Variation of the isotopic composition of dissolved organic carbon during the runoff cycle in the Amazon River and the floodplains

HAL Id: insu-01678204

https://hal-insu.archives-ouvertes.fr/insu-01678204

Submitted on 26 Feb 2018

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.

Variation of the isotopic composition of dissolved

organic carbon during the runoff cycle in the Amazon

River and the floodplains

Patrick Albéric, Marcela A.P. Pérez, Patricia Moreira-Turcq, Marc F.

Benedetti, Steven Bouillon, Gwenaël Abril

To cite this version:

Patrick Albéric, Marcela A.P. Pérez, Patricia Moreira-Turcq, Marc F. Benedetti, Steven Bouillon, et

al.. Variation of the isotopic composition of dissolved organic carbon during the runoff cycle in the

Amazon River and the floodplains. Comptes Rendus Géoscience, Elsevier Masson, 2018, 350 (1-2),

pp.65-75. �10.1016/j.crte.2017.11.001�. �insu-01678204�

Hydrology, Hydrogeology

Variation of the isotopic composition of dissolved organic

carbon during the runoff cycle in the Amazon River and the

floodplains

Patrick Albe´ric

^a,

*, Marcela A.P. Pe´rez

^b

, Patricia Moreira-Turcq

^c

,

Marc F. Benedetti

^d

, Steven Bouillon

^e

, Gwenae¨l Abril

^b,f,g

aUMR7327,InstitutdessciencesdelaTerred’Orle´ans,CNRS–BRGM,universite´ d’Orle´ans,1A,ruedelaFe´rollerie,45071Orle´anscedex2, France

bDepartamentodeGeoquı´mica,InstitutodeQuı´mica,UniversidadeFederalFluminense,OuteirodeSa˜oJoa˜oBatista,s/n,24020-007 Nitero´i,RJ,Brazil

cLaboratoire«Ge´osciencesetEnvironnement»deToulouse,Institutderecherchepourlede´veloppement,universite´ Paul-Sabatier,14, avenueE´douard-Belin,31400Toulouse,France

dUMR7154,InstitutdephysiqueduglobedeParis,universite´ SorbonneParisCite´,1,rueJussieu,75238Pariscedex05,France

eKatholiekeUniversiteitLeuven,DepartmentofEarth&EnvironmentalSciences,Celestijnenlaan200E,3001Leuven,Belgium

fLaboratoire«EnvironnementsetPale´oenvironnementsoce´aniquesetcontinentaux»(EPOC),CNRS,universite´ deBordeaux,1,avenuedes Faculte´s,33405Talence,France

gLaboratoired’oce´anographieetduclimat,expe´rimentationsetapprochesnume´riques(LOCEAN),centreIRDFrance-Nord,32,avenue Henri-Varagnat,93143Bondy,France

ARTICLE INFO

Articlehistory:

Received25May2016

Acceptedafterrevision9November2017 Availableonline28December2017 HandledbyFranc¸oisChabaux

Keywords:

DOC

13C/¹²C AmazonBasin Runoffcycle C4plantraft

ABSTRACT

Giventherelativescarcityofstableisotopedataondissolvedorganiccarbon(DOC)inthe Amazon Basin, we hypothesized that the variability in DOC sources may be underestimatedinsuchmajorriverbasins.Toexplorethelinksbetweenthemainstem andtributariesandthefloodplain,particulareffortsweremadeduringfivedistinctcruises at different stages of the hydrograph between October 2008 and January 2011, to documentthespatialandtemporalvariationofDOCconcentrationsandd¹³C-DOCinthe centralAmazonRiversystem(Brazil). Basedon morethan200data, thespatialand temporalvariabilityofd¹³C-DOCvalueswasfoundtobelargerthanpreviouslyreportedin the same area. Although a small range of variation was observed throughout the hydrologicalcycleintheupperreachofthestudiedsection(–29.2to–29.5%intheRio Negroand–28.7to–29.0%intheRioSolimo˜es),amuchlargerone(–28.0to–34.6%)was found in the lower reach of the river, as the proportion of open lakes increased downstreaminthefloodplains.Thelowvariabilityintheupperreachessuggestsconstant andhomogeneousDOCsourcesfromuplandsoilsandfloodedforest,whilelowerd¹³C- DOCvaluesrecordedinthelowerreachmainstemathighand fallingwaterscanbe attributedtoagreaterexportofplankton-derived¹³C-depletedDOCfromfloodedlakes.

Noteworthy are the higher d¹³C-DOC values measured in the RioMadeira and the associatedfloodedlakes(–26.5to–28.8%),whichmayreflecttheimprintfromupland headwatersandaweakerdensityoffloodedforestinthewatershed.Thehigherd¹³C-DOC valuesobservedinthelowerreach duringlowwaters arestill notfully understood.

* Correspondingauthor.

E-mailaddress:patrick.alberic@univ-orleans.fr(P.Albe´ric).

ContentslistsavailableatScienceDirect

Comptes Rendus Geoscience

w ww . sc i e nce d i re ct . co m

https://doi.org/10.1016/j.crte.2017.11.001

1631-0713/ ^C 2017Acade´miedessciences.PublishedbyElsevierMassonSAS.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://

creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

Ourunderstandingofthedrivingforcesofcarbonfluxes intheAmazonRiverneedstobeimprovedtobetterexplain CO2outgassingpathways(Abriletal.,2014;Richeyetal., 2002).Stableisotopedataarepowerfultoolstoconstrain thesourcesofrespiredcarbonfromuplandsoils,flooded forestC3plants,macrophyteC4plants,andphytoplankton, but themajority of

d

¹³Cdataweremeasured on either dissolved inorganic carbon (DIC) or particulate organic carbon (POC) (Ellis et al., 2012; Mayorga et al., 2005;

Moreira-Turcqetal.,2013;Mortillaroetal.,2011; Quay etal.,1992).Inexistingdata,

d

¹³C-DOCvaluesaregenerally slightlylowerthan

d

¹³C-POCvalues,andnaturalvariation seemed limited (Aufdenkampe et al., 2007; Ellis et al., 2012;Hedgesetal.,1994;Mayorgaetal.,2005).Limited naturalvariationsin

d

¹³C-DOCcomparedto

d

¹³C-POCmay beconsistentwithalimitedcontributionofinsituaquatic primaryproductionandtheadvanceddegradationstatus of the river-transported dissolved organic matter pool (Hedges et al., 1994; Quay et al., 1992), which is also generally more recent in origin (Marwick et al., 2015;

Mayorgaetal.,2005).Theaimofthisstudywastotestif thevariability in natural DOC sources is not underesti- matedduetotherelativescarcity ofDOCstableisotope dataforthemiddleAmazonBasinreach,withtheriskto incorrectlyidentifythecarbonfluxsourcesatthelevelof water–massexchangesinthefloodplainortounderesti- mate specific (production/transformation) mechanisms.

Thus,particulareffortsweremadetodocumentthespatial andtemporalvariationofDOC concentrationsand

d

¹³C- DOC in the central Amazon River, to explore the links betweenthemainstem,tributaries,andthefloodplain.We usedadetailedspatialandtemporalsamplingstrategyto describehowDOCfromvariouspotentialsourcesmightbe transportedintheriverinrelationwiththefloodpulse.We also take advantage of a river section located along a floodplain gradient to document potential contrasts between upland soils, C3 and C4 wetland sources, and phytoplankton.

2. Materialandmethods 2.1. Fieldcampaignandsampling

Watersampleswerecollectedona800-kmtransect alongthelowerAmazonRiverbasinfromManacapuruon the Solimo˜es River, to Santarem at the mouth of the Tapajo´sRiver,locatedinagradientofdecreasingflooded forestareaandincreasingopenlakearea(Fig.1).Thethree

characteristictypesofwaterfoundintheAmazon(Sioli, 1984)weresampledinthemainstemandtributaries:the

‘‘whitewaters’’characterizedbyhighlevelsofsuspended sediments, high nutrient concentrations, and high con- ductivityintheSolimo˜esandMadeirarivers,thetwomain tributariesoftheAmazonRiverflowingfromtheAndes;

the ‘‘black waters’’ with low conductivities in the ManacapuruandNegroriversdrainingthelowestAmazo- nian forest dominated by podzol soils and intensely coloredbyhumicorganicmatter;theorganicpoor‘‘clear waters’’withlowsuspendedsedimentconcentrations,and low conductivity and nutrient concentrations in the Tapajo´s, Urubu, and Trombetas. We also sampled five floodplainsalongthegradientfromfloodedforesttoopen lakes. Upstream, the Va´rzea of Cabaliana and Janauaca´

receive large inputs of whitewaters fromthe Solimo˜es Riverandaresurroundedbylargeareasoffloodedforest (Fig. 1). In themiddle region of the Amazon River, we sampled two floodplain lakes just downstream of the Amazon–Madeiraconfluence(Fig.1),LakesCanac¸ariand Miratuba,mainlyreceivingwhitewaterfromtheflooding of the Amazon and Madeira Rivers. Lake Miratuba, receiving Madeira’s waters from the south through a networkofsmallchannels,issurroundedbylargeflooded forests and dries up substantially (a 70% decrease in surfacearea)duringthelowwaters.Thisisnotthecasein Lake Canac¸ari, which is a very homogeneous lake, connected to smallflooded forests and which does not dryupbymorethan20%atlowwater(Fig.1).Canac¸ari mainly receives white waters from theAmazon, asthe clearwatersfromtheUrubuRiverthatdrainsthenorthern localbasinareby-passedtothenorthandrarelyenterthe lake.Finally,inthemostdownstreampartofthestudied area(Fig.1),theVa´rzeaofCuruaı´ iscomposedofmorethan 30interconnectedlakesofmostlywhitewatertemporally or permanentlyconnectedto theAmazonmainstem by small channels (Bonnetet al., 2008; Hess et al., 2003).

Moreover,theVa´rzeaofCuruaı´ isnotstronglyconnectedto floodedforests,mostofthefloodplainconsistingofopen lakes(Fig.1).

SamplingforDOCand

d

¹³C-DOCwasperformedduring fivecruisesatdifferentstagesofthehydrograph(Fig.2).

The low-water (LW) season was sampled in October 2008andOctober2009,therisingwater(RW)seasonwas sampledinJanuary–February2011,thehigh-water(HW) season in June–July 2009, and the falling water (FW) season in August–September 2010. During 2009, the lowestwaterstagewasdelayed,thustheOctobercruise sampledthesecondpartofthefallingperiod.Forthemore downstreamstations,theendofthisperiodshowswater Floating meadows principallyconsisting of C4 macrophytes were found to increase d¹³C-DOCvaluesby1.5%intheirvicinity,butthisimpactwasnolongernoticeableat distancesof10mfromtheplantrafts.Thisrathermodest¹³C-enrichmentsuggestsrapid decompositionand/ordilutionofthiswetland-derivedDOC.

C 2017Acade´miedessciences.PublishedbyElsevierMassonSAS.Thisisanopenaccess articleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/

4.0/).

levelsequivalenttothe2008LWperiod(Fig.2).Concern- ingthedatausedinthisstudy(DOCconcentrationsand

d

¹³C-DOC values), the values measured in October 2009wereingoodagreementwiththosefoundinOctober 2008(seeFig.3),andaredifferentfromthosefoundduring theFWperiodinAugust–September2010.Wesampled the major tributaries (Solimo˜es, Negro, Madeira, and Tapajo´s),andsixstationsintheAmazonmainstemduring thefourseasons(Fig.2),whereastheManacapuru,Urubu and Trombetas Rivers were sampled onlyoccasionally (Table1).Ineachfloodplainlake,asafunctionofthewater level,wesampled3to6stations.Densephytoplankton bloomsareknowntooccurinsomeofthesefloodplain

lakes(Moreira-Turcqetal.,2013);wethereforeinvesti- gatedwhetherdiurnal variationsinDOCand

d

¹³C-DOC occur by samplingduring 24hat a frequency of 1 to 4hours.24-hsamplingwasperformedinLakeJanauaca duringHWandFW,andinLakeCanac¸ariduringFW.In addition, we performed sampling along three vertical profilesinthefloodplainlakesduringHWconditionsto documentpotentialeffectsofwatercolumnstratification.

Finally,duringtheRWcruise,asetofspecificsampleswas takenforDOC,DIC,

d

¹³C-DOCand

d

¹³C-DICanalysis.The samplingtookplaceinthesouthernpartofLakeCanac¸ari (Fig. 1)from the openwater areainto a decaying and sulphide-richplant raftknownas ‘‘capim’’.During this period,thewaterdepthundertheplantraftwasabout 5m,asintheopenlakearea.Thesefloatingmeadowsare mainlycomposedof C4 aquaticmacrophyteswith

d

¹³C valuesaround–13%,buttoalesserextantofC3plantsthat have

d

¹³Cvaluesrangingfrom 28to 35%(Moreira- Turcqetal.,2013;Mortillaroetal.,2011).

Allwatersamplesweretakenfroma25-mvesselboard withamodifiedNiskinbottlesamplerormanuallyfroma smallerboat(Abriletal.,2014).

Samplesfor

d

¹³C-DOCwerefilteredinthefieldthrough precombusted GF/F glass fiber filters (0.7

m

m), either undervacuumorbyuseofsyringesandacidifiedtopH1–2 withSuprapur¹phosphoricacid.

Samples for

d

¹³C-DIC were collected using 120-mL glassesserum bottles sealedwitha rubberstopper and poisoned with0.3mLof HgCl2 at 20g
L ¹ to avoid any microbialrespirationduringstorage.Vialswerecarefully sealed,takingcarethatnoairremainedincontactwiththe samples, and stored in the dark to prevent photo- oxidation. In thecase of thecapim transect, additional Fig.1.StudyareashowingtheAmazonBasinmainstem,thetributariesandfloodplainlakes,thesamplingstations,andthelocalitiescitedinthetext.

Fig.2.DistributionofthefiveCARBAMAcruisesduringwhichdatawere collected,alongwiththeAmazonhydrographatO´ bidosduringyears 2008–2011(startandendplottedforeachcruise).

samplesofsurfacewaterweresampledmanuallywitha funnel,thenfilteredtrough a GF/Fmini-filter tippedon syringes, poisoned with one drop of a saturated HgCl2

solutionand stored in headspace-free 7-mL Exetainer¹ vials.

2.2. DOCand

d

¹³C-DOCanalysis

DOCand

d

¹³C-DOCvalues(2009to2011)wereobtained by a wet oxidation method (Albe´ric, 2011) using a commercially available liquid chromatography (LC) Fig.3. DOC(a)andd¹³C-DOC(b)valuesforCARMAMAcruises(October2008toJanuary2011)fromupstreamtodownstreamAmazonmainstemand tributarystationsaccordingtothehydrographicalregime.Analyticalstandarddeviationerrorbarsfortriplicatesaresmallerthanplotmarksize.Forthe low-waterperiods(2008and2009),themeanandSDvalueswereplotted,exceptfortheTapajo´sandtheAmazonupstreamoftheconfluencewiththe Madeira(only2009data).TheTapajo´shadaverylowd¹³C-DOCvalue( 34.6%)duringhighwaters,whichfallsoutsidetheY-axisrange.

interface (LC-Isolink, ThermoScientific) coupled with a continuousflow isotoperatiomassspectrometer (IRMS, Delta V, ThermoScientific).Briefly, for the1–10mg
L ¹ DOC range, 100

m

L of He purged water sample were injectedin-lineviatheLC-interface.MixingoftheMilliQ watermobilephasewithreactants(H3PO41.5M;Na2S2O8

0.4M)isperformedviatheLC-IsolinkinterfaceandDOCis oxidizedintoCO₂around1008Candthentransferredby the carriergas(He) tothe ion sourceof theIRMS.The analysisintriplicatestook10min;therepeatabilityand accuracywereestimatedtobewithintherangeof0.1and 0.3%,respectively(Albe´ric,2011).

ThesamplestakeninOctober2008wereanalyzedusing acustomizedwetoxidationanalyzer(ThermoHiperTOC) coupled witha ThermoDelta+XL IRMS(Bouillon etal., 2006).

2.3. DICand

d

¹³C-DICanalysis

WecalculatedDICfrompCO2,totalalkalinity(TA),and temperature measurements using the carbonic acid dissociation constants of Millero (1979) and the CO2

solubility from Weiss (1974), as implemented in the CO2SYSprogram.pCO₂wasmeasuredcontinuouslywith anequilibratorconnectedtoaninfraredgasanalyzer(Abril et al., 2014). TA was analyzed by automated electro- titration on 50-mL filtered samples with 0.1NHCl as titrant.TheequivalencepointwasdeterminedwithaGran methodfrompHbetween4and3.Theprecisionbasedon replicate analyses was better than 5

m

M (Abril et al., 2015).

The

d

¹³C-DICwasmeasuredfollowingtheprocedureof Gillikin and Bouillon (2007). The measurements were performed using an Isotope Ratio Mass Spectrometer (Micromass Isoprime), equipped with a manual gas injection port. To correct for the partitioning of CO2

between the headspace and the water phase in the samples,and tocalculate the

d

¹³Cof thetotalDIC, the isotopicfractionationofCO2atthewater–airinterfaceasa functionofthelabtemperatureofMiyajimaetal.(1995) wasapplied.

Inthecaseof thecapim transectadditionalsamples, DIC and

d

¹³C-DIC values were obtained by the same methodasabove,forDOCand

d

¹³C-DOC,butbymanual injection,pushing10

m

LoftheExetenaircontentthrough theinjectorandwithoutheatingtheLC-Isolinkinterface reactor(Brandes,2009).

d

¹³C-DICdataobtainedwiththis secondmethodwerediscardedbecauseofobviousisotope exchangesduringsamplinginsomeofthesmallvolume vials.Hence,onlythe

d

¹³C-DICvaluesacquiredbythefirst method(120mLsamples)wereretainedinthisstudy.

3. Results

3.1. Amazonmainstemandtributaries

DOCconcentrationsintheAmazonmainstemaveraged 3.90.9mg
L ¹(Table1).Higherconcentrationswerefound in the Rio Negro and Rio Manacapuru black waters (6.70.9 and 5.40.7mg
L ¹, respectively). The lowest concentrationswerefoundinsamplesfromRioSolimo˜esand RioMadeira(3.00.6and2.50.9mg
L ¹,respectively)and theclearwatersoftheTapajo´s(3.01.0mg
L ¹). Concen- trationsinthemainstemwerethehighestinthelowerreach andduringFW(Table1;Fig.3a).

Theaverage

d

¹³C-DOCvalues ( 29.20.6%, Table1) didnot varysignificantly inthemainstemandtributaries betweenManausandO´bidos-Santare´mirrespectivelyofthe typeofwater,exceptfortheRioMadeirathatshowedhigher

d

¹³C-DOC values ( 27.80.7%) (Table 1 andFig. 3b). A different pattern is noticed when looking at seasonal Table1

Mainstemandtributaries2008–2011.DOCaverageconcentrations(mg
L ¹)andd¹³C-DOCmeanvalues(%versusVPDB).

Site Watercolor

Cond./pH/TSS(mS
cm ¹/–/mg
L ¹)

Season DOC(mg
L ¹) d¹³C(%) n

Amazonmainstem

EncontrotoObidos White/57/6.7/39 LW 3.50.2 –28.90.2 11

White/70/6.8/146 RW 3.40.2 –29.10.3 6

White/52/6.3/22 HW 4.40.2 –29.70.4 6

White/53/6.8/36 FW 4.81.3 –29.30.9 6

All 3.90.9 –29.20.6 29

Encontro White/63/6.7/20–150 All 3.70.5 –29.00.5 5

UpstreamMadeira White/66/6.6/20–140 All 3.90.4 –29.00.4 4

UpstreamItacoatiara White/58/6.7/28–133 All 3.50.5 –29.00.3 5

DownstreamItacoatiara White/56/6.7/20–136 All 4.00.9 –29.50.6 5

Parintins White/54/6.7/25–131 All 4.11.2 –29.50.4 5

Obidos White/53/6.7/25–137 All 4.11.2 –29.20.8 5

Tributaries

Solimo˜es White/75/6.8/25–184 All 3.00.6 –28.90.1 5

Madeira White/65/6.8/24–353 All 2.50.9 –27.80.7 5

Negro Black/10/4.9/8 All 6.70.9 –29.30.1 5

Manacapuru Black/12/5.3/6 All 5.40.7 –29.80.4 4

Urubu Clear/15/5.9/8 All 4.10.9 –29.60.6 4

Trombetas Clear/14/5.3/4 HW 3.8 –30.1 1

Tapajo´s Clear/18/6.6/4 All 3.01.0 –30.52.5 4

LW:lowwater;RW:risingwater;HW:highwater;FW:fallingwater;Cond.:conductivityat258C;TSS:totalsuspendedsolid,seasonalrangesinsteadof averagevaluesareindicatedforwhitewaters.

variability:upstreamoftheconfluenceoftheRioSolimo˜es andtheRioNegro(EncontrodasAguas),the

d

¹³C-DOCvalues ofbothriverswerealsorelativelyconstantthroughoutthe hydrologicalcycle(Fig.3b).Themean

d

¹³C-DOCvalueswere –28.9and–29.30.1%forthe RioSolimo˜es andthe Rio Negrowaters,respectively(Table1).Incontrast,inthemost downstreampartofthestudiedriversection,alargeseasonal variabilitywasobserved,witharangeofupto2%atO´bidos (Table1andFig.3b).IntheAmazonmainstem,anopposite trend occurred upstream and downstream of Itacoatiara (Fig. 3b).Upstreamof Itacoatiara,

d

¹³C-DOCvalues higher thanthemeanvalueforeachstationweremeasuredduring FW;theyweresimilartothosemeasuredintheRioMadeira.

Incontrast,downstreamofItacoatiara,duringthesameFW period,lower

d

¹³C-DOCvalueswerefoundaswellasduring HW.Theselow

d

¹³C-DOCvaluescoincidedduringFWwith higherDOCconcentrations(Fig.3a).Inaddition,theMadeira waterswerecharacterizedby

d

¹³C-DOCvalueshigherthan anyvaluesreportedforotherstationsateveryperiod.The largestvariationwasobservedintheRioTapajo´sclearwaters between–34.6%duringHWand–28.0%duringLW.

3.2. Floodplainlakes

Theresultsforthefloodplainlakesaresummarizedin Fig.4andTable2.Withineachlake,3to8differentsites weresampled,leadingtoalargerspatialvariabilityinDOC and

d

¹³C-DOCwhencomparedtoriverinesites.Incontrast, diurnalandverticalprofilevariabilitieswerefoundtobe small,with

d

¹³C-DOCrangingfrom0.2to0.7%(Table2).

DOC concentrations were generally higher than in the mainchannels.Whilethehighestconcentrationwasfound during RW in the upper reach lake (Cabaliana), small seasonal variations are reported in the other lakes surrounded by large flooded forests (Janauaca and Miratuba).Incontrast,higherconcentrationsweremea- suredduringFWin thelakestheleastconnectedtothe flooded forest (Canac¸ari and Curuaı´). Concerning

d

¹³C- DOC,threeprincipalobservationscanbemade:(1)despite thewiderrangein

d

¹³C-DOC,theaveragevaluefoundin floodplainlakes(withtheexceptionofLagoMiratuba)was similartothemeanvaluemeasuredformainstems,i.e.– 29.20.3%; (2) during all seasons, the

d

¹³C-DOC values werehigherinLagoMiratubacomparedtothevaluesforthe otherlakes.Withanaveragevalueof 27.80.4%,

d

¹³C- DOCvaluesinLagoMiratubaresembletheonemeasuredin theRioMadeira.(3)Maximalseasonalvariabilitywasfound at the more downstream locations, i.e.for the Va´rzea of Curuaı´, where the highest

d

¹³C-DOC values were found duringLWandthelowestvaluesduringHW,whichissimilar towhatwasobservedintheTapajo´sandtoalesserextantin themainchannelatObidos(Figs.3band4b).

3.3. CapimtransectLagoCanac¸ari

The concentrations of DOC approximately doubled whengoingfromopenwatersintotheplantraft,while

d

¹³C-DOCvaluesincreasedfrom–30%to–28.5%(Fig.5a).

Thetrendlineinthe

d

¹³C-DOCversus1/DOCplot(insert Fig. 5a) points to a value around 27% for theadded carbon,clearlynotcorrespondingtoaC4plantsource.

Along the same transect, the concentrations of DIC increasedbyafactorof10(from2to20mg
L ¹),whilethe

d

¹³C-DICvaluesincreasedfrom–13.3%(pH6.5)to–10.7%

(pH6.3)(Fig.5b).

d

¹³C-DICvaluesmeasuredinthemainstem during this period (detailed resultsnot shown) typically ranged around –13.80.1% (pH6.9), whichisconsistent withpreviousdatafromthisregion(Quayetal.,1992).

4. Discussion

4.1. Comparisonwithprevious

d

¹³C-DOCand

d

¹³C-POCdata intheAmazonRiverbasin

Since theworkof Caiet al. (1988),it hasbeen well recognized(Aufdenkampeetal.,2007;Hedgesetal.,2000;

Mayorgaet al.,2005; Quayetal.,1992)thatPOCinthe upperreachesoftheAmazonRiverhaslessnegative

d

¹³C values than in the central and lower reaches. The mechanism behind that general down-river trend has been proposed to be the smaller degree of isotope fractionation during photosynthesis by theC3 plants at thelowerpCO₂foundinaltitude.Ouraverage

d

¹³C-DOC values (–29.30.6%) forthe lowerAmazon were inthe samerangeasthosereportedinpreviousstudies(Mayorga etal.,2005;Pe´rezetal.,2011;Quayetal.,1992).Inparticular, thelessnegativevalueswereportfortheRioMadeiraagree well with the high-water stage value (–28.0%) given by Hedges etal. (1994) in the same predominantly upland- draining river(Quayetal., 1992).The smallvariabilityof

d

¹³C-DOCvalueswemeasuredduringthehydrologicalcycle fortheRioNegroandtheRioSolimo˜esattheentryofthe studied river section suggests relatively constant and homogeneous DOC sourcesfromupland soilsandflooded forests, in agreement with the degraded pattern of the dissolved organicmatter which has beenreportedin the lowlandAmazonreach(Hedgesetal.,1994,2000).

Theseasonalvariability for

d

¹³Cvaluesof organic C- speciesarelessdocumented.Thedatafromourstudyshow aclearseasonaltrend,withlower

d

¹³C-DOCvaluesduring HW and FWdownstream ofItacoatiara.A similartrend was observed,in this area, withlower

d

¹³C-POC values duringtheHWperiodthanduringotherperiods(Moreira- Turcqetal.,2013).Moreover,inthelowlandmainchannel, Quayetal.(1992)foundlower

d

¹³C-POCvalues (forthe fine POC fraction) during the FW period, yet did not observeanyclearseasonalvariationfor

d

¹³C-DOC.These lower

d

¹³C-POCvalues wereinterpretedasan increased contribution of C₃ sources from tributaries draining lowlandregionsandfromfloodplainsoilsandvegetation.

In contrast,wereporthigher

d

¹³C-DOCvalues inthe MadeiraRiver,themainchannelupstreamofItacoatiara during FW and at the lower reach stations during LW (Figs.3band4b).Whilethepersistenceoftheuplandeffect duetoaweakerdensityoffloodedforestcomparedtothe Solimo˜es Basinmayexplain thehigher

d

¹³C-DOCvalues foundintheMadeiraRiverandtheassociatedlake(Lago Miratuba),thequestionariseswhetheralargerinfluenceof C4-grasses can explain the higher

d

¹³C-DOC values we observedinthelowerreachduringLW.Althoughdecaying matterinthesemassiveplantraftswasfoundtoincrease

Fig.4.DOC(a)andd¹³C-DOC(b)valuesforCARBAMAcruises(October2008toJanuary2011)fromupstreamtodownstreamfloodplainlakes.

d

¹³C-DOC values locally by 1.5%, reaching –28.5%

(Fig. 5a), the raise appears poorly linked to C₄ plants, since

d

¹³C-DOC values, in microcosm experiments in which C4 plants sorted out from these mats were incubated, reached –24% to –17%, depending on the amountofmacrophytebiomassesused(Mortillaroetal., 2016).Incontrast,theimpactofdecayingplantmaterialin floatingmeadowsonDICconcentrationswasfoundtobe very strong, with an almost 10-fold increase in DIC concentration, while

d

¹³C-DIC values raised to –10.7%

(Fig.5b),avalueclearlyhigherthanthatinopenlakeand mainstemwatersatpH6.3(Quayetal.,1992),whichmay likely correspond to the respiration input of C4 plant carbon.Wethereforehypothesizethatmostoftheorganic matterproducedbythesemacrophytematsisdecayedin situ,andthatitseffectontheDOCpoolinlakewatersis relatively small. This observation agrees with previous resultsshowingthesmallcontributionofC4macrophytes to particulate matter and food-web in the lake and mainstem waters despite their high biomasses in the ecosystems(Moreira-Turcqetal.,2013;Mortillaroetal., 2011,2016;Quayetal.,1992).

4.2. Newinsightsfromvariationsduringtherunoffcycle

Thedatainthisworkpointtoacontrastingseasonality of

d

¹³C-DOCvaluesinthemainAmazonchannelbetween

thereachesupstreamanddownstreamoftheconfluence withtheRio Madeira(fromupstreamtodownstreamof Itacoatiara).ApartthespecificimpactoftheRioMadeira waters, the upstream reach is characterized by values clustering around the whole lowland Amazon average value (i.e. –29%), while the downstream reach is characterized bylower

d

¹³C-DOCvalues duringHWand FWand,toalesserextentbyhighervaluesduringLW.The lower

d

¹³C-DOC values during FW coincided with an increase in DOC concentrations in the main channel (Fig.3b).Fig.6illustratestheupstreamversusdownstream trendbetweenDOCconcentrationsand

d

¹³C-DOCduring theFWperiod. Wehypothesize higherplanktonicinflu- encefromfloodedlakesdownstreamofItacoatiaraduring HW and FW.More ¹³C-depleted values foundinCuruaı´

water during HW (Fig. 3b) support our hypothesis although not much variation was observed in Lago Canac¸ari. As indicated by Archer (2005), significant differences in floodplain characteristics occur at the confluence of theRio Madeira with the Rio Amazonas.

UpstreamoftheRioMadeiramouth,therearerelatively few lakes in the floodplain, while downstream of the Madeira,theareaischaracterizedbylargemeandersand secondary channels withincomplete levees resulting in inundation of large floodplainlakes. The channel-flood- plain geomorphology and valley width differences on eithersideofItacoatiaraareshowninFig.7a,basedonthe Table2

DOCconcentrations(mg
L¹)andd¹³C-DOCvaluesinfloodplainlakes(%versusVPDB).

Locations Cond./pH/TSS(mS
cm ¹/–/mg
L ¹) Season DOC(mg
L¹) d¹³C(%) n

Va´rzeaCabaliana

75/6.9/10 LW 4.80.1 –29.70.6 4

120/7.0/14 RW 7.00.4 –29.10.1 2

74/6.5/4 HW 4.20.1 –29.10.1 5

69/6.7/6 FW 5.10.7 –29.60.7 3

Va´rzeaJanauaca´

52/7.1/15 LW 4.10.5 –29.40.7 7

60/6.6/75 RW 3.80.5 –28.60.1 3

41/6.2/6 HW 4.30.1 –29.20.1 5

60/6.7/6 FW 4.40.3 –29.20.2 5

Dailyvariation^a HW 4.40.2 –29.60.2 27

Dailyvariation^b FW 4.81.0 –29.20.7 7

Depthprofile(0.5–10m) HW 4.40.2 –29.80.2 6

Depthprofile(0.5–10m)^c HW 4.20.1 –29.10.2 6

LagoMiratuba

56/6.8/65 LW 4.10.3 –27.60.3 3

52/–/28 RW 3.20.1 –28.00.3 2

47/6.3/26 HW 3.60.1 –28.30.5 5

61/7.6/29 FW 4.00.8 –27.40.7 5

LagoCanac¸ari

44/7.9/23 LW 4.20.1 –28.80.1 2

13/6.2/19 RW 4.01.4 –29.50.2 3

46/6.4/11 HW 5.10.2 –29.40.2 4

53/–/11 FW 6.10.8 –29.60.3 3

Dailyvariation^b FW 6.50.7 –28.80.3 7

Va´rzeaCuruaı´

42/8.1/51 LW 4.40.7 –27.00.4 5

59/6.4/22 RW 3.00.3 –28.70.4 4

45/6.5/15 HW 3.90.1 –30.00.3 8

57/9.5/60 FW 6.70.5 –29.10.3 3

Depthprofile(0.5–2m) HW 4.10.1 –30.30.5 3

aSampledeveryhouratadepthof0.5matasinglelocation.

b Sampledevery4hatadepthof0.5matasinglelocation.

cProfileinthenorthernpartoftheva´rzea.

dataofMertesetal.(1996,theirFig.2A)andDunneetal.

(1998, their Fig. 3A). We hypothesize that the larger floodplainlakeareadownstreamItacoatiaraaccountsfor thelower

d

¹³C-DOCvaluesduringtheHWandFWperiods (Fig.7b).Thehighphytoplanktonbiomassfoundinthese largerfloodplainlakes,withlikelylow

d

¹³Cvalues(–33%

to–37%),assuggestedbyEllisetal.(2012),maybethe source explaining the trend observed in the mainstem duringtheseperiodsoflargeexportoforganicmatterfrom thefloodplaintotheriver mainchannel(Moreira-Turcq etal.,2013).However,the

d

¹³Cvaluesofplanktonsources are difficulttoconstrain over largerscales,since inthe same studied area, values ranging from –23 to –40%

during LW, –31 to –43% during HW (Mortillaro et al., 2011),andreachingupto–18%duringLWdense(cyano)- phytoplanktonblooms(Moreira-Turcqet al.,2013)have been reported. The large variation in

d

¹³C-DOC values observedintheTapajo´sclearwaters,whichmatchesthat reportedfor

d

¹³C-POCvaluesinthesameriver(Mortillaro etal.,2011),maybealsolinkedtoseasonaldifferencesin phytoplanktonnatureandbiomass.

5. Conclusions

Three reaches in the lowland Amazon mainstem betweentheRioNegroconfluenceandtheTapajo´smouth couldbedistinguishedwithregardto

d

¹³C-DOCvariations:

(1) upstream of the Rio Negro and the Rio Solimo˜es confluence, almost no seasonal variation was observed aroundtheregionalaverageof–29.20.6%;(2)upstream ofItacoatiara(RioMadeiramouth),higher

d

¹³C-DOCvalues wereobservedduringthefallingwaterperiod,buttheorigin ofthis¹³C-enrichmentremainstobedeterminedunambigu- ously;and(3)downstreamtheRioMadeiramouth,larger meanders secondary channelsandlakes inthe floodplain mayresultin anincreased plankton contributionto DOC, leadingto aseasonalpatternwithlower

d

¹³C-DOC values during high and fallingwater periods,and higher values duringlowwaters.Ourresultsarebased,ontheonehand,on 4to5discretesamplingdatesinthemainstemalong the hydrological cycle, and, on the other hand, on hourly samplinginlakesthathavenotsuggestedsignificantdaily variations.Togofurtherwoulddemandadifferentsampling strategy,couplingphytoplanktondynamicand

d

¹³Cstudies at a recording rate allowing a better follow of the river regime.

Acknowledgments

ThisresearchisacontributiontotheCARBAMAproject, fundedbytheFrenchNationalAgencyforResearch(grant number 08-BLAN-0221), the French INSU national pro- grammeEC2CO,andtheNationalCouncilofResearchand Development(CNPq),Brazil(UniversalProgramnumber 477655/2010-6).Itwasconductedundertheauspicesof theEnvironmentalResearchObservatoryHYBAM,suppor- tedbytheINSUand theIRD (InstituteforResearchand Development,France).G.A.benefitsfromaBrazilianCNPq

‘‘ScienceWithoutBorder’’fellowship.Particularthanksare addressedtothecrewandthecaptainoftheYane-Jose´-IV (Manaus)forassistanceduringthecruises.Thecomments Fig.5. Variationofd¹³C-DOCvaluesandDOCconcentrations(a)andd¹³C-

DICvaluesandDICconcentrations(b)frominsideaC4plantrafttoopen waterinLagoCanac¸ari(RW,January2011).Ad¹³C-DOCversus1/DOCplot for thesamples closest to theplantraft is inserted in Fig.5a.DIC concentrationsarefromtwosamplingsets;oneanalyzedbyamethod adaptedfromBrandes(2009)inblacksquaresandtheother(3data,open diamonds)bytheregularmethodusedduringthisprogramfor¹³C-DIC analysis (seesection2). Analyticalerrors are comprisedwithin the symbols.

Fig.6. d¹³C-DOCversusDOCconcentrationsintheAmazonmainstem andfloodplainlakesduringfallingwaters.

andsuggestionsofthetwoanonymousreviewerswerea welcomehelptoimprovethemanuscript.

References

Abril,G.,Martinez,J.-M.,Artigas,L.F.,Moreira-Turcq,P.,Benedetti,M.F., Vidal,L.,Meziane,T.,Kim,J.-H.,Bernardes,M.C.,Savoye,N.,Deborde, J.,LimaSouza,E.,Albe´ric,P.,LandimdeSouza,M.F.,Roland,F.,2014.

AmazonRivercarbondioxideoutgassingfuelledbywetlands.Nature 505,395–398.

Abril,G.,Bouillon, S.,Darchambeau,F., Teodoru,C.R., Marwick,T.R., Tamooh,F.,OchiengOmengo,F.,Geeraert,N.,Deirmendjian,L.,Pol- senaere,P.,Borges,A.V.,2015.Technicalnote:largeoverestimationof pCO2calculatedfrompHandalkalinityinacidic,organic-richfresh- waters.Biogeosciences12,67–78,http://dx.doi.org/10.5194/bg-12- 67-2015.

Albe´ric,P.,2011.Liquidchromatography/massspectrometrystableiso- topeanalysisofdissolvedorganiccarboninstreamandsoilwaters.

RapidCommun.MassSpectrom.25,3012–3018.

Archer,A.W.,2005.ReviewofAmazoniandepositionalsystems.Spec.

Publs.Int.Ass.Sediment.35,17–39.

Aufdenkampe, A.K.,Mayorga, E., Hedges, J.I., Llerena,C., Quay, P.D., Gudeman,J.,Krusche,A.V.,Richey,J.E.,2007.Organicmatterinthe PeruvianheadwatersoftheAmazon:compositionalevolutionfrom theAndestothelowlandAmazonmainstem.OrganicGeochem.38, 337–364.

Bonnet,M.P.,Barroux,G.,Martinez,J.-M.,Seyler,F.,Moreira-Turcq,P., Cochonneau,G.,Melack,J.M.,Boaventura,G.,Maurice-Bourgoin,L., Leo´n,J.G.,Roux,E.,Calmant,S.,Kosuth,P.,Guyot,J.L.,Seyler,P.,2008.

FloodplainhydrologyinanAmazonfloodplainlake(LagoGrandede Curuai).J.Hydrol.349,18–30.

Bouillon,S.,Korntheuer,M.,Baeyens,W.,Dehairs,F.,2006.Anewauto- matedsetupforstableisotopeanalysisofdissolvedorganiccarbon.

Limnol.Oceanogr.Methods4,216.

Brandes,J.A.,2009.Rapidandprecised¹³Cmeasurementofdissolved inorganic carbon innatural watersusing liquidchromatography coupledtoanisotope-ratiomassspectrometer.Limnol.Oceanogr.

Methods7,730–739.

Cai,D.-L.,Tan,F.C.,Edmond,J.M.,1988.Sourcesandtransportofparticu- lateorganiccarbonintheAmazonRiverandestuary.Estuar.Coast.

ShelfSci.26,1–14.

Dunne,T.,Mertes,L.A.K.,Meade,R.H.,Richey,J.E.,Forsberg,B.R.,1998.

Exchangesofsedimentbetweenthefloodplainandchannelofthe AmazonRiverinBrazil.Geol.Soc.Am.Bull.110,450–467.

Ellis, E.E., Richey, J.E., Aufdenkampe, A.K.,Krusche, A.V., Quay, P.D., Salimon,C.,BrandaodeCunha,H.,2012.Factorscontrollingwater- columnrespirationinriversofthecentralandsouthwesternAmazon Basin.Limnol.Oceanogr.57,527–540.

Gillikin,D.P.,Bouillon,S.,2007.Determinationofd¹⁸Oofwaterandd¹³Cof dissolvedinorganiccarbonusingasimplemodificationofanelemen- talanalyser-isotoperatiomassspectrometer:anevaluation.Rapid Commun.MassSpectrom.21,1475–1478.

Hedges,J.I.,Cowie,G.L.,Richey,J.E.,Quay,P.D.,1994.Originsandpro- cessing of organic matter inthe Amazon River asindicated by carbohydratesandaminoacids.Limnol.Oceanogr.39,743–761.

Hedges,J.I.,Mayorga,E.,Tsamakis,E.,McClain,M.E.,Aufdenkampe,A.K., Quay,P.D.,Richey,J.E.,Benner,R.,Opsahl,S.,Black,B.,Pimentel,T., Quintanilla,J.,Maurice,L.,2000.OrganicmatterinBoliviantributaries oftheAmazonRiver:acomparisontothelowermainstem.Limnol.

Oceanogr.45,1449–1466.

Hess,L.L.,Melack,J.M.,Novo,E.M.,Barbosa,C.C.F.,Gastil,M.,2003.Dual- seasonmappingofwetlandinundationandvegetationforthecentral Amazonbasin.RemoteSens.Environ.87,404–428.

Marwick,T.R.,Tamooh,F.,Teodoru,C.R.,Borges,A.V.,Darchambeau,F., Bouillon, S., 2015.The ageofriver-transported carbon:a global Fig.7.a:schematicillustrationofalong-streampatternofchannel-floodplaingeomorphologyandvalleywidthaccordingtoMertesetal.(1996)andDunne etal.(1998);b:d¹³C-DOCvaluesinthemainstemandtributariesbetweenManacapuruandSantare´matlowwaterandfallingwater,highlightingthe divergenceofvaluesdownstreamofItacoatiara(datafromthisstudy,Fig.3b).

perspective. Global Biogeochem. Cycles 29 , http://dx.doi.org/

10.1002/2014GB004911.

Mayorga,E.,Aufdenkampe,A.K.,Masiello,C.A.,Krusche,A.V.,Hedges,J.I., Quay,P.D.,Richey,J.E.,Brown,T.A.,2005.Youngorganicmatterasa sourceofcarbondioxideoutgassingfromAmazonianrivers.Nature 436,538–541.

Mertes,L.A.K.,Dunne,T.,Martinelli,L.A.,1996.Channel-floodplaingeo- morphologyalongtheSolimo˜es–AmazonRiver.Brazil.Geol.Soc.Am.

Bull.108,1089–1107.

Millero, F.J., 1979.The thermodynamicsofthe carbonate systemin seawater.Geochim.Cosmochim.Acta43,1651–1661.

Miyajima,T.,Miyajima,Y.,Hanba,Y.T.,Yoshii,K.,Koitabashi,T.,Wada,E., 1995.Determiningthestableisotoperatiooftotaldissolvedinorganic carboninlakewaterbyGC/C/IIRMS.Limnol.Oceanogr.40,994–1000.

Moreira-Turcq,P.,Bonnet,M.-P.,Amorim,M.,Bernardes,M.,Lagane,C., Maurice, L., Perez,M.A.P., Seyler,P., 2013.Seasonal variabilityin concentration, composition, age,and fluxesof particulate organic carbonexchangedbetweenthefloodplainandAmazonRiver.Global Biogeochem.Cycles27,119–130,http://dx.doi.org/10.1002/gbc.20022.

Mortillaro,J.-M., Abril,G.,Moreira-Turcq, P.,Sobrinho, R.,Pe´rez, M., Meziane,T.,2011.Fattyacidandstableisotopes(d13C,d15N)signa- tures of particulateorganicmatter in theLower Amazon River:

seasonalcontrastsandconnectivitybetweenfloodplainlakesand themainstem.OrganicGeochem.42,1159–1168.

Mortillaro,J.-M.,Passarelli,C.,Abril,G.,Hubas,C.,Albe´ric,P.,Artigas,L.F., Benedetti,M.F.,Thiney,N.,Moreira-Turcq,P.,Pe´rez,M.A.P.,Vidal,L.O., Meziane,T.,2016.ThefateofC4andC3macrophytecarbonincentral Amazonfloodplainwaters:insightsfromabatchexperiment.Lim- nologica59,90–98.

Pe´rez,M.A.P.,Moreira-Turcq,P.,Gallard,H.,Allard,T.,Benedetti,M.F., 2011.DissolvedorganicmatterdynamicintheAmazonbasin:Sorp- tionbymineralsurfaces.Chem.Geol.286,158–168.

Quay,P.D.,Wilbur,D.O.,Richey,J.E.,Hedges,J.E.,Devol,A.H.,1992.Carbon cyclingintheAmazoneRiver:Implicationsforthe¹³Ccompositionof particlesandsolutes.Limnol.Oceanogr.37,857–871.

Richey,J.E.,Melack,J.M.,Aufdenkampe,A.K.,Ballester,V.M.,Hess,L.L., 2002.OutgassingfromAmazonianriversandwetlandsasalarge tropicalsourceofatmosphericCO2.Nature41,617–620.

Sioli,H.,1984.TheAmazonanditsmainaffluents:hydrography,mor- phologyoftherivercourses,andrivertypes.In:Sioli,H.(Ed.), The Amazon,LimnologyandLandscapeecologyofamightytropicalriver anditsbasin.DrW.JunkPublishers,Dordrecht, pp.127–165.

Weiss,R.,1974.Carbondioxideinwaterandseawater:thesolubilityofa non-idealgas.Mar.Chem.2,203–215.