Skeletal correlates of coral density banding and an evaluation of radiography as used in sclerochronology

Abstract

The skeletal correlates of density band patterns seen on radiographs of coral skeletons is investigated by theoretical analysis of radiography, comparison of radiographs and skeleton, and modelling of skeletal slices of the massive Indo-Pacific coral Pontes lutea and the branching Caribbean coral Pontes porites. Radiography resolves finer detail of the coral skeleton than has previously been recognised. Small changes in path length of individual skeletal elements attenuating an X-ray beam, including dissepiments as thin as 3 to 10 pm, can generate contrast on radiographs as well as variations in skeletal bulk density. The annual high/low density band pattern was found to be correlated with areas of relatively thick/thin skeletal elements respectively. Secondary density variations seen on radiographs, 'fine' bands, were found to be correlated with the position of corallite walls oriented perpendicular to the the long axis of the skeletal slice or dissepiments. Modelling of skeletal slices showed that errors in the density band record may be introduced by (1) skeletal slices not following the growth axis of the colony, and (2) changes in corallite orientation which generate density variations that are not part of any growth record. The use of radiography as a tool for demonstrating growth records in coral skeletons is critically reviewed. It is proposed that the skeletal growth record consists of patterns of thickeninglthinning and spacing of skeletal elements. The illustrations of these patterns by analysis of the actual skeleton, rather than by radiography, may prove to be a very accurate method of determining growth records in corals.