Modelling strontium isotopes in past biospheres - Assessment of bioavailable ~(87)Sr/~(86)Sr ratios in local archaeological vertebrates based on environmental signatures

摘要

Sr-87/Sr-86 isotopic ratios in skeletal remains of archaeological vertebrates are used for provenance analysis since long. However, the definition of the past bioavailable isotopic ratio at the site of recovery is not known beforehand and geological maps can provide no more than gross expectations. Therefore, the assessment of the "local Sr isotopic signature" is still of crucial importance. In this study, we present a tool for the prediction of such local isotopic signatures by creating a concentration weighted mixing model that links lithospheric, biospheric, and atmospheric strontium per site. The major strontium sources and their input into an animal's body were assessed by choosing elemental strontium and its isotopic signature in groundwater, soil, vegetation, and precipitation as components for the mixing model, augmented by literature values. The model was applied to 24 sites located in the alpine transect of the Inn-Eisack-Adige-Brenner passage across the European Alps, a passage used since the Mesolithic. Predicted local bioavailable Sr-87/Sr-86 ratios were compared with measured values from locally excavated archaeozoological bone samples from three taxa of large and mainly residential vertebrates (cattle, pig, red deer) to verify the models' accuracy. With regard to the fact that the environmental samples predict the past local bioavailable Sr-87/Sr-86 at a specific site while the vertebrates had different and species-specific home ranges, thereby integrating strontium from a region of primarily unknown size, the model is capable of assigning reasonable expectation values. For 11 sites, up to 100% of the vertebrate isotopic signatures were correctly predicted. Mismatches at the remaining sites are explainable by special environmental factors, and also the fact that some import of animals can never be excluded beforehand. Suggestions for site-specific adjustments of the model are made. (C) 2018 Elsevier B.V. All rights reserved.