基于Logistic回归模型的西非沿海欧洲沙丁鱼渔场与环境因素关系模型的构建

A Logistic regression model on relationship between sardine Sardina pilchardus fishing ground and environmental variables in the coastal waters of West Africa

为研究西非沿海不同季节海洋环境因素对欧洲沙丁鱼Sardina pilchardus渔场形成的影响,基于Logistic模型,利用2007—2011年中国渔船在西非沿海捕捞欧洲沙丁鱼的渔业数据和网络获取的海洋环境遥感数据[包括海表温度(SST)、叶绿素a浓度(Chl)和海面高度异常(SSHA)],构建了春、夏、秋、冬4个季节欧洲沙丁鱼渔场与环境因素关系的模型,利用逐步回归法对自变量进行筛选,并对最优回归模型进行评价,最后利用该模型对4个季节影响欧洲沙丁鱼出现的各个环境要素进行了分析。结果表明,4个季节欧洲沙丁鱼的出现与地理位置有着密切的关系,而在4个季节中环境因素对欧洲沙丁鱼出现有着不同的作用,即春季影响欧洲沙丁鱼出现的显著因子为SST和SSHA,夏季为Chl和SSHA,秋季为SST,冬季则为SSHA。该研究结果对欧洲沙丁鱼渔场形成机制分析具有一定的应用价值。

In the study, a Logistic regression model on relationship between sardine Sardina pilchardus fishing ground and environmental variables was established by stepwise regression method based on fishing logbook, spatial information and oceanic environmental factors, including sea surface temperature( SST) , chlorophyll-a concentra-tion ( Chl) and sea surface height anomaly ( SSHA) in western Africa coastal region from 2007 to 2011 in order to evaluate the optimal regression model and each environmental variable impacting the fishing ground in different sea-sons. Results showed that there was a close relationship between the occurrence of sardine population and geograph-ical position. It was found that environmental factors played different roles in sardine distribution in different sea-sons:SST and SSHA showing significant affecting on the occurrence of the fish in spring, Chl and SSHA in sum-mer, and SST in autumn and SSHA in winter. The findings provide definite referential value to analyze the forma-tion mechanism of sardine fishing grounds.