A large-scale phylogeny of Amphibia including over 2800 species, and a revised classification of extant frogs, salamanders, and caecilians

The extant amphibians are one of the most diverse radiations of terrestrial vertebrates (>6800 species). Despite much recent focus on their conservation, diversification, and systematics, no previous phylogeny for the group has contained more than 522 species. However, numerous studies with limited taxon sampling have generated large amounts of partially overlapping sequence data for many species. Here, we combine these data and produce a novel estimate of extant amphibian phylogeny, containing 2871 species (40% of the known extant species) from 432 genera (85% of the 500 currently recognized extant genera). Each sampled species contains up to 12,712 bp from 12 genes (three mitochondrial, nine nuclear), with an average of 2563 bp per species. This data set provides strong support for many groups recognized in previous studies, but it also suggests non-monophyly for several currently recognized families, particularly in hyloid frogs (e.g., Ceratophryidae, Cycloramphidae, Leptodactylidae, Strabomantidae). To correct these and other problems, we provide a revised classification of extant amphibians for taxa traditionally delimited at the family and subfamily levels. This new taxonomy includes several families not recognized in current classifications (e.g., Alsodidae, Batrachylidae, Rhinodermatidae, Odontophrynidae, Telmatobiidae), but which are strongly supported and important for avoiding non-monophyly of current families. Finally, this study provides further evidence that the supermatrix approach provides an effective strategy for inferring large-scale phylogenies using the combined results of previous studies, despite many taxa having extensive missing data.

现生两栖类（extant amphibians）是陆生脊椎动物中物种辐射分化最为多样的类群之一，已知现存物种超过6800种。尽管近年来学界对其保护生物学、物种分化及系统分类学给予了大量关注，但此前针对该类群的所有系统发育（phylogeny）研究中，所涵盖的物种数均未超过522种。然而，诸多类群采样（taxon sampling）受限的研究已为大量物种积累了大量部分重叠的序列数据（sequence data）。本研究整合上述数据，构建了涵盖2871种现生两栖类的全新系统发育估计结果，该物种数量占已知现生两栖类物种总数的40%，涉及432个属，占当前已确认的500个现生两栖类属的85%。每个被采样物种的序列数据来自12个基因（3个线粒体基因、9个核基因），总长可达12712个碱基对（bp），单物种平均序列长度为2563个碱基对。本数据集为既往研究中确认的诸多类群提供了极强的支持证据，但同时也揭示多个当前公认的科并非单系群（non-monophyly），尤以雨蛙总科（hyloid frogs）类群为甚，例如角蛙科（Ceratophryidae）、盔头蛙科（Cycloramphidae）、细趾蟾科（Leptodactylidae）和暴蛙科（Strabomantidae）。为修正上述及其他相关问题，我们针对传统上以科和亚科等级划分的现生两栖类类群，提出了修订后的分类系统。该新分类系统纳入了多个当前分类体系中未被认可的科，例如阿尔索蛙科（Alsodidae）、巴特拉奇蛙科（Batrachylidae）、鼻蛙科（Rhinodermatidae）、钝头蟾科（Odontophrynidae）和泽蛙科（Telmatobiidae），这些类群均得到了强有力的支持，且对避免当前各科出现非单系群问题至关重要。最后，本研究进一步证实：尽管诸多类群存在大量缺失数据，但通过整合既往研究成果构建大规模系统发育树时，超级矩阵法（supermatrix approach）仍是一种行之有效的研究策略。