Stock Assessment For Indian Ocean Swordfish Based On Bayesian State-space Surplus Production Model

Swordfish（Xiphias gladius）is a large billfish species with high economic value,widely distributed in tropical to temperate waters of the three oceans,and is at the upper end of the food chain.As an important bycatch economic species in China’s longline fishery,it is important to assess its resources and develop management strategies for the sustainable use of resources and ecosystem conservation.This study assessed the resource of Indian Ocean swordfish based on the fishery logbook data of China’s longline fishery and the fishery data published by the Indian Ocean Tuna Commission（IOTC）,taking into account various uncertainties.In the field of stock assessment,Catch Per Unit Effort（CPUE）is generally regarded as a relative index of resource density,which is usually assumed to be positively correlated with the resource.In this paper,a Generalized Addictive Model（GAM）was used to standardize the CPUE data of Indian Ocean swordfish in the context of China’s longline fishery production data in the Indian Ocean.The best standardized model was selected according to the AIC criterion,and the maximum explanation rate of CPUE was27.3%.The model results indicated that the significant variables were weather,latitude,and month.The effect of latitude and month were more significant,explaining 7.87%and7.43%of the total deviation,respectively.The higher CPUE occurred between 10°N-10°S,which may be related to the preference of swordfish for warm waters.The standardized CPUE showed a trend of decreasing,then increasing,and then decreasing,with high fluctuations.As one of the important species in Indian Ocean commercial fisheries,it is necessary to do routinely stock assessment to judge the resource status.In this paper,based on the surplus production model JABBA（Just Another Bayesian Biomass Assessment）,we assessed the resource status of swordfish in the Indian Ocean based on historical catches from 1950-2018 and publicly available CPUE data,and combine the standardized CPUE data of Indian Ocean longline swordfish fishery in China,to compare and analyze the influence of standardized CPUE on the assessment results.The results showed that the model was more sensitive to the different standardized CPUE data.The RMSE（root mean squared error,RMSE）and DIC（deviance information criterion）values were smaller when the standardized CPUE data of three longline fleets from Japan,Portugal and South Africa were selected.The base case suggests that the 2018 Indian Ocean swordfish resource’s Maximum Sustainable Yield（MSY）was estimated of 30,900 t,which is higher than the current catch of 30,100 t.There is a 76.6%probability of being neither overfished nor overfishing.The assessment results were less sensitive to the a priori distribution of the intrinsic growth rates parameter r,and the parameter r was negatively correlated with the posterior distribution of the environmental carrying capacity K.The retrospective analysis showed that the model did not have significant retrospective problems and its results were closer to reality.The projection analysis shows that the resource remains in neither overfishing nor overfished until 2028 when the total allowable catch（TAC）is below 32000 t.JABBA-Select is an expanded version of JABBA that accurately accounts for the moderate changes in fishing selectivity and fleet dynamics over time,and distinguishes between Spawning Biomass（SB）and Exploitable Biomass（EB）.As a tool to achieve the management objectives of the fisheries ecosystem,the assessment of the resource in combination with this influencing factor can better approximate the actual situation and provide a reference for the development of management strategies.In this study,taking Indian Ocean swordfish as an example,eight model case were set up for analysis,stock assessment and projection based on the different fishing selectivity differences between fleets,relationships of EB and SB,and different priori distributions of steepness information.The results of the base case indicated that the Indian Ocean swordfish resource was in health with 98%probability,It estimate MSY 31700 t and reference point SB₂₀₁₈/SB_MSY=1.66 and F₂₀₁₈/F_MSY=0.59,which means either overfishing nor overfished.The TAC projection analysis for the next 10 years indicates that the stock will remain healthy as of 2028 even if catches increase to 36,000 t.The results of both the biological reference point estimates and the resource projection analysis of JABBA-Select were in high agreement with the IOTC assessment using the full assessment model SS3,indicating that in some cases of data-limit,JABBA-Select’performance can even reach the age structure assessment model.This study provides a reference for many fishery stock assessments with moderate data.