The Dire Wolf was a wolf in parallel

 

Mauricio Anton's depiction of gray wolves fighting dire wolves over a bison kill. This reconstruction is based upon the new dire wolf genome paper that reveals that many of North America's wolf-like canids were not particularly closely related gray wolves and coyotes.

Dire wolves have captured the public imagination for some time now. Game of Thrones fans know them well, and our popular understanding of them is that they were like very big gray wolves.  Game of Thrones embellished their size, but they were larger on average than a typical Northwestern gray wolf, the subspecies that ranges through the Pacific Northwest, most of Alaska, and much of the Northern Rockies.

My understanding was that a dire wolf was very much like a very big gray wolf that had special adaptations to hunting megafauna. It was very similar to the Beringian gray wolf, an extinct ecomorph of gray wolf that ranged from Alaska to Wyoming  It, too, was adapted for hunting very large prey.

Paleontologist had developed phylogeny of the dire wolf based upon comparative morphology.  Wang and Tedford contended that the dire wolf derived from a large Pliocene Chinese "wolf" (Canis chihliensis). Their work contends that this Chinese canid was the common ancestor of the gray wolf and dire wolf, a claim for which I have always been skeptical. European paleontologists have shown rather convincingly that ancestor of the gray wolf was the Mosbach wolf.

What I was always hoping for was that someone would be able to extract DNA from dire wolves and maybe answer my skepticism.

Even with my general view that the Mosbach wolf was the ancestor of the gray wolf, I still accepted that there was a North American wolf radiation that paralleled the evolution of the gray wolf in Eurasia.

But every time I would look more closely at the paleontology literature, I would be more confused. For example, one species posited in the dire wolf's lineage was Edward's wolf (Canis edwardii). This animal has been suggested as an ancestor the putative "red wolf," and Dan Flores made a big deal out it being the ancestor of the coyote in Coyote America, even though it was closer to the red wolf in size. This literature gets somewhat confusing because Edward's wolf is sometimes listed as Canis priscolatrans, but they are now generally regarded as the same species.

None of this makes much sense, especially with the new literature clearly showing a much closer genetic relationship between the coyote and gray wolf. If this closer relationship is true, the coyote is not an ancient New World canid. If that more recent origins date is accurate, the coyote is actually a form of gray wolf in much the same way that the domestic dog is.

So I have many questions about North America's wolf-like canids. I have found the paleontology literature not particularly enlightening, because the genomic data almost entirely contradicts its theses.

But a few days ago, a paper was published in Nature that vindicated my skepticism.  Genomes were sequenced from the subfossil remains of five dire wolves. They then compared that data to extant wolf-like canids using a Bayesian analysis called an MCMCtree. 

The authors discovered that dire wolves were not closely related to extant wolves or coyotes. Instead, they diverged 5.7 million years ago  They turn out to be most closely related to the endemic African jackals, now often classified in the genus Lupullela rather than Canis. 

This study has profound implications about wolf evolution. It clearly shows that coyotes originated from Old World ancestors, and they are not direct descendants of Canis lepophagus, the first Canis species that appeared about 5 million years ago in North America. It is also is also not descended from Canis edwardii/Canis priscolatrans.

 

This study suggests that a whole lineage of parallel wolves existed in North America that were quite distinct from modern ones. Canis edwardii might belong to this lineage, and Canis armbrusteri, the ancestor the dire wolf, most definitely does. 

In the same way that many experts are now using Lupullela as the genus for the endemic African jackals, the researchers associated with this paper have suggested that the dire wolf be called Aenocyon dirus, the "terrible wolf." Armbruster's wolf should also be reclassified in this genus, and it is possible that Edward's wolf will be too.

Further, if these dates of divergence hold, the common ancestor of the dire wolf and North American endemic wolf lineages would be somewhere in the genus Eucyon, which existed very near the divergence between wolf-like canids and the "South American clade" of canids, the Cerdocyoninae. So this famous reconstruction depicting the dire wolf as a robust bush dog might not be as far-fetched.

Two years ago, I wrote that some of our assumptions about dire wolves might be wrong, and I am honestly surprised that this layman produced some hypotheses that were later shown to be correct. Full disclosure: I never spoke a word to any of these researchers, nor did they contact me. I am not such a narcissist to assume they ever read what I write. I am not a scientist, but I am one of their stenographers and exponents.

I even introduced this whole idea through an April Fool's prank. I had no idea that my prank would actually turn out to be a major finding in a paper almost four years later. 

So, yeah, there is that.

I should note that the authors used a Bayesian statistical approach for the dire wolf genome study, which is a methodology that includes probability models that are based upon previous research, The coyote-gray wolf genome paper that I often refer to used a frequentist methodology, which does not set up such parameters. This is why the dire wolf paper shows a 1.1 million year divergence between the coyote and gray wolf,, and the other paper shows a 51,000 year divergence. I would like to see this paper recalibrated using either a frequentist analysis or a Bayesian analysis that includes this as one of the parameters. The dates will be quite a bit different, but they still show this extinct lineage of North American wolves.

The exact phylogeny and taxonomy of wolf-like canids is not entirely clear, but the relationships are coming more into focus. This paper is as much a revelation about the evolution of these creatures as the paper discovering more recent origins between the coyote and gray wolf.

The researchers of this dire wolf paper discovered no evidence of gene flow between the dire wolf and the gray wolf and coyote, which both apparently colonized North America from Eurasia.  It may have lacked the behavioral and morphological flexibility that allows gray wolves and coyotes to thrive in constantly changing environments. Further, it might have susceptible to diseases that proliferated from growing populations of gray wolves and coyotes.

Popular press coverage of the new placement of the dire wolf within the wolf evolution story often points out how surprised the researchers were by the discovery. One researcher even thought the dire wolf would come up as a subspecies of the gray wolf, a possibility that certainly made some sense. 

Because canids have a tendency toward morphological plasticity and convergent evolution, our ability to determine the taxonomy and phylogeny of these species based upon morphology will always be limited. This is why the paleontology got so much of this wrong.

Sometimes what you think you see really isn't there at the molecular level.