Where are the Bipeds in the Reptile Tree?

A new paper by Kubo and Kubo (2012) discussed bipedalism in archosaurs. They found basal dinosaurs and Poposaurus to be definite bipeds. Possible bipeds in their study included Euparkeria, Crocodylomorpha, Gracilisuchus, Ornithosuchidae and Pterosauromorpha (pterosaurs + Scleromochlus). Seeking to find a phylogenetic connection, Kubo and Kubo (2012) recovered over 12,000 MPTrees. They avoided pterosaurs and most of the bipdal croc-types without comment. They considered Poposaurus a “crurotarsan,” which is traditional thinking that has come into question here.

Figure 1. Click to enlarge. This is the large reptile tree with bipeds highlighted in white and possible bipeds in pink.

The large reptile tree (Fig. 1) includes 300+ taxa and recovers a single tree (when the two poorly known taxa are excluded). Several taxa were facultative bipeds. Others were obligate bipeds (highlighted in white).

In the new Lepidosauromorpha, only one clade was bipedal, the Fenestrasauria (pterosaurs and kin). However 19 living lizards (not listed) are facultative bipeds.

In the new Archosauromorpha several clades were at least faultatively bipedal, from Eudibamus and Lagerpeton + Tropidosuchus to Smok + Postosuchus and basal archosaurs, most of which (excepting theropods) had descendants that reverted to a quadrupedal configuration. In the large reptile tree the only two ornithischian dinosaurs reverted to quadrupedalism and both were armored. Lotosaurus had a finback, reason enough.

Figure 1. Silesaurus as a biped and occasional quadruped. Click for more info.

Kubo and Kubo (2012) considered Silesaurus a quadruped. I think it was rarely a quadruped. They discussed comparisons with certain mammals, which developed more propulsive powers of the forelimb, such as a mobilized scapula. Dinosaurs, with their heavy tails, had a center of balance further back than in mammals. All good thoughts.

By the large reptile tree it appears that reptiles like Lewisuchus and Gracilisuchus originated bipedal locomotion in the Archosauria. Bipedalism conferred certain advantages. Those advantages include breathing while running (avoiding Carrier’s restraint), elevating the skull above obstructions in order to see further, enabling the forelimbs to perform other tasks (such as flapping in Cosesaurus). Increased speed is not associated with bipedality. That many clades reverted to quadrupedalism is just in the nature of evolution, like losing limbs after they had evolved from fins.

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
Kubo T and Kubo MO 2012. Associated evolution of bipedality and cursoriality among Triassic archosaurs: a phylogenetically controlled evaluation. Paleobiology 38(3):474-485.

The Sanctity of the Peer Review Process

An older article (2006) from the New York Times reminds us that the sanctity of the academic journal peer review process cannot be sanctified. A quote from that article on medical review: “Virtually every major scientific and medical journal has been humbled recently by publishing findings that are later discredited. The flurry of episodes has led many people to ask why authors, editors and independent expert reviewers all failed to detect the problems before publication.” Perhaps, like democracy, peer review has its problems, but it remains (at its best) the best solution among many.

Unfortunately…
At its worst the peer review process maintains false paradigms and outdated traditions. Many have been highlighted here, from the nesting of Vancleavea with archosaurs to the nesting of Tetraceratops and Casea with synapsids and mesosaurs with basal reptiles. Clades of scientists review each other’s work and pass through commonly held beliefs unsupported by testing on the scale of the large reptile tree.

Getting Blackballed
I’m currently blackballed due to stepping on several professional toes with my observations and insights. I can’t get the simplest abstract published. Even when I’ve “seen the specimen” it doesn’t seem to matter to the keepers of the status quo. New ideas backed up with hard evidence coming from me are not acceptable.

Take for example the Cosesaurus issue. Here is a taxon with an antorbital fenstra, a pteroid, preaxial carpal, stem-like coracoids, strap-like scapulae, four sacrals, an elongated ilium, a fused pubis/ischium, a pteroid, appressed fibula, reduced metatarsal 5, elongated p5.1 and several other otherwise pterosaurian traits. A paper describing these traits (many had already been published) was rejected by referees in the pterosaur community. The paper included admission of past errors in observation highlighted here. Evidently the refs preferred that the origin of pterosaurs remain a mystery and that someone from their professional ranks should discover the origin of pterosaurs. That has so far fallen flat.

This is why reptileevolution.com and pterosaurheresies.com have come into being. Here, in this blog there has been a mini-paper per day for nearly a year.

I’m comforted by the fact that nothing can be discovered twice and that scientific publication is broadening its horizons to include online data and hypotheses. Science is where you find it, even though some choose to look the other way.

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.

Assembling the Squamate Tree of Life – part 3 – the Tritosauria and the Krypteia

Gauthier (2012) is the largest phylogenetic analysis of the Squamata. Earlier here and here we looked at various aspects of this powerful paper. Today we’ll finish up.

Unfortunately, Gauthier et al. (2012) ignored the descendants of a sister to the stem squamate Huehuecuetzpalli. The large reptile tree found the odd drepanosaurs, tanystropheids and pterosaurs were members of this previously ignored lizard clade, here called the Tritosauria. Having a large and encompassing reptile tree as a guide might have alerted Gauthier et al. (2012) to include tritosaurs, but alas, this clade was not on their radar.

Maybe someday it will be…

The Fossorial Taxa
Fossorial animals dig through dirt and are adapted to living underground. Among squamates, these include certain snakes and amphisbaenians. Gauthier et al. (2012) found that amphisbaenians and all snakes formed a natural clade, which they called the Krypteia (hidden ones).

On the other hand, the large reptile tree found amphisbaenians and dibamids nested within skinks while snakes were diphyletic, arising from two distinct varanid clades, one clade out of Heloderma and Lanthanotus and another clade of snakes out of Ardeosaurus and Adriosaurus.

As in other prior lizard/snake trees the very derived, very tiny snake, Leptotyphlops, nested at the base of all snakes in the Gauthier (2012) tree. I never understood this. Even the jaws don’t even move up and down like all other tetrapod jaws!! It’s hard to tell what’s what in Leptotyphlops with so many bones gone or fused. In the large reptile tree it’s basically the very last taxon in the new Lepidosauromorpha and more derived than the other burrowing snakes.

According to Gauthier et al. (2012) this most basal of snakes was related to a basal amphisbaenian, Spathorhynchus, despite the many basic differences. The large reptile tree recovered Spathorhynchus as a fossorial skink, distinct from all snakes.

The large reptile tree concentrated on taxa at the bases of reptilian clades and subclades in order to recover relationships. It cannot compete with the Gauthier et al. (2012) tree once one gets into the various squamate clades, but at the bases the Gauthier et al. (2012) tree lacks several key taxa that could prove to be important. From the Scleroglossa, the Gauthier et al. (2012) tree lacks Liushusaurus, Eolacerta, Yabeinosaurus, Tamaulipasaurus, Bahndwivici and Cryptolacerta.

What would their inclusion do to the Gauthier et al. (2012) tree?
It’s puzzling how such a large and carefully scored tree as the Gauthier et al. (2012) tree could arrive at so many key oddball (mis-matching) sisters. Perhaps more fossil taxa could have brought the two large trees to a closer accord.

One Last Hoohah!
Like the large reptile tree, Gauthier et al. (2012) found Estesia to nest closer to Varanus than Heloderma, confirming that the Monstersauria may be polyphyletic.

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
Gauthier, JA, Kearney M, Maisano JA, Rieppel O and Behkke ADB 2012. Assembling the Squamate Tree of Life: Perspectives from the Phenotype and the Fossil Record. Bulletin of the Peabody Museum of Natural History 53(1):3-308. online here.

Assembling the Squamate Tree of Life – part 2 – Tchingisaurus

Updated February 22, 2015 with a new image of Tchingisaurus.

Earlier the squamate tree of life by Gauthier et al. (2012) was introduced. In large part this tree resembled the large reptile tree of reptileevolution.com. Parts of the trees were different from each other. We’ll look at some of those today and later.

Figure 1. Click to enlarge. Tchingisaurus, a basal Gekkotan, according to the large reptile tree.

Tchingisaurus
The nesting of Tchingisaurus (Fig. 1) at the base of a basal scleroglossan clade that included Gilmoreteius (Macrocephalosaurus) is one such difference.  Previously I had not looked at Tchingisaurus, a Cretaceous specimen known from a partial skull preserved in 3D. The Gilmoreteius clade in the Gauthier et al. (2012) paper was nested close to the base of the clade that produced Adriosaurus and mosasaurs and also close to the clade that produced Eichstattisaurus and Gekkotans.

In the large reptile tree Tchingisaurus nested as a sister to Gekko (note the shared lack of any temporal bars), also near the base of the Scleroglossa. Eichstattisaurus nested closer to Ardeosaurus and Adriosaurus. So there was a comparative switch-off between the two clades with Tchingisaurus and Eichstattisaurus nearly trading places. The large reptile tree nested the mosasaurs closer to their traditional sisters, the varanids and snakes, not the gekkotans.

Excluded Taxa
Missing from the base of the Gauthier et al. (2012) tree were the basal scleroglossans Liushusaurus and Eolacerta. Also missing were the basal squamates the Daohugo lizard, Lacertulus, Meyasaurus, Tijubina, Homoeosaurus and Dalinghosaurus. Are these exclusions the cause of the differences in the two trees? And I’m not even including the third squmate clade, the Tritosauria, which we have covered earlier and I’ll touch on again later.

The Iguania
Only three taxa were recovered in the Iguania in the large reptile tree. I wasn’t so interested in their relationships, which I considered relatively uncontroversial, but the larger study by Gauthier et al. (2012) had little resolution at the base of the Iguania. So, maybe this clade is more mysterious and interesting than I first imagined. Or perhaps the Iguania needed to be anchored with the above named saurians.

We’ll take another look at the origin of the snakes and amphisbaenians in the next few days, noting the differences between the results recovered in the large reptile tree vs. the much larger study by Gauthier et al. (2012).

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
Gauthier, JA, Kearney M, Maisano JA, Rieppel O and Behkke ADB 2012. Assembling the Squamate Tree of Life: Perspectives from the Phenotype and the Fossil Record. Bulletin of the Peabody Museum of Natural History 53(1):3-308. online here.

The Limits of Phylogenetic Analysis

In short:

1. Individual variation – Within a species there are tall ones, short ones, robust, gracile, long-waisted, barrel-chested, strong-jawed and various skin colors, all ignoring gender differences and we’re just talking about humans here. As demonstrated earlier with two species of Rhamphorhynchus, as taxa become closer and closer in genetic proximity, the influence of individual variation becomes relatively greater. And that’s entirely natural.

2. Bad data – In fossil research, ancient or cartoonish drawings may be the only data. More rarely bones are mislabeled.

3. Incomplete data – Often in fossil research the specimen is incomplete. Crushing can sometimes be just as difficult because researchers find it difficult to identify individual bones. Here’s where DGS, the digital graphic segregation technique that first colorizes photographs of the bones in crushed skeletons, which enables the operator to digitally restore them to their original positions. This similar to CT scanning techniques which do the same thing but automatically and in three dimensions seeing beneath the matrix.

4. Lack of pertinent characters in the character list – If the character list cannot differentiate an unresolved clade, then more characters are needed to recover resolution.

5. Combinations of the above – The current large reptile tree cannot resolve the exact position of SMNS 12352, which is known from a partial rostrum only. Sister taxa are also incompletely known. Perhaps more characters could resolve this dilemma. Certainly more complete specimens would help.

6. Weird Convergence – Tetraceratops appears to be a synapsid, but when tested with Tseajaia, it nests with Tseajaia. Weird convergence occurs most often in incomplete taxa.

We all struggle with the mad scramble for more data. In Science it’s okay to work with scraps if that’s all you have, then build upon that as more data comes in.

Assembling the Squamate Tree of Life – part 1

Some of the heaviest hitters in paleontology joined forces to produce a 300-page paper (including tons of photos and the data matrix) of squamate phylogeny, including several fossil taxa. Gauthier et al. (2012) takes the reader through the history of squamate studies, discusses some of long standing problems and some of the new molecular studies. 141 extant and 51 extinct species were included. The outgroup consisted of three Rhynchocephalians. 610 characters were tested. 112 trees were recovered, chiefly at the base of the Iguania. The homoplasy index was 0.82, so a great deal of homoplay was present. This was a huge study and powerful due to its size.

Happily most of the Gauthier (2012) tree echoed the results of prior trees and the large reptile tree. At the base of both: Huehuecuetzpalli followed by Iguania and Scleroglossa with the latter divided into Gekkota, Scincomorpha and Anguimorpha. Major differences include: 1) Mosasaurs and their kin at the base of the Scleroglossa. 2) Eichstattisaurus at the base of the Gekkota, 3) Amphisbaenia as the sister to a 4) monophyletic Serpentes (snakes). The large reptile tree found 1) mosasaurs to nest with varanids, 2) Eichstattisaurus to nest with basal snakes close to mosasaurs and their kin, far from the Gekkota, 3) amphisbaenids as sisters to skinks, 4) and diphyletic clades of snakes arising from sisters to Lanthanotus and Adriosaurus.

Figure 1. Click to enlarge. The Gauthier et al. 2012 family tree of the squamates, color added for large clades.

We’ll look at these differences point by point in coming blogs and attempt to dissect the differences and why they occurred.

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
Gauthier, JA, Kearney M, Maisano JA, Rieppel O and Behkke ADB 2012. Assembling the Squamate Tree of Life: Perspectives from the Phenotype and the Fossil Record. Bulletin of the Peabody Museum of Natural History 53(1):3-308. online here.

Proterochampsia Paper

David Dilkes was kind enough to send his new paper on Proterochampsa (Dilkes and Arcucci 2012). His tree follows traditional nestings (Fig. 1). Dilkes and Arcucci (2012) also describe the long journey this odd branch has taken as new taxa were slowly added over the years. Makes interesting reading.

Figure 1. Proterochampsia tree by Dilkes and Arcucci (2012). Green added to highlight relationships recovered by the large reptile tree (Fig. 2). Some notes added in blue and red highlight missing and “by default” taxa that should not be included. Not sure why parasuchians don’t nest closer to proterochampsids here as they often do in other trees, including the large reptile tree. Euparkeria seems out of place there, but does nest close to Riojasuchus in other trees.

Unfortunately, one again, too few taxa were added to this tree to recover the same relationships recovered by the large reptile tree (Fig. 2) in which all sister taxa share larger suites of traits. In the Dilkes and Arcucci (2012) tree you get such odd pairings as Doswellia and Vancleavea, Riojasuchus and Aetosaurus, Euparkeria and Parasuchus among others. Only the taxa within the focus group, the Proterochampsia (node D), are true sisters also recovered by the large reptile tree.

Wisely, Dilkes and Arcucci (2012) left out pterosaurs, which are often nested close to parasuchians and proterochampsids. Unfortunately they left out Lagerpeton, members of the Choristodera and several Youngina/Youngoides specimens, all of which would have helped clarify relationships, according to the large reptile tree.

Figure 2. Left: A segment of the large reptile tree showing what happens when more taxa are included. The Pararchosauriformes form a branch separate from the Euarchosauriformes and develop an antorbital fenestra and foss by convergence. Right: Reducing the branch on the left to include only those taxa chosen by Dilkes and Arcucci (2012) with the addition of the thalattosaur, Vancleavea, mistakenly chosen for inclusion by Dilkes and Arcucci (2012). Here more parsimony in sister taxa, but several forced nestings further toward the base of the tree.

A segment of the large reptile tree (Fig. 2) recovers a different topology because more taxa are included. In the large reptile tree the Proterochampsia were more closely related to parasuchians and choristoderans. All share a dorsal, posteriorly-displaced naris (reversed in Champsosaurus as a snorkel), and several other synapomorphies.

A Distinct Convergent Antorbital Fenestra
We discussed earlier the four times the antorbital fenestra was developed. Check it out. We also earlier discussed the nesting of the large proterochampsid (Fig. 3),

Figure 3. A new specimen attributed to Proterochampsa alongside the holotype specimen.

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
Dilkes D and Arcucci A 2012. Proterochampsa barrionuevoi (Archosauriformes: Proterochampsia) from the Late Triassic (Carnian) of Argentina and a phylogenetic analysis of Proterochampsia.  Palaeontology (advance online publication) 1-33. doi: 10.1111/j.1475-4983.2012.01170.x

A call for disruptive innovation in science publishing

Cooper (2012) quotes: “A disruptive innovation is an innovation that helps create a new market and value network, and eventually goes on to disrupt an existing market and value network (over a few years or decades), displacing an earlier technology. The term is used in business and technology literature to describe innovations that improve a product or service in ways that the market does not expect.” -Wikipedia

Then proposes: Now that Nature Precedings is no more, a new disruptive open data-sharing platform (ODSP) for the life sciences is needed. Based, in part, by the model Nature Precedings established. Here I propose 5 qualities of an ideal ODSP and outline 10 benefits (see Table 1) to scientists for embracing such a potentially disruptive model. 

Table 1. Benefits of An Ideal Open Data-Share Platform

1. Negative data can be reported and shared

2. Preliminary data reporting can foster collaborations

3. Demonstration of feasability and preliminary data for grant applications with shrinking page limits

4. Students can publish their findings on small projects that enable them to establish themselves in scientific research

5. Novel findings can be established in a permanent and citable digital record

6. Findings from unfunded pilot projects can be reported

7. Free general public access to scientific findings

8. Copyright is retained by the creator of the work, the researcher, not the publisher

9. Fast (days) compared to the established peer review model (months)

10. Venue for early crowd-funding of small project 

The above list was published in a new online venue, figshare.com. Is this the wave of the future? Here’s a paragraph from the “about” page on figshare.com.

Figshare allows researchers to publish all of their research outputs in seconds in an easily citable, sharable and discoverable manner. All file formats can be published, including videos and datasets that are often demoted to the supplemental materials section in current publishing models. By opening up the peer review process, researchers can easily publish null results, avoiding the file drawer effect and helping to make scientific research more efficient. Figshare uses creative commons licensing to allow frictionless sharing of research data whilst allowing users to maintain their ownership.

Since each post in The Pterosaur Heresies is a mini-paper, I fully support the aims and objectives of this effort. With regard to palaeontology, so far only Andrew Farke and Greg Paul have made contributions to figshare.com.

References
Cooper D 2012. A call for disruptive innovation in science publishing with a new open data-sharing platform for the life sciences. figshare.com 3.0 2011 Nature Precedings. pp. 2

Zhejiangopterus growth series

Figure 1. Click to enlarge. There are several specimens of Zhejiangopterus. My reconstruction was scaled to the smallest one, the one with the nice lateral view complete skull and mandible. I overlooked the fact that the other specimens were more than twice as tall. That was brought to my attention by a sharp-eyed reader.

An Oversight
Earlier I scaled several azhdarchids. A sharp-eyed reader noted a size problem with the Zhejiangopterus specimen.

Five To Choose From
There are at five specimens of Zhejiangopterus. The smallest has the perfect skull. That’s the one I used to reconstruct the specimen – and I scaled the genus to that specimen. That, of course, overlooks the bigger ones (Fig.1). My bad.

Three Times Taller
The largest Zhejiangopterus was three times as tall as the smallest. If the largest was an adult, the smallest would have been a juvenile, not quite three times larger than a hypothetical hatchling.

Eight Times Smaller
Of the pterosaur adults we know, compared to real and hypothetical egg diameters and pelvic opening diameters, adults appear to have been a constant eight times taller than hatchlings.

Now the fun begins.

The Traditional View of Pterosaur Growth
Most pterosaur workers will stake their careers on the “fact” that baby pterosaurs had big eyes and a short rostrum. That’s why they continue to refuse to include small Solnhofen pterosaurs in phylogenetic analyses.

Here is a good test
The littlest Zhejiangopterus does not have a short rostrum or large eyes. We can rely on the isometric hypothesis of pterosaur growth because embryo pterosaurs have adult proportions, unlike most other tetrapods. This growth series of Zhejiangopterus appears to confirm that. Isometric scaling was used to produce Figure 1.

If someone wants to recreate this growth series with accurate tracings of the five specimens, that might be enlightening.

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
Cai Z and Wei F 1994. On a new pterosaur (Zhejiangopterus linhaiensis gen. et sp. nov.) from Upper Cretaceous in Linhai, Zhejiang, China.” Vertebrata Palasiatica, 32: 181-194.
Unwin D and Lü J. 1997. On Zhejiangopterus and the relationships of Pterodactyloid Pterosaurs, Historical Biology, 12: 200.

wiki/Zhejiangopterus

Is Glaucosaurus an Edaphosaurid?

Yes.
The  single, small, round, big-eyed, small-toothed skull of Glaucosaurus megalops (Fig. 1) was first described by Williston (1915) as most resembling the much larger Edaphosaurus. Broom (1932) linked it to Mycterosaurus from the same locality in Texas. Romer and Price (1940) followed Williston (1915). Reisz (1986) considered Glaucosaurus a synapsid of uncertain affinities, noting a lack of discernible sutures. Most recently, Modesto (1994) nested Glaucosaurus with Edaphosaurus. He noted “…many of the elements of the skull and roof are either absent or damaged…” and, despite the juvenile proportions, he wrote,“Glaucosaurus possesses a suite of autapomorphies which indicates that this form cannot be recognized as a juvenile of any other synapsid taxon.”

Ianthasaurus is a small pelycosaur close to Edaphosaurus, but is known from only post-cranial material, which is frustrating as Glaucosaurus is closely related, but the two taxa preserve different portions of their skeleton.

Figure 1. Left: Glaucosaurus megalops (FMNH UC 691) from Modesto (1994) alongside several  images of Edaphosaurus, Australothyris and Romeria primus for comparison. Despite resemblances to other taxa, Glaucosaurus nests with Edaphosaurus in the large reptile tree. What does the variety in the skull of Edaphosaurus tell us?

Glaucosaurus shares with Edaphosaurus a prefrontal ventral process transversely expanded,  similarly-sized teeth, a lack of canines and a lack of a pterygoid transverse process. In the large reptile tree Glaucosaurus nests with Edaphosaurus, but its worthwhile to also check it against Haptodus.

So, yes, Glaucosaurus does nest with Edaphosaurus in the large reptile tree. And it is a late survivor of an earlier radiation. Interestingly, the addition of this taxon shifted Milleropsis to the base of the Heleosaurus branch, but yesterday the additions of the two Varanosaurus specimens moved it back. It’s only a step away. That’s not unexpected as certain data from this clade is largely crushed and based on various low-rez drawings, rather than hi-rez tracings of the original materials.

In any case, it’s exciting learning about the various paths evolution took and the many, many dead ends that had to happen to bring us today’s living synapsids and diapsids.

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
Modesto SP 1994. The Lower Permian Synapsid Glaucosaurus from Texas. Palaeontology 37:51-60/