SVP 2021 abstracts – 17: Bristol’s pterosaur quad launch study from 2019 returns

Remember 2016
when Rayfield, Palmer and Martin-Silverstone of Bristol University created the following advertisement for a budding paleontologist/engineer student? “The main objective of this proposal is to investigate the effectiveness of the quadrupedal launch [of pterosaurs] and by comparing it with the bipedal launch of birds, test if it was one of the factors that enabled pterosaurs to become much larger than any bird, extant or extinct.”

The Bristol pterosaur workers
did not want to test a quad launch (Fig. 1) vs bipedal launch (Fig. 2) in pterosaurs. Rather this team assumed that pterosaurs took off on all fours based on the ability of tiny 1.2 oz vampire bats to leap and then unfold their wings, then flap, then fly.

Figure 1. Unsuccessful Pteranodon quad launch based on Habib (2008) in which the initial propulsion was not enough to permit wing unfolding and the first downstroke prior to crashing. A push-up is not enough to do this even with the elastic snap effect. The smallest bats manage this by starting with open wings.

Figure 2. Click to play. Successful heretical bird-style Pteranodon wing launch in which the slender hind limbs and huge wings provide the necessary thrust for takeoff in the manner of birds. This assumes a standing start and not a running start in the manner of lizards. Note three wing beats take place in the same space and time that only one wing beat takes place in the Habib/Molnar model shown in figure 1.

Worse yet,
the Bristol team assumed giant azhdarchids could fly, ignoring the vestigial distal phalanges that made azhdarchids flightless before becoming giants. Engineer and inventor, Paul Macready, had to shorten the neck and lengthen the wings of his model Quetzalcoatlus to make it fly (Fig. 3).

Figure 3. Paul MacCready’s flying pterosaur model had longer wings than Q. sp., with its vestigial distal wing phalanges. Here the model and its inspiration are shown to the same length.

I think the Bristol pterosaur team found the student they advertised for in 2016,
If so, he’s now a PhD: Benjamin W Griffin. This guess is based on the observation that this is the second time Griffin and his Bristol team have presented an abstract for pelvic range of motion during a hypothetical quad launch in pterosaurs.

From the Griffin et al. 2021 abstract:
“Pterosaurs were the first vertebrates to develop powered flight as a method of locomotion and subsequently attained sizes unseen in any other flying group.”

Only after becoming secondarily flightless, as described earlier here.

“Launch is the most power intensive aspect with regards to flight, requiring the generation of a launch impulse that will provide sufficient velocity and height to safely begin the flapping cycle. As flying animals increase in size, their capacity for muscular force generation does not increase at the same rate, making this launch impulse harder to generate.”

So far, that sounds sensible. That’s why birds combine a wing flap with a leg launch (Fig. 2). Unfortunately, the quad launch hypothesis separates these two methods for becoming airborne. Between the hind limb leap and the first flap of the pterosaur wing there is a long period of time in which there is no thrust and no lift (Fig. 1), only momentum.

“One hypothesis for how pterosaurs circumvented this issue is that they utilized a quadrupedal
launch, allowing them to incorporate the flight muscles in the launch cycle.”

The tiny bats that can leap into the air like this weigh only 1.2 ounces. Things are different when you are not so tiny (see video of fruit bat below). By contrast, birds the size of pterosaurs incorporate flight muscles in the launch cycle, combining the initial leap with energetic flapping to produce maximum thrust (Figs. 2, 4).

Griffin et al. continued:
“Ornithocheiraeans were a clade of Cretaceous pterosaurs including both Anhanguera and
Ornithocheirus. The largest members of this clade reached wingspans of 6 m. By using Range of Motion (ROM) mapping of the pectoral and pelvic girdles of a generalised 5 m wingspan ornithocheiraean model, we tested the ability of ornithocheiraeans to assume the poses required for quadrupedal launch.”

As the Griffin team did in 2019.

“Additionally, we were able to simulate the effects of soft tissues on the joint mobility of the girdles. This facilitates expansion of the ROM through cartilaginous offsets and restriction of the
ROM by constrictive soft tissues including ligaments and muscles. The ROM maps were then compared against published poses hypothesised to be used in pterosaur launch.”

Published poses? From paleoartists? Why not start from scratch and test quad poses against bipedal poses matched to tracks? See animated figure 4.

Figure 4. Pterodactylus animated matched to tracks. The stick figure in the background is also matched to tracks and this is another source for the Bristol team’s awkward and dangerous hypothesis.

Wonder if the Griffin team ever got rid of the morphological cheats employed by Witton and Habib 2010 (Fig. 5), chronicled here in 2011. You might remember Habib’s idea was to plant the wing finger into the substrate, then roll over it to develop an elastic slingshot effect. The problem is: the wing finger never touches the substrate in any pterosaur track. To make pterosaurs do this Habib and Witton 2010 reduced the three free fingers and put them on top of the wing digit to get them out of the way (Fig. 5). That’s also called cheating. Strangely, no one else noticed as this hypothesis became adopted by museums and artists worldwide.

Figure 5. Errors in the Habib/Molnar reconstruction of the pterosaur manus.

Griffin et al. continued:
“Over 197,000 potential poses were tested for the ball-and-socket joint of the pelvic girdle, and over 591,000 potential poses were tested for the more complex semihellar joint seen in the pectoral girdle.”

Methinks Griffin is ‘milking’ that 2016 offer made by co-authors Rayfield, Palmer and Martin-Silverstone and confessing to it! That is more poses than I would be interested in testing by about 196,900. BTW, the word ‘semihellar’ appears nowhere else in the English language according to a keyword search on Google.

“The ROM maps generated show that the ornithocheiraean model can assume the poses required for a quadrupedal launch.”

Apparently that is the resting (= pre-launch) pose (Fig. 6) for a quadrupedal ornithocheirid. Meanwhile, in the bipedal pose, the center of gravity remains over the extremely tiny (= vestigial) feet, which the animated version (Fig. 6 at left) illustrates much larger, again cheating morphology.

Figure 6. NHM Anhanguera compared to skeletal image in a more upright pose. Free fingers elevated and rotated to show the size and shape of the unguals. Note the tiny feet in the actual Anhanguera, overdeveloped in the NHM model, which also rotates the wing finger away from the plane of the wing.

Griffin et al. continued:
“Additionally, the ROM maps indicate that the model is incapable of using a bipedal posture to generate a launch impulse, even with the largest cartilage offset.”

This will have to be demonstrated, not just said. Methinks they are cheating again in order to make the bipedal model invalid (or to keep their paychecks coming). Note how the model (Fig. 6 left) directs the femora nearly straight down, as if pterosaurs were archosaurs. The femora should be directed much more laterally, as in lepidosaurs, keeping the axis of the femoral head aligned with the axis of the acetabulum.

“This study demonstrates that medium sized pterosaurs could assume the postures required for quadrupedal launch. Next steps will estimate whether sufficient muscular leverage and
power could be generated through a quadrupedal launch cycle.”

Since this 2021 presentation repeats the one presented in 2019, the only cycle apparent here is the recycling of old myths and reconstructions. Scientists should want to falsify their hypotheses in order to get rid of those idea that don’t stand up to scrutiny. That’s when the harshest critics become the greatest allies, whether that feels good or not.

PS
Sharp-eyed readers will note that the university-level textbook “Vertebrate Paleontology” was written by another Bristol University professor, Michael Benton. In addition, David Unwin, who mistakenly imagined a single uropatagium between the hind limbs and lateral toes of Sordes from a displaced wing membrane, also taught at Bristol from 1991 to 1997. So if you wonder why myths persist in paleontology, look no further than this center of influence.

References
Griffin BW, Martin-Silverstone E, Demuth O, Pegas R, Palmer C and Rayfield E 2021. Pectoral and pelvic range of motion constraints on Ornithocheiraen quadrupedal launch. Journal of Vertebrate Paleontology abstracts.
Griffin BW, Demuth OE, Martin-Silverstone E and Rayfield EJ 2019. Simulated range of motion mapping of different hip postures during launch of a medium-sized ornithocheirid pterosaur. Journal of Vertebrate Paleontology abstracts.
Witton MP and Habib MB 2010. On the Size and Flight Diversity of Giant Pterosaurs, the Use of Birds as Pterosaur Analogues and Comments on Pterosaur Flightlessness. PLoS ONE 5(11): e13982. https://doi.org/10.1371/journal.pone.0013982

Here’s a pterosaur proposal doomed to a crashlanding

SVP abstracts – Ornithocheirid hip range of motion (ROM)

SVP 2021 abstracts – 16: Stem gekko? Stem skink? Or neither?

Meyer, Brownstein and Gauthier 2021 report:
“Squamates, with more than 11,000 species, are a major portion of extant tetrapod biodiversity. However, their phylogenetic relationships remain highly contested between hypotheses generated from morphological versus molecular data.”

There is a long history of other workers are also seeing this difference.
Don’t trust genes in deep time studies. Use the hard evidence of traits and fossils.

“This may be due in part to poor sampling of stem members of disparate crown clades, particularly from the Jurassic, when most of the crown ‘backbone’ clades are estimated to have originated.”

No. We have a good sampling in the LRT. Add extant taxa to fossil cladograms, as in the large reptile tree (LRT, 1991 taxa, subset Fig. 2) to make your ‘poor sampling’ better.

Figure 1. Originally pictured as a generic lizard (below), here Eichstattsaurus scaled to the track size walks upright. This taxon nests in the lineage of snakes in the LRT.

Meyer, Brownstein and Gauthier 2021 continued:
“Here, we identify a stem gekkotan from the Kimmeridgian Brushy Basin member of the Morrison Formation. This new species replaces the Tithonian Eichstaettisaurus as the oldest
stem gekkotan currently known.”

Neither is a stem gekkotan. In the LRT Eichstaettisaurus (Fig. 1) is in the snake lineage (Fig. 2), a sister clade to the gekko clade. Just add pertinent taxa to find this out for yourself.

Figure 2. Subset of the LRT focusing on Squamata and the new addition: Paramacellodus (yellow).

Meyer, Brownstein and Gauthier 2021 continued:
“This identification is based on a reexamination of specimen DINO 19514. It consists of a disarticulated partial skull including the maxillae, prefrontals, frontals, parietal, left jugal, right postfrontal and squamosal, partial braincase, both dentaries, and the fused left postdentary
bones. Previous examination of DINO 19514 assigned it to the scincomorph Paramacellodus.”

DINO 19514 (Fig. 3) was described earlier by Evans and Chure 1998.

Figure 3. The DINO 19514 specimen of Paramacellodus from Evans and Chure 1998. Colors, layering, animation and DGS reconstruction added here.

In the LRT the DINO 19514 specimen of Paramacellodus nests with two other Early Cretaceous small to flat-headed lizards, equally small Tepexisaurus and tiny Retinosaurus (Fig. 4). These three are in their own clade basal to the large clade that includes Heloderma + varanids + mosasaurs + skinks + amphisbaemids (Fig. 2).

So Paramacellodus is neither gekko nor skink.

Figure 4. Tepexisaurus and Retinosaurus at full scale on a 72 dpi monitor. Skull of Tepexisaurus enlarged at right. Paramacellodus is similar in size to Tepexisaurus.

Meyer, Brownstein and Gauthier 2021 continued:
“We utilize μCT to re-examine the morphology of DINO 19514 in previously inaccessible detail and find it to be a new taxon that is neither Paramacellodus nor a scincomorph. This new taxon is diagnosed by an enlarged pineal foramen, a relatively wide inter-orbital portion of the frontals (more than 50% of the width of the frontoparietal suture), a postfrontal fused to the postorbital, and a wide parietal nuchal fossa. We incorporated DINO 19514 into a large squamate dataset (165 out of 791 characters x 169 species). Unconstrained maximum- and implied-weights (K=12) parsimony infer it as the earliest-diverging stem gekkotan, sister to a clade containing Eichstaettisaurus, Norellius and crown gekkotans.”

In the LRT (subset Fig. 2) Norellius is another lizard closer to snakes than gekkos. Meyer et al. make no mention of two basal gekkotans in the LRT: Tchingisaurus and Chometokadmon, so one wonders whether or not taxon exclusion is a problem here.

“DINO 19514 grants insight into the condition at the base of Pan-Gekkota. As in other stem gekkotans, it has paired frontals, in contrast to the fused condition of the crown. While incomplete, it is apparent that the subolfactory frontal processes are like those of other stem gekkotans in being intermediate between the ancestral squamate condition and the crown gekkotan condition in which they meet and fuse on the ventral midline. Likewise, the elevated marginal tooth count of DINO 19514 is intermediate between the low ancestral count and the high count of crown gekkotans. A lack of palatal dentition unites it with stem and crown gekkotans, pushing this loss to the Jurassic. The posterior teeth are unicuspid, a feature that it shares with other stem and basal crown-gekkotans.”

Evans and Chure 1998 published on DINO 19514 before the descriptions of and Tepexisaurus (2020). So they had no idea these two would someday be related. Meyer, Brownstein and Gauthier 2021 do not mention these taxa, so they may have omitted them, and the stem gekkotans and stem snakes listed above (Fig. 2). Wait for the paper.

References
Evans SE and Chure DJ 1998. Paramacellodid lizard skulls from the Jurassic Morrison Formation at Dinosaur National Monument, Utah. Journal of Vertebrate Paleontology. 18 (1): 99–114.
Hoffstetter R 1967. Coup d’oeil sur les Sauriers (=lacertiliens) des couches de Purbeck (Jurassique supérieur d’Angleterre). Coloques Internationaux du Centre National de la Recherche Scientifique 163:349–371.
Meyer D, Brownstein CD and Gauthier J 2021. Computed tomography reveals a Jurassic stem-gekkotan from the Morrison Formation. Journal of Vertebrate Paleontology abstracts.

wiki/Paramacellodus
wiiki/Tepexisaurus – not yet posted

SVP 2021 abstracts – 15: The Tetrapod Crown

This is a long abstract,
covering many bases, so hang in there.

Otoo and Coates 2021 report:
“Since its initial description, Whatcheeria [Fig. 1] has become the poster child of post-Devonian stem tetrapods.”

According to the LRT (Fig. 2), this is an inappropriate accolade
because Whatcheeria (Fig. 1) is not a stem tetrapod.

Figure 1. Whatcheeria fossil.

Stem tetrapods = Taxonomic proximal outgroup taxa to frogs, salamanders, caecilians.

Figure 2. Subset of the LRT focusing on basal tetrapods. Colors indicate number of fingers known. Many taxa do not preserve manual digits.

Simply by employing more taxa, in the Large Reptile Tree (LRT, 1991+ taxa; Fig. 2) Early Carboniferous Whathcheeria is a derived taxon with no known descendants. Therefore it is not a good ‘poster child’. It would be better to pick Tersomius (Fig. 3) the last common ancestor of all extant tetrapods.

Figure 3. Tersomius texensis, an amphibamid lepospondyl close to Dendrerpeton.

Otoo and Coates continue:
“But this consensus is matched by uncertainty about the broader membership of the apical stretch of the stem lineage: Whatcheeria, in general, is the ‘safe’ pick while other genera and families flip in and out of the crown.”

See figure 2 above. Taxa don’t flip in and out of the crown Tetrapoda in the LRT.

“Here, we take advantage of the new postcranial redescription of Whatcheeria to re-examine these relationships and explore patterns of tetrapod clades and characters close to the crown node. A new dataset assembled around Whatcheeria provides strong corroboration of the sister group relationship with Pederpes, and that this clade, the Whatcheeriidae, is, in turn, the sister group to all post-Devonian tetrapods (excluding finned relatives).”

The LRT (Fig. 2) does nest Whatcheeria with Pederpes,
but these are taxa without descendants.

“However, the Visean genus Ossinodus, although often presented as a whatcheeriid, is likely not a member of the same group and instead provides a glimpse of an as-yet undersampled
Mississippian- or earlier- tetrapod radiation.”

Ossinodus (Fig. 5) is not related to Whatcheeria (Fig. 1) in the LRT. Instead, Ossinodus is a more basal tetrapod closer to Trypanognathus and Greererpeton (Fig. 4)
the basalmost tetrapods in the LRT (Fig. 2).

Figure 4. Basal tetrapods are all flat, long and with short limbs. Ossinodus is derived in having a shorter body and large limbs.

Figure 5. Ossinodus and Gogonasus to scale with Tiktaalik and Elpistostege.

Otoo and Coates continue:
“Further to this, we find no close links between Whatcheeria and another group of putative close relatives, the anthracosaurs.”

By contrast, in the LRT the Whatcheeria clade is a sister clade
to the Anthracosaurus clade.

The definition of Anthracosauria has varied over time. If often includes Diadectes and kin, which are considered pre-amniotes by most prior workers. By contrast, the LRT nests Diadectes and kin within Reptilia (= Amniota). In the LRT (Fig. 2) Anthracosaurus nests in the clade that starts with Ichthyostega, a clade that left no descendants.

“Anthracosaurs (whole or in part), ‘lepospondyls’ (whole or in part), and colosteids are clustered with early members of the amniote and amphibian total groups. In summary, our results increase crown membership at the expense of the stem, and the baphetids remain as the only limbed clade other than the whatcheeriids excluded from the tetrapod crown.”

In the LRT the first three clades are all widely separated. Baphetids nest with anthracosaurs and whatcheeriids.

“In these trees, support for the tetrapod crown node remains weak and the branching structure throughout much of the tree is unstable.”

That’s what happens with taxon exclusion.

“In our estimates, Caerorhachis and the colosteids might not be far removed from conditions at the crown node.”

The LRT does not rely on ‘estimates’. The LRT does not use the word, “might.” In the LRT Caerorhachis (Fig. 6) nests as a reptilomorph within the Crown Tetrapoda, far from the much more fish-like colosteids, far from the crown group.

Figure 6. Caerorhachis is a reptilomorph and a member of the Crown Tetrapoda.

Otoo and Coates continue:
“Their current placement implies fewer steps between the basal branching members of the crown group and the most crownward stem members.”

The authors fail to mention ‘the most crownward stem members’ (yellow taxa in figure 2).

“Of particular interest, in results of reweighted analyses, colosteids are joined by Aytonerpeton, previously mooted as a colosteid, and these are coupled with the temnospondyls.”

Aytonerpeton (Fig. 7) entered the LRT (Fig. 2) as a collosteid. You heard that here in 2019. These are not related to temnospondyls in the LRT.

Figure 7. Aytonerpeton parts from Clack et al. 2016, restoration added. This taxon nests with Collosteus, Gogonasus in a clade that developed limbs in parallel to those in the Tetrapoda and left no extant descendants.

Otoo and Coates continue:
“The Tournaisian age of Aytonerpeton suggests that this more inclusive crown group is older than previous Visean dates, pegged by the diversity of tetrapod clades known from East Kirkton.”

Except that when taxa are added, Aytonerpeton is not related to Tetrapoda. It’s a tetrapod mimic from a more basal node.

“Correspondingly, this earlier, albeit tentative, minimum date implies a much greater range of Tournaisian tetrapods than currently recognized.”

Perhaps Otoo and Coates need to add taxa.

References
Otoo BK and Coates MI 2021. Old things are new again, Whatcheeria, Aytonerpeton and a Tournasian tetrapod crown. Journal of Vertebrate Paleontology abstracts.

Aytonerpeton enters the LRT as a derived collosteid

SVP 2021 abstracts – 14: Trawdenia endoskeleton

Caron AM, Tietien K and Coates M 2021 bring us new data
on the Middle Carboniferous fish preserved in 3D, Trawdenia planti (formerly Mesopoma planti; NHM P11656; Fig. 1). The endoskeleton has not been published (that I am aware of).

From the Caron, Tietien and Coates abstract:
“The actinopterygian crown group is thought to have diverged close to the Devonian–Carboniferous boundary, but a persistent failure to find Paleozoic taxa to populate
the stems of extant clades obstructs understanding the origin of modern biodiversity.”

By contrast, the large reptile tree (LRT, 1991 taxa) finds a dual origin for bony fish
deep in the Silurian following moray eels and kin.

Figure 1. Trawdenia (formerly Mesopoma) planti. NHM P11656,

From the Caron, Tietien and Coates abstract, continued:
“Here we present new data on the endoskeletal anatomy of Trawdenia planti from the Early Pennsylvanian of Lancashire, UK. This species has already revealed exquisitely preserved details of the hyoid arch, pectoral, axial, and dermal skeletons via microcomputed tomography (μCT), coupled with a partly exposed cranial endocast. The neurocranium and gill
skeleton are now revealed as similarly well preserved, and the newly-exposed morphologies demand revision of existing phylogenetic hypotheses.”

Apparently Trawdenia planti is a renamed Mesopoma planti (Figs. 1, 2),
a taxon we looked at earlier here.

“Unexpectedly, the Trawdenia braincase is reminiscent of the Devonian Moythomasia, but sports a paddlefish-like gill skeleton and a host of other morphological characters that beg the question: are these symplesiomorphies or might Trawdenia be a stem chondrostean?”

Several issues here: 1). Trawdenia/Mesopoma nests with Moythomasia in the LRT (Fig. 2), so similarities should be expected.

2) These taxa do not nest with traditional chondrosteans, like paddlefish (= Polydon, a freshwater basking shark with an operculum, according to the LRT).

3. No traditional chondrosteans (e.g. Acipenser, Polypterus, Amia) nest with one another in the LRT, making Chondrostei an invalid clade.

“In pursuit of answers, we built a dataset comprising broadly-applicable characters from the neurocranium, hyoid and gill arches, and parasphenoid. Preliminary analysis recovered a topology with an ‘ancient fish clade’ (in this instance Polypteriformes plus Chondrostei) branching from the base of the tree and Paleozoic taxa branching from the neopterygian stem, thereby conflicting both with accepted molecular phylogenetics as well as current paleontological consensus. To mitigate the influence of convergence, we then excluded characters contingent upon determinate growth, for example regarding the persistence of fissures and fontanelles, and repeated the analysis. This approach generated intriguing results including a populated chondrostean stem – a radical departure from alternative tree topologies that have dominated since the 1980s.”

While it’s always fun to brag about “a radical departure from alternative tree topologies,” these three authors are a year or two late, hundreds of taxa short and less radical than the LRT.

“We recover Trawdenia as a stem-actinopteran, polarizing downstream character state changes and shedding new light on the evolutionary trajectory of endoskeletal systems across Actinopteri.”

Not so fast, kids. The LRT (subset Fig. 2) employs dozens of omitted taxa and nests Trawdenia = Mesopoma far from the origin of Neopterygii + Chondrostei = Actinopteri.
None of these are valid clades in the LRT.

“By employing modern μCT methods on excellently preserved specimens such as Trawdenia and applying discriminate character choice, we might be finally uncovering the deep
Paleozoic roots of the major extant actinopterygian divisions.”

Better to just add more taxa. Use the LRT as your guide. Trawdenia = Mesopoma is closer to tetrapods than to basal bony fish in the LRT. Testing more taxa all at once can sometime be more illuminating than testing one specimen with µCT scans when trying to figure out hypothetical interrelationships.

References
Caron AM, Tietien K and Coates M 2021. Revisiting Trawdenia planti: Endoskeletal data and and a fresh look at Actionopterygian roots. Journal of Vertebrate Paleontology abstracts: 76.
Coates M and Tietjen K 2018. ‘This strange little palaeoniscid’: a new early actinopterygian genus, and commentary on pectoral fin conditions and function. Earth and Environmental Science Transactions of The Royal Society of Edinburgh , Volume 109 , Issue 1-2: Fossils, Function and Phylogeny: Papers on Early Vertebrate Evolution in Honour of Professor Jennifer A. Clack , March 2018 , pp. 15–31.
Traquair RH 1890. Observations on some fossil fishes from the Lower Carboniferous rocks of Eskdale, Dumfriesshire. Annals and Magazine of Natural History 6:493-493.

SVP 2021 abstracts – 12: Passerines

Steel et al. 2021 discuss the bird clade, Passeriformes
“The major crown bird subclade Passeriformes (= Passerines) comprises >6,000 extant species, making up over half of extant avian diversity.”

Maybe more. Most bird cladograms don’t include moas, chickens and parrots in this clade. Traditionally crows are considered Passeriformes. In the large reptile tree (LRT, 1991+ taxa) crows (Corvus) are basal members of this clade and Passer, the sparrow, is a derived, phylogenetically miniaturized chicken, basal to hoatzins, moas and parrots.

According to Wikipedia:
Passerines are divided into three suborders: Acanthisitti (New Zealand wrens), Tyranni (suboscines), and Passeri (oscines). The passerines contain several groups of brood parasites such as the viduas, cuckoo-finches, and the cowbirds. Most passerines are omnivorous, while the shrikes are carnivorous.

Steel et al. 2021 continue:
“Recent work provides a robust phylogenomic framework for extant passerine diversity, yet limited work has targeted passerine comparative anatomy on a broad phylogenetic scale.

The LRT is one of those ‘limited works’ that employs comparative anatomy on a broad phylogenetic scale. Genomic studies cannot be trusted in deep time studies.

This has contributed to a long-standing misconception that passerines exhibit morphologically “uniform” skeletons, which may have dissuaded workers from pursuing large-scale comparative osteological studies of passerines.”

Interesting, Wish I knew which taxa they nested within Passeriformes because the LRT usually breaks all prior taxonomic conventions. I’m going to guess that parrots and moas were not included because they never are included with sparrows and chickens.

“As such, many components of the passerine skeleton are understudied, and existing morphological matrices tend to identify relatively few phylogenetically informative characters for each skeletal component.

What is ‘relatively few’? And what is enough to lump and separate all included taxa? The authors don’t say, but at least they are studying traits.

“This has hindered interpretation of the passerine fossil record, as many isolated fossils of passerine bones remain unassigned beyond the ordinal or subordinal level. Isolated carpometacarpi are preserved relatively frequently, with numerous passerine carpometacarpi known from the Cenozoic.

“Here, we present a detailed analysis of the passerine carpometacarpus, towards the goal of improving comparative morphological knowledge across Passeriformes. We identify morphological synapomorphies for major passerine subclades, sampling 100 taxa distributed across extant passerine diversity.”

When an analysis keeps a focus on just the carpometatarsus
prior to establishing a valid cladogram that runs the risk of omission and convergence.

“Our approach marshalled both published characters and a substantial amount of previously undescribed anatomical variation, resulting in >50 phylogenetically informative characters that were optimized across a robust phylogenomic scaffold derived from recent studies. Our results show high levels of homoplasy within the passerine carpometacarpus, and that several frequently quoted “diagnostic” characters for particular clades exhibit previously unrecognized homoplasy.”

Sounds like the authors are being rather careful. Glad to see they are working with a last common ancestor method, rather than depending on ‘diagnostic’ characters.

“Despite this, we identify numerous diagnostic character combinations for key passerine clades. The character matrix from this study will aid in future diagnoses of isolated passerine
carpometacarpus fossils, enabling insight into when and where representatives of major passerine clades first appear in the fossil record. This work provides a starting
point for large-scale comparative analyses of the passerine skeleton, and reveals a substantial degree of previously unrecognized morphological variation among Passeriformes.”

Did the authors include moas? Parrots? Woodpeckers? Hoatzins? If not, it’s time to do so.

References
Steell E, Nguyen J, Benson RB and Field DJ 2021. Comparative morphology of the passeine carpometacarpus: implications for interpreting the fossil record of crown passeriformes. Journal of Vertebrate Paleontology abstracts.

wiki/Passerine

SVP 2021 abstracts – 13: Mosaic evolution during the rise of the tetrapods

Simoes and Pierce report,
“The fish-to-tetrapod transition is one of the most iconic events in vertebrate evolution, yet fundamental questions regarding the dynamics of this transition remain unresolved.”

Only due to taxon exclusion. The LRT (subset Fig. 3) resolves most such problems
by nesting flatter taxa with smaller limbs, morphologically closer
to panderichthyids (Fig. 1), just outside of the base of the Tetrapoda (Fig. 3).

Mosaic evolution has been thoroughly debunked (with regard to pterosaurs) here. After nesting more than 2000 taxa in the various cladograms, I haven’t seen it anywhere else either.

Figure 1. Dorsal and ventral views of Panderichthys and several basal tetrapods demonstrating the low, flat skulls and bodies with small limbs and relatively straight ribs.

Simoes and Pierce continue:
“These questions include precisely dating the time of origin of tetrapods, recognizing the closest fish-like relatives to elpistostegalians, and detecting the pace of phenotypic change that led to the origin of the tetrapod body plan.”

There are no such questions in the LRT because it recognizes (Figs. 1–3)
‘the closest fish-like relatives to elpistostegalians.’

“Here we use both recent and new advances in Bayesian morphological clock modeling to reveal the evolutionary dynamics of early tetrapodomorphs (tetrapods and their closest fish relatives) and to answer these long-standing questions. We combine both osteological and ichnological data and introduce a new method to automatically detect morphological character
partitions for inferring evolutionary trees. Further, we adapt a technique from molecular phylogenetics for measuring the strength of selection upon partitions based on rates of morphological evolutionary change.”

This is where workers get into trouble. Stick with traits. Forget genes in deep time studies.

“Our data show that combining osteological and ichnological data results in major shifts on the time of origin of all major groups of tetrapodomorphs (up to 25 Myr into the past) and that low rates of net diversification (not fossilization, as previously suggested) explain long ghost lineages in the early tetrapodomorph fossil record.”

360 + 25 = 385 mya, just after the time of Brindabellaspis.

“Further, our results show that most early tetrapodomorph lineages are characterized by extremely low rates of morphological change, indicating widespread stabilizing selection upon their “fish-like” morphotype.”

Hmmm. Traditional studies bunched most transitional taxa at 360mya or so. The LRT indicates quite a bit of radiation in the late Devonian skeletal record including Tulerpeton, a basal reptilomorph, also appearing at 360mya.

Figure 2. Colosteus relatives according to the LRT scaled to a common skull length. Their sizes actually vary quite a bit, as noted by the different scale bars. Only Pholidogaster and Colosteus are taxa in common with traditional colosteid lists.

Simoes and Pierce continue:
“This pattern was broken only by elpistostegalians, especially among early tetrapods, which underwent sustained high rates of morphological evolution for ~30 Myr.”

The LRT does not measure rates, per se, but note the tetrapod mimics that were phylogenetically more primitive, like Colosteus, extended into the Early Carboniferous as late survivors of an earlier radiation.

“The fastest rates detected were concentrated on the skull, including several adaptations for feeding, suggesting that as primary evolutionary driver towards the acquisition of the tetrapod
body plan.”

First come up with a valid phylogeny. Then discuss rates. Step one has only been documented in the LRT at present due to taxon exclusion in traditional studies.

“Fast phenotypic changes coupled with low taxonomic diversity and inferred low rates of net
diversification of early tetrapods during the Late Devonian provides a strong example of the decoupling between periods of taxonomic and phenotypic radiation.”

Unfortunately this sounds like academic noodling.
Get back to a simple, complete cladogram!

Figure 3. Subset of the LRT focusing on basal Tetrapods and their ancestors.

“This taxonomic-phenotypic rate duality has been recently detected for other vertebrate lineages and may thus be more common than previously assumed by traditional evolutionary theory.”

These workers sound like they have millions of fossils
from hundreds of strata at the base of the Tetrapoda. They don’t.
And I haven’t heard one of them mention Greererpeton (Fig. 1),
the basalmost tetrapod in the LRT (Fig. 3).

References
Simoes TR and Pierce S 2021. Detecting strength of selection and mosaic evolution during the rise of tetrapods. Journal of Vertebrate Paleontology abstracts.

SVP 2021 abstracts – 11: Evolution of dentition in Mammaliamorpha

Mammaliamorpha definition:
“The most recent common ancestor of Morganucodonta and the crown group mammals.”

Since Morganucodon (Fig. 1) is a basal marsupial in the large reptile tree (LRT, 1989+ taxa, subset Fig. 2). So Mammaliamorpha is a junior synonym for Mammalia.
Mammaliamorpha may be a junior synonym for Probainognathia (Martinez et al. 1996).
The definition depends on the taxon list and cladogram.

So when Sulej 2021 reports on “The evolution of dentition in Mammaliamorpha”, he’s actually talking about just Mammalia. If Sulej brings up any outgroups (Fig. 2), those should be beyond the scope of his study based on his headline.

Figure 1. Morganucodon skull in several views.

From the Sulej 2021 abstract:
“It seems that we know a lot about the evolution of cynodonts and the origin of mammals. But if we try to understand the evolutionary changes of the dentition in both groups, some gaps appear in the picture.”

Gaps? Reversals? Or a bad cladogram?
Make sure your phylogeny is correct and does not omit pertinent taxa.

“Brasilitherium and other Brasilidontidae are closest to Morganucodon and Sinoconodon according to recent phylogenetic analyses.”

In the LRT (subset Fig. 1) Brasilitherium is not related to Brasilodon.
Only two of the above taxa are related to one another.

Suggestion: Build your own cladogram.
Don’t depend on (= trust) “recent phylogenetic anlyses”.

Figure 2. Subest of the LRT focusing on basal mammals and outgroups. Morganucodon is in green. Other taxa found in the Sulej text are highlighted in yellow.

Sulej 2021 builds his case on a borrowed falacy based on taxon exclusion.
“However, as Martinelli and Bonaparte noticed in 2011, the postcanine tooth replacement in Brasilidontidae is in the posteroanterior direction, whereas it is in an anteroposterior direction in
basal mammaliaforms (e.g., Sinoconodon and Morganucodon).”

Sinoconodon is not related to Morganucodon in the LRT.

“All Brasilidontidae have the multicuspidated lingual cingulum shelf on the lower postcanines (in Morganucodon on lower and upper) and all postcanines are single rooted (double rooted in
Morganucodon). The recently described Kalallitkigun (only slightly younger stratigraphically than most Brasilidontidae) from Greenland has double roots, and advanced multicusped crowns in the lower postcanines, which further complicates the situation.”

Kalallitkigun is not a taxon listed on Google.

“The new study on the very small Carnian eucynodont Polonodon, that shows a small size difference between the smallest and largest postcanines, suggests that the first step to diphyodont dentition was made through the very small size of the skull, such as in Adelobasileus.”

Polonodon is known only from isolated teeth. Adelobasilius (Fig. 2) is known only from a braincase, similar to Therioherpeton a cynodont close to Sinoconodon (Fig. 1). None of these nest within Mammalia or close to Morganucodon.

Figure 3. Adelobasileus restored like Therioherpeton after first nesting together in the LRT.

Sulej 2021 builds his case
“Analyzing the size of the small teeth of Polonodon, Sulej and coauthors recently suggested that the fast replacement of teeth in very small eucynodonts was possibly the first step in the evolution to diphyodont dentition as an example of heterochrony.”

Mammals are defined by breast feeding. That is inferred in fossils by single replacement teeth (= diphyodonty) in which milk teeth first erupt as breast feeding ceases.

“However, we still do not know the sequence of the appearance of the apomorphies leading to the dentition of Morganucodon, such as double rooted teeth, multicuspidated lingual cingulum shelf, and diphyodont dentition.”

Sulej would know the sequence of the appearance of any and every trait if he had a valid phylogenetic framework, like the LRT. Build your own cladogram. Don’t borrow.

References
Martinez RN, May CL and Forster CA 1996. A new carnivorous cynodont from the Ischigualasto Formation (Late Triassic, Argentina), with comments on eucynodont phylogeny. Journal of Vertebrate Paleontology 16(2):271-284.
Sulej T 2021. Evolution of dentition in Mammaliamorpha. Journal of Vertebrate Paleontology abstracts.

SVP 2021 abstracts – 10: Apterodon taxonomic issues

Updated April 2, 2022
with additional taxa attracting Apterodon back into the marsupial dogs and cats clade.

Al Ashqar et al. 2021 bring us their report
on the most complete Apterodon (Fischer 1880; Fig. 1) cranium ever recovered. We looked at Apterodon earlier here (if you want more info on this taxon and its relatives).

Figure 1. Apterodon macrognathus is similar to the new skull described by these authors.

From the abstract:
“Hyaenodonta is the only group of carnivorous mammals known from the Fayum, and this group shows a broad range of body size and dietary specialization. Apterodontinae is a peculiar group among Hyaenodonta, with fusiform crania and a highly modified dental morphology unknown in other carnassial bearing mammals.”

In the large reptile tree (LRT, 1989+ taxa) Hyaenodon (Fig. 2) is a creodont, a carnivorous marsupial. In the LRT Apterodon does not nest with Hyaendon, but nearby (Fig. 4) within the clade of marspial dogs and cats. The traditional ‘carnassial’ of Apterodon is actually a carnassial-mimic first molar (Fig. 1). Hyaenodon has a carnassial-mimic tooth, too.

Figure 2. Hyaenodon horridus. This carnivorous marsupial was formerly considered a placental creodont. Moving Apterodon and Pterodon to Hyaenodon adds 24 steps to the LRT.

Al Ashqar et al. continue:
“Here, we report the most complete Apterodon cranium ever recovered from Quarry M (~29
Ma), one of the youngest terrestrial localities in the Jebel Qatrani Formation of the Fayum Depression. Based on craniodental morphology, the new cranium is very similar to Apterodon macrognathus [Fig. 1] from Quarry A (~ 33 Ma), also in the Jebel Qatrani Formation.”

“The new skull preserves the complete upper left tooth row aside from I1 and P1 and right I2, P2-3 and M2. Like other apterodontines, the carnassial cutting blade is very reduced, reflecting a possible mesocarnivorous diet.”

That’s a carnassial-mimic in Apterodon. That’s why it looks strange.

Figure 3. Apterodon, Pterodon, Hapalodectes and kin derived from a sister to Thylacinus.

Al Ashqar et al. continue:
“The skull is anteroposteriorly elongate with a narrow, crushed rostrum. The nuchal crest and basicranium are well preserved with two large occipitals. The nuchal crest has a clover shape and does not reach the mastoid process. The neurocranium is complete and elongated. The zygomatic arches are dorsoventrally tall, broad and robust. The petrosal is preserved in the right side. The discovery of this cranium allows detailed craniodental comparisons with the better-known A. macrognathus, better documentation of the evolution of the distinct
apterodontine bauplan, and closer examination of the transtethyan dispersal of Apterodon during the Paleogene.”

Also added to the LRT is the related Pterodon, aka Hyainailouros (Fig. 3). The Early Eocene IVPP V12385 specimen assigned to Hapalodectes (Fig. 3) is also a taxon close to Apterodon.

Once again, adding taxa clarifies interrelationships. With 1990 taxa, the LRT is able to lump and separate taxa that are traditionally confused with one another due to taxon exclusion.

References
Al Ashqar et al. 2021. The cranium of Apterodon (Hayainalouridae, Apterodontidate) from one of the youngest vertebrate bearing locatlies (Early Oligocene) in the Fayum Depression, Eqypt. SVP abstracts 2021: 44.
Fischer P 1880. Note sur un nouveau genre de mammifere fossile (Apterodon gaudryi) des Phosphorites du Quercy. Bulletin de la Société Géologique de France 8: 288-290.

wiki/Apterodon
wiki/Hyaenodon
wiki/Hyainailouros
wiki/Anagale

SVP 2021 abstracts – 09: Dinosaur diversity missing? Still, no.

Napoli 2021 reports:
“A large suite of research is predicated upon the assumption that the incomplete fossil record is
representative of the actual diversity that existed in the past. Based on several lines of evidence, I argue that this assumption is unjustifiable.”

The LRT disagrees. Completely new taxa that don’t fit in the present tree are no longer being discovered and described.

“Approximately 1,300 species of non-avian dinosaurs are known from the entire Mesozoic Era. This diversity pales in comparison to that of modern diverse amniote clades. The mean species
richness of diverse, disparate amniote clades in the present is ~104, while an average of only 7 non-avian dinosaurs are known from any 1 Ma interval in the Mesozoic.”

That doesn’t matter. No weird taxa are showing up anymore. Only more species.

“The vast quantity of missing species in the fossil record is problematic for macroevolutionary and paleoecological research, as the influence of unknown taxa is unquantifiable and consistent with an infinite number of scenarios.”

We will always find new species. No two Pteranodon or Triceratops are alike.

“Field exploration and the discovery of new fossils remain critical components of paleontological research.”

Agree.

“As methods for morphological species delimitation improve, it is likely that collections of individuals currently referred to a single species will prove to represent assemblages of
multiple taxa, especially when the specimens come from a wide spatial or temporal distribution. As species are continually discovered, phylogenetic topologies will likely change considerably, particularly within clades that comprise mostly small taxa such as Paraves.”

Disagree. A good phylogenetic topology will not shift as species are added, except at the species level, not at the larger branches.

References
Napoli JG 2021. The missing diversity of the dinosaur fossil record. Journal of Vertebrate Paleontology abstracts.

SVP 2021 abstracts – 08: Phylogeny of the Azhdarchoidea

This topic has already been documented
in the large pterosaur tree (LPT, 260 taxa), which minimizes taxon exclusion.

Traditionally
workers have mistakenly lumped azhdarchids (like Quetzalcoatlus) with tapejarids (like Tapejara) based on a convergent tall antorbital fenestra, toothless jaws and little else (Fig. 1).

Unfortunately, Thomas 2021 continues this tradition.
Taxon exclusion has been a traditional problem among university professors, their cladograms, textbooks and their students. Small transitional taxa (Figs. 2, 3) are traditionally omitted.

Figure 1. From 2012. Click to enlarge. Putative members of the Azhdarchoidea include Chaoyangopterus and Huaxiapterus. The larger study found no clade uniting just the Azhdarchidae and the Tapejaridae. Instead n42 was found to be basal to azhdarchids and Germanodactylus cristatus n61 was found to be basal to tapejarids.

Thomas 2021 reports:
“The phylogeny of Pterosauria remains a highly contentious topic.”

Not a problem here, in the LPT, where taxon exclusion has been minimized.

“Finer-scale phylogenetic relationships of Azhdarchoidea, and Azhdarchidae in particular, have
been poorly explored as of yet.”

Perhaps so if all the one-bone taxa are included. Otherwise, no problem in the LPT.

“The fragmentary or poorly-preserved nature of many azhdarchoid specimens has precluded their placement in large-scale cladistic analyses of pterosaurs.”

That’s a good thing. Try to employ as many complete specimens before attempting to nest the bits and pieces. Traditional pterosaur cladograms omit far too many taxa.

“The internal relationships of Azhdarchidae have yet to be tested in depth, as the majority of azhdarchid remains consist of isolated or scattered skeletal elements and are poorly covered by character sampling of existing phylogenetic datasets.”

Add taxa. Find out the outgroups to the Azhdarchidae. Professional paleontologists have traditionally omitted dozens since the first cladograms appeared in 2003. You’ll find virtually all, and certainly more, well-represented pterosaurs in the LPT than elsewhere.

“To investigate this, I constructed a phylogenetic dataset of pterodactyloid pterosaurs with a focus on Azhdarchoidea.”

…And that’s the problem. Azhdarchids arose from dorygnathid pterosaurs.
The traditional clade “Pterodactyloidea” is not monophyletic.
The ‘pterodactloid’ grade arose at least 4x (Figs. 2, 3). We’ve known this since Peters 2007.

Figure 2. Click to enlarge. The descendants of Sordes in the Dorygnathus clade and their two clades of pterodactyloid-grade descendants.

Figure 3. Click to enlarge. The base of the Scaphognathia illustrating the size reduction that preceded the size increase in the transition from Scaphognathus to several later, larger “pterodactyloid”-grade clades.

Thomas 2021 continued:
“The taxon sample includes every diagnostic species of azhdarchoid, numerous unnamed azhdarchoid specimens, and an extensive outgroup sampling.”

I have no confidence in the results here because dorygnathids (Fig. 2) are not mentioned.

“The character sampling focuses on the anterior skull, cervical vertebra, and humerus; these elements are commonly preserved in azhdarchoids, and thorough character sampling can help
resolve possible phylogenetic positions of isolated elements. Parsimony-based phylogenetic analysis recovers a dichotomy in Azhdarchoidea between a Tapejaridae-Thalassodromidae-Keresdrakon lineage and a Chaoyangopteridae-Azhdarchidae lineage.”

That’s not news. All prior cladograms recovered the same. Only Peters 2007 and the LPT discovered these clades arose from traditionally different ancestors (Figs. 2, 3) because it includes small transitional taxa that give rise to four pterodactyloid-grade pterosaurs.

“The analysis recovers a novel topology within Azhdarchidae, including a basal Gondwanan clade with prominent modifications of the upper and lower jaws, a predominantly European
clade of robust and sometimes short-necked taxa, and a cosmopolitan clade of derived taxa with particularly elongate cervical vertebrae. Several azhdarchoids of uncertain affinities, including Montanazhdarcho and Leptostomia, are recovered as basal azhdarchids.”

These are based on scaps. Add them AFTER validating the phylogeny of the more complete specimens by first adding the tiny Solnhofen pre-azhdarchid taxa (Fig. 4) that Thomas omits.

Figure 4. Click to enlarge. Here’s the 6 foot 1 inch President of the USA alongside several azhdarchids and their predecessors. Most were knee high. The earliest examples were cuff high. The tallest was twice as tall as our President. This image replaces an earlier one in which a smaller specimen of Zhejiangopterus was used.

Thomas 2021 continued:
“The earliest potential azhdarchids appear in the Aptian–Albian; isolated elongate cervical vertebrae from the Late Jurassic and earliest Cretaceous are recovered as
ctenochasmatids.”

Potential azhdarchids?? See the LPT for azhdarchid ancestry that goes back to Late Jurassic phylogenetically miniaturized transitional taxa (Fig. 4) then back to nonvolant Middle Triassic sprinters.

“After a decline in toothed pterosaurs across the Cenomanian–Turonian boundary, azhdarchids
become the most diverse component of the toothless Coniacian–Maastrichtian pterosaur record. Azhdarchids with giant wingspans (>9 meters) do not form a clade, suggesting giant size evolved multiple times within Azhdarchidae.”

Nesting scrappy material is difficult. It takes more than luck to nest a bill tip with a cervical. Young Henry N Thomas has no idea that azhdarchids arose from tiny Solnhofen taxa (Fig. 4). That’s not what they teach at the university level, nor what students will find in outdated textbooks. So just keep adding taxa to your growing cladogram. Or check out the large pterosaur tree to better understand pterosaur phylogeny.

References
Peters D 2007. The origin and radiation of the Pterosauria. Flugsaurier: The Wellnhofer Pterosaur Meeting, Munich: 27.
Thomas HN 2021. The phylogeny of Azhdarchoidea (Pterosauria) and the rise of toothless pterosaurs. Journal of Vertebrate Paleontology abstracts.