It Came From The Wastebasket #16: Catopsalis Catastrophe

The rodent-like multituberculates were a major lineage of mammals that were only distantly related to modern marsupials and placentals. They originated around the time of the mid-Jurassic (~168 million years ago), survived through the end-Cretaceous mass extinction, and went on to become one of the most diverse and successful types of mammal in the Paleocene. After that point they began to decline, and after anw over-130-million-year-long run they went extinct* in the early Oligocene (~33 million years ago).

(* Except, possibly, in South America, where an enigmatic fossil known as Patagonia peregrina may represent a multi surviving as recently as about 18 million years ago in the early Miocene.)

First discovered in North America in the 1880s, Catopsalis foliatus was part of a group of multituberculates called taeniolabidoids. These multis got significantly larger than the rest of their kind – averaging beaver-sized but with some species getting up to at least capybara-sized – and were some of the first mammals to evolve into relatively big herbivores after the extinction of the non-avian dinosaurs.

An illustration of Catopsalis, an extinct multituberculate mammal. It resembles a rodent, with a whiskery nose, large eyes, small rounded ears, short clawed legs with spurs on its ankles, and a long tufted tail. Its colored mostly brown with pale spotted markings along its sides.
Catopsalis foliatus

Catopsalis was named based on a partial jawbone and a few teeth, and over the next century or so various other similar-looking fossils from both North America and Asia were added into the genus as additional species. Eventually Catopsalis contained eight different species, ranging over about 10 million years from the late Cretaceous to the early Eocene – not especially big compared to some other wastebaskets we’ve looked at this month, but it was still a problem, muddying up attempts to understand the actual evolutionary relationships and biogeography of the taeniolabidoids.

Cladistic studies in the 1980s showed that Catopsalis was paraphyletic, made up of at least five separate lineages, and a few of them were subsequently renamed and reclassified. The Cretaceous Asian forms became Djadochtatherium and Catopsbaatar, and are now considered to be part of a different lineage of multis known as djadochtatherioids, while one of the remaining North American species then became Valenopsalis.

…But a couple of other new Catopsalis species have also been named in the meantime (one as recently as 2018), so there are still seven different species that need sorting out in this particular wastebasket.


Even for a fossil species from an isolated island, Adalatherium hui is very weird.

This mammal was part of an enigmatic group known as gondwanatheres, which were probably early members of the theriiform lineage – slightly closer related to modern marsupials and placentals than to monotremes. Found in the southern continents of Gondwana between the Late Cretaceous and the Miocene, these animals were adapted for herbivory with convergently rodent-like ever-growing front teeth that helped them chew through tough plant matter.

They were previously known mainly from isolated teeth and jaw fragments, with some rare full skull material, but Adalatherium is remarkable for being represented by a complete skeleton.

And it’s turned out to be far stranger than anyone expected.

Living in northwestern Madagascar during the Late Cretaceous, about 70-66 million years ago, Adalatherium was one of the larger known Mesozoic mammals at around 60cm long (2′) – although the one known specimen seems to have been a juvenile, so mature individuals were probably slightly larger.

(And based on its body proportions, its close relative Vintana may actually have been even bigger than previously thought. Whether this sort of large size was common in Cretaceous gondwanatheres or if this was just island gigantism is still unknown, though.)

It was probably a marmot-like digging animal, excavating burrows with its large claws and powerful limbs, and since it likely evolved from ancestors that had become isolated on Madagascar over 20 million years earlier it had developed a very unusual mixture of both “primitive” and highly specialized anatomical features. It had more back vertebrae than any other known Mesozoic mammal, upright forelimbs, sprawling hind legs with bowed-out tibias, strong back and leg musculature, and a therian-like pelvis with epipubic bones.

And then there’s the snoot.

The snout region of Adalatherium‘s skull was pockmarked with a large number of foramina, holes that allow the passage of nerves and blood vessels through the bone. It had more of these than any other known mammal, and their presence suggests that it probably had a very sensitive upper lip and whiskery snout. Most mammals with a lot of whiskers just have one very big foramina, but Adalatherium seems to have evolved a different solution to the same problem.

It also had one other bizarre feature – a hole in the top of its nose. A large “internasal vacuity” between its nasal bones is a unique feature not known in any other mammal, and its function is a total mystery.

Since this hole was also surrounded by many foramina it may have supported some sort of soft-tissue sensory structure on top of its nose. So I’ve speculatively depicted it here with a leathery horn-like “shield”.

Adalatherium skull
From fig 2 in Krause, D. W. et al (2020). Skeleton of a Cretaceous mammal from Madagascar reflects long-term insularity. Nature 581, 421–427.

Island Weirdness #10 – The Kogaionids

Multituberculates were a group of rodent-like mammals that originated back in the Early Jurassic, at a point on the mammalian family tree between the origin of monotremes and the earliest therians (represented today by marsupials and placentals).

While they were an incredibly successful group, found around the world in large numbers, in Late Cretaceous Europe multis had become incredibly rare and restricted to just a single place: Hațeg Island.

Isolated there, they evolved into a unique family known as the kogaionids, diverging from their ancestral mostly-herbivorous diet to instead become specialized insectivores with distinctly red iron-pigmented teeth and huge blade-like lower premolars.

Skull of Barbatodon, from fig 2 in Smith T, Codrea V (2015) Red iron-pigmented tooth enamel in a multituberculate mammal from the Late Cretaceous Transylvanian “Haţeg Island.” PLoS ONE 10(7): e0132550. doi: 10.1371/journal.pone.0132550  | CC-BY-4.0

Some of them also had oddly domed skulls and proportionally tiny brains, along with highly acute senses of smell, eyesight, balance, and motor control.

Kogaionon ungureanui was one of the first kogaionids to be discovered, and gives its name to the group as a whole. Although known only from a skull, it was probably rat-sized, around 30cm long (~12″).

Unusually for island species, which are often ecologically fragile and vulnerable, the kogaionids’ insectivorous habits allowed them to successfully survive through the end-Cretaceous mass extinction 66 million years ago while the Hațeg dinosaurs and pterosaurs perished. And when conditions changed and their island home became reconnected to the rest of Europe they rapidly spread out and became common across the entire region for a further 10 million years, only finally disappearing in the early Eocene about 56 million years ago.


Litovoi tholocephalos, a multituberculate mammal from the Late Cretaceous of Romania (~70-66 mya). Living on what was at the time the large offshore Hațeg Island, this rat-sized animal (about 25cm /10″ long) was part of a lineage of insectivorous multis called the kogaionids, with the same sort of red-colored enamel on its teeth as other species like Barbatodon.

Its brain was surprisingly tiny proportional to its size – one of the smallest known brain-to-body ratios of any mammal, and more similar to those of non-mammalian cynodonts – but it also seems have been highly specialized for processing sensory input, with relatively enormous regions associated with smell, eyesight, balance, and motor control. The olfactory bulbs of its brain were so enlarged, in fact, that they caused its skull to bulge out into an unusually dome-shaped forehead.

Its reduced brain size may have been due to limited food availability on its isolated island home. Brains are very metabolically expensive organs, and some other extinct island mammals like hippos, hominids, and goats are also known to have evolved smaller brain sizes. Modern shrews even seasonally shrink their own brains during winter for similar energy-saving reasons.

Almost-Living Fossils Month #04 – The Last Multis

Known almost exclusively from the southern continents of Gondwana – hence their name – the gondwanatheres were part of a widespread and very long-lived group of mammals known as multituberculates.

Although multis resembled placental rodents, and gave birth to tiny undeveloped young in a similar manner to marsupials, they originated much further back in the mammal evolutionary tree. They existed by at least the Early Jurassic (~183 mya), and their ancestry may go even further back into the Late Triassic (~220 mya) if they were descended from haramiyidans.

Multis survived the end-Cretaceous extinction and became very diverse through the first half of the Cenozoic, until a combination of factors such as climate shifts, new types of vegetation, and the evolution of new mammalian predators (and possibly also competition from placental rodents) resulted in most of them going extinct by the Early Oligocene (~33 mya) – with only the gondwanatheres surviving past that point in the then-isolated continent of South America.

Patagonia peregrina was the very last known gondwanathere in the fossil record, living just 21-17.5 million years ago in the Early Miocene of Argentina. Although only teeth and jaw fragments have been found so far, it was probably about 15cm long (6″) and would have been a burrowing herbivore similar to modern gophers or tuco-tuco. Its ever-growing rodent-like teeth were adapted for grazing on tough grasses in its savanna-like habitat, and it would have lived alongside several other now-extinct types of mammal – but we’ll be getting to those ones later in the month.

Since it seems like these last gondwanatheres had survived by retreating into a rather specialized ecological niche, they sadly probably didn’t persist for very long beyond the time of Patagonia. A wave of extinctions associated with sudden climate cooling about 14 million years ago may well have been the final blow to the once-successful lineage of the multituberculates.

Month of Mesozoic Mammals #20: Hidden Herbivores


For a long time very little was known about a group of mammals called gondwanatheres. Named for their occurrence in the southern continents that made up Gondwana, they were represented only by fossil teeth and jaw fragments, and it wasn’t even clear what type of mammal they actually were. But recent discoveries of more complete skulls (and a currently undescribed full skeleton*) are starting to reveal more information, and we now know they were actually part of the multituberculates, or at least very closely related to them.

* described and named as Adalatherium in 2020

They were also the latest-surviving of the multis, lasting well into the Cenozoic with the youngest known fossils dating to just 17.5 million years ago.

Vintana lived during the Late Cretaceous of Madagascar (70-66 mya), and is known from a single skull. It was a specialized herbivore with rodent-like teeth adapted for chewing tough plants, seeds, and roots, and huge powerful jaw muscles attached to downward-extended cheekbones – a feature convergently seen in a few other mammals.

It was one of the largest known mammals of its time, estimated to have been at least 60cm long (2′). It had relatively large eyes, well-developed inner ears, and an expanded area of its brain associated with processing scents, all features that indicate it had very keen senses and may have been quite a fast and agile animal.

Month of Mesozoic Mammals #19: Red In Tooth


Living during the Late Cretaceous of Romania (70-66 mya), Barbatodon was part of a group of European multituberculates known as the kogaionids. These multis originated on the then-isolated Hațeg Island alongside dwarf dinosaurs and giant pterosaurs, and adapted to a highly insectivorous diet. They even managed to survive through the end-Cretaceous mass extinction, and spread across Europe for a further 10 million years before going extinct in the early Eocene about 56 million years ago.

Barbatodon is mainly known from teeth and partial skull material, so its full size is uncertain, but it was likely rat-sized at around 25-30cm long (10-12″). In one specimen its teeth were also preserved with their original coloration – a distinctive “blood red”. This feature is seen in some modern rodents and shrews, and is caused by iron minerals in the enamel that are thought to add extra strength. Since multis didn’t have ever-growing teeth like rodents, this added durability would have been especially important to them.

Another group of multis, the taeniolabidoids, also had red teeth, but since fossil enamel is rarely so well-preserved and unaltered we don’t know whether this was a shared ancestral feature or due to convergent evolution.

skull of Barbatodon || from fig 2 in Smith T, Codrea V (2015) Red iron-pigmented tooth enamel in a multituberculate mammal from the Late Cretaceous Transylvanian “Haţeg Island.” PLoS ONE 10(7): e0132550. doi: 10.1371/journal.pone.0132550 || CC-BY-4.0

Month of Mesozoic Mammals #18: Leaps and Bounds


Known from the Late Cretaceous of Mongolia (85-70 mya), Catopsbaatar was a fairly large multituberculate, similar in size to a modern chinchilla at about 40-50cm long (1′4″-1′8″) with about half of that length being its tail.

It was part of a group of Asian multis called the djadochtatheriids, which lived alongside famous dinosaurs like Velociraptor in a sandy desert environment. They were mostly jerboa-like animals capable of bipedal hopping – although one member of the group, Mangasbaatar, was a burrower instead.

Although Catopsbaatar had features in its vertebrae and hindlimbs convergently similar to those of modern hopping mammals, the somewhat more sprawling posture of multis mean it wouldn’t have jumped in quite the same way. It may have actually launched itself upwards at a steeper angle, in a manner a little more like a frog.

Djadochtatheriids weren’t the first hopping Mesozoic mammals, however, since fossilized footprints are known from both the Mid-Jurassic of South America and the Early Cretaceous of Korea. We don’t know what types of mammals made these tracks, or what they looked like, but they show that similar styles of locomotion may have evolved multiple times in early mammals.

Month of Mesozoic Mammals #17: Awesome Ankles


Moving on to the next major group of the theriiform mammals, we have the multituberculates – or “multis” for short.

First appearing in the Early Jurassic, about 183 million years ago (possibly descending from haramiyidans), multis were one of the most successful and long-lived mammal lineages of all time, found throughout the world and making up more than half the known mammal species in some fossil deposits. They even made it through the end-Cretaceous mass extinction and became even more diverse in the Paleocene, although shifts in vegetation, climate change, and the rise of new predators seem to have sent them into decline by the mid-Cenozoic. The last surviving group of multis, the gondwanatheres, finally went extinct in the Miocene (~17.5 mya).

Multis had rodent-like teeth, except with huge blade-like lower premolars, and likely occupied similar ecological niches to their modern counterparts. The structure of their pelvises also suggests they were some of the earliest mammals to give live birth to tiny undeveloped young, similar to marsupials.

Ptilodus skull by Nobu Tamura || CC-BY-3.0

Rugosodon is known from the Late Jurassic of China (161-155 mya) and is one of the earliest multituberculates represented by near-complete fossil remains.

About 25cm long (10″), it was a ground-dwelling chipmunk-like animal with highly flexible ankle joints that would have made it very a fast and agile runner, capable of navigating uneven surfaces. These specialized ankles were a defining trait of multis, allowing later forms to adapt to lifestyles ranging from tree-climbing to burrowing to jerboa-like hopping.

And while many later multis were primarily herbivores, Rugosodon’s teeth show it was an omnivore, indicating that a more generalized diet was ancestral to the group.