Highly variable length-at-age in vermilion snapper (Rhomboplites aurorubens) validated via Bayesian analysis of bomb radiocarbon

Abstract

Vermilion snapper (Rhomboplites aurorubens) has been reported to exhibit highly variable length-at-age, while age estimates derived from otoliths are often imprecise (i.e., average percent error ≥ 8.0%). This has led to questions whether the observed variability in length-at-age is due to variability in growth or ageing error. The current study was conducted to address this by testing the accuracy of otolith-derived age estimates via age validation. Age was estimated by enumerating opaque zones in otolith thin sections observed under a dissecting microscope with transmitted light. Then, age validation was performed by applying the bomb radiocarbon (14C) chronometer, where vermilion snapper birth year Δ14C values were derived from eye lens cores for contemporary samples and otolith cores for archived samples. Otolith-derived age estimates were found to be accurate, with the maximum validated age being 28 years. Results of Bayesian regression modeling indicated there was no significant relationship between eye lens Δ14C variance and fish length (p = 0.93). Further Bayesian regression modeling demonstrated vermilion snapper do not exhibit sexually dimorphic growth (p = 0.91). Study results have important implications for vermilion snapper assessment and management, as variable growth reduces the accuracy in assigning age based on length. Moreover, the approach reported here could be applied to test the accuracy of highly variable length-at-age estimates in other marine fishes.