基于MAXENT模型预测西北太平洋秋刀鱼潜在渔场

Prediction of potential fishing ground for Pacific saury(Cololabis saira)based on MAXENT model

为了研究海洋环境因子对西北太平洋秋刀鱼渔场的影响机制,利用2013—2014年我国大陆秋刀鱼渔船的渔业统计资料和美国NOAA网站的海洋环境数据建立MAXENT模型,使用受试者工作特征曲线(ROC)验证MAXENT模型的精度,用刀切法(Jackknife)评价海洋环境因子的重要性,Arcgis软件制作秋刀鱼潜在渔场分布图。结果表明:MAXENT模型的预测精度较高,结果可靠。秋刀鱼盛渔期8—10月的主要渔场集中在北海道和千岛群岛东部附近海域,海表温(SST)是影响秋刀鱼渔场分布的最重要因子,对模型的平均贡献百分比为64.3%,叶绿素(Chl.a)的平均贡献百分比为20.6%,盐度(SSS)的平均贡献百分比为15.2%,8—9月叶绿素(Chl.a)比盐度(SSS)对秋刀鱼渔场分布的影响高,10月叶绿素(Chl.a)对秋刀鱼渔场分布的影响比盐度(SSS)低。

To study the influence mechanism of marine environmental factors on the potential fishing ground for Pacific saury in Northwestern Pacific,by using the fishery statistics data of Chinese Pacific saury fishing boats from 2013 to 2014 and the marine environment data of NOAA website,this paper set up MAXENT models:the accuracy of the MAXENT model was verified by using receiver operating characteristic(ROC)curve,the importance of oceanographic environmental factors was evaluated by Jackknife Method,the potential fishing area distribution map of the Pacific saury was made by Arcgis.The results show that MAXENT model was proved to be accurate and reliable:during the Pacific saury’s best fishing period from August to October,the potential fishing ground mostly occurred around the eastern waters of Hokkaido and Kuril islands.The fishing ground distribution for Pacific saury is influenced by SST mostly among several environmental factors,of which the average percentage contribution to the model is 64.3%,in the meantime,the average contribution percentage of chlorophyll(Chl.a)and slinity(SSS)are 20.6%,15.2%respectively.In August and September,chlorophyll(Chl.a)holds a higher model contribution to the distribution of the Pacific saury than salinity(SSS),whereas,the effect of chlorophyll(Chl.a)on the distribution of Pacific saury in October was lower than salinity(SSS).