中文摘要：

利用cpSSR和AFLP标记,比较了巴山松、黄山松、油松、马尾松以及云南松的遗传多样性与遗传分化结果.发现5种松树都表现出较高水平的遗传多样性与种间分化,但cpSSR比AFLP标记提供了更多的种内种间分化,而展示的遗传多样性相对要少.在对不同方法指标的对比研究后,认为将cpSSR的单倍体关系考虑在内,能够更有效地分析植物居群遗传分化造成的地理结构.对于显性标记AFLP,传统的平方根法会低估遗传参数,而Lynch和Milligen的方法因样本量的限制也可能高估遗传多样性,Bayesian法被认为是最精确的方法,提供了与分子方差分析（AMOVA）近似的遗传分化值.

英文摘要：

Genetic diversity and genetic differentiation of five Eurasian pines, Pinus henryi, P. taiwanensis, P. tabulaeformis, P. rnassoniana, and P. yunnanensis, were compared using amplified fragment length polymorphism （AFLP） and chloroplast simple-sequence repeat （cpSSR）. High genetic differentiation and median genetic diversity with cpSSR markers were found both at population and species level, while median differentiation and higher genetic diversity with AFLP data. Although the infinite allele model were preferred at most chloroplast gene loci,other model should not be rejected according to the similar support rates. Measures of subdivision that consider similarity between haplotypes offered better information of plant geographic structure than that of the conventional ones. Among several methods analyzed in AFLP, Sequare root method provided downwardly biased estimates for the genetic parameters, while Lynch and Milligen method normally over-estimated genetic diversity due to limited samples. Thus Bayesian statistics was the most accurate and popular method in those data with the dominant markers and its value of species differentiation （θ^B=0. 103 5） was close to the parameter given by AMOVA.