Phylogenomics of Eugenia. Phylogenetic relationships within the hyper-diverse genus Eugenia (Myrtaceae: Myrteae) based on target enrichment sequencing

Eugenia is one of the most taxonomically challenging lineages of flowering plants, in which morphological delimitation has changed over the last few years resulting from recent phylogenetic study based on molecular data. Efforts until now have been limited to Sanger sequencing of mostly plastid markers. These phylogenetic studies indicate 11 clades formalised as infrageneric groups. However, relationships between these clades are poorly supported at key nodes and inconsistent between studies, particularly along the backbone and within Eugenia sect. Umbellatae, which encompasses ca. 700 species. To resolve and better understand systematic discordance, 54 Eugenia taxa were subjected to phylogenomic Hyb-Seq sequencing using 353 low-copy nuclear genes. Twenty species trees based on coding and non-coding loci of nuclear and plastid datasets were recovered using coalescent and concatenated approaches. Concordant and conflicting topologies were assessed by comparing tree landscapes, topology tests, and gene and site concordance factors. Topologies are similar except between nuclear and plastid datasets. The coalescent trees better accommodate disparity in the intron dataset, which contains more parsimony informative sites, while concatenated trees recover more conservative topologies as they have a narrower distribution in the tree landscape. This suggests that highly supported phylogenetic relationships recovered in previous studies do not necessarily indicate overwhelming concordant signal. Congruence must be interpreted carefully especially in concatenated datasets. Despite this, congruence between the multi-species coalescent (MSC) approach and concatenated tree topologies recovered here is notable. Our analysis does not support Eugenia subg. Pseudeugenia or sect. Pilothecium as currently circumscribed, suggesting necessary taxonomic reassessment. Five clades are further discussed within Eugenia sect. Umbellatae, progress towards its division into workable clades. While targeted sequencing provides a massive quantity of data that improves phylogenetic resolution in Eugenia, uncertainty still remains in Eugenia sect. Umbellatae. The general pattern of higher site coefficient factor (CF) than gene CF in the backbone of Eugenia suggests stochastic error from limited signal. Tree landscapes in combination with concordance factor scores, as implemented here, provide a comprehensive approach that incorporates several phylogenetic hypotheses. We believe the protocols employed here will be of use for future investigations into the evolutionary history of Myrtaceae.

番樱桃属（Eugenia）是被子植物中分类学难题最多的类群之一，近年来基于分子数据的系统发育研究推动了其形态界定的多次修订。此前的相关研究多局限于以主要质体标记（plastid markers）为靶标的桑格测序（Sanger sequencing）。现有系统发育研究已明确划分出11个被正式确立为属下类群的支系（clades）。但这些支系间的系统发育关系在关键节点上支持度极低，且不同研究间结果存在分歧，尤其是在系统发育主干支系以及包含约700个物种的伞形番樱桃组（Eugenia sect. Umbellatae）中这一问题尤为突出。
为解析并深入理解该类群的系统发育不一致性，本研究对54个番樱桃属类群开展了基于353个低拷贝核基因的系统基因组Hyb-Seq测序。分别利用溯祖法（coalescent）与串联法（concatenated approaches），基于核基因组与质体基因组的编码及非编码位点，共构建了20个物种树。通过比较树谱、拓扑检验（topology tests）以及基因与位点一致性因子（gene and site concordance factors），评估了拓扑结构的一致性与冲突情况。结果显示，除核基因组与质体基因组数据集间存在差异外，其余拓扑结构整体相似。
溯祖树更适配内含子数据集（intron dataset）的演化差异——该数据集包含更多简约信息位点（parsimony informative sites）；而串联树则呈现出更为保守的拓扑结构，因其在树谱中的分布范围更窄。这表明以往研究中获得的高支持度系统发育关系，未必代表压倒性的一致性信号，对一致性的解读需格外谨慎，尤其在串联数据集中。尽管如此，本研究获得的多物种溯祖（multi-species coalescent, MSC）法与串联树拓扑结构间的一致性仍较为显著。
本研究结果不支持当前界定的番樱桃亚属（Eugenia subg. Pseudeugenia）以及Pilothecium组，提示有必要对其分类学界定进行重新评估。针对伞形番樱桃组，本研究进一步划分出5个支系并展开讨论，为将其划分为可操作的分类类群提供了研究进展。靶向测序（targeted sequencing）技术可产生海量数据，有效提升了番樱桃属的系统发育分辨率，但伞形番樱桃组内的系统发育不确定性仍未完全消除。
番樱桃属主干支系的位点一致性因子（CF）普遍高于基因CF，这一整体模式提示信号有限所带来的随机误差。本研究结合树谱与一致性因子评分的分析方法，可全面整合多种系统发育假说。我们认为本研究采用的实验方案，可为未来桃金娘科（Myrtaceae）演化历史的相关研究提供参考。