We spend a lot of time debating about how wolves became dogs. A huge debate exists in the archaeological and paleontological literature about how one can determine whether the remains of a canid represent a wolf, a dog, or a transitional form between wolves and dogs. This debate is why the oldest dog remains are dated to around 14,000 years ago and come from the Bonn-Oberkassel site. Anything older than that, a big debate exists among experts about what can be used to define a wolf, a dog, or a transitional form.
But this debate does not exist solely in relatively recent transition between wolves and dogs. The entire evolution of Canis lupus is a hotly contested and often contradictory, depending upon which source one reads and whether one is looking a source that relies upon paleontological and morphological analysis or one that looks at the molecular evolution of the species.
It is well-accepted in European paleontology that Canis lupus evolved from Canis mosbachensis. Mark Derr paid particular attention to this evolution in his How the Dog Became the Dog. He posits that the extinction of the large hunting dog, Xenocyon lycaonoides, created an ecological niche that could be filled by the Mosbach wolf evolving into the gray wolf.
Yes, the Mosbach wolf was smaller than the modern gray wolf. Individuals from Northwestern Europe were mostly about the size of a modern Indian wolf or a “red wolf.” Indeed, the similarities between some of these mosbachensis wolves and red wolves are the best evidence for a unique red wolf species, for one can argue that red wolves are just a relict form of the Mosbach wolf that held on in Eastern North America. Of course, the genetic data do not agree with this assertion, but it is an interesting idea nonetheless.
My reading is that the Mosbach wolf gave rise to Canis lupus in Eurasia between 300,000 and 500,000 years ago. The coyote, though often posited as a primitive Canis, is actually derived from a divergent form of Canis lupus that got marooned in the American Southwest some 50,000 years ago and evolved to fit a jackal-like niche on a continent already dominated by dire wolves.
The Mosbach wolf disappeared from the fossil record around 300,000 years ago, but there is always a debate as to the possibility that it held on longer. The red wolf and Indian wolf are certainly possibilities for its continued existence today, but as we’ve looked at more wolf genomes both of those don’t come out so distinctive. Every study that I’ve seen that uses Indian wolf genomes finds that they are divergent Canis lupus, and the red wolf is a cross between wolves that are of that coyote type and relict Southeastern gray wolves from a later invasion of the continent. I do think there is pretty good historical data that some smaller wolves that we would define as coyotes lived in the Eastern states at the time of contact, particularly the small brown wolf of Pennsylvania mentioned by Shoemaker and the small “wolues” of Jamestown mentioned by John Smith. My guess is that no one really took stock of what they were killing when they killed off the wolves of Eastern states. It is very possible that coyote-like wolves were killed off in great numbers along with the big ones, and later on, coyotes from the plains came East, crossing with wolves and even relict original Eastern coyotes to form the modern Eastern coyote. The red wolf and the larger Eastern coyote are thus recreations of the Mosbach wolf that have happened in modern times.
In Europe, one potential late surviving Mosbach wolf was thought to have been found in Apulia, Italy, at the Grotta Romanelli site. Wolf remains have been found in the cave that date to between 40,000 and 69,000 years ago and they were often described as belonging to a late surviving Mosbach wolf. A recent morphological analysis revealed that these remains were of a peculiar form of Canis lupus that lived in that part of Southern Italy, and they were not of any kind of Mosbach wolf.
However, the Mosbach wolf is particularly intriguing. Occasionally, it has been posited as a direct ancestor of the domestic dog, but because we don’t have an overlap between the signs of the earliest dog domestication and the existence of Canis mosbachensis in the fossil record, one should be very careful in making such an assertion.
This same caveat should be placed when one sees Canis variabilis posited as dog ancestor. For one thing, there is no such thing as Canis variabilis. Instead, all the specimens listed as this species that come from the Zhoukoudian site in China have now been reassigned to Canis mosbachensis. This reassignment posits them as Canis mosbachensis variabilis, so whenever one encounters that “Canis variabilis” in a paper, just remember that they are discussing a particular East Asian form of the Mosbach wolf.
From my own speculative meta-analysis, it seems that the Mosbach wolf is ancestral to the entire wolf/dog/coyote species complex, which may include the African golden wolf, and the Eurasian golden jackal. A genome comparison study that included dogs, wolves, and one Israeli Eurasian golden jackal found that the divergence between the golden jackal and the dog and wolf species happened just before the anatomically modern Canis lupus replaced Canis mosbachensis in the fossil record. The Eurasian golden jackal could potentially be derived from a diminutive form of Canis mosbachensis that moved toward a more generalist scavenger form.
We also have some evidence of small Mosbach wolves in Europe that could have potentially gone in the direction of the golden jackal. This specimen was found not far from the Grotta Romanelli wolf that were found to be anatomically modern and not Mosbach wolves. It was found at the Contrada Monticelli site in Apulia. It was unusual in that it was quite a bit smaller than the Mosbach wolves found in other parts of Europe, and the authors found that Mosbach wolves were as morphologically variable as modern wolves are.
In North Africa, we also have a recent discovery of a canid that was much like the Mosbach wolf. The authors thought it was a bit different from the Eurasian form, and they decided to call this species Canis othmanii. This African wolf-like canid was found at a site in Tunisia and dates to the Middle Pleistocene, and it could potentially be the basal gray wolf that hybridized with the Ethiopian wolf to make the African golden wolf. More work needs to be done on this specimen, for it very well could wind up like Canis variabilis, a regionally distinct form of the Mosbach wolf.
The really fuzzy part about Canis mosbachensis isn’t that it is the ancestor of the gray wolf. The educated speculations I make about its relationship to the golden jackal and the golden wolf could be debated, and we need lots more data to figure out if I am right or not.
The really fuzzy part is what came before the Mosbach wolf. Most scholars think that Etruscan wolf (Canis etruscus), which makes an appearance in the fossil record around 2 million years ago, is the ancestor of the Mosbach wolf. For years, there was a debate about whether the Mosbach wolf was a chrono-subspecies of the Etruscan wolf or a chrono-subspecies of the gray wolf. All these suggestions would be technically true, simply because we could regard the Etruscan wolf-Mosbach wolf-gray wolf as a species that lasted and evolved over this time period.
However, a bit of a debate now exists as to whether the Etruscan wolf is the ancestor of the Mosbach wolf. An extensive morphological analysis of Etruscan wolf remains and those of another Canis species called Canis arnensis, which compared both to the modern black-backed jackal, the gray wolf, the golden jackal, and the golden wolf, found that our previous delineation between arnensis as being jackal-like and etruscus as being wolf-like were over-simplifications. Some characters of arnensis are much more like modern gray wolves than etruscus is, and it is possible that arnesis gave rise to the Mosbach wolf. Still, the bulk of scholarship thinks that the Etruscan wolf is the ancestor of the Mosbach wolf.
However, because the Mosbach wolf was not included in the analysis, it might be difficult to make such a conclusion. However, maybe the Etruscan wolf or something like it is the ancestor of the Ethiopian wolf. The ancestral Ethiopian wolf must have had an extensive range in Northern Africa for it to have hybridized with Canis mosbachensis, Canis othmanii, or a basal modern gray wolf to form the African golden wolf.
I have focused most of this post on the origins of gray wolves in the Old World, but the first Canis species to evolve were found in North America. Canis lepophagus first appeared in the fossil record 5 million years ago. It was very similar to a coyote or a Canis arnensis of the Old World. This is the part of the story where the molecular data has largely shaken up what we used to believe about coyotes. Lepophagus is thought to have evolved into the larger Edward’s wolf (Canis edwardii), which is sometimes called Canis priscolatrans. These animals might have been the same species or very closely related to the Etruscan wolf. The modern coyote is thought to have derived from edwardii/priscolatrans/estrucus 1 million years ago, but genome-wide comparisons put the existence of most recent common ancestor of gray wolves and coyotes at less than 51,000 years ago.
The dire wolf derived from Armbruster’s wolf (Canis armbrusteri). Armbruster’s wolf derived from Canis edwardii/priscolatrans/etruscus 1.8 million years ago. The dire wolf then evolved from that species 125,000 years ago, which means the dire wolf’s most recent common ancestor with modern wolves and the coyote may have been as far back as 2 or even 3 million years ago.
This analysis is still being worked out. The molecular data is constantly throwing wrenches into the machinery of paleontology, especially the paleontology of canids. The most successful extant canid lineage are full of parallel evolution and phenotypic plasticity, and in this way, it has become quite a challenge to sort out the evolutionary history of these species. At various times, large wolf-like forms have evolved as have smaller coyote or jackal-like forms.
The story of Canis starts with a coyote-like lepophagus, but right now, its likely niche is adopted by the modern coyote, which also very similar to it. But the molecular data suggest that the coyote evolved to adopt this similar niche from a larger Eurasian gray wolf and that it did not directly descend from lepophagus over 5 million years in only North America. Instead, it evolved into wolves that wandered into Eurasia, becoming the Mosbach wolf and then anatomically modern gray wolf. Some of these wolves wandered back into North America and became generalist scavengers in the land of the dire wolf.
Very similar stories likely are lost to us, but we must understand that the history of wolves is not just about getting bigger and developing pack-hunting behavior. That is one part of the story, but another part is about evolving to fit niches, which sometimes means evolving a smaller size and more generalist diet.
Some of my ideas here are very speculative, but I think they are nested in my reading of the available literature. Do not assume that I have the final story of how these creatures evolved, but just understand that the molecular side is so rarely considered in paleontology literature that it is almost like we’re reading evolutionary history of two different lineages.
More work must be done to formulate a synthesis between these two disciplines. Otherwise, there will be continued conflict, and the one using an older methodology and often working with much more incomplete data-set will fall by the wayside. And that is not the one using full genomes.
If we know what problems exist using morphological studies on extant and recently extinct canids, it is very likely that we’re missing important data on many extinct species, one for which there is no DNA to test.
Still, paleontology has much to tell us about the way early wolves lived. It can tell us much about how the ecosystems were and why wolves evolved in the way they did. But its methodologies often miss relationships between extant forms and miss the tendency toward parallel evolution.
I tried for about two years to watch Joe Rogan’s interview with Dan Flores, who wrote a book on coyotes that I think is quite full of misunderstandings about canid taxonomy. When Rogan questioned him about the papers that show a recent origin for the red wolf, Flores pretty much just dismissed those papers because they didn’t look at fossil.
That’s not how it works. Within canids, we know that parallel evolution is a big thing, and it is very possible that coyote-like and red wolf-like canids have evolved more than once on this continent. Indeed, a careful reading of the paleontology and molecular data strongly suggests that this is the case.
In fact, it has always been the case with these wolf-like canids. Big ones evolved from small ones, but sometimes, the big ones become small, because it is a better fit for survival.