A coral Sr/Ca calibration and replication study of two massive corals from the Gulf of Mexico

This study examined the variations in the ratio of strontium-to-calcium (Sr/Ca) for two Atlantic corals (Montastraea faveolata and Siderastrea siderea) from the Dry Tortugas National Park (centered on 24.7°N, 82.8°W) in the Gulf of Mexico. Cores from coral colonies in close proximity (10s of meters) and with the same environmental conditions (i.e., depth and water chemistry) were micro-sampled with approximately monthly resolution and the resulting Sr/Ca variations were calibrated with local sea surface temperature (SST) records. Replication tests for coral Sr/Ca variations found high agreement between intra-colony variations and between individual colonies of S. siderea (a single M. faveolata colony was sampled). Regression analysis of monthly variations in coral Sr/Ca and local SST revealed significant correlation on monthly and inter-annual timescales. Verification of the calibration on different timescales found coral Sr/Ca–SST reconstructions in S. siderea were more accurate than those from M. faveolata, especially on inter-annual timescales. Sr/Ca–SST calibration equations for the two species are significantly different (cf., Sr/Ca = -0.042 SST + 10.070, S. siderea; Sr/Ca = -0.027 SST + 9.893, M. faveolata). Mean linear extension for M. faveolata is approximately twice that of S. siderea (4.63, 4.31, and 8.31 mm year⁻¹, A1, F1, and B3, respectively); however, seasonal Sr/Ca variability in M. faveolata is less than S. siderea (0.323, 0.353, and 0.254 mmol mol⁻¹, A1, F1, and B3, respectively). The reduced slope for M. faveolata is attributed to physical sampling issues associated with complex time-skeletal structure of M. faveolata, i.e., a sampling effect, and not a growth effect since the faster growing M. faveolata has the reduced Sr/Ca variability.