How Reliable are Growth Back-Calculations Based on Otoliths?

Abstract

Growth back-calculations from otoliths assume that the relationship between fish and otolith length is linear through time. The final (or observed) individual fish-otolith ratios are then combined to prepare a fish-otolith regression for the population, upon which the subsequent back-calculations are based. However, recent studies have demonstrated that the fish:otolith size ratio varies systematically with somatic growth rate, resulting in relatively large otoliths in slow-growing fish. Such a growth effect will result in a fitted fish-otolith regression which differs significantly from that of the mean of the individual fish-otolith slopes. Fraser–Lee growth back-calculations made from such a regression consistently underestimate previous lengths at age. The bias may explain the apparent ubiquity of Lee's phenomenon. Back-calculation bias was eliminated through use of an algorithm defining individual fish-otolith trajectories and a biologically determined, rather than a statistically estimated, intercept. Adaptations of the biological intercept back-calculations procedure accurately predicted previous lengths in the presence of both stochastic error and time-varying growth rates. When used to reevaluate some published back-calculations, the biological intercept procedure resulted in more accurate values than those previously estimated, and reduced or eliminated the presence of Lee's phenomenon.