运用SS3评估东太平洋大眼金枪鱼(Thunnus obesus)资源——复杂模型和简化模型的比较

摘要

大眼金枪鱼(Thunnus obesus)是东太平洋最重要的商业性金枪鱼鱼种,其资源评估采用的是结构复杂的Stock Synthesis 3模型(SS3).模型简化是提高资源评估效率的必要手段,但对大眼金枪鱼简化模型的效果尚未开展研究.本研究尝试从渔业数据结构的角度,将SS3复杂模型的23个渔业简化为仅含围网和延绳钓2个渔业,从而比较简化模型的评估能力.结果显示,简化模型能较为准确地描述大眼金枪鱼补充量、亲体量、捕捞死亡系数等主要时间序列的历史动态变化,对传统生物学参考点FMSY的估计也较为准确,且受陡度和自然死亡系数的影响较小,但对其他参考点的估算误差较大.陡度参数对简化模型基于Kobe图判断资源状态的准确性有重要影响,陡度较低时,简化模型能较为准确地判断资源状态.研究表明,权衡模型的评估能力和降低模型结构的复杂性,是大眼金枪鱼资源评估今后需要重点研究的任务之一.此外,对模型简化的效果评价,与采用的生物学参考点和资源状况判断标准的选择有关.%The big-eyetuna (Thunnus obesus) is the most commercially important tuna species in eastern Pacific Ocean.The fishery of this species has been managed by Inter-American Tropical Tuna Commission (IATTC).IATTC evaluates the stock of big-eye tuna annually using Stock Synthesis 3 (SS3).SS3 is one of the most commonly used models that assess age-structured fishery stock based on biological parameters and fishery data such as the annual catch,the age-composition,the length composition of catch and abundance indices.A recent assessment of the big-eye tuna stock in eastern Pacific Ocean indicated a small recovery of the spawning stock biomass (SSB) and the fishing mortality (F) with respect to biological reference points (BRPs).The integrated model effectively combines multiple data with various sources into single analysis,and propagates the uncertainties associated with model parameters and structures into model outputs such as the biomass time series and BRP.However,it is time-consuming to use full SS3 model to analyze the big-eye tuna stock as it has as many as 23 fisheries.Therefore,model simplification has become necessary to improve the analysis efficiency.In this study,we tried to simplify the SS3 modeling process by using 2 fisheries (purse seine and longline fisheries) instead of 23.Representative abundance index and size composition from the purse seine fisheries and longline fisheries of the full model were used in the simplified one.The outputs of the simplified model were compared to that of the full model to evaluate the performance of the former.The results showed that the simplified model effectively captured the historical trends of stock time series (e.g.,recruitments,SSB and F) and estimated FMSY with little interference from the steepness and natural mortality parameters.The reduced interference from steepness allowed the simplified model to correctly analyze the stock status based on the Kobe plot.However,the simplified model increased errors in the estimation of other BRPs.This study implied that selecting BRPs and defining stock status of big-eye tuna could affect the evaluation of the performance of the simplified model.Further studies are needed to achieve a balance between high performance and reduced complexity of the model.