Figuring out the upside-down skull of Yanoconodon

Figure 1. Yanoconodon fossil in situ. See the skull in closeup in figure 2.

Figure 1. Yanoconodon fossil in situ. See the skull in closeup in figure 2. The published tracing is distorted here to match the underlying photo.

Wikipedia reports, “Yanoconodon was a small mammal, barely 5 inches (13 centimetres) long. It had a sprawling posture, Yanoconodon was a Eutriconodont, a group composing most taxa once classified as “triconodonts” which lived during the time of the dinosaurs. These were a highly ecologically diverse group, including large sized taxa such as Repenomamus that were able to eat small dinosaurs, the arboreal Jeholodens, the aerial volaticotherines and the spined Spinolestes. Yanoconodon is inferred to be a generalized terrestrial mammal, capable of multiple forms of locomotion.

Figure 1. Yanoconodon is exposed in ventral view. Even so, if you employ DGS, even on a fuzzy photo, you can put together a reconstruction that shares several traits with Repenomamus.

Figure 2. Yanoconodon is exposed in ventral view. Even so, if you employ DGS, even on a fuzzy photo, you can put together a reconstruction that shares several traits with Repenomamus.

Mammal-like reptiles?
Wikipedia also reports, “The Yanoconodon holotype is so well preserved that scientists were able to examine tiny bones of the middle ear. These are of particular interest because of their “transitional” state: Yanoconodon has fundamentally modern middle ear bones, but these are still attached to the jaw by an ossified Meckel’s cartilage. This is a feature retained from earlier stem mammals, and illustrates the transition from a basal tetrapod jaw and ear, to a mammalian one in which the middle ear bones are fully separate from the jaw. Despite this feature Yanoconodon is a true mammal. It is thought that the feature was retained during early embryo development,[4] whereas it is lost in most other mammal groups. The intermediate anatomy of the middle ear of Yanocodon is said to be a “Rosetta Stone”[5] of mammalian middle ear evolution.”

In the large reptile tree (LRT, 1037 taxa) Yanoconodon, Repenomamus, Jeholodens and Spinolestes are not mammals, but very close to the base of the Mammalia. Both clades share Pachygenelus as last common ancestor. So that means the ‘transitional state’ mentioned above is indeed outside the Mammalia. Other paleontologists consider this list of taxa to be mammals, but here the mammal-like traits they had were developed in parallel and not quite to mammal standards.

Figure 4. Repenomamus reconstructed using DGS methods. The manus and feet are loose figments at present. Despite its predatory nature, note the reduction in canines, a clade trait.

Figure 4. Repenomamus reconstructed using DGS methods. The manus and feet are loose figments at present. Despite its predatory nature, note the reduction in canines, a clade trait.

The skull of Yanoconodon
(Fig. 2) can be largely, but not completely, reconstructed based on the visible bones. The skull is low and wide and without the typical constriction anterior to the jugals. The anterior teeth are large and spike-like while the posterior teeth are molariform. Large teeth typically require deep roots and deep bones to house those roots. The mandibles are as long as the skull. The small orbits are far forward on the skull and the temporal fenestra are correspondingly large.

Figure 2. The origin and radiation of stem mammals and crown mammals. Compare the LRT tree (above) to a recent cladogram by Close et al. 2015.

Figure 2. The origin and radiation of stem mammals and crown mammals. Compare the LRT tree (above) to a recent cladogram by Close et al. 2015.

With the new data on Yanocondon
several taxa within the LRT shifted places, but not far and still within the derived Cynodontia. Something about the Mammalia helped them survive several extinction events that the derived Tritylodontia (= Pseudomammalia) succumbed to. Pseudomammalia LOOK like mammals, but are not mammals. They continued to exist into the Early Cretaceous and some, like Repenomamus, were quite large.

References
Close RA, Friedman M, Lloyd GT and Benson RBJ 2015. Evidence for a mid-Jurassic adaptive radiation in mammals. Current Biology. 25(16): 2137–2142. 
Luo Z, Chen P, Li G, and Chen M 2007.
 A new eutriconodont mammal and evolutionary development in early mammals. Nature 446:15. online Nature

wiki/Yanoconodon

Is Phascolotherium a basal mammal? Perhaps not…

Yesterday we looked at the origin of mammals and noted the Rowe 1988 considered the fossil mandible Phascolotherium bucklandi (Middle Jurassic; Owen 1838; Fig. 1) one of the earliest known mammals. Unfortunately, the mandible specimen does not have enough traits to nest Phascolotherium in the large reptile tree (LRT, 1011 taxa) with complete resolution.

Nevertheless,
that doesn’t stop one from visually comparing Phascolotherium to more complete taxa.

Figure 1. Phacolotherium compared to the tritylodontid Jeholodens.

Figure 1. Phacolotherium compared to the tritylodontid Jeholodens. The smaller Jeholodens image is to scale with the much larger Phacolotherium specimen.

There’s a pretty good match for Phascolotherium
with Jeholodens  (Ji et al. 1999); non-mammalian cynodont/tritylodontid/mammaliaform known from the Middle Cretaceous. The Jeholodens mandible is smaller than Phascolotherium, It has an unerupted 4th molar, which would indicate immaturity if it was a mammal. Only mammals do not replace molars.

Of Jeholodens
Ji et al. 1999 reported, “The postcranial skeleton of this new triconodont shows a mosaic of characters, including a primitive pelvic girdle and hindlimb but a very derived pectoral girdle that is closely comparable to those of derived therians. Given the basal position of this taxon in mammalian phylogeny, its derived pectoral girdle indicates that homoplasies (similarities resulting from independent evolution among unrelated lineages) are as common in the postcranial skeleton as they are in the skull and dentition in the evolution of Mesozoic mammals.”

There’s a postscript
Take another look at the mandible of Jeholodens (Fig. 1). Note the giant incisor 1 and the robust jaw articulation. Where else do we see this combination in small mammals? In Multituberculata and Haramiyidae, but both nest with rodents, plesiadapids and carpolesteids in the LRT. Traditional cladograms nest Multituberculata and Haramiyidae either before Mammalia or between monotremes and therians as very basal mammals. I have long wondered, if this was so, which basal pre-mammals or mammals look most like multituberculates and might therefore be most closely related? Could it be Jeholodens? So it was time for a test. Shifting the multituberculates to Jeholodens currently adds 34 steps to the LRT. Let’s see what happens when multis are re-scored with prejudice toward Jeholodens

So what happened?
The multis and haramiyids did not shift. The tree topology did not change. Apparently any resemblance between Jeholodens and these two clades must have been by convergence. Or the amount of convergence is overtaking the true relationship. Since everything in science is provisional we’ll keep testing.

References
Ji Q, Luo Z and Ji S 1999. A Chinese triconodont mammal and mosaic evolution of the mammalian skeleton. Nature 398:326-330. online.
Owen R 1838
. On the jaws of the Thylacotherium prevostii (Valenciennes) from Stonesfield. Proceedings of the Geological Society of London 3, 5–9.

Yanoconodon: Proximal sister to the Mammalia

Updated September 22, 2018
with the addition of pertinent taxa.

This post was composed several weeks ago.
After all the intervening excitement I’m glad to bring Yanoconodon to your attention.

Yanoconodon alllini (Luo, Chen, Li and Chen 2007; Early Cretaceous, 122 mya; 13 cm in length; Figs. 1, 2) is known from a nearly complete and articulated crushed fossil. It is traditionally considered a eutriconodont, a clade that traditionally includes Spinoletes, Repenomamus, GobiconodonLiaoconodon and Jeholodens. Unfortunately that clade is paraphyletic in the large reptile tree (LRT) because other traditional members nest outside this clade in the LRT. Here Yanoconodon nests with Maotherium in that clade (Fig. 3).

Yanoconodon had a semi-sprawling posture
and a a long, robust torso with an unusually thick lumbar vertebrae provided with very short ribs. The limbs were short. The canines were quite narrow. The posterior jaw bones were still attached to the jaw. They had not yet become completely reduced to middle ear bones and completely separated from the jaw bones. So, by definition and cladogram (Fig. 3), Yanoconodon was not a true mammal. Wikipedia disagrees as that author reports, “Despite this feature Yanoconodon is a true mammal.”

FIgure 1. Yanaconodon nests as the proximal outgroup to the Mammalia in the LRT.

FIgure 1. Yanoconodon nests as the proximal outgroup to the Mammalia in the LRT. Even so it has several autapomorphies (differences from the actual  hypothetical ancestor.)

from the Luo, Chen and Chen abstract
“Detachment of the three tiny middle ear bones from the reptilian mandible is an important innovation of modern mammals. Here we describe a Mesozoic eutriconodont nested within crown mammals (1) that clearly illustrates this transition: the middle ear bones are connected to the mandible via an ossified Meckel’s cartilage. The connected ear and jaw structure is similar to the embryonic pattern in modern monotremes (egg-laying mammals) and placental mammals, but is a paedomorphic feature retained in the adult, unlike in monotreme and placental adults. This suggests that reversal to (or retention of) this premammalian ancestral condition is correlated with different developmental timing (heterochrony) in eutriconodonts. (2) This new eutriconodont adds to the evidence of homoplasy of vertebral characters in the thoraco-lumbar transition and unfused lumbar ribs among early mammals. (3) This is similar to the effect of homeobox gene patterning of vertebrae in modern mammals, making it plausible to extrapolate the effects of Hox gene patterning to account for homoplastic evolution of vertebral characters in early mammals.” (4)

Notes

  1. The LRT nests Yanoconodon just outside the crown mammals. Not sure why the authors say this, given what they report about the posterior jaw bones as posterior jaw bones.
  2. Curious that the retention of “this pre-mammalian ancestral condition” does not indicate to the authors that Yanoconodon is indeed a pre-mammal.
  3. Yanoconodon does not nest as an early mammal in the LRT.
  4. …or…not, if Yanoconodon is indeed a non-mammalian trithelodont. Other non-mammalian cynodonts lived alongside Jurassic mammals. Only one purported eutriconodont listed above is a mammal, Volaticotherium. It nests as a basal placental. Triconodon is a mammal, too, a monotreme known from just a dentary and teeth.
Figure 2. From Luo et al. the posterior jaw bones of Yanoconodon. These are not middle ear bones, so Yanoconodon is not a mammal.

Figure 2. From Luo et al. the posterior jaw bones of Yanoconodon. These are not middle ear bones, so Yanoconodon is not a mammal. The malleus is the articular. The incus is quadrate.

Yanoconodon is a great transitional fossil.
You can’t call it a mammal, because it nests outside the last common ancestor of all mammals.

Figure 3. Basal mammals begin with Ornithorynchus, the most primitive living mammal. Yanoconodon nests just outside this clade.

Figure 3. Basal mammals begin with Ornithorynchus, the most primitive living mammal. Yanoconodon nests just outside this clade.

References
Luo Z, Chen P, Li G, and Chen M 2007. A new eutriconodont mammal and evolutionary development in early mammals. Nature 446:15. online Nature

wiki/Yanoconodon

Repenomamus is not a mammal, but close…

Everyone boggled this one
including yours truly. Repenomamus (Li et al. 2001, Hu et al. 2005; Yixian Formation, Early Cretaceous; IVPP V 12549) never really found any ‘soul mates’ within the Mammalia within the large reptile tree (LRT). For one thing, mammals tend to have larger brains reflected in larger crania. Carnivorous mammals tend to not shrink the canines and they tend to lose the septomaxilla, but Repenomamus retains it. More below.

Figure 1. Three specimens of Repenomamus to scale with bones colorized.

Figure 1. Three specimens of Repenomamus to scale with bones colorized.

Now, in the LRT
Repenomamus nests with Pachgenelus and the Tritylodontids, Oligokyphus and Kayentatherium and this is a good match. Repenomamus represents the carnivorous half of this clade. O and K were herbivores. We can only assume (prior to analysis) that other gobiconodontids like Gobiconodon, are going to also nest with them. Pachygenelus had a reduced canine. These stem-mammals have always been tricky to work with.

Figure 1. Basal mammals and therapsid outgroups nest around Repenomamus in the proximal clade outside the Mammalia.

Figure 1. Basal mammals and therapsid outgroups nest around Repenomamus in the proximal clade outside the Mammalia.

 

Originally, Hu et al.
recognized the primitive and derived traits in Repenomamus. That’s why they named it Repen “reptile” + mam “mammal” as they note, “referring to the the animal’s both reptilian and mammalian features, and its large size.” 

He et al noted several derived (mammalian) characteristics:

  1. well-developed dentary/squamosal articulation
  2. reduced number of teeth
  3. differentiation of premolars and molars
  4. presence of dorsal process of the premaxilla not in contact with nasal
  5. closed medial wall of the orbit
  6. fingerlike promontorium on the petrosal
  7. structure of the cheek teeth is simple
  8. differentiation of cusps on upper molars is indistinct

Hu et al also noted
Repenomamus retained some primitive (non-mammalian) features, but did not list them in the abstract, but did list them in the text:

  1. large size, relative to other Early Cretaceous mammals
  2. upper dental formula 3.1.2.4
  3. upper canine subequal to third upper incisor
  4. several dental characters
  5. large septomaxilla
  6. sagittal crest short and low
  7. lambdoid (posterior) crests well-developed
  8. sloping occipital plate in dorsal view
  9. post dentary bar retained
  10. dentary lacks angular process
  11. most of these traits are common to tritylodontids, not mammals of the time.

I note 

  1. the scapula has a short spine along the anterior rim.
  2. The arc-like coracoid is retained
  3. the cranium is very narrow
  4. the lateral temporal arch is robust, as in tritylodontids
Figure 3. Repenomamus giganticus in situ. Here the overlooked coracoids are identified.

Figure 3. Repenomamus giganticus in situ. Here the overlooked coracoids are identified.

Tritylodontids
were proximal outgroups to the mammals in the Triassic. And they appear to have developed in parallel to mammals in several traits as they continued into the Jurassic and Early Cretaceous. This parallel development appears to be the source of the present confusion.

Figure 5. Repenomamus giganticus reconstructed.

Figure 5. Repenomamus giganticus reconstructed.

And once again,
a phylogenetic analysis that was not made for mammals and has too many taxa per character was able to lump and split taxa more accurately than larger studies — all without ever seeing the specimen first hand. Sometimes it just takes a long list of candidate taxa to provide adequate opportunities to make that happen.

I’ll update the pages in ReptileEvolution.com to reflect those changes shortly.

References
Hu Y, Meng J, Wang Y-Q and Li C-K 2005. Large Mesozoic mammals fed on young dinosaurs. Nature 433:149-152.
Li J-L, Wang Y, Wang Y-Q and Li C-K 2001. A new family of primitive mammal from the Mesozoic of western Liaoning, China. Chinese Science Bulletin 46(9):782-785.

wiki/Repenomamus