stable isotopes in benthic foraminiferal calcite from a...

Stable isotopes in benthic foraminiferal calcite from a river-influenced

Arctic marine environment, Kara and Pechora Seas

Leonid PolyakByrd Polar Research Center, Ohio State University, Columbus, Ohio, USA

Vladimir StanovoyArctic and Antarctic Research Institute, St. Petersburg, Russia

David J. LubinskiInstitute of Arctic and Alpine Research, University of Colorado, Boulder, Colorado, USA

Received 18 December 2001; revised 21 May 2002; accepted 28 June 2002; published 22 January 2003.

[1] Oxygen and carbon stable-isotope compositions of modern benthic foraminifera in the Kara and Pechoraestuarine regions of the Arctic continental shelf were compared with water d18O, d13CDIC, temperature, andsalinity. The foraminiferal d18O distribution is mostly similar to that of equilibrium calcite, primarily controlledby mixing of runoff and seawater in near-estuarine areas (depth < 15–20 m), with additional temperature controlseaward. The d13C composition is controlled by remineralization of organic matter, mostly imported by rivers,and by water mixing. Most analyses used the common Arctic foraminifer Elphidium excavatum forma clavata,which yields depleted d13C, as expected for an infaunal species. Isotope records for sediment cores off thePechora and Yenisey rivers show a decelerating rise in isotope values, of up to 10% in d18O, over the last 10 kyrin a pattern similar to that of rising sea level and concomitant retreat of river mouths, with local complexitiesprobably resulting from glacio-isostatic rebound and changes in sedimentation. INDEX TERMS: 4207

Oceanography: General: Arctic and Antarctic oceanography; 4219 Oceanography: General: Continental shelf processes; 4804

Oceanography: Biological and Chemical: Benthic processes/benthos; 4870 Oceanography: Biological and Chemical: Stable isotopes;

KEYWORDS: stable isotopes, benthic foraminifers, Arctic seas, near-estuarine environment

Citation: Polyak, L., V. Stanovoy, and D. J. Lubinski, Stable isotopes in benthic foraminiferal calcite from a river-influenced

Arctic marine environment, Kara and Pechora Seas, Paleoceanography, 18(1), 1003, doi:10.1029/2001PA000752, 2003.

1. Introduction

[2] The Kara Sea and the adjacent southeasternmost partof the Barents Sea (Pechora Sea) are characterized by auniquely high riverine input, which constitutes over 1.5 �103 km3 yr�1, more than a third of the total continentalrunoff into the Arctic Ocean (Figure 1) [Gordeev et al.,1996; R-ArcticNET, 2001]. This discharge has a profoundinfluence on hydrography, sedimentation, and biology onthe continental shelves adjacent to the river mouths. It alsocontrols the upper water mass structure in the entire ArcticOcean and, thus, the extent of Arctic sea ice [Aagard andCarmack, 1994; Stein, 1998; Forman et al., 2000]. Fur-thermore, low-salinity surface waters exported from theArctic Ocean affect overturning in the Nordic and LabradorSeas and, thus, the formation of North Atlantic Deep Water,which ventilates the World Ocean [e.g., Aagard and Car-mack, 1994]. During glacial periods, the Arctic hydro-graphic environments expanded southwards and dominatedeven larger areas than at present. Knowledge of the evolu-tion of the Arctic hydrology is therefore essential forunderstanding changes in the oceanic circulation and cli-

mate, an especially important topic in view of the potentialincrease in high-latitude precipitation with atmosphericwarming [Cattle and Crossley, 1995; Delworth et al.,1997].[3] Our study seeks to improve the reconstruction of

paleohydrographic environments in river-proximal areasusing oxygen and carbon stable isotope compositions(18O/16O and 13C/12C) recorded in biogenic calcite [e.g.,Anderson and Arthur, 1983; Erlenkeuser et al., 1999;Volkmann and Mensch, 2001]. Biogenic calcite d18O iscontrolled by the stable-isotopic composition of ambientwater, which is formed by mixing of waters from varioussources, and temperature during calcification [e.g., Shackle-ton, 1974]. Biogenic calcite d13C is primarily a function ofdissolved inorganic carbon (DIC), which is affected by watermixing on the continental shelves [Anderson and Arthur,1983; Erlenkeuser et al., 1995, 1999]. Additionally, benthiccalcite d13C is modified by the release of isotopically lightCO2 during organic matter decomposition [e.g., Grossman,1984]. By evaluating the modern distributions of calcitestable isotopes and hydrography, we can improve interpre-tations of past isotopic changes and, thus, better reconstructthe evolution of Arctic hydrographic environments.[4] Biogenic calcite in sediments from Arctic continental

shelves is mostly composed of benthic foraminifers [e.g.,

PALEOCEANOGRAPHY, VOL. 18, NO. 1, 1003, doi:10.1029/2001PA000752, 2003

Copyright 2003 by the American Geophysical Union.0883-8305/03/2001PA000752$12.00

3 - 1

Korsun et al., 1994; Wollenburg and Mackensen, 1998;Polyak et al., 2002a], which thus constitute the majorsource for stable-isotopic records in these environments.However, there have been few studies of d18O and d13C inmodern high-latitude shelf benthic foraminifers [Poole,1988; Vilks and Deonarine, 1988], and just one, verylimited data set from a river-proximal area [Erlenkeuserand von Grafenstein, 1999]. Our study elucidates theenvironmental controls on the stable-isotopic compositionof modern benthic foraminiferal tests from the Kara andPechora seas by comparing foraminiferal isotopic valueswith those of bottom waters. These results are then used tointerpret two Holocene (approximately 10 kyr) stable-iso-tope records from shallow Pechora and Kara Sea areas thatexperienced a large change in riverine inputs during sealevel rise after the Last Glacial Maximum [Polyak et al.,2000, 2002b].[5] A mostly infaunal foraminifer, Elphidium excavatum

forma clavata, was chosen for this study because it iscommon in a wide variety of Arctic shelf environments,both modern and Quaternary [e.g., Hald et al., 1994].Infaunal foraminifers adequately record bottom waterd18O, but tend to deviate from bottom water d13C, which

makes epifaunal species a preferable material for stable-isotope measurements [e.g., Grossman, 1984; McCorkle etal., 1990]. However, epibenthic species are rare in river-proximal environments [e.g., Murray, 1991; Polyak et al.,2002a] and would not provide a representative material forour study. For comparison, a mostly epibenthic foraminifer,Cibicides lobatulus, was analyzed where it co-occurs withE.e. clavata.

2. Oceanographic Settings

2.1. Hydrography

[6] Runoff from the Ob’, Yenisey, and Pechora rivers isapproximately 450, 580, and 130 km3 yr�1, respectively,and is mainly limited to summer, with a peak dischargeoccurring in June, after the snowmelt and ice break-up[Gordeev et al., 1996; R-ArcticNET, 2001]. The offshorespreading of riverine waters varies depending on the dis-charge and the prevailing winds and typically results insurface salinities below 30 psu over most of the Kara andPechora Sea area, with the active mixing of runoff reachingdepths of 15–20 m (Figure 2) [Burenkov and Vasil’kov,1995; Pavlov et al., 1996; Adrov and Denisenko, 1996].

Figure 1. Map of the Kara and Pechora seas showing 50 and 200 m water depth contours and samplelocation. Filled circles show surficial sediment samples: R/V Akademik Karpinskiy 1991 (four-digitlabels without lettering), R/V Boris Petrov 1997 (two-digit labels without lettering), and four coretops (L0987, DM4401, GF163, and GF134) (Table 2). Diamonds are sediment cores, B-212/218and DM-4401, used for the Holocene stable-isotope study (Table 3). Water samples shown weremeasured for d18O, as well as temperature and salinity (see text). See Brezgunov et al. [1983] formore water samples in the SE Barents Sea outside the map limits. Insert shows circum-Arctic rivernetwork of drainage, with line thickness of rivers representing relative runoff [R-ArcticNET, 2001](base map courtesy of R. Lammers).

3 - 2 POLYAK ET AL.: STABLE ISOTOPES IN BENTHIC FORAMINIFERAL CALCITE

Bottom salinities are less affected by riverine water, otherthan in shallow river-proximal areas, and are more stablethan at the surface (Figure 3a). The overall hydrographicpattern is characterized by a steady, decelerating increase ofsalinities with distance from the estuaries. The temperaturedistribution pattern has weaker gradients and is less regularthan that of respective salinities, but the warmest summertemperatures are consistent with low salinities, occurring inshallow, river-proximal areas (Figure 3B and Table 1).

Elevated temperatures are also characteristic of intermediatedepths of 200+ m in the deep troughs affected by theadvection of Atlantic-derived water (Figure 2) [Pfirman etal., 1994; Pavlov et al., 1996]. The Barents Sea is generallywarmer than the Kara Sea due to the Atlantic influence.

2.2. Biological Productivity

[7] As elsewhere in high-latitude seas, the productiveseason in the study area is limited to the hydrographic

Figure 2. Vertical profiles of salinity, temperature, and water d18O based on samples measured for d18O,without surface data (see Figure 1 for location). Upper panels show the upper 100 m expanded.

Figure 3. Multiyear means of (a) summer bottom salinity and (b) temperature in the Kara and Pechoraseas, with superimposed foraminiferal data points (cf. Figure 1). Shading highlights salinities >34 psu andtemperatures <�1�C. The fields in the Pechora Sea are detailed using the climatological monthly maps ofAdrov and Denisenko [1996].

POLYAK ET AL.: STABLE ISOTOPES IN BENTHIC FORAMINIFERAL CALCITE 3 - 3

summer, typically from May to September, and is triggeredby Spring sea-ice melt [Usachev, 1968; Druzhkov et al.,2001]. In the Pechora Sea, the ice starts melting in May,whereas in the Kara Sea this process is commonly delayeduntil June and may span the whole summer [Borodachev,1998]. Productivity throughout the Siberian seas is tightlycoupled with maximum river discharge, which occurs inJune–July and provides a significant source of food for theshelf biota by delivering nutrients and labile organic matter[Vedernikov et al., 1995; Nothig and Kattner, 1999].Accordingly, seasonal primary production is generally highin and near the estuaries, reaching >300 mg C m�2 day�1,and decreases to very low levels of <50 mg C m�2 day�1 inthe open sea. Somewhat enhanced productivity has alsobeen observed in the western Kara Sea near the coasts ofNovaya Zemlya, possibly in relation to glacial meltwater[Matishov, 1989; Druzhkov et al., 2001].[8] Data constraining the flux of organic carbon to the

seafloor are scarce, but the flux pattern can be approximatedby the benthic production. The total macrobenthic biomassin the Kara Sea has highest values of >150 g m�2 near theestuaries and decreases northward, consistent with thegeneral distribution pattern of primary productivity [Zenke-vich, 1963; Matishov, 1989]. Extremely low benthic bio-mass of <3 g m�2 characterizes the deep Novaya ZemlyaTrough in the western Kara Sea, whereas the shallow areaadjacent to Novaya Zemlya has somewhat elevated valuesof >25 g m�2. In the Pechora Sea, benthic productivity isgenerally high and reaches >500 g m�2 due to a longerphytoplankton vegetation season at lower latitudes and tothe proximity of the Polar Front in the Barents Sea [Zen-kevich, 1963; Denisenko et al., 1997].

2.3. Stable Isotopes

[9] The d18O composition in the runoff of the Pechora,Ob’, and Yenisey rivers is estimated as approximately �14,�16, and �17% versus SMOW, respectively [Brezgunov etal., 1983]. These results are based on summer to early fallmeasurements and closely represent the annual means forrespective precipitation d18O [Rozanski et al., 1993; Wolfe etal., 2000], because summer discharge constitutes over 85%of annual Arctic runoff [Gordeev et al., 1996]. These valuesfit into the broader longitudinal trend of riverine d18Odecrease across northern Eurasia from �13% in Scandina-via to �24% in the far east [Letolle et al., 1993], which isconsistent with the prevailing direction of atmospheric

vapor transport east of the North Atlantic. The mixing ofrunoff and marine, Atlantic-derived water (d18O composi-tion of near 0%) primarily controls the overall d18Odistribution in the Arctic seas [Ostlund and Hut, 1984;Schlosser et al., 2000]. This distribution may be locallymodified by the processes of sea-ice freezing and melting,which cause the rejection of isotopically light brines and therelease of d18O-enriched meltwater, respectively [e.g.,Strain and Tan, 1993].[10] Similar to the seawater d18O composition, DIC d13C

in the Siberian seas depends on the mixing of riverine andmarine waters characterized by DIC d13C values of approx-imately �7 and 1%, respectively [Erlenkeuser et al., 1995,1999]. Remineralization of organic carbon may additionallydecrease bottom water DIC d13C in and near the riverestuaries due to the oxidation of organic matter that ismostly exported from rivers [Erlenkeuser et al., 1999].

3. Materials and Methods

3.1. Foraminiferal Samples

[11] Most sediment samples used in this study werecollected by grabs from R/V Akademik Karpinskiy (AK)and by box cores from R/V Akademik Boris Petrov (BP) inAugust–September of 1991 and 1997, respectively, alongwith measurements of bottom hydrographic properties[Figure 1 and Table 2; Ivanov, 1995; Matthiessen andStepanets, 1998]. The top 2 cm of sediment was fixed byethanol and stained by Rose Bengal to determine live, or atleast protoplasm-containing faunal specimens. Stained for-aminifers, mostly of 125–250-mm diameter, were pickedfrom 36 samples, seven of which contained C. lobatulus inaddition to E.e. clavata. Four unstained, 2-cm-thick coretop samples were added for a denser geographic coveragein the Kara Sea from sites that have measured modernsedimentation rates of near 1 mm/yr [Polyak et al., 2002b],which indicate that core top samples generally representmodern environments.[12] To evaluate the use of modern foraminiferal stable-

isotope data for the study of paleohydrology of SiberianArctic, E.e. clavata tests were picked from down coresediment in borehole B-212/218 and core DM-4401 fromoutside the estuaries of the Pechora and Yenisey rivers,respectively [Figures 1 and 4, and Table 3; Polyak et al.,2000, 2002b]. The present influence of riverine waters onboth core sites has a similar magnitude, given similarsummer bottom salinities of 32–32.5 psu. Due to highsedimentation rates in these areas, sample spacing of 10–50cm in DM-4401, and mostly 50–100 cm in B-212/218provides a submillenial age resolution, little affected bybioturbation. Foraminifers were picked from the same sizerange as in surface samples, mostly 125–250 mm. As E.e.clavata was absent at some intervals in DM-4401, anotherforaminifer, Haynesina orbiculare, was also picked fromthis core and an adjacent surficial sample (eight samplescontained both species for a comparison of their stable-isotopic signatures).[13] Foraminiferal d18O and d13C compositions were

simultaneously measured in all analyzed samples, surfaceand down core, on a Finnigan MAT-252 mass-spectrometer

Table 1. Linear Correlation Coefficients Between d18O Values

(Water and Equilibrium Calcite), Temperatures, and Salinities, by

Data Subsetsa

SubsetsSurfacePechora

SurfaceKara

Subsurfaceand Bottom All

S/T �0.80 �0.56 �0.35 �0.51d18Ow/S 0.99 1.00 0.83 0.99d18Oe.c./d

18Ow 1.00 1.00 0.87 0.99d18Oe.c./S 0.98 0.99 0.81 0.98d18Oe.c./T �0.88 �0.63 �0.65 �0.62No. of samples 80 47 84 211

aThe Pechora subset includes adjacent data from the southeastern BarentsSea.


Table

2.Values

ofd1

8O

andd1

3C(versusPDB)Measuredin

Benthic

Foraminiferal

TestsFrom

SurficialSam

plesandRelated

Bottom

Water

Dataa

Sam

ple

Latitude

N,

deg

Longitude

E,

deg

Water

Depth,

m

Salinity

Mean,

psu

sSalinity

Mean,

psu

b

Tem

perature

Mean,

�C

sTem

perature

Mean,

�Cb

Salinity

Point,

psu

Tem

perature

point,

�C

E.e.clavata

d18O,%

E.e.clavata

d13C,%

C.lobatulus

d18O,%

C.lobatulus

d13C,%

AK-1905

68.75

50.23

57

33.4

2.8

2.8

3.6

33.4

1.2

2.87

�3.07

...

...

AK-1907

69.35

50.24

37

33.6

2.5

2.0

3.1

33.5

3.5

2.67

�1.91

...

...

AK-2006

69.63

60.12

15

34.3

0.7

1.1

1.5

28.0

4.2

0.48

�2.25

...

...

AK-2102

69.80

61.68

27

34.1

0.7

0.5

2.7

33.3

�1.3

2.50

�1.74

...

...

AK-2104

69.42

66.25

29

31.4

1.1

3.0

3.8

32.8

�0.3

2.05

�2.23

...

...

AK-2202

70.29

66.17

20

32.0

0.7

1.9

0.6

32.5

�0.8

2.19

�0.87

...

...

AK-2207

74.02

67.84

57

33.7

1.1

�1.1

0.7

33.7

�1.2

3.46

�2.22

3.13

0.86

AK-2209

73.11

72.69

25

25.2

6.7

�0.2

1.4

31.5

�1.6

1.87

�3.46

...

...

AK-2401

72.42

79.68

14

33.4

2.2

�1.4

0.9

<2

7.5

�1.03

�4.69

...

...

AK-2504

73.63

80.25

40

31.3

3.5

�1.1

1.0

30.3

�0.8

2.43

�3.20

...

...

AK-2507

74.34

81.39

43

32.4

2.3

�1.2

0.8

23.3

�1.5

0.18

�4.83

1.07

�0.68

AK-2601

75.15

77.22

44

33.1

1.5

�1.5

0.8

33.3

�1.5

2.98

�2.04

...

...

AK-2603

76.01

75.34

75

34.0

0.9

�1.5

0.8

33.7

�1.7

3.08

�2.75

...

...

AK-2701

76.41

69.94

70

34.7

0.3

�1.3

0.8

34.3

�0.9

2.16

�2.66

...

...

AK-2703

75.28

64.87

330

34.7

0.5

�1.4

0.7

34.3

�1.4

3.50

�1.88

3.10

1.44

AK-2704

75.29

62.11

72

34.9

0.4

�1.5

0.8

34.4

�1.4

3.06

�2.27

...

...

AK-2708

73.44

57.53

60

34.9

0.2

�1.6

0.8

33.9

�1.2

3.64

�2.77

...

...

AK-2709

73.06

56.72

75

34.7

0.3

�1.5

0.6

34.3

�1.3

3.02

�2.94

...

...

AK-2712

72.61

59.55

98

34.7

0.1

�1.5

0.4

34.0

�1.2

3.49

�1.91

2.39

1.17

AK-2715

72.35

56.84

322

34.5

0.3

�1.5

0.6

...

...

3.60

�1.31

...

...

AK-2717

71.59

55.85

53

34.2

0.8

�1.2

0.9

34.1

�1.1

3.32

�1.71

...

...

AK-2719

71.69

59.07

165

34.6

0.2

�1.4

0.9

32.9

�1.5

4.11

�2.92

...

...

AK-2804

70.27

58.51

32

34.5

0.3

0.7

1.2

31.3

4.6

2.54

�2.79

...

...

AK-2806

70.45

55.37

73

33.8

3.0

1.7

5.1

...

...

3.30

�2.69

2.77

0.75

AK-2808

70.72

52.99

110

33.5

4.0

2.1

6.9

32.8

�1.0

3.71

�2.28

3.34

0.60

BP-19

74.00

79.03

30

31.7

2.2

�1.3

0.7

...

...

4.15

�1.52

...

...

BP-24

73.53

79.92

41

30.6

4.2

�1.0

1.1

31.8

�1.2

0.93

�3.82

...

...

BP-27

72.89

80.09

19

29.1

3.3

�0.4

1.1

30.6

�0.3

1.32

�4.74

...

...

BP-21

74.00

81.01

41

32.2

2.3

�1.2

0.7

32.6

�1.4

1.69

�4.95

...

...

BP-42

73.90

81.67

32

31.8

2.5

�1.0

0.8

31.9

�1.2

2.02

�2.89

...

...

BP-46

74.00

77.20

27

31.2

1.7

�1.2

0.7

31.6

�1.3

1.42

�2.99

...

...

BP-50

73.61

72.95

28

29.7

2.2

�1.0

0.7

31.8

�0.1

1.64

�2.47

...

...

BP-52

74.00

72.66

30

31.3

1.7

�1.4

0.6

31.8

�1.4

1.60

�2.74

...

...

BP-55

73.22

75.62

14

27.6

2.0

�0.6

0.6

9.7

5.2

�2.31

�4.72

...

...

BP-56

72.89

75.48

14

25.6

2.3

�0.2

0.7

15.5

3.5

�4.53

�5.68

...

...

VA-31

73.31

50.08

268

34.8

0.2

0.2

1.4

...

...

3.79

�1.79

3.71

0.27

R-9

72.22

66.38

126

34.0

0.5

�0.6

1.1

...

...

3.38

�2.27

...

...

GF-134

72.30

57.64

341

34.6

0.2

�1.5

0.6

...

...

3.78

�2.06

...

...

GF-163

77.00

70.00

542

35.0

�1.3

...

...

3.53

�1.70

...

...

DM-4401

74.00

79.95

35

32.5

2.2

�1.2

0.8

...

...

2.26

�3.17

...

...

aSee

Figure

1forlocations.

bStandarddeviationsfrom

temperature

andsalinitymultiannual

means.


at the Woods Hole Oceanographic Institution. Samplescontained two to eight tests of E.e. clavata and one to fourheavier tests of C. lobatulus and H. orbiculare. The overallprecisions of d18O and d13C measurements were at least 0.1and 0.06% versus PDB, respectively [Ostermann andCurry, 2000].

3.2. Hydrographic Data

[14] The life cycle of benthic foraminifers may span overa year, but their reproduction, growth, and calcificationpresumably occur mainly during the productive season[Murray, 1991; Korsun et al., 1994; Graf et al., 1995;Wollenburg and Mackensen, 1998]. Therefore, it is assumedthat the stable-isotopic composition of biogenic calcite inthe study area primarily reflects summer conditions [cf.Erlenkeuser et al., 1999]. Because the exact timing offoraminiferal calcite growth is not known and the short-

term variability of hydrographic conditions in river-prox-imal areas is large, we use mean water characteristics inaddition to point measurements for the evaluation of fora-miniferal stable-isotope composition.[15] To characterize mean summer hydrographic environ-

ments in the study area, we have compiled salinity andtemperature measurements, collected mostly during July toSeptember, for the last approximately 20 years from histor-ical databases [National Snow and Ice Data Center, 1997,1998; Matishov et al., 1998]. "Bottom water" samples weretypically collected several meters above the seafloor, butthey closely represent the true bottom water where theinfluence of marine water near the bottom is substantial,as indicated by salinities above ca. 20 psu [Simstich, 2001],which is the case for almost all our foraminiferal samples.Data were included only when coverage for a particular yearspanned a majority of the data-collection area. Scattered

Figure 4. Stratigraphy of sediment cores and age-depth curves compared to the global sea level afterFairbanks [1989]. For more detail, see Polyak et al. [2000, 2002b]. 14C ages with asterisks weremeasured on plant detritus; others were measured on foraminiferal or mollusc calcite and reservoir-corrected by �460 yr [after Stuiver and Braziunas, 1993]. B-212/218 is a composite of several closelylocated boreholes (referred to as B-210-218 in the work by Polyak et al. [2000]). Most samples arefrom B-212, with four youngest samples from B-218 (see Table 3). 14C ages used for estimating agesof stable-isotope samples are from respective sections of B-212 and B-218; two additional 14C agesfrom another well, B-215, are shown in parenthesis. Two age models, ‘‘young’’ and ‘‘old,’’ areshown for the bottom parts of the Holocene section in both B-212/218, based on alternative 14Cages, and DM-4401, based on a comparison of apparent age-depth curve with the global sea level.


data from each year were gridded and then all yearly gridswere averaged to form mean multiyear fields (Figure 3).The averaged grid point values were interpolated to theforaminiferal sampling sites (Table 2). The generated meanmultiannual site values are mostly similar to point measure-ments of temperature and salinity, except for some river-proximal samples that show a large departure from themeans, grossly exceeding their standard deviations. Theseoffsets clearly result from short-lived fluctuations in waterproperties, a common feature in river-proximal environ-ments [e.g., Burenkov and Vasil’kov, 1995; Pavlov et al.,1996].[16] Our foraminiferal samples mostly characterize a

salinity range above 30 psu, due to a decline in foramini-feral numbers in the estuaries, which may result fromdissolution and/or a physiological effect of reduced salinity[Korsun, 1999; Polyak et al., 2002a]. Five samples havelower salinities, reaching 25 psu in multiannual mean valuesand as low as <2 psu in point measurements. The lattervalues probably result from short-lived discharge episodesand may not represent the actual salinity of the ambientwater in sediment [cf. Simstich, 2001]. Because live

(stained) E.e. clavata have been reported from other riverenvironments at salinities as low as 10–15 psu [Lutze,1965; Ellison and Nichols, 1976], future studies in thePechora and Kara estuary regions should help define thelocal salinity limit by continuing to sample near the rivermouths.

3.3. Seawater D18O and Equilibrium Calcite D

18O

[17] Seawater d18O distribution in the study area is basedon samples from the southern Kara Sea and southeasternBarents Sea, including the Ob’ and Pechora estuaries,collected during the summer to early fall season (JulyOctober) in 1976 and 1977 [Figure 1; Ferronskii, 1978;Brezgunov et al., 1983]. This data set was supplemented bynew samples from four vertical water profiles collected inAugust 1998 near northernmost Novaya Zemlya from R/VIvan Petrov and analyzed for d18O at the Institute of Arcticand Alpine Research (Figure 1 and Table 4). To increase thestatistical representation for subsurface (below 1 m) andbottom water, we included the respective October data,which have distribution patterns similar to those of summer.October surface water data were not used because they

Table 3. Stable-Isotopic Values Measured in Foraminiferal Tests From Down Core Sedimentsa

Well/CoreCore Depth,

mAge Model,

kabE.e. clavatad13C, %

E.e. clavatad18O, %

E.e. clavata corr.d18O, %c

H. orbicul.d13C, %

H. orbicul.d18O, %

H. orbicul. corr.d18O, %c

B-212/218, 69�150N, 57�180 E, 20 ±2 m Water Depth218 1.2 1.62 �0.65 2.01 1.99 . . . . . . . . .218 3.4 4.60 �0.23 2.54 2.49 . . . . . . . . .218 3.7 4.79 �1.47 2.67 2.61 . . . . . . . . .218 4.0 4.98 �1.27 2.72 2.66 . . . . . . . . .212 4.6 5.36 �1.96 2.43 2.36 . . . . . . . . .212 5.3 5.80 �2.09 1.55 1.46 . . . . . . . . .212 6.3 6.21 �2.64 1.95 1.84 . . . . . . . . .212 7.0 6.46 �3.27 1.31 1.18 . . . . . . . . .212 9.4 7.04 �3.28 1.17 1.00 . . . . . . . . .212 10.4 7.28 �4.86 1.77 1.57 . . . . . . . . .212 10.9 7.40 �3.03 1.96 1.74 . . . . . . . . .212 12.0 7.67 �3.32 0.00 �0.24 . . . . . . . . .212 12.4 7.76 �4.00 1.00 0.75 . . . . . . . . .212 12.9 7.88 �3.38 0.43 0.17 . . . . . . . . .212 13.3 7.98 �3.74 0.52 0.25 . . . . . . . . .212 14.5 8.24 �3.74 0.93 0.60 . . . . . . . . .212 15.1 8.36 �6.08 �1.22 �1.59 . . . . . . . . .212 15.6 8.45 �4.60 �0.62 �1.03 . . . . . . . . .212 16.2 8.56 �4.66 �2.85 �3.31 . . . . . . . . .212 17.8 8.86 �5.65 �0.53 �1.12 . . . . . . . . .212 18.2 8.93 �3.54 0.92 0.31 . . . . . . . . .

DM-4401, 74�000N, 79�570 E, 35 m Water DepthBP-24 surface �3.82 0.93 n/a �1.66 2.53 n/a

DM-4401 0.03 0.11 �3.17 2.26 2.26 �0.98 2.02 2.02DM-4401 0.18 0.74 �2.97 1.94 1.93 �0.68 1.98 1.97DM-4401 0.33 1.37 �2.93 2.23 2.22 �0.97 2.59 2.58DM-4401 0.73 3.06 . . . . . . . . . �0.60 2.41 2.38DM-4401 1.17 5.25 . . . . . . . . . �0.92 1.88 1.82DM-4401 1.53 7.04 . . . . . . . . . �1.38 1.45 1.28DM-4401 1.74 7.30 �3.50 2.24 2.04 �1.97 1.57 1.37DM-4401 1.84 7.43 �3.82 1.87 1.65 �1.46 1.80 1.58DM-4401 3.43 9.33 �4.74 1.24 0.72 �1.87 1.47 0.95DM-4401 3.73 9.69 �5.62 �5.23 �5.83 �4.70 �5.74 �6.34DM-4401 4.03 10.05 . . . . . . . . . �5.33 �5.56 �6.21DM-4401 4.08 10.12 �7.37 �6.96 �7.62 . . . . . . . . .

aSee Figure 4 for stratigraphy and age control.b‘‘Young’’ age models used below 14 m in B-212/218 and below 2 m in DM-4401 (see Figure 4).cd18O values corrected for global sea level by �0.011% per meter rise [Fairbanks and Matthews, 1978; Fairbanks, 1989].


show a temperature inversion due to the fall cooling[Brezgunov et al., 1983; Adrov and Denisenko, 1996].[18] The water d18O and temperature values (measured on

the same samples) were used to calculate equilibrium calcite(e.c.) d18O by the equation

d18O e:c: PDBð Þ ¼ d18O water SMOWð Þ � 0:27þ Temperature �Cð Þ � 16:9ð Þ=� 4:0ð Þ

rearranged from Shackleton [1974] with a conversion to thePDB scale by Hut [1987]. Shackleton’s equation may notideally reflect equilibrium calcite precipitation for all envi-ronments [e.g., Bemis et al., 1998], but its use makes iteasier to compare results with many previous estimatesbased on this equation [e.g., Poole, 1988; Erlenkeuserand von Grafenstein, 1999; Lubinski et al., 2001] or a verysimilar one by O’Neil et al. [1969] [Vilks and Deonarine,1988; McCorkle et al., 1990]. The more recent temperatureequation for oxygen-isotope calcite fractionation of Kimand O’Neil [1997] would yield d18O values lower by ca.0.5%.

4. Results

4.1. Modern D18O

[19] The distribution of measured water d18O valuesshows a strong relationship with salinity [Figure 5 andTable 1; cf. Brezgunov et al., 1983]. This is not surprisingas the study area is characterized by a large salinity gradientformed by mixing of riverine and marine waters withdistinctly different d18O compositions. Values of d18O inthe Pechora estuary are 2% heavier than those in the Ob’

estuary for the same salinities, which is consistent with awest–east gradient in d18O composition of meteoric waters.Away from the estuaries, at salinities above 20, this offsetdiminishes and d18O/salinity distribution shows a similar

Table 4. Values of d18O (Versus SMOW) Measured on Water Samples From the Northern Kara Sea (R/V Ivan Petrov 1998) at the Stable-

Isotope Laboratory of the Institute of Arctic and Alpine Researcha

StationLatitude N,

degLongitude E,

degWater Depth,

mSample Depth,

mTemperature,

�CSalinity,psu

d18O,%

IP-2 76.99 70.10 557 1 1.68 30.70 �1.132IP-2 76.99 70.10 557 60 �1.34 34.48 �0.195IP-2 76.99 70.10 557 120 �0.09 34.70 0.027IP-2 76.99 70.10 557 170 0.73 34.84 0.129IP-2 76.99 70.10 557 240 �0.38 34.78 0.100IP-2 76.99 70.10 557 300 �0.91 34.82 0.116IP-2 76.99 70.10 557 550 �1.38 34.95 0.094IP-3 77.75 67.96 470 1 0.65 27.80 �1.754IP-3 77.75 67.96 470 50 �1.38 34.56 �0.016IP-3 77.75 67.96 470 100 �0.35 34.69 0.008IP-3 77.75 67.96 470 125 0.11 34.74 0.037IP-3 77.75 67.96 470 200 �1.21 34.77 0.121IP-3 77.75 67.96 470 265 �0.82 34.87 0.147IP-3 77.75 67.96 470 350 �1.07 34.90 0.093IP-3 77.75 67.96 470 450 �1.36 34.93 0.089IP-4 76.98 70.07 550 1 1.98 31.17 �0.853IP-4 76.98 70.07 550 25 �0.84 33.85 �0.238IP-4 76.98 70.07 550 60 �1.49 34.42 �0.167IP-4 76.98 70.07 550 140 �0.45 34.76 0.012IP-4 76.98 70.07 550 250 �1.13 34.80 0.016IP-4 76.98 70.07 550 270 �0.78 34.44 0.066IP-4 76.98 70.07 550 400 �1.15 34.87 0.099IP-4 76.98 70.07 550 550 - 32.35 0.127IP-18 76.04 70.02 347 1 5.43 9.36 �12.395IP-18 76.04 70.02 347 330 �1.47 34.71 0.001

aOverall laboratory measurement precision is �0.1% and the reported precision for this data set is <0.05%.

Figure 5. Measured water d18O versus simultaneouslymeasured salinity, with regression lines for the Pechora andKara Sea samples (from Brezgunov et al. [1983], withadditions shown in Table 4).


mixing. Some scatter, resulting in offsets not exceeding±1%, indicates additional processes, such as ice freezing/melting [cf. Brezgunov et al., 1983]. The depth distributionof d18O shows a large gradient down to approximately 15–20 m, consistent with the active mixing of runoff, and amuch slower change at larger depths (Figure 2). A slightd18O depletion characterizes the depths of 250–350 m in theNovaya Zemlya and St. Anna troughs, which is probablyrelated to the descent of isotopically light brines producedby ice freezing.[20] The calculated equilibrium calcite d18O largely inher-

its a close relationship with salinity from the water d18O,with a slope of 0.46 up to the d18Oe.c. values of approx-imately 0%, which correspond to salinities below 30 psu(Figure 6A and Table 1). The slope steepens to 0.7 athigher salinities, reflecting a decrease in water mixinggradient away from river mouths. The threshold salinityof 30 psu, with corresponding d18Oe.c. of 0%, is attainedin surface water far from the estuaries, whereas in bottomwater it occurs much closer, at depths of 20 m or slightlyless (Figures 2, 3, and 7). This results in generally highsalinities and respective d18Oe.c. values of almost all ana-lyzed bottom water samples, which is most relevant forunderstanding the benthic foraminiferal d18O controls. Up tohalf of the 5% change in d18Oe.c. in these samples isattributed to a temperature control given a correspondingtemperature range reaching 10�C in the Pechora Sea(Figure 6D). A smaller gradient of 5�C in the Kara Seasuggests less temperature influence here.[21] Measured foraminiferal calcite d18O values show

similar relationships with salinity and temperature as thoseobserved for d18Oe.c., although interpretation is moredifficult given sparse foraminiferal samples at salinitiesbelow 30 psu and large differences between mean andpoint temperature and salinity values for the shallowestnear-estuarine sites (Figure 8). Both the Kara and Pechorasamples show a relationship with salinity. This relation-ship, however, is stronger in the Kara samples. Conversely,the Pechora samples show a stronger relationship withtemperature.[22] To help evaluate the measured versus expected

foraminiferal (E.e. clavata) calcite d18O values, we calcu-lated d18Oe.c. for foraminiferal sample sites using mean andpoint temperatures and salinities (Figure 9 and Table 2).Water d18O values were estimated from linear regressions ofd18O against salinity (Figure 5). Both mean and pointhydrographic data provide a good fit for d18Oe.c./d

18Ofor.

distribution at d18Ofor. values above 0%, with disequili-bria of �0.65 and �0.61% based on mean and point data,respectively. At d18Ofor. values below 0%, characterizingthe most shallow (<15 m) near-estuarine samples, meansalinity data show a much larger offset of �3 to �5%(Figure 9A). In contrast, corresponding d18Oe.c. valuesderived from point measurements are lighter than d18Ofor.,including two extremely deviating data points (Figure 9B).Measured C. lobatulus d18O values are lighter than those inE.e. clavata by 0.3% in average (Table 2), indicating aslightly larger disequilibrium. Values closer to the equili-brium are inferred for H.orbiculare based on a +0.6% offsetfrom E.e. clavata d18O in core DM-4401 (Table 2).

[23] The basic correspondence between measured andexpected foraminiferal d18O composition is also reflectedin the spatial distribution of E.e. clavata d18O values thatrise consistently with increasing distance from river estua-ries and with depth (Figures 7 and 10). Similar to the waterd18O pattern (Figure 2), there is a threshold at 15–20 mbelow which the gradient of foraminiferal and equilibriumcalcite d18O decreases drastically. Down to 100 m, thisgradient is still larger than that in the water d18O. Somedepletion in the calcite d18O is also evident at larger depths,notably between 250 and 350 m in the Novaya Zemlya andSt. Anna troughs.

4.2. Modern D13C

[24] The overall distribution of measured foraminiferald13C shows a relationship with salinity (Figures 10 and 11),as expected given increasing DIC d13C with the transitionfrom river to marine waters in front of the Ob’ and Yeniseyestuaries [Erlenkeuser et al., 1999]. However, the foramini-feral d13C:salinity gradient, demonstrated most fully by E.e.clavata, is much steeper than that of the DIC mixing models(Figure 11). E.e. clavata d13C values at high salinities aremostly 1–2.5% below the equilibrium DIC values in theKara Sea mixing model, whereas the low-salinity d13Cvalues are up to 5% lighter than the DIC. This accountsfor approximately 0.15–0.2% additional d13C change persalinity unit due to the oxidation of organic matter.[25] The d13C values in E.e. clavata are between 2 and 4

% lighter than in the epibenthic foraminifer C. lobatulusfrom the same samples (Table 2; Figure 11); a similar offsethas been reported on several samples from the Barents Sea(T. Dokken, personal communication). Most C. lobatulusd13C values in our data are closer to the ‘‘global’’ than to thelocal (Kara Sea) DIC mixing line (Figure 11). This isprobably related to the location of C. lobatulus sites farfrom the river mouths and under some influence of watersfrom the Barents Sea; especially noteworthy is the sole C.lobatulus sample near the Yenisey estuary which fits thelocal DIC mixing model.

4.3. Holocene Stable-Isotope Records

[26] Holocene sediments in B-212/218 overlay the ero-sional unconformity corresponding to the low sea level ofthe last glacial maximum (LGM) [Figure 4; Polyak et al.,2000]. Sand grades upward to finer-grained mud and thenback to sand, reflecting changes in sedimentary regime withchanging sea level and distance from the Pechora River.Core DM-4401 did not reach the erosional surface of theLGM regression, but its lowermost sedimentary unit con-tains abundant terrestrial plant detritus indicating proximityto the Yenisey river [Figure 4; Polyak et al., 2002b]. The14C ages obtained on these cores provide a generally robustchronostratigraphy, with some uncertainty for the lower-most part of the records. A date of 10.2 ka from the bottomof DM-4401 was obtained on plant detritus that may bepartially older than the enclosing sediment. By comparingthe DM-4001 age-depth patterns to that of the global sealevel, we expect the actual age of the core base to be notolder than 9 ka (Figure 4). In B-212/218, an age inversion


occurs near the base of the Holocene, which requires twoage models for this interval. A comparison with the globalsea level curve may not help to constrain these models,because the B-212/218 site is located close to the LGM ice

sheet limit and was possibly affected by glacioisostaticrebound in the early Holocene [Gataullin et al., 2001]. InFigure 12 we use a younger model with a more gradual age-depth curve.

Figure 6. (a, c) Equilibrium calcite d18O versus salinity and (b, d) equilibrium calcite d18O versustemperature. d18Oe.c. values have been calculated from measured water d18O and temperature using theequation of Shackleton [1974]. Best fit in Figure 6a is shown for salinities below 30 psu. Lowerpanels (Figures 6c and 6d) show subsurface and bottom samples only (axis ranges are indicated byboxes in Figures 6a and b).

Figure 7. Equilibrium calcite and foraminiferal d18O versus water depth (without surface d18Oe.c. data).Right panel is a blow-up for depths above 100 m.


[27] Both d18O and d13C measured in foraminifers downcore B-212/218 and DM-4401 show a dramatic changestarting in the early Holocene (Figure 12). The major trendof this change is a general decelerating increase in stable-isotope values until ca. 5 14C kyr BP (ka), totaling as muchas 10% in d18O in DM-4401. In B-212/218, this pattern iscomplicated by a preceding period of noticable d18Odecrease until ca. 9 or 8.5 ka, depending on the age model.In the late Holocene, after ca. 5 ka, d18O values remainedstable. Changes in d13C in DM-4401 agree with those ind18O, showing an increase of almost 5% until ca. 5 ka and a

subsequent stabilization, with a slight decrease during thelast 2–3 ka. In contrast, d13C values in B-212/218 show (1)a decrease in the early Holocene broadly corresponding tothat in d18O, but with a more complex structure and (2) anadditional sharp increase of 3% between ca. 6.5 and 4.5 ka.

5. Discussion

5.1. Modern D18O

[28] The pattern of foraminiferal and equilibrium calcited18O in the Kara and Pechora seas agrees with that estab-lished on mollusc calcite d18O in the near-estuarine area ofthe Ob’ and Yenisey rivers [Erlenkeuser et al., 1999]. Thesedata are also consistent with those obtained on mixedcalcareous faunas from the Laptev Sea [Erlenkeuser andvon Grafenstein, 1999] and on C. lobatulus from theLabrador shelf [Vilks and Deonarine, 1988], both areasstrongly affected by freshwater discharge. In the proximityof the shallow river mouths, where the mixing gradientbetween riverine and marine waters is very large, changes incalcite d18O are predominantly controlled by the water d18Ocomposition and are, thus, closely related to salinity (Fig-ures 6 and 10). When the salinities of 30 psu are attained,which typically corresponds to bottom water depth of 15–20 m, the mixing gradient decreases so that temperaturestarts playing a significant role in controlling the calcited18O (Figures 6 and 7). In the Pechora Sea, the warmest partof the study area, the temperature control is especiallystrong and accounts for approximately half of the d18Oe.c.

variability (Figures 6 and 8). The temperature controlexplains a somewhat larger gradient in calcite d18O thanin water d18O composition at depths below 20 m, notablydown to 100 m (cf. Figure 2). Some depletion in calcited18O at yet larger depths in the Novaya Zemlya and St.Anna troughs presumably results from a combined action ofbrines and the advection of relatively warm Atlantic-derivedwater [Pfirman et al., 1994; Pavlov et al., 1996].[29] The distribution of foraminiferal d18O at values

above 0%, when corrected for apparent disequilibrium,mostly matches the distribution of the expected equilibrium

Figure 8. (a, c) Foraminiferal d18O versus bottom salinityand (b, d) foraminiferal d18O versus temperature usingmultiyear summer means (Figures 8a and 8b) or pointmeasurements (Figures 8c and 8d). Note that severalsamples lack point data (Table 2).

Figure 9. Measured E.e. clavata d18O versus d18Oe.c. calculated from water d18O:salinity regressions

(Figure 5) and temperature using (a) multiyear summer means or (b) point measurements. Notethat several samples lack point data (Table 2).


calcite d18O in the study area (Figures 7, 9, and 10).Estimated disequilibria, between �0.6 and �0.65% forE.e. clavata and 0.3% larger for C. lobatulus, are similarto those suggested in previous studies. McCorkle et al.[1990] demonstrated that E.excavatum d18O is 1% lighterthan the equilibrium in the deep North Atlantic. A near�1% disequilibrium was also suggested for unspecifiedElphidium spp. from the Laptev Sea [Erlenkeuser and vonGrafenstein, 1999] and for C. lobatulus from the south-western Barents Sea and the Labrador shelf [Poole, 1988;Vilks and Deonarine, 1988]. We note that the use of sometemperature/d18O equations other than Shackleton’s wouldresult in expected d18Oe.c. values of 0.5–1% lower [e.g.,Kim and O’Neil, 1997; Bemis et al., 1998] and, thus,smaller foraminiferal disequilibria. However, this will notaffect the prediction of paleohydrography from foramini-feral d18O where the d18Oe.c. model is calculated using thesame equation (cf. Figure 6).[30] At d18Ofor. values below 0%, their relationship

with equilibrium calcite d18O does not show a straightfor-ward pattern. The use of point salinity data here producessuspiciously large offsets toward lighter d18Oe.c. values(Figure 9b), which likely result from local, short-livedfluctuations in riverine inputs and may not adequatelyrepresent the ambient water in sediment [cf. Simstich,2001]. The use of mean hydrographic data results in adifferent pattern, with E.e. clavata d18O offset from the 1:1line by �3 to �5% (Figure 9a). This apparently systematicoffset may result from a lag between the seasons ofcalcification and hydrographic data collection, which islikely to be more noticeable in the shallow near-estuarineareas because of their highly variable hydrography. Analternative explanation implies a dramatic increase in bio-logical fractionation due to high metabolic activity that maybe expected during short periods suitable for life in theseextreme environments [cf. Polyak et al., 2002a]. Moredetailed studies of calcite-producing organisms in Arctic

river-proximal sites are needed to address this problem.Until a reliable calibration set is established, low foramini-feral d18O values should be considered with caution whenreconstructing paleohydrography.

5.2. Modern D13C

[31] The lightest values of foraminiferal d13C are observedat the Ob’ and Yenisey estuary mouths (Figure 10b),

Figure 10. Geographic distributions of (a) d18O and (b) d13C measured in E.e. clavata tests. Thindotted and solid lines show 50 and 200 m water depth contours. Thick dashed lines in Figure 10aare isolines of equilibrium calcite d18O (versus PDB) calculated from d18O (versus SMOW) andtemperature measured in bottom water (see Figure 1 for sample location).

Figure 11. Foraminiferal d13C plotted against summersalinity. Solid and dotted curves show ‘‘global’’ (Norwe-gian-Greenland Sea) and local (Kara Sea) marine-riverinewater mixing models for the DIC d13C, respectively[Erlenkeuser et al., 1999].


reflecting the combined effect of riverine DIC and organicremineralization. The latter clearly prevails given the E.e.clavata d13C:salinity gradient three to four times larger thanthat of the equilibrium DIC d13C (Figure 11). A similardistribution of E.e. clavata d13C has been observed in theLong Island Sound with respect to both absolute values andthe d13C gradient controlled by remineralization [Thomas etal., 2000]. It is not surprising that remineralization is mostintense in near-estuarine areas because of extensive deposi-tion of river-imported organic detritus and heightened pro-ductivity. Although riverine inputs of organic materialpredominate in our study area, the authigenic marine com-ponent may contribute up to a third of the total organicmatter in sediment [Boucsein et al., 1999].[32] The light foraminiferal d13C values extend northeast-

ward from estuary mouths, presumably indicating the posi-tion of mixing zones between riverine and marine waters. Agroup of relatively heavy values of both d13C and d18Oimmediately north of the Yenisey estuary (Figure 10)coincides with high bottom water salinities that consistentlyoccur at this site and presumably indicate a subsurfacecountercurrent toward the estuary [cf. Stein and Stepanets,2000]. In addition to river-proximal areas, relatively lightd13C values characterize a narrow zone along the coast ofNovaya Zemlya (Figure 10b), in line with elevated produc-tivity in this area in comparison to the adjacent westernKara Sea [Matishov, 1989; Druzhkov et al., 2001].[33] Assuming an infaunal habitat for E.e. clavata, we

expect its d13C to be lighter than the bottom water DIC dueto the accumulation of d13C-depleted CO2 in pore waters[e.g., McCorkle et al., 1990]. Indeed, E.e. clavata in ourdata shows a systematic d13C offset of around �3% fromthe mostly epifaunal species C. lobatulus, which appears tohave d13C values close to the equilibrium, in agreement withearlier estimates [Poole, 1988] and observations on otherspecies of Cibicides or closely related genera [e.g., McCor-kle et al., 1990]. The offset of �2 to �4% between E.e.

clavata and C. lobatulus d13C values is consistent with thetypical disequilibrium range of �1 to �4% that character-izes various infaunal foraminifers [Grossman, 1984;McCorkle et al., 1990]. A comparison with H. orbiculareanalyzed in the Holocene record DM-4401 provides addi-tional, although indirect information on the habitat-con-trolled foraminiferal d13C composition; this information isespecially helpful because C. lobatulus characterizes onlyhigh-salinity environments. The d13C offset between E.e.clavata and H. orbiculare consistently falls between �1 and�3%, confirming the mostly infaunal habitat for E.e.clavata and suggesting a shallow infaunal to epifaunalhabitat for H. orbiculare, a species indicative of river-proximal environments [e.g., Korsun, 1999; Polyak et al.,2002b].

5.3. Holocene Records

[34] After the waning of the LGM ice sheets, shallowcontinental shelves were dramatically affected by the post-glacial sea level rise that came close to the present level byca. 5 ka [Fairbanks, 1989; Figure 4]. Local deglaciationprocesses, such as the release of meltwater masses andisostatic adjustment of the seafloor and adjacent coasts, alsohad a profound effect on sedimentary and hydrographicenvironments in the Barents and Kara seas [e.g., Hald andVorren, 1987; Lubinski et al., 2001; Gataullin et al., 2001].The investigated sites, B-212/218 and DM-4401, are situatedoutside the reconstructed LGM ice sheet limits and werelikely beyond the range of deglacial meltwater impact at thebeginning of the Holocene when little glacier ice remained[Polyak et al., 2000, 2002b]. Glacioisostatic rebound mayhave been significant in the southern Pechora Sea due to itsproximity to the LGM margin, but was outpaced by theglobal transgression sometime by the early Holocene[Gataullin et al., 2001]; the near-estuarine area in the south-ern Kara Sea was probably further from the ice sheet marginand was not affected noticeably by the rebound.

Figure 12. Foraminiferal stable-isotopic records in Holocene sediments of cores B-212-218 and DM-4401, plotted against 14C age using the ‘‘young’’ age models (Figure 4). H. orbiculare stable-isotopicdata are shown for DM-4401 intervals lacking E.e. clavata. H. orbiculare d18O and d13C values havebeen adjusted by �0.6 and �2.0%, respectively, for the offsets from E.e. clavata values (Table 3).The d18O values have been corrected for the global sea level using an estimate of �0.011% per meter rise[Fairbanks and Matthews, 1978; Fairbanks, 1989].


[35] With sea level rise, shorelines on the low-gradientshallow continental shelves, such as the southern Kara Sea,rapidly migrated landwards, accompanied by retreatingrivers [cf. Bauch et al., 2001]. This change clearly had aprofound effect on calcite d18O composition that is predom-inantly controlled by runoff inputs in the Kara and Pechoraseas at water depths above 15–20 m (Figures 6 and 7).Given the age-depth profiles in sediment cores (Figure 4),we assume that this water-depth level was not attained untilapproximately 7 ka in B-212/218 or slightly earlier in DM-4401. Therefore, we infer that the steep rise by 8.5% inforaminiferal d18O (sea level corrected) prior to ca. 8 ka inDM-4401 (Figure 12A) primarily reflects a dramatic declinein runoff inputs to the core site, associated with the retreat ofthe Yenisey mouth. Such a change, if due to temperaturealone would equate to a impossible increase of >30�C. Ourinterpretation is supported by a corresponding steep increasein d13C by 3% (Figure 12B) and by other paleoceano-graphic proxies, notably a high content of plant detritus andfreshwater diatoms in sediment with low foraminiferalstable-isotopic values [Figure 4; Polyak et al., 2002b]. Werecognize, however, that our current data set with a singlerecord from each estuary limits our ability to discriminatebetween the large influence of sea level rise on runoff inputto a given location and the potential influence of com-pounding factors such as variations in total river dischargeand shifts in the position of the river channel [e.g.,Sidorchuk et al., 2000]. A comprehensive reconstructionof this history requires investigating multiple sedimentaryrecords from the shelf areas affected by the riverine retreat.[36] In contrast to the early Holocene stable-isotopic trend

in DM-4401, d18O values in B-212/218 show a considerabledecrease of 3.5% prior to ca. 9 or 8.5 ka, depending on theage model, with a subsequent fast rise to values similar tothose in DM-4401. The respective d13C values in B-212/218also show a dip, although with a more complex pattern. Thisdeviation in B-212/218 stable-isotope record may reflect theeffect of glacioisostatic rebound in the southern PechoraSea, extended into the early Holocene. Gataullin et al.[2001] suggested that the rebound was outpaced by thesea level rise between 11 and 10 ka, but their age estimate istentative and could be easily off by ca. 1 kyr.[37] Despite a clearly predominant effect of diminishing

runoff inputs on both salinity and equilibrium calcite d18Ocomposition on the shallow shelf of the Kara and Pechoraseas in the early Holocene, estimating paleosalinities fromour foraminiferal d18O records is not straightforward. Thistask is especially complicated by an insufficient understand-ing of the d18Oe.c./d

18Ofor. relationship at low d18Ofor. values,below 0% in E.e. clavata (Figure 9). If we extrapolate thedisequilibrium between �0.6 and �0.65%, characteristic ofhigher d18Ofor. values, to the entire observed d18Ofor. range,then the estimate of paleosalinity change based on the earlyHolocene 8.5% d18O increase in DM-4401 (Figure 12)would exceed 15 units, starting with values of15 psu. Thiscalculation uses the d18Oe.c.:salinity slope of 0.46 associatedwith d18Oe.c. values below0% (Figure 6A), consistent withwater depths at the core site of less than 15–20 m prior toca. 8 ka (Figure 4). If we assume, however, that the apparentoffset of approximately �4% between d18Ofor. and d18Oe.c.

at d18Ofor. values below 0% represents a true disequili-brium (Figure 9B), then the estimated paleosalinity changewould be more modest, with initial values about 23 psu.Until the pattern of foraminiferal d18O associated with lowsalinities in the most river-proximal environment is clari-fied, we propose to use the above paleosalinity estimates asapproximate end-members.[38] Among other possibilities, these estimates may need

to be adjusted for a change in the distribution of water d18Oin the study area [Rohling and Bigg, 1998]. Based on arelative stability of the North Atlantic circulation after theend of the last deglaciation [e.g., Waelbroeck et al., 2001],we expect that the marine end-member of the d18O mixingline was generally stable during the Holocene, back to ca.10 ka. Meteoric d18O, however, could be more variable,with early Holocene values possibly higher than at presentby 1–2% in the lower Yenisey River region, due to anincreased flux of Atlantic air masses [Wolfe et al., 2000].This would decrease the d18O:salinity gradient resulting in alowering of estimated paleosalinity values by up to 3 units.[39] Interestingly, the lowest reported E.e. clavata d18O

values of �8.0 to �8.5%, corresponding to estimatedsalinities between 10 and 16 psu (Yoldia Sea) [Schoninget al., 2001], are similar to values at the base of DM-4401and may indicate the actual salinity limit for E.e. clavata(we note that our point measurement of <2 psu at samplingsite AK-2401 is suspiciously low and may not represent theaverage ambient salinity in sediment). A limit of 10–15 psuappears to be consistent with occurrences of live (stained)E.e. clavata reported from various localities, such as theBaltic Sea and the Chesapeake Bay [Lutze, 1965; Ellisonand Nichols, 1976].[40] After ca. 8 ka, foraminiferal d18O in both cores is

characterized by a modest overall increase of 1.5–2% andeven smaller superimposed fluctuations until ca. 5 ka andstable values thereafter (although, some short-term fluctua-tions may be overlooked due to sparser age resolution after5 ka). These, relatively minor changes may be attributed tothe combined effect of variable temperature and water d18Ocomposition given the positive d18O values and core depthsgreater than 15–20 m (cf. Figures 4 and 6). Because sealevel continued to rise noticeably until ca. 5 ka, we infer thatit was a more important control on calcite d18O thancontemporaneous temperature changes.[41] Foraminiferal d13C values after 8 ka also show an

overall modest increase until ca. 3 ka in DM-4401 and until6.5 ka in B-212/218, generally in line with a deceleratingriver-mouth retreat. The steep rise in E.e. clavata d13C in B-212/218 after ca. 6.5 ka is strikingly different from the trendin DM-4401, as well as from the d18O pattern in B-212/218.This sudden change can be better understood if we considerthat the respective lithology in B-212/218 switched fromrelatively fine-grained mud to sand (Figure 4), probably dueto the sea level driven retreat of estuarine mud depocenters[cf. Levitan et al., 2001], in contrast to DM-4401 which isstill located within the mud accumulation area off the Yeniseyestuary. The elevated permeability of sandy substrate pre-vents the accumulation of fine organic detritus and thus cutsdown carbon remineralization rates, which should haveraised calcite d13C values in B-212/218. The 2.5% value


of this rise is consistent with an observed gradient in fora-miniferal d13C controlled by remineralization (Figure 11).The persistence of E.e. clavata under a changed environ-ment illustrates an exceptional ecological adaptability ofthis foraminifer, which may help to explain its widedistribution on the high-latitude shelves. A slight decreasein d13C during the last 2–3 kyr coincides with an elevatedcontent of freshwater diatoms in DM-4401 [Polyak et al.,2002b], suggesting some increase in remineralization ratesdue to higher riverine organic inputs.

6. Conclusions

[42] The composition of equilibrium calcite d18O in theriver-proximal areas of the Kara and Pechora seas, notably atwater depths above 15–20 m, is shown to be predominantlycontrolled by the water d18O values due to a strong gradientin the mixing of riverine and marine waters. Accordingly,d18Oe.c. here is closely correlated with salinity. At largerdepths, d18Oe.c. is controlled by the combination of waterd18O composition and temperature, with the temperaturecontrol being especially significant in the Pechora Sea. Thedistribution of benthic foraminiferal (E.e. clavata and C.lobatulus) calcite d18O in the study area is mostly consistentwith that of d18Oe.c., with an average disequilibrium of �0.6to �0.65% for E.e. clavata and by 0.3% larger for C.lobatulus. Actual disequilibrium values, however, may belower, depending on the choice of the d18Oe.c.:temperaturemodel. In river-proximal samples from the shallowest waterdepths, the d18Oe.c./d

18Ofor. relationship is less clear and mayindicate a disequilibrium for E.e. clavata as high as �3 to�5%. Further investigation is needed for insight in thecontrols on foraminiferal d18O in this environment.[43] Foraminiferal calcite d13C shows a relationship with

salinity due to a combined effect of remineralization oforganic matter, which is strongest in sediments proximal toriver estuaries, and the mixing of the DIC in riverine andmarine waters. Remineralization is believed to be primarilyfueled by organic matter exported by rivers, but maypartially result from high seasonal productivity in theestuaries.[44] The benthic foraminiferal species analyzed are sim-

ilar in their d18O composition, but have systematic, largeoffsets in d13C values. E. excavatum forma clavata, the mostwidespread calcareous foraminifer on the Arctic shelves,has d13C around 3% lower than the epibenthic species C.lobatulus, which corroborates the mostly infaunal habitat ofE.e. clavata. The d13C content in H. orbiculare, a character-istic river-proximal foraminifer, is about 2% heavier than

E.e. clavata, suggesting a shallow infaunal to epifaunalhabitat for H. orbiculare.[45] Down core foraminiferal stable-isotope records off

the estuaries of the Pechora and Yenisey Rivers provide ahistory of hydrographic change during the Holocene. Theprevailing trend of d18O and d13C is a decelerating increasein their values, amounting to 10% in d18O, until ca. 5 ka.This trend generally matches the postglacial sea level risepattern and is interpreted to primarily reflect a retreat ofriver mouths from the core sites. The accompanying salinityincrease may have been as large as 15+ units, but thisestimate is only tentative given the paucity of modernbenthic foraminiferal samples at salinities below 30 psu,uncertainties with foraminiferal d18O disequilibrium in themost river-proximal environment, and potential changes inmeteoric water d18O. Deviating from the general trend in theearly Holocene, both d18O and d13C in the core from thePechora Sea decrease until ca. 8.5 ka. This pattern mayreflect glacioisostatic rebound of the seafloor because thecore site is located close to the LGM ice sheet limit. Anotherdeviation from the prevailing trend is an abrupt rise in d13Cvalues in the Pechora core after ca. 6.5 ka. We infer that thisrise, corresponding to a coarsening of sediment, indicates areduction in organic accumulation and remineralization.[46] A more detailed interpretation of the past environ-

ments requires the assembly of more comprehensive data onforaminiferal stable-isotopic compositions in a wide hydro-graphic range. This may be achieved by a comparison of thestable-isotopic signatures of species inhabiting river-proxi-mal areas, such as H. orbiculare and Elphidium incertum[e.g., Polyak et al., 2002a]. It is essential that foraminiferalcollection be accompanied with direct measurements ofstable-isotopic composition of ambient water [cf. McCorkleet al., 1990]. Better understanding of foraminiferal life cyclesand living strategies is needed to define the calcificationenvironments, whichwould require amultiseasonal samplingprogram. To improve reconstructions of paleohydrographyfrom stable-isotope records, more efforts should also con-centrate on comparing themwith independent proxies such asbiogeochemical markers and microfossil transfer functions.

[47] Acknowledgments. This research was supported by USANational Science Foundation awards OPP-9725418 and OPP-9818247.Foraminiferal samples were collected and processed by S. Korsun (R/VBoris Petrov) and O. Kijko (R/V Akademik Karpinskiy). D. Ostermanndirected stable-isotope analyses of foraminifers. L. Febo and A. Sharapovaassisted in picking foraminifers. We appreciate comments by S. Korsun andL. Febo. Formal reviews by E. Thomas, C. Hillaire-Marcel, N. Nørgaard-Pedersen, and an anonymous reviewer were very helpful for improving thepaper, especially the very detailed comments by the first two reviewers.This is Byrd Polar Research Center publication 1274.

ReferencesAagard, K., and E. C. Carmack, The ArcticOcean and climate: a perspective, in The Po-lar Oceans and Their Role in Shaping theGlobal Environment, edited by O. M. Johann-nessen et al., pp. 5–20, AGU, Washington,D. C., 1994.

Adrov, N. M., and S. G. Denisenko, Oceano-graphic description of the Pechora Sea, in Bio-geocoenoses of West Arctic Glacial Shelves (inRussian), edited by G. G. Matishov et al., pp.166–179, Russ. Acad. of Sci., Apatity, 1996.

Anderson, T. F., and M. A. Arthur, Stable iso-topes of oxygen and carbon and their applica-tion to sedimentologic and paleoenvironmentalproblems, in Stable Isotopes in SedimentaryGeology, SEPM Short Course 10, edited byM. A. Arthur et al., pp. 1–151, 1983.

Bauch, H. A., H. Kassens, O. D. Naidina, M.Kunz-Pirrung, and J. Thiede, Compositionand flux of Holocene sediments on the easternLaptev Sea shelf, Arctic Siberia, Quat. Res.,55, 344–351, 2001.

Bemis, B. E., H. J. Spero, J. Bijma, and D. W.Lea, Reevaluation of oxygen isotopic composi-tion of planktonic foraminifera: experimentalresults and revised paleotemperature equations,Paleoceanography, 13, 150–160, 1998.

Borodachev, V. Ye., L’dy Karskogo morya (Ice inthe Kara Sea) (in Russian), 182 pp., Gidrome-teoizdat, St. Petersburg, 1998.

Boucsein, B., K. Fahl, M. Siebold, and R. Stein,Quantity and quality of organic carbon in sur-face sediments of the Ob and Yenisey estuaries


and adjacent coastal areas: marine productivityvs. terrigenous input, Ber. Polarforsch., 300,116–126, 1999.

Brezgunov, V. S., V. Debolskii, V. V. Nechaev,V. I. Ferronskii, and T. V. Yakimova, Character-istics of the formation of the oxygen isotopiccomposition and salinity upon mixing of seaand river waters in the Barents and Kara seas,Water Resour. (USSR), 9, 335–344, 1983.

Burenkov, V. I., and A. P. Vasil’kov, The influenceof runoff from land on the distribution of hydro-logic characteristics of the Kara Sea (Engl.transl.), Oceanology (Russia), 34, 591–599,1995.

Cattle, H., and J. Crossley, Modeling Arctic cli-mate change, Philos. Trans. R. Soc. London,Ser. A, 352, 201–213, 1995.

Delworth, T. L. S., S. Manabe, and R. J. Stouffer,Multidecadal climate variability in the Green-land Sea and surrounding regions: a coupledsimulation, Geophys. Res. Lett., 24, 257–260,1997.

Denisenko, S. G., N. V. Denisenko, Y. A. Frolo-va, N. A. Anisimova Sandler, and S. Dale,Current state of bottom fauna and structureof bottom communities in the Pechora Sea,in Natural Conditions of the Kara and BarentsSeas, Proc. Russian-Norwegian Workshop 95,Norsk Polarinstitutt Rapportserie, 97, editedby V. Volkov et al., pp. 390–394, 1997.

Druzhkov, N. V., P. R. Makarevich, and E. I.Druzhkova, Phytoplankton in the south-wes-tern Kara Sea: composition amd distribution,Polar Res., 20, 95–108, 2001.

Ellison, R. L., and M. M. Nichols, Modern andHolocene foraminifera in the Chesapeake Bayregion, in First International Symposium onBenthonic Foraminifera of Continental Mar-gins, Part A, Ecology and Biology, MaritimeSediments Sp. Publ., 1, edited by C. T. Schaferand B. R. Pelletier, pp. 131–151, 1976.

Erlenkeuser, H., Stable carbon isotope ratios inthe waters of the Laptev Sea/Sept. 94, Ber.Polarforsch., 176, 170–177, 1995.

Erlenkeuser, H., and U. von Grafenstein, Stableoxygen isotope ratios in benthic carbonateshells of Ostracoda, Foraminifera, and Bival-via from surface sediments of the Laptev Sea,summer 1993 and 1994, in Land-Ocean Sys-tems in the Siberian Arctic: Dynamics and His-tory, edited by H. Kessens et al., pp. 503–514,Springer-Verlag, New York, 1999.

Erlenkeuser, H., R. Spielhagen, and E. Taldenko-va, Stable isotopes in modern water and bi-valve samples from the southern Kara Sea,Ber. Polarforsch., 300, 80–90, 1999.

Fairbanks, R. G., A 17,000 year glacio-eustaticsea level record: influence of glacial meltingrates on the Younger Dryas event and deep-ocean circulation,Nature, 342, 637–642, 1989.

Fairbanks, R. G., and R. K. Matthews, The mar-ine oxygen isotope record in Pleistocene coral,Barbados, West Indies, Quat. Res., 10, 181–196, 1978.

Ferronskii, V. I., Investigation of the distributionof oxygen-18 and tritium in waters of theBarents and Kara seas for the prognosis ofhydrologic changes in the estuarine areas ofthe Ob’ River and Kara and Pechora seas (inRussian), Tech. Rep., 101 pp., Inst. WaterStud., USSR Acad. Sci., Moscow, 1978.

Forman, S. L., W. Maslowski, J. T. Andrews, D.Lubinski, M. Steele, and J. Zhang, Researchersexplore Arctic freshwater’s role in ocean cir-culation, Eos. Trans. AGU, 81, 169 – 174,2000.

Gataullin, V., J. Mangerud, and J. I. Svendsen,The extent of the Late Weichselian ice sheet in

the southeastern Barents Sea, Global Planet.Change, 31, 451–472, 2001.

Gordeev, V. V., J. M. Martin, I. S. Sidorov, andM. V. Sidorova, A reassessment of the Eura-sian river input of water, sediment, major ele-ments, and nutrients to the Arctic Ocean, Am.J. Sci., 296, 664–691, 1996.

Graf, G., S. A. Gerlach, P. Linke, W. Queisser,W. Ritzrau, A. Scheltz, L. Thomsen, and U.Witte, Benthic-pelagic coupling in the Green-land-Norwegian Sea and its effect on the geo-logical record, Geol. Rundsch., 84, 49–58,1995.

Grossman, E. L., Stable isotope fractionation inlive benthic foraminifera from the southernCalifornia borderland, Palaeogeogr. Palaeocli-matol. Palaeoecol., 47, 301–327, 1984.

Hald, M., P. I. Steinsund, T. Dokken, S. Korsun,L. Polyak, and R. Aspeli, Recent and LateQuaternary distribution of Elphidium excava-tum f. clavata in Arctic seas, Cushman Found.Spec. Publ., 32, 141–153, 1994.

Hald, M., and T. O. Vorren, Stable isotope strati-graphy and paleoceanography during the lastdeglaciation on the continental shelf offTroms, northern Norway, Paleoceanography,2, 583–599, 1987.

Hut, G., Consultants’ group meeting on stableisotopic reference samples for geochemicaland hydrological investigations, Rep. to theDirector General, 42 pp., Int. Atomic EnergyAgency, Vienna, 1987.

Ivanov, G. I., Studying the geochemical variabil-ity on the western Arctic shelf for the assess-ment of technogenic inpact on the environment(in Russian), Tech. Rep., Research InstituteOkeangeologia, St. Petersburg, 1995.

Kim, S. T., and J. R. O’Neil, Equilibrium andnon-equilibrium oxygen isotope effects in syn-thetic carbonates, Geochem. Cosmochem.Acta, 61, 3461–3475, 1997.

Korsun, S. A., Benthic foraminifera in the Obestuary, west Siberia, Ber. Polarforsch., 300,59–70, 1999.

Korsun, S. A., I. A. Pogodina, G. A. Tarasov, andG. G. Matishov, Foraminifers of the BarentsSea (Hydrobiology and Quaternary paleoecol-ogy) (in Russian), 138 pp., Russ. Acad. Sci.,Apatity, 1994.

Letolle, R., J. M. Martin, A. J. Thomas, V. V.Gordeev, S. Gusarova, and I. S. Sidorov, 18Oabundance and dissolved silicate in the Lenadelta and Laptev Sea (Russia),Mar. Chem., 43,47–64, 1993.

Levitan, M. A., V. M. Kuptsov, E. A. Romanke-vich, and A. V. Kondratenko, Some indicationfor late Quaternary Pechora River discharge:Results of vibrocore studies in the southeasternPechora Sea, Int. J. Earth Sci., 89, 533–540,2001.

Lubinski, D. A., L. Polyak, and S. Forman, De-ciphering the latest Pleistocene and Holoceneinflows of freshwater and Atlantic water to thedeep northern Barents and Kara Seas: Forami-nifera and stable isotopes, Quat. Sci. Rev., 20,1851–1879, 2001.

Lutze, G.-F., Zur foraminiferen-fauna der Ostsee,Meyniana, 15, 75–142, 1965.

Matishov, G. G., (Ed.), Ecology and BiologicalResources of the Kara Sea (in Russian), 183pp., USSR Acad. Sci., Apatity, 1989.

Matishov, G., A. Zuyev, V. Golubev, N. Adrov,V. Slobodin, S. Levitus, and I. Smolyar, Cli-matic Atlas of the Barents Sea 1998: Tem-perature, Salinity, Oxygen [CD-ROM],Murmansk Mar. Biol. Inst. (Russia) and Natl.Oceanogr. Data Cent., NOAA, Silver Spring,Md., 1998.

Matthiessen, J. and O. Stepanets (Eds.), Scienti-fic cruise report of the Kara Sea expedition ofRV ‘‘Akademik Boris Petrov’’ in 1997, Ber.Polarforschung, 266, 102 pp., 1998.

McCorkle, D. C., L. D. Keigwin, B. H. Corliss,and S. R. Emerson, The influence of microha-bitats on the carbon isotopic composition ofdeep-sea benthic foraminifera, Paleoceanogra-phy, 5, 161–185, 1990.

Murray, J. W., Ecology and Palaeoecology ofBenthic Foraminifera, 397 pp., Addison-Wes-ley-Longman, Reading, Mass., 1991.

National Snow and Ice Data Center, Joint Amer-ican-Russian Atlas of the Arctic Ocean, Winter[CD-ROM], Environmental Working Group,Boulder, Colo., 1997.

National Snow and Ice Data Center, Joint Amer-ican-Russian Atlas of the Arctic Ocean, Sum-mer [CD-ROM], Environmental WorkingGroup, Boulder, Colo., 1998.

Nothig, E.-M., and G. Kattner, Distribution ofphytoplankton biomass and nutrient concentra-tions, Ber. Polarforsch., 300, 37–44, 1999.

O’Neil, J. R., R. N. Clayton, and T. K. Mayeda,Oxygen isotope fractionation in divalent metalcarbonates, J. Chem. Phys., 51, 5547–5558,1969.

Ostermann, D. R., and W. B. Curry, Calibrationof stable isotopic data: an enriched d18O stan-dard used for source gas-mixing detection andcorrection, Paleoceanography, 15, 353–360,2000.

Ostlund, H. G., and G. Hut, Arctic Ocean watermass balance from isotope data, J. Geophys.Res., 89, 6373–6381, 1984.

Pavlov, V. K., L. A. Timokhov, G. A. Baskakov,M. Yu. Kulakov, V. K. Kurazov, P. V. Pavlov,S. V. Pivovarov, and V. V. Stanovoy, Hydro-meteorological regime of the Kara, Laptev, andEast-Siberian seas, Tech. Memo., Appl. Phys.Lab., Univ. of Wash., Bellingham, Wash., 179pp., 1996.

Pfirman, S. L., D. Bauch, and T. Gammelsrød,The northern Barents Sea: Water mass distri-bution and modification, in The Polar Oceansand Their Role in Shaping the Global Envir-onment, edited by O. M. johannessen et al., pp.77–94, AGU, Washington, D. C., 1994.

Polyak, L., V. Gataullin, O. Okuneva, and V.Stelle, New constraints on the limits of theBarents-Kara ice sheet during the last glacialmaximum based on borehole stratigraphy fromthe Pechora Sea, Geology, 28, 611–614, 2000.

Polyak, L., S. Korsun, L. A. Febo, V. Stanovoy,T. Khusid, M. Hald, B. E. Paulsen, and D. A.Lubinski, Benthic foraminiferal assemblagesfrom the southern Kara Sea, a river-influencedarctic marine environment, J. ForaminiferalRes., 32, 252–273, 2002a.

Polyak, L., M. Levitan, T. Khusid, L. Merklin,and V. Mukhina, Variations in the influence ofriverine discharge on the Kara Sea during thelast deglaciation and the Holocene, GlobalPlanet. Change, 32, 291–309, 2002b.

Poole, D. A. R., Stable isotope fractionation inrecent benthic foraminifera from the southwes-tern Barents Sea, Cand. Sci. M.S. thesis, 128pp., Univ. Tromsø, Tromsø, Norway, 1988.

R-ArcticNET, A Regional Hydrographic DataNetwork for the Pan-Arctic Region, v2.1[CD-ROM], Water Sys.s Anal. Group, Univ.of N. H., Durham, N. H., 2001.

Rohling, E. J., and G. R. Bigg, Paleosalinity andd18O: A critical assessment, J. Geophys. Res.,103, 1307–1318, 1998.

Rozanski, K., L. Araguas-Araguas, and R. Gon-fiantini, Isotopic patterns in global precipita-tion, in Continental Isotopic Indicators of


Climate, Geophys. Monogr, 78, edited by P. K.Swart et al., pp. 1 –36, AGU, Washington,D. C., 1993.

Schlosser, P., B. Ekwurzel, S. Khatiwala, B. New-ton, W. Maslowski, and S. Pfirman, Tracer stu-dies of the Arctic freshwater budget, in TheFreshwater Budget of the Arctic Ocean, NATOSci. Ser. 2, 70, edited by L. L. Edward et al., pp.453–478, Dordrecht, Boston, Mass., 2000.

Schoning, K., F. Klingberg, and S. Wastegard,Marine conditions in central Sweden duringthe early Preboreal as inferred from a stableoxygen isotope gradient, J. Quat. Sci., 16,785–794, 2001.

Shackleton, N. J., Attainment of isotopic equili-brium between ocean water and the benthonicforaminifera genus Uvigerina: Isotopicchanges in the ocean during the last glacial,in Colloque International sur les MethodesQuantitatives d’Etude des Variation du Climatau Sours du Pleistocene, Coll. Int. C.N.R.S.,219, 203–209, 1974.

Sidorchuk, A. Y., A. V. Panin, O. K. Borisova, S.A. Elias, and J. P. Syvitski, Channel morphol-ogy and river flow in the northern Russianplain in the late glacial and Holocene, Int. J.Earth Sci., 89, 541–549, 2000.

Simstich, J., Is water from the Rosette Sampler asuitable proxy for bottom water? Comparisonof salinity in Rosette and Multi Corer, Ber.Polarforsch., 393, 18–20, 2001.

Stein, R., (Ed.), Arctic Paleo-River Discharge(APARD): A new research programme of theArctic Science Board (AOSB), Ber. Polar-forsch, 279, 128 pp., 1998.

Stein, R., and O. Stepanets (Eds.), Scientificcruise report of the joint Russian-GermanKara-Sea expedition of RV ‘‘Akademik BorisPetrov’’ in 1999, Ber. Polarforsch., 360, 139pp., 2000.

Strain, P. M., and F. C. Tan, Seasonal evolutionof oxygen isotope-salinity relationships inhigh-latitude surface waters, J. Geophys.Res., 98, 14,589–14,598, 1993.

Stuiver, M., and T. Braziunas, Modelling atmo-spheric 14C influences and 14C ages of marinesamples to 10,000 BC, Radiocarbon, 35, 137–189, 1993.

Thomas, E., T. Gapotchenko, J. C. Varekamp,E. L. Mecray, and M. R. Buchholtz ten Brink,Benthic foraminifera and environmentalchanges in Long Island Sound, J. CoastalRes., 16, 641–655, 2000.

Usachev, P. I., Phytoplankton of the Kara Sea, inPlankton Tikhogo Okeana (Plankton of the Pa-cific Ocean) (in Russian), edited by G. I. Syo-mina, pp. 6–28, Nauka, Moscow, 1968.

Vedernikov, V. I., A. B. Demidov, and A. I.Sudbin, Primary production and chlorophyllin the Kara Sea in September 1993 (Engl.transl.), Oceanology (Russia), 34, 693–703,1995.

Vilks, G., and B. Deonarine, Labrador shelfbenthic Foraminifera and stable oxygen iso-topes of Cibicides lobatulus related to the Lab-rador Current, Can. J. Earth Sci., 25, 1240–1255, 1988.

Volkmann, R., and M. Mensch, Stable isotopecomposition (d18O, d13C) of living plankticforaminifers in the outer Laptev Sea and the

Fram Strait, Mar. Micropaleontol., 42, 163–188, 2001.

Waelbroeck, C., J.-C. Duplessy, E. Michel, L.Labeyrie, D. Paillard, and J. Duprat, Thetiming of the last deglaciation in NorthAtlantic climate records, Nature, 412, 724–727, 2001.

Wolfe, B. W., T. W. D. Edwards, R. Aravena, S.L. Forman, B. G. Wagner, A. A. Velichko, andG. M. MacDonald, Holocene paleohydrologyand paleoclimate at treeline, north-central Rus-sia, inferred from oxygen isotope records inlake sediment cellulose, Quat. Res., 53, 319–329, 2000.

Wollenburg, J. E., and A. Mackensen, Livingbenthic foraminifers from the central ArcticOcean: faunal composition, standing stockand diversity, Mar. Micropaleontol., 34,153–185, 1998.

Zenkevich, L. A., Biology of the Seas of theU.S.S.R., 955 pp., Allen and Unwin, Concord,Mass. [c/o Paul & Co.], 1963.

��D. Lubinski, Institute of Arctic and Alpine

Research (INSTAAR), University of Colorado,C.B. 450, Boulder, CO 80309-0450, USA.([email protected])L. Polyak, Byrd Polar Research Center, Ohio

State University, 1090 Carmack Road, Columbus,OH 43210, USA. ([email protected])V. Stanovoy, Arctic and Antarctic Research

Institute (AARI), St. Petersburg 199397, Russia.([email protected])


stable isotopes in benthic foraminiferal calcite from a...

Documents