This section will briefly introduce the different kinds of evidence used to reconstruct metazoan phylogeny, and review their respective strengths and weaknesses. Until about 15 years ago all metazoan phylogenies were strictly based upon morphological evidence, including information about embryology. With the advent of molecular phylogenetics this situation has changed dramatically, and many workers now prefer to infer phylogenies solely with molecular data, or in combination with morphological evidence. In contrast, information derived from fossils plays only a minor role in metazoan phylogenetics, with the exception of several taxa for which a relatively rich and detailed fossil record is available, notably the arthropods, echinoderms, priapulids, brachiopods, and chordates.
1.02.1.1.1 Molecules Our understanding of the tree of life on all levels is increasingly based upon molecular evidence. Although many zoologists still endorse the value of morphological evidence, the molecular hegemony is increasingly gaining strength in metazoan phylogenetics as well.
The foundation for the new view of metazoan phylogeny is firmly based upon analysis of the 18S rRNA, or nuclear small ribosomal subunit (SSU), gene (Eernisse and Peterson, 2004; Halanych, 2004). The SSU gene was one of the first genes that was sequenced for species belonging to different animal phyla. For most of the traditionally recognized animal phyla the SSU gene has now been sequenced for at least one species, and for increasing numbers of phyla multiple species have been sequenced. Consequently, the SSU gene is currently the most broadly sampled gene within the Metazoa. Its relatively slow mutation rate makes it an appropriate gene to study the deep divergences of the Metazoa, but the SSU gene has not resolved all high-level phylogenetic questions.
To complement and test the results based on the SSU gene, researchers are now investigating the phy-logenetic utility of different molecular loci as well, including the 28S rRNA, or nuclear large ribosomal subunit (LSU) gene, the myosin heavy-chain type II gene, various mitochondrial genes, as well as mito-chondrial gene order, Hox genes, and the genes for elongation factor-1a and elongation factor-2 (Giribet, 2002, 2003; Ruiz-Trillo et al., 2002; Halanych, 2004).
In addition, the sequencing of the entire genomes of several model metazoans such as the nematode Caenorhabditis elegans, the fly Drosophila melano-gaster, and the primate Homo sapiens has allowed for a different strategy of phylogenetic analysis. Instead of sampling many taxa for the same gene, several phylogenomic studies have recently been published that have attempted to reconstruct metazoan phylogeny by comparing large numbers of homologous sequences for just the small set of animals for which genome sequence data are available (Copley et al., 2004; Wolf et al., 2004; Dopazo and Dopazo, 2005; Philip et al., 2005; Philippe et al., 2005). Interestingly, the results of several of these studies do not support the findings of phylo-genetic analyses based on fewer genes, but with larger samples of species. This apparent discrepancy focuses attention on two important issues.
First, these apparent phylogenetic conflicts indicate the importance of adequate taxon sampling. In order to represent properly phylogenetically important variation in gene sequences, and to prevent being misled by homoplasy, phylogeneticists should aim at a sufficiently large representation of metazo-ans in their phylogenies.
Second, recent research has shown that, for the same set of taxa, independent phylogenetic analyses based on only one or a few genes can be in significant conflict with each other (Rokas et al., 2003b).
However, in such cases the conflict may be resolved by combining all molecular evidence into a single phylogenetic analysis, which may then yield a single, well-supported phylogeny. This shows that it is crucial to study metazoan phylogeny with multiple molecular loci, and that one should be wary of accepting phylogenies based on just one or two loci.
As an illustration of the high pace of developments in the discipline, new studies published between submission and revision of this article show that the conflict between standard molecular phylogenetic analyses and phylogenomic studies is probably only apparent (Copley et al., 2004; Philippe et al., 2005). It is the probable result of long-branch attraction resulting from insufficient sampling in the first phylogenomic studies. The long-diverged gene sequences of the distantly related model organisms have accumulated so many mutations that chance similarities may cause them to be grouped together in a phylogenetic analysis. Increasing taxon sampling may break up such long branches, decreasing the conflict with the much-better-sampled analyses that focus on only one or a few genes.
Finally, molecular evidence has not only added greatly to our ability to reconstruct metazoan phylogeny, it has also allowed us for the first time to estimate the approximate divergence times of the major metazoan taxa, even when the fossil record is mostly mute about most of these divergence events (Smith and Peterson, 2002). This application of molecular evidence is among the most exciting, but also the most controversial of topics in evolutionary biology. Of particular interest are the problems of apparent discrepancies of divergence time estimates based on molecular and fossil evidence observed for many groups of organisms, ranging from vascular plants to birds, and the possibility that the major metazoan lineages diverged very rapidly, making it very difficult to reconstruct and precisely date the sequence of divergence events giving rise to modern crown groups (Smith and Peterson, 2002; Graur and Martin, 2004; Reisz and Muller, 2004; Rokas et al., 2005; Ho and Larson, 2006). However, through an increasing understanding of the relative strengths and weaknesses of molecular and fossil divergence times estimates, recent studies have been able to bring molecular estimates of divergence times and the metazoan fossil record increasingly in closer agreement, although a period of cryptic evolution of the major metazoan lineages is still suggested, about which the fossil record is silent.
1.02.1.1.2 Morphology Morphological characters, obtained from the study of all stages of the life cycle, from zygote to adult organism, had been the sole source of phylogenetic information until the late 1980s. Even today morphological evidence is habitually used for phylogenetic analyses of the Metazoa, either alone, or in combination with molecular evidence (Ax, 1995-2001; Zrzavy et al., 1998; Giribet et al., 2000; Nielsen, 2001; Peterson and Eernisse, 2001; Glenner et al., 2004; Jenner and Scholtz, 2005).
Certain types of morphological character have traditionally been imbued with great phylogenetic value. The most familiar major divisions in the animal kingdom reported in textbooks reflect many of these characters. For example, the possession of bilateral symmetry, with distinct anteroposterior, dorsoventral, and left-right axes has long been regarded as the principal synapomorphy of the Bilateria. The possession of a body composed of three germ layers, ectoderm, endoderm, and meso-derm, is reflected in the name Triploblastica, a synonym of Bilateria.
Within the Bilateria the clades Protostomia and Deuterostomia are typically diagnosed on the basis of different fates of the embryonic blastopore, which characteristically is said to give rise to the adult mouth in protostomes, and the anus in deuterostomes.
Great phylogenetic value has also been attached to the possession of characteristic cleavage patterns in the early embryo. For example, the widely recognized clade Spiralia is characterized by spiral cleavage, found in such phyla as the mollusks, annelids, and nemerteans.
The nature of body cavities has been of paramount importance in metazoan phylogenetics, and the bilaterian groups Coelomata, Pseudocoelomata, and Acoelomata have often been distinguished in textbooks, with the first possessing a coelom, and the latter two lacking a coelom.
In addition to these important characters, the organization of the central nervous system (brain structure and configuration of main nerve cords) and the nature of the life cycle (indirect development with larvae, or direct development from egg to adult without an intervening free-living larva) have also played important roles in generating phylogenetic hypotheses for the Metazoa. Currently available morphological data matrices may include hundreds of characters, and the largest published morphological data set for the Metazoa included more than 16000 data entries (Zrzavy et al., 1998).
Despite this wealth of morphological information and the publication of a significant number of morphological phylogenetic analyses of the Metazoa over the 15 years, no detailed consensus view of animal relationships has yet resulted on the basis of morphology alone. This is because the selection and interpretation of phylogenetic characters are fraught with difficulties, and different decisions feeding into data matrix constructions may lead to different phylogenies (Jenner and Schram, 1999; Jenner, 2001, 2003, 2004a). Even when a data set has been properly compiled, homoplasy of characters may mislead the phylogenetic analysis. Recent research increasingly shows that key phylogenetic characters once thought to have evolved only once may in fact be evolutionarily very labile.
1.02.1.1.3 Fossils Although fossils may be thought of as providing perhaps the most direct evidence of the course of evolution, fossils have nevertheless not played a leading role in establishing high-level metazoan phylogeny. So far the fossil record remains largely silent about the details of the origin of the animal phyla, but as the fossil record continues to be mined, new and valuable insights into metazoan evolution continue to accrue (Valentine, 2004).
The most important contribution of fossils to metazoan phylogeny is in supplying the information to reconstruct the stem taxa of modern or crown group taxa (Budd and Jensen, 2000). Of special relevance are various exceptionally preserved Cambrian faunas, such as the Burgess Shale fauna of British Columbia, the Sirius Passet fauna of northern Greenland, and the Chengjiang fauna of southwest China. Fossils collected at these and other localities have greatly informed the early evolution of groups such as the arthropods, priapulids, and chordates. However, the phylogenetic significance of many fossils remains elusive, in particular the highly problematic Ediacaran (Precambrian) fossils, which may be a diverse assemblage that contains genuine members of extant phyla, and real oddballs that cannot be placed inside the Metazoa.
The inclusion of fossils into phylogenetic analyses of living taxa can have a huge impact on tree topology (Wheeler et al., 2004; Lee, 2005). Consequently, ignoring available fossil evidence cannot be easily justified, even when one is only interested in the phylogenetic relationships of living taxa.
Was this article helpful?