Jeholodens and Spinolestes: two new tritylodontids

Revised October 4, 2016 with a shifting of Jeholodens and Spinolestes to the Tritylodontidae, which are pre-mammals arising from Pachygenelus. Tritylodontids replace their molars, something mammals do not do.

The Early Cretaceoous
included the first radiation of basal mammals. The vast majority of these, so far, have been multituberculates, small rodent-like taxa that actually nest with rodents in the large reptile tree. Traditional paleontologists nest multituberculates much more primitively, prior to the Theria (live-bearing mammals) despite their many rodent-like traits, like enlarged incisors followed by a diastema (toothless region) and flat cranial region.

With so many multituberculates
in the Cretaceous, I’m always looking for non-multituberculate mammals from the era.  Repenomamus was a tritylodont, pre-mammal. Vincelestes was a marsupial. Maotherium was an Early Cretaceous tritylodont. Liaoconodon was also a tritylodont.  I had high hopes that the next two Early Cretaceous mammals were, as advertised (i.e. something other than multituberculates.)

Another traditional clade of Early Cretaceous mammals
are the eutriconodonts. These include taxa with traditional tooth arcades, but lacking deep canines. Repenomamus is one traditional eutriconodont mammal, but nests in the large reptile tree with Pachygenelus in the tritylodontids. Spinolestes (Fig. 1) and Jeholodens (Fig. 2) are also listed at eutriconodonts, but they nest together in the large reptile tree both as sisters and as derived tritylodontids.

Figure 1. Spinolestes with bones colorized in DGS and both manus and skull reconstructed.

Figure 1. Spinolestes with bones colorized in DGS and both manus and skull reconstructed. Note the tooth pattern recovered here is different than as originally described. If your screen is 72 dpi, then this image is about half again as large as life size.

Spinolestes
has accessory neural articulations, like some shrews do. It also appears to have two sacrals and is otherwise robust overall. The entire foot is present, but scattered. I attempted a reconstruction of the lateral view of the skull based on published clues (Fig. 1).

Figure 2. Jeholodens holotype. Note the tip of the snout is missing, and so are the large anterior premaxillary teeth that characterize this clade.

Figure 2. Jeholodens holotype. Note the tip of the snout is missing, and so are the  anterior premaxillary teeth. If your screen is 72 dpi, then this image is almost twice as large as life size.

Jeholodens jenkinsi
(Ji et al. 1999) was also considered a triconodont, but the tip of the snout is missing. Not as robust as Spinolestes, Jeholodens (Fig. 2) was nevertheless a flat-bodied specimen able to slip into rock cracks. Wikipedia reports the eye was 5 cm across. The true figure is 5 mm.

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.”

The present analysis indicates
that Jeholodens nests with pre-mammal tritylodontids. The naris and small premaxillary teeth provided in the original reconstruction are imagined because that part is broken off the matrix (Fig. 2).

Be wary when you see
the terms ‘mosaic’ and ‘modular’. As we’ve seen before, that usually means the phylogeny is off. Evolution works by a gradual accumulation of traits all over the body, not in a modular or mosaic fashion. Trityodontids look like mammals because they are the proximal outgroup taxon. Multituberculates nest with rodents, with whom they share so many traits.

Figure 3. The tarsus of Jeholodens compared to that of a cynodont, mutituberculate and Didelphis, a marsupial (metatherian).

Figure 3. The tarsus of Jeholodens compared to that of a cynodont, mutituberculate and Didelphis, a marsupial (metatherian). Note that metatarsal 5 is missing in all taxa. Metatarsal 4 becomes a dual metatarsal in Didelphis and most eutherians, but not in Vulpavus, Onychonycteris,

 

Figure 4. Rattus pes. Note distal tarsal 4 only backs up pedal digit 4 and digit 5 rides alongside.

Figure 4. Rattus pes. Note distal tarsal 4 only backs up pedal digit 4 and digit 5 rides alongside.

 

 

 

Using Didelphis for a derived tarsus is a little misleading…
From the Ji et al. 1999 paper (Fig. 3) it looks like Jeholodens has a basal tarsus because distal tarsal 4 is not wide enough to double as a distal tarsal 5, as it does in the marsupial, Didelphis. A quick peek at Rattus (Fig. 4), Vulpavus and Onychonycteris shows that these placental taxa likewise do not widen distal tarsal 4 to back up pedal digit 5. And it is not clear how the Jeholodens tarsus actually stacks up. (Fig. 5). If the tarsus was loose, as it is in bats like Onychonycteris or Pteropus, then it doesn’t necessarily mean the tarsus was primitive, like that of a cynodont. That’s why the overall scores are more important than individual character scores.

Just because something is published in Nature or Science, doesn’t mean it’s necessarily right, as we’ve seen before with Yi, Cartorhynchus, Sclerocormus, Chilesaurus, dinosaur origins. pterosaur origins and turtle origins.

Figure 5. Tarsus of Jeholodens with elements colorized as in other taxa.

Figure 5. Tarsus of Jeholodens with elements colorized as in other taxa.

And while I’m thinking about it,
it may be that clades like “Triconodonta” and “Eutriconodonta” may be junior synonyms for long established taxa, as we looked at earlier here.

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.

Martin T et al. 2015. A Cretaceous eutriconodont and integument evolution of early mammals. Nature 526:380-384. online.

 

Advertisements

The origin of the amniote astragalus – Piñeiro et al. 2016

A new PeerJ paper
by Piñeiro, et al. (2016) attempts to shed light on the origin of the amniote astragalus by comparison to the ontogenetic development of the mesosaur tarsus. They claim that Mesosaurus is a very primitive amniote, following the thinking of traditional paleontologists.  By contrast, the  large reptile tree nests mesosaurs with highly derived thalattosaurs and ichthyosaurs, following basal pachypleurosaurs (sauropterygians) in the revived clade Enaliosauria, derived from marine younginiforms, diapsids, prodiapsids and basal archosauromorphs, all arising during the preceding Carboniferous.

In any case
The new paper seeks to answer the question, ‘Did the astragalus arise from one bone (the intermedium) or the fusion of several bones, including the tibiale, centralia and intermedium?’ Piñeiro et al. found that among embryo mesosaurs, a single four-part bone, creates the astragalus.

Two false paradigms affect the Piñeiro et al study 
1: They follow traditional beliefs that amniote skeletal structures should be present in basalmost amniotes. In reality, as we all know, amniotes are defined ONLY by the way they protect their embryos, with an amniotic sac, a structure lacking in amphibians. As we learned earlier, the basalmost amniote in the large reptile tree is Gephyrostegus bohemicus, a late-surviving member of a an earlier Viséan radiation. It has no traditional amniote traits. But a revised list of amniote traits can be seen here and in the six blogs that follow.

2: They believe that Mesosaurus is a basal amniote. The Early Permian is indeed early, but with basal reptiles already diversifying in the Viséan, some 40 million years earlier, the Permian is not early enough. Moreover a basal nesting is not supported in the large reptile tree. Along the same lines they do not understand that Diadectes, Tseajaia, Westlothiana and others nest within the Amniota (= Reptilia) in the only study that tests their relationships in a large gamut study of other tetrapods and amniotes, the large reptile tree.

Piñeiro et al. used an interesting graphic technique
of presenting the tarsal elements of several taxa as a series of interlocking hexagons. That’s fine on one level, but does not let us see the actual elements or reconstructions of the same, which is an unfortunate loss. They also mix up left and right, dorsal and ventral views of skeletal elements. It would be helpful to flip certain elements in order to present all the elements consistency for ready comparison. This should be standard operating procedure.

In the large reptile tree the intermedium remains a separate element
from the tibiae in stem (pre) amniotes like Proterogyrinus, Seymouria and the basal amniote Gephyrostegus (Fig. 1).

Figure 1. The intermedium remains separate from the tibiale in Proterogyrinus, Seymouria and Gephyrostegus.

Figure 1. The intermedium remains separate from the tibiale in Proterogyrinus, Seymouria and Gephyrostegus.

BTW, while researching Seymouria
I came across a bizarre reconstruction in Berman et al. 2000 (Fig. 1 under red circle) that did not match the bone tracings and added another row of central tarsals that no other tetrapods have.

Essentially
the Reptilia is, in reality, two clades, the Lepidosauromorpha and Archosauromorpha. Let’s take the former first. The tibiale does not fuse to the intermedium in all reptiles. So the astragalus is not present in every amniote.

Figure 2. Sample lepidosauromorph tarsi compared to Gephyrostegus. Here are Captorhinus, Emeroleter and Tjubina, a basal tritosaur lepidosaur.

Figure 2. Sample lepidosauromorph tarsi compared to Gephyrostegus. Here are Captorhinus,Orobates, Emeroleter and Tjubina, a basal tritosaur lepidosaur. Note the separate tibiae in Gephyrostegus, Orobates and Emeroleter. So an astragalus appears most of the time in this clade, not all of the time.

The appearance
and or fusion of tarsal elements varies within the Reptilia. And it varies with ontogeny within certain taxa, like mesosaurs. Older individuals often have more bones and more sharply defined bones. In the Archosauromorpha (Fig. 2), perhaps eight taxa precede the first appearance of the astragalus (fusion of tibiae and intermedium) in Casineria.

Figure 2. Comparison of archosauromorph tarsi, including Mesosaurus, the latter from Piñeiro et al

Figure 2. Comparison of archosauromorph tarsi, including Mesosaurus, the latter from Piñeiro et al Not to scale. Note the generally conservative pattern here, despite the liberal changes in relative bone sizes.

Here
(Figs. 1, 2) the astragalus (yellow/orange element) is only composed of the tibiale and intermedium in these taxa and a small perforation marks the division. The elements of the centralia may fuse together, but not with other elements in the above listed taxa. These fusion patterns occur by convergence in the two basal reptile clades.

According to Piñeiro et al
the astragalus changes greatly during the ontogeny of Mesosaurus. They interpret (“with doubts”) the embryo astragalus as the fusion of the tibiae, intermedum and two centralia. I don’t see any more than three centralia in the above illustrated taxa, and sometimes they fuse together, but not with the tibiale or intermedium (Fig. 2).

As a final note
I find it odd that workers are eager to change the names of some fused bones, like the astragalus and navicular, but are not interested in renaming other fused bones (like the postfrontal + postorbital). Instead, one bone is typically said to be present while the other is said to be absent. And that doesn’t make sense when both are present, just fused.

Let’s fix that in consensus.

References
Berman DS, Henrici AC, Sumida SS, Martens T. 2000. Redescription of Seymouria
sanjuanensis (Seymouriamorpha) from the Lower Permian of Germany based on complete mature specimens with a discussion of paleoecology of the Bromacker locality assemblage. Journal of Vertebrate Paleontology 20(2):253268
Piñeiro et al. 2016. The ontogenetic transformation of the mesosaurid tarsus: a contribution to the origin of the primitive amniotic astragalus. PeerJ 4:e2036; DOI 10.7717/peerj.2036

Pterosaurs Tarsals – More Evidence vs Padian 1983

Some pterosaurs (like Rhamphorhynchus and the new Painten pterosaur) had 4 or 5 tarsals. Others had only two (like Pteranodon, Figs. 1-3).

Figure 1. Pteranodon tarsals (in color). Blue = astragalus. Yellow + calcaneum.

Figure 1. Pteranodon tarsals (in color). Blue = astragalus. Yellow + calcaneum. YPM = Yale Peabody

The question is: 
In those pterosaurs with two tarsals is it more parsimonious that the 1) distal tarsals disappeared? or 2) the distal tarsals fused to the proximal tarsals? or 3) converging with birds, did the proximal tarsals fuse seamlessly to the tibia/fibula?

What does the evidence indicate?

There are pterosaur workers (Padian 1983, Bennett 2001, Nesbitt 2011, Witton 2013) who consider the tibia + fibula of pterosaurs a “tibiotarsus” because they say the proximal tarsals (astragalus + calcaneum) fused seamlessly to the distal tibia/fibula (Fig. 1). (We looked at this earlier here.) Birds have this sort of tibiotarsus. Padian 1983 compared bird tibiotarsi to Dimorphodon (Fig. 2) and the case looked pretty good back then.

However,
It’s important to remember that birds had a long ancestry as dinosaurs with distinct ascending processes of the astragalus that ultimately fused seamlessly to the tibia after the miniaturization that preceded and succeeded Archaeopteryx. Pterosaurs don’t have that long history, nor do they have ancestors with an ascending processes, nor did they undergo phylogenetic miniaturization prior to getting their wings. Even Archaeopteryx has a distinct ascending process — not seamless.

Under the Padian 1983 hypothesis 
the two tarsals found with Dimorphodon are distal tarsals. Likewise, Bennett (2001) proposed a tibiotarsus for Pteranodon. Eaton (1913, Fig. 1) called them podials, a general name form carpals or tarsals. We don’t see the same long ancestry progress in pterosaur ankles. In fact, there’s no ancestry for this type of ankle at all.

Figure 1. Pterosaur distal tibia. Left: Dimorphodon. Right Pteranodon.

Figure 2. Pterosaur distal tibia. Left: Dimorphodon. Right Pteranodon in anterior (above) and posterior (below) views. Padian (1983) and Bennett (2001) consider the bulbous parts to be the fused proximal tarsals. They are not. The proximal tarsals, astragalus (blue) and calcaneum (yellow) are distinct. Missing here are any distal tarsals. Padian identified this view of Dimorphodon as the anterior, because it looked so much like the anterior of the distal bird tibiotarsus (not shown here). But look again. It looks more like the posterior of the distal tibia of Pteranodon identified by Bennett.

Figure 4. Foot and tarsus of Pteranodon, FHSM-P-2062 and restored and relabeled. From OceansofKansas.com.

Figure 3. Foot and tarsus of Pteranodon, FHSM-P-2062 and restored and relabeled on top, from original online mislabeled image found at OceansofKansas.com (below). Note, the distal tibia bulge is posterior in Pteranodon, but bulges both ways in Dimorphodon and other pterosaurs, like the Painten pterosaur.

Rather, when you look at basal pterosaurs like Peteinosaurus (Fig. 4), you find four distinct tarsals.

Figure 4. Peteinosaurus and Dimorphodon BMNH4212 pedes. Four tarsals are present on both.

Figure 4. Peteinosaurus and Dimorphodon BMNH4212 pedes. Four tarsals are present on both. Yes the tarsals have moved in Dimorphodon with the distal tarsals rising to the level of the proximal tarsals. 

Same with the classic specimen of Dimorphodon. The engraving (Fig. 5) shows four tarsals.

Figure 6. Click to enlarge. The four tarsals identified on the the classic BMNH 41212 specimen of Dimorphodon.

Figure 5. Click to enlarge. The four tarsals identified on the the classic BMNH 41212 specimen of Dimorphodon. Non-foot bones are ghosted out. Calcaneum = yellow. Astragalus = blue. Distal tarsal 4 = pink. Centrale = magenta. Yes, they have moved during taphonomy, If you count four tarsals, that’s all I’m asking for now.

This is in contrast to Padian’s (1983) interpretation of BMNH 41212 (Fig. 6) where he adds a cylindrical joint to the distal tibia with a circumference smaller than in the other tibia at left.

Figure 6. Tarsals of Dimorphodon BMNH 41212 specimen according to Padian 1983. Figure 5 doesn't match.

Figure 6. Tarsals of Dimorphodon BMNH 41212 specimen according to Padian 1983. Figure 5 matches in most regards — except for the tarsals.

Padian 1983 removed tarsals from the matrix of two far less complete specimens attributed to Dimorphodon, YPM 350 and YPM 9182 (Figs. 7-9). Oddly, the smaller of the two specimens (YPM 9182) fused the two large tarsals to one another (the only such event I am aware of). The larger specimen (YPM 350) did not.

Figure 7. The YPM 350 specimen attributed to Dimorphodon. Note the tarsals fuse to one another despite the smaller size. The femora do not match, though similar in most regards. So there is some doubt that this is indeed Dimorphodon.

Figure 7. The YPM 9182 specimen attributed to Dimorphodon. Note the tarsals fuse to one another despite the smaller size. The femora do not match. The ventral maxilla is straighter. The jugal is deeper. M4.2 is shorter.  So there is some doubt that this is indeed congeneric with Dimorphodon. The question here is: did the calcaneum fuse to the fourth distal tarsal? And if so, did Padian get his tarsal backwards? With Padian’s orientation the tarsal has a posterior tuber. But no pterosaur ever developed a tuber, certainly not on any distal tarsals. And not on any calcaneum either. Let’s keep an eye out for further examples of this. 

Figure 8. About the size of the classic Dimorphodon, the YPM 350 specimen has unfused tarsals. Note the very few bones. The specimen is extremely disarticulated. The other two tarsals could have been easily scattered.

Figure 8. About the size of the classic Dimorphodon, the YPM 350 specimen has unfused tarsals. Note the very few bones. The specimen is extremely disarticulated. The other two tarsals could have been easily scattered. This specimen appears to be closer to the classic Dimorphodon in all regards.

Figure 9. Location of the tarsals (red circles) on the YPM 350 and YPM 9182 specimens attributed to Dimorphodon by Padian 1983. Do you think some other tarsals could have escaped?

Figure 9. Location of the tarsals (red circles) on the YPM 350 and YPM 9182 specimens attributed to Dimorphodon by Padian 1983. Do you think some other tarsals could have escaped?

Padian 1983 noted the cylindrical shape of the distal tarsals and their convergence with the bird tibiotarsus. But there are pterosaurs, like the Painten pterosaur (Fig. 10), that have a cylindrical distal tibia AND four tarsals.

Figure 10. The Painten pterosaur with tarsals colorized. There are four of them. Note the cylindrical shape of the distal tibia/fibula.

Figure 10. The Painten pterosaur with tarsals colorized. There are four of them. Note the cylindrical shape of the distal tibia/fibula.

So, the evidence for Dimorphodon having only two tarsals is fading. The evidence for cylindrical distal tarsals is strong. Pteranodon has only two tarsals. Whether they were created by fusion or reduction awaits further evidence. There is no evidence for a gradual evolution of fusion in the tarsals and tibia/fibula. Rather, there is plenty of evidence for the retention of paired distal and paired proximal tarsals. There is also evidence in YPM 9182 for the fusion of the proximal tarsals in certain pterosaurs.

Ramifications
Nesbitt 2011 fell prey to the idea of a fused tibiotarsus in pterosaurs when he wrote: “a few peculiar features in the hind limb of lagerpetids merit discussion and suggest that they may be more closely related to pterosaurs than to dinosaurs. Specifically, the ankle of lagerpetids is more similar to that of basal pterosaurs (in particular, Dimorphodon) than to basal dinosauriforms and early dinosaurs. The calcaneum and astragalus are coossified, the ventral surface of the calcaneum is rounded like that of the astragalus, there is no posterior groove of the astragalus, and the calcaneum lacks any sort of calcaneal tuber in both Dimorphodon and lagerpetids. These four character states shared between lagerpetids and Dimorphodon are absent in basal dinosauriforms (e.g., Marasuchus, Asilisaurus). Basal dinosauriforms have a separate calcaneum and astragalus, the ventral surface of the calcaneum, although rounded, is different from the ventral surface of the astragalus, they have a posterior groove of the astragalus, and the calcaneum bears a small calcaneal tuber. It is possible that pterosaurs and lagerpetids share additional ankle characters or differences; however, the ankle of Dimorphodon is heavily ossified, thus concealing the distal end of the tibia and the proximal surface of the astragalus.”

The large reptile tree demonstrates that pterosaurs have no relationship with Lagerpeton and neither do basal dinosaurs, which are distinct from both.

References
Bennett SC 2001. The osteology and functional morphology of the Late Cretaceous pterosaur Pteranodon. Part I. General description of osteology. Palaeontographica, Abteilung A, 260: 1–112. Part II. Functional morphology. Palaeontographica, Abteilung A, 260: 113–153.
Nesbitt SJ 2011. The early evolution of archosaurs: relationships and the origin of major clades. Bulletin of the American Museum of Natural History 352: 292 pp.
Padian K 1983. Osteology and Functional Morphology of Dimorphodon macronyx (Buckland) (Pterosauria: Rhamphorhynchoidesa) Based on New Material in the Yale Peabody Museum. Postilla 189 44pp.