Reports of the death of lipless theropods are greatly exaggerated (part 1)

A while ago (over 4 years! Where does the time go?), I wrote a journal on why I didn’t think theropods had lips, at my deviantArt website (  That post generated a number of comments, both in favor of my hypothesis and against.

My attention had been directed to a new abstract was recently published by a commenter on my journal regarding this and I wanted to discuss it. Presented at the 4th Annual Meeting, 2016, Canadian Society of Vertebrate Palaeontology, in it the authors favor giving theropod lips. When I first read it, I thought, “They have a pretty good case, looks like I am wrong, and will need to start drawing theropods with lips again.” However, upon re-reading it, I am not so sure.

I am copying the text of the abstract since it is relatively short:

(Oral Presentation)
Dental anatomy and skull length to tooth size ratios support the hypothesis that theropod dinosaurs had lips
Robert.R. Reisz1, D. Larson2
1Department of Biology, University of Toronto Mississauga, 3359 Mississauga Road,
Mississauga, Ontario, L5L 1C6,
2Department of Ecology and Evolutionary Biology, University of Toronto, 100 Queen’s
Park, Toronto, Ontario, M5S 2C6, Canada,

Two competing hypotheses, whether the large teeth of theropod dinosaurs were
exposed or covered by peri-oral tissues or lips when the mouth was closed, are tested
using phylogenetic bracketing, dental anatomy and development, and regression analysesusing skull length to tooth size ratios.
Two different anatomical patterns can be discerned in reptiles. In crocodiles, the
closest extant, toothed relatives of theropod dinosaurs, about ¼ of the tooth is covered by gingiva, but there are no lips, and the crowns are exposed permanently. In contrast, in extant squamates, a more distant reptilian relative to dinosaurs, teeth are covered by lips when the mouth is closed, and there is extensive gingiva. Phylogenetic bracketing, in the absence of evidence from birds and from fossils would tend to support the hypothesis that the large teeth of theropod dinosaurs would be exposed when the mouth is closed, although there is little reason to suggest that the same was the case for the small teeth of other dinosaurs, like the cheek teeth of ornithischians.

Dental anatomy and development offer a different perspective. As the hardest
vertebrate tissue, enamel has a low water content (Zheng et al. 2013), and is hydrated and maintained by glandular secretions in the mouth. We propose that this requirement of hydration is not possible to maintain if the tooth is exposed permanently. We tested this by examining the exposed teeth of terrestrial mammals (tusks), modified teeth that evolved independently in several mammal clades.

Histological thin sections show that tusks in mammals do not have enamel. At the initial stages of development, some enamel may be formed, but soon after eruption the enamel is worn away, and may be replaced by cementum. This suggests that the large teeth of theropod dinosaurs, all known to have well preserved and maintained enamel, even with specialized ziphodonty (Brink et al. 2015), were not exposed permanently, but covered by reptilian lips similar to those found in squamates.

Similarly, ordinary least squares regression analyses of skull lengths to tooth sizes
in varanid lizards and theropod dinosaurs of various sizes indicate that the teeth of
theropod dinosaurs conform to the same pattern as varanid lizards. This provides strong added support to the hypothesis that theropod dinosaurs had lip-covered teeth, as teeth in theropod dinosaurs are no larger than would be expected in a similarly sized varanid lizard. This conclusion has wider implications, suggesting that this may be the primitive condition for all terrestrial vertebrates, allowing us to test whether the large, tusk-like structures of some basal ornithischians (Weishample and Witmer 1990), or the large canine-like teeth of terrestrial vertebrates (Brink et al. 2015) were exposed or not.

Finally, we propose that the lip-covered dental pattern is primitive for terrestrial
vertebrates, and that of crocodilians is a derived condition related to their secondary
aquatic or semiaquatic adaptations. It should be noted that terrestrial stem crocodilians (Clark et al. 2001) have a dental anatomy very similar to that of theropod dinosaurs, and likely had lips too.


Since it is an abstract only, and the actual data hasn’t been published, I can only comment on problems with I see with how they frame the hypothesis and test it.

In the third paragraph they posit that enamel needs to be adequately hydrated to be maintained, and that this would not be possible with permanently exposed lips. They appeal to histological evidence in mammalian tusks that show that the teeth are covered in cementum, not enamel, and thus dinosaurs must have had lips.

There are several problems with this.

(A) Mammalian tusks are very specialized structures and are not used like other teeth. Mammals that have tusks tend to use them in specialized tasks such as rooting through dirt, fighting (as in elephants) or digging (see warthog image below).20378063-mud-shovel-warthog-boar-digging-a-hole-to-sleep-in-photographed-in-namibia-stock-photo

Why would we expect animals engaged in such specialized behavoiors to be histologically similar to theropods? I am not aware of any theropod that is hypothesized to have used it jaws (and thus teeth) in similar behaviors. Of course if an animals is using it’s teeth in ways that are being constantly abraded by rough dirt and clashing, I would expect the enamel to be replaced by cementum. I don’t think this is a good behavioral analogy to theropods at all, and thus is a very weak argument in support of lips.

(B) Having teeth that are partially exposed some of the time (for extended periods) and then completely submerged in water on a repeated basis would actually be worse.  Crocodiles spend a significant time sunbathing in order to modulate their internal temperature as they are ectothermic:


If hydration of the enamel were a problem, one would actually expect crocodiles to retain their lips. Anyone who has had to repeatedly wash their hands over an extended period of time knows how repeated washing and drying chaps the skin. For animals that are semiaquatic, one would think if lips were needed in order to protect the enamel, lips would be retained in order to avoid extremes of being dehydrated to being completely saturated.

(C) In addition to the problems with analogous behavior between theropods and tusked mamals, there is another problem: Mammals are, by-and-large, diphyodont, whereas theropods (as well as crocodiles) are polyphyodont. What this means is that most mammals only have two sets of teeth, the deciduous set (aka primary or “baby teeth” in humans), followed by the permanent set. Theropods, on the other hand, continuously shed their teeth throughout their lives. This rate varies, but appears to have been equivalent to modern day crocodiles (, about 1 to 2 a year. In at least some sauropods, they replaced their teeth every one to two months (

Unlike squamates whose teeth are directly fused implanted to the jawbone (i.e., no tooth sockets; this can take the form of acrodonty or pleurodonty – the latter in which they are fused or ankylosed to the jaw bone; see Jaime’s comments below on my choice of terminology, as well as my response), theropods and crocodiles also share in having in their dental anatomy tooth sockets (alveoli). What this means is that there is already a replacement tooth right behind when a tooth is lost and can quickly be replaced.

So theropods have the best of both worlds, continuous (see comment below) repeated replacement of their teeth as in squamates, as well as having tooth sockets like mammals.

I would hypothesize that this would enable theropods to overcome any concerns of eroding enamel.

(D) Other semi-aquatic and aquatic toothed animals have not lost their lips, or extraoral seals. If you looked at aquatic and semi-aquatic squamates, such as sea snakes, marine iguanas and and some lizards (such as the Chinese Alligator Lizard, see below), none have lost their lips:




Sea snake:

Marine iguana:



And a nice short video of a water monitor (Varanus mertensi):


So I don’t understand what the evolutionary reasons would be for crocodiles to lose their lips, since all other extant toothed reptiles that are aquatic or semi-aquatic have not lost their lips.

Nor have aquatic and semi aquatic mammals:

Elephant seal (female):



(Quick side note on whales: while whales no longer have muscular movable lips, they have retained the oral seal, and when the mouth is closed, the teeth are invisible).

It’s therefore not clear why crocodiles would lose their lips,  and stating that is because of their aquatic or predatory habits is falsified by the above examples. It is also interesting to note that some of the most derived members of the Trionychia (soft-shelled turtles) have lost (or nearly lost) their beaks (external premaxillary, maxillary and dentary rhampotheca, see Jaime’s comments and my response below; they have retained an internal palatine and tomian rhampotheca) and have actually independently evolved fleshy lip-like structures, and are aquatic:


Note that this really is an independent evolution, as the most basal members of crown Trionychia, such as Carettochelys , still have beaks, whereas derived members, such as Carettochelys Pelodiscus, have lost or nearly lost their beaks and replaced with lips (see cladogram from Wikipedia below). Turtles have interesting implications with regards to phylogenetic bracketing and the biological restrictions on lips, teeth and beaks. In the second part of this article, I want to discuss these evolutionary developmental reasons that, in my view, also suggest no lips – and not just for theropods, but likely for all dinosaurs.

Trionychia cladogram


  1. I am confused by a few parts of your article

    “So theropods have the best of both worlds, continuous replacement of their teeth as in squamates, as well as having tooth sockets like mammals.”

    But before you write: ” squamate teeth are directly fused to the jawbone”

    So which is it? Permanently fused to the skull or continuous replacement?

    Going back to my first citation, you also write that dinosaurs share with crocodilians their tooth sockets, and that they get best of both worlds. What are you hinting at? If they have tooth sockets and tooth replacement like crocodylians, they share both traits with crocodilians. there is no “best” they get from the world of squamates, according to you, their teeth are fused solid to the skull.

    You cite Carettochelys as both a primitive and derived member of the soft shelled turtles and link to the same wikipedia page twice. Did you mean to link to Trionychia as the advanced member?

    You’ve provided some interesting points that I agree with. aquatic lifestyle did not mean the loss of lips in most aquatic vertebrates, and the turtles independently evolving lips is especially peculiar. I guess the question is why is it that crocodile teeth are covered in cementum instead of enamel and why are their teeth exposed.

    Crocodiles don’t use their teeth like tusks, they use them to grasp prey. It’s also interesting that the abstract talks about stem crocodylians which had lips. were their teeth covered in enamel or in cementum? i wonder if this has been investigated. Another question that comes to mind is whether it is possible that crocodiles lost lips and enamel in their transition to aquatic lifestyle even though other secondarily aquatic animals did not go through this change.

    If it is indeed true that all animals with enamel have teeth sealed inside their mouth, dinosaurs with enamel-covered teeth (which often have a much more complex shape than just the blunt cone of a crocodile) would I think benefit from having the teeth sealed to protect them, even if they had the ability to replace them. Sharks have replacable teeth covered in enamel as well and most have what (to me) would pass as a sealed mouth (with the jaws closed) as well.

    I am interested to see the second part of this article though.


    1. My wording there is unclear, you’re right, and I apologize.

      Theropods (like crocodiles) have the “best of both worlds” in that they borrow two strategies, one from mammals, and one from squamates. Mammals have tooth sockets, but only replace their teeth in the vast majority of taxa twice in their lifetime, and have more durable teeth as a result. Squamates do not have tooth sockets but replace their teeth repeatedly throughout their lives (not continuously as in crocodiles, I misspoke, and I will correct the main text), but have the disadvantage of losing their teeth easily.

      Another typo on the Carettochelys. The second link was meant to be for Pelodiscus. I must not have correctly save the proofread version of my post. Again, apologies. Thanks for reading so carefully!

      Crocodiles do indeed have cementum, as do all toothed animals (as far as I am aware) (see: (1) and (2) But crocodiles also have enamel and dentin (see link (1)), they have not lost it, but it is reduced when compared to squamates. The interesting thing with the tusked mammals is that they had no enamel at all, and it was replaced instead completely be cementum. I view this as a result of their unique behavior in most tusked animals that use their teeth in either direct combat, gathering food (non-food processing related, such as digging for tubers in suids ) or as tools in shaping their environment (digging dens as in warthogs).

      They infer stem crocodylians had lips based on their similarity to theropods, not on any actual evidence other than provided in their abstract for arguing theropods had lips. Like I said, I think their arguments for theropods are very weak at best in my opinion.

      Your example of sharks reinforces my point that the reasons for crocodiles losing their lips are incorrect. First, in sharks enamel exists with lips in spite of living in an aquatic environment, which falsifies the main reason that crocodiles are said to have lost their lips – supposedly because living predominately in water provides adequate hydration for the tooth enamel. Second, two corollary arguments that I have seen made elsewhere is that either lips are needed to protect the teeth in predatory combat (which is why they are retained) or lost because they interfere with prey capture. The fact that we have large macropredators such as varanids and sharks that retain their lips in both aquatic and terrestrial environments falsifies both of those corollary arguments as well. So there does not appear to be an evolutionary reason why crocodiles would lose their lips.


  2. I’m chiming in and breaking my otherwise silence on this topic because, well, it is this topic.

    There are several factual inaccuracies involved here, which should be resolved in either an edited, updated form, or in part 2.

    1. Squamates in general do not posses teeth fused to their jaw bones. Only some squamates do so. Lepidosaurians in general all possess either pleurodont or acrodont implantation, and fusion typically occurs in the latter (but not always). In those that do, the teeth *are* fused wholly to the jaw bone, there is a form of replacement that requires the formation of cementum, it’s hardening, then eventually absorption and degradation as the tooth is shed and the replacement comes in. This is sometimes coincidental with development of alveolar bone in pleurodont or subthecodont implanation, where the interdental ridges separating teeth are made of this partial connective tissue. This is no different than thecodont implantation in snakes, mosasaurs, or crocodilians.

    2. Tronychid turtles possess beaks. Every single turtle since Proganochelys has a margin beak on the maxilla, premaxilla, and dentary, and trionychids like Apalone and Trionyx are no different. The exception is the presence of a fleshy tissue that derives from the face and jaw and cover the beaks. They are unique in possessing these tissues, amongst all turtles, but they did not lose beaks.

    You can see these beaks in the following images: and and

    While I also possess reservations on the argument presented, the subject at hand has not been subjected to substantive testing; merely lots and lots of hand waving and anecdotes. Anecdotes do not an argument make. Testing the water retention versus the structure of the tooth, and its function, seems to be the point of the Reisz/Larson study. The anecdotal observation, of the basal presence of fleshy tissues around the oral margin in terrestrial vertebrates, is also a true one: of those animals that lack them, they have beaks, or are modern crocodilians; and of those, their facial structure appears to indicate those tissue types. So we make assumptions and test them, and here we are.


    1. I’m sorry Jaime that I have roused you out of your paleo-sabbatical! I will try not to do it again, lest the paleo-gods strike me with fossilization 😉

      As for (1), I agree my wording was imprecise and could cause an incorrect viewpoint. However, squamates do not have tooth sockets (dental alveoli), unlike crocodiles, dinosaurs, and mammals which do. This was the point I was trying to make when I said ‘fused’, and I didn’t feel getting into the nitty gritty of acrodont and pleurodont (where the teeth are truly fused or ankylosed) dentition would greatly clarify the matter. Not sure what general term we be better (maybe ‘implanted’?), but thank you for your technical clarification, I will update the main text.

      On number (2), I agree that even derived Trionychian taxa such as Amyda for example retain a palatine rhampotheca and tomian rhampotheca (such as here: that is housed inside the mouth and is used in food processing. However, in these derived taxa, any external rhampotheca on the premaxilla, maxilla and dentary has indeed been reduced or lost (even if externally visible, as you show). I would not consider this to be a ‘beak’ in the colloquial sense of the term I was intending. But I take your point and will update the text.

      As for your last comments, I am not sure who you are accusing of hand-waving, but I will presume me, even though you know what they say about making assumptions. 🙂

      That said, I agree with your interpretation of what the focus of the study is: dental hydration and how it correlates with dental morphology. However, I am not saying their focus is wrong (I think it is a neat way to test these hypotheses). What I am saying is that several of their assumptions are faulty (i.e., crocs losing lips because they are semi-aquatic, even though other semi-aquatic and aquatic extant reptiles have not lost their lips), and the way they test them is not much better. I don’t consider this hand-waving, which is trying to obscure the issues.

      I am trying to clarify the issues, and these are the issues that need to be tested if one is to make any headway in a resolution. One of these is explaining why, if crocodiles were ancestrally lipped, they lost their lips. I think the ‘anecdotes’, as you call them, of showing living taxa that have similar habitats and/or habits to crocodiles that retain their lips is not anecdotal. It’s cold, hard facts. Another is that, to my eyes, the proposed osteological correlates of lips in living taxa do not seem to applicable to theropods; a third is that the behavioral analogies seemed flawed as well.

      I have covered the issue of foramina in my deviantArt journal post, which was in response to another published abstract, and think I have shown convincingly (at least in an informal setting), that the foramina morphologies in living squamates with lips, avian taxa with beaks and in crocodilian taxa without either, are not much different than the diversity that can bee seen in theropods, and are therefore not that helpful in testing the lips hypothesis.

      Now I am taking on the issue of dental structure and function under similar terms. I think you and I can both agree that using terrestrial mammals that retain lips, but have perpetually exposed, specialized teeth in the form of tusks, are not good analogues for testing tooth structure and function in theropods. Similarly, why are varanids singled out as theropod analogues? Many other lizards that have much more dissimilar dental morphologies and jaw mechanics to those taxa, and yet, also have lips. How is this really testing anything to find a specific group of lizards that have dental morphology close, in some respects, to theropods, and then correlating this with lips? Especially when many other lizards that have lips do not share these morphologies? To me this is what confuses and obscures the issues when certain arguments are focused on very specialized cases that can be easily confounded by other factors, and without regards to if it makes any sense to the bigger picture. That is what I would call hand-waving, but your mileage may vary.


      1. On the issue of trionychid rhamphotheca, the point was they weren’t lost; merely reduced in some form, if that, and overlain by extensive soft-tissue structures which are unique to them amongst turtles. And there’s osteological correlates to these tissues… Your use of the image doesn’t destract from this point, either: it shows a marginal tomia on upper and lower jaw, and these cap around the dorsal margins of the jaw bones, they are present on the outside of the jaws, and are visible in vivo while the mouth is closed, as I showed. Palatal rhamphothecae in turtles don’t really occur: the structures you are looking at, that are palatal, are cornified epithelium, and some of it does look pretty funky hard, such as some seaturtles, but they aren’t the cornified sheets that comprise typical rhamphotheca, even really soft ones in some ducks.

        (The hand waving comment was a general phrase, it wasn’t pointed at you. Of that, I don’t think you’re guilty)

        There’s probably a few explanations for why modern crocodiles lost the extensive fleshy facial tissues of their ancestors, but merely by admitting this, one is making a case that they ancestrally had them. This is the tricky part where some argue they never did. Modern crocodilian facial anatomy differs pretty extensively from that of even lizards, turtles, and birds, sufficiently to indicate that whatever they were doing, it was unique; and ancestral crocodilians, non-eucrocodilians mostly, have facial structure similar to lizards and snakes than they do turtles and birds. But that aside…

        All the prospective whys and whatnots and wherefores amount to nothing without testing. Testing the issue of foraminal concentration, innervation and vascularization relative to tissue type, the relation of tissue type to margin bone structure on the macro and micro and histological levels, and so forth; dental hydration and the form of dentition, whether enamel is viable without a dermal covering, and how hydration of that enamel is maintained, etc.; and of course, whatever is the perfect analogue? You ask why people single out varanids, but they don’t; or they shouldn’t. Varanids are, however, a large group of large reptile with extensive facial integument and variable dental morphologies and diets, affording a group of variable states with similar oral integument. If those aren’t valid comparatively, why would any other lizard be? Even the argument of the EPB is meant to be taken as a point from which to start, not the point at which to end, as some seem to suggest. The point is not to find the answer, but eliminate the unlikely or improbable outcomes, and the EPB is a simple explanation, but it raises more questions than the alternative does, and no one suggests a plan to test this route. So here we are, with a firm testable prediction; and work from Morhardt’s angle presents a potential second line of prediction, with work from Hieronymus’s thesis being a third. Reisz & Larson have presented the first experimental analysis on this subject, even though it is not the first *published* conjecture focusing on the subject (that being Ford’s).

        We are probably getting ahead of ourselves on this subject at the moment, trying to see what this study says, and assuming it says more than it does. There should be a dose of reality, that there are things people are using this to say that it doesn’t, and can’t, and that is what should be done. When it’s published.


Leave a Reply

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

You are commenting using your 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