Yesterday Liu et al. 2017 reported on
a pregnant Dinocephalosaurus (Figs. 1-5). This is wonderful and exciting news. However, the embryo is NOT in the process of passing through the cloaca, as we’ve seen in ichthyosaurs. The embryo is much higher in the abdomen, still in the uterus. So the headline “Live birth in an archosauromorph reptile” is… at best… premature. Live birth is still a possibility. A critical examination of the data reveals a few more major and minor problems.
Figure 1. The holotype (not the new specimen) of Dinocephalosaurus in resting, feeding and breathing modes. In breathing mode the throat sac would capture air that would not be inhaled until the neck was horizontal at the bottom of the shallow sea. Orbits on top of the skull support this hypothesis. Image from Peters 2005. The new specimen has a longer neck, a more robust tail, and a different pedal morphology.
the authors nested Dinocephalosaurus within the Archosauromorpha (Fig. 2). That is incorrect. Dinocephalosaurus nests within the new Lepidosauromorpha in the large reptile tree (LRT, 952 taxa), which minimizes the taxon exclusion problem suffered by the much smaller taxon list in the Liu et al. 2017 tree.
Figure 2. Cladogram from Liu et al. 2017 with colors added based on results from the LRT. Taxon exclusion is a major problem here. Note in the Liu et al. cladogram members of the Protorosauria are divided into three clades. In sympathy, members of the Tritosauria and Protorosauria do indeed converge with one another. More taxa clears up the problem shown here of cherry-picking taxa.
Dinocephalosaurus actually nests
within the lepidosaur clade, Tritosauria, a clade that also includes Tanystropheus, pterosaurs and several other taxa (Fig. 7) that had been mistaken for protorosaur relatives in the Liu et al and other prior studies.
As a lepidosaur,
Dinocephalosaurus would have been able to retain embryos within the mother far longer that in extant archosauromorphs. And based on the extreme thinness of pterosaur eggshells (closest known relatives with embryos, Fig. 7), those leathery eggshells only develop just prior to egg laying. So live birth is only one of a spectrum of options for Dinocephalosaurus. As in pterosaurs, the eggs could have hatched shortly after the female laid them on the shoreline.
Figure 3. The holotyype of Dinocephalosaurus. Although extremely similar, the new specimen is different in several ways. See below.
Liu et al. report that live birth is unknown in the Archosauromorpha.
However, in the LRT mammals and enaliosaurs (sauropterygians + ichthyosaurs) are both archosauromorphs that experience live birth. Hyphalosaurus, a choristodere archosauromorph, had extremely thin eggshells and retained developing embryos inside the mother until laying those eggs.
Figure 4. Hypothetical Tanystropheus embryo compared to part of an embryo of the new specimen attributed to Dinocephalosaurus.
More about that embryo
What little is preserved of the Dinocephalosaurus embryo (Fig. 4) is curled up in its amniotic sac, as one would expect for any reptile embryo still in utero. For comparison, note the hypothetical Tanystropheus embryo alongside it. That long neck has to go somewhere and Dinocephalosaurus provides further evidence that juvenile tritosaurs were isometric duplicates of their adult parents. That long neck did not develop with maturity. Among other tritosaurs we see juveniles similar in proportion to adults in the basal form, Huehuecuetzpalli, and all pterosaur embryos.
Liu et al. further report. “Despite the complexity of this transition, viviparity has evolved at least 115 times in extant squamates (lizards and snakes), in addition to a single time in the common ancestor of therian mammals. Moreover, viviparity is a common reproductive mode in extinct aquatic reptiles including eosauropterygians, ichthyosaurs, mosasauroids, some choristoderans and likely mesosaurs.” Since mosasauroids are extinct squamates that makes at least 116 times for lepidosaurs. Some living squamates produce eggs that hatch shortly after they are expelled, a sort of transition from oviparity to viviparity. That’s where pterosaurs fall and perhaps Dinocephalosaurus.
More cladogram issues
The Liu et al. figure 1 cladogram shows a polytomy of most reptilian clades arising during the Permian. No such polytomy appears in the LRT in which Archosauromorpha diverged from the Lepidosauromorpha tens of millions of years earlier in the Viséan (Lower Carboniferous). Liu et al. mistakenly report that trilophosaurs, rhynchosaurs and pterosaurs are archosauromorph reptiles. They are lepidosauromorph reptiles in the LRT.
Figure 5. The new Dinocephalosaurus has traits the holotype does not have, like a longer neck with more vertebrae, a robust tail with deep chevrons and a distinct foot morphology with a more elongate pedal digit 4. The partial embryo is in magenta at left.
The new specimen looks like a Dinocephalosaurus, but is it one?
Distinct from the holotype, the new specimen has a deep robust tail with deep chevrons (Fig. 5). They all share a common ancestor in one of the highly variable Macrocnemus specimens (Fig. 7). The toes of the new specimen are more asymmetric. The neck probably has more vertebrae (several are lost, but note the longest ones are NOT at the base of the neck in the holotype). Unfortunately little more can be said with so much of the mother lacking at present. We’ve already seen a Chinese Tanystropheus similar to, but not identical to the European Tanystropheus. We can imagine even greater variation within the available gamut of the present sparse fossil evidence.
It really is too much
to expect identical specimens to come from distinct fossil bearing strata. So variation within Dinocephalosaurus is a possibility.
The paleo-community needs to include more specimen-based taxa in their cladograms or the Liu et al. problem (not restricted to them!) is going to continue ad infinitum. I know that’s a lot of work. But it can be done (I’ve done it!) and it needs to be done. Just start with a large gamut analysis and keep adding taxa to it. That will make the current phylogenetic problems go away.
Images of tanystropheids and dinocephalosaurs swimming horizontally through open waters (Liu et al. 2017 their figure 3) may not be an accurate portrayals of their daily lives. Other options have been published (Fig. 1) or appear online (Fig. 8). Odd-looking tetrapods often have uncommon niches and atypical behaviors.
Figure 8. Tanystropheus in a vertical strike elevating the neck and raising its blood pressure in order to keep circulation around its brain and another system to keep blood from pooling in its hind limb and tail.
Li C, Rieppel O and LaBarbera MC 2004. A Triassic aquatic protorosaur with an extremely long neck. Science 305:1931.
Liu, J. et al. 2017. Live birth in an archosauromorph reptile. Nature Communications 8, 14445 doi: 10.1038/ncomms14445
Peters D, Demes B and Krause DW 2005. Suction feeding in Triassic Protorosaur? Science, 308: 1112-1113.