Comparison of age-length and growth-increment general growth models of theSchnute type in the Pacific Blue Mussel, Mytilus trossulus Gould

Models of Mytilus growth, based mostly on length-at-age data, have typicall y taken the form exemplified by the von Bertalanffy or Gompertz formulation s. These models require assumptions about the form of the growth curve. We used the Schnute general growth model to examine growth in the Pacific Blue Mussel, Mytilus trossulus, in Prince William Sound, AK. The Schnute model provides a convenient analytical method for selecting among all previously published growth models. Mussels were tagged with individually numbered tag s at 13 sites in July 1997 and were collected in July 1998. Age was determi ned from surface growth rings on the shell, and shell length at maximum ann ulus was measured. Annual deposition of the growth rings was verified throu gh radial sections of mussel valves, aided by acetate peels, in conjunction with in situ annual growth measurements. Growth was modeled with the Schnu te general growth model for age-length data or with an analog of the Schnut e model for growth-increment data. Bootstrap confidence intervals were obta ined for all parameters of the model and for model predicted lengths at eac h annulus. Confidence intervals of the between-annuli growth-increment mode l overlapped those of the age-length model at all annuli when growth over t he entire range of ages in the population was estimated. Differences in gro wth model parameters between the age-length model and the mark-recapture an alog could be accounted for solely by inherent differences in age-based ver sus length-based models. Growth estimates generated from between-annuli mea surements were equivalent to growth estimates obtained from mark-recapture measurements of annual growth. In general, mussel growth at our sites was b est described by the von Bertalanffy submodel. (C) 2001 Elsevier Science B. V. All rights reserved.