Abstract
This study addresses the inherent uncertainty when estimating growth from limited mark–recapture information. A selection procedure was developed utilizing 18 competing growth estimation methods. The optimal method for a given data set was identified by simulating the length at capture and recapture under different scenarios of measurement error and growth variability while considering the structure of observed data. This selection procedure was applied to mark–recapture data for 37 female and 16 male tope sharks Galeorhinus galeus obtained from tagging studies in the north‐east Atlantic Ocean. Parameter estimates differed strongly among methods, showing the need for careful method selection. The selection approach suggested that best estimates for males and females were given by James' weighted least‐squares approach with a fixed asymptote. Given an average total length ( L T ) at birth of 28 cm, the von Bertalanffy growth function of north‐east Atlantic G. galeus would be L T = 200·85 − (200·85 − 28) e − 0·076 t for females and L T = 177·30 − (177·30 − 28) e − 0·081 t for males. The resulting age estimates were up to 11 years lower when compared with previous estimates derived from highly uncertain vertebrae readings. More generally, this procedure can help identify optimal estimation methods for a given data set and therefore aid in estimating more reliable growth parameters from mark–recapture information.