Tradition used to say
that rhynchosaurs were related to rhynchocephalians. Then about forty years ago Carroll and Benton changed everything — without phylogenetic analysis or providing a closer outgroup relative. They said it was so — and it was so!
Carroll (1988) revisiting Carroll (1977) reported, “It was long thought that rhynchosaurs were closely related to modern sphendontids not the basis of general similarities of the skull and dentition. The common presence of primitive features such as the lower temporal bar only points to their common origin among early diapsids. Although the dentition appears to be vaguely similar, it is fundamentally different. Sphenodontids have only a single row of acrodont teeth in the maxilla, but rhynchosaurs have multiple rows of teeth set in sockets. Sphenodontids have a second row of teeth in the palatine, but this bone is edentulous in the rhynchosaurs. What appear to be long premaxillary teeth in the rhynchosaurs are actually processes from the premaxillary bones. Sphenodontids have true premaxillary teeth.”
Problem is, the rhynchosaur outer tooth row is indeed the maxilla (Fig. 1), the inner row is [surprise!] the palatine, as in rhynchocephalians, especially Priosphenodon. That the palatine fuses to the maxilla does not take away the identity of either. Thus, the palatine is not edentulous in rhynchosaurs.
The premaxillary is also toothless in Priosphenodon, currently considered a rhynchocephalian. Mesosuchus is considered a proto-rhynchosaur, not a rhynchocephalian, yet it has socketed teeth on the premaxilla. So there’s a transition zone developing between rhynchosaurs and rhynchocephalians, which are distinct, but decidedly related — because — no other taxa are closer to them than these two are to each other in the large reptile tree.
Benton (1983) reported, “Rhynchosaurs have no special relationship with the sphenodontids. The supposed shared characters are either primitive (e.g. complete lower temporal bar, quadratojugal, akinetic skull, inner ear structure, 25 presacral vertebrae, vertebral shape, certain character of limbs and girdles) or incorrect (e.g. rhynchosaurs do not have acrodont teeth, the ‘beak-like’ premaxilla of both groups is quite different in appearance, the ‘tooth plate’ is wholly on the maxilla in rhynchosaurs but on maxilla and palatine in sphenodontids).”
Benton did not realize the lower temporal bar was derived in sphenodontians. Early lepidosaurs don’t have it.
Acrodont teeth are also derived from socketed teeth, so all sphenodontids had to do was stop fusing their teeth to their skull in order to go back to the socketed teeth found in rhynchosaurs. Rhynchosaurs stop fusing their ankles and stop fusing their teeth to their jaws. That’s just what they do.
The beak-like premaxillae look a little different in rhynchosaurs and rhynchocephalians, but no other reptiles have anything closer in appearance. If you can find one, please let me know. It’s not among the tanystropheids or protorosaurs. They don’t look anything like rhynchosaurs.
Benton did not realize that the palatine fuses to the maxilla in rhynchosaurs and both include socketed teeth. And we can see the evolution of one to another, especially in transitional taxa like Clevosaurus, Mesosuchus and Priosphenodon.
In those early days, before computer-assisted phylogenetic analysis, Benton and Carroll eye-balled taxa and made decisions. Nowadays, those decisions have become traditions so rhynchosaurs are not longer in the same inclusion sets as rhynchocephalians — until reptileevolution.com came along.
Darren Naish (DN) aid it best in his blog on rhynchosaurs, “When the jaws are closed, these tusk-like premaxillae fit in between the dorsally curved anterior tips of the dentaries.” Funny thing, the rhynchocephalian, Priosphenodon, has those.
DN: “Mesosuchus and Howesia would have superficially resembled big lizards or tuataras in proportions.” Well, these traits also go into phylogenetic analyses — if allowed to do so.
DN: Numerous studies published since the 1980s have shown that rhynchosaurs and the members of the tuatara clade were not really closely related (R. Burckhardt had argued as early as 1900 that the supposedly similar premaxillary beaks of tuatara and rhynchosaurs were actually completely different). Tuatara and kin – now termed the Sphenodontia* – are lepidosaurs (and hence close to squamates), while rhynchosaurs are archosauromorphs: part of the same group as archosaurs and their relatives. Archosauromorpha has been defined as a branch-based taxon that includes all taxa closer to Protorosaurus than to Lepidosauromorpha** (Dilkes 1998, p. 528) and autapomorphies of the clade include slender cervical ribs, a notch on the leading margin of the interclavicle, and an ilium with a small anterior and large posterior process (Dilkes 1998).
Okay, again, no phylogenetic analysis here. Just eyeballing. it. Slender cervical ribs are also found in Tritosaurs. A notch on the leading margin of the interclavicle is also found on xxx. An ilium with a small anterior and large posterior process is also found in sphenodontids.
Benton 1985 performed an early cladistic analysis, though apparently not with a computer as no data matrix was presented. The following traits were considered to ally rhynchosaurs with other archosaurmorphs. I abbreviate where possible.
Character list C: Archosaurornorpha
(1) Premaxilla extends up behind naris. NOT in rhynchosaurs.
(2) Nares elongate and close to midline. Also in Clevosaurus.
(3) Quadratojugal located mainly behind the lower temporal
fenestra. Also in Sphenodon. If a lower temporal bar is present, it is formed largely by the jugal, and the squamosal has a short ventral process.
(4) Loss [fusion] of the tabulars. Also in Sphenodon.
(5) Stapes without a foramen. ??
(6) Vertebrae not notochordal. Sphenodon and geckos retain notochordal vertebrae. This trait is invisible in most reconstructions.
(7) Transverse processes on dorsal vertebrae project as distinctive narrow elongate processes. NOT in Mesosuchus, present in Hyperodapedon.
(8) Cleithrum absent. Also ABSENT in Sphenodon.
(9) No entepicondylar foramen in the humerus. NOT in lizards.
(10) Loss of foramen in carpus between ulnare and intermedium. Also ABSENT in lizards.
(11) Presence of a lateral tuber on the calcaneum. NOT in Hyperodapedon.
(12) Complex concave-convex articulation between the astragalus and
calcaneum. Most lepidosaurs have a fused ankle, but some do not.
(13) Fifth distal tarsal lost. ALSO absent in Hyperodapedon.
(14) Fifth metatarsal hooked without lepidosaur specializations. NOT in Hyperodapedon. A hooked 5th metatarsal is present in Sphenodon.
Character list M: Rhynchosauria
(1) Premaxilla bearing a small number of acrodont teeth, or none at all (Fig.8D, E). Also in Priosphenodon.
(2) Single median naris. Good trait.
(3) Fused parietals. Not described for Priosphenodon.
(4) Presence of three proximal tarsals, with the centrale closely associated with the astragalus
Character list N: Rhynchosauroidea
(1) Premaxilla beak-like and lacks teeth.
(2) Parietal foramen absent.
( 3 ) Teeth have ankylothecodont implantation. Rhynchosauroid teeth are implanted in deep sockets and fused to the jaw by bone of attachment. They show a mixture of the characters of acrodont and thecodont teeth (See not far from sphenodontids)
(4) Batteries of functional teeth on the maxilla and dentary. As in Priosphenodon to a lesser extent.
And the latest news
When you add Noteosuchus (Fig. 2), the earliest known rhynchosaur, to the mix, it nests with Sapheosaurus, a derived rhynchocephalian. This gives us our best clue as to what the skull of Noteosuchus looks like using phylogenetic bracketing. I’d like to get better data on Sapheosaurus if available, especially of the skull.
Benton M J 1983. TheTriassic reptile Hyperodapedon from Elgin: functional morphology and relationships. Phil. Trans. R. Soc. Lond. B 302, 605^717. Carroll, R. L. 1988 Vertebrate paleontology and evolution. New York: W. H. Freeman & Co.