摘要
CPUE标准化是渔业资源评估的重要研究内容。不同名义CPUE计算方法通常会影响到CPUE的标准化,进而影响到对资源丰度的评价。在CPUE标准化过程中,首先对名义CPUE进行计算,通常有以下3种方法:(1)直接将每个记录的产量除以其对应的捕捞努力量计算出CPUE值,然后输入模型;(2)先对所有的记录计算出CPUE,然后以一个空间尺度(小渔区,如0.5°×0.5°)统计数据,对每个小渔区内所有的CPUE直接求平均值(所有CPUE求和再除以记录的数量)输入模型;(3)以一个空间尺度统计数据,用小渔区内的所有记录的总产量除以总捕捞努力量计算的CPUE输入模型。本文以我国在西南大西洋的阿根廷滑柔鱼鱿钓渔业为例,分别使用上述3种CPUE计算方法,运用广义加性模型(Generalized Additive Model,GAM)模型进行CPUE标准化(对应的模型分别为GAMa、GAMb和GAMc),比较不同模型的结果。研究发现,3种模型得出的年标准化CPUE和月标准化CPUE存在差别,GAMa与GAMb、GAMc的结果相差较大,GAMb和GAMc所得的结果相差较小,而GAMa与GAMb、GAMc的标准化CPUE值、变异系数以及模型的各因子方差贡献率等相差较大。3种模型之间的差别主要由于样本数、捕捞努力量的假设、数据记录的时空尺度和模型中因子的选择等因素影响。因此在使用商业性渔业数据分析渔业资源状况时,需要考虑由于CPUE计算方法的不同带来的不确定性。
CPUE standardization is a role in fisheries stock assessment.Different calculating methods for nominal CPUE can affect standardized CPUE which usually is used to index fisheries abundance.In CPUE standardization modeling,nominal CPUE is as a response variable which usually can be calculated in three methods：（1） for every record,CPUE is calculated by catch dividing the corresponding effort,and the all CPUEs were input the model;（2） all fisheries data records are first grouped by one spatial scale（fishing grid,such as 0.5°×0.5°）,then for every grid,average CPUE is calculated by all CPUE dividing the number of records,all average CPUEs of fishing grids are as the model input;（3） the fisheries data grouping and the model input are same as（2）,but the average CPUE for every grid is calculated by total catch dividing total fishing effort.For evaluating impacts of different nominal CPUE inputs on CPUE standardization modeling,Chinese Illex argentinus fishery in the South Atlantic Ocean was,as a study case,the three above-mentioned nominal CPUEs calculated from this fisheries data were input generalized additive models which were used to standardize these CPUEs.The corresponding models were GAMa,GAMb and GAMc,respectively.The results derived from the GAMs were compared.From these analysis,there were differences among these standardized CPUEs derived from the three GAMs.GAMb and GAMc showed similar trends in standardized CPUEs.But there were significant variances between GAMa and the latter two GAMs in the values of standardized CPUE,coefficient of variance and the variance contribution rate of each explanatory variable in the models.The differences among these GAMs were mainly owing to sample numbers,the assumption for calculating fishing effort,the spatio-temporal scale for grouping fisheries data and model selection,etc.On the summary,the certainties due to different calculating methods for CPUE should be considered when commercial fisheries data were used to analyze the status of fisheries stock.
出处
《上海海洋大学学报》
CAS
CSCD
北大核心
2010年第2期240-245,共6页
Journal of Shanghai Ocean University
基金
国家科技"八六三"计划(2007AA092202)
国家科技"八六三"计划(2007AA092201)
国家科技支撑计划(2006BAD09A05)
上海市捕捞学重点学科(S30702)
关键词
CPUE标准化
广义加性模型
西南大西洋
阿根廷滑柔鱼
CPUE standardization
generalized additive model
Southwest Atlantic Ocean
Illex Argentinus fishery
