Empirical and theoretical approaches for assessing the local radiocarbon Marine Reservoir Effect

The radiocarbon activity of a reservoir is a function of both intra- and inter-reservoir phenomena. In general, any process removing 14C from a reservoir (or equivalently introducing pre-aged carbon) without its replenishment by a counteracting process will cause a reservoir effect. First defined in 1986, the 14C marine reservoir effect (MRE) still puzzles researchers worldwide. Its quantification is paramount for the application of the radiocarbon dating tool in marine environments, bearing important consequences for the chronology of coastal archaeological sites. Unfortunately, a significant lack of MRE data has hindered chronometric studies in regions such as the coast of Brazil, where thousands of prehistoric shell mounds remain poorly investigated. The geographic and temporal fluctuations of the MRE have led me to employ a two-pronged approach, using known-age and archaeological samples, in an effort to understand this variability for the Brazilian coast. In this kind of study, the importance of contextualization cannot be overstated and, therefore, my methodology has involved a range of accessory techniques (e.g., anthracology, stable isotopes and X-ray diffraction analyses) to ensure the suitability of samples for MRE quantification and thus to enhance the accuracy of the results. The pre-bomb local components of the MRE - i.e., ∆R values - are mostly within the range from -100 to 100 14C yr, but record the influence of phenomena such as upwelling and limestone dissolution at specific locations on the coast. The archaeological ∆R value of -102 ± 74 14C yr, in turn, was derived from the Late Holocene shell mound Sambaqui da Cabeçuda, a massive coastal site in Southern Brazil, where the archaeological evidence suggests a funerary function. The highly negative ∆R value is likely caused by the input of recently formed carbon species in the system and I discuss different Bayesian models for the calculation of this offset, as well as a series of complications for MRE studies in Brazilian shell mounds. In order to disentangle the complex relationships between archaeology and the environment, I have used stable isotopes of N and C together with Bayesian analysis to assess the diet of 5 humans from Cabeçuda. The quantification of the high proportion of marine diet (up to 80%) allowed me to calibrate bone 14C ages with an appropriate mixed curve. Despite various obstacles, the current practice of using ∆R and a global calibration curve remains the best available approach for marine calibration. Nonetheless, by acknowledging that discussion of other possibilities is paramount for the progress of radiocarbon research, I present a method for the achievement of local marine calibration curves. The advantages and limitations of this method are discussed at the end of the thesis. Overall, this research uses interdisciplinary techniques to tackle questions of archaeological and environmental significance and provides useful methodological insights for future studies.