Puijila is not an otter

I wasn't going to comment on Puijila, but I just can't take it any more. For those unaware, a recent Nature article by Rybczynski et al. described the fossil pinniped Puijila, which has four nice legs and is the "least specialized for swimming" among the well-known fossil pinnipeds. (Pinnipeds are seals, sea lions, and walruses.) From an evolutionary perspective, this makes a nice example of what the pinnipeds might have looked like as they evolved from land to sea creatures. Paul Garner thinks it fits in with Kurt Wise's new perspective on mammal baramins, (published in the Genesis Kinds book) that makes mammal baramins much larger than previously suspected. Last year, I found evidence that seals (Phocidae) were a holobaramin and therefore not genealogically related to other pinnipeds (that work is published in Animal and Plant Baramins), and that makes Puijila kind of a puzzle to me. I don't like to comment on things like this until I feel I have something worth saying, especially since Kurt made a good case for his perspective. Obviously one of us isn't right (or both of us, who knows?), but I'm not sure which one (and in case you're curious, I don't really care all that much).

In this case, I feel compelled to comment prematurely because I've noticed that some creationists are sort of writing off Puijila as just an otter. What evidence is there that Puijila is an otter? As far as I can tell, it seems to be the worst kind of (non-)evidence: headlines. Yes, some editors chose to call Puijila "otter-like" or some such (for example: "Otter-like fossil reveals early seal evolution"). They're referring to his superficial appearance. Based on the actual skeletal evidence, there is no doubt that Puijila is a pinniped. How do we know? Consider the following tree, from Figure 4 of the original publication:

Copyright 2009, Nature

I know what you're thinking: "Oh NO! It's an evolutionary tree!" Let's forget about the evolutionary implications of the tree topology and just think of this as nothing more than a clustering diagram (which it is). The question we want to know is how these critters are clustering together? You can see the diagram labels three main clusters: the Musteloidae (weaselly things), the Pinnipedia, and the Ursidae (the bears). Puijila clusters with the other pinnipeds. How does that tell us it's not an otter? Because there's an otter in the tree. It's called Lontra, the one in the Musteloidae with the otter-looking silhouette next to it. Notice how it's on a completely separate cluster from Puijila. The bears cluster closer to Puijila than the otter does! The only way we could justifiably call Puijila an otter is if we called all the Arctoidea otters, including bears, walruses, and wolverines. That would be silly.

Maybe you're still not convinced we should trust "evolutionary" trees. Fine, I examined the same skeletal information using baraminic distances, and I found the same answer: Puijila clusters with the other pinnipeds but not with the otters. Don't believe me? Get the supplemental data from nature.com and do the analysis yourself. Puijila is not an otter.

I could drone on about the specific synapomorphies that group Puijila with the pinnipeds, but I think that's unnecessary. Chances are, you don't care that much anyway, and I think I've made my point.

What's the moral of the story? Headlines are not evidence! Headlines are made up by some editor looking to catch your eye. In this case, I'd be willing to bet that the editors in question never even saw Puijila in person. Real evidence is based on some kind of actual object, observation, or analysis, which the authors of the article did. Editors don't do that sort of thing. It's not their job. They just want your attention.

So please stop saying that Puijila is an otter. Doing so makes you look otterly ridiculous. (That pun might make me look otterly ridiculous, but I'm willing to take that chance.)

Rybczynski et al. 2009. A semi-aquatic Arctic mammalian carnivore from the Miocene epoch and origin of Pinnipedia. Nature 458:1021-1024.