First, I should note this manuscript is a perfect example of what Makovicky and Dyke (2001) critiqued in their paper "Naive falsification and the origin of birds", where a single set of features is claimed to be difficult to reverse and thus credibly undermine the confluence of numerous characters that support the Birds are Dinosaurs hypothesis. I think any modern paper should address Makovicky and Dyke's arguments against using the naive falsificationist approach.
Second, you should address both Limusaurus and Psittacosaurus. Psittacosaurus is commonly preserved with gastroliths (e.g. Wings, 2007: fig. 2E), and so is dissimilar to the probable hadrosaurid examples you cite where indigestible fragments make their way into the intestine. Limusaurus also has a collection of gastroliths (Xu et al., 2009) and is universally agreed to be a ceratosaurian, which even Feduccia and Martin agree(d) are theropods. These examples show that the bird-like condition can evolve from the basal dinosaurian condition, and specifically from the theropod condition. So if it can happen in Limusaurus, why is it unlikely to have happened for maniraptoriforms?
References- Makovicky and Dyke, 2001. Naive falsification and the origin of birds. In Gauthier and Gall (eds.). New Perspectives on the Origin and Early Evolution of Birds: Proceedings of the International Symposium in Honor of John H. Ostrom. Peabody Museum of Natural History. 501-509.
Xu, Clark, Mo, Choiniere, Forster, Erickson, Hone, Sullivan, Eberth, Nesbitt, Zhao, Hernandez, Jia, Han and Guo, 2009. A Jurassic ceratosaur from China helps clarify avian digital homologies. Nature. 459, 940-944.
First, thank you for your attention and comments, Mickey! I think that information on Limusaurus and Psittacosaurus will be added to the new manuscript version. Nevertheless, please note that my discussion on the digestive physiology of dinosaurs is based on several lines of evidence (feces, gastric pellets, stomach contents, intestine contents, gastroliths).
The sole presence/absence of gastroliths in the guts wouldn’t be so conclusive. In theory, any animal can swallow some stone pieces independently on its stomach anatomy. Herbivores often swallow astonishing amounts of sediment and their stomachs sometimes act as sedimentation traps like in hippopotamuses (Wings et al., 2008; Bajdek et al., 2014). Such mineral clasts may accumulate in the food chain passing into the guts of carnivores (Bajdek et al., 2017).
We should rather ask what happens with the stones when they are swallowed. Gastrolith finds are rare in articulated skeletons of dinosaurs in contrast to those of birds. In compsognathids (Sinocalliopteryx) occasionally swallowed stones were simply excreted rather than retained in stomach (passed into the intestine), similarly as bone fragments (Xing et al., 2012). The region of body cavity which contains supposed gastroliths in Limusaurus is poorly preserved (is it an artifact caused by strong preparation?) and it is some hard to really interpret. Notably, other theropods such as coelophysids, allosaurids, and tyrannosaurids excreted indigestible items in feces and the same refers to hadrosaurids (see references in the text).
References
Bajdek, P., Owocki, K., Niedźwiedzki, G., 2014. Putative dicynodont coprolites from the Upper Triassic of Poland. Palaeogeogr. Palaeoclimatol. Palaeoecol. 411, 1–17. doi: 10.1016/j.palaeo.2014.06.013
Bajdek, P., Owocki, K., Sennikov, A.G., Golubev, V.K., Niedźwiedzki, G., 2017. Residues from the Upper Permian carnivore coprolites from Vyazniki in Russia – key questions in reconstruction of feeding habits. Palaeogeogr. Palaeoclimatol. Palaeoecol. 482, 70–82. doi: 10.1016/j.palaeo.2017.05.033
Wings, O., Hatt, J.M., Schwarm, A., Clauss, M., 2008. Gastroliths in a pygmy hippopotamus (Hexaprotodon liberiensis Morton 1844) (Mammalia, Hippopotamidae). Senckenberg. Biol. 88 (2), 345–348.
Xing, L., Bell, P.R., Persons, W.S.IV, Ji, S., Miyashita, T., Burns, M.E., Ji, Q., Currie, P.J., 2012. Abdominal contents from two large Early Cretaceous compsognathids (Dinosauria: Theropoda) demonstrate feeding on confuciusornithids and dromaeosaurids. PloS ONE 7, e44012. doi: 10.1371/journal.pone.0044012
Thank you for your quick reply.
Regarding Limusaurus, gastroliths are present in several specimens (Wang et al., 2017), and notably in older individuals in this taxon that loses teeth as it ages. They state- "Gastroliths are present only in L. inextricabilis specimens of stages III-VI. IVPP V20098 (Stage III) possesses sand-like gastroliths that are similar in quantity and size to those found in some Mesozoic birds [S67, S68], whereas in more mature individuals (e.g. IVPP V15923) the quantity and size of gastroliths is comparable to ornithomimosaurians [S69] and oviraptorosaurians [S70] at comparable body sizes to subadult and adult Limusaurus (stages IV-VI)." Thus the multiple specimens of Limusaurus and Psittacosaurus weren't merely incidental ingestion of rocks, and so provide as much evidence for having a bird-like gastric mill as Caudipteryx or Sinornithomimus do. They prove a "reappearance and strong development of a specific stomach functionality that either had been already lost in the ... dinosaurs or was never present in their lineage." And if it can happen to Limusaurus and Psittacosaurus, it could happen to birds, right?
With reference to the comments below on dromaeosaurid gut contents, what about Microraptor specimen QM V1002 (Xing et al., 2013)? That preserves a mass of fish bones and scales and the authors wrote "the position of the mass suggests it was located in the posterior portion of the digestive tract, most probably within or near the jejunum, and the alignment of the elongated elements likely reflects their prior passage through the anterior portion of the tract." They go on to compare this favorably to Sinocalliopteryx and Scipionyx.
If a ceratosaur and a ceratopsian doing things the bird way and a 'bird' doing things the theropod way don't disprove your hypothesis that digestive physiology is a fundamental difference between the groups and unlikely to reverse or converge, then what kind of example would disprove it?
References- Xing, Persons, Bell, Xu, Zhang, Miyashita, Wang and Currie, 2013. Piscivory in the feathered dinosaur Microraptor. Evolution. 67(8), 2441-2445.
Wang, Stiegler, Amiot, Wang, Du, Clark and Xu, 2017. Extreme ontogenetic changes in a ceratosaurian theropod. Current Biology. 27(1), 144-148.
Thanks again for comments… I am not going to dismiss “contrary evidence” of any kind. In the new manuscript version, I think to discuss examples given by you and also mention the case of dromaeosaurids proposed by Andrea Cau and more precisely define a “bird”.
But I think that I have already answered to your doubts. First of all, as I said the sole presence of gastroliths wouldn’t be too conclusive, and my hypotheses are based on multiple lines of evidence. Gastroliths are found in stomachs of a variety of animals such as hippopotamuses, elasmosaurids, and the Permian Hovasaurus, and what? There is no gastric mill in hippos despite gravel may reach 1/3 of the total wet stomach content! Gastroliths are generally absent in pterosaurian stomachs, but as you can notice there is known an "exception to the rule"--Pterodaustro guinazui (Codorniú et al., 2013).
What really matters is that (1) gastroliths are extremely rare in the guts of dinosaurs (and also pterosaurs, by the way) when compared to birds and crocodiles, and (2) fossils clearly show that in dinosaurs poorly-digestible items such bones and stones were expelled in feces, which is moreover supported by the fact that (3) gastric pellets are absent in dinosaurian stomachs in contrast to stomachs of fossil birds.
You mentioned the "Naive falsification (...) where a single set of features is claimed to (...) credibly undermine the confluence of numerous characters. Isn’t it precisely what you are doing right now, Mickey?
Regarding Xing et al. (2013), you should note the very minute micrometric size of the fish bone elements in the intestines of Microraptor. The case of Microraptor has nothing to do with Sinocalliopteryx, as seen in both Microraptor specimens. Possible gastric pellets composed of big bones are present in Microraptor stomach, whereas its intestine solely contains remains of very minute dimensions which are sub-millimetric. Thus, if you ask me to interpret the digestive physiology of Microraptor, it was comparable to that encountered among birds, crocodiles, and some early archosaurs rather than to typical dinosaurian physiology.
References
Codorniú, L., Chiappe, L.M. and Cid, F.D. 2013. First occurrence of stomach stones in pterosaurs. Journal of Vertebrate Paleontology 33(3): 647–654.
Thanks again for the quick reply. I do think that we're talking past each other a bit (much like Martin and I did in our email conversation on bird origins), so I think it would be helpful if you would answer my direct questions.
If you don't think gastroliths are conclusive evidence for a bird style digestive system for Limusaurus and Psittacosaurus, then they wouldn't be conclusive for Caudipteryx or Sinornithomimus either. It seems to me that your criteria have no way of demonstrating a bird-style system in a fossil unless you could somehow connect a regurgitated pellet to a particular taxon. Which is so rare that it hasn't happened in fossils to my knowledge. In other words, you can exclude a bird-style system by the presence of large materials in the intestine or coprolites, but you can't ever exclude a dinosaur-style system. So you would by your own criteria never be able to know if a dinosaur developed the bird-like system. What kind of evidence would convince you that Limusaurus or Psittacosaurus had a bird-style system?
I'd also say you shouldn't just cite Hippopotamus as an excuse for e.g. Psittacosaurus, since hippos have broad mouths and graze (at night, with poor eyesight), while Psittacosaurus has a narrow beak and much larger stones than expected for its size (Sereno et al., 2010). Those authors concluded "The unusual size and mass of the gastrolith mill in P. gobiensis and other psittacosaurs thus add support to the interpretation that nuts or seeds with a hard casing and high fibre content may have been a predominant component of their diet." That correlates with the cranial evidence they present, but what evidence supports your alternative that Psittacosaurus were ingesting large stones accidentally?
More generally I wonder why you think the rarity of gastroliths in dinosaurs matters. Your argument is that the bird-croc system is too complicated/fundamental to evolve twice, in the standard topology once at the base of Archosauria and again in maniraptoriforms/'birds' after it was lost in Dinosauria. So even if Limusaurus and Psittacosaurus are the only dinosaurs that re-evolved the gastric mill (which would still make gastroliths rare in dinosaurs), they would still prove the gastric mill CAN re-evolve in something. IF Psittacosaurus were shown to possess a gastric mill, would you think 'birds' re-evolving the gastric mill from theropods is plausible?
You misunderstand Makovicky and Dyke's concept of naive falsification. It's when a certain feature is claimed to be biologically implausible to evolve a certain way, and so every other feature supporting the relationship must be convergence. Among their examples- birds embryologically have manual digits II-III-IV and theropods lacked a functional digit IV, so birds cannot have evolved from theropods and all their other similarities must be convergence. Or- Scipionyx and Sinosauropteryx show theropods had a diaphagm and birds have an airsac system that could never evolve from a diaphragm, so that all their other similarities must be due to convergence. In your case, it's that the bird-croc gastric mill wouldn't have evolved twice, so birds couldn't have convergently evolved it from dinosaurs that lacked it, thus all the other similarities between birds and dinosaurs must be convergent. Since I already have the majority of character evidence on my side, which is why cladistics favors Birds Are Dinosaurs, I cannot be a naive falsificationist. Instead, cladists are "relative falsificationists" in Makovicky and Dyke's wording, which basically means we analyze all the evidence equally and don't think any 'key' character or group of characters can invalidate a phylogenetic hypothesis. In other words, even if every dinosaur lacked a gastric mill and had large fragments in its intestine and every 'bird' had the opposite, that would just be a few to several characters that are easily overcome by all the osteological characters. In modern phylogenetics, you need a big dataset incorporating or numerically overwealming prior datasets to overcome a hypothesis. Suggesting several characters don't match up just isn't objective enough for people to care about since the 1980s.
Reference- Sereno, Zhao and Lin, 2010. A new psittacosaur from Inner Mongolia and the parrot-like structure and function of the psittacosaur skull. Proceedings of the Royal Society B. 277(1679), 199-209.
I’ve repeated several times that just the presence of gastroliths cannot be considered too conclusive and also that my interpretations are based on several independent lines of evidence.
Please, note that Psittacosaurus was a ceratopsian and as you should know these animals had sophisticated masticatory apparatus. Gastric mill is a feature most typical of herbivorous birds, which as you know are toothless… I can see no need for re-evolving the gizzard in Psittacosaurus, can you? It would be more expected in dinosaurs of a poorly developed dentition, like ankylosaurs. But we know well-preserved gut content of Minmi and there are no stomach stones…
Second, if gastric mill re-evolved in ceratopsians, why gastroliths are not so common in other ceratopsian species? Animals ingest sediment for a variety of reasons, including e.g. mineral supplement and we know virtually nothing about the diet and lifestyle of ceratopsians.
I will discuss the case of Psittacosaurus and Yinlong in the new manuscript version. Yet, I honesty have no doubt that it will pretty obvious for any reader who bases his/her reasoning on PARSIMONY, that there is no reason to really think that dinosaurs had a gastric mill.
Molnar, R.E. and Clifford, H.T. 2001. An ankylosaurian cololite from the Lower Cretaceous of Queensland, Australia., in: Carpenter, K. [ed.], The Armored Dinosaurs. Indiana University Press, Bloomington, pp. 399–412.
Ah, so your solution is that not even Psittacosaurus had a gastric mill. Which would then imply that no evidence likely to be preserved in the fossil record could convince you any dinosaur or other animal had the bird condition. Thus your own criteria ensure that even if a dinosaur evolved the bird-style system, you would never know. Your hypothesis could never be falsified, which is unscientific.
As for why Psittacosaurus would evolve a gastric mill, ceratopsians didn't chew food, they sheared it. Again, see Sereno et al. (2010) who viewed psittacosaur jaws as functionally similar to psittaciform jaws among birds. If a ceratopsian was born with a mutation that kept more accidentally eaten stones in its digestive system instead of passing them through, it could eat harder nuts and seeds than its competitors, and pass on those genes. Natural selection takes place taking advantage of this new food source and you end up with populations of animals with gizzards and the instinct to eat large stones. It's a just-so-story, but seems as plausible as any other evolutionary scenario for a feature.
Why no gastric mill in ankylosaurs? Because animals don't evolve features that would be better for them. That's getting evolution backwards, Lamarkian instead of Darwinian. Natural selection just takes advantage of features that happen to be expressed due to mutations caused by the genetic part of reproduction. But if the right mutation(s) never happen, or they do and the individual(s) don't pass them on for any number of reasons (die before breeding, recombination eliminates mutation in offspring, etc.), then natural selection has nothing to work with. So apparently no ankylosaur zygote happened to get the mutation to start developing a gastric mill, or if it did that population/individual failed to succeed for any number of reasons (or hasn't been discovered/recognized yet, like ceratosaurs pre-Limusaurus).
It's a similar answer to why gastric mills aren't present in most other ceratopsians. If Yinlong is a psittacosaur (Han et al., 2017), then neoceratopsians happened not to have or usefully exploit that mutation. If it's not a psittacosaur, then both genera evolved it convergently, or it evolved once in ceratopsians and was lost in neoceratopsians because it was no longer advantageous to survival. Maybe the ancestral neoceratopsian population moved to a geographic area or habitat that no longer had many hard-shelled nuts and seeds. Who knows? There are so many possibilities and we should never think evolution could have only occured in a certain way. That's another part of Makovicky and Dyke's naive falsificationist paper- e.g. Feduccia and Martin (1998) claiming Velociraptor's furcula is probably not a furcula because they thought a furcula would have evolved in a flight context (and now we have furculae in coelophysids, abelisaurs, allosaurs...). You should never dismiss a hypothesis because you can't think of why it would have happened that way.
Finally, on the topic of biological reasons for structures, it seems pretty clear to me that gastric mills evolve in theropods that benefit from them by being more herbivorous, or at least not macrocarnivorous. They show up in Limusaurus, Caudipteryx, Shenzhousaurus, Sinornithomimus, Jeholornis and Sapeornis, but not coelophysids, Morrison large theropods, tyrannosaurids or dromaeosaurids. That's pretty good correlation, even with the odd exception like Yanornis. My impression is that gastric morphology was seemingly easy to change in theropods once diet favored it. I think most of our colleagues would agree with that, whereas I don't think they're going to take your ideas seriously once they see you claim Psittacosaurus didn't have a gastric mill. Nothing personal, but anyone familiar with the data is going to see that as special pleading and dismiss what you have to say.
References- Feduccia and Martin, 1998. Theropod-bird link reconsidered. Nature. 391, 754.
Han, Forster, Xu and Clark, 2017. 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. DOI: 10.1080/14772019.2017.1369185
Thanks for your comments on the ankylosaur zygote and natural selection, Mickey! Of course my hypotheses are falsifiable--for example, you can describe dinosaur (not maniraptoran) gastric pellets like those produced by birds.
