La pêche au thon indonésienne a souffert d'un problème de gestion dû à des statistiques de pêche incomplètes et peu fiabilité, ce qui entraîne un manque de compréhension de la dynamique des populations de thonidés de la région. Le gouvernement de l'Indonésie a lancé un programme de développement de l'infrastructure pour l'océanographie spatiale (INDESO) afin d'aider à la gestion et au suivi des ressources marines du pays. Une application du projet concerne les pêcheries thonières avec l'objectif ambitieux de modéliser en temps réel et avec des prévisions de quelques jours les distributions et abondances de trois espèces de thonidés: thon obèse, albacore et bonite. Le modèle utilisé est SEAPODYM (modèle spatial d'écosystème et de dynamique de population). La présente thèse traite uniquement du thon obèse et a trois objectifs majeurs: la préparation d'un ensemble de données de pêche géoréférencées, la production de conditions initiales pour la configuration du modèle régional INDESO et la simulation de la dynamique régionale des populations. L'ensemble de données sur les captures et les efforts de pêche géoréférencés du thon obèse de l'océan Indien a été standardisé selon cinq procédures: standardisation de la résolution spatiale, conversion et standardisation des unités de capture et d'effort, recalage des captures géo-référencées au niveau des captures nominales, et détection des principaux changements de capturabilité sur de longues séries temporelles de données sur la pêche.. L'ensemble de données de prises géoréférencées standardisées couvre les deux tiers de la capture nominale totale en raison du manque de références géographiques pour plusieurs flottilles de pêche. Le modèle régional a été configuré en trois étapes: paramétrage du modèle à résolution grossière sur une longue période historique, " downscaling " et paramétrage de la configuration globale opérationnelle et " downscaling " vers le modèle régional opérationnel. La première étape a permis de paramétrer le modèle sur le Pacifique puis l'Océan Indien sur une période de trente-neuf ans à une résolution mensuelle de 2°, permettant d'établir les conditions initiales de la population pour la deuxième configuration, à partir de 1998 et à une résolution hebdomadaire 1/4°. Cette deuxième configuration du modèle a nécessité une méthode de " downscaling " pour réviser le paramétrage et obtenir la même solution malgré quelques différences dans le forçage physique. Ce modèle opérationnel global a ensuite fourni des conditions initiales de la population et des conditions aux frontières ouvertes pour les flux de poissons traversant les frontières régionales du modèle INDESO (1/12 ° de résolution quotidienne). L'ensemble des données de pêche standardisées dans l'océan Indien a été utilisé pour inclure la mortalité par pêche et valider l'optimisation réalisée dans l'océan Pacifique. Les résultats des simulations de modèles suggèrent que le thon obèse est distribué en concentration plus élevée dans l'océan Indien Nord (au nord de 20 °S), avec une extension à travers le canal du Mozambique et le long d'un prolongement vers l'est entre 35 ° et 40 ° S. Les configurations du modèle opérationnel (mondial et régional) utilisent la production primaire nette (VGPM) et la profondeur euphotique dérivées des données satellitaires, ainsi que les cartes mensuelles climatologiques de l'oxygène dissous provenant de l'Atlas Mondial des Océans (WOA) comme forçages biogéochimiques. Une simulation régionale utilisant le modèle INDESO biogéochimique PISCES comme forçage alternatif aux produits dérivés des satellites a été testée. Les résultats préliminaires montrent que, une fois la production primaire PISCES ajustée à la valeur moyenne VGPM, les deux produits donnent des résultats similaires, ce qui suggère que des prévisions à plus long terme basées sur le modèle biogéochimique physique couplé peuvent être proposées. / The Indonesian tuna fishery has suffered from a management problem due to incomplete and less reliability of tuna data leading to lack of understanding about tuna population dynamics in its region. The government of Indonesia initiated the Infrastructure Development of Space Oceanography (INDESO) programme to support marine resource management and monitoring of the Country. One application concerns the tuna fisheries with a challenging objective of real-time and forecast modeling of three tuna species biomass distributions: bigeye, yellowfin and skipjack. The model used is SEAPODYM (Spatial Ecosystem and Population Dynamics Model). The present thesis is dealing with bigeye tuna only, and had three major objectives: the preparation of a geo-referenced fishing dataset, the production of initial conditions for the regional INDESO model configuration, and the simulation of regional population dynamics. The georeferenced fishing catch and effort dataset of the Indian Ocean bigeye tuna was standardized throughout five procedures: standardization of spatial resolution, conversion and standardization of catch and effort units, raising of geo-referenced catch to nominal catch level, screening and correction of outliers, and detection of major catchability changes over long time series of fishing data. . The standardized geo-referenced catch dataset covers two-third of total nominal catch due to lack of geographic references for several fishing fleets. The regional model was configured along three steps: the parameterization of coarse resolution model over a long historical period, the downscaling and parameterization of operational global configuration, and the downscaling to the operational regional model. The first step provided model parameterization over the Pacific and Indian Ocean for thirty-nine years period at 2° monthly resolution, allowing to establish initial conditions of the population for the second configuration starting in 1998 at resolution 1/4° weekly. This second model configuration required a downscaling method to revise the parameterization and achieve the same solution despite some differences in the physical forcing. This global operational model provided initial conditions of the population and open boundary conditions (OBCs) constraining the fluxes of fish through the regional borders of INDESO model (1/12° daily resolution). The standardized Indian Ocean fishing dataset was used for including fishing mortality and validate the optimization achieved in the Pacific Ocean. Model simulation outputs suggest that bigeye is distributed in higher concentration in the North Indian Ocean (north of 20°S), with an extension through the Mozambique Channel and along an eastward prolongation between 35° and 40°S.. The operational model configurations (global and regional) are using VGPM net primary production and euphotic depth derived from satellite data, and climatological dissolved oxygen monthly maps from the World Ocean Atlas (WOA) as biogeochemical forcings. A regional simulation using the INDESO biogeochemical PISCES model forcing as an alternative to the satellite derived products was tested. The preliminary results show that once the PISCES primary production is scaled to the VGPM mean value, both products provide similar results, suggesting that longer time scale forecast based on the coupled physical biogeochemical model can be proposed.

This study used the Gordon-Schaefer model to discuss the equilibrium levels for bigeye tuna in the Western and Central Pacific of open access and present value maximization. And then to compare the catches and the stocks on the two model¡¦s equilibrium value, the result shows the management of bigeye tuna in the Western and Central Pacific tend to present value maximization model. Following that, this paper did sensitivity analysis to understand the impact of the changed parameters on stock size and effort. Finally, using simulation analysis on open access model and present value maximization model. In open access model, the result shows that resources will face extinction crisis if the fishery is not controlled well. In the present value maximization model, the bigeye tuna would be sustainable management. This result is valuable for the fishery management authorities to maintain the development of fishery and cherishing ocean resources at the same time.

This study used the Gordon-Schaefer model which was extended to Open Access (OA) model and Present Value Maximization (MPV) model to discuss the equilibrium levels for bigeye tuna in the Western and Central Pacific Ocean by purse seine fishery data. I compared the catches and the biomass with the two model¡¦s equilibrium value, and the result showed that the operating mode of bigeye tuna in the Western and Central Pacific Ocean tended to the MPV model, and the bigeye tuna resources were diminishing. In addition, the sensitivity analysis was used in order to understand the impact of various parameter changes on the two fisheries model¡¦s equilibrium value. In the OA model, the change of the price, the cost per unit effort and the catch coefficient would have greater impact on the equilibrium biomass. In the MPV model, the equilibrium biomass was sensitive to the change of the environmental capacity. Finally, in order to understand whether the management of the Western and Central Pacific Commission (WCPFC) was effective, I used simulation analysis in accordance with the catches restriction and effort restriction management strategies. The result showed that the catches restriction strategy which WCPFC established was still not proper enough. It must cut a large number of catches immediately and continue for some time enough to make the resource recover. The effort restriction strategy could make good effect on the bigeye tuna resource recovery.

The objects of this study were to investigate the differences of total mercury(THg), organic mercury(OHg) concentrations in the muscles of albacore and bigeye tuna from the Pacific Ocean, and the THg and OHg concentration in the livers , and the trace elements (Cu, Fe, Mn and Zn) in the muscles and livers of albacore, in order to evaluate the bioaccumulative status of the oceanic migratory fishes. Meanwhile, The relationships between mercury concentration and size (length and weight) and age were established, and the differences of mercury concentrations in the tuna between species and geography were examined. And that the safety consumption level was also discussed. One hundred eighty-five albacore and 134 bigeye tuna were sampled from the period October 2001 to April 2006 and from September 2005 to November 2006, respectively, in the Western and Central Pacific Ocean. The results showed that the average THg (mg/kg flesh wt.) and OHg (mg/kg flesh wt.) of muscles were 0.435¡Ó0.145 and 0.279¡Ó0.087, respectively, for albacore, and 0.935¡Ó0.655 and 0.544¡Ó0.396, respectively, for bigeye tuna. The THg and OHg of livers of albacore were 0.426¡Ó0.363 and 0.193¡Ó0.121, respectively. The average OHg percentages of the THg in albacore and bigeye tuna were 67¡Ó19 % and 62¡Ó20 %, respectively. Except there was no significant correlation between OHg and age in the liver of albacore, the THg and OHg in the muscles, and THg in the livers of albacore, and THg in the muscles of bigeye tuna showed a positive correlation with size and age. The THg and OHg levels in the muscles of albacore were lower than those of bigeye tuna, and the accumulative rates of THg and OHg in albacore were lower than those in bigeye tuna. No significant different of muscles-THg and -OHg concentrations of albacore from the four samplind areas. However, the THg and OHg levels in muscle of bigeye tuna showed that north group higher than those in equator. The concentrations of THg were similar in muscle and liver of albacore, but those OHg, Cu, Fe, Mn and Zn concentrations in liver were significantly higher than those in muscle. A detoxificated mechanism may be commenced when the concentration of THg in liver of albacore exceed 0.8 mg/kg, to maintain the muscle concentration of THg at a level about 0.7 mg/kg. The OHg concentrations of all albacore and 88% of bigeye tuna were below WHO food safety standard for migratory fishes(1.0 mg/kg). The consumption amount of albacore and bigeye tuna were no more than 300 and 150g, respectively, will not exceesed the PTWI(provisional tolerable weekly intake) set by WHO, that will not cause any health threat for consumer.

Abstract Because the number of sharks is decreasing , the problem has been concerned in recent years. In this paper , I conduct a research about the shark bycatch problem¡Ataking the target specie of bigeye tunas and the bycatch specie of sharks as the research objects . Initially I collect from 2000 to 2007 Atlantic bigeye tuna and shark catches datas for resource assessment.and then making comparasion with resource equilibrium values of open access fishery model and net present value maximization fishery model. it is found during the period it has showed a upward trend for bigeye tuna resources¡Abut it has not showed the bigeye tuna resources achive the optimal level. It has showed a diminishing trend for shark resources, the result shows the resources will face extinction crisis if the fishery is not controlled well. Then, doing sensitivity analysis to understand the effects of exogenous parameters to bigeye tuna and shark catches, resources and efforts. Finally, facing with the sharks bycatch problem , discussing the effect of improved fishing gears on sharks bycatch control by doing the sensitivity analysis of fishing hook cost and bycatch coefficient to shark resources and catches¡OFrom the results , it shows that no matter affecting the fishing hook cost or bycatch coefficient , the amount of shark can be reduced effectively by adopting improved fishing gears.

Le thon obèse (Thunnus obesus), le thon alabore (Thunnus albacares) et le listao (Katsuwonus pelamis) comptent parmi les espèces de thons les plus utilisées en conserve. Lors de la fabrication de conserves de thon, la substitution d'espèce et/ou le mélange de différentes espèces de thon sont interdits par la réglementation européenne. L'authentification des espèces de thon reste complexe à cause du degré de similitude élevé entre les espèces de thon, ou encore, lorsque les caractéristiques morphologiques externes sont éliminées au cours du filetage et lors de la mise en conserve. Par conséquent, des substitutions involontaires ou frauduleuses peuvent se produire. Dans cette étude, le marqueur mitochondrial du gène de la sous-unité 2 de la NADH déshydrogénase a été utilisé pour identifier le thon obèse et le gène de la sous-unité II de la cytochrome c oxydase a été utilisé pour identifier le thon albacore et le listao en utilisant la PCR en temps réel basée sur la technologie TaqMan. Deux méthodes différentes basées sur la qPCR ont été développées pour quantifier le pourcentage de chair de chaque espèce présente au sein d'une boîte de thon. La première a été basée sur la quantification absolue avec standard externe réalisée avec les deux marqueurs. La seconde a été basée sur la quantification relative avec standard externe avec le gène endogène de l'ARN 12S. Sur la base de ces résultats, nous pouvons conclure que notre méthode peut s'appliquer pour quantifier les deux espèces de thon albacore et obèse génétiquement très proches lorsqu'elles sont utilisées dans un mélange binaire en conserve. / Bigeye tuna (Thunnus obesus), yellowfin tuna (Thunnus albacares) and skipjack tuna (Katsuwonus pelanis) are among the most widely used tuna species for canning purposes. Not only substitution but also mixing of tuna species is prohibited by the European regulation for canned tuna products. However, it can be difficult to authenticate the tuna species, due to their high degree of similarity or even when the external morphological characteristics are removed due to filleting before canning. Consequently, involuntary or fraudulent substitutions may occur during the canning process. In this study, the mitochondrial marker from NADH dehydrogenase subunit 2 gene was used to identify bigeye tuna and the mitochondrial marker cytochrome c oxidase subunit II gene was used to identify yellowfin tuna and skipjack tuna, utilizing TaqMan qPCR methodology. Two different qPCR-based methods were developed to quantify the percentage of flesh of each species used for can processing. The first one was based on absolute quantification using standard curves realized with these two markers ; the second one was founded on relative quantification with the universal 12S rRNA gene as the endogenous gene. On the basis of our results, we conclude that our methodology could be applied to authenticate the two closely related tuna species (bigeye tuna and yellowfin tuna) when used in a binary mix in tuna cans.