Saturday, March 12, 2022

Theropod Database Update March 2022 with SVP 2021 thoughts

And after a few months we're back.  This update incorporates taxa named since last time and also incorporates many of the theropod abstracts from SVP 2021.

Two new ceratosaurs were added, Guemesia and Berthasaura.  The latter is interesting in just how similar to tetanurines, and coelurosaurs and ornithomimosaurs/oviraptorosaurs in particular, the supposed noasaurid is.  Besides the obvious dentary similarities to e.g. Microvenator and highly pneumatic lacrimal, the typical ceratosaurian cervical postzygodiapophyseal lamina is reduced, the brevis fossa is reduced, and like Elaphrosaurus the femoral medial epicondyle is reduced and the fibular crest is separate from the proximal tibial condyles.  The authors recover it as the most basal noasaurid, but this seems unlikely given the Aptian-Albian age.

Among basal tetanurines, Iberospinus and "Maip" were added, each formerly in the Database as a questionable referred specimen of Suchosaurus? girardi and Megaraptoridae indet..  Dynamoterror was updated based on Yun's paper and the SVP abstract on the new specimen.  I've incorporated Paul et al.'s new Tyrannosaurus species as synonyms, but haven't gone into any detail examining the issue yet.

Left ilium of Ornitholestes hermanni holotype AMNH 619 showing plesiomorphic characters not present in pennaraptorans- highly convex supracetabular crest and deep brevis fossa (courtesy of the AMNH).

was famously recovered as an oviraptorosaur by Chapelle et al. (2021), but while their CT scanning is amazing and I can't wait for the resulting paper, I'm skeptical of that phylogenetic placement.  For one, some of their proposed synapomorphies are slightly closer to the oviraptorosaurian condition than neutral, but defining the state this loosely would result in other taxa also exhibiting the condition.  The supposedly pendant paroccipital processes for instance are slightly downturned, but not nearly as much as oviraptorosaurs' and e.g. Allosaurus has a more extreme condition.  They don't count as either short/deep or downturned using standard TWiG scoring or the Lori matrix.  The other issue is that the postcranium is so plesiomorphic that basically all pennaraptoran characters become convergences between oviraptorosaurs and paravians, which might be true, but does keep an oviraptorosaurian Ornitholestes at 18 steps longer than the shortest trees in the Lori matrix even with the new data from CT scans.  However, that data did lead to Ornitholestes moving crownward to be sister to Maniraptoriformes instead of the most basal maniraptoromorph.

"Fukuivenator" just got a huge redescription (Hattori et al., 2021), with excellent CT scans in every view.  They corrected most of what the original description got wrong (which was even more than I thought), and surprise(!) the weirdly large premaxilla WAS the other maxilla all along.  Note while they agreed with my reinterpretation of pedal phalangeal assignments, their materials list incorrectly says supposed IV-2 is actually III-2 and supposed III-1 is actually II-2, which besides just being wrong anatomically and contradicting their own figures would leave us with an extra III-2 and no III-1.  They list three "Withdrawn elements: left pterygoid; posterior caudal vertebra; left ischium" without comment, but Hattori (pers. comm. 3-4-2022) indicates the pterygoid and ischium cannot be identified in the existing materials and that the caudal is theropodan but not part of the holotypic block.  The Lori paper recovered "Fukuivenator" as the most basal alvarezsauroid but found it could be a therizinosaur with just two steps.  Hattori et al. actually recovered the latter topology, but scoring the new data in the Lori matrix leaves it sister to Ornitholestes just outside Maniraptoriformes instead.

Nqwebasaurus had new cranial material described by Radermacher et al. (2021), but they still found it in basal Ornithomimosauria using Choiniere's TWiG matrix.  Adding the new data to the Lori analysis resolved it as sister to the alvarezsauroid-therizinosaur clade, but details on the prootic and other elements in the eventual description could change things.  Pelecanimimus' postcrania were finally described by Cuesta et al. (2022), and incorporating this data moved it to Ornithomimosauria as in the consensus.  But it moves back to Alvarezsauroidea in only five steps (which moves Nqwebasaurus back there too), so it's more important than ever to get the skull (and braincase in particular) described.

Khulsanurus is a recently described alvarezsauroid (Averianov and Lopatin, 2021), but the element described as a distal pubis seems too small to be this bone, and differs from theropod pubes in the posteromedial flange and anteroposteriorly deep apron.  Considering its size and shape, this is more likely to be a proximal dorsal rib. Another new taxon is the parvicursorine Ondogurvel alifanovi described by Averianov and Loptatin (2022).  They reported it differs from Parvicursor "by the dorsally arcuate supraacetabular crest of the ilium, by the tibia less curved labially in transverse plane, by less expanded proximal part of the fibula, and by a relatively shorter pedal phalanx II-1."  While these differences are real, the femur was said to be "almost identical with the femur in Parvicursor" and the two emerge as sister taxa if entered into the Lori matrix.  As they are from the same formation and Parvicursor is a juvenile with a tibia 62% of the size of Ondogurvel, it is hypothesized here that the taxa are synonymous and the shorter pedal phalanx and fused metatarsus are ontogenetic differences while the others could be individual variation (tibial curvature is known to vary in Microraptor, for instance).  Averianov and Lopatin added Ondogurvel to a TWiG analysis and used implied weighting to recover it sister to Albinykus, but in their earlier Parvicursor redescription noted this was done because equal weighting led to large polytomies in Parvicursorinae.  Thus it does not strongly argue against synonymy with Parvicursor.  More material of the Bissekty alvarezsaurid were just described by Averianov and Sues (2022 online), which they name Dzharaonyx.  As with other Bissekty taxa, remains are isolated so may represent more than one taxon.  In this case, each referred ulna differs in that "The round facet for the aponeurosis tubercle of the radius is clearly present in the mature specimen ZIN PH 2020/16 but is absent in the juvenile ZIN PH 2021/16", and manual ungual I examples differ in that "In two specimens (ZIN PH 2445/16 and 2446/16), the collateral grooves are enclosed by notches whereas they pass through ventral foramina on the third specimen (ZIN PH 2618/16)."  While these could be individual variation, they could also indicate different taxa.

Holotype of Shuvuuia deserti IGM 100/975. Courtesy of the AMNH.

Eudromaeosaurs have gotten quite a bit of recent attention, with "Vectiraptor" from England and lots of new data on Mongolian taxa.  "Airakoraptor" was officially named Kuru (Napoli et al., 2021; Ruebenstahl et al., 2021), and again I express my displeasure for giving any taxon a genus name that is already a word because it's terrible for finding mentions in the literature or online.  Just call it "Kururaptor"- that word didn't exist before, so it would be very easy to search for every reference to it.  The Kuru talk and paper also clarified what's been a confusing issue, that supposed Velociraptor ulna IGM 100/981 was called that due to a clerical error, and is actually part of partial skeleton IGM 100/3503.  Although called Velociraptor sp. by those works, I would say it falls within the range of variation of Tsaagan.  Another Mongolian eudromaeosaur issue is the new long-snouted species of Velociraptor based on IGM 100/982, which has been hinted at since Kundrat (2004) and will be finalized once Powers' (2020) thesis chapter is turned into a paper.  What might not be as obvious is that the specimen is responsible for a good portion of what we thought was described V. mongoliensis anatomy thanks to Norell and Makovicky (1999), especially the manus.  This makes description of fighting dinosaurs V. mongoliensis skeleton IGM 100/25 even more important.

The new Crato bird pes Kaririavis was described by Carvalho et al. (2021 online).  It's weird in mixing euornithine characters like a plantarily displaced third metatarsal and hypotarsus with primitive characters like no distal fusion and a low intercondylar eminence.  The authors added it to O'Connor's bird analysis and stated it emerged in a large ornithuromorph polytomy with Piscivoravis, Yanornis, Yixianornis, Songlingornis, Iteravis, Gansus, hongshanornithids, Apsaravis, Ichthyornis, hesperornithines and neognaths in a reported 15 MPTs of 1247 steps.  However their supplementary info shows "the 15 Most Parsimonious Trees (MPTs) resulted in present phylogenetic analysis", but in all of these Kaririavis is the sister taxon to Ichthyornis.  Running the matrix actually results in 1104 shorter trees of 1246 steps where Kaririavis can go anywhere in Ornithothoraces except Hesperornithes+Aves and Schizoouridae sensu Wang et al., including trees where it is an enantiornithine.  Just to make it more complicated, Carvalho et al. claim that "35 characters [were] treated as ordered" when their included NEXUS file leaves all characters unordered by default, and when those characters (identified from the Mengciusornis analysis this was taken from) are ordered, 720 MPTs of 1274 steps are found where Kaririavis is the basalmost euornithine in each one.  When added to the Lori analysis Kaririavis can equally easily fall out as an enantiornithine or a neognath, and with a single extra step can be sister to all other euornithines with a plantarily displaced third metatarsal.  But the matrix isn't great at handling isolated metatarsi, so I'm not too confidant about any specific placement and just keep it as Ornithothoraces incertae sedis for now.

Sacrum of Bambiraptor feinbergi holotype AMNH 30556 in ventral view, anterior to left. Courtesy of the AMNH.

Two new enantiornithines were named- "Brevirostruavis" and Beiguornis.  The latter's description sadly lacks a measurements table, but the photos are better quality than usual for this kind of paper.  As for "Brevirostruavis", it was presented as a short-beaked bird with long ceratobranchials, but when reconstructed the beak isn't all that short, and no data is ever given as to shorter ceratobranchial length in other Jehol enantiornithines.  This seems to be a trend with Jehol enantiornithines, where the splashy headline is a supposed weird characteristic that doesn't end up having great evidence- horned Dapingfangornis, predatory-beaked Boluochia, Cathayornis caudatus with lots of unfused caudal vertebrae, tiny Liaoxiornis, Chiappeavis with a weird large premaxilla, heterodactyl Dalingheornis. etc..  I think that once Jehol enantiornithines are examined statistically with taphonomy and ontogeny in mind (e.g the huge changes in sternal ossification and thus shape throughout growth), a large number of genera are going to be sunk.

Finally, we have the euornithines of which the new taxa are O'Connor et al.'s (2021, 2021 online) "Meemannavis" and "Brevidentavis".  The former was found in a huge polytomy by O'Connor et al. in equally weighted strict consensus trees (note the trees in their figure are majority rule and implied weighting), but sister to Archaeorhynchus in the Lori analysis, which seems plausible.  I think this recent trend (also seen in the Ondogurvel paper, Parvicursor redescription and just this week the Dzharaonyx paper) to present implied weighting trees as a solution to a big polytomy is misleading.  As the new Musivavis paper shows (Wang et al., 2022- Fig. 14), weighting can give different results from the unweighted trees we've been using for decades and different strengths of weighting can give different topologies from each other.  The alvarezsaurid papers used a k value of 3, O'Connor et al. used a k value of 12, and none of them even try to justify their choice.  Thus instead of giving defensible cladograms, we get trees generated from arbitrarily skewing the data.  In any case, "Brevidentavis" was presented at SVP at "Brachydontornis" which also made it into a few parts of the resulting paper.  Creisler (DML, 2021) revealed "Brevidentavis" is the intended final name, and the paper will presumably be fixed once it gets out of Early View and the taxa are official.

References- Norell and Makovicky, 1999. Important features of the dromaeosaurid skeleton II: Information from newly collected specimens of Velociraptor mongoliensis. American Museum Novitates. 3282, 45 pp.

Kundrat, 2004. Two morphotypes of the Velociraptor neurocranium. ICVM-7 Abstracts. Journal of Morphology. 260(3), 305.

Powers, 2020. The evolution of snout shape in eudromaeosaurians and its ecological significance. Masters thesis, University of Alberta. 437 pp.

Averianov and Lopatin, 2021 online. The second taxon of alvarezsaurid theropod dinosaurs from the Late Cretaceous Khulsan locality in Gobi Desert, Mongolia. Historical Biology. Latest Articles. DOI: 10.1080/08912963.2021.2000976

Carvalho, Agnolin, Rozadilla, Novas, Ferreira Gomes Andrade and Xavier-Neto, 2021 online. A new ornithuromorph bird from the Lower Cretaceous of South America. Journal of Vertebrate Paleontology. e1988623. DOI: 10.1080/02724634.2021.1988623

Chapelle, Norell, Ford, Hendrickx, Radermacher, Balanoff, Zanno and Choiniere, 2021. A CT-based revised description and phylogenetic analysis of the skull of the basal maniraptoran Ornitholestes hermanni Osborn 1903. The Society of Vertebrate Paleontology Virtual Meeting Conference Program, 81st Annual Meeting. 81.

Hattori, Kawabe, Imai, Shibata, Miyata, Xu and Azuma, 2021. Osteology of Fukuivenator paradoxus: A bizarre maniraptoran theropod from the Early Cretaceous of Fukui, Japan. Memoir of the Fukui Prefectural Dinosaur Museum. 20, 1-82.

Napoli, Ruebenstahl, Bhullar, Turner and Norell, 2021. A new dromaeosaurid (Dinosauria: Coelurosauria) from Khulsan, central Mongolia. American Museum Novitates. 3982, 47 pp.

O'Connor, Lamanna, Harris, Hu, Bailleul, Wang and You, 2021. First avian skulls from the Lower Cretaceous Xiagou Formation, Gansu, China. The Society of Vertebrate Paleontology Virtual Meeting Conference Program, 81st Annual Meeting. 196.

O'Connor, Stidham, Harris, Lamanna, Bailleul, Hu, Wang and You, 2021 online. Avian skulls represent a diverse ornithuromorph fauna from the Lower Cretaceous Xiagou Formation, Gansu Province, China. Journal of Systematics and Evolution. Early View. DOI: 10.1111/jse.12823

Radermacher, Fernandez, de Klerk, Chapelle and Choiniere, 2021. Synchrotron μCT scanning reveals novel cranial anatomy of the enigmatic Early Cretaceous South African coelurosaur, Nqwebasaurus thwazi. The Society of Vertebrate Paleontology Virtual Meeting Conference Program, 81st Annual Meeting. 213-214.

Ruebenstahl, Napoli, Bhullar, Turner and Norell, 2021. Two new eudromaeosaurs from Khulsan (central Mongolia) reveal modern-like faunal predatory structure amoung non-avian dinosaurs. The Society of Vertebrate Paleontology Virtual Meeting Conference Program, 81st Annual Meeting. 222-223.

Averianov and Lopatin, 2022 online. A new alvarezsaurid theropod dinosaur from the Upper Cretaceous of Gobi Desert, Mongolia. Cretaceous Research. DOI: 10.1016/j.cretres.2022.105168

Averianov and Sues, 2022 online. New material and diagnosis of a new taxon of alvarezsaurid (Dinosauria, Theropoda) from the Upper Cretaceous Bissekty Formation of Uzbekistan. Journal of Vertebrate Paleontology. Latest Articles. e2036174. DOI: 10.1080/02724634.2021.2036174

Cuesta, Vidal, Ortega, Shibata and Sanz, 2022 (online 2021). Pelecanimimus (Theropoda: Ornithomimosauria) postcranial anatomy and the evolution of the specialized manus in Ornithomimosaurs and sternum in maniraptoriforms. Zoological Journal of the Linnean Society. 194(2), 553-591.

Wang, Cau, Luo, Kundrat, Wu, Ju, Guo, Liu and Ji, 2022. A new bohaiornithid-like bird from the Lower Cretaceous of China fills a gap in enantiornithine disparity. Journal of Paleontology. FirstView. DOI: 10.1017/jpa.2022.12


  1. As discussed by Madzia and me (2017:, Implied Weighting should be used to test the impact of homoplasy on relationships, and not just to produce more resolved topologies. In any case, trees produced using very aggressive K values (in particular, with K = 5 or less) should place fragmentary OTUs in very bizarre and unlikely placements (e.g., if I re-run my theropod tree with K = 3, Velociraptor osmolskae is usually placed in Ceratosauria).

  2. What's your preferred range of K values then, Andrea?

    1. In general, I explore a wide range of K values, then I select those results most interesting for the discussion (as I wrote above, each particular tree is not significant per se, but it might help illustrating the impact of homoplasy relative to other topologies). In some papers, I used 3-27, in others 5-20: usually, the differences between the two approaches are minimal: in general, the greatest changes in topology are with K values < 6; with K values higher the topologies usually (but not always) converge to the same tree. This happens because with large K values the weight of any extra step in topology is rapidly reduced respect an unweighted setting.
      So, the mere fact that beyond a certain K all trees are (almost) identical does not necessarily mean that that topology is the "true tree", it is just a consequence that the larger the K value the lowest the impact of homoplasy in producing a tree different from the unweighted setting.

  3. Mark Norell likes naming dinosaurs after Buddhist deities. Apsaravis ukhaana, Citipati osmolskae, Erketu ellisoni, Beg tse, Shri devi, Kuru kulla... at least the first three had unique specific names. How hard is it to use "Begtseia rugosus", "Shridevi macronychus", or "Airakoraptor kurukullae"?

  4. Cristatusaurus (Lacerda et al. 2021), “Megalosaurusingens (Soto et al. 2020), Phuwiangvenator (Samathi et al.; these authors also report a proximally wedge-shaped metatarsal III in Cristatusaurus, nullifying one of Chilantaisaurus’ alleged avetheropod characters), Cretaceous avians (Mayr 2022) & Piksi (Martin-Silverstone et al. 2016) also need to be updated.

    1. Thanks for informing me of Martin-Silverstone's comments on Piksi maybe being theropod after all. I was unaware of them and need to look into each side's arguments.

      As for Lacerda et al.'s study, I'm unconvinced by its underlying method's validity. In their morphometric analyses you can always (except for figure 2a) draw a boundary around Suchomimus and Cristatusaurus excluding other taxa. And in their morphometric phylogenetic analyses, I'd like to see a proof of concept before with e.g. Allosaurus or tyrannosaurids because I think premaxillary shape is quite variable within a theropod species. Thus I don't trust characters like "(1) the first premaxillary alveolus of C. lapparenti is more posteriorly positioned on the rostrum; (2) the fourth premaxillary alveolus of C. lapparenti is slightly larger, which is also (3) more spaced from the third premaxillary; (4) the edge of the third premaxillary alveolus is slightly more ventrally placed in C. lapparenti; and (5) C. lapparenti has the posterodorsal region of the premaxilla slightly less expanded, with a less prominent sagittal crest" to be taxonomically significant.

      Then when it comes to the claimed morphological differences, Lacerda et al. seem to get things wrong. They claim the proposed autapomorphy by Sales and Schultz (2017) was "a very convex secondary palate that extends below the line of the premaxillary teeth, a morphological condition that differs from both Suchomimus tenerensis and Baryonyx walkeri", but the latter paper actually says "This condition is also seen in Oxalaia, Baryonyx, and the African spinosaurines (Fig 4A and 4C), and possibly in Irritator and Angaturama [3,5,28,62]. On the other hand, the convex palate of Suchomimus can only be seen in lateral view through breaks in the referred rostrum MNN GDF501", which makes the Suchomimus paratype the outlier with this potential individual or taphonomic difference from other spinosaurids. Similarly they say "it is recovered as the basalmost
      taxon due to the absence of crown striations (character 142
      [0]) and by the presence of premaxillary teeth serrations (character
      149[0]), two conditions that differ from other spinosaurids (Sales
      and Schultz 2017)", but the latter paper never claims this (maybe it's scorings in their matrix?). Baryonyx has been described as having serrations on all teeth (Charig and Milner, 1997) and Ceratosuchops explicitly so on premaxillary teeth (Barker et al., 2021). Striations have not been shown to be taxonomically variable in spinosaurids and the premaxilla of Suchomimus (MNN GDF501) seems to lack preserved teeth anyway.

      So yeah, I'm not convinced.