Abstract:
Previous investigations of the δ18O of the skeletons of Florida specimens of the reef coralMontastraea annularis have failed to produce the full temperature range suggested by calibration studies of other corals. Explanations for this phenomenon include different relationships between temperature and the δ18O of skeletons of Floridian corals, changing δ18O of the water, physiological variables (''vital effects''), and an insufficient number of samples taken per year with consequent superposition of calcium carbonate precipitated at different times within an individual sample. In this study, we investigate all of these hypotheses, by measuring the δ18O of corals grown in the field which were periodically stained with alizarin-red S and where the δ18O of the water was measured and the temperature continuously recorded. We compare the effect of sampling the coral skeletons at different resolutions and the effect of sampling within different skeletal elements. Our study shows that discrete, high-resolution sampling of coral exotheca (fifty samples a year) is necessary to reproduce temperatures for this species in Florida waters. Coral skeletons sampled using lower resolution methods showed an artificial attenuation of the annual range in skeletal δ18O, with similar δ18O minima during the skeleton represented by the summer months, but larger differences in the winter δ18O maxima. Replicate isotope transects from fast and slow growing areas and different regions of the corallite were also compared. The δ18O of rapidly growing (8 mm/y) portions of the colony was 0.1 to 0.2%% heavier than the slowest growing (1.1 mm/y) portions of the colony. This difference as well as the difference between the skeleton sampled at high and low resolutions appears to result in part from the attenuation of the δ18O signal as a result of the reduced sampling rate in slower growing sections of the coral and is not solely a result of variable kinetic effects.