OXYGEN ISOTOPES OF MARINE DIATOMS AND RELATIONS TO OPAL-A MATURATION

Abstract

In order to develop the potential tool of diatom oxygen isotopes for paleoenvironmental studies we compared oxygen isotopes of natural marine diatoms sampled from ocean surface water, sediment traps and surface sediments with oxygen isotopic fractionations determined for laboratory diatom cultures. Freshly grown natural diatoms (phytoplankton samples and sediment trap material) and cultured diatoms reveal similar oxygen isotope fractionation factors. The fresh diatoms have 3 to 10%% lower isotope fractionation factors than fossil (sedimentary) diatoms. A temperature-related oxygen isotope fractionation could not be established for the laboratory cultures (and the natural phytoplankton samples), and there is evidence that diatom growth rate until reaching the stationary growth state also controls the measured silica-water oxygen isotope fractionation factor. It is possible, however, that slow diatom growth in sea surface water may well lead to a temperature-dependent silica-water oxygen isotope fractionation which is the prerequisite for a use of diatom oxygen isotopes in paleo-surface water studies.FTIR-spectroscopic analyses of various diatomaceous materials revealed that the ratio of integrated peak intensities for Si-O-Si/Si-OH correlates with the 3 to 10%% δ18Osilica increase from fresh to fossil diatoms. Open-system (flow-through) silica dissolution experiments suggest that the diatom frustules are isotopically homogenous and that the increase in 18O is therefore not due to dissolution of isotopically light surficial Si-OH groups. It is concluded that slow internal condensation reactions during silica maturation in surface sediments cause both an increase in the intensity ratio of Si-O-Si/Si-OH and the 18O content of framework oxygen. These findings also indicate that the oxygen isotope compositions of marine sediment diatoms do not indicate sea surface water temperature but rather reflect variable 18O contents of surface sediments.