4 Phylogeny of Animals
129
4.4.1 Battle over the Coelomata and the Importance
of Taxonomic Sampling
4.4.1.1 Early Phylogenomic Attempts Challenged the “New View”
The Ecdysozoa clade is the most controversial hypothesis of the “New View” of
animal phylogeny and early multigene analyses of bilaterians therefore attempted
to verify this hypothesis (Blair et al. 2002, Philip et al. 2005, Wolf et al. 2004).
These studies employed from 100 to 780 nuclear genes which were analyzed following concatenation or independent approaches and they were based on species
for which whole genome sequences were available (Blair et al. 2002, Philip et al.
2005, Wolf et al. 2004). Surprisingly, none of these studies recovered the ecdysozoan clade, indeed they all found that nematodes diverged prior to the radiation
of coelomate animals. However, these results are thought to have been influenced
by long-branch artefact because the genome of the nematode C. elegans exhibits
very fast evolutionary rates (Aboobaker and Blaxter 2003). Indeed, these findings were challenged by the publication of the work of Philippe et al. (2005a, b),
which successfully recovered the “new view” of animal phylogeny (Philippe et al.
2005a). By exploiting newly released EST data, these authors incorporated numerous new species belonging to groups such as the choanoflagellates, closest relatives
of metazoans, the cnidarians, sister-group of the bilaterians, and a broad range of
nematodes and platyhelminthes, some of which exhibit slower evolutionary rates
(e.g. Trichinella for nematodes). The inclusion of those taxa made it possible to
limit the phenomenon of long branch attraction – in phylogenetic jargon, to “break
the long branches” – and thus produced support for the Ecdysozoa clade (Fig. 4.3).
Notably, the platyhelminthes and the nematodes sometimes clustered together in
this study, but the authors showed that the removal of fastest evolving genes enabled
the recovery of the “new view” topology, with platyhelminthes being grouped with
annelids and molluscs. This contradicts previous results, which estimated that gene
sampling alone, and not taxon sampling, had the ability to increase the phylogenetic
accuracy (Rokas and Carroll 2005).
4.4.1.2 Coelomata and the Interpretation of Rare Genomic Changes
The importance of dealing with accurate taxon sampling was also stressed in a
recent debate about a possible revival of the Coelomata hypothesis. The origin of
this debate was two studies that investigated two different qualitative genomic characters: “rare amino acid replacements” and patterns of intron conservation. Rare
amino acid changes have been defined as amino acid substitutions that take place
in a limited subset of taxa and that are caused by several nucleotide changes. Those
substitutions are thus associated with a low probability of homoplasy, and 34 such
positions were initially found to be in agreement with the Coelomata topology
(Rogozin et al. 2007). However, subsequent addition of the cnidarian N. vectensis as
129
4.4.1 Battle over the Coelomata and the Importance
of Taxonomic Sampling
4.4.1.1 Early Phylogenomic Attempts Challenged the “New View”
The Ecdysozoa clade is the most controversial hypothesis of the “New View” of
animal phylogeny and early multigene analyses of bilaterians therefore attempted
to verify this hypothesis (Blair et al. 2002, Philip et al. 2005, Wolf et al. 2004).
These studies employed from 100 to 780 nuclear genes which were analyzed following concatenation or independent approaches and they were based on species
for which whole genome sequences were available (Blair et al. 2002, Philip et al.
2005, Wolf et al. 2004). Surprisingly, none of these studies recovered the ecdysozoan clade, indeed they all found that nematodes diverged prior to the radiation
of coelomate animals. However, these results are thought to have been influenced
by long-branch artefact because the genome of the nematode C. elegans exhibits
very fast evolutionary rates (Aboobaker and Blaxter 2003). Indeed, these findings were challenged by the publication of the work of Philippe et al. (2005a, b),
which successfully recovered the “new view” of animal phylogeny (Philippe et al.
2005a). By exploiting newly released EST data, these authors incorporated numerous new species belonging to groups such as the choanoflagellates, closest relatives
of metazoans, the cnidarians, sister-group of the bilaterians, and a broad range of
nematodes and platyhelminthes, some of which exhibit slower evolutionary rates
(e.g. Trichinella for nematodes). The inclusion of those taxa made it possible to
limit the phenomenon of long branch attraction – in phylogenetic jargon, to “break
the long branches” – and thus produced support for the Ecdysozoa clade (Fig. 4.3).
Notably, the platyhelminthes and the nematodes sometimes clustered together in
this study, but the authors showed that the removal of fastest evolving genes enabled
the recovery of the “new view” topology, with platyhelminthes being grouped with
annelids and molluscs. This contradicts previous results, which estimated that gene
sampling alone, and not taxon sampling, had the ability to increase the phylogenetic
accuracy (Rokas and Carroll 2005).
4.4.1.2 Coelomata and the Interpretation of Rare Genomic Changes
The importance of dealing with accurate taxon sampling was also stressed in a
recent debate about a possible revival of the Coelomata hypothesis. The origin of
this debate was two studies that investigated two different qualitative genomic characters: “rare amino acid replacements” and patterns of intron conservation. Rare
amino acid changes have been defined as amino acid substitutions that take place
in a limited subset of taxa and that are caused by several nucleotide changes. Those
substitutions are thus associated with a low probability of homoplasy, and 34 such
positions were initially found to be in agreement with the Coelomata topology
(Rogozin et al. 2007). However, subsequent addition of the cnidarian N. vectensis as