Stable carbon and oxygen isotope ratios of foraminifera in surface sediments of the South China Sea

The relationship between planktonic and benthic foraminiferal stable-isotope values and oceanographic conditions and factors controlling isotopic variations are discussed on the basis of oxygen and carbon isotopic analyses of 192 modern surface and Last Glacial Maximum (LGM) samples from the South China Sea (SCS). The harmonic variation of benthic delta18O in surface sediments with water depth and temperature implies that the temperature is the main factor influencing benthic delta18O variations. Planktonic delta18O fluctuates with sea surface temperature (SST) and salinity (SSS). The N-S temperature gradient results in planktonic delta18O decreasing from the northeast to the south. Cool, saline waters driven by the winter monsoon are interpreted to have been responsible for the high delta18O values in the northeast SCS. The East Asian monsoons not only bring nutrients into the South China Sea and maintain high nutrient concentration levels at the southwestern and northeastern ends, which cause depleted delta13C both in planktonic (surface) and benthic (bottom) samples but also reduce planktonic/benthic delta18O differences. The distribution of delta18O and delta13C in the surface and LGM samples are strikingly similar, indicating that the impact of SST and SSS has been maintained, and nutrient inputs, mainly from the northeastern and southwestern ends, have been controlled by monsoons since the LGM. Comparisons of the modern and LGM delta18O indicate a difference of about 3.6 °C in bottom-water temperature and a large surface-to-bottom temperature gradient during the LGM as compared to today.

Western Arabian Sea SST estimation

Millennial-scale records of planktonic foraminiferal Mg/Ca, bulk sediment UK37', and planktonic foraminiferal d18O are presented across the last two deglaciations in sediment core NIOP929 from the Arabian Sea. Mg/Ca-derived temperature variability during the penultimate and last deglacial periods falls within the range of modern day Arabian Sea temperatures, which are influenced by monsoon-driven upwelling. The UK37'-derived temperatures in MIS 5e are similar to modern intermonsoon values and are on average 3.5°C higher than the Mg/Ca temperatures in the same period. MIS 5e UK37' and Mg/Ca temperatures are 1.5°C warmer than during the Holocene, while the UK37'-Mg/Ca temperature difference was about twice as large during MIS 5e. This is surprising as, nowadays, both proxy carriers have a very similar seasonal and depth distribution. Partial explanations for the MIS 5e UK37'-Mg/Ca temperature offset include carbonate dissolution, the change in dominant alkenone-producing species, and possibly lateral advection of alkenone-bearing material and a change in seasonal or depth distribution of proxy carriers. Our findings suggest that (1) Mg/Ca of G. ruber documents seawater temperature in the same way during both studied deglaciations as in the present, with respect to, e.g., season and depth, and (2) UK37'-based temperatures from MIS 5 (or older) represent neither upwelling SST nor annual average SST (as it does in the present and the Holocene) but a higher temperature, despite alkenone production mainly occurring in the upwelling season. Further we report that at the onset of the deglacial warming, the Mg/Ca record leads the UK37' record by 4 ka, of which a maximum of 2 ka may be explained by postdepositional processes. Deglacial warming in both temperature records leads the deglacial decrease in the d18O profile, and Mg/Ca-based temperature returns to lower values before d18O has reached minimum interglacial values. This indicates a substantial lead in Arabian Sea warming relative to global ice melting.