Abstract

Changes in marine upwelling can affect the radiocarbon content of seawater and thus affect the marine radiocarbon reservoir age, R. These radiocarbon variations are preserved in mollusk shell carbonate. Shell-based estimates of R in a variable-upwelling environment can be biased by (1) changes in molluscan growth rates due to fluctuating environmental conditions and (2) time averaging during sampling due to homogenization of days or weeks of precipitated carbonate. We modeled the growth, radiocarbon content, and radiocarbon sampling of two Peruvian mollusks, Argopecten purpuratus (bay scallop) and Mesodesma donacium (surf clam), to quantify these potential biases. Argopecten purpuratus grows year round, but M. donacium prefers cold conditions and its growth rate decreases in summer. Radiocarbon assays by accelerator mass spectrometry on multiple ∼1 mg samples of a model A. purpuratus shell can capture the full range of annual R variation; similar sampling of a model M. donacium shell only captures ≤75% of this range. Given an annual R variation of 530 14C yr, the mean R calculated using a group of ∼1 mg carbonate samples from an A. purpuratus shell is within 30 14C yr of the actual mean R; that of a M. donacium shell may be skewed up to 140 14C yr older than the actual marine mean. If growth tolerances and parameters of mollusks used for R analyses are considered, it may be possible to correct for these biases and improve the accuracy of marine radiocarbon chronometry.

You do not currently have access to this article.