Aquatic Dinosaurs (besides birds)
The idea that dinosaurs never ventured into aquatic niches is perhaps one of the most frequent ways in that media distinguishes dinosaurs from marine reptiles like ichthyosaurs. It is not without merit; the only dinosaurs that produced marine forms were ornithurine birds, and within these only Hesperornithes, penguins and plotopterids became recognisable as cetacean-like animals (smaller, flightless seabirds like the Great Auk and the Flightless Cormorant are usually nested within clades of capable flyers).
However, we now know that dinosaurs did in fact venture into aquatic niches. Much like the Cenozoic saw several clades of placental mammals venture into freshwater ecosystems, the lakes, rivers and swamps of the Mesozoic were equally filled with several semi-aquatic dinosaurs, either hunting underwater like modern cormorants or seeking refuge from predators like modern swamphen. While a fully aquatic non-avian dinosaur most likely never existed, dinosaurs most certainly swam.
Ornithischians as a whole don’t seem to have ventured much into aquatic niches, nor did sauropods (ironically enough). One possible explanation is the absence of aquatic angiosperms through most of the Mesozoic, thus giving herbivores less reasons to take aquatic niches. Some primitive ceratopsians and hadrosaurs have been interpreted as semi-aquatic; while these opinions are usually considered delusional ramblings, several north american ceratopsid and hadrosaur taxa might have engaged in semi-aquatic behaviour, specially species found in the vicinity of the Western Interior Seaway. Both Laramidia’s and Appalachia’s shores were dominated by wetlands, which would have encouraged such behaviour.
One of the most logical candidates for aquatic ornithopods is Lurdusaurus arenatus, a large iguanodont (possibly a close relative of hadrosaurs) from the Aptian of North Africa. This ornithopod is noted as having a low, robust torso, similar to that of an hippo, and limbs that are short and robust, being poorly adapted for running, and also unusually dense, like those of modern diving birds. The short, slightly spreading metatarsus is backed by an enlarged foot pad, in keeping with the massive, spreading hand, a possible adaptation to move in soft substrates. The ecosystem it lived in, dominated by tropical swamps, would have certainly encouraged a semi-aquatic lifestyle, if only as means from protection from predators like carcharodontosaurs.
Thescelosaurids like Thescelosaurus itself have also been suggested as being semi-aquatic, having short, robust legs that are poorly suited for running and having lived in what appears to have been a wetland ecosystem. In these animals, like in modern hippos, capybaras and tapirs, terrestrial foraging was likely the main source of food, with semi-aquatic behaviour being mostly for protection, either directly evading predators or sleeping in the water/islands away from shore like in modern aquatic birds.
Evidence for aquatic behaviour in basal theropods is mostly composed of aquatic trackways made by an animal while swimming. Said trackways are usually attributed to a coelophysid, probably an animal similar to Megapnosaurus. It is possible that said trackways might have been made an animal that was swimming casually, not really being specialised to an aquatic lifestyle. However, coelophysids, with their long rostrums and stork-like necks and legs, might have lived like modern day wetland birds such as herons, wading in the shallows in search of prey.
Dilophosaurs are sometimes assumed to have been dedicated piscivores, though so far this has only been inferred from interpretations of their dental apparatus, which, while unusual, can also be interpreted as apropriate for flesh slicing.
Ceratosaurus is considered by Robert Bakker to be a specialised predator of aquatic prey. It’s tail is described as “alligator like”, making it a good swimmer, while the dentition is supposedly apropriate to capture aquatic prey like fish and crocodiles. It’s inferred lifestyle is overall similar to that of spinosaurs (see below), being a semi-aquatic predator that actually chased prey underwater like an otter or a cormorant, rather than just wading. However, given how frequent it’s teeth are found on terrestrial dinosaurs, this lifestyle is not taken seriously anymore (
as 70% of Bakker’s work anyway).
Abelisaurs have been inferred occasionally as aquatic animals, based on their robust limbs. Now that we know that many were actually fast predators like modern cheetahs, or sauropod specialists, an aquatic lifestyle is considered unlikely.
However, a close relative of abelisaurs, Masiakasaurus, is thought to have been a piscivorous animal, thanks to it’s bizarre dentition. Compare to other noasaurs, it’s body is also more robust, though nowhere as near as in spinosaurs, suggesting that this animal was transitioning from a wader like coelophysids to a diver, meaning it probably hunted more often in the shallows.
Spinosaurids were the apex of aquatic behaviour in non-avian dinosaurs. With robust skeletons, gharial like jaws (I’m sick of folks who say they had crocodile like jaws; spinosaur jaws were long and thin, like those of gharials and ornithocheirid pterosaurs, not wide like those of crocodiles; likewise, their bite force was also considerably weak, comparable to that of gharials, unlike the tremendous bite force of crocodiles), nostrils high up the jaws and limbs more adapted for swimming than running after prey, spinosaurids were essencially the closest dinosaurs came to seals and archaeocete whales, being diving predators.
Oxygen isotope ratio studies famously indicated that, rather than just wading like herons, spinosaurs actually swam after prey, spending most of their lives on water. Even the supposedly least aquatic spinosaur, Spinosaurus itself, had it’s isotope ratios much closer to those of modern crocodiles than to those of other theropod dinosaurs, indicating that, rather than a terrestrial carnivore or a “bear analogue”, this animal was living like modern day gharials and similar crocodyllians, spending most of it’s time swimming, coming ashore only to bask, to lay eggs and, occasionally, to sleep (which it could also do on the water, of course). Indeed, spinosaurids appearently outcompeted pholidosaurids, a lineage of gharial like crocodyllians. On other words, spinosaurs were more efficient aquatic predators than crocodyllians themselves! (of course, as indicated behaviour, terrestrial niches were a different matter entirely)
Spinosaurids were not above hunting terrestrial prey; one pterosaur fossil shows evidence of being attacked by a spinosaurid, while an iguanodont juvenile bears similar tooth marks. Modern gharials still do ambush terrestrial prey from time to time, however, so it is possible that such incidents happened on unlucky land animals being near the water. Indeed, such incidents appear to not have been common place; Spinosaurus itself has no evidence of having predated land animals, but there are many fossils that indicate a predator/prey relationship between Spinosaurus and several contemporary fish, such as Onchopristis.
The demise of spinosaurids is similarly fitting for aquatic animals; they were hit very hard by the Cretaceous Thermal Maximum, of which most tetrapod victims were aquatic (such as nearly all marine sauropsid clades with the exception of two plesiosaur clades and mosasaurs). Mid-Santonian remains in China indicate that spinosaurids did survive the CTM, but they appearently never returned to their original splendor, perhaps coinciding with the success of champsosaurs.
Aside from Avialae, aquatic lifestyles are not known in Maniraptora except for one clade.
Unenlagiines were as a whole stork like critters, with long legs and long rostrums, being adapted to hunt small prey. Much like modern wading birds, their legs appear to be less suited for running when compared to those of other deinonychosaurs like troodontids, indicating that they presumably spent most of their time wading on the shallows, safe from terrestrial predators. Likewise, competition with the most defenitely terrestrial azhdarchid pterosaurs would have also encouraged a piscivorous lifestyle (though azhdarchids do appear to have replaced unenlagiines in parts of South America).
One particular genus stands out among unenlagiines. The largest member of this clade, Austroraptor, appears to have occupied a similar niche to spinosaurids; it is far more robust than it’s relatives, and it became much larger, indicating that it too dived after prey instead of just wading. Unlike spinosaurids, it had small arms, indicating that it probably engaged in cormorant style swimming. Considering that unenlagiines used their arms as wings, having small arms is presumably indicative of it being analogous to the modern Flightless Cormorant, having given up flight for an aquatic lifestyle.