An empirical comparison of estimation procedures for the von Bertalanffy growth equation

Abstract

One non-linear and two linear methods of fitting the von Bertalanffy growth equation to length-age data were compared using Monte Carlo simulations of fish populations while varying the standard error of the length, total sample size, sampling time interval, von Bertalanffy growth parameter, and annual adult survival. The iterative, non-linear method usually produced the most accurate and precise parameter estimates. The non-linear method also provided asymptotic confidence intervals about point estimates, placed fewest constraints on data collection, and was the easiest to use. It is suggested that traditional linear solutions to the von Bertalanffy growth equation be abandoned.