Abstract: The height-growth and genetic differentiation of Bretschneidera sinensis seedlings was studied. The results indicated that colony growth distribution of B. sinensis was behaving of the Gaussian distribution from 13 to 92 cm. Using the standard deviation classification method of normal distribution, 8 types of height-growth of B. sinensis seedlings were classified with the interval value of colony standard deviation. The results indicated that 85% B. sinensis were distributed in the growth types of mean and adjacent interval. The ground diameter growth of genetic differentiation of B. sinensis was the same as the height growth. With the height growth and ground diameter growth of one-year-old seedling as selection-character, 4 elite treelets were selected according to selection-stands of t determination principle, another 5 elite treelets were selected by the comprehensive evaluation through using the genetic differentiation of seedling at the early stage. These 9 elite treelets were named Tianzhong No. 1—9 respectively. Comparing with average individual, the height growth of 9 elite treelets increased 95%, 97%, 105%, 105%, 142%, 113%, 111%, 100%, 118% respectively, and the ground diameter growth of those elite treelets increased 57%, 61%, 63%, 75%, 105%, 63%, 64%, 57%, 79% respectively. The 9 elite treelets had good growth quality, which would be selected as valuable gene resources of B. sinensis.

Key words: Bretschneidera sinensis, t determination principle, genetic differentiation, early stage selection, normal distribution standard deviation classification method

摘要： 对钟萼木Bretschneidera sinensis一年生苗高生长适应性与早期生长分化进行研究。结果表明，苗木群体高生长类型权重呈正态分布于13～92 cm区间；应用正态分布标准差分类法，以苗木群体高生长标准差值作为不同生长类型划分间隔值，苗木高生长划分为8个生长类型，85%以上个体的高生长量处于平均值与相邻两生长类型，地径生长习性遗传分化表现出与苗高相似特性。利用钟萼木苗木早期生长分化显著的特点，以苗木高和地径作为选择性状，依据t检验选择法确定选择标准，共筛选双重指标都能满足的个体4株，经综合评价，另选择5株优良个体，共计9株，分别命名为田钟1～9号，田钟1～9号高生长比平均值分别提高95%、97%、105%、105%、142%、113%、111%、100%、118%；地径生长比平均值分别提高57%、61%、63%、75%、105%、63%、64%、57%、79%。生长量表现较好的优良个体可作为该树种优良基因种质资源。

关键词: 钟萼木, t检验选择法, 遗传分化, 早期选择, 正态分布标准差分类法