基于线性混合模型的蓝点马鲛叉长和体重关系的月间及性别差异

为了研究蓝点马鲛（）生活史特征的异质性，根据2018年10月至2019年3月在东海外海渔场的拖网调查采样数据，对其叉长和体重关系的月间及性别差异进行了研究。依据收集的367尾蓝点马鲛样本，求得其叉长和体重关系（b的估计均值为2.794。本研究构建了广义线性模型和9个线性混合模型，用于研究蓝点马鲛的叉长和体重关系（）在时间及性别上的差异。贝叶斯信息准则（BIC）值和均方根误差值均表明，最复杂的线性混合模型（即月份和性别对两个参数P<0.01）。在最优模型中，a值则与此相反。本研究表明，月份和性别对蓝点马鲛叉长和体重关系具有显著的影响，线性混合模型能把月份和性别的异质性通过随机效应在单个模型中更准确、快速地体现，从而进一步证实了此模型在数据来源异质性研究中的优势。

Monthly and sexual differences in the growth of Japanese spanish mackerel (Scomberomorus niphonius) based on a linear mixed model

is an important economic fish species from coastal China. It is also one of the few large-scale catch species that maintain a high yield in China. Studies have shown that the relationship between fork length and body weight differs between seasons and areas. To better understand the seasonal and sexual variations in relationships between the fork length and weight (, we used the trawl survey data at the East China Sea fishing grounds from October 2018 to March 2019 and constructed a generalized linear model and nine linear mixed models. Based on the collected 367 samples, the relationship between fork length and body weight () is obtained, in which the estimated mean value of condition factor a is 3.022×10^-5, and the estimated mean value of allometric growth parameter b is 2.794. Both the Bayesian Information Criterion (BIC) value and the root mean square error (RMSE) value indicated that the most complex linear mixed model (i.e. month and gender have random effects on both parameters ) had the best performance. According to the results of the difference analysis between the best model and the generalized linear model, the relationship between the fork length and weight of <0.01). In the best model, was higher in October and November (feeding period) and lower in December and January (overwintering period), while was higher in males and lower in females. It can be concluded that the relationship between the fork length and weight of was significantly affected by season and sex. The linear mixed model allows the easier combination of season and sex, which would be of great value in the heterogeneity studies of fish life history traits.