研究桉树控制授粉后目标性状的基因作用方式是探索其基因重组规律的重要内容。 常规的数量统计分析精度往往不高, 而DNA分析的专业要求高, 且费时费力。 该研究利用近红外光谱(NIRs)研究不同基因型桉树杂交种、 亲本及杂交种与亲本间近红外光谱信息的关系, 探索NIRs用于桉树杂交种与其亲本判别的可行性和准确性。 以控制授粉的桉树亲本及其杂交F1代材料为对象, 每种基因型从各自田间试验分别选取10个单株, 采集树冠中上部新鲜健康叶片。 用手持式近红外仪Phazir Rx(1624)采集桉树杂交种与其亲本叶片的NIRs信息。 每单株选10片完全生理成熟的健康叶片, 避开叶脉扫描其正面光谱5次, 以50条NIRs信息的均值代表单个叶片的NIRs信息, 最终每个基因型获得10条NIRs信息。 对原始NIRs采用二阶多项式S.G一阶导数预处理。 预处理后的NIRs用于多元统计分析, 首先对桉树杂交亲本和子代样本进行主成分分析(PCA), 直观展示不同基因型的分类情况。 然后运用簇类独立软模式(SIMCA)和偏最小二乘判别分析(PLS-DA)两种有监督的判别模式验证NIRs用于桉树杂交种与其亲本树种的分类判别效果。 PCA结果显示, 不同的亲本间、 杂交种间及杂交种与亲本间样本的主因子得分可以清晰地将各基因型分开。 SIMCA模式判别分析中, 桉树杂交种样本到亲本PCA模型的样本距离显示, 待判别样本能够形成单独的聚类, 且能直观反映两者的遗传相似。 PLS-DA判别结果显示, 桉树杂交亲本的PLS模型能通过预测其杂交子代的响应变量将其与亲本准确分开。 结果表明, 桉树叶片的NIRs信息可以准确地反映桉树杂交子代遗传信息的传递规律, NIRs判别模型可以准确地将各种基因型予以区分。 因此, NIRs信息不仅可用于桉树杂交种和纯种的定性判别, 还可以分析桉树基因重组过程中加性遗传效应的大小, 从而为桉树遗传基础分析及其育种改良研究提供理论支撑。

Studying the gene control pattern of interested traits after control pollination is one of the key fields in exploring the law of gene recombination in eucalypt improvement. The accuracy of conventional quantitative analysis for that is often low, and the DNA analysis for that has high professional requirements, and is time consuming and laborious commonly. The aim of the current study is to study the relationship among different genotypes of hybrids, parents, hybrids and their parents in eucalypt based on the near infrared spectroscopy (NIRs) of foliage, and to discuss the practicability and the accuracy of the NIRs discriminant model for the classifying of eucalypt hybrids and their parents. The genetical materials in the study contained three eucalypt parents and their F1 progenies by control pollination. Fresh and healthy leaves from middle to upper crowns in a tree from their field testing trials were collected, and 10 individuals were chosen per genotype. The handheld portable near infrared spectrometer Phazir Px (1624) was used to scan the NIRs of foliage collected. 10 healthy current-year leaves were chosen per individual tree, five scans for NIRs from each side of the middle part of the frontal vein of the leaves were taken, calculated the average of 50 scans as the NIRs data of a leaf, thus 10 NIRs were got for every genotype in totally. The transform of S.G first derivative with second order polynomial fit was performed for the raw NIRs in present study. The successive multivariate analysis was conducted after NIRs pretreatment. To demonstrate the classification of different genotypes by the principal component analysis (PCA) with the NIRs data of hybrids and the patents in eucalypt. Then, two supervised discriminant models, soft independent modeling of class analogy (SIMCA) and partial least squares-discriminant analysis (PLS-DA) pattern recognition, were used to test the accuracy of NIRs model in the classifying for eucalypt hybrids and their cross parents. The scores plot of PC1 and PC2 in PCA demonstrated strong groups among different genotypes, such as cross parents, hybrids, and between hybrids and their parents. The sample distance to parents PCA model in the SIMCA analysis showed that the hybrids to be distinguished can form a clear group differentiated with their parents, and demonstrated the genetic similarity between parents and their progenies directly. The PLS-DA pattern recognition analysis indicated that the hybrids can be discriminated with cross parents by the response values of hybrids predicted by the parents PLS model. All the findings in present study showed that NIRs information of eucalypt leave truly reflects the transmission of genetic information occurring in the process of control pollination, and different genotypes including hybrids and their parents can be discriminated accurately by NIRs models, suggesting that NIRs can be used not only for qualitative identification between eucalypt hybrids and the cross parents, but also for assessing the extent of additive genetic effects in gene recombination, which can provide theoretical reference for the genetic basis analysis and breeding improvement in eucalypt.