monotremes. Well over 90% of identified living species fall into the first of those groups, which prompts the interesting question of how we divide those up - what, in short, are the next branches on the tree?
In the strict, old-style, system of classification, the next major level down is that of the "order", groups of broadly similar creatures that are in turn, divided into more than one hundred "families". Primates, for instance, are an order.
What the orders are, and how many of them there might be, has varied a little down the years, but nineteen or twenty would be a good, modern estimate. But, of course, the orders themselves are entirely arbitrary, and, even if they weren't, it isn't as if they all sprang simultaneously from the same stock. In evolutionary terms, there must have been some early branches in the placental family tree from which the "orders" that we know today descended.
The question of how to group the orders together into some higher level pattern is one that mammalogists have been been pondering for well over a hundred years. The problem is that the orders are pretty distinctive - is a moose more like a wolf than it is a monkey? When all you have to go in is physical anatomy, these are not the simplest questions to answer, and it's not surprising that people have come up with different answers.
Today, however, we have large quantities of genetic and biochemical data to add to the anatomical, and the ability to crunch vast arrays of possible similarities on powerful computers. Since roughly the end of the twentieth century, multiple different lines of evidence have allowed us to come to a reasonably clear consensus, and group all placental mammals into four main evolutionary branches. And they aren't necessarily the ones we would have expected.
The most diverse branch includes such varied creatures as horses, tigers, dolphins, and bats, and goes to show just how broad the evolutionary groups we're talking about are at this level. (And, yes, this means that a moose is, indeed, more closely related to a wolf than it is to a monkey. But it's even closer to a sperm whale...) The branch with the most species, on the other hand, is the one that includes the primates and rodents, along with a few of their relatives. The remaining two branches have only a handful of species each. One includes such things as elephants and manatees, and the other is the xenarthrans.
Which are weird.
Sufficiently weird, in fact, that, unlike the other three branches, we already knew that this one existed - even if we we had changed its exact membership down the years. Indeed, the term "xenarthran" was originally coined all the way back in 1889, by palaeontologist Edward Drinker Cope. Just a couple of years earlier, the zoologist Oldfield Thomas had proposed that the group (although not yet under that name) was so odd that they ought to be listed as a branch equal in status to the marsupials - in other words, as descendants of an entirely separate evolutionary path than that which led to all other placentals.
Today, of course, we know that these creatures are, indeed, true placentals, just as their reproductive biology would imply. However, there are at least some grounds to suppose that the xenarthrans actually represent the oldest of the four placental branches. This had long been suspected, but it's far from proven, and we still aren't entirely clear in what order the four branches divided from one another or quite how they interrelate.
So what are xenarthrans, and what's so odd about them?
The word "xenarthran" means "strange joints", and refers to the odd way that their backbone is put together - different from that in any other placental mammal. That's by no means the end of their oddities, though, since they also have an unusually low body temperature, and the males have no scrotum. (Their testicles are instead lodged between their bladder and their rectum, which would be a rather bad idea in most other placentals). There are only a little over 30 species of xenarthran alive today, almost all of them native to South America, and they are usually grouped into three living families: the armadillos, the sloths, and the anteaters.
Tracing the ancestry and evolution of the xenarthrans is not always easy. When, as so often in real life, we don't have a complete fossil skeleton, one of the key things we turn to in mammals is the exact shape and number of the teeth. One key indicator, for instance, is the dental formula, which represents the number of incisors, canines, premolars, and molars in each jaw. But no known xenarthrans have any meaningful dental formula, those that have cheek teeth don't have ones that look anything like those of other placentals, and only two known species even have tooth enamel as adults. Both of which are extinct.
The reason that this is important is that a large part of the reason we use teeth for this sort of thing is that tooth enamel preserves at least as well, and probably better than, bone in the fossil record. Thus, even if the rest of the skeleton is degraded away over millions of years, there's a pretty good chance that the tooth enamel will survive. Which is a bit of a bugger if the animal in question doesn't have any.
On the plus side, this peculiarity at least means that it's possible to identify that any fossil tooth you do manage to find at least belongs to an armadillo or sloth, even if it does make it harder to figure out how they're related to anything else. They are not, of course, unique in having these problems, since living dolphins and whales (among others) are also fairly odd in this particular respect... but, in their case, their ancestors were more normal, and those of xenarthrans really don't seem to be.
And, of course, even this much doesn't work with anteaters, because they haven't got any teeth at all.
The ultimate origins of the xenarthrans remain, therefore, shrouded in mystery. Even if we did have a fossil of one of their immediate ancestors, it might be rather hard to figure that that's what it was. Nonetheless, we can trace their history back at least some of the way to their origins in South America, since, for all the limited number of species alive today, there were rather a lot in the past.
The very oldest xenarthran fossils known belong to armadillos that lived over 50 million years ago. Unfortunately, they are just fragments of the armoured shell, the sort of thing that you can certainly tell belonged to an armadillo, but that tells you pretty much nothing else about the animal, let alone what its un-armoured ancestors might have been like. And there's a good chance that it's still 10 to 15 million years after the origin of the group, if not more. Plus, you'll note, it is an armadillo, meaning that the ancestors of the other two families had presumably already split off.
Slightly later than this we find the fossils of those two species that still had tooth enamel. Both are also armadillos, indicating that that family lost their enamel rather later than this, and, perhaps more interestingly, that sloths must have lost their enamel in an entirely separate evolutionary event, since the two had long since parted company by this time. The best armadillo fossils get progressively more complete after this, but even so, we don't get really good ones until around 20 million years ago.
The oldest sloth fossils are a little under 40 million years old, and from much further south than the earliest armadillos. They already seem somewhat specialised, and, by around 30 million years ago, we see rather more varied sloth fossils, including some related to the great ground sloths of later times. Finally, the anteater fossil record is particularly poor, dating back, at the very most 20 million years, and quite possibly rather less, and including only a handful of species.
So can we tell anything at all about the doubtless odd little beasts that the first xenarthrans evolved from? With no decent fossils from sufficiently far back, we have to fall back on other methods, and one of these is phylogenetic bracketing. The principle behind this technique is actually fairly straightforward. Take a look at the living creatures most closely related to the fossil you're interested in, including some that diverged before it evolved, and some that came after. Anything they all have in common is probably true of your fossil as well.
The reality is a little more complicated, and one has to bear in mind that what we're dealing with here is essentially educated guesswork... only with the word 'phylogenetic' in it to make it sound more scientific. There are a number of reasons why it might not work in any given situation, but it's actually not too bad as a starting point and is likely right more often than not.
A new study applies phylogenetic bracketing to a number of xenarthran species, both living and extinct, in an attempt to figure out what their original ancestor may have been like. To start with, what habitat did these ancestral xenarthrans live in? The answer: no clue. There's just too much variation among all the known forms to be able to guess.
Not a good start, then, but what about how they moved about? This also looks like it's going to be difficult to answer if we look only at living species, but once you add in the fossil forms, the researchers reckon that a pattern begins to appear. As they applied the bracket to ever wider groups within the Xenarthra, they found again and again that the ancestral forms seemed to be good at digging, with the only major exception being the giant ground sloths. Since those appear relatively late, the researchers conclude that the very first sloths were likely digging animals, just as anteaters and most armadillos are today. And, by inference, so was the very first xenarthran.
Similarly, both anteaters and sloths appear to be descended from animals that were good at climbing - although armadillos clearly aren't, which leaves the very base of the family tree unresolved in that respect. When it comes to diet, all known sloths, living or otherwise, are herbivores, and all anteaters... eat ants. But armadillos are more varied, which at first doesn't seem to help. In fact, the analysis ends up with two equally likely answers for the diet of first xenarthran: it was either herbivorous or ant-eating, with the omnivorous armadillos having apparently arisen later.
Here, however, the researchers come back once again to those peculiar teeth. Even today, anteaters are not the only mammals that live on a diet consisting almost entirely of ants or termites. And when we look at the other ant-eating mammals, we find that all of them have reduced, or even absent, teeth. So, the authors say, that would fit: the first xenarthrans ate ants, developed reduced teeth as a result, losing such things as their enamel, and only later evolved the alternative lifestyles of living sloths and armadillos.
So putting this together, and remembering once again that this is just inference, not proof, what have we got? An animal that eats ants, is good at digging, and possibly likes climbing trees. The first two make sense, since you have to get the ants out of the anthill somehow. The latter two might seem a stranger combination, but in fact there are animals that do both of these things today. One of them, the tamandua (a kind of anteater), is even a xenarthran.
We may never find this animal to determine whether this guess is right or not. But it's a plausible explanation, although, personally I'd suspect that omnivory or a more generalised insectivory are viable alternatives for the diet. Until we find better fossils, though, we're never going to know...
[Photo by Christian Mehlführer, from Wikimedia Commons. Cladogram adapted from Gaudin & Croft, 2015.]