基于产卵场环境因子的阿根廷滑柔鱼资源补充量预报模型研究  

Study on forecasting model of recruitment for Illex argentinus by using the environmental factors in the spawning ground

作　　者：汪金涛[1] 高峰[1,2,3] 雷林[1,2,3] 官文江[1,2,3] 陈新军[1,2,3] 

机构地区：[1]上海海洋大学海洋科学学院,上海201306 [2]大洋渔业资源可持续开发省部共建教育部重点实验室,上海201306 [3]上海海洋大学国家远洋渔业工程技术研究中心,上海201306

出　　处：《海洋学报》2014年第12期119-124,共6页

基　　金：国家863计划(2012AA092303);国家发改委产业化专项(2159999);上海市科技创新行动计划(12231203900);国家科技支撑计划(2013BAD13B01)

摘　　要：西南大西洋阿根廷滑柔鱼Illex argentinus是短生命周期种类,其资源量极易受到海洋环境变化的影响。根据2003—2011年我国鱿钓船队在西南大西洋的生产统计数据,以及产卵场海洋表面温度（SST）、海表温度距平值（SSTA）,计算分析了阿根廷滑柔鱼在产卵期产卵场各月最适表层水温范围占总面积的比例（用PS表示）以及表征海流强度的SST、SSTA等多种环境变量因子与单位捕捞量渔获量（CPUE）的相关性,建立多种基于主要环境因子的资源补充量预报模型,同时分析比较预报模型的优劣。相关性分析表明：6月份有3片连续区域的SST与CPUE之间存在强相关性,分别为38°～39°S、54°～55°W,40.5°～41.5°S、51°～52°W,39.9°～40.4°S、42.6°～43.1°W。利用6月份此3片连续区域SST与次年CPUE建立的三元线性模型,模型符合统计检验,偏差解释率为82.4%。在此基础上加入7月份PS影响因子建立3种方案下的误差反向传播（EBP）神经网络模型。结果认为,包含了福克兰寒流与巴西暖流表温信息的方案3模型优于其他两种模型,其准确率可以达到90%以上。Illex argentinus is short-life cycle squid and is sensitive responding to environment changes with great a- bundance fluctuations. In this study,according to the fishing production data from Chinese mainland squid jigging fleets from 2003 to 2011 in the southwest Atlantic,combined with sea surface temperature （SST） and sea surface temperature anomaly （SSTA） in the spawning ground, the relationships between the area occupied by favorable SST （defined as those with temperatures in the range from 16℃ to 18℃, expressed by Ps）, current strength （Characterized by SST,SSTA ） and catch per unit effort （CPUE） were calculated and analyzed by different meth- ods, and then the forecasting model of resources recruitment based on the above environmental factors were estab- lished. The results indicated that there are significant correlations between CPUE and SST at three key areas on June. The key areas are as follows： 38°--39°S and 54°--55°W,40.5°--41.5°S and 51°--52°W,39. 9°--40.4°S and 42.6°-43. 1°W, respectively. A multivariate linear model between the SST in three key areas and CPUE of the next year is established （p〈0.05）. The other models by using Error Baekpropagation Network （EBP） were also developed,which included SST in three key areas and Ps on July. It is found that the model with SST in the key areas indicationg Falkland current and Brazil current is better than the other models, the forecasting accuracy rate was more than 90%.

关 键 词：西南大西洋 阿根廷滑柔鱼 补充量预报 神经网络 

分 类 号：S931.41[农业科学—渔业资源]