Two of these South American groups would outlast the Miocene, only dying out relatively recently. These were the litopterns and the notoungulates, both of which bore most of their weight on the third toe of each foot, and which may, indeed, be related to the group containing the tapirs, horses, and rhinos, which does the same today.
At least, that was the general rule. One family within the notoungulates, known as the interatheres broke what otherwise seems to be a universal rule with the group and bore its weight equally on two toes, in the same manner that the (more distantly related) cloven-footed mammals do today. Miocochilius, known from Colombia to Bolivia and Peru, was an extreme example. Notoungulates typically had three or four toes on each foot, but Miocochilius had only two toes on its hind feet, and the third toe on the fore-feet was greatly reduced, perhaps used to steady itself, but probably not for walking on.
While this cloven-footed pattern does not in itself make the animal better at running (after all, horses don't work this way) the reduction of the other toes probably means that it was quite fast. Like all interatheres, and most of the later notoungulates in general, it had high-crowned cheek teeth and tough enamel on its incisors, suggesting an abrasive grazing diet that horses might also have approved of. It was, however, a good deal smaller than a modern horse, likely standing a little over 30 cm (one foot) tall at the shoulder.
This was, however, fairly large by the standards of interatheres, which were typically much smaller than most ungulates today and may have filled a similar niche to rabbits. Interatherium, for which the group is named, was a more typical size, perhaps weighing only around 2 kg (5 lbs). It lived further south than Miocochilius, in Patagonia but, apart from its size, and despite being the type genus of its family, it really didn't look much like other interatheres. That's largely due to its very short legs and short snout, which may have made it look somewhat like an otter when it was alive. It was another grazing animal, so why it would have wanted such an unusual shape for a grazer is unclear; it may have spent a lot of time in the water, or perhaps it squeezed into narrow burrows or... well, who knows?
These were not the only small notoungulates living in South America at the time. The hegetotheres, a long-lived group that would survive in some form into the Ice Ages, were also present, and likely in considerable numbers. A Middle Miocene example here is Hemihegetotherium, one of the larger members of its group - most were rabbit-sized, but it was about 80 cm (2' 8") in length. This was originally described from Bolivia, although recently a second species was discovered in Argentina. The bones of its lower hind limbs were solidly fused together, something we'd normally see in a digging animal. Since the limbs aren't otherwise the sort of shape we'd expect of a running animal, it's possible that it did a bit of both.
Altitypotherium, belonging to the related 'mesothere' group, was probably better at digging. The structure of its inner ear suggests that it would have been more agile than we'd expect for a burrowing animal, so perhaps it had a more generalised lifestyle, sometimes scratching in the ground to dig up roots, but more often feeding on low-lying vegetation. It may, perhaps, have been similar to living wombats, and was around the same size.
Near the opposite end of the scale, toxodonts such as Nesodon were among the largest notoungulates of the Early to Mid Miocene. Thought to have weighed around 550 kg (1,200 lbs) this somewhat resembled a small, hornless, rhino with a powerful body, short limbs and a comparatively heavy head. Like the smaller notoungulates mentioned above, it had high crowned teeth that seemed to imply a tough, grazing diet, along with powerful, almost rabbit-like incisors. However, analysis of the micro-wear on those teeth showed that they were not much used for this sort of food, and the animal was likely a browser or mixed feeder, eating softer leaves, albeit perhaps supplemented with something tougher such as tree bark.
Large though this was, and while other similarly-sized browsing animals such as Huilatherium also existed in the Early to Mid Miocene, they were not the largest South American mammals of their day. That honour likely goes to the astrapotheres, huge hoofed herbivores whose relationship to the notoungulates remains unclear - they are certainly a different group, but just how far apart they are in evolutionary terms is difficult to quantify.
The Middle Miocene, however, was the last gasp for the astrapotheres, whose survival into the Late Miocene is doubtful, and which certainly didn't live beyond that. They had been a more significant group in the previous epoch, and only a few survived - although these do include some unusually large-sized species.
The largest Miocene example, and perhaps the last overall, was Granastrapotherium, which some estimates put as weighing over two tons, roughly equivalent to a modern rhino. However, they would have resembled elephants more than rhinos, despite the shorter limbs and more rhino-like body. This is largely because they appear to have had a distinct and flexible trunk, so far as we can tell from the position of the bony nostrils, which are very high up, between the eyes. Whether it was as long as that of an elephant, or more like of that a tapir, is, however, harder to say.
Not only that, but it also had four long tusks, formed from the canine teeth. This combination of features, together with the complete lack of incisor teeth (there are normally at least some in astrapotheres) implies that they probably fed like elephants, too. Their diet likely consisted mostly of soft leaves, perhaps near rivers and lakes, since their bones are often found alongside those of fish, implying that they may have fallen in when they died.
It's likely that few of the local carnivores would have preyed on something the size of Granastrapotherium, at least once it approached adulthood, but lesser creatures were not so lucky. Next time I will look at some of the carnivroes living in South America at the time.
[Photo by "ghedoghedo", from Wikimedia Commons.]