This post is a shout-out to Fonseca et al. (2024), whose abstract states-
"Resolving the evolutionary relationships of early diverging (‘basal’) ornithischian dinosaurs is a challenging topic in palaeontology, with multiple competing hypotheses on the phylogenetic relationships of heterodontosaurids, ‘hypsilophodontids’, and other early-diverging forms. These hypotheses cannot be directly compared because they are derived from differently constructed datasets (i.e. distinct samples of taxa and characters). This study aims to address these issues by revising and combining the distinct datasets into a single analysis in order to create the most comprehensive dataset for the investigation of the phylogenetic relationships of early-diverging ornithischians."
This is the ideal for a phylogenetic analysis and exactly what I attempted for Lori, providing some useful data. They write "constrained analysis forcing a close affinity between Heterodontosauridae and Marginocephalia require at least 26 extra steps, while joining them and Pachycephalosauria require at a minimum 44 steps", so those are implausible. The weird little bipedal armored Jakapil comes out sister to Eurypoda, and "enforcing it as a marginocephalian requires 16 additional steps" so is unlikely. Making it parankylosaurian requires only 4 steps though, so is quite likely and matches geographically. For as much crap as I gave Norman's phylogenetic nomenclature, his Hypsilophodontia including Tenontosaurus (and other rhabdodontomorphs) only takes 6 more steps, so is less likely but not bad. While Fonseca et al.'s main results are interesting, I found the supplementary information where they commented on all the taxa they excluded equally informative.
 |
| Teeth of Ferganocephale adenticulatum- A-B paratype ZIN PH 31/42; C-F holotype ZIN PH 34/32; G-J paratype ZIN PH 36/42; K-N paratype ZIN PH 33/42. Apical- A, C, G, N; Mesial or distal- B, E, I, L; Labial or lingual- D, F, H, J, K, M. After Averianov et al., 2005. |
"Ferganocephale adenticulatum Averianov et al. (2005) displays no ankylothecodonty (ch388:0), low crowns (ch411:0), round crown apices (ch413:1), no apicobasal ridges (ch421:0), an asymmetric labiolingual expansion (ch422:1), a cingulum (ch423:1), and no denticles. While these features show the most similarity to Chaoyangsauridae, the lack of denticles and strength of the cingulum are very different from the clade, so a classification as intermediate Saphornithischia is most likely. The position of Ferganocephale adenticulatum requires further work done to assess the dental morphologies within Pan-Aves as a whole, especially due to its un-archosauriform lack of denticulation (see Ezcurra, 2016)."
 |
| Holotype right maxilla of Lusitanosaurus liassicus in lateral view (University of Lisbon coll.; destroyed). After Lapparent and Zbyszewski, 1957. |
"Lusitanosaurus liassicus Lapparent and Zbyszewski (1957) displays ankylothecodont teeth (ch388:1), low crowns (ch411:0), no recurvature (ch414:2), no apicobasal ridges (ch421:0), and no cingulum (ch423:0). The type specimen was lost in a fire (Mateus et al., 2022), so the only information available about the taxon is from the original description and plate of Lapparent and Zbyszewski (1957). The crowns are tricuspid, which is shared with members of Pterosauria (Dalla Vecchia, 2013), Squamata (Conrad, 2008), Tanystropheidae (Renesto & Dalla Vecchia, 2000; Spiekman et al., 2020), and basal Cynodontia (Hahn et al., 1984) but not Ornithischia. The lack of a cingulum also suggests Lusitanosaurus liassicus is not an ornithischian, as that was the only feature identified by Irmis et al. (2007) to distinguish early ornithischian teeth from similar but non-ornithischian phyllodont teeth. While it has been previously excluded from studies as a dubious basal thyreophoran (Antunes & Mateus, 2003; Norman et al., 2004a; Norman, 2020c), the distribution of anatomy instead suggests that Lusitanosaurus liassicus cannot be considered an ornithischian and should instead be considered Archosauromorpha indeterminate on the basis of ankylothecodont tooth implantation."
 |
| Teeth of Taveirosaurus costai- A paratype TV 16; B paratype TV 11; C paratype 14; D-E, J-K holotype TV 10; F paratype TV 8; G paratype TV 7; H paratype TV 13; I paratype TV 9. Lingual- A-D, F-J; Labial- E, K. After Antunes and Sigogneau-Russell, 1991. |
"Taveirosaurus costai Antunes and Sigogneau-Russell (1991) displays cheek tooth crowns without mesial and distal ‘kinks’ (ch412:0), low crown height (ch411:0), coarse denticles (ch416:2), and no cingulum (ch423:0), which would support a position just within or outside Ornithischia. However, the morphology of the crowns is very divergent from ornithischians, and the only features linking the taxon with Ornithischia are also found in non-ornithischian archosauromorphs (Irmis et al., 2007), mammaliforms (Luo et al., 2002) and even hybodontiformes (Stumpf et al., 2022). Taveirosaurus costai is considered to belong to Eutriconodonta here on the basis of linear molar cusps and a distinct, straight cingulum (Luo et al., 2002) as well as its Late Cretaceous age, until further work is done to establish possible shared dental morphology with Ornithischia."
 |
| Holotype tooth of Trimucrodon cuneatus (Lehrstuhl für Paläontologie coll.). Lingual- a; Labial- b. After Thulborn, 1973. |
"Trimucrodon cuneatus Thulborn (1973) displays sporadic wear facets (ch407:0), low crowns (ch411:0), kinked mesial and distal crown edges (ch412:1), triangular crown apices (ch413:0), no recurvature (ch414:2), mesiodistal expansion above the root (ch415:1), no apicobasal ridges (ch421:0), symmetric labiolingual expansion (ch422:0), and no cingulum (ch423:0). These features show significant conflict, especially the lack of a cingulum and asymmetrical labiolingual swelling which are exclusive synapomorphies of Parapredentata except for clades that are excluded for other characters (eg. lack of apicobasal ridges). Until further work is done to assess the dental morphologies within Pan-Aves as a whole, Trimucrodon cuneatus should be retained as a possibly valid taxon that cannot be classified beyond Ornithischia outside Parapredentata."
Several taxa like Alocodon, Phyllodon and Xiaosaurus are considered valid despite generally being ignored due to being fragmentary. They even include a large supplement illustrating their characters, and justifying which characters were not included. And you can tell they went up to the submission limit by commenting on recent taxa like Vectidromeus and Chakisaurus.
Things I wish Fonseca et al. would have done are list the possible positions of the taxa excluded WISELY a posteriori- "Daemonosaurus chauliodus; Diodorus scytobrachion; Gamatavus antiquus; Sacisaurus agudoensis; Mymoorapelta maysi; Nevadadromeus schmitti; Hualianceratops wucaiwaneisis; Beg tse; Ceratops montanus; the ‘Vegagete ornithopod’; Weewarrasaurus pobeni; Atlascopcosaurus loadsi; Galleonosaurus dorisae; Leaellynasaura amicagraphica; Qantassaurus intrepidus; Trinisaura santamartaensis; Isasicursor santacrucensis; Sektensaurus sanjuanboscoi; Callovosaurus leedsi; and Draconyx loureiroi", and scorings for the excluded taxa like Vectidromeus and "Hypsilophodon" wielandi. Also more constraint analyses for things like the position of Lesothosaurus and Albalophosaurus (here the earliest pachycephalosaur).
Now we need someone to do this for the 30+ incertae sedis sauropodomorph taxa everyone's ignored for over three decades since they were kicked out of Plateosaurus/Massospondylus/Euskelosaurus using Yates' matrix or something better.
References-
Lapparent and Zbyszewski, 1957. Les dinosauriens du Portugal. Mémoires des Services Géologiques du Portugal, nouvelle
série. 2, 1-63.
Thulborn, 1973. Teeth of ornithischian dinosaurs from the Upper
Jurassic of Portugal. Memórias dos Serviços Geológicos de
Portugal. 22, 89-134.
Antunes and Sigogneau-Russell, 1991. Nouvelles données sur les Dinosaures du Crétacé supérieur du Portugal. Comptes rendus de l'Académie des sciences. Série 2, Mécanique, Physique, Chimie, Sciences de l'univers, Sciences de la Terre. 313(1), 113-119.
Averianov, Martin and Bakirov, 2005. Pterosaur and dinosaur remains from the Middle Jurassic Balabansai Svita in the northern Fergana depression, Kyrgyzstan (central Asia). Palaeontology. 48(1), 135-155.
I am a bit puzzled. The paper mentions 5 supplementary files, which #5 the one including the data matrix, yet the supplementary material zip file includes only four files (none of which is the matrix). Am I the only one with this issue?
ReplyDeleteI asked Fonseca about this, the editors missed and forgot to upload the supplementary files 5 and 6
DeleteThere has been a slight issue with JSP missing two of our supp files, the matrix and scripts (Supps 5 and 6). Beyond that, the supplementary information is complete. It is being dealt with and hopefully will be done soon. Right now Andre is sending the matrix to those who request it until JSP remedies the situation.
DeleteThere will be more in the works on this analysis as we go, I'm happy to have been able to do the comments in the supplementary on the excluded taxa, but overall we are happy with being able to get this ~5 year long project published so others can comment and provide feedback as is being done here. More early dinosaurs, stem dinosauromorphs, and ankylopollexians are go tos for future additions. More commentary on pruned taxa or volatile taxa would have been a nice addition, though at the time of submission we were afraid of coming too close to the page limit for the paper (which is why the excluded commentary is its own supplement).
Thanks for the reply and congrats. As someone who has spent >20 years on an analogous project focusing on theropods, I appreciate much that someone has started something similar for ornithischians. Note that my latest version (Cau 2024: https://www.paleoitalia.it/wp-content/uploads/2024/05/01_Cau_2024_BSPI_631.pdf) placed silesaurids as basal ornithischians too.
DeleteBetter scan of Lapparent & Zbyszewski (1957) available here, FYI Mickey: https://docbase.lneg.pt/docs/Memorias/2.pdf
ReplyDeleteThanks! Still sad the specimen was destroyed in a fire. I'm sure there's a ton more than could have been done with it using micro-CT. Are there other known ankylothecodont tricuspid taxa, and from the Jurassic?
DeleteThe Early Cretaceous rhynchocephalian Ankylosphenodon has ankylothecodont teeth but it's unclear whether its teeth are bicuspid or tricuspid.
DeleteThis is a fantastic paper - well done to the authors. I'm especially intrigued by _Albalophosaurus_ coming up as a pachycephalosaur.
ReplyDeleteOne sour note is the inclusion of _Ceratops montanus_, a nomen dubium. According to the authors:
"Though the inclusion of Ceratops montanus could be argued against due to its incompleteness and dubious
identity (Mallon et al., 2016) it is an important specifier to include (Madzia et al., 2021), and current character
selection is sufficient to definitively resolve it together with the other ceratopsids even if it hampers internal
resolution of the clade."
Note the desultory disclaimer at the end of the sentence: "even if it hampers internal resolution of the clade." So _Ceratops montanus_ is included in the analysis, even though it's worse than useless - it's actually unhelpful .
Just get rid of _Ceratops montanus_ and the legalistic fiction that it's an OTU. It's only included simply because it has a bunch of family-level taxa named after it (like Ceratopsidae, Ceratopsoidea etc). We can rename the family-level taxa, and ignore _Ceratops montanus_ altogether - just like we ignore Deinodontidae in favor of Tyrannosauridae.
Not the authors' fault, BTW. _Ceratops montanus_ is a nuisance. Because the name is tied to higher-level taxa, there is an obligation to treat it as an OTU.
DeleteThere isn't a whole lot to support the placement of Albalophosaurus, which I suspect may move again if dental characters are further revised. Supplement 4(?) should include synapomorphies of that result.
DeleteWhile I agree that Ceratops is a poor taxon to include, at this point I don't think using alternative names for all the clades rooted on it would be beneficial, Ceratops-ia, -oidea, -idae. I would rather emphasize the importance of including all significant clade specifiers, even when dubious, so that names can be applied unambiguously, than stress about including a taxon that while problematic on a genus level, is definitively a member of the clade it is nominal genus for.
Ceratopsidae was so far from our focus that a small polytomy would not be of concern for character evolution and synapomorphy resolution. Including Ceratops as an OTU that is pruned a posteriori so we can objectively place it as a specifier while preventing it's impact on resolution may have been another way to go about it.
I agree with this approach about taxa like Ceratops. I and Madzia did something similar in 2018 including an OTU based on the (very limited) material of Troodon formosus holotype in our paravian analysis, in order to test if the taxon specifiers of Troodontidae/nae was "a troodontid" (that is, a member of the least inclusive clade containing Sinovenator, Sinornithoides and Saurornithoides).
DeleteI. Reid - I take your point re _Ceratops_. BTW, Ceratopia is safe whatever happens. It's only the coordinated family-level taxa that would be affected (Ceratopsidae, Ceratopsoidea).
Delete"Including Ceratops as an OTU that is pruned a posteriori so we can objectively place it as a specifier while preventing it's impact on resolution may have been another way to go about it."
ReplyDeleteYep that's what I thought too. It's a neat workaround. But personally, I'm not sure it's sensible policy to treat nomina dubia as OTUs, just for bookkeeping reasons. Again not a criticism, just a general observation.
On a separate note, I agree with Mickey - I found the supplementary information extremely informative. So thanks for the effort in examining these often obscure taxa. Among other things, it's good to see _Xiaosaurus_ as a valid taxon.
Finally, any thoughts on the suggestion that the _Leaellynasaura_ hypodigm includes some material?
Sorry I meant to write "Finally, any thoughts on the suggestion that the _Leaellynasaura_ hypodigm includes some theropod material?"
DeleteThanks. IIRC, there was a suggestion that the skull roof might come from a maniraptoran theropod - and the large orbits are consistent with this.
DeleteLarge orbits are an allometrically-controlled feature of every small-bodied vertebrate. Alone, they could not define a maniraptoran.
DeleteTim- that's vaguely familiar to me as well. If you can track down the reference, it would be useful for the Database to have Leallynasaura in the ex-theropod section.
DeleteAndrea - To rephrase, the large orbits of Leaellynasaura have been regarded as either being an adaptation to low-light polar conditions, or a juvenile character. However, IIRC the later interpretation was that the shape/size of the orbits was a maniraptoran character.
DeleteThis comment has been removed by the author.
ReplyDeleteMickey - I'll try and hunt down the source. Might have been a CAVEPS abstract from a while back (c. 10 years). I remember thinking 'WTF' when I read the suggestion/interpretation that some of the material assigned to Leaellynasaura might be non-ornithopodan. The skull roof and maybe the tail (which is extraordinarily long). Again, I'm going from memory.
ReplyDeleteThe specimen with hyper-elongate tail shows ornitischian-like caudal vertebrae (e.g., the shape and elongation of the zygapophyses is typical of small bipedal ornitischians and not much theropodan). In any case, the tail is associated to thoracic vertebrae bearing ossified epaxial tendons and to a femur with an asymmetrically pendant fourth trochanter, so the specimen it's surely ornithischian.
Delete