Estimating size-at-harvest from Indigenous archaeological clamshell assemblages in Coastal British Columbia

Shellfish have supported Indigenous lifeways on the Pacific Coast of North America for millennia. Despite the ubiquity of clamshells in archaeological sites, shell size measurements are rarely reported due to a lack of applicable basis for generating size estimates from fragmentary remains. We present a linear regression-based method for determining shell length from hinge and umbo measurements of littleneck (Leukoma staminea; n = 239), butter (Saxidomus gigantea; n = 274), and horse (Tresus nuttallii; n = 92) clams using both contemporary and archaeological shells collected from three regions in coastal British Columbia, Canada. We examine the accuracy of these size estimations, which indicate that 83%–97% of the variability in dorsal shell length is predicted by umbo thickness and hinge length. Hinge length generated higher R2 values yet exhibited greater intra- and inter-observer error. While the predicted dorsal length for each species differed by region, this size difference was smaller than intra- and inter-observer error, suggesting broad applicability for these simple measurements. We applied these formulae to a Tseshaht First Nation archaeological clamshell assemblage (n = 488) on western Vancouver Island spanning 3000 years and observed profiles that resemble contemporary legal size limits, which suggests the sustained use and maintenance of local shellfisheries. The accuracy of these regression models for determining shell length from fragments highlights the utility of this approach as a basis for assessing past shellfish management practices.