38. red sea geochemistry - deep sea drilling project

24
38. RED SEA GEOCHEMISTRY Frank T. Manheim 1 , U. S. Geological Survey, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts ABSTRACT The Red Sea drillings reveal a number of new facets of the hot-brine-metalliferous system and other geochemical aspects of the sea, its sediments, and its past history as follows: 1) Dark shales rich in organic material, and containing enhanced Mo and V concentrations, are characteristic of Plio-Pleistocene strata in the Red Sea. Values as high as 1500 ppm V and 500 ppm Mo were obtained in sediments containing up to 8 percent organic carbon. 2) Metalliferous sediments in the hot brine deep (Site 226) are similar in composition in both solids and interstitial water to previously analyzed sediments. However, one site (228) well south of the known hot-brine deeps shows zinc mineralization reaching 5 percent Zn in late Miocene shale-anhydrite breccias. 3) Pore fluid studies show that near-saturated (NaCl) brines having similar total salt concentration to the hot-brine fluids are associated with Miocene evaporites at Sites 225, 227, and 228. However, their chemical and isotopic composition precludes such fluids being part of the "hot brine plumbing system." Hydrogen and oxygen isotope studies demonstrate that fluids trapped between and among the evaporitic rocks have a strong meteoric water component, presumed to have entered the rocks during or shortly after formation in shallow evaporating pans. The composition of pore fluid at Site 227 suggests the presence of late-stage evaporite minerals of the tachyhydrite CaMg2Clg I2H2O series in the in situ rocks. 4) Diffusivity measurements show that the pre-Miocene strata permit dissolved salt or gas diffusion to the extent of from 1/2 to about 1/10 the rate in free solution. However, in anhydrites diffusivity is reduced more than 100-fold, and no diffusion could be detected through halite rock. The rates applied to interstitial salt gradients at Site 225 suggest that less than 1 meter of rock salt is removed per million years by diffusion processes. The diffusion of salt can already be detected a few meters below the sediment-water interface, and based on the interstitial water studies, one can affirm the presence of salt at depth at Sites 228, 230, and possibly 229, where rock salt was not encountered by the drill. 5) Isotopic measurements on leads show that both leads from Site 228 and the hot brine deep (Site 226) require input from igneous or volcanic sources (e.g., volcanic ash). Elsewhere, however, leads of sedimentary-pelagic origin are noted. 6) Isotopic and other evidence indicates that the long-distance transport of subterranean brines advocated by Craig (1969) is unlikely. Instead, it is proposed that the source of the hot brines is subevaporite clastic or other aquifers of early to middle Miocene age that have been disrupted by rifting. These discharge in the deeps by virtue of hydrodynamic continuity with heavy brines at higher positions on the nearby flanks of the Red Sea. In this case, the waters might be fossil (middle Miocene) Red Sea waters of relatively normal salinity that have acquired greater salt concentration by diffusion from overlying late Miocene Contribution No. 3213 of the Woods Hole Oceanographic Institution. Publication authorized by Director, U. S. Geological Survey. 975

Upload: others

Post on 01-Jan-2022

9 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

38. RED SEA GEOCHEMISTRY

Frank T. Manheim1, U. S. Geological Survey, Woods Hole OceanographicInstitution, Woods Hole, Massachusetts

ABSTRACT

The Red Sea drillings reveal a number of new facets of thehot-brine-metalliferous system and other geochemical aspects ofthe sea, its sediments, and its past history as follows:

1) Dark shales rich in organic material, and containingenhanced Mo and V concentrations, are characteristic ofPlio-Pleistocene strata in the Red Sea. Values as high as 1500 ppmV and 500 ppm Mo were obtained in sediments containing up to8 percent organic carbon.

2) Metalliferous sediments in the hot brine deep (Site 226) aresimilar in composition in both solids and interstitial water topreviously analyzed sediments. However, one site (228) wellsouth of the known hot-brine deeps shows zinc mineralizationreaching 5 percent Zn in late Miocene shale-anhydrite breccias.

3) Pore fluid studies show that near-saturated (NaCl) brineshaving similar total salt concentration to the hot-brine fluids areassociated with Miocene evaporites at Sites 225, 227, and 228.However, their chemical and isotopic composition precludes suchfluids being part of the "hot brine plumbing system." Hydrogenand oxygen isotope studies demonstrate that fluids trappedbetween and among the evaporitic rocks have a strong meteoricwater component, presumed to have entered the rocks during orshortly after formation in shallow evaporating pans. Thecomposition of pore fluid at Site 227 suggests the presence oflate-stage evaporite minerals of the tachyhydrite CaMg2ClgI2H2O series in the in situ rocks.

4) Diffusivity measurements show that the pre-Miocene stratapermit dissolved salt or gas diffusion to the extent of from 1/2 toabout 1/10 the rate in free solution. However, in anhydritesdiffusivity is reduced more than 100-fold, and no diffusion couldbe detected through halite rock. The rates applied to interstitialsalt gradients at Site 225 suggest that less than 1 meter of rocksalt is removed per million years by diffusion processes. Thediffusion of salt can already be detected a few meters below thesediment-water interface, and based on the interstitial waterstudies, one can affirm the presence of salt at depth at Sites 228,230, and possibly 229, where rock salt was not encountered bythe drill.

5) Isotopic measurements on leads show that both leads fromSite 228 and the hot brine deep (Site 226) require input fromigneous or volcanic sources (e.g., volcanic ash). Elsewhere,however, leads of sedimentary-pelagic origin are noted.

6) Isotopic and other evidence indicates that the long-distancetransport of subterranean brines advocated by Craig (1969) isunlikely. Instead, it is proposed that the source of the hot brinesis subevaporite clastic or other aquifers of early to middleMiocene age that have been disrupted by rifting. These dischargein the deeps by virtue of hydrodynamic continuity with heavybrines at higher positions on the nearby flanks of the Red Sea. Inthis case, the waters might be fossil (middle Miocene) Red Seawaters of relatively normal salinity that have acquired greater saltconcentration by diffusion from overlying late Miocene

Contribution No. 3213 of the Woods Hole OceanographicInstitution. Publication authorized by Director, U. S. GeologicalSurvey.

975

Page 2: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

F. MANHEIM

evaporites. The model is consistent with the isolated nature of thebrine deeps and suggests that flow might have been enhanced byincreased hydraulic gradients during periods of lowered Red Sealevels.

7) Interstitial water evidence indicates that Pleistocenelowerings of sea level did not cause evaporative conditions leadingto actual gypsum or other evaporite deposition in the deeperwater zones, as has been postulated. This in turn suggests thatsill depths were greater than have been assumed.

INTRODUCTION

In order to get a broad coverage of pertinent data on theRed Sea hot-brine-metalliferous systems and associatedgeochemical phenomena, a two-fold strategy wasemployed: an emission spectrograph was placed onboard toperform immediate shipboard analyses of more than 23elements. These data served both to orient geochemists andsedimentologists in the composition of various types ofphases as they were encountered and to guide selection ofsamples and types of further analyses to be performedon-shore in supplemental studies. The rapidity of the visual,semiquantitative spectrochemical technique, employed byD. Siems of the U. S. Geological Survey, permitted assay ofnearly every lithologic type of sediment or rock, as well assmall particles, minerals, and organisms picked out underthe microscope by paleontologists and petrologists. It alsopermitted early detection of anomalous boron andmagnesium concentrations in interstitial brines, by analysisof the dried brine salt. In addition to spectrography, addedemphasis was placed on the interstitial water and watersampling programs, and regular measurements of electricalresistivity of minimally disturbed cores were performed inorder to obtain an estimate of their diffusive permeabilityto dissolved species according to previously describedmethods (Manheim and Waterman, in press).

In addition to the geochemical interests of the shipboardparty, especially P. Staffers and R. Coleman, whoperformed special studies of evaporitic rocks and basalts,respectively, samples were distributed to shore laboratoriesfor the following surveys:

Chemical composition of sediments (major inorganicconstituents): F . T. Manheim, U. S. Geological Survey,Woods Hole; U.S.G.S. laboratories, Washington, D.C.;Exploration Geochem. laboratories, Denver)

Lead isotope studies on sediments and brines: B. Doe,U.S.G.S., Denver

Sulfur isotope studies: I. R. Kaplan, U.C.L.A.Uranium isotope studies: T. L. Ku, University Southern

CaliforniaCarbon and oxygen isotope studies on sediments and

interstitial waters: J. R. Lawrence, Lamont-DohertyGeological Observatory

Carbon and oxygen isotopes on small samples ofseparated foraminifera: W. A. Deuser, Woods HoleOceanographic Institution

Deuterium content of interstitial waters: I. Friedman,U.S.G.S., Denver

Interstitial water studies on minor and traceconstituents: B. J. Presley, Texas A & M University

Interstitial water studies, major constituents: F . T.Manheim, F. L. Sayles, and L. S. Waterman, Woods Hole

Light hydrocarbons in canned samples: R. D. Mclver,Esso Products Research Corp.

Carbon isotopes in gases, and general gas composition:I. R. Kaplan, U.C.L.A.

Studies of rare earths in evaporites: J. L. Bischoff,University of Southern California

Potassium-argon "age" determinations on sediments:Geochron Laboratories, Cambridge, Mass.

The available results of these workers and theircolleagues are included in this volume.

During the preparation of this Initial Reports volume, asummary of the main Valdivia discoveries of new brinepools and hydrothermal sediment sites appeared (Backerand Schoell, 1972); also, thin sections from basaltic rocksfrom drill holes off the Sudanese and Ethiopian coasts weremade available by C. Burk of Mobil Oil Co.

Special thanks are due to the Analytical Laboratories ofthe United States Geological Survey, Washington, D. C , theLaboratory of the Geochemical Exploration Branch,Denver, for their extensive analytical assistance, and B. F.Jones (Arlington, Va.) and Omer Raup (Denver) for theirhelp and discussion. Numerous discussions and communica-tions with colleagues, both on shipboard and in associatedshore laboratories, are acknowledged. Finally, I should liketo thank Rebecca Belastock for assistance in calculationsand D. A. Ross for providing manuscript materials.

COMPOSITION OF THE SEDIMENTS

Where sufficiently numerous and representative, theshipboard semiquantitative analyses (Manheim and Siems,this volume, Tables 1 to 5) may be averaged to make meanswith enhanced quantitative value: a rule of thumb is thatthe probable error of the mean is Ey/N~, where E is themean error of the individual analyses and N is the numberof analyses. This presumes absence of significant systematicerror; i.e., random scatter about the true values. Theshipboard spectrochemical data, together with supple-mentary data on Zn, As, and a few other constituents, havebeen utilized to prepare means characterizing the sitesstratigraphically and lithologically. Summary tables are alsoprovided to include the quantitative (rapid rock) analysesand other shore laboratory data (Manheim and Siems, thisvolume, Table 6).

Black Shales

One of the discoveries of the shipboard analyses was thatpre-Miocene black shales rich in organic material containunusual concentrations of vanadium (V) and molybdenum(Mo) similar to other V- and Mo-rich organic shales famousin the geologic record. Such shales and their elementalconcentrations include the following:

976

Page 3: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

RED SEA GEOCHEMISTRY

Mean cone. (max. and min.) (ppm)

Name andLocation

Kupferschiefer,N. Germany

Swedish "Alumshale," C

Pierre shale (U.S.western states)

Red SeaPleistocene shales

U. Cretaceousshales, Kansas

Average shale

Org.

6.5

17.1

-

3.3

3.8

0.88

V

1700(5300-200)

1500(3600-100)

1200(2000-200)

600(1500-70)

350(500-200)

130

Mo

300(1500-14)

89(300-25)

400(700-150)

120(500-10)

24(50-10)

2.6

Reference

Wedepohl(1964)

Landergrenand Manheim(1962)

Vine (1966)

This volume

Vine (1966)

Vine (1966)Clarke (1924:cited in Vine,1966)

Figure 1 shows the distribution of V and Mo in blackshale bands throughout Sites 225, 227, 228, and 229. Onemay see immediately the thinness and multiple distributionof these bands, which extend from the early-middlePliocene to the late Pleistocene. Within Miocene evaporiticstrata, nearly all shale and claystones are dark and containsignificant organic matter; however, these shales only rarelycontain concentrations of V and Mo comparable to those ofyounger sediments, even when the latter are stronglydiluted with calcium carbonate. The concentration anddistribution of the Mo-V-rich sediments apparently doesnot indicate stratigraphic continuity or any significantlateral zonation. Some bands are only a few centimeterswide. If the rates of deposition of the dark bands arecomparable to those of the post-Miocene sediments ingeneral, the zones rich in organic material represent periodsof a few hundred years duration. In areas of less rapidaccumulation than Site 229 within the central rift system,bands rich in organic material have been noted in sedimentsin piston cores (Manheim, in press); that is, in still youngerdeposits.

The mean composition of all dark, organic sedimentsanalyzed from Sites 225, 227, and 228 is summarized inTables 1 and 2 and in Figure 2. One may note that inpre-Miocene strata changes in composition are relativelyinsignificant, except that vanadium and possibly nickel aremost abundant in Pleistocene and Late Pliocene strata. Thisincludes such transition elements as Ti, Mn, Cr, and Co,suggesting that the type and relative quantity ofnoncarbonate detritus supplied to the Red Sea basin duringthis time was relatively constant. Their rather highconcentrations relative to aluminum suggest thatbasaltic-gabbroic (high-mafic) detritus played a significantrole among sediments introduced to Red Sea waters. Note,however, that Si and Al means may be biased to the lowside by the upper detection limit of the spectrographicanalyses.

Late Miocene (evaporitically associated) sediments showmarked differences in composition. There is a sharp drop inCa, with corresponding increase in Si, Ti, and Mg. Boronincreases to levels characteristic of salt-pan clays (Harder,1970; Valeyev et al., 1972) although boron is not

necessarily enriched significantly in the associated porefluids. This confirms the expectation that the sedimentswere laid down in late-stage evaporative brine bodies,regardless of the nature of fluids that may havesubsequently replaced original waters of deposition.

In comparing the mean compositions of the black shales(Table 2) to world shale averages (Table 3), we note that inpre-Miocene strata Mo is enriched from 20- to 60-fold overthe average shale, whereas V, Mn, Co, and Cu, Zn, Ni, andFe are significantly enriched in approximately decreasingorder. Ca and Sr are also enriched, but these elements areattributable to the substantial carbonate component of theshales.

B and Zr are depleted in the shales with respect toaverage world shales.

In the late Miocene sediments V is no longersignificantly enriched, Mo remains enriched, though muchless than higher in the section, whereas B reverses its roleand becomes enriched four-fold, notwithstanding thedilution of the boron-containing silicate component bycalcium carbonate. Magnesium also increases substantiallyin the section.

The substantial increase in Si/Al ratio noted in theevaporitic-associated shales may be related to theappreciable presence of cristobalite noted by Stoffers andRoss (this volume).

Finally, one should note that unlike V and Mo inpre-Miocene black shales, Zn, and to a much lesser extent,Cu, are locally and sporadically enriched. Zn appears to beprimarily enriched in the form of sphalerite and was sofound in bands and impregnations in the shatteredanhydrite-shale breccias in the lowermost parts of Site 228(Core 39). Its enrichment depends on a combination ofphysicochemical factors including fluid migration, heatgradients, and the presence of H2S, that cause localpost-depositional precipitation. In contrast, V and Moappear to be closely associated with uptake on organicmatter or organic-matter-linked sulfides during or shortlyafter deposition of the sediment and are relatively immobilethereafter. Numerous authors (Hodgson, 1965, andreferences cited) have pointed out that porphyrins derivedfrom degradation of plant matter (whether of terrestrial orphytoplankton origin) have a high affinity for vanadiumand hold it tenaciously even under oxidizing conditions.

Mo is deposited by two major mechanisms, a reducingmode and an oxidic uptake with iron and manganese oxidesas in deep ocean ferromanganese concretions. Since we arehere speaking of organic and often pyrite-rich sediments,the former mode is pertinent. Ordinary shales have onlyvery small detrital concentrations of Mo (about 2 ppm). Mois strongly taken up (coprecipitated) by fine-grained ironsulfides (Korolev, 1958). Since Mo in seawater is presentonly in quantities of about 10µg/l, and presumably far lessin oxygen-free pore fluids, molybdate being the stablemobile species, post-depositional diagenetic enrichmentduring burial is limited. Significant molybdenumenrichments are thus generally associated with stagnantwater bodies, in which sulfide-containing sediments at thesediment-water interface can extract Mo from relativelylarge volumes of bottom water. For example, pre-Mioceneblack shales show a mean of 77 ppm Mo, or approximately

977

Page 4: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

D-J00 SITE 225 SITE 227 SITE 228 SITE 229

50-

75-

100-

150-

175-

200-

225-

250-

275-

300-

325-

o o o o oo o o o o

Mo

SITE 225

-all concentrations in ppm-values for Mo (reversed scale) in parentheses

SITE 228

SITE 229

SITE 227

Figure 1. Concentration of V and Mo in black shale bands in Sites 225, 227, 228, and 229. Concentrations in ppm. Data from Manheim and Siems (this volume).

Page 5: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

TABLE 1Spectrochemical Analyses of (Calcareous) Black Shales by Site and Stratigraphic Horizon

Element

Si (%)Al (%)Fe (%)Mg (%)Ca (%)Ti (%)Mn(%)Ag (ppm)As (ppm)B (ppm)Ba (ppm)Cd (ppm)Co (ppm)Cr (ppm)Cu (ppm)Mo (ppm)Ni (ppm)Pb (ppm)Sn (ppm)Sr (ppm)V (ppm)Zn (ppm)Zr (ppm)

Site 225

L.Pleistocene

0-52 m

10.75.03.71.8

16.70.270.150.23

1520

533-

30907393

11610

<31170

500104

26

Na

33333333133—33333333323

E.Pliocene

52-115 m

12.24.84.01.8

20.00.280.14

<O.l1022

800—

167255

17097

8<3

1000725

7721

N

44444444244_44444444444

E.Pliocene

115-173 m

9.55.06.01;5

20.00.250.2

<O.l—

10600

—2070856085

8<3

8502005025

N

22222222_22_22222222212

L.Miocene

173-227 m

22.55.05.06.23.20.40.24

<O.l<5

367125

_6692

1754

976

< 3100

65205

85

TV

44444444144_44444444444

Site 227

L.Pliocene

8-65 m

15555

200.50.5

<O.l5

20200_200100700

30300

15< 3

1000500

<40050

TV

11111111111_11111111111

E.Pliocene

65-225 m

14.04.85.92.1

14.50.310.160.216

20164

—27

1019993

11710.0

>3858289

<IOO32

TV

1212121212121212

31

12_12121212

12121212

6

L.Miocene

225-239 m

243.65.13.32.60.290.12

< l<5

40084

_2068

1648

66<5<3

220200188

33

N

55555555255_55555555545

Site 228

E.Pleistocene

51-155 m

12.6551.5

150.20.5

< l<530

700_507070

150300

20< 3

1000700

9250

N

11111111111_11111111121

L.Pliocene

155-273 m

14.64.76.41.6

10.70.390.45

< l<531

686_

34129104

99124

16<3

1000514186

60

TV

77777777177——7777717757

E.Pliocene

273-291 m

25581.81.20.70.02

<O.l<557

500—

43183117<2

1508

<3150183134233

TV

33333333223—33333333343

L.Miocene

291-325 m

19.84.38.44.16.30.50.9

<O.l<5

390329<25

18117146

248657<3

1320197572

98

N

99999999299299999999969

a7V = number of analyses.

W

σCΛW>Omoomg

Page 6: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

F. MANHEIM

TABLE 2Mean Values, Sites 225, 227, 228

Element

Si (%)Al (%)Fe (%)Ti (%)Ca (%)Mg (%)Mn (%)Pb (ppm)Ag (ppm)Cu (ppm)Zn (ppm)Sn (ppm)Cd (ppm)Mo (ppm)Ni (ppm)Co (ppm)V (ppm)Ba (ppm)Sr (ppm)B (ppm)As (ppm)Zr (ppm)Cr (ppm)

Pleistocene

11.65.04.40.24

15.91.70.32

150.16

7198<3-

121208

40600616

108325

<IO3880

Na

444444444444

4444444244

L. Pliocene

13.94.85.10.39

16.92.80.36

13<O.l

286221<3-99

17383

579562

100024

644

100

N

12121212121212121212126

121212121212124

1212

E. Pliocene

16.24.96.60.42

11.91.80.138.60.14

10094<3

-52

11730

224421619

29<697

118

N

171717171717171717171117

171717171717165

1717

L. Miocene

22.14.36.20.404.04.50.42

23<O.l

162321<3

-128335

120179545385<57292

N

1818181818181818181814

181818181818185

1818

= Number of analyses.

TABLE 3Ratio of Black Shale Components to World Shale Average (Vinogradov, 1967)

Element

Si (%)Al (%)Fe (%)Ti (%)Ca (%)Mg (%)Mn(%)Pb (ppm)Ag (ppm)Cu (ppm)Zn (ppm)Mo (ppm)Ni (ppm)Co (ppm)V (ppm)Ba (ppm)Sr (ppm)B (ppm)As (ppm)Zr (ppm)Cr (ppm)

World ShaleAverage

(%)

24.09.523.320.432.501.390.070

20-

5780

29520

130800450100

-200100

Pleistocene

0.480.531.30.566.41.24.60.7-1.31.2

612.22.04.60.82.40.25-0.190.8

L. Pliocene

0.580.501.50.916.82.05.10.7-5.0

(2.8)50

1.84.24.40.72.20.24-0.211.0

E. Pliocene

0.680.512.00.984.81.31.90.4-1.8

(1.2)(26)

1.21.51.70.51.40.29-0.481.2

L. Miocene

0.920.451.80.931.603.26.01.1-2.84.0

(6.1)0.91.70.90.21.23.9-0.360.9

75 ppm Mo corrected for possible detrital contributions. Ata concentration of 10 µg/1 Mo, each gram of dry sedimentcontains the equivalent of all the Mo in 7.5 liters of water.Considering that probably efficiency of uptake frombottom waters is far lower than 100 percent, many tens orperhaps hundreds of liters of water must sweep past bottomsediments to provide the necessary Mo. One can conclude

that the dark bands of sediment represent times ofextensive surface productivity and bottom waterstagnation, resulting in oxygen-free bottom sediments, andthat the reducing zone includes the sediment-waterinterface. Incorporation of Mo directly in organic matter asa metal-organic complex would require similar environ-mental conditions.

980

Page 7: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

RED SEA GEOCHEMISTRY

7 0 0 -

6 0 0 -

5 0 0 -

400 -

300 -

200 -

100 —

0 -

8 0 0 -

7 0 0 -

6 0 0 -

5 0 0 -

4 0 0 -

3 0 0 -

2 0 0 -

100 —

0 -

700 -

600 -

5 0 0 -

4 0 0 -

3 0 0 -

2 0 0 -

100 —

0 -

TRACE ELEMENT COMP. (ppm)

SITE 225

smáLCu Zn Ni Mo V

SITE 227

"h Jl.Cu Zn Ni Mo

rffflrft

SITE 228

Cu Zn Ni Mo V B1 2 3 4 1 2 3 4 1 2 3*4 1 2 3 4 1 2 3 4 1 2 3 4

Figure 2. Concentration of Cu, Zn, Ni, Mo, V, and B inblack shales from Sites 225, 227, and 228, Red Sea.Asterisk indicates value determined by single, realtivelyhigh-metal content sample. Numbers denote: (1) Pleisto-cene; (2) late Pliocene; (3) early Pliocene; (4) lateMiocene (evaporite series).

According to Pratsch (1973, oral communication), blackshales are also widespread in the Pleistocene of drill holeson the Ethiopian continental margin of the Red Sea. Wemay presume that these are analogous to those in our drillcores. Therefore, the black shales appear to becharacteristics of the Red Sea Neogene in general.

Non-Black Shales and Carbonates

Mean composition of non-black shale sediments is givenby site and stratigraphic interval in Table 4, utilizing thespectrographic analyses. Results of both rapid rock("chemical") and spectrographic analyses are summarizedfor black shale and non-black shale samples in Table 5.There may be some overestimation of Fe by thespectrochemical method, this being primarily designed fortrace constituents, and less free from systematic effects ofmatrix for major constituents.

Enrichments in the black shales of V, Mo, and transitionmetals are marked in pre-Miocene strata. In theevaporite-associated shales, copper and zinc are enriched.However, the mean values for Cu and Zn are far less

impressive than local enrichments, reaching as much as 5percent in sphalerite-rich zones at Site 228.

The general composition of non-black shale sediments isconsistent with an admixture of basaltic-gabbroic and someless mafic (granitic) detritus, as described by Coleman (thisvolume) with biogenic carbonate material. The maficinfluence is suggested by relatively high concentrations ofNi, Co, Cr, Ti, and Mg, but low concentrations of elementsmore prominent in granitic (acidic) rocks such as Zr and Pb.Mg concentrations may also be enhanced by some detritaldolomite (Stoffers and Ross, this volume) and unusualconcentrations of palygorskite, a hydrous magnesiumsilicate believed by Stoffers and Ross (this volume) to bederived in large part from transport from the Arabiancontinent rather than by in situ alteration of volcanic glass.Still another source of Mg is magnesium calcite. In thelowermost portion of Site 227 magnesite is reported. Thisunusual mineral is probably related to the late-stageevaporitic minerals (tachyhydrite-bischofite) and magne-sium-rich brines associated with them that are found at thissite (Manheim et al., this volume).

Pyrite

Means of composition of semiquantitative analyses ofsmall pyrite grains and aggregates picked by thepaleontologists (especially R. Fleisher) are given in Table 6.The compositions are rather typical for sedimentary pyritesexcept with regard to the unusually low leadconcentrations. One might note that low leadconcentrations are also characteristic of sphalerite-richseparates noted previously (Manheim and Siems, thisvolume), and the Red Sea sediments generally. This appearsto be partly related to low lead content in detrital materialsentering the Red Sea, it being a rather low lead province ingeneral. The major constituents such as Ca, Ti, and othersare associated with impurities attached to the pyrite grains.

A significant difference in composition of pyrites,associated with color, was observed. Yellow, brassy pyritescontain low arsenic concentrations, whereas silver-whitepyrites, hinting at arsenopyrite, contain large concen-trations (over 1000 ppm As). The arsenic traces weredistributed in an erratic manner, however, and have notbeen linked to any given stratigraphic interval or sedimenttype.

Lithified Carbonates

As discussed in Milliman et al. (1969) and referencescited, lithified layers containing both aragonite and highmagnesian calcite are common in surficial sediments of theRed Sea. The drill cores reveal that aragonite and highmagnesian calcite occurrences are frequent throughout thePleistocene of Sites 225, 228, and 229, but decline withdepth. Deepest aragonites recorded in the X-ray data are at160 meters at Site 227 (early Pliocene). The chemical data(Table 7) reflect presence of both inorganically precipitatedaragonite and magnesian calcite in the lithified crusts fromPleistocene and Pliocene sections. Taken as a whole,however, Sites 225, 227, and 228 contain only smallproportions of aragonite (approximately 5 percent or lessof total carbonate). Since the mean Sr for all samples fromthese sites is usually below 1000 ppm, lower than that

981

Page 8: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

F. MANHEIM

TABLE 4Spectrochemical Analyses of Non-black Shale Sediments by Site and Stratigraphic Horizon

Element

Si (%)bAl (%)bFe (%)Mg (%)Ca (%)Ti (%)Mn(%)Ag (ppm)As (ppm)B (ppm)Ba (ppm)Cd (ppm)Co (ppm)Cr (ppm)Cu (ppm)Mo (ppm)Ni (ppm)Pb (ppm)Sn (ppm)Sr (ppm)V (ppm)Zn (ppm)Zr (ppm)

Site 225

Pleistocene

0-52 m

134.34.94.6

180.30.21

<O.l—

2068

-127732

2.769

8<3

129029

753

99999999_99-99999999999

L. Pliocene

52-115 m

134.73.84.0

180.350.22

<O.l_

1265

_1281392770

6<3

80042

—42

N

1313131313131313_1313_131313131313131313_13

E. Pliocene

115-173 m

175.45.34.0

160.430.2

<O.l_

2496

_309686<296

7<3

45051

_51

TV

77777777_77_777777777_7

L. Miocene

173-227 m

543.82.44.0

130.260.2

<O.l—

15087

_908347

441

6<3

45037

<5040

TV

33333333_33—33333333313

Site 227

Pleistocene

0-28 m

155.03.56.0

200.40.4

<O.l<520

100—

1510060<275

7.5<3

100030

_50

TV

22222222222—222222222—2

L. Pliocene

28-65 m

1454.24.8

200.40.38

<O.l_

1185

_259252

69012<3

67032

_42

N

44444444—44_444444444—4

E. Pliocene

65-225 m

165.04.62.2

140.420.24

<O.l—

23219

—21

1707814

102<50

<3710110

—51

N

1616161616161616—1616-161616161616161616—16

a7V = number of analyses.^Biased on core side by upper detection limits.

estimated to characterize fresh planktonic foraminifera orcoccoliths (about 1500 ppm according to Manheim et al.,1971, and references cited), the dominant contribution ofSr to pore fluid may be made by the recrystallization ofcalcite to lower-Sr forms, rather than the higher-Sr, but lessabundant aragonite (0.9% Sr).

The aragonitic lithified layers in piston cores have beenrelated to episodes of higher than normal salinity at 11,000to 20,000; 50,000; and 70,000 years (Müliman et al., 1969,and references cited). In principle, if these layers could bepreserved during burial to depths on the order of 200meters below the sea floor, with temperatures of about35 C or above, aragonite rather than low magnesian calciteought to be the stable phase. However, increasingtemperature, ionic strength of solutions, and Mgconcentrations in the pore fluids (the latter of evaporiticorigin) promote dolomite formation at the expense ofaragonite. For this reason, aragonite is presumably absent inthe lower levels of Sites 225, 227 and 228.

Mean Sr content of samples from Site 229 is about 2000ppm, considerably higher than at the other sites. This sitealso contains appreciably more aragonite in its carbonatefraction. Although much of this aragonite is attributable tolow-Sr pteropod tests, presumably a significant proportionis also due to post-depositional growth of inorganic(high-Sr) aragonite. The high proportion of Sr (about 100ppm) in the pore fluid may be due to recrystallization ofboth aragonite and biogenic calcite.

Evaporitic Rocks and Mineral Parageneses: PossibleUnrecovered Evaporite Minerals

Some partial analyses of anhydrite and halite, mineralsof overwhelming dominance in evaporite suites, are given inTable 8. However, Stoffers and Kuhn (this volume) showthat a potassium-bearing phase, judged to be polyhalitefrom petrographic observations, is frequently present inamounts over 1 percent (11 out of 29 halite samples). Brand K values, the latter recalculated from Stoffers andKühns's polyhalite values, are plotted for Site 227 in Figure3, along with data from two samples analyzed by the U.S.Geological Survey laboratories. The only sample from thesame lithologic interval (at 228 m) shows excellentagreement between the laboratories' results.

Stoffers and Kuhn discuss the significance of K, Br, andSr data in some detail, arriving at the following chiefconclusions:

1) Using Br/NaCl as an indicator of the degree ofevaporation of an original seawater parent brine, valuesrange from 0.004 to 0.017, corresponding to the boundaryof the anhydrite and halite precipitation zones and thehalite/polyhalite zones, are obtained. It is recognized thatpolyhalite (K2Ca2Mg(Sθ4)4*2H2θ is not normally aprimary evaporite mineral.

2) Bromine values have also been used by Kuhn (1953)to estimate depth of water in the evaporite depositional

982

Page 9: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

RED SEA GEOCHEMISTRY

TABLE 4 - Continued

Element

Si (%)bAl (%)bFe (%)Mg (%)Ca (%)Ti (%)Mn(%)Ag (ppm)As (ppm)B (ppm)Ba (ppm)Cd (ppm)Co (ppm)Cr (ppm)Cu (ppm)Mo (ppm)Ni (ppm)Pb (ppm)Sn (ppm)Sr (ppm)V (ppm)Zn (ppm)Zr (ppm)

Site 228

Holocene

0-7 m

175.06.8

1611

<O.l-

18230

-18

14070<28810<3

84084

-104

TV3

55555555-55—555555555-5

Pleistocene

7-155 m

194.97.63.6

150.520.36

<O.l<IO24

250_

25149

6712

1048

<31220

169—

138

N

2121212121212121212121_212121212121212121_21

L. Pliocene

155-273 m

194.86.12.2

100.560.20

<O.l_

12247

_is

13965<2836.1

<3530117

_116

N

99999999_99_999999999—9

E. Pliocene

273-291 m

22>5

744.20.60.2

<O.l_

30350

_22

17585(2)

110<5<3

150150

_150

N

22222222_22

222222222_2

L. Miocene

291-325 m

194.88.36.05.70.531.3

<O.l

490400

18113157

37254<3

480133

<500100

N

9999999

_99_9999999999

Site 229

Holocene

0-45 m N

14 184.9 183.9 185.0 18

18 180.47 180.06 18

<O.l 18_ _

19 1884 18

_ _12 18

179 1825 18<7.8 18

103 186 183 18

2040 1846 18

_ _190 18

Pleistocene

45-210 m

154.63.43.9

170.420.10

<0.2_

1685

_11

16341<291

8<3

197056

_144

N

1010101010101010_1010_101010101010101010_10

aN = number of analyses.^Biased on core side by upper detection limits.

basin. Using this technique for halite sequences at Sites 225and 228, a water depth of only a few meters was deduced.

3) Dirty gray halites have abundant fluid inclusions incontrast to clear white halites. The latter are presumed tobe recrystallization products. However, the clear halitetended to have "enigmatically" higher bromine concentra-tions.

4) Sr/Ca lOOO values associated with anhydrites varyfrom 3.5 to 25. Values between 1 and 3 are held to denoteisomorphous Sr-Ca replacements. Discrete strontiumminerals are found at values greater than 7, whereas onlycelestite should occur in anhydrite for values above 20.Primary anhydrite ought to have greater Sr content (0.23)than gypsum (0.23), according to Usdowski, cited inStoffers and Kuhn (this volume). About half of theanhydrites showed Sr content greater than 0.23. Based onthis, as well as the anhydrites' massive habit, the authorsconcluded most anhydrite to be primary, rather thanrecrystallized gypsum.

5) Although Stoffers and Kuhn hold to a shallow originof the observed evaporites, they suggest that earlierevaporites might have been deposited in deep wateraccording to the model of Schmalz (1969); i.e., the latestdeposits are only the culmination of salt buildup in a deepbasin. The waters were presumably supplied from theMediterranean during "waterfall" episodes betweenintermittent desiccation of both basins. This would be an

extension of the concept of Hsü et al. (1973) for theMediterranean Sea.

Before commenting here and in succeeding sections onStoffers' and Kühn's concept, one should recognize that therecovered evaporites and associated rocks represent onlyabout one half or less of total penetrated intervals;moreover, the halites had been reduced to about half ofwireline core diameter by leaching during core retrieval.Thus, it is quite possible that other, more soluble evaporiticphases were lost in the coring process. Polyhalites would beretained since this sulfate mineral is more akin to anhydritethan halite in its solubility.

Fortunately, interstitial waters in shales associated withthe evaporitic rocks yield data from which one can estimatethe nature of original evaporites at depth. This is becausediffusivity measurements discussed later indicate that in thecourse of geologic time, pore fluid within associated shalesshould be in virtually complete equilibrium with evaporitessome meters away. However, the shales have very poor bulkpermeability, so that only their outer layers (including themargins of breaks and cracks) will be contaminated bydrilling fluid during core recovery. By removing these outerlayers prior to squeezing, contamination can be minimized.

At Site 225 the interstitial fluids in the deepest sampleshow marked increases in Mg, K, and SO4, relationshipsthat could signal approach to equilibrium with a mineralassemblage of halite, anhydrite, and polyhalite. However,Site 227 shows a quite different picture. Here the non-NaClcomponent of the brine is dominated by Mg-Ca chlorides,

983

Page 10: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

F. MANHEIM

TABLE 5Chemical ("Rapid Rock") and Spectrochemical Averages for Major Sediment

Types for Red Sea Strata, Excluding Hot Brine Deposits

Element

SiO2

SiA12O3

AlFe2θ3FeTiO2

TiCaOCaMgOMgMnOMnSrOSrNa2ONaK 2 OK

P2O5

OrganicCarbon

PbAgCuZnSnCdMoNiCoVBaSrBAsZrCr

Black Shale

Pre-Miocene

7V=33Spectro.

31.714.8

6.84.878.35.80.650.39

19.914.2

3.562.150.310.24

11<0.13

164124

<3-

77149

50398496814

267

70107

N=3Chern.

43.520.313.2

9.53.84.8a0.760.466.504.655.333.210.110.0850.120.0103.012.231.931.600.240.10

3.3b

6NAd

132104

NANANANANANANANA

53NANA

NA

5

Miocene

N=18Spectro.

47.322.1

6.54.38.96.20.670.405.604.07.464.50.540.42

0.00

23<O.l

161321< 3

—128335

120179545385<57292

N = 1Chern.

46.521.710.6

7.636.807.7ia0.640.381.10.798.24.950.210.160.010.0082.291.701.811.500.170.074

1.7C

NA__NA

NA——

NA_

NANA

_NANANA

Nonblack Shales

Pre-Miocene

7V=116Spectro.

34.015.9

6.744.857.335.130.730.44

21.815.6

7.154.310.280.22

7<O.l54

_

(3)_

(9)921884

153114

19—

89112

iV=22Chern.

29.413.7

8.96.42.93.34a0.580.35

22.616.2

3.892.350.140.110.0830.0701.451.081.201.000.260.11

5NA

4943

NANA

1157NA

107NANA

41NANANA

Miocene

N=12Spectro.

59.527.8

6.324.559.776.830.770.46

10.57.539.125.501.331.03

42<O.l

129_

(3)——

6416

109320470404

_85

105

N = 1Chern.

39.218.3

8.66.22.62.75a0.430.269.66.868.75.250.080.0620.020.0172.591.921.401.160.370.16

5NA

7651

NANA

(2)20

NA9NANA

280NANANA

Note: A refers to number of samples for major constituents. Values are in percent dry weight formajor constituents and ppm for trace elements (shown as element form only). Note that semi-quantitative analysis method is primarily directed toward trace elements and is less reliable formajor constituents. In some cases, Al and Si spectrochemical means may be biased on the low sidebecause of the upper limit of technique. Organic carbon determined at DSDP laboratory, ScrippsInstitution of Oceanography, La Jolla (G. W. Bode).aTotal iron.

Mean of 20 analyses.

Mean of 6 analyses.

NA = not analyzed.

suggesting a bischofite (MgCl2 6H2O)-tachyhydrite(Mg2CaCl6 12H2O) mineral assemblage. Table 9 sum-marizes the non-NaCl components in Site 225 and 227brines, together with corresponding components of

comparative minerals possibly present in the evaporiticrocks. Site 228 had not yet reached the halite-stage rocksand the interstitial waters here are not in equilibium withthem, though they rapidly approach saturation with NaCl

984

Page 11: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

RED SEA GEOCHEMISTRY

TABLE 6Pyrite Samples, Sites 225-229

TABLE 7Composition of Lithified Carbonates

Element

Si (%)Al (%)Fe (%)Ti (%)Ca (%)Mg(%)Mn (%)Pb (ppm)Ag (ppm)Cu (ppm)Zn (ppm)Sn (ppm)Cd (ppm)Mo (ppm)Ni (ppm)Co (ppm)V (ppm)Ba (ppm)Sr (ppm)B (ppm)As (ppm)Zr (ppm)Cr (ppm)

Pleistocene

4.31.1

>200.0121.70.020.032

(5)(0.1)60

(140)(3)

-33327341.711.31016.833

(1000)10

7.3

N

333333333323-3333333133

Late Pliocene

5.31.5

>200.043.00.0760.067

(5)(0.1)

125-

(3)-

185293

8813

(11)8319

(150)5.8

41

N

6666666666-6—6666666666

Early Pliocene

8.02.7

>200.0354.31.20.055

130.17

157417

(3)

103250

4113

(10)6718

(360)(5.8)8.3

N

666666656626—6666666666

with depth.Another mechanism is suggested by laboratory

experiments of Raup (1973, oral communication) which areconsistent with suggestions of D'Ans (1961), field data ofHolser (1966), and particularly pertinent to thepolyhalite-tachyhydrite problem: this is the reaction ofgypsum or anhydrite with tachyhydrite-bischofite-relatedsolutions to form polyhalite and calcium chloride attemperatures greater than room temperature.

CaMg2Cl6 12H2O+ 4KC1(tachyhydrite)

(polyhalite)

8CaSO4

5CaCl2 + 4H2O

The available evidence suggests a qualitative scheme forthe development of mineral parageneses found at Sites 225and 227. The evaporitic rocks in both areas were laid downunder conditions that irregularly permitted accumulation ofminor late-stage evaporites, such as carnallite, sylvite,tachyhydrite, and bischofite, as well as other phases,together with concentrated chloride-rich brines. Possiblysome secondary polyhalite was already formed at this stage.However, additional polyhalite may have formed fromreaction between interstitial brine and anhydrite at asomewhat later, post-burial stage, when temperaturesincreased further. Recrystallization of earlier formed halitesand crystallization of polyhalites formed the clearer, laterbands, whereas the early formed halite may be representedby the gray phase described by Stoffers and Kuhn (thisvolume) as containing abundant fluid inclusions andchevron and hopper structures. This relationship wouldexplain the higher Br concentrations in the clear halite

Element

SiO2

Si

A12O3

Al

Fe2O3

Fe

TiO2

Ti

CaO

Ca

MgO

Mg

MnO

Mn

Na2O

Na

K2O

K

P2O5

P

Pb (ppm)

Ag (ppm)

Cu (ppm)

Zn (ppm)

Sn (ppm)

Mo (ppm)

Ni (ppm)

Co (ppm)

U (ppm)

Ba (ppm)

Sr (ppm)

B (ppm)

As (ppm)

Zr (ppm)

Cr (ppm)

Pre-Miocene

N = 9Spectro.

19.5

9.13

5.22

3.76

2.09

1.46

0.43

0.26

26.4

18.9

5.37

3.24

0.084

0.065

« 5 )

<O.l

16

<óO

<3

<2

35

7

26

30

3410

13

-

37

86

N = 2Chern.

22.6

10.56

6.70

4.82

2.25

2.10a

0.59

0.35

27.8

19.9

5.0

3.02

0.12

0.093

0.69

0.51

0.87

0.72

0.48

0.21

layers, since these may have formed in equilibrium with abrine partly enriched with late-stage components due toboth mother liquors and dissolution of solid phases. Sinceboth carnallites and tachyhydrites can contain more than4000 ppm Br in their structures, dissolution of suchminerals could yield very high Br (and also B, Li, etc.)concentrations. A considerable part of the originallytrapped interstitial solutes may have diffused away duringthe immediate post-burial stage of the evaporites, so thatthe interstitial waters now remaining represent a closerapproach to equilibrium with the evaporite minerals.

985

Page 12: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

F. MANHEIM

TABLE 8Analyses of Evaporitic Rocks

σ o o oin halite

Anhydrites

Major Constituent 225-24-1135/140 227-37-2 228-37-2

CaOSrOMgOK 2 OSO3SUM

Traces (ppm)

FeMnTiBBiU

(38.9)0.130.530.05

55.595.1

—< l

<IOa β

(38.9)0.154.50.04

55.597.1

<20205034

<IO0.8

(39.2)

σ.zσ0.530.02

55.995.6

—< l

<IO0.6

Halite Rock

Major Constituent 225(%) "200 m" 227-27-2 227-31, CC 227-93-4

CaOa

SrOMgOK 2 OSO3

Traces (ppm)

BBrU

(3.8)0.020.080.044.6

160<IO

0.3

(1.1)0.010.030.011.34

16033

0.4

(L04)<O.Ol

2.80.631.25

200110

0.4

(0.47)

0.020.040.57

28044

0.4

Note: Fe, Mn, and Ti from semiquantitative spectrochemicalanalyses. Other components by U. S. Geological Survey, AnalyticalLaboratories, Washington, D. C.aCa corresponding stoichiometrically to SO4; does not include other

(e.g., carbonate) forms.

According to Manheim et al. (this volume) the driedinterstitial salt in the bottom of Site 227 (350 meters)would correspond to 66 percent halite, 32 percenttachyhydrite having a composition of CaQ.6Mg2.4Cl6 12H2θ, 1.5 percent carnallite, and 0.5 percentanhydrite. Although tachyhydrite is a hitherto rarely foundmineral, there is evidence that members of theMgCl2-CaCl2-H2θ mineral series may be far moreimportant than commonly realized. Highlighting thisconcept is the recent discovery of bedded primarytachyhydrite aggregating more than 100 meters thickness inLower Cretaceous evaporite deposits on the east coast ofBrazil (Wardlaw, 1972). One should note that tachyhydriteis not stable at room temperature and is extremelyhygroscopic (CaCl2 being commonly used as a desiccant).Thus, in the Sergipe (Brazil) deposits some coresimperfectly coated with paraffin were found later to havecomplete liquified (Raup, 1973, oral communication).Therefore, many evaporite deposits, especially thoseassociated with high-CaCl2 brines, may have containedminor admixtures of tachyhydrite at depth (and under

Depth (m)

200

, 0 -

20 -

30 -

40 -

250i

60 -

70 -

80 -

90 -

300 -

10 -

20 -

30 -

40 -

350

0 .2 .4 .6 .8 1.0 1.2 1.4 1.6 K % t o t a l rock

SITE 227

High Mg-Cachloride brines

• K data of Stoffers & Kuhn (1974)D K data USGS Laboratoriesx Br data Stoffers & Kuhn (1974)O Br data USGS Laboratories

Figure 3. Distribution of Br and K with depth in Site 227.

TABLE 9Mole Percent of Components In Evaporite Minerals and

Residual Interstitial Brine Salts (Excluding Halite)

Mineral K Mg Ca Ch SO4

Polyhalite (K2Ca2Mg(SO4)4 2H2O) 22.2 11.1 22.2 44.5

Carnallite (KMgCl3

Bischofite (MgCl2

•6H2O)

6H2O)

Tachyhydrite (CaMg2Clé • 12H2O)

Site 225a

Site 226

Site 227b

Seawater

20

-

-

9.5

8.4

0.9

6

20

33.3

22.2

22.2

4.6

26.1

31

-

-

11.1

4.2

23.9

6.4

6

60

66.6

66.6

51.8

61.8

65.2

40

-

-

-

12.2

1.2

0.5

17

a28-l(H), 212 meter depth. Sample extracted while heated to approxi-mate in situ conditions.

t>44, CC, 350 meter depth. Sulfate estimated from equilibrium data forsaturation with anhydrite at 50° C, as indicated in Manheim et al., thisvolume, Table 2.

higher temperatures), which mineral was lost in the courseof core recovery and manipulation.

The pathways by which dominantly calcium chloridebrines are formed have been much discussed, since excessesof calcium over magnesium are much more common thanthe reverse. This is a serious problem, since late-stagemarine brines should be expected to be depleted in calcium.Discussions of this problem, with references to othersources are offered by Zaitsev (1968) and Wardlaw (1972).

986

Page 13: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

RED SEA GEOCHEMISTRY

Sources of calcium pertinent to the present case includedolomitization

MgCl2 + 2CaCO3 MgCa(CO3)2 + CaCl2

ion exchange with clays, or exchange reactions with silicates

2NaCl CaAl 2Si 20 8

(anorthite)2SiO2

(silica)2H2O

2NaAlSi2O6 H2O + CaCl2

(analcime)

After burial and consolidation of the layers, the almostcompletely impermeable halite beds should have shieldedthe deeper intraevaporite units from interaction with laterrocks or from appreciable fluid or ion loss.

Where late-stage evaporites are precipitated andredissolved, one might expect a great variability in traceelement enrichment in the resulting liquids. Halites andpolyhalites that are precipitated or recrystallized inequilibrium with such solutions could show quite extremevariations that might influence calculations using the Brdepth and Sr indicators previously mentioned.

Composition of Hot Brine Sediments (Site 226)

The disturbed nature of the sediments from this site didnot offer much promise of improvement over previousdetailed studies of the metalliferous brines (Bischoff, 1969;Hackett and Bischoff, 1973). A general comparison of thesemiquantitative analyses with previous averages is given inTable 10.

Taking into account the experimental error, the onlypossibly significant difference between the presentlyrecovered sediment and previous studies is in theappreciably higher Pb and lower Mn concentrations. Nolead values were found in the previous studies as high as the

TABLE 10Comparative Data for Hot Brine Sediments

SiO2

A12O3

F e 2 θ 3

Mn3θ4

CaO

ZnO

CuO

PbO

AgO

Fe-Mont.

24

1.7

35

2

4.8

4.7

0.7

0.1

0.0062

Goeth.

24

1.1

49

2.8

3.4

1.0

0.5

0.1

0.0033

Site 226 (l-6m?)

15-27

5

20-30

0.4-1.0

5-22a

1

0.7-1.0

0.7-0.5

0.003-0.007

Note: "Fe-Mont'.' and "Goeth." refer toFe Montmorillonite and geothite faciesof Bischoff (1969); Site 226 data fromTable 2 in Manheim and Siems (thisvolume). Values are in percent air-dry,largely brine-free, sediment.

aReflects admixture of anhydrite.

0.5 percent registered here. However, the spectrochemicalestimates were not confirmed by a sample from Site 226(1-4) analyzed by Delevaux and Doe (this volume) andhence should be taken cautiously.

Carbon and Oxygen Isotopic Composition of theSediments

Isotopic relationships in bulk carbonates (Lawrence, thisvolume) are subject to a variety of influences: individualrelationships characteristic for given carbonate tests ,partial equilibration with Red Sea bottom water on deathof organisms, and equilibration with interstitial water andother solid phases of buried sediment. Lithified carbonatesfrom a variety of subeuphotic environments (Milliman andMuller, in preparation) show a consistent correlationbetween δ C 1 3 and δ θ l 8 (line A-A' in Figure 4). It issuggested that this apparent equilibration is dominated bythe nature of Cθ3=ion at the time of precipitation of thecalcium carbonate, rather than protracted equilibration ofthe solid phase with water or carbonate-water systems aftersediment burial. In contrast, a scattering area roughlyindicated by line B-B' shows the behavior of freshplanktonic foraminifera and coccolithophorids and manybiologically formed shallow-water limestones. Still otherpatterns characterize volcanic or other unusual carbonates.

The isotopic data for Red Sea bulk carbonates fallclearly on the "lithified sediment" line in Figure 4,indicating substantial crystallization of the carbonates

13C (POB) in

α

7oo

©

A Q >α

+ 8 -

7 -

6 -

5 -

4 -

3 -

2 -

+1 -

0

-1 -

2 -

3 -

4 -

5 -

6 -

7 -

1 0 9 8 7 6 5 4 3 2 - 1 0 + 1 2 3 4 5 6 7 8 9 1 0

18 nδ 0 (PDB) m /oo

O SITE 225 (LAWRENCE, this volume)× SITE 228 (LAWRENCE, this volume)Q SITE 225, 227, and 228 (SUPKO, et al., this volume)• Intra-evaporite dolomites (SUPKO, et al., this volume)Globigerinas, Site 228 (Deuser; this volume)

Figure 4. C13 and O18 data for carbonate sediments fromSites 225, 227, and 228, Red Sea. Line A-A' refers totrend line for subeuphotically lithified carbonates, worldoceans. Line B-B' roughly indicates region of scatter forshallow water (euphotic zone) carbonates, virtuallyexclusively of biogenic origin.

987

Page 14: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

F. MANHEIM

below the photic zone. The great spread in data reflects inpart the variety of environments, both "evaporated," orotherwise heavy-isotope-enriched, and depleted (fresh watertype), characterizing the Red Sea sediments.

A comparison of the oxygen isotopic data from bulkcarbonates and interstitial waters shows a wideningdivergence toward greater depths. Both bulk carbonate andinterstitial water isotopic ( δ θ 1 8 ) ratios tend to becomelower with depth, reflecting the meteorically influencedwater input in the evaporite sequence, but the interstitialwaters show the more pronounced change. Lawrence (thisvolume) points out the possibility that weathering ofvolcanic ash or igneous detritus ( δ θ 1 8 +6 to +13 % o

SMOW) to authigenic silicates such as montmorillonite( δ θ l 8 + 22 to +28 % o ) could influence the isotopiccomposition of recrystallized carbonate. However, nogeneral increase in δ θ 1 8 of the bulk silicate with depth isfound ( δ θ 1 8 approximately +12 to +18 °/O0 for Site 228).Isotopic ratios for some intraevaporitic dolomites alsoshowed equilibration of the carbonates with light,"meteorically influenced" waters. We may reject theconcept that the enrichments in light isotopes originate bylocal fractionation effects (e.g., reduction of sulfate) owingto the consistent δD - δ O 1 8 relationships in the pore fluids(Figures 6 and 7 and later discussion).

Uranium, Lead, and Sulfur Isotopic Data

Ku (this volume) points out that Red Sea sediments ingeneral have relatively low Th/U ratios (1.6-2.7). Uraniumconcentrations, except for Site 229, are comparable tothose of average shales and pelagic clays (2-3 ppm). Site229 showed U values from 8 to 19 ppm of the bulksediment, or up to 60 ppm on an assumed carbonate-freebasis. Th/U ratios were only 0.13 to 0.48. Post-depositionalmigration of U is assumed on the basis of the T h 2 3 0 / U 2 3 4

data. Although radiometric dating did not prove feasible forthe Site 229 materials, in spite of the rapid accumulationthere, an estimate for Site 228 (Core 1, Section 1) providesa rough figure on the order of 10 cm/103 years, in generalagreement with previous radiometric data from the area aswell as paleontologic estimates.

Lead isotope data (Delevaux and Doe, this volume)generally confirm previous findings relating leads fromAtlantis II deep metalliferous sediments to a mantle(possible leached volcanics) origin, and establish theiraffinity to ores found at Rabigh, on the Saudi Arabiancoast. Similarly, lead from Site 227 (Core 20, Section 4)showed leads derived from pelagic sediments, comparableto the most radiogenic sediments from the Red Seareported earlier.

A very important result was shown by the lead isotopiccomposition of the brecciated anhydrite-shale sequence(Core 39, Section 1) at Site 228. The lead-enriched sampleis isotopically related to the Atlantis II and Discovery deepsmetalliferous samples, and thus provides the first geneticlink between metals in the known hot-brine deeps andother Red Sea sites. This point is discussed further later.

δS 3 4 values of -33 to -36 from Site 227, Cores 3 and 13,are typical for H2S and associated metal sulfides producedfrom reduction of sulfates by bacteria (Shanks et al., thisvolume) in recent marine basins. On the other hand,

sulfides in heavy mineral separates from the brecciatedshales and anhydrites in the bottom of Site 228 (Core 39),showed δS 3 4 values ranging from -12 to +22 °/oo. Themean, +3 % o , is not far from the mean of the range foundfor hot brine sediments by Kaplan et al. (1969) andHartmann and Nielsen (1966). The new data for Site 228do not contradict an origin of the sulfides related to that ofthe hot brine deep sediments, but the exact source andmode of origin of the sulfides in both remains controversialand not fully understood.

Potassium-Argon Measurements on EvaporiteAssociated Rocks

Measurements of Ar40 and K on anhydrite fromMiocene strata at Site 227 and on shale from Site 228, Core37, Section 1, 110 cm were performed on whole-rockspecimens.2 The anhydrite could not be dated owing toinsufficient (negligible) potassium content. The shale, froma depth of about 305 meters, provided the following data:

K(%) Ar4 0(ppm) Ar40/Tot. Ar apparent age (m.y.)3.35 0.0404 0.700 162 ± 6

The apparent age yielded by this measurement is far toogreat to correspond with the real age of the horizon, whichis presumably Miocene. There are two possible explanationsfor this anomaly. First, the very fine clay material of whichthe sample is chiefly constituted may be detritus from oldersediments and other rocks eroded from the Sudanesemainland. Coleman (this volume) reports that heavyminerals from Plio-Pleistocene rocks at this site suggest anaffinity to granitic metamorphic rocks, possiblyPrecambrian rocks and their weathering productsdistributed through the former delta region on theSudanese continental margin near Site 228. A secondpossiblity is that excess argon migrating from potassicmaterials in deeper strata has been incorporated into thefine-grained clays at this site. Although the evidence offluid migration is strong, the first explanation is probablyto be preferred, partly because most of the indigenous clayminerals associated with the deeper Red Sea strata tend tobe montmorillonite and chlorite (Stoffers and Ross, thisvolume) which have considerably lower K content than thepresent shale.

FLUIDS AND GASES

Hot Brine Deeps

Extensive information is available on the properties ofhot brines and sediment interstitial waters from the hotbrine sediments. These data and current information arecompiled in Tables 11 and 12. With some exceptions, mostof the interstitial water data fall within analytical ormanipulative error limits of the precise data for Atlantis IIand Discovery deeps brines reported by Brewer and Spencer(1969). Most variable are Ca and SO4 both in Brewer andSpencer's Discovery Deep samplings, as well as in interstitialwater studies of Hendricks et al. (1969). These may beexpected, since interstitial waters are often associated withsolid anhydrite, and changes in temperature and other

'Krueger Enterprises, Inc., Geochron Laboratories Division.

988

Page 15: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

RED SEA GEOCHEMISTRY

TABLE 11Major Element Composition of Fluid Phases in Red Sea Sediment Hot Brine Systems

Element

α

ßr

so4<CO2 as HCO3

Na

K

Ca

Mg

Total dissolved solids

Red SeaDeep Water

22.5

0.076

3.14

0.14g

12.5

0.45

0.47

1.49

40.6

Atlantis II Deepa

156.3157.3d

0.128

0.840

0.075-0.132g

92.6

1.87

5.15

0.764

257.3

Discovery Deep

155.3

0.119-0.123

0.75-0.70

0.032h

92.8-93.0

2.15

4.7-5.1

0.81

256.4

A-II Deep, 1965Interstitial Waters^

154-160

0.95-0.13

-0.5

-

93-97

2.2-2.51

4.7-5.7

0.66-0.73

-

A-II Deep, 1972Interstitial Waters0

157.5e

0.10f

0.79

(alk HCO3) <0.03

93.1

2.10

6.0

0.70

260.4

NOTE: Red sea bottom water is estimated from normal oceanic salt composition (Culkin, 1965) adjusted to normal bottom waterchlorinity for the rift valley positions as given in Miller et al. (1966). Surface water salinities vary. Atlantis II Deep chlorinityfrom (d) refers to sampling of 1971, whereas other value represents sampling of 1965. Major element interstitial water data aretaken chiefly from Core 84K data of Hendricks et al. (1969) since their extractions were performed onboard. Some major elementdata (g) also available in Brooks et al. (1969), but their extractions were performed on stored cores and are accordingly lesspreferred. K values in data of Hendricks et al. (1969) are consistently much too low and may be a typographic error or faulty calibra-tion. All values in g/hg.aBrewer and Spencer (1969). Shishkina and Bogoyavlenskii (1970) also agree with the authors' results.bHendricks et al. (1969).cManheim et al. (this volume) mean of two samples from Site 226, SW A-II Deep: 226-1-2 and 226-1-4 (IW)dBreweretal. (1971).eIncludes Br and I.tPresley et al. (this volume).

gSchoell and Stahl (1972); Weiss (1969).hWeiss (1969).

Kaplan et al. (1969) for Core 84K.JHighly variable; data of Kaplan et al. (1969) also show scatter.

reactions in the sediments may cause shifts insediment-water reactions involving this phase. Not includedin the table are the interstitial values of Hendricks et al. forDiscovery Deep. In spite of scatter, these values also tend toreflect compositions in the Discovery Deep waters. A pointto be made is that there are few indications of anyconsistent changes in interstitial constituents with depth.It would rather appear that the sediments are permeatedwith brine of the composition currently discharging intothe rift basins and show little evidence of retaining brine ofan earlier, different type. As will be shown later, however,the hot brine basin fluids bear little resemblance tointerstitial waters associated with evaporite deposits at Sites225, 227, and 228, and one may conclude that they have adifferent history.

Trace and minor constituents, compiled in Table 12, aresubject to much greater analytical and manipulative scatterthan the major constituents. Nevertheless, examination ofthe data by various investigators reveals several consistentpatterns and invites the following conclusions.

1) Fe and Mn concentrations are in the range 50-80ppm for both the Atlantis II Deep brines and sedimentinterstitial waters and have remained so over a period of 7years. Fe, but not Mn, is depleted in the waters of theDiscovery Deep, presumably due to oxidation and

precipitation phenomena. To a lesser degree it is alsodepleted in interstitial waters.

2) Br and B are only moderately enriched in the hotbrines with respect to Red Sea bottom water and interstitialwaters of the metalliferous sediments (less than a factor of2), which implies a sharp depletion with respect to Cl ortotal dissolved salts. F is almost totally depleted in theAtlantis II Deep brines; this has been used by Craig (1969)to show that there has been little contribution by Red Seabottom water to the hot brine composition.

3) Sr is enriched five-fold in both hot brines andinterstitial waters over Red Sea bottom waters, but thisrepresents a depletion with respect to Cl or Na. On theother hand, Li, forming few difficult soluble components,maintains concentrations on the order of 5 ppm in both hotbrines and interstitial waters; this represents a modestenrichment for Li/Cl with respect to normal seawater (3 XI0'5 vs. 1 X105).

4) Zn, Pb, and Cu show very substantial enrichmentswith respect to normal or Red Sea bottom waters ofapproximately 5, 0.5 and 0.3 ppm, respectively. Excludingthe stored-core samplings of Brooks et al. (1969) (their veryhigh Zn and Cl values are possibly artifacts of storage andreaction with galvanized iron box core tubes), these valuesare not far from means of the interstitial waters, even

989

Page 16: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

F. MANHEIM

TABLE 12Minor and Trace Constituents in Fluid Phases of Red Sea Sediment-Hot Brine Systems

Element

AlAsAgBBaBrCdCo

Cu

F

Fe

ILi

Mn

MoNH3-NNO2-NNO3-NNi

Pb

PO4-PSeSiθ4-SiSr

UZn

Red SeaBottom Water

_—

5.3—76

1.6

0.21

_

0.01-

<0.001 c

O.lic—

_

0.02c_

0.34c9.2

3.3—

AtlantisII Deep

<O.lb_

10 c

0.9C128e

0.16e

<0.001-0.017g0.26^

0.03-0.52g

<0.02h8ie

83-100g82i_

4.1-4.78

82e83-100g

_-

(0.001 )c(0.011)c

0.3-0.52g0.6210.5J

(0.02)c0.15b

(12.3)c4 8 e

45i0.48k5.4e

4.6-10g

DiscoveryDeep

<O.lb_

7.54<0.3c

119-123e

0.000-0.005g0.13e

O.OOlg0.075e

0.06-0.34g0.014h0.05ie0.27e

0.05-0.15g

0.03"4.0-4.1g

0.26h54.6e

56-66g_-

(0.001)c(<O.OOl)c

0.002g0.34h

0.1710.08-0.21g

<O.Olc<0.05b(15.7)c

46e44i_

0.77^0.8-2.8g

Interstitial WaterAtlantis II Deep

1.7-5.8a

_

0.008-0.077.5-10a

_108-142a0.5-1.78

0.75-2.9a0.3-4.78

0.25-1.4a

16-115a

46.6-6.7g54-89g59-88a

O.O3-O.3a---

0.5-0.8a2.3-3.5e

0.25-0.83a3.0-4.2g

2-13a_

4.9a43-5 8a55-72g

_

l-13a5-400g

Interstitial WaterDiscovery Deep

1.7-6.7a<o.ii

0.03-0.047.5-13a

_92-125a0.5-0.7g

1.2-2.2a0.3-0.6g

0.17-0.58a

4.5-29a

4.6-5.9825-5 9815-6 8a-(0.3)a

---

0.6-0.75a2.4-2.9g

0.4-0.75a3.0-3.98<l-10a<0.053-5 a

53-83a58-728

l-13a1-1688

Interstitial WaterAtlantis II

_

--

8 4 0 d

l.Od9 8 f

0.20*

--

0.45

-

52f

-

4_7d3.15

70*—

<0.55--2.2f

----

45d-

15*

Note: All values in ppm.aHendricks et al. (1969) (chiefly chemical data cited).bKaplan et al. (1969).cMiller et al. (1966).

Manheim et al. (this volume).e Brewer and Spencer (1969).1 Presley et al. (this volume.)SBrooks et al. (1969).hBrewer et al. (1966)1 Faure and Jones (1969).j Delevaux et al. (1967).kKu (1969).

though the latter show a great deal of scatter. There is asharp depletion in the heavy metals on moving from theAtlantis II to Discovery deep waters, probably owing tocoprecipitation of the metals with iron.

Selective precipitation of the brine metals is clearlyresponsible for the formation of the heavy metal deposits inthe Atlantis II Deep. Miller et al. (1966) and Craig (1969)stressed that the metal concentrations in the brines areprobably not unusual for sedimentary brines associated

with evaporite deposits and require no input of volcanicemanations or other unusual sources. On the other hand,the lead isotope data appear to require some mantle orvolcanic associations for the brine and sediment lead(Delevaux and Doe, this volume, and references cited). Anumber of the constraints on the system have beendiscussed in great detail by Craig (1969), but in the writer'sopinion no subsequent evidence has been uncovered thatclarifies the uncertainty regarding possible metal source.

990

Page 17: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

RED SEA GEOCHEMISTRY

Interstitial Waters from Non-Hot Brine Sites andTheir Significance

As has been shown by Manheim et al. (this volume),Sites 225, 227, and 228 reveal marked interstitial increasesin chlorinity and total salt content with depth, reachingnear-saturation (NaCl) values in the late Miocene evaporitestrata (Figure 5). The 16 percent increase in interstitialchlorinity at Site 229 is much more gradual than at thepreceding sites, but in view of the extraordinarily rapidaccretion rate of the sediments, is believed to signifyprobable presence of evaporites at depth at this site as well.Likewise, a single sample from Site 230 already showedsuch strong enrichment in chlorinity above that of bottomwater (33.0 versus 22.4 Cl °/oo) at a depth of a few meters,that halite deposits could not have been far below. Thus, itappears that salt deposits and associated saturated brinesare omnipresent in the southern half of the Red Sea.

Though the "non-hot brine deep" brines resemble thehot brines in total salt content, their ionic composition ismarkedly different as may be seen in Table 13. Here, thesodium chloride component corresponding to total sodiumin the brines is subtracted from the values and the molarproportions of the remaining major components is listed. Itis significant that each site has its own chemical"signature."

Below continuous halite beds, we can anticipate littlecommunication between the fluids and those above, unlessfaulting or fracturing rupture the salt structure sufficientlyto create breaches in it. Thus, the fluids are representativeof interstitial waters either buried with the evaporites oremplaced before final consolidation and sealing of thesediments. At the upper margin of the evaporites, leachingand diffusion away of solubles create solution breccias andother disconformities, while allowing soluble constituentsto permeate overlying strata. Some constituents, such as Kand Mg, are clearly removed from solution by reaction withoverlying rocks, whereas nonreactive species such asinterstitial Cl more accurately reflect the salt migrationprogress (Manheim et al., this volume). The fact thatdolomites within or just above the evaporites are morestoichiometrically perfect than Ca-rich protodolomiteshigher in the section may reflect the greater exposure toMg-rich interstitial solutions at the evaporite juncture.

Though trace metal (Table 14) analyses are not availablefrom the most concentrated interstitial brines, they alreadyexhibit significant enrichments in Cu and Zn atintermediate salt concentrations.

Isotopic measurements on the interstitial waters areparticularly significant because they offer the potential oftracing the source and history of the waters, in a wayunrelated to the salt concentrations they may have acquired

SALINITY (°/oo)

<X>OO

o o o o o o o σ o o oO C \ J * d - < . 0 0 0 O C \ J « 3 - t £ > 0 0 O•— i— i— ,— ,— N CM CM CO N OO

1 0 0 -

O -<L>Q

200-

3 0 0 -

Figure 5. Distribution of interstitial salinity (total dissolvedsalt) with depth in Sites 225, 227, 228, and 229 (RedSea).

TABLE 13Mole Percent of Components in Non-NaCl Residue of

Atlantis II Hot Brine Salts and InterstitialWaters of Sites 225, 226, and 227

Mg

Ca

K

Cl

so4

"Hot Brine"Atlantis II Deepa

5.1

20.6

7.7

59.9

1.4

Site 226

4.6

23.9

8.4

61.8

1.2

Interstitial Waters^

Site 225

22.2

4.2

9.5

51.8

12.2

Site 227

26.1

6.4

0.9

65.2

0.5

Seawater

31

6

6

40

17

aFrom Brewer and Spencer, 1969.

Manheim et al., this volume.

TABLE 14Selected Minor Constituents in Interstitial Waters

from Presley et al. (this volume)

Depth (m)

214

166

0.5a

Sample

225-28-1

228-21-2

226-1-1

Fe

2.6

3.0

52.0

Mn

0.8

3.1

70.0

ppm

Cu

0.5

0.4

0.48

Zn

4.6

1.6

15.0

Cd

0.20

0.07

0.20

Ni

5.0

1.0

2.0

aAtlantis II Deep.

991

Page 18: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

F. MANHEIM

through leaching of solid evaporites. The deuterium(Friedman and Hardcastle, this volume) and O 1 8 data(Lawrence, this volume) both show indications ofdecreasing heavy isotope content with depth at Sites 225,227, and 228. This behavior does not reflect fresherwater environmental stages during post-Miocene time(Fleisher, this volume), judged by the paleontological data.In fact, the change of isotopic ratios to "lighter" valuesproceeds continuously with increasing interstitial salinity(Figure 6). From this we must conclude that the waters arediffusing upward from a "lighter" isotopic reservoirassociated with the evaporitic strata, in the same mannerthat soluble salts are diffusing upward driven by theconcentration gradient. In the former case, however, therecannot be a steady-state system, for there is not unlimitedwater reservoir, as there is concentrated brine that can becontinuously formed by leaching the halite, anhydrite, andother solid evaporite minerals.

Deuterium concentrations (ratios) at Sites 225 and 227(Figure 6) show very close correspondence to salinitiesapproaching saturation with halite (at the boundary ofoccurrence of halite in the section). At this point, however,interstitial fluids from Site 227, which penetrated another166 meters into the evaporites, drop very sharply in δD toreach relatively constant levels of -2 to -4 °/oo within theevaporite strata. The sharpness of the drop is attributed tothe very poor diffusional permeability of the anhydritic-halite strata. For all practical purposes diffusive or bulkfluid movements within the evaporitic strata at Site 227 areprobably interdicted (see diffusion studies later). Probably,evaporites had not yet been penetrated deep enough to find"original" interstitial brines unaffected by diffusion fromabove; that is, at a still greater depth "lighter fluids" mighthave been found.

Figure 7, plotting δD against δ θ 1 8 shows clearly that theinterstitial fluids from Sites 225, 227, and 228 are not partof the Atlantis II Deep hot brine system, as delineated by

Craig (1969). Lawrence (this volume) has pointed out thatthe deeper samples at Site 227 show a highly anomalousbehavior, falling to very low δ θ 1 8 values withoutcorresponding depletion of δD. Since there is no significantreservoir of H within the sediments, Lawrence was forcedto conclude that the δ θ l 8 values had been depressed bythe addition of low δ θ 1 8 waters from reaction with basaltsor other volcanic rocks at depth. Similar decrease ofoxygen isotopic ratios with depth has been observed else-where (Lawrence, this volume), attributed to diageneticreaction with silicate rocks. This relationship would seemto render it highly unlikely that long-distance underseatransport of brines could have taken place without changingδD/δO1 8 relationships.

Diffusion and Migration of Ions and Gases

In order to measure the capacity of the sediments androcks to permit ionic diffusion through their intercon-nected pore spaces, special electrical resistivity measure-ments were undertaken at Sites 225, 227, 228, and 229 (seesite reports and Manheim, this volume, GeochemicalMethods). Electrical resistivity of pore fluids had been usedsuccessfully on Leg 22 for measurement of diffusionproperties of unconsolidated sediments, and these proper-ties were regarded as particularly important to theunderstanding of intrastratal migration of dissolved speciesin the Red Sea geochemical systems. It has been shownpreviously that "formation factor" or the ratio of wholerock resistivity to resistivity of the entrapped pore fluidsalone is inversely proportional to the diffusion constantfor ions and molecules moving within interconnected porespaces in rocks, provided that clays or other solid particlesdo not carry an appreciable proportion of the currentthrough the bulk rock. At the same time it has also beenlong known that formation factor, (F) varies with porosityaccording to the relationship F = a/Φn where a is a constant

22-

20-

18-

16-

14-

12-

10-

8-

6-

4-

+2-

0-

-2-

-4

EVAPORITES

"HOT BRINES"ATLANTIS II DEEP(CRAIG, 1969)

282 m

0 20 40 60 80 100 2 00

350 m

300

S u/oo

Figure 6. Distribution of interstitial δ D/H (SMOW) withdepth, Sites 225, 227, and 228. Arrows refer tosalinity-δ D/H relationships applicable to normal RedSea waters (Craig, 1969).

+2

o ( / o o SMOWJ

Figure 7. Deuterium-oxygen isotopic diagram for waters ofRed Sea region (modified from Craig, 1969) withinterstitial water data for Sites 225 and 228 fromFriedman (this volume) and Lawrence (this volume).

992

Page 19: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

close to 1, 0 is porosity, and n is a factor depending largelyon the cation exchange capacity (clay content) of the rock.The exponent may vary from less than 2 for "clean"particle systems such as sands or limestones to 5 formontmorillonitic clays or shales.

Plots of F against depth and porosity are given for Sites225, 227, 228, and 229 in Figures 8 through 11. These bearout in a general way the decreasing diffusive permeabilityof the post-Miocene carbonate-rich sediments with depth.Values of F range from between 2 and 3 in the uppermostsediments to more than 10 at depth. Taken at face value,this means that the uppermost sediments permitteddiffusion of the dominant dissolved species (i.e., NaCl) at arate of about 1/2 to 1/3 that in free solution, whereas withdepth the rate decreased to less than 1/10. In the cementedpalagonites (rocks, which though still having considerableporosity, had become rigidly cemented and ringing hard tothe hammer) at Site 229, diffusive permeability fell tcabout l/30th of free solution values.

The diffusive properties of rocks at Site 229 areinteresting because of gas occurrence strong enough tocause abandonment of the site. Mclver has pointed out thatthe gas contents in the canned samples were among thehighest found in the DSDP program to date. As much as16 percent methane (by volume) was found at 61meters, total values falling off with depth. We shouldrecognize, however, that the total amounts recovered in thecans are probably subject to artifacts in the canning processand certainly do not represent original gas content, becauseof the opportunity for loss of gas during raising of cores.Moreover, on-deck observation of core recovery showed nosignificant diminution of total gas content, judged by theexcess pressure released when removing the core nose andseparation of core sections in their liners while on deck andin storage racks preparatory to core splitting. Thehydrocarbon to bitumen ratios of 0.2 to 0.45 (Mclver, thisvolume) suggest that the petroleum-hydrocarbon genera-tion, though not very far along, has begun, even though theprobable in situ temperatures have not reached valuesnormally regarded as necessary for these processes.

Gas chromatographic (shipboard) analysis of theexsolved gas showed a continuing increase in theethane/me thane ratio with depth. Although the palagonitelayer is demonstrably less permeable to diffusion than over-and underlying layers, it equally clearly does not seem toform an absolute barrier to gas permeation (Figure 12). Incontrast, CO2 increase with depth is interrupted by thepalagonite, probably owing to uptake of CO2 o n

weathering silicate rocks (glassy tuffs) to form CaC03Below the palagonite the CO2 content of the gas againincreases irregularly with depth. Regrettably, these datarefer only to the gas recovered with the cores, and we haveno way of knowing what the total original gasconcentrations were, nor whether distortion of theproportions of gases occurred as a result of preferential lossof some of the constituents. Construction of a pressure corebarrel to capture sediments under in situ pressures has beenauthorized by the DSDP and should permit a majorbreakthrough in knowledge about gas distribution andmigration in marine sediments.

As one might expect from their low porosity, rockswithin the evaporite sequences showed correspondingly

RED SEA GEOCHEMISTRY

SITE 225FORMAT/ON FACTOR [FJ

I0 100 1000 POROSITY (P)I0 (%) 100

200

holite(F>>6000)

Figure 8. Distribution of formation factor and porositywith depth at Site 225.

SITE 227

FORMATION FACTOR (FJI0 100

.1

1000 POROSITY (PJI I0 % 100

100

Figure 9. Distribution of formation factor and porositywith depth at Site 227.

SITE 228FORMATION FACTOR (FJ

I0 100 1000 POROSITY (PJI0 (%) 100

300L

CARBON OOZE,MΔRL AND CLAYS

Figure 10. Distribution of formation factor and porositywith depth at Site 228.

993

Page 20: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

F. MANHEIM

SITE 229.229A

FORMATION FACTOR (FJI I0 100

100

300

POROSITY (PJI0 % 100

CARBONATE CEMENTED^TlslLTSTONE-PALAGONITE

100

3 0 0

Figure 11. Distribution of formation factor and porositywith depth at Hole 229.

F (Formation Factor) j CO2 P^ok height (arbitrary units)

1 10 100 1000

Hard polaqonite tuffite

Cemented siltstone

C 2 H 6

CHΔ / C~Hc•area ratio

Figure 12. Distribution of formation factor and core gaswith depth at Hole 229A.

high formation factors or low diffusion coefficientscalculated from these. Nonanhydritic shales within thesestrata did not display any significant change in diffusionproperties from shales overlying the evaporites. This mightbe expected in one sense, since these shales are unlikely tolose any further moisture (porosity) by consolidation proc-esses. On the other hand, cementation processes that mighthave been enhanced in the strong brine zone have not mate-rially changed permeability of the shales. Diffusive permea-bility within the anhydrites appears to be proportional totheir content of clay minerals. It ranges from somewhat lessthan 100 to several hundred times less than free in solution.On the other hand, halite specimens tested were off scaleon our resistivity meter, corresponding to F values greaterthan 6000. This corresponds to literature data showing thathalites are among the most impermeable rocks known,exceeding granites and crystalline basalts in this respect.

The above discussion has been restricted to theproperties of the rocks as measured onboard ship. We mustrecognize that the increase in temperature with depth in thehole will increase the diffusion constants by factors of upto two or more depending on the temperature reached. Toa less predictable degree, the release of pressure on thecores will also change the resistive properties.

Assuming a steady-state salt diffusion profile has beenattained by the evaporitic sites (Figure 5) and utilizing theformation factor data to get limits on diffusion coefficients,we may calculate rough limits on the volume of dissolvedsalts removed from the evaporite layers by leaching andupward diffusion. At Site 225, for example, we take agradient from 180 meters to 210 meters as 100 g/kg totalsalt/30 meters or about 4 × I0"5 g/cm3 per cm of depth.Formation factors at a presumed temperature of about50°C are about 5 and 120 for shale and anhydrite,respectively. Then, given NaCl diffusion coefficient ofabout 2 × I0"5 cm2/sec for free solution, the appropriatediffusion coefficients are 4 X I0"6 and 1.7 × I0"7 cm2/secfor shale and anhydrite, respectively. Then, / = -kdc/dx,where / is the flux of salt in g/sec across a 1-cm2 area, k isthe diffusion coefficient for NaCl, and dcfdx is theconcentration gradient with depth. The fluxes are 4 × I0"4

and 1.7 × I0"5 g/yr for shale and anhydrite, respectively.This means that at the present time from 0.1 to 1.8 metersof rock salt are leached away from the evaporite boundaryper million years at Site 225. These magnitudes are perhapslarge enough to give rise to some of the crinkly beddedanhydrites with thin shale interbeds in the uppermost partof the evaporite sequences; that is, some of the anhydritebeds are lag remnants of anhydrite dispersed in largeramounts of halite. The total amount of interstitial salt in a1-cm2 column above a depth of 180 meters at Site 225 isapproximately 540 g. At the fluxes calculated above, itwould require about 1 million years to replace all theinterstitial salt in the column for shales at the 180-210meter zone, whereas it would require 31 million years todo so if anhydrites characterized this boundary. If thelatter were so, it would mean that the assumption ofsteady state is not correct; that is, that the amount ofinterstitial salt in the column is still increasing. In actualfact, the strata between 180 and 210 meters are intermixedanhydrites and shales, and the shape of the interstitial saltgradient does suggest that steady state (closer approach to astraight-line gradient, somewhat steeper with depth) has notbeen attained. A more straight-line (steady-state-like)gradient is evident for Sites 227 and 228 (Figure 5).

ORIGIN OF THE EVAPORITES

Overwhelming evidence points to a shallow-waterevaporite pan origin of the evaporites penetrated at the Leg23B sites.

1) Both petrographic evidence and high Sr/Ca values(Staffers and Kuhn, this volume) indicate that much of theanhydrite is primary. Stability considerations for gypsum-anhydrite indicate that the anhydrite must have beendeposited in very warm conditions such as those commonlyfound in shallow evaporating pans.

2) The bromine indicator of evaporite environment(Kuhn, 1953) yields water depths of a few meters depth(Stoffers and Kuhn, this volume).

3) Nodules and oncolites in the anhydritic layersresemble remnants of stromatolitic deposits typical of ashallow-water brine (Stoffers and Kuhn, this volume).

4) The chemical nature of the evaporites, and theinterstitial waters associated with them, shows greatvariation from site to site. This is typical for sabkha-like

994

Page 21: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

RED SEA GEOCHEMISTRY

deposits. Moreover, late-stage evaporites or their leachatesare present (chlorides of the tachyhydrite-bischofite series).These deposits cannot readily be preserved underdeep-water conditions (Braitsch, 1971).

5) Hydrogen and oxygen isotopic ratios of theinterstitial waters, especially those trapped below substan-tial halite beds, and hence protected from mixing withoverlying waters, show strong indications of meteoric waterinfluence. This influence also extends to carbon and oxygenisotopes in carbonate phases that have equilibrated orpartially equilibrated in brines within or near the evaporiticstrata. A deep-water evaporite basin the size of the Red Seawould necessarily contain strongly evaporated (seawater)brines having correspondingly heavier, not lighter, isotopiccontents.

Stoffers and Kuhn (this volume) have suggested thatonly the last evaporites deposited in the Red Sea basin mayhave been shallow water in origin. Earlier evaporites in thevery thick series already known from drill holes arepostulated by the above authors to have been laid down ina deep basin such as was proposed by Schmalz (1969). Thisconcept agrees with the views of Hassan and El-Dashlouty(1970), who attributed a deep origin to evaporites in theGulf of Suez by virture of their concentric distribution,with late-stage evaporites in the center. Their dominantevaporite basin occurs in lower-middle Miocene(?) and isopen ended to the Red Sea proper. However, the evaporitesextend into Aquitanian (basal Miocene and possiblyOligocene). Total evaporitic thickness aggregates more than3600 meters.

The deep-water concept conflicts with that of Ross andSchlee (in press) who envisage doming of the Red Sea areain Paleogene time, as evidenced by widespread absence ofOligocene and early Miocene strata in the deeper drill holes.Since the rock salt-evaporite series was preceded by anormal marine Middle Miocene phase, indicated by theplanktonic fauna of the Upper Globigerina group (Ahmed,1972, and references cited), one would presume that adeep basin should have also allowed more complete andthicker early and early-middle Miocene sequences. More-over, thin salt, anhydrite, and gypsum found in boreholeson land in Ethiopia extend continuously to thickerevaporites in the Red Sea (Pratsch, 1973, oral communica-tion). This suggests that the environments of depositionboth in the Red Sea and in present shore areas at the timeof evaporite formation were similar, probably closer to thesagging shallow-water model of Ross and Schlee (in press).

One independent possibility for determination of thedepth of water for earlier evaporites in the Red Seasequence is the vesicle-size technique as elaborated byMoore (1965) and Jones (1969) for submarine basalts.Through the courtesy of J. C. Pratsch and C. Burk (MobilOil Co., Princeton, N. J.) thin-sections of basic igneousrocks from Mobil/Exxon's Red Sea B-l well, in theEthiopian sector of the Red Sea (see figure in Ross andSchlee, in press), were received and examined by G.Thompson (Woods Hole Oceanographic Institution). Theserocks consist of submarine, chloritized basalts havingunusually large quantities of coarse-grained pyroxenes,partly altered to secondary calcite, chlorite, and zeolites.The basalts underlies the evaporites at depths of about2100 to 2800 meters and contain vesicles. The

interpretation of depths from such vesicles is complicated,however, by observations of Aumento (1971) who foundthat vesicles were not only found in shallow igneous rockson the Mid-Atlantic ridge, but also in deep water zones,evidently through the agency of excess volatiles. Thus, untilthese relationships are clarified, basalts cannot be used as adefinitive indication of water depth of extrusion.

In conclusion, the writer regards the origin of depositionof the sequence of evaporites in the Red Sea unsettled onpresent evidence. No doubt a careful examination ofevaporitic rocks at depth from more centrally located drillholes, bearing in mind the lines of exploration outlinedearlier, would clear up this question, and there may bestructural or other lines of argument, not currently knownto the writer, that might help decide the matter.

ORIGIN OF THE HOT METALLIFEROUS BRINES

This treatment will not recapitulate extensive previousdiscussions of the controversial question (e.g., Craig, 1969)but will add the new information and constraints gained inthe course of the Red Sea drillings to provide a briefsynthesis of the genesis of the hot brines. Major facts andconclusions that bear on the genesis follow:

1) As summarized by Craig (1969), the hot brines'water has a hydrogen and oxygen isotopic compositionthat falls in the range of normal Red Sea waters. Thevolume and heat of discharging brines in the central RedSea area, as well as their consistent composition over aperiod of years, indicate that they are issuing frompermeable reservoirs of considerable extent.

2) Strongly saline interstitial waters associated withevaporites at Sites 225, 227, and 228 have chemical andisotopic composition very different from those of the hotbrines. The evaporite-associated waters show evidence ofcontributions of light "meteoric" waters to the fluidstrapped with the evaporites. Moreover, the very poorpermeability of the rocks in the late Miocene evaporiticsequence precludes discharge of even minor amounts ofbrines from them, if similar lithologies characterize theevaporitic zones consistently. "Cold" brine deeps such asthose reported by the Valdivia expeditions (Backer andSchoell, 1972) may owe enhanced salt concentrations toleaching of salt from outcropping salt formations in localbasinal traps, or diffusion from salt beds at shallow depthsbelow the sea floor.

3) During R/V Chain Cruise 100 to the Red Sea in theSpring of 1971, piston cores on the flanks of the hot brinedeep, 300 meters above the present brine zone, yieldedmetal-rich layers (up to 2% Cu) at about 8 meters below thesediment surface (Manheim, in press). Stratigraphicevidence suggested an age of 100,000 years or more. Thesedata imply existence of brine discharge well beforeemplacement of the known metalliferous deposits, whichbegan about 13,000 years ago (Hackett and Bischoff,1973), though perhaps at different discharge points and/orin lower volumes than during the latest period.

4) Whereas the concept by Craig (1969) that the hotbrines formed from normal Red Sea water seems welldocumented, his extraordinary long-distance plumbing, orsubterranean transport of brines from the southern end ofthe Red Sea, up to 900 km to the hot brine deeps is not.

995

Page 22: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

F. MANHEIM

His placing of a presumed "source" at this point dependson the coincidence of isotopic measurements of present-daywaters here with similar values of δD/δO 1 8 in the hotbrines. A subsidiary argument based on Ar and N in thebrines is subject to a number of assumptions and is notentirely clear. The present isotopic studies on interstitialbrines at Sites 225, 227, and 228, show that one shouldexpect influence on interstitial isotopic ratios by diageneticreactions between fluids and host rocks, including bothcarbonates and silicates. The long-distance plumbingconcept postulates virtually complete lack of reactivity ofthe brines and must be rejected as unrealistic (Baturin et al.,1969). Moreover, there is no indication of how normal RedSea water could penetrate subsurface formations in thesouthern end of the Red Sea against density gradients (i.e.,increasingly saline fluids at depth, as has been demonstratedin this study). No hydrodynamic mechanism has beenadduced which would permit such a flow direction, even ifthe unprecedented rapidity of fluid movement over thegreat distance were allowed. Another difficulty in Craig'singenious but improbable plumbing system is that hisgeographic relationships linking salinity and isotopic dataapply only at present, whereas the waters now discharginginto the rift deeps must have entered the system at aminimum, many thousands of years ago. At the earlier datean identical hydrochemical pattern for the Red Sea wouldhave been highly unlikely, particularly in the light ofpaleoenvironmental evidence (Deuser and Degens, 1969;Müliman et al., 1969).

5) Origin of the hot brines and metalliferous sedimentsmust take into account the origin of the leads. Evidently,some contact between the brines and mantle-associatedrocks is needed to provide the lead isotopic signature(Delevaux and Doe, this volume). Basalts recovered in theLeg 23 operations showed little significant evidence ofchemical leaching (Coleman, this volume). However,water-laid volcanic tuffs are known from drill holes throughthe Miocene evaporite sequence (Pratsch, 1973, oralcommunication) and could provide metals within sedimen-tary horizons.

The writer favors the hypothesis that the brines aredischarging from ruptured subevaporite clastic or permeablecarbonate aquifers, possibly corresponding in stratigraphicposition to the Globigerina group of middle Miocene age,Overlying evaporites of late Miocene age formed thenecessary hydrodynamic seal (aquiclude). The connatewaters of these strata (Neumann and Chave, 1965) weremarine waters of relatively normal Red Sea salinityoriginally, but in time became very salty through leachingof soluble evaporitic minerals and ionic diffusion fromoverlying (later-deposited) evaporitic rocks. The strata arepresumably in hydrodynamic continuity from the shallowshelf area of the Red Sea to the deep central rift zone.Thus, when the evaporite layers were breached by thecentral Red Sea rifts, a powerful hydrodynamic gradientwas mobilized and enabled discharge of the heavier brinesinto the bottom of the deeps. Figure 13 illustrates this inschematic fashion for a section from the Saudi Arabianshelf near Jiddah to the Atlantis II Deep.

The contrast in brine density between the Atlantis II andDiscovery deeps brines and normal Red Sea water is about

0.10 Sp.G. units. The actual configuration of pre-evaporite(less likely, intraevaporite) aquifers is not known, but forthe purpose of illustration, an extension of theheavy-brine-containing aquifer 500 meters above thedischarge point (but below the evaporitic cover strata)would correspond to a normal artesian head of 50 meters(i.e., elevation of aquifer recharge point 50 meters abovesea level). Such a head would be more than adequate tomaintain powerful artesian flow, including erosion ofsubmarine strata by sapping. Actual heads might be stillgreater, for elevation differences might be as large as 2000meters.

A noteworthy fact about the known hot brine deeps isthat they are located at a minimum distance between riftzone and both shores for the southern half of the Red Sea;that is, the hydraulic gradient for subevaporite aquifers is amaximum here. No supraevaporite, subsurface flow is

20°-

15°-

D H O T BRINE DEEPS(SITES 225, 226,227)

^DISCHARGE FROMFLANK (PROPOSED)

<*JLONG DISTANCE* MIGRATION (CRAIG, 1969)

HYPOTHESIS

ETHIOPIA

~~r35C

Depth (m) A.

π DEEP

40c

SEA LEVEL JIDDAH

1000

2000-

3000 -

4000

rift zone and basalticintrusives

M. Mioceneelastics andcarbonate

40

Km

Figure 13. Schematic diagram showing proposed brinedischarge path from flanks of Red Sea to Atlantis IIDeep. Crosses with numbers refer to drill sites.

996

Page 23: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

RED SEA GEOCHEMISTRY

regarded as likely, because sealing beds adequate to confinefluid flow and maintain a hydraulic head are absent. Hereagain, the factors favoring a flank source of dischargingbrines make it easy to explain isolated discharge points incentral positions in the Red Sea. In a long-distance travelfrom the southern terminus of the Red Sea, one must askthe question of how a pressure gradient greater than theflanking system could be developed, considering the"frictional" losses with distance, and why hot brines arenot found in intervening parts of the central rift system.Careful study of this area by the Valdivia expedition haslocated only several "cold" brine deeps (e.g., Sudan, Suakindeeps) that show no evidence of forcible discharge of deep(and hot) brines (Backer and Schoell, 1972).

As mentioned earlier, the "normal" character ofentrapped seawater in the presumed aquifer strata wouldaccount for its isotopic character, somewhat modified byexchange with evaporites. Mg would be reduced to lowlevels by reaction with carbonates to form dolomite, andsoluble Ca, whereas the sulfate content would be controlledby equilibrium with solid anhydrite. Minor and traceconstituents in the brine would be leached from thesediments and possibly from pyroclastics by the high ionicstrength brines. Temperatures can easily be accounted forby normal thermal gradients. For example, if we assume thedischarging brines originally resided at a depth of 2000 metersbelow the sediment surface, a gradient of only 4°C/meters(low for Red Sea conditions) would yield an in situ temper-ature of 103°C, assuming surface sediment temperature of23°C. We may presume some cooling took place duringtransit from original positions to the discharge point; on theother hand, heat could also have been contributed by newlyemplaced basalts, but would not necessarily be required.

During past evaporative epochs associated with eustaticfalls in world sea levels, the relative effect of hydraulicheads would probably be increased, since there is evidencethat although sea levels were probably lowered andsalinities increased, the latter remained far from concentra-tions needed to precipitate gypsum (Manheim et al., thisvolume). Were a hydraulic level above sea level present atthe upper end of the continuous aquifer system beneath(possibly also between) the poorly permeable evaporiticcovering strata, hydraulic heads would be greatly enhanced,perhaps even doubled. One might suggest that past "hotbrine" discharge episodes were triggered during suchregressions of sea level and should thus correlate with thelithified zones discussed by Milliman et al. (1969). If thispicture is correct, we may be on a declining end of adischarge cycle, even though short-term fluctuations indischarge may temporarily alter brine deep temperaturesand volumes. Alternatively, breaching of a larger fluid pathmight compensate for a lowered head and increase fluidflow over the long term.

The above hypothesizes large-scale flow and hence metaltransport from nonevaporitic strata, presumably below theevaporites, conducted through evaporites in areas ofseparation or disruption, as in the rift zone. The hypothesislends itself to testing especially in boreholes that penetratethe evaporite sequence, where fluids and evaporite samplescan be recovered for investigation. Recharge of the fossil(M. Miocene) fluids by seepage from the present Red Sea is

probable, but the recharging fluids entering at outcrop orsubcrop positions on the flanks may or may not havereached the brine deeps up to the present.

REFERENCES

Ahmed, S. S., 1972. Geology and petroleum prospects ofeastern Red Sea: Am. Assoc. Petrol. Geol. Bull., v. 56, p.707.

Aumento, F., 1971. Vesicularity of mid-oceanic pillowlavas. Canadian J. Earth Sci., v. 8, p. 1315.

Backer, H. and Schoell, M., 1972. New deeps with brinesand metalliferous sediments in the Red Sea: Nature, v.240, p. 153.

Baturin, G. N., Kochenov, A. V., and Trimonis, Ye. S.,1969. Composition and origin of the iron ore sedimentsand hot brines in the Red Sea: Oceanology (Okeano-logiya in transl.), v. 9, p. 360.

Bischoff, J., 1969. Red Sea geothermal brine deposits: theirmineralogy, chemistry, and genesis. In Degens, E. T. andRoss, D. A. (Eds.), Hot Brines and recent heavy metaldeposits in the Red Sea: New York (Springer-Verlag), p.368401.

Braitsch, O., 1971. Salt deposits, their origin andcomposition: New York-Heidelberg (Springer-Verlag).

Brewer, P. G. and Spencer, D. W., 1969. A note on thechemical composition of the Red Sea Brines. In Degens,E. T. and Ross, D. A. (Eds.), Hot brines and recentheavy metal deposits in the Red Sea: New York(Springer Verlag), p. 174.

Brewer, P. G., Riley, J. P., and Culkin, F., 1965. Thechemical composition of the hot salty water from thebottom of the Red Sea: Deep-Sea Res., v. 12, p. 497.

Brewer, P. G., Wilson, T. R. S., Murray, J. W., Munns,R. G., and Densmore, C. D., 1971. Recent hydrographicobservations on the Red Sea brines and a markedtemperature increase: Nature, v. 231, p. 37.

Brooks, R. R., Kaplan, I. R., and Peterson, M. N. A., 1969.Trace element composition of Red Sea geothermal brineand interstitial water. In Degens, E. T. and Ross, D. A.(Eds.), Hot brines and recent heavy metal deposits in theRed Sea: New York (Springer-Verlag), p. 180.

Craig, H., 1969. Geochemistry and origin of the Red Seabrines. In Degens, E. T. and Ross, D. A. (Eds.), Hotbrines and recent heavy mineral deposits in the Red Sea:New York (Springer-Verlag), p. 208.

Culkin, F., 1965. The major constituents of seawater. InRiley, J. P. and Shirrow, G. (Eds.), Chemical oceanog-raphy, vol. 1: London (Academic Press), p. 121-161.

D'Ans, J., 1961. Uber die Bildungsmöglichkeiten desTachhydrits in Kalisalzlagerstàtten: Kali and Steinsalz, v.3, p. 119.

Delevaux, M. H., Doe, B. R., and Brown, G. F., 1967.Preliminary lead isotope investigations of brine from theRed Sea, galena from the Kingdom of Saudi Arabia, andgalena from United Arab Republic (Egypt): EarthPlanet. Sci. Lett., v. 3, p. 139.

Deuser, W. G. and Degens, E. T., 1969 O1 8 /O1 6 andCl3/çl2 ratios of fossils from the hot-brine deep area ofthe central Red Sea. In Degens, E. T. and Ross, D. A.(Eds.), Hot brines and recent heavy metal deposits in theRed Sea: New York (Springer-Verlag), p. 336.

Faure, G. and Jones, L. M., 1969. Anomalous strontium inthe Red Sea brines. In Degens, E. T. and Ross, D. A.(Eds.), Hot brines and recent heavy mineral deposits inthe Red Sea: New York (Springer-Verlag), p. 243.

997

Page 24: 38. RED SEA GEOCHEMISTRY - Deep Sea Drilling Project

F. MANHEIM

Hackett, J. P. and Bischoff, J. L., 1973. New data on thestratigraphy, extent and geologic history of the Red Seageothermal deposits: Econ. Geol., v. 68, p. 553-504.

Harder, H., 1970. Boron content of sediments as a tool infacies analysis: Sediment. Geol., v. 4, p. 153.

Hartmann, M. and Nielsen, H., 1966. Sulfur isotopes in thehot brine and sediment of Atlantis II Deep (Red Sea):Marine Geol., v. 4, p. 305.

Hassan, F. and El-Dashlouty, S., 1970. Miocene evaporitesof Gulf of Suez region and their significance: Am. Assoc.Petrol. Geol. Bull., v. 54, p. 1686.

Hendricks, R. L., Reisbick, F. B., Mahaffey, F. J., Roberts,D. B., and Peterson, M. N. A., 1969. Chemical composi-tion of sediments and interstitial brines from theAtlantis II, Discovery and Chain Eeeps. In Degens, E. T.and Ross, D. A. (Eds.), Hot brines and recent heavymetal deposits in the Red Sea: New York (Springer-Verlag), p. 407.

Hodgson, G. W., 1965. Vanadium, nickel, and iron tracemetals in crude oils of western Canada: Am. Assoc.Petrol. Geol. Bull., v. 49, p. 2537.

Holser, W. T., 1966. Diagenetic polyhalite in recent saltfrom Baja California: Am. Mineralogist, v. 51, p. 99.

Hsü, D. J., Cita, M. B., and Ryan, W. B. F., 1973. Theorigin of the Mediterranean evaporites. In Ryan,W. B. F., Hsü, K. J. et al., Initial Reports of the DeepSea Drilling Project, Volume XIII: Washington (U. S.Government Printing Office), p. 1203.

Jones, J. G., 1969. Pillow lavas as depth indicators: Am. J.Sci.,v. 267, p. 181.

Kaplan, I. R., Sweeney, R. E., and Nissenbaum, A., 1969.Sulfur isotope studies on Red Sea geothermal brines andsediments: In Degens, E. T. and Ross, D. A. (Eds.), Hotbrines and recent heavy metal deposits in the Red Sea:New York (Springer-Verlag), p. 474.

Korolev, D. F., 1958. The role of iron sulphides in theaccumulation of molybdenum in sedimentary rocks ofthe reduced zone: Geokhimiya, v. 4, p. 452 (EnglishEd.).

Ku, T-L., 1969. Uranium series isotopes in sediments fromthe Red Sea hot-brine area. In Degens, E. T. and Ross,D. A. (Eds.), Hot brines and recent heavy metal depositsin the Red Sea: New York (Springer-Verlag), p. 512.

Kuhn, R., 1953. Tiefenberechnung des Zechsteinmeeresnach dem Bromgehalt der Salze: Z. deutsche geol.Gesellschaft, v. 105, p. 646.

Landergren, S. and Manheim, F. T., 1962. Uber dieAbhàngigkeit der Verteilung von Schwermetallen vonder Fazies: Forschr. Geol. Rheinl. u. Westfalen, v. 10, p.173.

Manheim, F. T., in press. Geochemical studies in the RedSea. Shipboard analysis of rock and sediment samples:Saudi Arabia government publication, Kihlsgaard, T. H.(Ed.).

, in press. Interstitial waters of marine sediments.In Riley, J. P. and Chester, R. (Eds.), ChemicalOceanography: London (Academic Press).

Manheim, F. T., Sayles, F. L., and Waterman, L. S., 1971.Interstitial waters studies on small core samples, DeepSea Drilling Project, Leg 8. In Tracey, J. I., Jr., et al.,Initial Reports of the Deep Sea Drilling Project, Volume8: (U. S. Government Printing Office), p. 857.

Manheim, F. T. and Waterman, L. S.,in press. Diffusimetry(diffusion constant estimation) on sediment cores byresistivity probe. In von der Borch, C. C , Sclater, J. G.,et al., Initial Reports of the Deep Sea Drilling Project,Volume 22: Washington (U. S. Government PrintingOffice).

Miller, A. R., Densmore, D. C , Degens, E. T., Hathaway,J. C , Manheim, F. T., McFarlin, P. F., Pocklington, R.,and Jokela, A., 1966. Hot brines and recent irondeposits in deeps of the Red Sea: Geochim. Cosmochim.Acta,v. 30, p. 341.

Milliman, J. D., Ross, D. A. and Ku, T-L., 1969.Precipitation and lithification of deep-sea carbonates inthe Red Sea: J. Sediment. Petrol., v. 39, p. 724-736.

Moore, J. G., 1965. Petrology of deep-sea basalt nearHawaii: Am. J. Sci., v. 263, p. 40-52.

Neumann, A. C , and Chave, K., 1965. Connate originproposed for hot salty bottom water from a Red Seabasin: Nature, v. 206, p. 1346.

Richter-Bernburg, G., 1953. Uber salinare sedimentation:Zeitschrift Deutsch, geol. Gesellschaft, v. 105, p. 593.

Ross, D. A. and Schlee, J. S., in press. Shallow structureand geologic development of the southern Red Sea:Geol. Soc. Am. Bull.

Schmalz, P. R., 1969. Deep water evaporite deposition: agenetic model: Am. Assoc. Petrol. Geol. Bull., v. 53, p.798.

Schoell, M. and Stahl, W., 1972. The carbon isotopiccomposition and the concentration of the dissolvedanorganic carbon in the Atlantis II deep brines/Red Sea(sic): Earth Planet. Sci. Lett., v. 15, p. 206.

Shishkina, D. V. and Bogoyavlenskii, 1970. O solevomsostave glubinnykh vod Krasnogo morga: Geokhimiya, v.5, p. 632.

Valeyev, R. N., Ozol, A. A., and Tikhvinsky, I. N., 1972.Genetic characteristics of the halide-sedimentationaltype of borate deposits: Intern. Geol. Rev., v. 15, p.165.

Vine, J. D., 1966. Element distribution in some shelf andeugeosynclinal black shales: U. S. Geol. Survey Bull., v.1214E,p. 31.

Vinogradov, A. P., 1967. Vvedenie v geokhimiyu okeana(Introduction to the geochemistry of the oceans),Moscow (Nauka).

Wardlaw, N. C , 1972. Unusual marine evaporites with saltsof calcium and magnesium chloride in Cretaceous basinsof Sergipe, Brazil: Econ. Geol., v. 67, p. 156.

Wedepohl, J. H., 1964. Untersuchungen am Kupferschieferin Nordwestdeutschland: Ein Beitrag zur Duetung derGenese bituminòser Sedimente: Geochim. Cosmochim.Acta, v. 28, p. 305-364.

Weiss, R. F., 1969. Dissolved argon, nitrogen and totalcarbonate in the Red Sea brines. In Degens, E. T. andRoss, D. A. (Eds.), Hot brines and recent heavy mineraldeposits in the Red Sea: New York (Springer-Verlag), p.254.

Zaitsev, I. K., 1968. Analysis of main hypotheses forformation of salt water and brine in light of recent data:Int. Geol. Rev., v. 10, p. 1037.

998