Glacial-Holocene calcareous nannofossils and paleoceanography in the northern South China Sea





Marine Micropaleontology 32 (1997) 95-l 14


calcareous nannofossils and paleoceanography


the northern South China Sea

Kuo-Yen Wei *, Tien-Nan Yang, Chi-Yue Huang

Department of Geology, National Taiwan University, Taipei, Taiwan Received

I June 1996; accepted 1 December 1996


Last glacial to Holocene paleoceanography of the northern South China Sea was inferred from nannofossil variations and several hydrographic proxies from a piston core, SCS90-36 (17”59.7O’N, 111”29.64’E, water depth 2050 m). The upper part of the sedimentary sequence (dated 15.5 to 1.2 ka) provided a high-resolution record whereas part of the sediments older than 15.5 ka was lost due to erosion. A correspondence analysis of the nannofossil succession suggests that the paleoceanography developed

in four stages. The first stage (26-13.3 ka) has a fairly well-preserved diverse nannoflora

dominated by



Florisphaera projkda. The floral composition together with high concentration of ketones (C,,) and organic carbon indicates high surface-water fertility. The second stage, the deglacial period (13.3-10.7 ka), had an increased surface-water turbidity and a stronger influence of Pacific open-ocean waters as evidenced by the decrease of Florisphaera profunda

and increase of

Emiliania huxleyi,

respectively. A preservation peak of calcareous

microfossils centered at 12 ka correlates to the global preservation event of Termination I. The third stage, early Holocene

(10.74.4 ka), is marked by a gradual increase of F:

profunda and small placolith taxa at the expense of E. huxleyi. The floral composition indicates that conditions were more oligotrophic compared to the pre-Holocene. The preservation of nannofossils became progressively worse, indicating a rise of the nannofossil lysocline. In sediments deposited at 5.5 and 4 ka, nannofossil preservation improves, probably reflecting a local cooling event. During the last stage, from 4.4 to 1.2 ka, E. huxleyi, Umbilicosphaera

and large


increased their relative abundance to replace small placoliths.

Further stratification of the surface water column may have been responsible for this floral succession.

Keywords: Pleistocene; Holocene; South China Sea; nannofossil; paleoceanography

1. Introduction

The extensive development

of glaciers in polar ar-

eas during the last glacial maximum

(LGM) resulted

in a sea-level

drop by 120 m (Fairbanks,


Most straits that bound the South China Sea (SCS)


too shallow

to allow exchange

of waters

*Corresponding author. Tel.: +886-2-3691143; fax: +866-2- 3636095; e-mail:

between the South China Sea and the open oceans.

The Bashi Strait, located between Taiwan and Lu-

zon Island (Fig. l), with a sill depth of 2500 m,

was the only passageway

to the West Philippine


during the last glacial. The exposure of continental

shelves, especially

in the southern part, also caused

a l/5 reduction

of the surface area of the South

China Sea (Wang, 1990). The lower sea level might

also have caused the Kuroshio to shift offshore, east

of its current position

(Ujiie et al., 1991; Ahagon

0377~8398/97/$17.00 C 1997 Elsevier Science B.V. All rights reserved.


96 K.-Y; Wei et d/Marine Micropaleontology 32 (1997) 95-114

llZ@ 114O 116' 11e 120@


ll? 114O 116' 1w 120'

Fig. 1. Locations of Core SCS90-36 studied and Core RC26-16 mentioned in the text.

et al., 1993) putting the South China Sea under a

stronger influence

of the Continent

Coastal Current

(Wang and Wang, 1990). In the northwestern


the Kuroshio Front advanced


during the

last glacial (Moore et al., 1980; Thompson,


The Kuroshio Front then moved northward at about

13-10 ka (Chinzei

et al., 1987), synchronous


the post-glacial

warming of the North Atlantic Ocean

(Duplessy et al., 1981). There were, however, two in-

tervening retreats of the Kuroshio Front to the south

during the post-glacial

period; the first was at 1 l-10

ka, corresponding

to the Younger Dryas Event while

the second took place after 5.5 ka. Both events were

associated with brief cooling (Chinzei et al., 1987).

The winter sea surface temperatures

(SSTs) in the

northern part of the South China Sea were estimated

to be4-6”C

lower than today on average (at RC26- 16

by Wei et al., 1996; at SCS 90-36 by Huang et al.,

1997; and to the south, near the Mindoro


by Miao et al., 1994). The summer SSTs were es-

timated to be quite similar to today’s or 2°C lower

(Wei et al., 1996; Miao et al., 1994; Huang et al.,

1997). The larger seasonal temperature

contrast has

been attributed to reduced surface water exchanges

across the straits and to a stronger winter monsoon

(Miao et al., 1994). This stronger winter monsoon

during the glacial might also have resulted in better


of the surface water and so have induced

higher biological


(Huang et al., 1997).

A subsequent

drop of surface productivity


the transition

from the glacial to the postglacial


also been documented

in other western Pacific low-

latitude areas and in the SCS (Herguera and Berger,

1991; Herguera, 1992; Thunell et al., 1992). Both the

mixing rate and nutrient


of the deep

waters were lowered during the glacial-postglacial

transition (Herguera and Bergen 1994).

Huang et al. (1997) documented

a rich set of

last glacial to Holocene



and interpretations

from the piston core SCS90-36

located on the northern

slope of the South China

Sea (Fig. 1). Here we document a quantitative


of the calcareous



from the

same core. We attempt to integrate the nannofossil


K.-E Wei et al. /Marine Micropaleontology 32 (1997) 95-114 97

evidence with the other available data to further illus-

trate the deglacial warming and climatic fluctuations

during the Holocene

in the northern

South China



Samples and procedures

The piston core SCS 90-36 was recovered from

17”59.70’N, 111”29.64’E, at a water depth of 2050 m

on the southern slope of the Xisha Trough (Fig. 1).

The sequence


of fine silty clay without

any detectable


layers. The chronology


the sequence

was constrained

by eight radiocarbon


of planktic


using accelerator

mass spectrometry


(Fig. 2; Huang

et al.,

1997). The radiocarbon

age of the bottom of the

studied sequence is about 26 kyr. The extremely low


rates (~2 cm/kyr) in the lowermost

part (Fig. 2), however, have led us to suspect that

most of the sediments


prior to 15.5 ka

were lost due to erosion.


were taken every 3 cm from the top

103 cm of the core, corresponding

to a temporal


of -500

yr per sample for the Holocene

and -200

yr for the deglacial period. A total of 32

samples were analyzed. At least 700 coccoliths were

identified and tallied per sample under a Zeiss Pho-


at a magnification

of 1250x.


samples were examined

under a scanning



to facilitate



Age W

0 5 10 15 20 25 30


6 o



17.2 12oJ

Fig. 2. Depth-age diagram of Core SCS90-36, showing sedi- mentation rates estimated from eight AMS 14C dates of planktic foraminifera.

Table 1

List of tallied nannofossils

Calcidiscus leptoporus a (Murray and Blackman) Loebhch and Tappan, 1978

Calcidiscus macintyrei b (Bukry and Bramlete) Loeblich and Tappan, 1978

Calciosolenia murrayi Gran, 1912 Ceratolithus cristatus a Kamptner, 1950

Coronocyclus nitescens b (Kamptner) Bramlette and Sullivan, 1961

Crenalithus sp. Chen, 1978 Cricolithus jonesi Cohen, 1965 Dictyococcites perplexa Bums, 1975

Dictyococcites productus (Kamptner) Backman, 1980 Discoaster spp. b

Emiliania huxleyi (Lohmann) Hay and Mohler, 1967 Florisphaera ptojiatda profunda Okada and Honjo, 1973 Florisphaera profunda elongata Okada and McIntyre, 1979 Gephyrocapsa caribbeanica Boudreaux and Hay, 1967 Gephyrocapsa oceanica Kamptner, 1943, see Matsuoka and

Okada, 1989

Helicosphaera carteri carteri a (Wallich) Kamptner, 1954 Helicosphaera carteri hyalina a Gaarder, 1970

Helicosphaera carteri wallichii a (Lohmann) Okada and McIntyre, 1977

Neosphaera coccolithomorpha ’ Lecal-Schlauder, 195 1 OolithotusfragilisC (Lohmann) Okada and McIntyre, 1977 Pontosphaera discopora Schiller, 1925

Pontosphaera japonica (Takayama) Bums, 1973 Pontosphaera multipora (Kamptner) Roth, 1970 Pseudoemiliania lacunosa b (Kamptner) Gartner, 1969 Reticulofenestra haqii Backman, 1978

Reticulofenestra minuta Roth, 1970

Reticulofenestra minutulu (Gartner) Haq and Berggren, 1978 Rhabdosphaera clavigera a Murray and Blackman 1898 Rhabdosphaera longistylis a Schiller, 1925

Syracosphaera histrica ’ Kamptner, 1941 Syracosphaera lamina ’ Lecal-Schlauder, 195 1 Syracosphaera pulchra ’ Lohmann, 1902 Syracosphaera spp. ’

Thoracosphaera spp. ’

Umbellosphaera irregularis ’ Paasche, 1955 Umbellosphaera tenuis ’ (Kamptner) Paasche, 1955 Umbilicosphaeru sibogae a (Weber-van Bosse) Gaarder, 1970 a Dissolution-resistant.

b Reworked species. ’ Dissolution-susceptible.

The nannofossil

taxa tallied are listed in Table 1.



and nomenclatural

notes on

some selected species are given in Appendix A.

To monitor

the preservation

state of the nanno-

fossil assemblages,

a dissolution

index, defined as

the ratio of resistant

species to resistant

+ suscep-


98 K.-E Wei et al. /Marine Micropaleontology 32 (1997) 95-114

tible species, was calculated.

We adopted Roth and

Coulboums’ ranking (Roth and Coulboum,

1982) of

the north tropical Pacific nannofossil




to assign the various species into the two


(Table 1). Only minor and rare species

were included in the calculation

in order to avoid be-

ing overwhelmed

by dominant

species and hopefully

to reduce artifacts introduced

by factors other than


A rich set of paleoceanographic

proxies, including

weight percentage

of carbonate

and organic carbon,

relative abundances

of planktic foraminifera,


and oxygen isotopic ratios of planktic

and benthic


U,k index of long-chain



well as abundance of Cs7 ketanes, were measured by

Huang et al. (1997).


Multivariate analysis

To recognize

the down-core



tion pattern,

we applied

a multivariate







CA) (Hill and Gauch, 1980), to the nannofloral


data. This method allows taxa and samples to be

ordinated simultaneously

and therefore shows the re-

lationship between them. The DCA is a modification

of regular correspondence

analysis to eliminate


‘arch effect’ that normally

distorts the projection


data points on the first axis. This modified method

can adjust the scores so that the data points are more

linear along the first axis and more evenly spread

out. The DCA assumes that the samples come from

a gradient in which different variables

(taxa) char-

acterize different parts of the gradient. In applying

this method, we explicitly seek an ordination


along the chronological

axis. In other words, we are

trying to identify taxa which show most age-depen-

dent variation and therefore characterize

the various

stages in the paleoceanographic


All reworked

taxa and taxa with relative abun-

dances less than 1% were excluded. To reduce skew-

ness and to improve normality

of data distribution,

the original percentage

data were natural log trans-

formed prior to analysis.

The transformation



weights the minor and rare taxa relative

to the dominant

taxa so that variations

in the minor

taxa will not be overwhelmed

by the dominant ones.

4. Results and interpretations







the studied section and display a diverse


with good preservation.

Between 28 and

39 species/subspecies

were recorded

per sample.


species are illustrated

in Plates I-IV.

The nannofossil


were dominated


three taxa:

Gephyrocapsa spp., Emiliania huxleyi

and a deep dwelling species,

Florisphaera profunda

(Fig. 3). The latter two are considered

to be pelagic



their relative



water depth, whereas

Gephyrocapsa has a prefer-

ence for neritic environments

in the marginal seas of

the western Pacific (Chen and Shieh, 1982; Okada,

1983, 1992; Cheng, 1992; Chen and Huang, 1995).

The compositions

of the studied samples fall gen-

erally in the bathypelagic


of open sea in

the western

Pacific as defined

by Okada (1983).

The occurrence


Syracosphaera, Rhabdosphaera,

Oolithotus, Ceratolithus and Umbellosphaera, al-

though rare in abundance,

is typical for the bathy-

pelagic environments

in the western Pacific marginal

seas (Chen and Shieh, 1982; Okada, 1983; Wang and


1983; Cheng, 1992).

The correspondence

analysis yielded a quite inter-

pretable pattern (Fig. 4). The first two axes account

for 44 and 11% of the data variation,


Samples exhibit a well-phased



tion. The floral succession

can be subdivided


four stages: Stage I, dated 25.9-13.3

ka, represented

by the glacial assemblage;

Stage II, from 13.3 to 10.7

ka, marked by the deglacial flora; Stage III, 10.74.4

ka, the early Holocene;

and Stage IV, 4.4-1.2


the late Holocene.

The glacial and deglacial


ples are clustered

and distributed

on the left half

of the first axis, with a progressive

upward migra-

tion along the second axis. In contrast, the Holocene

samples occupy the right half of the first axis and

show a downward


along the second axis.

The boundaries

of stages are characterized

by major

shifts in the data trajectories

(Fig. 4a). The charac-

teristic nannofossils

of each stage can be found in

the corresponding

positions on the projection

of the

species scores on the two eigen-axes

(Fig. 4b).

The glacial stage (Stage I, 25.9-13.3

ka) is char-


by the dominance


Gephyrocapsa and


K.-I: Wei et al. /Marine Micropaleontology 32 (1997) 95-114 99 96




% Emiliania huxlevi

0 10 20304050 0 1020304050 0 20 40 60 10 ;zz; l2 14 16 18 20 22 24 26



oceanica (L diag) n G. oceanica (L vert) @ G. oceanica (M diag) E3 G. oceanica (M vert ) G. ericsonii m G. caribbeanica ( L ) E] Geph. caribbeanica ( M ) - ... ... ... ... 1 6 - ::;:::::::;::i ... ... ... ...

Fig. 3. Downcore variation of the relative abundances of three dominant nannofossil taxa. Horizontal lines mark the boundaries of four stages recognized from the correspondence analysis shown in Fig. 4.

Florisphaeru (Fig. 3). The diverse flora and more



of dissolution-susceptible


(such as

Umbellosphaera, Syracosphaera and Cal-

ciosoknia murruyi) relative to other stages suggests

that the nannofossil


in the glacial in-

terval is generally better. Since the intermediate


deep waters of the SCS during the glacials were only


with Pacific Intermediate

waters through

the Bashi Strait, the bottom water chemistry

at the

studied site is believed

to be mainly



that of the Pacific Intermediate

waters. The good


of carbonate during glacials was due to

reduced production

of North Atlantic

Deep Waters


and probably

the formation

of nutrient-


Pacific intermediate


at that time

(Thunell et al., 1992).

Gephyrocapsa tends to dominate in highly fertile

waters. Winter (1982) found that the distribution


G. oceanica was closely related to the concentration

of phosphate in the Gulf of Elat (‘Aqaba), Red Sea.

The abundance

of G.

oceanica in the northern South

China Sea also shows a decreasing

trend with con-


of phosphate from south to north (Cheng,

1992). The predominance

of large

Gephyrocapsa in

the nannoflora

(Fig. 6), in association

with the high


of organic carbon and ketones (CY)


100 K.-Y Wei et al. /Marine Micropaleontology 32 (1997) 95-114












A ;\,’

4.8 ka 0


10.2 ka a 30



13.2ka AXIS II ?? ??W . 20 ?? 13.4 ka ??Last Glacial ??



ka 4.1 ka 0 0 0 ” I ‘Y I




45 60


SDecies Score


0 Crenalithus 200 - G. oceanica Umbellosphaera .


??(L.V) Caldosolenia ??G’c ?? C. jonesi ’ U-.D) ?? E. h ??small Reticulofenestra










G. oje&a (M)



small Gephrocapsa . ??C. leptoporus Umbilicosphaera -200 - . large Reticulofenestra . Syacocphaera


-400 I ’ I ’ I I ., . -200 -100 0 100 200 300 400



Fig. 4. Scores of samples (upper panel) and species (lower panel) on the plane of the first two eigen-axes resulting from detrended correspondence analysis.

(Fig. 7; Huang et al., 1997) suggests the existence

of a less maritime


with high fertility


during the last glacial. The surface-water


in the SCS was estimated

to be two

times higher during the last glacial maximum


in the Holocene (Thunell et al., 1992). Furthermore,


K.-l! Wei et al. /Marine Micropaleontology 32 (1997) 95-114 101

the high frequency


Rhabdosphaera spp. (Fig. 6),

a characteristic


species (Winter et

al., 1994), is also indicative

of colder conditions



to the post-glacial

stages II-IV.

The increase


Emiliania huxleyi during the



{(Stage II, 13.3-10.7

ka, Fig. 3)


an increasing


of the open-ocean

waters due to the concomitant

rise of global sea level


1989). At the same time, in the north-


Pacific the Kuroshio

Front moved north-

ward (Chinzei

et al., 1987), synchronous

with the



of the North Atlantic


(Duplessy et al., 1981).

The major drop of

Florisphaera between 13.3

and 9 ka (Fig. 3) may either indicate a rise of the


or an increase

of turbidity

of the sur-

face waters. Among the various coccolithophorids,

Florisphaera is distinctive

in living in the lower


zone with low light and high nutrients.

Its habitat is restricted to waters below 100 m and

warmer than 10°C (Okada and Honjo, 1973; Honjo

and Okada,

1974; Okada

and McIntyre,



and McIntyre



a con-

ceptual model that when the nutricline

is deeper the

growth of

Florisphaera would be enhanced relative

to surface dwelling


The relative



Florisphaera therefore is indicative of


of the nutricline

in the euphotic zone; the

higher the relative abundance,

the deeper the nutri-

cline. On the other hand, Ahagon et al. (1993) argued

that this model is only applicable

to the equatorial

areas where the nutricline

depth is mainly governed

by the equatorial


Ahagon et al. (1993)

instead documented

that in the marginal seas around

the Japanese islands, there was a close relationship


the F:

profunda abundance

and seawater


(Secchi depth). Given the proximity


the northern South China Sea to the continent and the


high sedimentation

rates of the sequence,

we follow Ahagon et al. (1993) in suggesting that the

decrease of the relative abundance



profunda is an indication of increasing turbidity of

the surface waters rather than a shallowing

of the




is probably


by increased lateral transportation

of fine sediments

carried by the newly established







Strait and conti-

nental shelves off the Pearl River (Zhujiang)



Wang, pers. commun.,

1995). The rise of

the sea level due to the global meltwater discharges

during the deglacial period (13.0-l 1.5 and 10.0-9.0

ka, Fairbanks,

1989) would have caused the Taiwan

Strait submerged

and become a passageway


which the Continent

Coastal Current flew southward

and flushed previously





shelves. The fine particles were probably


by contour currents, forming nepheloid

layers with high turbidity.

On the other hand, the timing (13-9 ka) of the

drop of the relative abundance

of E

profunda is con-

sistent with that recorded in the equatorial



and McIntyre,

1990b). If this reduction

of E

profunda abundance is a global phenomenon,


not only in the Atlantic



gence but also in the western Pacific marginal


then it might suggest either a global shallowing



depth, or a global increase of cloudiness


in less penetration

of sunlight).

Such a

global mechanism,

however, is difficult to imagine.

The intervals

dated between

12.5 and 11.5 ka

showed the best nannofossil



ing many

Umbellosphaera and Oolithotus fragilis

(Figs. 5 and 6). Planktic


also showed


good preservation

in the same interval

(Fig. 5; Huang et al., 1997). The dissolution


sity of the bottom water during the deglacial


is inferred to be the lowest throughout

the past 15.5

kyr (Fig. 5), synchronous

with the global deglacial


spike of Termination

I (Berger, 1977).

Boyle (1988) suggested that during the deglaciation

carbonate preservation

was enhanced. During Termi-

nation I the circulation

of the intermediate

waters of

the NSCS already appears closely linked with that

of the Pacific intermediate

waters, whose corrosive-

ness, in turn, was governed by the relative strength

of North Atlantic

Deep Water and Antarctic


tom Water. The occurrence of this preservation


agrees with the ‘nutrient


model’ of

Boyle (1988).



change during the deglacial

period is shown by

Gephyrocapsa. The total abun-

dance of

Gephyrocapsa (Fig. 3) continuously


creased from the glacial level through the deglacial

to the late Holocene. Furthermore,

within the


rocapsa complex, the dominance of the large forms

of G.

oceanica was first displaced by the medium-


102 K.-Z Wei et al. /Marine Micropaleontology 32 (1997) 95-114

Susceptible Species

? ?

Umbellosphaera tenuis

? ?

Oolithotus fragllis cavum

? ?

U. irregularis

? ?

0. fragilis fragllis

? ?



N. coccolithomorpha H Syracosphaera pulchra

0 2 4 6 8 10 12 14 16 18 20 22 24 26

-- Risistant/(S+R)

- frag/(fr&+foram)

0 2 4 6 8 10 12 14 16 18 20 22 24 26

AGE (Ka)

Fig. 5. Summary of microfossil indicators of carbonate dissolution in comparison with down-core variations in %CaCO, and CaCOj mass accumulation rate.

sized G. oceanica

during the deglacial

stage, and

later during the Holocene

by small forms, collec-

tively designated

as small Gephyrocupsa

(cf. Mat-

suoka and Okada, 1989). This diminution

trend is

also mirrored by the progressive


of small


forms in the upper section

of the core.

These small placoliths


Cricolithus jonesi,


sp., Dictyococcites,

as well as small


forms such as R. minuta (Fig. 4b

and Fig. 6).


104 K.-Y Wei et al. /Marine Micropaleontology 32 (1997) 95-114




G. sacculifer



16 18 24 26




22 ,


23 24 25 26 ; I. I , I. I , I


200 600 1000 1400 I . I , I . I I

Fig. 7. Time-series of 6’8O of planktic foraminifera Globigerinoides sacculifer, 6”O between surface-dwelling (G. sacculifer) and subsurface-dwelling (Globoroculia manardii) species, sea-surface paleotemperature derived from unsaturation ratio of long-chain methyl alkenones ((I$), and ketone (C37) concentration. Data are from Huang et al., 1997.

The early Holocene


ka) assemblage




Florisphaera, second by E. huxleyi

and third by

Gephyrocupsa (Fig. 3). Apparently,

the post-deglaciation


has been more mar-

itime with less terrigenous




the previous

periods. Small placoliths

became rel-

atively abundant.

These small forms are probably


elements of the flora.


which inhabits tropical oligotrophic

pelagic surface

waters (McIntyre et al., 1970; Young, 1994), also be-

came relatively more abundant. Such a floral compo-

sition indicates that it was a warmer, more maritime,




during the

early Holocene.

This interpretation

is supported by

the increase

in 6180 gradient


the surface

water (represented

by the planktic foraminifera


bigerinoides sacculijkr) and the subsurface water


m, represented


Globorotalia menardii)

(Fig. 7; Huang et al., 1997). This larger gradient

is indicative

of stronger


of the water

column. Strong monsoon winds might intermittently

deepen the mixed layer and cause nutrient injection

into the euphotic

zone, inducing


of op-


species such as small placoliths




At about 5.5 and 4.0 ka the preservation

of cal-

careous nannofossils

slightly improved. The dissolu-

tion index of planktic foraminifera

shows also a pe-

riod of better preservation

centered at 5 ka (Fig. 5).

This short period predated slightly the episodic cool-

ing of 1 to 2°C between 4.5 and 3.0 ka (Fig. 7).


K.-E Wei et al./Marine Micropaleontology 32 (1997) 95-114 105

iths decreased during the last 4000 yr while that

of larger forms such as CuZcidiscus Zeptoporus,




R. haqii

increased (Figs. 4 and 6). The

vertical 6180 gradient also increased (Fig. 7), sig-

nifying a more stratified surface-water column. The

frequency and/or intensity of seasonal mixing might

have become lower, which in turn would have re-

duced the yield of :small placoliths in the flora. By 3

ka, sea surface temperatures had become similar to

today’s (Fig. 7).

Throughout the core, the variations of nannofos-

sil preservation bear no obvious relationship to the

variations in CaCOs content or CaC03 accumulation

rate (Fig. 5). This suggests that although the preser-

vation of nannofossils may have been governed by

the general corrosiveness of the intermediate waters

of the SCS, the influence is not necessarily reflected

in the accumulation rates of CaCOs or %CaCOs.

The depth of this site (2050 m) is slightly above

the present nannofossil lysocline (2300 m, Chen and

Shieh, 1982). The nannofossil assemblage therefore

is considered to be less sensitive to the glacial deep-

ening of the lysocline and carbonate compensation

depth (CCD, at 3000 m at present and 4000 m during

the LGM; Rottman, 1979; Thunell et al., 1992) than

other deeper sites. Owing to this site’s proximity to

the continent and riverine sediment input sources of

the Pearl River (Zhujiang) and the Hanjiang River,

the influx of terrestrial sediments is considered to

be the primary controlling factor of sedimentation

of this site, while the bottom water chemistry is


5. Summary and conclusions

(1) A quantitative study was performed on last

glacial to Holocene (25.9-1.2 ka) nannofossils from

a piston core collected from the northern South

China Sea. Integrated with other proxies, the time-

progressive change in the nannoflora reveals the

paleoceanographic evolution of the area since the

last glacial.

(2) A detrended correspondence analysis of the

nannofloral data suggests that the floral succession

can be subdivided into four stages, corresponding

to the major paleoceanographic developments in this

area: Stage I (25.9-13.3 ka), glacial stage; Stage II

(13.3-10.7 ka), deglacial stage; Stage III (10.74.4

ka), early Holocene and Stage IV (4.4-1.2 ka), late


(3) The glacial stage is characterized by the domi-

nance of





strong continental influence with high fertility. Good

preservation of nannofossil and high concentration

of ketone and organic carbon are caused by both

high biological productivity and less corrosiveness

of the bottom waters.

(4) The increase of

E. huxleyi

in the deglacial

interval suggests an increasing influence of the open-

ocean, corresponding to the global sea-level rise. The

decrease of a deep-dwelling taxa,

Florisphaera pro-


is interpreted as responding to the increased

turbidity in surface waters due to the increased in-

put of fine, flushed sediments during the deglacial

period by the Continental Coastal Current. A preser-

vation spike of nannofossils centered at 12 ka is also


(5) A time-transgressive

diminution trend of


existed through the glacial-Holocene,

parallel to the decreasing trend of the relative abun-

dance of


This trend is accompanied

by the increase of other small placoliths, espe-

cially during the Holocene. This might suggest that

the NSCS became more maritime and oligotrophic

through time, while the surface waters became more

stratified as indicated by the increase of the vertical

6180 gradient. These small placoliths were oppor-

tunistic elements generated during blooming seasons

when seasonal mixing of the surface waters took


(6) Two episodic cooling events were identified

in the post-glacial stages, one at about 11-10 ka,

the other at 5-3 ka. Both are characterized by bet-

ter preservation of nannofossil assemblages and a

drop in sea-surface paleotemperature of l-2°C as

indicated by the U,k index.


We would like to thank Tom Marchitto, Pinxian

Wang, Jeremy Young and an anonymous reviewer

for improving the manuscript with their comments

and criticisms. This paper is a contribution of the

ROC PAGES project. Research funding was pro-

vided by Grant No. NSC84-261 l-M-002~OO4GP of

the National Science Council, ROC.


106 K.-E Wei et al. /Marine Micropaleontology 32 (1997) 95-114




K.-E Wei et al. /Marine Micropaleontology 32 (1997) 95-114 107

Appendix A. Thxonomic remarks on selected


Small elliptical placoliths

Small placoliths belonging to Reticulofenestra, Dictyococ- cites, Gephytocapsa, Emiliania and Crenalithus ate often hard to adequately differentiate under the light microscope due to their small sizes. For the sake of stability and reproducibility of taxonomic identification, we mainly followed strategies outlined

Plate I

Electron micrographs of nannofossils from Core 90-36. Scale bar =3pm.

1. Emiliania huxleyi (Lohmann) Hay and Mohler, proximal view. Subdepth = 17 cm.

2. Emiliania huxleyi (Lohmann) Hay and Mohler, distal view. Subdepth = 17 cm.

3. Reticulofenestra minutula (Gartner) Haq and Berggren, proxi- mal view. Subdepth = 62 cm.

4. Reticulofenestra minuta Roth, proximal view. Subdepth = 62 cm.

5. Cricolithus jonesi Cohen, distal view. Subdepth = 98 cm. 6. Umbellosphaera irregularis Paasche, proximal view. Subdepth = 62 cm.

7. Umbellosphaera tenuis (Kamptner) Paasche, distal view. Sub- depth = 17 cm.

8. Umbilicosphnera sibogae (Weber-van Bosse) Gaarder, distal view. Subdepth = 8 cm.

9. Umbilicosphaera sibogae (Weber-van Bosse) Gaarder, proxi- mal view. Subdepth = 44 cm.

10. Gephyrocapsu oceanica Kamptner (medium-sized, diagonal bar), distal view. Subdepth = 17 cm.

11. Gephyrocupsa oceanica Kamptner (medium-sized, vertical bar), distal view. Subdepth = 17 cm.

12. Calcidiscus leptoporus (Murray and Blackman) Loeblich and Tappan, proximal view. Subdepth = 98 cm.

13. Uolithotus fragilis (Lohmann) Okada and McIntyre, distal view. Subdepth = 17 cm.

14. Oolithotus fragilis (L.ohmann) Okada and McIntyre, proximal view. Subdepth = 98 cm.

15. Syracosphaera pulchra Lohmann, distal view. Subdepth = 98 cm.

16. Syracosphaera lamina Lecal-Schlauder, proximal view. Sub- depth = 17 cm.

17. Syracosphaera lumina Lecal-Schlauder, distal view. Sub- depth = 17 cm.

18. Syrucosphaera sp. distal view. Subdepth = 98 cm. 19. Syracosphaeru mediterrunea Lohmann, distal view. Subdepth = 62 cm.

20. Helicosphneru carteri carteri, (Wallich) Kamptner, distal view. Subdepth = 44 cm.

21. Helicosphaera carteri hyalina, proximal view. Subdepth = 44 cm.

22. Helicosphaera carteri wullichii (Lohmann) Okada and McIn- tyre, proximal view. Subdepth = 44 cm.

by Rahman and Roth (1989), Matsuoka and Okada (1989) and Biekart (1989). The common spirit of these taxonomic schemes is their simplicity and practicality. Differentiation of the vari- ous genera is based on features that are recognizable under the light microscope, such as the overall size, presence/absence of a bridge, relative size of the central opening and extinction pattern under crossed-nicols

Elliptical placoliths showing the typical extinction pattern of Gephymcapsa but without visible bridges were counted as species of Reticulofenestra and Dictyococcites. Specimens with a more or less open central area were assigned to Reticulofenestra, while those with a completely closed central area were identified as Dictyococcites (Matsuoka and Okada, 1989). Two Dictyococ- cites species were recognized based upon size criteria (Backman, 1980; Matsuoka and Okada, 1989): D. productus (smaller than 4 pm) and D. perplexa (larger than 4 pm = D. antarcticus Haq, 1976).

Forms smaller than 4 bm showing a bright inner rim and faint birefringence along the outer margin were considered to be Emiliunia huxleyi. The slits between the ‘T-shaped’ elements on the distal shield are often visible under parallel nicols. However, small etched specimens of Reticulofenestru and Dictyococcites, as well as Gephyrocapsa with its bridge totally dissolved, may possibly be mis-identified as E. huxleyi. Forms similar to E. huxleyi in size, but showing stronger birefringence and a similar extinction pattern to that of Reticulofenestra, are considered to be Crenalithus sp. (Chen, 1978). For some authors, however, Crenalithus is considered to be a synonym of Reticulofenestra (Backman, 1980) or Dictyococcites (Biekart, 1989) (discussed later).

Forms close in size to R. minutula but showing a large central opening surrounded by a narrow rim of strong birefrin- gence were designated as Cricolithus jonesi (Cohen, 196.5; Chen, 1978, 1979). However, dissolved small syracosphaerids might be mistaken for this species under the light microscope.

Crenalithus and Reticulofenestra

Roth (1973) erected the genus Crenalithus for the small el- liptical placolith with the type species originally described as Coccolithus domnicoides Black and Barnes (1961). This genus is characterized by its small size, serrate margin and non-imbri- cated elements in the distal shield. From observation of recent coccolithophorids, Okada and McIntyre (1979) adopted the con- cept of this genus and erected new species and subspecies, C. parvulus parvulus, C. pan&us tecticentrum, C. punctatus and C. sessilis. They considered these living species to resemble the extinct C. doronicoides. Nishida (1979) and Steinmetz (1991) also reported living Crenalithus species. The proximal side of the central area of these living Crenalithus species is covered by either a solid plate or reticulate grid.

In this study, we classified those specimens which resemble R. minuta but with a larger central opening as Crenalithus sp.

(sensu Chen, 1978). However, these Crenalithus forms should

considered as small Reticulofenestra if one follows Backman’s argument (Backman, 1980). These forms are also possibly some dissolved forms of Dictyococcites productus.


108 K.-Y Wei et al. /Marine Micropaleontology 32 (1997) 95-114


K.-I: Wei et al. /Marine Micropaleontology 32 (1997) 95-114 109 cussed above, instead, we tried merely to differentiate as many

forms as possible to explore the paleoenvironmental significance of these small placoliths. Therefore, we tentatively retain the species concept of Crenalithus sp. proposed by Chen (1978) to encompass all the small elliptical placoliths (t3 pm) that have a large central opening and show a thin rim resembling the extinc- tion pattern of Reticulofenestra under crossed polarized light. We observed that such forms do exist in our samples and they could be consistently differentiated from other reported Reticulofenes- tra and Dictyococcites species, although the true identities of these Crenalithus awaits further detailed study.

Crenalithus sp. Chen (F’late III, 16)

Basionym: Chen, 1978, p. 139, pl. 4, figs. l-2.

Remarks: Small elliptical placolith (13 ,um) with non-imbri- cated or slightly overlapping elements. Under scanning electron microscope, it shows a smooth surface with little relief on the distal side. The elements are relatively wide and show a serrate margin. Central area is open. Under crossed polarized light, it

Plate II

Electron micrographs of nannofossils from Core 90-36. Scale bar =3pm.

I. Culciosolenia murrayi Gran. Subdepth = 17 cm.

2. Florisphaera profunda elongata (the larger one) and Florisphaera profunda profunau (the smaller one). Subdepth = 17 cm.

3. Discosphaera tubiferu (Murray and Blackman) Ostenfeld. Subdepth = 98 cm.

4. Discoaster sp. reworked. Subdepth = 62 cm.

5. Rhabodosphaera clavigera Murray and Blackman. Subdepth = 44 cm.

6. Rhabdosphnera clavigera Murray and Blackman. Subdepth = 62 cm.

7. Neosphaera coccolithomorpha Lecal-Schlauder, distal view. Subdepth = 17 cm.

8. Pontosphaeru multipora (Kamptner) Roth, distal view. Sub- depth = I7 cm.

9. Pontosphaera discopora Schiller. Subdepth = 98 cm. 10. Pontosphaera japonrca (Takayama) Nishida, proximal view. Subdepth = 98 cm.



Ceratolithus crisratus Kamptner. Subdepth = 62 cm. 12. Coccosphere of Gephyrocapsa oceanica Kamptner. Subdepth = 98 cm.

13. Thoracosphaera heimii (Lohmann) Kamptner. Subdepth = 62 cm.

14. Scyphosphaera sp. Subdepth = 98 cm. Backman (1980) considered Crenalithus to be a junior synonym of the genus Reticulofenestra. He rejected the specific epithet doronicoides and identified three species of small Reticulofenestru for the late Neogene: R. mint&da (3-5 wrn with a relatively large central opening), R. haqii (3-5 pm with a relatively smaller central opening) and R. minuta (~3 pm). We did not differentiate R. parvula (~2 pm in size = Crenalithus parvulus Okada and

McIntyre, 1979, sensu Biekart, 1989) from R. minutu.

shows a bright rim with an extinction pattern resembling that of Reticulofenestra. It differs from Dictyococcites productus and Reticulofenestra minuta by its open central area and narrow rim. It is distinguished from Gephyrocapsa by lacking a crossbar in the central opening.

Cricolithus jonesi Cohen (Plate I, 5; Plate III, 10)

Basionym: Cricolithus jonesi Cohen, 1965, p, 16, pl. 2, figs. J. K; pl. 16, figs. a-c.

References: Chen, 1978, p. 140, pl. 5, figs. 1-3; Chen, 1979, fig. 5; Chen and Shieh, 1982, pl. 1, figs. 5-8.


The proliferation of taxa assigned to the genus Gephyro- capsa and the associated variable concepts of species have caused a lot of controversies among specialists (Rahman and Roth, 1989; Matsuoka and Okada, 1989; Biekart, 1989). One of the problems associated with this group is the small size of its various forms such that the morphological characteris- tics of many species/subspecies are only recognizable under the scanning electron microscope. To facilitate a stable taxonomic practice under the light microscope, we have adopted the size criteria of Matsuoka and Okada (1989) to classify the Gephyro- capsa into three size classes: large forms (maximum diameter >5 pm), medium forms (between 5 and 2.5 pm) and small forms (~2.5 wm). Medium and large Gephyrocapsa were classified as G. caribbeanica or G. oceanica. The former has a smaller cen- tral area spanned by a robust bridge, whereas G. oceanicu has a relatively larger central opening. Specimens of G. oceanica were further assigned into two varieties: one with a vertically oriented bridge against the vertical axis and the other with a diagonally oriented bridge. The large G. oceanica with a diagonal bridge [abbreviated as G. oceanica (large, diagonal)] is similar to G. lumina Bukry (1973) while the large G. oceanica (vertical) is close to G. omega Bukry (1973). All small Gephyrocupsa forms were lumped as small Gephyrocapsa without further splitting. Some of these small Gephyrocapsa specimens, however, can be easily identified as G. ericsonii.

Genus Helicosphaera Kamptner, 1954

Some authors have proposed that Helicosphaeru wallichii and H. hyalina should be regarded as varieties of H. carteri (Theodoridis, 1984, Okada, 1992). We followed the decision made in the Florence Meeting of INA to include these forms as intraspecific varieties of H. carteri (Jordan and Young, 1990). Three varieties have been recognized as shown in Plates I and IV.

Reticulofenestra minuta Roth (Plate I, 4)

Basionym: Reticulofenestra minuta Roth, 1970, p. 850, pl. 5, figs. 3, 4.

Synonyms: Dictyococcites minutus (Haq) Haq, Lohmann and Wise, 1976, p. 759. Gephyrocapsa doronicoides (Black and Barnes) var. 4 Pujos, 1985, p. 563, pl. 2, fig. 1.

References: Biekart, 1989, pl. 1, fig. 4a, b; pl. 6, fig. 6a-c; pi. 7, fig. 2a, b.


110 K.-Y Wei et al. /Marine Micropaleontology 32 (1997) 95-114















Plate III






Optical micrographs of nannofossils from Core 90-36. Scale bar = 5 pm.

1. Gephyrocupsa oceanica Kamptner (large-sized, vertical bar). Subdepth = 62 cm. (a) Transmitted light. (b) Cross-polarized light. 2. Gephyrocapsa oceanica Kamptner (large-sized, diagonal bar). Subdepth = 62 cm. (a) Transmitted light. (b) Cross-polarized light, 3. Gephymcupsa oceanica Kamptner (medium-sized, vertical bar). Subdepth = 62 cm. (a) Transmitted light. (b) Cross-polarized light. 4. Gephyrocapsa oceanica Kamptner (medium-sized, diagonal bar). Subdepth = 62 cm. (a) Transmitted light. (b) Cross-polarized light. 5. Gephyrocapsa caribbennica Bordeaux and Hay. Subdepth = 62 cm. (a) Transmitted light, (b) Cross-polarized light,

6. Small Gephyrocapsa sp. Subdepth = 62 cm. Cross-polarized light.

7. Small Gephyrocapsa spp. (the small one = Gephyrocapsa ericsonii McIntyre and Be). Subdepth = 62 cm. Cross-polarized light, 8. Emiliania huxleyi (Lohmann) Hay and Mohler. Subdepth = 62 cm. Cross-polarized light.

9. Emiliania huxleyi (Lohmann) Hay and Mohler. Subdepth = 62 cm. Cross-polarized light. 10. Cricolithus jonesi Cohen. Subdepth = 62 cm. Cross-polarized light.

11. Reticulofenestru minutula (Gartner) Haq and Berggren. Subdepth = 62 cm. (a) Transmitted light, (b) Cross-polarized light. 12. Reticulofenestra haqii Backman. Subdepth = 62 cm. Cross-polarized light.

13. Reticulofenestra minuta Roth. Subdepth = 62 cm. Cross-polarized light. 14. Dictyococcites perplexu Burns. Subdepth = 62 cm. Cross-polarized light.


K.-Y: Wei et al. /Marine Micropaleontology 32 (I 997) 95-114 111 Syracosphaera

Gaarder and Heimtdal (1977) provided a detailed review of the systematic paleontology of this genus and lucid descriptions of species. We basically adopted their taxonomic scheme except that we followed Jordan and Young (1990) and Jordan and Kleijne (1994) in retaining all Coronosphaera species in the genus Syracosphaera.

and Paleoclimatic History of Late Neogene Sediments of the Northern Florida Continental Shelf. Ph.D. Diss., Texas A&M Univ., College Station, TX (unpubl.).

Chen, M.-P., 1979. Late Pleistocene calcareous nannoplankton in southern Okinawa Trough. Acta Oceanogr. Taiwan. 10, 95-l 18.


Chen, M.-P., Huang, C-K., 1995. The distribution of calcareous nannofossils and grain size in the surface sediments of the East China Sea and their relationship to the current pattern. TAO 6 (l), 129-150.

Ahagon, N., Tanaka, Y., Ujiie, H., 1993. Florisphaera profunda, a possible nannoplankton indicator of late Quaternary changes in sea-water turbidity at the northwestern margin of Pacific. Mar. Micropaleontol. 22, 255-273.

Backman, J., 1980. Miocene-Pliocene nannofossils and sedimen- tation rates in the Hatton-Rockall Basin, N.E. Atlantic Ocean. Stockholm Contrib. Geol. 36 (1), 91 pp.

Berger, W., 1977. Deep sea carbonate and the deglaciation preservation spike in pteropods and foraminifera. Nature 269, 301-304.

Chen, M.-I?, Shieh, K.-S., 1982. Recent nannofossil assemblages in sediments from Sunda Shelf to abyssal plain, South China Sea. Proc. Nat. Sci. Count. (Taiwan, ROC) A 6, 250-285. Cheng, X., 1992. Calcareous nannofossils in surface sediments

of the central and northern parts of the South China Sea. J. Micropalaeontol. 11, 167-176.

Biekart, J.W., 1989. The distribution of calcareous nannoplank- ton in late Quaternary sediments collected by the Snellius II Expedition in some southeast Indonesian basins. Proc. K. Ned. Akad. Wet. Ser. B 92 (2), 77-141.

Black, M., Bames, B., 1961. Coccoliths and discoasters from the floor of the South Atlantic Ocean. J. R. Microsc. Sot. 80 (2). 137-147.

Chinzei, K., Fujioka, K., Kitazato, H., et al., 1987. Postglacial environmental change of the Pacific Ocean off the coast of central Japan. Mar. Micropaleontol. 11, 273-291.

Cohen, C.L.D., 1965. Coccoliths and discoasters from Adriatic bottom sediments. Leidse Geol. Med. 35, 14.

Duplessy, J.C., Delibrias, G., Turon, J.L., Pujol, C., Dupart, J., 1981. Deglacial warming of the northeastern Atlantic Ocean: Correlation with the paleoclimatic evolution of the European continent. Palaeogeogr., Palaeoclimatol., Palaeoecol. 35, 121-


Boyle, E.A., 1988. The role of vertical chemical fractionation in controlling Late Quatemary atmospheric carbon dioxide. J. Geophys. Res. 93, 15701-15714.

Bukry, D., 1973. Coccolith stratigraphy, eastern equatorial Pa- cific, Leg 16, Deep Sea Drilling Project. Init. Rep. DSDP 16, 653-711.

Fairbanks, R.A., 1989. A 17,000-year glacio-eustatic sea level record: Influence of glacial melting rates on the Younger Dryas event and deep sea circulation. Nature 342, 637-642. Gaarder, K.R., Heimdal, B.R., 1977. A revision of the

genus Syrucosphaeru Lohmann (Coccolithineae). ‘Meteor’ Forsch.-Ergeb. D 24, 54-71.

Chen, M.-P., 1978. Calcareous Nannoplankton Biostratigraphy

Plate III (continued)

15. Dictyococcites productus (Kamptner) Backman. Subdepth = 62 cm. Cross-polarized light.

16. Crenalithus sp. Subdepth = 62 cm. Cross-polarized light. 17. Syracosphaera hisrrica Kamptner. Subdepth = 62 cm. Cross- polarized light.

18. Syracosphaera pulchra Lohmann. Subdepth = 26 cm. Cross- polarized light.

19. Syracosphaera sp. Subdepth = 62 cm. (a) Transmitted light. (b) Cross-polarized light.

20. Syracosphaera sp. Subdepth = 23 cm. Cross-polarized light. 21. Syracosphaera lamina Lecal-Schlauder. Subdepth = 26 cm. Cross-polarized light.

22. Discosphaera tubqera (Murray and Blackman) Ostenfeld. Subdepth = 62 cm. (a) Transmitted light. (b) Cross-polarized light.

Herguera, J.C., 1992. Deep-sea benthic foraminiferal and bio- genie opal: Glacial to postglacial productivity changes in the western equatorial Pacific. Mar. Micropaleontol. 19, 79-98. Herguera, J.C., Berger, W.H., 1991. Paleoproductivity from ben-

thic foraminifera abundance: Glacial to postglacial change in the west-equatorial Pacific. Geology 19, 1173-l 176. Herguera, J.C., Berger, W.H., 1994. Glacial to postglacial drop

in productivity in the western equatorial Pacific: Mixing rate vs. nutrient concentrations. Geology 22, 629-632.

Hill, M.O., Gauch, H.G.Jr., 1980. Detrended correspondence analysis: An improved ordination technique. Vegetatio 42, 47-58.

Honjo, S., Okada, H., 1974. Community structure of coccol- ithophores in the photic layer of the mid-Pacific. Micropaleon- tology 20, 209-230.

Huang, C.-Y., Wu, S.-F., Zhao, M., Chen, M.-T., Wang, C.-H., Tu, X., Yuan, P.B., 1997. Last glacial to interglacial surface ocean variability in the South China Sea: A high-resolution record of sea-surface temperature, productivity, and South- east Asia monsoon. Mar. Micropaleontol. 32 (l-2) (this vol- ume).

23. Rhabodosphaera clavigeru Murray and Blackman. Subdepth Jordan, R.W., Kleijne, A., 1994. A classification system for liv- = 62 cm. (a) Transmitted light. (b) Cross-polarized light. ing coccolithophores. In: Winter, A., Siesser, W.G. (Ed%), 24. Rhabodosphaera clavigera Murray and Blackman. Subdepth Coccolithophores. Cambridge Univ. Press, Cambridge, pp. = 23 cm. (a) Transmitted light. (b) Cross-polarized light. 83-105.


112 K.-Y. Wei et al. /Marine Micropaleontology 32 (1997) 95-114 ‘: * *, +‘ * ;












Plate IV






Optical micrographs of nannofossils from Core 90-36. Scale bar = 5 pm.

1. Helicosphaera carter-i hyalina Gaarder. Subdepth = 23 cm. (a) Transmitted light. (b) Cross-polarized light.

2. Helicosphaera carteri wallichii (Lohmann) Okada and McIntyre. Subdepth = 62 cm. (a) ‘Transmitted light. (b) Cross-polarized light. 3. Helicosphaera carteri carteri, (Wallich) Kamptner. Subdepth = 26 cm. (a) Transmitted light. (b) Cross-polarized light.


Fig.  1.  Locations  of  Core  SCS90-36  studied  and  Core  RC26-16  mentioned  in  the  text
Fig. 1. Locations of Core SCS90-36 studied and Core RC26-16 mentioned in the text p.2
Fig.  2.  Depth-age  diagram  of  Core  SCS90-36,  showing  sedi-  mentation  rates  estimated  from  eight  AMS  14C  dates  of  planktic  foraminifera
Fig. 2. Depth-age diagram of Core SCS90-36, showing sedi- mentation rates estimated from eight AMS 14C dates of planktic foraminifera p.3
Fig.  3.  Downcore  variation  of  the  relative  abundances  of  three  dominant  nannofossil  taxa
Fig. 3. Downcore variation of the relative abundances of three dominant nannofossil taxa p.5
Fig.  4.  Scores  of  samples  (upper  panel)  and  species  (lower  panel)  on  the  plane  of  the  first  two  eigen-axes  resulting  from  detrended  correspondence  analysis
Fig. 4. Scores of samples (upper panel) and species (lower panel) on the plane of the first two eigen-axes resulting from detrended correspondence analysis p.6
Fig.  5.  Summary  of  microfossil  indicators  of  carbonate  dissolution  in  comparison  with  down-core  variations  in  %CaCO,  and  CaCOj  mass  accumulation  rate
Fig. 5. Summary of microfossil indicators of carbonate dissolution in comparison with down-core variations in %CaCO, and CaCOj mass accumulation rate p.8
Fig.  7.  Time-series  of  6’8O  of  planktic  foraminifera  Globigerinoides  sacculifer,  6”O  between  surface-dwelling  (G
Fig. 7. Time-series of 6’8O of planktic foraminifera Globigerinoides sacculifer, 6”O between surface-dwelling (G p.10



相關主題 :