Incorporating heterogeneity into growth analyses of wild and captive broadnose sevengill sharks Notorynchus cepedianus

Growth estimates for shark species are mostly derived from length-at-age data, where age information is commonly obtained from hard structures such as vertebrae and dorsal-fin spines. Unfortunately, hard structures cannot be used for estimating the age of many shark species, particularly those occurring in deep water or belonging to ancient groups. Alternative methods are therefore required for the estimation of growth for these shark groups. The broadnose sevengill shark Notorynchus cepedianus has very poorly calcified vertebrae that cannot be used for age and growth estimations. Three stochastic versions of the von Bertalanffy-Fabens growth model with random parameter k were fitted to length-increment data of captive and of wild tagged and recaptured N. cepedianus from southeastern Australia, California and southern Africa. The model based on a gamma distributed parameter k provided the best fit to the data. Captive females and wild females showed the highest and lowest values of mathematical expectation E[k], whereas wild males and wild females showed the lowest and highest values of mean maximum length L-infinity, respectively. For a time interval of 1. yr, captive females had the highest length increment, whereas wild females had the lowest length increment. Our approach allows accounting for the natural heterogeneity of growth in the estimation of the growth parameters of N. cepedianus, which has not been done previously for captive and wild sharks. This the first study to provide a representative set of growth parameters for male and female N. cepedianus, which is crucial information for the modelling of the population dynamics of this top predator.