基于DNA条形码的物种界定算法比较研究——以北京周边地区舟蛾科为例

The reliability of DNA barcoding as a means of identifying Notodontid moth species(Insecta:Lepidoptera, Notodontidae)in Beijing, China

[目的]为了探究3种常用物种界定方法(jMOTU、ABGD、GMYC)的界定效果。[方法]本研究以中国北京周边地区10个采样点483个舟蛾科样品为例,利用线粒体细胞色素C氧化酶Ⅰ亚基基因(Cytochrome c oxidase subunitⅠgene,COⅠ或COX1,约650 bp)的部分序列,进行3种物种界定算法(jMOTU、ABGD、GMYC)的实例比较研究。[结果]3种物种界定方法的鉴定效力存在差异,与形态学结果相比较,ABGD方法划分物种的准确率为100%,基于BEAST的GMYC模型结果与形态学结果一致,产生的置信区间(64~68)覆盖了形态学的结果(67)。然而,基于d8tree/MPLtree的GMYC方法倾向于高估MOTUs,jMOTU方法倾向于低估物种数目。[结论]结果显示,ABGD方法和基于BEAST的GMYC模型方法对于本文研究对象舟蛾科能够较好地划分,可以对基于形态学的物种界定进行有效补充。

[Objectives] To evaluate the performance of three species identification algorithms （jMOTU, ABGD, GMYC） with respect to the Notodontidae （Insecta, Lepidoptera）. [Methods] CO I genes from 483 notodontid moth specimens collected from 10 collection sites in Northern China around Beijing were amplified using universal barcoding primers. [Results] Results obtained using each of the three algorithims differed. Using the results obtained by morphological species identification methods as a reference, the ABGD algorithm was 100% accurate. The results obtained using GMYC_BEAST were also generally consistent with those obtained by morphological methods, having a confidence interval of 64-68 c.f. 67 species identified by morphological methods. However, the GMYC_d8tree/MPLtree method tended to overestimate MOTUs and the jMOTU method tended to underestimate the number of species. [Conclusion] ABGD and GMYC_BEAST can reliably identify species within Notodontidae on the basis of CO I gene variation.