COMPARISON OF OTOLITH-BASED BACK-CALCULATION METHODS TO DETERMINE INDIVIDUAL GROWTH HISTORIES OF LARVAL STRIPED BASS, MORONE-SAXATILIS

In rearing studies on 6- to 22-d-old larval striped bass, Morone saxatilis, we applied several back-calculation methods to known-growth larvae. A growth effect occurred on otolith diameter - standard length relationships, where slower growing larvae had relatively larger otoliths. Otolith growth was less affected by feeding regime than was somatic growth. Due to the conservative nature of otolith growth, proportional based (Biological Intercept Method) and simple linear regression methods linearized somatic growth transitions and did not estimate periods of negative growth. A quadratic regression method which used age as an additional predictor resulted in the accurate back-calculation of size at age in all groups of laboratory-reared larvae. However, when model coefficients were applied to a test population of pond-reared larvae, the quadratic model performed poorly. While differences in relative otolith size between pond- and laboratory-reared larvae could be ascribed to a temperature effect, the inability to apply the model also indicates a problem specific to regression-based methods. Theoretical rationale and experimental proof provided evidence for the inclusion of age in back-calculation models, but parameterization will have to occur for each field application.