Quantitative modelling of the spatial dynamics of South Pacific and Atlantic albacore tuna populations

Abstract

We present a modelling approach to estimate and predict the dynamics of highly migratory albacore tuna populations in the South Pacific and Atlantic oceans. We use the previously developed model SEAPODYM (Spatial Ecosystem And POpulation DYnamics Model) with its parameter estimation approach based on the maximum likelihood estimation and adjoint method. The model describes temporal and spatial population dynamics using continuous advection-diffusion-reaction equations with an ageing term. We introduce a method to account for seasonal spawning migrations as movements of population density modelled in a Eulerian framework. The aim of this study is to explore the capacity of SEAPODYM to predict complex spatial dynamics of highly migratory species that are consistent with available data and existing knowledge on albacore biology. The geo-referenced fisheries dataset used for model calibration includes effort, catch and length-frequency distributions, constituting the total fishing pressure on the South Pacific albacore population over more than thirty years of exploitation. The model is then validated on the fishing data for Atlantic albacore, thus providing evidence of estimated model parameter invariance with respect to space and time. Given the robustness of the results, the quantitative approach presented here could be used to assist stock assessment and to improve management advice.