Maximilien Simon

Models for fisheries stocks assessment of tunas mostly depend on commercial fisheries information. Bayesian modeling framework is particularly suitable for taking into account observation errors resulting from these data and the uncertainty of diagnoses resulting from models. In this context, the use of informative probability distributions (“priors”) for some parameters can facilitate the inference by integrating information on biology of the species. We examined the possibility of eliciting informative priors for 2 demographic parameters for bluefin tuna (*Thunnus thynnus*), and yellowfin tuna (*Thunnus albacares*): *(i)* the steepness, which is a parameter involved in the relationship between spawning biomass and the number of individuals entering in a fishery exploitable stage. *(ii)* The intrinsic population growth rate, one of the key parameter of the biomass production model. The assumption that a probability distribution of each of these parameters can be characterized by life history traits is discussed. A review of the literature for various biological and ecological information for tuna has been achieved. From the synthesis of this information, probability distributions for fecundity and mortality rates at age are estimated. 20000 random drawings are made in these distributions and for each drawing, an age-structured population model is used to determine the 2 parameters of interest. An empirical distribution of each parameter is obtained and the sensitivity of the result to assumptions on fertility and mortality is evaluated. We showed that the largest uncertainty in the life cycle of tuna is on the survival of eggs and larvae and hence of mortality of young of the year. Furthermore, the distribution of the steepness and the population growth rate are highly sensitive to this mortality parameter. Finally, the lack of understanding in recruitment dynamics induce strong limitations to the construction of informative priors for these parameters.

**Keywords:** Fisheries; Prior distribution; Demographic variables; Tunas

**Biography:** PhD student at the french institute for marine research (IFREMER). Working on fisheries modelisation for application to tuna stocks assesments