Haya 2021: still suffering from taxon exclusion

Barta and Norell 2021
give us a detailed look at every bone in the basal ornithischian, Haya griva (Fig. 1). We looked at Haya earlier here and nested it close to Pisanosaurus in the large reptile tree (LRT, 1810+ taxa).

Figure 1. Haya in lateral view.

Figure 1. Haya in lateral view.

For reasons unknown
Barta and Norell did not include Chilesaurus and Daemonosaurus (Fig. 2) in their text or phylogenetic analysis.

Figure 1. Skulls of Daemonosaurus, Haya and Jeholosaurus to scale.

Figure 1. Skulls of Daemonosaurus, Haya and Jeholosaurus to scale.

The hypothesis of interrelationships 
that nested Chilesaurus and Daemonosaurus as phytodinosaurs basal to Ornithischia (Fig. 2) has been online since 2011.

Figure 2. Subset of the LRT focusing on the Phytodinosauria with Buriolestes at its base.

Figure 2. Subset of the LRT focusing on the Phytodinosauria with Buriolestes at its base.

No matter how much detail you put into your study of a taxon
it is all for naught if you decide to exclude pertinent taxa. You will not understand the phylogeny of that taxon, how it relates to others. Haya is a basalmost ornithischian in the LRT, an hypothesis of interrelationships not confirmed by Barta and Norell due to taxon exclusion. They had a chance to deliver big news and muffed it.

The Barta and Norell cladogram suffered from massive loss of resolution
at many nodes. Never a good sign. If you can tell two taxa apart generically, as fossils, you should be able to lump and separate them in a cladogram.

Perhaps too many incomplete taxa were tested.
Don’t include incomplete taxa until you have your tree topology all worked out first.


References
Barta DE and Norell MA 2021. The osteology of Haya griva (Dinosauria: Ornithischia) from the Late Cretaceous of Mongolia. Bulletin of the American Museum of Natural History 445: 1-111.

 

Isaberrysaura: close to Stegosaurus in the LRT

Missed this one three years ago.
Salgado et. al 2017 documented a stegorsaur-like skull (Fig. 1) and some post-crania preserved with a gut full of seeds they named Isaberrysaura.

Figure 1. Isaberrysaura bones and diagram from Salgado et al. 2017. Colors added. Long-legged bipedal restoration is the original.

Figure 1. Isaberrysaura bones and diagram from Salgado et al. 2017. Colors added. Long-legged bipedal restoration is the original, revised here to more closely match Stegosaurus.

Originally this taxon was considered
a long-legged biped (Fig. 1), but here in the large reptile tree (LRT, 1747+ taxa), and elsewhere (Han et al. 2018), Isaberrysaura nests with Stegosaurus (Fig. 2).

Figure 3. Jurrasic Park 4 giant Stegosaurus (above, highlighted by Photoshop) and to scale with President Obama (below).

Figure 3. Jurrasic Park 4 giant Stegosaurus (above, highlighted by Photoshop) and to scale with President Obama (below).

Figure 3. Subset of the LRT focusing on Ornithischia.

Figure 3. Subset of the LRT focusing on Ornithischia. A few taxa have shifted since several years ago.

References
Han F, Forster CA, Xu X and Clark JM 2018. Postcranial anatomy of Yinlong downsi (Dinosauria: Ceratopsia) from the Upper Jurassic Shishugou Formation of China and the phylogeny of basal ornithischians. Journal of Systematic Palaeontology. 16 (14): 1159–1187. [added after publication]
Salgado L et al. (6 co-authors) 2017. A new primitive Neornithischian dinosaur from the Jurassic of Patagonia with gut contents. Nature.com/scientificreports 7:42778 DOI: 10.1038/srep42778

 

A new ornithischian, Changmiania, enters the LRT

Meet a new fossorial fossil dinosaur,
perfectly preserved in its own burrow.

Yang et al. 2020 bring us
self-buried specimens of Changmiania liaoningensis (Yixian, Early Creteceous; Figs. 1–3).

Figure 1. Changmiania in situ and illustrated in situ from Yang et al. 2020.

Figure 1. Changmiania in situ and illustrated in situ from Yang et al. 2020.

The authors nest this ornithischian
as the basalmost member of the clade Ornithopoda (= duckbills, etc.; Fig. y).

Figure y. Cladogram of the Ornithischia from Yang et al. 2020. Colors added. Green= related taxa in the LRT. Yellow = taxa share with the LRT.

Figure y. Cladogram of the Ornithischia from Yang et al. 2020. Colors added. Green= related taxa in the LRT. Yellow = taxa share with the LRT. Compare to figure x which includes more outgroup taxa to polarize the basal taxa.

By contrast,
the large reptile tree (LRT, 1733+ taxa; subset Fig. x) nests Changmiania (Figs. 1–3) with Kulindadromeus (Fig. 4) and Heterodontosaurus (Fig. 5). That’s several nodes apart from the clade Ornithopoda in the LRT (Fig. x).

Figure x. Subset of the LRT focusing on Ornithischia. This cladogram differs considerably from that published in Yang et al. 2020.

Figure x. Subset of the LRT focusing on Ornithischia. This cladogram differs considerably from that published in Yang et al. 2020. Here Haya is a basal ornithischian. In Yang et al. Haya is deep within the Ornithopoda.

Digging (fossorial behavior)
traits found in Changmiania proposed by the authors include:

  1. fused premaxillae
  2. nasal laterally expanded overhanging the maxillas
  3. shortened neck formed by only six cervical vertebrae;
  4. neural spines of the sacral vertebrae completely fused together, forming a craniocaudally-elongated continuous bar;
  5. fused scapulocoracoid with prominent scapular spine;
  6. and paired ilia symmetrically inclined dorsomedially, partially
    covering the sacrum in dorsal view.
Figure 2. Changmiania skull in dorsal view.

Figure 2. Changmiania skull in dorsal view from Yang et al. 2020. Colors added here.

Figure 3. Changmiania skull in lateral view from Yang et al. 2020. Colors added here.

Figure 3. Changmiania skull in lateral view from Yang et al. 2020. Colors added here.

Several ornithischian taxa
(Orodromeus, Oryctodromeus, and Zephyrosaurus) have also demonstrated (or suggested by phylogenetic bracketing) fossorial behavior. Heterodontosaurus (Figs. 5, 6) shares traits #1, 2 and 3. Heterodontosaurus has longer fingers #2 and #3 (for digging?).

Figure 1. Kulindadromaeus, a sister to Heterodontosaurus with proto-feathers. Images from and traced from Godefroit et al. 2014. Since theropods and heterodontosaurs both had something like feathers, if they were the same kind of feathers, their last common ancestor had feathers. That last common ancestor was a herrerasaur or its proximal predecessor. Note the Godefroit et al. skull does not match their description but has a standard maxilla ascending process. See color overlays for correct ed interpretation.

Figure 4. Kulindadromaeus, a sister to Heterodontosaurus with proto-feathers. Images from and traced from Godefroit et al. 2014. Since theropods and heterodontosaurs both had something like feathers, if they were the same kind of feathers, their last common ancestor had feathers. That last common ancestor was a herrerasaur or its proximal predecessor. Note the Godefroit et al. skull does not match their description but has a standard maxilla ascending process. See color examples for correct ed interpretation. Click to enlarge.

Figure 7. Heterodontosaurus skull. Note the fused premaxillae, overhanging nasals and pmx/mx notch for a lower fang.

Figure 5. Heterodontosaurus skull. Note the fused premaxillae, overhanging nasals and pmx/mx notch for a lower fang. The general layout of the skull is very much like that of Changmiania. See figures 2 and 3.

Figure 1. Heterodontosaurus with feather quills arising from the lower back, sacrum and proximal tail.Figure 1. Heterodontosaurus with feather quills arising from the lower back, sacrum and proximal tail.

Figure 6. Heterodontosaurus with feather quills arising from the lower back, sacrum and proximal tail.

It’s worth noting
that many extant birds dig burrows, too. They use their bills to peck and their feet to sweep. Progress is often slow. Here’s an online article discussing birds that use and dig burrows.

Figure 7. Changmiania, both specimens in situ.

Figure 7. Changmiania, both specimens in situ.

Here’s a YouTube video
of a belted kingfisher working on its own burrow. Much is not shown, but the feet are back-kicking out the rubble, presumably produced by the pecking of the strong beak at the back wall of the burrow.

References
Yang Y, Wu W, Dieudonne P-E and Godefroit P 2020. A new basal ornithopod dinosaur from the Lower Cretaceous of China. PeerJ 8:e9832 DOI 10.7717/peerj.9832

Changmiania publicity:
naturalsciences.be/en/news/item/19274

Müller and Garcia 2020 propose an alternate origin for Ornithischia

Müller and Garcia 2020 cast doubt
on all prior hypotheses of origin for the herbivorous dinosaur clade, Ornithischia.

From the abstract:
“Whereas ornithischian dinosaurs are well known from Jurassic and Cretaceous deposits, deciphering the origin and early evolution of the group remains one of the hardest challenges for paleontologists.”

No, it’s not. That problem was solved nine years ago. In the large reptile tree (LRT, 1729+ taxa; subset Fig. 1) Ornithischia + Sauropodomorpha arise from primitive phytodinosaurs like Eodromaeus, Barberenasuchus and Buriolestes. Tell Müller and Garcia to keep adding taxa. They have too few.

Figure 1. Subset of the LRT focusing on dinosaur and ornithischian origins.

Figure 1. Subset of the LRT focusing on dinosaur and ornithischian origins.

The cladogram in Müller and Garcia 2020
suffers from massive taxon exclusion. Incredibly, the basal ornithischians, Chilesaurus and Jeholosaurus (Figs. 1, 2) are not mentioned in the text. Surprising how anyone can discuss the origin of a clade without including its basalmost members?

To their credit, the authors do not mention pterosaurs.

To their discredit, the authors are still trying to push the invalidated relationship between the poposaur dino-mimic, Silesaurus, and the clade Ornithischia.

To their discredit, the authors omit all of the basal bipedal crocodylomorph sisters to Lagosuchus.

Due to taxon exclusion, the authors recover a traditional and invalid Saurischia / Ornithischia dichotomy. Add taxa to solve that problem.

A few reconstructions would have helped the authors identify several taxonomic mismatches.

Figure 1. Chilesaurus and kin, including Damonosaurus and basal phytodinosauria.

Figure 2. Chilesaurus and kin, including Damonosaurus and basal phytodinosauria, from 2015.

Figure 1. Skulls of Daemonosaurus, Haya and Jeholosaurus to scale.

Figure 3. Skulls of Daemonosaurus, Haya and Jeholosaurus to scale. Illustration from 2015.

You can access the Müller and Garcia 2020 paper
and get a free PDF here.


References
Müller RT and Garcia MS 2020. A paraphyletic ‘Silesauridae’ as an alternative hypothesis for the initial radiation of ornithischian dinosaurs. Biology Letters 16(8): 20200417. doi: 10.1098/rsbl.2020.0417 

wiki/Ornithischia

Sanxiasaurus calls into question ornithischian interrelations 

The announcement of the discovery
of the ornithischian, Sanxiasaurus (Ning et al. 2019; Fig. 1) calls into question the topology of the Ornithischia. What are the outgroup taxa and clades? Which are the basalmost taxa?

Figure 1. Ornithischia from Ning et al. Overlay lists missing taxa and outgroups from the LRT (see Fig. 2).

Figure 1. Ornithischia from Ning et al. Overlay lists missing taxa and outgroups from the LRT (see Fig. 2). Evidently the name Sanxiaosaurus above is a typo from the original graphic.

The trouble is
Ning et al. failed to include a long list of outgroup and basal taxa (Figs. 1, 2). So, what are the ramifications?

Figure 2. Subset of the LRT focusing on the Phytodinosauria.

Figure 2. Subset of the LRT focusing on the Phytodinosauria. Take a look at the outgroup for the Ornithischia and basalmost taxa, then see how many are included in figure 1.

As usual,
topologies can change when you add pertinent taxa and delete irrelevant taxa. You can only know which is which when you start with a wide enough gamut cladogram. Only then can you have confidence that your subsets are indeed monophyletic.


References
Ning Li, et al. (11 co-authors) 2019.
 A neornithischian dinosaur from the Middle Jurassic Xintiangou Formation of Yunyang, Chongqing, China: the earliest record in Asia. Historical Biology (advance online publication)
doi: https://doi.org/10.1080/08912963.2019.1679129
https://www.tandfonline.com/doi/full/10.1080/08912963.2019.1679129

wiki/Sanxiasaurus

Convolosaurus enters the LRT basal to pachycephalosaurs

Originally it was considered a basal ornithopod.

FIgure 1. Convolosaurus from Andrzejewski, Winkler and Jacobs 2019, re built from a flock of juvenile specimens.

FIgure 1. Convolosaurus from Andrzejewski, Winkler and Jacobs 2019, re built from a flock of juvenile specimens.

Andrzejewski, Winkler and Jacobs 2019 report,
“The new ornithopod, Convolosaurus marri gen. et sp. nov., is recovered outside of Iguanodontia, but forms a clade with Iguanodontia exclusive of Hypsilophodon foxii. The presence and morphology of four premaxillary teeth along with a combination of both basal and derived characters distinguish this taxon from all other ornithopods.” 

Figure 2. Subset of the LRT focusing on the clade Phytodinosauria. Convolosaurus nests closer to the dome head dinosaurs, not the ornithopods.

Figure 2. Subset of the LRT focusing on the clade Phytodinosauria. Convolosaurus nests closer to the dome head dinosaurs, not the ornithopods.

By contrast
the large reptile tree (LRT, 1419 taxa) nests Convolosaurus basal to the Agilisaurus + Stegoceras at the base of the Pachycephalosauria (dome-head dinos). All these taxa were included in the original paper, but they did not nest together. Andrzejewski, Winkler and Jacobs 2019 did not nest their cladogram on Chilesaurus and Daemonosaurus, two taxa missing from their cladogram. This may have played a part in the different tree topologies.

Then again…
the LRT presently does not include Hypsilophodon or Thescalosaurus, taxa that nest with Convolosaurus in Andrzejewski, Winkler and Jacobs 2019. Soon they will be added. Then we’ll revisit this. 

Figure 3. Convolosaurus cladogram from Andrzejewski, Winkler and Jacobs 2019. Note the complete lack of consensus between the tree topology and figure 2.

Figure 3. Convolosaurus cladogram from Andrzejewski, Winkler and Jacobs 2019. Note the complete lack of consensus between this tree topology and the LRT in figure 2. In the LRT Haya and Pisanosaurus nest together near the base of the Ornithischia, but not here. 

Convolosaurus marri (Andrzejewski, Winkler and Jacobs 2019; SMU 72834; 2.5m in length) informally nicknamed the “Proctor Lake hypsilophodont”, this specimen is known from a flock of subadults.


References
Andrzejewski KA, Winkler DA and Jacobs LL 2019. A new basal ornithopod (Dinosauria: Ornithischia) from the Early Cretaceous of Texas. PLoS ONE. 14 (3): e0207935. doi:10.1371/journal.pone.0207935.

Pampadromaeus in exquisite detail

Langer et al. 2019
bring us new data on the basal phytodinosaur, Pampadromaeus (Cabriera et al. 2011; Fig. 1). Pampadromaeus was a small (1m length) biped with a generalized basal dinosaur morphology.

The authors produced a cladogram
(Fig. 2) focusing on Pampadromaeus and kin. Due to taxon exclusion, they considered Pampadromaeus a basal sauropodomorpha. The LRT (subset Fig. 3) nests it as the Late Triassic last common ancestor (LCA) of Sauropodomopha + Ornithischia.

Figure 2. Basal ornithischia and Pampadroameus, a sister to their common ancestor. Daemonosaurus likely resembled Pampadromaeus, with its long neck.

Figure 1. Basal ornithischia and Pampadroameus, a sister to their common ancestor. Daemonosaurus likely resembled Pampadromaeus, with its long neck.

Unfortunately,
Langer et al. are using an old taxon list that is missing many taxa. In the large reptile tree (1406 taxa, LRT, subset Fig. 3) Marasuchus is a basal theropod, not a dinosaur outgroup. Silesaurus is an outgroup poposaur (a dinosaur-mimic), not a sister to Ornithischia. Ornithischia is a dinosaur in-group, nesting with basal phytodinosaurs and sauropodomorphs. Crocodylomorpha (not included in Langer et al.) is the outgroup for the Dinosauria in the LRT.

Figure 2. Cladogram from xx 2019, colors added. Marasuchus is not an outgroup to the Dinosauria. In the LRT it nests as a basal theropod.

Figure 2. Cladogram from xx 2019, colors added. Marasuchus is not an outgroup to the Dinosauria. In the LRT it nests as a basal theropod.

The differences between the tree topologies
in the Langer et al. cladogram and the LRT appear to result largely from the choice of outgroup. In the LRT the outgroups are not chosen, but are recovered from a list of 1400+ taxa. Oddly, some taxa in the LRT are not included in the Langer et al. study. These include Leyesaurus, Barberenasuchus, Eodromaeus, and one specimen of Buriolestes. The LRT includes no suprageneric taxa, like Ornithischia (Chilesaurus, Daemonosaurus, Jeholosaurus and their descendants).

Figure 4. Subset of the LRT focusing on the Phytodinosauria. Three sauropods are added here.

Figure 3. Subset of the LRT focusing on the Phytodinosauria. Three sauropods are added here.

Experiment with outgroups
If the taxon list for the LRT is reduced to more or less match that of Langer et al. 2019, AND Marasuchus is chosen as the outgroup, the result more closely approaches the Langer et al. tree topology (Fig. 4).

Figure 3. Matching the taxon list to that of xx 2019 and choosing Marasuchus for the outgroup results in this cladogram of basal dinosaur relationships.

Figure 4. Matching the taxon list to that of xx 2019 and choosing Marasuchus for the outgroup results in this cladogram of basal dinosaur relationships using LRT data and taxa.

Adding Euparkeria to this list
(Fig. 4) to create a more distant outgroup for  all included taxa moves Silesaurus outside the Dinosauria and moves Marasuchus and Guabisaurus into the Dinosauria. Basically this is what the LRT recovers with a taxon list similar to Langer et al. 2019.

Figure 5. Adding Euparkeria to figure 4 results in this tree where Silesaurus nests outside the Dinosauria, as it does in the LRT. Theropods nest together, as in the LRT. So do phytodinosaurs.

Figure 5. Adding Euparkeria to figure 4 results in this tree where Silesaurus nests outside the Dinosauria, as it does in the LRT. Theropods nest together, as in the LRT. So do phytodinosaurs.

Pampadromaeus barberenai (Cabriera et al. 2011) is a new dinosaur from the Late Triassic of Brazil. It was originally described as a stem sauropodomorph known from a partial disarticulated skeleton and most of the skull bones. The authors reported, “Based on four phylogenetic analyses, the new dinosaur fits consistently on the sauropodomorph stem, but lacks several typical features of sauropodomorphs, showing dinosaur plesiomorphies together with some neotheropod traits.”


References
Cabreira SF, Schultz CL, Bittencourt JS, Soares MB, Fortier DC, Silva LR and Langer MC 2011. New stem-sauropodomorph (Dinosauria, Saurischia) from the Triassic of Brazil. Naturwissenschaften (advance online publication) DOI: 10.1007/s00114-011-0858-0
Langer MC, McPhee BW, Marsola JCdA, Roberto-da-Silva L, Cabreira SF 2019. Anatomy of the dinosaur Pampadromaeus barberenai (Saurischia—Sauropodomorpha) from the Late Triassic Santa Maria Formation of southern Brazil. PLoS ONE 14(2): e0212543.
https://doi.org/10.1371/journal.pone.0212543
Martínez RN and Alcober OA 2009. A basal sauropodomorph (Dinosauria: Saurischia) from the Ischigualasto Formation (Triassic, Carnian) and the early evolution of Sauropodomorpha (pdf). PLoS ONE 4 (2): 1–12. doi:10.1371/journal.pone.0004397. PMC 2635939. PMID 19209223. online article

wiki/Panphagia
wiki/Pampadromaeus

Looking for a vestigial toe 5 on Jeholosaurus

Jeholosaurus is a small Early Cretaceous sister
to the Late Jurassic Chilesaurus and Late Triassic Daemonosaurus. All three nest as basalmost Ornithischia in the large reptile tree (LRT, 1399 taxa).

Phylogenetic bracketing indicates
a likely pedal digit 5 with a few phalanges should be found on all three taxa. Prior studies failed to reveal it. Current data does not include the pes for Daemonosaurus, nor show the ventral aspect of Chilesaurus, but Jeholosaurus does present the view we’re looking for (Fig. 1). I failed to notice pedal 5 before. I think others have overlooked it as well. Here it is:

Figure 1. Jeholosaurus pes in ventral aspect. DGS colors identify parts of pedal digit 5 disarticulated and broken on the sole of the foot and reconstructed at right.

Figure 1. Jeholosaurus pes in ventral aspect. DGS colors identify parts of pedal digit 5 disarticulated and broken on the sole of the foot and reconstructed at right. This observation is awaiting confirmation or refutation. Phylogenetic bracketing indicates this foot had a pedal digit 5 in vivo.

Finding pedal digit 5 on Jeholosaurus
was made a bit more difficult due to the vestige nature of the digit and its crushed and broken pieces, disarticulated from its traditional alignment lateral to pedal digit 4. This observation based on this photo awaits confirmation or refutation.


References
Han F-L, Barrett PM, Butler RJ and Xu X 2012. Postcranial anatomy of Jeholosaurus shangyuanensis (Dinosauria, Ornithischia) from the Lower Cretaceous Yixian Formation of China. Journal of Vertebrate Paleontology 32 (6): 1370–1395.
Xu X, Wang and You 2000. A primitive ornithopod from the Early Cretaceous Yixian Formation of Liaoning. Vertebrata PalAsiatica 38(4:)318-325.

wiki/Jeholosaurus
wiki/Daemonosaurus

 

 

SVP 2018: A super-matrix for an invalid ‘Thyreophora’

Raven, Maidment and Barrett 2018 report,
“The individual lineages, Ankylosauria and Stegosauria, have been studied thoroughly, but there has never before been a comprehensive whole-group cladistic analysis of Thyreophora.”

According to Wikipedia:
“Thyreophorans are characterized by the presence of body armor lined up in longitudinal rows along the body. Threophora (Nopsca 1915) has been defined (Sereno 1998) as the group consisting of all species more closely related to Ankylosaurus than to Triceratops. Thyreophoroidea was first named by Nopcsa in 1928 and defined by Sereno in 1986, as “ScelidosaurusAnkylosaurus, their most recent common ancestor and all of its descendants”. Eurypoda was first named by Sereno in 1986 and defined by him in 1998, as “Stegosaurus, Ankylosaurus, their most recent common ancestor and all of their descendants”.

Raven, Maidment and Barrett 2018 report,
“Here, the first species-level phylogenetic super matrix of the whole-group Thyreophora is presented, incorporating all previous known cladistic analyses of ankylosaurs, stegosaurs and basal thyreophorans and including all valid species within Thyreophora, for a total of 89 taxa and 338 characters.

Unfortunately
in the large reptile tree (LRT, 1308 taxa, Fig. 1) the last common ancestor of Ankylosauria and Stegosauria also includes among its descendants: heterodontosaurids, lesothosaurs, duckbills and horned dinosaurs.

So ‘Goodbye, Eurypoda and Thyreophora’
(unless these clades someday become more inclusive by consensus). The super-matrix of Raven, Maidment and Barrett 2018 probably does not include the LRT or the taxonomic tree it finds within the clade Ornithischia. So use their study to learn about the stegosaurs and the ankylosaurs, but not the last common ancestor of stegosaurs and ankylosaurs. That would be a sister to late-surviving Late Cretaceous Haya and/or Late Triassic Pisanosaurus.

Figure 4. Subset of the LRT focusing on the Phytodinosauria. Three sauropods are added here.

Figure 1. Subset of the LRT focusing on the Phytodinosauria. Scelidosaurus through Minmi are basal ankylosaurs. Lesothosaurus through Stegosaurus are basal stegosaurs.

References
Raven TJ, Maidment SC and Barrett PM 2018. The first phylogenetic super-matrix of the armored dinosaurs (Ornithischia, Thyreophora). SVP abstracts.

wiki/Thyreophora

Leptoceratops repairs an unresolved node

Earlier a skull only taxon, Auroraceratops, nested with a skull-less taxon, Stenopelix in the Ornithischia. Such non-overlapping trait pairings cannot be resolved. Does one nest basal to another? Or vice versa? Do they nest together? The cladogram cannot tell you, so you end up with 3 possible trees. Sometimes that happens.

Figure 1. Leptoceratops in situ 2x. This taxon resolves the headless and head only node at the base of the ceratopsians in the LRT.

Figure 1. Leptoceratops in situ 2x. This taxon resolves the headless and head only node at the base of the ceratopsians in the LRT. Those ribs are oddly straight. Did they produce a horned-lizard like torso (wider than deep)?

What helps is taxon addition.
Here Leptoceratops (Fig. 1), known from complete material (skull + post-crania) is added to the LRT and returns the basal ceratopsian node to resolution.

Leptoceratops gracilis (Brown 1914; Latest Cretaceous; 2m long) is a late surviving member of an early radiation of bipedal ceratopsians that probably browsed while quadrupedal. The ventral maxilla was convex. The skull was narrower than most. The ribs appear to have flared widely, not deeply, creating a horned-lizard-like torso.

References
Brown B 1914. Leptoceratops, a new genus of Ceratopsia from the Edmonton Cretaceous of Alberta. Bulletin of the AMNH 33, article 36.


Dr. David Unwin reports
the untimely death of Dr. Lü Junchang, author and co-author of dozens of publications coming out of China. I met Dr. Lü long ago when he was working in Texas, early in his career and studies. The pterosaur, Beipiaopterus, was the subject of an early paper in 2002. He was kind enough to send PDFs on his work at every request.