Two centuries of streamflow behavior inferred from freshwater pearl mussel shell δ18O in Northern Sweden

Abstract

Global warming alters snowmelt timing and soil frost cycles, impacting storage-discharge dynamics of subarctic catchments. Assessing how these changes affect streamflow requires quantifying the relative contribution of winter versus summer precipitation. While this can be achieved based on oxygen isotope data (δ18O) in precipitation and streamwater, the full potential of this approach is hindered by short and truncated time-series. Here we overcome this limitation by combining modeled precipitation δ18O time-series with streamwater δ18O data from freshwater pearl mussel shells. Based on more than 2,700 shell-derived streamwater δ18O values reconstructed from four Northern Swedish catchments we estimated contributions of precipitation to streamflow since 1836. The reconstructed δ18O streamwater records, which typically capture summer low-flow conditions during the growing season of the mussels, showed pronounced differences between catchments. In a monitored reference catchment, a negative correlation between baseflow and δ18O-derived contributions of summer precipitation suggests that these differences reflect catchment-specific variations in baseflow during low-flow conditions. These variations in seasonal precipitation contributions to streamflow were linked to physiographic catchment characteristics, which also affected long-term summer low flow conditions. Reconstructions based on bivalve shells therefore provide insights into how landscape features (e.g., catchment size, soil conductivity, lakes) influence water transit times and delay the release of winter precipitation across catchments over centuries. With this transferable and scalable approach, isotope data sets can be extended into pre-instrumental times, supporting adaptive water-resource management in data-scarce regions.