No, it's not a dire wolf

Genes do not make the essence of a species

Okay, I guess I have to weigh in on this one. If you do not have any access to the internet but Substack, you may not have heard that a company, Colossal Biosciences, who have a garish and hyperbolic website at the least, claim to have “de-extincted” dire wolves (Aenocyon dirus; will nobody adopt my much better term “extanted”?). The reports are at best a PR nightmare, but on their own account they inserted fifteen genes into a wolf (or maybe domestic dog) embryo. Out of some 20,000 genes a mammal species usually has, that means around three quarters of 1/100th of the genome has been changed. This means it is still just a Canis lupus, whether wild or domestic.

Mammalian species can exchange genes across large phylogenetic distances such as Endogenous Retroviral (ERV) insertion, as well as the usual introgression or hybridisation. As much as 8% of our genomes are ERV, and mammals as a whole may have 40% ERV derivation in the past (that’s an old figure, from the 90s, so it may be wrong). In sum, these changes are less than normal, natural, gene exchange.

But this is for the scientists to debate and settle, not a philosopher. What I would like to note instead is the notion of “species” being employed here. In a more balanced report, Gizmodo reports questions being asked by scientists, and this passage in particular caught my eye:

Jeremy Austin, Director of the Australian Centre for Ancient DNA, told ScienceAlert that all Colossal has done is create a genetically engineered gray wolf that looks like what the company thinks a dire wolf might have looked like. And even that is up for debate; canids are morphologically similar, making it hard to verify from fossil remains the exact appearance of an extinct member of the family.

Evolutionary biologist Beth Shapiro of Colossal has given a definition of a species that Austin finds misleading.

"Species concepts are human classification systems, and everybody can disagree and everyone can be right," Shapiro told Michael Le Page at New Scientist.

"I think that the best definition of a species is if it looks like that species, if it is acting like that species, if it's filling the role of that species, then you've done it," she told ABC News.

But looks aren't everything. Cryptic species, for instance, are organisms that are almost indistinguishable but are genetically distinct and do not typically interbreed.

Austin compares Shapiro's lenient definition to the literary folktale, the Emperor's New Clothes.

"If you say you've done something, and enough people believe you, then, well, you've done it," Austin told ScienceAlert.

"Whereas I think a lot of scientists are going to be scratching their heads, saying, 'Look, you've got a white, gray wolf.' That's not a dire wolf under any definition of a species ever… I don't think that this represents de-extinction in any way, shape, or form."

So the question is: what would a novel organism need to have to be an old species? First off, beware words like “organism” here. What you need at a minimum for sexually reproducing organisms to be a species is a population. This may, or may not be true of asexual and chromosomal (polyploid) species, but in this case where the reference is the familiar furry vertebrates we usually think of, a population of reproducers is needed.

Secondly, and this is a view repeated from Darwin to David Hull, once a species goes extinct, it cannot be resurrected (to use Jay Odenbaugh’s term). There is no reproductive linkage between them, which is and always has been the sine qua non of specieshood. Maybe, if you had an entire genome, you’d have a morphospecies analogue, but even that is unlikely. Why? Because genes do not determine the nature of species. Not alone, that is. Genes do not equal essence. Ecosystems, population structure including migration, individual adaptation, and a host of non-genetic factors come into play. Apropos of wolves, ecological structure is one of the reasons populations diverge (Leonard 2014).

The “species problem” is an attempt to consider what makes a species a species, which presumes quite a lot in just posing the question. Do “species” exist naturally, or are they, as Beth Shapiro says, just a human convention? If not, when does the older kind (call it whatever you like) get resurrected? Is 0.00075 of a genome enough? Now I do not think species are “just” human conventions, or at least some species. But the rank, or category, of species most certainly is. Specialists in a group converge on something similar (like the orangutan) even if they may subdivide that group (into three species of orangutan). Cryptic species, which are morphologically identical, have different behaviours, ecological functions, population dynamics, mating strategies, and, yes, genomes, so that when the test arises for each trait or property, there can be agreement for further names. What makes a species for wolves, orangutans and other furry mammals isn’t what makes a species for ferns, bacteria, fungi or even other vertebrates. But one thing that is necessary, although not sufficient, is reproduction, or descendency, even if we do not know it.

This notion that there are essential properties that makes some animals the same species as others rests on a category mistake, in my opinion (see previous post). It is the same thing as saying that there is some observer-independent property of constellations of stars that makes them constellations. We see the groups, but it is a matter of solid evidence and prior scientific understanding to determine whether there is a taxon there, or just resemblance, and this is something many molecular geneticists seem not to understand.

Finally, any species shares, perhaps not uniquely, a developmental process, or system. Now, canines have a developmental system that permits quite a bit of hybridism, but there’s nothing too unusual about that in most groups of sexual vertebrates. But Canis and Aenocyon are not all that close [see this excellent slide deck], although coyotes and gray wolves do interbreed (that’s the infamous red wolf).

So, Canis and Aenocyon lack shared ecologies, genetics, develomental systems, histories (since the split some 5 or 6 million years ago) and probably even species modalities. But to illustrate the absurdity of this claim, suppose humans were extinct, and the intelligent meerkats that succeed us attempted to resurrect us by modifying 15 genes of a common chimp, also 6ish million years separate from us. At best you’d get a weaker, and possibly noisier, chimp. You know, traditionalists.

Ref.

Leonard, Jennifer A. “Ecology Drives Evolution in Grey Wolves.” Evolutionary Ecology Research 16 (2014): 461–73.