A Lesson in Taxonomy Starring Puijila darwini

Phylogenetics is kind of like the mythical hydra: for every gap closed by a new fossil, three new gaps appear. This makes paleontology very exciting, but it also makes the science easy to pick on by Creationists and other science-deniers. Why are three gaps opened instead of just two? Well, because we weren't around to see one species branch directly from another, taxonomy assumes that all common ancestors are unknown. All you can really do is say that Velociraptor and Deinonychus share certain features inhereted from a common ancestor. Even if Tyrannosaurus evolved directly from Daspletosaurus, it's just not something we can ever know. So because every animal in the fossil record (and in the modern world) exists on its own "sidebranch" to the main line of faunal evolution, then three gaps appear:

1) The gap between the new taxa and the common ancestor;
2) The gap between the common ancestor and its the next rung down;
3) The gap between the common ancestor and the next rung up.

Here's a simplified Dinosauria cladogram illustrating this point:

When I say "shared characteristics," I mean things like number of fingers or length of cervical ribs, things like that. Some of these relationships are based on very technical measurements and features. At any rate, the tree is designed to show that Sauropodomorpha is closer to Theropoda than either is to Ornithischia. That doesn't mean thatt Dinosauria is polyphyletic, it just means that Saurischia is a more exclusive group than Dinosauria. You could say that the features diagnostic of "Dinosauria" are entirely arbitrary, and you'd be right. The whole point is to show relationships, though. If we decide one day that silesaurids should be dinosaurs, the label "Dinosauria" would be moved one rung down to include Silesauridae, which is currently considered more or less an outgroup to the Dinosauria proper.

So what happens when a new fossil fills a critical gap in an otherwise poorly-known transitional series? Let's use dinosaurs and birds as an example. The relationship was realized back in Sir Richard Owen's time. In fact, he favorably compared the femur of Megalosaurus to that of an ostrich. Anyway, watch what happens when we toss Archaeopteryx between Allosaurus and Gallus:

Because Archaeopteryx shows features unique to it and exclusive to both birds and Allosaurus, it cannot be suitable common ancestor between the two outgroups. However, because it has more features in common with Gallus than Allosaurus, it must be closer to the chicken. But because it has its own unique suite of characters, it is not a suitable ancestor. So it split off from a theoretical ancestor between itself and Gallus. Now, we'll probably never find an actual ancestor-decendant relationship in the fossil record, but we can come darn close and make predications as to what features that ancestor should have. Eoraptor, for instance, is the basalmost saurischian dinosaur known. It's got a few unique characters (apomorphies), but on the whole it's a fine model for the common ancestor between Sauropodomorpha and Theropoda. That is, Eoraptor has lots of features common to both groups.

Here's a simplified theropod cladogram using manual digits as an example of this principle:

So when did theropods lose that fourth digit? More fossils are necessary for that answer. You'd have to find a theropod that, morphologically, falls between Herrerasaurus and Allosaurus (like Carnotaurus) and count its fingers (it has four). Then go farther up the tree to, say, Spinosaurus (who is between Carnotaurus and Allosaurus) and you see that has three fingers. So you've further narrowed the taxa gap for that finger to be lost. Likewise, when did tyrannosaurs lose their third finger? Basal tyrannosaurs like Guanlong and Eotyrannus have three fingers, but later ones like Gorgosaurus and Tyrannosaurus have two. Someday we'll find a tyrannosaur with two full fingers and the third finger has been reduced to a splint of bone (as the fourth one is in Herrerasaurus).

The same process has recently been done with whales. Whales used to be a total mammalian mystery. Aside from the fact that they gave birth to live young and nursed their babies, modern whales were far too derived to pinpoint a fossil ancestor. Because of their bizarre skeletons, shaped by a millenia of marine adaptation, comparisons to other modern mammal groups was virtually useless. The cladogram basically looked like this:

But then, paleontologists discovered Ambulocetus natans, and the whole cladogram changed. It just takes one fossil to demonstrate relationships between any two groups.

Aside from being a vicious shallow-water predator with whale-sized dentition, Ambulocetus had a unique ankle structure shared by it and the Artiodactyla, an enormous group that also includes goats, cows, hippos, pigs, and antelope (and more). But because Ambulocetus also had a very unique inner ear shared only by whales, scientists knew that it was a ridiculously basal whale. More finds would come later further solidifying the artiodactyl link (Rodhocetus) and show that early whales hauled themselves onto land to give birth (Maiacetus). The big mystery now is exactly how far back toward the origin of the Artiodactyla whales go, and also when they lost their hindlimbs (are there fossils that document this? Anybody?).

So now the whale cladogram is a lot clearer:

So with all that information in mind, I present to you Puijila darwini, a new basal pinniped that sheds light on the transition from mustalid-like carnivore ancestor to flipper-finned seal. Pinnipeds are a group of extremely diverse and successful mammalian carnivores. They include seals (eared and otherwise), sea lions, and walruses. Their closest non-marine relatives are bears and mustalids (weasels, otters). The ancestor of modern seals has been very difficult to nail down, though. The oldest known seal is Enaliarctos, which already has well-developed flippers. So there exists a big morphological gap between mustalids and seals. That gap has now been partially...um...sealed!

It's less than a meter in length and was found on Devon Island, one of those horribly cold islands above Canada near Greenland. Puijila has a seal's head, including the osteological correlates indicating large, sensative whisker pads. However, its body and feet are more like an otter in that it has, you know, feet instead of flippers. The fingers and toes are flattened, though, indicating extensive webbing (same correlates are in beavers, otters, and Casterocauda). And check out that bacculum! I would like to point out, however, that the feet and hands of Puijila are longer than any otter or beaver.

Look at that muzzle--I would not want to come toe-to-snout with Puijila, though it's small enough you could potentially kick a field goal with it. At any rate, a few more interesting points about Puijila. First, it was discovered well within the arctic circle, lending some credence to the notion that pinnipeds originated in the colder northern waters of the arctic. What's more, its swimming style was more toward pinnipeds than mustalids: quadrupedal "doggie paddling" rather than pelvic paddling. Puijila's long tail probably did very little in the water, and the limb-centric swimming style of Puijila could have easily developed into pinniped swimming, which is similarly dominated by the limbs. Finally, Puijila was not an oceanic swimmer. It preferred freshwater lakes and streams, which indicates that pinnipeds went through a freshwater "phase" before braving the deep blue sea. This is to be expected given other marine mammal's ancestral records (Ambulocetus, Pezosiren).

So what's the moral of the story? Puijila represents a good model for the common ancestor between it and modern pinnipeds, both morphologically and ecologically. It helps to close the distance between mustalids (and bears) and walruses, and future fossil discoveries may show other transitions, such as the loss of caudal vertebrae, the perpetual retraction of the hindlimbs, or the enlargement of manual digit I. So think of Puijila as a seal-headed, long-footed otter, and a virtually perfect representative of common ancestry. Feel free to put together your own cladogram!Oh, and here's a picture I whipped together from the description.

Tune in next time for: "The Headache Of Convergence," or, "Ornithischians: a Poor Choice!"