Eritherium: ‘most primitive and smallest known proboscidian ancestor’

Gheerbrant 2009
announced the discovery of Eritherium azzouzorum (late Paleocene; Morocco; Fig. 1) as the incomplete remains of the ‘most primitive and smallest known proboscidian’ and as the ‘oldest modern ungulate related to the elephant order.’

2014 brought us an older, but overlooked, proboscidian ancestor.
Radinskya yupingae (McKenna et al. 1989, Holbrook 2014; IVPP V-5255; Middle Paleocene; China; Fig. 1) was originally considered a perissodactyl-like herbivore, but never with confidence. Here Radinskya nests between Ectocion and Procavia, at the base of the elephant clade. Note the tiny procumbent second incisor. That’s the germ of what will become a giant tusk in elephants, mammoths and mastodons… and hyraxes (see Fig. 1). As in Procavia the tabulars contact one another medially, posterior to the central postparietal. The premolar is almost molar-like in appearance.

Gheerbrant’s cladograms
(Fig. 2) suffer from taxon exclusion and inappropriate taxon inclusion, as noted therein. The large reptile tree (LRT, 1346 taxa) nests elephants apart from each of the ungulate orders, which nest apart from each other. Artiodactyla (Owen 1848) is the basal clade, so all are artiodactyls in an expanded cladistic sense. Ungulata was coined earlier (Linneaus 1766), so there is that to consider. Hippos nest with Mesonyx outside this clade, so Mesonychidae is basal to all these clades. Phenacodontidae (Phenacodus is the last common ancestor in the LRT) is more basal still.

There is an old tradition
embraced by Gheerbrant and not confirmed by the LRT that nests embrithopods (like Arsinoitherium), anthracobunids and desmostylians (both mesonychids) with hyraxes, manatees and elephants (Fig. 2), but see the large reptile tree (LRT, 1346 taxa, subset Fig. 3). This tree topology continues unchanged even though more taxa are added every week. This is a measure of its strength.

Unfortunately,
there is not enough material to add Eritherium to the LRT, but placing Notostylops (not mentioned by Gheerbrant) over what little remains of Eritherium is a pretty good match (Fig. 1), except for the narrowness of the skull and much smaller teeth, as also seen in Radinskya, a taxon that is a little older and just as small as Eritherium.

Traditionally
Notostylops has been considered a notoungulate, which is why it is almost never included in elephant studies, as we learned earlier here.

References
Gheerbrant E 2009. Paleocene emergence of elephant relatives and the rapid radiation of African ungulates. DOI: 10.1073/pnas.0900251106 http://www.pnas.org/content/106/26/10717. PDF
Holbrook LT 2014. On the Skull of Radinskya (Mammalia) and Its Phylogenetic Position. Journal of Vertebrate Paleontology34(5):1203–1215.
McKenna MC, Chow M, Ting S and Luo Z 1989. Radinskya yupingae, a perissodactyl-like mammal from the late Paleocene of China. In Prothero DR, Schoch RM The evolution of perissodactyls. Oxford monographs on geology and geophysics. 15. New York: Oxford University Press. pp. 24–36.

wiki/Radinskya
wiki/Eritherium

NatGeo PR
Science Daily PR