Résumé
This paper describes the model FASST. FASST is a fully structured age, size, space and time stock assessment model. It is expressed as a deterministic system of partial differential equations continuous in age, size and time. Space is represented using large discrete zones interconnected by exchange rates. All the parameterizations used to represent recruitment, growth, mortality and movements are chosen to be size-dependent and age independent. Consequently, for simplicity purposes, the age dimension can be simplified. The model is integrated numerically to simulate the stock dynamics. A complex process-and observationerror framework is used to estimate the model parameters, and hence the past stock status, in a Bayesian context. The model uses simultaneously most fishery data currently available for tuna stock assessment such as catch by fleet in number and weight, length-frequency samples, otolith analysis, tag-recapture data. It provides a new way to integrate that information into a consistent stock assessment framework.