Herbivores All the Way Down

Back in the 1980s and 1990s, a revolution in the way dinosaurs were conceived to have evolved occurred, in which a basal stock of croc-ish animals (this was during the “thecodont origin of … everything (?)” phase of archosaur relationship hypotheses) developed into the dinosaurs we know today, basically divided into three groups:

Theropod dinosaurs, including potentially birds; sauropods and their kin, the “prosauropods”; and ornithischians. While the latter group were considered herbivorous almost exclusively, with some omnivorous basal forms (like “fabrosaurs”), and the first group was almost exclusively considered carnivorous with some omnivorous forms, the herbivory/carnivory aspects never crossed over. In Sauropodomorpha, the group containing sauropods and “prosauropods,” basal forms were considered mostly herbivores, with some carnivorous trends (some of it based on confused fossils of giant croc-like rauisuchids or postosuchids). But this meant that theropods were nested within a group that trended, at its base, towards herbivory.

All dinosaurs, then, would have had a little plant-eater in their near-ancestry.

Many years later, as fossils were found and data refined, phylogenetic reconstructive methods improved, and interpretive processes wrung through the logic wringer, we’ve come to a better understanding of what the first dinosaurs were probably like, and how they diversified.

herbivory to carnivoryBy the time theropods were well and truly settled, herbivory is rare. Until recently, all herbivorous taxa were confined to birds and the remarkable therizinosauroids, giants with scythes and sickles for claws. With the advent of more refined hypotheses, new theories for recognizing herbivory came about, focusing on how one determines herbivory in other amniotes when one cannot examine their diet directly, or its remains (e.g., stomach, colon, or fecal remains, termed gastrolites, cololites, and coprolites, respectively). This technique, for instance, allowed Lindsay Zanno and Peter Makovicky (2011) to formulate a series of tests of congruence amongst various morphological traits to place several odd and ambiguous taxa into their place on the spectrum of carnivory to herbivory.

Surprisingly, the odd eternal outliers that are oviraptorosaurs end up as probably herbivores, along with ornithomimosaurs. The combination of a small head and a long neck are strong indicators for effective upper level foraging in moderate to small animals, so it also stands to reason that the snouts of these animals should be relatively narrow or pointed, and for the most part this is true. Being a browser seems a common adaptation for herbivorous theropods, occurring in diverse clades, but all of them maniraptorans: ornithomimosaurs, therizinosauroids, and oviraptorosaurs. And in all of them, the basal forms seem to have been larger headed and more toothsome, so the development of herbivorous features is convergent. In 2008, Xu Xing and colleagues described a new basal abelisauroid, Limusaurus inextricabilis, that was remarkably toothless, small-headed and long-necked, seeming similar to the elaphrosaurids that have been known for a while but always based on head-less material. Its toothlessness, amongst many toothy relatives, provoked much thought on the plasticity of keeping being a meat-eater amongst early theropods, otherwise a “sure thing” to bet on.

But perhaps it was an interesting outlier. Or not…

Chilesaurus diegosuarezi skeleton sm

Meet Chilesaurus diegosuarezi , a strange convergence of … well … convergence. It puts the lie to adage that you can’t have two non-coelurosaurian neotheropodan dinosaurs that were herbivorous … that I just made up. Chilesaurus diegosuarezi was just recently described by Fernando Novas and colleagues, and comes from eastern Aysén, Chile (from whence it gets its name), in a volcanic deposit called the Toqui Formation, which is dated to the end of the Jurassic, in the Tithonian. Thus, Chilesaurus diagosuarezi predates all other herbivorous theropod dinosaurs by a significant time period, in which almost all are known at least from the middle of the Cretaceous or later, save the Yixian Formation taxa, like Incisivosaurus gauthieri.

Initially, one might be tempted to call this animal a “prosauropod,” and indeed it resembles one strongly. Many features of the jaws, arms, and vertebrae recall the basal sauropodomorphans in which dietary variation must have played a role in at least some speciation amongst them. But there similarities end. The vertebrae bear distinct anterior and posterior pneumatic pleurocoels, foramina enlarged as the cervical diverticulae penetrated the bones; these pneumatic openings show further evidence of division as they were excavated, a feature unknown in any non-theropod to date. As in some ornithischians, herrerasaurids, and a variety of wholly maniraptoran theropod dinosaurs, the pubis is retroverted and much more slender than the ischium, and the like many ornithischians, the ischium has a large “boot,” but the pubis does not. This combination at once recalls only ornithischians, and likely had the specimen been composed only of this element it might have seemed as such, except that the pubes bear a distinct apron that extends more than halfway up the bone, and the two elements are firmly conjoined along this apron; as in the paired pubes, the paired ischia have a similar apron, and both diverge at their distal ends from the apron, forming a notch in the end, features found only in theropods.

Holotype skeleton of Chilesaurus diegosuarezi, SNGM-1935, as recovered in the field prior to preparation. Elements modified from Novas et al. (2015) (rotation of labels and scale bar).

Holotype skeleton of Chilesaurus diegosuarezi, SNGM-1935, as recovered in the field prior to preparation. Elements modified from Novas et al. (2015) (rotation of labels and scale bar).

And while many specimens were found, these include nearly complete skeletons (such as the holotype specimen, shown above), in articulation and life pose (as is common in tuffaceous siliciclastic deposits, in which volcanic ash accumulates rapidly around the skeleton or mixes with sand in water to form a fluffy, but ultimately fatal, preservative). There is no question these disparate parts are elements of the same animal. They point to many interesting things. Among these remains comes confirmation, the jaws of an animal with apparent theropod attributes and an herbivore’s teeth: flatted, not conical, with denticulate margins that disappear with wear, arranged en echelon, partially overlapping the adjacent teeth, and angled towards the front of the jaw (mesially). Teeth like these are otherwise known in Incisivosaurus gauthieri in shape, but only by arrangement in ornithischians. The sheer number of features suggests an entire clade of similar animals, of which only Chilesaurus diegosuarezi is known.

Perhaps most surprising is how many herbivorous adaptations have been found in this new animal. Zanno and Makovicky identified 21 distinct characters commonly found amongst herbivorous dinosaurs or inferred to be present in them, and their distribution holds up strongly in tests of living dinosaurs (birds) as well as mammals, non-mammalian synapsids, several basal amniotes, lizards, and even some varieties of crurotarsan croc-relatives (Reisz & Sues, 2000). They referred to these as extrinsic evidence of herbivory, or correlated herbivorous traits (CHTs), characters that grouped with others that occur in putative herbivores.

1, dentary turned downwards at the tip; 2, ventral margin of the dentary concave in aspect; 3, dentary with complete tooth loss; 4, dentary with only caudal tooth loss; 5, dentary symphysis projects rostrally; 6, anteriormost dentary teeth conical; 7, premaxillary teeth conical or almost conical; 8, rostral teeth in upper and lower jaws conical or subconical; 9, teeth symmetrical (not exhibiting reflexing, or strong curvature); 10, loss of ziphodonty (“blade-shape” lost); 11, dentary with anterior tooth loss [significant correlation]; 12, premaxillary teeth larger in relation to maxillary teeth; 13, premaxilla with complete tooth loss [significant correlation]; 14, premaxillary teeth lack denticulations [significant correlation]; 15, mandibles don’t show patterns of tooth replacement, most teeth largely the same size [significant correlation]; 16, “cheek” tooth shape is “lanceolate,” or oval with a pointed apex; 17, dentition with varying shapes, or sizes, of crowns, along the tooth row (heterodonty) [significant correlation]; 18,  premaxillary teeth lean forward (procumbent); 19, dentition “packed,” or arranged en echelon (see above) [significant correlation]; 20, mandibular joint depressed below the tooth row; 21, cervical vertebral series increased beyond 10 vertebrae.

Zanno & Makovicky further considered additional characters and termed them putative herbivorous traits (PHTs), amongst which the strongly correlated characters above (CHTs) were drawn. These are: 22, dentary symphysis U-shaped in aspect; 23, pubic shaft concave anteriorly, or “bowed caudally,” in aspect; 24, ischia elongated relative to ilium or pubis; 25, pubes retroverted. These characters are considered less strongly correlated as they tend not to show up in theropods. For instance, the pelvic features are common in ornithischians and birds, but not non-alvarezsaurian nonavian theropods, and medial curvature of the dentary symphysis (U-shaped) occurs much more rarely but is otherwise known in ornithischians, and birds only weakly as a product of fusion of the mandibular symphysis.

Of these 25 characters Chilesaurus diegosuarezi exhibits seven: 1, 2, 9, 10, 18, 19, 24, and 25. Character 19 correlates strongly with herbivory, and taken with the pelvic characters gives us several strong lines of data from which to infer herbivory in a theropod dinosaur. Character 17 involves one of the problems of the term “heterodonty” I bring up here, which is conflation of shape-based and size-based variations in teeth. Were we only to consider one aspect of this character, size, we could then add an eighth PHT to the list. Overall, this is more than even ornithomimosaurs and oviraptorosaurs exhibit, and their case for herbivory, or omnivory at least, is robust.

Chilesaurus diegosuarezi also appeals to us in other ways. Like Limusaurus inextricabilis, the small head and slender neck are paired with short arms and a robust shoulder girdle. They lend credence to an idea that the much, much larger Cretaceous Deltadromeus agilis (Sereno et al., 1996), recently considered a relative of noasaurids and of elaphrosaurids, as a basal abelisauroid, may very well be a giant, edentulous herbivore, the largest theropod herbivore from Africa. Chilesaurus, like Limusaurus, has also been reconstructed with a small hand, reduced to two functional digits. Like “Ingeniayanshini and other “ingeniine” oviraptorids, the first manual digit (or thumb) is very large and almost as long as the second digit. A third metacarpal bears a single phalanx preserved, but it is possible this digit was more complete; if so, it would be much shorter than the second, and the manus would have a stub or a thing functional third digit, resembling the “ingeniine” and likely basal alvarezsaurian conditions.

And while preservation of the jaws of Chilesaurus diegosuarezi are incomplete, they reveal some curious things. Using data from Zanno and Makovicky, we can’t readily predict that the anteriormost teeth were much larger than the others, as the dental alveoli are intact and the same diameter as those following, and the same is true between premaxilla and maxilla; posterior maxillary teeth become reduced in size, but not apparently in shape, and this is typical in theropods, but the expected extremes of shape, or of size, occurs in the very first few tooth positions, and then diminishes distinctively. That size diminishes is apparent (Novas et al., 2015, extended data figure 2, d), but shape is obscured by wear.

The shape of the jaw, the shapes of the teeth, are interesting in that they involve circumspectly the inference specifically of one odd Jurassic dinosaur, Eshanosaurus deguchiianus (Xu, Zhao & Clark, 2001). Based on an isolated mandible with several teeth, the jaw was inferred first to be a coelurosaurian theropod by Zhao & Xu (1998) prior to being identified (and named) as a therizinosauroid by Xu et al. (2001). This identification has not come without question (e.g., Rauhut, 2003). The “prosauropod”-like features of the jaw were noted by Matthew Lamanna (pers. comm. in Kirkland & Wolfe, 2001) shortly after the material was described, and since then, the material has failed to be subjected to intensive study, including phylogenetic analysis. I have been trying to get work done on this subject, but stalled for relatively personal reasons. I intimated a few years back I’d get around to this, but the desire to keep things under-wraps while I refined my arguments/accessed data kept it down. Now that time has come.

Like Chilesaurus diegosuarezi, Eshanosaurus deguchiianus bears numerous correlated herbivorous traits, and not unsurprisingly several of these are found in basal sauropodomorphans, which merely tells us that many features noted by the authors are convergent due to diet. Some, however, may not be; and Eshanosaurus deguchiianus exhibits features of the mandible that relate to potentially other aspects of herbivory not addressed by Zanno & Makovicky (2011), including the presence of a lateral thickening of the dentary, or “ridge” and the shelf that the upper half of the lateral dentary surface becomes when this ridge is present. The most important characteristic that has remained problematic in this case (as can be observed in discussion of Barrett, 2009) is the shape of the teeth. The oddity of lanceolate, slightly denticulate, leaf-shaped teeth in a non-coelurosaurian theropod dinosaur removes the burden this “problem” presents, and as it occurs in the Jurassic indicates that it is possible Eshanosaurus deguchiianus is part of that Chilesaurus clade, rather than as originally conceived a coelurosaurian, or later a therizinosauroid, dinosaur. This has the additional effect of removing the putative time gap and morphological similarities with derived, but not basal, therizinosauroids, and as a non-coelurosaurian, the temporal problem of an Early Jurassic “therizinosaur.” Indeed, being “merely” a tetanuran theropod in the Lower Jurassic is far less of a stretch. Additional comments would require examination of the material to resolve certain problems that the description brings up, including relative preservation of the mandible.

Theropod herbivores, excluding birds, form a large diversity of shapes and sizes. From the bottom and working clockwise from the outside: Caudipteryx zoui, Khaan mckennai, Deinocheirus mirificus, Erlikosaurus andrewsi, Struthiomimus altus, Limusaurus inextricabilis, Chilesaurus diegosuarezi.

Theropod herbivores, excluding birds, form a large diversity of shapes and sizes. From the bottom and working clockwise from the outside: Caudipteryx zoui, Khaan mckennai, Deinocheirus mirificus, Erlikosaurus andrewsi, Struthiomimus altus, Limusaurus inextricabilis, Chilesaurus diegosuarezi.

Herbivory in theropods seems to be the reversal of many, many states, and the process looks involuted. The gaps between similar small cursorial theropods in maniraptoriforms and their herbivorous or omnivorous relatives suggests that there are yet more small, cursorial members to find amongst therizinosauroids, ornithomimosaurs, oviraptorosaurs; amongst elaphrosaurids and possibly noasaurids within abelisauroids; and other clades otherwise seemingly “normal.” Ornitholestes hermanni bears some of these PHTs in its teeth and jaws, while alvarezsaurians bear others, especially postcranially. They all seem to be relatively small, gaining size only over time and in diverse ecosystems.

Once it seemed that herbivores amongst theropods were confined to the Maniraptoriformes. Later, it was at least Coelurosauria, then with Limusaurus it was found to be Neotheropoda. A gap between the neotheropodan and the earliest maniraptoriform having been filled with Chilesaurus diegosuarezi, it seems that herbivores really seem to go all the way down. It is quite possible more basal taxa, more sauropodomorphan like, as in Eoraptor, will come out as theropod herbivores.

I am very excited for the future.

Barrett, P. M. 2009. The affinities of the enigmatic dinosaur Eshanosaurus deguchiianus from the Early Jurassic of Yunnan Province, People’s Republic of China. Palaeontology 52 (4): 681-688.
Kirkland, J. I. & Wolfe, D. G. 2001. First definitive therizinosauroid dinosaur (Dinosauria: Theropod) from North America. Journal of Vertebrate Paleontology 21 (3): 410-414.
Novas, F. E., Salgado, L., Suárez, M., Agnolín, F. L., Ezcurra, M. D., Chimento, N. R., de la Cruz, R., Isasi, M. P., Vargas, A. O. & Rubilar-Rogers, D. 2015. An enigmatic plant-eating theropods from the Late Jurassic period of Chile. Nature (online before print: DOI: 10.1038/nature14307)
Rauhut, O. W. M. 2003. The interrelationships and evolution of basal theropod dinosaurs. Special Papers in Palaeontology 69: 1-213.
Reisz, R. R. & Sues, H.-D. 2000. Herbivory in Late Paleozoic and Triassic terrestrial vertebrates. pp.9-41 in Sues (ed.) Evolution of Herbivory in Terrestrial Vertebrates. Cambridge University Press (Cambridge).
Sereno, P. C., Dutheil, D. B., Iarochene, M., Larsson, H. C. E., Lyon, G. H., Magwene, P. M., Sidor, C. A., Varricchio, D. J. & Wilson, J. A. 1996. Predatory dinosaurs from the Sahara and Late Cretaceous faunal differentiation. Science 272: 986-991.
Xu X., Clark, J. M., Mo J., Choiniere, J., Forster, C. A., Erickson, G. M., Hone, D. W. E., Sullivan, C., Eberth, D. A., Nesbitt, S., Zhao Q., Hernandez, R., Jia C.-k., Han F.-l. & and Guo Y. 2009. A Jurassic ceratosaur from China helps clarify avian digital homologies. Nature 459: 940–944.
Xu X., Zhao X.-j. & Clark, L. M. 2001. A new therizinosaur from the Lower Jurassic Lower Lufeng Formation of Yunnan, China. Journal of Vertebrate Paleontology 21 (3): 477-483.
Zanno, L. E. & Makovicky, P. J. 2011. Herbivorous ecomorphology and specialization patterns in theropod dinosaur evolution. Proceedings of the National Academy of Sciences, Philadelphia 108 (1): 232-237.
Zhao X.-j. & Xu X. 1998. The oldest coelurosaurian. Nature 394: 234-235.

This entry was posted in Art, Biological Comparison, Paleoecology, Paleontology, Reconstruction, Science Reporting and tagged , , , , , , , . Bookmark the permalink.

17 Responses to Herbivores All the Way Down

  1. ijreid says:

    Nice skeletal. The cladogram is unique, on the basis that sauropodomorpha nests paraphyletic with relation to ornithischian, with the later in the former.

  2. ijreid says:

    My first thoughts when I was this was that it was a prosauropod, and then ornithischian, then theropod.

  3. Jaime, would you mind discussing the armored look you’ve given Limusaurus inextricabilis in this restoration. Were there any fossil evidences of dermal scutes or plates? Thanks.

    • I’d say this is a product of it being an abelisaur, and that the only firm evidence for dermis outside of the Megalosauroidea-Bird clade is for squamous skin and large scute-like scutellae in the skin. So far, filamentous integument in theropod dinosaurs in unknown this far down, though I did restore Masiakasaurus with it a while back.

      • Alex says:

        Megalosauroidea? I’m sorry but wasn’t the classification of Sciurumimus as a Megalosaur dismissed because of the fact that it was based on extremely weak phylogenetic evidence? The fact that it took them THREE tries to get Sciurumimus into Megalosauroidea is rather fishy to me. But of course, it’s hard to tell with baby dinosaurs.

      • I like that Masiakasaurus! ya knever know, filimentous integ might be old as the hills! I certainly do like all the scutellae in abelisaurs and I think you made a very unique and plausible look for the Limusaurus.

  4. Alessio says:

    Really interesting, the “plant-eaters panoply” at the end is a real masterpiece, if i may add.

  5. Andrea Cau says:

    Just a note: Limusaurus is older than Chilesaurus (Oxfordian vs Tithonian).

  6. Jaime, adding Daemonosaurus and Jeholosaurus to your tree, along with several other taxa listed at reptileevolution.com/reptile-tree.htm will add new insight to your tree. I found those two to be sisters to Chilesaurus. Novas will do the same next time around.

    • Dave, adding those two to the tree would be useless, as the cladogram is nothing more than an ad hoc creation, not one pooled by personal analysis. It was created from the combined consensus of the last 15 years of cladistic analysis of theropods.

  7. Pingback: PaleoNews #12 | Thetetanuraeguy

  8. Zach Miller says:

    Great article, Jaime. And wonderful pictures, too. You’ve given Chilesaurus a longer snout than in the description, which I prefer. The oddly abbreviated bulldog face in the paper’s skeletal reconstruction came off as wrong to me, given that the bones in the middle of the skull are missing.

  9. Cristopher.S Bonilla says:

    I love your work, I have a question: What other animal , dinosaur , mammal, anuro , plant, tree or pterosaur may have lived the CHILESAURIO ?

  10. tato says:

    Nobody is talking about Eshanosaurus. It would be great to finally find a place for those isolated taxons

  11. James A. Stearns says:

    I’ve been suggesting Deltadromeus (which is almost certainly the same as Bahariasaurus) was an herbivore for a while now; it would seem to solve the problem of “how did Spinosaurus, Bahariasaurus, and Caravharodontosaurus coexist?” fairly well, with each giant theropod specializing in a different area (piscivore, herbivore/omnivore, traditional carnivore).

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s