Sonar and the Fossil Brain

Most species of large, placental herbivore belong to a group broadly defined as the "ungulates". There are exceptions; animals such as elephants, which were once thought to be related to the ungulates, but later turned out to have a completely different origin, some of the larger primates, and the occasional oddity such as the panda. In general though, outside of Australia, the ungulates have been the most successful of the large mammalian herbivores. The group includes two, related, orders of mammals. The even-toed ungulates, including antelopes, cattle, deer, and pigs, among others, are by far the larger group today, while the odd-toed ungulates, which include horses, rhinos, and tapirs, are less numerous, but just as familiar.

Both of these groups, and possibly some of the weird-looking prehistoric herbivores of South America, are descended from a general type of animal called "condylarths". The condylarths are probably not a real evolutionary group of animals, in that they aren't all necessarily closely related, and even if they were, they would just be a rough assemblage of things related to the modern groups, rather than a group in their own right. They may well be what is called a "wastebasket taxon" - a handy place to put some animal whose relationship to other animals you aren't really sure of yet.

Nonetheless, regardless of the status of the condylarths as a whole, at least some of them must have been the ancestors or close relatives of the ungulates that followed. They lived during the Paleocene and Eocene, the first two epochs of the Age of Mammals, and they are relatively unspecialised animals, only beginning on the path that would later lead to the herbivores we have today. For example, many of them still had claws, something not found on any living ungulate, which generally have hooves, but they all had elongated toes that, at the very least, had hoof-like features.

Among the last condylarth families to die out were the hyopsodontids. These were, for their day, a remarkably successful group of animals; in North America they are easily the most common mammalian fossils from the mid to late Eocene, and fossils have also been recovered from Britain, France, and Mongolia. In evolutionary terms, they appear to be closer to the odd-toed than the even-toed ungulates, although they can't have been their direct ancestors - we know of some odd-toed ungulate fossils that predate them.

The above is just one possible interpretation of how these groups relate to one another, and many others exist. Aside from the ungulates themselves, all of these groups can be considered 'condylarths', and for simplicity, some condylarths are omitted from the diagram, including some related to the even-toed ungulates, and others branching off even before the acrtocyonids. That 'condylarths' appear at three different points on the diagram does, however, show that they cannot be a single evolutionary unit of animals.

The best known genus within the family is that for which it is named, Hyopsodus. Indeed, it is the only one for which we have a reasonably complete skeleton, and it so common that it represents 60% of all mammalian fossils recovered from the major Bridger formation in Wyoming, as well as being found elsewhere across the northern hemisphere. It was typically about the size of a hedgehog or large rat, although there was some variation among the different species, of which at least nineteen are known.

Hyopsodus had a slender, somewhat elongated, body, with short limbs, making it look a little like a polecat. It had gripping claws and flexible ankle joints that suggest it may have been at home in the trees, although the shape of the chest suggests powerful muscles suitable for digging. The forelimbs don't have the sort of adaptations we would expect to find in a burrowing animal, but it might at least have scratched at the ground in search of food. Unlike modern ungulates, it has a full set of teeth, but the canines are tiny, suggesting that it may have preferred a herbivorous diet, although its certainly plausible that it would also have eaten the odd insect or worm.

But how much else can we tell about its lifestyle and habits? We may have some complete and well-preserved skeletons, but many clues about its life will be hidden in the soft tissue parts, most of which we can know nothing about. Most... but not quite all.

The brain is unusual among bodily organs in being more or less directly encased in bone. The inner surface of the skull doesn't trace out every tiny fold on the surface of the brain, but it does outline it in a fair amount of detail. With the aid of modern CT scanning it is possible to examine the inside of a fossil skull and build up a 3-dimensional image of the brain that once lay within. This has, in fact, been done for Hyopsodus before, but a new study by Maeva Orliac of the University of Montpellier, and colleagues took advantage of a better preserved skull than ever before.

One of the resulting images is shown to the left, and illustrates the deduced shape of the brain as seen from above. Together with similar endocasts taken from other fossils, and with what we know of the brains of living mammals, this can provide us with significant information about how Hyopsodus lived, and how it relates to other animals.

Firstly, what about the size of the brain? Obviously, this is a small animal, so the brain is not going to be huge, and the total volume comes out at less than 3 cm³. What exactly that means rather depends on how big we think the rest of the animal was. Since there isn't anything very much like it alive today, that has resulted in a range of opinions as to how much Hyopsodus would have weighed when it was alive. Still, it's probably easier than trying to guess the weight of, say, dinosaurs, and the authors of this study came out towards the lower end of the range of estimates, roughly that of a large rat.

If that's right, then the brain is fairly large for the animal's size, at least as compared with other condylarths such as Phenacodon or Arctocyonix, and it's perhaps comparable with that of some of the earlier true ungulates, such as Hyracotherium. But the relative proportions of the brain can also tell us a lot.

In general, the brain seems to show a mixture of features that are, for ungulates, both primitive and advanced. For example, the olfactory bulbs ('ob' on the diagram) are large. These are the structures involved with the sense of smell, and they're generally quite small in ungulates, suggesting a more primitive lifestyle closer to that of shrews than that of horses or deer. On the other hand, the cerebrum ('np' on the diagram), the part of the brain in humans that is involved, among other things, with sensation and conscious thought, is relatively large, albeit primitive in its shape. This may suggest an animal that was - at least for the era in which it lived - fairly intelligent. It's not very impressive by the standards of living mammals, perhaps, but, back then, it didn't have to be.

Brain (purple) and sinuses (blue) of Hyopsodus lepidus

The most dramatic prediction of the paper, however, comes from the unusual size of a structure called the inferior colliculus ('ci' on the diagram). This is a relay and processing centre for information concerning sounds, providing output to sound-related reflexes (such as the 'startle' response), and to the higher centres of the brain responsible for the actual sensation of hearing. So, presumably, this was an animal with a good sense of hearing, and that probably relied on sound and smell more than it did on vision. But might it mean more than that?

Another group of mammals with similarly developed brain structures are the bats. Bats presumably need a large inferior colliculus because of their echolocation abilities, so might that also be true of Hyopsodus? Of course, regardless of what else it might have been doing, it's a safe bet that Hyopsodus couldn't fly, and, while their fossils are often found near old lake beds, they clearly weren't as aquatic as dolphins, either. Yet we have known since the 1960s that bats and whales are not the only mammals to make use of echolocation: shrews and tenrecs do, too.

These nocturnal ground-living animals use echolocation primarily to find tunnels in which to hide from predators, and probably also to find their away about inside such tunnels without bumping into things. Hyopsodus didn't have the adaptations to dig it's own tunnels, but it did have an elongated, weasel-like, body that it could have used, just as weasels do, to crawl into narrow passages dug by something else. Significantly, shrews and tenrecs also have a large inferior colliculus, with the brains of tenrecs, in particular, apparently being similar to those of Hyopsodus.

Did this supposedly primitive animal really use ultrasonic clicks to navigate its away about at night or underground? That's not something we can prove either way, and there are other possible explanations for its brain anatomy. But it is a possibility, and, at the very least, the large size and proportions of the animal's brain do seem to indicate that it was more intelligent and "advanced" than we had previously given it credit for.

Perhaps that is part of the reason that it outlasted so many of its fellow condylarths.
.
[Reconstruction picture by "Apokryltaros" from wikimedia commons, endocast image from Orliac et al. 2012.]