Accurate estimates of tag-induced mortality rates are contingent on the number of tagged and recaptured lobsters

Tag-induced mortality (TIM) biases many capture-recapture studies, leading to abnormally high mortality estimates in the first-year post-tagging. Although models exist to account for this bias, estimating TIM has been problematic and restricted to artificial environments. Here, we use a method for estimating Jasus edwardsii (Hutton, 1875) TIM in situ and demonstrate the conditions under which accurate estimates can be achieved. We use a long-term capture-mark-recapture study conducted since 2000 at the Crayfish Point Scientific Reserve (CPSR), Hobart, Tasmania, Australia, to estimate the rate of in situ tag induced mortality and demonstrate the assumptions relating to sampling design that are required to achieve accurate estimates. TIM estimates were high and relatively similar for both males and females. The similarity between sexes would indicate that for this species, combined sex estimates may be sufficient, which requires substantially less effort. Estimates of TIM were sensitive to the number of recaptured lobsters and at least 15 lobsters, tagged in an initial survey, had to be captured in two subsequent surveys. As recapture rates for lobsters over two subsequent recapture events are relatively low, this resulted in a large number of lobsters needing to be tagged in the initial survey. Given that most tagging studies have at least three surveys, we suggest that the design incorporate the ability to also estimate TIM. This is particularly important if tagging studies are used to estimate population parameters for exploited species, as not accounting for TIM would lead to overestimation of resources and inappropriate catch allocations.