The importance of terrestrial organic carbon inputs on Kara
Sea shelves as revealed by
n
-alkanes, OC and
d
13
C values
M.B. Fernandes
a,1, M.-A. Sicre
b,*
aInstituto de QuõÂmica, Universidade Federal do Rio de Janeiro, Centro de Tecnologia, Bloco A, Sala 607, 21949-900, Rio de Janeiro/RJ, Brazil bLaboratoire des Sciences du Climat et de l'Environnement, Domaine du CNRS, Avenue de la Terrasse, 91191 Gif-sur-Yvette Cedex, France
Abstract
Organic carbon (OC) andn-alkanes were investigated in the suspended matter (SM) and sur®cial sediments of the Arctic Ob and Yenisei Rivers and the adjacent Kara Sea. Organic nitrogen andd13C values were determined as
addi-tional parameters in the sediments. OC andn-alkane concentrations were highly variable in space. This variability was interpreted as the result of a complex sedimentation pattern involving the discharges of these two rivers and the sea-sonal occurrence of a shore-ice front along the coast. Using the constant ratio of high molecular weight oddn-alkanes to OC in both rivers, we estimated that more than 70% of the OC preserved in shelf sediments is terrestrially-derived. This result was substantiated by the overall dominance of land-derivedn-alkanes. A second approach usingd13C values
and assuming binary dilution of riverine and marine OC led to comparable estimates.#2000 Elsevier Science Ltd. All rights reserved.
Keywords:Organic carbon;d13C;n-Alkanes; Ob River; Yenisei River; Kara Sea
1. Introduction
Rivers are the prime sources of fresh water and sedi-ments to the ocean, representing 70% of the world's solid inputs to the marine realm (Milliman, 1991). However, river discharges can vary substantially with changes in climate and human land-use. In the twentieth century, the worldwide development of industrial and urban centers has motivated the construction of dams for hydroelectric plants and irrigation conduits inducing major alterations of river ¯ow and discharge of sus-pended solids reaching coastal margins (e.g. River Nile). It is estimated that 90% of riverine particulate inputs remain trapped in coastal areas (Mantoura et al., 1991). The extent to which these areas can retain or export ¯u-vial material can aect the global carbon mass balance.
The understanding of the transfer of ¯uvial organic carbon (OC) to the marine environment on a global
scale requires an extended geochemical database. Apart from rivers discharging into temperate North-American and European shelves, regions such as the Eurasian shelves of the Arctic Ocean are poorly documented (e.g. Dmitriev and Pivovarov, 1983; Fernandes, 1996; Yun-ker et al., 1996; Broyelle, 1997). The present study aims to partly ®ll these gaps. Samples were collected in Sep-tember 1993 during the SPASIBA 3 cruise of the R/V Dimitriy Mendeleev, as part of the international Joint Global Oceanic Flux Studies (JGOFS) and the Scienti®c Program on the Arctic and Siberian Aquatorium (SPA-SIBA) (Lisitsyn and Vinogradov, 1995). Elemental organic carbon and nitrogen (OC and N), carbon stable isotope (d13C), andn-alkanes (ALKs) were determined
in the suspended matter (SM) and surface sediments of the Ob and Yenisei estuaries and Kara Sea to estimate the amount of terrestrially-derived OC buried in shelf sediments.
2. Study area
The Ob and Yenisei rivers are the main source of fresh water into the Arctic Ocean. Their water discharges reach 400 and 630 km3/year, respectively (Khublarian,
0146-6380/00/$ - see front matter#2000 Elsevier Science Ltd. All rights reserved. P I I : S 0 1 4 6 - 6 3 8 0 ( 0 0 ) 0 0 0 0 6 - 1
Organic Geochemistry 31 (2000) 363±374
www.elsevier.nl/locate/orggeochem
* Corresponding author.
E-mail addresses: Marie-Alexandrine.Sicre@lsce.cnrs-gif.fr (M.-A. Sicre), milena@ensp.®ocruz.br (M.B. Fernandes).
1 Present address: CESTEH/ENSP, FundacËaÄo Oswaldo
1994), as a result of a higher runo in the Yenisei basin (218 mm/year) than in the Ob (164 mm/year) (Spitzy and Ittekkot, 1991). Water and suspended sediment dis-charges show a strong seasonality. Up to 90% of the Ob and Yenisei's solid suspended in¯ux is delivered during spring ¯oods (P®rman et al., 1995). The Ob mean solid discharge, around 504 kg/s, is almost 4 times higher than that of the Yenisei, averaging 130 kg/s (Dai, 1995). Sedimentation rates in the Yenisei estuary (3±100 cm/ 1000 years; Lein et al., 1995) are low in comparison to other major world estuaries. The vast catchment areas for these Arctic rivers (52.5106km2) combined with
low rainfall results in extremely low discharge/drainage basin area ratios (< 0.2 m/year) (Milliman, 1991). As a consequence of low-lying terrain and low precipitation, sediment yields (river load/drainage basin area) are also correspondingly low (4±9 t/km2/year) (Milliman, 1991).
Signi®cant quantities of dissolved and colloidal organic matter (OM), in particular humic substances, enter the waterways during freshet, as soils and peat deposits are ¯ushed by melt waters (Telang et al., 1991). As a result, dissolved organic carbon (DOC) inputs into Siberian rivers are more than 3 times higher than particulate organic carbon (POC) yields, which represent less than 0.5 t/km2/year (Spitzy and Ittekkot, 1991).
Low nutrients and low water temperatures do not favor primary production in the Kara Sea, which varies from mesotrophic to oligotrophic (Vedernikov et al., 1994, 1995; Vinogradov et al., 1994, 1995). At the time of the SPASIBA 3 cruise, diatoms were the dominant phytoplankton species (Lisitsyn et al., 1994; Dai and Martin, 1995).
3. Experimental
3.1. Sampling
Samples were collected in September 1993, well after the peak river ¯ow (June±July). Collection sites are located between 68 and 76N, along the Ob and Yenisei
rivers and in the adjacent Kara Sea. Three longitudinal sections of the estuary (riverine zone, mixing zone and marine zone) were distinguished based on the location of frontal zones and salinity values (Fig. 1b; Fernandes and Sicre, 1999). Salinity values ranged from 0 to 32 throughout the study area.
Surface water samples were collected above the halo-cline, between 3 and 9 m. 15 to 80 L of water were pumped into glass containers and then ®ltered on-board ship on pre-extracted Whatman GF/F glass ®ber ®lters (147 or 290 mm diameter, 0.7mm porosity). Filters and sur®cial sediments (0±3 cm) were wrapped in aluminum foil and kept frozen atÿ18oC until extraction. All
sur-faces in contact with the samples were of Te¯on or stainless steel.
3.2. Analytical procedure
Filters for hydrocarbon analysis were extracted three times by ultrasonication in 200 mL of a mixture of CH3OH:CH2Cl2(1:3 v/v) for 15 min. The three extracts
(600 mL) were combined in a round bottom ¯ask and dried with MgSO4overnight. The total lipid extract was
®ltered through glass wool, concentrated by rotary eva-poration and then transferred to a 4 mL vial. Sediments were freeze-dried and homogenized. An aliquot was extracted in a re¯ux system with 200 mL of a mixture of CH3OH: CH2Cl2 (1:3 v/v) for 2 h followed by
ultra-sonication for 15 min. The total lipid extract was ®ltered, concentrated by rotary evaporation and dried with MgSO4overnight. These extracts were then treated in the
same way as those of the particulate phase. Particulate and sedimentary non-aromatic hydrocarbons were isolated from the lipid mixture by elution with hexane through a pipette ®lled with silica gel (Fernandes et al., 1997).
3.3. Hydrocarbon identi®cation and quanti®cation
Quanti®cation of individual n-alkanes was based on comparison of peak areas with those of perdeuterated tetracosane added prior to extraction. The quanti®ca-tion limit for individual compounds was estimated at around 1±3 ng (signal/noise 5 10) and the reproduci-bility of the method was estimated at 10%. Blanks showed no detectable contamination. Hydrocarbons were analyzed using a CP-Sil-8 CB Chrompack capillary column (50 m0.32 mm i.d.0.25mm ®lm thickness)
mounted in a Girdel 3000 gas chromatograph (GC) equipped with a glass needle injector heated at 250
C and a Flame Ionization Detector (FID). The GC oven was programmed from 80 to 300
C at 2
C/min. Helium was used as carrier gas. Data acquisition and processing was performed on a Varian Vista CDS 402 integrator.
The identi®cation of individual n-alkanes was per-formed by comparison of their retention times with those of authentic standards. Structural assignment was con®rmed by gas chromatography/mass spectrometry (GC/MS) on selected samples. GC/MS analyses were carried out on a Varian 3400 GC coupled to an Ion Trap Varian Saturn mass spectrometer. Samples were injected using an on-column, Septum, Programmable Injector (SPI) on a DB5 J & W Scienti®c capillary
col-umn (30 m 0.25 mm i.d. 0.25 mm ®lm thickness)
programmed from 80 to 300
C at 3
C/min. Operating conditions were: ion source temperature 140
C, electron energy 70 eV, scanning from 40 to 600 amu, 0.6 scan/s.
3.4. Bulk analyses
Fig. 1. Study area (a) and sampling locations along the riverine, mixing and marine zones of the Ob and Yenisei River estuaries and Kara Sea shelf (b).
organic carbon (OC), organic nitrogen (N) and d13C
were analyzed following the experimental procedure by Girardin and Mariotti (1991). Brie¯y, samples were decarbonated in 25 mL of HCl (0.001 M), followed by the addition of HCl (0.5 M) till a constant pH around 3 was reached. They were centrifuged, washed with deio-nized water and dried for 12 h at 50
C. After being homogenized and weighed, aliquots were combusted in an elemental analyzer (Carlo Erba, NA 1500 NC, Fisons). The CO2 and N2 produced were
chromato-graphically separated. OC and N were then measured by thermal conductivity. The analytical reproducibility of these measurements was estimated at 3%. Combus-tion gases were then cryogenically puri®ed through sev-eral temperature traps and introduced into the isotopic ratio mass spectrometer ®tted for rapid switching (VG Sira 10 Spectrometer). Samples were compared to Pee Dee Belemnite (PDB) as reference to determine their d13C values. The experimental error was estimated to be
better than 0.1%obased on successive injections of
the tyrosine standard.
4. Results and discussion
4.1. Suspended matter
SM concentrations (Dai, 1995) are in the lower range of concentrations reported for world rivers (Meybeck, 1982; Ittekkott, 1988): they range from 4 to 143 mg/L (average of 30 mg/L,n=7) in the Ob River, and from 1 to 23 mg/L (average of 5 mg/L,n=11) in the Yenisei River, while < 1 mg/L in the Kara Sea. OC percentages (Dai, 1995) vary from 3 to 5% in the Ob River (average of 4%,n=7), and from 4 to 11% in the Yenisei (average of 7%,n=11). They are higher in the Kara Sea, ranging from 3 to 19% (average of 12%,n=5).
All concentration data for the particulate n-alkanes are given in Table 1. Total particulate odd carbon number n-alkanes from C25 to C31 (PALK25±31) and
OC concentrations (in ng/L and mg/L, respectively) are higher in the Ob River than in the Yenisei. Total n -alkanes (in mg/L) decrease seawards in both estuaries (Fig. 2). Except for some stations of the northern Kara Sea,n-alkane particle loads (inmg/g) are relatively con-stant, generally around 10 mg/g (Fig. 2), indicating the rather homogeneous composition of riverine SM. This translates into a high correlation betweenn-alkane con-centrations and OC or SM values (r>0.95,P<0.001, for both transects) and suggests that soil erosion and riverine hydrodynamics are the major factors control-ling the aquatic distribution of n-alkanes. Enhanced concentrations (inmg/g) are found in the northern Kara Sea, mainly at stations 4396, 4398 and 4399 (> 20mg/g). P®rman et al. (1995) reported elevated DDT levels in this area of the Kara Sea that could be attributed to
transport by drifting-ice. The Yamal Current ¯owing towards the northeast at the time of sampling can also be envisaged as a potential vector of additional SM inputs (Fig. 1b; Burenkov and Vasil'kov, 1995). How-ever, this hypothesis cannot be substantiated without samples collected within this current ¯ow. Prevailing odd and high molecular weight (5C25)n-alkanes
indi-cate major vascular plant wax inputs.n-C27is dominant
and carbon preference index values (CPI24±32) are
gen-erally > 3 (Table 1). The occurrence of alkylated poly-cyclic aromatic hydrocarbons, such as octahydro-chrysenes and tetrahydropicenes, generated from the aromatization of 3-hydroxy triterpenoids and other con-stituents of higher plants, further supports the hypothesis of a dominant terrestrial origin (Fernandes, 1996).
Total n-alkane particle loads in the eastern Eurasian Lena Delta and Laptev Sea are substantially higher than values found in this work (Table 2). Compositional fea-tures indicate signi®cant planktonic inputs (Broyelle, 1997) that are not observed in the Kara Sea, at least not during the SPASIBA 3 cruise. Such inputs may account partially for the observed dierences in concentrations with respect to our data. At the freshet of the Mackenzie River, with a similar basin cover (vegetation and soil), values are lower than ours (Table 2), butn-alkane and SM levels are also signi®cantly correlated (Yunker et al., 1991).
4.2. Sediments
Only a small portion of the OC transported by the Ob and Yenisei SM is preserved in the sediments. OC per-centages in both rivers are relatively constant at around 1.5% (n=17) (Fig. 3). OC percentages in sediments of the Kara Sea are signi®cantly lower (average of 0.7%,
n=6), but close to the average OC concentrations of ®ne-grained shelf and delta sediments (0.75%) calcu-lated for world rivers by Berner (1982). Besides direct inputs from soil erosion, the OC content of shelf sedi-ments may also re¯ect better preservation of OM in ®ne ¯uvial particles which bear the highest mineral surface areas (Prahl and Carpenter, 1983; Prahl, 1985; Mayer, 1994a,1994b; Keil et al., 1998).
d13C values determined from sediments of both rivers
(Table 3) are close to the isotopic signature of con-tinental C3 vegetation (e.g. GonÄi et al., 1997, 1998; Ruttenberg and GonÄi, 1997). High C/N ratios suggest that the contribution of plankton (C/N<6; McCon-naughey and McRoy, 1979a; Meybeck, 1982) is minor (Fig. 3). Both d13C (from
ÿ29.0 to ÿ24.4) and C/N
(from 7 to 13) values reported here cover the same ran-ges as those found by Krishnamurthy et al. (in press) for sediments in the same area (d13C from
ÿ28.5 toÿ24.4,
Totaln-alkane concentrations are on average11mg/
g in both Ob and Yenisei sediments. Values in the Kara Sea are around 4mg/g, i.e. lower than those recorded in the Canadian Mackenzie Shelf and in the Eurasian Lena
Delta, but higher than in Alaskan Arctic shelves where river inputs are smaller (Table 2). As for SM, the ratio of odd n-alkanes from C25 to C31 to OC is relatively
constant in the estuary. The correlation between these
Table 1
C25, C27, C29and C31individual and total n-alkane concentrations (inmg/g and ng/L) in the suspended matter along the Ob and
Yenisei estuaries and Kara Sea shelf; CPI24±32values are also reported
Sample C25
(mg/g) C27
(mg/g) C29
(mg/g) C31
(mg/g) Total (mg/g)
Total (ng/L)
CPI24±32 Sample C25
(mg/g) C27
(mg/g) C29
(mg/g) C31
(mg/g) Total (mg/g)
Total (ng/L)
CPI24±32
Marine zone (Kara Sea)
4397 2.1 3.6 2.4 2.3 10.4 9.4 3.0 4398 6.0 11.6 7.0 6.0 30.5 13.7 3.3 4396 5.9 12.0 7.2 7.4 32.5 18.2 4.2 4399 4.2 8.3 4.8 4.3 21.6 12.5 3.9 4395 2.4 4.7 3.4 3.1 13.6 7.4 2.5 4400 ± ± ± ± ± ± ±
Ob mixing zone Yenisei mixing zone
4414 1.1 1.7 1.2 1.1 5.2 20.7 3.7 4401 2.7 4.6 3.1 2.8 13.3 14.1 3.7 4415 1.3 1.6 1.3 1.1 5.3 31.1 3.2 4402 2.3 4.0 2.8 2.6 11.6 11.9 3.6 4420 3.6 4.6 3.4 2.9 14.4 128.5 4.0 4403 3.3 6.2 4.9 4.6 18.8 23.2 4.0 4416 2.4 3.4 2.2 1.9 9.8 92.4 4.2 4404 7.8 13.7 9.7 9.3 40.5 86.7 5.4 4419 3.0 4.1 2.6 2.3 12.0 191.7 4.3
Ob riverine zone Yenisei riverine zone
4417 3.1 4.4 2.9 2.5 12.9 259.1 4.4 4413 2.1 3.2 2.7 2.3 10.3 51.3 3.8 4418 2.0 2.9 2.0 1.7 8.5 1220.4 4.5 4411 ± ± ± ± ± 37.3 3.2 4412 3.2 4.9 3.8 3.1 15.0 49.5 3.7 4405 2.1 3.6 2.7 2.6 11.0 76.8 5.6 4410 2.6 4.2 3.5 2.8 13.0 59.2 4.4 4406 2.5 4.1 3.0 2.8 12.3 278.3 5.5
4409 ± ± ± ± ± ± ±
4407 1.9 3.3 2.9 2.4 10.5 56.4 4.8 4408 2.0 3.4 2.8 2.5 10.7 61.6 5.1
±, Data not available.
Fig. 2. Total particulate concentrations of oddn-alkanes from C25to C31(ALK25±31inmg/L andmg/g) along the Ob (a) and Yenisei
(b) transects. POC data (in mg/L) from Dai (1995) is also shown from upriver the Ob (c) and Yenisei (d) rivers to the Kara Sea.
two parameters is highly signi®cant (r=0.95,P<0.001). However, concentrations are quite variable in space (Fig. 3). High values, mainly in the mixing zone, indi-cate an area of preferential deposition (Fig. 3). River runo and estuarine mixing are, in general, the domi-nant factors controlling particle deposition/resuspension episodes. However, the sedimentation pattern in this region is complicated by ice dynamics. Areas of unfro-zen sea water surrounded by ice (polynyas) are known to have a recurrence of 80% o the mouths of the Ob and Yenisei in late winter (P®rman et al., 1995). These polynyas develop in an area lying approximately between stations 4414 and 4396 and between stations 4401 and 4399 (Fig. 1b; Pavlov and P®rman, 1995). In late winter, the water column in the polynyas along the fast-ice front is destabilized by brine rejection during ice formation. This leads to water convection and con-sequent resuspension of ®ne organic-rich sediments, which can become trapped into the newly-forming ice. The OC content of particles in Arctic coastal ice can reach as much as 4% (Stein and Fahl, in press). As a consequence, areas under the polynyas may be depleted in organic-rich sediments (Fig. 3).
n-C27 is the dominant sedimentaryn-alkane, as also
observed in other Eurasian shelves such as the Barents Sea (Yunker et al., 1996) and the Laptev Sea (Broyelle, 1997). At the northernmost marine stations (4397 and 4398),n-C31prevails. Yunker et al. (1991, 1993)
repor-ted the dominance of n-C31 in peat samples of the
Mackenzie Basin and explained the shift fromn-C27to n-C29/n-C31 in the outer estuarine sediments of the
Fig. 3. Total sedimentary levels of oddn-alkanes from C25to C31(ALK25±31inmg/g andmg/gOC) along the Ob (a) and Yenisei (b)
transects. OC levels (in mg/g) and C/N ratios in sediments collected along the estuaries of the Ob (c) and Yenisei (d) rivers are also plotted from upriver to the Kara Sea.
Table 2
Total n-alkane levels recorded in Eurasian, Canadian and Alaskan shelves of the Arctic
Study area Totaln-alkanes (mg/g)
Suspended matter North America
Mackenzie Rivera 12.44.3
Siberia
Lena Delta/Laptev Seab 11±217 (65)
Kara Seac 22±54(38.414.7)
Ob Riverc 10±26 (16.96.2)
Yenisei Riverc 19±68 (27.114.2)
Sediments North America
Mackenzie Shelfa 9.81.7
Beaufort Sea (Alaska)d 1.4±5.0
SE Bering Sead 0.3±2.9
Norton Soundd 0.01±5.4
Navarin Basind 0.25±2.6
Gulf of Alaskad 0.06±2.0
Kodiak Shelfd 0.01±0.8
Cook Inletd <0.01±3.6
Siberia
Lena Delta/Laptev Seab 0.3±38.5(18)
Kara Seac 2±7(4.02.0)
Ob Riverc 4±16(11.84.9)
Yenisei Riverc 3±17 (10.73.8) a Sum ofn-C
13-n-C36(Yunker and Macdonald, 1995). b Sum ofn-C
15-n-C36(Broyelle, 1997). c Sum ofn-C
16-n-C36(this work). d Sum ofn-C
Table 3
C25, C27, C29and C31individual and totaln-alkane concentrations (inmg/g dry weight andmg/gOC), CPI24±32values, OC concentrations (in mg/g) andd13C values (in%PDB) in
sediments along the Ob and Yenisei estuaries and Kara Sea shelf
Sample C25
(mg/g) C27
(mg/g) C29
(mg/g) C31
(mg/g) Total (mg/g)
Total (mg/gOC)
CPI24±32 OC
(mg/g)
d13C
(%)
Sample C25
(mg/g) C27
(mg/g) C29
(mg/g) C31
(mg/g) Total (mg/g)
Total (mg/gOC)
CPI24±32 OC
(mg/g)
d13C
(%)
Marine zone (Kara Sea)
4397 0.8 1.2 1.2 1.2 4.5 375 4.9 11.9 ÿ24.4 4398 0.3 0.5 0.5 0.5 1.9 305 5.0 6.1 ÿ24.4 4396 0.3 0.5 0.4 0.4 1.7 385 4.8 4.3 ÿ25.2 4399 0.7 1.0 0.9 0.9 3.5 296 5.3 11.9 ÿ24.6 4395 0.2 0.4 0.3 0.3 1.2 400 4.9 3.1 ÿ25.4 4400 0.3 0.5 0.4 0.4 1.8 367 5.0 4.8 ÿ25.4
Ob mixing zone Yenisei mixing zone
4414 2.0 2.8 2.2 2.2 9.3 490 5.1 19.0 ÿ26.8 4401 1.4 2.3 2.0 2.0 7.7 473 5.4 16.2 ÿ25.9
4415 1.4 2.0 1.5 1.4 6.2 509 4.9 12.2 ÿ27.5 4402 1.4 2.5 2.1 2.3 8.4 545 5.9 15.4 ÿ26.5
4420 ±a ± ± ± ± ± ± ± ± 4403 1.5 2.7 2.3 2.3 8.7 431 6.3 20.3
ÿ26.7 4416 2.2 3.1 2.2 1.9 9.4 478 4.8 19.6 ÿ29.0 4404 1.9 3.5 2.7 2.6 10.7 456 6.1 23.4 ÿ27.0
4419 ± ± ± ± ± ± ± ± ±
Ob riverine zone Yenisei riverine zone
4417 0.5 0.7 0.5 0.5 2.3 357 4.6 6.4 ÿ28.4 4413 1.1 2.5 2.0 2.0 7.6 486 8.0 15.6 ÿ26.8
4418 1.6 2.4 1.7 1.4 7.1 470 5.2 15.2 ÿ28.7 4411 1.1 2.4 1.7 1.6 6.8 403 7.6 16.9 ÿ27.0
4412 0.3 0.5 0.4 0.4 1.6 399 6.5 4.0 ÿ26.5
4405 1.4 3.0 2.2 2.1 8.7 504 7.2 17.2 ÿ27.2 4410 0.9 2.2 1.6 1.5 6.3 340 8.1 18.4 ÿ26.8 4406 0.7 1.4 1.0 0.9 4.0 510 7.4 7.8 ÿ27.2 4409 1.2 2.8 2.0 2.0 8.0 440 8.2 18.1 ÿ27.4
4407 ± ± ± ± ± ± ± ± ±
4408 0.7 1.7 1.3 1.3 5.0 465 7.7 10.7 ÿ26.5
a Data not available.
M.B.
Fernandes,
M.-A.
Sicre
/
Organic
Geochemistry
31
(2000)
363±374
Mackenzie River by selective deposition of peat parti-cles towards the outer shelf. Transport by nepheloid brines and drifting-ice formed in the polynyas in late winter (P®rman et al., 1995) could also account for the presence of ®ne low density peat inputs oshore. The dominance of n-C31 in sediments of the Arctic Ocean
has been attributed to similar mechanisms by Schubert and Stein (1996). According to Yunker et al. (1991, 1993) peat inputs would be a seasonal phenomena thus, more easily observed in the sediments than in the SM.
Sedimentary CPI24±32andd13C values (Fig. 4) clearly
discriminate the OM in both rivers. Sediments from the mixing zone have CPI24±32values intermediate between
those of the Ob and Yenisei while thed13C values tend
to be heavier than they should be by mixing only. Fur-ther oshore, OC is even more enriched in13C (Table 3).
Admixtures of marine constituents likely account for the carbon isotope composition changes, as also indicated by decreasing C/N ratios seawards (Fig. 3). These fea-tures suggest the production and settling of OM enri-ched with N-containing compounds and in 13C, most
probably autochthonous. However, the CPI24±32 and d13C values indicate that the OM deposited in shelf
sediments mainly originates from land erosion.
4.3. Terrestrial deposition
All the parameters used for this study emphasize the accumulation of continentally-derived material over the Kara shelf. In order to quantify the percentage of ter-restrial OC sequestered in the sediments of the Kara Sea (%TERR), two dierent methods were employed (Prahl et al., 1994, and references therein):
The ®rst method is based on the correlation earlier evidenced between sedimentaryn-alkanes and OC. The equations of the linear ®t between OC (in g/g dry wt)
and terrigenous odd n-alkanes from C25 to C31
(ALK25±31 in mg/g dry wt) for both rivers is given
below and shown in Fig. 5:
ALK25ÿ31451OC0:117 r0:95;P<0:001
This equation indicates that in the absence of terrige-nous n-alkanes, OC concentrations tend to zero in the river, which implies that sedimentary ¯uvial OC is almost entirely terrigenous (Fig. 5). It also suggests that plantwax n-alkanes and terrestrial OC are exported from the mouth of both rivers in a ®xed ratio, which equals the slope of this linear regression. The change of this ratio in shelf sediments would then re¯ect the dilu-tion with marine OC that does not contain plantwaxn -alkanes. Accordingly, the %TERR for each coastal sediment of the Kara Sea can be estimated by the fol-lowing proportion:
%TERR ALK25ÿ31=OCsample slope x100
The slope de®ned by the Ob and Yenisei dataset (450) is steeper than for the temperate Columbia River
discharging along the Washington coast, around 280 (Prahl et al., 1994). %TERR in the Kara Sea calculated from this method are comprised in the range 65±90% (Table 4).
The second approach uses d13C values and assumes
that OM in coastal sediments derives from the binary mixing of riverine and marine OC. The %TERR in this case is calculated by the following equation:
%TERR
13C
sampleÿ 13Cmarine
13C
riverineÿ 13Cmarine x100
Fig. 4. d13C vs. CPI
24±32values in sediments along the Ob and
Yenisei transects into the Kara Sea.
Fig. 5. Terrigenous oddn-alkanes from C25to C31(ALK25±31)
The d13C
riverine endmember is estimated taking into
consideration the solid discharges of each river, 130 kg/s for the Yenisei and 504 kg/s for the Ob (Dai, 1995):
13Criverine
Yeniseiis the mean value for Yenisei sediments
from stations 4408 to 4401 (ÿ26.8%) andd13CObis the
mean value for Ob sediments from stations 4418 to 4414 (ÿ28.1%). The riverine endmember calculated from this
equation isÿ27.8%. The drawback of this estimation is
the lack of knowledge on the isotopic composition of the river-borne suspended particles that are actually being deposited on the shelf on a year-round basis. The absence of odd high molecular weight n-alkanes from C25to C31in the marine OM is a crucial assumption in
the estimation of d13C
marine. The assumption of an
almost exclusive terrestrial origin for these compounds in the Kara Sea is supported by sedimentary alkane, alkanol and fatty acid ®ngerprints reported by Belyaeva et al. (1995). The marine d13C endmember can be
ten-tatively predicted from thex-intercept (fory0) of the
linear regression of ALK25±31/OC vs. d13C values in
the marine sediments (Fig. 6). The equation obtained from the linear ®t of the six Kara Sea sediments is:
ALK25ÿ31=OC ÿ61:413Cÿ1175 r ÿ0:68;P<0:5
The marined13C endmember derived from this
equa-tion is ÿ19.1%o. However, this regression line passes
well over the riverine endmember (Fig. 6) and the cor-relation is poor. If we include the mean riverine end-member in the data set (d13C=
ÿ27.8%), the correlation
coecient rises to ÿ0.84 (P<0.05), but the regression
line (SALK25±31/OC=ÿ39.5 d13C-630) leads to a
heavier marine endmember (ÿ16.0%). For comparison,
average values previously reported for Bering Sea are
ÿ24.4%ofor phytoplankton,ÿ22.1% for zooplankton
andÿ22%in sediments (McConnaughey and McRoy,
1979a,1979b). However, planktonic d13C values are
known to span a wide range (ÿ10 toÿ31%) depending
on species and environmental conditions (Wong and Sackett, 1978; Showers and Angle, 1986, and references therein). At high latitudes thed13C of phytoplankton is
expected to be more depleted in 13C than found from
our calculations due to a higher solubility of dissolved CO2in cold waters and thus enhanced isotopic
fractio-nation (Sackett et al., 1965; Fontugne and Duplessy, 1981). However, in coastal environments temperature may not be the only or dominant factor aecting the carbon isotopic ratio of autochthonous matter. The isotopic composition of dissolved inorganic carbon
(DIC) may also play a part. Cifuentes et al. (1988) reported values as high as ÿ16.6% in the SM of the lower Delaware estuary in spring, where biomass and primary production rates were high. This result was attributed in part to the d13C value of the DIC, but
mostly to decreasing isotopic fractionation as diusion of CO2across the cell becomes the rate-limiting step of
photosynthesis. These factors possibly explain the heavy d13C values of the marine endmember calculated here.
However, inventories of d13C signatures of
auto-chthonous organisms growing in coastal environments would be necessary to better constrain binary dilution models.
Being aware of the diculties associated with accu-rate de®nition of endmembers, we calculated the %TERR in coastal sediments using the estimated river-ine endmember (ÿ27.8%) and the two marine
end-members (ÿ19.1 and ÿ16.0%o) derived from our
calculations (Table 4). Estimates of the %TERR are only about 10% dierent depending on the marine
endmember used. Those derived from d13C
marine= ÿ16.0%fall in the same range as the values calculated
by the ®rst method. They indicate that approximately 70±80% of the OC present in Kara Sea sediments is of terrestrial origin. Krishnamurthy et al. (in press) calcu-lated %TERR >50 for the Kara Sea based on sedi-mentary isotopic signatures. Recent work has also shown that OM accumulating in other Eurasian shelves (e.g. Kuptsov and Lisitsin, 1996; Yunker et al., 1996; Broyelle, 1997; Fahl and Stein, 1997; Stein and Fahl, in press) and on the west Beaufort Sea (Shaw et al., 1979; Steinhauer and Boehm, 1992) is mainly terrigenous although this has not been quanti®ed. The %TERR
Fig. 6. Terrigenous oddn-alkanes from C25to C31(ALK25±31/
OC) vs.d13C values in Kara Sea sediments. The bold line
indi-cates the linear regression obtained from the six Kara Sea sediments. The dashed line indicates the linear regression obtained from the six Kara Sea sediments and the mean river-ine endmember.
values calculated for the Columbia shelf using terrige-nous n-alkane and OC data average 59%, while those calculated fromd13C values average 67% (Prahl et al.,
1994). On the tropical Amazon shelf, %TERR estimates based on sediment accumulation rates, OC and d13C
values are around 70 % (Showers and Angle, 1986).
5. Conclusions
Geochemical data provide evidence of the strong in¯uence of the Ob and Yenisei discharges more than 100 km oshore. These rivers supply terrigenousn-alkanes to the Kara Sea in a relatively constant ratio to OC. n -Alkane/OC values are not statistically dierent for the Ob and Yenisei, but values previously reported for the tem-perate Columbia River are notably lower. These results point out the geographical speci®city of this ratio, which needs to be empirically evaluated for each coastal system. Based onn-alkane, OC andd13C values, the contribution
of terrestrial OC in shelf sediments was estimated to be around 70% or more of the total sedimentary OC. Our values fall within the range calculated for other tropical or temperate continental margins under the in¯uence of large rivers, despite dierences in the composition of river-borne OM (e.g. dierentn-alkane/OC ratios). This result corroborates the hypothesis that the amount of terrestrial OC preserved in coastal sediments might be largely controlled by interactions between OM and mineral surfaces (e.g. Keil et al., 1994; Mayer, 1994a,b; Keil et al., 1997, 1998; Mayer et al., 1998; Hedges and Keil, 1999). Autochthonous inputs do not seem to be sig-ni®cantly preserved in the sediments despite the shallow water depth of the Kara Sea (<200 m).
Acknowledgements
We would like to thank the Conselho Nacional de Des-envolvimento Cientõ®co e TecnoloÂgico (CNPq, BrasõÂlia, Brazil) for a research scholarship to M.B.F. We also
acknowledge ®nancial support provided by FundacËaÄo de Amparo aÁ Pesquisa do Estado do Rio de Janeiro (FAPERJ, Rio de Janeiro, Brazil) and CoordenacËaÄo de AperfeicËoamento de Pessoal de NõÂvel Superior (CAPES, BrasõÂlia, Brazil). We are grateful to Dr. Minhan Dai for supplying POC and SM data. Dr. Andre Mariotti is acknowledged for providing access to CHN and isotope ratio monitoring equipment. Micheline Grably and Cyril Girardin are thanked for help during sedimentary OC andd13C analysis. We also appreciated Anne Lorre's help
during analytical procedure. We acknowledge Dr. Kirsten Fahl, Dr. Fred Prahl and an anonymous reviewer, who provided constructive comments on this manuscript. This is LSCE contribution number 0343.
References
Belyaeva, A.N., Madureira, L.A.S., Eglinton, G., 1995. Assessment of various source lipids in the Kara Sea sedi-ments. In: Grimalt, J.O., Dorronsoro, C. (Eds.), Organic Geochemistry: Developments and Applications to Energy, Climate, Environment and Human History. AIGOA, San SebastiaÂn, Spain, pp. 670±672.
Berner, R.A., 1982. Burial of organic carbon and pyrite sulfur in the modern ocean: its geochemical and environmental sig-ni®cance. American Journal of Science 282, 451±473. Broyelle, I., 1997. Contribution aÁ l'eÂtude de la biogeÂochimie de
la matieÁre organique en milieu estuarien: cas des deltas du RhoÃne et de la LeÂna. Ph.D. thesis, Universite Pierre et Marie Curie (Paris 6), Paris, France.
Burenkov, V.I., Vasil'kov, A.P., 1995. The in¯uence of runo from land on the distribution of hydrologic characteristics of the Kara Sea. Oceanology 34, 591±599.
Cifuentes, L.A., Sharp, J.H., Fogel, M.L., 1988. Stable carbon and nitrogen isotope biogeochemistry in the Delaware estu-ary. Limnology and Oceanography 33, 1102±1115.
Dai, M.-H., 1995. RoÃle des colloõÈdes dans le transfert du car-bone organique et des eÂleÂments meÂtalliques associeÂs en milieu estuarien et coÃtier. Ph.D. thesis, Universite Pierre et Marie Curie (Paris 6), Paris, France.
Dai, M.-H., Martin, J.-M., 1995. First data on the trace metal level and behaviour in two major Arctic river/estuarine systems Table 4
Assessment of the terrestrial contribution to organic carbon (%TERR) in sediments of the Kara Sea using two dierent approaches:
n-alkanesOC, andn-alkanes/OCd13C (with two dierent marine endmembers)
(Ob & Yenisey) and in the adjacent Kara Sea. Earth and Planetary Science Letters 131, 127±141.
Dmitriev, F.A., Pivovarov, S.V., 1983. Hydrocarbons in snow, ice, and water in the northern part of the Kara Sea. Soviet Meteorology and Hydrology 5, 67±71.
Fahl, K., Stein, R., 1997. Modern organic carbon deposition in the Laptev Sea and the adjacent continental slope: surface water productivity vs. terrigenous input. Organic Geochem-istry 26, 379±390.
Fernandes, M.B., 1996. Contribution aÁ l'eÂtude du comporte-ment des hydrocarbures non aromatiques et aromatiques en milieu estuarien. Ph.D. thesis, Universite Pierre et Marie Curie (Paris 6), Paris, France.
Fernandes, M.B., Sicre, M.-A., 1999. Polycyclic aromatic hydrocarbons in the Arctic: Ob and Yenisei Estuaries and Kara Sea Shelf. Estuarine, Coastal and Shelf Science 48, 725±737.
Fernandes, M.B., Sicre, M.-A., Boireau, A., Tronczynski, J., 1997. Polyaromatic hydrocarbon (PAH) distributions in the Seine River and its estuary. Marine Pollution Bulletin 34, 857±867.
Fontugne, M.R., Duplessy, J.-C., 1981. Organic carbon iso-topic fractionation by marine plankton in the temperature rangeÿ1 to 31C. Oceanologica Acta 4, 85±90.
Girardin, C., Mariotti, A., 1991. Analyse isotopique du13C en
abondance naturelle dans le carbone organique: un systeÁme automatique avec robot preÂparateur. Cahiers Orstom SeÂrie PeÂdologie 26, 371±380.
GonÄi, M.A., Ruttenberg, K.C., Eglinton, T.I., 1997. Sources and contribution of terrigenous organic carbon to surface sediments in the Gulf of Mexico. Nature 389, 275±278. GonÄi, M.A., Ruttenberg, K.C., Eglinton, T.I., 1998. A
reas-sessment of the sources and importance of land-derived organic matter in surface sediments from the Gulf of Mexico. Geochimica et Cosmochimica Acta 62, 3055±3075.
Hedges, J.I., Keil, R.G., 1999. Organic geochemical perspec-tives on estuarine processes: sorption reactions and con-sequences. Marine Chemistry 65, 55±65.
Ittekkot, V., 1988. Global trends in the nature of organic mat-ter in river suspensions. Nature 332, 436±438.
Keil, R.G., MontlucËon, D.B., Prahl, F.G., Hedges, J.I., 1994. Sorptive preservation of labile organic matter in marine sediments. Nature 370, 549±552.
Keil, R.G., Mayer, L.M., Quay, P.D., Richey, J.E., Hedges, J.I., 1997. Loss of organic matter from riverine particles in deltas. Geochimica et Cosmochimica Acta 61, 1507±1511. Keil, R.G., Tsamakis, E., Giddings, J.C., Hedges, J.I., 1998.
Biochemical distributions (amino acids, neutral sugars, and lignin phenols) among size-classes of modern marine sedi-ments from the Washington coast. Geochimica et Cosmo-chimica Acta 62, 1347±1364.
Khublarian, M.G., 1994. Quality of continental and marine waters in the Arctic Basin of Russia. Arctic Research of the U.S 8, 266±269.
Krishnamurthy, R.V., Machavaram, M., Baskaran, M., Champ, M., Brooks, J., in press. Organic carbon ¯ow in the Ob, Yenisey Rivers and Kara Sea of the Arctic region. Mar-ine Pollution Bulletin.
Kuptsov, V.M., Lisitsin, A.P., 1996. Radiocarbon of Qua-ternary along shore and bottom deposits of the Lena and the Laptev Sea sediments. Marine Chemistry 53, 301±311.
Lein, A.Y., Miller, Y.M., Namsaraev, B.B., Pimenov, N.V., Rusanov, I.I., Savvichev, A.S., Ivanov, M.V., Pavlova, G.A., 1995. Biogeochemical processes of sulfur turnover in the early stages of sediment diagenesis along the Yenisey River-Kara Sea section. Oceanology 34, 619±629.
Lisitsyn, A.P., Vinogradov, M.E., 1995. International high-latitude expedition in the Kara Sea (the 49th cruise of the R/ V Dmitriy Mendeleev). Oceanology 34, 583±590.
Lisitsyn, A.P., Shevchenko, V.P., Vinogradov, M.E., Severina, O.V., Vavilova, V.V., Mitzkevich, I.N., 1994. Particle ¯uxes in the Kara Sea and Ob and Yenisei Estuaries. Oceanology 34, 758, (in Russian).
Mantoura, R.F.C., Martin, J.-M., Wollast, R., 1991. Intro-duction. In: Mantoura, R.F.C., Martin, J.-M., Wollast, R. (Eds.), Ocean Margin Processes in Global Change. John Wiley & Sons, Chichester, England, pp. 1±3.
Mayer, L.M., 1994a. Surface area control of organic carbon accumulation in continental shelf sediments. Geochimica et Cosmochimica Acta 58, 1271±1284.
Mayer, L.M., 1994b. Relationships between mineral surfaces and organic carbon concentrations in soils and sediments. Chemical Geology 114, 347±363.
Mayer, L.M., Keil, R.G., Macko, S.A., Joye, S.B., Ruttenberg, K.C., Aller, R.C., 1998. Importance of suspended particu-lates in riverine delivery of bioavailable nitrogen to coastal zones. Global Biogeochemical Cycles 12, 573±579.
McConnaughey, T., McRoy, C.P., 1979a. Food-web structure and the fractionation of carbon isotopes in the Bering Sea. Marine Biology 53, 257±262.
McConnaughey, T., McRoy, C.P., 1979b.13C label identi®es
eelgrass (Zostera marina) carbon in an Alaskan estuarine food web. Marine Biology 53, 263±269.
Meybeck, M., 1982. Carbon, nitrogen, and phosphorus transport by world rivers. American Journal of Science 282, 401±450. Milliman, J.D., 1991. Flux and fate of ¯uvial sediment and
water in coastal seas. In: Mantoura, R.F.C., Martin, J.-M., Wollast, R. (Eds.), Ocean Margin Processes in Global Change. John Wiley & Sons, Chichester, England, pp. 69±89. Pavlov, V.K., P®rman, S.L., 1995. Hydrographic structure and variability of the Kara Sea: implications for pollutant dis-tribution. Deep-Sea Research II 42, 1369±1390.
P®rman, S.L., KoÈgeler, J., Anselme, B., 1995. Coastal environ-ments of the western Kara and eastern Barents Seas. Deep-Sea Research II 42, 1391±1412.
Prahl, F.G., 1985. Chemical evidence of dierential particle dispersal in the southern Washington coastal environment. Geochimica et Cosmochimica Acta 49, 2533±2539.
Prahl, F.G., Carpenter, R., 1983. Polycyclic aromatic hydro-carbon (PAH)-phase associations in Washington coastal sediment. Geochimica et Cosmochimica Acta 47, 1013±1023. Prahl, F.G., Ertel, J.R., GonÄi, M.A., Sparrow, M.A., Ever-smeyer, B., 1994. Terrestrial organic carbon contributions to sediments on the Washington margin. Geochimica et Cos-mochimica Acta 58, 3035±3048.
Ruttenberg, K.C., GonÄi, M.A., 1997. Phosphorous distribu-tion, C:N:P ratios, andd13C
ocin arctic, temperate, and
tro-pical coastal sediments: tools for characterizing bulk sedimentary organic matter. Marine Geology 139, 123±145. Sackett, W.M., Eckelmann, W.R., Bender, M.L., Be, A.W.H.,
1965. Temperature dependence of carbon isotope composi-tion in marine plankton and sediments. Science 148, 235±237.
Schubert, C.J., Stein, R., 1996. Deposition of organic carbon in Arctic Ocean sediments: terrigenous supply vs marine pro-ductivity. Organic Geochemistry 24, 421±436.
Shaw, D.G., McIntosh, D.J., Smith, E.R., 1979. Arene and alkane hydrocarbons in nearshore Beaufort Sea sediments. Estuarine and Coastal Marine Science 9, 435±449.
Showers, W.J., Angle, D.G., 1986. Stable isotopic character-ization of organic carbon accumulation on the Amazon continental shelf. Continental Shelf Research 6, 227±244. Spitzy, A., Ittekkot, V., 1991. Dissolved and particulate organic
matter in rivers. In: Mantoura, R.F.C., Martin, J.-M., Wol-last, R. (Eds.), Ocean Margin Processes in Global Change. John Wiley & Sons, Chichester, England, pp. 5±17. Stein, R., Fahl, K., in press. Holocene accumulation of organic
carbon at the Laptev Sea continental margin (Arctic Ocean): sources, pathways, and sinks. Geo-Marine Letters.
Steinhauer, M.S., Boehm, P.D., 1992. The composition and dis-tribution of saturated and aromatic hydrocarbons in nearshore sediments, river sediments, and coastal peat of the Alaskan Beaufort Sea: implications for detecting anthropogenic hydro-carbon inputs. Marine Environmental Research 33, 223±253. Telang, S.A., Pocklington, R., Naidu, A.S., Romankevich,
E.A., Gitelson, I.I., Gladyshev, M.I., 1991. Carbon and mineral transport in major American, Russian, Arctic and Siberian Rivers: the Yukon, the Arctic Alaskan Rivers, the Arctic Basin Rivers in the Soviet Union, and the Yenisey. In: Degens, E.T., Kempe, S., Richey, J.E. (Eds.), Biogeochem-istry of Major World Rivers. John Wiley and Sons, New York, pp. 75±104.
Vedernikov, V.I., Demidov, A.B., Sud'bin, A.I., 1994. Primary production and chlorophyll in the Kara Sea in September 1993. Oceanology (in Russian) 34, 703.
Vedernikov, V.I., Demidov, A.B., Sud'bin, A.I., 1995. Primary production and chlorophyll in the Kara Sea in September 1993. Oceanology 34, 630±640.
Venkatesan, M.I., Kaplan, I.R., 1982. Distribution and trans-port of hydrocarbons in surface sediments of the Alaskan Outer Continental Shelf. Geochimica et Cosmochimica Acta 46, 2135±2149.
Vinogradov, M.E., Shushkina, E.A., Lebedeva, L.P., Gagarin, V.I., 1994. Mesoplankton in the eastern part of the Kara Sea and Ob and Yenisei Rivers estuaries. Oceanology 34, 723, (in Russian).
Vinogradov, M.E., Shushkina, E.A., Lebedeva, L.P., Gagarin, V.I., 1995. Mesoplankton in the east Kara Sea and the Ob and Yenisey River estuaries. Oceanology 34, 653±660. Wong, W.W., Sackett, W.M., 1978. Fractionation of stable
carbon isotopes by marine phytoplankton. Geochimica et Cosmochimica Acta 42, 1809±1815.
Yunker, M.B., Macdonald, R.W., 1995. Composition and ori-gins of polycyclic aromatic hydrocarbons in the Mackenzie River and on the Beaufort Sea shelf. Arctic 48, 118±129. Yunker, M.B., Macdonald, R.W., Fowler, B.R., Cretney, W.J.,
Dallimore, S.R., McLaughlin, F.A., 1991. Geochemistry and ¯uxes of hydrocarbons to the Beaufort Sea shelf: a multi-variate comparison of ¯uvial inputs and coastal erosion of peat using principal component analysis. Geochimica et Cosmochimica Acta 55, 255±273.
Yunker, M.B., Macdonald, R.W., Cretney, W.J., Fowler, B.R., McLaughlin, F.A., 1993. Alkane, terpene, and polycyclic aromatic hydrocarbon geochemistry of the Mackenzie River and Mackenzie shelf: riverine contributions to Beaufort Sea coastal sediment. Geochimica et Cosmochimica Acta 57, 3041±3061.