28).

Bianchi PNAS | December 6, 2011 | vol. 108 | no. 49 | 19475

Dow

nloa

ded

by g

uest

on

July

2, 2

020

to assimilate in the riverine productivitymodel (58). More specifically, the riverineproductivity model suggests that highlyand moderately labile autochthonous OCare subsidies from the riparian zone.Moreover, these subsidies compensate forthe typically greater abundance of re-calcitrant TerrOC from upstream leakageand floodplain inputs (58). In contrast, theRCC suggested that terrestrially derivedfine particulate OM leaking from head-waters was the most important sourceof carbon for food webs in large rivers,although the increased abundanceof phytoplankton in larger rivers wasacknowledged (56).

Riverine Inputs of TerrOC to Global Ocean.Rivers play a major role in exportingTerrOC from the continents to the oceans(1, 3). Approximately 0.3 Pg C·y−1 of DOCand 0.2 Pg C·y−1 of POC are discharged byrivers to the global ocean (59); these OCfractions represent approximately 50% oftotal C (organic and inorganic inputs; 0.9Pg C) export by rivers (48) (Fig. 1). Beforereaching the open ocean, many of theserivers enter into estuarine zones andshelves that have also been shown to benew sources of CO2 to the atmosphere(refs. 6, 48 and refs. therein). In particular,the water-to-air flux of CO2 in estuarieshas been shown to be 0.3 Pg C·y−1 (6). Incontrast, there can be rapid and efficientdelivery of fossil OC to the coast with littledegradation in regions of highly erosivesteep mountain building areas such asTaiwan (60).

Changing Paradigm in AquaticEcosystemsFreshwaters. Based on a sample of 85 riversand 2,000 lakes, it has been established thatmany inland freshwater systems are netheterotrophic, based on supersaturation ofCO2 (61). Moreover, linkages betweenupper pelagic phytoplankton and bacterialbiomass only explain approximately 20%of bacterial production; once again, thisimplies the importance of allochthonoussources in fueling ecosystem stability inthese freshwater systems. Another widelyheld paradigm is that riverine OC repre-sents highly degraded remnants of vascularplant sources aged in soils. However, re-cent work has shown that riverine DOC inmost systems is modern and much youngerthan POC, and therefore not likely aged insoils (54, 62). In fact, lignin compositionaldifferences lignin in soils versus coastalsediments may be entirely caused by se-lective fractionation of OC during leachingphases and not degradation. Such frac-tionation processes have already been ob-served for free amino acids (63) andpeptides (64). This may suggest that riv-erine DOM is less degraded than pre-viously thought, and thus is more labile,

allowing for more rapid mineralization incoastal waters. There are many anthropo-genic factors that are contributing to ob-served changes in river and coastal OC,such as damming and land-use changes infarming (ref. 65 and refs. therein). Forexample, the Amazon mainstream hasvery low rates of algal production (66),whereas more than 50% of the POCin the upper Mississippi River is algal-derived (67).

Coastal and Open Ocean“Missing” TerrPOC. There is at least twice asmuch TerrPOC delivered by rivers eachyear than can be accounted for by burial ofOC in marine sediments (68). Assuminga river POC (OC >0.2, 0.45, or 0.7 μm insize) discharge of approximately 0.2 PgC·y−1 (69) and a TerrOC burial in sedi-ments of approximately 0.1 Pg C·y−1, thereis greater than 0.1 Pg C of TerrOC that isremineralized every year between riverdischarge and marine sediment burial(Fig. 1). More recent calculations esti-mated that 30% of the TerrOC beingburied in marine sediments is of terrestrialorigin (3). Ninety percent of the OC burialin the ocean occurs in deltaic and marginsediments and is associated with mineralparticles largely of clay-silt sizes (68). ThisTerrOC associated with minerals can beexchanged with marine OM as it entersthe coastal ocean, which also may be whywe do not see as much as expected fromriverine inputs. OC preservation is signifi-cantly enhanced by the formation of mi-croaggregates and has recently beenshown to encapsulate colloidal sized par-ticles of OC with minerals, creating poresthat are filled with OC (70). More specif-ically, this has reaffirmed that OC is notprotected by small mesopores on clayparticles (2–10 nm) too small to excludeexoenzymes that degrade OC (68). Re-suspension of highly fluidized muds indeltaic regions results in greater oxygenexposure compared with most other fine-grained sediments in shelf environments(71). Although low oxygen concentrationis generally assumed to limit microbialprocesses in sediments, it should also benoted that the lack of availability of labileOM (LOM) limits bacterial heterotrophicactivity in aquatic ecosystems even in thepresence of adequate oxygen levels (72).The “missing” TerrPOC in coastal sedi-ments suggests that this material is beingremineralized faster than previouslythought and/or is being transported toother regions offshore.

“Missing” TerrDOC. Although the 0.3 Pg ofDOC (which has been defined in the lit-erature as OC < 0.2, 0.45, or 0.7 μm insize) that is discharged annually to theglobal ocean can be accounted for inthe turnover time of all oceanic DOC

(ca. 4,000–6,000 y) (73), chemical bio-marker and stable isotopic data (74, 75)indicate that there is very little TerrOC inthe global ocean. The average residencetime of TerrDOC in the open ocean is lessthan 100 y (76), suggesting the removal canbe as short as decadal time scales. TheArctic Ocean receives approximately 10%of the global river discharge, with Arcticrivers delivering approximately 0.03 Pg C·y−1

TerrDOC to shelf waters (77). The agingof deepwater in the ocean may in part berelated to processes that occur in coastalregions. Rivers that empty into the NorthAtlantic have been found to have old (i.e.,14C-depleted) and young (i.e., 14C-en-riched) TerrDOC with predominantly oldPOC (62). Selective degradation over theresidence time of the estuarine and riversystems may leave more recalcitrant ma-terial OM, resulting in preaging materialbefore reaching the oceans (62).There is growing evidence that photo-

chemical transformations enhance the lossof TerrDOC in the coastal ocean (78). Infact, in a mesocosm experiment, as muchas 75% and 72% of dissolved lignin wasremoved after 28 d and 10 d, respectively,of exposure to sunlight in MississippiRiver waters (79). Recent work in theCongo River, which delivers 5% of theglobal riverine DOC to the ocean (80), hasshown that the DOC has molecular sig-natures similar to carboxyl-rich alicyclicmolecules [CRAM; a complex mixture ofcarboxylated and fused alicyclic structureswith a carboxyl-C–to–aliphatic-C ratio of1:2 to 1:7 (81)], and was highly photo-la-bile (82). Many of these studies haveshown molecular signatures of CRAMderived from lignin, suggesting a linkage inpart with TerrOC. CRAM also representsthe most abundant component of very oldDOC in the deep ocean (81). Moreover,this work showed that some of the missingriverine-derived DOC in the global oceanis, in part, being removed at the river/ocean boundary in river plume environ-ments by photochemical processes. Link-ages between the formation of CRAM(82) in shallow coastal waters via photo-chemical reactions, and possibly microbialdecay, may prove to be critical in resolvingthe absence of TerrOC biomarkers inthe open ocean, as well as the “old” re-calcitrant pool observed in the oceandeepwaters.Finally, it should be noted that other in

situ open ocean processes, not related tocoastal photochemical transformations ofriver-derived DOM, can form recalcitrantDOM via (i) production of recalcitrantDOM from phytoplankton (e.g., acylhe-teropolysaccharides); (ii) formation ofmelanoidins through condensation re-actions; (iii) contribution from cell wall-derived material; (iv) adsorption of labileDOC to colloidal material; (v) formation

19476 | www.pnas.org/cgi/doi/10.1073/pnas.1017982108 Bianchi

Dow

nloa

ded

by g

uest

on

July

2, 2

020

of colloids from microzooplankton graz-ing; (iv) release of metabolites; and (vii)the microbial carbon pump (ref. 83 andreferences therein).

Priming Effect: Bridging the GapBetween Terrestrial and Aquatic SciencesPriming is formally defined by Kuzyakovet al. (11) as “strong short-term changes inthe turnover of soil organic matter causedby comparatively moderate treatmentsof soil.” Such treatments might includeinput of organic and mineral fertilizer,exudation of organic substance by roots,mechanical treatment of soil, and rewet-ting and drying effects on soils. Horvath(84) defined co-metabolism, a subcom-ponent of priming, as “degradation ofa recalcitrant substance by microorganismsin the presence of a readily degradablesubstrate without the ability for these mi-croorganisms to utilize the products ofrecalcitrant OC oxidation.” Priming istypically divided into “real” priming,whereby soil OM is decomposed, and“apparent” priming, whereby there arechanges in microbial biomass turnover butno effects on SOC decomposition (15).

Priming in Soils.A review by Jenkinson et al.(85) on priming sparked the beginning ofmany new publications on priming in thesoil literature that continues today. Ingeneral, there are three different hypoth-eses explaining the priming effect (86)(Fig. 3). In one case, LOM-degrading en-zymes are produced by LOM decomposersthat are also capable of degrading re-calcitrant OM (ROM) into intermediatecatabolites, which could not otherwise beused by LOM decomposers (Fig. 3A). Inanother situation, a community of de-composers mineralizes catabolites (i.e.,cometabolism), whereby LOM degrada-tion by LOM decomposers brings energyto the ROM decomposers. This energyallows the ROM decomposers to produceROM-degrading enzymes, releasing nu-trients for both ROM and LOM decom-posers, creating a net mutualism betweenthese two microbial communities (Fig.3B). Finally, there may be a single pop-ulation capable of producing enzymes ableto degrade LOM and ROM; LOM deg-radation provides energy for the synthesisof ROM-degrading enzymes (Fig. 3C). Allthe aforementioned scenarios representthe simplified pathways established forpriming in soil types. Similar ones mayexist in aquatic systems, but further work isneeded for such comparisons.Many priming experiments in soils have

showed that CO2 evolution could be in-creased from four- to 11-fold higher afterthe addition of labeled 13C-plant residues(ref. 15 and refs. therein). Some of thecommon priming substrates in soils aresimple sugars and amino acids, with com-

bined additions having synergistic effectson priming (87). Other common substratesin soils making up a significant fraction ofDOC are oxalic acid, acetate, and catechol(88). Deep SOC persists because it isbound to soil minerals and is in a chemicalform that is not easily assimilated by mi-crobes (89), and perhaps lacks contactwith priming substrates from root tips inthe upper soils (Fig. 4).According to Kuzyakov et al. (17), what

is needed to quantify priming effects is theapplication of both 14C and 15N labeling ofsoil OM or soil microorganisms; simulta-neous monitoring of nitrogen-gross min-eralization and CO2 efflux; experimentsthat examine interactions of plants, mi-croflora, and soil fauna; and quantificationof nutrients mobilized or immobilized inpriming effects (Fig. 4). The mineraliza-tion of charcoal, a very recalcitrant form ofOC, was shown to increase between 36%and 600% (vs. control) when primed withglucose (87). Priming has been shown toincrease the loss of soil carbon by 169%after priming by sucrose, 44% followingchopped maize, and 67% following groundmaize additions (90). In another primingexperiment, soil samples from deep hori-zons (0.6–0.8 m) were incubated with 14C-labeled cellulose at amounts reflectinga rate representing approximately 25% ofthe annual fresh plant litter carbon de-position into the surface layer by plantroots; the results showed an increase indecomposition of SOC, as indicated by an

increase the amount unlabeled of soil-originated CO2 (87).

Priming in Aquatic Systems.Benzoic acid hasbeen shown to act as a priming substancefor the assimilation of fulvic acids by thebacteria Arthrobacter spp. in lake waters(91). In one of the few examples of prim-ing in marine sediments, it was shown thatalgal OC additions induced levels ofbackground remineralization (i.e., prim-ing) by as much as 31% (92), suggestingthat OC priming in marine sediments islikely to be more important than pre-viously thought. Earlier work in marinesediments only implied the importance ofco-metabolism (a component of priming;Fig. 3) as a mechanism in breaking downrecalcitrant material (93). Other studieshave shown that sediments can be primedwith fresh substrates, typically from algalsources (94) (Fig. 4). It has also been ar-gued that the enhanced remineralizationof OC observed in sediments as a result ofbioturbation, and the downward transportof oxygen is also likely linked with the in-jection of fresh surface substrates (92).Although bioassays are still consideredperhaps the best method for predicting thebiological utilization of OC in ecosystemswhen combined with proper formulationof biodegradation kinetics (ref. 95 andrefs. therein), it is important to include thepriming potential. In sediments, bulk POCcharacterization (e.g., C/N molar ratio)does not reflect the actual reactivity andavailability for substrates to the microbialcommunity; it is likely fueled by a smallfraction of POC that is not detected bybulk measurements because it is rapidlycycled through the DOC pool in porewaters (28).Although the majority of DOC produced

in the surface ocean is used very quickly bymicrobes, approximately 20% of the an-nual net community respiration in theglobal ocean escapes the euphotic zone andenters into the large pool of recalcitrantDOC in deepwaters (97). This may be anexcellent location in which to look forpriming effects between these two impor-tant global pools of OC, and does haverelevance in part to TerrOC because someof the deeper OC pool has been shown tohave TerrOC material (e.g., CRAM).Dissolved free amino acids can turn overon the order of hours to days, and there-fore contribute to the rapid turnover ofopen ocean DOC (97). By-products fromzooplankton grazing on primary pro-duction are an important source of DOCto bacterial production in the open ocean;this labile fraction may turnover as quicklyas 2 to 6 d. This young, labile fraction canmix with the semilabile and recalcitrantcomponents in the open ocean, resultingin possible priming effects (Fig. 4).By using nucleic acids as a bacterial

Community of LOM decomposers

Community of LOM decomposers

LOM

LOM

ROM

ROM

Community of decomposers of ROM degradation products

Community of decomposers

LOM ROM

Population of decomposers

A

B

C

Ener

gy, s

truc

tura

l C,

and

nut

rient

s

LOM-degrading enzymes

Ener

gy, s

truc

tura

l C,

and

nut

rient

s

Ener

gy a

nd

stru

ctur

al C

Ener

gy a

nd

stru

ctur

al C

LOM

-deg

radi

ng

enzy

mes

ROM

-deg

radi

ng

enzy

mes

Lim

iting

nut

rient

s an

d st

ruct

ural

CLi

miti

ng n

utrie

nts

and

stru

ctur

al C

Ener

gy a

nd

stru

ctur

al C

LOM

-deg

radi

ng

enzy

mes

ROM

-deg

radi

ng

enzy

mes

Lim

iting

nut

rient

s an

d st

ruct

ural

C

Fig. 3. Three different priming effects involvingLOM and ROM, respectively. These proposedpriming effects are mainly derived from work insoils. Modified from Guenet et al. (86).

Bianchi PNAS | December 6, 2011 | vol. 108 | no. 49 | 19477

Dow

nloa

ded

by g

uest

on

July

2, 2

020

biomarker and compound specific-radio-carbon measurements, it was speculatedthat the older, more recalcitrant compo-nent of the DOC pool in the North Pacificwas made available for bacterial utilizationand assimilation via co-metabolism and/orphotooxidation (98). Surface water fromthe Sargasso Sea showed that bacterialgrowth only occurred when the DOC ex-ceeded ambient background levels ofDOC; in fact, only the surplus DOC,which represented approximately 6% to7% of the total DOC, was used by bacteria(99). Other work showed that bacteriaserve as remineralizers of a small fractionof bulk DOM during short periods whenlabile DOM becomes available (97).Making better linkages among ecosys-

tem science, analytical chemistry, molecu-lar microbiology, and computer infor-matics has recently been espoused assomething that is needed to fill the gaps inour knowledge of OM cycling in aquaticsystems (83). One avenue that is verypromising in addressing the issue of whatmicrobes are investing into enzymes and

what the return is in a complex mixture ofPOC and DOC is the use of transportergenes (100). In this process, random li-braries of expressed genes from coastalbacteria communities have been used toidentify sequences reflective of DOC-transporting proteins (i.e., mRNAs). Manyof the transport genes showed that car-boxylic acids, polyamines, and lipids werekey substrates in the biologically activepool of coastal DOC. Many bacteria hadgenes for DOC components commonlyfound in DOC such as amino acids,whereas other bactera (e.g., Roseobacter,SAR11, Flavobacteriales, and γ-Proteo-bacteria clades) had genes that were usedfor specific components of DOC. Recentwork has also shown that a greater numberof genes are produced for polysaccharidedegradation by microbial populations liv-ing in deep oceanic water than in surfacewaters (101). The role of Archaea andBacteria also remains an interestingquestion with regard to possible primingeffects and utilization of different fractionsof OC in the coastal and open ocean (102)

(Fig. 4). For example, Crenarchaeota havebeen shown to be less active in assimilatingamino acids and glucose, yet twice as ac-tive as bacteria in assimilating protein anddiatom exudates (102).

Recalcitrance and Lability: It’s All inthe WordIn the context of the aforementioned poolsof labile and recalcitrant OC in soils,sediments, freshwaters, and oceanicwaters, there appears to be an emergingconsensus that may modify our meaning ofthe words “recalcitrant” and “labile.” Thatis, these terms, which are commonly usedto describe differences in the “quality”of OC (particularly TerrOC) as food re-sources for heterotrophs, may only bevalid in the context of the ambient envi-ronment. Moreover, the potential assimi-lation capability of the local microbialcommunity, where such substrates mayreside, also needs to be considered. Todate, these terms have been defined largelybased on a chemocentric and anthropo-centric view that is in part based on our

Fig. 4. Sources of possible priming substrates in different terrestrial and aquatic environments. Modified from Raymond et al. (114).

19478 | www.pnas.org/cgi/doi/10.1073/pnas.1017982108 Bianchi

Dow

nloa

ded

by g

uest

on

July

2, 2

020

view of the chemical composition of Ter-rOC and/or petrogenic OC. Although thistopic has been mentioned previously in theliterature (e.g., see reviews in refs. 28, 83and refs. therein), it was not in the broadercontext of TerrOC and its origins. Con-sider for a moment that recent experi-mental work has shown that culturedprokaryotes were 14C-dead, illustratingthat they were living off of “recalcitrant”OC from 365-Myr-old black shale (103).Similarly, functional exoenzymes of meta-bolically active bacteria have been found in124,000-y-old sapropel of the easternMediterranean (104). Viable bacteria havealso been found in some of the oldestpermafrost layers in the Northern hemi-sphere (105). In addition to the “power” ofthe ever-evolving enzyme capability of mi-crobial communities, there are physicalgradients (e.g., photochemical examplesmentioned previously) that also control theavailability of “labile” and “recalcitrant”OC substrates. Finally, many of the in-cubation experiments in aquatic studiestypically exclude light when examining theefficiency of uptake of added or naturalconcentrations of TerrPOC and/or Terr-DOC by heterotrophs, which prevents thepossible interactive effects of priming be-tween algal exudates on TerrOC uptake.I propose that future carbon cycling

studies in aquatic ecosystems need to bemore cognizant of the importance of thepriming effect potential (PEP) in designingTerrOC utilization experiments and mod-eling efforts. For example, regions such ascoastal deltaic region, where there are highinputs of TerrOC, as well as, in many cases,high primary production; steep nutrient,light, and salinity gradients; and microbialcommunities that are highly adaptable tochanging physicochemical gradients, willhave a high PEP (Fig. 4). Similarly, up-welling regions, whether it be at the equa-tor or along coastal margins, are locationswhere “old” deepwater DOC comes incontact with surface waters that have someof the highest primary production rates inthe global ocean, and hence a high PEPbecause of the presence of phytoplanktonexudates (Fig. 4).

Impact of Global Change on Role ofTerrOC in Aquatic EcosystemsThe Arctic has and will remain one of thefocal points in climate change research asa result of the visibility of its rapid eco-system changes. For example, because ofglobal warming, Arctic rivers have allshown trends of increasing water discharge,as well as total DOC and TerrDOC tothe Arctic Ocean (106). TerrDOC has beenshown to be correlated with summer sea-

ice retreat in certain years in the Arctic.Increasing freshwater inputs will decreaseresidence time of TerrOC (both POC andDOC) to the Arctic Ocean, likely in-creasing inputs to the northern Atlantic;understanding the consequence of suchchanges will be a key part in understandingthe changing global carbon cycle (72, 107).In Arctic sediments, approximately 40%

to 50% of the OC being buried is derivedfrom petrogenic OC, likely derived fromweathering of ancient sedimentary rocksin the region (108). This is likely to beenhanced with greater overland flow pat-terns, as described earlier for the DOCand POC pools with ensuing climatechange. Although significant changes areoccurring with enhanced ice-melting in theArctic, there was little evidence for thetransport of TerrOC in Hudson Bay (109).The release of labile DOC from estuariesto the coast has been shown to be in partrelated to changing temperatures overdifferent seasons and not to the inherent“recalcitrance” of the material (62).The Arctic is likely to be an excellent lo-cation for observing the PEP on rapidlychanging inputs and export of TerrOC inthe advent of climate change. Unliketemperate and tropical systems, the Arcticis a place with greater inputs of freshwaterand TerrOC, and also where bacterialactivity is slower as a result of lower tem-peratures and photochemical reactionsreduced by the light/dark cycle (106).Bacterial diversity has been shown todecline with increased climatic severity, somany of the aforementioned changes inthe flux of TerrOC to the Arctic are alsolikely to be impacted by such changes inmicrobial communities with a high PEP.Globally, wetland loss rates are in-

creasing with sea-level rise (65), which willresult in greater inputs of both TerrPOCand TerrDOC to the coastal ocean. Re-cent work has also shown that the role ofTerrOC in coastal marine sedimentsis likely more important than previouslythought for the long-term sequestrationof atmospheric CO2 (65). For example,land-use changes were found to be moreimportant than changes in climate andplant CO2 fertilization to increasesin Mississippi River water and carbon ex-port from the watershed over the past 50 y(111). Similarly, with the changing view ofmore efficient consumption of TerrOC ininland freshwater systems (61), it is likelythese systems will continue to see greaterconsumption and CO2 efflux with in-creasing global temperature.Carbon leaching (i.e., biogenic Terr-

DOC and DIC) can increase the net lossesfrom croplands by as much as 24% to 105%

in United Kingdom soils (112). Thesedifferences between natural forested andcropland regions have important im-plications for the cycling of TerrOC withthe advent of expanding land-use changesin world. Future methods in soil spectralproperties using remote sensing such asthe cellulose absorption index, the lignin-cellulose absorption index, the normalizeddifferent till index, the normalized differ-ence senescent vegetation index, and nor-malized difference vegetation indices 5and 7, will prove to be extremely usefulwith the rapidly occurring land-usechanges in the world (113).

ConclusionsIn summary, I propose that, if we are ex-amine the role of PEP on the utilization ofTerrOC in global ecosystems, we need todevelop (i) an understanding between theevolution of terrestrial organisms (e.g.,fungi) that evolved biochemical pathwaysto consume recalcitrant molecules (e.g.,lignin) TerrOC on land with the pathwaysfor consuming TerrOC in aquatic systems;(ii) a greater understanding of the com-plex role of enzyme kinetics as it relatesto OM cycling in natural systems; (iii)a broader view of the terminology used todefine the “quality” of TerrOC; (iv) betterexperimental design that allows for pri-mary producer exudates to interact withTerrOC; (v) better incorporation of mo-lecular microbiological techniques to un-derstand the role of short-term strategiesused by microbial populations to use Ter-rOC in different environments; (vi) moreintegrative thinking among classic analyti-cal chemistry, molecular microbiology, andevolutionary ecology; (vii) more cross-breeding of ideas between terrestrial andaquatic ecosystems; and (viii) models thatallow for the inclusion of priming effects inthe global carbon cycling and climatechange. The integration of the foundationsof ecological/evolutionary theory in thecontext of priming combined with the newanalytical advancements in the chemicalcharacterization OC and molecular micro-biology will prove vital in our under-standing of OC cycling in a changingclimate.

ACKNOWLEDGMENTS. Jo Ann Bianchi, KathrynSchreiner, Daniel C. O. Thornton, Richard Smith,and Stella Woodard all read early drafts of thismanuscript. The author also thanks all his col-leagues who are actively working the fluxesand cycling of terrestrially derived organic mat-ter to the global ocean. Much the work fundedto study terrestrially derived organic matter inthe author’s laboratory was supported by theNational Aeronautics Space Administration,Department of Energy, Office of Naval Research,and National Science Foundation.

1. Syvitski JPM, Vörösmarty CJ, Kettner AJ, Green P

(2005) Impact of humans on the flux of terrestrial

sediment to the global coastal ocean. Science 308:

376–380.

2. Hedges JI (1992) Global biogeochemical cycles:

progress and problems. Mar Chem 39:67–93.

Bianchi PNAS | December 6, 2011 | vol. 108 | no. 49 | 19479

Dow

nloa

ded

by g

uest

on

July

2, 2

020

3. Burdige DJ (2005) The burial of terrestrial organiccarbon in marine sediments: A reassessment. GlobalBiogeochem Cycles, 10.1029/2004GB002368.

4. Houghton RA (2007) Balancing the global carbonbudget. Annu Rev Earth Planet Sci 35:313–347.

5. Hansell DA, Olson D, Dentener F, Zamora L (2009)Assessment of excess nitrate development in thesubtropical North Atlantic. Mar Chem 106:562–579.

6. Cai WJ (2011) Estuarine and coastal ocean carbonparadox: CO2 sinks or sites of terrestrial carbonincineration? Ann Rev Mar Sci 3:123–145.

7. Galy V, et al. (2007) Efficient organic carbon burial inthe Bengal fan sustained by the Himalayan erosionalsystem. Nature 450:407–410.

8. Drenzek NJ, et al. (2009) A new look at old carbon inactive margin sediments. Geology 37:239–242.

9. Hedges JI, Oades JM (1997) Comparative organicgeochemistries of soils and marine sediments. OrgGeochem 27:319–361.

10. Bingemann CW, Varner JE, Martin WP (1953) Theeffect of the addition of organic materials on thedecomposition of an organic soil. Soil Sci Soc AmProc 17:34–38.

11. Kuzyakov Y, Friedel JK, Stahr K (2000) Review ofmechanisms and quantification of priming effects.Soil Biol Biochem 32:1485–1498.

12. Löhnis F (1926) Nitrogen availability of greenmanures. Soil Sci 22:253–290.

13. Broadbent FE, Bartholomew WV (1948) The effect ofquantity that then died or became dormant as thecellulose was of plant material added to soil on itsrate of decomposition. Soil SOM Sci Soc Am Proc 13:271–274.

14. Broadbent FE (1956) Tracer investigations of plantresidue de-composition in soil. Proceedings of theConference on Radioactive Isotopes in Agriculture,12-14 January 1956 (US Atomic Energy Commission,Washington, DC), pp 371–380.

15. Blagodatskaya E, Kuzyakov Y (2008) Mechanisms of realand apparent priming effects and their dependenceon soil microbial biomass and community structure:Critical review. Biol Fertil Soils 45:115–131.

16. Banks HP (1975) Reclassification of Psilophyta. Taxon24:401–413.

17. Kendrick P, Crane PR (1997) The origin and earlyevolution of plants on land. Nature 389:33–39.

18. Delwiche CF, Graham LE, Thomson N (1989) Lignin-like compounds and sporopollenin coleochaete, analgal model for land plant ancestry. Science 245:399–401.

19. Martone PT, et al. (2009) Discovery of lignin inseaweed reveals convergent evolution of cell-wallarchitecture. Curr Biol 19:169–175.

20. Goodell B, et al. (1997) Low molecular weightchelators and phenolic compounds isolated fromwood decay fungi and their role in the fungalbiodegradation of wood. J Biotechnol 53:133.

21. Eastwood DC, et al. (2011) The plant cell wall-decomposing machinery underlies the functionaldiversity of forest fungi. Science 333:762–765.

22. Hilldebrant U, et al. (2001) Arbuscular mycorrhizalcolonization of halophytes in Central European saltmarshes. Mycorrhiza 10:175–183.

23. Newell SY, Fell JW (1997) Competition amongmangrove oomycetes, and between oomycetes andother microbes. Aquat Microb Ecol 12:21–28.

24. Takii S (2003) Amino acids as main substrates forsulfate-reducing bacteria in surface sediment ofa eutrophic bay. J Gen Appl Microbiol 49:329–336.

25. Balba MT, Evans WC (1977) The methanogenicfermentation of aromatic substrates. Biochem SocTrans 5:302–304.

26. Benner R, et al. (1984) Anaerobic biodegradation ofthe lignin and polysaccharide components oflignocellulose and synthetic lignin by sedimentmicroflora. Appl Environ Microbiol 47:998–1004.

27. Vainio EJ, Hantula J (2000) Direct analysis of wood-inhabiting fungi using denaturing gradient gelelectrophoresis of amplified ribosomal DNA. MycolRes 104:927–936.

28. Arnosti C (2011) Microbial extracellular enzymes andthe marine carbon cycle. Ann Rev Mar Sci 3:401–425.

29. Weiss MS, et al. (1991) Molecular architecture andelectrostatic properties of a bacterial porin. Science254:1627–1630.

30. Hoppe H-G (1991) Microbial extracellular enzymeactivity: A new key parameter in aquatic ecology.Microbial Enzymes in Aquatic Environments, edChrost RJ (Springer, Berlin), pp 60–83.

31. Mayer LM, Schick LL, Hardy KH, Wagai R, McCarthy JF(2004) Organic matter in small mesopores insediments and soil. Geochim Cosmochim Acta 68:3863–3872.

32. McBride MJ, et al. (2009) Novel features of thepolysaccharide-digesting gliding bacterium Flavo-bacterium johnsoniae as revealed by genomesequence analysis. Appl Environ Microbiol 75:6864–6875.

33. Velicer GJ (2003) Social strife in the microbial world.Trends Microbiol 11:330–337.

34. Allison SD (2005) cheaters, diffusion and nutrientsconstrain decomposition by microbial enzymes inspatially structured environments. Ecol Lett 8:626–635.

35. Moldreup P, Olesen T, Komatsu T, Schjonning P,Rolston DE (2001) Tortuosity, diffusivity, andpermeability in the soil liquid and gaseous phases.Soil Sci Soc Am J 65:613–623.

36. Grossart H-P, Tang KW, Kiørboe T, Ploug H (2007)Comparison of cell-specific activity between free-living and attached bacteria using isolates andnatural assemblages. FEMS Microbiol Lett 266:194–200.

37. Gram L, Grossart H-P, Schlingloff A, Kiørboe T (2002)Possible quorum sensing in marine snow bacteria:Production of acylated homoserine lactones byRoseobacter strains isolated from marine snow. ApplEnviron Microbiol 68:4111–4116.

38. Turner RD (1966) A survey and illustrated catalogueof the Teredinidae (Mollusca; Bivalvia) (Museum ofComparative Zoology, Cambridge).

39. Coleman DC, Crossley DA (1996) Fundamentals of SoilEcology (Academic, San Diego).

40. Lee SY (1989) The importance of Sesarminae crabsChiromanthes spp. and inundation frequency on thedecomposition of mangrove (Kandelia candel (L.)Druce) leaf litter in a Hong Kong tidal shrimp pond.J Exp Mar Biol Ecol 131:23–43.

41. Ekschmitt K, Liu M, Vetter S, Fox O, Woters V (2005)Strategies used by soil biota to overcome soil organicmatter stability - why is dead organic matter left overin the soil? Geoderma 128:167–176.

42. Darwin CR (1881) The Formation of Vegetable Mould,Through the Action of Worms, with Observations onTheir Habits (John Murray, London).

43. Davidson EA, Janssens IA (2006) Temperature sensitivityof soil carbon decomposition and feedbacks to climatechange. Nature 440:165–173.

44. Hedges JI, Keil RG, Benner R (1997) What happens toterrestrial organic matter in the ocean? Org Geochem27:195–212.

45. Schöning I, Totsche KU, Kögel-Knabner I (2006) Smallscale spatial variability of organic carbon stocks inlitter and solum of a forested Luvisol. Geoderma136:631–642.

46. Keil RG, Tsamakis E, Fuh CB, Giddings C, Hedges JI(1994) Mineralogical and textural controls onorganic composition of coastal marine sediments:Hydrodynamic separation using SPLITT fractionation.Geochim Cosmochim Acta 58:879–893.

47. Guggenberger G, Kaiser K (2003) Dissolved organicmatter in soil: Challenging the paradigm of sorptivepreservation. Geoderma 113:293–310.

48. Cole JJ, et al. (2007) Plumbing the global carbon cycle:Integrating inland waters into the terrestrial carbonbudget. Ecosystems (NY) 10:171–184.

49. Butman D, Raymond P, Oh NH, Mull K (2007) Quantity14C age and lability of desorbed soil organic carbon infresh water and seawater. Org Geochem 38:1547–1557.

50. Kaiser K, et al. (2002) The comparison of dissolvedorganic matter in forest soil solutions: changesinduced by seasons and passage through themineral soil. Org Geochem 33:307–318.

51. Raymond PA, Saiers JE (2010) Event controlled DOCexport from forested watersheds. Biogeochemistry100:197–209.

52. Dalzell BJ, Filley TR, Harbor JM (2005) Flood pulseinfluences on terrestrial organic matter export froman agricultural watershed. J Geophys Res 110:G02011.

53. Raymond PA, Caraco NF, Cole JJ (1997) Carbondioxide concentration and atmospheric flux in theHudson River. Estuaries 20:381–390.

54. Mayorga E, et al. (2005) Young organic matter asa source of carbon dioxide outgassing from Amazonianrivers. Nature 436:538–541.

55. Battin TJ, et al. (2008) Biophysical controls on organiccarbon fluxes in fluvial networks. Nature 1:95–100.

56. Vannote RL, Minshall GW, Cummings KW, Sedell JR,Cushing CE (1980) The River Continuum Concept. CanJ Fish Aquat Sci 37:130–137.

57. Hullar MA, Kaplan LA, Stahl DA (2006) Recurringseasonal dynamics of microbial communities instream habitats. Appl Environ Microbiol 72:713–722.

58. Junk WJ, Bayley PB, Sparks RE (1989) The flood-pulseconcept in river-floodplain system. Proceedings of theInternational Large River Symposium (Dodge DP, ed).Special publication. Can J Fish Aquat Sci 106:110–127.

59. Meybeck M (1982) Carbon, nitrogen, and phosphorustransport by world rivers. Am J Sci 282:401–450.

60. Hilton RG, Galy A, Hovius N, Horng MJ, Chen H (2011)Efficient transport of fossil organic carbon to theocean by steep mountain rivers: An orogenic carbonsequestration mechanism. Geol 39:71–74.

61. Dodds WK, Cole J (2007) Expanding the concept oftrophic state in aquatic ecosystems: It’s not just theautotrophs. Aquat Sci 69:427–439.

62. Raymond PA, Bauer JE (2001) Riverine export of agedterrestrial organic matter to the North AtlanticOcean. Nature 409:497–500.

63. McKnight DM, et al. (1992) Sorption of dissolvedorganic carbon by hydrous aluminum and ironoxides occurring at the confluence of Deer Creekwith the Snake River, Summit County, Colorado.Environ Sci Technol 26:1388–1396.

64. Aufdenkampe AK, Hedges JI, Richey JE, Krusche AV,Llerena CA (2001) Sorptive fractionation of dissolvedorganic nitrogen and amino acids onto fine sedimentswithin the Amazon Basin. Limnol Oceanogr 46:1921–1935.

65. Bianchi TS, Allison MA (2009) Large-river delta-frontestuaries as natural “recorders” of global environ-mental change. Proc Natl Acad Sci USA 106:8085–8092.

66. Wissmar RC, Richey JE, Stallard RF, Edmond JM (1981)Plankton metabolism and carbon processes in theAmazon River, its tributaries, and floodplain waters,Peru-Brazil, May-June 1977. Ecology 62:1622–1633.

67. Kendall C, Steven RS, Kelly VJ (2001) Carbon andnitrogen isotopic compositions of particulateorganic matter in four large river systems across theUnited States. Hydrol Process 15:1301–1346.

68. Hedges JI, Keil R (1995) Sedimentary organic matterpreservation; an assessment and speculative synthesis.Mar Chem 49:81–115.

69. Schlunz B, Schneider RR (2000) Transport of terrestrialorganic carbon to the oceans by rivers: Re-estimatingflux- and burial rates. Int J Earth Sci 88:599–606.

70. McCarthy JF, et al. (2008) Protection of organiccarbon in soil microaggregates via restructuring ofaggregate porosity and filling of pores withaccumulating organic matter. Geochim CosmochimActa 72:4725–4744.

71. Aller RC (1998) Mobile deltaic and continental shelfmuds as suboxic, fluidized bed reactors.Mar Chem 61:143–155.

72. Kristensen E, et al. (2000) Carbon and nitrogenmineralization in sediments of the Bangrongmangrovearea, Phuket, Thailand. Aquat Microb Ecol 22:199–213.

73. Druffel ERM, Williams PM, Bauer JE, Ertel JR (1992)Cycling of dissolved and particulate organic matter inthe open ocean. J Geophys Res 97:15,639–15,659.

74. Meyers-Schulte KJ, Hedges JI (1986) Molecularevidence for a terrestrial component of organicmatter dissolved in ocean water. Nature 321:61–63.

75. Opsahl S, Benner R (1997) Distribution and cycling ofterrigenous dissolved organic matter in the ocean.Nature 386:480–482.

76. Hernes PJ, Benner R (2003) Photochemical andmicrobial degradation of dissolved lignin phenols:implications for the fate of terrigenous dissolvedorganic matter in marine environments. J GeophysRes: Oceans 108:3291.

77. Rachold V, et al. (2004) Modern terrigenous organiccarbon input to the Arctic Ocean. The Organic Carbon

19480 | www.pnas.org/cgi/doi/10.1073/pnas.1017982108 Bianchi

Dow

nloa

ded

by g

uest

on

July

2, 2

020

Cycle in the Arctic Ocean, eds Stein RS,Macdonald RW (Springer, New York), pp 33–55.

78. Miller WL, Moran MA (1997) Interaction ofphotochemical and microbial processes in thedegradation of refractory dissolved organic matterfrom a coastal marine environment. Limnol Oceanogr42:1317–1324.

79. Benner R, Kaiser K (2011) Biological and photochemicaltransformations of amino acids and lignin phenolsin riverine dissolved organic matter. Biogeochemsity102:209–222.

80. Coynal A, Seyler P, Etcheber H, Meybeck M, Orange D(2005) Spatial and seasonal dynamics of totalsuspended sediment and organic carbon species inthe Congo River. Global Biogeochem Cycles 19:GB4019.

81. Hertkorn N, et al. (2006) Characterization of a majorrefractory component of marine dissolved organicmatter. Geochim Cosmochim Acta 70:2990–3010.

82. Abdulla HA, Minor EC, Hatcher PG (2010) Using two-dimensional correlations of ¹³C NMR and FTIR toinvestigate changes in the chemical composition ofdissolved organic matter along an estuarinetransect. Environ Sci Technol 44:8044–8049.

83. Kujawinski EB (2011) The impact of microbialmetabolism on marine dissolved organic matter.Ann Rev Mar Sci 3:567–599.

84. Horvath RS (1972) Microbial co-metabolism and thedegradation of organic compounds in nature.Bacteriol Rev 36:146–155.

85. Jenkinson DS, Fox RH, Rayner JH (1985) Interactionsbetween fertilizer nitrogen and soil nitrogen - the so-called ‘priming’ effect. J Soil Sci 36:425–444.

86. Guenet B, Danger M, Abbadie L, Lacroix G (2010)Priming effect: Bridging the gap between terrestrialand aquatic ecology. Ecology 91:2850–2861.

87. Hamer U, Marschner B, Brodowski S, Amelung W(2004) Interactive priming of black carbon andglucose mineralisation. Org Geochem 35:823–830.

88. Guggenberger G, Zech W (1993) Dissolved organiccarbon control in acid forest soils of the Fichtelge-birge (FRG) as revealed by distribution patterns andstructural composition analysis. Geoderma 59:109–129.

89. Baldock JA, Skjemstad JO (2000) Role of the soilmatrix and minerals in protecting natural organicmaterials against biological attack. Org Geochem31:697–710.

90. Fontaine S, et al. (2007) Stability of organic carbon indeep soil layers controlled by fresh carbon supply.Nature 450:277–280.

91. de Haan H (1977) Effect of benzoate on microbialdecomposition of fulvic acids in Tjeukemeer(Netherlands). Limnol Oceanogr 22:38–44.

92. van Nugteren P, et al. (2009) Seafloor ecosystemfunctioning: The importance of organic matterpriming. Mar Biol 156:2277–2287.

93. Canfield DE (1994) Factors influencing organic carbonpreservation in marine sediments. Chem Geol 114:315–329.

94. Turnewitsch R, Domeyer B, Graf G (2007)Experimental evidence for an effect of early-diagenetic interaction between labile and refractorymarine sedimentary organic matter on nitrogendynamics. J Sea Res 57:270–280.

95. Vähätalo AV, Aarnos H, Mäntyniemi S (2010)Biodegradability continuum and biodegradationkinetics of natural organic matter described by thebeta distribution. Biogeochemistry 100:227–240.

96. Carlson CA, Hansell DA, Tamburini C (2011) DOCpersistence and its fate after export within theocean interior. Microbial Carbon Pump in theOcean, eds Jiao N, Azam F, Sanders S (AAAS,Washington, DC), pp 57–59.

97. Cherrier J, Bauer JE, Druffel ERM (1996) Utilizationand turnover of labile dissolved organic matter bybacterial heterotrophs in eastern North Pacificsurface waters. Mar Ecol Prog Ser 139:257–279.

98. Cherrier J, Bauer JE, Druffel ERM, Coffin RB,Chanton JP (1999) Radiocarbon in marine bacteria:Evidence for the ages of assimilated carbon. LimnolOceanogr 44:730–736.

99. Carlson CA, Ducklow HW (1996) Growth ofbacterioplankton and consumption of dissolvedorganic carbon in the oligotrophic Sargasso Sea.Aquat Microb Ecol 10:69–85.

100. Poretsky RS, Sun S, Mou X, Moran MA (2010)Transporter genes expressed by coastal bacterio-plankton in response to dissolved organic carbon.Environ Microbiol 12:616–627.

101. DeLong EF, et al. (2006) Community genomics amongstratified microbial assemblages in the ocean’sinterior. Science 311:496–503.

102. Kirchman DL, Elifantz H, Dittel AI, Malmstrom R,Cottrell MT (2007) Standing stocks and activity ofArchaea and Bacteria in the western Arctic Ocean.Limnol Oceanogr 52:495–507.

103. Petch ST, et al. (2001) 14C-dead living biomassevidence for microbial assimilation of ancientorganic carbon during shale weathering. Science252:1127–1131.

104. Coolen MJL, Overmann J (2000) Functional exo-enzymes as indicators of metabolically activebacteria in 124,000-year-Old sapropel layers of theeastern Mediterranean Sea. Appl Environ Microbiol66:2589–2598.

105. Vishnivetskaya TA, et al. (2006) Bacterial communityin ancient Siberian permafrost as characterized byculture and culture-independent methods. Astro-biology 6:400–414.

106. Spencer RGM, et al. (2009) Utilizing chromophoricdissolved organic matter measurements to deriveexport and reactivity of dissolved organic carbonexported to the Arctic Ocean: A case study of theYukon River, Alaska. Geophys Res Lett 36:L06401.

107. Hernes PJ, Benner B (2006) Terrigenous organicmatter sources and reactivity in the North AtlanticOcean and a comparison to the Arctic and Pacificoceans. Mar Chem 100:66–79.

108. Drenzek NJ, et al. (2007) Constraints on the origin ofsedimentary organic carbon in the Beaufort Sea fromcoupled molecular 13C and14C measurements. MarChem 103:146–162.

109. Kuzyk ZZA, Goni MA, Stern GA, Macdonald RW(2008) Sources, pathways and sinks of particulateorganic matter in Hudson Bay: Evidence from lignindistributions. Mar Chem 112:215–229.

110. Chmura GL, Anisfeld SC, Cahoon DR, Lynch JC (2003)Global carbon sequestration in tidal, saline wetlandsoils. Global Biogeochem 17:22-1–22-14.

111. Raymond PA, Oh NH, Turner RE, Broussard W (2008)Anthropogenically enhanced fluxes of water andcarbon from the Mississippi River. Nature 451:449–452.

112. Kindler R, et al. (2010) Dissolved carbon leachingfrom soil is a crucial component of the netecosystem carbon balance. Glob Change Biol 17:1167–1185.

113. Serbin G, Daughtry CST, Hunt ER, Brown DJ,McCarty GW (2009) Effect of soil spectral propertieson remote sensing of crop residue cover. Soil WaterManag Conserv 73:1545–1558.

114. Drenzek NJ (2007) The temporal dynamics ofterrestrial organic matter transfer to the oceans:Initial assessment and application. PhD thesis (MIT-Woods Hole Oceanographic Institution, WoodsHole, MA).

Bianchi PNAS | December 6, 2011 | vol. 108 | no. 49 | 19481

Dow

nloa

ded

by g

uest

on

July

2, 2

020

Correction

PERSPECTIVECorrection for “The role of terrestrially derived organic carbon inthe coastal ocean: A changing paradigm and the priming effect,”by Thomas S. Bianchi, which appeared in issue 49, December 6,2011, of Proc Natl Acad Sci USA (108:19473–19481; first publishedNovember 21, 2011; 10.1073/pnas.1017982108).The author notes “Although I included a citation to and a figure

fromGuenet et al. in my article, the importance of this reference indeveloping the argument of linkage between terrestrial and aquaticcarbon was not properly detailed. Guenet et al. was the first to notethe potential interactions between terrestrial and aquatic carbonand the importance of priming effect (PE) in aquatic ecosystems.Additionally, the work for Guenet et al. should be cited in thefollowing sections of the article:“On page 19477, left column, second full paragraph, lines 27–

30, ‘All the aforementioned scenarios represent the simplifiedpathways established for priming in soil types’ should instead ap-pear as ‘All the aforementioned scenarios represent the simplifiedpathways established for priming in soils and natural waters (86).’“Also on page 19477,middle column, the legend for Fig. 3 should

instead appear as ‘Three different priming effects involving LOMand ROM, respectively. These proposed priming effects are soilsand natural waters derived from work in soils. Modified fromGuenet et al. (86).’“Lastly, on page 19477, middle column, first full paragraph, lines

10–14, ‘The mineralization of charcoal, a very recalcitrant form ofOC, was shown to increase between 36% and 189% (vs. control)when primed with glucose (87)’ should instead appear as ‘Themineralization of charcoal, a very recalcitrant form of OC, wasshown to increase between 36% and 600% (vs. control) whenprimed with glucose (87).’”

www.pnas.org/cgi/doi/10.1073/pnas.1202757109

5134 | PNAS | March 27, 2012 | vol. 109 | no. 13 www.pnas.org