Data congruence, paedomorphosis and salamanders

Identifying and addressing methodological incongruence in phylogenomics: A review

The availability of phylogenetic data has greatly expanded in recent years. As a result, a new era in phylogenetic analysis is dawning-one in which the methods we use to analyse and assess our data are the bottleneck to producing valuable phylogenetic hypotheses, rather than the need to acquire more data. This makes the ability to accurately appraise and evaluate new methods of phylogenetic analysis and phylogenetic artefact identification more important than ever. Incongruence in phylogenetic reconstructions based on different datasets may be due to two major sources: biological and methodological. Biological sources comprise processes like horizontal gene transfer, hybridization and incomplete lineage sorting, while methodological ones contain falsely assigned data or violations of the assumptions of the underlying model. While the former provides interesting insights into the evolutionary history of the investigated groups, the latter should be avoided or minimized as best as possible. However, errors introduced by methodology must first be excluded or minimized to be able to conclude that biological sources are the cause. Fortunately, a variety of useful tools exist to help detect such misassignments and model violations and to apply ameliorating measurements. Still, the number of methods and their theoretical underpinning can be overwhelming and opaque. Here, we present a practical and comprehensive review of recent developments in techniques to detect artefacts arising from model violations and poorly assigned data. The advantages and disadvantages of the different methods to detect such misleading signals in phylogenetic reconstructions are also discussed. As there is no one-size-fits-all solution, this review can serve as a guide in choosing the most appropriate detection methods depending on both the actual dataset and the computational power available to the researcher. Ultimately, this informed selection will have a positive impact on the broader field, allowing us to better understand the evolutionary history of the group of interest.

Insights into skull evolution in fossorial snakes, as revealed by the cranial morphology of Atractaspis irregularis (Serpentes: Colubroidea)

Comparative osteological analyses of extant organisms provide key insight into major evolutionary transitions and phylogenetic hypotheses. This is especially true for snakes, given their unique morphology relative to other squamates and the persistent controversy regarding their evolutionary origins. However, the osteology of several major snake groups remains undescribed, thus hindering efforts to accurately reconstruct the phylogeny of snakes. One such group is the Atractaspididae, a family of fossorial colubroids. We herein present the first detailed description of the atractaspidid skull, based on fully segmented micro-computed tomography (micro-CT) scans of Atractaspis irregularis. The skull of Atractaspis presents a highly unique morphology influenced by both fossoriality and paedomorphosis. This paedomorphosis is especially evident in the jaws, palate, and suspensorium, the major elements associated with macrostomy (large-gaped feeding in snakes). Comparison to scolecophidians-a group of blind, fossorial, miniaturized snakes-in turn sheds light on current hypotheses of snake phylogeny. Features of both the naso-frontal joint and the morphofunctional system related to macrostomy refute the traditional notion that scolecophidians are fundamentally different from alethinophidians (all other extant snakes). Instead, these features support the controversial hypothesis of scolecophidians as "regressed alethinophidians," in contrast to their traditional placement as the earliest-diverging snake lineage. We propose that Atractaspis and scolecophidians fall along a morphological continuum, characterized by differing degrees of paedomorphosis. Altogether, a combination of heterochrony and miniaturization provides a mechanism for the derivation of the scolecophidian skull from an ancestral fossorial alethinophidian morphotype, exemplified by the nonminiaturized and less extreme paedomorph Atractaspis.

A New Basal Salamandroid (Amphibia, Urodela) from the Late Jurassic of Qinglong, Hebei Province, China

A new salamandroid salamander, Qinglongtriton gangouensis (gen. et sp. nov.), is named and described based on 46 fossil specimens of juveniles and adults collected from the Upper Jurassic (Oxfordian) Tiaojishan Formation cropping out in Hebei Province, China. The new salamander displays several ontogenetically and taxonomically significant features, most prominently the presence of a toothed palatine, toothed coronoid, and a unique pattern of the hyobranchium in adults. Comparative study of the new salamander with previously known fossil and extant salamandroids sheds new light on the early evolution of the Salamandroidea, the most species-diverse clade in the Urodela. Cladistic analysis places the new salamander as the sister taxon to Beiyanerpeton, and the two taxa together form the basalmost clade within the Salamandroidea. Along with recently reported Beiyanerpeton from the same geological formation in the neighboring Liaoning Province, the discovery of Qinglongtriton indicates that morphological disparity had been underway for the salamandroid clade by early Late Jurassic (Oxfordian) time.

TreSpEx-Detection of Misleading Signal in Phylogenetic Reconstructions Based on Tree Information

Phylogenies of species or genes are commonplace nowadays in many areas of comparative biological studies. However, for phylogenetic reconstructions one must refer to artificial signals such as paralogy, long-branch attraction, saturation, or conflict between different datasets. These signals might eventually mislead the reconstruction even in phylogenomic studies employing hundreds of genes. Unfortunately, there has been no program allowing the detection of such effects in combination with an implementation into automatic process pipelines. TreSpEx (Tree Space Explorer) now combines different approaches (including statistical tests), which utilize tree-based information like nodal support or patristic distances (PDs) to identify misleading signals. The program enables the parallel analysis of hundreds of trees and/or predefined gene partitions, and being command-line driven, it can be integrated into automatic process pipelines. TreSpEx is implemented in Perl and supported on Linux, Mac OS X, and MS Windows. Source code, binaries, and additional material are freely available at http://www.annelida.de/research/bioinformatics/software.html.