A 2020 primer on reptile tarsals

This blogpost had its genesis 
with a new paper by Blanco, Ezcurra and Bona 2020, who studied archosaur and stem archosaur ankles. They reported, “Here, we integrate embryological and palaeontological data and quantitative methodologies to test the hypothesis of fusion between the centrale and astragalus, or the alternative hypothesis of a complete loss of this element.”

More on the results of that paper
after this short primer.

Not sure how much readers know about this subject.
Ankles used to be ‘the thing’ back in the 1980s when terms like “crocodile normal” and “crocodile reversed” were common and influential. Today, with over 230 tested traits in the large reptile tree (LRT, 1658+ taxa), a few ankle traits fade to insignificance. Tiny details, like pegs and sockets, are ignored here. Instead this primer will start with broader, readily visible patterns of presence, absence and fusion.

Figure 1. Basal tetrapod ankles with tarsal elements identified by color.

Figure 1. Basal tetrapod ankles with tarsal elements identified by color.

The origin of carpals preceded the origin of tarsals.
The most primitive (but late appearing) tetrapods, like Trypanognathus, had poorly ossified tarsals and tiny limbs and digits. The most primitive appearances of tarsals in the LRT comes tentatively with Early Carboniferous Pholidogaster (Fig. 1) and completely with Early Carboniferous Greererpeton (Fig. 1). Both are more primitive in the LRT than the traditional fin-to-finger transitional Late Devonian taxa, Acanthostega and Ichthyostega, with their robust limbs and supernumerary digits. We don’t have fossils yet, but we have tracks of Middle Devonian tetrapods. Five is the primitive number of pedal digits in Pholidogaster. Four remains the primitive number of manual digits.

Remember, the first reptile,
Silvanerpeton, is from coeval Early Carboniferous strata. That means we’re missing many intervening taxa from earlier (Late Devonian) layers.

Starting with the sub-equal distal tarsals of Greererpeton
the medial distal tarsals of Gephyrostegus shrink, matching the smaller diameters of the medial metatarsals. The medial centrales also shrink. Two proximal tarsals fuse to become the astragalus and together with the new calcaneum (Pholidogaster lacks one) form the largest tarsal elements, strengthening the tarsus into a tighter, stronger set.

Note the slight rotation of the hind limb of Gephyrostegus
(Fig. 1) relative to the axis of the toes, corresponding to the increased asymmetry of the digit lengths. Distinct from the more fish-like Greererpeton, short-bodied, big-footed Gephyrostegus was able to clamber about on a myriad of landforms, stems and branches with its belly raised off the substrate (Fig. 2).

Gephyrostegus in anterior view

Figure 2. Gephyrostegus in anterior view demonstrating the need for shorter medial toes in tetrapods with a sprawling gait. This insures the toes to not scrape the substrate during the recovery phase and also assures that all the toes contribute to the propulsive phase.

Immediately following Silvanerpeton in the LRT
the Reptilia (=Amniota) split to form two major clades, the Archosauromorpha and the Lepidosauromorpha. At first, both were amphibian-like reptiles (laying amnion-layered eggs) with many traditional amphibians in their number here transferred to the Reptilia based on the last-common ancestor in the LRT method of classification, rather than a list of traditional skeletal traits.

Archosauromorpha step one: Gephyrostegus to Petrolacosaurus
Gephyrostegus (Fig. 3) is the proximal outgroup to the clade Reptilia. Five distal tarsals are present. 4 and 5 are larger than 1, 2 and 3. Four centrale are present. Two proximal tarsals are present, the astragalus (= tibiale) and calcaneum (= fibulae). No intermedium is present.

Petrolacosaurus (Fig. 3) is a basal diapsid. Here Two medial centrale fuse together. Another centrale fuses to the astragalus. The lateral centrale fuses to distal tarsal 4 doubling its size. Distal tarsal 5 shrinks to half. That makes pedal 5 not line up with the other four tarsals.

We’ll return to archosauromorphs below
after dealing with the lepidosauromorphs in order. But first compare the minor differences between the two major reptile clades following Gephyrostegus (Figs. 3, 4).

Figure 1. Gephyrostegus, a reptile outgroup, compared to Petrolacosaurus, a Late Carboniferous archosauromorph basal to archosauriforms and archosaurs.

Figure 3. Gephyrostegus, a reptile outgroup, compared to Petrolacosaurus, a Late Carboniferous archosauromorph basal to archosauriforms and archosaurs. Compare to figure 2.

Lepidosauromorpha step one: Gephyrostegus to Nyctphruretus
Compared to Gephyrostegus, in the owenettid Nyctiphruretus (Fig. 4) two medial centrale fuse together by convergence. Two lateral centrale fuse to distal tarsal 4 tripling its size. Distal tarsal 5 enlarges. Distinct from Petrolacosaurus (Fig. 3), all five metatarsals remain aligned proximally and the hind limb becomes more aligned with the axis of the toes again.

Figure 2. Gephyrostegus compared to the basal lepidosauromorph. Note the fusion of the some centrales into distal tarsal 4.

Figure 4. Gephyrostegus compared to the basal lepidosauromorph. Note the fusion of the some centrales into distal tarsal 4.

Lepidosauromorpha step two: Nyctiphruretus to Huehuecuetzpalli
Huehuecuetzpalli (Fig. 5) is a more arboreal late-survivor in the Early Cretaceous from an Early Triassic radiation of tritosaur lepidosaurs. All distal tarsals are reduced. One and two are absent. Five is fused to metatarsal five creating a twisted ‘hook’. The last centrale is fused to the astragalus and the hind limb is strongly rotated relative to the toes.  The calcaneum is smaller and able to detach from the fibula.

Are you starting to see that bones have a history of homology? Most tarsal fusion is not at all apparent unless comparisons are made in a phylogenetic context. Of course, this affects scoring in analysis.

Figure 1. Lepidosauriform tarsals. The centrale is larger than the distal tarsal 2 in Late Permian Nyctphuretus. Huehuecuetzpalli is Early Cretaceous, so like fenestrasaurs, its ankle also evolved since the Early Triassic split to lose smaller tarsals.

Figure 5. Lepidosauriform tarsals. The centrale is larger than the distal tarsal 2 in Late Permian Nyctphuretus. Huehuecuetzpalli is Early Cretaceous, so like fenestrasaurs, its ankle also evolved since the Early Triassic split to lose smaller tarsals.

Lepidosauromorpha step three: Macrocnemus to Peteinosaurus
Compared to taxa above (Fig. 5), Middle Triassic Macrocnemus (Fig. 6) loses distal tarsal 2 and retains the medial centrale.

Increasingly bipedal Langobardisaurus (Fig. 6) loses distal tarsal 1. The centrale is larger. The other tarsals are smaller.

Increasingly bipedal and sometimes flapping Cosesaurus (Fig. 6) loses distal tarsal 2. The centrale mimics distal tarsal 2. Distal tarsal 4 is not much larger than the centrale. The tibia migrates back in line with the axis of the pes.

Completely bipedal and flapping Peteinosaurus (Fig. 6) has a simple hinge ankle joint with alll four tarsal elements relatively larger and more similar in size.

Figure 3. Tritosaur lepidosaur tarsals. Note how the centrale moves distally to replace distal tarsal 1 and 2.

Figure 6. Tritosaur lepidosaur tarsals from Peters 2000. Note how the centrale moves distally to replace or fuse with distal tarsal 1 and 2. Or is the centrale really distal tarsal 2?

Archosauromorpha step two: Petrolacosaurus to Protorosaurus
Compared to the basal archosauromorph diapsid, Petrolacosaurus (Figs. 3, 7) In Protorosaurus (Fig. 7) distal tarsal 5 fuses to metatarsal 5, as in the lepidosaur tritosaur, Huehuecuetzpalli (Fig. 5) by convergence. The astragalus moves to a more central position as the medial centrale articulates directly with the tibia in a short-lived experiment that does not continue with taxa more directly in the archosauriform lineage (Fig. 8).

Figure 2. Petrolacosaurus and Protorosaurus pedes to establish homologies.

Figure 7. Petrolacosaurus and Protorosaurus pedes to establish homologies.

Archosauromorpha step three: Protorosaurus to Archosauriforms
Compared to Protorosaurus (Fig. 8), the tarsals are little changed in the basal archosauriform, Proterosuchus, with the note that, as mentioned above, the centrale does not contact the tibia. The distal tarsals are sub-equal in size. The calcaneum is laterally extended, backing up the similarly extended mt5.

In Euparkeria (Fig. 8) the tarsus is similar with symmetrical and block-like proximal tarsals. The calcaneum does not back up mt 5. Pedal digit 3 longer than 4 signaling a less sprawling, more upright, form of locomotion with a simple hinge ankle joint.

In increasingly bipedal PVL 4597 (Fig. 8), basalmost archosaur, distal tarsals 1 and 2 are absent, 3 and 4 are fused. The calcaneum has a posterior process, the ‘heel’. Mt 1 is longer creating a more symmetrical pes with a simpler hinge ankle joint.

Figure 8. Archosauriform pedes compared to Protorosaurus.

Figure 8. Archosauriform pedes compared to Protorosaurus.

Archosauromorpha step four: PVL 4597 to higher Archosauria
Compared to PVL 4597 (Fig. 9), the tarsus of the basal dinosaur Herrerasaurus (Fig. 9) is further reduced, and so is the calcaneum. The astragalus develops an anterior ascending process that also seen in Crocodylus (Fig. 9), which no longer has a simple-hinge ankle joint. Here fused distal tarsal 4/5 is thicker and the astragalus contacts mt 1, slightly rotating the tibia medially for a more sprawling configuration. Distinct from other archosaurs, mt 1 is the most robust in the set and mt 5 becomes a robust vestige in Crocodylus.

Figure 7. Archosaur tarsals compared.

Figure 9. Archosaur tarsals compared.

Blanco, Ezcurra and Bona 2020 report, 

  1. “the astragalus developed ancestrally from two ossification centres in stem archosaurs 
  2. the supposed tibiale of bird embryos represents a centrale.
  3. The tibiale never develops in diapsids.”

In counterpoint,

  1. Figure 1 indicates the two ossification centers go back to the basal tetrapod, Greererpeton. Protorosaurus was an oddball with the centrale contacting the tibia.
  2. Figure 1 indicates it is inappropriate to call the proximal tarsal in any reptile the ‘tibiale’ as the astragalus is present prior to basalmost taxa.
  3. See above.

Blanco et al. report,
“The proximal tarsus of archosaurs is ancestrally composed of a medial astragalus that articulates proximally with the tibia and fibula and a lateral calcaneum that articulates proximally with the fibula.” The authors do not identify the owner of such a tarsus, but let us presume it is that oddball Protorosaurus (Figs. 7, 8).

Blanco et al. reach back to captorhinids
to suggest an outgroup to archosaurs. Unfortunately, captorhinds are basal lepidosauromorphs in the LRT. So the authors are looking where they should not be looking for progenitors.

The Blanco et al. membership list of Archosaurs is over extended.
Blanco et al. employ the invalid clades ‘Pseudosuchia‘ and ‘Avemetarsalia‘, which includes the lepidosaur pterosaurs. They also employ the lepidosaur, Macrocnemus (Fig. 6) as an archosauriform outgroup. Their Archosauromorpha include the archosauriforms, Proterosuchus and Erythrosuchus and the archosaurs Caiman, Lewisuchus and Rhea. They are not aware that the old definition of Archosauromorpha now includes synapsids when given the taxon list of the LRT.

Figure 8. Basal pterosaur and basal dinosaur pedes (feet) compared. While convergent in many respects, certain traits separate these two unrelated clades.

Figure 10. Basal pterosaur and basal dinosaur pedes (feet) compared. While convergent in many respects, certain traits separate these two unrelated clades.

Pterosaurs are traditionally considered archosaurs, 
but that was shown to be invalid twenty years ago. Perhaps it would help if a basal archosaur dinosaur and a basal lepidosaur pterosaur were shown side-by-side (Fig. 10). We’ve already seen many instances of convergence in the tarsal evolution of archosauromorphs and lepidosauromorphs. This is just one more instance of the same. It is time for paleontologists to stop dragging their tails and get up to speed in this arena.


References
Blanco MVF, Ezcurra MD and Bona P 2020. New embryological and palaeontological evidence sheds light on the evolution of the archosauromorph ankle. Nature Scientific Reports (2020)10:5150. https://doi.org/10.1038/s41598-020-62033-8
Peabody FE 1951. The origin of the astragalus of reptiles. Evolution 5(4):339–344.