An attempt at reconstructing a phylogenetic tree of the Ciliophora using parsimony methods

Abstract

Summary Phylogenetic trees were constructed for the Ciliophora using a parsimony analysis that applies the Camin-Sokal method to characters of known polarity and the Wagner method (which requires no knowledge of the ancestral state) to the other characters. The data covered 56 species and 23 morphological, nuclear and ultrastructural multistate characters. Since no real-world outgroup can be assumed with certainty to root the ciliophoran tree, we used three hypothetical ancestor hypotheses; only one of them (hypothesis 3: somatic kinetosomes in pairs considered ancestral; no character transformation series assumed for the position of the buccal area or for the organization of the buccal infraciliature) produced interesting trees. Two trees, called A and D, have been retained because they were shorter than the others and were equally optimal for different codings of the hypothetical ancestor. In tree A, there is an early separation in two main branches. The first one contains two groups: the Karyorelictea-Heterotrichea (Postciliodesmatophora) and the Hypotrichea-Oligotrichea (Spirotricha) on the one hand, and the colpodids (Transversala) on the other. The second branch leads to 3 groups containing all other ciliates. In tree D, the Postciliodesmatophora and Spirotricha are first separated from all other ciliates; this is in agreement with molecular phylogenies. Despite these differences, the same five major groups appear in both trees; the main difference is in the position of the colpodid group. Class Karyorelictea appears to be polyphyletic, with (a) a Loxodia-Trachelocercia line whose genera share the same type of somatic cortex and nuclear organization, and (b) a Protoheterotrichia-Protocruziidia line which is closer to the Heterotrichia. Nyctotherus is closer to the hypotrichs than to the heterotrichs. As in the molecular trees, the heterotrichs are closer to some of the Karyorelictea, with which they share the same main type of cortical cytoskeleton (postciliary ribbons), than to the hypotrichs and oligotrichs, where the cortical microtubules are not postciliary fibers. So, there are two competing types of reinforcement of the cell cortex by microtubules, and these were selected as early as the first (in tree D) or the second branching (in tree A); this is justification enough to consider the subphylum Tubulicorticata as totally artificial. The validity of the subphylum Filocorticata is also discussed, considering the cortical cytoskeleton of some of the Vestibuliferea (Blepharocorythida and Entodiniomorphida). The Litostomatea, Vestibuliferea and Phyllopharyngea emerge as a sister-group of the Oligohymenophorea. In the phyllopharyngids, macronuclear DNA is gene-sized, as in the hypotrichs; this means that DNA fragmentation occurs independently in different lineages. Macronuclear characters concerning chromatin organization that depend on the size of the DNA molecules have become diversified into paraphyletic lines such as the phyllopharyngids, oligotrichs and hypotrichs for the character “DNA duplication in replication bands”. Nassula is separated from the Furgasonia-Pseudomicrothorax group, which is close to the scuticociliates. Nassula is close to Coleps . The peniculids branch away markedly from the tetrahymenids and are closer to the scuticociliates and peritrichs. The results are discussed with reference to some other new data, phylogenetic reconstructions and molecular trees.