Geological respiration of a mountain belt revealed by the trace element rhenium

HAL Id: insu-02163277

https://hal-insu.archives-ouvertes.fr/insu-02163277

Submitted on 24 Jun 2019

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of

sci-entific research documents, whether they are

pub-lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diffusion de documents

scientifiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

Geological respiration of a mountain belt revealed by

the trace element rhenium

Robert Hilton, Jérôme Gaillardet, Damien Calmels, Jean-Louis Birck

To cite this version:

Robert Hilton, Jérôme Gaillardet, Damien Calmels, Jean-Louis Birck. Geological respiration of a

mountain belt revealed by the trace element rhenium. Earth and Planetary Science Letters, Elsevier,

2014, 403, pp.27-36. �10.1016/j.epsl.2014.06.021�. �insu-02163277�

Earth and Planetary Science Letters 403 (2014) 27–36

Contents lists available atScienceDirect

Earth

and

Planetary

Science

Letters

www.elsevier.com/locate/epsl

Geological

respiration

of

a

mountain

belt

revealed

by

the

trace

element

rhenium

Robert

G. Hilton

a

,

b

,

∗

,

Jérôme Gaillardet

b

,

Damien Calmels

c

,

d

,

Jean-Louis Birck

b

a_{DepartmentofGeography,DurhamUniversity,ScienceLaboratories,SouthRoad,Durham,DH13LE,UK}

b_{EquipedeGéochimieetCosmochimie,InstitutdePhysiqueduGlobedeParis,SorbonneParisCité,UnivParisDiderot,UMR7154CNRS,F-75005Paris,France} c_{EquipedeGéochimiedesIsotopesStables,InstitutdePhysiqueduGlobedeParis,SorbonneParisCité,UnivParisDiderot,UMR7154CNRS,F-75005Paris,France} d_{DepartmentofEarthSciences,UniversityofCambridge,Cambridge,CB23EQ,UK}

a

r

t

i

c

l

e

i

n

f

o

a

b

s

t

r

a

c

t

Articlehistory:

Received7January2014

Receivedinrevisedform9June2014 Accepted12June2014

Availableonline9July2014 Editor: T.M.Harrison Keywords: carboncycle organiccarbon rhenium weathering erosion mountainrivers

Oxidationofrock-derived,petrogenic,organiccarbon(OCpetro)duringweathering ofsedimentaryrocks

isamajorsourceofcarbondioxide(CO2)totheatmosphere.Thisgeologicalrespirationisthought to

beenhancedbyphysicalerosion,suggestingthatmountainbeltscouldreleaselargeamountsofCO2to

counter theCO2 sequestrationachievedbythe erosion,riverinetransferandoceanic burialoforganic

carbonfromtheterrestrialbiosphere.However,OCpetrooxidationratesinmountainbeltshavenotbeen

quantified.Hereweuserhenium(Re)asaproxytotrackOCpetrooxidationinmountainrivercatchments

ofTaiwan,whereexistingmeasurementsofphysicalerosionrateallowthecontrolsonOCpetrooxidation

tobeassessed.RehasbeenshowntobecloselyassociatedwithOCpetroinrocksandfollowingoxidation

duringchemicalweatheringformsasolubleoxyanion(ReO₄−)whichcontributestothedissolvedloadof rivers.Soilsonmeta-sedimentaryrocksinTaiwanshow thatRe lossiscoupledtoOCpetro lossduring

weathering, confirming previous observations from soil profiles on sedimentary rocks elsewhere. In Taiwan rivers, dissolvedRe fluxincreaseswith thecatchment-average sedimentyield, suggestingthat physicalerosionrateisamajorcontrolonOCpetrooxidation.BasedonourcurrentunderstandingofRe

mobilityduringweathering,thedissolvedRefluxcanbeusedtoquantifyanupperboundontheOCpetro

oxidationrate andthe associatedCO2 transfer.TheestimatedCO2 releasefromthismountainbelt by

OCpetro oxidationdoesnot negateestimatesof CO2 sequestration byburial ofbiosphericOC offshore.

Thefindingsare comparedtoOC transfersestimatedforthe Himalaya,whereOCpetro oxidationinthe

mountainbeltremainsunconstrained.Together,thesecasessuggestthatmountainbuildinginthetropics canresultinanetsinkofOCwhichsequestersatmosphericCO2.

CrownCopyright©2014PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBY license(http://creativecommons.org/licenses/by/3.0/).

1. Introduction

Organic matter within sedimentary rocks constitutes a vast stockofcarbonthatwas sequesteredfromtheatmosphereinthe geologicalpast,containing

∼

15

×

1015_{gC whichis}

_∼

_{25 000times}

the carbon content of the pre-industrial atmosphere (Sundquist andVisser, 2005). Oxidation of this rock-derived, or ‘petrogenic’, organiccarbon(OCpetro) during weatheringatEarth’ssurface isa

major source of CO2 to the atmosphere and sink of O2 (Berner

andCanfield,1989; DerryandFrance-Lanord,1996).Better under-standingthebalancebetweenOCpetro oxidationandthe

sedimen-tary burial ofrecently photosynthesizedorganiccarbon fromthe terrestrialbiosphere (OCbiosphere)isfundamental toassessinghow

*

Correspondingauthorat:DepartmentofGeography,DurhamUniversity,Science Laboratories,SouthRoad,Durham,DH13LE,UK.Tel.:+4401913341970.

E-mailaddress:r.g.hilton@durham.ac.uk(R.G. Hilton).

atmosphericCO2andO2concentrationshaveevolvedover

geolog-icaltime(Hayesetal.,1999; HayesandWaldbauer,2006).OCpetro

oxidationisthoughttooccurwhensedimentaryrocksareexposed toaqueousandgasphase O2 (Chang andBerner, 1999), withthe

rate of CO2 release controlled by the supply of OCpetro to react

(Petschet al.,2000; Bolton etal., 2006).As such, mountainbelts wherehighratesofphysicalerosioncansupplyabundant OCpetro

to the surface (Galy et al., 2008a; Hilton et al., 2011) may be locationswherethisCO2sourceismostpotent.Iftrue,OCpetro

ox-idationmaycountertheCO2consumptionachievedbytheerosion

ofOCbiosphereanditsefficientpreservationandburialin

sedimen-tary deposits of these settings (France-Lanord and Derry, 1997; Stallard,1998; Galyetal.,2007; Kaoetal.,2014).Inordertoassess thenetimpactofmountainbuildingonCO2fluxestoandfromthe

atmosphere(Caldeiraetal.,1993; DerryandFrance-Lanord,1996; Gaillardet andGaly, 2008), the rates of OCpetro oxidation during

weatheringinmountainbeltsmustbequantified.

http://dx.doi.org/10.1016/j.epsl.2014.06.021

OCpetro oxidation hasbeenshownto be significantduring the

floodplaintransportofclasticsedimentserodedfromtheHimalaya and Andes mountain ranges in the Ganges and Amazon rivers, respectively (Galy et al., 2008a; Bouchez et al., 2010). However themethodemployed,whichusedtheradiocarbon contentofOC in riversuspended load todetermine OCpetro loss fromthe solid

phase, did not have the resolution necessary to quantify OCpetro

oxidation rate in the mountain catchments where it should be mostrapid.Thisisbecausechemicaldenudationratesare usually

∼

0.01–0.1 times lower than physical denudation rates in moun-tain belts (Jacobsonand Blum, 2003; West et al., 2005), so sus-pended sediments in mountain rivers have a geochemical com-position similar to bedrocks (Hilton et al., 2010). Therefore it is difficultto usethesolid products ofweatheringcarriedby rivers toquantifyOCpetrooxidationinmountaincatchments(Hiltonetal.,

2011).Analternativeapproachistotracktheproductsofchemical weatheringcarriedinthedissolvedloadofrivers(Meybeck,1987; Gaillardetetal.,1999) whichhasprovideddetailedconstraintson chemical weathering processes and inorganiccarbon fluxes from mountaincatchments(JacobsonandBlum,2003; Westetal.,2005; Calmelsetal.,2011).

Thetraceelementrhenium(Re) isastrongcandidatetoactas a proxyfor OCpetro oxidation during weathering assuggested by

previous work(Dalaiet al.,2002; Jaffe etal., 2002) dueto: i)its associationwithOCpetroinrocks;andii)itsredox-dependent

solu-bility,factorswhicharecloselylinked.Inoxygenatedwaters(Eh

>

0 V),Reisdominantlypresentasthesolubleperrhenateoxyanion (ReO−₄)overawiderangeofpHvalues(5.5to9.5)(Brookins,1986) and in seawater Re behaves conservatively (Anbar et al., 1992; Colodner etal., 1993). In marine depositional environments, OC-bearing sediments experience reduction of ReO−₄ contained in seawater and Re forms a solid phase (Colodner et al., 1993; Crusiusetal., 1996; CrusiusandThomson,2000).Isolationof or-ganic matter from marine sedimentary rocks has suggested that ReiscomplexedtoOCpetrointhedepositionalenvironment,

favor-ing organic chelating sites to co-existing sulfide phases, both in marinesedimentswithhighorganiccontents(Cohenetal.,1999; Selby and Creaser, 2003) and relatively organic-poor sediments (Pierson-Wickmannetal.,2002; Rooneyetal.,2012).

When these sedimentary rocks are re-exposed, for example by erosion andexhumation during orogenesis (e.g. Hilton et al., 2011), Re-bearing OCpetro is subject to chemical weathering in

oxic surface environments (Petsch et al., 2000; Jaffe etal., 2002; Pierson-Wickmannetal.,2002).Previousweatheringstudiesfrom soilson OC-richrocks havedemonstrated that Re loss tracksthe OCpetroloss(Jaffeetal.,2002).Incontrast,oxidativeweatheringof

sulfidemineralscontinuedmuchdeeperintothesoilprofilewhere OCpetroandRewerestillfoundatconcentrationssimilartothe

un-weatheredbedrock (Jaffe etal., 2002). While the fateof gaseous CO2 releasedbyOCpetrooxidationisdifficulttotraceatthe

catch-mentscale(KellerandBacon,1998),Rewillbeoxidizedtothe sol-ubleReO−₄ (Brookins,1986).Reisnotthoughttobecycledthrough the modern terrestrial biosphere (cf. to molybdenum which is used in nitrogenase), meaning that the mobilized ReO−₄ enters the hydrological network (Brookins, 1986; Colodner et al., 1993; Milleretal.,2011).Riversintegratechemicalreactionsand hydro-logicalsources acrossthe landscape(Gaillardet etal., 1999), thus thedissolvedRe fluxcanbe interpretedastheresultofoxidative weatheringreactionsoccurringupstream(Dalaietal.,2002). Build-ingonpreviouswork,whichhasestablishedthecloseaffinityofRe andOCpetro insedimentaryrocks andtheir coupledbehavior

dur-ingoxidativeweathering,OCpetrooxidationratesmaybequantified

fromthedissolvedRe flux,providedtheRe toOCpetro ratioofthe

rockswhichhaveundergoneweatheringisknown.

Here we examine the transfer of Re in mountainrivers with higherosionrateswithaviewtoassessingthecontrolsonOCpetro

oxidation, while estimating the associated CO2 release across a

mountain belt for the first time. We focus on major rivers in Taiwan, whereriver gauging recordsconstrain therates and pat-terns ofphysicalerosion(Dadsonetal., 2003) andprevious work has quantified theerosion ofOCbiosphere (Hilton et al., 2012) and

tracked its preservationin marine sedimentsoffshore (Kaoet al., 2014).Forthefirsttime,thisallowstherole oferosiononOCpetro

oxidation tobe assessed,andmeansthat anOC budgetiswithin reachforthemountainbelt.WeexaminethemobilityofRefrom the solidphase duringchemical weatheringinsoilsfromTaiwan, and measure the dissolved Re concentration inthe major rivers. WediscussindetailtheassumptionsrequiredtouseReasa quan-titative proxyof OCpetro oxidation, inlight of thenew datafrom

TaiwanesesoilsandpreviousworkonRemobilityduring weather-ing.OurestimatesofdissolvedRefluxprovidenewinsightonthe major controls onOCpetro oxidationand providean upper bound

estimateoftheassociatedCO2release,allowingustobetterassess

theimpactoforogenesisonthecarboncycle. 2. Materialsandmethods

2.1. Setting

TaiwanformedbythecollisionoftheLuzonArc onthe Philip-pine Sea Plate with the Eurasian continental margin. Rapid tec-tonicupliftandseismicitycombinewithatropicalcycloneclimate to drive high physical erosion rates, coupling river incision and bedrock landsliding on steep slopes (Hovius et al., 2000). Sus-pended sediment yields havebeen quantified from1970 for the major rivers and vary around the island, reaching amongst the highestintheworldat

>

10 000t km−2yr−1 (Dadsonetal.,2003). Rapidphysicalerosionandturnoverofthelandscapebylandslides resultsinthinsoils(

<

0.8 m) (Tsai etal.,2001) andhighratesof erosionandfluvialexportofOCbiosphere (Hiltonetal.,2012) which

isexported totheoceanandpreservedefficientlyinmarine sedi-ments(Kaoetal.,2014).

Exhumation in theCentral Range has exposed passivemargin sedimentaryrockswhichcontain

∼

0.2–0.4weight%OCpetro(Hilton

et al., 2010). Peakmetamorphic temperatures decreasefrom east (

∼

500◦C)towest(

<

150◦C)acrosstherange(Beyssacetal.,2007). RamanSpectroscopyrevealsthatOCpetrocanbepresentasgraphite

intheTananaoSchistandPilushanformations(Fig.S1)inthe east-ern flank of the Central Range mountains (Beyssac etal., 2007). Deep-seatedbedrocklandslidescan mobilizeunweatheredOCpetro

along with surface soilsandcontribute OCpetro to the suspended

loadofTaiwaneserivers(Hiltonetal.,2010),asinothermountain riversystems(Clarketal.,2013).Onceinthefluvial network,the suspendedsedimentandOCpetro israpidlyexportedwithminimal

timeforsubsequentoxidation(Hiltonetal.,2008a, 2011).Asa re-sult, thesolid loadofmountainrivers inTaiwanisdominatedby unweatherederosion products(SelvarajandChen,2006) and Tai-wan delivers

∼

1.7 MtC yr−1 _ofOC

petro tothe oceanin solidform

(Hiltonetal.,2011).

2.2. Samples

River water samples were collected for this study and com-bined withexisting samplesofriver water(Calmels etal., 2011), river bed materials (Hilton et al., 2010) and soils (Hilton et al., 2013). River waters for this study were collected from major rivers in Taiwan during the typhoon season in September 2007 following several days rainfall (Table S1). To establish the hy-drological controls on Re transfer, four additional samples were analyzed from a suite collected in 2004 (Calmels et al., 2011) during a typhoon flood when daily runoff (water discharge nor-malized by the upstream area, mm day−1_{) reached}

_>

_{10 times}

R.G. Hilton et al. / Earth and Planetary Science Letters 403 (2014) 27–36 29

the long-term mean. Surface waters from turbulent rivers were collected using established methods for Re (Dalai et al., 2002; Milleretal., 2011) inthoroughlyrinsedLDPEbottlesandfiltered immediatelythrough 0.2 μm nylon filtersinto acid-cleaned trace analysisgradeNalgene LDPEbottlesrinsedwithfilteredriver wa-ter,thenacidifiedtopH

∼

2 (trace analysisgradeHNO3).An

addi-tionalun-acidified aliquotwas collectedformajoranions.Filtered samples were stored at 4◦C in the dark. Daily water discharge (m3/s)wasobtainedfromtheWaterResourcesAgency,Taiwan.

Toassessthegeochemicalcompositionofrocksatthescaleof rivercatchments(e.g.Tipperetal., 2006; Galyetal.,2008b)river bedmaterials previously collected from three major river catch-ments(Hiltonetal., 2010) were analyzed.Thesedrainarangeof geologicalformations which arerepresentative ofthe wider Cen-tralRange,Taiwan(Fig.S1;TableS2).AsdescribedbyHiltonetal. (2010),

∼

500cm3 _{ofsandmaterial were collectedfromalluvium}

deposited in bedrock channels at low flow in March 2006. The majority of these sediments were deposited following the flood associatedwith Typhoon LongWang (October 2005) (Turowskiet al.,2008) and were likelytohave beensourced by100–1000s of bedrocklandslidesupstream(Hoviusetal.,2000).Giventhe tran-sient nature of sediment storage in Taiwan rivers and the short fluvialtransittimes(Hiltonetal.,2008a),theseriverbedmaterials can provide constraint on catchment-averaged bedrock composi-tion.

ToassesswhetherthecoupledlossofReandOCpetro observed

elsewhereinOC-rich bedrocksoccursduring weatheringofrocks in Taiwan, solid weathering products collected from upper soil horizonsinthe Wulucatchment were used (Fig.S1). These were collectedbyHiltonetal. (2013),as

∼

500 cm3 _{ofsoiloveradepth}

of

∼

10 cm, corresponding to A and E Horizons of variable hu-mifiedorganicmatterwithcoarseandfinefractionsbearinglittle structureofexisting bedrock. Priorto analysis, soilandriverbed materials samples were each homogenized to an integrated bulk sample.

2.3.Rheniumconcentration

Re concentrations in river waters ([Re]diss) are typically low

(pmol L−1_{) and so an established anion-exchange column}

chem-istry technique was used to pre-concentrate Re andremove the sample matrix(Huffmanetal., 1956; Birck etal., 1997). In sum-mary,100mlofsamplewasloadedonto3 mlofAG1-X8(200–400 mesh) resin at 0.1 NHNO3 andeluted to cleaned Teflon at4 N

HNO3. Re recovery was quantified using standard solutions

(Re-filament dissolved in HNO3) and matrix effects assessed by Re

addition to riversamples. Purified eluted residues were re-taken in 3% HNO3 and analyzed by quadrupole

inductively-coupled-plasma mass spectrometry (Q-ICP-MS). The full procedural blank was1.8 pg,

∼

2%ofthetypicalsamplemassofRe,similarto pub-lishedmethods(Dalaietal.,2002).Standardswithsimilar concen-trationsas samples not used to constructcalibration curves had average precision of 3%. Sample reproducibility was 5% which is taken asthe precision on the analyses. Recovery during column chemistrywas100%withinthisprecision.Toanalyzesmaller vol-umewater samples (

<

20 ml) we developed a standard-addition (SA)methodforwhich pre-concentrationisnot requiredandthe useof HNO3 isminimized, the major blankcontributor (Bircket

al.,1997).AsetofsamplesanalyzedbyQ-ICP-MSfollowing anion-exchangechemistry were also analyzedby SA, usinga minimum ofthree additions, withstrongagreement betweenthe measure-mentsacrossthedataset( y

=

0

.

96

±

0

.

02x,r

=

0

.

99, P

<

0

.

0001) suggesting[Re]dissbySA-Q-ICP-MSisprecisetobetterthan10%in

thesesamples.

Insolid samples,[Re] was determinedby isotope dilution fol-lowing established methods (Birck et al., 1997). 50 mg of

pow-der was weighed to Teflon bombs and a 185_{Re enriched-spike}

was addedtoeach sample. Followingequilibration,digestionwas achievedwith4mlofconcentrateddistilledHF:HNO3at150◦C.Re

was extractedusing3-methyl-1-butanol (iso-amylol)liquid–liquid extraction(Bircketal.,1997).Thepurifiedsamplewasre-takenin 3% HNO3,analyzedbyQ-ICP-MS andcorrectedfortheprocedural

blank.

2.4. Majorionandelementalconcentrations

Inriver water,major ionconcentrationswere analyzedby Ion Chromatography (Table S1).In riverbedmaterials andsoils,total sulfur concentrationswere determined by combustion andmajor elements(CaOandAl2O3)determinedbyICP-OESatSARM,CRPG,

Nancy (Table S2). The OC and nitrogen concentrations (%) were measured in previous work following inorganic carbon removal (Hiltonetal.,2010, 2013).Theradiocarbonactivityofsoilsamples was alsodetermined by Hilton etal. (2013) andisreportedhere asthefractionmodern(Fmod)correctedto

−

25

h δ

13CVPDB based

on measured stable isotope composition, where Fmod

=

1 is the 14_{C contentof1950atmosphereandF}

mod

=

0isa sample which

containsnomeasurable14C(StuiverandPolach,1977).TheOCpetro

content ofsoil sampleswas quantified using Fmod measurements

andabinarymixingmodelasoutlinedinAppendix A. 3. Results

3.1. DissolvedrheniuminTaiwanrivers

Inriverwaters,[Re]dissrangefrom4

.

7

±

0

.

2 pmol L−1 to25

.

4

±

1

.

2 pmol L−1 (TableS1)whichare towardthehigherendof pub-lished data from global rivers (Fig. 1A) (Colodner et al., 1993; Dalai etal., 2002; Miller etal., 2011;Rahamanetal., 2012). Dif-ferences in [Re]diss between catchmentsin Taiwan (RSD

=

53%)

are greater than the mean variability of[Re]diss ina single river

catchment (RSD

<

22%). Some ofthis variabilityis linked tothe runoff atthe time ofsampling, with[Re]diss positively correlated

with runoff (Fig. 1B; r

=

0

.

65, P

=

0

.

02). However, samples col-lectedfromtheLiwuRiverduringafloodeventhave[Re]dissvalues

whichare relativelyinvariant (8%RSD)whilerunoffvaried byup to

>

10the longtermmean(Fig. 1B).The findings areconsistent withmeasurementsfromArcticrivers,whichdidnotshow a sig-nificant relationshipbetween waterdischarge and[Re]diss (Miller

et al., 2011). This contrasts strongly with some major elements which tend to show dilution at high runoff (e.g. Tipper et al., 2006), such as[Na] which is strongly negatively correlated with runoffinthesampleset(Fig. 1C;r

= −

0

.

88,P

=

0

.

004)andduring thefloodevent(Calmelsetal.,2011).Thedilutionof[Na] is con-sistent withelements mobilizedpredominantlyby acidhydrolysis reactionsduringsilicateweathering(Calmelsetal.,2011;Maheret al.,2011).

In theabsence of amajor hydrological control on [Re]diss,we

examined other controlson [Re]diss.Incontrast toprevious work

(Dalai et al., 2002; Miller et al., 2011), there was no signifi-cant correlation between [SO4] and [Re]diss in Taiwan (Fig. 1A;

P

=

0

.

25). [Re]diss was correlated with[K] (r

=

0

.

62, P

=

0

.

008),

supporting recent findings from Indian rivers with minimal an-thropogenicdisturbance(Rahamanetal.,2012).Inthefloodevent samples where the instantaneous suspended sediment concen-tration (mg L−1) was measured (n

=

4), there was no significant correlation with [Re]diss ( P

=

0

.

07). The most significant

corre-lation with [Re]diss was with the catchment average suspended

sediment concentration measured over 30 yr (Fig. 2; r

=

0

.

84,

P

<

0

.

0001). This is a measure of decadal physical erosion rate normalizedbycatchmentrunoff(Dadsonetal., 2003). Inaddition tothistrend,catchmentsdraininghighergrademetamorphicrocks

Fig. 1. A. Dissolvedrheniumconcentration,[Re]diss(pmol L−1),versussulfateconcentration,[SO4](μmol L−1),forTaiwanrivers(blackcircles)andaglobalrivercompilation

(Dalaietal.,2002; Milleretal.,2011;Rahamanetal.,2012).Acrossallmeasurementsthereisapositivecorrelation(r=0.44;P<0.0001)butconsiderablescatter,and valuesarenotcorrelatedfortheTaiwandataset( P=0.25). B. [Re]dissversusinstantaneousrunoff(mm day−1)forTaiwanriversamples(labeledbycatchment). C. Dissolved

sodiumconcentration,[Na](μmol L−1_{),versusinstantaneousrunoffforthesamesamplesshowninpartB.Barsindicatetheprecisionontheanalysiswherelargerthanthe}

pointsize.

inthe northeastflank ofthe CentralRange (Beyssacet al.,2007; Hiltonetal.,2010),suchastheLiwu,Hualien,Hsiukuluan(Fig.S1), appear to have slightly higher [Re]diss than catchments draining

lowergrademetamorphicrocks atsimilarerosionrates(Laonung, Chihpen).However,thescatterintherelationship(Fig. 2)mayalso be controlled by hydrological variability in [Re]diss (Fig. 1B) and

longer-termrecords of[Re]diss are neededto examine second

or-dertrendsindetail.

3.2. Rheniuminriverbedmaterialsandsoils

The river bed materials in Taiwan represent the solid prod-ucts of erosion and have [Re] values which range from 0.60 to 0.96 ppb (Table S2). These are lower than measured in one of the first studies to propose Re as a tracer of OCpetro oxidation,

with

[

Re

]

=

75–116 ppb in a Devonian black shale (Jaffe et al., 2002). The difference is mirrored by the low OCpetro content of

riverbedmaterialsinTaiwan,

[

OCpetro

]

=

0

.

17–0

.

23%,comparedto

[

OCpetro

]

=

6

.

0–7

.

8% intheDevonianshale(Jaffeetal.,2002).

How-ever,[Re]inTaiwaneseriverbedsedimentsaresimilartoriverbed sedimentsmeasuredintheHimalayaof

[

Re

]

<

1

.

2 ppb (Rahaman etal.,2012)andHimalayanrocks(Pierson-Wickmannetal.,2002) where OCpetro contents of river bedloads are similar to Taiwan,

with

[

OCpetro

]

<

0

.

2% (Galyetal.,2008b).The[Re]ofriverbed

ma-terialsfromTaiwan are alsosimilar tomodern marine sediments in sub-oxicwaters (Crusius andThomson, 2000), whereOC con-tentsand[Re]arelowerthansedimentsinanoxicbasins(Ravizza etal., 1991). We note that [Re] may vary withthe grain size of riverbedmaterials inanalogy tomajor elements (Bouchez etal., 2011),however,alargersamplesetisnecessarytoestablishthese controls.

Themean ratioof[Re]to [OCpetro]in thethreeriverbed

ma-terialsamplesis 3

.

7

×

10−7 andthevaluesrangewithin a factor oftwofrom2

.

6

×

10−7 to4

.

6

×

10−7.Theseratiosarelowerthan thosemeasuredontwosetsofOC-richDevonianblackshale,with mean[Re]:[OCpetro]

=

12

.

8

±

3

.

4

×

10−7(n

=

5,

±

2 SE)(Jaffeetal.,

2002) andmean[Re]:[OCpetro]

=

12

.

2

±

5

.

0

×

10−7(n

=

8,

±

2 SE)

(Selby and Creaser, 2003). However, they are similar to OC-rich Jurassicshale(mean[OCpetro]

=

20

±

8%)withmean[Re]:[OCpetro]

=

3

.

5

±

0

.

9

×

10−7 _(n

₌

_5,

_±

_{2 SE)(Cohenetal., 1999). The}

dif-ferencesin[Re]:[OCpetro]betweenlocalities isexpectedgiventhe

different redox conditions, sedimentation rates and bioturbation

Fig. 2. Dissolved rhenium concentration ([Re]diss inpmol L−1)for Taiwan rivers

sampledin2004and2007versusthecatchmentaveragesuspendedsediment con-centrationfor1970–1999(Dadsonetal.,2003) (errorsnotprovided)ameasureof physicalerosionperunitofwaterrunoff.Verticallinesshowanalyticalprecision(if largerthanthepointsize).

rateslikely tohaveoccurredatthetime ofburial, whichwill in-fluence the [Re] and [OCpetro] (McKay et al., 2007). In addition,

the[Re]:[OCpetro]mayvarywithmetamorphicgradeandthe

ther-malmaturity ofOCpetro,althoughthisremains poorlyconstrained

(Rooneyetal., 2012; Cummingetal.,2014).Nevertheless,the re-sultsconfirm theimportanceofconstraining alocal[Re]:[OCpetro]

inupstream lithologieswhenusingRe toexamineOCpetro

weath-ering(Dalaietal.,2002).

The thin surface soils have low [OCpetro]

<

0.05% (Table S2),

consistent with oxidative loss of OCpetro during weathering. The

bulksoilshave

[

Re

]

=

0

.

07–0

.

19 ppb,lowerthantheriverbed ma-terials.SoilsarealsodepletedinCaandS whencomparedtobed materials(TableS2),consistentwithcarbonateweatheringand sul-fideoxidationofthesedimentaryrocks(Calmelsetal.,2011).The coupledlossofRe andOCpetro relativetoan immobileelementis

evident inTaiwanesesoils(Fig. 3).Theresultsareconsistentwith observationsfromsoilprofiles elsewhere,whererocks hadhigher

R.G. Hilton et al. / Earth and Planetary Science Letters 403 (2014) 27–36 31

Fig. 3. Solidphasesamples,riverbedmaterials(blacksquares)andhomogenized

surface(AandE)soilhorizons(whitecircles)fromTaiwan.Rheniumconcentration, [Re],andpetrogenicorganiccarbonconcentration,[OCpetro],havebeennormalized

toanimmobileelement(aluminum,Al2O3)totrackweatheringprocesses.Arrow

indicatescoupledlossofReandOCpetrofromsoilhorizonsfromTaiwanconfirming

workonOCpetro-richlithologieselsewhere(Jaffeet al.,2002).Verticallinesshow

propagateduncertaintyon[OCpetro](Appendix A)iflargerthanthepointsize.

[OCpetro],higher[Re]andalowermetamorphicgradethanTaiwan

(Jaffeetal.,2002). 4. Discussion

4.1.Insightsonthesourceofdissolvedrhenium

The association of Re and OCpetro in sedimentary rocks and

the solubility of Re upon oxidation has lead previous work to suggest that [Re]diss may provide a robust tracer of OCpetro

oxi-dationreactions (e.g. Dalaietal., 2002). However,it isimportant to assess the source of [Re]diss and address the validity of

as-sumptionswhichlinkdissolvedRe fluxtoOCpetro oxidation.First,

the solid residue of chemical weathering provides insight on Re mobilization to the dissolved load. Bulk soil samples from Tai-wan have lower [OCpetro] and [Re] when compared to the river

bed materials (Fig. 3), which are likely to have a composition closetounweatheredbedrock followingtheir recentmobilization by bedrock landslides (Hovius et al., 2000). This resultis consis-tent with observations of coupled Re and OCpetro loss made on

soil profiles elsewhere (Peucker-Ehrenbrink and Hannigan, 2000; Jaffeetal.,2002) andsuggestscoupledoxidativeweatheringofRe andOCpetro. Once oxidized to the perrhenate anion ReO−4, Re is

solublein oxygenated waters (Eh

>

0 V) withpH ranges 5.5–9.5 (Brookins,1986).Asaresult,Relossfromsoilsshouldberecorded inthe[Re]dissofriversanditsconcentrationinfluencedbytherate

ofweatheringoftheRe-bearingsubstrate(Colodneretal.,1993). In river waters, previous work has suggested that a correla-tionbetween[Re]diss and[SO4]mayimplyacommonsource(i.e.

sulfide minerals) during weathering (Miller et al., 2011). In Tai-wan, sulfide oxidation is an important weathering reaction and the dominant source of SO2₄− to rivers (Yoshimura et al., 2001; Calmels etal., 2011) andso the weak correlation between [SO4]

and[Re]diss inTaiwan (Fig. 1A) maysuggest sulfide minerals are

not the dominant host of Re (Selby and Creaser, 2003). While thisisconsistentwiththeavailable measurementsfromsoilsand riverbed materials in Taiwan (Fig. 3), it contrasts with previous conclusions drawn from observations linking [Re]diss to [SO4] in

Himalayan rivers (Dalai et al., 2002) and globally (Miller et al., 2011).Thediscrepancycanbereconciledbynotingthatwhilethe global trend between [Re]diss and [SO4] is significant (r

=

0

.

44;

P

<

0

.

0001), it is extremely scattered (Fig. 1A). For [Re]diss

val-ues similar to Taiwan, the global data shows [Re]diss varies by

>

50 pmol L−1 for a given [SO4]. Therefore, we propose that the

globalcorrelation doesnot indicate a common source of Re and SO2₄−,but insteadreflects changes in evaporationand/or dilution which set a first order control on all dissolved ion concentra-tions,especiallyinlargerivers(Gaillardetetal.,1999).Thiswould also explain why[Re]diss is better correlatedwith[Ca] (r

=

0

.

57;

P

<

0

.

0001) in the same compilation of data shown in Fig. 1A. We are therefore wary of using correlations between dissolved ionconcentrationstointerpretacommonsourceduringchemical weathering(cf.Milleretal.,2011).

The significant correlation between [Re]diss in Taiwan rivers

andthecatchment-averagesuspendedsedimentconcentration be-tween 1970 and 1999 (Fig. 2) suggests that the rate of physi-cal erosion over recentdecades plays a major role incontrolling [Re]diss of a catchment. This observationsheds newlight on the

weathering processes which mobilize Re from rocks to the dis-solvedphase.Physical erosioninTaiwanisdominatedbybedrock landslideswhichscourdeep(

>

1 m)intotherockmass(Hoviuset al., 2000). Anincrease in the decadalaverage erosionrate there-fore reflects an increase in the frequency at which unweathered bedrock isexposedatthesurface by landslides.Ifweathering re-actions are mainly limited by the supply of new minerals, then increasedphysicalerosionrateshouldresultinanincrease inthe concentrationofdissolvedproductsinriverwaters.Previouswork hashypothesizedthatOCpetro oxidationinsoilsunderpresent

at-mospheric O2 levels are limited by the supply of OCpetro to an

oxidative weathering zone near the surface (Petsch et al., 2000; Bolton etal., 2006) and maybe enhanced by microbial assimila-tion insurface soils(Petsch et al.,2001). Therefore,theobserved increase in[Re]diss witherosionrate(Fig. 2) supportthis

hypoth-esis andsuggest that supply limits OCpetro oxidative weathering,

evenattheveryhighratesofphysicalerosioninTaiwan.

The decoupling between[Re]diss andmajor elements with

in-creasing runoff, such as [Na] (Figs. 1B andC), is also consistent with this explanation and suggests Re is mobilized from sur-face soils. Having taken rainwater inputs into account, dissolved Na (and Mg, Ca) is predominantly sourced from the inorganic rock matrix by acid hydrolysis reactions (Gaillardet et al., 1999; Tipperet al., 2006; Calmelsetal., 2011). Laboratory experiments on freshminerals show that acidhydrolysis reactions occur

<

10 times slower than oxidation weathering reactions (Chang and Berner, 1999; White and Brantley, 2003). Therefore longer flow paths whichlengthen thetimescalesoffluid–rock interactionare requiredtoelevate[Na](Maheretal.,2011)andso[Na]isstrongly diluted withincreasing runoff (Fig. 1C). As a result, groundwater inputs which are most significant at low runoff can make up a large proportion(

∼

70%)ofthe Naflux inTaiwan (Calmels etal., 2011). Incontrast, Re is not diluted withrunoff (Fig. 1B)andso groundwater inputsare likely to be relatively minor. Overall, the behaviorof[Re]diss(Figs. 1Band2)ismoreconsistentwithitbeing

liberatedfromanoxygenatedsurfaceweatheringzonebyoxidation reactionswhichoccurmuchquickerthanacidhydrolysis.

Theobservationthat oxidativeweatheringofOCpetrois

supply-limitedmayseematoddswithmeasurementsofsignificantexport ofunweatheredOCpetro inthesuspendedloadofTaiwaneserivers

(Hiltonetal.,2010, 2011).However,thisreflectsthedominantrole ofbedrocklandslidesintheerosionofmountainbeltslikeTaiwan (Hovius et al., 2000). Landslides erode the weathering products fromthinsoils,whichcanbefullydepletedinOCpetro (Fig. 3)and

deliveredthemtoriverchannels.However,thevolumeofbedrock landslidesincreasesasapowerlaw(exponent

>

1

.

2)ofthe land-slidearea(Larsenetal.,2010) andsolarge,deep-seatedlandslides can deliver significant amounts of unweathered OCpetro to

Table 1

Rhenium-derivedOCpetrooxidationrateestimatesforrivercatchmentsinTaiwan.

Catchment Drainage area (km2₎ Suspended sedimentyield (t km−2_yr−1₎a Water discharge (g yr−1₎a Decadal averageSSC (g L−1₎a nb _[Re] diss (pmol L−1₎ RSD c (%) Reflux

(g yr−1₎ OC_ratepetrooxidation

(tC km−2_yr−1₎d Wulu 639 17 371 1.41E+15 7.9 3 14 10 3629 12−22 Yenping 476 19 118 1.00E+15 9.1 1 13 − 2417 11−19 Hsiukuluan 249 12 851 6.23E+14 5.1 1 10 − 1192 10−18 Hualien 1506 20 850 3.16E+15 9.9 2 20 21 11 634 17−30 Liwu 435 33 103 1.04E+15 13.8 6 21 22 4010 20−35 Chenyoulan 367 8719 6.97E+14 4.6 3 9 2 1130 7−12 Upper Chenyoulan 205 2927 4.10E+14 1.5 1 5 – 358 4−7 Laonung 853 10 785 1.62E+15 5.7 1 5 – 1505 4−7 Taimarli 190 2105 4.37E+14 0.9 1 6 – 458 5−9 Chihpen 166 21 687 3.98E+14 9.0 1 8 – 571 7−13 Taiwan average 35 980 10 672 4.98E+16 – 10 13 53 122 308 7−13

a _{Averagesfrom1970to1999 (Dadsonetal.,2003),withsuspendedsedimentconcentration(SSC)calculatedfromdecadalmeansedimenttransferandwaterdischarge.} b _Numberof[Re]

dissanalysespercatchment. c _{RelativeStandardDeviation(RSD)on[Re]}

dissmeasurements. d _{Rangeofestimatesderivedfromtherangeofmeasured[Re]:[OC}

petro]ratiosinriverbedmaterials(TableS2).

proportionoftheexhumedOCpetrodoesnotspendsignificanttime

inthe weathering zone before beingexported by rivers. Bedrock landslidescanthereforeexplainboththehighratesofOCpetro

ox-idationbyincreasingmineralsupplyatlandslidesites(Fig. 2)and alsotherapidfluvialexportandre-burialofun-weatheredOCpetro

offshoreTaiwan(Kaoetal.,2014).

4.2. DissolvedrheniumfluxandestimatesofOCpetrooxidation

Theobservedlinkbetween[Re]diss andsedimentyield(Fig. 2),

along with the coupledloss of Re and OCpetro fromsoils in

Tai-wan (Fig. 3), suggest that [Re]diss provides a powerful proxy of

OCpetro oxidation inTaiwan. Here we describe howthe dissolved

Re flux can provide insight onthe CO2 release by OCpetro

oxida-tion,andconsidertheassumptionswhichneedtobeapplied.The firststep isto quantifythe dissolved Re flux,accountingfor sys-tematicchangesinconcentrationwithrunoff(Calmelsetal.,2011). [Re]diss wasnotdiluted withincreasingrunoff (Fig. 1B)andsoRe

fluxcanbequantifiedrobustlyusinganaverage[Re]diss.Refluxes

forindividualcatchmentswerecalculated,basedonbetween1and 6samples per catchment (Table 1). While thisis relatively small number of samples andcatchment-scale Re flux maybe refined by longer records, the numberis similar to work quantifying Re fluxes in large rivers (Miller et al., 2011). At the mountain belt scale,thearea-weightedmean[Re]diss inthesampledcatchments

is13

.

2

±

3

.

7 pmol L−1 (n

=

10,

±

2 SE,Table 1)andforan annual water discharge of 49.8 km3yr−1 (Dadson et al., 2003) the dis-solved Re flux from Taiwan to the ocean is 700

±

200 mol yr−1_.

TheReyieldestimatesfromindividualrivercatchmentsinTaiwan were correlated to the sediment yield measured independently from1970to1999 (Fig. 4A).

Havingquantified the oxidizedRe flux from rivercatchments, theOCpetro oxidationratecanbe estimatedprovided thatthe

ini-tial[Re]:[OCpetro]ofrocksisknown.Inaddition,twoassumptions

arenecessary:i)thatReishostedprimarilyinOCpetro;andii)that

oxidation mobilizes [Re]diss and releases CO2 fromOCpetro atthe

same rate. The first assumption would appear to be valid based on the high Re contents of organic matter isolated from sedi-mentary rocks, with OCpetro shown to dominate the total mass

of Re (Cohen et al., 1999; Pierson-Wickmannet al., 2002; Selby andCreaser,2003; Rooneyetal.,2012).However, someRecanbe sourcedfrominorganicphases(Dalaietal.,2002),whichmayhost

∼

30% ofRe inmarinesediments(Selby andCreaser,2003).A Re-derived OCpetro oxidation ratewillbe an overestimationifthisis

the case. Regarding the second assumption, weathering products from soil horizons in Taiwan show coupled loss of OCpetro and

Re during weathering (Fig. 3). However, soil profiles from else-where have shown that oxidation can result in slightly higher losses of Re from the solid phase (

∼

99%) compared to losses of OCpetro (

∼

80%)(Jaffeetal.,2002; Pierson-Wickmannetal.,2002).

Inaddition,rocksinTaiwancancontaingraphitic-OCpetro (Beyssac

et al., 2007) which has been shown to be resilient to chemical and physical breakdown in river catchments (Galyet al., 2008a; Bouchezetal.,2010) butitisnotknownwhethergraphitic-OCpetro

contains Re. Based on our currentunderstanding ofRe mobility, theseassumptionsmeanthatthedissolvedRefluxislikelyto rep-resentanupperboundontheOCpetro oxidationrate.Thesedetails

warrant futureresearch torefine theRe-proxy, howeverournew findingsonthemobilityofReduringweatheringinTaiwan(Figs. 2 and3)suggestthataRe-derivedestimateofOCpetro oxidationrate

offersarobust,upperboundquantification.

To calculate OCpetro oxidation rates, the dissolved Re flux

(Fig. 4A)mustbecombinedwithmeasurementsoftheinitial[Re] to[OCpetro]ratioofsedimentaryrocksinthecatchment.Individual

bedrock samplesmaysignificantly overestimatetheheterogeneity present at the catchment scale and instead, river bed materi-als can provide a robust average sample of the major geological formationsupstream(Galyetal.,2008a; Hiltonetal.,2010).In Tai-wan,[Re]:[OCpetro]ofthethreeriverbedmaterialsamplesranged

within afactoroftwo, from2

.

6

×

10−7 _{to 4}

_.

₆

_×

₁₀−7_.Whilethe

sample set of river bedmaterials is small (n

=

3), we note that thesamplesanalyzedherehavenitrogentoOCratiosbetween0.1 and 0.2 (Table S2) which covers a range of compositions repre-sentative ofOCpetro inthe major geological formations (Hilton et

al., 2010).The samplesprovidea plausiblerangeof[Re]:[OCpetro]

of the source and allow us to assess a range of CO2 emission

estimates. With the knowledge that OCpetro oxidation rates

de-rivedfromdissolvedRe fluxarelikelytobean upperbound,and thattheinitial[Re]:[OCpetro]ofrocksisthemainsourceof

uncer-tainty, we estimate amaximum CO2 release by OCpetro oxidation

ofbetween0.27and0.47 MtC yr−1 _{fromthismountainbelt.}

Nor-malizedovertheislandsourcearea(35 980km2)theCO2 yieldis

7.4–13.0 tC km−2_yr−1_{. Thisis asignificant carbontransfer andis}

higherthanestimatesofCO2 drawdownbysilicateweatheringin

mountain belts of 3–4 tC km−2yr−1 (Jacobson and Blum, 2003). The results suggest that in Taiwan the chemical denudation of OCpetro byoxidationrepresents

<

20%ofthetotalOCpetro

denuda-tion(weatheringpluserosion)with

∼

1.7 MtC yr−1exportedtothe oceaninsolidform(Hiltonetal.,2011).Incontrast,theextensive floodplainsdownstreamoftheHimalayaandAndesmountains(of theGangesandAmazonrivers,respectively)providetimefor addi-tionalOCpetrooxidation,removingallbutthemostresilientOCpetro

R.G. Hilton et al. / Earth and Planetary Science Letters 403 (2014) 27–36 33

Fig. 4. A. Dissolvedrhenium(Re)yield(g km−2_yr−1_{)versusaveragesuspendedsedimentyield1970–1999(Dadsonetal.,2003) forindividualcatchmentsinTaiwan(symbols}

asinFig. 2).VerticalbarsdenotethemeanstandarddeviationindissolvedReconcentrationinthesampledcatchments(Table 1). B. Petrogenicorganiccarbon(OCpetro)

oxidationrates(tC km−2_yr−1_{)calculatedusingthedissolvedReyieldforTaiwancatchments(gray rectangles)versussuspendedsedimentyield.Therangeofestimates}

representuncertaintybasedonthemeasuredrangeofRetoOCpetroratiosinriverbedmaterials.TheblackstarshowsOCpetrooxidationrateestimatedpreviouslyusing

dissolvedRefluxfromtheYamunaRivercatchment,Himalaya(Dalaietal.,2002).

(graphite)fromthe solid load(Galy etal., 2008a;Bouchez et al., 2010,2014).InTaiwan,thecombinationofbedrocklandslides(see Section4.1),rapidfluvialtransittimesandstrongland–ocean cou-plingmeanthatdespitetherapidOCpetro oxidationrates(Fig. 4B)

the majority ofexhumed OCpetro (

>

80%) is eroded and reburied

offshoreTaiwan(Kaoetal.,2014).

TheRe-derivedOCpetrooxidationratesforindividualcatchments

(Table1)arepositivelyrelatedtothedecadalsuspendedsediment yield (Fig. 4B). The catchment-scale OCpetro oxidation rates may

be refinedusing longer time-seriesof [Re]diss measurements and

catchment-specific[Re]:[OCpetro]ratiosofriverbedmaterials.

Nev-ertheless, therelationship isconsistent witha Re-derivedOCpetro

oxidationratemadeusingthesamemethodsintheYamunaRiver drainingtheHimalaya(9600km2),whereOCpetrooxidationis

esti-matedtorelease

∼

4 tC km−2yr−1 ofCO2 (Dalaietal.,2002) with

denudation

∼

1700 t km−2_yr−1 _{(Lupkeretal.,2012) (Fig. 4B). Our}

new data confirm the hypothesis that physical erosion rate is a majorcontrolonOCpetro oxidationandmoderatesCO2release

dur-ingweathering(Petsch etal.,2000),evenattheveryhigherosion ratesexperiencedinTaiwan.

4.3.ThenetOCbudgetofamountainbelt

The Re-derived OCpetro oxidation rate can be used to assess

thenet OCbudget ofamountain belt forthefirst time. Physical erosionin Taiwan results in the mobilization of OCbiosphere from

soilsandvegetationanditstransportedtotheoceansinriver sus-pended load(Hilton etal., 2008a). The erosion ofthe OCbiosphere

resultsintheexportof

∼

0.5 MtC yr−1(1012gC yr−1)ofOCbiosphere

fromTaiwan,avalue whichis thoughttobe conservative(Hilton et al., 2012). Recent observations show that OCbiosphere is

effi-cientlypreserved(70–100%)insedimentsoffshoreTaiwan(Kaoet al.,2014),inanalogywithtectonicmarginselsewhere(Galyetal., 2007; Blair andAller, 2012), with an estimate of CO2

sequestra-tionof0.5–0.6MtC yr−1byterrestrialOCbiosphereburial(Kaoetal.,

2014). Additional OC burial occursasmarine organic matterand hasnotbeenquantified.Therefore,giventhattheestimateofCO2

releaseprovidedhereislikelytobeanupperbound(Section4.2), andthattheestimatesofCO2drawdownarelikelytobe

conserva-tive(Kaoetal.,2014),theassessmentofnetOCsequestrationwill alsobeconservative.Takingtheestimatedboundsintoaccount,we

findthat cyclingofOCduring theerosionandweatheringof Tai-wanpresentlyactsasanetsinkofatmosphericCO2 (Fig. 5).

The findings can be compared to the Ganges–Brahmaputra river system draining the Himalaya, where source to sink sedi-mentandOCtransfershavebeenpartlyquantified(France-Lanord and Derry, 1997; Galy et al., 2007, 2008a; Gaillardet and Galy, 2008). There terrestrial OCbiosphere is buried efficiently in rapidly

accumulating, O2 depleted waters of the Bay of Bengal and

se-questers 3.7 MtC yr−1 derived froma continental source area of

∼

1

.

6

×

106 km2 (Galyetal., 2007). Incontrast,

∼

0.8 MtC yr−1 of OCpetro isthoughttobe oxidizedduring floodplaintransportonce

OCpetro hasleftthemountainbelt (Galyetal., 2008a). OCpetro

ox-idation rates have not been quantified in Himalayan catchments where CO2 emissions may be higher due to faster physical

ero-sion(e.g.Fig. 4B). However,denudationratesintheHimalayaare lower thanTaiwan (Dadsonetal., 2003; Lupkeretal., 2012),and theupwardcorrected OCpetro oxidationrateisunlikelytocounter

theCO2 sinkbyburialofOCbiosphere.

Together with our new data, these findings suggest that in mountain belts located in the subtropics, the CO2 sequestration

achieved by OCbiosphere export and burial is greater than the

CO2 emissions by geological respiration during oxidative

weath-ering of OCpetro (Fig. 5). The frontal Himalaya andTaiwan

expe-rience high physical erosion rates driven by fluvial incision and bedrock landsliding, which can supply abundant OCpetro for

ox-idation (Hilton et al., 2011). However, both orogenic belts fea-tureproductiveterrestrialbiomasswhichiserodedrapidlyandis buriedwithclasticsediment(Galyetal.,2007; Hiltonetal.,2012; Kaoetal.,2014).ThebalanceoftheseOCfluxespresentlyactasa sinkofCO2(Fig. 5).However,thenetOCcycleinTaiwanappearsto

besensitivetosmallincreasesintheCO2releasebyOCpetro

weath-ering, which will vary spatially (and temporally) in response to changesintherateofphysicalerosion(Fig. 4B).Itisalsosensitive to changes in thedistribution of mountainforest and soil which might reduce the CO2 sequestration associated witherosion and

burial ofOCbiosphere (Hilton et al., 2012). These factorsshould be

centralincontrollingtheimpactofmountainbuildingonthe long-term sequestration of CO2 in the lithosphere(Derry and

France-Lanord, 1996) and the influence of orogenesis on the evolution of atmosphericCO2 andO2 concentrations(Caldeiraet al., 1993;

im-Fig. 5. NetorganiccarbonbalanceinTaiwanexpressedintermsofCO2fluxesto

(positive)andfrom(negative)theatmosphere.Petrogenicorganiccarbon(OCpetro)

oxidationratederivedfromdissolvedReflux(Fig. 4A,thisstudy)isdifferencedfrom thetransferandburialofOCfromtheterrestrialbiosphere(OCbiosphere)transferand

burial(Hiltonetal.,2012; Kaoetal.,2014) toassessthenetOCcycle,withupper (dashed)andlower(dotted)boundsfornetCO2sequestration.

pactofchangingphysicalerosionratesontheglobalcarboncycle has received renewed attention via the process ofsulfide oxida-tive coupled to carbonate weathering (Calmels et al., 2007) over interglacial–glacialcycles(Georgetal.,2013) andduringthe Ceno-zoic(Torreset al., 2014). Ourfindings suggest a moredirectlink via theoxidative weatheringofOCpetro andCO2 release (Fig. 4B)

andbetterunderstandingthe ratesandpatternsofOCpetro

oxida-tionremainsaresearchpriority. 5. Conclusions

The ratesand controlson CO2 releaseby OCpetro oxidation in

a mountain belt were assessed using the trace element Re. In solid weathering products from Taiwan, Re loss was coupled to OCpetroloss(Fig. 3),confirmingpreviousworkfromsoilprofileson

OCpetro-richrocks. Inriverwaters fromTaiwan,[Re]diss was

posi-tively,significantlycorrelatedwiththedecadalaveragesuspended sediment concentration (1970–1999), a measure ofphysical ero-sion rate normalized by runoff (Fig. 2). We estimate Re flux for thesampledcatchmentsandfindthattheRe yieldalsoincreases withsedimentyield(Fig. 4A).Anincreasein[Re]diss andRefluxin

basinswhicherodemorerapidlyisconsistent withthe dominant sourceofRefromasurficialweatheringzone,whereoxidation re-actionsarelimitedbymineralsupply.

TheRefluxmeasurementsareusedtoestimatetheCO2 release

byOCpetrooxidationwithknowledgeoftheinitial[Re]:[OCpetro]of

bedrocks,andassuming:i)that Reishosted primarilyinOCpetro;

andii)congruent release of[Re]diss andCO2 fromOCpetro occurs

duringoxidative weathering.Wenotethat thefirstassumptionis validatedbypreviouswork, butmayleadtoan overestimationof CO2 fluxifsome dissolvedRe isderived frominorganicminerals.

Thesecondassumptionappearstobevalidbasedonour measure-ments of OCpetro and Re loss in weathering products in Taiwan

(Fig. 3)andissupportedbyprevious observationsofcoupledloss ofRe andOCpetro in soil weatheringprofiles. However, basedon

ourcurrentunderstandingofRemobility,itmayleadto an over-estimationofOCpetro oxidationrateifReisliberatedmorerapidly

during weathering.The patternsofCO2 releaseacrossTaiwanare

controlledby physicalerosionrate(Fig. 4B),confirming mountain beltsashotspotsofCO2 releaseby geologicalrespiration.The

ab-solute rates of CO2 emission in river catchmentsestimated from

dissolvedRefluxarelikelytobeanupperbound.Nevertheless,at themountainbeltscale,theOCpetro oxidationrateisnotsufficient

tonegatetheestimatedCO2drawdownbyerosionofOCfromthe

terrestrial biosphere, its fluvial transport and marine burial off-shore. MountainbuildinginTaiwan presently actsasan OCsink, sequesteringatmosphericCO2 duringweatheringanderosion.

Acknowledgements

Data associated with this research can be found in the sup-plementary materials. The research was funded by a Natural Environment ResearchCouncil(NERC), UK,NewInvestigatorGrant to R.G. Hilton (NE/I001719/1) and by a CNRS EC2CO grant (‘OXYMORE’) to J. Gaillardet. R.G. Hilton acknowledges Research Leave fromDurham University andfunding fromIPG, Paris, as a Visiting Professor. We thank F. Campas and J. Moureau for lab-oratory and analytical assistance and A. Galy, A.J. West, E. Tip-per, D. Selby and J. Bouchez for discussions, and L. Derry and threeanonymousrefereesfortheircommentswhichimprovedthis work.

Appendix A. QuantificationofOCpetrocontentinsoilsamples The OCpetro content (%) of soil samples was determined

us-ing Fmod measurements (Hiltonetal., 2013) and abinary mixing

modelwhichassumescontemporaryinputs(OCbiosphere)mixwith

radiocarbon depletedOCpetro (e.g.Galyetal., 2008a; Hiltonetal.,

2008a; Clarketal.,2013):

(

Fmod

)

petro

×

fpetro

+ (

Fmod

)

biosphere

×

fbiosphere

= (

Fmod

)

sample

(A.1) where f denotes the fraction of OC derived from petrogenic and biosphere sources ( fpetro and fbiosphere, respectively) where

fpetro

+

fbiosphere

=

1 forabinarymixture.Theradiocarbonactivity

ofpetrogenicOC

(

Fmod

)

petro

=

0 bydefinition(radiocarboncontent

indistinguishablefrombackground).TheC-weightedmeanFmodof

uppersoilhorizonsinTaiwanhasbeenmeasured byHiltonetal. (2008a, 2008b) as Fmod

=

0

.

98

±

0

.

07 (n

=

10,

±

SD) and this is

taken as the radiocarbon activity of the OCbiosphere end member

composition

(F

mod

)

biosphere.

Eq.(A.1)canbesolvedfor fpetro andthemeasured fpetro

com-bined withmeasured total organiccarbonconcentration (Corg,%)

to quantity the [OCpetro], in% (Kaoet al., 2014). The uncertainty

on[OCpetro]contentofsoilsisderivedfromthepropagationofthe

variabilityonthe endmembercompositions(Fig. 3).Theanalysis providesanupperestimateofOCpetrocontentbecauseagingofsoil

organic matter can also deplete 14C content, rather than OCpetro

addition (Hilton etal., 2013). Previous work hasestablished that OCintheriverbedmaterialsamplesisdominatedbyOCpetro

us-ingelementratiosandstablecarbonisotopes(Hiltonetal.,2010). Appendix B. Supplementarymaterial

Supplementarymaterialrelatedtothisarticlecanbefound on-lineathttp://dx.doi.org/10.1016/j.epsl.2014.06.021.

References

Anbar, A.D., Creasar, R.A., Papanastassiou,D.A., Wasserburg, G.J., 1992. Rhenium inseawater:confirmationofgenerallyconservativebehaviour.Geochim. Cos-mochim.Acta 56,4099–4103.

Berner, R.A., Canfield, D.E., 1989. A new model for atmospheric oxygen over Phanerozoictime.Am.J.Sci. 289,333–361.

Beyssac,O.,etal.,2007.LateCenozoicmetamorphicevolutionandexhumationof Taiwan.Tectonics 26,TC6001.http://dx.doi.org/10.1029/2006TC002064.

R.G. Hilton et al. / Earth and Planetary Science Letters 403 (2014) 27–36 35

Birck,J.L.,Barman,M.R.,Capmas,F.,1997.Re–Osisotopicmeasurementsatthe fem-tomolelevelinnaturalsamples.J.Geostand.Geoanal. 20,19–27.

Blair,N.E.,Aller,R.C.,2012.Thefateofterrestrialorganiccarboninthemarine en-vironment.Annu.Rev.Mar.Sci. 4,17.1–17.23.

Bolton,E.W.,Berner,R.A.,Petsch,S.T.,2006.Theweatheringofsedimentaryorganic matterasacontrolonatmosphericO2:II.Theoreticalmodeling.Am.J.Sci. 306, 575–615.

Bouchez,J., et al., 2010.Oxidation ofpetrogenicorganic carbon intheAmazon floodplainasasourceofatmosphericCO2.Geology 38,255–258.

Bouchez,J.,Gaillardet,J.,France-Lanord,C.,Maurice,L.,Dutra-Maia,P.,2011. Grain-sizecontrolsofriversuspendedsedimentgeochemistry:insightsfromAmazon Riverdepthprofiles.Geochem.Geophys.Geosyst. 12,GC00380.

Bouchez,J.,etal.,2014.Source,transportandfluxesofAmazonRiverparticulate or-ganiccarbon:insightsfromriversedimentdepth-profiles.Geochim.Cosmochim. Acta 133,280–298.

Brookins,D.G.,1986.Rheniumasananalogforfissiogenictechnetium:Eh–pH dia-gram(25◦C,1bar)constraints.Appl.Geochem. 1,513–517.

Caldeira,K.,Arthur,M.A.,Berner,R.A.,Lasaga,A.C.,1993.Coolinginthelate Ceno-zoic.Nature 361,123–124.

Calmels,D.,Gaillardet,J.,Brenot,A.,France-Lanord,C.,2007.Sustainedsulfide ox-idationbyphysicalerosionprocesses intheMackenzieRiverbasin:climatic perspectives.Geology 35,1003–1006.

Calmels,D.,etal.,2011.Contributionofdeepgroundwatertotheweatheringbudget inarapidlyerodingmountainbelt,Taiwan.EarthPlanet.Sci.Lett. 303,48–59. Chang,S., Berner,R.A.,1999.Coalweatheringandthegeochemicalcarboncycle.

Geochim.Cosmochim.Acta 63,3301–3310.

Clark,K.C.,etal.,2013.Newviewson“old”carbonintheAmazonriver:insight fromthesourceoforganiccarbonerodedfromthePeruvianAndes.Geochem. Geophys.Geosyst. 14,1644–1659.http://dx.doi.org/10.1002/ggge.20122.

Cohen,A.S.,Coe, A.L.,Bartlett, J.M.,Hawkesworth,C.J., 1999.PreciseRe–Os ages oforganic-richmudrocksandtheOsisotopecompositionofJurassicseawater. EarthPlanet.Sci.Lett. 167,159–173.

Colodner,D., et al.,1993. Thegeochemicalcycles ofrhenium:areconnaissance. EarthPlanet.Sci.Lett. 117,205–221.

Crusius,J.,Thomson,J.,2000.ComparativebehaviorofauthigenicRe,U,andMo dur-ingreoxidationandsubsequentlong-termburialinmarinesediments.Geochim. Cosmochim.Acta 64,2233–2242.

Crusius,J.,Calvert, S.E.,Pedersen,T.F.,Sage,D., 1996.Rheniumand molybdenum enrichmentsinsedimentsasindicatorsofoxic,suboxicandanoxicconditions ofdeposition.EarthPlanet.Sci.Lett. 145,65–79.

Cumming,V.M.,Selby,D.,Lillis,P.G.,Lewan,M.D.,2014.Re–Osgeochronologyand OsisotopefingerprintingofpetroleumsourcedfromaTypeIlacustrinekerogen: insightsfromthenaturalGreenRiverpetroleumsystemintheUintaBasinand hydrouspyrolysisexperiments.Geochim.Cosmochim.Acta 138,32–56. Dadson,S.J.,etal.,2003.Linksbetweenerosion,runoffvariabilityandseismicityin

theTaiwanorogen.Nature 426,648–651.

Dalai,T.K.,Singh,S.K.,Trivedi,J.R.,Krishnaswami,S., 2002.Dissolvedrheniumin theYamunaRiverSystemandtheGangaintheHimalaya:roleofblackshale weatheringonthebudgetsofRe,Os,andUinriversandCO2intheatmosphere. Geochim.Cosmochim.Acta 66,29–43.

Derry,L.A.,France-Lanord,C.,1996.Neogenegrowthofthesedimentaryorganic car-bonreservoir.Paleoceanography 11,267–275.

France-Lanord,C.,Derry,L.A.,1997.Organiccarbonburialforcingofthecarboncycle fromtheHimalayanerosion.Nature 390,65–67.

Gaillardet, J., Galy, A., 2008. Himalaya – carbon sink or source? Science 320, 1727–1728.

Gaillardet,J.B.,Dupré, B.,Louvat,P.,Allègre,C.J.,1999. Globalsilicateweathering andCO2consumptionratesdeducedfromthechemistryoflargerivers.Chem. Geol. 159,3–30.

Galy,V.,etal.,2007.EfficientorganiccarbonburialintheBengalfansustainedby theHimalayanerosionalsystem.Nature 450,407–410.

Galy,V.,Beyssac,O.,France-Lanord,C.,Eglinton,T.I.,2008a.Recyclingofgraphite duringHimalayanerosion:ageologicalstabilizationofcarboninthecrust. Sci-ence 322,943–945.

Galy,V.,France-Lanord,C.,Lartiges,B.,2008b.Loadingandfateofparticulate or-ganiccarbon fromthe Himalaya tothe Ganga–Brahmaputradelta. Geochim. Cosmochim.Acta 72,1767–1787.

Georg,R.B.,West,A.J.,Vance,D.,Newmand,K.,Halliday,A.N.,2013.Isthemarine osmiumisotoperecordaprobeforCO2releasefromsedimentaryrocks?Earth Planet.Sci.Lett. 367,28–38.

Hayes,J.M.,Waldbauer,J.R.,2006.Thecarboncycleandassociatedredoxprocesses throughtime.Philos.Trans.R.Soc.Lond.B 361,931–950.

Hayes,J.M.,Strauss,H.,Kaufman,A.J.,1999.Theabundanceof13_{Cinmarineorganic} matterandisotopicfractionationintheglobalbiogeochemicalcycleofcarbon duringthepast800Ma.Chem.Geol. 161,103–125.

Hilton,R.G.,etal.,2008a.Tropical-cyclone-drivenerosionoftheterrestrialbiosphere frommountains.Nat.Geosci. 1,759–762.

Hilton,R.G.,Galy,A.,Hovius,N.,2008b.Riverineparticulateorganic carbonfrom anactivemountainbelt:importanceoflandslides.Glob.Biogeochem.Cycles 22, GB1017.http://dx.doi.org/10.1029/2006GB002905.

Hilton,R.G.,Galy, A.,Hovius,N., Horng,M.-J.,Chen,H.,2010. Theisotopic com-positionofparticulateorganiccarboninmountainriversofTaiwan.Geochim. Cosmochim.Acta 74,3164–3181.

Hilton,R.G.,Galy,A.,Hovius,N.,Horng,M.-J.,Chen,H.,2011.Efficienttransportof fossilorganiccarbontotheoceanbysteepmountainrivers:anorogeniccarbon sequestrationmechanism.Geology 39,71–74.

Hilton,R.G.,etal.,2012.Climaticandgeomorphiccontrolsontheerosionof ter-restrialbiomassfromsubtropicalmountainforest.Glob.Biogeochem.Cycles 26, GB3014.

Hilton,R.G.,Galy,A.,West,A.J.,Hovius,N.,Roberts,G.G.,2013.Geomorphiccontrol ontheδ15_{N ofmountainforests.Biogeosciences 10,1693–1705.}

Hovius,N.,Stark,C.P.,Chu,H.-T.,Lin,J.-C.,2000.Supplyandremovalofsediment inalandslide-dominatedmountainbelt:CentralRange,Taiwan. J.Geol. 108, 73–89.

Huffman, E.H., Oswalt, R.L., Williams, L.A., 1956. Anion-exchange separation of molybdenum and technetium and of tungstenand rhenium. J. Inorg. Nucl. Chem. 3,49–53.

Jacobson,A.D.,Blum,J.D.,2003.Relationshipbetweenmechanicalerosionand at-mosphericCO2 consumptionintheNewZealandSouthernAlps. Geology 31, 865–868.

Jaffe,L.A.,Peucker-Ehrenbrink,B.,Petsch,S.T.,2002.Mobilityofrhenium,platinum groupelementsandorganiccarbonduringblackshaleweathering.EarthPlanet. Sci.Lett. 198,339–353.

Kao,S.J.,etal.,2014.Preservationofterrestrialorganiccarboninmarinesediments offshoreTaiwan:mountainbuildingandatmosphericcarbondioxide sequestra-tion.EarthSurf.Dyn. 2,127–139.

Keller,C.K.,Bacon,D.H.,1998.Soilrespirationandgeorespirationdistinguishedby transport analysesof vadose CO2, 13CO2 and 14CO2. Glob.Biogeochem. Cy-cles 12,361–372.

Larsen,I.J.,Montgomery,D.R.,Korup,O.,2010.Landslideerosioncontrolledby hill-slopematerial.Nat.Geosci. 3,247–251.http://dx.doi.org/10.1038/ngeo776.

Lupker,M., et al.,2012. 10_{Be-derivedHimalayan denudationrates and sediment} budgetsintheGangabasin.EarthPlanet.Sci.Lett. 333–334,146–156. Maher,K.,2011.Theroleoffluidresidencetimeandtopographicscalesin

deter-miningchemicalfluxesfromlandscapes.EarthPlanet.Sci.Lett. 312,48–58. McKay,J.L., Pedersen,T.F.,Mucci,A.,2007. Sedimentaryredoxconditionsin

con-tinental margin sediments (N.E. Pacific)–influenceon the accumulationof redox-sensitivetracemetals.Chem.Geol. 238,180–196.

Meybeck,M.,1987.Globalchemical weatheringofsurficialrocksestimatedfrom riverdissolvedloads.Am.J.Sci. 287,401–428.

Miller,C.A.,Peucker-Ehrenbrink,B.,Walker,B.D.,Marcantonio,F.,2011.Re-assessing thesurfacecyclingofmolybdenumandrhenium.Geochim.Cosmochim.Acta 75, 7146–7179.

Petsch,S.T.,Berner,R.A.,Eglinton,T.I.,2000.Afieldstudyofthechemicalweathering ofancientsedimentaryorganicmatter.Org.Geochem. 31,475–487.

Petsch,S.T.,Eglinton,T.I.,Edwards,K.J.,2001.14_{C-deadlivingbiomass:evidencefor} microbialassimilationofancientorganicmatterduringshaleweathering. Sci-ence 292,1127–1131.

Peucker-Ehrenbrink,B.,Hannigan,R.E.,2000.Effectsofblackshaleweatheringon themobilityofrheniumandplatinumgroupelements.Geology 28,475–478. Pierson-Wickmann,A.C.,Reisberg,L.,France-Lanord,C.,2002.BehaviorofReand

Osduringlow-temperaturealteration:resultsfromHimalayansoilsandaltered blackshales.Geochim.Cosmochim.Acta 66,1539–1548.

Rahaman, W.,Singh,S.K.,Shukla, A.D.,2012.Rhenium inIndianrivers:sources, fluxes, and contribution tooceanic budget.Geochem. Geophys.Geosyst. 13, Q08019.

Ravizza,G.E.,Turekian,K.K.,Hay,B.J.,1991.Thegeochemistryofrheniumand os-miuminrecentsedimentsfromtheBlackSea.Geochim.Cosmochim.Acta 55, 3741–3752.

Rooney,A.D.,Selby,D.,Lewan,M.D.,Lillis,P.G.,Houzay,J.P.,2012.EvaluatingRe–Os systematicsinorganic-richsedimentaryrocksinresponsetopetroleum gen-eration usinghydrous pyrolysisexperiments. Geochim. Cosmochim.Acta 77, 275–291.

Selby,D.,Creaser,R.A.,2003.Re–Os geochronologyoforganic richsediments:an evaluationoforganicmatteranalysismethods.Chem.Geol. 200,225–240. Selvaraj,K.,Chen,C.T.A.,2006.ModeratechemicalweatheringofsubtropicalTaiwan:

constraintsfromsolid-phasegeochemistryofsedimentsandsedimentaryrocks. J.Geol. 114,101–116.

Stallard,R.F.,1998.Terrestrialsedimentationandthecarboncycle:coupling weath-eringanderosiontocarbonburial.Glob.Biogeochem.Cycles 12,231–257. Stuiver,M.,Polach,H.A.,1977.Discussion:reportingof14_{Cdata.Radiocarbon 19(3),}

355–363.

Sundquist,E.T.,Visser,K.,2005.In:Schlesinger,W.(Ed.),Biogeochemistry.Elsevier, Oxford,pp. 425–472.

Tipper,E.T.,etal.,2006.Theshorttermclimaticsensitivityofcarbonateandsilicate weatheringfluxes:insightfromseasonalvariationsinriverchemistry.Geochim. Cosmochim.Acta 70,2737–2754.

Torres,M.A.,West,A.J.,Li,G.,2014.Sulphideoxidationandcarbonatedissolutionas asourceofCO2overgeologicaltimescales.Nature 507,346–349.

Tsai,C.C.,Chen,Z.S.,Duh,C.T.,Horng,F.W.,2001.Predictionofsoildepthusinga soil–landscaperegressionmodel:acasestudyonforestsoilsinsouthern Tai-wan.Proc.Natl.Sci.Counc.,Repub.China,PartB 26,34–39.

Turowski, J.M., Hovius, N., Hsieh, M.L., Lague, D., Chen, M.C., 2008. Distribu-tion of erosion across bedrock channels. Earth Surf. Process. Landf. 33, 353–363.

West,A.J.,Galy,A.,Bickle,M.,2005.Tectonicandclimaticcontrolsonsilicate weath-ering.EarthPlanet.Sci.Lett. 235,211–228.

White, A.F.,Brantley,S.L., 2003.Theeffectoftimeonthe weatheringofsilicate minerals:whydoweatheringratesdifferinthelaboratoryandthefield?Chem. Geol. 202,479–506.

Yoshimura,K.,etal.,2001.Geochemicalandstableisotopestudiesonnatural wa-terintheTarokoGorgekarstarea,Taiwan–chemicalweatheringofcarbonate rocksbydeepsourceCO2andsulfuricacid.Chem.Geol. 177,415–430.