Phascolarctos cinereus) is, by many standards, a fairly odd animal. Like most medium to large Australian marsupials, it is herbivorous, but it goes further than that by eating nothing but eucalypt leaves, which are not only low in nutrients, but actually poisonous. Presumably, they do this simply because nothing else does, which means that food should always be available, and they deals with the low nutrient content by spending around two thirds of their life snoozing, and with the poison with unusually efficient detoxifying liver enzymes. Like horses, they are hind-gut fermenters, which also helps them to extract what nutrition they can from food that, frankly, isn't very good.
Koalas have no close living relatives. There is only one living species (with no subspecies), and its considered different enough from everything else to be given its own family - one of a number of mammalian "families" that contain just one species. It's now agreed that the closest relatives koalas do have are the wombats, and even a brief glance at the respective animals tells you that they can't be that close.
So the only opportunity we have for really unravelling the history of koalas comes from examining fossils. For, while koalas are the only member of their family alive today, there were, of course, others in the past. But, still, not terribly many. Some other single-species families, such as that of the pronghorn antelope, represent the last surviving member of a group that was one much larger. Once, there were lots of species of North American antelope, many of them pretty weird to modern eyes, but now only the pronghorn remain, a solitary reminder of a once more diverse group.
Koalas, on the other hand, never seem to have been very numerous. The oldest remains of the living species date back to the beginning of the Ice Ages; while Australia was never covered with huge glaciers, the climate would have changed at around that time, and that may be connected with the emergence of modern koalas. Still, for much of that time, our familiar koala had at least two close relatives in the Australian forests. One of these was the giant koala (Phascolarctos stirtoni), which died out around 50,000 years ago, not long before humans reached the continent. It reached about twice the weight of the living species, which, arguably, isn't all that giant, but is still pretty big for a koala.
Most of the fossil species, however, are much older, dating back to the late Oligocene or early Miocene. At least eight species have been described from that time, although there are probably more. One of the better known is Litokoala from the early Miocene of Queensland. Only around half the size of modern koalas, there are indications from the animal's teeth that, while certainly herbivorous, they ate more than just eucalypt leaves, something that is probably true in general of early koala species.
However, even these older fossils are pretty identifiably koalas and not, for example, wombats. A new analysis by Karen Black and co-workers, of the University of New South Wales, looked at all the known species of fossil koala to try and put them into a clear framework. They also described a new species, Priscakoala lucyturnbullae (the first half of the name means simply "ancient koala") from the early Miocene of Queensland. The new fossil is just a portion of the upper jaw, and some scattered teeth from the lower jaw, but that may be more useful than it sounds, because teeth are often useful features in determining the precise differences between fossil species.
The jaw in question is about the same size as that in living koalas, so, assuming they had roughly the same bodily proportions, the rest of the animal probably was, too. Which makes it much larger than the contemporary Litokoala, and presumably the largest tree-dwelling herbivore in those particular forests. Because koalas have such a strange diet, their cheek teeth are fairly distinctive, but those on this species, while similar, lacked the shearing blades that modern (and most other fossil) koalas develop from a life-time of chewing tough leaves. So, presumably, they ate something softer - most likely still leaves, but perhaps the softer, more nutritious ones of tropical rainforest trees, rather than the tough fibrous leaves of eucalypts.
Indeed, especially given what we know of the climate of the day, it is likely that all early koalas lived in tropical rainforest, and that they switched to eucalypts as the climate changed and Australia became the drier place that (most of) it is today. While they seem to have switched to their new diet early on in the Miocene, at least some rainforest koalas lived on, because we know of one species that survived right through until the Pleistocene.
As for the overall analysis of the koala family, that showed that Priscakoala may be its most primitive known member. However, since it certainly isn't the oldest, that means it must represent an early branch in the koala family tree, one that found a comfortable niche in which to survive, even as its kin went on to explore new habitats. The analysis also looked at the enigmatic genus Koobor.
The name "koobor" is an aboriginal word for "koala", and that is originally what these animals were thought to be. Unfortunately, the key defining traits that would confirm this belong to a part of the skeleton that just isn't included in the few fossil remains we have, and there has been considerable debate about whether Koobor is a koala at all, or a member of some family of koala or wombat-like animals that are now entirely extinct. The new study seems to show that it it, indeed, a koala, albeit a primitive one, but without more complete remains, it's still hard to know how accurate that is.
Part of the problem is that, according to our best guess, the first koalas appeared about 40 million years ago, during the late Eocene. But there are no good fossil beds in Australia from that time period, leaving the earliest history of the koalas - and, for that matter, the wombats - shrouded in mystery. What we do know, however, is that when multiple species of koala did live at the same time they did not, by and large, live in the same place. There are a few exceptions, but, for the most part, any given forest only had one species of koala in it.
The different koalas of the past apparently specialised in particular types of forest, even if their diet was not quite as restricted as their modern day counterparts. As Australia dried out, and the eucalypt forests took over, only the species adapted to that type of woodland survived. Given that perspective, the existence of just one species of koala in the modern world is, perhaps, not so surprising.
[Picture from by "Fir0002" from Wikimedia Commons.]