12个日本落叶松家系扦插苗造林后的高生长表现及其稳定性分析

Behaviors of height growth of Larix kamepferi cuttings of 12 families after afforestation and its analysis of stability

文章分析了 12个日本落叶松家系扦插苗造林后的高生长表现、遗传参数和稳定性参数。得出 :在经过一次选择后 ,优良家系 (超过日本落叶松丰产林标准 15 % )只占全部家系的 1/3 ,并且仍有少量的家系低于此标准 ,可能的原因是基因型×环境 (年份 )互作的敏感性 ;连续测定期间 ,树高总生长量的广义遗传力变化不大 (74 5%～78.0 % ) ,说明高生长受中上等遗传效应的控制 ;但遗传变异系数和表型变异系数均随年龄增长而下降 ,显然再行选择需加大选择强度 ;然而 ,变异系数下降有利于为生产提供品质均一的无性系材料。稳定性分析表明 ,家系×年龄的互作是明显的 ,即高生长性状是对环境敏感的性状 ;多数家系能够利用模型进行回归预测 ;如欲在树木发育的幼龄期进行选择 ,当缺少多点试验数据时 ,只能通过连续测定 ,对稳定性好的家系作出可靠评价。文中还对稳定性方法在早期测定中的作用进行了探讨。

In this paper, behaviors of height growth, genetic and stability parameters of Larix kaempferi cuttings of 12 families after afforestation were analyzed. The results showed that good families (15% above the standard of Larix kaempferi fast growing high yield plantations)constituted only 1/3 of the total; Furthermore, there were a few families which were lower than the standard, the reason might be sensibility of interaction between gene types and environment (years). The generalized heritability of total growth of height changed little (among 74.5%～78.0%)during continuous determination, that indicated height growth was under the control of superior genetic effect; but both genetic variation coefficient and phenotype coefficient decreased with the increasing of the age, it is obvious that selection intensity need strengthening when selection was carried out once more; however, the decreasing of variation coefficient was beneficial to provide with clone materials of even quality for production. The results of analysis of stability showed that interaction between families and ages was remarkable, namely, the characters of height growth were sensible to environment; most families can make regression prediction by using models; if selection was implemented in young stage during tree development and experimental data of many places were lacked, the reliable evaluation on stable families could be worked out only through continuous determination. The effect of stability method in early stage determination was also studied in this paper.