Sediment deposition in the Late Holocene abyssal Black Sea with climatic and chronological implications

Deep-Sea Research.Vol. 38. Suppl 2. pp. SI211-S1235. 1991.Printed in GreatBntain.

019~149191 $300 + 0.00© 1991 Pergamon Press pic

Sediment deposition in the Late Holocene abyssal Black Sea withclimatic and chronological implications

BERNWARD J. HAy,*t MICHAEL A. ARTHUR,:j: WALTER E. DEAN,§ ERIC D.NEFF:j: and SUSUMU HONJO*

(Received 12 January 1990; in revised form 31 January 1991; accepted 25 February 1991)

Abstract-The temporal sedimentary patterns in the Late Holocene central eastern and westernBlack Sea are very similar. The sedimentary history was most visibly affected by the coccolithophorid species Emiliania huxley; which briefly invaded the Black Sea for the first time ("FirstInvasion Period"), nearly disappeared again shortly afterwards ("Transition Sapropel"), butreturned permanently several centuries later ("Final Invasion Period"). The temporary neardisappearance of E. huxleyi was probably caused by a temporary drop in salinity. Accumulation ofE. huxley; was on average about 40% higher in the western than in the eastern Black Sea. Highestcoccolithophorid production occurred basin-wide during part of the Little Ice Age. The accumulation of terrigenous matter was generally higher in the eastern than in the western Black Sea byabout 20%.

INTRODUCTION

THE Black Sea is the largest anoxic environment in the world. However, because of thedelicate balance between the inflow of saline Mediterranean water and fresh river water itis very sensitive to change. Although a very controversial issue, one such change may bethe postulated shallowing of the oxic-anoxic interface over the last few decades due toreduction of river influx from the Soviet Union (e.g. BRYANTSEV et al., 1988; MURRAyet al.,1989). Reconstruction of past changes in the biogeochemical conditions in the Black Seaare possible with high temporal resolution due to the preservation of the seasonal particleflux as laminated sediments across the basin. These laminations provide excellent basinwide time horizons and, should they indeed represent annual varves, provide a detailedtime scale.

The most notable event in the history of the Late Holocene Black Sea was the changefrom oxic to anoxic conditions at the sediment-water interface at about the time of thePostglacial Climatic Optimum (e.g. Ross and DEGENS, 1974; DEGENS etal., 1980a). Atthistime, the gradual rise in global sea level in post-glacial times led to incursion of salineMediterranean water into the Black Sea through the Bosporus, creating a pycnoclinewhich reduced mixing between the fresh river inflow and the dense Mediterranean water.

*Woods Hole Oceanographic Institution, Woods Hole, MA 02543, U.S.A.tPresent address: Louis Berger & Associates, Inc., 303 Bear Hill Road, Waltham, MA 02154, U.S.A.*School of Oceanography, University of Rhode Island, Narragansett, RI 02882, U. S.A.§U.S. Geological Survey, Federal Center, MS 939, Denver, CO 80225, U.S.A.

S1211

S1212 B. J. HAY et al.

26° E 28° 30° 32° 34° 36° 38° 40° 42°Fig. 1. Map showing the locations of cores Be21 (43°0S'N, 32°02'£) and BesS (42°4S'N.37°35'N). The locations of these sites correspond to sediment trap sites BSKI and BSK3,

respectively (HAY and HONJ o , 1989).

Continued inflowof Mediterranean water increased the salinity in the surface water of theBlack Sea to about 180/00 at present.Given the comparatively rapid salinity change in the mack Sea in the Late Holocene,the biogeochemistry also changed rapidly. One of the most significant events occurredwhen Emiliania huxleyi invaded the Black Sea a few millennia ago (more detaileddiscussion on ages below) . The establishment of E. huxleyi defines the boundary betweenthe older sapropel (Unit II) and the more recent coccolith ooze (Unit I) (Ross and

DEGENS, 1974). Prior to the invasion, the planktonic flora was dominated by diatoms anddinoflagel1ates (WALL and DALE, 1974). The extensive core record from the R.Y. Knorrcruises in 1988documents that the E. huxleyi invasion was not a one-time occurrence, butrather required two major attempts. Currently. E. huxleyi is the overwhelmingly dominantcoccolithophorid species in the Black Sea.Influences controlling the biogeochemistry of the Black Sea are the proximity to landmasses, the extreme restriction of supply of seawater, the large drainage area of the riversthat flow into the mack Sea , and the active climatic variability in Eurasia . Those influencesmost likely left their imprint in the laminated sedimentary sequences. By correlating thelaminated sequences of Unit I precisely across the entire abyssal basin of the Black Sea, itis possible to investigate the temporal as well as the regional variability. In this paper, wefocus on the variability in terrigenous and biogenic sedimentation over the last severalthousand years (Unit I and uppermost Unit II). The samples are derived from two

representative cores from the centers of the eastern and western Black Sea (Fig. 1). Thediscussion integrates published climatological and geographical observations from this

Sediment deposition in the Late Holocene Black Sea S1213

region to understand better the forcing mechanisms that influenced the depositionalhistory in the Black Sea.

METHODS

Sediment cores were collected during the R V. Knorr Black Sea cruise 134-8, leg 1,1988. Details about core collection and sampling are described in HONJO et al. (1988).Sediment samples from the two sites discussed in this study (Fig. 1) were collected with abox corer with a surface area of 50 x 50 em and a depth of 50 cm. The box cores weresubcored immediately after core recovery by gently inserting thin-walled plastic tubes of10em diameter. The subcores were then stored in a permanently cooled refrigerator van ata temperature of 9°C, the ambient bottom water temperature in the Black Sea. Onesubcore was split within a few hours of collection for shipboard X-ray radiography,photography and description. Subcores used for the analyses described in this paper werenot opened until the refrigerated van returned with the RV. Knorr to Woods Hole.

The X-ray radiographs taken on-board the RV. Knorr allowed good identification ofthe individual black and white laminae in Unit I (NEFF et al., in preparation). Accordingly,Unit I in the subcores was continuously sampled at intervals of25 black and white laminaecouplets, identified on the radiographs (Fig. 2). (The laminations were later recountedmore precisely using thin sections, which resulted in different, generally higher number ofcouplets for each sample.) Unit II is also laminated, but the laminae are very thin and notas easily distinguishable. Therefore, Unit II was sampled at 0.5 ern intervals.

We redefined the boundary between Unit I and Unit II. Based on cores collected duringthe R. V. Atlantis cruise in 1969, Ross and DEGENS (1974) placed the Unit lIII boundary atthe base of the permanent occurrence of E. huxleyi. This boundary corresponds to the topof sample 36 (Fig. 2). However, we believe that because Unit I was defined as alithostratigraphic unit, the base of this unit should start with the first occurrence of theconditions characteristic for the unit. By this criterion, we define the boundary betweenUnit I and Unit II as the base ofthe first invasion of E. huxleyi, which occurred at the sametime basin-wide (base of sample 43), as seen in all cores. The period of the first appearanceof E. huxleyi willbe referred to as "First Invasion Period"; the period of the return to moresapropelic deposition with high terrigenous matter concentrations will be referred to as"Transition Sapropel" (Fig. 2). Both periods were previously part of Unit II (or sapropel).The period from the time of permanent return of E. huxleyi to the present, as originallydefined by Ross and DEGENS (1974), will be referred to as "Final Invasion Period".

In addition to the cores, one bulk sample was analysed from the uppermost fluffy layerfrom the top of BC21 and one sample from the core top immediately below the fluff layer(0---2 em). These samples were collected onboard ship immediately after recovery of thebox core.

Elemental analyses were carried out by Inductively Coupled Plasma Atomic EmissionSpectrometry (ICPES) as described in FLOYD et al. (1980). A 30 mg sample was accuratelyweighed into a graphite crucible. The organic matter was removed by heating the crucibleat 550°Cin a muffle furnace for 2 h. After cooling, 210 mg LiB02 were added to the samplein the crucible. The mixture was thoroughly mixed with a clean stainless steel spatula andthen heated at 950°C in the muffle furnace for 15 min. The resulting liquid bead wasimmediately dissolved in a Teflon beaker with 30 ml of 4% nitric acid after removingthe crucible from the furnace. The wavelengths used for element identification were

51214 B. J. HAY et al.

422.67 nm (Si), 396.14 nm (AI), 337.26 nm (Ti) and 212.41 nm (Ca). As standards,synthetic aqueous solutions in a matrix of 0.56% w/v LiB02 in 4% nitric acid were used.The values of 10replicate analyses of the Si, Ti , AI and Ca concentration were within ±2%of the values of the first analyses .

The concentration of E. huxleyi in the sediment was estimated using the Ca:AI ratio ofUnit II. In Unit II, the occurrence of coccoliths in the sediment is limited to reworkedUpper Cretaceous and Eocene species supplied to the basin floor by river discharge andcoastal abrasion processes (BUKRY et al. , 1970). Holocene coccoliths are essentially absent ,except for a thin lamina of Braarudosphaera bigelowi about 10em below the First InvasionPeriod (Ross and DEGENs, 1974). This lamina was observed in core 1474 of the R.Y.Atlantis cruise in 1969, which corresponds to the location of core BC55. Similarly, thecontribution of calcium in the form of chemogenic carbonate is considered low (TRIMONIS,1974). Assuming that the Ca:Al ratio in Unit II is indeed representative of the terrigenouscarbonate fraction , and assuming that this is invariant in time and did not changesignificantly during the deposition of Unit I, it is possible to calculate the concentration ofE. huxleyi in Unit I by subtracting the assumed background detrital carbonate from thetotal carbonate . In Unit II of core BC21, the average Ca:AI ratio was0.977 (n =21; S.D. =0.20); in core BCSS it was 1.217 (n = 10; S.D. = 0.20). The concentration of E. huxleyicoccoliths for each sample of each core is then determined as

coccolith (%) = {Ca(% )Unill sample - {Ca/AIUnil II X Al (%) Unit I sample}] X 2.5,

where 2.5 is the conversion factor from Ca to CaC03•

The average ratio of detrital non-carbonate material (wt%) to Al (wt%) is 17.3 (ARTHURet al. , in preparation) . Detrital material was determined indirectly by subtracting themeasured carbonate and organic carbon contents from 100%. Carbonate and organiccarbon were determined by the coulometric method (ARTHUR et al., in preparation). Forcomparison, the ratio of terrigenous matter (wt%) to AI(wt%) in average shale is 18(WEDEPOHL,1967) .

Accumulation rates of the major sediment components were calculated using theelement concentrations, bulk density and laminae counts as data input. Bulk densitysamples were collected immediately after the cores were opened in the laboratory (N EFF etal., in preparation). Ages were based on counts of black and white laminae couplets fromthin section, assuming that one couplet represents one year. This assumption is presentlycontroversial and will be discussed further below. Given the uncertainty of the age datingat this time , we will limit the presentation of results to concentrations and ratios, since theyare independent of a specific time scale. Accumulation rates through time will be discussedlater.

CHRONOLOGY

The first bulk 14C ages available for Holocene Black Sea sediment were obtained fromfour cores collected during the R.Y. Atlantis cruise in 1969 (Ross and DEGENS, 1974).Ross and DEGENS (1974) measured the base of the Final Invasion Period as 3090 years BP,

the base of the First Invasion Period as 3450 years BP, and the base of Unit II as 7090 yearsBP (Fig. 3). Other ages from the carbonate and organic matter fractions in Unit I ofdifferent cores had considerable scatter. In fact, the ages obtained from the First InvasionPeriod and the beginning of the Final Invasion Period from core 1462K were identical at

Sediment deposition in the Late Holocene Black Sea S121S

Unit IFinal Invasion Period

PNl<BloomPeriod

Unit II

.-.-.-.-.! . ~

:Iu

LI') 0 LI') 0 Sampl e # LI') 0 LI') 0 LI') 00LI') LI') <0LI') N N M

1'0 " '; " "( " " ' 1" . 1, " 11" .,I I I , , I I I II 1 « ,I' I Il l !

"! ! , !

I I I I I I I I I Iii ' i

~ 8 § 8 § # of Laminae Couplets ~ 8 8v <0 -e- ~-ltl 5!

t~

I

gI

o e p t h

oV

I

ltlvt

( e m)

s,

Fig. 2. Photograph of subcores of box cores BC21 and BeSS, showing correlation of key laminae.The number of laminae couplets assign cd to the individual samples arc based on laminae countsfrom thin section (NEFF et a!., in preparation). Both cores were topped with several centimeters of a

low-density fluff layer, not shown on these photographs.


CHRONOLOGY OF LAMINATED BLACK SEA SEDIMENTS

Carbon-14 Dating lVarve CountsROSS ••d JONES . CALVERT DEGENS rt NEf F tl .1..

DEGENS.1914 \990 ".1.• \911 .l.1980. i. prep.

('/ ROSS..4 DEGENS"14

1633

1256998

50836600

1600

7450

3200

7090±180

34SO±UO

3090±140

UalllllJlI.......,

t, 0 " FlulTLa1tf

l~ll·I·II_~,~': J6 ':' boundar1l"))

;: 4 ;:; Unil unbounda.,.

Finlt:l Invuion_ Period

·c;:J

FinalInyasion

... Period....~

Fig. 3. Relationship of ages determined for the laminated Late Holocene Black Sea sediments.

3450 years BP. The scatter of the first few 14C ages could have been caused by noticeablecontributions of terrigenous organic matter and carbonate to the radiocarbon age , whichwere performed on bulk carbon in the sediment.

Revised ages for the Holocene sediments were presented by DEGENS etal. (1980a) basedon varve counts. The base of the Final Invasion Period was counted as 998 years BP and thebase of Unit II was counted as 5083 years BP in two cores from the western basin slope andthe abyssal plain of the Black Sea, respectively.

The difference between the radiocarbon ages and the varve ages was re-examined byCALVERT et al. (1987) with samples from the 1969 R.V. Atlantis cruise. The radiocarbonage of 2000 years of the available core top (presumed sediment-water interface) of agravity core from the center of the Black Sea (core 1432)was interpreted in the absence ofbetter core material, as a consistent offset of the age throughout the Holocene sediment.Radiocarbon ages of deeper samples from Units I and II suggested an age of 1600years BP

for the onset of Unit I and an age of 6600 years BP for the beginning of Unit II, after thesubtraction of the 2000 years correction. However, due to loss of the upper 20 ern duringthe coring operation (Ross and DEGENS, 1974), the top of core 1432 is stratigraphicallylocated only in the middle of Unit I. Updated 14C ages are forthcoming (CALVERT, personalcommunication), however.

Recently, JONES (1990) measured the 14C ages in more detail on the carbonate andorganic carbon fraction of total samples of our box cores , using the accelerator massspectrometer technique. JONES (1990) determined ages of 3200 years BP for the Unit IIIIboundary, and 7450 years BP using organic carbon and 14,000 years BP using carbonatecarbon for the Unit IIIIII boundary. The true core top ("fluff layer") was dated as abovezero after consideration of the bomb radiocarbon contamination (JONES, personal communication) .

Sl2lS B. 1. HAY et al.

, ,Fi",,' lrwanun Pmocl :':w.s.,~_ Finl lrrvasion Pmod

Unit I I Unit IIACES

V_",e. 0I

IJ33NEFF d al~ In prep. 1%5514 C ROSSANDDECENS, 1974 3090 34SO 709014C J0NES,lggo 3200 7450

A1wnlaum NEFF d al~ In !'"P. 1.39 I1m'I"!"u -AR, BCI ROSS AND DECENS, 1974 056 -" - 1.58 I1m'I"!"u

J0NES,lggo 1.36 -"---an 1.47 -"-Aluminum NEFF et a1~ In prep. 1.5. J/ml/"!"u -AR.BC5S ROSS AND DECENS. 1974 0.67 -" - 1.8% J/m'I"!"u

JONES,lggo 1.56 -" -.vel~ -"-

Calculated AJft u ...miDJ the ..me BOI: 110% (ROSS _ DECENS, 1974)a1umJnwn accwnul.doD ralos In the 1280 (JONES,I990)FlnallDvuioD Period u In UDit II(In yeare B.P.) BC55: 1144 (ROSS _ DECENS. 1974)

1335 (JONES,I990)

Fig. 4. Calculation of the AI accumulation rates for Unit II and the Final Invasion Period.Assuming the Al accumulation rates in the Final Invasion Period were identical to the rates in UnitII. the age for the base of the Final Invasion Period would be between about 1100and 1350years BP.

The laminations were recently recounted by NEFF et al. (in preparation) using X-rayradiographs and large thin sections. Thin sections allowed a higher resolution. specificallyin the Transition Sapropel with thinner laminae and less mineralogical contrast. A highernumber of couplets of black and white laminae could be counted from the thin sectionsthan from the radiographs. The base of the Final Invasion Period was counted from thinsections as 1256 years BP and the base of the First Invasion Period as 1633 years BP (Fig. 3).assuming coupled laminae represent varves (i.e. annual accumulation).

Another indication of the age of Unit I comes from the 210Pb activity. CRUSIUS andANDERSON (in press) examined 11 samples from the top 2 em of the laminated sedimentcolumn in a core collected at the same location as core BeSS . Based on the 210Pb activitythey calculated a mass accumulation rate of about 50 g m-

2y-I. In comparison. th~

average mass accumulation rate of BeSS from laminae counts (N EFF et al.• in preparation)was 70 g m-2 y-I for the Final Invasion Period. Thus. assuming the 210 Pb ages arerepresentative for the entire Final Invasion Period. the age for the beginning of this periodwould be about 1750 years BP. Unfortunately. the inte~valwith detectable 210pb activity isfar too short to be statistically significant for all of Unit 1.

An age of about 3000 years BP for the Unit VII boundary would imply that the averagealuminum accumulation rate in the abyssal Black Sea decreased by a factor of 2.6 from anaverage ~l accumulation rate of _;bo~lt . 1.6 ~ m-~ y-I in Unit II to an average Alacc~mulatlo.n rate of abou~ O.6.g m y 10 Unit I (FIg. 4) . Certainly, periodic changes interngenous input ~ue t~ cllll~atlccha?ges .are common in the Black Sea region and typicallywell-documented 10 climatic and historic records sometimes even f \ 1. . • or annua unusuaclimatic events. It therefore should be expected that a substannat d . t ." ecrease In erngenousinput by a factor of 2.61astmg several millennia is reflected in id f li .h· . . I a WI e range 0 c imatic and

istonc records from the large drainage area of the rivers draini' h BI k Srecords. however. were not found In f . . . ng Into t e ac ea . Such

. act , It IS more likely that the riverine sediment load


increased rather than decreased due to increased agricultural activity over time (DEGENS etal. , 1980b). In contrast, with an age of 1633years sr-tor the Unit IIII boundary , the averageAl accumulation rate in Unit I was on average 1.5 g m? y-l, similar to the average Alaccumulation rate in Unit II (Fig. 4). Terrigenous matter constitutes about two-thirds ofthe abyssal sediment in Unit II and about one quarter of the sediment in Unit I (HAY,1988).

The Al accumulation rates discussed above and in Fig. 4 were determined as follows.There appears to be good agreement on the length of deposition of Unit II (from the UnitIII/II boundary to the base of the First Invasion Period) at about 4000 years. Ross andDEGENS (1974) estimated 3640years, JONES (1990) estimated 4250 years, and DEGENS et al.(1980a) estimated a time span of 4085 years from the Unit IIlIII boundary to the base of theFinal Invasion Period. The average accumulation rates for Units I and II at core sites BC21and BCS5 were computed (Fig. 4) . For Unit II , data inputs consisted of the 14C ages byJONES (1990) and by Ross and DEGENS (1974). Sediment thickness , bulk density and AIconcentrations were determined from cores GGC19 and GGC38, collected at the samesite as BC21 and BCSS, respectively (HoNJO et al., 1988). For Unit I, accumulation rateswere computed as discussed above, using laminae ages (NEFF et al., in preparation) and14C ages (Ross and DEGENS, 1974) for comparison. The average Al accumulation ratesfrom Unit II were then compared to average Al accumulation rates for the Final InvasionPeriod of Unit I (samples 6-35). The Transition Sapropel and First Invasion Period werenot included in this comparison because they represent a brief transition period withtransitional concentrations of biogenic and terrigenous matter. Assuming that the Alaccumulation rate in the Final Invasion Period of Unit I was the same as in Unit II. theestimated age for the beginning of the Final Invasion Period ranges from about 1100 to1350 years BP (Fig. 4) .

The current state of dating the laminated sediment sequences in the Black Sea suggeststhat the exact ages may not be determined for some time in the future . The 14C agespossibly represent the upper limit due to contributions of "dead" Cretaceous carbonateand "old" terrigenous matter in the sediment. Substantial terrigenous contributions to theorganic matter in the Black Sea sediments are possibly reflected by carbon isotope data(DEUSER, 1972; CALVERT and FONTUGNE, 1987; ARTHUR et al., in preparation) and byorganic compounds (WAKEHAM et al . , in press). Terrigenous carbonate in the FinalInvasion Period averages 8% of the total carbonate fraction in BC21 and 14% in BCSS,assuming the Ca:AI ratio of the terrigenous matter in Unit II is the same as in Unit I(Fig. 5). A concentration of 8% of "dead" carbonate adds about 700 years to the absoluteage, a concentration of 14% about 1200 years. Also uncertain is the pre-nuclear"apparent" 14C age of the surface water of the Black Sea that would have been locked inthe biogenic carbon and carbonate at the time of primary production. MURRAY et al. (1991)argues that due to regular entrainment of "older" deeper Black Sea water into the surfacewater, the pre-nuclear "apparent" age of the surface water may have been 1430 years.

The question as to whether the couplets of black and white laminae are annual varvescan be answered more confidently when we have a full understanding of the laminaeformation in the Black Sea. White laminae contain over 90% coccoliths of E. huxleyi;black laminae contain mostly terrigenous matter (HAY, 1988). White laminae are deposited during the E. huxleyi bloom in summer and autumn ; the black laminae areprobably deposited during the diatom bloom in spring (HAY et al. , 1990). Possibly due todissolution of the biogenic silica at the basin floor, the terrigenous matter becomes

S1220 B. J. HAY et al,

Terrigenous Carbonate of the Total Carbonate

Unit IFinal Invasion Period

50 BC21. Wastetn Black Sea

45 •••• -1- ••• BeSS.Eastern Black Sea

\'* 40c 350 30'';:;tV 25......c 20Q)o 15

~c0 10

U 50

0 S 10 15 20 25 30 35 40 45Sample number (downcore-+)

# of Laminae Couplets

Fig. 5. Terrigenous carbonate concentration in Unit I (First and Final Invasion Periods) of BC2tand BCS5 as percentage of the total carbonate concentration . as calculated from the Caconcentrations. The remaining carbonate constitutes mostly coccoliths of E. huxleyi. Carbonate iscomprised dominantly of CaC03• as reflected by the correlation factor of 0.996 between Ca

analyses (ICP ) and carbon ate carbon analyses (coulometric method ) from identical samples.

comparatively enriched in the black laminae. During 8 years of sediment trap studies in thesouthwestern and central Black Sea , only one E. huxleyi bloom was not collected,although this does not necessarily mean that there was not an E. huxleyi bloom in the BlackSea during that year which would have formed a white lamina in the sediment column.Blooms may be regional (HAY and HONJO, 1989), but the settling particles may bedistributed rapidly throughout the Black Sea basin by large eddies, extending from theshelf to the central regions (MURRAY and IZDAR, 1989; UNLUATA and LAVIOLElTE. 1990).The fact that the laminae over the last millennia can be correlated over 1000 kmthroughout the Black Sea basin indicates that distribution must be a factor in the laminaeformation.

In light of the existing information, we believe that a chronology based on counts of theblack and white laminae reflects the absolute age more closely than a it chronology.

RESULTS

In both cores, Unit II consists largely of terrigenous matter and a high concentration oforganic matter (samples 64-44) (Fig. 6b; Table 1). Unit I consists of a -I-em-thickcoccolith ooze layer at the base (First Invasion Period; samples 43 and 42), overlain by atransitional interval approximately 3 em thick and composed largely of terrigenous matter(Transition Sapropel; samples 41-36) (Fig. 2) . The Transition Sapropel is overlain bycoccolith ooze deposited after the permanent invasion of E. huxley; ofthe Black Sea (FinalInvasion Period; samples 35-6).

The sediments of the First Invasion Period contain about 50 couplets of light and darklaminae. The average thickness of one laminae couplet is about 0.3 rnm, although the


Table l. A verage accumulation (accum .) rates and concentrations ofthe major sediment components . The LateHolocene sediment sequence collected in the box cores is subdivided into the upper part ofUnit II . the First In vasionPeriod , the Transition Sapropel, and the Final Invasion Period . The Peak Bloom Period with comparatively high

calcium deposition is summarized as well

Unit I

Final Peak FirstInvasion Bloom Tran sition invasion of UpperPeriod Period Sapropel E. huxleyi Unit II

Samples (6-35)* (9-12) (36-4 1) (42-43) (>43)Laminae couplets (175-1256) t (267-428)t (1256-1556)t (1556-1633)t (> 1633)t

Western Black SeaConcentration (%) (BC21)

Ca 24.2 27.9 7.56 22.4 4.75Si 4.75 2.84 13.8 5.52 12.8AI 1.69 0.977 5.00 2.06 4.86Ti 0.083 0.051 0.241 0.099 0.233Organic carbon 5.41 4.22 8.82 6.52 13.3E. huxleyis 56 67 7 51Terrig. matter§ 29 17 86 36 84

Accum . rate (g m- 2 y-l) (BC21)Total sediment AR 85.2 114 39.7 45.2CaAR 21.5 32.0 3.1 10.0SiAR 3.6 3.1 5.5 2.6AlAR 1.3 1.1 2.0 1.0TiAR 0.06 0.06 0.10 0.05Organic carbon AR 4.4 4.7 3.4 3.0E. huxleyi ARj: 51 77 2.9 23

Terrig. matter AR§ 22 19 34 17 17

Eastern Black SeaConcentration (%) (BC55)

Ca 22.7 28.0 7.53 17.2 5.32Si 6.34 3.90 13.8 8.65 11.8AI 2.38 1.45 5.31 3.20 4.38Ti 0.113 0.069 0.253 0.162 0.216Organic carbon 5.10 4.34 6.89 7.01 13.7E. huxleyi; 49 66 3 33

Terrig . matter§ 41 25 92 55 76

Accum. rate (g m-2 y-l) (BCSS)Total sediment AR 70.2 83.0 40.9 39.8CaAR 16.5 23.3 3.4 6.8SiAR 4.2 3.2 5.4 3.4AlAR 1.6 1.2 2.1 1.3TiAR 0.08 0.06 0.10 0.06Organic carbon AR 3.5 3.6 2.8 2.8E. huxleyi AR * 36 55 2.4 13Terrig. matter AR§ 28 21 36 22

*For accumulation rates , samples 8-35 were averaged.t Number of laminae couplets based on counts of couplets of black and white laminae from thin sections (NEFF

et al., in preparation).

*Computed after formula given in Methods.§Computed by multiplying the AI concentration and accumulation rate, respectively , by factor 17.3.


thickness varies considerably. Thick white laminae deposited at the base of this periodreflect a strong initial appearance of E. huxleyi. These laminae gradually thin towards theend of this period (Fig. 2), suggesting that the E. huxley; blooms were becoming lessintense . Throughout the Tran sition Sapropel, E. huxley; is largely absent except for a fewthin laminae indicating very brief periods of return (e.g. within sample 40). E. huxley;permanently returns in sample 36, more gradual this time than at the beginning of the FirstInvasion Period, as indicated by the thinner white laminae . Throughout the Final InvasionPeriod, the thickness of the individual white and black laminae varies considerably(Fig. 2) .

The laminated sedimentary sequences in the abyssal Black Sea can be traced throughoutthe entire basin (Fig. 2). The sequences are continuous in both cores, despite theinterbedding of a turbidite in BC21. The deposition of the turbidite did not causedetectable erosion at the base indicating low energy deposition . The contact is smooth,and comparison with core BCSSshows no missing laminae (Fig. 2). The thickness of Unit Ifrom the first invasion to the top of the fluff layer is 49 em in BC21 (or 37 em excluding the12 em thick turbidite) and 30 em in BCSS. That difference translates into a -20% highersedimentation rate in Unit I in the western Black Sea .

The carbonate in Unit I (Fig. 6a) consists almost entirely of remains of E. huxley; (Fig.S). In contrast, carbonate in Unit II is mainly of detrital origin (TRIMONIS, 1974).Coccoliths are particularly concentrated in an interval in the upper third of the core(samples 9-12; Peak Bloom Period; Fig. 2). On average, the Ca accumulation rate duringthe Final Invasion Period in BC21 is30% higher than in BCSS(Fig. 7a). The accumulationrate of E. huxley; in BC21 is even higher by 40% than in BCSS. In the Transition Sapropel,the average Ca accumulation rate is slightly higher in BCSS than in BC21, reflecting thedominance of detrital carbonate.

Periods of comparatively high Ca concentrations in both cores (Fig. 6a) are alsoreflected in high Ca:AI ratios (Fig. 8a) . The Corg:AI ratio of both cores is similar to thepattern of the Ca :AI ratio (Fig. Sb), reflecting the higher contribution of marine organicmatter during time periods of higher E. huxley; deposition . The Corg:AI ratio during thedepos ition of the Transition Sapropel is unusually low, possibly due to dilution byterrigenous material.

Other biogenic components are not present in significant concentrations. Diatomswere measured to contribute between 9 and 39% of the annual particle flux at 1200 m inthe Black Sea water column (HAY et al., 1990). Once settled to the basin floor,however, the biogenic silica is dissolved in the upper few centimeters of the sedimentcolumn. This dissolution is indicated by biogenic silica measurements of the fluff layer(7%) and the o-Z em sediment layer (4%) (PILSKALN , 1989), and biogenic silicameasurements in the deeper sediment column «1%) (SHIMKUS et al. , 1973). In coresBCZl and BC55, the only indication of higher diatom concentrations may be in the twouppermost samples, where Si:AI ratios are highest (Fig. 8c). The average Si:AI ratio is2.8 in BC2l (samples 3S-6) and 2.7 in core BC5S. In comparison , Si:AI ratios in thefluff sample and the 0-2 em sample from BC2l are 3.5 and 3.2 , respectively. Theaverage Si:AI ratio is 3.9 in particles collected in the water column at 1200 moverseveral years (HAY, 1987).

The ratio of the Al accumulation rate in the western Black Sea (BC2l) to that in theeastern Black Sea (BCSS) indicates a change from about equal rates in the Transition


5 10 15 20 25 30 35 40 45 50 55 60 65Samplenumber (downcore-)

Calcium

(a)

(b)60 6555

Unit II

5045403510 15 20 25 30

~ BC21, Wos'o,n BI,d<So,.. . ........ Bess. E,storn Black Sea

5

2

4

6

20

30

co~c:QlUCoo

/I of LaminaeCouplets

5Si '

40

soI ' i

30 35

T

30,125!

2'0

20I

151S

I,'0

S 10, ,S

BC21 0BCSS Jt-.....J.,,..-J.---,rL---r'--t--u.-........----,--1-..--r-l----

T • Infe/beddedTurt>ldne fl2 em) o e p I h (cm)

Fig. 6. Concentrations of (a) Ca and (b) Al for cores BC21 (western Black Sea) and BCSS(eastern Black Sea) . The 'T' on the depth scale for core BC21 marks the position of a 12-cm

thick turb idite.

Sapropel to higher rates in the eastern Black Sea during the Final Invasion Period(Fig. 7c). The difference is most pronounced between samples 31 and 25, when the Alaccumulation rate was as much as 35% lower in the western Black Sea than in the easternBlack Sea (Fig. 7c). After about sample 27, the Al accumulation rate ratio graduallyincreases. On average, the Al accumulation rate in the eastern Black Sea was about 20%higher than in the western Black Sea during the Final Invasion Period even though thewestern Black Sea receives most terrigenous matter supplied to the Black Sea basin (Fig.1). A substantial part of the river-borne sediment must be deposited on the Danube Fanand does not reach the abyssal plain of the western Black Sea.

Changes in the terrigenous matter input between the eastern and western Black Sea alsoare reflected in the Si:AI and Ti:Al ratios (Fig. 8c,d) . In the western Black Sea, both ratiosremain roughly constant throughout the core . In the eastern Black Sea (BCSS), however,both ratios are distinctly lower between samples 39 and 2S; they decline sharply after theFirst Invasion Period and then increase again to their previous level and to the levelobserved consistently in the western Black Sea.

S1224 B. J. HAY et al,

,+:

!..~~ l

\¥4-

Org. Carbon Ace. Rate Ratio (b)

- Ca Ace. Rate Rallo (a)••••-+0... f. hlJ)( Ace. Rate Rallo

10 15 20 25 30 35 40 45

Moans /.....",.. 3-35},oleaB E '-"'YI

5

mon

Unill

0.7

0.8o

III

a:

~

in1.4III

w<,

in

~ 1.2

U'lU'l

U 1035 40 4S0 5 10 15 20 25 30

III(e)<, 11 AI Ace. Rate Rallo

N10

UED 0.9

1.8

1.6;:;-'lS 1.4 B:

! A..:- 1.20E>. 1.0<:l'tl..5 0.800 0 5E 1.6~

i30

20I

1'5

# of LaminaeCouplets

T

~5 30,1 4,5

do 2'5

5 10 15I ( I

~ 1'0

BC21 01-1-L,---L----,r-'----r--r--....u..--r-'-__~

BCSS 0

T. Inre<baddadTurbld/la t12 em) o e p , h (em)

Fig. 7. Ratios of the accumulation of (a) Ca and E. huxleyi, (b) organic carbon and (c) AI.between cores BC21 and BC55 for the Final Invasion Period. Data were smoothed by using a

3·point moving average to emphasize the major trends.

REGIONAL CLIMATE HISTORY: PREVIOUS STUDIES

Climatic events during the Late Holocene varied considerably within the drainage areasof the Black Sea rivers (Fig. 9). These drainage areas are located in central andsoutheastern Europe (Danube River), European Russia (Dnestr, Bug. Dnepr and DonRivers), and Northern Turkey and the Caucasus (Yesil Irmak, Kizil Irmak, Rioni andKoruh Rivers. and others). The Danube presently contributes about 56% of the detrital

Sediment deposition in the Late Holocene Black Sea 51225

65

(c)60 65

60

55

55

Unilll

50

50

BC21,We"_ Black St.... .... ... Be55, Easlem Blaek Sea

T"ms/li"" ~ FhI '1''''''''''' """od (a)Sapropel

35

:...... ....... :

...............rt .~

30252015105

ot-"'"""T""-......-"'"""T""-......-...........-..---~~~~~~~~=~o

2

4

3

5

10

20

3.2

24

.Q 3.0iiia: 2.8

«~ 2.8

o~a:«.........IIIU

o~a:«.........co.0iiioeno

0.055

.Q

~ 0050

«.........F 0.045

55 60 65

(d)

65605520 25 30 35 40 45 50Samplenumber (downeore_)

151050.040 +-----.--,..-...--~--,--..--~-_ ___.-_,..-~-..._~

o

# of LaminaeCouplets

55I

i40

50! i

3'0 35

( em)

45!

i

25

T

30.125\

20

o e p I h

15!

"010

i5I

AT. Inlttbedded

TlJrl>/dn. (12 CIIl)

BC21 0BeSS Jt---'-Jr-"-r:'----,.l--+--..u.....~--,_J._r-_r_''----

Fig. 8. Ratios between the Al concentrations and the concentrations of (a) Ca, (b) organiccarbon, (c) Si and (d) Ti. Organ ic carbon concentrations are from ARTHUR et al. (in preparation).

S1226 B. J. HAVer al.

6000

.. _-...

sooo

_..-

"

·r"......... ·· ..

-._...

Years BP3000 4000

/I

"""" ./... ,

20001000

CASPIANSEA

LEVEL

VARVETHICKNESS .t-1Il.......a.u...IU..~<;;:J"V"\;;~~1Jtfv;~-M"'"7i];~q\~,..--..LAKE SAKI

EUROPEANCLIMATE

FROMARCHEOL.EVIDENCE

TUlkey 1'1

PRECIPITATION 1••_C.nl,,1 Europe _ ..

Europe (generall - .. _. ,.. - .

___T_Uf_ke-jY _... _

GLACIEE'::: Alps 1_. _ ..... .. .

Cllm,tlc P.lled I!:UEl ~ ", '.

SUB·ATLANTIC PERIOD SUB·BOREAL PERIOD ATLANTIC PERIOD

o 1000 2000 3000 4000 5000Years 8 P

6000 7000

Fig. 9. Overview of major climatic events in Europe and Turkey. The time scale for the Atlantic,Sub-Boreal Period, and Sub-Atlantic Periods is based on LAMB (1977, p. 372). All ages reported inthe literature as years AD were converted into years BP using the conversion of zero years AD equals2000 years BP. Data sources: Caspian Sea level (CHEPALYGA, 1985). Black Sea level (FEDOROV,1978, stippled line; CHEPALYGA, 1985, solid line). Varve thickness Lake Saki (SCHOSTAKOVICH,1934; smoothed by moving averages of three). Climate from archeological evidence (BOUZEK,1983. This climatic pattern is derived from evidence such as settlement density, cave settlementsand datable soil developments in limestone sequences in central Europe. Scale is qualitative).Precipitation (bars indicate moister periods with thicker bars indicating higher rainfall, the dottedlines indicate drier periods. A diagonally striped bar indicates uncertainty about the event from theliterature), Central Europe (LAMB, 1977, pp. 372f, 449, 461f), Europe (general) (FRENZEL, 1966;STARKEL, 1966), Turkey (BUTZER, 1958; BINTLIFF, 1983; GRISWOLD, 1979; EtSMA, 1978; ERINC,1978). Glaciers (bars indicate periods of glacial advances with thicker bars indicating majorperiods; the dotted lines indicate absence of glacial advances), European Alps (LAMB, 1977,pp.214, 463; ERtNC, 1978; HEUBERGER, 1954; HARDING, 1983), Turkey (ERtNC, 1952, 1978).Climatic periods: LIA (Little Ice Age); LO (Little Optimum); peo(Post-glacial Optimum) (after

LAMB, 1977).

matter supplied to the Black Sea, the European Russian rivers about 8%, and theCaucasian and Turkish rivers 36% (SHIMKUS and TRIMONIS, 1974). The Danube drainagebasin represents about 35%, the drainage basins of the European Russian rivers about47%, and the drainage basins of the Caucasian and Turkish rivers about 18% of the totaldrainage area (OWENS et al., 1980b).

The level of the Caspian Sea is a good indirect indicator of the moisture level on theRussian plain (CHAPPELL, 1970; GERASIMOV, 1978). The Caspian Sea receives over 75% of


its water from river discharge and only 25% from precipitation (GERASIMOV, 1978). Almostall the water entering the Caspian Sea (97%) remains in the basin and evaporates. Over80% of the river discharge is supplied by the Volga River alone. The drainage basin of theVolga is immediately adjacent to the basins of the Don and Dnepr Rivers.

The Post-glacial Climatic Optimum in the Atlantic Period reached its maximum at about6000 years BP with moister conditions north of the Alps (FRENZEL, 1966; LAMB, 1977,p. 372; BOUZEK, 1983). The Atlantic Period ended with a cooling event that led to theexpansion of Alpean glaciers. As in Europe, precipitation during the Climatic Optimumwas higher on the Russian plain (KHOTINSKIY, 1985), which is reflected in the high level ofthe Caspian Sea (Fig. 9). In contrast , Turkey was very dry between 7000 and 5000 years BP(ERINC, 1952; BRICE, 1978).

The Sub-Boreal Period was characterized in Europe by warmer and generally drierconditions, coupled with considerable shorter-term variability in precipitation (LAMB,1977, p. 372). Higher precipitation in central Europe apparently occurred around 4000years BP (FRENZEL, 1966; BOUZEK, 1983). Higher river discharge may have caused the highBlack Sea level during this time (CHEPALYGA, 1985). Periods of major glacial advances inthe European Alps during the Sub-Boreal Period are not reported. On the Russian plain,precipitation levels were probably lower until about 3500 years BP, when the Caspian Seaapproached levels similar to the levels during the Post-glacial Climatic Optimum (KHOTINSKIY, 1985; CHEPALYGA, 1985). Information is limited for the climatic conditions inTurkey during the Sub-Boreal. Contrary to the northern latitudes , the climate in Turkeybecame moister at 5000 years BP probably due to a latitudinal shift in circulation patterns(BINTLlFF, 1983). BUTZER (1957, 1958) suggested drier conditions in the Near Eastbetween 4400 and 2850 years BP.

The beginning of the Sub·Atlantic Period (3000/2500 years BP) is marked by colderglobal temperatures. In central Europe, glaciers expanded (HEUBERGER, 1954) andprecipitation increased (FRENZEL, 1966; LAMB, 1977, p. 374). At about 2500 years BP drierconditions returned until about 1600 years BP. After about several centuries of moisterconditions, warmer and drier conditions followed from about 1200 to 800 years BP, knownas the Little Optimum (LAMB, 1977, p. 374). This event was characterized by warmer anddrier conditions in most of Europe, including central European Russia , but moisterconditions in the Mediterranean region. LAMB (1977) explains the weather patterns duringthe Little Optimum as due to a shift of the subtropical anticyclones to the north withweaker westerly winds at mid-latitudes (about 4Q-600N).

On the Russian plain, vegetation changes suggest generally moister conditions duringthe Sub-Atlantic (KHOTINSKIY, 1985). The level of the Caspian Sea oscillated from about2700 years BP to the Present , with generally lower levels between 1600 and 800 years BP(CHEPALYGA, 1985). This period includes the Little Optimum, for which warmer and drierconditions are reported in Russia (LAMB, 1977, p. 374). The Caspian Sea level rose againafter 800 years BP with the movement of the storm track back north (CHAPPELL, 1970).

Limited information exists about the climate in Turkey from the early part of the SubAtlantic period. The Black Sea was at its lowest level between 3000 and 2500 years BP,similar to the Caspian Sea levels, despite supposedly higher precipitation in Europe .Perhaps the inflow of Mediterranean water through the Bosporus was reduced at the timedue to a lower Mediterranean water level. Drier conditions in Turkey are suggested by thethinner varve record of Lake Saki on the Crimean Peninsula (SCHOSTAKOVICH , 1934; Fig .9) . These presumably annual laminations are considered more reflective of the climatic


conditions of Turkey than of Europe or European Russia (LAMB, 1977, p. 408) . At 1900years BP, climate conditions in Turkey were reportedly similar to present conditions(ERINC, 1978). Shortly thereafter, from about 1700to 1200years BP, the Near East sufferedlong periods of drought (LAMB, 1977, p. 146). During the Little Optimum. conditions inTurkey were moister and colder from about 1250 to 700 years BP (BUTZER, 1958; ERlNc,1978). In fact, ice masses reportedly floated on the Marmara Sea in the years 1261, 1247,1245,1238, 1072,1066,989 and 787 BP (ERINC, 1978). Moister conditions in the Black Searegion also are suggested by SCHOSTAKOVICH (1934), who measured thicker varves between1200and 750 years BP. Archeological evidence indicates that the Mediterranean and NearEastern regions were moister than at present between about 1000and 800 years BP, whichsimultaneously resulted in more frequent stream flows in wadis in African and Arabiandeserts (LAMB, 1977, p. 439). At least temporary moister conditions are suggested also byreports of historic floods further south at Baghdad in 1958,926 and 741 years BP (LAMB,1977).

After 800 years BP, the climate in Europe generally deteriorated, becoming colder andwetter. Caspian Sea levels also rose sharply at about 800 years BP (LAMB, 1977, p. 439; Fig.9). The deteriorating climate conditions culminated in the Little Ice Age from about 450 to150 years BP, with the main phase occurring in most of Europe from 450 to 300 years BP(LAMB , 1977, p. 461). The climatic patterns during the Little Ice Age varied regionally, andthe major climax came at different times in different parts of the world. In the EuropeanAlps , glaciers advanced between 460 and 300 years BP and again between 220 and 150yearsBP (LAMB, 1977). While the earlier surge has been attributed mostly to a lengthening ofthesnow season and a shortening of the summer ablation period, the later glacial advance wascaused by an increase in precipitation. Temperatures during the latter advance fluctuatedless than a degree over the course of 30 annual means (MANLEY, 1966; LAMB, 1977).Similar to the Alpine glaciers , advances in Turkish glaciers occurred from about 390 to 310years BP, and briefly again at around 270 years BP and around 140 years BP (ERINC, 1952).

DISCUSSION

Some of the most prominent features in the laminated sediment record of the Black Seaare the events that permanently established E. huxleyi, and the variability in biological andterrigenous production during the Final Invasion Period . These events will be discussed inthe context of the major regional climatic patterns. The link between the climatic recordsof the region and the laminated sediment sequence may give us further indications for thechronology of the depositional events .

First invasion ofE. huxleyi

The Black Sea is very sensitive to changes in the salinity balance between the inflow offresh river water and saline Mediterranean water through the Bosporus. It is thereforeconceivable that comparatively small changes in water level , evaporation and precipitation in the drainage basins of the Black Sea rivers may have led to second-ordervariability during the rise in salinity since the first inflow of Mediterranean water. Indeed.DEUSER (1972) previously interpreted fluctuations in the OBC record of organic carbon inUnit II as an indication of an unsteady salinity rise.

At the time of the invasion into the Black Sea, E. huxleyi was well established in theMediterranean (e .g. ERBA, 1989). E . huxleyi probably invaded the Black Sea initially

Sediment deposit ion in the Late Holocene Black Sea S1229

because the salinity had reached a level above its tolerance threshold (e.g. BUKRY, 1974).The required minimum salinity for the growth of E. huxleyi in laboratory cultures is 16%0(MJAALAND, 1956). In nature , however, the lowest salinity at which E. huxleyi is known toexist is 110/00 in the Sea of Azov (BUKRY, 1974). BOUDREAU and LEBLOND (1989) predictedthat it would take about 1600-3100 years to increase the salinity to 11%0 from fresh waterconditions, although their model is admittedly a very simplified first approach . Previouschronological studies estimated a time period of about 4000 years from the beginning ofthe Unit II deposition to the beginning of the final E . huxleyi invasion (Ross and DEGENS,1974; DEGENS et al. , 1980a; JONES , 1990; Fig. 3).

During the First Invasion Period the terrigenous supply pattern changed in the easternBlack Sea , as shown by the sharp drop of the Si:Al and Ti:AI ratios (Fig. 8c,d). Possibleexplanations for this change include (1) a change in the grain size of the particulate matter ,(2) a change in the dominant provenance of the terrigenous particles , and (3) a change inthe circulation pattern and thereby the particle supply pattern . SPEARS and KANARISSOTIRIOU (1976) found a positive relationship between quartz content and the Ti :AI ratio,because most Ti minerals such as ilmenite, rutile, anatase and sphene are also fairlyweathering-resistant, like quartz. Thus, the drop in the Ti:AI and the Si:AI ratios in theeastern Black Sea following the first invasion of E. huxleyicould have been caused by finergrain sizes of the terrigenous material. A change in provenance is another possibility,because Ti :Al ratios in common rocks can range widely from about 0.18 in basalts to 0.02in granites (TUREKIAN and WEDEPOHL, 1961). "Average" shale has a ratio of 0.058, averagesandstone a ratio of 0.060, and average carbonate a ratio of 0.095. Thus, the lower butvariable Ti :AI ratios in the Black Sea sediment could reflect the change in the predominance of a river source, caused by changes in precipitation in the drainage area of the majorrivers draining into the eastern Black Sea (Kizil Irmak, Yesil Irmak, Coruh and Rioni) orchanges in the distribution of the terrigenous matter in the Black Sea.

The scenario suggested from these observations is that salinity in the Black Sea hadincreased gradually to a level tolerable for E. huxleyi. Shortly after the invasion of E.huxleyi , however, climate conditions changed and river discharge led to a reduction insalinity to a level which became largely unacceptable for the existence of E. huxleyi. Thesupply of terrigenous matter to the Black Sea increased (Fig. 6b) with the increase in riverdischarge . E. huxleyi permanently returned to the Black Sea when the decrease in rainfaIland river discharge decreased .

Chronology

Based on the extent of the drainage areas, a decrease in salinity due to higher rainfalland runoff that interrupted the invasion of E. huxley; would most likely have been causedby an increase in rainfall in Europe and European Russia rather than an increase in rainfallin Turkey. In the historical record, the period that is most suitable for the deposition of theTransition Sapropel based on other existing information lies between 2000 and 1000 yearsBP. The beginning of this period is marked by the end of the most severe regression of theBlack Sea since the Post-glacial Optimum ("Fanagorian Regression"). This regressionmay have accelerated the first invasion of E. huxleyi .The regression occurred from 3500 to1500 years BP (FEDOROV, 1978) or from 3000 to 2200 years BP (CHEPALYGA, 1985) (Fig. 9).Reported decreases in Black Sea levels vary from about 5 to 15 m (OSTROVSKIY et al., 1977;FEDERov, 1978; CHEPALYGA, 1985; BALABANOV and IZMAlLOV, 1989). The sea levels of the

S1230 B. 1. HAyer al.

world oceans at the time of the Fanagorian Regression ranged from zero to 2 m below thepresent level (e.g. MORNER, 1971). The difference in global sea level and the Black Sealevel should have led to relatively higher inflow of Mediterranean Sea water through theBosporus. The subsequent transgression in the Black Sea ("Nymphaean Transgression")increased the levels of the Black Sea to about 1-2 above present levels (Fig. 9). Theelevated Black Sea level may have led to relatively higher outflow of Black Sea water to theMediterranean Sea and simultaneously a reduced inflow from the Mediterranean Sea.This change could have resulted in lower salinity levels in the surface water of the BlackSea, explaining the temporary near-disappearance of E. huxleyi.

High precipitation levels in Europe generally coincided with the Nymphaean Transgression. Higher precipitation occurred between about 1700/1600 to 1200 years BP (Fig. 9).Archeological evidence suggests higher rainfall between 2000 and 1400years BP (BOUZEK,1983; Fig. 9) , but these data are very qualitative . Rainfall in Turkey was low before andduring this moist period in Europe, and only increased during the Little Optimum from1200 to 800 years BP (Fig. 9), when dry conditions prevailed in the northern latitudes.

The available climatic records and sediment data from the Black Sea region matchreasonably well with the ages based on counts of laminae couplets (NEFF et al., inpreparation) . The low Black Sea level between 3000 and 2000 years BP may have beeninduced by lower rates of freshwater inflow. Less freshwater input could have in tumresulted in a comparatively more rapid increase in salinity , allowing E. huxleyi to invadearound 1600 years BP. Comparatively moist conditions in Europe over the next 400 yearsincreased the Black Sea level and lowered the salinity shortly after the invasion, forcing E.huxley; to retreat. The Transition Sapropel was deposited. E. huxley; returned with thebeginning of the Little Optimum at around 1200years BP, when drier conditions in Europeled to an increase in salinity. River discharge increased in Turkey, but not sufficiently tomake up for the decrease from the European and Russian rivers. The shift in rainfall tosouthern regions may have caused the temporary drop in the Si:AI and Ti:Al ratios in theeastern Black Sea until about 850 years BP.

There is no indication among the existing scientific and historical records of a change inclimatic condition at the Unit 1/11 boundary that could result in a reduction in riverinesediment load throughout Unit I by a factor of 2.6, as required to validate the existing 14Cchronology (Fig. 4) . This factor represents a substantially lower accumulation rate ofterrigenous matter and should have resulted from noticeably drier conditions in Europe inUnit I. While shorter-term climatic fluctuations existed and are well documented in thehistoric records, noticeably drier conditions in Europe were not observed. Clearly , thisissue should be investigated in more detail.

Sediment accumulation in the Final In vasion Period

Accumulation rates of major sediment components were investigated assuming achronology based on the deposition of annual laminae couplets (NEFF et al. , in preparation). The interpretation of the accumulation rates should provide relevant information of the major trends , even if the chronology of the Black Sea sediment becomesfurther refined in the future.]te_tl~tal sediment accumulation rates during the Transition Sapropel are about 40 g

m y 10 the eastern (BeSS) and western Black Sea (BC21) (Fig. lOa;Table 1). The totalaccumulation rates approximately double in the Final Invasion Period with about 20%


Unit IFlnal/rMI"'" Period T.S.

160 Total Ace. Rate (a)

140120100

eo6040

BC21.WI"_ BlockSII20 .... ....-... BC5S.Elster" BlockSI'0

0 5 10 15 20 25 30 35 ~o ~5

Calcium (b)ffi 40Ql>-'<,N

30E<,

.920

a>10

ell

ex: 0~o ~50 5 10 15 20 25 30 35

C Eo huxleyl (c)0 100

III 80

::l 60E::l 40ou 20-c

0~50 5 10 15 20 25 30 35 ~o

3.0 Aluminum (d)

2.0

1.0

5O.o+--~-~-~-.,.....-.,.....-~-~-~-

o

~~$§~~~§~~~~~~~II 01 Laminae Couplets

TBC21 0 5 10 15 20 25 30,1 45

BC55 JI

iI

i i iI I I

/ I I5 10 15 20 25 30T. "".rt>edded o e p I h (cm)Turbldll' (IZ em)

Fig. 10. Total accumulation rates (a) and accumulation rates of (b) Ca, (c) E. huxley; and (d) AI,for cores BC21 and BCS5. The E. huxley; accumulation rate is near-identical to the Caaccumulation rate. but varies depending on terrigenous matter content of the sample. The E.huxley; accumulation rate was particularly high in samples 9-12 ("Peak Bloom Period") in both

cores.


.. ,.. .......I I, I , I , I

1200 1300 1400 1500 1600

-------_ .

F iltoillfPOSlOli 't'lod

i t

_?---.:IJ Liul. Ie. Agr fL

.Ii ,,:,.: JLittl. Optimum d":.

Y • a r • A. D.1'" aMI .... 11M ... 1_ I" UN U" 11M ... ...

Ii 11M. 21M. joo .Iio 5IM. 61io ,Iio 11M. 9Iio tiloo 1100

Yea r s n. P.

Tim.

CJar;" SUa" 'illl' In Age[urope•• Alp. (IITurkish gl.don (2)

IIjgbcr BajnCi'JI EuroPC (1)

'Ugb,r BalaCal! Tucker I Iule OpclmumPh,s. rnyironmrnl, Turkey (J) .A.chaool . ..Id.nee, No•• Ea.. (u ' , , .Lako S.kl ..r.... CromoanP. (4) " " ' " ..Phys. .n,lronm.nl, Nur Easl (5)Hislorlc noods, B.gdad (I)

Timo Period. (I)

£. Aud" , eeeum , puks

Terrlg. m.uer .ctum. po.ksCh.nl' In I.rrlg.nous

pro..n.nee, Easl. BIKk S••

SlraUgnphy

Fig. 11. Summary of the major events recorded in the sediment record of the Late HoloceneBlack Sea. The major depositional events within Umt I may have occurred as follows: after the firstinvasion of E. huxleyi around 1600 years DP. the Transition Sapropel was deposited during thesubsequent four centuries as a result of high ramfall in Europe . Despite moister cond itions inTurkey. dry conditions in Europe during the Little Optimum led to an increase in salinity and apermanent return of E. huxley; around 1200 years DP. Cold temperatures between 450 and 300years BP during the Little Ice Age may have led to the Peak Bloom Period. [References: 1. LAMB

(1977); 2, ERINC (1952); 3, ERINC (1978); 4, S CHOSTAKOVICH (1934); 5. B UTZER (1958»).

higher rates in the western Black Sea than in the eastern Black Sea . The total accumulationrates increase slightly from the beginning of the Final Invasion Period towards the Present .

The Al accumulation rates are highest during the Transition Sapropel (Fig. lOc;Table 1)and lower during the Final Invasion Period, particularly in the eastern Black Sea. AIaccumulation rates are highest within the Final Invasion Period between samples 19 and13. During the Little Ice Age, AI accumulation rates are comparatively low.

The Ca accumulation rates in both cores typically range between 10 and 30 g m-2 y-I inthe Final Invasion Period (Fig. lOb). Ca accumulation rates in the eastern and westernBlack Sea are on average over 40% higher between samples 9 and 12 (Peak BloomPeriod) . High Ca accumulation rates are also recorded in samples 23 and 33 in the westernBlack Sea (BC21). The Ca accumulation rates are on average 30% higher in the westernBlack Sea than in the eastern Black Sea . Similar relationships exist for the accumulationrate of E. huxley; (Fig. lOc).

Laminae counts suggest that the Peak Bloom Period occurred during the first cold spellof the Little Ice Age (450-300 years BP) (Fig. 11). This cold spell was not just a latitudinalshift of the climatic belts as it happened at other times in the Late Holocene , but a result ofcolder temperatures in the entire Black Sea drainage region. Glaciers advanced in theEuropean Alps as well as in the Caucasus and in Turkey (Fig. 9). Winter storms during thistime period may have agitated the water column more strongly, bringing nutrients fromthe nutrient-rich deeper water to the surface . Colder temperatures during this time periodare also suggested by sharply lower temperatures from alkenone measurements (S.WAKEHAM , personal communication).

Sediment deposition in the Late Holocene Black Sea 51233

The variability in biogeochemical parameters in the laminated Black Sea sedimentsdocuments the responsiveness of the biogeochemical conditions in the basin to climaticchanges. Undoubtedly, the biogeochemical system in the Black Sea will remain sensitivein the future as the inflow of fresh water is reduced by such man-made causes as riverdamming for irrigation (TOLMAZIN, 1985) and if a global sea level rise leads to a greatersaline water inflow through the Bosporus.

Acknowledgements-We thank J. Broda and A. Gagnon, Woods Hole Oceanographic Institution, S. Derman,Turkish Petroleum Corporation, T . Konuk and M. Duman , Dokuz Eyltil University, Izmir , and other membersof the scientific shipboard party of the R.V. Knorr Black Sea cruise 134-8, as well as the very co-operative officersand crew of the R.V . Knorr , for their support in the successful coring operation. Inorganic geochemicallaboratory work was accomplished by D. Bankston (deceased) and L. Ball (ICPES) , S. Manganini and S. Carter(sample preparation) , Woods Hole Oceanographic Institution. This research was partially funded by theNational Science Foundation Grants Nos OCE-8614363 (Honjo) and OCE-8711741 (Arthur), and the U .S.Geological Survey Evolution of Sedimentary Basins Program.

REFERENCES

BALABANOV I. P. and YA. A. IZMAILOV (1989) Change in level and hydrochemical regime of the Black and AzovSeas during the past 20,000 years . Water Resources, 15,539-546.

BINTLIFF J . L. (1983) Climatic change. archaeology and Quaternary science in the eastern Mediterranean region.In : Climatic change in Later Prehistory, A. HARDING, editor, Edinburgh University Press, pp . 143-161.

BOUDREAU B. P. and P. H . LEBLOND (1989) A simple evolutionary model for water and salt in the Black Sea .Paleoceanography, 4. 157-166.

BOUZEK J . (1983) Climatic changes in central European prehistory. In: Climatic changes in later prehistory , A. F.HARDING. editor, Edinburgh University Press, pp. 179-191.

BRICE W. C. (1978) The desiccation of Anatolia. In: The environmental history of the Near and Middle East, W .C. BRICE, editor, Academic Press, New York, pp. 141-147.

BRYANTSEV V. A., D . YA. FAsHcHuK,T . A. AYZATULLlN, S. V. BAGOTSKIY and A. V. LEONOV (1988) Variation inthe upper boundary of the sulfide zone in the Black Sea: analysis of field observations and modeling results.Oceanology, 28, 180--185.

BUKRY D. (1974) Coccoliths as paleosalinity indicators-evidence from the Black Sea. In: The Black Seageology, chemistry, and biology, E. T. OWENS and D. A . Ross, editors. Memoirs of the AmericanAssociation of Petroleum Geologists, 20, 353-363.

BUKRY D., S. A. KING. M. K. HORN and F. T. MANHEIM (1970) Geological significance of coccoliths in fine-grained carbonate bands of postglacial Black Sea sediments. Nature, 226, 156-158.

BUTZER K. W. (1957) Late glacial and past glacial climatic variation in the Near East. Erdkunde, 11,21-35.BUTZER K. W. (1958) Quaternary stratigraphy and climate in the Near East. Bonner Geogr. Abh., 24, 1-157 .CALVERT S. E. and M. R. FONTUGNE (1987) Stable carbon isotopic evidence for the origin ofthe organic matterin

the Holocene Black Sea sapropel. Chemical Geology, 66, 315-322.CALVERT S. E., J. S. Vocst.and J. R . SOUTHON (1987) Carbon accumulation rates and the origin of the Holocene

sapropel in the Black Sea. Geology, 15,918-921.CHAPPELL J. E. JR (1970) Climatic change reconsidered: Another look at "the pulse of Asia". Geographical

Review, 60, 347-373.CHEPALYGA A. L. (1985) Inland Sea Basins. In: Late Quaternary environments of the Soviet Union, A. A .

VELICHICO, editor, University of Minnesota Press, Minneapolis, pp . 229-247 .CRUSIUS J. and R. F. ANDERSON (in press) A study of the mobility of 21OPb in Black Sea sediments. Geochimica et

Cosmochimica Acta.DEGENS E . T .• W. MICHAELIS, C. GARRASI. K. MOPPER, S. KEMPE and V. A. ITTEKKOT (198Oa) Warven

Chronologie und fruhdiagenetische Umsetzungen organischer Substanzen holozaner Sedimente desSchwarzen Meeres. Neues lahrbuch Geologisch-Palaeontologischer Monatshefte, 5, 65-86.

DEGENS E. T., P. STOFFERS, S. OLUBIC and M. D. DICKMAN (198Ob) Varve chronology: Estimated rates ofsedimentation in the Black Sea deep basin. In: Initial Reports ofthe DeepSea Drilling Project, D. A. Ross etal., editors, Vol. 42B, Washington, DC, U .S. Government Printing Office, pp , 499-508.

51234 B. J. HAY et al .

DEUSER W. G . (1972) Late-Pleistocene and Holocene history of the Black Sea as indicated by stable-isotopestudies. Journal of Geophysical Research, 77, 1071-1077 .

EISMA D. (1978) Stream deposit ion and eros ion by the eastern shore ofthe Aegean. In : The environmental historyof the Near and Middle East, W. C. BRICE. editor , Academic Pre ss, New York . pp . 67-81.

Ell.BA E . (1989) Stratigraphy of cruise TYRO-87 cores from the Bannock. Tyro . and Kretheus basins . In : Anoxicbasins of the eastern Mediter ranean . Results from the Third Conference. Bergamo, Italy. 14-16 December1988, M. B. OrA, A. CAMERLENGHI and C. CORSELL! , editors, Ricerca Sctentifico ed Educazione Permanente, 72. 20--23.

Ell.INC S. (1952) Glacial evidence of the climatic variations in Turkey. Geogra]. Annaler (Stockholm ),)4, 8997.

ERINC S. (1978) Changes in the physical environment in Turkey since the end of the Last Glacial, In : Theenvironmental history of the Near and MIddle East, W. C. Bll.ICE. editor, Academic Press, New York,pp .87-110.

FEDOROV P. V. (1978) The Pleistocene of the Ponto-Caspian , Nauka Press, Moscow.FLOYD M. A ., V. A. FASSEL and A . P. d'SILVA (1980) Computer-controlled scanning monochromator for the

determination of 50 elements in geochemical and environmental sampl es by inductively coupled plasmaatomic emission spectrometry. Analytical Chemistry , 52. 2168-2173 .

FRENZEL B. (1966) Climatic change in the Atlantic/sub-Borcal transition on the Northern Hemisphere : botanicalevidence. In : World climate from 8000 to 0 B.C., Proceedings of the International Conference held atImperial College, London, Royal Meteorological Society, pp . 99-123.

GERASIMOV I. P. (1978) The past and the future of the Aral and the Caspian Seas . In : The l'nvironml'ntal history ofthe Near and MIddle East , W. C. BRICE, editor, Academic Press, New York, pp. 335-349.

Gll.ISWOLD W. J . (1979) Climatic variation and the SOCial revolutions in seventeenth century Anatolia. Papergiven at the 1979 International Conference on Climate and History; University of East Anglia.

HARDING A. F. (1983) Introduction: climatic change and archaeology . In: Climatic changes tn later prehistory; A.F. HARDING, editor. Edinburgh University Press. pp . 1-10.

HAy B. J. (1987) Part icle flux in the western Black Sea in the present and over the last 5000 years : Temporalvariability. sources , transport mechanisms. Ph .D. thesis, WHOI·87-44. 201 pp .. MIT-WHOI JointProgram.

HAy B. J . ( 1988) Sediment accumul ation in the central western Black Sea over the past 5100 years.Paleoceanography, 3. 491-508 .

HAyB. J. and S. HONIO (1989) Particle sources in the present and Holocene Black Sea . Oceanography. 2 , 26-31.HAy B. J ., S. HONIO, S. KEMPE, V. A . IrrEKKoT , E . T . DEGENS, T . KON uK and E . lZOAR (1990) Interannual

variability in particle flux in the southwestern Black Sea . Deep-Sea Research, 37, 911-928.HEUBERGER H. (1954) Gletschervorstosse zwischen Daun und Fernau-Stadium in den nordlichen Stubaier

Alpen . Zeit. Gleich , u. Glazialgeol .. 3, 91.HONJO S., B. J. HAY and Members of the Shipboard Party (1988) Temporal and Spatial Variability in

Sedimentation in the Black Sea: Cruise Report , R.V. Knorr 138-4. Leg 1 April 16-May 7, 1988. Black SeaSedimentation Data File . Vol. 2, WHOI Technical Report WHOI-88-35, and Piri Re is InternationalContribution Serie s No .6, 156 pp .

JONES G. A . (1990) AMS radiocarbon dating of sediments and waters from the Black Sea. E05, 71, 152.KHOTINSKlY N. A. (1985) Holocene clima tic change. In: Late Quaternary environments ofthe Sov iet Union, A . A .

VELlCHKO, editor, University of Minnesota Press, Minneapolis , pp. 305-309.LAMB H. H. (1977) Climate-s-past, present and future, Vol. 2, Climatic history and the future , Methuen, London,

835 pp .MANLEY G . (1966) Problems ofthe climat ic optimum: the contribution of glaciology. In: World climate from 8000

to 0 B.C.• Proceedings of the International Symposium, Imperial College , London, 1&-19 April 1966,pp.34-39.

MJMLAND G . (1956) Some laboratory experiments on the coccol ithophorid CoccolithlIS huxleyi. Oikos, 7,251-255 .

MORNER N.-A . (1971) The Holocene eustatic sea level problem. Geologie en Mijnbouw. SO. 699-702.MURRAY J. W. and E . lZDAR (1989) The 1988 Black Sea oceanographic expedition: overview and new discovenes.

Oceanography , 2,15-21.

MURRAY J. W.• H . W. JANNASCH, S. HON10, R . F. ANDERSON, W. S. REEBURGH . Z. Top , G . E . FRtEDRICH,L. A .CODISPOTI and E . IZDAR (1989) Unexpected changes in the oxidanoxic interface in the Black Sea. Nature,338,411-413.


MURRAY I. W. , Z. Top and E. Ozsor (1991) Hydrographic properties and ventilation ofthe Black Sea . Deep-SeaResearch, 38 (Supp!.) , S663-S689.

OSTROVSKIY A. B ., YA. A .lzMAYLOV, A . P . SHCHEGLOV, Kh. A . ARSLANOV, N.l. TRETICHNYV, N . A . GEY, T . Tu .PIOTROVSKAYA. Y . M . M URATOV. Y. YEoSHCHELlNSKIY,l. P. BALABANOV and S. I. SKIBA (1977) New data onthe stratigraphy and geochronology of Pleistocene marine terraces of the Black Sea coast, Caucasus, andKerch-Taman region . In : Paleogeograph y and deposits ofthe Pleistocene of the Southern Seas of the USSR,P. A . KAPLIN and I. A . SHCHERBAKOV. editors, Nauka Press, Moscow, pp. 61-<i9.

PILSKALN C . H . (1989) Geochemical composition of surface sediments and fluff layer. Black Sea Expedition1988. Unpublished data report, Monterey Bay Aquarium Research Institute .

Ross D . A . and E. T . DEGENS (1974) Recent sediments of the Black Sea . In : The Black Sea-geology . chemistry,and biology , E . T . DEGENS and D. A . Ross, editors, Memoirs of the American Association of PetroleumGeologists, 20, 183-199.

SCHOSTAKOVICH W . B . (1934) Bodenablagerungen der Seen und periodische Schwankungen der Naturerscheinungen (in Russian with German summary). In : Memoirs of the Leningrad Hydrological Institute, pp . 95110.

SHIMKUS K. M. and E . S. TRIMONIS (1974) Modern sedimentation in Black Sea . In : The Black Sea-geology,chemistry, and biology, E . T . DEGENS and D . A . Ross , editors, Memoirs of the American Association ofPetroleum Geologists , 20 , 249-278 .

SHIMKUS K. M., V. Y . MUKHINA and E . S. TRIMONIS (1973) On the role of diatoms in the Late Quaternarysedimentation in the Black Sea. Oceanology, 13,882-887.

SPEARS D. A. and R. KANARIS-SOTIRIOU (1976) Titanium in some carboniferous sediments from Great Britain.Geochimica et Cosmochimica Acta, 40, 345-351.

STARKEL L. (1966) Post-glacial climate and the moulding of European relief. In: World climate from 8000 tooB.C. , Proceedings of the International Conference held at Imperial College, London, Royal Meterological Society, pp . 15-33.

TOLMAZIN D . (1985) Changing coastal oceanography of the Black Sea . I: Northwestern shelf. Progress inOceanography, IS , 217-276.

TRIMONIS E . S. (1974) Some characteristics ofcarbonate sedimentation in Black Sea. In : The Black Sea-geology,chemistry, and biology , E . T . DEGENS and D . A. Ross , editors, Memoirs of the American Association ofPetroleum Geologists , 20, 279-295 .

TUREKIAN K. K. and K. H. WEDEPOHL (1961) Distributions of the elements in some major units of the earth'scrust. Geological Society ofAmerica Bulletin, 72,175-192.

UNLUATA U . and P. E . LAYIOLETTE (1990) Eddies and filaments associated with the coastal circulation in theBlack Sea . EOS, 71.135 .

WAKEHAM S. G .• J. A . BEIER and C. H. CLIFFORD (in press) Organic matter sources in the Black Sea as inferredfrom hydrocarbon distributions.

WALL D . and B. DALE (1974) Dinoflagellates in late Quaternary deep-water sediments of Black Sea. In: TheBlack Sea-geology, chemistry, and biology, E. T. DEGENS and D . A . Ross , editors, Memoirs of theAmerican Association of Petroleum Geologists, 20, 364-380. .

WEDEPOHL K. H. (1967) Geochemie. Walter de Gruyter & Co . Berlin. 220 pp.

Sediment deposition in the Late Holocene abyssal Black Sea with climatic and chronological implications

Documents