Radiocarbon Reservoir Effects on Shells from SE Arabia in the Context of Paleoenvironmental Studies

Shells are among the most frequently used materials for dating coastal sites with radiocarbon, especially when no organic matter, such as bone collagen, is preserved. Shells are frequently used to constrain chronologies from archaeological sites or within palaeoenvironmental research. Sites throughout Arabia tend to lack organic material suitable for dating. Marine shells in upwelling areas, such as the Arabian Sea, show an older apparent age than contemporaneous terrestrial material, hence are most likely depleted in radiocarbon. In radiocarbon dating, this is referred to as marine reservoir effect. Correcting for this effect for a given period and location is required to establish reliable chronologies. The aim of this research is to explore an understanding of the marine reservoir effect with respect to site-related factors and temporal variation. The suitability of shells for environmental monitoring is investigated using sclerochronology and stable isotopes and interpreted on the basis of background information on diet and habitat of respective mollusc species. A species-specific reservoir effect is then determined for radiocarbon chronologies and temporal variations are evaluated by assigning a speciesspecific reservoir effect to each period investigated. High resolution radiocarbon measurements along the shell are tested with the aim to support carbon isotope (δ13C) interpretation and determine a minimum amount of carbon necessary for reliable data. Shell foraging and subsequent heating provides an important aspect of this work with respect to its influence on the shell material and possible shifts in data, especially radiocarbon. A shift in data due to heating would prevent the use of heated shells for environmental monitoring and radiocarbon dating. Additionally, a pilot study sheds light on possible variations in hardwater effects on land and freshwater gastropods depending on the geological setting. Sampling sites range from archaeological excavations along the coast (Kalba) and inland (Wadi al-Hilo) to wadis throughout the Hajar Mountains in the United Arab Emirates. The site of Kalba was chosen as the location for investigating the reservoir effect of marine shells and environmental conditions during their lifetime. Kalba is located between the Gulf of Oman and the Hajar Mountains and provides shells of different species from a mangrove forest that has existed since at least the mid-Holocene (ca. 7000 cal BP). Archaeological sites in Kalba dating to Neolithic and to the Bronze Age offer sediment layers rich in shells. Additionally, Kalba offers the unique possibility to monitor seasonal ocean circulation patterns where upwelling from the Arabian Sea to the South alternates with saline, dense outflow of Arabian Gulf waters. The Hajar Mountains reveal a diverse geology with ophiolites located southwest of Kalba and carbonates to the north. The Dibba Zone shows outcrops of shelf carbonates and metamorphic rocks and provides the greatest diversity of lithologies in the Hajar Mountains. For the study of terrestrial shells, this diversity conditioned the sampling sites in order to derive a dependence of possible hardwater effect on lithology involved. The first part of the study concentrates on two shell species from Kalba, the bivalve Anadara uropigimelana and the gastropod Terebralia palustris, because they are often found in archaeological contexts. Shell and charcoal pairs were recovered from four layers representing two periods, Neolithic and Bronze Age, to evaluate temporal changes in the species-specific reservoir effect. For A. uropigimelana VIII a drop in ΔR from 576 ± 90 to 112 ± 44 years over this period was calculated, whereas for T. palustris the values decreases from 389 ± 66 to -19 ± 36 years. This species-specific reservoir effect likely reflects differences in habitat and/or dietary habits. T. palustris with a mainly terrestrial diet yielded a significantly lower ΔR compared to the mainly marine influenced A. uropigimelana. The temporal changes are probably explained by reduced upwelling in the Arabian Sea as observed in marine sediment cores, due to a decrease in summer monsoon strength. High resolution 14C measurements were performed in combination with stable isotopes δ13C and δ18O on two specimens of A. uropigimelana. Here, 14C served to support the interpretation of the inconclusive δ13C signal that has its origin in diet as well as water dissolved inorganic carbon (DIC). Sample sizes below 30 μg C did not result in reliable data due to elevated backgrounds for 14C measurement and cross contamination from the previous sample. Nevertheless, trends in data could be identified and provided insights to help interpret changes in amplitude or extreme drifts of the isotopes as a result of changes in environmental conditions, such as freshwater input, or diet. Shell foraging often includes heating of shells that inflicts a transformation from aragonite to calcite and also destroys the microstructural pattern of the shell material. Data indicates that the 14C measurement is not shifted by heating of the shell material. Therefore, heated shell material like unheated shell carbonate provides reliable material for chronological purposes. Land snails of species Zootecus insularis and freshwater snails of species Melanoides tuberculata were recovered from different geological substrates. A strong link between lithology and hard-water effect of the terrestrial species was found, even though distinct values for the hard-water effect could not be determined for all sites involved. Unfortunately, no live specimens could be found to resolve this problem. Therefore, archaeological samples of known context provide the most reliable basis for interpretation. In conclusion this study shows that shells of A. uropigimelana and T. palustris, heated or unheated, are a valuable tool for evaluating chronologies. The drop in marine reservoir effect ΔR of the shells mirrors local changes in ocean circulation and upwelling and is directly linked to changes in climate and sea level during the Holocene. On a regional as well as global scale it coincides with significant changes in other paleoclimate archives. Regarding the pilot study on the terrestrial hardwater effect of Z. insularis and M. tuberculata, however, promising new results for some locations are presented. As expected, it varies with lithology, but still needs to be verified by known age samples before being used in this region. In brief, the results of this thesis show that marine reservoir effects are not only dependent on location but also vary through time and by species. Whilst this has been demonstrated here for the Southeastern part of the Arabian Peninsula, similar variations in time and species are likely to also occur in other regions of the globe. To establish reliable radiocarbon chronologies using shells, the reservoir effect has to be determined considering all of these factors. Whilst time consuming, and in part challenging, this approach, offers numerous possibilities for evaluating environmental data from a single material