delta O-18, Sr/Ca and Mg/Ca records of Porites lutea corals from Leizhou Peninsula, northern South China Sea, and their applicability as paleoclimatic indicators

Combined seasonal to monthly resolution coral skeletal delta(18)O, Sr/Ca, and Mg/Ca records are reported for one modem and two late Holocene Porites lutea corals from a fringing reef at Leizhou Peninsula, the northern coast of the South China Sea (SCS). All the profiles for the period 1989-2000 reveal annual cycles well correlated with instrumental sea surface temperatures (SST), and display broad peaks in summer and narrow troughs in winter, reflecting seasonal growth rate variations. Calibration against instrumental SST yields the following equations: delta(18)O=-0.174(+/- 0.010)xSST(degrees C)-1.02(+/- 0.27) (MSWD=5.8), Sr/Ca-(mmol/mol)=-0.0424(+/- 0.0031)xSST(degrees C)+9.836(+/- 0.082) (MSWD=8.6), and Mg/Ca-(mmol/mol)=0.110(+/- 0.009)XSST(degrees C)+ 1.32(+/- 0.23) (MSWD=55). The scatter in the Mg/Ca-SST relationship is much larger than analytical uncertainties can account for, suggesting the presence of SST-unrelated components in the Mg/Ca variation. Calculated Sr/Ca-SST values for two later Holocene Porites lutea samples (U-series ages similar to 541 BC and similar to 487 AD, respectively) from the same reef suggest that SST in the SCS at similar to 541 BC was nearly as warm as in the 1990s (the warmest decade of the last century), but at similar to 487 AD, it was significantly cooler. This observation is consistent with climatic data reported in Chinese historic documents, confirming that the Sr/Ca-SST relationship is a reliable thermometer. Removing the SST component in the delta(18)O variation based on calculated Sr/Ca-SST values, the residual delta(18)O reflects the deviation of the Holocene seawater delta(18)O from the modem value, which is also a measure of the Holocene sea surface salinity (SSS) or the summer monsoon moisture level in mainland China. Such residual delta(18)O was close to zero at similar to 541 BC and -0.3 parts per thousand at similar to 487 AD, suggesting that it was as wet as in the 1990s at similar to 541 BC but significantly drier at similar to 487 AD in mainland China, which are also consistent with independent historic records. Calculated Mg/Ca-SST values for the two late Holocene corals are significantly lower than the Sr/Ca-SST values and are also in conflict with Chinese historic records, suggesting that coral Mg/Ca is not reliable proxy for SST. At comparable Sr/Ca ranges, fossil corals always display negative Mg/Ca offsets if compared with the modem coral of the same site. We interpret this observation as due to preferential loss of Mg during meteoric dissolution of cryptic Mg-calcite-bearing microbialites in the exposed fossil corals. Microbialites (MgO up to 17%, Sr only 100-300 ppm) are ubiquitous during reef-building processes and their presence in only a trace amount will have a significant impact on coral Mg/Ca ratios without detectable influence on coral Sr/Ca ratios. (c) 2004 Elsevier B.V. All rights reserved.