Integrating recapture-conditioned movement estimation into spatial stock assessment: A South Australian lobster fishery application

The practical aim of this study was to estimate biomass in two subregions, inshore (<60 m depth) and offshore (>60 m), for application to rock lobster management. To account for lobster movement between these subregions, movement estimation was integrated into the length-structured stock assessment model. Tagged lobsters were released inshore and offshore in the mid-1990s. For fitting to these conventional single tag-recoveries, a recapture-conditioned movement rate estimation method was adapted and integrated into the assessment model. The individually-tailored probabilities of survival during time-at-large and of subsequent capture in each subregion needed to estimate movement rates are provided by the over-arching stock assessment estimator. Calculating a probability specific for each lobster recapture permits the estimator to account for naturally varying times and locations of tag release and recapture. Seasonality of tagged lobster movements indicated that inshore-to-offshore migration occurred predominantly in winter. Yearly movement rates of lobsters were relatively low. Biomass was approximately 3–4 times higher inshore than offshore. Testing using simulated data confirmed the accuracy of model-integrated recapture-conditioned movement estimates. Using these same simulated data, conventional movement rate estimates, calculated as raw proportions of recaptured lobsters that moved, were positively biased. Incorporating survival and recovery rates in movement estimation can thus substantially improve the accuracy of movement, and thus biomass, estimates in spatially-resolved stock assessments.