Oligocene (pt 15): Land of the Fire-Beasts

Pyrotherium

Even at the dawn of the Oligocene, South America had already long been isolated from the northern continents, giving it the opportunity to evolve its own distinct mammalian fauna, with many animals quite unlike those seen elsewhere. Most of these strange animals would die out millions of years later when the Isthmus of Panama finally formed. The armadillos are among the exceptions.

Armadillos first evolved on the continent during the previous epoch, if not earlier, but most of the older fossils are incomplete, making it difficult to trace their detailed relationships. The oldest reasonably complete armadillo skulls belong to Kuntinaru, first described in 2011. This lived in Bolivia towards the end of the Oligocene, around 27 million years ago and would already have looked much like modern armadillos, albeit somewhat smaller than the species most familiar to North Americans. 

Indeed, the skull is sufficiently well preserved that we can place it in the same subfamily as the living three-banded armadillos that still live in Bolivia today (among other places). Since genetic evidence suggests that the subfamily may have originated no later than 29 million years ago, this is likely a very early example of its kind, confirming just how old the group is.

Not all armadillos of the day would have been so familiar, however. Peltephilus was about the size of the largest nine-banded (North American) armadillos of today, and likely had a similar body form. However, it was unusual in that two of the bony plates on the snout were conical, rather than flat, effectively giving a pair of side-by-side horns. 

Long thought to be a carnivorous predator on the basis of its teeth, more recent analysis of the rest of the body shows that it was adapted for digging rather than running, and thus probably had a lifestyle closer to modern armadillos. It likely fed on tough plant food rather than meat. Quite what it wanted the horns for remains a mystery; they might have helped with digging, but it's at least as likely they were defensive, as is now thought to be the case for the "horned gophers" of the later, Miocene, epoch.

Beyond this, most armadillo remains from the Oligocene are fragmentary. Many species have been named, but we often have little more than some of their bony scales, giving us little clue as to what the animal they belonged to looked like. A notable example here is Glyptaltelus, thought to be an early, and much smaller, relative of the massive glyptodonts of later epochs, although it likely already had the solid dome-like shell of interlocking plates. 

The oldest known sloth fossils date to the Oligocene, and this also accords with what we have gleaned from genetic analysis, which suggests that they first arose early in the epoch. (This, it should be noted, is using the 'crown group' definition, meaning anything descended from the last common ancestor of the living animals; sloths are weird enough that that last common ancestor must itself have had a long, and distinct, evolutionary history). Of course, here, we're not talking about tree sloths, which are relatively modern, but the ground sloths from which they evolved.

Once again, however, we don't have a great record of fossil sloths this far back in time. Even so, we can place some of these animals within groups known from later times. Orophodon and Octodontotherium, for example, date back to the Late Oligocene, and are related to the mighty Mylodon ground sloths of the Ice Ages (and thus, more distantly, to the two-toed sloths of today). Not only were they much smaller, however, but they also had a different diet. While their later relatives, and even some contemporaries, were primarily grazing animals, these early examples were still unspecialised herbivores, eating leaves, twigs, and the occasional tough seed.

Perhaps the earliest known sloth, however, is the misleadingly named Pseudoglyptodon. Identified from part of the lower jaw, its teeth look remarkably like those of the glyptodonts, but possess key differences that hint at a much closer relationship to sloths than to armadillos. While these primarily relate to details of their internal structure, it's also the case that Pseudoglyptodon had far fewer teeth than any sensible glyptodont - just sixteen in total, with four in each quarter of the jaw. Furthermore, only the six pairs of teeth at the back of the jaw were glyptodont-like, with the four teeth at the front of the mouth having sharpened points that made them look like biting canines. (They're probably modified premolars, though, with the animal lacking incisors and true canines altogether). 

More complete fossils have since been found, allowing us to say that the animal weighed somewhere between 65 and 90 kg (140 to 200 lbs), equivalent to a very large dog, such as a Newfoundland or St Bernard. We don't know much about its lifestyle, beyond the fact that it would have been herbivorous, probably feeding on bushes and other low-growing plants. The current consensus is that it probably isn't a member of the crown group, but represents an earlier branch of the sloth family tree that split off before the two-toed and three-toed lineages separated. Which, by some definitions, makes it the closest known relative of sloths that isn't actually a sloth, but, by others, just makes it an odd early example of its kind.

Other large herbivores of the time are harder to place from a modern perspective. While we now know that the two main groups of large South American ungulates, the notoungulates and the litopterns, are related (albeit distantly) to rhinos and horses, the same is not necessarily true of the astrapotheres. While it does seem the most likely explanation, other relationships are at least possible, and the details can be hard to unravel for such broad groups living so long ago. 

Astrapotheres are better known from the later Miocene epoch, but they already had a long history at the dawn of the Oligocene. The best-known astrapothere from this time period is probably Parastrapotherium, which in many respects resembled its later relative, Astrapotherium. It was a squat, heavily built animal with flat feet, four large tusks, and possibly a tapir-like proboscis. Weight at least one ton, it was one of the largest of its kind, with most other known species being much smaller. Maddenia, for instance, was no larger than a sheep, although it already seems to have had many of the specialisations of the later forms.

Large though they may have been, the astrapotheres were not the largest animals in South America at the time. That honour probably goes to a species of Pyrotherium, an animal the size of a rhinoceros and estimated to have weighed around two tons. This was the largest and most famous of the eponymous pyrotheres, a group of large herbivores whose affinities are at least as debatable as those of the astrapotheres.

The name "Pyrotherium" translates as "fire beast" because the first fossils were discovered in a lair of volcanic ash, and it remains the only pyrothere species known from much more than its teeth. This makes it difficult to tell exactly how they relate to other animals, since many of the features we might want to use might be examples of parallel evolution. The feet, for example, closely resemble those of the mighty arsinotheres of Africa, but that may just be because of their similar size and build; if we had more complete fossils of early pyrotheres that's the sort of thing we could rule out. It's perhaps more likely that they are related to the notoungulates - and perhaps are even an early branch within that group.

Regardless, Pyrotherium would have had some external resemblance to astrapotheres in life. It was heavily built, with short, pillar-like legs and remarkably dense bones, presumably to help carry its great weight, although similar adaptations are seen in some semi-aquatic animals. It had a narrow snout with the nostrils placed unusually far back; again, this can be an adaptation to a semi-aquatic lifestyle, but it's more likely in this case to be due to the presence of a short trunk like that of a tapir. 

It had two pairs of large tusks at the front of the mouth and powerful grinding teeth with a unique structure that meant they did not wear down as much as those of similar mammals. For example, elephants have to replace their teeth periodically through life, dying once they have used up the last pair, while fossils of apparently elderly pyrotheres have been discovered with their teeth still relatively intact. Since pyrothere fossils have been found in habitats thought to have been semi-arid at the time they were alive (which rather counts against any "semi-aquatic" hypothesis) they probably ate tough leaves and branches, possibly tearing them down with their tusks and manipulating with a short trunk, in a manner not unlike that of modern elephants.

But for all that the great herbivorous "southern ungulates" would have been a dominant part of the South American ecosystem during the Oligocene, they were far from the only mammals on the continent. I've already taken a brief look at the rodents, but monkeys have also been on the continent for a long time, while another group of mammals have been there even longer...

[Picture by Nobu Tamura, from Wikimedia Commons.]