Compositional variation in planktonic Foraminifera shells:
S. Eggins, P. De Deckker* and J. Marshall
*Department of Geology, ANU
The incorporation of Mg into the tiny calcitic shells (tests) that are secreted by planktonic foraminifera is extraordinarily sensitive to temperature, with large exponential increases of about 10% per °C occurring in shell Mg/Ca ratio. Thermometers calibrations have now been made for a small but expanding number of planktonic foraminifera species and form the basis of a rapidly emerging tool for reconstructing paleoseawater temperature from fossil shell material. Precise seawater temperature estimates are now routinely derived by bulk analysis of between 10 and 50 whole fossil foraminifera shells using modern instrumental techniques. Unfortunately, the reliability of these estimates is compromised by the compositional variation that occurs within a population of foraminifera shells (due to seasonal and interannual seawater variability), subsequent modification of preserved bulk shell compositions due to preferential dissolution on the seafloor of more Mg-rich shells and shell parts and, the effects of cleaning to remove contamination by diagenetic/detrital phases.
To illuminate the extent of these problems we have investigated the extent of compositional variation within several planktonic foraminifera species that are commonly used in paleocean reconstruction. Using in-house developed laser ablation ICP-MS technology, we are able to analyse individual chambers and to profile chamber
|walls with sub-micron depth resolution (Figures 10 & 11). We have found correlated Mg/Ca, Mn/Ca, Ba/Ca and Zn/Ca variations but relatively unfirm Sr/Ca compositions through the shell walls of Globigerinoides sacculifer, Globigerinoides ruber, Globigerina bulloides, Neogloboquadrina pachyderma and Neogloboquadrina dutertrei. Distinct chamber and chamber-wall layer compositions occur within individual shells, particularly in species that migrate within/through the thermocline. This compositional heterogeneity appears to be consistent in detail with seawater temperature changes that accompany habitat migration during the adult life-cycle stages of the investigated species. However, an anomalous, Mg-rich (<1-6 mol% Mg) surface veneer occurs on each of the analysed species and is found on both fossil and modern shells. These shell surface coatings are of primary biogenic origin and bias the analysed bulk shell compositions toward higher Mg/Ca values by between 5 and 35%. Subject to the ability to exclude the surface veneer during analysis as possible with laser ablation ICPMS anlaysis, bulk shell Mg/Ca compositions may provide reliable mean seawater temperature estimates only for species that calcify under uniform (near-surface?) conditions. Laser ablation ICPMS shows considerable promise as a means for reconstructing paleoseawater temperatures and has the potential to add considerable value to deep-sea core paleocean and climate records by yielding: (1) the range and variability of past seawater temperature from a population of shells, (2) the thermocline and compositional structure of the water column through analysis of specific shell parts in species that migrate over large vertical ranges; and (3) integration with _18O analysis of the remaining shell material to constrain past seawater _18O from the temperature dependent variation of _18O and therefrom estimate past seawater salinity and density. The technique may also be used to track habitat changes occurring during the adult life-cycle stages of planktonic foraminifera.
Figure 10: SEM image of a N. dutertrei shell in which multiple (14 separate), 29µm diameter, laser ablation ICPMS analyses have been made with a pulsed ArF Excimer laser (_=193nm). Each analysis requires <60 seconds and consumes only 10-20 ng of shell material, or about 0.1% of the total shell mass, leaving the remainder available for (18O or other analysis.
Figure 11: High-resolution compositional depth profile obtained through the lower large chamber wall of the N. dutertrei specimen shown in Figure 1. Note the distinct compositions of the inner and outer wall layers, the highly Mg-rich surface veneers, and the sub-micron scale compositional resolution achieved. Total analysis time was 60 seconds.