【目的】揭示木荷(Schima superba)多地点家系变异及环境互作规律,为木荷良种选育提供参考。【方法】以2019年在浙江龙泉、建德、开化和江西分宜营建的木荷优树家系林为材料,调查树高、胸径、冠幅、枝下高、最大分枝粗等的遗传变异规律及立地效应,估算所试材料的现实增益及遗传增益等。【结果】多点4年生木荷优树的树高、胸径、冠幅、最大分枝粗、最大分枝角等具有极显著的地点、家系和家系×地点的互作效应,家系生长受到遗传和环境因素共同影响。不同种植地点家系生长性状差异较大,江西分宜点的胸径和分枝性状的变异系数相对较高,其次为浙江建德和浙江开化,最低为浙江龙泉点。木荷的树高与胸径呈显著中至强的正相关(0.66< R<0.93,P<0.01),与冠幅和分枝粗呈显著中至强的正相关(0.60 < R<0.87, P<0.05),与分枝角相关性不显著。各地点的树高、胸径、冠幅等生长性状均受到中度以上的遗传力(0.50~0.78)控制。4个地点的胸径和树高性状均值与其育种值均呈现良好的线性关系(R > 0.90, P<0.001)。通过对木荷树高和胸径的BLUP-GGE分析,发现胸径的BLUP-GGE较为可靠,其中浙江建德点最具有区分力,浙江开化点最具有代表性。根据速生性和稳定性初步选出7个生长优良的家系,遗传增益(2.30%~12.12%,平均7.37%)和现实增益(3.48%~18.30%,平均11.12%)较高。【结论】木荷生长性状受家系以及种植点环境影响,以胸径进行遗传分析和品种筛选较为可靠,综合多地点木荷生长表现初步选出7个优良家系,分别来自福建建瓯和浙江龙泉产区。

【Objective】In order to treveal the genetic variation of Schima superba families in multiple sites and their interaction with the environment, we performed the forest survey, the tree growth and branching performance. The results would provide a theoretical basis for selecting S. superba varieties with strong adaptability, stability, and fast-growing among multiple sites. At the same time, these super varieties could be applied in multiple locations. 【Method】A systematic investigation was conducted on the growth and branching traits of S. superba family forests built in Longquan, Jiande, Kaihua Cities (County) in Zhejiang Province, and Fenyi County in Jiangxi Province in 2019, including tree height (H), breast height diameter (DBH), crown width (C), maximum branch diameter (MBD), and maximum branch angle (MBA). Using a combination of GGE (genotype main effects and genotype-by-enviroment interactions) and BLUP (best linear unbiased prediction), this study revealed genetic differences among families of S. superba, evaluated the effects of environmental factors on growth and branching traits, explored the adaptability, stability, and rapid-growth among these families in multiple sites, and estimated the genetic and real-gains of these tested materials. 【Result】The H, DBH, C, MBD and MBA of the 4-year-old S. superba families in Longquan, Jiande, Kaihua in Zhejiang Province and Fenyi in Jiangxi Province showed highly significant interactions among sites, families, and family × sites. The growth and branching traits of these families were influenced by both genetic and environmental factors. There were significant differences in growth traits among families from different sites. For example, the coefficient of variation in DBH and branching traits were relatively high among families in Fenyi, followed by Jiande and Kaihua, and relatively low in Longquan. There was a significant medium to strong positive correlation between H and DBH among families (0.66<R<0.93, P<0.01). At the same time, a significant medium to strong positive correlation showed between H and C and MBD (0.60<R<0.87, P<0.05), but no significant correlation found between H and MBA. The growth and branching traits of families were controlled by moderate to high heritability across different sites, with a heritability range from 0.50 to 0.78. That was indicating that genetic factors play an important role in the growth of S. superba. A linear regression model was established between the mean value of DBH and H of families from four locations and their breeding values (BLUP). The results showed a good linear relationship (R>0.90, P<0.001). The BLUP-GGE combined method was used to evaluate the H and DBH of S. superba, and the results showed that the BLUP-GGE model for DBH was more reliable. Among them, the families from the Jiande showed strong differentiation, followed by Longquan. While the families from Longquan had the most representativeness, followed by Kaihua and Jiande. Based on the fast-growing and stability results, we selected 7 excellent growth performing families, whose genetic gains ranged from 2.30% to 12.12% (average 7.37%), and the real-gains ranged from 3.48% to 18.30% (average 11.12%). These families demonstrated a high breeding potential and application value. 【Conclusion】The growth and branching traits of 4-year-old multi-site S. superba families were influenced by the family and planting environment. It was more reliable to use DBH as a key indicator in genetic analysis and selection. Based on the genetic gain performance, we had selected 7 superior families with excellent growth and stable traits.