论文部分内容阅读
基于2008—2010年黄海南部近海(SYS)、东海北部外海(NECS)和东海中部近海(MECS)小黄鱼体长和体质量数据,采用均值回归和分位数回归模型,解析了小黄鱼幼鱼和成鱼群体体长-体质量关系的空间变异.结果表明:协方差模型和线性混合模型的残差标准误基本一致,线性模型残差标准误最高.从线性混合模型对特定区域和总体区域平均体质量计算的相对比值来看,SYS和NECS幼鱼群体的平均体质量高于总体平均值,但MECS低于总体平均值;成鱼群体则为NECS平均体质量高于总体平均值,MECS和SYS低于总体平均值.分位回归估计的肥满度和异速生长指数结果显示,幼鱼群体在不同分位的估计参数呈显著变化,SYS异速生长指数均值为2.85,在0.1~0.95分位的估计值变化范围为2.63~2.96.MECS和NECS参数估计值和置信区间在各分位数呈异质性变化,低分位时,NECS估计值在3个调查区域中最低,MECS最高;高分位时,MECS和NECS均高于SYS.对低分位0.25、中分位0.5和高分位0.75分位数的异速体长体质量关系的方差分析结果显示,低分位和高分位数之间体长-体质量关系极为显著(0.25∶0.75,F=6.38,df=1737,P<0.01),低分位数和中分位数之间为显著(0.25∶0.5,F=2.35,df=1737,P=0.039),中分位数和高分位数之间接近显著(0.5∶0.75,F=2.21,df=1737,P=0.051).成鱼群体SYS异速生长指数均值为3.01,在0.1~0.95分位的估计值变化范围为2.77~3.10.低分位和高分位数之间体长-体质量关系差异达到显著水平(0.25∶0.75,F=3.31,df=2793,P=0.01),低分位和中分位之间差异不显著(0.25∶0.5,F=0.98,df=2793,P=0.43),而高分位和中分位之间则差异极显著(0.5∶0.75,F=3.56,df=2793,P<0.01).
Based on the data of body length and body mass of small yellow croaker (SYS), NECS and MECS of the Yellow Sea from 2008 to 2010, the mean and quantile regression models were used to analyze the effects of The results showed that the standard error of the residual error of the covariance model and the linear mixed model were basically the same, and the standard error of the residual error of the linear model was the highest.Using the linear mixed model, the spatial heterogeneity of the specific region and the total regional average body The average body mass of SYS and NECS juveniles was higher than the overall mean, but MECS was lower than the overall mean. For adult fish, the average body mass of NECS was higher than the overall mean. The average body mass of MECS and SYS Lower than the overall average.Fractional regression estimated fatness and allometric growth index results showed that the estimated parameters of juvenile populations in different quantiles showed significant changes, allometric mean growth rate of SYS 2.85, at 0.1 ~ 0.95 points The estimated values of bits range from 2.63 to 2.96.The estimates and confidence intervals of the parameters of the MECS and NECS show heterogeneity among the quantiles. In the low quantile, the NECS estimates are the lowest among the three surveyed areas and the highest is the MECS. High score , MECS and NECS were higher than that of SYS.Analysis of ANOVA on allograft body mass ratios of 0.25, 0.5 and 0.75 of high school showed that low and high scores (0.25: 0.5, F = 2.35, F = 6.38, df = 1737, P <0.01) between body weight and body weight. df = 1737, P = 0.039), and the mean interquartile range and high quantile were significant (0.5:0.75, F = 2.21, df = 1737, P = 3.01, with a range of 0.77 to 3.10 in the range of 0.1 to 0.95.The differences of body length and body mass between low and high quantiles reached significant levels (0.25: 0.75, F = 3.31, df = 2793 , P = 0.01). There was no significant difference between low and high quintiles (0.25: 0.5, F = 0.98, df = 2793, P = 0.43) (0.5: 0.75, F = 3.56, df = 2793, P <0.01).