30. STABLE ISOTOPE INVESTIGATIONS OF THE MIOCENE … · 2007-05-17 · 30. STABLE ISOTOPE INVESTIGATIONS OF THE MIOCENE EVAPORITES AND PLIOCENE AND PLEISTOCENE SEDIMENTARY ROCKS AND

30. STABLE ISOTOPE INVESTIGATIONS OF THE MIOCENE EVAPORITES AND PLIOCENEAND PLEISTOCENE SEDIMENTARY ROCKS AND OOZES

PREFACE

During the thirteenth cruise of the Deep-Sea DrillingProject, carbonate oozes, diagenetically altered sediments,sedimentary rocks, and evaporites have been sampled in theMediterranean. The purpose of our investigation is to apply

the stable oxygen- and carbon-isotope techniques in orderto provide some data useful for the interpretation of thedepositional and diagenetic histories of these deposits. Thematerials were made available to us in connection with thepreliminary research necessary to assist in the preparationof the Initial Cruise Report of DSDP Leg 13.

30.1. PRELIMINARY ISOTOPIC INVESTIGATIONS OF SAMPLES FROM DEEP-SEA DRILLINGIN THE MEDITERRANEAN SEA

R. M. Lloyd, Shell Development Company (A Division of Shell Oil Company),Exploration and Production Research Center, Houston, Texas1

andK. J. Hsü, Swiss Federal Institute of Technology, Zurich, Switzerland.

ABSTRACTThe oxygen and carbon isotope composition of nine carbonate

and three sulphate samples from DSDP Leg 13 were analyzed toprovide data for an interpretation of the environments of theirdeposition and diagenesis. Analyses of dolomite and anhydritesamples yielded data in support of the geological deduction byRyan, Hsü, and others that the upper Miocene Mediterraneanevaporites were formed in desiccated inland basins after the Straitof Gibraltar was closed in the late Miocene.

INTRODUCTION

During Leg 13 of the Deep Sea Drilling Project in theMediterranean, a number of diagenetically altered sedi-ments and sedimentary rocks were sampled. (See Chapter 38,this volume, for location map.) The purpose of thisinvestigation is to apply stable oxygen- and carbon-isotopetechniques to provide some data useful for the interpreta-tion of the depositional and diagenetic histories of thesedeposits. The materials were made available to us inconnection with research to assist in the preparation of theInitial Cruise Report of DSDP Leg 13. Our investigationshave been jointly supported by the Shell DevelopmentCompany and the Swiss Federal Institute of Technology.

MEDITERRANEAN EVAPORITES

One of the most significant results of the DSDPMediterranean cruise is the discovery of an upper Miocene

1 (EPR Publication No. 608).

evaporite formation, which underlies most parts of theMediterranean. The origin of the evaporite is, however, aquestion of some controversy. The following three alterna-tive hypotheses have been advanced.

1) Deep-water, deep-basin model. This hypothesisassumes the deposition of evaporite minerals in a deep-water basin (Schmalz, 1969). At the time of evaporitedeposition, the Mediterranean is assumed to have been adeep-water basin, not isolated from the Atlantic, butseparated from the latter by a shallow sill. However, thecirculation was sufficiently reduced to cause an increasedsalinity in the Mediterranean. Eventually, carbonates, sul-fates, and halite were crystallized out of brine andaccumulated on the deep basin floor to form the suppos-edly deep-water evaporite.

2) Shallow-water, shallow-basin model. This hypothesisassumes the deposition of evaporite minerals on the bottomof a shallow restricted shelf sea, which may or may nothave an open connection with the Atlantic. The presentdepth of the Mediterranean was related to post-Miocenesubsidence, subsequent to formation of the upper Mioceneevaporites.

783

R.M.LLOYD, K.J.HSU

3) Desiccated deep-basin model. This hypothesisassumes the deposition of evaporite minerals on inlandplayas, whose flat basin floors were thousands of metersbelow the Atlantic sea level. The playas owed their origin tothe desiccation of the late Miocene Mediterranean, when itwas completely isolated from the Atlantic. The return ofnormal marine conditions in early Pliocene led to thedeposition of deep-water pelagic oozes on top of the playaevaporites.

The arguments, pro and con, have been discusssed inother sections of this volume. The evidence on the wholemay be considered to favor the desiccated deep-basinmodel. Isotopic analyses were carried out to throw additi-onal light on this problem.

Oxygen and carbon isotope data for Miocene dolomitesand dolomitic calcites from the Mediterranean evaporiteformation and calcites from overlying Pliocene sedimentsare given in Table 1. These data are compared in Figures 1and 2 with analyses reported by Fontes et al. (Chapter 30.2this volume) from the Mediterranean evaporite and withvalues for Holocene carbonates from lacustrine and marineMiocene environments. Oxygen isotope data on mineralsulfates from the Miocene evaporite formation are given inTable 2. These data are compared with values for marineand lacustrine sulfates in Figure 3.

TABLE 1Isotopic Composition of Upper Miocene and Lower Pliocene

Carbonate Minerals, Mediterranean

SampleLocation No. Description

PDB

Mineral δθ18 δC 13

AlboranBasin

LevantineBasin

TyrrhenianBasin

121-19-1

121-21-1

129-2-1

132-21-2

132-22-1

UM dolomiticlimestone

UM dolomiticlimestone

UM densedolomite

UM Sucrosedolomite

Pliocene redooze

Pliocene redooze

Calcite

Calcite

Dolomite(Ca5 3)

Dolomite(Ca5 3)

Calcite

Calcite

-1.6

0.5

1.6

3.4

1.3

1.2

-3.4

-0.4

-10.2

-35.8

0.1

0.2

MIOCENE CARBONATES

Dolomite occurs in the Mediterranean Miocene evaporitedeposits in quantities varying from traces to greater than 90weight per cent. The dolomite is fine-grained and containsexcess calcium carbonate, up to 55 mol percent. It is foundin both soft marly sediments and in indurated crusts.Layers of relatively pure dolomite are rare and it is usuallya minor component of the sediment, on the order of 10 percent or less (Fontes et al, Chapter 30.2 this volume). Theremaining sediment consists of varying admixtures ofcalcite, quartz, feldspar, gypsum, and a variety of clayminerals.

The most striking feature of the carbonate isotope datais its variability (Figure 1). This variability contrasts sharply

with the rather narrow range of values for Holocenedolomites and dolomitic calcites found in tidal flat depositsassociated with shallow-water marine deposition. The varia-bility compares favorably with data reported for dolomitesand calcites from lacustrine evaporite deposits, especiallythe data from a group of Pleistocene lakes from West Texasdescribed by Parry et al. (1970).

The West Texas samples are from a group of isolatedlake basins which formed in depressions eroded intoPliocene deposits of the Texas high plains. The basins havemaximum dimensions on the order of a few miles and arescattered over an area of 5000 square miles. The depositsrange in age of from 12,000 to over 37,000 years. Dolomiteand calcite are found mixed with quartz, feldspar, clays,and minor amounts of celestite and gypsum. There isalmost as much scatter in the isotopic data for individuallake basins as there is among all of the basins. The authorsconclude that: "the wide scatter in isotopic compositionsof the dolomites indicates that they formed from solutionsof widely differing isotopic compositions and temperatures;conditions which could be expected in isolated desiccatingfluvial lake systems in which evaporation is extreme" (Parryet al, 1970, p. 830).

According to a recent proponent of the deep-basin,deep-water evaporite formation hypothesis: 'There are noactive deep evaporite basins today,.. ."(Schmalz, 1969, p.822). It is not possible, therefore, to make simple isotopiccomparisons between this environment and the Mediterra-nean carbonates as we did above for the shallow-marine andplaya lake environments. However, we may speculate onpossible isotopic variations on the basis of the proposedmodel for deep-water evaporite formation.

The essential element of this model is the existence of ashallow sill at the rim of the basin which allows newseawater to enter in a surface layer and dense brine toescape in a lower layer in a continuous refluxing system.Though there are details to the model which are necessaryto explain the sequence of salts deposited, the essentialisotopic factors are:

1) The source of water and salts is seawater from theoceanic reservoir.

2) Concentration of salts is by evaporation from afree-water surface.

3) Because of density stratification, the floor of thebasin can, at times, be stagnant.

Ocean water has a constant oxygen isotopic compositionof about 0.0 per mil on the PDB scale. At earth surfacestemperature (~25° C), calcites formed in equilibrium withsuch water would have δ θ 1 8 values of -2 per mil anddolomites δ θ 1 8 values of -1 to +3 per mil.2 Evaporationfrom a free-water surface causes isotopic enrichment ofseawater to an upper limit of about 6 per mil (Lloyd,1968a). Such enrichment would effect carbonate valuesdirectly. Thus, calcites might range from -1 to +4 anddolomites from 0 to +9 per mil2 in a deep evaporite basinenvironment.

The range is given to accommodate the unresolved differences ofopinion among isotope workers as to how much, if any, isotopicfractionation exists between dolomite and calcite.

784

30.1. ISOTOPIC INVESTIGATIONS

LACUSTRINE

WEST TEXAS (1

UTAH (2) —

-25 -20 -15

136C PDB

-10 -5 10

gogoo o

MARINE

BAHAMAS (3)

FLORIDA (3 ) -

BONAIRE — • ü •OOD<X>Q O O

O

• DOLOMITEo CALCITE

MEDITERRANEAN EVAPORITE

THIS REPORT--35

FONTES, e t al oS ooo* 8S> $

Figure 1. Oxygen isotope data for dolomite and calcite from lacustrine, marine, and Mediterranean evaporite samples. Datafrom the literature: (1) Parry, et al. (1970); (2) and (3) Degens and Epstein (1964); Fontes, et al. (this volume).

These ranges accommodate the more positive values forthe Mediterranean Miocene carbonates, but fail to accountfor the very negative oxygen isotope values found in manyof the samples from the same unit.

Some of the observed variation in carbon isotope valuesof the Mediterranean samples might be accommodated bythe deep-water, deep-basin hypothesis if we accept thepossibility of periods of stagnation of bottom waters.During such periods the normal bicarbonate carbon ac-quired by the water mass at the surface could be contami-nated by more negative carbon from CO2 generated byorganic decay. It should be pointed out, however, thatperiods of stagnation were proposed by Schmalz (1969) toaccount for euxinic sulfide-rich deposits in ancient evapor-ite basins. We are unaware of similar sediments in theMediterranean sequence.

The large negative C13 value of the Levantine (Site 129)dolomite sample is very unusual. The closest match to thisis one from a deep-water open marine dolomite, whosegenesis was probably related to the bacterial breakdown ofhydrocarbons (Russell et al., 1967). Limestones associatedwith Sicilian sulfur deposits in the Solfifers Formation (anequivalent of the Mediterranean Evaporite) likewise havehighly negative C13 values (-8.8 to -43). These limestonesare also believed to have derived their carbonate primarilyfrom oxidized methane enriched in C12 (Jensen, 1968). Inany event, such extreme values speak neither for nor againsta deep-water origin.

PLIOCENE CARBONATES

The Miocene evaporite series is directly overlain byPliocene carbonate-rich open marine sediments with abun-dant pelagic organisms and a normal marine benthonicostracod fauna. Our isotope values (Table 1) and elevenreported by Fontes et al. (Chapter 30.2 this volume) all fallwithin a narrow range of 0.2 to 1.3 per mil for δ θ 1 8 and-0.1 to +1.2 mil for carbon relative to PDB.

The isotopic composition of the oozes may be construedto indicate formation in cool normal marine waters, on theorder of 10° C. However, partial evaporation of theMediterranean water may be partly responsible for thepositive δ θ 1 8 values. The present-day waters in theMediterranean have a δ θ 1 8 value of about 1.2. If thePliocene oozes were deposited from waters of similarisotopic composition, our results would suggest that thelargely planktonic skeletons were crystallized from a watercolumn with an average temperature of about 15°C. Benson(Chapter X, this volume) studied the ostracod fauna andfound the ostracod assemblage "most likely to occur livingin open ocean between 1000 and 1500 meters (bottomtemperatures between 4° and 6°C)." This is consistent withthe postulate that the Pliocene bottom waters were muchcooler than those of the present-day Mediterranean andthat the first Pliocene waters flooded deep, but dessicatedMediterranean basins.

785

R. M. LLOYD, K. J. HSU

LACUSTRINE

WEST TEXAS (1)

-5

1860 PDB

10T I I 1 I I I I I

99

UTAH (2)• %o oo

MARINE

BAHAMAS (3)•

FLORIDA (3)-

BONAIRE —

• DOLOMITE

o CALCITE

MEDITERRANEAN EVAPORITE

THIS REPORT-

FONTES, e t al- o o o o o oo

oo m m —t

Figure 2. Carbon isotope data for dolomite and calcite from lacustrine, marine, and Mediterranean evaporitesamples. Data from the literature: (1) Parry et al. (1970); (2) and (3) Degens and Epstein (1964); Fontes et al.(this volume).

TABLE 2Isotopic Composition of Sulphate from Mediterranean Evaporite

LocationSample

No. Description Mineral SMOW

BalearicBasin

TyrrhenianBasin

124-8-1

124-10-1

132-23-1

UM laminatedanhydrite

UM nodularanhydrite

UM laminatedanhydrite

Anhydrite

Anhydrite

Gypsum

16.7

15.8

3.8

MIOCENE SULFATES

The most interesting feature of the sulfate isotope valuesfrom the Miocene evaporite is again their variability (Figure3). The fact that the values appear to avoid the range ofpresent-day marine sulfates is only an accident of sampling.Fontes et al (Chapter 30.2, this volume) report a value of14.3 for a Miocene sulfate from Site 122.

Our knowledge of oxygen isotope variation in sulfatas isvery limited. Marine oceanic sulfate has a very constantδ θ 1 8 value of about 10 per mil relative to the SMOWstandard (Longinelli and Craig, 1967; Lloyd, 1967). Lloyd(1968b) has proposed that this represents a steady-statevalue in the oxidation-reduction sulfur cycle of the ocean

rather than the true equilibrium exchange value with oceanwater. Exchange rates between water and sulfate areextremely slow at earth surface conditions. Therefore,mineral sulfates precipitated from evaporated marine waterswould tend to reflect the constant value of marine sulfate(plus a fractionation of crystallization) rather than theconditions of temperature and water isotopic compositionat the time of precipitation, which could vary substantiallyfrom place to place (Lloyd, 1968b). This accounts, then,for the fact that gypsum derived from marine wateroccupies a narrow range from 13 to 15 per mil relative toSMOW (Figure 3).

Gypsum samples from lacustrine environments have awider spread of values and are, in general, more positive.This may represent inherited values from older sulfatesdissolved from outcrop (Longinelli, 1968), or perhaps, adifferent isotopic balance determined by local varieties inthe sulfur cycle oxidation-reduction systems.

The fact that the Miocene sulfate samples show notendency to group near the values for marine-derivedsulfates argues against both the shallow-basin, shallow-waterand the deep-basin, deep-water models. For mineral sulfatesto match the one very negative Miocene sample (9 per mil),there is a report of a thermal spring water containingdissolved sulfate with a δ θ 1 8 of 4.8 per mil (Lloyd, 1967).Gypsum precipitated from this solution would have a valuenear 9 per mil.

786

30.1. ISOTOPIC INVESTIGATIONS

10

δO ' SMOW

15 20I \ \ I

HOLOCENE

MARINE-

LACUSTRINE-

MEDITERRANEAN

EVAPORITE

Figure 3. Oxygen isotope data for oxygen in sulfates. Marine data from Lloyd (1967). Lacustrinedata from Play a deposits West Texas and New Mexico.

DISCUSSION

Because of the variability of data, we cannot define , inany exact sense, the environment of formation of anyportion of the Mediterranean Miocene evaporite deposits.However, it is this great variability which support thehypothesis that deposition occurred in a desiccated basinconsisting of playas, residual salt ponds, and isolatedephemeral lakes covering a broad flat area of manythousands of square miles. That fresh water from rain andrun-off was an important contributor to the basin isdemonstrated by the negative carbon and oxygen isotopevalues for many of the carbonate samples.

The desiccated playa environment might have existed aspart of a shallow basin. However, the rapid onset over theentire basin of uniform open marine conditions, indicatedby the Pliocene carbonate isotope values, supports theconclusion that the basin was deep.

REFERENCES

Degens, E. T. and Epstein, S., 1964. Oxygen and carbonisotope ratios in coexisting calcites and dolomites fromrecent and ancient sediments. Geochim. Cosmochim.Acta. 28,23.

Lloyd, R. M., 1967. Oxygen isotope composition of oceanicsulfate. Science. 156, 1228.

, 1968a. Oxygen isotope enrichment of sea waterby evaporation, Geochim. Cosmochim. Acta. 30, 801.

Jensen, M. L. 1968. Isotopic geology of gulf coast andSicilian sulfur deposits Geol. Soc. Am., Special PaperNo. 88. 525.

, 1968b. Oxygen isotope behavior in the sulfate-water system,/. Geophys. Res. 73, 6099.

Longinelli, A. and Craig, H., 1967. Oxygen-18 variations insulfate ions in sea water and saline lakes. Science. 156,56.

, 1968, Oxygen isotopic compositions of sulfateions in water from thermal springs. Earth Plan. Sci.Letters. 4, 206.

Parry, W. T., Reeves, C. C, Jr. and Leach, J. W., 1970.Oxygen and carbon isotopic composition of West TexasLake carbonates. Geochim Cosmochim. Acta. 34, 825.

Russell, K. L., Deffeyes, K. S., Fowler, G. A. and Lloyd, R.M., 1967. Marine dolomite of unusual isotopic composi-tion. Science. 155, 89.

Schmalz, R. F., 1969. Deep-water evaporite deposition: Agenetic model, Bull. Am. Assoc. Petrol. Geologists. 53,798.

787

J-CH. FONTES, R. LETOLLE, W. D. NESTEROFF, W. B. F. RYAN

30.2. OXYGEN, CARBON, SULFUR, AND HYDROGEN STABLE ISOTOPES IN CARBONATE ANDSULFATE MINERAL PHASES OF NEOGENE EVAPORITES, SEDIMENTS, AND IN

INTERSTITIAL WATERS

J-Ch. Fontes, R. Letolle, W. D. Nesteroff, Laboratoire de Géologie Dynamique, University of Paris, Veand

William B. F. Ryan, Lamont-Doherty Geological Observatory of Columbia University, Palisades, New York

INTRODUCTION

The relative abundance of certain elements and theirstable isotopes sometimes indicates the environmentalconditions under which discrete mineral phases originate. Inthe case of the sediments and sedimentary rocks recoveredduring Leg 13, we were particularly anxious to examine theenvironment of formation of the late Miocene (Messinian)evaporites.

The evaporite minerals calcite, dolomite, anhydrite, andgypsum were available from Sites 122, 124, and 134 in thewestern Mediterranean; Site 132 in the central TyrrhenianBasin; and Sites 125 and 129 in the eastern Mediterranean(see Figure 1). The materials consist of unaltered authigeniccomponents (fine-grained anhedral dolomite in the graymarls of Cores 6, 7, 10, 11, and 13 of Hole 124) andminerals formed through secondary replacement and hydra-tion (crystals of selenite in the Core Bit sample of Holes122 and 125A). The scope of this preliminary researchincluded determination of the source of water whichsupplied the brine basin during precipitation of at least partof the evaporite series, and elucidation of the conditionsleading to replacement of anhydrite by gypsum or to therecrystallization of gypsum.

A second objective was an evaluation of the isotopiccomposition of carbon and oxygen in the bulk carbonate ofthe Serravallian open-marine sediments below the evapor-ites, and the Pliocene and Pleistocene pelagic oozes above.The purpose of this effort was to determine the environ-mental setting of the Mediterranean Basin before and afterthe late Miocene "crisis of salinity" (Ruggieri, 1967).

Sample Coverage

In the time allowed for this study we were able to make*8 and δ C i 3 analyses on 7 samples of Serravallian

sediments from Holes 126 and 129, 12 samples of thePliocene and Pleistocene pelagic oozes from Holes 125A,132 and 134, and 32 samples of Messinian calcites anddolomites. The results of these analyses, grouped bydrillhole, are listed in Table 1. Also included is themineralogy of these samples and an approximation as to theamount of calcite and dolomite present in each sample.

Two complete δ θ 1 8 and δ S 3 4 analyses of sulfate ions ingypsum from drill bit samples of Holes 122 and 125A weremade, as well as δ θ 1 8 and δD analyses of water ofhydration (Table 2). δ θ 1 8 and δD analyses of interstitialwaters from dolomitic marls of the evaporite sequence inHole 132 were also made (Table 3).

ISOTOPIC COMPOSITION OF THE EVAPORITES

The evaporites illustrated in Figure 1 include layers ofcarbonate marls, anhydrite, gypsum, and halite. The mostcomplete sections are from Site 124, on the Balearic Rise,and Site 132, in the Tyrrhenian Basin. With the exceptionof some minor amounts of gypsum in Core 7, the calciumsulfate phase at Site 124, is almost entirely anhydrite (seechapters 6, 21 and 22.1). At Site 132, however, similaranhydrite has been completely replaced by gypsum (seeChapter 22.1). Present at both of these sites, and also atSites 125A, 129, and 134, are interbeds of steel blue togray marls, which are occasionally finely laminated, andpossess a variable amount of fine-grained anhedral dolo-mite. Similar dolomitic and calcitic marls were recoveredfrom Holes 125A, 129A, 134 and 134D.

Oxygen and Carbon Isotopes in the Carbonates

At times calcite and dolomite occur together m theevaporite marls. When fossil tests are present they arealways calcitic. Siderite was identified in Sample132-26-1-20 cm. The total carbonate content is usually lessthan 30 per cent; the other components consist ofterrigenous clastic minerals (fine-grained quartz, clay miner-als), pyrite, and disseminated organic matter. Smectites arethe most important clay minerals. The relative proportionsof other minerals do not seem to change significantly withchanges in the abundance of gypsum and anhydrite, as ifthey represent some kind of background sedimentation.

The isotopic compositions of the carbonates are dis-persed between two extremes: pure calcite and puredolomite.

Calcites: The pure calcites, depicted by the solid dots inFigure 2, exhibit: (1) low values of δ θ 1 8 and δ C 1 3 incomparison to normal marine pelagic oozes (open circles),(2) distinctly lower δ θ 1 8 than the Messinian dolomites,and (3) a large scatter of δ θ 1 8 and δ C 1 3 values, with allthe values grouped in the quadrant of negative oxygen andnegative carbon ranges.

Because the Pliocene and Quaternary pelagic oozes allcluster in the quadrant of positive oxygen and carbonranges, it is isotopically easy to delineate the Miocene/Plio-cene boundary (see also Figure 2 of Chapter 30.3). In Hole132 where the Messinian marls in Core 21-2 are calcitic, theevaporite marl/pelagic ooze contact is accompanied by apronounced decrease in both δ θ 1 8 and δ C 1 3 . In Hole 134there is a similar change. However, in Hole 125A, from theMediterranean Ridge in the Ionian Basin, the underlyingmarls are dolomitic, with no calcite, and thus we observea decrease only in δ C 1 3 (i.e. a trend toward more negativevalues).

788

30.2. CARBONATE AND SULFATE ISOTOPE STUDIES

Figure 1. Location of the drill sites from which samples have been taken for the measurements of stable isotopes.

In the apparently continuous section at Site 132 in theTyrrhenian Basin, the isotopic shift at the Miocene/Plioceneboundary, in Core 21-2, from negative to positive values isobviously related to the abrupt termination of evaporiteconditions and the filling of the basin with normal marinewater. The shift can be explained as being the result of aninflux of cold North Atlantic deep water, replacing awarm water body of meteoric origin, already greatly de-pleted in the heavy isotope of oxygen. The concept thatthe terminal phase of the evaporite epoch was accompaniedby the widespread occurrence of alkaline lakes is collabo-rated by the finding of the "Melanopsis-Cyprideis" faunain the upper Messinian "lago mare" facies of northernSicily (Ruggieri and Greco, 1965). We doubt that the iso-topic shift can be entirely due to temperature becausethere is also a marked change in the carbon isotope ratiofrom negative to positive values. Yet the C 1 3 fraction-ation with temperature is known to be much lower thanO 1 8 fractionation (Mook, 1971).

Dolomites: The marls characterized by the presences ofdolomite without calcite (solid triangles in Figure 2) are, inthe vast majority of cases, isotopically distinct from thecalcites. Except for three samples from Core 25 of Hole132, all of the dolomite samples have quite highly positiveδ θ 1 8 values, with a mean of about + 4 °/oo PDB. With oneexception, the Site 124 (Balearic Rise) dolomites all grouprather tightly (i.e. 3.34 > δ θ 1 8 > 2.14 and 2.16 > δ C 1 3 >0.27) in the positive quadrant of Figure 2 within the rangeof dolomites related to evaporite environments (Clayton etal, 1968).

The one negative measurement is from Core 7 of thisdrill hole; the sample has a δC l 8 of -4.73 ‰ and isassociated with secondary gypsum replacement in the form

of selenite (see Figure 18A of Chapter 22.1). In this casethe carbonate is a mixture of calcite and dolomite and thelow O 1 8 content can be explained by the predominance ofan isotopically light calcite (calcisparite). In each of thesesamples the δ C 1 3 is also negative, reaching -6.34 °/oo inSample 132-25-2-70 cm. This suggests that meteoritic waterhas probably played an important role in the formation ofthe dolomite—perhaps as secondary dolosparite.

Some very low δ C i 3 values have been obtained insamples from the eastern Mediterranean Basin. Both sam-ples, 129-2-1 (Strabo Trench) and 125A-DB (MediterraneanRidge — Ionian Basin), contain faunas indicative of abrackish water environment (see documentation in Chap-ters 7 and 10). Thus it is easy to accept that these ratiosreflect the input of fresh water into the eastern basins andthe development there of Alkaline lakes in late Messiniantime (see Chapter 36.2).

However, the low δC13 values may also be attributed tothe contribution of carbon dioxide of biogenic origin in thebicarbonate ion system. In fact, sample 134-10-1 (δC13 =-9.57 °/oo) is from an interval rich in gasoline rangehydrocarbons (see Chapter 32). The suggestion of biologicalinfluence has been presented by Dessau et al. (1960) andCheney and Jensen (1965) for carbonates connected withthe Sicilian evaporite beds and by others (Russell et al.,1967; Hathaway and Degens, 1970; Deuser, 1970) forvarious sedimentological systems at other locations.

Whereas the primary isotopic composition of calcitesmay change during diagenesis, dolomites generally do notundergo very much exchange (Fritz, 1971). The negativeδC13 values for the dolomitic marls of Sites 125A and 129are lower than those given by Fritz (1970) for secondarydolomite and suggest that some of the very fine anhedral

789

J-CH. FONTES, R. LETOLLE, W. D. NESTEROFF, W. B. F. RYAN

TABLE 1Oxygen and Carbon Isotopic Composition of Selected Samples

of the Evaporitic Marls and Pelagic Sediments

WESTERN MEDITERRANEAN DRILLSITES

Site 122 - Valencia Trough Balearic Basin

Age Sample

162 my Core Bit

£83S 8

Per Cent

Calcite Dolomite

T

Mineralogical Content andOther Components

n.d.

δO δ CvsPDB

-3.24 +0.02

Site 124 - Balearic Rise

-350m.y. 67

„ 10-1 (113)g 10 CCo 10 CC

§ 11-1 (124)13-2 (35)

0+00000

60++++

85++

I, Ch, Q, M, GG, Q + n.d.n.d.n.d.n.d.Q, I, M, Kn.d.

Balearic Basin

+2,14X

-4,73+3,34X

+2,42X

+2,56X

+3,08x

+3,29X

+1,27-2,43-2,16

+2,15+2,22+1.74+0,27

Site 132 - Tyrrhenian Rise

g 19-3 (100)8 20-4 (40)a 20-4 (90)‰ 21-2 (60)

-88 21-2 (76)25 my 21-2 (100)

22-1 (78)22-1 (140)23-1 (120)25-1 (50)25-1 (90)

g 25-2 (70)8 25-2 (110)| 25-2 (135)S 25-2 (140)

26-1 (20)26-1 (24)26-1 (110)

60806030

612

4+20+0

n.d.000

n.d.n.d.

0000

020008+2

n.d.100

15n.d.n.d.

I, K, QI, Qi> QI, Q

M, I, Chi, GI, M, Chi, G, Kn.d.G + n.d.G, Q, I, ChiG, I, Q, Chi, K, Mn.d.G, Q, I, Chi, Mn.d.n.d.M, I, Q, K, Chi, FI, M, Q, K, Chi, Sn.d.n.d.

Tyrrhenian Basin

+0,24+1,33+1,05+1,12

-1,59-1,09-2,67-3,72-6,37

-2,44-4,87-5,85-1,17X

+0,19x

-2,70-1,11

+0,77+1,25+1,09+0,59

-1,37-1,66-1,29-4,23-6,53

-2,60-6,34-4,0-2,66

0,00-6,66-4,49

Site 134 — Balearic Abyssal Plain

g 3-1 (126)8 3-2 (130)a 3 CC‰ 5CC

-326m.y. 7-5 (140)6 g 7 CC (24)

§ 8 10-1 (115)

++++

+150

0000

+15++

n.d.n.d.n.d.n.d.

n.d.Q, I, M, Chi.A, H, Q

Balearic Basin

+0,68+0,91+0,99

+0,10-1,76

+4.60x

+0,25+0,55+0,82

+0,08-1,32-9.57

Site 134 - Western Sardinia Slope

-183 my 1-1 (90)ó g 1-1 (144)

14T

2T

G, M, I, Ch, QG, Q, I, Chi, M

Balearic Basin

+2.41+2.96

+1.93+1.92

dolomite in the evaporite marls might indeed be an earlydiagenetic product.

A more detailed petrographic analysis is needed to fullyevaluate the problems of diagenesis. However, if, as webelieve, some of the dolomites and calcites have preservedtheir original primary isotopic composition it must be

accepted that the Mediterranean Sea at certain times wasfed predominantly from rivers and rainfall and that theevaporite basins were isolated from the open sea. The greatvariations in the isotopic composition of the calcite indicatethat the influence of fresh water was intermittent. In thecase of the dolomites of Site 124, the mineral phase was

790

30.2. CARBONATE AND SULFATE ISOTOPE STUDIES

TABLE I-Continued

EASTERN MEDITERRANEAN DRILLSITES

Site 125 A

Age

Plio

-ce

ne

-81 my

- Mediterranean Ridge

Sample

2-2 (66)5-3 (48)6-1 (25)

6-1 (150)7 CC (140)9-1 (113)Drillbit

Calcite

606465

0000

Per CentDolomite

04

10++25++++

Mineralogical Content andOther Components

G, Q, II, Q, KI, Q,K

n.d.I, Q, K, Fn.d.A + n.d.

Ionian Basin

δ θ 1 8

+0,37+1,10+1,48

+5,20x

+4,3 8*+5,23*+5,09*

δ C 1 3

vsPDB

-0,04+0,91+1,23

-1,08-0,92-3,28

-11,66

Site 126 - Cleft in Mediterranean Ridge Ionian Basin

1 1-4 (90)

1!-106m.y. 5-1 (130)

. e 5 C C

g J 6-1 (130)$ | 6CC

60

12634

4

0000

I, Q, K, Chi

M, K, Q, IM, K, QM, K. I, QQ, + n.d.

+1,35

+0,48+0,72+2,71+0,19

+0,49

+0,27+0,33+0,65-2,83

Site 129 - Strabo Trench Levantine Basin

g 2-Centerbit8 2-1 (136)| 3-1 (49)S 3CC

++

0n.d.15

025n.d.13

n.d.Q, M, Chi, I, Gn.d.M, Q, I, Chi, K

-6,81+l,60 x

-0,83+0,17

-1,92-26,90

-8,22-7,51

Site 129A - Strabo Mountains

| g 2CC 15 10 M, Q, I, Chi

Levantine Basin

-2.21 -1.97

LEGEND: T = trace, n.d. not measured. G = gypsum, I = illite, M = montmorillonite, Q = quartz, K = kaolinite, F = feldsparsChi = chlorite, A = clays, H = halite, S = siderite.

The mineralogical components are given in the order of their relative importance. The isotopic values for pure dolomites (x)have been corrected for isotopic fractionation effect as given by Sharma and Clayton (1965).

TABLE 2Isotope Composition of Gypsum

Site

122125A

Location

Valencia troughIonian Basin

Sample

Core BitCore Bit

Facies

Selenite gravelSelenite chips +++traces Fe

Crystallization Water

δ θ 1 8 a

+4.07

+3.89

δ D a

-16.0

-8.5

Sulfate Ion

δS34t>

+22.0

+22.6

δ θ l 8 a

+ 14.3

vs SMOW

TABLE 3Isotope Composition of Interstitial Waters

Site

132132

Location

Tyrrhenian BasinTyrrhenian Basin

Sample

25-2125 cm25-1,15 cm

Dry Residue

53.8 0/0055.0 0/00

Pore Water

δ θ l 8 a

-0.78-0.88

δ D a

+6.4+4.3

1 vs SMOW

'vs CD

791

J-CH. FONTES, R. LETOLLE, W. D. NESTEROFF, W. B. F. RYAN

© 1

• 2

• 3

Δ 4

x 5

- 2 5

Λ

Figure 2. A plot of δ O 1 8 against bC13 for samples investigated. 1 = Pliocene and Quaternary pelagic oozes, 2 = UpperMiocene (Messinian) dolomites, 3 = Upper Miocene (Messinian) calcites, 4 = A mixture of the calcites and dolomites, and 5= Calcites of Middle Miocene (Seπavallian) age. The dashed line represents the function 2(bC13 + 50) - 0.5(òO18 + 50)given by Keith and Weber (1964) to separate carbonates of marine origin (right of the line) from carbonates of fresh-waterorigin (left).

apparently crystallized from concentrated water with amodified ionic content.

Oxygen, Sulfur, and Hydrogen Isotopes in the Sulfates

For natural samples, the effects of isotopic fractionationbetween the sulfate ion (SO4") and the water of crystalliza-tion are difficult to relate to experimental data (Longinelliand Craig, 1967; Lloyd, 1967; and 1969; and Misutani andRafter, 1969).

Oxygen 18 in the Sulfate Ion: Studies of the data δ O 1 8

values of brines and of gypsum in saline ponds borderingthe Mediterranean has shown that the data δC l 8 content ofthe sulfate ion has been enriched; also precipitated gypsumhas an enriched data δ O 1 8 of+14 to 15 °/oo (Fontes andSchwartz, unpublished data). We have obtained values of14.3 and 15.1 % o in the sulfate of core bit samples of Site122 and 125A. We interpret those values as an indication of

δ θ l 8 enrichment due in the processes of brineconcentration.

It is generally admitted (Longinelli and Craig, 1967) thatin the range of biological temperatures and pH, isotopicexchange between the oxygen in sea water and that in thesulfate ion is very slow, with a half reaction time in therange of 100 to 1000 years. It appears that under peculiarconditions of confinement (i.e., inhibited circulation,strong evaporation, high temperatures and high brineconcentration) sulfates should more rapidly initiate anisotopic equilibration with the water.

Consequently, and as a present working hypothesis, onecan reasonably assume that the measured values of + 14.3and 15.1 °/oo correspond to precipitation of calcium sulfatein situations well isolated from the open sea for which,however, the original supply of sulfate ions was from theopen sea.

792

30.2. CARBONATE AND SULFATE ISOTOPE STUDIES

Sedimentwaters

smowM.W.

50C.W.

_ 5 + 5 δ1 8/smow %o

Cristallization water

Figure 3. A plot ofδD against δθ18 for the interstitial waters of Cores 132-25 and 132-26 (+ signs) and the crystallizationwaters of the Core Bit samples from Sites 122 and 125 A (solid dots). The insert in the upper left hand corner shows thecorrection (because of fractionation) to be applied to convert the isotopic value of crystallization water (C.W.) to itsmother water (M.W.) value. As an example, if present Mediterranean sea water (M.S.) is concentrated by evaporation toprecipitate gypsum, the M. W. circle shows the isotopic range of the saturated mother water and the C. W. circle the rangeof the crystallization water in the newly formed gypsum. At the lower left are the localization lines for worldwideprecipitation (solid line) and eastern Mediterranean precipitation (dashed line).

Sulphur 34 in the Sulfate Ion: Two samples of gypsum,one from Site 122 and one from Site 125A, have provided6S34 values of + 22.0 and + 22.6 %>o vs. Canon Diablo;these are similar to measured values from recent marinesulfates. Because δS34 values have remained quite constantsince the beginning of the Tertiary (Nielson, 1965 andNielsen and Rambow, 1969), and since the older Mesozoicand Paleozoic salt deposits of Europe and North Africa arecharacterized by much lower δS34 values, it is unlikely thaterosion of older sedimentary rocks has supplied the sulfateto the Messinian basin. It is more probable, in light of whathas already been concluded about the oxygen isotopes, thatthe sulfate ions were derived directly from the open ocean.

Oxygen 18 and Deuterium in the Water of Crystalli-zation: Isotope fractionation during the hydration ofanhydrite to gypsum is known to be an equilibrium ornear-equilibrium process (Gonfiantini and Fontes, 1963;Fontes and Gonfiantini, 1967). In the interval of + 15°C to+ 60°C, which covers the ordinary range of temperaturesfor the precipitation of evaporites, the fractionation is alsoknown to be independent of temperature.

The following enrichment factors of the isotopes be-tween the mother water (Vine) and crystallization water canbe defined: e θ l 8 +4.0 ‰ e D = -16 °/oo wi the-δ(mother water ) -δ (crystallization water ).

Present-day values close to +8%o for δ θ 1 8 and +50%ofor δD vs. SMOW are obtained for mother water of directlyprecipitated gypsum from the concentration of Mediterra-nean sea water, when the original liquid has been concen-trated to one-fifth its original volume (Fontes, 1966).

As shown in Table 2, crystallization waters of theselenite (gypsum) crystals from Sites 122 and 125A, thoughsimilar in isotopic composition, cannot correspond toevaporatively concentrated, calcium-sulfate-saturated seawater. These evaporite specimens have quite obviouslyundergone recrystallization. The macro- and microscopicfabric of the sample from Site 125A (see Figures 14 and 15of Chapter 7), which indicates complete alabastrization,supports this conclusion.

The question now arises as to the origin of the waterleading to the crystallization. If, as shown in Figure 3, wetake the isotopic values of the measured crystallizationwater (solid dots) and correct these data for fractionation

793

J-CH. FONTES, R. LETOLLE, W. D. NESTEROFF, W. B. F. RYAN

(open circles) we arrive at a composition for the ancientmother water far different from what would be obtained bytaking present day Mediterranean water (circle with a cross)and evaporating it. A clue to the origin of the ancientmother water comes from an examination of interstitialwater in the evaporitic marls.Oxygen 18 and Deuterium in Interstitial Waters: It turnsout that the isotopic composition of connate waters ofMessinian age, extracted by squeezing sediments from Cores25 and 26 at Site 132 (see Table 3 and plus sign in Figure3), lies close to the values of the crystallization watercorrected for fractionation during the formation of gypsum(open circles of Figure 3). The isotopic composition of theinterstitial waters can be attributed to that of seawaterwithout any sign of enrichment by evaporation (Craig andGordon, 1965). It is believed to antedate the generalenrichment of marine water masses in heavy isotopes due tothe balancing effect of the glaciations and the relativeisolation of the Mediterranean basins. We consider that themother water, now locked within the sediment both as porewater and crystallization water, is most likely AtlanticOcean water brought in during brief marine invasionsduring the Messinian and, in the case of the Site 122 sampleat the upper contact of the evaporite layer, by the marinewaters of the Pliocene (see Chapter 47). It is unlikely thatthe ancient mother water was solely meteoric because itscorrected values lie well off the trend (dashed line of Figure3) of rain water from arid or semi-arid areas (Craig, 1961and IAEA, 1969 and 1970).

If we consider that the mother water was suppliedduring marine invasions from the Atlantic Ocean, it will beof further interest to compare its inferred isotopic composi-tion (open circles of Figure 3) with that of the Pliocenecarbonate sediments deposited at a time when we knowthere was open communication between the Mediterraneanand the Atlantic (Benson, 1971).

ISOTOPIC COMPOSITION OF THE PELAGIC OOZES

Oxygen and carbon isotope determinations were madeon the carbonate fraction of pelagic open-marine sedimentsof Sites 125A, 126, 129, 132 and 134. The samples fromSites 126 and 129 include sediment of pre-evaporite age(middle Miocene), and the remainder are samples of earliestPliocene to Quaternary age.

Oxygen and Carbon Isotopes in the Pliocene andQuaternary Sediments

The Pliocene-Quaternary pelagic oozes are generally richin calcium carbonate (30% < CaCθ3 < 70%), as shown inFigure 1 of Chapter 30.3.

The clustering of the isotopic compositions (+0.24 <δO18 < + i.35 and-0.04 < 6 C 1 3 < + 1.25), which appearsas the group of open circles with dots in Figure 2, is relatedto the crystallization of carbonate skeletal material in alarge mass of water, which acts as an isotopic buffer.Sample 126-1-4-90, assigned to the "pre-glacial" Pleisto-cene, is slightly heavier than the Pliocene carbonates anddoes not show the important variation correlated with the"temperature and/or water-isotopic-composition effect"known for the cold climate epochs of later glacial times. A

similar observation has been made by Emiliani et al. (1961)on pericontinental Pliocene/Calabrian deposits.

The most positive values of δ θ l 8 in the pelagic oozesoccurs in Sample 125 A-6-1-25 cm, just above the Mio-cene/Pliocene contact where a significant amount of thecarbonate (l/6th) is in the form of detrital dolomitereworked from the subjacent evaporitic marls with a δ θ * 8

of +5.20 °/oo. Detrital dolomite has been observed to havea similar effect on the bulk isotopic composition ofcarbonate sediment from the Persian Gulf (Sugden, 1963).

The values of -0.78 and -0.88 ‰ vs. SMOW for theoxygen isotopic composition of the interstitial waters oflate Miocene age suggest that the ancient sea water wasperhaps 1.5 to 2.3 o / o o lighter in 0 1 8 than today. Thisassumed value of -0.8 °/oo for Mediterranean Sea waterimplies that the budget of the Mediterranean Sea was quitedifferent from that of the present. Therefore the values of+0.24 < δO18 < +1.33 °/oo U.S.P.D.B. for the earliestPliocene oozes of Cores 19 to 21 of Site 132 suggest thatthe temperature was near 8°C to 13°C at maximum —rather close to that of the recent Mediterranean bottomwater, and perhaps lower. This problem will be discussed ingreater detail in Part V of Chapter 47. However, we canstate that the situation revealed in the present study is quitedistinct from the compositional balance of the oceanicmasses as discussed by earlier authors (i.e. Emiliani, 1966;Shackleton, 1967; Dansgaard and Tauber, 1969; Letolle etal, 1971).

Oxygen and Carbon Isotopes in theMiddle Miocene Sediments

The carbonate content of the middle Miocene (Serraval-lian) sediments of Site 126 is only 3 to 12 per cent byweight but it is entirely calcitic—mostly nannofossils. Theisotopic data show a much broader scatter (+ signs in Figure2) than for the Pliocene and Quaternary sediments with aδC13 value of -2.83 °/Oo for the core catcher sample ofCore 6. Perhaps the light values are related to densitystratification during Serravallian time, leading to oxygendepletion in the bottom water mass of the Ionian Basin (nobenthic fauna was present). It is also possible that somepart of the Serravallian calcite was recrystallized duringMessinian immersions.

Similar light values have been recorded in Quaternarysediments from layers of sapropel mud rich in organicmatter and influenced by exchanges of water with theBlack Sea.

Markedly negative values of the δC13 were observed inthe samples from Core 3 of Hole 129. These samplescontain the euryhaline ostracod Cyprideis (see Chapter36.2), which is suggestive of an alkaline lake deposit.

CONCLUSIONS

The carbonates of the evaporite layer exhibit a largescatter in their δ θ l 8 and δCl3 values. The calcites are allnegative, whereas the dolomites have generally positiveδO18 values with δC13 values that reach -26.9 °/oo PDB.Some of the fine-grained dolomite is interpreted as aprimary precipitate or a very early diagenetic product. As awhole, the δ θ l 8 range for the carbonates is -6.81 < δ O l 8 <+ 7.13. We believe these measurements indicate that at

794

30.2. CARBONATE AND SULFATE ISOTOPE STUDIES

times during the Messinian "crisis of salinity" the Mediter-ranean basins became isolated from the open ocean andwere supplied to a great extent by river and rain water.

The markedly negative values of 5C13 in the dolomitesmight be related to diagenesis associated with carbondioxide derived from organic matter depleted in the heavierisotope. The positive values might correlate with theintervention of atmospheric carbon dioxide at a time whenthe basins were very shallow and abiotic.

The sulfates examined appear to be of marine origin,indicating periodic influxes of appreciable amounts ofwater from the open ocean. This conclusion is corroboratedby the presence of marine fossils (foraminifera and cocco-liths) in interbeds within the evaporite sequence.

The hydration of anhydrite to gypsum and the forma-tion of selenite at Sites 122 and 125A in both the westernand eastern Mediterranean basins was caused by integrationof marine sea water in the crystal lattice.

The late Miocene/early Pliocene water was perhapssignificantly lighter (perhaps 1.5 to 2.3 °/oo) in O l 8 thanpresent day Mediterranean water. The earliest Pliocenesurface water at Site 132 in the Tyrrhenian Basin wasprobably as cold or colder than present day Mediterraneanbottom water.

Negative δ C 1 3 values from the middle Miocene (Ser-ravallian) pre-evaporite calcite sediments are correlated withevidence that the Mediterranean basins were stagnant ornearly stagnant at that time, leading to an enrichment oforganic matter in the stratified surface water mass.

ACKNOWLEDGMENTS

We are greatly indebted to the U.S. National ScienceFoundation, sponsor of the Deep Sea Drilling Project. TheCentre National D'Exploitation des Oceans and the CentreNational de la Recherche Scientifique provided the Frenchparticipants to the shipboard staff of Leg 13.

We are grateful to Mrs. L. Merlivat who analyzed watersfor deuterium at the C.E.N. in Saclay (France) and M.F.Melières (Université de Paris) who carried out a good partof the X-ray diffraction analyses.

We owe thanks for most of the analytical part of thiswork to Miss A. Filly and Miss M. C. Sichère,, Financialsupport was provided by Laboratoire Associe n°13 duCentre National de la Recherche Scientifique.

Discussions with K. J. Hsü and J. R. Lawrence have beenhelpful.

REFERENCES

Benson, R. H., 1971. Ostracods as indicators of thresholddepth in the Mediterranean during the Pliocene. Ab-stract. 83 Int. Sedim. Congr. Heidelberg.

Cheeney, E. S. and Jensen, M. L., 1965. Stable carbonisotopic of biogenic carbonates. Geochim. Cosmochim.Acta. 29 1331.

Clayton, R., Jones B. F. and Berner R. A. 1968. Isotopestudies of dolomite formation under sedimentary condi-tions. Geochim. Cosmochim Acta. 32, 415, 432.

Craig, H., 1961. Isotopic variation in meteoric waters,Science. 133, 1702.

, 1965. The measurements of oxygen isotopepaleotemperatures. Spoleto 1965, Consiglio Nazionaledelle Ricerche. E. Tongiorgi ed., 161-182.

Craig, H. and Gordon L. I., 1965. Deuterium and oxygen18 variations in the Ocean and the Marine Atmosphere.

In Stable Isotopes in Oceanographic Studies and Paleo-temperatures. Spoleto 1965. Consiglio Nazionale delleRicerche. E. Tongiorgi ed. 9-130.

Dansgaard, W. and Tauber H., 1969. Glaciers Oxygen 18Content and Pleistocene Ocean Temperatures. Science.166,499.

Dessau, G., Gonfiantini R. and Tongiorgi E., 1960. L'ori-gine dei giacimenti solfiferi italiani alia luce delleindagini isotopiche sui carbonati della serie gessoso-solfifera della Sicilia. Lo Zolfo., 15-16, 9.

Deuser, W. G., 1970. Extreme variations in Quaternarydolomites from the continental shelf. Earth and Planet.Sci. Letters, 8, 118.

Emiliani, C, Mayeda T., and Selli R. 1961. Paleotempera-ture Analysis of the Plio-Pleistocene Section at theCostella, Calabria, Southern Italy. Bull. Geol. Soc. Am.,72, 679.

Epstein, S., Buchsbaum R., Lowenstam H. A. and Urey H.C, 1953. Revised carbonate waterisotopic temperaturescale. Bull. Geol. Soc. Am. 64, 1315.

Fontes, J-Ch., 1966. Intérët en géologie d'une etudeisotopique de 1'évaporation. Cas de l'eau de mer. C. R.Ac. Se. 263, 1950.

Fontes, J-Ch., and Gonfiantini R. 1967. Fractionnementisotopique de 1'hydrogene dans Feau de cristallisation dugypse. CR. Ac. Sci., Paris. 265, 4.

Fritz, P., 1971. Geochemical characteristics of dolomitesand the 0*8 content of middle Devonian oceans. Earthand Planet. Sci. Letters. 11, 277.

Fritz, P., and Smith D. G. W., 1970. The isotopiccomposition of secondary dolomites. Geochim. Cosmo-chim. Acta. 34, 1161.

Gonfiantini, R. and Fontes J-Ch., 1963. Oxygen isotopicfractionation in the water of crystallization of gypsum.Nature. 200, n° 4907, 624.

Hathaway, J. C. and Degens E. T., 1968. Methone-DerivedMarine Carbonates of Pleistocene Age. Science. 165,690.

I.A.E.A., 1969, 1970. Environmental Isotope Data WorldSurvey of Isotope Concentration in Precipitation n° 11953-1963; n° 2 1964-1965.1.A.E.A., Vienne, technicalreports n° 96 and 117.

Keith, M. L. and Weber J. N., 1964. Carbon and oxygenisotopic composition of selected limestones and fossils.Geochim. Cosmochim. Acta, 28, 1787.

Letolle, R., De Lumley H., Pillard F. and Vergniaud-Grazzini C. 1971. Essai d'échelle paléoclimatique de laMéditerranée occidental basée sur 1'analyse isotopiquedes tests marins fossiles. Abstr. 8e Int. Sedim. Congr.,Heidelberg. A paraitre dans Quaternaria.

Lloyd, R. M. 1968. Oxygen isotope behavior in thesulphate water system. /. Geol. Res. 73, 6099.

Longinelli, A. and Craig H., 1967. Oxygen 18 variations insulfate ions in sea water and saline lakes. Science. 156,n° 3771 56.

Mizutani, Y. and Rafter T. A., 1969. Oxygen isotopicfractionation in the bisulphate ion-water system. N. Z. J.Sci. 12, 1,54.

Mook, W. G. 1971. Paleotemperatures and chlorinites fromstable carbon and oxygen isotopes in shell carbonate.Paleogeography, Paleoclimatology, Paleoecology. 9, 245.

Nielsen, H. 1965 Schwefelisotopes im marinen Kreislaufund das S34 der früheren Meere. Geol. Rundschau. 55,160.

Nielsen, H. and Rambow D. 1969. S. Isotopen unterSuchungen an sulfaten nessischer Mineralwàsser. Notizbl.hessL. Amt. Bodenforsch. 97 352.

795

J. R. LAWRENCE

Shackleton, N., 1967. Oxygen Isotope Analyses and Pleisto-cene Temperatures reassessed, Nature. 215, 15.

Sugden, W., 1963. Some aspects of sedimentation in thePersian Gulf. Sediment Petrol. 33, 2, 355.

Sharma, T. and Clayton, R. N., 1965. Measurements of018/016 ratios of total oxygen from carbonates. Geo-chim. Chosmochim. Acta. 29, 1347.

Ruggieri, G. 1967. The Miocene and later evolution of theMediterranean Sea. Sipt. Assoc. Publ. 1, 283.

Ruggieri, G. and Greco, A., 1965. Studi geologici epaleontologici su capo Milazzo con particulare regardo alMilazziano. Geol. Romona. 1, 41.

Russell, K. L. Deffreyes, K. S., Fowler, G. A. and Lloyd, R.M., 1967. Marine Dolomite of Unusual Isotopic Com-position, Science. 155 n°3759, 189.

30.3. STABLE OXYGEN AND CARBON ISOTOPE VARIATIONS IN BULK CARBONATESFROM LATE MIOCENE TO PRESENT, IN TYRRHENIAN BASIN - SITE 132

James R. Lawrence, Lamont-Doherty Geological Observatory, Columbia University, Palisades, New York

INTRODUCTION

This investigation was undertaken to determine to whatdegree the 018/016 and C l 3 / C 1 2 ratios of full carbonatesediments represent depositional conditions versus dia-genetic conditions. If appreciable diagenesis can be dis-counted, then the measured isotopic values should reflectthe surface water temperature at the time of growth of thecarbonate organisms and the 018/016 and Cl3/Cl2 contentof the surface waters. If, on the other hand, significantisotopic exchange has occurred during diagenesis, the iso-topic values should yield information about the tempera-tures and mechanisms of diagenesis.

The sediments from Site 132 consist of evaporites ofMiocene age overlain unconformably by a continuous sec-tion of pelagic oozes of Pliocene to Recent age (see Chapter13). The Miocene evaporites consist mostly of gypsum,dolomite, calcite and silicate detritus and contain, in severalhorizons, brackish water fossils. These sediments haveundergone lithification and recrystallization to a moderatedegree since deposition. The Plio-Pleistocene oozes consistof coccoliths (70-95%) and foraminiferal (5-30%) shellsformed in surface and near surface waters, plus a variableamount of silicate detritus (see Figure 1). Very littlelithification and recrystallization has taken place in theseyounger sediments (see Chapter 13).

Bulk samples, representing one centimeter horizons at19 intervals throughout the entire core, were analyzed for018/016and C13/C12 content of the carbonate fraction bythe method described by McCrea (1950). A11C13/C12 and018/016 analyses are reported in the δ-notation withδ-values being reported with respect to PDB and SMOW,respectively.

RESULTS AND DISCUSSION

Variations in the CaC03 content of sediments from Site132 are shown in Figure 1. There is an abrupt increase in

WEIGHT % CoCO3

20 30 4 0 70 SEDIMENTATION' RATE

2 0 0 -

Figure 1. Weight per cent of CaCOß in the sediments fromSite 132, Leg 13 plotted as a function of depth. Theaverage sedimentation rates in centimeters per thousandyears are also shown.

CaC03 content of the sediments across the Miocene-Pliocene unconformity. This is not unreasonable consider-ing the drastic change from an evaporite sequence tonormal marine pelagic oozes.

The CaC03 content of the Pliocene sediments is rela-tively uniform, with a decrease in the upper Pliocene whichcontinues into the Pleistocene. Comparison of the changes

796

30.3. STABLE ISOTOPES OF BULK CARBONATE, SITE 132

in CaC03 content with the changes in total sedimentationrate suggests, on an average basis, a twofold increase in theamount of silicate detritus going from the late Pliocene tothe Pleistocene. Also, the magnitude of shorter term fluctu-ations of CaCθ3 content are much greater in the Pleisto-cene than in the Pliocene. These variations in CaC03content probably reflect the effects of continental glacialcycles in Europe and Asia during the Pleistocene.

The isotopic data and CaCC>3 contents of the analyzedsamples are illustrated in Figures 2 and 3. In the discussionwhich follows, it is assumed that there are no large changesin O l 8 / 0 1 6 or C13/C12 due to changes in the biologicmakeup of the carbonate.

The largest changes in O 1 8/Ol6 and C13/C12 are ob-served across the Miocene-Pliocene unconformity. The018/016 and C13/C12 of most of the samples above theboundary are in the range expected for CaCU3 in equilib-rium with surface waters similar in temperature and iso-topic composition to those of the Mediterranean Sea today(see Figures 2 and 3). The samples below the boundary are,in contrast, distinctly depleted in 0*8 and C13, whichstrongly suggests that the water in which the carbonateswere formed, or recrystallized, was meteoric in origin. Also,the fact that the isotopic values display a sharp break over a

5 0

100

150

2 0 0

26 23

H-+4

80 of CαCO, in equilibriumwith H,0 with 8 0 * = 1.5 %.at indicated temperatures ( β C)

8 0 " of CoCO, in equilibriumwith H,0 with S0" = 0 . 0 ‰at indicated temperatures C O

5 +26o +27 +28 +29 +30 +318Olβ ( smow)

+32 +33

-5 -4 -3 -2 -I 0

8Olβ(RD.B.)

+1 +2 +3

Figure 2. δθ18 of the bulk carbonate from Site 132, Leg13 sediments. The temperature-δO18 scales shown werecalculated using the carbonate temperature scale ofEpstein et al. (1953). One is based on δO^HjO = +1.5percent representing CaCOß in equilibrium with present-day surface waters in the Mediterranean; the other on8018Hj0 = 0.0 per mills representing CaC03 inequilibrium with Atlantic Ocean Deep Water.

50

*> 100

150

200O -3 -2.0 -1.0 00

8C,, (PDB)

1.0 0 20 40 60 80 100

Wt % COCO3

Figure 3. 8C13 of the bulk carbonate from Site 111, Leg13 sediment. The weight per cent of CaCOj in the iso-topically analyzed samples is also shown.

4 cm interval (the two samples at the Mio-Pliocene bound-ary are only 4 cm apart) suggests that no large degree ofisotopic change as a result of diagenesis has occurred sincethe deposition of the Pliocene sediments.

Three samples in the Pleistocene, two sapropel horizons1

(at 2 and 53 m) and a layer containing a high percentage ofsilicate detritus (at Im, the peak of the Wisconsin glacialstage), are depleted in 0*8 and C13 compared to the bulkof the Plio-Pleistocene samples. These large decreases areprobably due mostly to influxes of 0 1 8 and CI3 depletedglacial melt waters into the surface waters of the Mediterra-nean. Part of the decrease inO18/Olb may be due to awarming of surface waters, since the sapropel horizons canbe correlated with warming trends following glacial maxi-mums (see Chapter 46).

The remainder of the Plio-Pleistocene samples exhibit noobvious trends in C13/C12 but do show significant changesin Ol8/O16. There is a decrease in 6018 of lo/oo over aperiod of 300,000 years, starting with the sudden appear-ance of pelagic oozes in the late Pliocene. This could beattributed to a warming of the surface waters of theMediterranean of from 4° to 5°C, to a decrease in theδθ!8of the surface waters of l % o , or to a combination of thetwo factors.

Paleontological evidence suggests the sudden appearanceof pelagic oozes was the result of a sudden influx of cold

Sapropels are organic-rich sediments which may have been formedduring eutrophication of Mediterranean surface waters, perhapsassociated with an influx of water from the Black Sea.

797

J. VAN DONK, T. SAITO, N. J. SHACKLETON

Atlantic Deep Water. The initialO18/016 of the Mediterra-nean surface waters would be expected to be free ofinfluence from the Atlantic source. Subsequent excessevaporation in the Mediterranean would increase t h e δ O 1 8

of the surface waters. The observed progressive decrease intheδOlδ of the carbonate from +32°/Oo to +31%o couldonly be brought about if temperature effects overrodethose of the isolation. Consequently the 4° to 5°C inferredwarming is a minimum value.

From the Late Pliocene to the Pleistocene, there is a 1.5to 2°/oo increase in the δ O 1 8 of CaCθ3. If no large isotopicchanges had taken place as a result of diagenesis, this couldbe attributed to: (a) an increase in the δ O 1 8 of the surfacewaters of 1.5 to 2°/oo as a result of progressive isolation ofthe Mediterranean from the Atlantic (see Chapter 47.4),(b) to a cooling of the surface waters by 6° to 9°C, or (c) toa combination of both factors. It is probable that bothcooling and an increase in t h e δ O 1 8 of the surface watersoccurred with an increase in glaciation. An increase inglaciation is consistent with dropping temperatures. Butalso, increased glacial uptake of water would drop sea leveland decrease circulation between the Atlantic and theMediterranean. Increased isolation of the Mediterraneanwould have a greater effect on increasing the O 1 8 / O 1 6 ofthe Mediterranean surface waters.

The variations in 018/016 and C13/C12 of the bulkcarbonate at Site 132 are consistent with changes in surfacewater temperatures or with 018/016 and C13/C12 varia-tions in the surface waters as inferred from physical orpaleontological changes in the sediment. In addition, sharpchanges in 018/016 and C13/C12 occur over short timeintervals in the sediments. Isotopic exchange or recrystalli-zation of the bulk carbonate would tend to smooth suchsharp isotopic gradients. These facts suggest that the iso-topic values of most of the carbonate, with the exceptionof the Miocene sediments, is largely unaffected by dia-genesis. Isotopic studies on individual carbonate fossilsshould verify this and determine to what degree changes inbiologic makeup of the carbonate portion of the sediments(ignored in this discussion) have determined the 018/016and C13/C12 of the bulk carbonate.

Lamont-Doherty Geological Observatory ContributionNo. 1860.

REFERENCES

Epstein, S., Buchbaum, R., Lowenstam, H. A. and Urey, H.C, 1953. Revised carbonate-water isotopic temperaturescale. Bull Geol. Soc. Am 64, 1315.

McCrea, J. M., 1950. The isotopic chemistry of carbonatesand a paleotemperature scale. /. Chern. Phys. 18, 849.

30.4. OXYGEN ISOTOPIC COMPOSITION OF BENTHONIC AND PLANKTONIC FORAMINIFERAOF EARLIEST PLIOCENE AGE AT SITE 132 - TYRRHENIAN BASIN

J. Van Donk and T. Saito, Lamont-Doherty Geological Observatory of Columbia University, Palisades, New Yorkand

N. J. Shackleton, Sub-Department of Quaternary Research, University of Cambridge, Cambridge, England

INTRODUCTION

Although it has been shown that some foraminifera donot deposit calcium carbonate in isotopic equilibrium withthe surrounding water, the deviations noted are relativelysmall (up to-1.0°/oo) For this reason it was consideredworthwhile to analyse specimens of both benthonic andplanktonic foraminifera from a single sample of pelagicooze of earliest Pliocene age from Site 132 (40° 15.7'N;11° 26.5'E, depth 2835 m). It was hoped to obtain infor-mation about the structure of the Mediterranean watermass following re-introduction of marine conditions at theend of the late Miocene "crisis of salinity".

A quarter core section from 52 to 75 cm in Section 4 ofCore 20 was sampled and washed. The sample interval isfrom a level only a few meters above the Miocene/Plioceneboundary (see Chapter 47).

RESULTS

Benthonic Foraminifera

The sample of benthonic foraminifera had a weight ofabout micrograms. It was roasted in vacuo before analysisat the University of Cambridge. Carbon dioxide was re-leased by the action of 100 per cent orthophosphoric acidin vacuo at 50°C, and it was analysed in a double-collectingmass spectrometer. Results were calibrated by the analysisof standard carbonates in identical conditions. The result(relative to the PDB standard) was:

δO 1 8=(+0.3±0.1)°/oo

δ C 1 3 =(+0.7±0.1)7oo

The analytical method is routine in this laboratory; thequoted uncertainty represents the standard deviation

798

30.4. ISOTOPIC COMPOSITION OF FORAMINIFERA, SITE 132

among replicate analyses of small samples of calciumcarbonate under the same conditions (1-σ).

Planktonic Foraminifera

Two species of foraminifera, Globigerina bulloides andSphaeroidinellopsis seminulina were isolated from thewashed residues and analysed separately. The specimenswere crushed and then shell fragments were washed ultra-sonically to remove any chamber fillings present. They werethen vacuum dried at room temperature. The carbonatematerial was reacted with 100 per cent orthophosphoricacid in vacuo at 25°C without having been roasted. Theanalysis was carried out in the Geochemistry Laboratory ofthe Lamont-Doherty Geological Observatory in a double-collecting mass spectrometer, and, as in the case of thebenthonic foraminifera, the results are reported with res-pect to the PDB standard.

G. bulloides

S. seminulina

δO 1 8=(+1.0±0.15) o/oo

δ C 1 3 =(+0.3±0.15)%o

δO 1 8 =(+0.4±0.03)7oo

δ C 1 3 =(+1.39±0.09)7oo

Measurements on the specimens of Sphaeroidinellopsisseminulina are considered to be the most reliable. Resultsshown above are an average of three separate analyses. Thesample comes from the Sphaeroidinellopsis Acme-zone (seeChapter 47.1) where this taxon sometimes comprises up to70 per cent of the foraminifera population. The substantialnumber of tests of Globigerina bulloides seems to suggestcold temperatures of the surface waters since, according toBe and Tolderlund (1971), this is a dominant species insubarctic and subantarctic waters with a peak abundance inwater masses having a temperature range of from 3° to19°C.

DISCUSSION

The most obvious, and probably the most significant,observation is that the oxygen isotopic composition of thebenthonic foraminifera is extremely close to that of theplanktonic foraminifera. The higher value of +1.0°/oo fromGlobigerina bulloides perhaps indicates that this taxon livedat shallower depths in the surface water layer than Sphaero-idinellopsis seminulina (personal communication, M. B.Cita). At these depths the O1 8 content is slightly enrichedas the result of evaporation and higher salinity.1

It is somewhat hazardous to translate the δ θ 1 3 valuesdirectly to paleotemperatures without knowledge of theisotopic composition of the Mediterranean water mass ofearly Pliocene age. The value of+0.3 would correspond to atemperature of about 15°C if the fossil tests were inisotopic equilibrium with present day North Atlantic DeepWater.2 The true value would be lower by a few degrees, atmost, if the species analysed did not deposit carbonate in

The oxygen isotope composition of the surface layer of theCaribbean is measurably heavier than that of the bottom water forthe same reason (Craig and Gordon, 1965).

2According to Craig and Gordon (1965) North Atlantic Deep Waterhas a present composition of+0.1 o/ o o (SMOW).

isotopic equilibrium. It would also be lower if account weretaken of a change in the isotopic composition of the oceanssuch as the one which would result from a reduction in theamount of ice on Antarctica. This last correction mayperhaps be ignored for Pliocene material.

At present the surface and deep waters in the Mediter-ranean have an isotopic composition about 1.2°/oo heavierthan North Atlantic Deep Water in consequence of theexcess evaporation (Epstein and Mayeda, 1953; correctedby Craig and Gordon, 1965). Benthonic foraminifera livingat 13°C today would be expected to have an isotopiccomposition of about +2.0°/Oo The Pliocene sampleanalysed would represent a temperature of around 21°C inpresent-day Mediterranean water, an improbably high value.However, a lower temperature than this would be derived ifthe salinity excess were much less than today as theconsequence of a more open exchange with the AtlanticOcean.

If we allow ourselves to examine the working hypothesisof a more open exchange in early Pliocene times betweenthe Mediterranean and the Atlantic (see Chapter 47.4), wealso have to ask how this exchange was initiated in the firstplace. After all, during the late Miocene evaporite deposi-tion, the passageway, to all intents and purposes, wasblocked.

From stratigraphic considerations (see Chapters 47.1 and47.2), the Mediterranean sample of pelagic ooze investi-gated here was deposited within approximately 100,000years after the first truly effective opening of the Gibraltarportal, this opening established permanent open marineconditions in the Mediterranean. The isotopic value ofδ O 1 8 = 0.3 for the benthonic foraminifera could imply thatthe Mediterranean then communicated with the Atlanticover a sill at a depth corresponding to a temperature ofabout 15°C or a little less, and that bottom water in theMediterranean entered over this sill. This implies a wideenough sill to permit free exchange in both directions.

The similarity of the isotopic composition of the plank-tonic and benthonic foraminifera indicates a thermal homo-geneity in the Mediterranean water structure. A speculativemodel which would account for both the similarity in theoxygen isotopic composition of the planktonic and ben-thonic foraminifera and its composition at 0.3 to 0.4°/oo,invokes a catastrophic flooding of the Mediterranean basinsfollowing the late Miocene evaporite epoch. In this model,water entering the Mediterranean would be comprised ofboth cold North Atlantic Deep Water (down to the level ofthe sill eroded away) and warmer surface water, well mixedas it poured into the partially desiccated basin (see Chapter43). The oozes in Section 4 of Core 20 at Site 132 couldhave been deposited either while the basin was still floodingor before it established a steady-state thermocline.

The depth of the sill at Gibraltar necessary to produce awater column in the Mediterranean with a mean tempera-ture of around 15°C is extremely difficult to estimate.There is evidence, though, that bottom water temperaturesin the open oceans were warmer in early Pliocene timesthan they are today.

From similar measurement of benthonic foraminifera incores from the Pacific Ocean, Emiliani (1954) reported acooling trend in the Pacific Bottom Waters from 10.4°C in

799

J. VAN DONK, T. SAITO, N. J. SHACKLETON

the middle Oligocene to 7.0°C in early-middle Miocene to2.2°C in the late Pliocene. We have been able to sub-stantiate that this cooling trend also was present in theAtlantic Ocean. A study by two of us (T.S. and J.V.D.) ofthe isotopic composition of benthonic foraminifera of lateMiocene and early Pliocene age in DSDP cores from theSouth Atlantic (Site 15 at 31°S) reveals that the Atlanticwaters were as warm as 5°C at the time the pelagic oozeswere deposited in Core 20 of Site 132.

The present day mean surface water temperature of theAtlantic west of Gibraltar is 22°C in August. It reaches24°C in the Mediterranean near Site 132. Since flooding,therefore, would have had to involve a considerable amountof North Atlantic deep water to bring the average tempera-ture down to 15°C, we can roughly infer that the incisionof Gibraltar might have been as deep as 1000 meters inorder to effect the required exchange.

We can conclude that, although the measurement ofbenthonic foraminifera alone does not exclude the possi-bility of bottom water formation locally within the Plio-cene Mediterranean, the measurements on both the ben-thonic and planktonic foraminifera can be more easilyunderstood in terms of entry of water into the Mediterra-nean from the Atlantic over a sill at Gibraltar of appreciabledepth and width.

ACKNOWLEDGMENTSThe study was initiated at the suggestion of William B.

F. Ryan. Discussions with K. J. Hsu, who carried thesamples from Lamont to Europe, have been very helpful.The research at Lamont-Doherty Geological Observatoryhas been supported by National Science Foundation grantGA 28507. Lamont-Doherty Geological Observatory Con-tribution Number 1809.

REFERENCES

Be, A. W. H. and Tolderlund, D. S., 1971. Distribution andecology of living planktonic foraminifera in surfacewaters of the Atlantic and Indian Oceans. In Micro-paleontology of Oceans; Funnell, B. M. and Riedel, W.R., (Editors), Cambridge University Press, London, 105.

Craig, H. and Gordon, L. I., 1965. Isotopic Oceanography:Deuterium and Oxygen 18 variations in the ocean andthe marine atmosphere. Proc, of the Symposium onMarine Geochemistry; The University of Rhode Island,Occasional Publication No. 3, 277.

Emiliani, C, 1954. Temperatures of Pacific Bottom Watersand Polar Superficial Waters during the Tertiary.Science. 119, (3013), 853.

Epstein, S. and Mayeda, T., 1953. Variation of O1 8

Content of Waters from natural sources. Geochim.Cosmochim. Acta, 4, 213.

800