multituberculates, like Late Jurassic Rugosodon (Figs. 1, 2), have been considered rodent-like mammals nesting near the origin of mammals.
In the pre-cladistic era Simpson 1945
nested multituberculates close to Prototheria.
In the cladistic era Novacek 1992 wrote:
“Neither molecular nor morphological probes have been successful in resolving major sectors of the mammalian tree. Both molecular and morphological results seem to converge on the idea that New World edentates (sloths, armadillos and anteaters) represent an early (perhaps even the earliest) branch in the placental mammal tree. Although this case is rather tenuous from an anatomical standpoint…” Certainly there are no marsupials or edentates that resemble one another.
in the large reptile tree (LRT, 1305 taxa) multituberculates nest as derived rodent carpolestids, despite their Jurassic antiquity. Among extant taxa, they are most closely related to today’s aye-aye (Daubentonia). Aye-ayes are not rodent-like primates, but living rodent carpolestids in the LRT. Even their tarsals are similar (Fig. 2). Earlier here and here we looked at related multituberculate, rodent and carpolestid skulls and teeth.
Based on LRT results, Novacek 1992 wrongly nested:
- All marsupials as a separate clade
- Marsupial Creodonta with placental Carnivora.
- Multituberculata apart from Rodentia, close to Monotremata
- Pholidota apart from Chiroptera
- Scandentia with Primates and Dermoptera
- Insectivora apart from Rodentia
- Cetacea as a monophyletic clade
- Cetacea with Artiodactylia and Tubulidentata
- Hippos with Artiodactyla
- Tubulidentata apart from Edentata
- Embrithopoda with Proboscidea and Desmostylia
- Novacek 1992 also omitted several relevant taxa, like Tenrec and Caluromys
Novacek did correctly nest
- Condylarthra and all hooved placentals as derived taxa.
- Monotremata was correctly nested as a basal clade.
- Palaeoryctoids, then including Asioryctes, were correctly nested between Metatheria and Eutheria.
With that introduction
Luo et al. 2016 studied the ankle of the multituberculate, Rugosodon, “the earliest-known postcranial fossil of a multituberculate mammal.“ They report, “Multituberculates as a group can now be diagnosed by derived features of the astragalus, the navicular, and the entocuneiform.” Those are the light red, bright green and yellow bones in figure 2. “These features are correlated with the mobility of the upper ankle joint and pedal digit I, indicating that early multituberculates acquired new locomotor functions of the limb and foot. However, the standing pedal posture of the basal multituberculates is plantigrade, typical of primitive mammaliaforms. The digitigrade posture appeared later in derived multituberculates.” (Like Kryptobaatar, a taxon with a digitigrade pes we looked at earlier here.)
Rugosodon has a large, plate-like parafibula
proximal to the fibula itself, but not co-osified. Luo et al. note most placentals do not have a parafibula, but Rattus and Daubentonia do. The parafibula is co-ossified to the fibula in prototherian mammals.
The Luo et al. paper does not include reference or comparisons to rodents, carpolestids or Daubentonia, yet another case of taxon exclusion revealed here. Clearly these taxa need to be added to multituberculate studies in the future. Paulchoffatia, mentioned in the Luo et al. title, nests basal to a sister clade to the multituberculates.
Luo Z-X, Meng Q-J, Liu D, Zhang Y-G and Yuan C-X 2016. Cruro-pedal structure of the paullchoffatiid Rugosodon eurasiaticus and evolution of the multituberculate ankle.Palaeontologia Polonica 67:149-169.
Novacek MJ 1992. Mammalian phylogeny: shaking the tree. Nature 356:121–135.
Simpson GG 1945. The principals of classification and a classification of mammals. Bulletin of the American Museum of Natural History 85:350pp.