Data from: Bayesian morphological clock methods resurrect placoderm monophyly and reveal rapid early evolution in jawed vertebrates

The phylogeny of early gnathostomes provides an important framework for understanding one of the most significant evolutionary events, the origin and diversification of jawed vertebrates. A series of recent cladistic analyses have suggested that the placoderms, an extinct group of armoured fish, form a paraphyletic group basal to all other jawed vertebrates. We revised and expanded this morphological data set, most notably by sampling autapomorphies in a similar way to parsimony-informative traits, thus ensuring this data (unlike most existing morphological data sets) satisfied an important assumption of Bayesian tip-dated morphological clock approaches. We also found problems with characters supporting placoderm paraphyly, including character correlation and incorrect codings. Analysis of this data set reveals that paraphyly and monophyly of core placoderms (excluding maxillate forms) are essentially equally parsimonious. The two alternative topologies have different root positions for the jawed vertebrates but are otherwise similar. However, analysis using tip-dated clock methods reveals strong support for placoderm monophyly, due to this analysis favoring trees with more balanced rates of evolution. Furthermore, enforcing placoderm paraphyly results in higher levels and unusual patterns of rate heterogeneity among branches, similar to that generated from simulated trees reconstructed with incorrect root positions. These simulations also show that Bayesian tip-dated clock methods outperform parsimony when the outgroup is largely uninformative (e.g., due to inapplicable characters), as might be the case here. The analysis also reveals that gnathostomes underwent a rapid burst of evolution during the Silurian period which declined during the Early Devonian. This rapid evolution during a period with few articulated fossils might partly explain the difficulty in ascertaining the root position of jawed vertebrates.

早期有颌脊椎动物（gnathostomes）的系统发育关系，为理解演化史上最具里程碑意义的事件之一——有颌脊椎动物的起源与辐射分化——提供了重要框架。一系列近期的分支系统学分析（cladistic analyses）表明，盾皮鱼（placoderms，一类已灭绝的带甲鱼类）构成了一个并系群（paraphyletic group），位于所有其他有颌脊椎动物的基干位置。本研究对该形态学数据集进行了修订与扩充，最关键的改进是以与简约信息性状（parsimony-informative traits）一致的方式纳入了独有衍征（autapomorphies），从而使得本数据集（与多数现有形态学数据集不同）满足贝叶斯末端定年形态钟（Bayesian tip-dated morphological clock）方法的一项重要前提假设。同时，本研究发现支持盾皮鱼并系群的性状存在若干问题，包括性状相关性（character correlation）与编码错误（incorrect codings）。对本数据集的分支分析显示，核心盾皮鱼类（不含具颌骨类群）的并系起源与单系群（monophyletic group）起源在简约性上基本相当。这两种备选拓扑结构仅在有颌脊椎动物的根位置上存在差异，其余部分基本一致。但采用末端定年钟方法（tip-dated clock methods）进行分析时，结果强烈支持盾皮鱼单系起源——这是因为该分析更倾向于演化速率更均衡的系统发育树。此外，强制设定盾皮鱼为并系群会导致分支间演化速率异质性（rate heterogeneity）程度更高且模式异常，与以错误根位置重建的模拟树所产生的异质性模式相似。这些模拟实验还表明，当外类群（outgroup）基本无信息（例如因存在大量不适用性状（inapplicable characters））时，贝叶斯末端定年形态钟方法的表现优于简约法，这与本研究的实际情况相符。本分析还显示，有颌脊椎动物在志留纪（Silurian period）经历了快速的演化爆发，这一演化速率在早泥盆世（Early Devonian）逐渐放缓。在几乎没有关节连接完好化石的志留纪发生的这一快速演化事件，或许可以部分解释学界难以确定有颌脊椎动物根位置的原因。