A dual bivalve approach for interpreting past sea surface temperatures and seasonality from shell midden sites using oxygen isotope sclerochronology

Abstract

Stable oxygen isotope (δ¹⁸O) analysis of archaeological shellfish remains combined with sclerochronology can be used to precisely reconstruct past sea surface temperature (pSST), season(s) of shellfish collection, and thus the season(s) of archaeological site occupation. Our study tests if δ¹⁸O_shell and sclerochronological analyses of marine bivalve species Leukoma staminea – with previously unassessed seasonality and pSST potential – can provide additional insights not captured by the better studied species, Saxidmous gigantea in British Columbia (BC, Canada). We analyzed live-collected L. staminea and S. gigantea shells from Sechelt, BC, and compared results to archaeological shell data from Powell River, BC, in the territory of the Tla'amin First Nation (1065 to 797 cal. B.P.). The seasonality of shellfish harvest differed between species, with S. gigantea preferentially collected in the spring, whereas L. staminea collected year-round. This highlights that sole-species seasonality studies may miss important variability in harvesting strategies. Additionally, comparisons between instrumental (5.7 to 20.4 °C) and reconstructed SST from modern L. staminea δ¹⁸O_shell (5.6 to 18.4 °C) showed good agreement in annual range. Our results indicate that archaeological S. gigantea record a wider range of pSST (−1.6 to 22.9 °C) than archaeological L. staminea (5.8 to 25.9 °C), suggesting that S. gigantea may be a more sensitive palaeotemperature recorder. Further, we found that using the same reconstructed δ¹⁸O_water value for both species in pSST reconstruction caused an overestimation of temperature. Accordingly, we argue that it is critical to calibrate the equation with corresponding species-specific δ¹⁸O_shell values, regardless of both having aragonite mineral structures. Our study highlights the potential of L. staminea to broaden seasonality interpretations while clarifying that its use for pSST reconstruction may not capture the lower range of temperature.