Studying climate signals through microscale chemical variability in diatoms.: A high-resolution chemical imaging study of biogenic silica

Abstract

Studying the chemical signature of fossil biogenic material is of great interest as it can provide information about the climate of the past. In the last few decades, the chemical signature of fossil biogenic silica has been receiving increased attention. This increased interest can be linked to the scarcity or complete absence of carbonates, a traditional target material for paleoproxy research, in sections of the world’s sediment records (particularly in high-latitude regions). Fortunately, these records are often dominated by silica microfossils, in particular the remains (termed frustules) of a unicellular microalga called 'diatom'. This research aims to apply high-resolution chemical imaging techniques in an attempt to understand better and thus improve existing paleoproxies based on biogenic silica. We predominantly focussed on the use of the Nanometer Secondary Ion Mass spectrometer (NanoSIMS), which allows us to create chemical maps on a nanometer scale. We also used other complementary high-resolution techniques such as the SIMS and Raman microspectroscopy. We applied these techniques to fossil and cultured diatoms in order to study their chemistry and their use for studying climate signals. For example, we looked at the presence of iron (Fe) in fossil frustules as changes in Fe-content can provide information on changes in past ocean conditions and indirectly on CO2 uptake by oceans. With the nanoSIMS, we were able to improve the existing proxy methods and remove some of the uncertainties. Another proxy (d15N) in fossil diatom frustules is based on the assumption that organic matter is retained in the frustule and protected over geological timescales. With the nanoSIMS and Raman imaging, we were, for the first time, able to image this organic presence in the silica of the cleaned fossil frustules. The oxygen isotope composition of diatom silica also has the potential to provide information on past climate conditions. However, concerns exist that secondary processes can cause the original signal to become overprinted, preventing robust interpretations. In this thesis, we show that this signal can be altered on the timescale of days to weeks. Both under seawater conditions as well as in the sediment. Meaning that the measured oxygen isotope composition is likely a combination of different signals ranging from the original growth conditions to porewater changes, potentially over geological timescales. Finally, as there is very little precedent concerning the application of the nanoSIMS to biogenic silica, we evaluate the different approaches that were used to study biogenic silica. Both the successful and unsuccessful attempts are evaluated, which will hopefully aid future researchers interested in unraveling the many mysteries still surrounding the chemistry of diatom frustules.