The extinct pycnodonts were a group of mostly circular-shaped fish, convergently similar to modern reef fish like marine angelfish or butterflyfish – but some of them developed much much weirder appearances.

Rostropycnodus gayeti here was one of the especially odd-looking forms, known from the mid-Cretaceous of Lebanon about 100-95 million years ago.

It had an elongated snout with the upper jaw longer than the lower, a pointed spiky horn on its forehead, and a massive pectoral region that bulged out at the front of its body. Meanwhile its pectoral fins were modified into big immobile spines, and its pelvic fins were highly reduced and positioned beneath another set of large spines.

And it was tiny, only about 5.5cm long ~(2″).

It would have been a slow swimmer, relying on its spikiness to deter larger predators, and it’s currently unclear what it ate with its unusual spiny snout. Many other pycnodonts had mouths full of round crushing teeth, but Rostropycnodus’ jaws seem to have been mostly toothless – so perhaps it used its snout to probe around in cracks or sediment for small soft-bodied invertebrates.


Did you know butterflies weren’t the first insects to look like butterflies?

Lepidopterans (the group of insects containing moths and butterflies) have been around since the Late Triassic – but it wasn’t until the diversification of flowering plants during the Cretaceous that recognizable moths would have evolved, and true butterflies didn’t actually appear until the early Cenozoic.

Before then, back in the mid-Jurassic about 165 million years ago, a completely different group of insects convergently evolved remarkably butterfly-like features such as large colorful scaled wings and long sucking proboscises.

Known as the kalligrammatids, these insects were giant members of the lacewing group, related to modern forms like antlions and owlflies. But unlike their predatory relatives the kalligrammatids were specialized pollinators, possibly having a mutualistic relationship with the flower-like cones of bennettitales or the pollination drops of some types of conifers. They seem to have originated in China and were found across Asia and Europe by the Late Jurassic, but a few fossils from South America suggest they were even more widespread and may just have a poor fossil record.

They reached wingspans of up to 16cm (~6″), comparable to some of the largest modern butterflies, and often sported conspicuous anti-predator markings on their wings such as stripes and eyespots – so it’s not surprising that they’re often nicknamed the “butterflies of the Jurassic”.

A fossil of a butterfly-like insect. Stripes and eye-spot markings are preserved on its wings.
Markings preserved on the wings of Oregramma illecebrosa, from Yang et al (2014) | CC BY 2.0

Rather ironically, the extinction of the kalligrammatids was probably linked to the rise of the flowering plants that the true butterflies would later be so dependent on. As flowers diversified and plants like the bennettitales declined, the kalligrammatids dwindled and disappeared, with the last known fossil record coming from the mid-Cretaceous of Brazil about 113 million years ago.

But while they were around, I do wonder if they also exhibited some similar behaviors – such as mud-puddling for extra nutrients, and specifically the habit of drinking the tears of larger animals that we see in some species. Perhaps some non-avian dinosaurs like this Dilong occasionally put up with kalligrammatids sitting on their faces!


This is not a deer.

In Africa during the Eocene and Oligocene, the main terrestrial herbivores were a different type of mammal entirely: hyraxes, the close relatives of elephants and manatees. Although their only modern representatives are small climbing rodent-like animals, hyraxes were once a much more diverse and widespread group, filling a variety of ecological niches and ranging from the size of rats up to the size of rhinos.

Antilohyrax pectidens was a mid-sized example of these diverse hyraxes, standing about 50cm tall at the shoulder (1′8″) and living around 34-28 million years ago in Egypt. It had a deer-like snout and long slender limbs adapted for running and leaping, with leg bones incredibly similar in size and proportion to modern springbok.

Its incisor teeth were comb-shaped and resembled those of colugos, so it was probably a similar sort of selective browser eating soft leaves and shoots.


Elasmosaurids are often depicted with noodly snake-like or swan-like necks, but they were probably actually quite stiff and inflexible in life. And while we know from fossilized gut contents that they ate relatively small prey like fish, crustaceans, and cephalopods, exactly how they used their distinctive long necks is still uncertain.

There’s some variation in the sizes and shapes of their teeth, so it’s likely each species was specialized for slightly different feeding styles – we’ve even found a filter-feeding one! – and the recently-named Leivanectes bernardoi here adds in a little more diversity, too.

Living about 115-112 million years ago during the mid-Cretaceous of Colombia, Leivanectes would have been fairly large at around 9m long (29′6″), slightly bigger than the other elasmosaurid species known from the same ancient marine deposits. It had a reduced number of teeth in its jaws, but these teeth were also proportionally larger, suggesting that it may have been tackling bigger tougher prey than its relatives.

Unfortunately it’s currently only known from a single partial skull, so we don’t have any other clues about its ecology.