Paleoceanographic interpretations of coccoliths and oxygen-isotopes from the sediment surface of the southwest Indian Ocean

Marine Micropaleontology, 13 (1989) 325-351

325

Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands

Paleoceanographic Interpretations of Coccoliths and Oxygen-Isotopes from the Sediment Surface of the Southwest Indian Ocean MARK J. FINCHAM and AMOS W I N T E R Marine Geoscience Unit, University of Cape Town, Rondebosch 7700 (South Africa) (Revised and accepted March 7, 1988)

Abstract Fincham, M.J. and Winter, A., 1989. Paleoceanographic interpretations of coccoliths and oxygen-isotopes from the sediment surface of the southwest Indian Ocean. Mar. Micropaleontol., 13: 325-351. The distribution of forty-four coccolithophore species in one hundred deep-sea core-tops from the southwest Indian Ocean is described. Three coccolith assemblages have been recognised (Maputo, Agulhas Current and deep water) by the relative abundances of four ecologically significant coccolithophore species (Gephyrocapsa oceanica, Emiliania huxleyi, Calcidiscus leptoporus and Umbilicosphaera sibogae). Their biogeographical distribution appears to be related to water temperature, nutrient concentration and dissolution. The degree of preservation of coccoliths and foraminifera indicates that the carbonate lysocline lies somewhere between 3500 and 4000 m, resulting in the concentration of dissolution-resistantmicrofossils below this depth. Stable oxygen isotope ratios of the planktonic foraminiferal species Globigerinoides sacculifer range between - 1.5 to -1.0%o PDB (equal to 22.8-25.1°C) and occur in a narrow band on the sea floor beneath the "A" route of the Agulhas Current. These values are about 0.5 per rail PDB lighter than samples analyzed on either side of this band and can be explained by the Agulhas Current's elevated temperature at the ocean surface of 2-3 ° C. Thus an oxygen isotope imprint of the Agulhas Current exists beneath it on the sea floor. The Agulhas Current is probably the major factor influencing sedimentation, sediment-distribution patterns and geological features in the study area. At present it is voluminous and fast flowing, possibly eroding sediments up to 2500 m below the surface. The oxygen-isotope ratios and nannoplankton counts obtained in this study indicate, however, that the majority of samples are most probably recent or at least not older than 85,000 years. This implies that sediments are accumulating on the ocean floor and that the Agulhas Current does not have a pronounced erosional influence, at least in areas from which cores were retrieved for this study.

Introduction

It is often difficult and impractical (due to patchiness, sampling limitations and expense) to correlate empirically the distribution of living coccolithophores with environmental factors. However, studying coccoliths from surface sediments on the ocean floor, where biological 0377-8398/89/$03.50

and oceanographic fluctuations are integrated, can overcome this limitation. Patterns of coccolith accumulation on the ocean floor reflect environments of production in the overlying surface waters {Roth and Berger, 1975; Geitzenhauer et al., 1977) and maps of species distribution in the sediment surface correspond to plankton maps (Honjo and

© 1989 Elsevier Science Publishers B.V.

326

Okada, 1974) and to the location of water masses, The distribution patterns of calcareous nannofossils in surface sediments of the Atlantic and Pacific have been the subject of extensive interest (McIntyre, 1967; McIntyre and Bd, 1967; McIntyre et al., 1970; Ushakova, 1970; Roth and Berger, 1975; Geitzenauer et al., 1977; Roth and Coulbourn, 1982). Only one research paper has dealt with the distributions of coccoliths in marine sediments in the southwest Indian Ocean. Siesser (1975) studied Pleistocene nannofossils in sediment cores from the South African continental margin, This study describes the composition and distribution of core-top coccolithophore assemblages, and their correlation with ecological conditions in the overlying water masses. We also attempt to extent the investigations of Vincent and Shackleton {1980) to the continental margin of southeast Africa. Vincent and Shackleton (1980) investigated oxygen isotopes of planktonic foraminifera from the sediment surface just north of the present study area. Their results enabled t h e m to construct isotherms from isotopically estimated temperatures which reflect t h e imprint of the Agulhas Current on the underlying oceanic sediments, Information obtained in this report should contribute to our understanding of the ecological tolerances and interpretation of coccolithophores from the southwest Indian Ocean and help in palaeoenvironmental reconstruction of the Agulhas Current region of the southwest Indian Ocean. Oceanography The Agulhas Current (Fig. 1) is the most prominent oceanographic feature of the study area. It is the western boundary current of the South Indian Ocean subtropical gyre and may be divided into four different oceanographic regimes which parallel the coast of southern Africa (Pearce, 1977): (1) The inshore region which lies predominantly on the continental

shelf and consists of relatively cool and low velocity water. (2) The current core, usually only a few tens of kilometers wide. It is warm (varying seasonally in temperature between 22 ° and 27°C), fast (surface velocity usually exceeds 1 meter per second though there is no pronounced seasonal variation in core velocity) and has a tendency to meander with a period of a few days. (3) Between the core and inshore region lies the western boundary of the current which is a region of intense horizontal shear. (4) The region seaward of the eastern boundary of the current core is a region of weaker currents and cooler temperatures. The core of the Agulhas Current divides into two major flow paths (commonly termed Routes A a n d B : F i g . 1) atapproximately26°S (Harris and Van Foreest, 1977). Route A, the inshore branch, travels from the northeast to southwest following the edge of the continental shell Route B flows down the eastern Central Terrace (Fig. 1) following the Mozambique Ridge southward, until at approximately 32 ° S it takes a sharp westerly turn rejoining route A near the coast (Martin, 1984). The Agulhas Return Current is formed from Agulhas Current water that has turned eastward and meanders east into the South Indian Ocean gyre (Shannon et al., 1973). Intermittent inshore counter-currents flow in a northeasterly direction (Malan and Schumann, 1979 ) and are associated with the formation of cyclonic eddies near Durban and Maputo. Counter-currents cause dynamic upwelling which results in high nutrient and low oxygen values (Martin, 1984). Evidence uncovered by satellite imagery and synoptic hydrographical data suggest that important changes occur in the system over periods of 3-4 weeks, rather than seasonally as was previously thought (Harris and Van Foreest, 1977). A selection of satellite photographs was taken by the European Space Agencies' "Meteosat" satellite from the 3rd of January to the 16th of March, 1984 and has been used to compile a composite map which roughly outlines the mean

BASIN

F

I

=//

R

C

J

j

A

ASI N

.i:" .¢

""°'°'

i.

/

g

.. ....

-

~

I

.~..

:i.

/.

2...::

~r

2

'~ ~'~

]

MADAG

e~..~

A SC AR

~e~.

2

CRRAOsZ I EN I

/

V

Fig. 1. Circulation p a t t e r n s of the southwest Indian Ocean. Surface/near-surface currents = black arrows. 1 = N o r t h Equatorial Current; 2 = South Equatorial Current; 3 = East Madagascar Current; 4 = Mozambique Current; 5 = Agulhas Current. 6 = West Wind Drift at subtropical convergence. From Martin ( 1981 ).

AGULHAS

/

A

D ~nFACE/,~,~SU.FACECURRE,rs

.

i

~

328

route of the Agulhas Current during this period (see Fig. 3a). For most of this time the Agulhas Current followed a constant route down the coast of southern Africa, its penetration south varying from week to week. Route B is only intermittently present during this period and is not as clearly defined, The Antarctic Bottom Water (ABW; Fig. 1 ) flows NE through the Agulhas Passage at depths greater than 4500 m, and may be represented in the Agulhas Basin by the I ° C isotherm (Westall, 1984 ). ABW originates in the Weddell Sea and spreads from the Atlantic-Indian Basin in two directions: one path spreads easterly, through the Crozet Basin into the Madagascar, Mascarene, Somali and Arabian Basins (Kolla et al., 1976). The second branch flows northward along the western boundary of the South Atlantic and passes through the SW Indian Ocean Ridge south of the Agulhas Plteau (Fig. 1 ) about 25 oE (Kolla et al., 1976; Westall, 1984) into the East Agulhas Basin. This component then flows northwestward into the West Agulhas Basin, and then northeastward through the Agulhas Passage. ABW is relatively rapid (622cm/s;Westall, 1984)andturbulentinitsflow through the Agulhas Passage (Westall, 1984) and bottom photographs showing well-developed ripples and a strong nephaloid layer confirm the importance of ABW in sediment transport in the Agulhas Passage. Materials and methods The study area in the southwest Indian Ocean is approximately 2000 km in length from off Maputo in the north to off the Agulhas Bank in the south and varies in width from 300 km in the north to 1500 km in the south. Figure 2 shows the bathymetric relief of the region and identifies the most prominent features (see Dingle, 1978; Goodlad, 1978; Martin, 1984 and Dingle and Robson, 1985 for a more detailed description), In total, one hundred core-top samples from water depths ranging between 248 and 4845 m

were investigated (Fig. 2 ). Forty-four samples were obtained from piston cores, forty from trigger cores, and sixteen from gravity cores, recovered during several cruises of the Thomas B. Davie. Samples are distributed along the entire length and breadth of the Agulhas Current Systern, although some areas have a better sample coverage than others. Because of this the study area has for convenience been divided into four sample sectors: (1) Maputo (M); (2) Durban (D); (3) East London (EL); (4) Port Elizabeth (PE) (Fig. 3d). Two to three grams of material from the top 2 cm of each piston core was analysed. The relative coccolith abundance in each sample was determined by polarizing light microscopy. The coccolith percentage per field of view at 1000 X magnification was then determined according to Winter et al. {1979). Organic matter was removed from all sampies which were wet sieved into three size fractions ( < 6 3 / l m , 63-150/~m, and greater than 150 /~m). These fractions were then dried at 50 ° C and weighed afterwards, coccoliths were centrifuged from the fine fraction. Approximately 300 coccoliths from each stub were counted and identified by SEM and the percent abundance of each species calculated. The degree of coccolith preservation in each sample was described according to a preservation index (on a scale from one to five; see Fig. 4d). Whenever possible, the percent of carbonate from each sample was determined from approximately 1 g of crushed, dry, organic-free material. A Karbonat bombe was used and the procedure of Muller and Gastner (1971) and Birch ( 1981 ) was followed. Six samples (no. 26, 45, 73, 97, 99 and 100) consisted of less than 0.25 g of sample and were not considered for carbonate analysis, since use of quantities of sample less than this resulted in large errors in carbonate percentage calculations. The > 150 /~m size fraction was examined under a binocular microscope. The fraction was then split using a dry sediment splitter. Between 300 and 600 planktonic foraminifera

329 '

'

'

'

'

'

'

'

'

'

t~/~/(c:)~

' ~'~.,'-7~7~, ~

,??;

°

~

.

Durb~

.


<:

ID

Fig. 2. The study area. Bathymetric contours are in 250 m intervals, core-top sample locations are numbered. AB = Almiranti Leite Bank; LC-Limpopo Cone; CT= Central Terrace; IT= Inharrime Terrace. (both whole and incomplete specimens) were counted from the split and the relative abundance of foraminiferal fragments from the split was estimated, Globigerinoides sacculifer specimens were chosen for isotopic determination since they were a b u n d a n t in most samples examined and since they have been shown by Erez and Luz (1983) to deposit their carbonate shells in isotopic equilibrium with sea water over a wide range of temperatures. About 0.5 mg of well preserved G. sacculifer were weighed from the

212-300 # m fraction (equivalent to about 45 specimens). The method of isotopic determination followed that of Shackleton (1974). Instrumental corrections of the results were calculated according to Craig (1957). The results are reported in per mil deviations relative to P D B and calibrated with standard NBS-20 according to the equation: is 18 180__ ( ( O / ~O)sample 1 × 1000 In order to compare our results with those of

330 Vincent and Shackleton (1980), temperature estimates of oxygen isotope values were calculated according to their equation:

T= 16.9- 4.38 (xso c - 1sOw) + 0.1 (lsO c - 1sOw) 2 Where lsOc is the oxygen isotope composition of the carbonate material, and 1so w is the oxygen isotope composition of water (0.3 p p m PDB ) in which G. sacculifer lived, Results The core-top material consisted mostly of nannoforaminiferal ooze (Kennett, 1982) in the north of the study area (sectors M and D; Fig. 3d) and foram-nannofossil ooze in the south (sectors EL and PE). Glauconite, quartz and biogenic silica were found in sediments in a narrow band at the base of the continental slope and down the length of the Natal Valley (Fig. 2). Sediment textures range from muds to muddy sands with samples in the north being generally coarser t h a n elsewhere. However, a particularly sandy group of samples was found in the Agulhas Passage. Fine material was conc e n t r a t e d o f f E a s t L o n d o n ( s e c t o r E L ; F i g . 4a).

Preservation of core-top material Three related parameters: Carbonate percentages (Fig. 4b) relative coccolith abun' dances (Fig. 4c ) and coccolith preservation (Fig. 4d) have been studied in order to ascertain the condition of the material in the study area. Carbonate percentages in the sediment surface samples varied greatly, ranging from 0% to 86% with an average of 46%. Relative coccolith abundance in each sample varied from absent to abundant. Most samples contained numerous coccoliths which under SEM inspection were generally well preserved though some were occasionally dissolved, Generally carbonate percentages and coccol-

ith preservation improve equatorwards whereas relative coccolith abundances increase in the opposite direction. In the region of the Agulhas Passage (sector PE) and at the base of the continental slope off Durban (sector D) sediments are mostly below the CCD. A number of factors seem to indicate that the lysocline begins at about 3500 m in the Agulhas Passage. These are the decrease between 3500 and 4000 m in relative abundance of two coccolith species, C./eptoporus (Fig. 5b) and E. huxleyi (Fig. 5d); the increase in foraminiferal fragments (Figs. 5c and 8a) and the decrease in carbonate percentages (Fig. 5a). Poorest coccolith preservation and lowest carbonate percentages were also recorded from this area. Furthermore, fine fraction percentages (Fig. 4a) and relative coccolith abundances are low, reflecting the corrosive action of the ABW which is forced through the Agulhas Passage at depths of about 4000 m. Off Maputo (sector M) where the sea floor is well above lysocline, coccoliths were very well preserved and percent carbonate values were elevated, whereas highest carbonate values were concentrated on the flanks of and at the base of the Mozambique Ridge (sectors D and EL) between 2000 and 4000 m (Fig. 4b ). However, foraminiferal fragments (Fig. 8a) contribute between 40 and 60% of the coarse fraction ( > 150 #m) found in the sediments off Maputo. Relative abundance of coccoliths generally increased in abundance with southerly latitude (Fig. 4c). Highest abundances were recorded off East London (sector EL), and low values were observed in the Agulhas Passage south of Port Elizabeth (sector PE) and in shallow water in the North (sector M).

Taxonomic notes The two "ecophenotypic variants" of Emiliania huxleyi, a "cold-water" form with fused blade-like elements in the proximal shield and a "warm water" variety with T-shaped elements in both shields (Winter, 1985 ), have not been separated into two groups, but have been placed together since it has been shown that

•~op q~!ta po~iaeul eas suo!~sao I aldureS -~ao'~aa~anoj o~u! pop!a!p oae soldUXeSuo!ssnas!p jo o~o aoj :suo!s!^!p ao~aaS .p •u~oa0 Lre!puI ~so,'aq~nosaq~jo aooB ~os aq~ uo se~elqu~o~e q~!ioaaoa ao.~maooaq~ aq~jo uo!~nq!a~s3.(i"a "sI!~Op aoj ~xo'4oa8 •~uoaano s ~ q l ~ V aq~ zapun ~oau~eaadm~ pea~ap aaq~!q ~ o H "op~a~.~uoosoa~ap m. ua^t8 oas s ~ n ~ z e d ~ ~!do'~os!pe^.u -a(i "Ou.u aa~aeI/~l~m.s'ea~ ,~q po~uaso~doa (sazn~eaadu~ aou~ae~) uo!~aldap a!do~osI .zoo[3 ea~ aq~ uo aaBfnao~a sap?ou.~ -aa~.ctOIDjo (l!ua aod s'4.red '4t(IcI 0 ~ ) so!~ea odo~os! u ~ x 0 .q "£~!suaom.~uoaana so~ouap oz!s ,~oaav "mndeI~1340~u!Ilo~dn ~obI .~oae~u! ~ ! i i o ~ ~uoa~poaaa~m.~ ~,86! '9! q~xeI~Io~ ~,06I '~ "uef po.uacIaq~ aa^o q ~ d ~uaaan0 s~qin~ ueaIA[ "e "8 "~!,~[

SE

..........

SE

N-

0$'

OZ

cj~

~

. . . . .

. . • • • ""

•" ,

/

OP

.- ...........

. . . . . . . . . •

o~

, !.,v: ~ . . . . . . .

i" •

£E _ _

30'

..~i::" 0

O~

£2~

NNNN

0$'

gE:

o6aIq~eeeV ~

aa~oM

i7,i

I

.

~

~ ~ .

I

.Z, I

": .:;:i;';:!'"' ' I .

,.:<.::

.

.

.

• ': ~::;': . " t' • ' •" "

;!~

¢

,,s,:q'q,~,,.,ov ~

-

1

'

OE:

o~ ~ue,.-lJn

i

.

ii... iii;

£~

£E

~'!':'

OE

0

eoq[n I q~e

I O N 3 0 3~

~z

o#

OS

eEa

s~6o, [ q,,,eoe v q-,., I. ! o o o o 3 ..

£E: ov ~

£6

o#

£8

£Z

OE

£~

I£B

332

25

30

35

25

.......

30

MI

35

U

I

.:#',,'~

/,,~-,--~)

40

_

~

I

,~/',,"~.

~I 3oPll

35

M 30

/ W 35

^ ~

~

25

4 30

~.,---~'°

0

25

:;~.!.f/'_ _

40

90

35

25

4O ~ 25

~

4 • .........

MI

I

35

30

35 ....

/ W 35W

~

25

30

A 30MI

L~

25

3o

25 25

.....

o

Fo~r

!~,-/,x~.,

0

25

-

Pl 3 o

~ 25

4O 30

~5

Fig. 4. a. Percentage of material in each sample from the sediment-surface < 3/~m. b. Percentage of carbonate material in samples, c. Relative abundance of coccoliths in each sample as determined by light microscopy; absent= < 5%; rare= 530%; common = 30-60%; abundant = > 60%. d. Preservation of coccoliths in each sample: Excellent = no evidence of dissolution or breakage; Good=some etching of the more solution susceptible species eg Syracosphaerids; Fair=etching and breakage of more fragile coccolith species; Poor = extensive etching; central grills and bridges often absent even on dissolution resistant species, eg Emiliania huxleyi, Gephyrocapsa oceanica and Calcidiscus leptoporus; Very poor = extreme etching and breakage, some species develop into unrecognizable rings, large numbers of coccolith fragments.

•:l•

Percent

Abundance

of Foram

Fragments

Percent

Calcium

Carbonate

(Tq"

o 0

O I:=

0

i

i

i

o i

t

o I

o i

o i

o

o

i

o

~ i

0

01

o i

o

o !

01

i

i

=

i

!

~

1

o

i

-,I

~'~

=

'= .

~

==

=

~

- -

g .~

Percent

o ~¢~

o i

o

o i

/~und=nce

o i

o I

of E.huxleyl

o i

o i

o i

Percent

o i

O i

o

o

o

o

o

i

Abundance

o

i

u

i

O

i

of

o

!

C.leptoprue

o

i

o

i

i

O

t~

E4

®

t~

~t~

O

i

® O Io

~

~ " ~ j t ~ IO K~

~

M

,

®

~

p

u~

UI

334

there is little correlation between placolith type and temperature, and their ecological affinities have not been confirmed (Winter, 1985). The orientation of the central bridge is a crucial factor in distinguishing between Gephyrocapsa oceanica and Gephyrocapsa caribbeanica, G. oceanica' s bridge is perpendicular to the long axis of the central area, while that of G. carribeanica is obliquely aligned. Specimens which were similar in appearance to one of these two species, but which were lacking a central bridge or lying on the SEM stub proximal side upwards (bridge obscured) were probably mostly G. oceanica specimens since identifiable G. carribeanica comprised less t h a n 1% of the assemblage composition in our samples.

Calcidiscus leptoporus and Umbilicosphaera sibogae (Fig. 6). Each of these species contributed at least 10% to the total assemblage composition and together they were responsible for about 86% of the total coccolith assemblage (Fig. 6). Four species: Helicosphaera

carteri, Rhabdosphaera clavigera, Syracosphaerapulchra, and Gephyrocapsa ericsoni each made up between 1% and 10% of the assemblage composition and contributed, together, about 7% to the total coccolith assemblage. Thirty-six species (some relict), each with an average abundance of less than 1%, contributed to the remaining 7% of the total coccolith population in the samples.

Major species Assemblage composition Thirteen of the one hundred core-top samples studied by SEM contained very few or no coccoliths and were therefore excluded from further investigation (Fig. 4c). Forty-four identified and fifteen unidentified coccolithophore taxa were recognised from the core-tops (Table I). Thirty-four of the identifiable taxa are modern and ten extinct. Twentysix species of coccolithophores found in the water column of the southwest Indian Ocean by Friedinger and Winter (1985) are represented in the core-top samples by their coccoliths (marked with an * in Table I). Six coccolithophore species known to live in the oceans at present (Coccolithus pelagicus, Crenalithus ses-

silis, Gephyrocapsa protohuxleyi, Discolithina japonica and Calyptrosphaera papillifera) were found in core-top sediments but are absent from water samples. Most of the species present in the water column but absent from the sediments were relatively fragile and probably dissolved on the journey to, or once at, the sea floor, Coccolith assemblages from the southwest Indian Ocean are dominated by epipelagic, placolith-bearing species. Four species dominated the coccolith assemblages at the sediment surface: Gephyrocapsa oceanica, Emiliania huxleyi,

E. huxleyi occurs in all but one sediment surface sample (no. 13) and contributed up to 50% of the coccolith assemblage. The relative abundance of E. huxleyi in core-tops increases equatorwards (Fig. 7a), reaching lowest values south of Port Elizabeth. High values on the Limpopo Cone and Central Terrace (sector M) are separated by a narrow band of low values lying beneath the path of the Agulhas Current, running northeast to southwest. Highest percentage occurrences are found at depths less than 3500 m. G. oceanica is present in all of the samples. Its maximum relative abundance is about 92% and its mean percentage of the total population is about 35%. Highest concentrations of G. oceanica occur on the Limpopo Cone off Maputo, decreasing with easterly longitude (Fig. 7b ). High abundances of G. oceanica also occur in the Agulhas Passage (sector PE), and low percentages are concentrated east of Durban. G. oceanica and E. huxleyi, the two dominant species in the study area, are also the two most important coccolithophore species in tropical to subtropical waters (McIntyre and B~, 1967). Their biogeographical distribution is of great interest since they are the two most abundant coccolithophore species in Quaternary sediments. E. huxleyi is the most abundant and

335 TABLE

I

Taxonomic

list of extant and extinct coccolithophores

Most recent or best available micrographs column

identified from the sediment surface of the southwest

are quoted as reference. Coccolithophore

Indian Ocean.

species that were present in the water

(Friedinger and Winter, 1987) are marked with an asterix. K i n g d o m P L A N T A E ; Division C H R Y S O P H Y T A Pascher,1914; Class P R Y M N E S I O P H Y C E A E Manton,1964; O r d e r COCCOLITHOPHORALES Schiller,1926. C E R A T O L I T H A C E A E Norris,1965. * 1.Ceratolithus cristasus K a m p t n e r , I950 ................................................................Wang and Samtleben,1983.PI.2(164). COCCOLITHACEAE Kamptner,1928. 2.Coccolithus pelagicus (Wallich),1892 .................................................................Siesser,1975.Pl.6(e,f). 3.Crenalithus sp.cf. Crenalithus sessilis ( L o h m a n n ) , 1 9 1 2 ................................. Nishida,1979.Pl.3(3). * 4.Calcidiscus leptoporus ( M u r r a y a n d Blackman),1890 ..................................... Wang a n d Samtleben,1983.Pl.l(10). * 5.Emiliania huxleyi ( L o h m a n n ) , 1 9 0 2 ....................................................................Winter et al.,1979.Pl.l(3). * 6.Geph),rocapsa caribbeaniea B o u r d r e a u x and Hay,1967 ................................... Winter, 1982.Pl.l(5). * 7.G. ericsoni M c l n t y r e a n d B~,1967 ....................................................................Winter et al.,1979.Pl.l(6}. * 8.G. oceanica K a m p t n e r , 1 9 4 3 ................................................................................O k a d a and McIntyre,1977.Pl.3(l). * 9.G. ornata H e i m d a l , 1 9 7 3 ......................................................................................O k a d a a n d McIntyre,1977.PI.3(l). 10.G. protohuxleyi M c l n t y r e , 1 9 6 9 ...........................................................................Winter et a l . , 1 9 7 $ . P l . l ( l - 6 ) . *ll.Oolithus fragilis ( L o h m a n n ) subsp, cavum O k a d a and M c l n t y r e , 1977 ....... O k a d a a n d Mclntyre,1977.PI.4(4,5). * 12.Umbilicosphaera hulbertiana G a a r d e r , 1 9 7 0 ........................................................Wang and Sarntleben, 1983.P1.1 (13). *I3.U. sibogae ( W e b b e r - v a n Bosse),1901 .................................................................O k a d a a n d Mclntyre,1977.PI.4(2). H E L I C O S P H A E R A C E A E Black,1971 emend. J a f a r and Martini,1975. * 14.Helicosphaera carteri (Wallich), 1877 ...................................................................Nishida, 1977.PI.9(4). *I5.H. pavimentum O k a d a a n d M c l n t y r e , 1 9 7 7 ........................................................O k a d a and M c l n t y r e , 1977.PI.4(6,7). PONTOSPHAERACEAE Lemmermann,1908. 16.Discolithina japonica T a k a y a m a , 1 9 6 7 ................................................................O k a d a a n d Mclntyre,1977.PI.6(3). 17.Pontosphaera discopora Schiller,1925 .................................................................Borsetti a n d Cati,1979.PI.2(3,4). *I8.P. wracusana L o h m a n n , 1 9 2 0 ...............................................................................O k a d a a n d Mclntyre,1977.PI.5(7). R H A B D O S P H A E R A C E A E L e m m e r m a n n in B r a n d t a n d Apstein,1908. * 19.Acamhoica quattrospina L o h m a n n , 1903 ..............................................................Nishida, 1979.P1.13( 1). *20.Discosphaera tubifera ( M u r r a y a n d Blackman),1898 ....................................... Nishida,1979.Pl.l 3( 1). *21.Neosphaera coccolithomorpha L e c a I - S c h l a u d e r , 1 9 5 0 .......................................... Wang a n d Samtleben,1983.PI.2(4). *22.Rhabdosphaera clavigera M u r r a y and Blackman,1898 .................................... Winter et al.,1979.Pl.12(l). *23.Umbellosphaera irregularis Paasche,1955 ...........................................................Wang a n d Samtleben,1983.PI.2(2). ' 2 4 . U . tenuis ( K a m p m e r ) , 1 9 3 7 ...................................................................................Winter et al.,1979.Pl.lll(l,2). SYRACOSPHAERACEAE Lemmermann,1908. *25.Anoploselenia brasiliensis ( L o h m a n n ) , 1 9 1 9 ........................................................Borsetti a n d Cati,1972.Pl.16(la,b). *26.Syracosphaera corolla Leccal,1965 ......................................................................Borsetti a n d Cati,1979.PI.3(5). "27.S. lamina L e c a l - S c h l a u d e r , 1 9 5 1 ..........................................................................Nishida,1979.PI.8(3). "28.S. mediterranea L o h m a n n , 1 9 0 2 ...........................................................................O k a d a a n d Mclntyre,1977.Pl.10(4,5). "29.S. mediterranea L o h m a n n var. binodata K a m p m e r , 1 9 3 7 ............................... O k a d a a n d Mclntyre,1977.Pl.10(6). "30.S. pulchra L o h m a n n , 1 9 0 2 ....................................................................................O k a d a a n d M c l n t y r e , 1 9 7 7 . P l . 1 0 ( l l , 1 2 ) . 31.Syracosphaera sp.cf, variablis (Halldal and Markali),1955 ........................... O k a d a a n d Mclntyre,1977.PI.9(7,8). C A L Y P T R O S P H A E R A C E A E B o u d r e a u x and Hay,1969. 32.Calyptrosphaera papillifera B o u d r e a u x and Hay,1969 .................................. Nishida,1979.PI.23(3). G E N U S O F U N C E R T A I N O R I G I N (lncertae sedis). 33.Hyaster perplexus (Bramlette a n d Riedel),1954 ............................................. Wang a n d Samtleben,1983.PI.2(l). T H O R A C O S P H A E R A C E A E , Schiller,1930. *34.Thoracosphaera cf. T. heimii ( L o h m a n n ) , 1 9 1 9 ................................................Winter et aI.,1979.PI.V(II). D I S C O A S T E R A C E A E Vekshina,1959. 35.Discoaster broweri Tan Sin Hok,1927 ................... ;...Bukry,1971.PI.3(2). R a n g e : L a t e Miocene to late Pliocene. 36.D. exilis Martini a n d Bramlette,1963 ....................... Martini a n d Bramlette,1963.Pl.104(l-3). Epoch:Middle Miocene. 37.D. deflanderi Bramlette a n d Reidel,1954 ................. Bukry,1971.PI.4(4). Range:mid Eocene to mid Miocene. 38.D. siapanensis Bramlette a n d Reidel,1954 ................ Haq,1980.PI.23(f). R a n g e : m i d to late Eocene. OTHER REWORKED COCCOLITHS 39.Dictyococcites daviesi (Haq),1976 ............................... Haq,1976.PI.XI(11). Epoch:Cenozoic. 40.Ericsonia cava ( H a y a n d Mohler), 1967 .................. P e r c h - N i e l s e n , l . H a y a n d Mohler,1967. 41.Pseudoemiliama lacunosa ( K a m p t n e r ) , 1 9 6 3 ............... Sarntleben,1978.PI.3(l). 42.Aperlapetra gronosa (Stover),1966 .............................. Bukry,1970.PI.6(6). R a n g e : A l b i a n - C a m p a n i a n . 43.Prediscophaera cretacea ( A r k h a n g e l s k y ) , 1 9 1 2 ........... Haq,1980.Pl.l 5(d) a f t e r Buckry,1969. R a n g e : A l b i a n - C a m p a n i a n . 44.Cyclicargolithus floridanus ( R o t h and H a y ) , 1 9 6 7 . . . B u k r y , 1 9 7 1 . Roth,1973.PI.6(2+3). R a n g e : E o c e n e - M i o c e n e .

2

;

10

U.llbogoe

.

.

.

.

.

.

.

.

.

.

.

.

.

Fig. 6. Stacked bar graphs for each sector, showing the relative abundances of the four most abundant species (Gephyrocapsa oceanica, Emiliania huxleyi, Calcidiscus leptoporus and Urabilicosphaera sibogae ), total minor species and reworked/unidentified coccoliths in each sample.

~

Reworked ¢occollthe

31 32 3~3 ,34 35 36 37 38 40 41 4~2 4~B 49 67 60 69 74 66 75 Sample Number

/

//

.

~0~ Sector 4 ~, (66+75)

L~;X3 Minor l p e c l e i

o

['1'

15116171~192~0272~8297'7707~OO81828'3848'~Br.875889909192939495969798 Sample Number

~ ~

o

a , n , i , ~ ' I ' i ' l ' l ' t ' t ' l ' l ' l '

Reworked coc¢ollthl

C.leptopOnJl

~' ~'

I~

43414 15:6;7~0511 52535455565758,596"061 ' i ' I ' J ! i ' 6"2836"470 J ' i ! 7172 J ' 737699100 I ' I Sample Number G.oceenl©o ~ E.huxleyl ~ C.Ioptoporue

/

Sector 2

Minor l p o ¢ l l l

~

1'3 14 2' 22 23 2'4 25 21~

Sample Number Ir.huxleyl

1°"

;

1o

~

~

7

,

//

;

so

Soctor 3

//

5

~

I O0

20 -

4

.....

m

20

U.llbogae

17"7J G.ocoonlco

,o

Sector I

~

O~

337

30

25

35

25

30

25

.........

MI (

35 W

~ 80

oercQntl

~

I

'~fE/~-

~..~x

/

°

I W 35W

30

~ "

•l: ~

~ 3O

~

"~

I W 35

o

oOl_bj

3Q

35_

o 2D percent 0 4D percent { ~ 60 percent Q 8D percent

U. s i b o g o e o 20 percent 0 40 percent ~ 60 pQrcent 0 8D percent

/ ~ ...;::

~P

°~

. ,.~::~

...>,~

Z

o ~

co

35

.,~,~

/

o~

35

30

o ~

.. ~

"~p o~

.

-:~

4 30

-ii S ~

o

~

: ~ 25

40 25

35

LEGEND, C. I e p t o p o r u s

40

o~r c ~nt

25

40 25

~

)80

I

" W

40 ~

30

35 .......

M 301~1 (

"~

"7

30

25

•

. /

o

0

35

~ 25

40 30

35

Fig. 7a. Relative percent abundance in each sample of the four major species, a. Emiliania huxleyi, b. Gephyrocapsa oceanica.

c. Calcidiscus leptoporus, d. UrnbiUcosphaera sibogae.

ubiquitous coccolithophore species living in today's oceans, occurring at relative abundances of between 60% and 80% (McIntyre and Bd, 1967; Okada and McIntyre, 1979). It has the widest biogeographic range and is one of the most euryhaline and eurythermal coccolithop-

hore species (McIntyre and Bd, 1967; McIntyre et al., 1970; Winter et al., 1983). The distribution of G. oceanica is not as well documented. It apparently has a more limited range than E. huxleyi, currently preferring warmer waters and marginal seas (Okada and Honjo, 1975). G.

338

oceanica has been reported to be abundant in upwelling areas of low latitude and in warm highly saline and fertile waters where it often replaces E. huxleyi as the dominant species (Winter, 1982, 1985; Mitchell-Innes and Winter, 1987). C. leptoporus seems to prefer subtropical waters and has a wide biogeographical range, similar to but slightly smaller than, that of E. huxleyi (McIntyre and B~, 1967; Schneidermann, 1977). This species usually contributes only a small percentage to living coccolithophores in the world's oceans (McIntyre and B~, 1967). Like G. oceanica, C. leptoporus seems to prefer high fertility waters (Roth and Berger, 1975). Winter (1982) also reported that C./eptoporus increased in abundance during presumed times of high fertility in the Red Sea. In the study area C. leptoporus has been observed in every sample, contributing on average about 15% ofthecoccolithassemblagereaching a maximum of 43%. C. leptoporus increases in percent abundances from north to south and highest abundances are recorded at depths below 3000 m (Fig. 7c), in contrast to E. huxleyi which shows the opposite trend, U. sibogae is usually found in medium to high fertility regions (Roth and Berger, 1975) and occurs in water between 18 and 24 ° C (McIntyre and B~, 1967; Okada and McIntyre, 1979). In our samples U. sibogae contributes, on average, just over 10% of the total coccolith assemblage, it is present in all but one of the samples and has a maximum abundance of 24%. Highest values are found to the north, parallel to the coast (Fig. 7d), lying in a narrow band under the mean flow path of the Agulhas Current. Maximum values occur at depths of between 1000 and 3000 m. Minor species H. carteri is most abundant in subtropical waters of medium to high fertility (Roth and Berger, 1975 ) but may extend into transitional waters (McIntyre and B~, 1967). H. carteri is

found in all but one of our samples, has a maximum occurrence of about 7%, and a mean of 2.6%. Maximum abundances are found in the southern part of the study area (sectors EL and PE), close to the continental margin. The highest relative percentage abundances ofH. carteri are found at depths between 3000 and 4000 m. R. clavigera is one of the common rhabdoliths in today's oceans, preferring subtropical and transitional waters (McIntyre and B~, 1967) and reaching highest values in low fertility waters (Roth and Berger, 1975). R. clavigera is absent from seven of the core-top samples, has a maximum relative abundance of about 7%, and a mean of 1.8%. High relative values are concentrated on the continental shelf off Durban and East London. Highest percentages of R. clavigera occur between 1500 and 2500 m. S. pulchra, the only Syracosphere species found in significant abundances in sediments of the southwest Indian Ocean, occurs occasionally in the equatorial to transitional zones of the Pacific and North Atlantic (Okada and McIntyre, 1977) and seems to prefer low fertility waters (Roth and Berger, 1975). S. pulchra is found in all but five of the core-top samples, has a mean abundance of 1.8% and a maximum of about 6%. Highest relative percentages were found in the north of the region (sectors M and D) and is most abundant at depths less than 4000 m. G. ericsoni, although less abundant, displays a similar surface water distribution pattern to that of G. oceanica (McIntyre and B~, 1967 ). G. ericsoni is also one of the most abundant species of coccolithophore presently living in the Gulf of Aqaba, suggesting a higher than normal salinity tolerance (Winter et al., 1979). G. ericsoni was absent from twenty-one of the samples, has a maximum relative abundance of about 8% but an average of only 1.1%. Highest abundances are concentrated in the north (sectots M and D) close to the coast, though some high values exist on the flanks of the Mozambique Ridge (sector EL). Generally speaking G. ericsoni becomes less abundant with depth.

339

Reworked coccoliths

Oxygen-isotope analysis

Small percentages of relict coccoliths were found in several samples, especially in the southern half of the study area, particularly in the Agulhas Passage (Fig. 8c). Four discoaster species and eight extinct coccolithophore species were identified. An additional fifteen forms were too dissolved and broken to be identified, Small numbers of discoasters, extinct since the Tertiary (Bukry, 1971), were found in seven samples. Concentration of discoasters in any one sample was only about 4% and it is therefore unlikely that these samples originated from outcrops of older material, especially since they contain high percentages of relatively modern species. Extinct coccolith species were also found in a few samples, but usually in extremely low abundances. Sample 67 (Sector PE, in the extreme southwest of the study area) is anomalous since it contains abundant (24%) C. pelagicus as well as a large concentration of relict coccolith species such as C. floridanus (39%),E. cava (11.1%) andD. daviesi (5%). C. pelagicus is a robust species which first appeared in the Upper Paleocene and its co-occurrence in samples with species that ranged through the Tertiary and Cretaceous suggest sediment reworking. Samples containing unidentified coccoliths were concentrated in the Agulhas Passage (Fig. 8d) and are probably from turbidite or relict sediments that have been scoured by ABW.

Specimens of Globigerinoides sacculifer were rare or absent from samples taken from depths greater than 4000 m. These samples were mostly located in the Agulhas Passage and could not be considered for isotopic analysis (Fig. 8b). Isotope ratios and derived temperatures are plotted geographically in Fig. 3b. Ninety percent of the samples had isotope ratios in the range - 1.5 and 0.0%o. Eight of the eighty-two samples run have isotope values that fall outside of this range. These anomalous values were found at depths in excess of 3500 m or in unstable areas, such as in the paths of mass gravity flows. Derived temperature values ranged from l8.5 to 25.1°C. This range of values is close to surface-water temperatures in the Agulhas Current region (Wyrtki et al., 1971). Warmest isotopic temperatures lie in a narrow band located parallel and close to the coast and temperatures decrease with distance from the coast a n d t o w a r d s t h e south.

Foraminiferal counts Samples containing high relative percent abundances of foraminiferal fragments are located in the Agulhas Passage (sector PE) and in shallow water sediments east of Maputo. Relative percent abundances of foraminiferal fragment are low off Durban and East London, with distance from the coast (Fig. 8a ) and with depth (Fig. 5c). Lowest values occur between 2000 and 3000 m. However, a sharp increase in fragments occur at about 3650 m.

Coccolith core-top assemblages The area of this study encompasses approximately 15 ° latitude (25-40 ° S) and should include subtropical (10-30 ° S ) and transitional (30-45 ° ) coccolith floral zones similar to the Atlantic and Pacific Oceans (McIntyre and B~, 1967; Okada and Honjo, 1973). Thus coccolith assemblages at the sediment surface from this region would be expected to differ from north to south as ecological ocean boundaries are crossed. Three different coccolith assemblages have been recognized from the study area, although some overlap occurs (Fig. 3c): Maputo, Agulhas Current and Deep Water (Table II). Each is associated with a specific oceanographic area and conditions. However, preferential dissolution and integration of seasonal fluctuations in our samples are probably the reasons that the

340 25

30

35

25

30 PII

4D ~

I

.

.

.

.

.

0~

I

cocco]iths

were

?ound.

~

i

7~

::

!

½cJl

"

35 •

0-1%

.

4D

30

Samp1e~ whQre umidentIfiad

Extinct Spp. Extinct Sppo ]-10 .

R 3D

::::i:.~f/'

4~

Ext|nct Spp. >IOZ .

25

~ 25

ReworNQd cocco]Ithm 0

=xolanatton.

0

35

LEGEND,

30

on

4

30

35 .......

M 3ORl?nr

.!.!~fo

25

30

25 25

_

30

l

oF~ 1

25 3o as 25 30 ~s Fig. 8a. Percent foraminiferal fragments in each sample from the greater than 150 #m fraction, b. Core-tops in which not enough foraminiferal specimens were available for isotopic determination (30-1,35,37-41,48-9, 67,69,73,78 and 97-100) or samples which gave isotopic ratios outside the glacial-interglacial range (no. 32, 36, 57, 58, 61, 76 and 87) due to poor reservation of sample, c. Distribution of reworked coccoliths, d. Distribution of unidentified coccoliths. Core-top locations are marked by sample numbers. three assemblages do n o t c o n f o r m to the floral z o n e s of M c I n t y r e a n d B6 ( 1 9 6 7 ) . T h e M a p u t o assemblage is f o u n d o n the shal-

low water c o n t i n e n t a l shelf, inshore of the A g u l h a s Current (sector M ) a n d is characterized by G. oceanica (averaging 42 % ), E. huxleyi

341 TABLE II

sibogae contribute about 15% and 8% respec-

Sea-surfacetemperatureand phosphatevaluesforthe Maputo,Agulhas Currentand Deep Water assemblages(after Wyrtki et al.,1971). Mean valuesof the relativepercent abundances of the significantcoccolithophorespeciesare also given, co~ ~ ~(°C) _~ ~(Mic ats/l) ~ la~ Assemblage Composition ~ ~ ~.~ species ~ 1~ 21 25 0.2 <0.8 G . o ¢ ~ 41.9~

tivelytowards this assemblage. Since all of the deep-water samples occur at or below the lysocline (3500-4000 m), dissolution is probably the major influence determining assemblage cornposition. The most abundant species found in the assemblage, G. oceanica and C. leptoporus, are also the most dissolution resistant (Berger, 1973)

SLmmer winter

25 21

26 22

>0.2 0.6

0.4 <0.8

20 24 20

26 26 22

<0.2 <0.2 0.4

0.6 0.2 0.6

Mean Depth = 925m 2 A ~ CLVRRI~P SLmvaer Winter

De~ : 2143m 3 D ~ P WATER SIml~er Winter

DePth= 3653m

E.huxlevi 32.5 % U.sibocrae 11.5 % C. leDtoDorus 5 . 1 % G.e~icsoni 1.7 % E.htDclevi G.oceanica C.le~rus U. siboc~e

S.~-~

15 19

23 23

<0.2 <0.2

0.6 0.2

15

19

0.4

0.6

31.2 % 50.0 % 14.3 % ll.1%

2.5 ~

"

Discussion Sedimentology

G.oceanica 34.8 % ¢.~eotoDorus 16.5 %

U.sit~uaeE'huxlevi25.310.5* ~.¢~i

3.1 ~

(33%), and lower concentrations of U. sibogae ( 12% ), and C. leptoporus (5 % ). The high abundance of G. oceanica in this area is probably related to inshore return currents and the associated upwelling of nutrient rich waters. The controlling oceanographic features of this area are the shallow water depths, low water temperatures (where there is upwelling) and the proximity of the warm Agulhas Current and its associated gyres, The Agulhas Current assemblage is restricted to a narrow offshore band parallel to the coast, directly beneath the mean flow path of the Agulhas Current through sectors M, D and EL. E. huxleyi dominates the assemblage with an average abundance of 33%, G. oceanica only comprises 28% of the assemblage (10% lower than in the other two assemblages) whiie C. leptoporus and U. sibogae occur at average abundances of 13% and 12 % respectively, The deep-water assemblage is situated in sectors E L and P E at depths greater than 3500 m, with the deepest samples being found in the Agulhas Passage (4879 m). G. oceanica occurs at abundances of about 41%, C. leptoporus is the next most abundant species, occurring at an average abundance of 23% (five times more abundant than in the north). E. huxleyi and U.

Carbonate percentages tend t o i n c r e a s e with distance from the southeast African coast as well as beneath the main flow path of the Agulhas Current. These trends are probably due to the reduced dilution by terrigenous material (Fig. 4b) and by the increased role played by the bioclastic sediments. Off the east coast of southern Africa continental shelf sediments are either terrigenous on the inner shelf or bioclastic on the outer shelf. Shelf sediments are moved to the deep-sea floor initially by the activity of the Agulhas Current system which transports sediments along the outer shelf (Flemming, 1981; Martin, pers. comm., 1986). When submarine canyons are encountered, these sediments are diverted seawards and transported by turbidity currents to the Natal Valley (Flemming, 1981; Dingle and Robson, 1985). The abundance ofcoccoliths which are found in the fine fraction of sediments may be reduced by winnowing. This results in the accumulation of sandy lag deposits (Goodlad, 1986) and is a process which influences the distribution and sorting of marine sediments. Sandy lags have been correlated with high velocity currents by Huang and Watkins (1977). Goodlad (1986) found that the distribution of the greater than 150/Lm fraction in sector D showed a close relationship with bottom current flow. Sectors M and D display lowest concentrations

342

of fine material (coccoliths included) and may be explained by the winnowing effect of the Agulhas Current which can erode surface sediment material down to depths of 2500 m (Duncan, 1970). Low coccolith abundances were also found in the Agulhas Passage (sector PE), probably as a result of winnowing by ABW, however, the increase of coccolith abundances in some of the deepest samples in the Agulhas Passage is puzzling (Fig. 4c). Perhaps these samples originated from Recent turbidites that were buried quickly and the coccoliths were thus protected from dissolution, With the exception of these low values in the Agulhas Passage, coccolith and fine fraction abundances increase with southerly latitude (Fig. 4a,c ). This is probably due to diminished winnowing associated with increasing depth of sea floor, Fine fraction percentages and coccolith ahundances are greatest in sector EL probably because the sea floor is too deep to be winnowed by the Agulhas Current and mostly too shallow to be influenced by the ABW.

Coccoliths Two species, E. huxleyi and C. leptoporus, display significant differences in percent abundances from sediment surface samples from the north to the south of the study area. E. huxleyi decreases in relative percent abundance from north to south while C. leptoporus shows the reverse trend. Temperature and salinity are thought to be two important environmental factors controlling the distribution of these two species. However, the temperature range of 15 and 26°C in the southwest Indian Ocean (Wyrtki et al., 1971) is well within the tolerance limit of both species (Okada and McIntyre, 1979). Sea surface salinities vary only slightly, ranging between 35 and 35.6%o (Wyrtki et al., 1971 ). These salinity ranges are also well within the tolerance limits for both E. huxleyi and C. leptoporus, Phosphate concentration in seawater may play an important role in influencing the dis-

tribution ofcoccolithophores (Reid, 1962; Krey and Babenerd, 1976; Winter, 1982). Phosphate values in the study area range between about 0.2 and 1.0 ttg at/1 (Wyrtki et al., 1971 ). Generally values increase from north to south with occasional pockets of relatively high phosphate concentrations associated with dynamic upwelling off Maputo and Durban (Martin, 1984). The strong nutrient gradient from north to south and the increase of C. leptoporus (Fig. 7c) in the same direction, to the detriment of E. huxleyi (Fig. 7a), seems to indicate fertility control in the distribution of these species. Other coccolithophore species which have been reported to be sensitive fertility indicators such as G. oceanica and H. carteri (see community structure section ) are also found in areas of high nutrient concentrations in the study area. G. oceanicais most abundant in the shallow sediments off Maputo. Both species are abundant in the deeper, southern part of the study area. The high abundances of G. oceanica off Maputo support the affinity of this species for warm waters and high nutrient concentrations (Winter, 1982, 1985). The ocean floor distribution pattern of E. huxleyi and C. leptoporus may not only be due to latitudinal environmental changes occurring in the water column but also to post-depositional dissolution. This is because there is a distinct sample-depth gradient from north to south, ranging from 248 m off Maputo to 4879 m south of Port Elizabeth. The coccoliths of C. leptoporus and E. huxleyi are both relatively dissolution resistant. However, C. leptoporus is somewhat more resistant to calcium carbonate dissolution than E. huxleyi (McIntyre and McIntyre, 1971; Berger, 1971; Roth and Berger, 1975 ) probably due to its larger and more tightly packed elements (Schneiderman, 1977). C. leptoporus is concentrated in samples below 3500 m (Figs. 5b and 6). On the other hand, E. huxleyi seems to remain relatively abundant in sediments shallower than 3500 m (Figs. 5d and 6), but decreases considerably in relative percent abundance at depths greater than this. Disso-

343 lution, therefore, seems to play a more important role in samples taken from the southern part of the study area where depths are greatest but cannot explain the increase in the abundance of C. leptoporus with depth (Fig. 5b) in the majority of samples taken from the northern two-thirds of the study area where we believe environmental factors play a more important role.

Core-top age determination Emiliania huxleyi is believed to have evolved from Gephyrocapsaprotohuxleyi somewhere between 270,000 and 250,000 y B.P. (McIntyre, 1969) and became abundant in the late Pleistocene between 73,000-85,000 y B.P. (Stages 4-5a: Thierstein et al., 1977). Thus samples containing abundant E. huxleyi should be younger than 85,000 y B.P. Indeed, all but one of the core-top samples taken in the study area contain E. huxleyi indicating that samples should all be younger than 270,000 y B.P. Most of the samples, especially those from the north of the study area, contain relative percent abundances of E. huxleyi between 30 and 52% and can be assumed to be less than 85,000 y B.P. In the south samples from the deepest region (the Agulhas Passage ) contain relative percent abundances of E. huxleyi of less than 10%, although the majority of samples contain relative abundances in the 10-30% range. Since samples from the south are generally from greater depths, dissolution probably diminishes percentages of E. huxleyi, thereby biasing results in favor of C. leptoporus and other resistant species. Had these samples been well preserved, E. huxleyi might have made up a larger fraction of the coccoliths identified, yielding values representative of 85,000 y B.P. or younger,

Oxygen isotopes Seven samples (no. 32, 36, 57, 58, 61, 76 and 87) for which isotope ratios were evaluated gave anomalously high or low ratios outside the tem-

perature range expected for Holocene samples (Dansgaard, 1964) for G. sacculifer. The locations of these cores (Fig. 8b ) in areas of accelerated dissolution and winnowing (in the Agulhas Passage) or at the bases of steep topography indicate that these samples may be highly dissolved or contaminated with glacial material ofPleistoceneage (seebelow). Remaining isotope ratios from core-top samples taken in the study area range from - 1.5 to 0.0 parts per mil PDB. Lightest values of - 1.49 to - 1 . 0 parts per rail PDB (equals 23-25°C, Vincent and Shackleton, 1980) occur in a narrow band on the sea floor beneath the "A" route of the Agulhas Current, in the northern part of the study area (Fig. 3b). These values are about 0.5 parts per rail PDB lighter than samples analyzed on either side of this band. The isotopic temperatures fall within the summer temperature range of the Agulhas Current, 20-26°C (Wyrtki et al., 1971 ). The 0.5 parts per mil enrichment in isotopic ratios of G. sacculifer specimens underneath the Agulhas Current can be explained by its elevated temperature at the ocean surface of between 2-3°C (one per mil change in 1so is equivalent to approximately 4 ° C; Emiliani, 1955 ). Thus an oxygen isotope imprint of the Agulhas Current exists beneath it on the sea floor. Isotopic temperatures of G. sacculifer lying beyond the influence of the Agulhas Current range between 20.5 ° and 22.8 ° C and also reflect the summer average for sea-surface temperatures. This would seem to indicate that G. sacculifer is more productive in the summer months and is in agreement with the findings of Vincent and Shackleton (1980). South of about 28 ° S the occurrence of isotopically light samples decreases. This may be a result of decreasing temperatures of the Agulhas Current as it flows southward or due to carbonate dissolution. Dissolution of less resistant foraminifera species such as G. sacculifer tends to make them isotopically heavier (Berger and Killingley, 1977) and many specimens of this species from depths greater than 3500 m were partly dissolved.

344 Some samples give isotope values greater than 0.5 parts per rail PDB (less than 20.5 ° C) and indicate temperatures below the seasonal range of modern surface temperatures. However, we believe that these values are positively biased since the samples are from depths equal to or greater than 3500 m and have probably experienced some dissolution. On the other hand, a few of these samples also occur on the slopes of topographic highs which are vulnerable sedimentary environments and thus may be Pleistocene outcrops material which have been exposed by mass gravity flows or current activity, A few samples yielding unexpectedly warm estimated temperatures were found beyond the influence of the Agulhas Current o r n o t d i r e c t l y related to its mean flow path. These may be core-tops of relict, reworked, or disturbed interglacial material, they occur chiefly off Durban near turbidite channels (Goodlad, 1986). Most of the oxygen isotope ratios obtained from the carbonate tests of foraminifera indicated depletion of 1sO, yielding Holocene values similar to those of Vincent and Shackleton (1980). This supports the belief that most of our coretops, especially north of 32 ° S, are probably of Recent age. -

Implications for sedimentology and the Agulhas Current The Agulhas Current is probably the major factor influencing sedimentation, sediment distribution patterns, and geological features in the study area (Martin, 1981 ). Evidence exists that the Agulhas Current may have influenced changing sedimentation throughout the Neogene with major current activity occurring between 65 myr B.P. and 38-22 myr B.P. (Martin, 1981). In the Pleistocene, fluctuations in the current probably occurred in response to sealevel changes and changing physiography (Martin, 1981 ). Vertical and lateral movements of the Agulhas Current during the past and particularly

during glacial intervals are of special interest because of the current's role in transporting heat from the equator to the poles. Any shift in its temperature structure and velocity would affect inter- and intra-ocean heat transfer and the climate of southern Africa. The farthest southerly penetration of the Agulhas Current is directly associated with the position of the subtropical convergence (STC), which separates warm saline subtropical water from cool low-salinity sub-polar water (Prell et al., 1979). Currently p o s i t i o n e d a t a b o u t 4 0 ° S (Fig. 1), the location of the subtropical convergence is linked to circulation patterns in the southern Indian Ocean (Duncan, 1970). At present about 20% of Agulhas Current water is lost from the Indian to the Atlantic Ocean (Gordon, 1985). The cold ABW which flows north through the Agulhas Passage at depths greater than 4000 m probably replaces some of this water but there is a net heat loss from the Indian to the Atlantic Ocean. A greater than 5 ° shift equatorwards in the position of the Subtropical Convergence would prevent Agulhas Current spillage into the Atlantic Ocean and thus alter the inter-ocean heat budget. Borehole and seismic data (Dingle et al., 1978; Dingle and Camden-Smith, 1979; Martin et al., 1982; Martin, 1984; Goodlad, 1986) have suggested that surface sediments in the study area are largely Cretaceous and Tertiary outcrops of relict material that have been exposed by the Agulhas Current during periods of intense flow. At present the Agulhas is voluminous and fast flowing and may erode and transport sediments as much as 2500 m below sea level (Duncan, 1970). Percent E. huxleyi present and oxygen-isotope ratios indicate that the majority of samples are most probably Recent or at least not older than 85,000 years except for sediments found in the Agulhas Passage. This suggests that sediments have been accumulating on the ocean floor faster than they have been eroding at least for the last 85,000 years and that the

345 Agulhas Current does not have a pronounced erosional influence, at least in areas from which cores were retrieved for this study, The data presented here demonstrate that the position of the Agulhas Current can be determined by analysing oxygen isotopic ratios from surfaces sediments from the Southwest Indian Ocean. it should therefore be possible to determine past changes in current position by studying isotopic ratios in cores underneath the current, The path of the Agulhas Current in the north of the study area is dictated by the location of its western boundary (the African continent) and therefore will not have changed much in the past. In the southwest of the study area, however, where the current has no well defined boundary, its position may have changed during the Pleistocene. Winter and Martin, 1988 (in prep.) have found that the glacial-interglacial range in oxygen isotope ratios of G. sacculifer from cores taken east of East London can be explained solely by changes in ice volume. No temperature effect is superimposed on the isotope signal, suggesting that temperature (and perhaps intensity) of the Agulhas Current remained more or less constant, at least since stage 5. Undissolved down-core sample material from the Agulhas Passage seems to indicate that the current core had not shifted in position or changed in intensity. This would imply that the inter-ocean heat transfer between the Indian and Atlantic Oceans (Gordon, 1985) could not have changed considerably. The position of the subtropical convergence during glacial time could therefore not have shifted northward more than a few degrees.

Conclusions Core-top samples from the southwest Indian Ocean show that: (1) Dissolution seems to be important in deretraining coccolith species abundances in the south where depths are greatest, whereas envi-

ronmental factors (e.g. nutrient concentration and water temperature) seem to play a more important role in the majority of samples taken from the northern two-thirds of the study area. (2) Dilution of calcareous sediments is related to the flux of terrigenous material to the ocean floor, and is therefore greatest close to the continent. Winnowing occurs in the shallow sediments off Maputo and Durban beneath the Agulhas Current and in the Agulhas Passage beneath ABW. (3) Three coccolith assemblages (Maputo, Agulhas Current and Deep water) have been recognized in the study area and are delineated by the relative percent distribution of four ecologically significant coccolithophore species (G. oceanica, E. huxleyi, C. leptoporus and U.

sibogae). ( 4 ) Lightest isotope values ( - 1.5 to - 1.0 per mil PDB) of G. sacculifer on the ocean floor occur beneath the "A" route of the Agulhas Current and are about 0.5 per rail PDB lighter than samples analyzed beyond the current's influence, thus reflecting the Agulhas Current's elevated temperature of between 2-3 ° C. (5) E. huxleyi percentages and isotope ratios from the carbonate tests of foraminifera indicate that sediments appear to have been accumulating over the last 85,000 years at least in areas from which the cores were taken. This suggests that the current is not particularly erosive at these sites at present and demonstrates the useful application of these data in reconstructing the paleohistory of the Agulhas Current.

Acknowledgements Special thanks to Matthew Smith, Lesley Lowe, and Rob Noble for their assiatance with sample preparations and analytical laboratory techniques. Thanks to Drs. Steve Goodlad, John Rogers, Bill Siesser, Frances Westall, Prof. Richard Dingle and Simon Robson for fruitful discussions. Dr. Keith Martin's assistance and comments from the conception of this project are greatfully acknowledged. Dr. Rich Johnson

346

is thanked for seemingly limitless information and computer expertise.The criticalreading of this manuscript by Prof. Richard Dingle and Dr. John Rogers is much appreciated. The staffsof the Electron Microscope Unit, Archaeometry Department, and Computer Science Department are also gratefullyacknowledged for their professional advice and technical assistance.Dr Grundlingh of National Research Institute of

Oceanography supplied satellitephotos. Samples were collected on various cruises by merebers of the Marine Geoscience Unit: R.V. Dingle,A.K. Martin, S. Goodiad, S. Robson, and F. Westall. This projectwas fundedby the South African National Committee for OceanographicResearch ( S A N C O R ) and the University of Cape Town Research Committee.

1 2 3 4 5 6 7 8 9 i0 ii 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

5088 5089 5090 5091 5097 5098 5104 5105 5109 5110 5111 5112 5113 5114 5115 5116 5117 5119 5121 5127 5729 5737 5743 5744 5745 5746 5751 5752 5753 5760 5761 5762 5764 5765 5766 5767 6326 6328 6331 6333 6334 6336 6571 6572 6575 6576 6577 6671 6672 6673

G G G G G G G G G T T T T G G G G G G G P P P P T P P P P P P T P P P P P P P P T T P P P P P T P P

mmber

Core

Core

Number Type

Sample

10YRS/2 2.5Y7/2 10YRB/I 2.5Y8/2 2.5Y8/2 IOYRS/I 2.5Y7/2 10YRS/2 10YR7/2 2.5Y8/2 2.5Y7/2 2.5Y8/2 5Y6/2 2.5Y7/2 10YR6/4 5Y7/2 10YR7/3 10YRS/I 2.5Y7/2 IOYP~/I 5Y8/I 2.5Y7/2 10YRB/I 10YRB/2 10YR7/2 2.5Y8/2 10YR7/I 10YRS/2 10YRB/2 10YRB/I 5Y8/I 10YRS/I 2.5Y7/2 5Y8/2 10YR8/2 10YRS/I 10YR8/I 2.5Y8/2 2.5Y7/2 10YR8/I 10YRS/2 2.5Y8/2 10YRS/2 10YR7/3 10YR8/2 10YR7/2 IOYR7/2 10YRB/2 2.5Y8/2 10YR8/2

Colonr

Rock

27"13'3 27"07'8 27°02'2 27°20'5 27°11'5 26"57' 26"43'6 26"42'6 26°34'1 26"49'8 26°07'6 25°57'2 25"32'2 25°22 ' 29°54'7 30°14'7 30°13'6 30°18'9 30°35'8 29"19'1 27"57'2 26"36'4 25"41'6 25"51'9 25"37'3 26°08'2 30"39'1 30"24'23 30"32'5 35"58'3 36"09'5 36"19'8 36°40'3 36°45'8 36°28'4 36°10'9 36"30'7 36"26'6 35"42'6 35"11'5 35"33'5 36"13'4 33"45'4 33"41' 33"50'9 33°57'8 34°02'8 34°54'7 34°42'6 34"13'3

Degrees

iat.

34°10 ' 33°41'5 33°30'5 33°38'9 33°14'3 33°11'5 33°16 ' 33"18'8 33°29'1 34°03'8 34°35'7 34°11'7 34°50 ' 34"13'3 31"40'5 31°54'9 31"54'7 32°32'7 33"07'6 33"12' 32"58'7 34°12'1 35"24' 34"25' 33"33' 33°27'7 34°19'3 33"54'4 32"42'5 25"25'3 25"19'6 25"14' 23°59'9 24°00'7 23"50'2 23"41'6 24"41' 24"46'1 25"04'1 25"42'6 24"58'7 24"33' 28"34'4 28"24' 28°08'3 28"02'6 27°58'5 26°55'4 26°46'3 27°35'2

Degrees

~.

SW of A1 Liete Bank St of Ponta do Our. ESt of Ponta do Our. E of Lake sibaya E of Lake Sibaya E of I~si Bay ~NE of ~ do Our. ~NE of Porfca do (Afro NE Of P ~ t a do Our. ENE of Ponta do Our. W of A1 Leite Bank E of Maputo NW Of A1 ieite Bank NW of A1 ieite Bank E of Durban SE of Durban SE of Durban ESE of Durban ESE of D/rban ~NE of Durban E of Lake St Tncia SW of A1 Liete Bank N of A1 Liete Bank NW of A1 Liete Bank S of Limpopo River ESE of Maputo ~ E of Durban ESE of [ilrban ESE of Durban S of Port Elizabeth SSW of Port Elizabeth S of Gamtoos River S of Storms River S of Storms River S of Plettenberg Bay S of Plettenberg Bay S of Cape St Francis S of Cape St Frar~is S of Gamtoos River S of Port Elizabeth S of Cape St Francis S of Cape St Francis SE of East Lor~k~n SE of East London SSE of East London S of East London S of East London SE of Port Elizabeth SE of Port Elizabeth SSW of~Ea~t London

Imcaticn

51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 i00

Number

Sample

Core

6674 6675 6677 6679 6680 6681 6682 6683 6686 6689 6690 6691 6693 6694 A1229 NIOEI5 NIOE39 NIOE40 NIOE45 NIOE49 NIOE50 NIOE51 NIOE55 NIOE56 NIOEI36 NIOE8 ~i ~4 ~5 ~9 ~12 ~13 MNI4 ~15 ~16 ~18 ~R~I9 ~20 MN21 ~23 ~24 ~25 ~26 ~#27 ~29 MN30 MNI531 ~32 6759 6761

P P T T T T P P P T T T T P P P P P P P P P P P P P T T T T T T T T T T T T T T T T T T T T T T P T

N ~ l e r Type

Core

IOYR8/2 10YR8/2 10YRB/I 10YRS/2 10YR8/2 19YR8/2 10YR8/2 10YR8/I 10YRB/2 5Y8/I IOYRS/I IOYRS/I IOYR8/2 10YR7/3 10YRB/2 5Y8/2 5Y8/2 IOYRB/I 10YR8/I 2.5Y8/2 10YRS/I 10YRS/2 5Y8/I 10YRB/2 10YR8/2 10YRB/2 2.5Y7/2 10YR8/2 10YRB/2 10YR8/2 10YRS/2 10YRS/I 2.5Y8/2 10YRS/I 10YRB/2 10YRS/I 10YRB/2 10YRB/2 10YRB/3 10YR7/4 10YR6/3 IOYRS/2 10YR8/2

IOYRS/2

10YR7/2 2.5Y8/2 10YR8/2 IOYRS/2 IOYRS/2 IOYRB/2

Colour

Rock

34°00'3 33°47'6 33°12 ' 33°10'9 33°12, 33"14'5 33°10'8 33"09'4 33°09'5 34"00'3 33°59'3 33°59'8 33°57 ' 34°00'1 26°12 ' 39"19'5 36"42'2 37"33' 36"44'1 32°17'8 31"51'7 33"10'4 34°49'1 34°35'1 39"10' 32"11'8 30"46'1 29"10'5 29"08'8 29°00 ' 28"54'3 29"02'3 29°12 ' 29"15'2 29"16'7 29"37'2 29o36 ' 29"29'7 29"27'4 29"24'9 29"27'2 29"30'3 29°28'8 29"33'4 29"40'9 29°44'2 29"40'9 29"42'8 32°49 ' 33"08'5

Degrees

Lat.

27"38'2 27"42'5 29"04'7 29"32'5 30"14'3 31"18'8 33"09'1 33"01'6 33"14'7 32"08'4 32"00'3 31"29'5 30"14'3 29"33'9 34"12' 30"06'5 22"29' 24°51 ' 21°40'7 29"28'7 30"29'2 31"46'9 31"42'9 26°13 ' 33"35' 35"25'1 33"43'6 34"45'9 34°38'4 34"13'8 33"08'7 33"34'1 33o29 ' 33°26'8 33"27' 33°22'1 33"10'7 33"13' 33"04'6 32°51'2 32°48'9 32"40'7 32"32'8 32"27'1 32"13' 31"58'2 31"55'6 31"44'9 33"49'25 32"33'4

D~/rees

Long.

SSW of East London SSW of East ~ n E of East London E of East London E of Fast E of East London E of East E of East Lor~__on E of East E of Port Elizabeth E of East London E of East London E of East London E of East London W of A1 ieite Bank E of East London S of George S of Cape St Francis S of Gourits River NE of East NE of East London E of East Lotion SE of East London SSE of Port Elizabeth SE of Port Elizabeth ~ of East London E of Durban E of [Alrban E of Durban E of Durban E of Durbln E of Durban E of Durban E of Durban E of DLtrban E of Durban E of Durban E of Durban E of Durban E of Durban E of Durban E of [ilrban E of Durban E of Durba~ E of Dtlrban E of Durban E of Durban E of Durban E of East Lotion E of East London

Location

Appendix I - Surface sample locations and descriptions: P = piston; G= gravity; T = trigger; 10YR7/2, 2.5Y7/2 = light grey; 5Y6/2 = light olive grey; 10YR6/4 = light yellowish brown; 10YR6/3 = pale brown.

%o

348 Appendix II - Relative percent abundances of the four most abundant coccolith species occurring in the study area ranked in order of decreA.qing mean percent abundances. Sample Ge~yrocapsa ~3niliania Calcidiscus Umbilicos~m~l-a Nt~ber oceanica b~xleyi leptopor~ sibogae 1 2 3

44.9 29.6 28.3

6.7 50.6 45.6

4

21.9

46.8

38.3 46.3 26.4 35.4 34.1 41 •4 38.5 50.6 88.6 48.3 HA 27.3 34.9 30.8 25.2 22.0 44.8 45.6 40.2 32.3 79.6 NA 15.4 NA 28.5 29.1 35.5 28 •6 45.1 36.9 35.9 46.6 NA 39.8 39.2 37.9 NA 36.0 64.9 27 .6 38.3 NA 42.3 50 •0 41.5 33.5

5 6 7 8 9 i0 ii 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

39.6 29.5 45.7 43.9 30.5 28 •2 38.5 30.1 0.0 35.8 NA 37.3 43.6 43.6 41.7 34.6 17.8 16.9 28.8 39.6 38.2 NA 47.3 NA 27.3 44.4 9.7 26.4 16.4 5.5 14.2 7.4 NA 4.6 6.5 8.1 NA 2.1 10.7 19.4 20.1 NA 4.3

8.8 0.9 3.7 2.7 2.2 4.1 4.6 1.4 3.8 5 •7 2.1 5.8 2.8 0.9 NA 8.6 4.8 ii.0 19.5 ii.i 11.2 8.8 7.3 4.0 6.9 NA 9.7 NA 14.5 4.7 42.1 17.7 15.5 40.0 31.1 ii.0 NA 25.9 43.2 30.1 NA 35 •0 8.9 25.1 24.5 NA 32.6

24.3 8.1 ii. 7 11.4 13.7 11.5 14.5 5.3 12.9 12.9 10.2 4.2 3.8 6.8 NA 15.6 9.9 7.8 6.9 14.1 14.5 14.0 15.4 17.5 6.6 NA 8.1 NA 20.1 7.1 5.9 12.1 14.8 6.5 9.4 15.1 NA 12.1 0.0 ii.i ~i~ 6.6 5.9 13 •4 6.3 NA 3.8

0 •0

1 6 •5

2 0 •1

16.2 12.4

23.6 27.9

8.1 11.8

Ge~hyrocapsa Emiliania Calcidiscus Umbilicosphaera Nlm~er oceanica huxleyi leptoporus sibogse Sample

51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100

34.1 38.0 20.1 32.1 31.4 25.1 74.5 37.9 80.8 NA 46.2 51.7 43.3 34.0 23.0 27.5 1.9 25.4 52.7 31.5 41.2 29.0 NA 91.6 31.0 36.2 33.6 51.4 20.9 24.5 38.2 26.2 24.1 NA NA NA 16.5 19.3 22.1 24.4 24.7 15 •9 27.0 32 .0 27.7 34.2 NA NA 17.3 41.2

31.8 22.8 18.4 41.1 14.9 32.3 1.6 26.7 3.8 NA 10.7 17.4 11.4 18.2 37 •7 16 •2 8.9 19.2 7.8 27.4 16.9 25.9 NA 0.9 11.2 19.0 27.3 23.5 36.2 37.7 24.3 34 •9 45.2 NA NA NA 50 •5 25.1 36.3 42.0 35.1 37.2 30.0 25.1 42.5 38.1 HA NA 45.8 6.6

14.0 13.7 29.3 13.1 30.0 25.7 13.5 16.4 5.6 NA 22.8 16.8 24.9 27.0 13 •5 30.8 0.2 34.9 15.6 10.9 15.5 22.5 NA 1.2 40.6 15.4 16.4 9.7 11.9 10.4 14.6 15.7 10.2 NA NA NA 9 •4 17.5 19.9 5.7 9.7 15.4 13.0 15.0 4.9 5.8 NA NA 12.7 6.6

7.3 9.7 17.7 4.4 8.9 9.1 3.8 8.7 2.5 NA 9.0 3.7 9.2 8.2 ii. 3 11 •7 0.0 7.0 8.4 8.1 11.6 9.8 NA 1.5 8.2 10.4 9.8 6.3 14.5 12.9 5.8 15.4 9.0 NA NA NA 12 •7 22.3 9.5 11.5 9.1 16.7 17.0 i0 •8 10.6 ii.i NA NA 7.8 8.2

349

Appendix III - Relative percent abundances of foraminiferal species and fragments, oxygen isotope ratios and derived

temperatures. Sanlole Number

Fragment Percent

CaCO 3 Percent

Oxygen 18 PDB

1 2 3 4 5 6 7 8 9 i0 ii 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

39.6 21.5 39.9 23.5 33.6 54.6 41.5 64.0 44.1 41.0 56.0 44.3 48.7 59.0 39.2 36.6 24.9 15.4 19.5 21.6 49.1 46.2 44.4 54.4 52.6 40.5 12.0 21.0 25.0 89.0 65.2 22.8 I00.0 96.0 84.8 14.0 66.2 91.2 88.4 90.7 73.5 69.4 32.7 14.1 14.6 45.5 10.7 74.4 59.9 33.8

36.5 39.7 54.9 59.9 35.9 61.8 52.8 43.9 44.5 39.7 32.4 59.4 37.0 42.5 18.5 32.6 11.8 39.2 78.8 62.3 43.6 34.5 46.4 44.6 60.5 NA 76.6 53.7 46.9 32.2 21.3 50.5 13.2 33.2 34.2 63.1 32.3 15.9 28.6 30.5 33.0 41.2 52.6 42.5 NA 54.4 36.2 26.3 25.2 35.2

-0.694 -1.127 -0.882 -1.246 -0.645 -0.521 -1.060 -1.089 -1.085 -1.205 -1.193 -0.647 -1.489 -0.668 -1.334 -1.006 -1.072 -1.014 -0.665 -1.125 -i.ii0 -1.429 -1.196 -0.514 -0.726 -0.425 -0.693 -0.503 -0.831 NA NA 0.533 NA -1.064 NA 0.522 NA NA NA NA NA -0.243 -0.879 -0.734 -1.334 -0.640 -0.721 NA NA -1.142

Derived Temp. 21.4 23.4 22.2 23.9 21.1 20.6 23.0 23.2 23.2 23.7 23.7 21.1 25.1 21.2 24.3 22.8 23.1 22.8 21.2 23.3 23.3 24.8 23.7 20.5 21.5 20.1 21.3 20.5 22.0 18.2 18.2 15.9 18.2 23.1 18.2 15.9 18.2 18.2 18.2 18.2 18.2 19.3 22.2 21.5 24.3 21.1 21.5 18.2 18.2 23.4

Sample Number

Fragment Percent

51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 I00

45.3 26.7 16.2 36.9 29.5 12.3 55.8 19.3 46.6 23.4 54.7 31.8 23.5 45.8 14.0 18.8 42.2 20.3 51.0 27.0 28.0 11.3 NA 50.5 16.0 25.6 16.0 86.3 4.0 10.7 83.1 31.2 20.3 18.4 32.3 13.9 13.9 18.7 21.3 20.7 24.3 27.8 14.4 28.4 24.2 29.6 NA 48.5 NA NA

CaCO 3 Percent 22.4 37.0 41.6 43.8 48.8 51.9 54.7 82.2 55.4 59.4 76.7 63.4 37.0 36.9 68.7 54.3 30.9 73.0 53.7 41.0 32.1 58.9 NA 35.9 82.7 75.3 63.1 66.6 85.7 71.7 53.8 44.3 53.4 46.2 47.4 69.7 71.4 82.2 38.9 60.3 57.5 59.2 57.8 69.2 46.5 24.1 NA 13.2 NA NA

Oxz~en 18 PDB

Derived Temp.

-0.119 -0.223 -1.285 -0.157 -1.228 -0.379 0.094 0.338 -0.257 -0.333 0.095 -0.152 -0.503 -1.154 -0.542 -0.816 NA -0.310 NA -0.249 -0.975 -0.621 NA -0.594 -0.055 0.584 -0.209 NA -0.664 -0.924 -1.045 -0.762 -1.312 -1.065 -0.888 -0.325 -1.648 -0.606 -1.322 -1.343 -1.093 -1.447 -1.135 -0.898 -0.817 -1.025 NA -1.936 NA NA

18.8 19.2 24.1 18.9 23.8 19.9 17.8 16.7 19.4 19.7 17.8 18.9 20.5 23.5 20.7 21.9 18.2 19.6 18.2 19.3 22.6 21.0 18.2 20.9 18.5 15.7 19.2 18.2 21.2 22.4 23.0 21.7 24.2 23.1 22.2 19.7 25.8 21.0 24.3 24.4 23.2 24.9 23.4 22.3 21.9 22.9 18.2 27.2 NA NA

350

References

Berger, W.H., 1973. Deep sea carbonates: evidence for a coccolith lysocline.Deep-Sea Res., 20: 917-921. Berger, W.H. and KiUingley, J.S., 1977. Glacial-Holocene transition in deep sea carbonates: selectivedissolution of the stable isotope signal.Science, 197: 563-566. Birch, G.F., 1981. The karbonate bombe: a precise, r a p i d and cheap method for determining calcium carbonate in sediments and rocks. Trans. Geol. Soc. S. Aft., 84: 199-203. Bukry, D., 1971. Discoaster evolutionary trends. Micropaleontology, 17 (1): 43-52. Craig, H., 1957. Isotopic standards for carbon and oxygen and correction factors.Geochirn. Cosmochim. Acta, 12: 133-149. Dansgaard, W., 1964. Stable isotopes in precipitation.Tellus, 16: 436-468. Dingle, R.V.D., 1978. South Africa. In: M. MouUade and A.E.M. Nairn (Editors), The Phanerozoic Geology of the World, 2. The Mesozoic. Elsevier, Amsterdam, pp. 401-433. Dingle, R.V.D., Goodlad, S.W. and Martin, A.K., 1978. Bathymetry and stratigraphy of the Northern Natal Valley (Southwest Indian Ocean): A preliminary report. Mar. Geol., 28: 28-106. Dingle, R.V.D. and Camden-Smith, F., 1979. Acoustic stratigraphy and current generated bedforms in deep ocean basins offsouth-eastern Africa.Mar. Geol.,33: 239-260. Dingle, R.V.D. and Robson, S.R., 1985. Slumps, canyons and related features on the continental margin off East London, SE Africa ( S W Indian Ocean). Mar. Geol.,67: 37-54. Duncan, C., 1970. The Agulhas Current. Thesis. Univ. Hawaii. Emiliani, C., 1955. Pleistocene temperatures. J. Geol., 63: 538-578. Erez, J. and Luz, B., 1983. Experimental paleotemperature equation for planktonic foraminifera. Geochim. Cosmochim. Acta, 47: 1025-1031. Flemming, B.W., 1981. Factors controlling shelfsediment dispersal along the southeast African continental margin. Mar. Geol., 42: 259-277. Friedinger and Winter, A., 1985. Distribution of modern coccolithophore assemblages in the area of the Natal Valley (southwest Indian Ocean). Joint Geol. Surv./ Univ. Cape Town Mar. Geosci.Unit Tech. Rep., 14: 146150. Geitzenauer, K.R., Roche, M.B. and McIntyre, A., 1977. Coccolith biogeography from North Atlantic and Pacific surface sediments. In: A.T.S. Ramsay (Editor), Oceanic Micropaleontology. Academic Press,N e w York, N.Y., pp. 973-1008. Goodlad, S.W., 1978. The bathymetry of the Natal Valley off the Natal and Zululand coasts (southern Africa).

Joint Geol. Surv./Univ. Cape Town Mar. Geosci. Unit Tech. Rep., 10: 96-105. Goodlad, S.W., 1986. Tectonic and sedimentary history of the Mid-Natal Valley (S.W. Indian Ocean). Joint Geol. Surv. Univ. Cape Town Mar. Geosci. Unit Bull., 15: 1415. Gordon, A.L., 1985. Indian-Atlantic transfer of thermocline water at the Agulhas Retroflection. Science, 227: 1030-1033. Harris,T.F.W. and Van Foreest, D., 1977. The Agulhas Current system. Univ. Cape Town, Dep. Oceanogr., 36 PP" Honjo. S. and Okada, H., 1974. Community structure of coccolithophores in the photic layer of the mid Pacific. Micropaleontology,20: 209-230. Huang, T.C. and Watkins, N.D., 1977. Contrasts between the Brunhes and Matayama. Sedimentary records of bottom water activityin the south Pacific.Mar. Geol., 23: 113-132. Kennett, J.P., 1982. Paleo-oceanography: Global Ocean Circulation.U.S. Natl. Rep. Int. Union Geodesy Geophys., 1979-1982. Rev. Geophys. Space Phys., 21: 12581274. Kolla, V., Sullivan, L., Streeter,S.S. and Langseth, M.G., 1976. Spreading of the A B W and itseffectson the floor of the Indian Ocean inferred from bottomwater potential temperature, turbidity and sea floor photography. Mar. Geol., 21: 171-189. Krey, I.and Babenerd, B., 1976. Atlas of the International Indian Ocean Expedition. Landesvermessungsamt Schleswig-Holstein,Kiel. Malan, O.G. and Schumann, E.H., 1979. Natal shelf circulation revealedby Landsat imagery. S. Afr.J. Sci.,75: 136-137. Martin, A.K., 1981. The influence of the Agnlhas Current on the physiographic development of the northernmost Natal Valley (SW Indian Ocean). Mar. Geol., 39: 25976. Martin, A.K., 1984. Plate tectonic status and sedimentary basin in-fillof the Natal Valley (S.W. Indian Ocean). Joint Geol. Surv./Univ. Cape Town Mar. Geosci. Unit Bull., 14: 1-209. Martin, A.K., Goodlad, S.W. and Salmon, D.A., 1982. Sedimentary basin in-fillin the northernmost Natal Valley, hiatus development and Agulhas Current palaeoceanography. J. Geol. Soc. Lond., 139: 183-201. Mclntyre, A., 1967. Coccoliths as paleoclimatic indicators of pleistocene glaciation.Science, 158: 1314-1317. McIntyre, A., 1969. The coccolithophorida in red sea sediments. In: E.T. Degens and P.A. Ross (Editors), Hot Brines and Recent Heavy Metal Deposits in the Red Sea. Berlin, pp. 229-305. McIntyre, A., Bd, A.W.H. and Roche, M.B., 1970. Modern Pacific coccolithophorids a paleontological thermometer.N.Y. Acad. Sci.Trans., 32: 720-731. McIntyre, A. and Bd, A.W.H., 1967. Modern coccolitho-

351 phoridae of the Atlantic Ocean. 1. Placoliths and c y r toliths.Deep-Sea Res., 14: 561-597. Mclntyre, A. and McIntyre, R., 1971. Coccolith concentrations and differentialsolution in oceanic sediments. In: B.M. Funnel and W.R. Reidel (Editors), The Micropaleontology of the Oceans. Cambridge Univ. Press, pp. 253-261. Muller, G. and Gastner, M., 1971. The "karbonate bombe", a simple device for the determination of carbonate content in sediments, soilsand other minerals. Neues Jahrb. Mineral., 10: 466-469. Okada, H. and Honjo, S., 1973. The distributionof oceanic coccolithophorids in the Pacific.Deep-Sea Res.,20: 355374. Okada, H. and Honjo, S., 1975. Distribution of coccolithophores in marginal seas along the western Pacific Ocean in the Red Sea. Mar. Biol., 31: 271-285. Okada, H. and McIntyre, A., 1977. Modern coccolithophores of the Pacific and north Atlantic oceans. Micropaleontology, 23: 1-55. Okada, H. and McIntyre, A., 1979. Seasonal distribution of modern coccolithophores in the western north Atlantic Ocean. Mar. Biol., 54: 319-328. Pearce, A.F., 1977. Some features of the upper 500 m of the Agulhas Current. J. Mar. Res., 35: 731-751. Prell, W.L., Huteon, W.H. and Willianis, D.F., 1979. The southern tropical convergence and late Quaternary circulation in the southern Indian Ocean. Mar. Micropaleontol., 4: 225-234. Reid, J.L., 1962. On circulation, phosphate-phosphorous content and zooplankton volumes in the upper part of the Pacific Ocean. Limnol. Oceanogr., 7: 287-36. Roth, P.H. and Berger, W.H., 1975. Distribution and dissolution of coccoliths in the south and central Pacific. Cushman Found. Foraminiferal Res., Spec. Publ., 13: 87-113. Roth, P.H. and Coulbourn, W.T., 1982. Floral and solution patterns of coccoliths in surface sediments of the north Pacific. Mar. Micropaleontol., 7: 1-52. Shackleton, N.J., 1974. Attainment of isotopic equilibrium between ocean water and the benthonic Foraminifera genus Uvigerina: isotopic changes in the ocean during the Pleistocene. In: Les m~thodes quantitatives d'$tude des variations du climat au cours du Pl$istoc~ne. Colloq, Int. C.N.R.S., 219: 203-209.

Shannon, L.V., Stander, G.H. and Cambell, J.A., 1973. Oceanic circulationdeduced from plasticdriftcards. Invest. Rep. S. Aft. Div. Sea Fish., 108, 31 pp. Schneiderman, N., 1977. Selective dissolution of recent coccolithsin the Atlantic Ocean. In: A.T.S. Ramsay (Editor), Oceanic Micropaleontology. Siesser,W.G., 1975. Calcareous nannofossils in Pleistocene sediment cores from the South African continental slope. Trans. R. Soc. S. Aft.,42: 108-147. Thierstein,H.R., Geitzenauer, K.R., Molfino, B. and Shackleton, N.J., 1977. Global synchroneity of late Quaternary coccolith datum levels:validation by oxygen isotopes. Geology, 5: 400-404. Ushakova, M.G., 1970. Coccoliths in suspension and in the surface layer of sediment in the Pacific Ocean. In: Funnal and Reidel (Editors). The Micropaleontology of the Oceans. Cambridge Univ. Press. Vincent, E. and Shackleton, N.J., 1980. Agulhas Current temperature distribution delineated by oxygen isotope analysis of foraminifera in surface sediments. Cushman Found. Foraminiferal Res. Spec. Publ., 19: 89-95. Westall, F., 1984. Current-controlled sedimentation in the Agulhas Passage, SW Indian Ocean. Joint Geol. Surv./ Univ. Cape Town Mar. Geosci. Unit Bull., 12: 1-276. Winter, A., 1982. Paleoenvironmental interpretation of Quaternary coccolith assemblages from the Gulf of Aqaba (Elat), Red Sea. Rev. Esp. Micropaleontol., 14: 291314. Winter, A., 1985. Distribution of living coccolithophores in the California Current system, southern California borderland. Mar. Micropaleontol., 9: 385-393. Winter, A., Reiss, Z. and Luz, B., 1979. Distribution of living coccolithophore assemblages in the Gulf of Elat (Aqaba). Mar. Micropaleontol., 4: 197-223. Winter, A., Almogi-Labin, A., Erez, Y., Halicz, E., Luz, B. and Reiss, Z., 1983. Salinity tolerance of marine organisms deduced from Red Sea Quaternary record. Mar. Geol., 53: M17-M22. Wyrtki, K., Bennet, E.B. and Rochford, D.J., 1971. Oceanographic Atlas of the International Indian Ocean Expedition. Natl. Sci. Found. Washington, D.C., U.S. Gov. Printing Office, 531 pp.