Iron Age migration on the island of Öland: Apportionment of strontium by means of Bayesian mixing analysis

Abstract

Migration is a complex subject to approach in archeology and the new materials and methods available, such as isotope analysis and DNA, make it possible, and necessary, to ask new questions. The objective of this paper is to highlight the possibilities with using a new approach to migration on a population level by applying Bayesian mixing analysis of strontium isotopes. The selected case, the island of Öland in the Baltic, was based on 109 human samples dated to the Early (500 BC–AD 400, n = 71) and Late (AD 400–1050, n = 38) periods. The results from both periods demonstrate that the distribution of Strontium (Sr) is multimodal with several peaks not associated with the local variation. Our results show a large immigration to Öland from other geological areas, with 32% of the population in the Early period and 47% in the Late period being nonlocal. In order to unravel these distributions, we use a Bayesian mixing analysis. The Bayesian mixing analysis provides us with a mean to disentangle the distribution of Sr that is not uninformed. The gravity model, however simplistic, is relevant for explaining the strontium variation in the population in Öland both in the Early and Late period. Our results indicate a significant internal migration in Scandinavia that is increasing in the Late Iron Age at the same time as the Viking expansions (the more well studied external migration), which is usually the only migration discussed. We argue that the method proposed and tested on the case of Öland adds new perspectives for approaching migration patterns in general on a population level, a perspective that is hitherto lacking in archaeology.