Inaechelys pernambucensis by Julio Lacerda. Once upon a time, the sea was full of side-necked turtles, sticking to the sea floor while the ancestors of modern sea turtles swam above them.

Sea turtles are best known as the sea reptiles that got away, the only survivors of the Mesozoic assemblages of ichthyosaurs, plesiosaurs and all that. Technically this is not true (dyrosaurids, gavialids, several sea snake taxa and neochoristoderes would like to have a word), but it also trivialises one small detail about modern sea turtles: they are all members of a single lineage, Chelonioidea. As turtles have dominated aquatic niches since the Triassic, surely one would expect more than one group to have become adapted to life in the sea.

And indeed, that is the case.

Several extinct lineages of turtles have conquered the seas. Depending on who you ask, the number may be as little as 4 times and as high as 8, still a bit short depending on your expectations for an aquatic group but nonetheless a much larger menagerie than the modern survivors would imply. Some of these groups were just as cosmopolitan and diverse as modern sea turtles, while others were apparently short-lived explorations of the sea. Regardless, these animals attest to several guilds of marine shelled reptiles comparable to the modern freshwater diversity of niches, and a few even rank as some of the largest turtles of all time.

Their absence in modern times is quite sad, and sadder still is that some did live up until very recently.

The first mariner

Odontochelys semitestacea by Brian Choo, hitching a ride from giant crinoids floating near the sea surface.

One of the very earliest turtles was already specialised to a life at sea. Odontochelys semitestacea is best known for its decidely basal, un-turtle-like absence of a carapace and presence of a solitary plastron and teeth (its in the name, after all), but a little known fact is that this animal occurs in deep sea marine deposits otherwise dominated by pelagic or abyssal taxa. , its unlikely that this critter was dragged from freshwater or even coastoal sites.

So one of the very first turtles lived far out in the open ocean.

As Odontochelys semitestacea lacks speciations for a pelagic lifestyle like flippers, a lifestyle proposed is that this animal hitchhiked rides on the giant crinoids that occupied the main filter feeding niche in Triassic seas, sticking close to the sea surface. Given that turtles are among the few amniotes , however, its not inconceivable that this was a benthic critter unlike any other vertebrate known, spending most of its life in the depths. Either way, the half-shelled toothed turtle spent its days gazing into the abyss, only leaving the darkness of the sea to lay eggs on land.

Sea turtles like this appear to have been extremely rare, as Triassic oceans were instead home to another clade of armoured marine reptiles, the . It seems it took the extinction of these animals to allow turtles to expand in the Jurassic, though given the presence of an already specialised turtle living far away from shore it is just as likely that these critters co-existed just fine with their shelled bretheren.


Leyvachelys cipadi by Jorge Blanco, enjoying a crab snack as it watches an ammonite die. Note the massive hands.

The seas of the Jurassic, Cretaceous and Palaeocene were home to an unique lineage of turtles, the . This assemblage is composed by Thalassochelydia, a clade composed by several Late Jurassic turtles of Europe and possibly Argentina, and Sandownidae, a clade ranging from the Cretaceous to Palaeocene with a more or less cosmopolitan distribution. The exact classification of these turtles has been controversial with several members having popped up across the turtle phylogenetic tree in previous studies, but the consensus is that they form a monophyletic group just outside the clade leading to crown-group turtles. They however so strongly converged with chelonioidean sea turtles on a variety of traits that the possibility that they’re actually cryptodires and closely related/ancestral to them can’t be fully ruled out yet.

Solnhofia parsonsi by “Ghedoghedo”. Again, notice the massive hands.

Like modern sea turtles, angolachelonians had enlarged anteorbital fenestrae allowing for large salt glands and shell fontanelles to increase the hydrodynamic value of their shells (Anquetin 2017). Unlike modern sea turtles, however, they did not possess flippers, but did they have enlarged digits, meaning that they were still competent swimmers and that in life their limbs probably resembled the mini-bat wings of modern soft-shelled and pig-nosed turtles (Anquetin 2017). Because of this it is assumed that, unlike modern sea turtles, angolachelonians did not swim much in the open sea and were mostly near-shore specialists, a few species even having secondarily returned to brackish water biomes (Billon-Bruyat 2005). Nonetheless, most species were in fact fully marine (Billon-Bruyat 2005, Anquetin 2017) and given the fact that a few are found in deep sea deposits (Mateus 2009) and the aforementioned adaptations for efficient swimming I wonder if a few like Angolachelys mbaxi weren’t in fact fully pelagic.

Angolachelys mbaxi by “BoneSharpe”. The artist seems to have agreed with my beliefs .

Both groups of angolachelonians have been recovered as monophyletic true clades, though a ghost lineage for sandownids must have existed for 30 million years or so must have existed to explain the temporal discrepancies between the two groups’ ranges. Thalassochelydians occur both in open sea and marginal marine environments including estuary habitats, putting them as more generalised in terms of ecology, while sandownids are known from both shallow and deep sea sites. Sandownids survived the KT mass extinction much like chelonioidean sea turtles; the exact date of extinction is unclear, but they likely disappeared either at the Eocene Thermal Maximum or in the Miocene alongside bothremyids, both climatic events that deeply affected sea life.


Caribemys oxfordiensis by Dmitri Bogdanov. One of the oldest crown-group turtles known, it seems to suggest that side-necked turtles were originally marine in habits.

Though side-necked turtles are nowadays restricted to the lakes, rivers and swamps of South America, Africa, Madagascar, New Guinea and Australia, in the past they were a cosmopolitan group occuring not only in freshwater bodies in Europe and Asia but also in the seas all over the world. Marine pleurodires are typically treated in literature as multiple different invasions of the sea by various lineages, but given that the earliest known pleurodire, Caribemys oxfordiensis, is a marine animal (Iturralde-Vinent 2001) as well as the fact that extinct pelomedusids do occur in marine settings (Ferreira 2015) it is just as likely that pleurodires were ancestrally marine and later returned to freshwater environments.

Pleurodires are rather distinctive turtles, so unlike angolachelonians they don’t suffer from taxononomical ambiguity in regards to their relations to chelonioids. Marine species don’t seem to differ much from freshwater species, some speciations like enlarged salt glands probably unecessary given how modern pleurodires . Like angolachelonians pleurodires never seem to have developed flippers and seem to be most common in near-shore environments, but even today some modern species , attesting to some moderately efficient swimming capacities ; one species, Araripemys barretoi, . Like modern species, most were probably benthic, ambushing prey from the sea bottom or probing crevaces.

A fictional marine pleurodire species from The Speculative Dinosaur Project, drawn by Brian Choo. Real marine pleurodires did not have flippers (though enlarged hands come close enough), but the “ass-gills” might be truer to form.

The most successful lineage of marine pleurodires were the bothremydids, which occured in marine environments in all continents from the mid-Cretaceous to the Miocene. Some marine pelomedusids . , but at least Stereogenyina species , implying that competition from the latter did them in.


Meiolania platyceps defecating in the sea by Joschua Knüppe. I have nothing to say.

Yes, Meiolania platyceps .

As it turns out, the supposed terrestrial habits for at least this species are erroneous, ranking closer to aquatic turtle species than land dwelling ones. Several unique adaptations of meiolaniid turtles, such as the enlarged nasal area and supposed tail club, are thus interpreted as having evolved in an aquatic context, the former housing salt glands and the latter being simply a reinforced tail, a more extreme version of the armoured tails of freshwater turtles like snappers. As there were no large freshwater bodies in Lord Howe island, Meiolania platyceps is interpreted as having been a coastoal herbivore akin to the modern marine iguana; its extinction is even connected to the declining sea levels of the Pleistocene, which would have surely favoured a terrestrial species but not a coastoal amphibious one.

Of course, a few mysteries remain. As noted in the paper, the short hand of Meiolania platyceps could suggest an ancestry from a terrestrial ancestor, implying that other meiolaniids could have been in fact terrestrial herbivores and that this particular species adopted unique marine habits based on its deficient island environment much like the marine iguana. Conversely, it could imply that other meiolaniids were in fact aquatic and that the Meiolania species of Melanesia are in fact the world’s youngest non-chelonioid sea turtles, having become extinct with the arrival of human beings a few hundred years ago.

Regardless, meiolaniids are nowadays agreed to be outside the turtle crown-group, making them also unique for being a truly ancient turtle lineage that survived into moern times. Assuming marine habits for all recently extinct meiolaniids, these are some of the most un-sea turtle like sea turtles of all time, being weird water dragons rather than graceful udnerwater flyers.


Another speculative pleurodire by Brian Choo.

Fossil turtles are underappreciated enough as it is. Hopefully this little trip through the prehistoric seas has illustrated why the very image of the sea turtle as we know it does not reflect the true diversity of these amazing animals.


Anquetin et al 2017. A review of the fossil record of turtles of the clade Thalassochelydia. Bulletin of the Peabody Museum of Natural History, 58, 317–369.

Octávio Mateus; Louis Jacobs; Michael Polcyn; Anne S. Schulp; Diana Vineyard; André Buta Neto; Miguel Telles Antunes (2009). “The oldest African eucryptodiran turtle from the Cretaceous of Angola” (PDF). Acta Palaeontologica Polonica. 54 (4): 581–588. :10.4202/app.2008.0063.

A new pleurodiran turtle from the Jagua Formation‭ (‬Oxfordian‭) ‬of western Cuba.‭ ‬Journal of Paleontology‭ ‬75‭(‬4‭)‬:860–869.‭ ‬-‭ ‬M.‭ ‬S.‭ ‬de la Fuente‭ & ‬M.‭ ‬Iturralde-Vinent‭ ‬-‭ ‬2001.

Ferreira, G.S., Rincón, A.D., Solórzano, A. & Langer, M.C. (2015) The last marine pelomedusoids (Testudines: Pleurodira): a new species of Bairdemys and the paleoecology of Stereogenyina.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store