Can We Reproduce The Woolly Mammoth?

Note: This essay is based on a paper given at the 2019 Eastern APA in New York City.

Media reporters often announce that we are on the verge of bringing back the woolly mammoth, even while there is growing consensus among scientists that resurrecting the mammoth is unlikely. To resurrect a member of an extinct species, one must have a method for doing so, and it’s not clear what method could be used to resurrect a member of Mammuthus primigenius. Cloning requires an intact genome, and no environment on earth is cold enough to freeze a large animal like the mammoth quickly enough to prevent DNA decay. That’s not to say that it is impossible for DNA to survive for many hundreds of thousands of years, only that such accidental preservation would likely consist exclusively of DNA fragments, not an entire genome.  This is why current “de-extinction” efforts are not aimed at bringing back the mammoth. Instead, the aim is to use fragments of mammoth DNA to recreate mammoth traits in close relatives. For example, Harvard scientists are merging mammoth and elephant DNA to create an Asian elephant with the thick coat of a mammoth. Such a creature would look mammoth like, but it would remain an elephant according to the scientific consensus. But is that consensus right?  How should we classify creatures that have both mammoth and elephant DNA? Are they elephants, mammoths, or both?

Since Darwin, we’ve known that membership in a species isn’t determined by similarity but is, rather, dependent on spatiotemporal continuity. In other words, for two objects encountered at different locations to be members of the same individual species, there must be a spatiotemporally continuous path between them. Given that fact, can a creature created by merging ancient fragments of DNA with modern elephants be spatiotemporally continuous with its long-dead conspecific? Well, we know that it’s possible for offspring of a species to be born even after all the living members have died. Consider, for example, species of annual plants. During winter, when the plants die, there are no living members of the species. Yet, when spring rolls around, the seeds that froze over winter give rise to new members of the same species. The point is made by Alastair Gunn when he writes:

Certainly we are in no doubt about the continued existence of species such as annual plants, whose entire population dies off each year. The seeds, which are all that survive the winter, are genes, not plants. Similarly, it is possible to imagine a species which survives the winter only in the form of unfertilized ova and sperm, relying on vector organisms or other environmental factors to arrange fertilization. In such a case, even though no individual member of the species exists, not even in embryonic form, we would surely not want to say that a new species evolved each spring.

If a plant species can survive the winter even though its living members do not, why resist the idea that a new mammoth could belong to the species M. primigenius even though there was a lengthy period of time when there were no living members? I don’t believe there is good reason for such resistance. Plausibly, then, it seems possible that the spatiotemporal continuity of extinct species can be preserved or maintained even when all living members are no more.

But is this theoretic possibility practically possible? What potential is there for reviving members of a species that has been extinct for thousands of years? This is a question that can only be answered by thinking about what it would mean to reproduce a member of an extinct species. What requirements must be met for a process to qualify as one of reproduction? In the case of the mammoth, since neither sexual nor asexual reproduction is possible, could processes used to merge fragments of mammoth DNA with modern elephants meet such requirements?

On the account of reproduction for which I’ve argued (I call it the ODP account), reproduction occurs when there is material overlap between parent and offspring (Overlap), the inherited parts contribute to the development of the offspring’s own reproductive capacities (Development), and the material parts, or their descendants, are passed to the next generation (Persistence). I’ve also argued that there is at least one method available for replacing elephant DNA with portions of mammoth DNA called Polymerase Chain Reaction (PCR) that can fulfill all the requirements. As a result, I believe it’s practically possible (though very unlikely) to resurrect a member of M. primigenius.

Suppose a scientist recovers a sequence of mammoth DNA and uses a method of replication like PCR to amplify the sequence. Suppose further that she takes the DNA product and inserts it into the genome of an elephant embryo, creating material overlap between the extinct mammoth and the embryo. This process would mean meeting the first requirement of my ODP account of reproduction. Since the DNA merger would happen at a very early stage of embryonic development, the mammoth DNA would likely contribute to the full development of the offspring, including its reproductive organs. If so, the second requirement of my ODP account would also be met. Further, and maintaining these assumptions, if the mammoth DNA were passed to further generations, all three requirements of my ODP account would be met. In such a case, amplifying and inserting mammoth DNA into an elephant egg would preserve the spatiotemporal continuity required of species membership and the newly created creature would meet the only widely conceded essential condition for qualifying as a member of the species M. primigenius.

Now, it’s perfectly reasonable to resist this conclusion by saying, “Surely, it’s absurd to believe that an organism born of an Asian elephant, created using an elephant’s egg, and containing only a small fraction of mammoth DNA is a member of M. primigenius.” I agree. It’s perfectly reasonable to have that response. Even so, the only widely conceded necessary condition for species membership is that two creatures are spatiotemporally connected. And no matter how small the fraction of mammoth DNA used to produce a new “mammoth” using an elephant as host, as long as there is overlap, development, and persistence, the spatiotemporal requirement of species membership is fulfilled.

Of course, this point doesn’t allay the unease that may come with classifying an organism born of an Asian elephant, created using an elephant’s egg, and containing only a small fraction of mammoth DNA as a member of M. primigenius. What could? Maybe nothing. But if that’s so, it’s because spatiotemporal continuity is a necessary and not a sufficient condition for species membership. There may be other conditions needed for determining when an organism belongs to a particular species. As Kevin De Queiroz points out, what those conditions are continues to be a matter of controversy, “each of several contemporary species concepts adopts a different property… as its cutoff for considering a separately evolving lineage to have become a species.” Thus, failure to meet some further condition of species membership over and above spatiotemporal continuity may justify excluding some particular organism (in our case, the creature resulting from merging elephant and mammoth DNA) from a species. And I suspect this is what explains the unease that would come from classifying an organism born and bred using an elephant host as a member of the species M. primigenius. It’s not that such an organism would fail to be sufficiently similar to its extinct ancestor. Rather, it would be that it failed to meet some further condition of species membership, even if we can’t say definitively what that further condition is or ought to be.

Monika Piotrowska

Monika Piotrowska is Assistant Professor at the University at Albany, SUNY. Her research focuses on conceptual and ethical issues arising from advances in genetics and biotechnology.