Times are displayed in (UTC-05:00) Central Time (US & Canada) Change

Towards a method of measuring uncertainty in molluscan oxygen isotope season of capture analyses.

Session: Timestamped Biomineralized Structures in Coastal Environmental Monitoring and Cultural Research

Presenting Author:

Dr. Charles F. T. Andrus

Authors:

Andrus, C. Fred T.¹, Goodwin, David H.²

(1) Geological Sciences, University of Alabama, Tuscaloosa, AL, USA, (2) Earth and Environmental Sciences, Denison University, Granville, OH, USA,

Abstract:

Archaeological season of capture has been determined by measuring sequential oxygen isotope values (δ¹⁸O) in mollusk shells for over 50 years, but no method of quantifying uncertainty in these determinations has yet been devised. Possible sources of error are diverse and include variability that may be impossible to quantify such as ancient people accidentally collecting previously deceased mollusks and depositing them in the sites along with live collected shells. In contrast, uncertainty associated with sclerochronological sampling techniques and shell growth variability may be approximated in certain taxa and habitats. Estimating uncertainty could benefit archaeologists because season of capture is often employed to reconstruct subsistence strategies and determine the minimum possible seasonal duration of site occupation, and even a small number of samples incorrectly attributed to a season may profoundly alter interpretation of past site activities.

Here we report results of two quantitative approaches to assess the accuracy of biogeochemical season of capture estimates. First, serial δ¹⁸O samples were collected from modern specimens with different but known collection dates by students who were not told the specimen’s dates of collection. Second, we used a mathematical model of shell growth coupled with a Monte Carlo simulation of a population of bivalve mollusks. Modelled shell δ¹⁸O values were constructed using known patterns of environmental variation and shell growth (i.e., patterns of senescence, seasonal shutdowns, and optimal growth temperatures). Digital “specimens” were then serially sampled at the commissure to estimate collection dates. In each experiment, δ¹⁸O values and patten of variability from the growing margin were then used to estimate collection dates. Our results highlight the importance of sample resolution and spacing for accurate season of capture estimates. Furthermore, biologically controlled patterns of growth can obfuscate true season of capture.