Calculation of the ‘Marine Reservoir Effect’ from the Dating of Shell-Charcoal Paired Samples from an Aboriginal Midden on Great Glennie Island, Bass Strait

Abstract

In recent years, increasing use has been made in Australia and the Pacific of samples of marine and freshwater  prehistoric shell samples as materials for obtaining radiocarbon dates for dating sites. This has sometimes been the case because only shell was present, as for example in a midden dump where charcoal was originally absent or where the charcoal had been washed or leached away subsequent to deposition. However even when charcoal is present in a deposit, shell may be preferred because of the structure of the midden, where the shell lenses are packed with greater stability and demonstrable stratigraphic integrity than the surrounding deposits; or where it is believed that charcoal specks may have moved through the midden matrix subsequent to initial deposition. This is particularly the case for example in northern Australia where some large shell middens consist almost entirely of shells with little or no 'earthy' deposit between them. Although the middens themselves are stable, their matrices have many interstices down which charcoal specks may be washed during the periodic heavy cyclonic rains which occur during the wet season. In a major dating programme on middens on the coastal plain of the Blyth River area of Arnhem Land, Betty Meehan (pers. comm.) and two of us (RJ and JH) have found that the shell date series have produced coherent and consistent results, whereas the charcoal samples have often been erratic, both compared with the parallel shell samples and with one's general expectations based on geomorphic and ethnographic evidence.