The co-evolution of Black Sea level and composition ...intranet.geoecomar.ro/rchm/wp-content/uploads/... · The co-evolution of Black Sea level and composition through the last deglaciation

ARTICLE IN PRESS

0277-3791/$ - se

doi:10.1016/j.qu

�CorrespondE-mail addr

Quaternary Science Reviews 25 (2006) 2031–2047

The co-evolution of Black Sea level and composition through the lastdeglaciation and its paleoclimatic significance

Candace O. Majora,�, Steven L. Goldsteinb, William B.F. Ryanb, Gilles Lericolaisc,Alexander M. Piotrowskid, Irka Hajdase

aDepartment of Geology and Geophysics, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USAbLamont-Doherty Earth Observatory and Department of Earth and Environmental Sciences, Columbia University, Palisades, NY 10964, USA

cIFREMER—Centre de Brest, DRO/GM—BP 70, F-29280 Plouzane cedex, FrancedDepartment of Earth Sciences, University of Cambridge, Downing Street, Cambridge CB2 3EQ, UKeInstitut fur Teilchenphysik (IPP), HPK H 27, ETH-Hoenggerberg, CH-8093 Zurich, Switzerland

Received 7 October 2005; accepted 28 January 2006

Abstract

The Black Sea was an inland lake during the last ice age and its sediments are an excellent potential source of information on Eurasian

climate change, showing linkages between regionally and globally recognized millennial-scale climate events of the last deglaciation.

Here, we detail changes from the last glacial maximum (LGM) through the transition to an anoxic marginal sea using isotopic (strontium

and oxygen) and trace element (Sr/Ca) ratios in carbonate shells, which record changing input sources and hydrologic conditions in the

basin and surrounding region. Sr isotope records show two prominent peaks between �18 and 16 kaBP cal, reflecting anomalous

sedimentation associated with meltwater from disintegrating Eurasian ice sheets that brought Black Sea level to its spill point. Following

a sharp drop in Sr isotope ratios back toward glacial values, two stages of inorganic calcite precipitation accompanied increasing oxygen

isotope ratios and steady Sr isotope ratios. These calcite peaks are separated by an interval in which the geochemical proxies trend back

toward glacial values. The observed changes reflect negative water balance and lake level decline during relatively warm periods

(Bølling–Allerød and Preboreal) and increasing river input/less evaporation, resulting in higher lake levels, during the intervening cold

period (the Younger Dryas). A final shift to marine values in Sr and oxygen isotope ratios at 9.4 kaBP cal corresponds to connection with

the global ocean, and marks the onset of sedimentation on the Black Sea continental shelf. This date for the marine incursion is earlier

than previously suggested based on the appearance of euryhaline fauna and the onset of sapropel formation in the deep basin.

r 2006 Elsevier Ltd. All rights reserved.

1. Introduction

The Black Sea (Fig. 1) is a large and deep body of waterat the landward extreme of a string of marginal basinsseparated from each other by shallow sills (Fig. 2). Duringglacial sea level lowstands, the marine connection wassevered and the Black Sea transformed into a giant lake(Arkhangel’skiy and Strakhov, 1938; Ross et al., 1970;Chepalyga, 1984; Mangerud et al., 2001) similar to themodern Caspian Sea. The evolving composition of the lakewater reflected the hydrological, erosional, and geographi-cal changes within its vast drainage area in Europe and

e front matter r 2006 Elsevier Ltd. All rights reserved.

ascirev.2006.01.032

ing author. Tel.: +1508 289 2460; fax: +1 508 289 2187.

ess: [email protected] (C.O. Major).

Asia. Thus, the Black Sea is well situated to record climatechanges in the continental interior. Prior studies haveinvestigated the lithology and mineralogy of the glacialBlack Sea lake sediments (Muller and Stoffers, 1974;Stoffers and Muller, 1978; Shopov et al., 1986; Major et al.,2002), but have focused primarily on the lake-to-marinetransition (Deuser, 1972; Wall and Dale, 1974; Hay et al.,1991; Lane-Serff et al., 1997; Ryan et al., 1997).Recent studies have begun to capitalize on the Black

Sea’s potential as a high-resolution climate archive(e.g., Ryan et al., 2003; Bahr et al., 2005). This study tiesproxy records of provenance and hydrologic changes in theBlack Sea to regional and hemispheric changes during thelast deglaciation. We show that oxygen and strontiumisotope ratios along with trace element chemistry of

dx.doi.org/10.1016/j.quascirev.2006.01.032

mailto:[email protected]

ARTICLE IN PRESS

60˚N

55˚NDonRiverDniepr

River

DanubeRiver

Manychdepression

SakaryaRiver

50˚N

45˚N

40˚N

40˚E30˚E20˚E10˚E

AC

B

Fig. 1. Map of the Black Sea and surrounding region, with rivers of present-day drainage area (major river systems with geochemical data discussed in this

paper are labeled). The location of the Manych depression, the spill path from the Caspian Sea, is also indicated. Sediment cores come from three survey

areas: the Romanian margin (A), West Crimea (B), and Kerch Strait (C).

Aegean/Mediterranean

SeaMarmara Sea Black Sea

modern Sea level

Bosporus Sill(-35 m)

DardanellesSill

(-70 m)

-120 m

Rivers

Fig. 2. Schematic of the interbasin connections along the Black Sea–Mediterranean corridor. When external (glacioeustatic) sea level lies below the

dividing sill depth (70mbsl, Dardanelles; 35mbsl, Bosporus) then the maximum sea level within the marginal basins is determined by the spill point and

changes in sea level by the hydrologic balance (precipitation+runoff–evaporation). Interbasin exchange can only occur when external sea level lies above

the sill depth(s).

C.O. Major et al. / Quaternary Science Reviews 25 (2006) 2031–20472032

radiocarbon dated carbonate samples record the history ofregional climate changes through this interval and theinflux of marine water. The data point to a previouslyundocumented highstand of Black Sea level early in thedeglaciation that was likely associated with outflow to thedownstream Marmara and Mediterranean Seas. The

timing of the freshwater input responsible for this high-stand indicates a major linkage to the disintegration oflarge ice masses in the Eurasian continental interior(Grosswald, 1998; Denton et al., 1999; Grosswald andHughes, 2002), as well as to meltwater and associated icerafting events in the northwestern Atlantic that preceded

ARTICLE IN PRESSC.O. Major et al. / Quaternary Science Reviews 25 (2006) 2031–2047 2033

the collapse of the North American ice sheets (Knutz et al.,2002) and the associated decrease in strength of themeridional overturning circulation (McManus et al.,2004). We also show that the inundation of the Black Seaby the Mediterranean began at �9.4 kaBP cal, earlier thanprevious estimates of �7.6 kaBP cal which were based onfaunal changes and sequence stratigraphic relationships(Ryan et al., 1997) and the onset of sapropel deposition inthe Black Sea (Ross and Degens, 1974).

2. Geological background

The Black Sea is connected to the global ocean via theBosporus and Dardanelles Straits, and receives runoff fromover 2 million km2 of eastern Europe and western Asia.The density contrast between Mediterranean sourced waterand fresh continental runoff results in water columnstratification, anoxic conditions and the deposition oflaminated, organic-rich sediments at depth greater than�200m below modern sea level (mbsl) (Ross and Degens,1974; Calvert, 1990; Jones and Gagnon, 1994; Arthur andDean, 1998). The modern water balance reflects an excessof input (precipitation+runoff+marine inflow) over eva-poration (Shimkus and Trimonis, 1974; Unluata et al.,1990), which results in an outflow of intermediate salinitywater (16–18 psu) to the basins downstream (Gunnersonand Ozturgut, 1974).

Over the past 3 million years, as world sea levelfluctuated relative to the sill depths with the growth anddecay of large ice sheets, the Black Sea alternated from acompletely isolated interior lake to a brackish-marineenvironment several times, with four extended periods ofhigh salinity corresponding to peak interglacial times inEurope (Schrader, 1979; Zubakov, 1988). The most recentlacustrine stage, persisting during the last glacial periodfrom before 25 kaBP cal until the early Holocene,is evidenced by low-chlorinity sediment pore fluids(Bruyevich, 1952; Manheim and Chan, 1974), brackish tofreshwater biotic assemblages (Nevesskaya, 1965; Wall andDale, 1974; Shcherbakov and Babak, 1979; Aksu et al.,2002), and light oxygen isotope ratios (d18O) in calciumcarbonate (CaCO3) (Deuser, 1972; Major et al., 2002).In fact, it has been estimated that the Black Sea waslacustrine for 90% of its Pleistocene history (Ross, 1978),because the connection with the oceans across the shallowdividing sill requires particularly high sea levels. Drownedcoastal and alluvial features and erosional unconformitieson the continental shelves provide solid evidence of pastBlack Sea lake lowstands of up to 155mbsl (Shcherbakovet al., 1978; Khrischev and Georgiev, 1991; Ryan et al.,1997; Ballard et al., 2000; Aksu et al., 2002), though thepresent depths of these features may be influenced byneotectonism in some areas. These lowstands indicate thatthe hydrologic balance of the ‘‘Black Sea lake’’ must havebeen negative at times in the past, analogous to theCaspian Sea. The relative freshness of the Black Sea priorto the most recent marine incursion required that water

from the prior marine connection (probably during theKarangatian Stage, equivalent to Marine Isotope Stage 5e,the most recent full interglacial) was flushed out of thebasin (Kvasov, 1968), necessitating periods of outflow (i.e.,positive hydrologic balance), and thus highstands, duringthe lacustrine stage.The Holocene incursion of marine water into the Black

Sea basin was accompanied by progressive introduction ofsalt-tolerant species whose CaCO3 shells show secularchanges in d18O (Ryan et al., 1997; Aksu et al., 2002). Deepbasin and continental slope cores document a pronouncedshift in the d18O of bulk carbonate from light freshwatervalues (��6%) to heavier values approaching the modernmarine range (0–+2%) between 9000 and 8000 yr 14CBP(10.1–8.85 kaBP cal) (Deuser, 1972; Major et al., 2002).Over this interval, the mollusk stratigraphy shows a changein assemblage reflecting a transformation to increasinglybrackish environments (Popov, 1973; Shcherbakov andBabak, 1979). Marine species of mollusks (Ryan et al.,1997) and dinoflagellates (Wall and Dale, 1974) appearsubsequently close to the onset of sapropel formation at7160750 yr 14C (�7.6 kaBP cal assuming a 400 yr radio-carbon reservoir age for the modern Black Sea) (Jones andGagnon, 1994; Siani et al., 2000).Based on sedimentological and faunal evidence tied to

shelf-wide unconformities in seismic reflection profiles,Ryan et al. (1997) suggested that the most recentreconnection of the Black Sea with the Mediterraneanoccurred at 7.5 kaBP cal when rising glacioeustatic sea levelbreached the Bosporus Sill and poured into a Black Sealake depressed below its spill point. They concluded thatresulting ‘‘flooding’’ of the Black Sea with marine watersquickly brought its surface up to the level of the globalocean. Aksu et al. (2002) challenged the ‘‘flood’’ hypoth-esis, citing evidence of back-stepping barrier bars on thesouthern Black Sea shelf, whose age they infer byextrapolation to indicate a transgression already underwayby 13 kaBP cal. Gorur et al (2001) report evidence ofestuarine sediments in the Sakarya River valley nearly 20mhigher than the Bosporus Sill depth shortly before theproposed time of flooding, and conclude that the Black Sealake must have been outflowing over its sill at the time ofthe marine reconnection, thus precluding a ‘‘flood’’.Freshwater outflow from the Black Sea was inferred fromobservations of sapropelic sediments and less salinity-tolerant biotic assemblages in the ‘‘downstream’’ basins(Marmara, Aegean, and eastern Mediterranean Seas)shortly before the marine connection of the Black Sea(Aksu et al., 2002). However, these salinity inferences havebeen questioned by Sperling et al. (2003), who show anincrease in d18O values in carbonates of the Marmara Seaduring the proposed period of Black Sea outflow, i.e., theopposite sense of change to what would be expected from amajor freshwater supply.Knowing the level of the Black Sea at the time of the

marine incursion is crucial in determining the mode of themarine input (i.e., gradual or catastrophic). Part of

ARTICLE IN PRESSC.O. Major et al. / Quaternary Science Reviews 25 (2006) 2031–20472034

determining this level involves understanding the mechan-isms that controlled Black Sea level prior to the marinereconnection. This paper discusses the changes in thefreshwater sources to the Black Sea during its isolationfrom the ocean, and shows linkages between regionally andglobally recognized millennial scale climate events of thelast deglaciation. We also show that the first marine inputto the Black Sea occurred simultaneously with flooding ofthe Black Sea continental shelf, consistent with thehypothesis of Ryan et al. (1997) but occurring more thana thousand years earlier than their proposed ‘‘flood’’ event.

3. Isotopic proxies for Black Sea hydrologic and source

changes

Changes in the oxygen isotopic compositions of naturalwaters can result from different processes, most impor-tantly through mixing of waters with different composi-tions, and kinetic and biological fractionation. Deuser(1972) interpreted d18O variations in Black Sea bulkcarbonates in terms of mixing between isotopically heavymarine water and light freshwater, and showed that theisotopic change followed lithologic changes indicatingdensity stratification (Fig. 3a). Other bulk carbonate d18Orecords (Major et al., 2002) show a more complicatedpattern prior to sapropel formation (Fig. 3b). Bulk

(A) (B)

0

100

200

300

400

500

600

-8 -6 -4 -2 0 2

9.6

16.5

20.1

0

100

200

300

400

500

600

700

800

-10 -8 -6

12.

15.

(cm)

(cm)

δ18Obulk carbonate (‰, PDB)

δ18Obulk c(‰, P

Fig. 3. Comparison of oxygen isotope records from the Black Sea: (A) d18Obul

(B) d18Obulk carbonate from core BLKS9810 (378m; Major et al., 2002), and (C)

(mixed cores). Note that (A) and (B) are plotted vs depth in core (cm), while (C

are indicated for (A) and (B). Correlations between records are indicated with d

incursion.

carbonate d18O records differ substantially from benthicshell d18O (Fig. 3c), which begins to rise several thousandyears prior to the bulk carbonate d18O. Clearly, interpreta-tion of d18O simply in terms of mixing will underestimatethe influence of changes in the sources themselves, orprocesses that fractionate oxygen isotopes such as evapora-tion, both of which can be important in an isolated systemsuch as the glacial Black Sea lake. The base level changesinferred from drowned shorelines point to negative shifts inthe hydrologic balance of the basin, indicating a relativeincrease in the influence of evaporation. Furthermore,shifts in meteoric water d18O, related to temperaturevariations, influence the overall lake isotopic compositionthrough runoff and precipitation (von Grafenstein et al.,1999). All of these factors result in a shift toward higherd18O—the same sense of change expected from marineinput. In order to distinguish between changes due tomarine input and changes related to source and hydrologicchanges, we must employ a tracer that is insensitive to theselatter effects.In contrast to oxygen isotopes, strontium isotope ratios

show no measurable fractionation from biological effects,temperature, or other physical environmental changes.Although the 87Sr/86Sr ratio of seawater has changedover geologic time, the value has remained constant overthe several thousands of years studied in this paper

(C)

0

5

10

15

20

25

30

-9 -7 -5 -3 -1 1-4 -2 0

7

15.32

18.0

21.1

103 Years BP(cal)

arbonate DB)

δ18Oshell(‰, PDB)

k carbonate from core 1474K (2117m, 42123.30N, 37136.20E) (Deuser, 1972),

d18Oshell measured on mollusk shells in cores from the three survey areas

) is plotted vs age (103 yrBP cal). Calibrated ages (from radiocarbon dates)

ashed lines; gray shaded bars indicate depth (A and B) or age (C) of marine


(Henderson et al., 1994). Sr isotope ratios of continentalwater bodies, on the other hand, reflect the catchmentgeology and can differ substantially between drainagebasins (e.g., Stein et al., 1997; Krom et al., 1999). Biologicaland inorganic precipitates record the changes, which resultfrom input and mixing of different water sources, providedthe different sources have sufficiently different isotoperatios. Seawater (currently 87Sr/86Sr ¼ 0.709155 for theglobal ocean, Henderson et al., 1994; 0.709157 forthe Aegean Sea; 0.709150 for the Marmara Sea—allindistinguishable from each other in terms of measurementerror) has a significantly different Sr isotope ratio fromthe average river water feeding the Black Sea(87Sr/86Sr ¼ �0.7088; Palmer and Edmond, 1989), as wellas a factor of �30 higher concentration of dissolved Sr(Broecker and Peng, 1982; Palmer and Edmond, 1989).Therefore, a small fraction of seawater can impart ameasurable Sr isotope signal. The salinity of modern BlackSea surface water, at 18%, represents roughly a 1:1 mixtureof marine and river water. The predicted 87Sr/86Sr ratio ofthis mixture is �0.709144, nearly that of seawater,reflecting the influence of the high proportion of marine-sourced Sr. The average 87Sr/86Sr of mollusks in the marinesediments of the Black Sea is 0.709133 (70.00002)indicating a slightly greater influence of riverine Sr thanthe simple 1:1 mixture prediction.

Early studies of Sr isotopes in Black Sea carbonatesshowed marine values in the uppermost coccolith-bearingsapropel and lower values (averaging 0.707370.0003) inthe lower lacustrine sediment (Cox and Faure, 1974). Thesemeasurements were made on bulk carbonates rather thanshells, and the differences were not interpreted directly interms of changes in water chemistry. Rather they wereinterpreted as reflecting Sr in reworked Tertiary (low87Sr/86Sr) carbonates during the glacial lacustrine phaseand in the marine Emiliani huxleyii coccolith during theHolocene. More recent studies in other marginal basinshave relied on more precise measurements in carefullyselected biogenic carbonates to detect changing marineinfluence and drainage reconfigurations in estuarine areas(Ingram and Sloan, 1992; Israelson and Buchardt, 1999).These studies reveal the sensitivity of the 86Sr/88Sr signal insuch environments, and the great utility in combiningrecords with other geochemical proxies, such as oxygenisotopes and trace element concentrations, that incorporateresponses to important environmental variables liketemperature and salinity. We show that the geochemicalevolution in the Black Sea is associated with lithologic andfaunal changes, and the timing of these changes is relatedto the climatic variations associated with the deglaciation.

4. Methods

Materials used in this study were recovered during thejoint French–Romanian BLaSON mission in 1998 (coresfrom the Romanian margin) and a joint Russian–Americanexpedition in 1993 (cores from the West Crimea margin

and Kerch Strait areas) (Fig. 1, Table 1). Radiocarbon agesfor all Romanian margin cores (BLKS- and BLVK-), aswell as a subset of the West Crimea and Kerch area cores,were measured at the ETH-Hoenggerberg AMS facility inZurich, Switzerland, with the exception of a single sample(BLVK9814, 130 cm) that was dated at both ETH and theLaboratoire des Sciences du Climat et de l’EnvironnementAMS facility in Gif-sur-Yvette, France. The remainingWest Crimea and Kerch area core dates were obtainedfrom the NOSAMS facility at the Woods Hole Oceano-graphic Institution (see also Ryan et al., 1997). Molluskshells were leached prior to measurement in order todecrease the effects of diagenetic carbon contamination.Nearly half of the samples analyzed for isotopes and traceelements come from dated horizons (see SupplementaryData); the remaining ages are assigned according to linearextrapolations between dated points. The dated shellsthemselves were not used for the other geochemicalanalyses.d18O was measured on a Micromass Optima CO2-source

mass spectrometer at the Lamont-Doherty Earth Observa-tory (LDEO) with a Multiprep carbonate preparationdevice. Calibration to VPDB is via NBS-19 and an in-house standard. Reproducibility of the standards was70.08% (1s). For Sr isotope analyses, mollusk shells weresuccessively leached using a procedure modified fromBailey et al. (2000), first with 0.1M HCl to removediagenetic surface recrystallization, followed by a washwith double-distilled water, and then with 0.02M HCl topartially dissolve the shell carbonate. Ostracod samples(Candona sp.) were cleaned with 5% H2O2, and 10–15individual valves (75–300 mg total weight) were picked foranalysis. Clay samples were isolated with the followingprocedure: (1) removal of CaCO3 by agitation in a weakacetic acid solution, (2) removal of amorphous iron oxidesby reaction with sodium citrate, sodium sulfate, sodiumchloride, and acetone, (3) removal of amorphous silica andaluminum oxides by reaction with sodium carbonate in anultrasonic bath, and (4) separation of the o2 mm sizefraction by settling. The clay fraction was then dissolved inhydrofluoric and nitric acid. Ostracod samples weredissolved directly in 3N HNO3 acid. Sr was extractedusing Eichrom Sr-spec resin. Sr was loaded onto tungstenfilaments with TaCl5 (Birck, 1986), and Sr isotope ratioswere measured by dynamic multi-collection on a VG Sector54 thermal ionization mass spectrometer at LDEO.87Sr/86Sr ratios are normalized to 86Sr/88Sr ¼ 0.1194. Beamsize was maintained at close to 5� 10�11A for 88Sr.Instrument performance was monitored through analysisof NBS987, which gave 87Sr/86Sr ¼ 0.710288 (70.000015,2s external reproducibility, n ¼ 16). All 87Sr/86Sr ratios arefurther corrected to an NBS987 value of 0.710235.Reported errors are the in-run 2s error of the mean. The87Sr/86Sr of modern seawater carbonate was measured at0.709150 (70.000016, 2s external reproducibility, n ¼ 6),which is indistinguishable from that of the modern ocean(Henderson et al., 1994). Sample preparation and analyses

ARTICLE IN PRESS

Table 1

Core locations

Core name Survey area Water depth (m) Core length (cm) Latitude (E) Longitude (N)

BLVK9801 Danube (BLaSON) 92 49 44115.20 30124.680

BLKS9804 Danube (BLaSON) 101 81 44112.00 30132.210









BLKS9830 Danube (BLaSON) 70 187.5 4410.640 29153.720

BLKS9831 Danube (BLaSON) 75 188 4411.90 29154.840 0

BLKS9834 Danube (BLaSON) 76 253.5 4410.660 29153.710



BS1 Dniepr 68 134 44157.60 3215.50

BS3-2 Dniepr 49 58 45121.90 31149.70

BS7 Dniepr 108 145 44153.40 3219.30

BS8 Dniepr 99 159 44154.40 3218.50

BS9 Dniepr 123 170 44153.00 3219.20

BS10 Dniepr 106 175 44153.50 3218.80

BS11 Dniepr 91 197 44155.20 3218.00

BS12 Dniepr 78 146 44158.80 32111.10

BS13 Dniepr 165 190 44155.30 32116.50

BS14 Dniepr 140 305 44151.60 32121.30

BS22 Kerch 129 118 44139.20 36135.20

BS24 Kerch 110 347 44140.00 36134.50


for Sr/Ca were carried out at LDEO using a Jobin-YvonPanorama ICP-AES with autosampler attachment follow-ing the techniques described in Schrag (1999) and usingcharacteristic spectral lines of 317 nm for calcium and407 nm for strontium. Data were calibrated by comparisonto the Mercenaria-3 standard. The average error (1 s)for the standards was 3.6%, and replicate percent errorwas 1.6%.

The changing environment in the Black Sea over the past25 000 years and the broad depth range of our cores(49–378mbsl) covering mid-shelf to upper slope environ-ments meant that no single species could be used for theentire time period. While this is potentially significant forthe d18O (Wefer and Berger, 1991) and Sr/Ca (Rosenthaland Katz, 1989), it is not a problem for 87Sr/86Sr (Faure,1998). Though all Sr/Ca data are presented in theSupplementary Data Table, we focus on data fromDreissena species in identifying trends. In the lacustrinesediments (prior to �9.4 kaBP cal), the d18O data areprimarily from Dreissena species, with the exception ofthree measurements made on gastropods. Some of thescatter in the d18O younger than �9.2 kaBP cal probablyreflect interspecies offsets, but these differences are minorcompared with the amplitude of the temporal changeswithin the Dreissena data.

In our discussion below, we assume that different proxiesfrom a depth in a core represent the conditions at thatpoint in time; however, this assumption is not valid if a

sedimentary unit includes a significant amount of reworkedmaterial. The lacustrine–marine boundary on the con-tinental shelf and upper slope is commonly marked by awinnowed and reworked unit, referred to here as the ‘‘shellhash’’. Although each proxy data point from this unit isassigned an age based on the 14C age of a shell from thesame depth, due to the possibility of reworking, the real agecould be younger. Therefore, we do not view data from theshell hash as representing a strict time series but rather arange of compositions within a discrete time window.Finally, the isotope and trace element records integratedata from three survey areas (Kerch Strait, West Crimea,and the Romanian margin); to distinguish these multiplecore data from the single core data (BLKS9810), we referto them as ‘‘mixed-core’’ results. The coherency of thegeochemical trends indicates that the major changesobserved are basin scale, not local, although localphenomena may also account for some of the scatter inthe data.

5. Results

5.1. Chronology

Radiocarbon ages are converted to calendar ages(kaBP cal) using the Calib 4.4 software (Stuiver et al.,1998; Stuiver and Reimer, 2004) assuming no reser-voir correction (Major et al., 2002). The reservoir age


(or the ‘‘hard water effect’’ resulting from the input ofradioactively ‘‘dead’’ carbon to the basin by the weatheringof carbonate rocks) for the lacustrine Black Sea is notknown. Although other investigators (Ostlund, 1974;Guichard et al., 1993; Bahr et al., 2005) have foundevidence for reservoir ages of 1000 years or more for thedeep waters (4400mbsl) of the Black Sea in both recentand glacial time, shallower depths do not appear to exhibitthis offset (Bahr et al., 2005). As all the cores used in thisstudy come from depths less than 400mbsl, we calculatecalendar ages using no reservoir correction for thelacustrine sediments and obtain good temporal relation-ships to other calibrated climate records (Major et al.,2002). The observation that the radiocarbon offset relativeto atmospheric 14C at shallow water depths in the modernBlack Sea (415790 y, Siani et al., 2000), at a time ofpersistent density stratification, is comparable to that in theNorth Atlantic suggests that a hard-water effect did notsignificantly increase the age of Black Sea lake surfacewaters. Age models for cores BLKS9809 and BLKS9810,as well as for other cores with multiple 14C analyses, wereconstructed by linear interpolation between dated levels(Major et al., 2002).

5.2. Isotope and trace element time series

The main features of the isotopic records are described inseveral stages occurring over the past �25 000 years(Fig. 4). These stages (labeled a through j) correspond closelywith major lithologic units noted in sediment cores from theshelf and slope (Major et al., 2002; Ryan et al., 2003).

5.3. �25–18 ka BP cal (stage j)

In the earliest part of this Black Sea record, whichincludes the Last Glacial Maximum (LGM), both oxygenand strontium isotope ratios show low variability(d18O ¼ �6.1 to �7.2%, 87Sr/86Sr ¼ �0.70865–0.70875)and Sr/Ca ratios also show fairly low variability( ¼ 2.4–3.3 in mg/g, Ca/Sr ¼ 319–418 in g/g), althoughthe resolution of the record is poor due to the scarcity ofshell material. The 87Sr/86Sr values of shells older than18 kaBP cal and associated with the LGM(0.7087470.00008 2s, N ¼ 13) are close to a weightedaverage of the major rivers entering the Black Sea: theDanube, Dniepr, Don, and Sakarya rivers (�0.70879;Table 2). Sr/Ca ratios in mollusk shells are offset from theaverage river input due to preferential incorporation ofcalcium relative to strontium (Turekian and Armstrong,1960); the partition coefficient for Dreissena in the BlackSea lake would be �0.6, which is high relative to otheraragonitic freshwater species (Rosenthal and Katz, 1989).Other high-resolution proxy records (Major et al., 2002),such as clay mineralogy, carbonate content, and bulkcarbonate stable isotopes, also suggest little variability instage j. Sedimentation is mainly homogeneous gray mud.

5.4. �18–15.9 ka BP cal (stages g through i)

Beginning at �18 kaBP cal, 87Sr/86Sr ratios undergo amarked rise (stage i) from LGM values. The changecorresponds with the appearance of two reddish-brownclay layers, which have also been noted in other cores fromthe western Black Sea (Khrischev and Georgiev, 1991;Bahr et al., 2005). These clays are characterized by aseveral-fold increase in illite and kaolinite (Major et al.,2002), which are the dominant clay types of the modernnorthern Black Sea shelf (Muller and Stoffers, 1974), andmuch higher 87Sr/86Sr values (0.751) than the gray claysabove and below (0.723–0.730) (see Supplementary Data).d18O data show a slight but progressive decrease of �1% toa minimum value of –8.1% at �17.6 kaBP cal, and Sr/Caremains in the same range as in stage j. After a brief returnto gray mud similar to the LGM compositions andassociated with falling 87Sr/86Sr (stage h), a second peakin 87Sr/86Sr occurs with the appearance of another thinnerreddish-brown clay layer (stage g). CaCO3 reaches itslowest values within the reddish-brown clay layers (o7%).

5.5. �15.9–15.2 ka BP cal (stage f)

87Sr/86Sr values drop sharply in the gray muds above thered-brown clay zone, as does the ‘‘northern component’’signature in the clay mineralogy (stage f). There is adoubling in carbonate content through the interval, reach-ing values higher than at any time previous in the record.

5.6. �15.2–9.4 ka BP cal (stages c through e)

Between 15.2 and �13.5 kaBP cal (stage e, whichincludes the dramatic warming of the Bølling oscillationand the subsequent mild Allerød oscillation), d18O climbsprogressively, reaching approximately �4%, and Sr/Caincreases to �4.5mg/g (Ca/Sr decreases to �220 in g/g),while 87Sr/86Sr remains constant or decreases slightly.Sediments on the upper slope of the western Black Seashow much higher percentages of CaCO3 (e.g., in coreBLKS9810), peaking at over 30% between �14.7 and13.3 kaBP cal (Major et al., 2002). d18O values between13.3 and 11.6 kaBP cal (stage d, encompassing the YoungerDryas) are intermediate to those in the intervals before andafterward (stages e and c), although the data within thisinterval do not show a clear trend. 87Sr/86Sr ratios andSr/Ca show a slight decrease in stage d, carbonate contentssignificantly decrease, and silt-sand fractions significantlyincrease in continental slope settings (Major et al., 2002;Bahr et al., 2005). Between �11.6 and 9.4 kaBP cal (stage c,when temperatures in Greenland show a sustained increasetoward full Holocene conditions), CaCO3 in the upperslope core (BLKS9810) shows a pronounced peak(to 460%), which is even higher than the Bølling–Allerødpeak (stage e).A single 87Sr/86Sr value of 0.709071 lies well off the trend

defined by the other samples (�0.7089) in stage c.

ARTICLE IN PRESS

YearsBP (cal.)

δ18 O

carb

onat

e (‰

, PD

B)

δ18 O

ice

(‰, S

MO

W)

% c

arbo

nate

% (

illite

+ka

olin

ite)

87S

r/86

Sr

Sr/

Ca

(mg/

g)

a b c d e f g-i j

0.7091

0.7089

0.7087

2500020000150001000050000

2

4

6

8

-7

-5

-3

-1

20

40

60

-44

-40

-36

-32

(A)

(B)

(C)

(D)

(E)

YDB/A

H1

Fig. 4. (A) GISP2 d18Oice (Grootes et al., 1993), with Heinrich event 1 (H1), Bølling–Allerød (B/A) and Younger Dryas (YD) intervals marked; (B) %

carbonate (circles) and % illite+kaolinite (dominant northern-source clay minerals; asterisks) from core BLKS9810; (C) d18Omollusk, (D) 87Sr/86Srcarbonate(dashed line connects data from core BLKS9810), and (E) Sr/Ca of Dreissena sp., all mixed-core data. Shaded areas indicate the marine section (green),

carbonate peaks (blue), and reddish-brown clay layers (pink). Data symbols for the d18Omollusk,87Sr/86Srcarbonate, and Sr/Ca denote the lithologic unit from

which the samples were taken (green: shelly shelf muds with marine mollusks; white: shell hash; blue: carbonate-rich muds; brown: shelf channel fill; gray:

grey mud; brown: reddish-brown muds); (a) through (j) denote unit labels used in Fig. 5. Data are included in the Supplementary Data Table.


ARTICLE IN PRESS

Table 2

Sr inputs to the Black Sea

Modern flow % of total

river input1

87Sr/86Sr (Sr) (ppm) (Ca) (ppm) Sr/Ca (Ca/Sr) mg/g (g/g) Data source

Danube 53 0.7089 0.24 49 4.9 (203) 1, 2

Dniepr 14 0.7085 0.22 45 4.9 (205) 1, 2

Don 7 0.7085 0.22 53 4.2 (241) 1, 2

Sakarya �4 0.7089 This study

River average 0.708792 0.24 48.6 4.8 (207)

Seawater (global average) 0.709155 7.62 413 18.5 (54) 3, 4

Aegean Sea (modern) 0.709157 This study

Marmara Sea (modern) 0.709150 This study

Caspian Sea (modern) 0.7082 5

Data sources: 1: Shimkus and Trimonis (1974); 2: Palmer and Edmond (1989); 3: Henderson et al. (1994); 4: Broecker and Peng (1982); 5: Clauer et al.

(2000).

C.O. Major et al. / Quaternary Science Reviews 25 (2006) 2031–2047 2039

This outlier value reflects five measurements on twodifferent shells of euryhaline mollusks (Cardium andAdacna) from the same horizon at a present water depthof 55mbsl. The 14C ages of both shells, measured at twodifferent labs, were 9580780 and 9580790 14C yr(�10.85 kaBPcal). The sedimentary unit from which thissample was taken is a grayish-brown silty clay with organic-rich laminae. This unit is bounded at the top and bottom bysharp erosional contacts, and is seen in seismic profile tocorrespond to infill of a local depression in the shelf.

5.7. 9.4–0 ka BP cal (stages a and b)

At 9.4 kaBP cal (stage b, close to when Greenlandtemperatures reach full Holocene values), there is a sharptransition to lower CaCO3 and a pronounced increase ind18O, 87Sr/86Sr, and Sr/Ca. These changes signal the marineincursion, which commences just after the upper carbonatepeak. Samples from the mixed cores indicate a slightlyearlier timing for the marine input (9.5–9.2 kaBP cal). Thedifference reflects uncertainty in the BLKS9810 age model,in which the transition is dated by interpolation between a14C date of 10 640780 yr at 94.5 cm and 7100 yr at 38.5 cm,the base of the sapropel (Jones and Gagnon, 1994)(see Table 1). The geochemical transition is associatedwith a change in lithology on the continental shelf andslope: a thin (3 cm), sharply defined gray-green mud lyingbetween the gray carbonate-rich muds and the base of thesapropel in BLKS9810, and a unit of crushed and wholeshells (shell hash) in the mixed-core samples. The shell hashunconformably overlies stiff, barren muds on the mid-shelf(from o92mbsl to at least 50mbsl) and Dreissena-bearingmuds on the outer shelf, and is characterized by a morediverse and salinity-tolerant assemblage of mollusks (Ryanet al., 2003) and dinocysts (Mudie et al., 2004) than olderlithologic units. The geochemical transition is complete by�9 kaBP cal. After this time and until the present day, thed18O and 87Sr/86Sr vary minimally, and are associated with

the presence of Mediterranean-type mollusk species (e.g.,Cardium and Mytilus).

5.8. Interproxy relationships

Each stage and its associated lithologic unit has acharacteristic geochemical signature, and the evolution ofwater compositions can be illustrated through the relation-ships between proxies (e.g., 87Sr/86Sr vs d18O in Fig. 5a andc and 87Sr/86Sr vs Ca/Sr in Fig. 5b and d). Note that87Sr/86Sr is plotted vs Ca/Sr and not Sr/Ca in order tohighlight a linear relationship in the binary mixing betweenendmember compositions (Faure, 1998); 87Sr/86Sr vs d18Omixing trends will follow hyperbolic paths. Time periodscorresponding to the glacial muds through the gray mudsoverlying the reddish-brown clays (stages f through j) showvariability in 87Sr/86Sr with no significant change in d18O orCa/Sr, whereas the carbonate-rich sediments and theintervening gray mud (stages c through e) show aprogressive increase in d18O without significant change in87Sr/86Sr. In contrast, 87Sr/86Sr, d18O, and Ca/Sr changetogether through the shell hash (stage b), signaling thetransition to marine conditions in the basin.

6. Discussion

6.1. Evidence for marine incursion at �9.4 ka BP cal

Although the lake–marine transition in the Black Sea isthe most dramatic environmental change in the basin in thelast 25 000 years, there has nonetheless been wide debateabout when it occurred. We start with a discussion ofevidence for the first marine incursion into the Black Sea.The unique combination of geochemical, geological, andfaunal changes associated with marine transition allow usto distinguish it from earlier environmental changes in theBlack Sea lake.The 87Sr/86Sr ratios of the biogenic carbonates, rising

abruptly from 0.70891 to 0.70901 in stage b, clearly show

ARTICLE IN PRESS

Fig. 5. 87Sr/86Sr vs d18O data cross plot (A) and schematic (C), and 87Sr/86Sr vs Ca/Sr data cross plot (B) and schematic (D). Schematic axes are identical

to data axes. Colored symbols in (A) and (B) represent average values within each stage, and error bars (and colored shading) representing 2s standard

deviation; (a) through (j) refer to time slices of the geochemical data corresponding to the major lithologic changes in the shelf and slope cores (see Fig. 4).

(C) and (D) highlight the direction of geochemical change within each stage. Ca/Sr is used to highlight the linear mixing relationship between endmembers

with Sr in the denominator. Although all mixing trajectories for 87Sr/86Sr vs d18O will be somewhat curvilinear, the hyperbolic relationship is shown only

for the most extreme case—the lacustrine–marine transition. Heavy dashed lines in (C) and (D) denote the 10–50% marine influence (transition);

the 87Sr/86Sr vs Ca/Sr trajectory through the marine endmember is offset because of the lower D for marine shells (�0.3) than fresh-brackish Dreissena

shells (�0.6).


the linkage between the Black Sea and the global ocean(Fig. 4). The shell d18O record also shows an offset in thesame time period—values prior to the transition are��2.6%, and the amplitude of the d18O shift is �2.1%.Modern-day values are reached with the appearance ofMediterranean fauna at the beginning of stage a,�9 kaBP cal, suggesting that exchange between the BlackSea and Mediterranean was similar to today by that time.Mixing relationships of both 87Sr/86Sr vs d18O and87Sr/86Sr vs Ca/Sr (Fig. 5) show a trend in the shell hashtoward the marine endmember (stage a) from thecompositions characterizing the uppermost lacustrine unit(stage c, CP-1). The ‘‘shell hash’’ lies directly along amixing line between the river and marine endmembers(Table 2), indicating that the mollusks of this unit grewduring the marine transition. Dinocyst assemblages in thecorresponding transitional unit (B1c; Mudie et al., 2004)from the deep southwestern basin also indicate increasingsalinity and the presence of seawater.

The timing of first marine input inferred from the Srisotope and other geochemical data is earlier than the ageof �7.5 kaBP cal suggested by Ryan et al. (1997) on thebasis of the first appearance of euryhaline marine fauna(e.g., Cardium exiguum). Thus, the Black Sea containedmarine water prior to the onset of sapropel deposition at�7.5 kaBP cal. This is consistent with models suggesting atime lag between the first marine inflow and the depletion

of bottom water oxygen required for sapropel formation(Deuser, 1972; Lane-Serff et al., 1997). With the appear-ance of Mediterranean-type mollusk species such asCardium exiguum and Mytilus galloprovincialis at�7.5 kaBP cal, dinoflagellates shifted to full euryhalineassemblages (Wall and Dale, 1974; Atanassova, 1995),suggesting progressive salinification and a threshold con-trol on the appearance of marine species.While shell hash is found in cores recovered from a range

of depths between 49 and 129mbsl, those from coresshallower than 80mbsl are never older than 9.4 kaBP cal(Fig. 6). Below 100mbsl, this unit contains older Dreissena

shells, reworked from the underlying shell-rich mud. Atdepths shallower than 80mbsl, the shell hash sits on anunconformable surface and organic carbon dates from thebarren unit below are 428 kaBP (Major, 2002). The lackof older intact shells at these shallower depths suggests thatsubaerial reworking, erosion, and abrasion of the Dreisse-

na-rich mud produced the shell hash matrix that laterserved as a substrate for more salt-tolerant speciescolonizing the Black Sea after the marine incursion. Suchsubaerial exposure is reflected in the low water contents ofthe shell hash and underlying sediments cored on the BlackSea shelf (Ryan et al., 1997). The timing of the first marinesignal indicates that the Bosporus sill was breached whenexternal (Mediterranean) sea level reached �30mbsl. Infillof the Sakarya River estuary at a depth of 28mbsl began

ARTICLE IN PRESS

0

20

40

60

80

100

120

140

160

180

0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000

Years BP (cal)

Dep

th b

elow

pre

sent

sea

leve

l (m

)

9.4

ka B

P c

al

Preserved coastalbedforms

Fig. 6. Sample age vs core depth. Symbols correspond to lithologic unit and/or mollusk assemblage (black circles: euryhaline (marine) fauna; white

triangles: brackish fauna in shell hash; grey circles: lacustrine sediments; gray square: euryhaline fauna in channel fill; black star: peat deposit (Gorur et al.,

1998)). Thin dashed line is ice-equivalent global sea level (Fleming et al., 1998). The marine linkage at 9.4 kaBP cal indicates a sill depth of �30m. Grey

bar indicates the depth range of well-preserved coastal bedforms, which lie well below the depth of contemporaneous glacio-eustatic sea level from

14 kaBP cal (Ryan et al., 2003). The channel fill sample is several tens of meters shallower than immediately older and younger samples.

C.O. Major et al. / Quaternary Science Reviews 25 (2006) 2031–2047 2041

shortly after 9 kaBP cal, characterized by the same molluskassemblages as are found in the shell hash (Gorur et al.,2001), suggesting a more gradual rise of sea level followingthe reconnection.

The simultaneous colonization of three areas of theBlack Sea continental shelf by a diversified brackish waterfaunal assemblage at a large depth range suggests that:(1) the rise in Black Sea level across the shelf coincided withthe marine incursion, and (2) the rise was abrupt enough tonot result in onlapping sediments. These observationsfurther suggest that the Black Sea was at a lowstand at thetime of the incursion, most likely corresponding to thedrowned shoreline features at 70–90mbsl observed on boththe Romanian and Russian margins (Ryan et al., 2003).The preservation of such features is further evidence for arapid base level rise, since the bedforms were not destroyedby ravinement during the transgression. Other evidence fora powerful reworking event associated with the marineincursion includes a prominent peak in detrital input (Bahret al., 2005) and grain size (Major et al., 2002) on thewestern continental shelf as well as a widespread ‘‘wash-out’’ surface between the Black Sea lake sediments and thebase of the sapropel (Khrischev and Georgiev, 1991).

Simple modeling of two scenarios for the incursion ofmarine water into the Black Sea (a ‘‘flood’’ case with rapidfilling vs a gradual inflow case with the Black Sea already atits spill point prior to the marine incursion) show a good fitto the 87Sr/86Sr data (Fig. 7). The ‘‘flood’’ case assumes anabrupt input of marine water equal to the volumedifference between a lowstand Black Sea (80mbsl) and aBlack Sea at its spill point (35mbsl), accounting for a 4%volume increase, and shows a slightly better fit to the

apparent early rise of 87Sr/86Sr ratios seen in the shell hash.However, because of the uncertainties in the dating andstratigraphic order as well as the presence of olderreworked material within this unit, it is not possible todefinitively distinguish between the two scenarios basedsolely on the geochemistry. It is clear, however, that themarine input caused a significant and geologically abruptchange in sedimentation as well as in the Black Sea waterchemistry.The shells dated at 9580 14C yr (10.9 kaBP cal) that lie

well off the trend defined by stage c (Fig. 4) are from coreBLVK9814, which lies at 55mbsl in a shelf depression. Thesample contains small specimens of the marine mollusksCardium and Adacna not found in any of the time-equivalent lithologic units on the deeper shelf or slope.These shells have d18O values within the range ofcontemporaneous data, but 87Sr/86Sr close to modernseawater. We consider this to be coincidental, because amarine water influence at that shallow water depth wouldhave left a recognizable imprint in the deeper parts of thebasin. These shells also point to a different environmentthan in the open lake: the assemblage comprises onlydwarfed specimens of Mediterranean-type species. Shcher-bakov and Babak (1979) describe the same faunalassemblage in facies they interpret to be perched salineponds along the Black Sea shelf and Sea of Azov, whichwere terrestrial settings during periods of lowered BlackSea level. The chemistry of such a small water body issensitive to the local conditions, driven by evaporation,algal blooms, and groundwater leakage. The 87Sr/86Sr ratiois distinct from the Black Sea water at that time, which issignificant because it requires that the level of the Black Sea

ARTICLE IN PRESS

(A)

(B)

(C)

0.7091

0.7090

0.7089

60

40

20

0

530

510

4906 7 8 9 10 11 12

103 Years BP (cal)

87S

r/86

Sr

Vol

ume

mar

ine

inflo

w(1

03 k

m3 /

200

y)V

olum

e B

lack

Sea

(103

km

3 )

Fig. 7. (A) Model comparison with data for a gradual inflow scenario

(grey solid line) and a flood or ‘‘dam break’’ scenario (black dashed line).

The model volume of marine inflow (B) and total Black Sea volume

(C) are shown for gradual inflow and flood scenarios.


lake lay below 55mbsl at 9580 14C yr (10.9 kaBP cal) andcould not have been outflowing over a shallow sill.

6.2. Implications of the pre-Bølling record (415 ka BP cal)

Increases in 87Sr/86Sr prior to the Bølling warming couldnot be caused by marine input because they occurredbefore the Mediterranean Sea reached the bedrock depth ofthe Dardanelles Sill (75–80mbsl, Lambeck and Bard, 2000;Yaltirak et al., 2002) and invaded the Marmara Sea, thefirst gateway to the Black Sea–Marmara corridor. Salini-fication of the Marmara Sea began much later, at �12000 14C yr (�14 kaBP cal; C- agatay et al., 2000; Sperlinget al., 2003). Furthermore, the decreases of 87Sr/86Sr inBlack Sea shells in stages h and f would require a cut-off ofmarine input. Once a marine connection was establishedthen, barring significant vertical motion at the sill,exchange cannot be shut off if global sea level continuedto rise through the deglaciation (Fig. 6, c.f. Fairbanks,1989; Siddall et al., 2003). These early variations in87Sr/86Sr, associated with changes in sedimentary input tothe Black Sea, must therefore be due to changes in the

freshwater sources. Rivers currently account for about one-third of the total water input (Unluata et al., 1990) and lessthan 4% of the Sr input to the basin, but in a Black Sealacking Mediterranean inflow the rivers would havecontributed more than half the water balance (theremainder being from precipitation) and all of thestrontium. Changes in relative fluxes of different riversduring the Black Sea lake phase, or in the Sr isotope ratiosof individual rivers, would have had a measurable impacton the 87Sr/86Sr of the Black Sea lake.The short, two-peaked interval of high 87Sr/86Sr corre-

sponding to the deposition of the reddish-brownclays (stages g and i, lasting from �18.3–17.0 to�16.5–15.9 kaBP cal, prior to the Bølling warming) mustrecord short-lived inputs from a high 87Sr/86Sr sourceassociated with anomalous runoff into the Black Sea.There is no present-day Sr source that can simply explainthese shifts. However, the contemporaneous occurrence ofhigh 87Sr/86Sr illite-rich reddish-brown clays suggests thatthe 87Sr/86Sr increase of Black Sea water was related to achange in the sediment load brought in by the northernrivers. Bahr et al. (2005) note the same reddish-brown claysin their cores from the Romanian margin, and with theirhigh-resolution data can identify four pulses of input (theoldest three of which are not individually distinguished inour isotopic data). An increase in the manganese content ofthe reddish-brown clays further points to decreased deepventilation of the Black Sea lake during enhanced fresh-water input. Greater erosion and river sediment loadsduring deglaciation are documented in Alpine flood plains(Hinderer, 2001), where they are linked to the abundanceof unconsolidated sediment and scarce vegetation coverwithin the drainage basin as well as higher transportcapacity of meltwater-bearing rivers. It is also possible thatsome preferential release of 87Sr during the early stages ofweathering following deglaciation (c.f. Blum and Erel,1995) contributed to the high river water 87Sr/86Sr ratios.It has been suggested that high water levels in the

Caspian Sea, caused by meltwater delivered to the Volgaand low evaporation rates resulting from cold tempera-tures, allowed spillover into the Black Sea via the ManychDepression (Chepalyga, 1984; Bahr et al., 2005).A ‘‘chocolate clay’’ deposit in the Lower Volga valleymarks the maximum transgression (early Khvalyn) of theCaspian Sea (Kroonenberg et al., 1997); the highstand(50m above modern sea level) has been tied to the mostlikely period of significant overflow into the Black Seaduring the past 25 000 years (Menabde and Svitoch, 1990).There is wide disagreement among dating methods on theages of the early Khvalyn terraces, but the most reliable14C dating points to a maximum highstand soon after15 kyr 14CBP (�18 kaBP cal) (Kroonenberg et al., 1997;Svitoch, 1999). This timing is consistent with the appear-ance of the high 87Sr/86Sr reddish-brown clay in the westernBlack Sea, though whether the Caspian and Black Seaclays are the same remains to be determined by geochem-ical analyses. The 87Sr/86Sr of the modern Caspian is too


low (0.7082; Clauer et al., 2000) to account for the changesseen in the Black Sea during the deposition of the reddish-brown clays, though the same processes that produced themore radiogenic runoff to the Black Sea likely alsoinfluenced the isotopic composition of the Caspian.Whether the high 87Sr/86Sr clays on the western BlackSea margin are sourced from the Caspian or northwesternBlack sea rivers, the explanation for the 87Sr/86Sr changemust lie in the change in river sediment load (affecting thedissolved Sr). Furthermore, the excess input (runoff+pre-cipitation) over output (evaporation) of the region,reflected by high Caspian Sea levels and the overall increaseinput of northern source waters to the Black Sea, wouldhave raised the level of the Black Sea, probably to itsoutflow. Such a base level rise is recorded on the Black Seashelf by channel infilling commonly seen in seismic profiles(Ryan et al., 2003) and inferred from sediment cores (Skibaet al., 1975), though sediment from these channels are notyet reliably dated (Kaplin and Selivanov, 2004).

The timing of the high 87Sr/86Sr pulses is consistent withchanges in the European geomorphological systems andwarmer climate occurring �3 ka prior to the Bøllingwarming. These include retreat of valley glaciers in theAlps (C. Schluchter, personal communication), and theincision of braided stream systems in the upper reaches ofrivers feeding the Black Sea indicating marked increase inrunoff (Kalicki and Sanko, 1998; Huhmann et al., 2004).Moraines indicate that the leading edge of the Scandina-vian ice sheet advanced into the watersheds of the DnieprRiver during the last glacial period (Kalicki and Sanko,1998) and rerouted into the Black Sea and Caspian smallerrivers that had formerly drained north to the Baltic andArctic Seas (Kvasov, 1968; Grosswald, 1980). The pulse ofclay delivery to the Black Sea thus can be linked to a majormelting phase of European ice (c.f. Denton et al. 1999).The �1% drop in d18O within the reddish-brown claylayers marks the input of 18O-depleted meltwaters to theBlack Sea.

The meltwater event in the Black Sea occurs close in timeto Heinrich Event 1 (H1), the last major ice-rafting event ofthe last glacial period in the North Atlantic, and estimatedat �17 kaBP cal (c.f. Hemming 2004). The strong doublepeak of the Black Sea 87Sr/86Sr record (Fig. 4) mirrors thedouble peak in ice-rafted detritus observed in highsedimentation rate records of H1 from the Nordic seas(Elliot et al., 2001). Although the uncertainty in our agemodel does not allow a precise determination of lead–lagrelationships, several authors have argued for a lead of theEuropean deglaciation relative to the North Americandeglaciation (e.g., Scourse et al., 2000), and have suggestedthat the enhanced flux of freshwater into the NorthAtlantic from melting of the large Scandinavian ice sheetscould have been the mechanism for the decrease in deepwater formation that cooled the north Atlantic during H1(Knutz et al., 2002). The meltwater pulses seen in the BlackSea record indicate that the European ice sheet responseoccurred in the continental area as well as the coastal

marine areas. In addition, it suggests that the forcing thatcaused the disintegration of the ice was a widespreadphenomenon beyond a coastal area response to sea levelrise (Jones and Keigwin, 1988; Grousset et al., 2000).Thus, the Black Sea geochemistry and stratigraphy

documents the pre-Bølling history of deglaciation inEurope, including two major pulses of melting that flushedfreshwater through the basin. This early European warm-ing is consistent with rapidly increasing Northern Hemi-sphere summer insolation beginning around 20 kaBP cal(Bard et al., 1990; Berger and Loutre, 1991), and suggests agreater sensitivity of the smaller volume European icesheets than the Laurentide ice sheet (Clark and Mix, 2002)to orbital forcing.

6.3. Climatologically controlled changes after the Bølling

warming

Following the intense melting of the Fennoscandian icesheet and alpine glaciers between �18 and 16 kaBP cal, theice had retreated far enough that it was no longer asignificant source of water to the Black Sea. Furthermore,no significant freshwater input to the Nordic seas occursafter the onset of the Bølling (Jones and Keigwin, 1988;Lehman et al., 1991; Koc and Jansen, 1994), with thepossible exception of a light d18O event associated with thedrainage of the Baltic ice lake (Lehman and Keigwin,1992), which would not have affected drainage to the BlackSea. Meltwater thus cannot be called on as a mechanism tobring the Black Sea to its spill point later on, to delivercontinuous downstream flow of freshwater to Marmaraand the Aegean throughout the deglacial and post-glacialphases, as some studies have suggested (Stanley andBlanpied, 1980; Lane-Serff et al., 1997; Aksu et al., 1999,2002; Hiscott et al., 2002).The increase in d18O of mollusk shells following the

Bølling warming (Fig. 4, stage e) occurs along with theappearance of inorganic carbonate in core BLKS9810 fromthe western shelf. The low 87Sr/86Sr at this time rules outthe possibility of marine inflow being responsible for theincrease in the d18O. The inorganic carbonate peaks recorda much lighter d18O signature (�6.5 to �8.9%) (Majoret al., 2002) than do mollusks (�4.3 to �2.5%) of the sameage. Carbonate precipitation is common in temperatelakes, where it commonly results from seasonal orpermanent supersaturation of surface waters. It is oftenrelated to biological activity (Kelts and Hsu, 1978), or‘‘chemical delta’’ formation in the vicinity of bicarbonate-rich river outflows, such as the Danube (Shimkus andTrimonis, 1974), mixing with lake waters (Hardie et al.,1978). The relatively light d18O of the carbonate peakssuggests precipitation in warmer and/or fresher surfacewater. Two factors, either separately or in combination,could account for the progressive shift in the mollusk d18Orecord: (i) a change in the meteoric water compositionlinked to northerly migration of the climate zones(Rozanski et al., 1992; von Grafenstein et al., 1999),


or (ii) a change in the hydrologic balance of the basin,namely an increase in evaporation relative to precipitation(Buchardt and Fritz, 1980). In the case of a change in themeteoric water composition, we would expect a gradualchange in the d18O of the Black Sea, reflecting the relativelysmall influx rate compared to the large reservoir volume. Inthis context, Central European lakes show a 4% shifttoward heavier d18O at the time of the Bølling warming(von Grafenstein et al., 1999), which is consistent with theprogressive change from ��7% to �–3% between 15 and9.4 kaBP cal in the Black Sea. This suggests that much ofthe d18O shift in Black Sea biogenic carbonates may beexplained by changes in meteoric water oxygen isotopecompositions, although evaporative processes could alsoplay a role in changing the d18O of the basin (Major et al.,2002; Bahr et al., 2005).

Nevertheless, there is abundant evidence that the waterbalance of the Black Sea basin must have reversedfollowing the meltwater-driven highstand. The presenceof preserved shoreline deposits on the Black Sea outer shelfand shelf-wide and upper slope unconformities post-datingthe meltwater pulse (Ryan et al., 2003) indicate a late stagelowstand, and gypsum-rich sediments in the Marmara Sea(Stanley and Blanpied, 1980) indicate evaporative condi-tions in that basin. There is no continuous sedimentaryrecord of a late glacial to early Holocene highstand in anyof the three survey areas of the northern and westerncontinental Black Sea shelves. Bahr et al. (2005) note alarge (at least four-fold) upslope increase in sedimentationrates in the lacustrine units above the reddish-brown claylayers which they ascribe to a more proximal location oftheir core sites on the continental slope to the DanubeDelta during times of lowered Black Sea level. Conversely,the older red clay layers do not show such a dramaticincrease in thickness with decreasing depth, even thoughterrigenous input is very high (Bahr et al., 2005), which isconsistent with a more hemipelagic sedimentation on thecontinental slope during high Black Sea lake level. Finally,87Sr/86Sr barely changes during the deposition of thecarbonate peaks (Fig. 4, stages c and e), in contrast to thelarge shifts associated with the reddish-brown clays (Fig. 4,stages g and i). Either decreased runoff (such that riverinput was insufficient to impart significant changes to the87Sr/86Sr ratio of the Black Sea) or a change in the averageriver source to an intermediate value of between 0.7088 and0.7089 could explain the minimal 87Sr/86Sr shift. However,the 87Sr/86Sr of the clays suggest that if anything the87Sr/86Sr ratio of river input to the Black Sea was lowerthan at the LGM. In all, the observations are bestexplained by a high Black Sea lake level fed by anomalousrunoff during the meltwater event(s), followed by adecreasing Black Sea lake level with reduced freshwaterinput in the warmer climatic period of the Bølling/Allerød.

In contrast to stages c and e, respectively the warmperiods of the Preboreal and Bølling/Allerød, the coolYounger Dryas period (stage d) shows a marked decreasein 87Sr/86Sr and Sr/Ca back toward compositions typical of

the LGM (Figs. 4 and 5d). The Younger Dryas-agedsediments are most closely associated with the Dreissena-rich muds of the typical ‘‘Neoeuxine’’ deposits of the BlackSea (Kuprin et al., 1974; Shimkus et al., 1978; Shopovet al., 1992). Shcherbakov et al. (1978) map the extent of atransgression of such Neoeuxine deposits up to the 30mbslisobath, though shallower than 80mbsl it is identified onlyas a winnowed shell gravel (the matrix of the ‘‘shell hash’’).Although pollen assemblages throughout Europe identifythe Younger Dryas as the driest period of the last glacialcycle (Rossignol-Strick, 1995), nonetheless it appears thatthe Black Sea level may have been rising through this timeinterval, perhaps as high as its outflow at 30mbsl. A latePleistocene highstand of the Caspian (Late Khvalyn;Kroonenberg et al., 1997), the last major transgressionbefore the drop to low Holocene water levels, reflects apositive regional water balance associated with the latestglacial period. Freshening of the Caspian and Black Seasduring the Younger Dryas is also indicated by the presenceof the diatom Stephanodiscus astrea (Shimkus et al., 1973).Higher Black Sea lake levels during the Younger Dryasmay then be related to greatly reduced evaporation ratesbecause of cold temperatures and ice cover, even thoughprecipitation and runoff may have also been reduced.The late glacial sequences across the Black Sea

continental shelf and slope are truncated by an erosionalsurface (Khrischev and Georgiev, 1991; Major et al., 2002;Ryan et al., 2003). The sediment overlying the erosionalsurface has a euryhaline faunal assemblage at shallowerdepths or sapropelic sediments at depths greater than�200mbsl. The presence of highly CaCO3-rich sedimentsin Preboreal time (stage c) suggests an environmentcomparable to that of the Bølling–Allerød warm period,i.e., increased evaporation and a lowering of Black Sealevel in the period just prior to the marine incursion.

7. Conclusions

Oxygen and strontium isotope ratios, in combinationwith Sr/Ca and lithological and stratigraphic considera-tions, yield a deglacial history of environmental changes inthe Black Sea from the LGM through the present day.Marine influence on Black Sea water chemistry began�9.4 kaBP cal, pointing to a Bosporus sill depth of �30mbelow modern sea level at the time of reconnection. Themarine incursion triggered changes in mollusk assemblagesand the first faunal colonization atop a shelf-wideunconformity. The concurrence of abrupt increases in87Sr/86Sr and d18O to marine compositions in molluskssampled from a thin shell gravel horizon atop theunconformity suggests that the incursion of Mediterraneanwater may have been the causative agent for the drowningof the lake margin. The date of 7.5 kaBP cal reported byRyan et al. (1997) is not the marine incursion itself, buttime when salinity rose to a sufficient threshold for anintroduction of euryhaline taxa following the first brackishspecies.


Geochemical variability in Black Sea biogenic carbo-nates since 25 kaBP cal and prior to 9.4 kaBP cal are theresult of changes in freshwater sources and do not reflectany marine influence. Influx of meltwater to the Black Seafrom the disintegrating Eurasian glacial ice resulted in ahighstand of the Black Sea lake early in the deglaciation.By 15 kaBP cal, the retreat of the Eurasian ice sheets andAlpine glaciers removed meltwater sources from the BlackSea drainage. The Sr and O isotope data from Black Seasediments and shells are consistent with sedimentologicaland morphological evidence of a lowstand, which wouldrule out continuous downstream delivery of freshwater toMarmara and the Aegean, and consistent with rapid andearly salinification of the Sea of Marmara at �14 kaBP cal(C- agatay et al., 2000; Sperling et al., 2003).

These source changes are correlated to major climateepisodes recognized throughout the northern hemisphere,such as the deglaciation of northern Europe and the Alps,Heinrich Event H1, the Bølling–Allerød and Preborealwarmings, and the Younger Dryas cold episode. The inputof freshwater into the north Atlantic from the retreat ofEuropean ice may indicate a link between the timing of themeltwater pulse into the Black Sea and Heinrich Event H1.

If the association of calcite precipitation before and afterthe Younger Dryas along with increasing d18O and acorrelation in time with lowstand shorelines around theBlack Sea’s lake is indicative of excess evaporation oversupply, then the Black Sea behaved similarly to theCaspian in response to regional climate warming (i.e.,expanded when cool, shrunk when warm). Highstands inwarm periods may then be limited to special times of eitherrapid melting of ice sheets or times of marine connection asglobal sea level approached its maximum elevation.

Acknowledgments

The cores were collected aboard the French Le Suroit aspart of the BLaSON program with the collaboration ofIFREMER and GEOECOMAR in Bucharest, and aboardthe R/V Aquanaut of the Southern Branch of the P. P.Shirshov Insitute of Oceanology. We thank Namik C-agatay, Franc-ois Guichard, Wallace Broecker, TaroTakahashi, Walter Pitman, Peter DeMenocal, JeanLynch-Stieglitz, Richard Fairbanks, Kazimeras Shimkus(now deceased), and Nicolae Panin for helpful discussions,as well as Frank Lamy and an anonymous reviewer fortheir thoughtful and constructive reviews. Post-cruiseanalyses were supported by grant OCE 97-11320 of theUS National Science Foundation and a Grant in aid ofresearch from Sigma Xi (to Major).

Appendix A. Supplementary Data

Supplementary data associated with this article can befound in the online version at doi:10.1016/j.quascirev.2006.01.032.

References

Aksu, A.E., Hiscott, R.N., Yasar, D., 1999. Oscillating Quaternary water

levels of the Marmara Sea and vigorous outflow into the Aegean Sea

from the Marmara Sea–Black Sea drainage corridor. Marine Geology

153, 275–302.

Aksu, A.E., Hiscott, R.N., Kaminski, M.A., Mudie, P.J., Gillespie, H.,

Abrajano, T., Yasar, D., 2002. Last glacial-Holocene paleoceanogra-

phy of the Black Sea and Marmara Sea: stable isotopic, foraminiferal

and coccolith evidence. Marine Geology.

Arkhangel’skiy, A.D., Strakhov, N.M., 1938. Geological structure and

history of the evolution of the Black Sea. Izvestiya Akademii Nauk

SSSR 10, 3–104.

Arthur, M.A., Dean, W.E., 1998. Organic-matter production and

preservation and evolution of anoxia in the Holocene Black Sea.

Paleoceanography 13, 395–411.

Atanassova, J. (Ed.), 1995. Palynological Data of Three Deep Water

Cores from the Western Part of the Black Sea. Advances in Holocene

Paleoecology in Bulgaria. Pensoft, Sofia pp. 68–93.

Bahr, A., Lamy, F., Arz, H., Kuhlmann, H., Wefer, G., 2005. Late glacial

to Holocene climate and sedimentation history in the NW Black Sea.

Marine Geology 214, 309–322.

Bailey, T.R., McArthur, J.M., Prince, H., Thirwall, M.F., 2000.

Dissolution methods for strontium isotope stratigraphy: whole rock

analysis. Chemical Geology 167, 313–319.

Ballard, R.D., Coleman, D.F., Rosenberg, G.D., 2000. Further evidence

of abrupt Holocene drowning of the Black Sea shelf. Marine Geology

170, 253–261.

Bard, E., Hamelin, B., Fairbanks, R.G., Zindler, A., 1990. Calibration of

the 14C timescale over the past 30,000 years using mass spectrometric

U-Th ages from Barbados corals. Nature 345, 405–410.

Berger, A., Loutre, M.F., 1991. Insolation values for the climate of the last

10 million years. Quaternary Science Reviews 10, 297–317.

Birck, J.-L., 1986. Precision K–Rb–Sr isotopic analysis; applications to

Rb–Sr chronology. Chemical Geology 56, 73–83.

Blum, J.D., Erel, Y., 1995. A silicate weathering mechanisms linking

increases in marine 87Sr/86Sr with global glaciation. Nature 373,

415–418.

Broecker, W., Peng, T.-H., 1982. Tracers in the Sea. Lamont-Doherty

Geological Observatory, Palisades, NY, 690pp.

Bruyevich, S.V., 1952. Increasingly fresh waters under recent sediments of

the Black Sea. Akademii Nauk SSSR Doklady 84, 575–577.

Buchardt, B., Fritz, P., 1980. Environmental isotopes as environmental

and climatological indicators. In: Fritz, P., Fontes, J.C. (Eds.),

Handbook of Environmental Isotope Geochemistry. Elsevier, Am-

sterdam, pp. 473–504.

C- agatay, M.N., Gorur, N., Algan, O., Eastoe, C., Tchepalyga, A., Ongan,

D., Kuhn, T., Kuscu, I., 2000. Late glacial-Holocene palaeoceano-

graphy of the Sea of Marmara: timing of connections with the

Mediterranean and the Black Seas. Marine Geology 167, 191–206.

Calvert, S.E., 1990. Geochemistry and origin of the Holocene sapropel in

the Black Sea. In: Ittekkot, V., et al. (Eds.), Facets of modern

biogeochemistry. Springer-Verlag, Berlin, pp. 326–352.

Chepalyga, A.L., 1984. Inland sea basins. In: Velichko, A.A. (Ed.), Late

Quaternary Environments of the Soviet Union. University of

Minnesota Press, Minneapolis, MN, pp. 229–246.

Clark, P.U., Mix, A.C., 2002. Ice sheets and sea level of the Last Glacial

Maximum. Quaternary Science Reviews 21, 1–7.

Clauer, N., Chaudhuri, S., Toulkeridis, T., Blanc, G., 2000. Fluctuations

of Caspian Sea level: beyond climatic variations? Geology 28,

1015–1018.

Cox, J.M., Faure, G., 1974. Isotope composition of strontium in

carbonate phase of two cores from Black Sea. In: Degens, E.T., Ross,

D.A. (Eds.), The Black Sea—Geology, Chemistry and Biology.

American Association of Petroleum Geologists, Tulsa, OK, pp.

566–569.

Denton, G.H., Heusser, C.J., Lowell, T.V., Moreno, P.I., Anderson, B.G.,

Heusser, L.E., Schluchter, C., Marchant, D.R., 1999. Interhemispheric




linkage of paleoclimate during the last glaciation. Geografiska Annaler

81, 107–153.

Deuser, W.G., 1972. Late-Pleistocene and Holocene history of the Black

Sea as indicated by stable-isotope studies. Journal of Geophysical

Research 77, 1071–1077.

Elliot, M., Labeyrie, L.D., Dokken, T., Manthe, S., 2001. Coherent

patterns of ice-rafted debris deposits in the Nordic regions during the

last glacial (10–60 ka). Earth and Planetary Science Letters 194,

151–163.

Fairbanks, R.G., 1989. A 17,000-year glacio-eustatic sea level record:

influence of glacial melting rates on the Younger Dryas event and

deep-ocean circulation. Nature 342, 637–642.

Faure, G., 1998. Principles and Applications of Geochemistry. Prentice

Hall, Upper Saddle River, NJ, 600pp.

Fleming, K., Johnston, P., Zwartz, D., Yokoyama, Y., Lambeck, K.,

Chappell, J., 1998. Refining the eustatic sea-level curve since the Last

Glacial Maximum using far- and intermediate-field sites. Earth and

Planetary Science Letters 163, 327–342.

Gorur, N., C- agatay, M.N., Emre, O., Alpar, B., Sakinc- , M., Islamoglu,

Y., Algan, O., Erkal, M., Akkok, R., Karlik, G., 2001. Is the abrupt

drowning of the Black Sea shelf at 7150 yrBP a myth? Marine Geology

176, 65–73.

Grootes, P.M., Stuiver, M., White, J.W.C., Johnsen, S., Jouzel, J., 1993.

Comparison of oxygen isotope records from the GISP2 and GRIP

Greenland ice cores. Nature 366, 552–554.

Grosswald, M.G., 1980. Late Weichselian ice sheet of Northern Europe.

Quaternary Research 13, 1–32.

Grosswald, M.G., 1998. Late-Weichselian ice sheets in Arctic and Pacific

Siberia. Quaternary International 45–46, 3–18.

Grosswald, M.G., Hughes, T.J., 2002. The Russian component of an

Arctic ice sheet during the Last Glacial Maximum. Quaternary Science

Reviews 21, 121–146.

Grousset, F.E., Pujol, C., Labeyrie, L., Auffret, G., Boelaert, A., 2000.

Were the North Atlantic Heinrich events triggered by the behavior of

the European ice sheets? Geology 28, 123–126.

Guichard, F., Carey, S., Arthur, M.A., Sigurdsson, H., Arnold, M., 1993.

Tephra from the Minoan eruption of Santorini in sediments of the

Black Sea. Nature 363, 610–612.

Gunnerson, C.G., Ozturgut, E., 1974. The Bosporus. In: Degens, E.T.,

Ross, D.A. (Eds.), The Black Sea: Geology, Chemistry, and Biology.

American Association of Petroleum Geologists, Tulsa, OK, pp.

99–114.

Hardie, L.A., Smoot, J.P., Eugster, H.P., 1978. Saline Lakes and their

Deposits: A Sedimentological Approach. Special Publications of the

International Assembly of Sedimentology, pp. 7–41.

Hay, B.J., Arthur, M.A., Dean, W.E., Neff, E.D., Honjo, S., 1991.

Sediment deposition in the Late Holocene abyssal Black Sea with

climatic and chronological implications. Deep-Sea Research 38,

S1211–S1235.

Hemming, S.R., 2004. Heinrich events: massive late Pleistocene detritus

layers of the north Atlantic and their global climate imprint. Reviews

of Geophysics 42, RG1005.

Henderson, G.M., Martel, D.J., O’Nions, R.K., Shackleton, N.J., 1994.

Evolution of seawater Sr-87/Sr-86 over the last 400-ka: the absence of

glacial interglacial cycles. Earth and Planetary Science Letters 128.

Hinderer, M., 2001. Late Quaternary denudation of the Alps, valley and

lake fillings and modern river loads. Geodinamica Acta 14, 231–263.

Hiscott, R.N., Aksu, A.E., Yasar, D., Kaminski, M.A., Mudie, P.J.,

Kostylev, V., MacDonald, J.C., Lord, A.R., 2002. Deltas south of the

Bosporus record persistent Black Sea outflow to the Marmara Sea

since �10 ka. Marine Geology.

Huhmann, M., Kremenetski, K.V., Hiller, A., Bruckner, H., 2004. Late

Quaternary landscape evolution of the upper Dnister valley, western

Ukraine. Palaeogeography Palaeoclimatalogy Palaeoecology 209,

51–71.

Ingram, B.L., Sloan, D., 1992. Strontium isotopic composition of

estuarine sediments as paleosalinity–paleoclimate indicator. Science

255, 68–72.

Israelson, C., Buchardt, B., 1999. Strontium and oxygen isotopic

composition of east Greenland rivers and surface waters: implication

for palaeoenvironmental interpretation. Palaeogeography, Palaeo-

climatalogy, Palaeoecology 153, 93–104.

Jones, G.A., Gagnon, A.R., 1994. Radiocarbon chronology of Black Sea

sediments. Deep-Sea Research 41, 531–557.

Jones, G.A., Keigwin, L.D., 1988. Evidence from Fram Strait (781N) for

early deglaciation. Nature 336, 56–59.

Kalicki, T., Sanko, A.F., 1998. Palaeohydrological changes in the Upper

Dneper Valley, Belarus, during the last 20,000 years. In: Benito, G.,

et al. (Eds.), Palaeohydrology and Environmental Change. Wiley,

Chichester, England, pp. 125–135.

Kaplin, P.A., Selivanov, A.O., 2004. Late glacial and Holocene sea level

changes in semi-enclosed seas of North Eurasia: examples from the

contrasting Black and White Seas. Palaeogeography Palaeoclimatal-

ogy Palaeoecology 209, 19–36.

Kelts, K., Hsu, K.J., 1978. Freshwater carbonate sedimentation. In:

Lerman, A. (Ed.), Lakes: Chemistry, Geology, Physics. Springer, New

York, pp. 295–323.

Khrischev, K., Georgiev, V., 1991. Regional washout on the Pleistocene–

Holocene boundary in the western Black Sea depression. Comptes

Rendus de l’Academie Bulgare des Sciences 44, 69–72.

Knutz, P.C., Hall, I.R., Zahn, R., Rasmussen, T.L., Kuijpers, A., Moros,

M., Shackleton, N.J., 2002. Multidecadal ocean variability and NW

European ice sheet surges during the last deglaciation. Geochemistry

Geophysics Geosystems.

Koc, N., Jansen, E., 1994. Response of the high-latitude Northern

Hemisphere to orbital climate forcing: evidence from the Nordic Seas.

Geology 22, 523–526.

Krom, M.D., Michard, A., Cliff, R.A., Strohle, K., 1999. Sources of

sediment to the Ionian Sea and western Levantine Basin of the Eastern

Mediterranean during S-1 sapropel times. Marine Geology 160, 45–61.

Kroonenberg, S.B., Rusakov, G.V., Svitoch, A.A., 1997. The wandering

of the Volga delta: a response to rapid Caspian sea-level change.

Sedimentary Geology 107, 189–209.

Kuprin, P.N., Scherbakov, F.A., Morgunov, I.I., 1974. Correlation, age

and distribution of the postglacial continental terrace sediments of the

Black Sea. Baltica 5, 241–249.

Kvasov, D.D., 1968. Paleohydrology of eastern Europe in late Quaternary

time, Yezhegodnkkh Chetniyakh Pamyati L.S. Berga Doklady Izd.

Nauka, Leningrad, pp. 65–81.

Lambeck, K., Bard, E., 2000. Sea-level change along the French

Mediterranean coast for the past 30 000 years. Earth and Planetary

Science Letters 175, 203–222.

Lane-Serff, G.F., Rohling, E.J., Bryden, H.L., Charnock, H., 1997. Post-

glacial connection of the Black Sea to the Mediterranean and its

relation to the timing of sapropel formation. Paleoceanography 12,

169–174.

Lehman, S.J., Keigwin, L.D., 1992. Sudden changes in North Atlantic

circulation during the last deglaciation. Nature 356, 757–762.

Lehman, S.J., Jones, G.A., Keigwin, L.D., Andersen, E.S., Butenko,

G., Ostmo, S.-R., 1991. Initiation of Fennoscandian ice-sheet retreat

during the last deglaciation. Nature 349, 513–516.

Major, C.O., 2002. Non-eustatic controls on sea-level in semi-enclosed

basins. Ph.D. Thesis. Columbia University, New York, NY, 223 pp.

Major, C.O., Ryan, W.B.F., Lericolais, G., Hajdas, I., 2002. Constraints

on Black Sea outflow to the Sea of Marmara during the last

glacial–interglacial transition. Marine Geology 190, 19–34.

Mangerud, J., Astakhov, V., Jakobsson, M., Svendsen, J.I., 2001. Huge

Ice Age lakes in Russia. Journal of Quaternary Science 16, 773–777.

Manheim, F.T., Chan, K.M., 1974. Interstitial waters of Black Sea

sediments: new data and review. In: Degens, E.T., Ross, D.A. (Eds.),

The Black Sea: Geology, Chemistry, and Biology. American Associa-

tion of Petroleum Geologists, Tulsa, OK, pp. 155–180.

McManus, J.F., Francois, R., Gherardi, J.-M., Keigwin, L.D., Brown-

Leger, S., 2004. Collapse and rapid resumption of Atlantic meridional

circulation linked to deglacial climate changes. Nature 428, 834–837.


Menabde, I.V., Svitoch, A.A., 1990. About the Character of the

Connection Between the Caspian and Black Seas in the Late

Pleistocene (in Russian), Caspian Sea: Problems of Geology and

Geomorphology. Nauka, Moscow, pp. 34–41.

Mudie, P.J., Rochon, A., Aksu, A.E., Gillespie, H., 2004. Late glacial,

Holocene and modern dinoflagellate cyst assemblages in the Aegean–

Marmara–Black Sea corridor. Review of Palaeobotany and Palynol-

ogy 256, 1–26.

Muller, G., Stoffers, P., 1974. Mineralogy and petrology of Black Sea

sediments. In: Degens, E.T., Ross, D.A. (Eds.), The Black Sea:

Geology, Chemistry, and Biology. American Association of Petroleum

Geologists, Tulsa, pp. 200–248.

Nevesskaya, L.A., 1965. Late Quaternary bivalves of the Black Sea, their

systematics and ecology. Nauka, Moscow, 176 pp.

Ostlund, H.G., 1974. Expedition ‘‘Odysseus 65’’: radiocarbon age of Black

Sea deep water. In: Degens, E.T., Ross, D.A. (Eds.), The Black Sea:

Geology, Chemistry, and Biology. American Association of Petroleum

Geologists, Tulsa, OK, pp. 127–132.

Palmer, M.R., Edmond, J.M., 1989. The strontium budget of the modern

ocean. Earth and Planetary Science Letters 92, 11–26.

Popov, G.I., 1973. New data on the stratigraphy of Quaternary marine

sediments of the Kerch Strait. Doklady Akademii Nauk SSSR 213,

84–86.

Rosenthal, Y., Katz, A., 1989. The applicability of trace elements in

freshwater shells for paleogeochemical studies. Chemical Geology 78,

65–76.

Ross, D.A., 1978. Summary of results of Black Sea Drilling. In: Degens,

E.T., Ross, D.A. (Eds.), Initial Reports of the Deep Sea Drilling

Project, vol. 42. US Gov’t Printing Office, Washington, DC,

pp. 1149–1178.

Ross, D.A., Degens, E.T., 1974. Recent sediments of the Black Sea. In:

Degens, E.T., Ross, D.A. (Eds.), The Black Sea: Geology, Chemistry,

and Biology. American Association of Petroleam Geologists, Tulsa,

OK, pp. 183–199.

Ross, D.A., Degens, E.T., MacIlvaine, J., 1970. Black Sea: recent

sedimentary history. Science 170, 163–165.

Rossignol-Strick, M., 1995. Sea–land correlation of pollen records in the

eastern Mediterranean for the glacial–interglacial transition: biostrati-

graphy vs radiometric time-scale. Quaternary Science Reviews 14,

893–915.

Rozanski, K., Araguas-Araguas, L., Gonfiantini, R., 1992. Relation

between long-term trends of oxygen-18 isotope composition of

precipitation and climate. Science 258, 981–985.

Ryan, W.B.F., Pitman III, W.C., Major, C.O., Shimkus, K., Moskalenko,

V., Jones, G., Dimitrov, P., Gorur, N., Sakinc, M., Yuce, H., 1997. An

abrupt drowning of the Black Sea shelf. Marine Geology 138, 119–126.

Ryan, W.B.F., Major, C.O., Lericolais, G., Goldstein, S.L., 2003.

Catastrophic flooding of the Black Sea. Annual Reviews of Earth

and Planetary Sciences 31, 525–554.

Schrader, H.J., 1979. Quaternary paleoclimatology of the Black Sea basin.

Sedimentary Geology 23, 165–180.

Schrag, D.P., 1999. Rapid analysis of high-precision Sr/Ca ratios in corals

and other marine carbonates. Paleoceanography 14, 97–102.

Scourse, J.D., Hall, I.R., McCave, I.N., Young, J.R., Sugdon, C., 2000.

The origin of Heinrich layers: evidence from H2 for European

precursor events. Earth and Planetary Science Letters 182, 187–195.

Shcherbakov, F.A., Babak, Y.V., 1979. Stratigraphic subdivision of the

Neoeuxinian deposits in the Black Sea. Oceanology 19, 298–300.

Shcherbakov, F.A., Kuprin, P.N., Potapova, L.I., Polyakov, A.S.,

Zabelina, E.K., Sorokin, V.M., 1978. Sedimentation on the Con-

tinental Shelf of the Black Sea. Nauka Press, Moscow, 211pp.

Shimkus, K.M., Trimonis, E.S., 1974. Modern sedimentation in Black

Sea. In: Degens, E.T., Ross, D.A. (Eds.), The Black Sea—Geology,

Chemistry and Biology. American Association of Petroleum Geolo-

gists, pp. 249–278.

Shimkus, K.M., Mukhina, V.V., Trimonis, E.S., 1973. On the role of

diatoms in Late Quaternary sedimentation in the Black Sea.

Oceanologia, Akademii Nauk CCCP 13, 1066–1071.

Shimkus, K.M., Komarov, A.V., Grakova, I.V., 1978. Stratigraphy of the

upper Quaternary deep sea sediments in the Black Sea. Oceanology 17,

443–446.

Shopov, V., Chochov, S., Georgiev, V., 1986. Lithostratigraphy of Upper

Quaternary sediments from the northwestern Black Sea shelf between

the parallels of the Cape Emine and Danube River mouth. Geologica

Balcanica 16, 99–112.

Shopov, V., Bozilova, E., Atanassova, J., 1992. Biostratigraphy and

radiocarbon data of the upper Quaternary sediments from the western

part of the Black Sea. Geologica Balcanica 22, 59–69.

Siani, G., Paterne, M., Arnold, M., Bard, E., Metivier, B., Tisnerat, N.,

Bassinot, F., 2000. Radiocarbon reservoir ages in the Mediterranean

Sea and Black Sea. Radiocarbon 42, 271–280.

Siddall, M., Rohling, E.J., Almogi-Labin, A., Hemleben, C., Meischner,

D., Schmelzer, I., Smeed, D.A., 2003. Sea-level fluctuations during the

last glacial cycle. Nature 423, 853–858.

Skiba, S.I., Scherbakov, F.A., Kuprin, P.N., 1975. On paleogeography of

the Kerch-Taman region in late Pleistocene and Holocene. Oceano-

logia 15, 862–867.

Sperling, M., Schmiedl, G., Hemleben, C., Emeis, K.-C., Erlenkeuser, H.,

Grootes, P.M., 2003. Black Sea impact on the formation of eastern

Mediterranean sapropel S1? Evidence from the Marmara Sea.

Palaeogeography, Palaeoclimatalogy, Palaeoecology 190, 9–21.

Stanley, D.J., Blanpied, C., 1980. Late Quaternary water exchange

between the eastern Mediterranean and the Black Sea. Nature 285,

537–541.

Stein, M., Starinsky, A., Katz, A., Goldstein, S.L., Machlus, M.,

Schramm, A., 1997. Strontium isotopic, chemical, and sedimentologi-

cal evidence for the evolution of Lake Lisan and the Dead Sea.

Geochimica et Cosmochimica Acta 61, 3975–3992.

Stoffers, P., Muller, G., 1978. Mineralogy and lithofacies of Black Sea

sediments, Leg 42B Deep Sea Drilling Project. In: Usher, J.L., Supko,

P. (Eds.), Initial Reports of the Deep Sea Drilling Project, vol. 42, Part

2. Ocean Drilling Program, College Station, TX, pp. 373–411.

Stuiver, M., Reimer, P.J., 2004. CALIB 4.4.

Stuiver, M., Reimer, P.J., Bard, E., Beck, J.W., Burr, G.S., Hughen, K.A.,

Kromer, B., McCormac, F.G., van der Plicht, J., Spurk, M., 1998.

INTCAL98 radiocarbon age calibration. Radiocarbon 40, 1041–1083.

Svitoch, A.A., 1999. Caspian Sea level in the Pleistocene: hierarchy and

position in the paleogeographic and chronological records. Oceanol-

ogy 39, 94–101.

Turekian, K.K., Armstrong, R.L., 1960. Magnesium, strontium, and

barium concentrations and calcite–aragonite ratios of some recent

molluscan shells. Journal of Marine Research 18, 133–151.

Unluata, U., Oguz, T., Latif, M.A., Ozsoy, E., 1990. On the physical

oceanography of the Turkish Straits. In: Pratt, L.J. (Ed.), The Physical

Oceanography of Sea Straits. Kluwer, Deventer, The Netherlands, pp.

25–60.

von Grafenstein, U., Erlenkeuser, H., Brauer, A., Jouzel, J., Johnsen, S.J.,

1999. A Mid-European decadal isotope-climate record from 15,500 to

5000 years BP. Science 284, 1654–1657.

Wall, D., Dale, B., 1974. Dinoflagellates in the late Quaternary deep-water

sediments of the Black Sea. In: Degens, E.T., Ross, D.A. (Eds.), The

Black Sea–Geology, Chemistry and Biology. American Association of

Petroleum Geologists, Memoirs, Tulsa, pp. 364–380.

Wefer, G., Berger, W.H., 1991. Isotope paleontology: growth and

composition of extant calcareous species. Marine Geology 100,

207–248.

Yaltirak, C., Sakinc, M., Aksu, A.E., Hiscott, R.N., Galleb, B., Ulgen,

U.B., 2002. Late Pleistocene uplift history along the southwestern

Marmara Sea determined from raised coastal deposits and global sea-

level variations. Marine Geology 192, 283–305.

Zubakov, V.A., 1988. Climatostratigraphic scheme of the Black Sea

Pleistocene and its correlation with the oxygen isotope scale and glacial

events. Quaternary Research 29, 1–24.

Documents

The co-evolution of Black Sea level and composition ...intranet.geoecomar.ro/rchm/wp-content/uploads/... · The co-evolution of Black Sea level and composition through the last deglaciation