An assessment of the Sr/Ca ratio in shallow water hermatypic corals as a proxy for sea surface temperature

The high precision measurement of the Sr/Ca ratio in corals has the potential for measuring past sea surface temperatures at very high accuracy. However, the veracity of the technique has been questioned on the basis that there is both a spatial and temporal variation in the Sr/Ca ratio of seawater, and that kinetic effects, such as the calcification rate, can affect the Sr/Ca ratio of corals, and produce inaccuracies of the order of 2-4 °C. In the present study, a number of cores of the massive hermatypic scleractinian coral Porites, from the central Great Barrier Reef, have been analyzed for Sr/Ca at weekly to monthly resolution. Results from a 24 year record from Myrmidon Reef show an overall variation from 22.7 °C to 30.4 °C. The record shows a warming/cooling trend with maximum warming centred on the 1986-1987 summer. While some bleaching was reported to have occurred at Myrmidon Reef in 1982, the Sr/Ca record indicates that subsequent summer temperatures were much higher. The 4.5 year record from Stanley Reef shows a maximum SST of 30 °C during the 1997-1998 El Niño event. The calibrations from Myrmidon and Stanley Reefs are in excellent agreement with previously published calibrations from nearby reefs. While corals do not calcify in equilibrium with seawater due to physiological control on the uptake of Sr and Ca into the lattice of coralline aragonite, it can be argued that, provided only a single genus such as Porites sp. is used, and that the coral is sampled along a major vertical growth axis, then the Sr/Ca ratio should vary uniformly with temperature. Similarly, objections based on the spatial and temporal variability of the Sr/Ca activity ratio of seawater can be countered on the basis that in most areas where coral reefs grow there is a uniformity in the Sr/Ca activity ratio, and there does not appear to be a change in this ratio over the growth period of the coral. Evidence from several corals in this study suggest that stress can be a major cause of the breakdown in the Sr/Ca-SST relationship. Thermal stress, resulting from either extremely warm or cool temperatures, can produce anomalously low Sr/Ca derived SSTs as a result of the breakdown of the biological control on Sr/Ca fractionation. It is considered that other stresses, such as increased nutrients and changes in light intensity, can also lead to a breakdown in the Sr/Ca-SST relationship. Two of the main issues affecting the reliability of the Sr/Ca method are the calibration of the Sr/Ca ratio with measured SST and the estimation of tropical last glacial maximum (LGM) palaeo-temperatures. Instead of producing a constant calibration, just about every one published so far is different from the others. What is obvious is that for most calibrations while the slope of the calibration equation is similar, the intercepts are not. While the cause for this variation is still unknown, it would appear that corals from different localities around the world are responding to their own particular environment or that certain types of environments exert a control on the corals' physiology. Sr/Ca derived SST estimates for the LGM and deglaciation of 5 °C-6 °C cooler than present are at odds with estimates of 2 °C-3 °C cooling by other climate proxies. The apparent lack of reef growth during the LGM suggests that SSTs were too cold in many parts of the tropics for reefs to develop. This would lend support to the idea that tropical SSTs were much cooler than what the CLIMAP data suggests.