Gallery and print store

Thursday, 27 April 2017

Walking with ichthyosaurs: the amphibious ichthyosaur hypothesis

Benjamin Waterhouse Hawkin's (1858?) sketch of amphibious marine reptiles, including a large shambling ichthyosaur. Image borrowed from Frank T. Zumbach's Mysterious World.
One of the most charming aspects of mid-19th century palaeoart are those amphibious marine reptiles: depictions of ichthyosaurs and plesiosaurs that hauled themselves onto rocks or beaches to rest, or lunge with open jaws at passers by (above). To modern eyes these images look naive and quaint, a clear reminder of how far our understanding of fossil animals has progressed in the last two centuries.

Of course, art has a habit of imitating life and, a good 150 years after amphibious marine reptiles became unfashionable in palaeoartworks, Ryosuke Motani and colleagues (2014) published a new marine reptile suggested to be capable of locomotion on land as well as in water: the ichthyosauriform Cartorhynchus lenticarpus. This Chinese, Early Triassic species is anatomically remarkable in several respects. Although reminiscent of early ichthyosaurs in overall shape, it has a considerably reduced snout, seems to lack teeth, is just 20 cm from snout to vent despite indications of osteological maturity, and bears enormously long forelimbs. Though unique when first discovered, another, much larger Cartorhynchus-like species has since been found in the same deposits, Sclerocormus parviceps. Together, these animals form a clade at the base of Ichthyosauriformes known as Nasorostra, the 'nose beaks', referring to a defining feature where their nasal bones reach the jaw tip (Jiang et al. 2016).

Holotype specimen of Cartorhynchus lenticarpus. Note the enormous forelimbs with their expansive unossified wrists, indicated by the distal phalanges being well posteriorly displaced from the upper arm bones. From Motani et al. (2014).
The amphibious habits of Cartorhynchus are primarily based on its unusually large forelimbs and small body size, it being reasoned that Cartorhynchus could drag or propel itself over exposed sediments like a mudskipper, turtle or pinniped. I find this idea fascinating: an ichthyosauriform that was at home outside of water? Cartorhynchus certainly deviates from ichthyosaur anatomy and evolutionary trends enough to inspire inquiry about its weird bauplan - if it was not amphibious, it might be doing something else equally unexpected. The amphibious Cartorhynchus hypothesis has received surprisingly little detailed attention online, save for coverage of a 2014 press release and this excellent primer article at Tetrapod Zoology, so there's scope for a closer look at this idea. What is the evidence for amphibious habits in Cartorhynchus, and how does this concept fit models of early ichthyosaur evolution?

The functional basis for an amphibious lifestyle in Cartorhynchus

Motani et al. (2014) present a fairly detailed argument in favour of amphibious habits in Cartorhynchus. The chief lines of evidence are those expansive forelimbs, but it's not just their size that matters: their enormous, unossified carpal regions are also significant. Several early ichthyosauriforms have poorly ossified carpal bones but the unossfied region in Cartorhynchus flippers is proportionally bigger by some margin. This would allow these ordinarily-rigid marine reptile flippers an unusual degree of flexibility and optimise them for terrestrial locomotion. Flipper-based terrestrial motion is surprisingly tricky because its users tend to be suboptimally designed for movement out of water and they almost always have to overcome drag forces acting on the body as well as shove themselves around. Moreover, substrates associated with coasts and waterways tend to be unstable, yielding under pressure and being challenging for even proficient terrestrial animals. These factors mean flippers can easily dig into substrate or slip across it rather than propel their owners about, and it's easy to see why beaching is fatal for so many specialised aquatic species.

Studies (using robot turtles!) suggest that rigid flippers are generally poor at terrestrial locomotion and may even be incapable of moving animals over some surfaces (Mazouchova et al. 2013). A bendy flipper, in contrast, works well, allowing the forelimb to flex before the substrate moves, spreading the weight of the animal over the distal limb and allowing the proximal flipper region to elevate and support the body (Mazouchova et al. 2013; Motani et al. 2014). The unusually expanded flexion zone in Cartorhynchus forelimbs would be well suited to this purpose, and certainly much better at this task than those of other ichthyosaurs. We might note, as an aside, that the lack of flexion zones in other marine reptile flippers, such as those of plesiosaurs, might be good reason to doubt their ability to crawl over land.

Did I mention the robot turtles? There are robot turtles. Supplementary video data from Mazouchova et al. (2013).

The downside of having lots of cartilage in a long flipper is that they are weaker against bending than a more ossified one, so their utility as a walking limb lessen as the forces involved in moving the body increase. It's here where the small size of Cartorhynchus comes into play. Small size equates to low body masses and smaller forces associated with lifting the body, less structural demand on the flipper, and reduced drag effects from the sliding belly. As is so often the case in evolution, small body size might be an enabler for evolutionary experimentation in Cartorhynchus, allowing it to perform feats that its bigger relatives just couldn't even if they were also equipped with giant, bendy fins.

The tail of Cartorhynchus is incompletely known but it's anatomical and phylogenetic proximity to the completely-known Sclerocormus suggests that its tail was long, flexible, and lacked any sort of fin or fluke (Jiang et al. 2016). A relatively simple tail lessens the risk of it dredging sediment or catching on debris during terrestrial locomotion and its flexibility might have permitted its use as a prop or even propulsive organ: fish such as the Pacific leaping blenny show how a long, bendy tail can be used to powerful effects in semi-terrestrial locomotion (Heish 2010, also below). Combinations of fin and axial motion in land-crawling fish can be surprisingly effective over a range of substrates (Standen et al. 2016) and we might assume similar options were available to Cartorhynchus.

Leaping blennies, robot turtles... is this the best blog post ever? From Wikipedia, source: Hsieh (2010).

The torso of Cartorhynchus is also of interest for this hypothesis. In contrast to some other Triassic ichthyosaurs, Cartorhynchus has a broad, stout torso rather than a long, laterally-compressed one (Carrol and Dong 1991). Though a wider torso would impart more drag during terrestrial crawling, it would aid stability when crawling over land. Moreover, torso drag can be lessened by shortening the body overall, giving new significance to the low Cartorhynchus pre-sacral vertebral count of 31 vertebrae, instead of a more typical ichthyosaurian count of 40-80 (Motani et al. 2014). Short, narrow hindlimbs, rather than the broad pelvic flippers of some other early ichthyosaurs, might have further aided drag reduction.

Cartorhynchus in context

It seems there's a prima facie argument for considering Cartorhynchus as equipped with some amphibious features. However, we should not get carried away - a suite of evidence for an aquatic lifestyle suggests it wasn't it a specialist denizen of shallow, partly-exposed habitats, but more of an animal able to exploit two realms. It has pachyostotic bones, true flippers rather than webbed walking limbs, and is adapted for suction-feeding: a mechanism where the combination of a small mouth and a large oral cavity creates a pressure differential during feeding, literally sucking small prey into the mouth if it's opened quickly (Motani et al. 2014). This foraging strategy cannot work outside of water so is strong support for Cartorhynchus foraging in fully aquatic settings.

Cartorhynchus also stems from the Nanlinghu Formation, a mudrock and limestone marine deposit rich in fossils of aquatic reptiles and marine invertebrates: ammonoids, bivalves and conodonts. We might take these data as signs that Cartorhynchus was quite happy in water and maybe spent most of its time there, visiting coastlines and beaches on occassion, rather than living there permanently. We should also regard it as a marine animal, not an inhabitant of rivers or swamps (though it would be extremely cool if one turned up in such deposits!).

Holotype of Hupehsuchus nanchangensis, a marine reptile seemingly more closely related to the ancestor of ichthyosaurs than Cartorhynchus. These guys surely deserve their own blog post and painting at some point. From Carroll and Dong (1991).
The relationships of Cartorhynchus to other marine reptiles is also interesting in light of the amphibious hypothesis. You could be forgiven for interpreting Cartorhynchus as some sort of bridge between ichthyosaurs and terrestrial reptiles, but, no, the nasorostran clade seems to nest above the root of the ichthyosaur line between 'true' ichthyosaurs and the fully marine, ichthyosaur-like hupehsuchians (Motani et al. 2014; Jiang et al. 2016). The ichthyosaur + hupehsuchian clade, Ichthyosauromorpha, may be further allied to another group of marine reptiles, the amphibious thalattosaurs (Motani et al. 2014 - Darren Naish has an excellent overview of this topic here). This surrounds Cartorhynchus with lineages that had taken to water in a significant way and we should conclude that any amphibious adaptations of Cartorhynchus do not represent an ichthyosaurian invasion of the sea, but ichthyosaurs returning to land.

Some might consider this surprising evolutionary scenario evidence against the amphibious hypothesis - why would a lineage of marine reptiles start retracing their adaptive steps to become landworthy, when the rest of the group is pressing ahead with more specialised aquatic lifestyles? In response, perhaps we should ask if a potentially amphibious marine reptile is really that surprising. A huge number of vertebrates have transferred between terrestrial and aquatic lifestyles in the last 400 million years, sometimes contrasting with wider adaptive trends taking place in closely related species. Well-understood evolutionary 'transitions' also show that large-scale adaptive phases are often complex with all manner of evolutionary experimentation and dead-end offshoots. We know that bridging aquatic and terrestrial realms can be advantageous to aquatic species - refuge from predators or rough seas, access to food off-limits to other marine species, access to safe habitats for rest or reproduction, etc. - and there's no reason to think ichthyosaurs were incapable of capitalising on these advantages, or immune to their selective draws. With all this in mind, the concept of a marine reptile exploiting semi-exposed habitats isn't really that radical. Maybe the key question here isn't 'why would a marine reptile go rouge and turn landward?' but is 'why aren't we seeing more of this sort of thing?'.

What about Sclerocormus?

A question currently unaddressed in technical literature is whether the other currently known nasorostran, Sclerocormus, might have also bear amphibious hallmarks. It has virtually all the same features that we likened to amphibious adaptations above, the only distinctions being marginally enhanced ossification of the forelimb (though it still retains a comparatively enormous unossified carpal region) and greater size overall (body length of 160 cm, representing an animal about 3.3 times larger than Cartorhynchus). In lieu of a detailed, quantified assessment it's difficult to say whether Sclerocormus was too heavy to pull itself along on land, but we can note that it is not especially big compared to the truly massive aquatic animals we have scampering over beaches today - leatherback turtles, giant pinnipeds, the odd manatee (Motani et al. 2014) and so on. Some of these animals weigh several tonnes and, if they can haul themselves out of water, maybe Sclerocormus could too.

Holotype specimen of the larger nasorostran species, Sclerocormus parviceps. From Jiang et al. (2016).
I find this question particularly interesting given how similar Sclerocormus and Cartorhynchus are in virtually all aspects (above). Is nasorostra a clade of potentially amphibious ichthyosaurs, or are we actually looking at growth stages of one oddball species? Their proportions are near identical, and they are only separated by fine details of anatomy (Jiang et al. 2016). Many proposed differences might be attributable to intraspecific variation, too. For instance, the significance of their slightly different vertebral counts is questioned through populations of living snakes, limbless lizards and fish with variable numbers of axial elements (Tibblin et al. 2016). Individually variable vertebral counts seem common in species with large numbers of axial elements, and this might have been true for ichthyosaurs. Ontogeny and scaling effects could explain other differences, including overall size, greater ossification of the postcranial skeleton, and subtle arrangements of skull bones. It can't be overlooked that these near identical species, unique in morphology in the grand scheme of ichthyosaur evolution, also happen to occur in the same member of the same formation, separated by only 14 m of strata (Jiang et al. 2016). For the time being, the identification of 'adult' skull fusion and textures in Cartorhynchus suggests they aren't the same species, but the marine reptile trait of retaining poorly fused skeletons into adulthood makes identifying adult forms especially tricky, especially with so few specimens to look at (Motani et al. 2014). It also seems worryingly difficult to tease fossil adults from juveniles without histological assessments, even with large sample sizes and good growth series (e.g. Prondvai et al. 2009). Perhaps we're waiting on histological examinations and more specimens to make a call on this.

So, walking with ichthyosaurs?

And finally, a painting: Cartorhynchus goes for a drag around a Triassic lagoon.
Putting all the strands of the amphibious Cartorhynchus hypothesis together, I don't see reason for excessive suspicion about the idea of beach hauling nasorostrans. At the core of the pro-amphibious argument is that Cartorhynchus (and perhaps, by extension, Sclerocormus) has weird anatomy that requires an explanation - it's just too different from other ichthyosauromorphs to pretend it wasn't doing something unusual, maybe even unexpected. Amphibious behaviours are an explanation that seem to chime well with provisional form-function investigations and seem a sensible hypothesis at this time. That said, we should be appropriately cautious in our interpretations of these animals: our understanding of nasorostrans is in its infancy and alternative, currently-unexplored functional hypotheses could explain their anatomy as well, or better, than the amphibious concept in future. Fingers crossed that these animals receive more dedicated functional investgiations in years to come.

Or maybe more robot turtles. Either is good with me.

Enjoyed this blog post? Want free art, books and other cool stuff? Consider supporting this work with a monthly micropayment

This blog is sponsored through Patreon, the site where you can donate a few dollars to your favourite online content creators to help them make a living. It takes time to produce the paintings and write the essays you see here and, if you enjoy this content, please consider becoming a patron for as little as $1 a month. In return, you get regular updates on projects I'm working on, including research, papers, books and paintings. Currently, this includes two palaeoart-heavy books, and my Patreon pages are filling with advance previews of their content. Plus, you get free stuff - prints, high quality images for printing, books, competitions - as way of thanking you for your support. As always, thanks to everyone who already sponsors my work!


  • Carroll, R. L., & Zhi-Ming, D. (1991). Hupehsuchus, an enigmatic aquatic reptile from the Triassic of China, and the problem of establishing relationships. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 331(1260), 131-153.
  • Hsieh, S. T. T. (2010). A locomotor innovation enables water-land transition in a marine fish. PloS one, 5(6), e11197.
  • Jiang, D. Y., Motani, R., Huang, J. D., Tintori, A., Hu, Y. C., Rieppel, O., ... & Zhang, R. (2016). A large aberrant stem ichthyosauriform indicating early rise and demise of ichthyosauromorphs in the wake of the end-Permian extinction. Scientific reports, 6, 26372.
  • Mazouchova, N., Umbanhowar, P. B., & Goldman, D. I. (2013). Flipper-driven terrestrial locomotion of a sea turtle-inspired robot. Bioinspiration & biomimetics, 8(2), 026007.
  • Motani, R., Jiang, D. Y., Chen, G. B., Tintori, A., Rieppel, O., Ji, C., & Huang, J. D. (2015). A basal ichthyosauriform with a short snout from the Lower Triassic of China. Nature, 517(7535), 485-488.
  • Prondvai, E., Stein, K., Ősi, A., & Sander, M. P. (2012). Life history of Rhamphorhynchus inferred from bone histology and the diversity of pterosaurian growth strategies. PLoS One, 7(2), e31392.
  • Standen, E. M., Du, T. Y., Laroche, P., & Larsson, H. C. (2016). Locomotor flexibility of Polypterus senegalus across various aquatic and terrestrial substrates. Zoology, 119(5), 447-454.
  • Tibblin, P., Berggren, H., Nordahl, O., Larsson, P., & Forsman, A. (2016). Causes and consequences of intra-specific variation in vertebral number. Scientific reports, 6, 26372.


  1. Wonderful post! I've been interested in the possible life appearance of these snub-nosed critters since they were first described. I also wonder if Sclerocormus is simply an adult Cartorhynchus. It would be unusual to see two virtually identical taxa in the same place at the same time. On the other hand, hupehsuchians are all from pretty much the same place at the same time and they're surprisingly diverse.

    Speaking of, I would love to see your take on Hupehsuchus or Nanchangosaurus. In the meantime, I'll toot my own horn here and link to my own two-part blog post on those bizarre critters, one of which features a life restoration by Ethan Kocak!

    I'd like to see that polydactylous hupehsuchian described at some point.

  2. I agree on the distinct liklihood of Sclerocormus just being an adult Cartorhynchus. So often this happens where a supposed new radiation is all found in the same formation, so conveniently- 'sapeornithids', pengornithids, the Archaeopteryx and Microraptor complexes.

    However, your statement that "This surrounds Cartorhynchus with lineages that had taken to water in a significant way and we should conclude that any amphibious adaptations of Cartorhynchus do not represent an ichthyosaurian invasion of the sea, but ichthyosaurs returning to land" is incorrect. If we assume the topology you and Naish describe is correct, then it takes just as many steps (2) to get a secondarily amphibious Nasorostra as it does to have huphehsuchians and ichthyosaurs convergently evolve to be fully aquatic from an amphibious ancestor.

    Great drawing, btw. Reminds me of a ratfish.

  3. Excellent article, really expands the context of how we think of these weird guys. Towards the end of the article I couldn't help but think of the epaulette shark which has a unique niche of hauling itself over exposed corals at the ebbing tide to hunt in tidal pools. In doing so it can exploit such pools without competition from other fish that have to retreat at low tide. It also can expand it's buccal cavity to suck in small prey (mainly invertebrates), is of a similar size, and is tolerant of low oxygen.