Estimating fish growth for stock assessments using both age–length and tagging-increment data

In age-structured stock assessments it would be useful to be able to include all available information on growth, including age–length observations and length increments from tagging experiments. However, it was suggested in 1988 that combing the growth information from these two sources was problematic because the age- and length-based growth information they contain are not directly comparable. We evaluate two approaches that have since been made to this problem and conclude that though both approaches achieve comparability the simpler method was better suited for use in stock assessments, in part because of lesser computational demands. We show how the simpler approach is improved by allowing for correlation between length deviates at tagging and recapture, which increases biological plausibility and corrects a negative bias in estimates of variability in length at age.