ИСТИНА

Molecular analyses suggest that Enoplia is the most divergentmajor clade within the phylumNematoda. Disparities in rates of sequence evolution between lineages, lack of intermediate taxa to break possible long branches and the limited amount of phylogenetic information shared by sampled taxa make phylogenetic inference for Enoplia a difficult task. The dataset was carefully selected to maximally encompass enoplian biodiversity and utilise basal representatives of other clades, thus breaking long branches at all levels of the tree. It includes several genera for which molecular evidence has not been previously published: Campydora, Leptosomatum, Thoracostomopsis, Bathylaimus, Tripyloides, Calyptronema, Viscosia, Halalaimus, Ironus, Alaimus, Xenella, Rhabdodemania, Tobrilus, Aporcelaimellus and Nygolaimus. The data was examined for presence of phylogenetic signal using approaches based on sequential removal of categories of variable sites, with phylogeny reconstruction at each step. This case study proves that the NJ method performs highly inconsistently with this problematic dataset and that bootstrap analyses (both NJ and MP) greatly underestimate repeatability of inferring even basal nodes of the nematode tree. Instead, comparing equally parsimonious topologies obtained in a series of MP runs with a number of statistical tests under the ML optimality criterion allowed the reconstruction of the major lineages within Enoplia.