34. INVESTIGATIONS OF TERTIARY CLAY MINERAL DISTRIBUTIONSAROUND TASMANIA, DSDP, LEG 29

Victor A. Gostin, Department of Geology, University of Adelaide, South Australiaand

Kevin C. Moriarty, Discipline of Earth Sciences, Flinders University of South Australia

ABSTRACT

Clay minerals from Tertiary samples of DSDP Sites 280, 281, 282,and 283 were analyzed mainly by X-ray diffraction and cation ex-change methods. Site 282, off the Tasmanian west coast, contains apersistent Australian clay mineral assemblage that reflects deepweathering conditions in Tasmania and basic volcanism increasingfrom Eocene into the Oligocene. Site 280, south of the South TasmanRise, was probably situated close to Antarctica in the Eocene-Oligocene and shows a change from temperate source area con-ditions in the early (?) and middle Eocene, to cold-glacial conditionsin the latest Eocene and Oligocene. This is supported by Oligocene-Miocene data from cores near Antarctica.

It is unlikely that any deep oceanic connection existed between thebasins on either side of the South Tasman Rise until probably thelate Oligocene when the circum-Antarctic current developed.

INTRODUCTION

This investigation forms part of a larger studyinvolving the detailed clay mineralogy of Recent andTertiary sediments in the Southern Ocean between Aus-tralia, New Zealand, and Antarctica. It concentrates onthe clay minerals of the Tertiary sequences in the Leg 29DSDP sites that surround Tasmania (Sites 280 to 283).For significant interpretations it is obviously necessaryto compare these results with those obtained from sitesof DSDP Leg 28, south of New Zealand and in the Aus-tralian region of Antarctica.

Among other factors, clay mineral types and propor-tions are influenced by the nature of the source rocksand their degree of weathering. The <2-µ fraction in theocean should become relatively dispersed within anydepositional area and therefore should show the inte-grated effect of its provenance and conditions ofweathering. As a consequence, clay mineral proportionsmay be used to outline certain provinces that havecharacteristic clay mineral distributions. Such analysesmay be usefully compared to proposed paleoclimaticand plate tectonic hypotheses.

The clay mineral results obtained are shown in Table1 and illustrated on Figure 1. Results from DSDP X-rayanalyses (Appendix I, this volume) are shown in Figure2. The vertical scale on the figures is a direct time scale,and lithologic columns are shown for each site. Thedrilling sites in the figures are arranged in such a waythat Site 281 on the South Tasman Rise separates Site282 in the ocean basin west of Tasmania, from Sites 280and 283 which are to the east or south of the rise. If, as isproposed, the separation of Australia and Antarctic oc-curred along a set of transform faults that extend southfrom along the west coast of Tasmania, then this

arrangement is appropriate for mineral analyses aroundthe area of final separation, i.e., the South Tasman Rise.

Since widely different results may be obtained by us-ing differing analytical methods, details of the methodsused here are described.

ANALYTICAL METHODSThe samples were washed free of salts. A subsample

was used to determine bulk chemical composition by X-ray fluorescence (XRF) spectroscopy. The sand fractionwas removed by wet seiving in a 63µ sieve. Carbonatewas removed from the <63µ fraction by glacial aceticacid. The <2µ fraction was separated by repeated cen-trifugation, and clay/silt proportions obtained byweighing the fractions after washing in alcohol, acetone,and ether, and air drying. Amorphous silica was re-moved from the <2µ fraction following the method ofHashimoto and Jackson (1960). The solute was savedfor analysis of Si and Al content in order to test for claymineral dissolution. Amorphous iron is removed follow-ing the method of Mehra and Jackson (1960) and thesolute saved for analysis of Fe, Al, and Si. A knownamount of clay is sedimented onto corundum tiles over avacuum saturated with Mg, glycerine, and air dried. X-ray diffraction is carried out in the range of 3° to 26°,with variations for particular samples. A separate sam-ple is sedimented onto a corundum disc over a vacuum.The disc is saturated with 2N BaCb, washed, dried, andanalyzed for Ba, K, and chlorine content by XRF. Thebarium content is used to calculate the cation exchangecapacity.

The aim of these preparation procedures is to obtain asediment fraction in which the component minerals arereadily identifiable and their relative quantities deter-minable. It is expected that an amount (usually low) of

1077

V. A. GOSTIN, K. C. MORIARITY

TABLE 1Clay Mineral Ratios as Determined by X-Ray Diffraction

and by Cation Exchange Methods

Sample(Interval in cm)

280-1-3, 122280A-1-3, 122280A-4-4, 96280A-6-2, 97280A-8-2, 105280A-11-2, 70280A-13-2, 109280A-15-2, 34280A-17-6, 54280A-19-5, 54280A-20-2, 53280A-22-1.91

281-6-5, 70281-11-6, 100281-134, 100281-13, CC (top)281-14-6, 110281-15-3, 60281-16-6,120

282-1-1,70282-4-3,110282-8-2, 66282-11-3, 85282-13-2, 89282-14-3, 83282-16, CC282-17-3, 81282-18-1, 109

283-5-2, 101283-8-2, 62283-13-2, 81283-17-4, 94

ChloriteXRD

46.819.727.747.938.225.8

7.212.6

5.48.3

13.70

16.519.512.7

016.618.512.8

trtrtrtr

12.8tr000

16.65.40

24.6

CEC

47.921.422.245.834.022.825.325.114.926.530.0

0

17.2

0000

11.40000

30.9

022.4

KaoliniteXRD

02.52.100

8.70.50000

9.6

23.25.40.30.4

22.413.615.7

39.039.625.241.536.051.867.857.833.7

13.94.30.63.9

CEC

02.81.700

7.71.70000

15.2

0.4

12.6

39.346.330.232.032.162.054.161.656.1

26.0

7.83.6

Montmor-illonite

XRD

3.934.4.9.42.6

13.66.1

77.852.680.479.258.936.513.013.44.70

20.335.533.0

15.949.464.147.428.432.922.029.455.356.377.893.841.9

CEC

1.728.727.4

6.823.117.122.3

5.946.333.610.0

0

40.1

15.240.956.959.536.319.737.824.825.418.3

21.547.0

IlliteMuscovite

XRD

49.343.460.849.548.259.414.534.814.212.527.453.947.361.782.399.640.732.438.545.111.010.711.122.815.310.212.811.013.212.5

5.629.6

CEC

50.447.248.747.442.952.450.769.038.839.860.085.2

99.6

30.1

45.512.912.9

8.620.218.3

8.113.618.424.8

70.727.0

the clay minerals is removed by these treatments, andthe removal will be a differential process according tothe size and nature of the mineral particles. To minimizethis effect, all samples were subjected to exactly the sametreatments. The clay mineral groups, kaolinite, illite-muscovite, montmorillonite, and chlorite, constitute thebulk of the <2µ fraction in all cases and are affectedvery little by the treatments. Zeolites are removed by thetreatments, allowing quantification of the mont-morillonite by cation exchange methods.

MINERAL IDENTIFICATION ANDQUANTIFICATION

Montmorillonite was determined both by the heightof the 17 Å peak obtained after glycerine saturation andfrom the cation exchange capacity (CEC). In thecalculations it is assumed that the CEC of mont-morillonite is 100 meq/100 g. This has been shown to bereasonable by analysis of nearly pure montmorillonitefrom the surface sediments in the area.

Chlorite was determined from the height of the 14 A(001) peak, with the presence of a (003) peak at 4.72 Å toconfirm the abundance. There are various methods todetermine the relative contributions of kaolinite andchlorite to the 7 Å peak. Those of Biscaye (1964) and

Schultz (1960) were applied without consistent successand were thus not used any further. A refinement of themethod of Rex and Murray (1970) and Pow-foong Fanand Rex (1972) was used. This assumes a constant ratioof the (001), (002), and (003) peak heights for the deep-sea chlorites. The height of the (003) reflection is used todetermine the relative abundance of chlorite, and thecontribution of this chlorite to the 7 Å peak is calculatedfrom a constant ratio. Since the (001) and (003) reflec-tions of chlorite are very sensitive to the variations iniron content and position in the lattice, any variation inthe composition of the chlorites will give a misleadingabundance. This will especially affect the kaolinitequantification. Accordingly, this study tried to deter-mine the possible variations in chlorite composition atSite 280 by measuring ratios of the (001/002/003) reflec-tions.

The results, shown in Table 2, indicate that the ratiosvary in a way that cannot be explained by variation inthe kaolinite proportion. Sample 280-1-3 containsnegligible kaolinite, and the (001/002) ratio is 2.1. Theaverage ratio for all other samples with high chloritecontent is 0.83. These results are consistent with thoseobtained from Mg-chlorites. The variation in (001/003)ratios shows that the assumption of constant composi-

1078

SITE

1-

282 100% 281

18.V VV

Core No. ALithology-i(DSDP symbols)CLAY MINERAL RATIOS

b0% continental^ontmorillonite basementdetermined separately^by cation exchange

Crystallinity of 1(decreases from 1 to 30)

CLAY MINERAL RATIOSC = ChloriteK = KaoliniteM = MontmorilloniteI = I H i t e and MuscoviteDetermined by peak heights of oriented mounts.

A2

2831

MI0.M

22.5

OLIG.

M

37.5

EOC.M

PAL.

E

- 65

Figure 1. Qay mineral ratios.

90H>

nt-1

>

rö

253H

δCA

o00

o

SITE 282 K 2811-

1-

v v v continentalbasement

Clay minerals determined as random powders on decalcif ied,<2 µm f ract ion.Clay minerals arranged in order from l e f t .C=Chlorite, K=Kaolinite, M=Montmorillonite,I = I l l i t e and Muscovite.MIXL=Mixed layer mica-montmorillonite.

Figure 2. Gay mineral ratios.based on DSDP analysis.

2801-

C K M

1-

K M I

. PLEIST.

PLIO.r

L

—

MIO.M

<>

GO

ST]

z

p

MOR]

[AR

ITY

-22.5

OLIG.M

37.5

53.5

TERTIARY CLAY MINERAL DISTRIBUTIONS

TABLE 2ooß Ratios for High Chlorite

Samples, Site 280

Sample

280-1-3280A-1-3280A-4-4280A-6-2280A-8-2280A-11-2280A-13-2280A-15-2280A-17-6

001/002

2.10.70.71.01.10.50.80.90.9

001/003

4.93.42.53.84.81.83.03.04.0

tion is not justified. However, it is assumed that allsamples had a (001/002) ratio of >0.83 for the chlorite.If the ratio in any particular sample was higher, then allof the 7 Å peak was attributed to chlorite. If the ratiowas less than 0.83, the 7 Å peak height was dividedproportionately between kaolinite and chlorite. Takinga higher ratio than 0.83 may be justified from theevidence, thereby raising the kaolinite proportion. Thisshould be taken into account in interpreting results withlow kaolin proportions in the presence of chlorite.

Relative percentages using peak heights werecalculated by assuming an equivalence in concentration,if the intensities of 7 Å kaolinite, 10 Å illite-muscovite,14 Å chlorite, and 17 Å montmorillonite, were in theratio of 2:1:1:4. The kaolinite, illite, and mont-morillonite equivalences are in common use. Thechlorite to illite equivalence was obtained experimental-ly from calibration curves of Mg-chlorite and illite mix-tures.

The CEC method of calculating relative percentagesuses the same equivalence ratios of intensities as stated,but ignores the montmorillonite peak height. The CECof the sample is divided between the constituentminerals by assuming a CEC for each mineral: 100 formontmorillonite, 20 for illite, 12 for kaolinite, and 10 forchlorite (units in meq/100 g). The relative concentrationgiven by the peak height is used to determine the actualcontribution of the minerals kaolinite, illite, and chloriteto the total CEC. The remainder is assigned to mont-morillonite. The potassium content is used to check theproportion of illite-muscovite obtained by the methods.It aids in determining the presence of high-potassiummica in the 10 Å fraction. Crystallinity of this fractionwas measured as the width at half peak height minus theinstrument factor.

Comparison Between Different MethodsThe results obtained by X-ray diffraction show a

general agreement between the fully treated clays onoriented mounts (Figure 1) and the decalcified randomoriented mounts (Figure 2). The main difference appearsto be in the Neogene calcareous ooze samples at Site281, where the proportion of montmorillonite and illiteis quite different. Analysis of the glauconite sample(Core 13) on the oriented mount showed an excellentsymmetrical 10 Å peak with a CEC of 19.2, whereas theDSDP analyses of random mounts indicate a mixed-layer illite-montmorillonite. Perhaps different samples

were used for these analyses. The calcareous ooze aboveCore 13 at Site 281 contains scattered glauconite pellets,so that some of the illite in Figures 1 and 2 may repre-sent authigenic glauconite.

The determination of montmorillonite by cationexchange method gave similar results to that obtainedby X-ray diffraction, except for samples adjacent to theextrusive or intrusive basalts at Sites 280 and 282, theorganic carbon-rich sediments in the lower part of Site280, and the very well-crystalline montmorilloniticsediments at Site 283. In all these cases, the mont-morillonite values determined by CEC were signifi-cantly lower than those determined by X-ray diffraction.These samples are being investigated further.

RESULTS AND DISCUSSION

Site 282 is closest to an Australian continentalinfluence, and the dominant clays are montmorilloniteand kaolinite. The increase in montmorillonite from lateEocene to middle Oligocene is accompanied by adecrease in kaolinite. Illite and muscovite have a fairlyconstant and low percentage (but high potassium con-tent), and chlorite is present in trace amounts only. Thelimited early Miocene and Pleistocene results are similarto the above. Such an assemblage suggests a persistingsupply from a deeply weathered landmass, with the in-crease of montmorillonite probably related to increasingvolcanic activity from the Oligocene into the Miocene.This is observed in the volcanism of Tasmania (Suther-land et al., 1973). Some decrease in kaolinite suggestsreduced weathering (colder) conditions on Tasmaniaand southeast Australia during the Oligocene.

Clay minerals at Sites 280, 281, and 283 are distin-guished by the presence of significant amounts ofchlorite, sometimes to the exclusion of kaolinite as atSite 280. A much higher illite-muscovite content is alsopresent, some of which shows the good crystallinity ofmuscovite (low numbers in the I column of Figure 1).These observations suggest an Antarctic influence, es-pecially at Site 280. The potassium content of the illitesbelow Core 280A-11 is very low, which suggests thatthese illites and a soil(?) chlorite was being derived fromtemperate Eocene conditions on Antarctica, without theintense leaching which was producing kaolinite in Aus-tralia at the same time. In contrast, the clays above Core280A-11 show a high potassium content (8%-10% K) in-dicating a muscovite-type mica. This, together with ahigh chlorite content, suggests cold to glacial conditionsin the source area, here postulated as Antarctica. Clayminerals of Oligocene and Miocene sediments in theAustralian and New Zealand sections of the oceanaround Antarctica show surprisingly little variation withtime and consist dominantly of mica and mont-morillonite with a significant proportion of chlorite, butvirtually no kaolinite (DSDP Leg 28, Sites 266 to 269,and 274). This suggests that the mica and chlorite mayhave been steadily supplied from the cold or glaciatedAntarctic continent, and that this influence reached Site280 which was quite close to that continent during theOligocene.

The high Eocene montmorillonite content at Site 280is associated with the highly organic, "euxinic" con-

1081

V. A. GOSTIN, K. C. MORIARITY

ditions at that site. The montmorillonite probablyformed from alteration of volcanic material which wasderived from the active spreading center that must haveexisted a short distance to the south. Perhaps some ofthis fine-grained material also found its way north intothe Tasman Sea as recorded by the clays at Site 283. Ingeneral, however, montmorillonite does not show an in-crease towards the contact with the underlying oceanicbasement. This would be expected from the proximity ofthe volcanically active spreading centers. Sedimentsfrom the nearby continents obviously dominate the claymineral assemblages.

Tertiary sediments to the southwest of New Zealand(Site Report Chapters 277, 278, 279, this volume) aredominantly biogenic, with only a small clay mineralcontent. This consists of dominant montmorillonite andmica from the Paleocene onwards, with kaolinite andchlorite both present in small, but significant amountsonly from the early Oligocene. The Paleocene-Eoceneassemblage is difficult to explain, but the Oligocene-Miocene clays appear to be thoroughly mixedAustralian and Antarctic assemblages.

CONCLUSIONS1. The clay mineral assemblage at Site 282 is charac-

teristically Australian in type, and this assemblage per-sists from the late Eocene.

2. A significant increase in montmorillonite occurs inthe Oligocene at Site 282 and is probably related to theincreasing volcanism in the Tasmanian region (Suther-land et al., 1973).

3. The clay mineral assemblages at Sites 280, 281, and283 are broadly similar. They are distinct from those atSite 282 and are probably related partly to an Antarcticsource. The Antarctic influence is most noticeable atSite 280, which must have been quite close to that conti-nent during the Oligocene.

4. Sediments in cores below Core 280A-11 (before thelate Eocene) indicate temperate conditions on Antarc-tica. Sediments of the Oligocene in cores of Hole 280A

and in cores near Antarctica (DSDP Leg 28) indicatecold to glacial conditions.

5. The clay mineral data support the hypothesis thatthere was no deep oceanic connection between thebasins on either side of the South Tasman Rise, that is,no circum-Antarctic current until probably the lateOligocene.

ACKNOWLEDGMENTSThe authors wish to thank the Division of Soils, C.S.I.R.O.,

Adelaide, for use of their analytical facilities.

REFERENCESBiscaye, P. E., 1964. Distinction between kaolinite and chlorite

in Recent sediments by X-ray diffraction: Am. Mineral., v.49, p. 1281.

Hashimoto, I. and Jackson, M. L., 1960. Rapid dissolution ofallophane and kaolinite-halloysite after dehydration. InSwineford, A. (Ed.), Clays and clay minerals, 7th Proc:London (Pergamon Press), p. 102-113.

Mehra, O. P. and Jackson, M. L., 1960. Iron oxide removalfrom soils and clays by a dithionate-citrate system bufferedwith sodium bicarbonate. In Swineford, A. (Ed.), Clays andclay minerals, 7th Proc: London (Pergamon Press), p. 317-327.

Pow-foong Fan and Rex, R. W., 1972. X-ray mineralogystudies—Leg 14. In Hayes, D. E., Pimm, H.C.,et al., InitialReports of the Deep Sea Drilling Project, Volume 14:Washington (U.S. Government Printing Office), p. 677.

Rex, R. W. and Murray, B., 1970. X-ray mineralogy studies,Leg 4. In Badev, R. G., Gerard, R. D., et al., Initial Reportsof the Deep Sea Drilling Project, Volume 4: Washington(U.S. Government Printing Office), p. 745.

Schultz, L. G., 1960. Quantitative X-ray determination ofsome aluminous clay minerals in rocks. In Swineford, A.(Ed.), Clays and clay minerals, 5th Proc: London (Per-gamon Press), p. 216-224.

Sutherland, F. L., Green, D. C, and Wyatt, B. W., 1973. Ageof the Great Lake basalts, Tasmania, in relation toAustralia Cainozoic volcanism: J. Geol. Soc. Australia, v.20, p. 85.

1082