Effects of intrapopulation variability on von Bertalanffy growth parameter estimates from equal mark-recapture intervals

The effects of intrapopulation variability on the parameter estimates of the yon Bertalanffy growth equation have received discussion in the literature. Here we evaluated the effects of intrapopulation variability, using computer simulations, on four commonly used methods for estimating the yon Bertalanffy growth parameters: the Ford-Walford plot, Ricker's method, Bayley's method, and Fabens' method. Intrapopulation variability in growth rates (k) and maximum sizes (L.,) plus initial size distributions and measurement error, were tested for their effects on the accuracy of the parameter estimates using simulated mark-recapture data with equal recapture intervals. Fabens' method and a modified Ford-Walford plot provided the most accurate estimates in all cases, but when intrapopulation variability was large, they performed poorly. With moderate intrapopulation variability, the bias in estimates was small although between-sample variance was quite large. Biased initial size distributions without either small or large size classes cause a magnification of the estimation errors. Without knowledge of the degree of intrapopulation variability in a natural population, large errors of unknown magnitude in parameter estimation can result, and care should be taken when interpreting these estimates. However, if this variability can be quantified, then approximate parameter estimate errors can be obtained.