摘要
茎柔鱼是世界上重要的经济头足类,也是我国大陆鱿钓渔业最重要的捕捞对象之一。根据中国大陆鱿钓船2003-2004年、2006-2012年9年的渔业生产统计数据和卫星遥感获得的海洋环境数据,利用广义线性模型(generalized linear model,GLM)和广义加性模型(generalized additive model,GAM)对秘鲁外海茎柔鱼的资源丰度进行CPUE标准化,分析秘鲁外海茎柔鱼资源丰度年际变化。经显著性检验,得到最终选择加入模型的自变量有年、月、纬度、经度、叶绿素浓度、年与经度交互项、年与纬度交互项、月与经度交互项、月与纬度交互项共9个自变量。GAM结果表明,模型自变量对因变量的决定系数高达42.3%。标准化后CPUE与名义CPUE变化趋势相同,年平均值略低于名义CPUE,2003-2012年秘鲁外海茎柔鱼资源丰度年间变化较大,资源丰度最高的年份为2004年,GLM标准化后的平均CPUE为7.94 t/d;资源丰度最低的年份为2007年,GLM标准化后的平均CPUE为3.28 t/d。
Dosidicus gigas is one kind of important economic fish in the world and also one of the most important squid jigging species of Chinese Mainland. According to the fishery-dependent data of 2003 - 2004 and 2006 -2012 by squid jigging boats of Chinses Mainland and the environment data from satellite remote sense, we standardised the CPUE of Dosidicus gigas off Peru by generalized linear model (GLM) and generalized additive model ( GAM ). We analysed the inter-annual variation of Dosidicus gigas off Peru. By significant testing, the 9 variables chosen into the model are year, month, latitude, longitude, concentration of chlorophyll-a, interaction of year and longitude, interaction of year and latitude, interaction of month and longitude and interaction of month and latitude. The results of GAM indicate that coefficient of determination (R2) is 42.3% which is high. The standardized CPUE has the same trends with nominal CPUE and the average value is a little lower than nominal CPUE. The inter-annual variation of Dosidicus gigas abundance off Peru during 2003 - 2012 is severe, the highest value of abundance appeared in 2004, the average CPUE after standardizing by GLM was 7.94t/d ; the lowest value of abundance appeared in 2007, the average CPUE after standardizing by GLM was3.28t/d.
出处
《上海海洋大学学报》
CAS
CSCD
北大核心
2015年第2期280-286,共7页
Journal of Shanghai Ocean University
基金
国家高技术研究发展计划(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)