We present a molecular phylogeny of the Proteocephalidea based on 28S rDNA sequence data that is a follow-up to the paper by Zehnder & Mariaux (1999). Twenty-three new sequences, including three outgroups are added in our new data-set. The Gangesiinae Mola, 1929 and the Acanthotaeniinae Freze, 1963 appear to be the most primitive clades. They are followed by a robust clade comprising the Palaearctic Proteocephalinae Mola, 1929 from freshwater fishes. The structure of the more derived clades, comprising most Neotropical and Nearctic species, is less resolved. At the nomenclatural level, we erect a new genus, Glanitaenia n. g. for G. osculata (Goeze, 1782) n. comb., previously Proteocephalus osculatus, and define an aggregate for the Palaearctic Proteocephalus Weinland, 1858. After a re-examination of all of the studied taxa, we identify two types of uterine development and show the importance of this character for the systematics of the order. Our phylogeny does not support the classical view of a Neotropical origin for the Proteocephalidea but rather favours an Old World origin of the group either in saurians or Palaearctic Siluriformes.

Cladistic analysis based on comparative morphology was used to examine the subfamily-level relationships within the cestode order Proteocephalidea. A single most parsimonious tree (70 steps, CI = 0.571; RC = 0.295; HI = 0.471) is consistent with monophyly for the Proteocephalidea and showed a relatively high consistency at the family level with the diagnosis of two major subclades.

We describe a new species of Nomimoscolex from the Amazon siluriform fishes Brachyplatystoma filamentosum, B. flavicans and B. vaillanti. It differs from N. piraeeba in a lower mean number of testes, the paramuscular position of the vitelline follicles, the ovarian width/proglottis width ratio and the cirrus-pouch length/proglottis width ratio. Protein electrophoresis assays performed for 25 enzymatic systems showed that specimens of N. suspectus n. sp. from the three host species form a homogenous population which was genetically isolated from N. piraeeba and N. dorad. Moreover, the latter two species, synonymised by Rego (1991) because of their close morphological similarity, could be separated at eight loci. We thus restore N. dorad as a valid species. We finally examined the composition of the genus Nomimoscolex using DNA sequences from the 5.8S rRNA, ITS-2 and 28S rRNA nuclear ribosomal genes and a matrix of 24 morphological characters. Phylogenetic relationships were inferred for nine species of the genus, five members of other monticelliid genera and two outgroup species. The results of the phylogenetic analyses performed on morphological and molecular characters converged with those from allozyme studies and showed that N. suspectus, N. piraeeba and N. dorad clustered in a distinct clade that excluded other members of the genus. We therefore recognised them as an aggregate of species to reflect an isolation supported by both morphological and genetic data. Because relationships among the remaining Nomimoscolex representatives and other genera were generally poorly resolved, regardless of the database analysed, no action was taken to reorganise them into alternative groupings.

No abstract