Paired ¹⁴C and ²³⁰Th/U Dating of Surface Corals from the Marquesas and Vanuatu (Sub-Equatorial Pacific) in the 3000 to 15,000 Cal Yr Interval

Abstract

ABSTRACT. Paired radiocarbon and 230 Th/U dating was performed on 13 surface corals from submerged reefs in the Marquesas and from raised terraces in Vanuatu. The absolute ages of the corals analyzed ranged from 3000 to 15,000 cal yr. Estimates of the difference between the absolute and 14 C ages of these corals are in agreement with previous determinations up until 11,500 cal yr. The resulting mean sea surface reservoir age R is determined at 390 ± 60 yr for the Marquesas region (9°S), which is slightly higher than the R value at 280 ± 50 yr for the Tahiti Islands (18°S). Multiple 14 C analyses of 2 corals from the Marquesas present scattered 14 C ages at ~12,000 and ~15,100 cal yr. This could be attributed to rapid changes of the 14 C content of surface waters around the Marquesas Islands or to a subtle submarine diagenesis.

Radiocarbon calibration curve spanning 0 to 50,000 years BP based on paired 230Th/234U/238U and 14C dates on pristine corals

“…These pristine drill core samples and rigorous quality control criteria for vadose (zone exposed to percolating rain water) exposed samples are especially essential to the accuracy of the older radiocarbon dates, where a trace amount of calcite (i.e., greater than 0.2%) can result in unacceptably large offsets in radiocarbon ages (Chiu et al, 2004(Chiu et al, , 2005. The ''less than 0.2% calcite'' is the single most important screening criterion we have adopted, and it explains many of the differences between our calibration curve and published coral data that typically use 1% calcite detection limits and measure between 1% and 5% calcite in their samples (Yokoyama et al, 2000;Paterne et al, 2004). …”

“…Radiocarbon ages spanning the last 11,900 years are calibrated by making radiocarbon age determinations on tree rings of known age (Damon and Long, 1962;Damon et al, 1963;Stuiver et al, 1998a, b;Spurk et al, 1998;Friedrich et al, 1999;Reimer et al, 2002Reimer et al, , 2004. For the age interval between 12,000 years and 50,000 years before present, radiocarbon ages are calibrated by less precise and less accurate methods, such as varved sediments (Hughen et al, 1998(Hughen et al, , 2000(Hughen et al, , 2004bSchramm et al, 2000;Goslar et al, 2000a, c;Kitagawa and van der Plicht, 2000;Hughen et al, 2004b;van der Plicht et al, 2004), correlation of distinct fluctuations in ocean/climate proxies dated by radiocarbon with similar features in the Greenland ice cores dated by layer counting and flow models (Hughen et al, 2000(Hughen et al, , 2004aVoelker et al, 2000), 230 Th/ 234 U/ 238 U dating of speleothems (Vogel and Kronfeld, 1997;Goslar et al, 2000b;Beck et al, 2001) and corals (Fairbanks, 1990;Edwards et al, 1993;Bard et al, 1990Bard et al, , 1998aBurr et al, 1998;Yokoyama et al, 2000;Cutler et al, 2004;Paterne et al, 2004;van der Plicht et al, 2004).…”

“…The rapid burial of these corals minimizes the potential for microboring and encrusting organisms to degrade the sample quality, such as has occurred with dredged samples that resided on the sea floor for tens of thousands of years (Paterne et al, 2004). Most Barbados samples older than 30,000 yr. were subaerially exposed during the last glacial maximum lowstand, but have remained below sea level for the past 14,000 yr. A few samples older than 30,000 yr grew in deep water and were not subaerially exposed during the subsequent glacial maximum lowstand.…”

“…Speleothems provide long records and are very useful for identifying general trends and maybe large 14 C production anomalies, but speleothems are potentially limited by the dating errors due to variable groundwater uranium-series and carbon chemistry. 230 Th/ 234 U/ 238 U dating of corals can be reasonably precise and accurate, but samples of appropriate age are hard to acquire and the coral 230 Th/ 234 U/ 238 U ages can be altered by freshwater diagenesis (Hamelin et al, 1991;Henderson et al, 1993;Gallup et al, 1994Gallup et al, , 2002Ribaud-Laurenti et al, 2001), and in some locations by marine cements (Ribaud-Laurenti et al, 2001;Paterne et al, 2004). Even trace amounts of diagenetic calcite deposited in subaerially exposed corals will bias the radiocarbon measurements (Chiu et al, 2005a).…”

“…Our calibration meets the requirements that each data point in the calibration has a measured calendar age ( 230 Th/ 234 U/ 238 U) and radiocarbon age with known errors that are independent of each other. We have chosen not to include coral data from other studies (Yokoyama et al, 2000;Paterne et al, 2004;Cutler et al, 2004;Hughen et al, 2004b) because coral samples reported in the literature typically contain between 1% and 5% calcite, a contaminant which is significantly concentrated during the sample-etching step prior to radiocarbon analyses, thereby corrupting the radiocarbon data (Chiu et al, 2005). In addition, we choose not to include coral samples that reported calcite detection limits above 0.2% (Bard et al, 1998), even if no calcite was reported in a sample, because samples contaminated by more than 0.2% are generally unsuitable for calibration purposes (Chiu et al, 2005), particularly for older samples.…”