The dire wolf is still extinct.
Last edited Thu Apr 10, 2025, 05:25 PM - Edit history (4)
There have been headlines – and even the cover of Time Magazine – touting the alleged “de-extinction” of the dire wolf, Aenocyon dirus, based on genetic engineering.
Please let me explain why this is bullshit, and why the headlines should be (mostly) disregarded.
Full disclosure – I’m a professional vertebrate paleontologist who also uses DNA (including DNA from fossils). I don’t work on mammals; rather, I work on animals that eat mammals. And I use the DNA to reconstruct evolutionary relationships - not for resurrection. Still, I have a decent idea what did (and did not) happen in this case.
The dire wolf (which used to be recognized as Canis dirus until analyses of ancient DNA showed it to be more closely related to other canids) is known from Pleistocene (“Ice Age”) deposits across North and South America. It was long viewed as basically a much larger relative of the modern gray wolf (Canis lupus), but it’s not actually a member of Canis, and the largest subspecies of C. lupus are about the same size as A. dirus. Still, it would probably have looked more or less like a wolf - a big dog, at any rate - and would have been an impressive predator.
It's best sampled, I think, from the La Brea Tar pits. There’s an entire wall at the George Page Museum in Los Angeles (which is one of the best paleo museums in North America) dedicated to dire wolf skulls. Animals would be stuck in the tar, drawing predators and scavengers that, in turn, would also get stuck in the tar. (The image many have of deep pits of tar is mythical; in most cases, it would be a thin, but extremely sticky, film of tar at the surface, or maybe on the bottom of a shallow lake. Animals stuck in the tar would actually remain mostly exposed at the surface while they were scavenged; they wouldn’t “sink into the tar.” But I digress.)
Anyway – several companies are trying to “resurrect” recently extinct animals. This is very controversial among scientists; many of us (myself included) think the enormous amount of money being spent to bring some of these animals back (which is extremely unlikely to happen, though I’d love to be proved wrong) would be better used to conserve existing habitat and preventing future extinctions. I have some other issues that I’ll raise later.
One of these companies, Colossal Biosciences, is claiming they’ve done this with the dire wolf.
Except they haven’t. Not even close.
What they did, from a straightforwardly scientific point of view, is actually very cool. They edited some gray wolf genes. 20 edits for 14 genes. Yes, it’s an accomplishment – but the cubs are not dire wolves. They’re gray wolves with a bit of genetic tinkering.
I know GMOs are a sensitive topic here. I will neither defend nor decry them, and the debate is irrelevant to my point, which is that when genes from different organisms are spliced into something like maize or tomatoes, the resulting plants are still maize and tomatoes. They just have some tinkered genes. (Again – whether this is a good thing or a bad thing is completely beside the point here.) This is pretty much identical to what’s been done here; they’ve bred gray wolves with some tinkered genes.
Although the people at Colossal Biosciences have accomplished something of note, they haven’t resurrected anything.
Is it a species?
In response to published criticism of their claims, Colossal issued a statement arguing that we shouldn’t (or don’t) use genetic criteria to recognize species, and that we should use phenotype. This opens that great oaken door to the Pandora’s Box that is species concepts, but it’s also completely wrong.
In high school biology, we learn that species are populations of interbreeding individuals that produce fertile offspring, but can’t do so with individuals from other populations. This is known as the Biological Species Concept (BSC), and the dirty secret is that very few biologists actually use it. I, for one, usually can’t – fossils, being dead, generally don’t mate. There are asexual species, and for some groups (including mine – crocodiles and alligators) – hybridization between what would generally be considered distinct species is very common. Following the strictest application of the BSC possible, the 12 or 13 recognized species of Crocodylus would become one.
So what’s a species?
Here’s a problem I’ve encountered. Maybe you have, too. I sometimes encounter groups of biologists, and they’re all alive. What do I do about that? The ancient Greeks used hemlock, but there are way too many biologists for that nowadays. So I lock them all in a room and say, “no one gets out until you agree on what a species is.” They will all die.
That said, the myriad concepts out there boil down to three. The BSC is one of them, but there’s also the Evolutionary Species Concept (ESC), in which a species is a lineage of populations evolving over time; and the Phylogenetic Species Concept (PSC), in which a species is basically the smallest unit of biodiversity circumscribed by unique combinations of morphological and genetic features.
These aren’t mutually exclusive of each other. In effect, the ESC is what a species actually is; the BSC explains how species come into being for sexually-reproducing organisms; and the PSC explains how we know we have a species. The ESC defines the conceptual species that we hope our operational (or inferred) species, which are smallest diagnosable units, approximate as closely as possible. And reproductive isolation is the best possible evidence that a smallest diagnosable unit is a species.
What Colossal described is kind of like the PSC, but the PSC as used by actual biologists would still (in concept, at least) represent evolutionary units. These cubs don’t fit that definition. So the statement they released is nonsensical and should be ignored.
Again, what Colossal Bioscience did is impressive. But they haven’t resurrected anything.
This also brings up a broader discussion – what, exactly, is “resurrection?”
The purest form would involve the re-creation of an organism from the complete genome of another. This is cloning. And for extinct animals, it’s very unlikely to happen. We have DNA from a comparatively small number of fossils, and it’s usually degraded, so we usually don’t have the complete genome.
We obviously have complete genomes for many recently extinct species (Tasmanian wolf, great auk, dodo, etc.), and we do have complete genomes from some fossils, but unless it’s in an intact cell nucleus, it won’t do much for cloning.
And we would want this in the nucleus of an ovum or stem cell.
There’s the issue of surrogacy. Woolly and Columbian mammoths (Mammuthus primogenius and M. columbi, respectively) were much larger than their closest living relative (Asian elephant, Elephas maximus), and their calves would have been larger; would a female Asian elephant be able to carry a mammoth calf to term? And would her body recognize it as a foreign entity with different DNA and reject it? Colossal Bioscience itself ran into this problem, which is why they switched from mammoths and dodos to the dire wolf - although without a close living relative, this is still a big problem for dire wolves.
(There have been claims in the media that the dire wolf and gray wolf are each other's closest relatives. They are not.)
This is why I think efforts to clone the Tasmanian wolf are doomed; their closest living relatives are no more than half the size of an adult thylacine. It's also why I think if we're going to clone a mammoth, we should try for one of the miniature species that lived on islands in the Mediterranean or off the coast of California. (I want those resurrected because I think they'd make awesome pets.)
The closest we've come to actually cloning something extinct was with the Pyrenean ibex, which is considered a subspecies of the Iberian ibex (Capra pyrenaica). The last known individual died in 2000. A cloned individual was born in 2003, but it only lived for a few hours. (This makes the Pyrenean ibex the only extinct species to be cloned and the only species to have gone extinct twice.)
This is also the reason we'll never resurrect Tyrannosaurus rex. The oldest DNA we have is around 2 million years in age, and it's degraded. DNA just doesn't last that long.
Assuming we can’t actually clone extinct animals, we could genetically engineer them, which is what Colossal Bioscience is trying to do here. Or we could selectively breed modern animals to look like their extinct relatives. I wouldn’t try this with a mammoth – generation times are way too long – but we could, in practice do so.
But are these true de-extinction? Or have we created something new that might resemble something extinct?
Then there’s the cost. A lot of money is being thrown at the problem. Meanwhile, land use and climate change are increasing the risk of further extinctions that could be prevented.
And what do we do with these animals? Put them in a zoo? “Re-wilding” is nonsense – I’d be happy to explain why in the comments – so releasing them into the wild is not an option. How do we maintain them, given that the environments they lived in may no longer exist?
There's also a deep ethical concern if someone tries to resurrect Australopithecus, Paranthropus, or an earlier species of Homo. What rights would these individuals have? Would they be considered human in the eyes of the law? Would they be treated with the same dignity as modern humans? And where would they live? This, I think, should be flat-out forbidden.
It hurts me to say these things, because it would be totally cool to see a real dire wolf.
Anyway, my thoughts on it. The dire wolf is still extinct, and I suspect it’s going to stay that way.
= new reply since forum marked as read
She was a wonderful companion but...a handful. We had a 6' fence, she managed to hop it a few times. Daily walks, at least 3 miles, many longer. She only shed twice a year, but it lasted six months each 
.
I am not a scientist but always interested.
Greetings from Germany
Are two reasons why I keep coming back to DU. I want to have a beer with you and I don't even like beer.
Thanks so much for sharing this. Please share more.
❤️pants
I just drove my husband bonkers trying to read it to him — this is the man who finds obscure branches of mathematics so much fun he insists on sharing with me. 
Anyway — I’m just as glad the dire wolf isn’t going to be roaming the hills any time soon. Thanks! 
I've had some thoughts on the species concept for several reasons (mostly because I've always been interested in paleontology and evolution, since preschool, despite eventually choosing an unrelated profession), especially including the very different approaches to the species concept based on interbreeding populations vs fossil remains, which usually include almost exclusively skeletal remains for vertebrates. I remember how the term "missing link" kept showing up in the popular press (ever since Darwin) despite the vagueness of the term -- the Futurama episode "A Clockwork Origin" satirized this concept to extinction.
I was also driven to reflection based on a discussion of the meaning of "equivalence relations" in mathematics -- and how concepts of "approximately equal to", if taken to be transitive, can lead to obviously invalid conclusions. Thus two members of a living species have "approximately equal" genomes, and could even be immediately related. But if A is genetically equivalent to B, and B is equivalent to C, and C .... equivalent to ZZZZ, at what point do we encounter a genome that differs enough from A to make viable offspring impossible ? Obviously this is usually discussed in terms of geographic separation of populations, and how many generations of separation are required for genetic drift to lead to speciation. (Cue discussion of Darwin's finches, Galapagos tortoises, and island populations in general.) But the same concept applies to successive generations, as well. No one denies that an organism is related to its ancestors in some way, but at what point in the succession do you drive in a survey marker and say "here lies the boundary between species, because generation 1 and generation 99,999 cannot interbreed/have wholly distinct type specimens ? And if that's true, wouldn't it be true of generation 101 and 100,100 as well ? Or 201 and 100,200 ? How to justify drawing a sharp line in the midst of a continuous gradient ? And how would you ever test the concept of interbreeding between living and dead generations, without complete knowledge of their genomes and absence of at least one breeding example ?
I'm sure I'm not the only one ever to wonder about such things, but I keep finding myself forced to these conclusions:
1) The species concept is a very useful tool, but ultimately an intellectual creation at least as much as a natural phenomenon, and we shouldn't allow it to restrict our thoughts too incautiously.
2) Whatever definition we choose is based on its utility here and now, in the case we are examining, and may not be generalizable even to apparently similar cases.
3) Ultimately, evolution happens, and the evolved and evolving species are under no obligation to conform to our labels (another of our useful, but human-devised, intellectual tools), our behave in line with our theories. General concepts can be extracted (and are powerful within their domains), but are not universally binding. After billions of years of evolution, and millions of recognized species, almost any possible mechanism of natural selection is likely to have occurred at least once, if not repeatedly. So I don't really have a bet on either dog in the Punctuated Equilibrium/Phyletic Gradualism debate, other than to say, "yes, I'm sure that has happened many times; examples abound" to both. Cool names, though, which sometimes seems to matter a lot to their proponents.