High Development

NOTES

Despite its great success and importance, the TIMS

Timing of Holocene sea-level U-series dating method has not been well developed in

highstands by mass spectro- China. Till now, only China University of Science and

Technology is known to have established this technique

metric U-series ages of a coral and published some preliminary results on corals, speleo-

thems, marine sediments and recent volcanics[9]. Follow-

reef from Leizhou Peninsula, ing the purchase of new-generation TIMS and

South China Sea MC-ICPMS, however, we predict that this method will

soon be developed in a number of other laboratories in

1 2 China. In this note, we report TIMS U-series dates for

ZHAO Jianxin & YU Kefu

Holocene Porites corals collected from Leizhou Peninsula

1. Department of Earth Sciences, University of Queensland, Brisbane,

in the South China Sea, with an attempt to obtain some

Qld 4072, Australia (e-mail: *.****@*****.**.***.**);

insight into post-glacial sea level changes and coral reef

2. South China Sea Institute of Oceanology, Chinese Academy of Sci-

ences, Guangzhou 501301, China (e-mail: ******@*****.**.**) development. It is expected that the analytical procedure

described in this study will be beneficial to colleagues in

Abstract Systematic mass spectrometric 230Th ages are

reported for a Porites coral reef terrace from Dengloujiao, China in the establishment of mass spectrometric U-series

Leizhou Peninsula, South China Sea. Seven episodes of coral dating techniques in the near future.

growth were recognized in this terrace: 7125 96, 6764 29,

1 Geographic location of coral reefs and sample se-

5826 37, 5006 54, 2543 24, 1915 15, and 1513 22 calendar

lection

years before present (cal. aBP). 50% of the coral age popula-

tion fall between 7200 and 6600 cal. aBP, marking

Southwest Leizhou Peninsula is located at the

post-glacial stabilization of global sea level. Considering the

south-most margin of Chinese continent, where China s

facts that Dengloujiao reef flat was measured at 1.6 2.5

best-preserved Porites corals occur. The area represents

m above modern tidal datum plane; modern Porites

the northern margin of the South China Sea and the world

corals in the South China Sea are living at least ~1 m below

coral reef distribution belt, and is an ideal base for inves-

the modern tidal datum plane; the top 20 30 cm of the

tigation of Holocene environmental change in China. This

reef was eroded; and crustal subsidence in the region

since mid-Holocene was negligible, we conclude that the area experienced sustained subsidence from Tertiary,

which was stabilized since mid-Holocene[10 12]. During

above age groups record at least two major periods (7200

5000 and 2500 1500 cal. aBP) of high sea-level at least 2.9 the last few decades, minor subsidence was also record-

3.8 m above the present-day level.

ed[10]. Based on measured Holocene sedimentation rate in

the vicinity[13], the total subsidence during the last 5000

Keywords: coral, South China Sea, U-series dating, sea level, Holo-

cene.

years should be less than 0.2 m.

Systematic investigations of coral reefs in the region

High-precision and accurate dating of coral reefs, a

have been undertaken since 1993, which allows recon-

spectacular feature in tropical and sub-tropical oceans, is

struction of a 18O-based sea surface temperature (SST)

important for the study of ocean ecosystem evolution, sea

record from modern Porites coral[14] and 14C dating of

level change, sea surface temperature and salinity varia-

Holocene corals from two coral reef terraces at Denglou-

tion, and for understanding the impact of human activities

jiao (20 12 N, 109 56 E) and Shuiweicun (20 24 N,

on global climate and ecosystem. Recent development of

109 52 E)[10]. Field measurements indicate that the ele-

thermal ionization mass spectrometric (TIMS) techniques

vations of two reef terraces are ca. 1.6 2.5 m and 1.2

for rapid and high-precision measurements of 230Th and

2.3 m above the modern tidal datum plane, respectively.

234

U isotopes in the U-series decay chain has revolution-

Based on the facts that modern Porites corals in the

ized Quaternary geochronology, and resolved many fun-

South China Sea are living at least ~1 m below the mod-

damental issues in palaeoclimate change, human evolution

ern tidal datum plane; the top 20 30 cm of the reef

and magma genesis over the past 500000 years before

flats has been eroded; and crustal subsidence in the

present (aBP). This method is particularly successful for

region since mid-Holocene is negligible, Nie et al.[10] ar-

dating corals. Important outcomes of such studies include

gued that the sea level at the time of the coral reef forma-

accurate calibration of 14C time-scale beyond the range of

tion was 2.5 3.8 m above the present tidal datum plane.

dendrochronology[1], and accurate timing and characteri-

A large number of Porites samples have been col-

zation of the last and previous interglacial sea-level high

lected from both reef terraces. In this study, we report new

stands[2 5], and characterization of sea level fluctuations

TIMS U-series ages for 13 Porites samples from Den-

during the last glacial-interglacial cycles[6 8].

348 Chinese Science Bulletin Vol. 47 No. 4 February 2002

gloujiao. The results allow recognition and precise timing spike was calibrated against HU-1 standard, whose iso-

of major episodes of mid to late-Holocene sea-level high- topic systematics are known to be in secular equilibrium,

with 230Th/234U activity ratio = 1. This calibration method

stands or reef-building stages in the region.

has the advantage of minimising errors propagated from

2 Analytical procedure

decay constant uncertainties for 230Th and 234U.

Detailed analytical procedure was described in ref.

3 Results

[15]. Sample chips weighing ~0.5 g were ultrasonically

cleaned, spiked with a 229Th-233U mixed tracer, and com- A total of 16 U-Th ages were determined for 13

Porites samples and the results are presented in table 1

pletely dissolved in double-distilled HNO3. After the total

dissolution, concentrated H2O2 was added to decompose and illustrated in fig. 1. All samples contain >99% arago-

organic matter and to ensure complete spike-sample mix- nite and show no visual evidence of diagenesis or altera-

tion. Eleven samples have 234U(T) values of 146 3

ing. After standard Fe hydroxide co-precipitation to

pre-concentrate U and Th, the precipitates were re-diss- 156 2, within the range for modern seawater (144 7) and

corals (149 4)[17, 18], suggesting that diagenetic influence

olved in 0.2 mL 7 mol/L HNO3 and loaded onto pre-con-

ditioned 0.8 mL Dowex AG-1 X8 200-400# anion col- on these samples is minimal and their U-Th ages are reli-

umns held in universal pipette tips. After being washed able. However, two samples (DL-01 and DL-07) have

significantly lower 234U(T) values (9 3 and 117 3) and

four times with 0.8 mL 7 mol/L HNO3, Th was eluted

three times with 0.8 mL 6 mol/L HCl and U three times elevated U (3.83 and 3.57 g/g), reflecting some degrees

with 0.8 mL H2O. The Th fraction was further purified of U mobility and U-Th isotopic disturbance. Compared

using the same column by washing six times with 0.8 mL with corals from other marine settings such as Vanuatu,

HNO3 and then collected two times with 1 mL 6 mol/L Huon Peninsula or Great Barrier Reef, corals in this study

contain quite high detrital Th (as reflected by their 232Th

HCl. U and Th fractions were loaded separately onto

zone-refined rhenium single filaments and sandwiched abundances), probably because coral colonies here are

into two graphite layers. A Fisons VG Sector 54 30 close to the continent where the water mass may have

thermal ionization mass spectrometer equipped with a received a significant detrital input. Because of this, all

WARP filter and an ion-counting Daly detector was used calculated U-Th ages are corrected for detrital contribu-

to determine U and Th concentrations and isotopic com- tion, assuming that the detrital component s U-Th isotopic

positions. Th isotopes were measured semi-automatically compositions are similar to average upper continental

in the sequence of 232Th-230Th-229Th and analytical preci- crust[16]. This correction results in a large age error magni-

sion of 0.1% 0.3% at 2 level for 229Th/230Th was usu- fication for some samples with high detrital Th (e.g.

ally achieved for a 230Th load of 1 30 pg. U isotopic ra- DL-04-2, DL-10 and DL-13). Nevertheless, among 13

samples analyzed, 10 samples have 232Th

tios were determined fully automatically in two steps. In

the first step, 233U-235U-238U peaks were measured when trital correction on these samples are not too high and

238

U signal was about (1 2) 10 13 A and 233U/235U and their age precision is still much better than 14C dates. Our

233

U/238U ratios were corrected for mass fractionation by correction scheme is accurate, as reflected by replicate

normalizing to 238U/235U = 137.88. In the second step, analyses for DL-05 and DL-09. These two samples have

233

U-234U-235U signals were measured at a higher tem- variable detrital 232Th but after large detrital correction

perature with a 235U signal of (0.5 5) 10 14 A. Mass duplicate analyses for each sample yield nearly identical

fractionation factor for this step was calculated by nor- ages, respectively. In addition, the reliability of our dating

malizing the measured 233U/235U in the second step against method is also verified by the high reproducibility for

the corrected 233U/235U in the first step. This new mass samples ranging from 100 to 6700 a of age (table 1).

fractionation factor was then used to correct the 234U/235U

4 Discussion

ratio determined in this step. This fully-automated proce-

dure achieves an analytical precision of 0.1% 0.3% (2 ) There is a general consensus that sea-level rise de-

for the 234U/238U ratio and

Contact this candidate