Sea surface temperature and salinity reconstruction from coral geochemical tracers

Massive scleractinian corals secrete an aragonitic skeleton which incorporates a large array of chemical tracers. Corals present several advantages for palaeoclimate research: they grow continuously, and can live up to 1000 years; they are easy to date; and they can be sampled at high resolution (weekly to monthly resolution). Both live and fossil corals can be collected in the field. In the past two decades, significant efforts have been made to identify robust tracers of sea surface temperature (SST) and salinity (SSS) in corals. To date, Sr/Ca and δ18O are considered to be the most reliable SST tracers, although changes in seawater δ18O can significantly alter SST reconstructed from coralline δ18O. Because these variations in seawater δ18O can be linked to SSS changes, this initial problem can in fact be turned into an advantage and provide us with an SSS tracer. The SST component in the coral δ18O signal can either be evaluated through Sr/Ca measurements, or in some case simply filtered out. However, there is still much uncertainty concerning the exact mode of incorporation of trace elements and stable isotopes into the coral skeleton. The effects of growth rate, light intensity, feeding habits, pH and water chemistry are still poorly documented. A review of the strength and weaknesses of Sr/Ca and δ18O is presented, together with some examples of SST and SSS reconstructions. Other potential SST tracers are also reviewed. It is expected that the ability to grow corals in aquarium under controlled conditions, and that the development of sophisticated analytical techniques at the micrometric level should help us understand better the robustness of each tracers and the factors controlling their incorporation in coral aragonite.