Even so, armadillos were more diverse in the Late Miocene than they are today. Probably the most distinctive were the glyptodonts, which had solid, often domed, shells without the flexible bands seen in living armadillos. As the sweltering heat of the Middle Miocene faded towards temperatures closer to those we see today, grasslands began to extend across much of the southern part of the continent, and the glyptodonts adapted to it, switching towards an even more heavily grass-based diet than before. The resulting group of what we might loosely call "advanced glyptodonts" was the one that would later briefly cross over into the north and includes the most famous examples. In addition to certain features of their shell, they were distinguished by having teeth especially suited for chewing, with the high crowns suitable for grinding up grass and other tough vegetation.
This group probably did not arise in the southern Pampas as one might expect. Instead, the oldest known example is Boreostemma, which lived in what is now Venezuela and Colombia around 10 million years ago. This region had previously been isolated from the southern parts of the continent by an extensive series of marshy wetlands, but the development of the Amazon in its modern form during the Late Miocene may have opened up a land corridor, allowing the glyptodonts to move south to the much wider grasslands of Argentina, and develop into later forms, such as Glyptodontidium and Kelenkura, the latter of which is estimated to have weighed around 160 kg (350 lbs).
When they reached the south, however, they would have found other armadillos already there, including some from the wider glyptodont group, such as Asterostemma. Others were more distant relatives.
In many respects, Neoglyptatelus, from Uruguay and Brazil, also looked much like an early glyptodont, although it was closer in size to a large modern armadillo, at around 60 cm (2 feet) in length and 15 kg (33 lbs) in weight. It was originally assumed to be such, a late survivor of the most primitive subgroup. More complete fossils, however, showed a considerable number of differences. For one, the shell did not form a solid shield, but was split into separate plates over the shoulders and pelvis, although there was just a single joint in between them rather than the series of rings that we'd expect in living armadillos. That's weird enough to place it in a distinct branch of the armadillo family tree, but it also turns out that it lacked the chewing teeth of true glyptodonts, and, indeed, didn't have many teeth at all. Unlike them, it probably ate ants or other small insects, rather than vegetation.
The "horned" armadillos, better known from earlier in the epoch, and distinguished by having two of the bony plates on their heads shaped into horn-like cones, barely struggled on into the Late Miocene. The last known example, Epipeltephilus, while it was probably never common, at least lived across a wide stretch of the continent, from Patagonia to northern Chile and Bolivia. Somewhat larger than its earlier relatives, the "horns" would have at least made it distinctive in life.
Other non-glyptodont armadillos were also large. For instance, the pampatheres, which also survived until the Ice Ages, were represented in the Late Miocene by animals such as the Brazilian Vassalia. This was around a metre (3 feet) in length and, with its armoured shield, would likely have weighed up to 90 kg (200 lbs). The shape of its teeth and the strength of its jaws suggest that, while it was smaller than even later pampatheres, it was already a dedicated herbivore, chewing on tough vegetation.
At the opposite end of the dietary scale, Macroeuphractus, although better known from the following, Pliocene, epoch, made its first appearance in the Late Miocene. A giant armadillo much more closely related to the living forms than either pampatheres or glyptodonts, this had a powerful bite and some of its cheek teeth were modified to look like the sharp canine teeth that true armadillos otherwise lack. It's hard to imagine that it didn't at least include a high proportion of meat in its diet.
Part of the reason that this may have been successful is that there was something of a dearth of large mammalian carnivores at the time to act as competitors. With most of the large mammals on the continent during the Miocene having been herbivores, this role was taken by the sparassodonts, and they were in decline towards the end of the epoch, although they didn't die out altogether until about halfway through the Pliocene. They were marsupials, or, more accurately, members of the same branch of the mammalian family tree, since they were not descended from the last common ancestor of the living forms.
Lycopsis is an example of one of the sparrasodonts that only just made it into the Late Miocene, dying out 9 million years ago, not long after the arbitrary Middle/Late dividing line. Known from across the continent, it was similar in size and shape to a fox but has most often been compared to the extinct thylacines ("Tasmanian wolves") of Australia, probably because of its marsupial affinities. The shape of its leg bones suggests that it would not have been good at running and, while it might have been able to climb trees to escape larger predators, it would not have been particularly agile once up there. Instead, it was likely an ambush predator, leaping out on passing prey, which apparently included the 3kg (7 lb) rodent Scleromys, whose remains have been found inside the body cavity of a Lycopsis fossil.
Much better known is Thylacosmilus, a highly specialised predator from what is now Argentina, that managed to survive into the Pliocene and was likely both the last and the largest of the sparassodonts overall. It was about the size of a jaguar and would have had an approximately cat-like appearance, although the head was proportionately larger. Most obviously, it had huge sabre-like teeth that fitted into bony flanges extending downwards from the jaw. Since it was a marsupial-relative and not a real cat, this suggests a strong case of parallel evolution with true cats such as Smilodon.
There are, however, a number of differences between Thylacosmilus and the true sabretooth cats. If anything, it appears to be even more committed to the sabretooth lifestyle than they were, with comparatively weak jaws and powerful neck muscles allowing it to stab down with its canines rather than killing by biting. The claws were not retractile, unlike those of most cats, although it has been argued that the forelimbs were strongly muscled and would have been effective at taking down prey allowing the animal - which was probably not very effective at either chasing or pouncing - to swat at its prey and then overbear it with its weight.
Not all researchers are convinced by the comparison however. A recent paper argued that the shape of the bones and the presumed associated musculature just doesn't fit with what we know of the hunting and killing strategy of Smilodon and its kin. They note, for instance, the very weak incisor teeth, which would have made it difficult to tear into dense flesh and don't find the claws effective as killing tools, either. In this view, Thylacosmilus was unlike any animal alive today (something that's not unusual among fossil creatures) and was instead a scavenger that fed on the soft parts of animals, disembowelling an existing carcass and slurping up the innards with the aid of a long tongue.
Whatever its feeding style, Thylacosmilus had an unusually large brain compared with other members of its wider group, on a par (after adjusting for body size) with some modern marsupials such as kangaroos. It has also been suggested that they continued to care for their young for some time after they were weaned, although this supposition relies on this being necessary in order to teach specialised hunting tactics - something that might be less true if they were simply scavengers.
The Late Miocene also saw the arrival of an entirely new kind of mammal in South America, one never seen on the continent before that time. Cyonasua was a member of the raccoon family and its oldest remains date back about 7 million years, when it lived in Argentina. Quite how it got there, however, is less clear.
That's because raccoons are placental carnivorans and first appeared in North America, not the south. Evidently, Cyonasua, or, more likely some undiscovered ancestor, somehow crossed the Caribbean a few million years before the creation of the Panamanian land bridge, an early herald of the Great American Biotic Interchange to come. That's not so much the puzzle, although it is still impressive, since the continents were approaching one another at the time, and there would have been islands in the gap between them. What's more surprising is that the oldest fossils are from Argentina, and, while we do have some from Venezuela, they are much younger, suggesting that they may have descended from animals living further south. At some point, the ancestors of those animals must have crossed the same area, travelling in the other direction, but if they left any fossils, we have yet to find them.
Cyonasua was widespread, with remains known also known from Uruguay, Bolivia, and Peru, the latter hinting that just maybe their original trek south was down the Pacific coast rather than across the developing Amazon. They were much larger than living raccoons, anything from two to three times the size of the most familiar living species. One might think that their arrival and subsequent success might have had something to do with the decline of the native mammalian carnivores such as Thylacosmilus, but this doesn't seem to be the case.
In fact, they probably didn't compete much at all, since Cyonasua, while probably more carnivorous than living raccoons, was still predominantly an omnivore, with a diet radically different from the sparassodonts of the day. Its limbs suggest that it lived primarily on the ground, although it might have been able to climb trees in a pinch, and may well have been moderately skilled at digging.
Sparassodonts went extinct before any more effective placental carnivores reached South America, and today the largest marsupials on the continent are opossums. Elsewhere in the world, however, marsupials would manage to keep their top position for longer, and it is to that continent that I will turn next...
[Photo by Ryan Somma, from Wikimedia Commons.]