Hou, Sukuan; Danowitz, Melinda; Sammis, John; Solounias, Nikos
Dead ossicones, and other characters describing Palaeotraginae (Giraffidae; Mammalia) based on new material from Gansu, Central China.
In: Zitteliana, Vol. B 32: S. 91-98
While the identity and validity of the extant families of ruminants are undoubted, there are significant problems with the determination of the interrelationships among the families, notably within the families of the Pecora, or horned ruminants. The morphological features used to construct ruminant phylogeny have been a source of controversy: many features used over the past century have been shown to be highly homoplastic and related to functional similarities. Ruminants evolved in the context of the later Cenozoic climatic changes, and many lineages adopted functional morphological adaptations related to feeding on more abrasive diets (resulting in the parallel evolution of a greater extent of loph development in the molars and, in some lineages, hypsodonty) and locomotion in open habitats (resulting in the parallel evolution of fused metapodials and reduction and/or loss of lateral digits). The fact that the molecular phylogeny shows a very different pattern from the currently accepted morphological one is of particular cause for concern, especially as molecular data are of no use for understanding the relationships of extinct lineages. Here we review the morphological data used in ruminant phylogenetics, and show even many of the less obviously functional features (e.g., number and position of the lacrimal orifices) are subject to homoplasy and variation, especially when fossil taxa are included. In addition, many morphological features treated as independent traits in phylogenetics are correlated (e.g., cranial morphology associated with hypsodonty). Some potentially reliable features are identified, but these do not help to sort out relationships within the Pecora. We advocate for the investigation into better morphological features, possibly derived from basicranial and ear region characters (although these features are not without their own issues of homoplasy), and for caution in character consideration in performing phylogenetic analyses.