目的 对越南金花茶 Camellia insularis 化学成分及表型性状进行分析，明确其化学成分差异及基于表型性状的亲缘关系，为越南金花茶种质分类和良种选育提供依据。方法 采用超高液相色谱法建立样品化学指纹图谱，目测和直尺观测表型性状，基于液相色谱峰数量和面积，采用主成分分析、相关分析和聚类分析方法，对越南金花茶化学成分差异及亲缘关系进行评价。结果 主成分分析结果表明，以色谱峰数量为评价指标， YN3-78、YN3-77、YNX3-15、YN3-72、YN3-62 等综合评分较高；以色谱峰面积为评价指标， YN3-60、YNX3-05、YNX3-01、YNX3-14、YNX3-04 等综合评分较高。相关分析结果显示，各表型性状与化学成分间存在显著相关。聚类分析结果表明， 77 株越南金花茶可分为 4 个类群，其中，第 1 群 13 株，属大叶型，化学成分出峰时间多在 19.32 min 以前，色谱峰数量较少、峰面积较高。第 2 群 26 株，包含所有线形叶植株，属大叶型，化学成分出峰时间多在 15.21 min 之前，色谱峰数量最少。第 3 群 19 株，叶面积较小，化学成分种类少，峰面积低，综合指标最差。第 4 群 19 株，叶面积最小， 该群 23.33 min 以后有出峰，是区别于其他类群的重要特征； 该群化学成分数量最多，峰面积均较低。结论 越南金花茶表型和化学成分均具有较大差异，通过聚类分析所得 4 个类群均有各自明显的表型性状特征；表型与化学成分存在显著相关，可通过表型性状预测化学成分种类和含量。该结果可为越南金花茶亲缘关系分析和优良种质筛选提供重要依据。

Objective To compare the chemical components and phenotypic characters of Vietnamese golden camellia (Camellia insularis) to investigate the difference in their chemical composition and their phenotypic characteristics-related genetic relationships and provide a basis for germplasm classification and screening of good germplasm of that species. Methods The chemical fingerprint of the samples was established by ultra-high liquid chromatography, and their phenotypic characteristics were observed and measured using a straightedge. Principal component analysis, correlation analysis, and cluster analysis were performed to evaluate the difference in chemical composition and genetic relationships of the samples based on the chromatographic peak number and area. ResultsThe results of principal component analysis showed that the comprehensive scores of YN3-78, YN3-77, YNX3-15, YN3-72 and YN3-62 were higher when the number of chromatographic peaks was used as the evaluation indicator, and those of YN3-60, YNX3-05, YNX3-01, YNX3-14 and YNX3-04 were higher when the chromatographic peak area was used as the evaluation indicator. The correlation analysis reveals that there was significant correlation between phenotypic traits and chemical composition. The cluster analysis results indicated that the 77 C. insularis samples could be divided into four groups, among which groups 1, 2, 3, and 4 consisted of 13, 26, 19, and 19 individuals, respectively. Samples in group 1 had large leaves, and fewer chromatography peaks with higher peak areas, most of which appeared before 19.32 min. Samples in group 2 were large-leaf, including all linear-leaf individuals, and had the minimum chromatographic peaks that mostly appeared before 15.21 min. Samples in group 3 had a smaller leaf area, fewer peaks and a lower peak area, and had the poorest composite index. Samples in group 4 had the smallest leaf and lowest peak areas but maximum chemical components; peaks appearing after 23.33 min in this group made it easy to be distinguished from other groups. Conclusion The phenotypes and chemical components of all the samples were different considerably, and the four groups as formed after the cluster analysis had their own distinct phenotypic characteristics. The phenotype was significantly correlated with the chemical composition, suggesting that the type and content of the chemical composition can be predicted by phenotypic traits. Our results may provide an important basis for the genetic relationship analysis and the screening of good germplasm of C. insularis.

R282.12

广西创新驱动发展专项资金项目（桂科AA18118049-7）；广西自然科学基金重点项目（2020GXNSFDA297029）；中药学广西一流学科（桂教科研[2018]12号）