Abstract
The marine scuticociliate Paratetrahymena parawassi n. sp. is described on the basis of morphology, especially infraciliature, and the sequence of its small subunit (SSU) rRNA gene to become the second known member of its genus. Paratetrahymena and other ciliates in the order Loxocephalida possess a mixture of morphological and morphogenetic features characteristic of the subclasses Hymenostomatia and Scuticociliatia. Accordingly, we used SSU rRNA sequences to analyze the phylogeny of Paratetrahymena and three other loxocephalid genera. Paratetrahymena and Cardiostomatella vermiformis formed a moderately well-supported clade that diverged at a deep level from all other scuticociliates, supporting separation of loxocephalids from other scuticociliates as a suprafamilial taxon. Sathrophilus holtae was a sister taxon to Paratetrahymena and Cardiostomatella in a poorly supported, unresolved relationship; nevertheless, association of all three genera into a single clade was supported by an approximately unbiased (AU) test. Any association of these genera singly or as a group with the Hymenostomatia was rejected decisively by AU tests and by a complete absence in the loxocephalids of the unique nucleotide identities that distinguish hymenostomes. Therefore, the morphological and morphogenetic similarities of loxocephalids to hymenostomes may be plesiomorphies, and the conflicting mix of scuticociliate and hymenostome characteristics seen in loxocephalids may result from differing rates of character evolution. Dexiotrichides pangi and Urocentrum, which is currently classified as a peniculid, formed a small clade that associated with hymenostomes and peritrichs. Monophyly of the Loxocephalida with Dexiotrichides and/or Urocentrum included was not rejected by AU; however, inclusion of Urocentrum in the Peniculia was rejected by AU tests. A hypothesis is offered to explain the lack of resolution of loxocephalid ciliates and Urocentrum in phylogenetic trees, namely that their phylogenetic positions are influenced by a combination of heterogeneous data and long-branch attraction caused by poor representation of taxa in analyses. The well-known genus Cyclidium, a member of the order Pleuronematida, was revealed to be polyphyletic as a byproduct of our analyses of loxocephalids. In particular, Cyclidium porcatum appears to fall outside the clade containing typical members of the subclass Scuticociliatia and thus invites investigation as a possible member of the order Loxocephalida.