Study Of Population Dynamics For The Southern Atlantic Albacore (Thunnus Alalunga)

The albacore (Thunnus alalunga) is a species of migratory tuna in the familyScombridae which widely distributed in temperate and tropical oceans. In the AtlanticOcean, the International Commission for the Conservation of Atlantic Tunas (ICCAT)has divided it into three stocks: the northern and the southern Atlantic stocks(separated at5oN) and the Mediterranean stock. Although the status of the southernstock is better than the northern stock, the2011Albacore assessment meeting byICCAT suggest that the southern stock may be facing with overfishing. InternationalSeafood Sustainability Foundation (ISSF) indicated the southern albacore stock hadsuffered overfishing at the beginning of21stcentury, although the current yield isbelow MSY level, this stock must be taken care of. Therefore, the populationdynamics models were applied on the assessment of this stock in order to give someuseful management recommendations.CEDA and ASPIC softwares were applied on the data of the southern Atlanticalbacore fishery in Chapter2. Results show that in CEDA the Fox surplus productionmodel with log-normal error assumption produced the highest R2value and resultsclose to those of ASPIC. CEDA is sensitive to the choice of initial proportion. InASPIC the Logistic production model is more sensitive to the choice of initialproportion than Fox model. MSY estimated by these two softwares is28300tons.With the development of fishery population dynamics, the uncertainty of themodels and risk analysis were gradually applied in fish stock assessment. Therefore,the study of Chapter3applied a Bayesian Schaefer model on the data of the southernAtlantic albacore fishery, taking the results of estimated parameters of carryingcapacity, intrinsic rate of increase and catchability coefficient from ASPIC and CEDAas the prior information. Monte Carlo Markov-Chain (MCMC) algorithm was used tocompute the posterior distribution of the three parameters. This study set eight harveststrategies to do risk assessment for this stock. The biomass of the year2025is less than BMSYwith a great probability of0.61. It concluded that the catch of27970t andthe harvest rate of0.15seem to be the best management measures.The relationship between stock and recruitment (SR) is considered as anessential part of fish population dynamics and it can be used to determine thebiological reference points. Chapter4applied a Bayesian analysis on the SR data ofthe southern Atlantic albacore stock. The posterior distributions of the parameters inboth the Shepherd and the Beverton-Holt (B-H) SR models are given. This study alsofits the SR data using three two parameters SR models based on the least squaremethod and then verify the posterior distribution of the “degree of compensation”parameter in the Shepherd SR model. According to the results of the Bayesiananalysis, this paper estimates the uncertainty on MSY. For the B-H SR model, the80%confidence interval of MSY is (19168,25860) t, with a median of22085t. Forthe Shepherd SR model, the80%confidence interval is (21414,29044) t, with amedian of24799t.The delay-difference models are intermediate between simple surplus productionmodels and complex age structured models. Being biologically meaningful and lessdata requirement than age-structured models are their main advantages. Chapter5applied a delay-difference model to fit the CPUE and catch data from1975-2011forthe southern Atlantic albacore stock. Results showed that the predicted CPUE usingthe delay-difference model can capture the inter-annual variations better than the Foxmodel. In the Monte Carlo simulation analysis this study superimposed white noises(CV) in observed CPUE data with four CV levels and applied the relative estimateerror (REE) to compare the results of the estimated, β in Ricker SRR models and thecatchability q to the true values. Results showed that the parameters are moresensitive to the white noises than β and catchability q. The southern Atlantic albacorestock had experienced overfishing from1985to2005. After that, this stock rebuildsgradually. Because the estimated ratios of catch against MSY are near1and thefishing mortality coefficient was above the MSY level in recent years, this studyconsidered that the albacore fishery in the southern Atlantic Ocean may sufferoverfishing. Therefore, this stock must be exploited with precaution. The delay-difference model has given a good fit for the data of the southern Atlanticalbacore stock. Due to the lack of biological realism for surplus production modelsand the complicated data requirement for age-structured models, the delay-differencemodel may be a useful choice.Above all, this study suggests that the southern Atlantic albacore stock issuffering slight overfishing and the yield must keep below22490t for sustainabledevelopment. With the work of this study, the author hopes to supply some worthyinformation to the pelagic fishery and to provide useful contribution to the fisheryassessment and management of our country.