樟科山胡椒属（Lindera Thunb.）是核心月桂族的关键类群，为东亚—北美间断分布格局。先前对山胡椒属系统发育的研究报道，由于相关类群的采样不全面和分子片段偏少等原因，其与近缘属（核心月桂族；木姜子属、新木姜子属、黄肉楠属、月桂属、拟檫木属和华檫木属）间以及属内各组之间系统关系并未得到解决。本研究通过分子系统发育学和形态特征综合分析，对东亚—北美分布的山胡椒属及其近缘属的系统发育学进行研究，并对东亚—北美间断分布的多系类群生物地理学问题进行初步探讨。 本研究在东亚—北美分布范围内选取了山胡椒属及其近缘属110个植物样品，其中，山胡椒属82个样品，近缘属28个样品；樟族及鳄梨族11个样品作为外类群。选取17对低拷贝核基因片段作为本研究的分子标记，基于贝叶斯法的多物种溯祖模型构建低拷贝核基因的物种树，最大似然法和贝叶斯法构建串联核基因系统发育树，使用贝叶斯宽松分子钟法推测山胡椒属及其近缘属物种树内每个进化枝的分歧时间。 低拷贝核基因多态性信息位点变异从15.84%到37.18%，单独对低拷贝核基因构建贝叶斯系统发育树基本分清鳄梨族与樟族、核心月桂族的族间系统发育关系。编码区比例较高的LPD、INTE和STP没有明显的系统发育树分支结构。基因串联构建系统发育树具有相似的拓扑结构，均支持对山胡椒属及其近缘属分成五个组合群。 研究结果强烈支持山胡椒属及其近缘属是一个单系类群，檫木属与樟族亲缘关系较近。山胡椒属与木姜子属并没有很好地属间界限区分，黄肉楠属红果黄肉楠表现为与新木姜子属聚为一支。物种树明显划分为12个分支的属内组间关系。山胡椒属三出脉类型叶片具很高的支持率，多蕊组，山胡椒组（狭叶山胡椒和山胡椒），长梗组，杯托组和掌脉组在物种树上各自强烈聚为一支。北美分布的山胡椒属物种没有清晰的系统发育位置。结果表明叶片三出脉、花序总梗类型、着生花序缩短枝、果托扩展程度等性状具有较强的系统发育信号。 分子钟法进化树表明核心月桂族最近共同溯祖时间在2.60 Ma（95% HPD：2.25–2.93 Ma），远低于化石估计的核心月桂族分化时间（约45 Ma）。分子钟法对物种树分化时间的严重低估可能由于核心月桂族物种间形态相似物种可能来源于庞大的祖先种群，且种内或种间具有杂交或基因流，间接表现在物种树上属间支持度低。 综合上述结果，认为山胡椒属和木姜子属并不是严格意义上的东亚—北美间断分布属，很可能属间或属内类群间间断分布。这种分布格局是山胡椒属和木姜子属在东亚—北美的间断分布具有复杂的网状进化过程以及多次接触，更可能的原因是祖先种群存在基因流，快速演化以及形态停滞进化等因素影响。

Lindera Thunb. (ca. 100 species) is a key group of core Laureae, which are distributed mainly in subtropical to tropical regions of Eastern Asia and North America. Due to incomplete sampling and limited genetic markers in previous phylogenetic studies, the relationships among Lindera and its related genera remain contentious and if they are monophyly requires further examination. In this study, we reconstructed the molecular phylogenetic and morphological analysis in order to clear up the systematic relationships among Lindera and its related genera. And some preliminary conclusions on the biogeography of Eastern Asia and North America disjunction among polyphyly or paraphyly taxa were infered. In this study, we totally sampled 121 samples, including 82 from Lindera, twenty-eight from close genera, and 11 were treated as outgroup. Seventeen nuclear low-copy genes (nLCGs: ACY, FASP, GPN, HET, INTE, ISOM, LEP2, LPD, MALA, MPD, PORI, PRUP, SPT2, STOP, STP, TPP, and VEST) were used. Maximum likelihood and Bayesian analysis were used to construct phylogenetic trees based on concatenation matrix of nLCGs. Multispecies coalescent model based on Bayesian method was used to construct species tree of nLCGs based on all loci partitioned. Divegence time were estimated using uncorrelated relaxed molecular clock. The information sites of nLCGs ranged from 15.84% to 37.18%. Bayesian analyses of nLCG resulted in better resolved relationships among Perseae, Cinnamomeae and core Laureae group. LPD, INTE and STP had low phylogenetic signals because few variations were observed in codon regions. The maximum likelihood and Bayesian phylogenetic trees had similar topological structures base on concatenation matrix, all of them support the division of five major groups among Lindera and its related genera. However, based on IQ-TREE with two methods for measuring datasets, the gene concordance factor (gCF) and the site concordance factor (sCF), indicated that gene sequences were short and conflicting signals limited the ability to resolve bootstrap values. Species tree show that the core Laureae was monophyletic, wchich was close with Cinnamomeae group. Additionally, Sassafras J.Presl merged within Cinnamomeae group would be reasonable. Phylogenetic analyses revealed the paraphyly of Lindera with the inclusion of Litsea Lam., Neolitsea (Benth.) Merr., Laurus L., Iteadaphne Blume. All samples of Neolitsea form a monophyletic clade, which was sister to Actinodaphne Nees (only one species in this study). Twelve well-supported clades were found in the species tree. All trinerved Lindera species that were belonging to sect. Daphnidium had the closest relationship, with Iteadaphne caudata (Nees) H.W.Li and Lindera tienchuanensis W.P.Fang & H.S.Kung were nested within them. The close relationships in the species of sect. Aperula, sect. Cupuliformes, sect. Lindera (L. angustifolia Cheng and L. glauca Blume), sect. Palminerviae, sect. Polyadenia were identified from species tree. These clades showed that the importance of leaves veins, inflorescences types, shortened brachyblasts and fruiting cupules among Lindera and its related genera. The estimated divergence times indicated that Lindera and its related genera were diverged at 2.60 Ma (95% HPD: 2.25–2.93 Ma), which was much more young than fossil recordings. This underestimated time probably reflected Lindera and its related genera may be derived from a large ancestral population, gene flows within or between species, thus showed low support intergeneric in the species tree. Based on the species tree, it was supposed that Lindera was not strictly Eastern Asia and North America disjunct distribution taxa, also suggested that the disjunct patterns were established between-genus or inter-genus. Species tree and molecular clock estimation supported multiple species attachment in network evolution, Because of rapid radiation, gene flow, incomplete lineage sorting, and morphological stasis.