Salt-marsh sea-level indicators and transfer function development for the Magdalen Islands in the Gulf of St. Lawrence, Canada

Late Holocene salt-marsh sediments allow the reconstruction of pre-industrial relative sea-level changes. Typically, reconstructions are derived by quantifying a relationship between contemporary proxies and elevation which is then applied to down-core fossil assemblages via a transfer function. Common sea-level indicating proxies are salt-marsh foraminifera and diatoms, however, testate amoebae represent an alternative complementary proxy. With current sea-level research increasingly invoking (multi-)decadal timescale processes such as ocean current strength or wind stress to explain relative sea-level changes, there is a need for increased vertical precision in reconstructions. This paper presents new surface assemblages of salt-marsh foraminifera and testate amoebae with respect to elevation from two coastal locations in the Magdalen Islands in the Gulf of St. Lawrence, eastern Canada. We analyse the suitability of each proxy for reconstructing sea level using multivariate statistics, ordination and regression modelling. Testate amoebae are found in greater abundance and diversity than foraminifera and their assemblage variations can be better accounted for by changes in elevation. The contemporary training sets from the Magdalen Islands were able to predict marsh-surface elevations (viz. sea level) to precisions of ± 0.08 m and ± 0.12 m for testate amoebae and foraminifera respectively. There were no benefits from combining the two proxies into a unified training set. It was possible to improve the performance of the foraminifera-based transfer function by incorporating a second dataset from Newfoundland which served to increase the range of the sampled environment and improve the correlation coefficient between sample elevations and model predictions. In accordance with existing literature, we find transfer function performance to be better at locations with smaller tidal ranges.