Fidelity of molluscan life and death assemblages on sublittoral hard substrata around granitic islands of the Seychelles

Abstract

This study quantified the degree of coincidence between living and dead molluscan faunas in a shallow-water coral reef environment in the Indian Ocean. The results were compared with those from a similar life:death study in the northern Red Sea, and with those published for reef corals and soft substrata molluscs. The proportions of quantitatively important taxa are robust to sampling intensity, but fidelity indices and rank-order correlations are strongly influenced by quantitatively unimportant taxa. Distinct differences between life and death assemblages were recognized, which are due to distinct biases in the death assemblage. Bivalves that lived in close contact with living corals are preferentially overgrown after death and should provide considerable temporal and ecological information in a potential fossil record, as they will be preserved within a rapidly growing reef framework. Some gastropod taxa are preferentially transported into surrounding soft substrata postmortem. Here they will be affected by time-averaging and taphonomic disintegration typically occurring in sediments resulting in the associated loss of much temporal information. Most gastropod shells, however, are inhabited by hermit crabs postmortem, which may strongly alter the fossil gastropod community structure. The results are similar to a comparable study in the northern Red Sea, with a major exception being the strong dominance of hermit crab-inhabited gastropod shells in the death assemblage of the Seychelles. Comparison of life:death assemblages between hard and soft substrata, in keeping with the northern Red Sea study, showed the strong dominance of dead shells in the soft substrata with the converse on the hard substrata. This results from different accumulation conditions for dead shells in soft substrata. Fidelity indices are well suited to demonstrate that sedimentary death assemblages are typically remarkably robust reflections of local community composition but they do not record the strong biases in death assemblages of coral reef associated hard substrata molluscs and are therefore unsuitable for comparisons of life and death assemblages in reef environments.