<正>首次揭示了群体遗传分化度量ＦＳＴ与典型相关分析有一定内在联系，进而指出了以往广泛采用的ＦＳＴ方法在区分遗传群体时的弱点，从理论上证实了多元分析方法是研究群体遗传变异分布的有效方法。运用典型相关分析，对杉木分布区内１６个群体的２２个同功酶基因位点数据进行了分析，给出了杉木群体较为详细的遗传变异分布模式，从而揭示了杉木群体间也存在较高程度的遗传分化。

The relationship between FST and canonical correlation analysis was revealed in this paper.The weakness of FST as a measurement of genetic differentiation among subpopulations was also indicated. The theoretical results demonstrate that multivariate statistical analysis is an effective method to study the distribution of population genetic variation. By using Canonical correlation analysis, we analyzed the data of 22 allozyme loci from 16 populations of Cunnlinghamia lanceoslata (Lamb.) Hook in People’s Republic of China.A detailed distribution pattern of gnentic variation was given, and greater differentiation among these populations was revealed.