Resurrected mammoths and dodos? Don't count on it

A Japanese team of scientists plans on cloning a woolly mammoth within five years. Photograph: Ria Novosti/AFP/Getty

I recently spoke at Revive & Restore's TEDx DeExtinction event at the National Geographic headquarters in Washington DC. Most of the speakers were brilliant geneticists working on ways to revive species that no longer live on earth by injecting DNA from extinct species into eggs of living relatives. The atmosphere was electric with the hopes and claims of top scientists bent on bringing back the woolly mammoth, the passenger pigeon, and other vanished species. I was the invited skeptic, and here's what I told them, more or less.

The poster child for de-extinction is the passenger pigeon. The first European visitors to North America saw flocks so huge that they darkened the skies from horizon to horizon. Even in the 19th century, when the pigeons were starting to decline, observers estimated over a billion birds in some flocks. A market hunter with a shotgun could kill 50 or 100 with a single shot. The combined weight of the pigeons could bring down giant tree limbs with a sound like cannon fire. Yet the last passenger pigeon, a bird named Martha, died in the Cincinnati Zoo in 1914.

What happened? The easy answer is that it was probably a combination of forest loss and market hunting. But in the late 1800s, there were still thousands of pigeons left, some in protected areas, so why didn't any populations rebound? The likely answer is that the birds, which laid only one egg each, needed hundreds of thousands of other pigeons around to properly mate and nest, and to overwhelm their predators. This is a well-known phenomenon, which biologists call the Allee Effect.

So the de-extinction experts are going to try to revive one the most social birds in history, at great cost, and will bring back four, five or even 50? Even though they might go extinct again?

Another problem, which applies to many of the proposed de-extinction animals, including the woolly mammoth, Tasmanian tiger, and passenger pigeon, is what animal will mother the babies. It would have to be a similar species, like a band-tailed pigeon for the passenger pigeon, but we know that each behaves very differently: they fly and mate differently; they eat different foods; and they have different calls. Maternal care is critically important for an infant's normal development. That baby bird will live in a cage labeled "passenger pigeon", but it might not be a real passenger pigeon, and it isn't going to bring them back from extinction.

And genetics might prevent this from even being a passenger pigeon. In the first college course I took in molecular biology, taught by Francis Crick, we learned that DNA is a book of instructions, telling us how to make a specific organism. "DNA makes RNA makes protein" was the mantra. But a half-century later, we now know that things are not so simple. A DNA strand doesn't tell you how to make an organism – it's more like a database, or a dictionary, than a book of instructions. All the words in Hamlet are in my dictionary, but if I scan my dictionary, Hamlet doesn't fall out.

The emerging science of epigenetics tells us that a strand of DNA can be read in hundreds, maybe thousands of ways. The internal and external environments of an egg cell can tell the it how to use the DNA to make an organism. That's why identical twins of animals, or even clones of plants, can have the same DNA but be turn out quite differently.

With de-extinction, the epigenetic departures are likely to be more extreme than in identical twins, because the donor animal (the species still around) will have an egg that responds to different environmental factors than the extinct animal's did. The DNA of a band-tailed pigeon, for example, will react to different stimuli than the passenger pigeon's would, but you need the former's egg to even attempt bringing back the latter. How the passenger pigeon's DNA will be read by the band-tailed pigeon's egg cell is anybody's guess.

I described this problem to a molecular biologist and he had a clever suggestion: take two clearly different but related living species, pretend one is extinct, and put its DNA into an egg cell of the other species. When you get a viable organism, you can see how well it matches the DNA donor – there will be plenty of living examples to compare it to. Use a black rat as the faux extinct DNA donor, and its genetic cousin the brown rat as the surrogate mother; if the animal you get from the pairing looks and behaves just like a black rat, great. If not, time to do some rethinking.

Finally, if it ever happens, de-extinction will only target a very few species, and is extremely expensive. Can we afford to divert dollars from tried and true conservation measures, which are already short of funds? Strategies such as creating wildlife sanctuaries and public campaigns to protect endangered species are proven successes. At this moment, brave conservationists are risking their lives to protect dwindling groups of surviving African forest elephants from armed poachers. Here we are, safely talking about bringing back the woolly mammoth.

The conservation movement has based much of its appeal on the power of the idea that "extinction is forever." We could undercut conservation if the public (including members of Congress) is led to believe that extinction is only temporary, a technological bump in the road – nothing to worry about.

Right now, de-extinction is just an interesting idea, what we might call recreational conservation. Pursue the dream if you like, but ease off the hype.