Deep Metazoan Phylogeny: The Backbone of the Tree of Life

J. Wolfgang Wägele, Thomas Bartolomaeus

• Publisher: De Gruyter
• Year: 2014
• ISBN:
  9783110372960
• Edition: 1
• Topic: Biological Sciences
• Subtopic: Evolution
• Index: Biological Sciences

The growing success of molecular methods has challenged traditional views of animal evolution and a large number of alternative hypotheses are hotly debated today. For the deep metazoan phylogeny project, data sets of hitherto unmatched quality and quantity were compiled and analysed with innovative bioinformatics tools. The book begins at the base of the tree of life to discuss the origin of animals and early branches of the phylogenetic tree. The following section presents special data sets gained from mitochondrial genomes and from morphology, with a focus on nervous systems. The final section is dedicated to theoretical aspects of data analysis and new bioinformatics tools. The book closes with a unique general discussion of all hypotheses contained in previous chapters. This work provides the most comprehensive overview available of the state of the art in this exciting field of evolutionary research.

J. Wolfgang Wägele and Thomas Bartolomaeus

1 Introduction

Animal phylogeny: riddles and answers

This book summarizes major results of a priority program funded by the German Research Foundation during a period of six years (2005–2011) to study the backbone of the metazoan tree of life. More than 30 teams from different institutes and universities contributed new data and innovative phylogenetic analyses. New bioinformatics tools were developed to overcome some methodological problems. During this time a large number of young scientists were trained in phylogenetics, working in an unprecedented multidisciplinary network of specialists.

This major effort was motivated by mutually incompatible phylogenetic hypotheses published recently, which left at a loss all those who write textbooks, teach systematics, or need a phylogeny as a basis to understand evolution. The contradictions between morphology-based hypotheses and molecular phylogenies, and the incompatibility of various “well supported” molecular analyses called for a major endeavor. Also, a closer look at poorly studied morphological characters was commendable to get independent evidence. Therefore, the morphological part of the priority program had a focus on the anatomy of brains, nervous systems, and mesoderm derivatives.

The reader will find here several up-to-date reviews which give an overview over various hypotheses. We did not aim at a harmonization of these hypotheses to produce one fully resolved tree: available data and current methods of data analysis do not allow the selection of a “best” hypothesis covering all nodes in the backbone of the animal tree of life. This book rather cautions against accepting the latest ideas, even when they are published in high ranking journals and adopted by a large audience, and it offers starting points for additional research.

Starting the priority program, one major assumption was that by increasing the data sets (number of sequenced genes, number of taxa) the phylogenetic signal in molecular data sets would become more distinct. While this is in general true for the signal-to-noise ratio, we had to realize that in molecular phylogenetics hidden systematic errors can occur independently of the quantity of sampled data (see the chapters on theory and methods). These errors are not caused by simple noise or by absence of signal in the data. They are probably the reason why morphological evidence strongly contradicts molecular phylogenies (see e.g. the arthropod chapter by Wägele and Kück). To better understand these sources of error, several theoretical studies were conducted. Their results are also presented in this book.

It became clear that new tools needed to avoid systematic errors still have to be developed. Therefore current users of popular software (see also examples in this book) must be aware of some dramatic pitfalls which remained unnoticed until recently.

The book is divided into two major parts: the first part contains all chapters with phylogenetic analyses of different taxa of Metazoa, the second part is dedicated to methods and tools.

The first two chapters discuss the origin of and relationships between diploblas-tic metazoans. Both studies are based on phylogenomic data, but they differ in the composition of data sets and in the way data were analyzed. While Wörheide et al. conclude that sponges are monophyletic and represent among the extant fauna the earliest lineage of Metazoa, a concept also supported by functional and morphological characters (e.g. retention of the feeding mechanism also seen in Choanoflagellata, absence of digestive exoenzymes etc.), Eitel et al. favor a clade Diploblasta, composed of Cnidaria, Ctenophora, Porifera and Placozoa as sister-taxon to Bilateria, because this clade got high support in their analyses of sequence data. This conflict is very typical for the current situation of phylogenetics. The solution is to search for sources of error, to better understand the footprints left by evolution in genomes at the level of site patterns. We have to strike out on new paths of data analysis. We also have to check the plausibility of hypotheses derived from phylogenetic graphs, which are supposed to represent the historical course of macroevolution. Plausibility means that a scenario based on a tree hypothesis must also be able to explain the evolution of modes of life, transitions between body constructions and ecological specializations, also including all available morphological and fossil evidence. Such considerations are lacking in the literature of the past 20 years.

In some cases it is more honest to state that we do not have sufficient evidence. An example is the analysis of the origin of Chaetognatha, discussed by Perez, Müller, and Harzsch. They consider all available data (sequences, HOX genes, embryonic development, anatomy, new neuroanatomical data), point out to possible contradictory interpretations and tentatively propose that chaetognaths may be a separate lineage which evolved parallel to other protostomes.

Further neuroanatomical studies of protostomes provided surprising evidence for the homology of brain characters present in annelids and arthropods, including Onychophora. Loesel demonstrates that these brains have – with their mushroom bodies, central body and glomeruli – a complexity not seen elsewhere. The homology of this type of brain is highly probable. Schmidt-Rhaesa and Rothe present new data on the brains of Gastrotricha and Cycloneuralia (Scalidophora and Nematoida). These brains are quite simple. Assuming that molecular phylogenies are without error, the Ecdysozoa hypothesis would imply a drastic reduction of brain structure (and of other anatomical features) in Gastrotricha and Cycloneuralia in comparison with arthropods, which is unusual in non-sessile animals. Alternatively, one has to assume a parallel evolution of highly similar brains in annelids and arthropods, which also is very unlikely. The search for systematic errors in molecular phylogenies (chapter by Kück, Misof, and Wägele) shows a way out of this dilemma.

Hankeln et al. focused on the Platyzoa hypothesis (Platyhelminthes + Gastrotricha + Gnathifera). They analyzed molecular data sets and review the literature. Their Figure 7.3 represents a typical result obtained with currently available methods, where Platyzoa are monophyletic (though not strongly supported) and related to a large group containing the Lophotrochozoa. The authors caution that in inferred tree graphs many lineages exhibit long branches and that these graphs may still contain artifacts. On the other hand, some details like the close relationship between parasitic Acanthocephala and Seisonidea seem to be plausible in view of the implied evolution of anatomy and modes of life.

The Lophophorata (= Tentaculata) hypothesis (Phoronida, Brachiopoda, Ectoprocta) was studied by Nesnidal et al. using phylogenomic data. The Lophophorata is a typical example for the careening of molecular phylogenies. While many molecular phylogeneticists agree that these taxa belong to a clade that also contains annelids and mollusks (the Lophotrochozoa), many previous studies rejected the monophyly of Lophophorata. Published ideas include paraphyly of Brachiopoda or a closer relationship between Ectoprocta (Bryozoa) and Entoprocta (Kamptozoa). Nesnidal et al. confirm the monophyly of Lophophorata and of the studied Brachiopoda. Kamptozoa are not included in Lophophorata.

Struck et al. provide a deeper inside into the phylogeny of Annelida, also based on molecular data. Zoologists must get used to the idea that other taxa with the traditional rank of phyla are now considered to be annelids, namely the Sipuncula, Echiura, and Myzostomida. Also, it is clear now that the traditional separation of Polychaeta and Oligochaeta is an artifical, non-phylogenetic classification. Arguments for the placement of Myzostomida are reviewed in detail by Bleidorn et al. Studies into the fate of the embryonic coeloms in Annelida and Panarthropoda revealed differences in almost all details of coelomogenesis except for the fact that, if present, these structures are topologically identical in both groups. Since no obvious function could be assigned to these organs in arthropods, Koch, Quast, and Bartolomaeus conclude that most likely historical constraints underlie their existence. In this chapter they modify the traditional concept on coelomic sacs as mesoblastic organs and demonstrate the significance of high-resolution electron microscopy in studying coelomic cavities and nephridia.

Arthropods are the most species rich taxon of Metazoa and their phylogeny is especially difficult to understand due to the old age and strong divergence of many lineages. Wägele and Kück focus on the Mandibulata and summarize morphological evidence (including fossils) that strongly supports the idea that insects are crustaceans. Crustacea are paraphyletic, as suggested by many previous molecular studies. However, in contrast to molecular phylogenies, the available morphological information shows that Remipedia (relic cave crustaceans) conserve a bauplan with many derived characters seen also in Myriapoda and Hexapoda. A new evolutionary scenario suggests that Tracheata (= Atelocerata) are monophyletic and that they evolved from a remiped-like marine ancestor. An explanation for the failure of molecular phylogenies that do not recover these relationships is suggested. These observations would suggest that Pancrustacea is a junior synonym of Mandibulata.

This case is similar to the contradiction between those neuroanatomical data that support the Articulata hypothesis, and the molecular phylogenies that suggest a clade Ecdysozoa. Life would be easier if we ignored the morphological evidence, but this attitude will not uncover the history of life.

Simon and Hadrys present a closer look at the phylogeny within Pterygota using molecular data. They do not propose a final solution but discuss analyses of their own new data in the light of the recent literature. Some of the currently supported ideas are the monophyly of Chiastomyaria (Ephemeroptera and Neoptera), with Odonata as a separate lineage at the base of the pterygote tree. The monophyly of Polyneoptera (orthopteroid taxa, Dictyoptera, Dermaptera and others) is confirmed, and within Holometabola the sister-group relationship between Hymenoptera and the remaining taxa.

The phylogeny within Crustacea is still a riddle. Stegner, Fritsch, and Richter provide new neuroanatomical data. They studied the central complex in the brain of representatives of several major taxa and conclude that this is a homologous part of the brain inherited from the last common ancestor of all crustaceans and insects. This brain complex has been reduced subsequently in several crustacean lineages. Further evidence for relationships cannot be derived from these data. The molecular analyses presented by von Reumont and Wägele confirm several results of previous publications and suggest that stability can be achieved when data and substitution models are selected carefully. For example, in recently published tree graphs Remipedia are the crustacean taxon closest to insects, an idea that is also partly compatible with the scenario presented by Wägele and Kück. However, stability does not imply absence of systematic errors. A suspicious detail seen in molecular phylogenies is that the morphologically highly evolved Malacostraca, which share e.g. several special brain characters with insects, are placed near Cirripedia and Copepoda.

The remaining last arthropod clade, the Chelicerata, is discussed by Dunlop, Borner and Burmester. Their updated phylogeny is basically consistent with previous morphology-based hypotheses, however, the available molecular data do not allow a finer resolution of relationships within Arachnida.

Due to the comparatively good congruence in published literature, the phylogeny of Deuterostomia was not a major theme in our program. However, Schlegel, Weidhase and Stadler conducted an analysis of new molecular data to understand the major nodes in the deuterostome tree. The authors provide a review of recent and some new hypotheses, like the subdivision into the clades Xenacoelomorpha, Ambu-lacraria and Chordata, and within Chordata into the Cephalochordata and Olfactores. The corresponding morphological perspective is summarized by Stach in a critical way. In comparison with the molecular trees, the conclusions are not always the same.

The first part of the book concludes with a review on the information contained in mitochondrial gene arrangements (Podsiadlowski et al.). Gene order is quite variable and not devoid of misleading parallel mutations. It is however possible to infer some node patterns for higher taxa.

The second part of this book begins with the presentation of a new tool that is especially valuable for the management of morphological evidence (Grobe and Vogt). MorphDBase is a database for morphological data and metadata. It implements for the first time standards for content, formats, nomenclature, terminology, and concepts. This web-based system can be used by any registered user, its data pool is growing rapidly. The necessity to standardize the terminology and to conceptualize characters used in phylogenetic analyses is elaborated for a case study on neuroanatomical characters. Together with many colleagues, Loesel and Richter worked out definitions and concepts that are the basis for neurophylogeny, the study of the evolution of nervous systems along the metazoan tree of life.

Misof et al. explain new tools useful for the assessment of molecular data quality prior to tree inference. These include measures of tree-likeness, the search for noise in alignments, or the selection of informative data subsets in large phylogenomic data matrices. Bernt et al. give an overview on computational methods for the analyses of gene order rearrangements, which are mostly applied for mitochondrial genomes.

The RNAsalsa software package is presented by Stocsits et al. This software is relevant to consider secondary structure information of RNA molecules in phylogeny inference. RNAsalsa infers the secondary structure for each sequence and builds alignments based on this information. The results can be used for phylogeny inference with RNA-specific substitution models.

Searching for additional types of molecular data that might be useful for phylogenetic analyses, Lehmann et al. tested near intron pairs (NIPs). The authors explain the workflow, summarize hitherto published results, and present new analyses. Donath and Stadler provide a wider overview on higher order molecular characters (molecular morphology) such as repetitive elements, NUMTs, or microRNAs. Though this type of characters is not free of homoplasies, it is obvious that NIPs and other “rare genomic events” bear relevant phylogenetic information.

Kück, Misof and Wägele caution against a too credulous acceptance of results of maximum likelihood analyses (ML). Though in theory the ML method is infallible when certain assumptions are fulfilled, the reality of incompatible results and the implausibility of some hypotheses prove that there exist several undetected sources of error besides the well-known model misspecifications and weak data sampling. Simulations help to identify the conditions that produce systematic errors as a basis for the development for new tree building algorithms.

Since the period of single gene sequencing is rapidly being succeeded by genome sequencing, growing masses of data have to be handled. The thorough review by Ebersberger and von Haeseler is dedicated to challenges and pitfalls of phyloge-nomics , pointing to the near future of our science.

The main progress that had been achieved during the “Deep Metazoan Phylogeny” project are detailed morphological data based on a broader taxon sampling including previously neglected groups, and a critical evaluation of sequence data. If it were possible to infer phylogeny from those molecular and morphological traces that during the course of evolution remained in all extant taxa, one would expect that identical phylogeny hypotheses should result from analyzing these traces. As long as this is not the case, it is much more the structure of the data that poses conflicts on the analyses than priority of one data source over the other. This book actually presents some examples where additional data and modified analytical tools reveal congruence between molecular and morphological studies, occasionally corroborating traditional views (e.g., monophyly of the Lophophorata) or weakening classical ideas (e.g., Myzostomida as isolated phylum). This book shows the present progress on our long way to a uniform picture of metazoan phylogeny.

Part I: New Data and Phylogenies

dp n="34" folio="" ? dp n="35" folio="" ?

Gert Wörheide, Tetyana Nosenko, Fabian Schreiber, and Burkhard Morgenstern

2 Progress and perspectives of the deep non-bilaterian phylogeny, with focus on sponges (Phylum Porifera)

Abstract: Sponges (Phylum Porifera) are benthic aquatic animals of great ecological and pharmaceutical importance. The phylum is arguably the earliest branching metazoan taxon and of considerable significance for the recon...

Table of contents

• Deep Metazoan Phylogeny: The Backbone of the Tree of Life
• Title Page
• Copyright Page
• List of Contributing Authors
• Table of Contents
• 1 Introduction
• Part I: New Data and Phylogenies
• Part II: New Tools and Methods
• References
• Index