2008 Joint Annual Meeting (5-9 Oct. 2008): Reporting Marine Reservoir Ages in a Variable-Upwelling Environment

The marine radiocarbon reservoir age is typically reported as a single value for a given location at a given time in the geologic past. Presently, the global marine reservoir age averages about 400 yr. Our work suggests that in a variable-upwelling environment like coastal Peru, seasonal upwelling variability makes a single value-plus-error-term an insufficient descriptor of reservoir age, and the associated error is misleadingly small.

Generally, marine reservoir age calculations using molluscan carbonate assume homogeneous intrashell radiocarbon content. In regions with intense and variable upwelling, this assumption is invalid. Using AMS dating of small (~2 mg) samples from Peruvian pre-bomb Argopecten purpuratus, Mesodesma donacium, Protothaca asperrima, and Donax marincovichi shells, we have measured > 300 ¹⁴C yr of radiocarbon age variation within single years of growth in individual shells.

The concept of a single “mean” marine reservoir age from bulk shell analysis is not meaningful due to species-specific effects. Mollusks tend to grow more rapidly and precipitate more shell mass under favorable conditions. For example, mollusks that prefer cold water will precipitate more shell during periods of strong cold upwelling, and the bulk radiocarbon age of shell will be older than that of the surrounding water integrated over the same period. Likewise, warm-water-favoring mollusks will precipitate more shell during reduced upwelling, producing a younger bulk shell radiocarbon age. Bulk analyses of different species growing under identical conditions may thus produce different mean reservoir ages.

For a marine radiocarbon reservoir age to be most meaningful, we suggest that, in addition to the shell species and provenance currently reported, the following data should be included: 1) the number of samples used in the analysis, 2) the reservoir age calculated from each shell sample, and 3) the size of the samples used, preferably in terms of the duration of shell growth they represent.