I did not know of
the early mammal, Shenshou (Bi et al. 2014; Oxfordian, earliest Late Jurassic; Fig. 1), when producing an earlier blog you can read here that pushed back the origin of mammals to the Triassic. Little did I know, that hypothesis originated two years ago by the Bi et al. team describing their find (see below). On close examination, however, there were problems with that description that have bearing on their phylogenetic placement of Shenshou.
From the Bi et al. abstract
“The phylogeny of Allotheria, including Multituberculata and Haramiyida, remains unsolved and has generated contentious views on the origin and earliest evolution of mammals. Here we report three new species of a new clade, Euharamiyida, based on six well-preserved fossils from the Jurassic period of China. These fossils reveal many craniodental and postcranial features of euharamiyidans and clarify several ambiguous structures that are currently the topic of debate. Our phylogenetic analyses recognize Euharamiyida as the sister group of Multituberculata, and place Allotheria within the Mammalia. The phylogeny suggests that allotherian mammals evolved from a Late Triassic (approximately 208 million years ago) Haramiyavia-like ancestor and diversified into euharamiyidans and multituberculates with a cosmopolitan distribution, implying homologous acquisition of many craniodental and postcranial features in the two groups. Our findings also favour a Late Triassic origin of mammals in Laurasia and two independent detachment events of the middle ear bones during mammalian evolution.”
This blog post was prompted by a comment made by reader M. Mortimer, who responded to my challenge, ““Where among the pre-eutherians is there a better match for multituberculates”? by writing “Where they go in the consensus (e.g. “sister” to Shenshou in your terminology- Bi et al., 2014), based on more characters and taxa than you use.”
First of all
the challenge was to find a mismatched sister and a better match WITHIN the taxon list of the large reptile tree. Bringing in an outside taxon, does not answer the challenge. That Shenshou was already considered a member of the clade under question establishes that M. Mortimer clearly did not understand or chose to ignore the parameters of the challenge. Be that as it may…
The Bi et al. tree
nests Allotheria (haramiyids and triconodonts) traditionally between monotremes and spalacotheres + eutherians + megatherians. Bi et al. used 113 taxa and 495 characters, and several magnitudes more dental traits than used here. I found several mistakes made in the description of Shenshou, so we’ll start with those.
I was able to trace some elements of the skull using Photoshop (Fig. 2). Shenshou has the basic buck-toothed rodent appearance with a narrow lateral temporal bar (zygomatic arch) and broad skull (as determined by the triangular displaced post parietal. I was not able to recover the same posterior dentary shape that Bi et al. produced (Fig. 7). The one traced here is more traditional.
Scapula and coracoid issues
Bii et al. reported a scapula with a spine along the anterior ridge (as in pre-mammal tritylodontids like Oligokyphus and Kayentatherium). Unfortunately I place the ridge at mid scapula. They also report the presence of a small coracoid (Fig. 7), but here that is interpreted as a series of half-buried vertebral centra (Fig. 3). If present, these traits would tend to push Shenshou toward more primitive taxa, but here they only produced autapomorphies in an otherwise firm nesting.
Manus and pes
re-reconstructions. The manus was pretty easy to reconstruct (Fig. 4. As it was preserved almost intact. Not so the pes (Fig. 5), which was scattered about and the proximal portions of two metatarsals were lost during excavation. In any case, the long fingers and toes indicate that Shenshou was arboreal, as Bi et al. originally described it.
should be separated from the tibia (Fig. 7) as it is in the fossil (Fig. 1). The anterior caudals have shallow, but long centra that are not shown in the Bi et al. reconstruction (Fig. 7). The femur was drawn too long with regard to the spinal column and the pelvis was drawn to short. The ectotympanic ring was ignored.
In the large reptile tree
(subset Fig. 8) Shenshou nests within the Eutheria, basal to rodents and multituberculates + haramiyids and not far from the branch that produced the other arborealist, Plesiadapis, an ancestral arboreal rabbit with similar long fingers and toes. A comparison of Shenshou with Rattus, the rat (Fig. 6), shows the two taxa have much in common despite the minor mistakes made by Bi et al. in their attempt at fulfilling an erroneous paradigm that Shenshou should nest far from arboreal eutherian mammals.
If you can show that
Allotheria belongs outside of the Theria, please bring that data to my attention. So far, based on traits from all over the body and not so much the teeth, those rodent-like taxa are nesting with rodents.
Thanks, Mickey, for bringing this paper to my attention. Sorry it didn’t hold up as evidence for your negative POV. Next time, please bring up two taxa within the LRT.
Bi S, Wang Y-Q, Guan J, Sheng X and Meng J 2014. Three new Jurassic euharamiidan species reinforce early divergence of mammals. Nature 514:579-584. online here.