Celebrating 300 Taxa in the Large Reptile Tree

Science is all about constants.  
The evolution of reptiles has been set in stone (literally). What we’re looking for now is a constant model that reconstructs that family tree with clues provided by the fossils we know …and how well we know them. We want to remove all the former mystery from this pursuit. After all, it is 2012. Everything else has been figured out by now…why not the reptile?

So when the taxon list in the large reptile tree rises to 300, as it just did, its worth a look to see what changes have occurred. (It’s a milestone, after all).

First of all, nothing earth-shattering.
In fact, none of the new taxa have changed the tree topology (= original order (going back several tree generations)). This tree continues to grow in good health, becoming fuller as more leaves and nodes are added. All sisters continue to look alike, which is the key after all. That’s how evolution works. Descendants look like ancestors. Sisters look like sisters. In the large reptile tree former enigmas have been nested. Mistaken nestings created by smaller earlier and concurrent studies have been rectified. Additional taxa have been added without changing the tree topology. No other study on reptile relations can claim the same certainty and consistency.

I believe we have a very good tree here.
And I’m looking forward to the next milestones at 350 and 400 taxa (so far not counting the pterosaurs nor the therapsids). The present tree provides several tests of relationships, of bone identities and of sutures vs. cracks. Errors have been acknowledged and repaired. Oversights have been reexamined. Hopefully we all will never rest on our laurels, traditions or paradigms. I look forward to making changes, but only if it gets us closer to modeling more accurately the actual reptile family tree.

Keep those discoveries coming! 
And as I mentioned earlier, if ANYONE finds two taxa in the large reptile tree that should not be sisters, please bring these to my attention. So far no one has despite some earlier comments. If there are any mismatches, please let me know.

We’ll talk about it.

Tapejara (Pterosaur) Growth Pattern

The traditional paradigm is that pterosaurs experienced allometric growth, like mammals, birds and crocodilians. If so the rostrum should be shorter and the eye sockets (orbits) should be proportionally larger.

The heretical view holds that pterosaurs experienced isometric growth with little proportional differences in embryos, juveniles and adults. Currently this is supported by the evidence of embryos and the few juveniles known, including Tupuxuara.

Now we have another juvenile, a Tapejara. Here (Fig. 1) the published tracings of the juvenile Tapejara have been enlarged to best fit the adult Tapejara. Here the proportions are virtually identical, with no rostral shortening or orbit enlargement. The main difference is the depth of the nasal and the angle of the jaw.

Figure 1. Tapejara (Pterosaur) growth patterns demonstrating, once again, isometric growth in which virtually no proportional differences distinguish juveniles from adults. In grey, the juvenile to scale with an adult. Overlay, the juvenile enlarged to the size of the adult. The depth of the nasal, over the orbit, appears to be the biggest difference here, but then again we’re dealing with drawings here. Could that nasal be rotated slightly?

Go with the evidence. Don’t follow traditions and paradigms.

As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.

Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.

References
Eck K, Elgin RA and Frey E 2011. On the osteology of Tapejara wellnhoferi KELLNER 1989 and the first occurrence of a multiple specimen assemblage from the Santana Formation, Araripe Basin, NE-Brazil. Swiss Journal of Palaeontology, doi:10.1007/s13358-011-0024-5.
Kellner AWA 1989. A new edentate pterosaur of the Lower Cretaceous from the Araripe Basin, northeast Brazil. Anais da Academia Brasileira de Ciências 61, 439-446.

wiki/Tapejara
wiki/Tupandactylus

A [Formerly] Unknown Tapejarid

The website Pterosaur.net published an image of an “unknown tapejarid” (aka the Rio tapejarid, Fig. 1), which is a largely articulated torso missing most of its ribs and gastralia. Here we’ll try to figure out what sort of tapejarid this torso is.

[Added a day late, I found the paper Sayão and Kellner (2006) that describes this specimen, MN 6588-V. Apologies. Still the authors were not able to identify it. ]

Figure 1. In situ tracing of the unknown “Rio” tapejarid MN 6588-V from Pterosaur.net

I was surprised to see this image published as “unknown.” Here we’ll try to lend a hand in identifying this headless specimen.

Figure 2. The ‘Rio’ tapejarid reconstructed. The overall size, the length of the torso and the distinctly narrow neck on the postacetabular process of the ilium all indicate affinity with Tapejara. Prepubes are rare in this clade, so good to see them here, with a shape distinct from cousin Huaxiapterus.

A Probable Tapejara
The elongated torso, the shape of the posterior ilium and the overall size indicate a close affinity with Tapejara in lieu of a phylogenetic analysis. As in other tapejarids, the relatively short coracoids produce a “bottom decker” wing attachment to the torso in which the wing attached relatively lower on the torso. The prepubes were robust with a relatively short stem and tiny perforation. The anterior ilium was quite gracile.

Figure 1. Click to enlarge and see the unknown tapejarid (Not shown here) to scale with the others. This is the Tapejaridae, including Sinopterus, Huaxiapterus, Tapejara, Tupandactylus, Tupuxuara and Thalassodromeus

As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.

Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.

References
Eck K, Elgin RA and Frey E 2011. On the osteology of Tapejara wellnhoferi KELLNER 1989 and the first occurrence of a multiple specimen assemblage from the Santana Formation, Araripe Basin, NE-Brazil. Swiss Journal of Palaeontology, doi:10.1007/s13358-011-0024-5.
Kellner AWA 1989. A new edentate pterosaur of the Lower Cretaceous from the Araripe Basin, northeast Brazil. Anais da Academia Brasileira de Ciências 61, 439-446.
Sayão JM and Kellner AWA 2006. Novo esquelito parcial de pterossauro (Pterodactyloidea, Tapejaridae) do membro Crato (Aptiano), Formação Santana, Bacia do Araripe, Nordeste do Brasil. Estudos Geológicos 16(2):16-40.

wiki/Tapejara

Doin’ the “Pterosaur Walk”

Figure 1. Man acting like a pterosaur on the beach with ski poles (click for blogspot). Note the vertical backbone and ski poles contributing to stability, not thrust. His mate (or daughter) is walking bipedally.

Footprints from around the world tell us that pterosaurs of all sizes walked on prehistoric beaches. Today we have two competing configurations hypothesizing how pterosaurs did this. I like the configuration demonstrated above (Fig. 1) and it is supported by matching tracks to trackmakers, following the thrust vectors and considering methods by which pterosaurs might launch themselves into the air.

The traditional and majority view (Bennett 1997, Fig. 2) posits that the backbone was nearly horizontal and the fingers impressed the sediment far ahead of the shoulders (Fig. 1). This has been criticized here and here on several grounds.

Figure 1. Click to animate. Walking pterosaur according to Bennett 1997. Note the forelimbs provide no forward thrust, but merely act as props.

The heretical minority opinion posits that the backbone was held much more vertically (diagonally) such that the toes impressed beneath the center of balance while standing, slightly behind while walking (as in humans, Fig. 2) and that the fingers impressed much closer to the center of balance – when they could (in Pteranodon, for instance, they could not).

Figure 2. Click to animate. Plantigrade and quadrupedal Pterodactylus walk matched to tracks. Note the more vertical backbone and the forelimbs not contributing thrust to locomotion, only stability.

Otherwise the fingers did NOT contribute thrust to forward progression while walking. In this pose pterosaurs could merely lift their ski-pole like arms to launch themselves bipedally. Such a pose also follows from bipedal ancestry and a secondary acquisition of a quadrupedal pose as demonstrated, among other traits, by a backward-pointing finger #3.

As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.

Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.

References
Bennett SC1997. Terrestrial locomotion of pterosaurs: a reconstruction based on Pteraichnus trackways. Journal of Vertebrate Paleontology, 17: 104–113.

Mesosaurus Mistake?

The recent interpretation of a lateral temporal fenestra in Mesosaurus (Piñeiro et al. 2012, Fig. 1) was accompanied by a new interpretation of the premaxilla/maxilla suture. Added to this problem, the published restorations of the Mesosaurus skull in lateral and dorsal views don’t match (Fig. 1). Note the shifting of the fenestrae among other sutural changes.  Unfortunately, no sister taxa have the premaxilla/nasal configuration that Piñeiro et al. (2012) propose.

Figure 1. Mesosaurus skull with lateral temporal fenestra. From Piñeiro et al. 2012. Note the many differences in bone shape not attributable to the two distinct views. More rigorous attention would have aligned the fenestra and sutures in both views of these apparently freehand sketches.

Figure 2. Mesosaurus skull reconstructed based on data from Modesto (2006).

Modesto
had it right.
Modesto (2006) delineated and identified the premaxilla and nasal of Mesosaurus (Fig. 2) and these sutures also agree with those of sister taxa, as recovered by the large reptile tree. With or without the skull, mesosaurs don’t shift within the tree.

Figure 3. New tracings of the Piñeiro et al. (2012) rostral skull of Mesosaurus. On the left the bones are colored in the new interpretation. At right blue indicates published interpretation of nasals, unlike those of any other sister.

The split or separation of the nasals seen in thalattosaurs and sauropterygians (by convergence) has its origin in mesosaurs. In these taxa the elongated premaxilla contacts (sometimes just barely) the frontals. In ichthyosaurs, like wumengosaurs and mesosaurs, the premaxilla does not contact the frontals. Even so, it does split the nasals at least anteriorly.
Don’t Forget the von Huene (1940) Interpretation.
von Huene (1940) spent a considerable amount of time examining several Mesosaurus specimens and he was the first to interpret a lateral temporal fenestra (Fig. 4). He also interpreted nasals extending anteriorly a short distance beyond the nares as in the Piñeiro (2012) interpretation.

Almost overlooked, von Huene’s interpretations of the rostrum are much narrower than those of Piñeiro et al. (2012) indicating some variation in the specimens they were looking at. Certainly these are distinct species. Perhaps distinct genera?

Figure 4. Mesosaurus according to von Huene 1940. Color added to nasals in blue and premaxilla in yellow. This interpretation of the premaxilla/nasal suture matches that of Piñeiro et al. (2012).

Which is Correct?
Neither Piñeiro et al. (2012) nor von Huene (1940) compared their specimens to sisters recovered by the large reptile tree in which the premaxilla bisects the nasals. The image below is recolorized from Piñeiro (2012) showing two interpretations: 1) a longer premaxilla and 2) a possible interdigitating nasal and premaxilla. Hard to say given the data (Fig. 2).

Figure 5. Two reinterpretations of the premaxilla/nasal suture in a recolorized restoration of Mesosaurus by Piñeiro et al. (2012). See Fig. 1 for the original. Also more squamosal has been restored on one side. It is not only possible, but likely that there is variation in several aspects of Mesosaurus from quarry to quarry and from species to species.

What’s Interesting
Here you see variation in interpretation between three paleontologists (one being the Piñeiro team). Look around and you’ll see the maxilla extends beyond mid orbit in two interpretations, but not the Modesto (2004) one. The tooth count differs in all three. The rostral proportions differ. The lateral temporal fenestra differs. The premaxilla/nasal suture differs. The frontal/nasal suture differs. Granted, they are not all looking at the same specimens. And now you can see why it is so important in phylogenetic analysis to use specimens for taxa, even if incomplete.

The Good Thing Is…
Mesosaurus will never again be lumped with basal anapsids like pareiasaurs and captorhinids. Hopefully…

As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.

Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.

References
Gervais P 1865. Du Mesosaurus tenuidens, reptile fossile de l’Afrique australe. Comptes Rendus de l’Académie de Sciences 60:950–955.
Laurin M and Reisz RR 1995. A reevaluation of early amniote phylogeny. Zoological Journal of the Linnean Society 113:165-223.
Modesto SP 1999. Observations on the structure of the Early Permian reptile Stereosternum tumidum Cope. Palaeontol. Afr. 35, 7–19.
Modesto SP 2006. The cranial skeleton of the Early Permian aquatic reptile Mesosaurus tenuidens: implications for relationships and palaeobiology. Zoological Journal of the Linnean Society 146 (3): 345–368. doi:10.1111/j.1096-3642.2006.00205.x.
Modesto SP 2010.The postcranial skeleton of the aquatic parareptile Mesosaurus tenuidensfrom the Gondwanan Permian. Journal of Vertebrate Paleontology 30 (5): 1378–1395. doi:10.1080/02724634.2010.501443.
Piñeiro G, Ferigolo J, Ramos A and Laurin M 2012. Cranial morphology
of the Early Permian mesosaurid *Mesosaurus tenuidens* and the evolution of
the lower temporal fenestration reassessed. Comptes Rendus Palevol.
von Huene F 1941. Osteologie und systematische Stellung von Mesosaurus. Palaeontogr. Abt. A. 92, 45–58.

“Protopterosaurus” – still the official poster child for pterosaur ancestry

Updated January 31,
with a revision. The ‘Protopterosaurus’ illustration comes from Wild 1978, the Eudimorphodon paper, not 1984, the Preondactylus paper.

Wild 1978 had just finished describing the first Triassic pterosaurs when he proposed a hypothetical pterosaur ancestor (Fig. 1) that was re-illustrated by Wellnhofer (1991) in his famous pterosaur encyclopedia. Today, nearly 30 years later, most pterosaur experts are still crossing their fingers, hoping for just such a creature to appear in the fossil record, having dismissed and rejected (Hone and Benton 2007, 2008) several actual candidates proposed by Peters (2000), now some 12 years ago, and supported by a greatly expanded phylogenetic analysis here.

Figure 1. “Protopterosaurus” the hopeful hypothetical ancestor of pterosaurs created by Wild (1978, not 1984) and supported by Wellnhofer (1991). Not sure why the tail was so thick and the hind limbs were so short. Cosesaurus demonstrates that the flying membrane aft of the forelimb originated distally, not medially, with a narrow-gauge occasionally bipedal trackway in which pedal digit 5 impressed far behind the other pedal digits, known as Rotodactylus. So, this should be a runner and flapper first, then a tree climber.

So the question is, how well does the hypothetical ancestor stack up against the real McCoys, as determined by phylogenetic analysis? [Think Cosesaurus.] 

And what  would “Protopterosaurus?” be? A lizard or an archosaur?

What is “Protopterosaurus”?
Wild (1978) and Wellnhofer (1991) both illustrated “Protopterosaurus” as an unarmored arboreal quadrupedal reptile with laterally extended limbs and digit 4 the longest on all four digits. Thus it can’t be an archosaur! Hmm. With that morphology it would have nested in the lizard or protorosaur clades with a strong lean toward the former. In several regards “Protopterosaurus” resembled something between Huehuecuetzpalli and Macrocnemus, (the latter Wild was quite familiar with), but with a much thicker tail.

Unfortunately we can’t peer beneath the skin here. “Protopterosaurus” was illustrated ‘in vivo.’

The Head
The naris was displaced from the tip of the snout in “Protopterosaurus” as in fenestrasaurs, not archosaurs. The naris preceded a depression, presumably an antorbital fenestra, as in Cosesaurus and archosaurs. The snout was pointed in both dorsal and lateral views (more like Cosesaurus). The eyes were raised above the midline of the back of the skull, probably due to a deep jugal. That’s not duplicated in any candidate, all of which had a larger orbit expanded nearly to the jawline, with a very shallow jugal.  The retroarticular process extended far behind the cranium, producing an anteriorly leaning quadrate. Likewise, that is not found in any candidate except, perhaps, Jesairosaurus.

The Neck, Torso and Tail
The cervical series in “Protopterosaurus” was as long as the skull and the neck was robust, as in most candidates. No S-curve was given to the dorsal series. No armor was present either. The caudals must have been provided with wide transverse processes and deep chevrons as the tail is illustrated to be thick and meaty, like that of most archosaurs and lizards — and unlike the tritosaurs and fenestrasaurs, including pterosaurs, all of which had a very thin bony tail.

The Appendices
The forelimbs of “Protopterosaurus” extended laterally with fingers of increasing length laterally, especially digit 4, which was as long as the humerus+ulna. Cosesaurus had fingers 3 and 4 virtually subequal when it developed a prepubis, pteroid, a locked down coracoid, sternal complex, membranes and an antorbital fenestra. So the flight digit came later, in Longisquama. Finger 5 was a vestige in “Protopterosaurus“. Finger 5 in reality was always short, but did not become a vestige until the pterosaur stage. The metacarpals of “Protopterosaurus” were illustrated longer medially than laterally, which is not duplicated in any candidate.

The femur was no longer than the humerus in “Protopterosaurus” and the tibia was no longer than the femur. These proportions are not found in present candidates, all of which had a longer hind limb than forelimb and relatively larger hind limbs compared to the torso. All candidates, from Scleromochlus to Cosesaurus, have been illustrated as bipeds and narrow-gauge tracks are known for a sister to Cosesaurus. Thus the lateral limbs and belly down configuration are not duplicated in any candidate. The foot in “Protopterosaurus” was about the length of the tibia, which is correct. Pedal digit 5 extended no further than p4.1 or p4.2, which is also correct with regard to Cosesaurus. Scleromochlus and its sisters had a vestige to no pedal digit 5, which essentially removes them from pterosaur sisterhood candidacy.

Soft Tissue
Flying squirrel-like membranes in “Protopterosaurus” extended from finger 4 to the knee, remaining shallow until reaching the elbow. Current evidence (Peters 2009) indicates that membranes developed distally first, just the opposite of the hypothetical traditional model. Current evidence indicates that flapping as a secondary sexual behavior also preceded gliding, so “Protopterosaurus” had it backwards. Minor membranes appear in “Protopterosaurus” behind the acetabulum and a propatagium (strangely without a pteroid and preaxial carpal) appeared in front of the forelimb. We know Cosesaurus had a pteroid and preaxial carpal (Peters 2009), but no propatagium is known in relation to it.

Wellnhofer (1991) considered the development of wing membranes as an aid in falling, parachuting and surviving, then in gliding to expand the range of escape and exploitation. Last to develop would be the enlarged sternum and pectoral girdle to add thrust to gliding. This is the opposite of what actually happened, as reported earlier and demonstrated in Cosesaurus, which used flapping as a secondary sexual behavior after becoming bipedal.

So, the imagined creature, “Protopterosaurus,” had many traits later found in the actual ancestors of pterosaurs, but missed several others. Not sure why current studies in pterosaur ancestry are not at least considering the possibility that fenestrasaurs and tritosaurs are good candidates for this honor. The first PhD to do so will be much less embarrassed by the last twelve years than the last PhD to do so will be.

References
Peters D 2000. A Redescription of Four Prolacertiform Genera and Implications for Pterosaur Phylogenesis. Rivista Italiana di Paleontologia e Stratigrafia 106 (3): 293–336.
Peters D 2009. A reinterpretation of pteroid articulation in pterosaurs. Journal of Vertebrate Paleontology 29: 1327-1330
Wild R 1978. Die Flugsaurier (Reptilia, Pterosauria) aus der Oberen Trias von Cene bei Bergamo, Italien. Bolletino della Societa Paleontologica Italiana 17(2): 176–256.
Wild R 1984.  A new pterosaur (Reptilia, Pterosauria) from the Upper Triassic (Norian) of Friuli, Italy. Gortania Mus Friulano Stor Nat 5: 45-62.
Wellnhoffer P 1991.  The Illustrated Encyclopedia of Pterosaurs. London: Salamander. 192 pp.

Pterosaurs landing in trees – part 3 – Longisquama

Earlier we demonstrated pterosaurs grappling tree trunks and perching on tree branches. Today we’ll report on the origin of this behavior in Longisquama (Fig. 1).

Figure 1. Longisquama on a tree trunk.

My What Long Fingers You Have Grandma!
The origin of tree clinging likely occurred in sister to Longisquama, the current outgroup for the Pterosauria. The fingers of Longisquama were larger than those of any pterosaur. Fingers one through three and five became smaller as the wing finger, #4, became larger.

The PILs Have Something to Say
In Longisquama the PILs of manual digit 4 were not aligned with those in digits 1-3. This indicates they no longer worked as a set and that metacarpal 4 + digit 4 likely were rotated into the plane of the wing, as in pterosaurs. This means, like pterosaurs, Longisquama was able to flex (or hyperflex) digit 4 in the plane of the wing and while clinging to trees, the wing would have created yet one more display trait, opening and closing tangential to the diameter of the tree trunk. Pterosaurs emphasized the wings and de-emphasized the dorsal frill.

Short Arms
The relatively short arms of Longisquama meant that the long legs had to crouch more, which created a pre-loaded spring ready to release for the next leap/glide. The arms lengthened in pterosaurs.

Finger 4 – Not Quite Ready for Flight, but Great Looking!
Before finger 4 became large enough to support gliding and flapping flight it served as yet ANOTHER decoration on Longisquama, already at no shortage for display traits. As in pterosaurs, finger four was free to flex in the plane of the wing, on a tangent to the diameter of the tree (Peters 2002). I did not realize at the time of that publication how large the fingers of pterosaur predecessors, like Longisquama, had become.

New discoveries bring new insights.

As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.

Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.

References
Peters D 2002. A New Model for the Evolution of the Pterosaur Wing – with a twist. Historical Biology 15: 277-301.

Pterosaurs landing in trees – part 2 – perching with their opposable digit

The traditional view holds that pterosaurs did not perch on branches, largely because they did not have an opposable digit. This is wrong. Pterosaurs had an opposable digit. It was just not like that of birds. Instead it was uniquely pterosaurian — working by extension, not flexion. Peters (2002, 2010) described how pedal digit 5 in basal pterosaurs acted like a universal wrench, extending while digits 1-4 flexed around a branch of any diameter (Fig. 1), thereby opposing the flexing digits (note the white arrows in the branch cross-section).

Figure 1. The pterosaur Dorygnathus perching on a branch. Above the pes of Dorygnathus demonstrating the use of pedal digit 5 as a universal wrench (left), extending while the other four toes flexed around a branch of any diameter and (right) flexing with the other four toes. As in birds, perching requires bipedal balancing because the medially directed fingers have nothing to grasp.

Pterosaur perching requires a bipedal capability
Pterosaurs could perch on horizontal branches — although not quite like birds. As in birds, the anterior toes were located below the center of balance in flight, the shoulder glenoid, so a bipedal configuration was possible without shifting or lifting the torso.

In the most primitive pterosaurs pedal 5.2 (actually a fused p5.2+p5.3) was straight. In Dorygnathus and certain other derived clades, pedal 5.2 was bent or bowed in various ways (Peters 2011). Bent or straight had little effect on the universal wrench.

Certain derived pterosaurs without a large pedal digit 5 (e.g. ctenochasmatids, pterodactylids) probably did not perch on tree branches. Footprints indicate they were beachcombers. Only those with large fingers and trenchant finger claws could also have found safe haven on tree trunks, upon which pedal digit 5 had less of an importance.

---

References
Peters D 2002. A New Model for the Evolution of the Pterosaur Wing – with a twist. – Historical Biology 15: 277–301.
Peters D 2010. In defence of parallel interphalangeal lines. Historical Biology iFirst article, 2010, 1–6 DOI: 10.1080/08912961003663500

Pterosaurs landing in trees – part 1 – grappling the trunk like a lemur

In the past, pterosaurs were usually pictured hanging from cliffs (remember Pteranodon in Fantasia?). Smaller pterosaurs were usually illustrated hanging beneath branches (Wellnhofer 1991) by their feet alone or inverted quadrupedal. Today we’ll start a short series on pterosaurs and trees demonstrating how some pterosaurs (not all) could interact with a tree trunk. If anyone finds this convergent with Archaeopteryx and kin, you’re probably right.

Figure 1. Dorygnathus on a tree.

Basal pterosaurs had long fingers with trenchant claws pointed medially when the wings were folded (Fig. 1), as in basal birds. These claws were likely used to grapple tree trunks, clinging to them in the manner of lemurs and other primates (Peters 2002), arms on both sides of the trunk. In that configuration the feet were planted side-by-side beneath the pelvis, toes pointing anteriorly to antero-laterally with the dorsal surface of digit 5 putting extension pressure on the trunk, which enabled the toe claws to dig just a little deeper, much like a church key can opener.

A Chance to Show Off
In that configuration the wing finger would have been free to fully open as it would have been set at a tangent to the circumference of the tree trunk, no matter the diameter (Peters 2002, Fig. 2). Opening and closing the large wing fingers would have created a large display device, much larger than any anole dewlap, but serving the same purpose — finding a mate (Peters 2002), which was the original purpose of wings on predecessors like Longiquama.

Figure 2. Pterosaur on a tree demonstrating the ability of the wing to open unimpeded on a tangent to the tree trunk. This would have served as a display mechanism. Modified from Peters (2002).

The large pectoral muscles would have provided sufficient adduction power to enable clinging and vertical walking.

Landing on tree trunks likely drove the first major change in pterosaurs, once they became volant. That forged the evolution of a longer forelimb. The most primitive pterosaur, MPUM 6009, had relatively longer legs and shorter arms (like its phylogenetic predecessor, Longisquama), but virtually all later pterosaurs had longer forelimbs (see Raeticodactylus for an example).

While some pterosaurs, perhaps anurognathids, were able to find insects on trees, it appears that most pterosaurs found trees a safe haven and a good take-off point. Unless facing down (which appears somewhat doubtful for the larger ones because they were unable to turn their feet backward in the manner of lizards, squirrels and bats, but they were able to point their feet laterally), tree-climbing pterosaurs would have taken off by launching themselves backwards, aided by gravity, twisting quickly to a normal flight configuration, like a woodpecker might do.

Due to a relatively stiff neck, a pterosaur would likely be taking a very close look at the tree trunk while grappling it. Certainly some left or right was possible, but more than 90 degrees would have been unlikely. Pterosaurs could have slept in this position, their forelimbs prevented from overextending by their joints and the propatagium. Since not all trees are vertical, many pterosaurs could have adopted this walking configuration on unobstructed horizontal to diagonal branches and fallen trunks.

As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.

Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.

References
Peters D 2002. A New Model for the Evolution of the Pterosaur Wing – with a twist. Historical Biology 15: 277-301.
Wellnhofer P 1991. The Illustrated Encyclopedia of Pterosaurs. London, Salamander Books, Limited: 1-192.

Fenestration in the Mesosaurus Skull – Piñeiro et al. 2012

Not an Anapsid After All
Traditionally Mesosaurus (Gervais P 1865) has been considered a basal reptile, an anapsid (Laurin and Reisz 1995, Modesto 1999) lacking any temporal fenestration. Most skulls are crushed and slightly scattered, requiring reassembly. It’s been confusing. Now a new paper by Piñeiro et al. (2012) provides strong evidence of a lower termporal fenestra in undisturbed material.

Figure 1. Graciela Piñeiro, lead author on the new mesosaurus paper.

The Piñeiro et al. (2012) Abstract
The Early Permian mesosaurids are the oldest known aquatic amniotes with an exclusively Gondwanan distribution. Although several hundred of complete skeletons have been discovered and intensively studied, the anatomy and taxonomic composition of the group, as well as its phylogenetic relationships remain controversial. Several well-preserved
mesosaurid specimens found in Uruguay justify a new anatomical reconstruction of the skull of Mesosaurus tenuidens, differing from earlier ones especially in the presence of a lower temporal fenestra. The significance of this structure for the evolution of temporal fenestration in amniotes is evaluated according to the two most recent phylogenetic hypotheses, in which mesosaurids are basalmost sauropsids or basalmost parareptiles. A synapsid-like fenestration may be the primitive condition for Amniota, and it may be also a basal condition for parareptiles, because recent phylogenies suggest a basal position for mesosaurids and lanthanosuchoids within that group, and both possess a lower temporal fenestra. Our results also give a moderately strengthened support for diapsid affinities of turtles.

Figure 2. Mesosaurus skull with lateral temporal fenestra. From Piñeiro et al. 2012.

The New Mesosaurus Skull
The lateral temporal fenestra of this Mesosaurus (Fig. 2) occurs between embayments of the jugal and squamosal. Some skulls may have closed off this fenestra. Others may produce a slightly different fenestra shape. This trait may be universal or not. It’s hard to tell, but at least some specimens of Mesosaurus had this trait.

Vindication
Earlier I reconstructed the skull of Mesosaurus with a lateral termporal fenestra based on reassembling  in situ tracings by Modesto (2006). With or without this trait, mesosaurs nested between Claudiosaurus and kin and enaliosaurs (sauropterygians, ichthyosaurs and thalattosaurs). That was not the first time someone proposed a lateral temporal fenestra (reference escapes me at the moment [von Huene 1941), but finding a lateral temporal fenestra broke with current paradigm. Further study of the more primitive and shorter-toothed Stereosternum will hopefully vindicate the appearance of an upper temporal fenestra as well.

Missing Sister Taxa 
Pineiro et al. (2012) attempted to nest Mesosaurus in a pseudoclade of select taxa virtually all of which possessed a lateral temporal fenestra, whether bound ventrally by a jugal/ quadratojugal/ quadrate bar or not. This tree assumed the lateral temporal fenestra appeared only once without convergence. Turtles (a taxon without temporal fenestration) were also included. Sauropterygians and several basal, mesosaur-like, aquatic forms with temporal fenestration, including Claudiosaurus, Hovasaurus, Thadeosaurus and Wumengosaurus were not included. Also missing were Adelosaurus and Acerosodontosaurus closer to the araeoscelids. Importantly, the large reptile tree found these six taxa to be the closest sisters to mesosaurs (represented by Stereosternum). Excluding these taxa is an unfortunate oversight that could have taken the study to the next level because these six taxa displayed various forms of temporal fenestration and mesosaurs nested in the midst of them.

Disagreements on the Phylogenetic Placement
Piñeiro et al. (2012) placed (not nested) Mesosauridae between Synapsida (Eothyris at the base) and Sauropsida (Romeria, etc. at the base), several nodes away from Petrolacosaurus and Ichthyosauria. Piñeiro et al. (2012) also placed Mesosauridae between Milleretidae and the aforementioned Romeria, etc.) following traditional topologies by Laurin and Reisz (1995) for basal amniotes, Reisz et  al. (2007) for “parareptiles,” and Laurin (2004) for other taxa, from Modesto (1999) for mesosaurids. Missing from the above studies were more undulating aquatic reptiles (listed above) like mesosaurs.

What is Happening in This Clade?
The examples of sister taxa in the large reptile tree indicate that the initial appearance of the diapsid configuration in taxa like Eudibamus and Petrolacosaurus, was modified in derived taxa like Araeoscelis (loss of lateral temporal fenestra), Mesosaurus (loss of upper temporal fenestra) and Wumengosaurus and sauropterygians (loss of lower temporal bar).

What the Large Reptile Tree Indicates
The large reptile tree indicates that the diapsid configuration appeared at least twice by convergence and variations thereof also appear by convergence. The lower temporal bar appeared in certain lines and disappeared in others.

Why were turtles included in this study?
Turtles are anapsids derived from diadectomorphs like Stephanospondylus and have nothing to do with diapsids like Petrolacosaurus or Mesosaurus, according to the recovered tree. Rather than elevating mesosaurs to the ranks of derived taxa, the authors proposed lowering the synapsid-like fenestra trait to the basal ranks (see abstract above), suggesting (to them) that turtles lost their diapsid configuration during their evolution. That hypothesis is not supported by the present tree based on a magnitude more taxa that encompasses the entire Reptilia.

As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.

Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.

References
Gervais P 1865. Du Mesosaurus tenuidens, reptile fossile de l’Afrique australe. Comptes Rendus de l’Académie de Sciences 60:950–955.
Laurin M and Reisz RR 1995. A reevaluation of early amniote phylogeny. Zoological Journal of the Linnean Society 113:165-223.
Modesto SP 1999. Observations on the structure of the Early Permian reptile Stereosternum tumidum Cope. Palaeontol. Afr. 35, 7–19.
Modesto SP 2006. The cranial skeleton of the Early Permian aquatic reptile Mesosaurus tenuidens: implications for relationships and palaeobiology. Zoological Journal of the Linnean Society 146 (3): 345–368. doi:10.1111/j.1096-3642.2006.00205.x.
Modesto SP 2010. The postcranial skeleton of the aquatic parareptile Mesosaurus tenuidensfrom the Gondwanan Permian. Journal of Vertebrate Paleontology 30 (5): 1378–1395. doi:10.1080/02724634.2010.501443.
Piñeiro G, Ferigolo J, Ramos A and Laurin M 2012. Cranial morphology
of the Early Permian mesosaurid *Mesosaurus tenuidens* and the evolution of
the lower temporal fenestration reassessed. Comptes Rendus Palevol.
von Huene F 1941. Osteologie und systematische Stellung von Mesosaurus. Palaeontogr. Abt. A. 92, 45–58.