单 位：
中国水产科学研究院东海水产研究所农业部海洋与河口渔业重点开放实验室
关键词：
刺鲳;底拖网调查;平均资源密度;调查设计法;Δ-分布
摘 要：
海洋生物的分布具有不均匀性,从而造成底拖网调查数据显著偏斜。对于该类型调查数据,国外学者提出了统计模型法加以处理;而在中国,通常采用调查设计法加以处理。本研究利用2005年东海区渔业资源底拖网调查资料,分析了刺鲳的资源密度空间分布特征,引用Δ-分布模型法估算了刺鲳的平均资源密度,并与调查设计法估算结果进行比较分析。结果表明,东海区刺鲳各季节的资源密度空间分布具有明显的右偏性,非零值资源密度的自然对数转换服从正态分布;利用Δ-分布模型法估算的春、夏、秋、冬季的平均资源密度均高于调查设计法的估算,依次高出30.8%、33.3%、67.7%、14.0%;通过本研究发现,调查设计法估算的资源密度均值偏低,且调查数据不服从正态分布,而使用Δ-分布模型法转换后服从正态分布,符合估算资源密度均值和置信区间的条件,可作为我国今后渔业资源调查数据标准化的模型。
译 名：
Comparison on difference of the stock density of Psenopsis anomala in the East China Sea by means of different estimating methods
作 者：
YUAN Xing-wei,JIANG Ya-zhou,YAN Li-ping(East China Sea Fisheries Research Institute,Chinese Academy of Fishery Sciences,Key and Open Laboratory of Marine and Estuarine Fisheries,Ministry of Agriculture,Shanghai 200090,China)
关键词：
Psenopsis anomala;bottom trawl survey;the mean of stock density;design-based methods;Δ-distribution
摘 要：
Since the spatial distributions of marine biota are patchy,data collected from bottom trawlers exhibit a skewed distribution with many zero catches and some very large catches.A lot of model-based methods abroad have been developed to estimate the mean of this distribution.However these model-based methods were rarely used in China.Based on the data collected from bottom trawl survey conducted in the East China Sea in 2005,present paper dealed with the spatial distribution characteristics of Psenopsis anomala,and estimated the mean of stock density used with two methods,namely the Δ-distribution and the design-based method.It could be found that the survey data were highly skewed to the right,but the nonzero values were lognormally distributed.The estimator of the mean of stock density in spring,summer,autumn and winter based on Δ-distribution were 30.8%,33.3%,67.7%,14.0%,respectively,higher than those based on the design-based method.The paper represented that the method based on Δ-distribution was more efficient in estimating the mean of survey data from bottom trawlers as compared with the traditional design-based method.There were two reasons to support the conclusion,on the one hand,the original data did not follow the normal distribution;and the mean of stock density of Psenopsis anomala was underestimated by using the traditional design-based method,and on the other hand,the nonzero values were lognormally distributed,which met the standards for estimating the mean and confidence intervals.So Δ-distribution could be suggested to be widely used in the further study.
