大西洋中部大眼金枪鱼的时空分布  

Temporal and Spatial Distribution of Thunnus obesus in the Central Atlantic

作　　者：林泓羽 汪金涛[1,2,3,4,5] 陈新军 雷林[1,2,3,4,5] 邓荣成 隋恒寿[6] 李彬 LIN Hong-yu;WANG Jin-tao;CHEN Xin-jun;LEI Lin;DENG Rong-cheng;SUI Heng-shou;LI Bin(College of Marine Sciences of Shanghai Ocean University Shanghai;Key Laboratory of Oceanic Fisheries Exploration,Ministry of Agriculture;National Engineering Research Center for Oceanic Fisheries,Shanghai Ocean University;Key Laboratory of Sustainable Exploitation of Oceanic Fisheries Resources,Ministry of Education,Shanghai Ocean University;Scientific Observing and Experimental Station of Oceanic Fishery Resources,Ministry of Agriculture,Shanghai 201306,China;CNFC Overseas Fishery Co.,LTD,Beijing 123445,China)

机构地区：[1]上海海洋大学海洋科学学院 [2]农业部大洋渔业开发重点实验室 [3]国家远洋渔业工程技术研究中心 [4]大洋渔业资源可持续开发教育部重点实验室 [5]农业部大洋渔业资源环境科学观测实验站,上海201306 [6]中水集团远洋股份有限公司,北京123445

出　　处：《广东海洋大学学报》2022年第5期123-133,共11页Journal of Guangdong Ocean University

基　　金：国家自然科学基金(NSFC41876141)。

摘　　要：【目的】探究垂直温度梯度对大眼金枪鱼时空分布的影响,为大西洋中部大眼金枪鱼的渔场预报提供科学的参考依据。【方法】根据2016-2019年中国水产集团远洋延绳钓船队在大西洋的大眼金枪鱼生产数据,结合垂直水温及其相邻垂直梯度水温差值,深度范围为0~300 m,采用广义加性模型(GAM)对单位捕捞努力量渔获量(CPUE)进行预测,由累计解释偏差(Accumulation of deviance explained)以及赤池信息准则(Akaike Information Criterion,AIC)选取最优GAM模型。【结果】模型结果表明,累计解释率达到52.9%,各个模型因子效应对CPUE的贡献率产生主要影响依次为经纬度交互效应、月份、深度150 m的水温t_(150)、深度100 m的水温t_(100)和深度300 m的水温t_(300)等。预测CPUE与名义CPUE的时空分布叠加结果表明,整体分布规律基本一致,0°-20°N、30°W-50°W的大西洋中部的西北区域海域为预测的高值区,是适宜大眼金枪鱼捕捞作业的海域。【结论】建立最优GAM模型能够较准确预测大西洋中部大眼金枪鱼的渔场分布,垂直水温结构对其渔场时空分布具有影响。【Objective】To investigate the effect of vertical temperature gradient on the spatial and temporal distribution of Thunnus obesus,and to provide a scientific reference for the fishing forecast of Thunnus obesus in the mid-Atlantic.【Method】Based on the Thunnus obesus production data of China Fisheries Group's ocean-going longline fleet in the Atlantic Ocean from 2016 to 2019,we combined vertical water temperature(t)and its adjacent vertical gradient water temperature difference(D)with a depth range of 0 to 300 m.We used the Generalized Addictive Models(GAM).GAM was used to predict Catch per Unit Effort(CPUE),and the optimal GAM model was selected by Accumulation of deviance explained and Akaike Information Criterion(AIC).【Result】The model results show that the cumulative explained rate reaches 52.9%,and the high-to-low order of each model factor in their contribution to CPUE is latitude and longitude interaction effect,month,water temperature t_(150)for depth of 150 m,water temperature t_(100)for depth of 100 m and water temperature t_(300)for depth of 300 m.The superimposed results of spatio-temporal distribution of predicted CPUE and nominal CPUE show that the overall distribution pattern is basically consistent,and the sea area in the northwest region of the mid Atlantic at 0°-20°N and 30°W-50°W is the predicted high value area,which is suitable for Thunnus obesus fishing.【Conclusion】The established optimal GAM model can predict the distribution of Thunnus obesus fishing grounds in the mid-Atlantic more accurately,and the vertical water temperature structure has influence on the spatial and temporal distribution of its fishing grounds.

关 键 词：大眼金枪鱼 时空分布 大西洋 GAM模型 

分 类 号：S932[农业科学—渔业资源]