We conducted leaf geometric morphometric analysis (GMMs) for five Quercus species (Section Quercus) of Fagaceae. In total, 887 leaves chosen from 182 individuals of 20 natural popu-lations were marked with GMMs. Leaf morphological characteristics of these samples were digitized to visualize leaf morphological differences. Generalized Procrustes analysis could effectively exclude the influence of leaf position and size on leaf shape. Results of principal component analysis at tree-level showed that the leaf morphology of Q. dentata was different with Q. aliena and Q. serrata. Canonical variates analysis at tree-level showed that leaf morphology of Q. aliena could be accurately distinguished from the other four species in leaf symmetric components. The results of multivariate analysis of asymmetrical components in leaves showed no distinction among the five species. The analysis at leaf-level showed that the two groups with a higher degree of discrimination were Q. aliena vs. Q. dentata (99.5% vs. 100%) and Q. dentata vs. Q. serrata (99.0% vs. 100%), which could be accurately distinguished by leaf shape. The two groups with a slightly lower degree of discrimination were Q. fabri vs. Q. serrata (90.5% vs. 86.8%) and Q. dentata vs. Quercus mongolica (85.1% vs. 82.9%). Our results provided new insights for the leaf shape identification among species with frequent hybridization and introgression.
以壳斗科栎属白栎组的5个种作为研究对象进行叶片几何形态学分析(GMMs)。选取了20个天然种群,共采集182个个体,887张叶片,对每张叶片采用几何形态学的标点法,数字化叶形态特征,使5个种的叶形态差异可视化。结果表明: 广义普氏分析可以有效地排除叶片位置和大小对叶片形状的影响。基于个体水平的主成分分析表明,槲树分别与槲栎和枹栎2个种的叶片形态差异明显,基于个体水平的典型变量分析表明,叶片的对称组分可以将槲栎与其他4个种准确区分。叶片非对称组分的多变量分析表明,5个种彼此间区别不明显。基于叶片水平的分析表明: 区分度较高的两组是槲栎vs.槲树(99.5% vs. 100%)和槲树vs.枹栎(99.0% vs. 100%),可以通过叶形准确区分;区分度略低的两组是白栎vs.枹栎(90.5% vs. 86.8%)和槲树vs.蒙古栎(85.1% vs. 82.9%)。本研究为存在频繁杂交渐渗的物种间叶型鉴定提供了新思路。.
Keywords:
Fagaceae; geometric morphometric analysis; leaf morphological characteristics; section