中亚热带赤皮青冈天然种群表型多样性分析

赤皮青冈(Quercus gilva)是我国中亚热带地区极具经济价值的珍贵用材树种。为揭示其表型多样性与变异规律以及影响表型变异的关键地理气候因子，该研究以14个赤皮青冈天然种群115个单株为材料，对其15个树体和叶片性状进行测量，并采用巢式方差分析、表型分化系数分析、多样性指数分析、相关性分析、主成分分析以及聚类分析等方法，探究赤皮青冈的表型变异规律及其与地理气候因子的相关性。结果表明：(1)赤皮青冈表型多样性水平较高，15个表型性状的变异系数和Shannon-Wiener指数平均值分别为35.070%和1.998,14个种群的变异系数在14.94%(洞口)～35.56%(龙山)之间、Shannon-Wiener指数在1.127(松阳)～1.980(常宁)之间。(2)15个表型性状在种群间和种群内均存在极显著差异(P<0.01),种群平均表型分化系数为41.491%,表型变异主要来源于种群内。(3)相关性分析发现，赤皮青冈部分性状间存在显著或极显著相关性，表型性状与地理因子间相关性不明显，降水是影响赤皮青冈表型性状的主要气候因子。(4)主成分分析发现，基于11个表型可提炼出4个...

Phenotypic diversity of Quercus gilva natural populations in middle subtropical China

Quercus gilva is a precious timber tree species in middle subtropical China. In order to investigate the phenotypic diversity, phenotypic variation pattern and the major geographic and climatic factors influencing phenotypic variation of Q. gilva populations, 15 growth and leaf traits of 115 individuals from 14 natural populations were measured, and then the phenotypic diversity, variation pattern and the correlation between phenotypic variation and geographic and climatic factors were studied using nested analysis of variance, phenotypic differentiation coefficient, diversity index, correlation analysis, principal component analysis and cluster analysis. The results were as follows:(1)The average variation coefficient and the average Shannon-Wiener index of 15 phenotypic traits were 35.070% and 1.998, respectively. Variation coefficient and Shannon-Wiener index of 14 populations ranged from 14.94%(Dongkou)to 35.56%(Longshan)and from 1.127(Songyang)to 1.980(Changning), respectively. These results indicated that Q. gilva showed a relatively high level of phenotypic diversity.(2)Significant differences in 15 phenotypic traits were found among and within populations(P<0.01), with an average phenotypic differentiation coefficient of 41.491%. This result elucidated that the phenotypic variation within populations was greater than that among populations.(3)Correlation analysis showed that there were significant or entremely significant correlations among parts of the traits, whereas no correlation was found between phenotypic traits and geographic factors. Precipitation was the major factor that influenced the phenotypic traits of Q. gilva.(4)Principle component analysis indicated that the first four principal components explained 82.961% of total phenotypic contribution rate of Q. gilva.(5)Cluster analysis revealed that 14 Q. gilva populations were divided into three groups according to their different leaf and petiole size, such as small-leaved micropetiole, large-leaved macropetiole and medium-leaved micropetiole, which was not related to the geographic distribution of populations. This study provides scientific basis for the protection and utilization of Q. gilva germplasm resources, and lays an important foundation for its improved variety breeding.