The earliest baleen whales didn’t actually have any baleen plates in their mouths, and the evolutionary origin of these unique filter-feeding structures is still poorly understood.

It was thought to have been a fairly simple linear process from toothed ancestors to a mix of teeth and baleen and then to fully toothless with just baleen, but more recent discoveries have begun to cast doubt on that idea. The teeth of ancestral baleen whales weren’t suited to filter-feeding at all, instead still being adapted for predatory piercing and chewing – actions which would have been constantly interfering with and damaging any proto-baleen forming alongside them, and making it seem much more unlikely that there would have ever been a transitional form that had both teeth and baleen at the same time.

But then how did baleen whales get their baleen?

Maiabalaena nesbittae here provides a possible solution. Discovered in Oregon, USA, this early baleen whale dates to the early Oligocene, around 33 million years ago, and compared to most of its modern relatives it was comparatively tiny, only about 4.6m long (15′).

And it had no teeth at all, but possibly also no baleen.

Baleen rarely fossilizes, so it’s unclear whether Maiabalaena actually had any or not, but the shape of its skull suggests it probably didn’t – it lacked the broad thickened upper jaw associated with supporting racks of baleen plates. It instead seems to have been adapted for suction feeding similar to modern belugas and beaked whales, using muscular cheeks and tongue to manipulate water pressure and pull small prey like fish and squid straight into its mouth.

Since it lived at a time when the Antarctic Circumpolar Current was forming and cooling the oceans, changing ecosystems and prey availability, it may represent a previously unknown stage in baleen whale evolution – a point when they’d moved towards specializing for suction feeding and lost their teeth entirely, before transitioning again over to filter-feeding with baleen in a completely separate evolutionary development a few million years later.


Fragmentary fossils of huge azhdarchid pterosaurs have been found in Canada since the early 1970s, and for a long time they were assumed to belong to Quetzalcoatlus. But more recently these remains were re-examined and shown to actually represent an entirely new genus and species.

Cryodrakon boreas – an excellent name meaning “icy dragon of the north wind” – was officially described in late 2019. With a wingspan of around 10m (32’10”) it was similar in size to its close relative Quetzalcoatlus, but it dates to about 10 million years earlier making it one of the oldest azhdarchids ever found in North America.

It lived about 76 million years ago in Alberta, with its fossils coming from the Dinosaur Park Formation, an area that at the time would have been a coastal plain near the northern parts of the Western Interior Seaway. Despite Alberta being located somewhat closer to the Arctic Circle than it is today, the climate was warm-temperate and temperatures rarely dipped below freezing, with short nights in the summers and only a few hours of daylight in the winters.

Like other azhdarchids Cryodrakon would have spent a lot of its time on all fours on the ground. While moving like that it would have been almost 5m tall (16’5″), similar in size to a modern giraffe, stalking smaller animals and eating whatever it could catch and fit into its mouth.


Much like how hyraxes were once far more diverse than their modern representatives, some ancient members of the tapir lineage were similarly weird.

Lophialetes expeditus was one of these odd tapir-relatives, living in Mongolia and China during the mid-Eocene about 48-37 million years ago. Standing around 50cm tall at the shoulder (1’8″) it had a build more resembling a deer or a horse than its pig-like modern cousins, and it was adapted for fast running in open plains, with long slender legs and three-toed hoofed feet that bore most of its weight on the middle digit.

Its skull had a nasal region similar to both modern tapirs and saiga antelope, suggesting the presence of a short trunk-like nose – but since some of its closest relatives didn’t have nearly such well-developed snouts, it seems that Lophialetes evolved its trunk separately to modern tapirs.


Pseudosuchians – the evolutionary lineage whose only surviving modern representatives are crocodilians – first originated in the early Triassic and were once an incredibly diverse group. These croc-relatives experimented with fully erect limbs and bipedalism quite a few separate times, and on several occasions ended up evolving remarkably similar body plans to their distant cousins the theropod dinosaurs.

One of the earliest branches of the pseudosuchians to do this were the ornithosuchids, the best known of which is Riojasuchus tenuisceps here.

Living in Argentina during the Late Triassic, about 217-215 million years ago, Riojasuchus had a distinctive “hooked” upper jaw and two rows of osteoderm armor plates along its back.

It was only around 1.5m long (4’9″), much smaller than some of the other pseudosuchians and early theropod dinosaurs it lived alongside. Its front limbs were shorter than its hind limbs and it was probably a facultative biped – moving slowly on all fours, but getting up on just its hind legs for bursts of high speed running – which would have helped it avoid being eaten by those larger predators.

Like other ornithosuchids it had very strangle ankles, with the bones in the joint articulating with each other the opposite way around compared to any other type of archosaur. The claws on its hind feet were also unusually tall and narrow, especially on the inner toes.

Its jaws were capable of delivering strong but somewhat slow bites, and the relative structural weakness of its narrow notched jaw would have made it difficult for it to deal with large struggling prey. It likely mostly hunted smaller vertebrates, and may also have been an opportunistic scavenger taking bites out of larger predators’ kills whenever it got the chance.


Most ichthyosaurs were similar in size to the modern dolphins they convergently resembled, but during the Late Triassic some of them got much much bigger.

Shonisaurus popularis lived about 222-212 million years ago, in Nevada, USA — a region that’s currently made up of dry deserts, but which was submerged under a tropical inland sea at the time.

At around 15m long (49′) it was roughly the same size as a modern humpback whale, with a long narrow snout, a fairly deep fusiform body, and four equally-sized flippers. Unlike many other ichthyosaurs it doesn’t seem to have had a dorsal fin, and its tail fluke shape was rather “primitive” indicating it was probably a slow cruising swimmer.

Juveniles had a few small teeth at the tips of their jaws, but larger adults were entirely toothless, suggesting that they may have specialized in different ecological niches at different stages of their lives. Fully-grown Shonisaurus probably mostly fed on prey such as soft-bodied cephalopods and small fish, which must have been incredibly abundant in the ancient Nevadan sea to support a population of such huge marine reptiles.

But Shonisaurus popularis wasn’t even the biggest of the Late Triassic giant ichthyosaurs. Further north in British Columbia, the closely related species Shonisaurus sikanniensis reached lengths of up to 21m (69′), and fragmentary remains from England hint at something even larger still, estimated at around 25m (82′) – close in size to the modern blue whale, and potentially being the largest non-dinosaurian reptile to ever live.