14. IRON-RICH BASAL SEDIMENTS FROM THE INDIAN OCEAN:SITE 245, DEEP SEA DRILLING PROJECT

Theodore B. Warner, Naval Research Laboratory, Washington, D. C,and

Joris M. Gieskes, Scripps Institute of Oceanography, La Jolla, California

ABSTRACTBasal beds of black manganiferous sediment (early Paleocene)

were recovered from the southwest Indian Ocean in the southernMadagascar Basin. Thirty samples were selected from the lowest26 meters of core, representing about 3 X 10^ years ofaccumulation and spreading. Concentrations of Ca, Al, Si, Fe,Mn, K, Ti, Sr, Cu, and Zn were examined. These sediments arechemically similar to basal sediments in the Atlantic and Pacificoceans, as well as to those presently being deposited on the EastPacific Rise, but they have a noticeably lower metal content.Over the examined period of sediment deposition, iron andmanganese accumulation rates decreased significantly. Inter-bedded ash layers have a distinctive chemical signature and aresometimes enriched with heavy metals. The ash layers have amuch higher Ti/Al ratio than the surrounding sediment, whichmakes the origin of these ashes somewhat uncertain.

INTRODUCTION

Sediments near the basaltic "basement" that presentlyare far from any active ridge or spreading center and arecovered by hundreds of meters of normal pelagic sedimentsoften resemble active-ridge sediments in many ways. Thishas been shown in the Atlantic and Pacific oceans(Boström, 1970; von der Borch and Rex, 1970; von derBorch et al., 1971; Cronan et al., 1972; Boström et al.,1972a). Such results can best be understood in terms of theconcept of sea floor spreading and, at the same time, offerdirect evidence supporting this hypothesis. To date, similarinformation has not been available from the Indian Ocean.

During the course of drilling in the southern MadagascarBasin northwest of the southwest Indian Ridge, evidence ofbasal iron-rich sediment was found. Holes 245 and 245A,drilled in August 1972 at 31°32.02'S, 52°18.1l'E in 4857meters of water, covered the entire stratigraphic column.The lowest sedimentary unit had many sharply defined ashlayers and a basal bed of black manganiferous clayey nannoooze underlain by a relatively young diabasic sül (Valuer,this volume).

The purpose of this report is to provide a generaldescription and some preliminary chemical analyses of thisiron-rich deposit. Emphasis was placed on obtaining a highsample density (30 samples in the lowest 26 meters ofsediment) in order to give data concerning, on the onehand, the normal variability of composition with time(about 3 × I06 years) and, on the other hand, the kinds oflarge-scale changes caused by different input events. In thisway, it was hoped that the different components could beseparated and their chemical signatures identified, i.e.,changes in the input rates of metal-rich sediments will relate

to the steadiness and duration of ridge crest emanations, ifthis is, in fact, the source at this location (Natland, 1973).

Long cores reaching to basement offer a particularlyattractive way to study both variations in ridge activitywith time and, also, changes in the influence of ridgeemanations with distance from the ridge. Samples taken atvarious places up the core should closely resemble a suite ofsurface cores taken at increasing distances from the ridge,such as those examined by Boström and Peterson (1966),Bender et al. (1971), Boström et al. (1972b), and Piper(1973). The deep samples differ from such a surface suiteby being older—they represent the conditions that existednear the ridge previously, and, thus, differences betweenold and recent crestal sediments will be due to a change inthe kind of crestal activity, to increasing distance, or todiagenesis. Long cores are also a very efficient way toobtain material for examination, for one core can besampled almost as closely spaced as desired, yielding theequivalent of a very large number of surface cores.Moreover, samples can be taken at particularly interestinglocations.

The manganiferous sediments of Hole 245 are among theoldest yet examined, being at least early Paleocene (Danian)in age, and thus serve as a reference point for the kinds ofcompositional changes that have occurred in suchsediments, as a function of time, since deposition. Datafrom even older Upper Cretaceous (Campanian) sedimentsfrom the Atlantic have been reported by Boström et al.(1972a).

METHODS

Each sample (about 2 g) was dried at 105°C, ground topass 200 mesh, pressed into pellets containing 0.75-g

395

T. B. WARNER, J. M. GIESKES

sediment mixed with 0.25g chromatographic cellulose, andexamined, using X-ray fluorescence, for Ca, Al, Si, Fe, Mn,K, Ti, Sr, Cu, and Zn. Al, Si, and K were determined with aconventional unit and the other elements by X-rayfluorescence, using energy spectrometry for detection. Inaddition, calcium carbonate was determined by the thermaldecomposition method (LECO) (Figure 1).

No attempt was made to remove sea salts from thesediments. The contribution of such salts to the totalcomposition is negligible, and it was felt that the risk ofchanging composition, because of possible leaching, was toogreat. The amount of any element present due to porefluids can be estimated from extrapolated pore water dataof Gieskes (this volume) and the water content. Watercontent is described elsewhere in this volume. In brief, itvaried linearly from 15 percent at 370 meters to 40 percentat 70 meters and from there up to about 65 percent at thesediment surface on a steeper gradient.

The precision of the data, given as the relative standarddeviation of an individual determination (Table 1), wasdetermined on separately prepared replicate samples andincludes errors due to sample preparation and sampleinhomogeneity, as well as measurement errors. Theaccuracy, given as estimated limit of relative error (Table1), represents the sum of estimated bias due to possibledifferences between the sample matrix and the standardsplus three times the relative standard deviation. Thisrepresents maximum estimated error from the true value.Relative errors from sample to sample will be defined betterby the precision, since matrix differences are quite smallbetween nearly all samples.

Bias was estimated by using as standards a series ofsamples of USGS diabase W-l diluted with 0 to 80 percentCaCO3 (J. T. Baker, Utrex grade) and spiked with 2 percentMn. A similar set of basalt (BCR-1) and a granite (G-2), pluscalcium carbonate, were measured as unknowns, anddeviations from expected values were determined.Carbonate-rich diabase and basalt spanned the composi-tional ranges probable in these samples, while the graniterepresented a "worst-case" variation unlikely to beencountered and hence defined as outside error limits dueto matrix variations.

Sediment Description

Hole 245 was drilled and intermittently cored to a depthof 396.5 meters, which included 7.5 meters of basalt. Hole245A was drilled to sample the 20 to 150 meter interval.Four major lithologic units are summarized here anddescribed in detail elsewhere in this volume (Chapter 7).Unit I (63 m), mostly of middle Miocene and late Eoceneage, consists of brown silt-rich clay and brown silty clay.Unit II (63 m), of late and middle Eocene age, is a brownnanno-bearing clay with interbeds of nanno ooze and siltyclay. Unit III (83 m), late early Eocene in age, is composedof orange to pinkish-gray nanno ooze. Unit IV (180 m),early Eocene and Paleocene in age, is a nanno chalk withsilicified chalk and chert beds in the upper part, devitrifiedvolcanic ash layers in the lower part, and a basal bed ofblack manganiferous clayey nanno ooze. Leclaire (thisvolume) notes that these Fe/Mn deposits appear not ascoatings but as amorphous agglomerates of very fine grains,

suggesting very fast formation of these oxides. A diabasicsill of an age of 27 ±3 m.y. occurs at 389 meters. Nosediment was recovered between 382 and 389 meters, butthe paleontological age of the sediment at 382 meters wasat least early Paleocene (Danian, ca. 63 m.y.). Themagnetically predicted age of basement was LateCretaceous, ca. 68 m.y. (Schlich, this volume). Thus, the sillintersects the sediment column not more than 50 metersabove true basement, assuming a uniform sedimentationrate of 1 cm/lOOOy.

This sediment column fits a model (Chapter 7, thisvolume) which is consistent with the sea floor spreadinghypothesis, involving lateral spreading and subsidencebelow the carbonate compensation depth.

RESULTS

The compositional changes observed (Tables 1 and 2,Figure 2) cover the sediment accumulated during a periodof 3 m.y. (approximately), based on an estimatedsedimentation rate of 10 m/m.y. (Chapter 7, this volume),lines have been drawn between samples in Figure 2 only toguide the eye, and they do not imply knowledge ofintervening layers.

If the XRF data are converted to CaCC»3 (assuming all Cato be present as calcium carbonate) and these data areplotted against the LECO calcium carbonate values, analmost linear relationship is obtained (Figure 1). The ratioof the two CaCO3 values is not equal to unity, the XRFyielding higher values in almost all cases. (Note that theaccuracy is about 2% for both methods.) The difference,i.e., the higher XRF values, is due to the calcium associatedwith the noncarbonate fraction of the sediment. Theexperimental inaccuracies do not allow us to presentmeaningful figures on noncarbonate calcium in the tables,but it appears that this calcium usually constitutes lessthan 3 percent of the noncarbonate fraction. Sayles andBischoff (1973) found similar values in East Pacific Risedeposits.

DISCUSSION

There are three main regions in the cored interval. Thelowest third (Samples 22-29) is enriched in Fe and Mn andcontains 50 to 55 percent CaCO3. The middle thirdincludes four samples taken from visually distinctdevitrified ash layers (Samples 9, 10, 16, and 18). Theupper third (Samples 1-8) has about 70 percent CaCO3 anddistinctly less Fe and Mn thani the lowest third. The entirelayer is predominantly Al-poor and Fe- and Mn-rich,characteristics of sediments influenced by active ridgecrests.

Sample 30 was from the core catcher and does notresemble nearby samples, but seems more like samples fromthe upper third. It is probable that this is a contaminatedsample, and in the discussion following it has not beenconsidered as part of the lowest third. It has been observedin previous studies that core catcher samples are oftennonrepresentative.

Mineralogy

Preliminary smear slides (Chapter 7, this volume),summarized in Tables 3 and 4, indicate that at least three of

396

IRON-RICH BASAL SEDIMENTS FROM THE INDIAN OCEAN, SITE 245

110

120

I30

I40

% CaCO3

l50

(XRF)

I60

I70

i80 90

Figure 1. Relation between "calcium carbonate" obtained from X-ray fluorescence data (assuming all Ca to be in calciumcarbonate) and calcium carbonate obtained from thermal decomposition measurements (LECOJ.

397

00

SampleNo.

12345

6789 (ash)

10 (ash)

1112131415

16 (ash)1718 (ash)1920

2122232415

2627282930

Basement,

Core, Section,Interval(in cm)

13-5, 19-2013-5,149-15013-6, 104-10514-1, 118-11914-2, 109-110

14-3, 77-7814-4, 64-6514-5,58-5914-6, 3-414-6, 65-66

14-6,97-9814, CC15-1,113-11415-2, 1-215-2,49-50

15-2,84-8515-2, 98-9915-2,135-13615-2,149-15015-3, 3940

15-3,80-8115-3, 145-14615, CC16-1,11-1216-1,56-57

16-1,56-5816-1,83-8416-1, 129-13016-1, 143-14416, CC

basalt

Relative standard deviation (%)

TABLE 1Chemical Composition of Sediments Obtained at Site 245 in the Indian Ocean (on a Whole-Sediment Basis)

Depthin Core

(m)

356.2357.5358.5360.2361.6

362.8364.1365.6366.5367.2

367.5368.0369.1369.5370.0

370.3370.5370.9371.0371.4

371.8372.5375.5380.1380.6

380.6380.8381.3381.4381.8

389.0

Estimated limit of error, relative (%)

Ca(Wt %)

30.827.630.920.429.8

29.624.630.8

3.64.3

29.522.827.826.326.6

14.226.616.327.326.7

27.723.022.222.920.1

18.720.922.017.328.6

1.44

Al(Wt %)

1.001.300.832.260.98

1.021.371.077.156.83

1.111.381.291.401.34

4.271.472.181.431.40

1.191.941.831.431.68

2.191.712.052.861.22

2.0

12

Si(Wt.%)

4.008.965.43

11.906.28

6.408.865.06

26.0025.58

6.709.507.768.858.21

18.088.227.958.188.18

7.4411.1011.259.40

11.60

14.1711.0512.2616.037.07

1.5

10

Fe(Wt %)

2.072.861.924.972.42

2.462.581.884.784.17

2.224.652.462.803.01

4.032.842.022.902.70

2.624.924.574.565.83

6.205.194.935.892.36

2.3

10

Mn(Wt %)

0.510.660.521.100.62

0.600.550.550.080.24

0.521.430.580.650.70

0.680.667.70.760.96

0.741.361.411.411.69

1.721.701.341.580.66

0.6

10

K(Wt %)

0.450.650.340.900.47

0.500.660.430.640.45

0.450.810.520.630.58

0.440.490.630.530.47

0.450.770.780.750.82

1.080.740.750.700.52

0.9

15

Ti(Wt %)

0.110.170.120.250.12

0.140.230.090.520.44

0.130.200.130.170.15

0.250.190.690.160.22

0.200.230.230.210.23

0.300.260.210.300.15

4.2

20

Sr(ppm)

530500490380480

470400450250260

430540420390410

360410800390390

390440490490530

530550510560420

0.9

30

Cu(ppm)

4090_

100-

__

110170

1011040

10060

110308040

< I O

< I O70

120100110

80100709040

11

50

Zn(ppm)

7090_80-

___

240240

50120909080

16080808060

4080

140110140

11013090

13050

8

50

WA

RN

wfa•—

g

o

w

Determined on separately prepared replicate samples randomly interspersed in the order of analysis; includes errors in sample preparation as well as measurement. Sample 11 wasprepared in duplicate, Sample 14 in triplicate, and Sample 26 in quadruplicate; the compositions given are the means of the replicated measurements.

TABLE 2Chemical Composition of Sediments Obtained at Site 245 in the Indian Ocean (Carbonate-Free Basis) and Selected Ratios on a Whole-Sediment Basis

aAssuming all Ca present as CaCO.,.

oo

CaCO3% S i *0 50 100 20 30

385-

390- BASALT

A l *3 5 7 5 7 9 11 0 2 4 0 1 2 3

\

,- i*

355-

360-

365-

370-

375-

380-

385-

390-

TiX0.4 0,6

1 J 1 l _ _

j\

;"J

J

•

BASALT

Sr ppm0 1000 2000

• —

/

\

×/*

\

5A

!

1JL

Cu ppm100 200 300

* ^

J1J

J• — —

t-

/

• — . .

\

IJ

Zn ppm100 200 300

/

\

\

*

\/

\

-~"\•'\\//

Sr/Ca*1010 20 30

*\/*

\

/

I

I

.*-

-

_ ^60

E==49

\

•

A l / K1 2 3

\

J\\

.. *

/

\

>-

4

~~—•-π.215.3

.

Al/Ti5 10 15

100 Al

20 30

100 Mn 100 Fe(Al+Mn+Fe) (Al+Mn+Fe)

0 10 20 40 50 60S i 0 2 / A l ? 0 3

4 6 8

V

Figure 2. iSiYe 245: Sediment composition on a carbonate-free basis and selected ratios on a whole-sediment basis.

IRON-RICH BASAL SEDIMENTS FROM THE INDIAN OCEAN, SITE 245

TABLE 3Smear Slide Studies of Samples from High-Ash Layers

Sample 9 (14-6, 3-4)ClayOpaquesCalcareous nannos

Smear 64, 6-6 (devitrified volcanic ash)70%20 (Fe-Mn oxides)10

Sample 16 (15-2, 84-85) Smear 2-84 (devitrified volcanic ash)Clay 90%Forams 3Calcareous nannos 5Micro crystalline carbonate 3Rhombs 1

TABLE 4X-Ray Studies of Samples from High-Ash Layers

Calcite Quartz Plagioclase Montmorillonite

Sample 16 (15-2,Bulk2-20µ<2µ

Sample 10 (14-6,Bulk2-20µ< 2 µ

84-85);25.1

—-

65-66);--

-

X-ray 2-830.93.7-

X-ray 6-641.20.7—

—5.2-

--

-

74.091.0

100.0

98.899.3

100.0

Note: Bulk X-ray studies (after removal of CaC03) on nonashlayers gave the following qualitative data:

Sample 5 (14-2, 100-110); montmorillonite, palygorskite, mica,clinoptillolite.

Sample 13 (15-1, 113-114); montmorillonite.

Sample 27 (16-1, 83-84); montmorillonite.

the four high-ash samples were mostly montmorillonite. Inthe surrounding sediments, montmorillonite is also the dom-inant mineral in the noncarbonate fraction, though somequartz, Plagioclase, clinoptilolite, and palygorskite can becontributors. More detailed mineralogical studies on allsamples have been planned.

Ash Layers

The ash layers show the presence of devitrified glassshards and generally contain only small amounts of calciumcarbonate. Vallier (this volume) reports many more ashlayers in the 20 meters of core above the segment reportedon in this report.

Most ash layers were fairly thin. Samples 9 and 10 didnot come from one thick ash layer, as might be assumedfrom Figure 2, but from two distinct layers about 7 cmthick, separated by 55 cm of sediment without visual ashlayers. Sample 18 has some somewhat unusualcharacteristics, and we suspect that we may have ground upa manganese nodule in the sample preparation.

The Siθ2/Al2θ3 ratio in these ash layers is about equalto 4, a value which is still twice that expected in puremontmorillonite. Some Plagioclase is also present, as well assome (detrital?) quartz. No precise quantitative data areavailable yet on the mineralogy of these particular samples,but it should be remarked that the S1O2/AI2O3 ratio isoften high for unaltered volcanic ashes, cf. for instance theanalyses for ashes in Hole 178 (Pratt et al., 1973) whichshow high ratios (>4). The Al/Ti ratios in the ash layers

(Samples 9,10, and 16) are significantly higher than in thesediments above and below them. In fact, the Al/Ti ratiosare characteristic of those found in the continental crust.Yet we believe that the source of these ashes is ridge crestyolcanism or associated island arc volcanism. Fe, Mn, and Kvalues are significantly lower (except, of course, in Sample18) in the ash layers than in the surrounding sediments.However, the Al/K ratio is significantly higher and the Al/Feratio significantly lower in the ash layers.

We found it of some interest to investigate the Sr/Caratio in the various samples and found that this ratio is veryhigh in the ash layers. Whereas in the upper third of thecore interval the Sr/Ca ratio is about 18 × I0"4, in the ashlayers values of more than 50 X I0"4 are found. Much ofthis strontium in the upper portion is associated with thecalcium carbonate fraction. In fact, Manheim and Sayles (inpress) estimate that recrystallized carbonate calcite shouldcontain about 0.05 percent strontium (cf., Katz et al.,1972), whereas we obtain an average value of 0.07 percent.Gieskes (this volume) found that the interstitial Sr++

concentrations increased linearly with depth in this hole,which is evidence of a recrystallization process of calciumcarbonate. Some of this mobilized Sr++ may be taken up inthe montmorillonites, thus raising the Sr/Ca ratio in the ashlayers. One could, therefore, consider the Sr/Ca ratio to bea diagnostic parameter of the presence of small admixturesof volcanic ash to the lowest third of the core interval. Theslightly higher Sr/Ca ratio and also the slightly lowerCaCO3 content would tend to support this view. Thedecreased SiO2/Al2θ3 ratios and increased Al/Ti ratio inSamples 24-29 would also indicate higher volcanic ashcontributions to the sediment.

Other Sediments

The upper and the lower third of the cored interval ischaracterized by uniform contents in Si, Al, K, and a fairlyconstant Al/Ti ratio. The latter ratio is similar to thosefound in the surface sediments in this same geographic area(Boström et al., 1972b). This is rather interesting becausethe nature of the clay mineral changes considerably in thishole (Matti et al., this volume). Boström et al. (1972b)noted that the Al/Ti ratio shows a regional distributionrelated not to oceanic ridges but, rather, to latitude.

The SiO2/Al2θ3 ratio of about 7 is quite high, whichdoes suggest the presence of appreciable amounts ofamorphous silica (no opaline skeletal material wasobserved) or of low aluminum silicates (e.g., palygorskite orNa-feldspar). More detailed mineralogical work has beenplanned.

Of interest is the rather low Al/K ratio in thesesediments. Such low ratios are often observed inactive-ridge sediments (Sayles and Bischoff, 1973). Onlysmall amounts of K-feldspar have been detected in thesesediments, but one possible explanation may be the uptakeof K+ ions from seawater or interstitial waters in possiblemixed-layer clay minerals. However, no significant changewas found in interstitial K+ values in the overlying sedimentcolumn.Fe, Mn, and Al

The ratios with respect to (Al + Fe + Mn) are very closeto those found by Boström et al. (1972a) as characteristic

401

T. B. WARNER, J. M. GIESKES

of basal metalliferous sediments in the Atlantic Ocean(Table 5). The data suggest (Figure 2) that Al and Mnbehave differently from Fe, for their ratios change sensiblyand gradually in the first 26 meters of the core, whereasFe/(A1 + Fe + Mn) is more or less constant. This impliesthat Fe and Mn may arrive in the sediments via differentpaths, as has also been discussed by Bender et al. (1971)and Cronan and Garrett (1973). It is interesting to notethat these ratios do not change much, whereas the actualpercent Fe drops 10 percent (11.5 to 9) and the percent Mndrops 25 percent (3.2 to 2.4) toward the top of the coredinterval. The ratios appear to be more useful identifiers forridge crest inputs than are the actual amounts, while theactual amounts may better describe the rates of input. Inview of the fact that about 7 meters of sediments aremissing above the diabasic sill, and also because of theunknown thickness of sediment below this sill to actualbasement, we can only speculate that actual contents in Mnand Fe will continue to rise to values substantially higherthan those found in this study (cf., Table 6, and Boströmand Fisher, 1971; Boström et al., 1972b).

It is difficult to assess the probable distance over whichridge crest activity had an influence on the sediments.

TABLE 5Comparison of Mn and Fe Enrichment in Basal

Metalliferous Sediments of theIndian and Atlantic Oceans

Although spreading rates are roughly known, spreading haschanged direction in this particular region (Schlich, thisvolume; McKenzie and Sclater, 1973). The genesis of adiabasic sill in this region of intense interaction between thesouthwestward trend of basement and the relatively youngocean floor spreading in the northwest direction from thesouthwest Indian Ridge again complicates interpretations.As distance above basement may be considerable, it isworthwhile to speculate about the possible source of thehigh iron in the sediments. Cronan and Garrett (1973)concluded that most of the iron did not seem to beassociated with the carbonate or ferromanganese oxidephases in the Pacific basal sediments. Sayles and Bischoff(1973) found that much of the iron in their samples fromthe East Pacific Rise was associated with the clay fractionin the form of nontronites. Although hydrothermalleaching of basalts (Corliss, 1971) has been suggested as themajor source of this iron, we submit that the iron mayoriginally have been associated with the volcanic glass andash deposited in these sediments, either interspersed in thecarbonate ooze or as distinct ash layers. Later diageneticalteration associated with devitrification processes mayhave mobilized some of this iron. Further chemical andmineralogical analysis of the samples should clarify thisconcept.

The manganese occurs mostly in the form of dispersedmanganese oxides. This observation led Leclaire (thisvolume) to suggest that deposition occurred rapidly.Perhaps the association between rapidly depositingbiogenous calcium carbonate and intermixed volcanic ashled to favorable conditions for rapid precipitation of MnC^phases from seawater (Michard, 1969; Cronan and Garrett,1973).

CONCLUSIONS

Studies of the iron-manganese-rich deposit in the lowestpart of Hole 245 drilled during DSDP Leg 25 have revealeda close correlation between these deposits and otheriron-manganese-rich deposits reported on active oceanridges. The present studies suggest that the provenance ofthe iron, and perhaps the manganese, is indeed associated

TABLE 6Comparison of Some Metal Concentrations in Metalliferous Sediments from Deep Cores in the

Pacific and Indian Oceans, in Surface Metalliferous Sediments from the East Pacific Rise,and in Normal Pelagic Clays from the Indian and Pacific Oceans (CaC03-Free Basis)

(Al

(Al

(Al

Ratio

100 Al+ Fe + Mn)

lOOFe+ Fe + Mn)

lOOMn+ Fe + Mn)

near ridge crestfar from ridge crest

near ridge crestfar from ridge crest

near ridge crestfar from ridge crest

Characteristic Valuesof the Ratio

Atlantic OceanBoström et al.

(1972a)

< 2 0>60

>60<40

>160-2

Indian Ocean(this study)

20

60

18

Average of 11 ferruginous sedimentsfrom the DSDP Leg 16, Pacific Ocean(Cronan et al., 1972)

East Pacific Rise, surface(Boström and Peterson, 1969)

Indian Ocean, DSDP, Leg 25(This study)

Average deep-sea clay(Turekian and Wedepohl, 1969)

Average surface Pacific pelagic clay(Cronan, 1969)

Average deep-sea surface IndianOcean clay (Marchig, 1972)

Age

Early Miocene tomiddle Eocene

Recent

Early Paleocene

Recent

Recent

Recent

Fe

17.5

18.0

11

6.5

5.1

6.5

Mn

4.5

6.0

3.2

0.7

0.5

0.6

Cu

920

730

200

-

320

Zn

360

380

250

-

—

Ti

-

0.02

0.5

0.5

—

0.7

402

IRON-RICH BASAL SEDIMENTS FROM THE INDIAN OCEAN, SITE 245

with volcanic activity on active ridge crests but that thesource may be associated with the deposition of volcanicash and glass, rather than with "hydrothermal" emanations.The distinct ash layers analyzed in this study have a verydifferent chemical signature, compared to the surroundingsediments. The latter are characterized by low contents inAl and high contents in Fe and Mn, as well as by low Al/Kand Al/Ti ratios. SiO2/Al2O3 ratios are high in thecarbonate portions of the core interval, suggesting thepresence of amorphous silica (not biogenous), quartz, andlow alumino-silicates (e.g., palygorskite). Sediments foundnear ridge crests are often characterized by such highSiθ2/Al2O3 and low Al/K ratios. Mineralogical studies, aswell as further chemical studies, are in progress, and thesewill clarify some of the observations made in this study.

ACKNOWLEDGMENTS

One of us (TBW) thanks the Naval Research Laboratoryfor generously granting a sabbatical leave during which thisstudy was performed, Scripps Institution of Oceanographyfor providing a stimulating environment and many valuablefacilities, and the many colleagues at Scripps who offeredgracious and enthusiastic assistance on many occasions. Weparticularly thank Tracy Vallier, Kathe Bertine, and MiriamKastner for frequent stimulating and helpful discussions;Ron La Borde and James Greenslate for help, advice,consultation, and the use of their X-ray spectrometers; andJohn Stevens for help with preparation of some of thesamples. Jerry Bode kindly carried out the LECO calciumcarbonate measurements.

This research was generously supported by NSF GrantGA-33229.

REFERENCES

Bender, M., Broecker, B., Gornitz, V., Middel, U., Kay, R.,Sun, S., and Biscaye, P., 1971. Geochemistry of threecores from the East Pacific Rise: Earth Planet. Sci. Lett.,v. 12, p. 425-433.

Boström, K., 1970. Geochemical evidence for ocean floorspreading in the South Atlantic Ocean: Nature, v. 227,p. 1041.

Boström, K. and Peterson, M. N. A., 1966. Precipitatesfrom hydrothermal exhalations on the East Pacific Rise:Econ. Geol., v. 61, p. 1258-1265.

Boström, K. and Peterson, M. N. A., 1969. The origin ofaluminum-poor ferromanganoan sediments in areas ofhigh heat flow on the East Pacific Rise: Mar. Geol., v. 7,p. 427-447.

Boström, K, and Fisher, D. E., 1971. Volcanogenicuranium, vanadium, and iron in Indian Ocean sediments:Earth Planet. Sci. Lett., v. 11, p. 95-98.

Boström, K., Joensuu, O., Valdes, S., and Riera, M., 1972a.Geochemical history of South Atlantic Ocean sedimentssince Late Cretaceous: Mar. Geol., v. 12, p. 85-121.

Boström, K., Farquharson, B., and Eyl, W., 1972b.Submarine hot springs as a source of active ridgesediments: Chem. Geol., v. 10, p. 189-203.

Corliss, J. B., 1971. The origin of metal-bearing submarinehydrothermal solutions: J. Geophys. Res., v. 76, p.8128-8138.

Cronan, D. S., 1969. Average abundances of Mn, Fe, Ni,Co, Cu, Pb, Mo, V, Cr, Ti and P in Pacific pelagic clays:Geochim. Cosmochim. Acta, v. 33, p. 1562-1565.

Cronan, D. S. and Garrett, D. E., 1973. Distribution ofelements in metalliferous Pacific sediments collectedduring the Deep Sea Drilling Project: Nature, Phys. Sci.,v. 242, p. 88-89.

Cronan, D. S., van Andel, T. H., Heath, G. R., Dinkelman,M. G., Bennet, R. H., Bukry, D., Charleston, S., Kaneps,A., Rodolfo, K. S., and Yeats, R. S., 1972. Iron-richbasal sediments from the eastern Equatorial Pacific: Leg16, Deep Sea Drilling Project: Science, v. 175, p. 61-63.

Katz, A., Sass, E., Starinsky, A., and Holland, H. D., 1972.Strontium behavior in the aragonite-calcite transforma-tion. An experimental study at 40-98°C: Geochim.Cosmochim. Acta, v. 36, p. 481-496.

Manheim, F. T. and Sayles, F. L., in press. Composition andorigin of interstitial waters of marine sediments, basedon Deep Sea Drill Cores. In Goldberg, E. D. (Ed.), TheSea: ch. X, v. 5.

Marchig, V., 1972. Zur Geochemie rezenter Sedimente desIndischen Ozeans: "Meteor" Forsch., Reihe C, No. 11,p. 1-104.

McKenzie, D. P. and Sclater, J. G. 1937. The evolution ofthe Indian Ocean: Sci. Am., v. 228, no. 5, p. 63-72

Michard, G., 1969. Depot de traces de manganese paroxydation: Comptes Rend. Acad. Sci. France, v. 269, p.1811-1814.

Natland, J. H., 1973. Basal ferromanganoan sediments atDSDP Site 183, Aleutian Abyssal Plain, and Site 192,Meiji Guyot, Northwest Pacific, Leg 19. In Creager, J. S.,Scholl, D. E., et al., Initial Reports of the Deep SeaDrilling Project, Volume 19: Washington (U.S.Government Printing Office), p. 629-636.

Piper, D. Z., 1973. Origin of metalliferous sediments fromthe East Pacific Rise: Earth Planet, Sci. Lett., v. 19, p.75-82.

Pratt, R. M., Scheidegger, K. F., and Kulm, L. D., 1973.Volcanic ash from DSDP Site 178, Gulf of Alaska. InKulm, L. D., von Huene, R., et al., Initial Reports of theDeep Sea Drilling Project, Volume 18: Washington (U. S.Government Printing Office), p. 833-834.

Sayles, F. L. and Bischoff, J. L., 1973. Ferromanganoansediments in the equatorial East Pacific: Earth Planet.Sci. Lett., v. 19, p. 330-336.

Turekian, K. K. and Wedepohl, K. H., 1969. Distribution ofthe elements in some major units of the earth's crust:Geol. Soc. Am. Bull., v. 72, p. 175-192.

von der Borch, C. C. and Rex, R. W., 1970. Amorphousiron oxide precipitates in sediments cored during Leg 5,Deep Sea Drilling Project. In McManus et al., InitialReports of the Deep Sea Drilling Project, Volume V:Washington (U. S. Government Printing Office), p.541-544.

von der Borch, C. C, Nesteroff, W. D., and Galehouse, J. S.,1971. Iron-rich sediments cored during Leg 8 of theDeep Sea Drilling Project. In Tracey, J. I., Sutton, G. H.,et al., Initial Reports of the Deep Sea Drilling Project,Volume VIII: Washington (U. S. Government PrintingOffice), p. 829-833.

403