【目的】 利用立地条件相对均一的圃地建立短期测定林,通过分析67个参试杉木无性系苗期生长性状的遗传变异情况及遗传-环境互作效应对各性状选择的影响程度,探讨无性系苗期超早期选择策略,进一步对大量候选无性系开展快速初筛和超早期选择,以降低长期测定成本,提高无性系选育效率。【方法】 利用杉木第3代种子园子代实生群体选择优良单株,扦插繁育成无性系,于圃地做性状短期测定。参试无性系67个,重复10次,12株小区,完全随机区组设计。造林1 a后,测定苗高、地径、侧枝数和最长侧枝长度共4个生长性状指标,通过构建表型方差分析模型,估算遗传方差分量,以及遗传与环境互作效应方差分量的值,并利用ASReml软件分别估算遗传力和重复力。【结果】 在圃地栽植1 a后,参试无性系的苗高、地径、侧枝数和最长侧枝长度均值分别为0.640 m、1.010 cm、10.30条和0.28 m,4个观察性状的表型变异系数分别为12.86%、14.88%、21.34%和14.89%;参试67个无性系在苗高、地径、侧枝数和最长侧枝长度性状上存在显著遗传差异,测定性状的重复力均达0.74左右,遗传力估值均稳定在0.48左右;所测4个性状的遗传与环境互作方差分量占总遗传方差的35%左右;地径与苗高、侧枝数和最长侧枝长度存在显著相关,遗传相关系数均达0.9以上;以地径性状为指标进行选择,苗高、地径、侧枝数和最长侧枝长度的遗传增益估值随着入选率的降低逐渐增高,但苗高、侧枝数和最长侧枝长度重复力、遗传力以及遗传-环境互作方差比例均保持在较为稳定的范围内,呈现出一定程度的波状变化;而地径则随着入选率的降低,重复力和遗传力估值下降,遗传-环境互作方差比例增大。当入选率降低到40%以下时,杉木无性系的苗高、侧枝数和最长侧枝长度3个性状的遗传-环境互作方差比例分别达到41.18%~48.61%、37.82%~40.13%和39.61%~54.37%,但地径的遗传-环境互作方差比例由45.91%快速上升至94.33%,当入选无性系数量由19个降至16个时,地径遗传力和遗传-环境互作方差比例发生显著变化,地径遗传力由0.226 3降至0.091 4,遗传-环境互作方差比例由63.09%迅速增加到83.26%;取30%左右入选率,筛选出19个无性系用于山地造林长期测定,初选材料的苗高、地径、侧枝数和最长侧枝长度的均值分别为0.73 m、1.20 cm、12.4条和0.33 m,遗传增益均值分别为10.81%、15.45%、16.66%和13.88%,分别比群体均值高出14.06%、18.81%、20.39%和17.86%。【结论】 遗传-环境互作效应对杉木无性系表型性状的影响不可忽视,其互作方差在总遗传方差中具有较大的占比;杉木无性系苗高生长和侧枝生长受遗传与环境互作作用影响相对较小,地径可能对圃地微环境变化等因素更为敏感,因而将苗高和地径性状综合起来进行杉木无性系超早期选择能够取得较为理想的结果;降低入选率并不能剔除遗传-环境互作效应对地径和最长侧枝长度的影响,高强度的选择反而会增加遗传-环境互作的影响,但适当的入选强度既能保留无性系间目标性状遗传变异的丰富度,又能固定大部分的遗传-环境互作效应;短期圃地测定,能对大量待测杉木无性系进行快速初筛,缩小长期测定林面积,降低测定成本;对参试无性系性状遗传与环境互作效应特征进行早期解析,可为充分利用遗传与环境的有益互作效应提供重要依据。

【Objective】 The efficiency of selection and long-term testing costs for clonal propagation candidates of Cunninghamia lanceolata were improved by implementing a short-term nursery test with 67 clonal propagation candidates. By analyzing the genetic variation of growth traits and the impact of genetic environmental interactions on the selection of various traits of clones during the seedling stage, this study explores strategies for ultra early selection of clone seedlings. 【Method】 A selection procedure was conducted from a population of two million seedlings, with 275 well performing individuals selected for further prorogation. The seeds were collected from a local third generation C. lanceolata seed orchard.The selected plants were propagated into clones by hedged cutting. Of the propagated clones, 67 individuals with a fine rooting ability were selected for further testing under a completely random block design with 12 plants per plot and 10 replications. Four traits (seedling height, diameter above ground, number of branches and the length of the longest branch) were measured after one year’s growth. Furthermore, a phenotypic analysis of variance model was constructed to estimate the values of genetic variance component and genetic environmental interaction effect variance component, and ASReml software was used to estimate in heritance and repeatability, respectively. 【Result】 After planting in the nursery for one year, the average seedling height, ground diamete, number of lateral branches and longest lateral branch length of the tested clones were 0.640 m, 1.010 cm, 10.30 and 0.28 m, respectively. The phenotypic variation coefficients of the four observed traits were 12.86%, 14.88%, 21.34% and 14.89%, respectively. There were notable genetic differences found in the traits of seedling height, diameter above ground, number of lateral branches, and length of the longest lateral branches among the tested clones, and the repeatability of the measured traits exceeded 0.74, and the estimated heritability remained stable at around 0.48. The variance component of the genetic and environmental interaction accounted for about 35% of the total genetic variance. There is a significant correlation between ground diameter and seedling height, number of lateral branches, and length of the longest lateral branch, with genetic correlation coefficients above 0.9. The genetic gain estimates of seedling height, number of lateral branches, and longest lateral branch length gradually increase with the decrease of selection rate based on the ground diameter trait. However, the variance ratios of repeatability, heritability, and genetic environmental interaction of seedling height, number of lateral branches, and longest lateral branch length remain within a relatively stable range, exhibiting varying degrees of wavy fluctuations. As the selection rate decreases, the value of repeatability and heritability of ground diameter decrease, while the variance ratio of genetic environmental interaction increases. When the selection rate decreased to below 40%, the genetic environmental interaction variance ratios of the three traits of seedling height, number of lateral branches, and longest lateral branch length of C. lanceolata clones reached 41.18%-48.61%, 37.82%-40.13% and 39.61%-54.37%, respectively. However, the genetic environmental interaction variance ratio of diameter rapidly increased from 45.91% to 94.33%.When the number of selected clones decreased from 19 to 16, the genetic environmental interaction variance ratios of ground diameter heritability and genetic environmental interaction variance ratios changed significantly, with diameter heritability decreasing from 0.226 3 to 0.091 4 and genetic environmental interaction variance ratios rapidly increasing from 63.09% to 83.26%. Based on a selection rate of approximately 30%, 19 clones were selected for further evaluation in multiple sites in a long-term afforestation project in a mountain area. The average seedling height, ground diameter, number of lateral branches, and longest lateral branch length of the selected clones were 0.73 m, 1.20 cm, 12.4 branches and 0.33 m, respectively. The estimated average genetic gains of the four observed traits were 10.81%, 15.45%, 16.66% and 13.88%, which were 14.06%, 18.81%, 20.39% and 17.86% higher than the population average, respectively. 【Conclusion】 The effect of genetic environmental interaction on the phenotypic traits of C. lanceolata clones cannot be ignored, and its interaction variance accounts for a large proportion of the total genetic variance. The growth of height and lateral branches of C. lanceolata clones are relatively less affected by the genetic environmental interaction effect, while the growth of ground diameter are more sensitive to changes in the microenvironment of the nursery or from unknown factors. Therefore, combining the growth performance of tree height and ground diameter of C. lanceolata clones for short-term testing can achieve ideal of selection. Reducing the selection rate does not eliminate the influence of genetic environmental interaction on ground diameter and longest lateral branch length. High intensity selection can actually increase the influence of genetic environmental interaction. Appropriate selection intensity can not only retain the richness of genetic variation in target traits between clones, but also fix most of the genetic environmental interaction effects. Short-term nursery testing can serve as a rapid preliminary screening technique, especially when there is a large amount of clonal candidates to be tested. Several benefits were apparent, including forest-land use and the long-term cost efficiency of testing. The clonal traits, genetic components, and interaction between genetics and environment could be evaluated in the super-early stage of clonal evaluation.