Burleigh J Gordon, Kimball Rebecca T, Braun Edward L
Department of Biology, University of Florida, United States.
Mol Phylogenet Evol. 2015 Mar;84:53-63. doi: 10.1016/j.ympev.2014.12.003. Epub 2014 Dec 27.
Birds are the most diverse tetrapod class, with about 10,000 extant species that represent a remarkable evolutionary radiation in which most taxa arose during a short period of time. There has been a tremendous increase in the amount of molecular data available from birds, and more than two-thirds of these species have some sequence data available. Here we assembled these available sequence data from birds to estimate a large-scale avian phylogeny. We performed an unconstrained maximum likelihood analysis of a sparse supermatrix comprising 22 nuclear loci and seven mitochondrial regions from 6714 species. We inferred a phylogeny with a backbone remarkably similar to that obtained by detailed analyses of multigene datasets, yet with the addition of thousands of more taxa. All orders were monophyletic with generally high support. While most families and genera were well supported, a number of them, especially within the oscine passerines, had little or no support. This likely reflects problems with the circumscription of these genera and families. Our results indicate that the amount of sequence data currently available is sufficient to produce a robust estimate of the avian tree of life using current methods of inference. The availability of a tree that is unconstrained by prior information, with branch lengths that have a direct connection to the underlying data, should be useful for comparative methods, taxonomic revisions, and prioritizing taxa that should be targeted for additional data collection.
鸟类是最多样化的四足动物类群,现存约10000个物种,代表了一次显著的进化辐射,其中大多数分类群在短时间内出现。可获得的鸟类分子数据量有了巨大增长,超过三分之二的鸟类物种有一些序列数据。在此,我们整合了这些来自鸟类的可用序列数据,以估计一个大规模的鸟类系统发育树。我们对一个稀疏超级矩阵进行了无约束最大似然分析,该矩阵包含来自6714个物种的22个核基因座和7个线粒体区域。我们推断出的系统发育树主干与通过对多基因数据集的详细分析所获得的主干非常相似,但增加了数千个更多的分类单元。所有目都是单系的,通常具有较高的支持度。虽然大多数科和属得到了很好的支持,但其中一些,特别是在鸣禽雀形目内,支持度很低或没有支持度。这可能反映了这些科和属的界定存在问题。我们的结果表明,目前可用的序列数据量足以使用当前的推断方法对鸟类生命树进行可靠的估计。一个不受先验信息约束、分支长度与基础数据有直接关联的系统发育树的可用性,应该对比较方法、分类修订以及确定应针对其进行额外数据收集的分类单元的优先级很有用。