Sunday, 3 November 2013

Spanish Painted Dogs (and What They Can Tell Us)

Members of the dog family can be broadly divided into two groups. One goup are the vulpines, including such familiar animals as the red fox and its Arctic relative. The other are the true canines, a rather more diverse group found wild on every continent bar Antarctica. While species in the first group are all basically fox-like, those in the latter are not necessarily "wolf-like". Indeed, the second group includes the zorros, or South American foxes, which descend from a more wolf-like animal that nipped south when the two American continents collided.

Still, the second group does include animals that are more obviously wolf-like, such as the wolves themselves, along with coyotes, jackals, dingos, and so on. But there are also a number of species that are more distinctive, and while (in most cases) clearly dogs of some kind, aren't that much like our normal ideas of either wolves or foxes.

One such is the African hunting dog (Lycaon pictus), also known as the "painted dog" or "African wild dog", among a number of other variants. These animals are unusual in a number of ways, not least including their piebald colour scheme and their remarkably effective hunting strategies and complex pack life. Unfortunately, those are beyond the scope of today's post, although maybe some day I'll cover the dogs in the same way I have already covered weasels and goats.

When it comes to figuring out quite where hunting dogs came from, however, there are two features that are particularly important. One is the fact that, even compared with other dogs, they have a high-meat diet. Wolves, coyotes, and the like, are obviously carnivorous, but that's not an all-or-nothing affair. Like most land-dwelling carnivores, they will eat plant material from time to time, especially if they're hungry. But, in this respect, African hunting dogs are more carnivorous than their kin - closer, in fact, to cats, which are about as pure a carnivore as one could wish for.

The second feature is that they are good at running, perhaps more so even than wolves. These two features - "hypercarnivory" and running ability - are significant because we can see traces of them in the hunting dog's skeleton. The former, for instance, affects the shape of their teeth, which are particularly well adapted for tearing meat and crushing bone. The second affects the shape of their feet.

Take a look at the front paws of a domestic dog. You'll see that your dog has four toes, each with a pad and a claw. Look a little further back, though, and, to one side, in between the pad on the dog's palm and the one on the wrist, you'll see a fifth claw. This is the dewclaw, and it might not even reach the ground when the dog is standing. In fact, it's your dog's thumb, and it's a feature found in all species of dog... except the African hunting dog.

The dewclaw doesn't do a heck of a lot, although at least some dogs may be able to use it to hold onto bones while they're chewing at them. In the case of the hunting dogs, it has probably disappeared to give them greater speed, in much the same way that most hoofed animals have lost all but one or two of their toes.

Armed with these two facts - oddly shaped teeth and a missing dewclaw - we should be able to trace the ancestry of hunting dogs back through time by examining the skeletons of fossil dogs. Before we do that, though, another approach would be to look at the animals' genetics and biochemistry. We know, after all, that humans are most closely related to chimpanzees largely because we share more of our DNA with them than with anything else. What does the African hunting dog share most of its DNA with?

Unfortunately, we don't have the complete genomes of African hunting dogs, or, indeed, anything much that they might be related to. That doesn't stop us using the genetic data we do have to make a good guess, but, when we try to do so, it turns out that we don't get a consistent picture. Genetics and evolution can be complex things, and they don't always produce neat answers, especially if a number of species appeared at more or less the same time, or one or more of them have evolved at different rates since they parted company.

We can say, from the genetics, that African hunting dogs are definitely canines, not vulpines. Nor are they particularly closely related to South American foxes, which isn't a great surprise. It's also clear that they represent one of the earliest branches in the canine family tree, probably first appearing around the early Pliocene, but, beyond that, a number of possibilities abound. One 2012 study showed that they seem to be closest to animals called bush dogs, which is a bit odd, since they only live in South America and look about as unlike as an African hunting dog as it's possible to imagine a dog could do.

Others have shown that they might be related to dholes (also called "Asian wild dogs") which makes a sort of sense, since these are among the few other dogs to have similarly shaped teeth to the African species. One from 2005 was unable to make firm conclusions, but, as a best guess, figured they represent an entirely separate line related to the common ancestor of wolves, coyotes, jackals, and dholes.

So much for that. What do the fossils say? Skeletons belonging to the modern species have been found as far back as the late Pleistocene, and as far north as Israel - quite a way beyond their current range in central and southern Africa. There are also fossils of animals closely resembling hunting dogs from the early Pleistocene of Eurasia, although relatively few from the time in between. Quite what this means depends on how much like hunting dogs you think those early fossils are.

One theory places these early fossils in the genus Xenocyon and suggests that they might represent the common ancestor of both African hunting dogs and dholes. The first definite dhole fossils date from the middle Pleistocene of Italy, so that fits, but there is another possibility, one that suggests these fossils represent hunting dog chronospecies.

Chronospecies are members of a single lineage of animals, where one evolves directly into another, and so on. They are distinguished, in other words, only by time and by gradual evolution, not by one species having branched into multiple others. They are, in effect, an actual view where we can see one species slowly changing into another.

According to this view, there are at least three species of hunting dog. The first is Lycaon falconeri, which lived in Europe at the very beginning of the Pleistocene - before the Ice Ages - and probably back into the tail end of the preceding epoch, the Pliocene. This evolved into Lycaon lycaonoides, which, in turn, evolved into the modern species. Since the theory was first proposed, a possible fourth species, Lycaon sekowei, has been discovered in southern Africa, perhaps representing an early southern emigration of hunting dogs from their original home in Europe, or perhaps being the original ancestor, rather than L. falconeri.

An interesting feature of the African fossil, wherever it fits, is that it has dewclaws. In other respects, including the teeth, it looks quite like living hunting dogs, which suggests that "hypercarnivory" evolved first, and rapid running later. If that's the case, and if these really are chronospecies, then the species in between - L. lycaonoides - ought to be intermediate in appearance as well.

And that is where new Spanish fossils of L. lycaonoides come in. They are just 0.8 million years old, the youngest fossils of their species known, dating from what must be just before the last hunting dogs went extinct in Europe. What we'd really like to see are their front feet: if these animals had no dewclaw, or an extremely small one, it's hard to believe that they could be the ancestors of dholes as well as hunting dogs (and thus, should really be called Xenocyon lycaonoides). Unfortunately, all we really have is the lower jaw.

But even that is suggestive. Some people think of species as a fixed thing, that doesn't really change over time, at least until it changes into something else. But evolution is, much of the time, a gradual process, and we do often see changes as a chronospecies slowly morphs into whatever replaces it. That we don't see this more often is at least partly due to the fragmentary nature of the fossil record, where we just see isolated snapshots in time. But, particularly with more recent fossils, sometimes we can see a wider picture, amost watching evolution as it happens.

According to the researchers who discovered the new fossils, that's what's happening here. These, the very latest members of their species known look, they say, more like living hunting dogs than do older fossils of the same species. They represent one step in that slow change from the extinct forms we already knew about to the ones that are alive today.

To fill in the gap completely, we'd need to know what was going on the mid Pleistocene, between 0.8 million years ago, when these dogs died, and 0.2 million years ago, when the earliest fossils of the living species appear. But this finding at least supports the chronospecies idea, and may give us some idea of what "painted dogs" looked like when they lived in Europe, and not just in Africa.

[Photo by Greg Hume, from Wikimedia Commons].


  1. Of course, the picture of the evolution *Lycaon*, *Cuon* and *Canis* is more complex than you describe here. There's these factors to consider:
    - Weird *Cynotherium* of Sardinia was likely close to *Lycaon* and a *Xenocyon* descendant itself.
    - There are the very obscure fossil dogs *Megacyon* and *Mekecyon* that are considered relatives or descendants of *Xenocyon* as well.
    - According to some studies, *Lycaon* and *Cuon* fall within traditional *Canis*, meaning all the aforementioned genera should be merged into *Canis* (I've seen them placed there in old sources, actually) or the Blackbacked and Striped jackal should be split off into their own genera.
    -*Canis* itself may be a *Xenocyon*-derivative, if the aforementioned studies are correct.
    - *Xenocyon* itself is clearly non-monophyletic if several genera are considered to have arisen from it. Then, in the study of fossil mammals there seems to be a far greater tolerance of such rampant paraphyly compared to, say, among dinosaur experts.

  2. Indeed; as I say, the idea that these are chronospecies is merely one theory. If, as the alternative suggests, the early species are *Xenocyon*, then the latter would be the ancestor of *Cuon* (that is, the dholes) at the very least, and possible *Canis* (wolves/coyotes/jackals) as well. This new fossil supports the chronospecies idea, but, as I think I also said, it's far from proof. Especially without the feet!

  3. The division into "vulpines" and "true canines" may be an oversimplification. There was an article in "American Scientist" a few years back (early 2000s?) reviewing the phylogenetics of the Canidae: as I recall, there were about five main "branches" of the family tree, with animals of fox-like morphology (often with "fox" as part of their common names) on at least four of them, and big dog/wolf critters only on one: so maybe "foxes," if you think of them as a group, are paraphyletic w.r.t. "true canines." And a dyed-in-the-wool cladist would say "there is no such thing as a fox"! The "Gray Fox" (Urocyon) was, if I recall correctly, on the earliest-diverging branch: equally distantly related to both "true canines" and to the "Red Fox" (Vulpes): no natural group (clade) unites the Gray and Red Foxes to the exclusion of "true canines." (B.t.w.: I REALLY, REALLY enjoy your blog: it's one of the things I have bookmarked for frequent visits!)

    1. For these purposes, I'm including the South American foxes (*Cerdocyon*, *Pseudalopex*, that sort of thing) as "true canines", in the sense that they're usually regarded as members of the Caninae, although not the Canini. By "vulpines" I'm really only referring to *Vulpes* itself, and probably *Otocyon*. As you say, the position of *Urocyon* is debatable - the 2005 study I linked to makes it a vulpine, the 2012 one has it (as you describe) as neither. But, yes, "fox" is not a thing :)