Fitting growth with the von Bertalanffy growth function: a comparison of three approaches of multivariate analysis of fish growth in aquaculture experiments

Three approaches for multivariate analysis of fish growth in aquaculture experiments with Nile tilapia (Oreochromis niloticus niloticus L.) based on the von Bertalanffy growth curve are presented and compared. The approaches are: an extended Gulland-and-Holt (GH) plot, a forced extended GH plot and a multilinear regression analysis for the growth parameter K. All three models provide valuable insight into the major environmental factors influencing the daily growth rate and explain 28-46% of the variance of the observed daily growth rate of the used data set. For all three methods, the modelled parameter is significantly related to the net yield of Nile tilapia and can, therefore, be used for the predictive modelling of management scenarios. The extended GH plot loads the influence of environmental parameters upon L-infinity, while the forced extended GH plot and Direct fitting of K load the influence on the growth parameter K. The latter is more in the tradition of aquaculture research. But the forced extended GH plot and Direct fitting of K can only be applied if L-infinity of the cultured species is known, as the selected L-infinity influences the variance in the regression variables.