The unique adaptation of Eucalyptus benthamii to low temperatures coupled to fast growth and versatile wood quality has made it a valued plantation species in frost-prone areas worldwide, but little is known on its quantitative genetic parameters for key industrial traits. We used GBLUP additive (GA), additive-dominant (GAD), single-step (HBLUP), and pedigree-based predictive models to estimate lignin, extractives, carbohydrates, and wood density at age 4 and tree volume at age 6. By capturing hidden relatedness and correcting pedigree errors, SNP data disentangled non-additive from additive variance providing more realistic estimates of narrow-sense heritability than pedigrees, and more accurate predictions of trait values. Predictive abilities (PAs) ranged from 0.12 for volume (pedigree-based model) to 0.44 for wood density (models H, GA, and GAD). Considerable dominance variance was seen for all traits, growth was the trait most influenced by it, resulting in PAs 48.9% higher when this effect is considered, a result with important consequences both for clonal propagation and overall selection efficiency (S_eff). Using a HBLUP model, phenotypes of non-genotyped trees increased PAs by increasing sample size and provided realized relationships with reduced genotyping cost. In a recurrent selection program, the preclusion of progeny testing provides an increase in S_eff between 232% and 299%. In a clonal selection program, the elimination of both progeny and initial clonal trial may increase S_eff between 134% and 277%. Increasing selection intensity by genomic prediction resulted in an additional impact on S_eff. This study provides groundwork to implement genomic selection in E. benthamii breeding.

Eucalyptus bethamii的独特适应性低温加上快速生长和多才多艺的木材质量使其成为世界范围内易受霜冻地区的有价值的种植园物种，但对其关键工业特性的定量遗传参数知之甚少。我们使用 GBLUP 添加剂 (GA)、添加剂显性 (GAD)、单步 (HBLUP) 和基于谱系的预测模型来估计 4 岁时的木质素、提取物、碳水化合物和木材密度以及 6 岁时的树木体积。捕捉隐藏的相关性并纠正系谱错误，SNP 数据将非可加性与可加性方差分开，提供比系谱更真实的狭义遗传力估计，以及更准确的性状值预测。预测能力 (PA) 的范围从体积的 0.12（基于谱系的模型）到木材密度的 0.44（模型 H、GA 和 GAD）。S _eff )。使用 HBLUP 模型，非基因分型树的表型通过增加样本量增加了 PA，并提供了降低基因分型成本的已实现关系。在循环选择程序中，子代检测的排除提供了232% 和 299% 之间的S _eff增加。在克隆选择程序中，消除后代和初始克隆试验可能会使S _eff增加134% 到 277%。通过基因组预测增加选择强度导致对S _eff的额外影响。本研究为在本氏大肠杆菌育种中实施基因组选择提供了基础。