MORPHOLOGY, ADAPTATIONS, AND RELATIONSHIPS OF PLESIORYCTEROPUS, AND ADIAGNOSIS OF A NEW ORDER OF EUTHERIAN MAMMALS

Plesiorycteropus, an extinct mammal known only from the Quaternary of Madagascar, is conventionally regarded as a tubulidentate and therefor e as closely related to extant Orycteropus. However, the shared derive d traits that might support such an association have never been adequa tely identified or critically evaluated. The character analyses presen ted in this paper reveal that many of the identifiably derived traits of the skeleton of Plesiorycteropus are related to adaptations for dig ging. Aardvarks possess many of the same adaptations, but so do fossor ial members of a broad diversity of other eutherian groups, including Dasypodidae, Manidae, Myrmecophagidae, Lipotyphla, and many others. Id entifiably derived traits of Plesiorycteropus that have no obvious con nection with digging are few, but the ones that can be adequately docu mented are by no means unique to aardvarks. Indeed, several of the app arently derived cranial and postcranial traits of Plesiorycteropus spe cifically echo conditions encountered in primitive ungulates, includin g various members of the paraphyletic assemblage Condylarthra. Accordi ngly, the view that Plesiorycteropus is unambiguously aardvarklike in its morphology and adaptations is not supported in this study. To exam ine how a parsimony analysis of a stated character set might specify a placement for Plesiorycteropus, a 30-character, 16-taxon data matrix was formatted for the program Phylogenetic Analysis Using Parsimony. T wo additional characters, based on morphological assessments of key ch aracters made by Bryan Patterson, were also used in some runs. Althoug h the scale of morphological variation in Plesiorycleropus requires th e recognition of two species, P. madagascariensis and P. germainepette rae n. sp., for the characters under consideration interspecific polym orphism was usually found to be negligible. Although in most manipulat ions of the data matrix Plesiorycteropus tended to group with ungulate s sensu lato (including Tubulidentata), its placement was unstable, an d an exclusive sister-group pairing of Plesiorycteropus + Tubulidentat a was rarely encountered. On the other hand, close pairings with xenar thrans, manids, and lipotyphlans did not occur unless the data matrix was purposely biased in those directions. As a minimum hypothesis, it may be concluded that Plesiorycteropus is apparently part of the great ungulate ''bush,'' but a more exact placement is not convincingly pro vided by any of the cladistic solutions investigated. One resolution o f this problem would be to refer Plesiorycteropus to superorder Ungula ta as incertae sedis, but this would make it the only Recent mammal la cking a recognized ordinal affiliation. An alternative would be to con sider Plesiorycteropus to be the sole known member of its own order. T his is the resolution preferred here, on the ground that Plesioryctero pus is as morphologically distinctive as any eutherian group currently granted ordinal status. The new order Bibymalagasia is created for it s reception.