The phenomenon of apparent change of growth rate in gummy shark (Mustelus antarcticus) harvested off southern Australia

The hypothesis of the 'Phenomenon of Apparent Change in Growth Rate' caused by length-selective fishing mortality is adopted for explaining observed d ifferences in published von Bertalanffy growth curves for gummy shark (Must elus antarcticus) determined from length-at-age data. Three independent pie ces of evidence are presented in support of this hypothesis to explain the differences in the curves between 1973-76 and 198-87 in Bass Strait and bet ween Bass Strait and South Australia during 1986-87: (1) Mean length of the sharks in each of the ages-classes 3-7 years are shown to be different bet ween the two periods and the two regions, but not different for the 2-year age-class which is affected less than the older age-classes by the fishing gear deployed in the fishery. (2) Avoiding the pitfalls of 'back-calculatio n', Rosa Lee's Phenomenon was detected by directly comparing the radii of g rowth-increment bands visible on the faces of the articular cups of vertebr al centra from sharks of various ages caught in the two periods and the two regions. Statistical tests indicated that the degree of Rosa Lee's Phenome non did not vary significantly between the periods or regions. (3) Through development of an appropriate model, the effects of length-selective fishin g mortality oa the mean length of sharks in the wild population for age-cla sses 2-16 years were simulated for a range of levels of hook and gillnet fi shing effort, with separate mesh-sizes of 6 and 7 in. for the gillnets. Sim ulated changes in mean length for sharks older than two years tended to be smaller than the differences observed in the published von Bertalanffy grow th curves but they were generally consistent with the observed trends. The simulations demonstrated how the magnitude of the von Bertalanffy growth pa rameters L-infinity and t(0) tend to increase and K tends to decrease as gi llnet fishing effort increases, and hence explain how these types of biases , commonly appearing in the scientific literature for gillnet shark fisheri es, can occur. The simulations also demonstrate that hook fishing effort in a fishery with legal minimum lengths has to be very high to cause distorti on of growth curves. (C) 1998 Elsevier Science B.V. All rights reserved.