He died of the cold. His name was Benjamin, the thylacine, Ben, the last Tasmanian Tiger – only we didn’t know that when he was captured and put in a zoo in 1933.

In grainy black-and-white footage at the time, Benjamin paces his enclosure, yawning and baring his jaws. He lies down, he sniffs the concrete. At one point (off-screen) he even gives the cameraman a cheeky bite on the bum.

He died three years later, locked out of his backroom shelter one freezing night, just weeks after his species was at last granted protected status in Tasmania following decades of hunting. Eventually, the world came to realise that Benjamin really was the last of Australia’s great striped marsupial. But, when he died, they saw only an animal too damaged to be preserved in a museum. His body was tossed in a dumpster.

Benjamin’s story now haunts ecologist Euan Ritchie as he maps accelerating extinction rates around the world. “What a perfect metaphor,” he says. “Benjamin died of neglect and then we threw him out. These species took millions of years to evolve and now they’re disappearing by the thousands because we don’t care enough. It’s like going into a museum and lighting all of its precious paintings on fire.”

Paleontologist Michael Archer is also haunted by the story. But he has a plan to make sure Benjamin is not the last thylacine. Archer isn’t one of Australia’s infamous Tasmanian Tiger hunters, the ones who trek into the bush convinced they may yet find a survivor of the extinct species. Though he has diligently DNA-tested suspected thylacine excrement sent in by such “spotters”, Archer says sightings always turn out to be “fascinating bullshit”.

Instead, he is part of the “de-extinction club”: a growing group of scientists working to harness genetic engineering and cloning to reach into the past – and resurrect extinct animals. Top of the list are the thylacine and the woolly mammoth. Archer and others say the unnatural pace of climate change and habitat destruction mean bringing back key species may now be the only way to stop ecosystems from collapsing. Harvard University’s renowned geneticist George Church, himself working to return the mammoth to the Arctic tundra, says reviving some species could even help combat the effects of global warming. But others, such as Ritchie, warn that it might put the wild in jeopardy all over again, or pull vital focus from the urgent work underway to save those species we do have left.

So how does de-extinction work? Would a woolly mammoth cooked up in a lab be a real mammoth or just a funny-looking elephant? How do we choose which species get a second chance? And is there any dino DNA left to get us to Jurassic Park?

Why bring back extinct animals?

There’s a story humans tell of saving animals in an ark – two of each kind to survive a great catastrophe. Today, the great catastrophe is here, at least for wildlife. Humans are burning through the planet’s resources with an unprecedented appetite, changing the climate, concreting over the wild. Archer says extinction rates are as high as they were during the Cretaceous period, when 75 per cent of species, including the dinosaurs, were wiped from the map. In half a century, the World Wildlife Fund calculates, we have lost more than half of the planet’s biodiversity. “We’ve now entered the planet’s sixth mass extinction event,” Ritchie says. “And, ultimately, humans need these ecosystems to survive, too. They’re our life support system.”

Archer argues that conventional conservation efforts “aren’t cutting the mustard” and the time has come for extraordinary intervention - an ark of sorts. “Normally, nature fills the vacancies from a big extinction event like this. But we’re not leaving any room for that this time. So we’re really in uncharted territory. And we have to be smart.”

For some scientists, being smart means conserving sperm, egg and tissue samples from endangered species in cryogenically frozen arks, the same way conservationists might keep breeding pairs in captive populations, in case they can one day be returned to the wild. For Archer and others, being smart means using technology not just to slow down extinction but to reverse it.

Of course, unlike Noah, you will need more than two of a species to bring it back. If de-extinction is to mean more than a few curiosities in a lab or a zoo, scientists recommend a gene pool of at least 50 to 1000 animals to start. And you need to make sure that both the species and the wild you’re sending them back into can cope with their return. And that’s even before we get to the technology itself.

How do you “de-extinct” something?

It’s not quite enough to thaw a frozen mammoth from a block of ice. Scientists need either tissue to clone an animal or enough of its DNA, its genetic blueprint, to engineer it. In Jurassic Park, that source code came from a preserved mosquito with a belly full of dinosaur blood. In real life, that wouldn’t actually be enough DNA to rebuild a dinosaur (the little molecule is hardy enough to survive at crime scenes but, after about 1.5 million years, it’s too decayed to read anymore). Still, carcasses of mammoths and Neanderthals preserved in the icy permafrost at the top of the world, some of them a million years old, have yielded enough of their genetic code for scientists such as Church to rebuild, and edit. Even the mysterious virus behind the deadly Spanish flu of 1918 was recreated in a lab from the frozen lungs of one of its victims, unearthed from an icy grave in Alaska.

“No-one’s advocating for the de-extinction of viruses, of course,” Church says. “But there’s a lot more possible than we first imagined.”

Since the development of better gene-editing tools such as CRISPR (which borrows the precision of ancient bacteria immune systems to find and edit specific genes), Archer says museum collections, too, have become “the flavour of the month”. “Suddenly everyone wants to go in and sample shrivelled toes [for] the DNA.”

Scientists are even learning how to wind back the clock on a living animal’s family tree, searching for dormant genes switched off over their evolution, such as a tail or bigger teeth, to help revive extinct ancestors, gene by gene. That’s how the paleontologist who inspired Michael Crichton’s Jurassic Park in the first place, Jack Horner, is hoping to build a dinosaur from its decidedly less scaly descendant: the chicken. Birds are the dinosaurs that escaped extinction, after all.

“Genes even older than [those of] the dinosaurs can be brought back, too,” Church says. “But it’s limited. It’s hard to reconstruct a [species’] entire genome that way. It’s not like you have a 3D printer where you say, ‘Print out this organism’ because the rules are way more powerful than that. They’re more mysterious. And the question is always, why do it?”

Horner himself says he’d be looking to turn on only a few lost traits in his dino-chicken (“the claws, teeth, arms, scales and the tail to keep the fourth-graders happy”). Already, a beak has become a snout in chicken embryos. The tail has proven the most difficult but in recent months “we’ve made headway understanding how [it] evolved from dinosaur to bird”. Even so, Horner’s chickenosaurus wouldn’t really be an extinct animal. “It’d be a new kind of dinosaur-like bird,” he says.

Church himself has viable DNA of his mammoth but will still need to pair it with the genome of its closest living relative, the Asian elephant, to try to bring it back (he sometimes calls the project the “mammophant”).

Likewise, Ben Novak at the genetic rescue and de-extinction group Revive & Restore has big plans to re-engineer and breed an extinct line of North American bird known as the passenger pigeon using existing flocks (he’s even done some experiments on birds at the CSIRO’s secure lab in Melbourne). And Archer plans to turn to the Tasmanian devil as a template to recreate the Tasmanian tiger, after the thylacine’s genome was at last sequenced from DNA found in a specimen at the Melbourne Museum.

These animals, if they ever come blinking and growling to life in the lab, will be hybrids of the past and the present. But there is a way to bring back an “100 per cent extinct animal”, Archer says, and that’s cloning.

You can clone extinct animals?

More than 20 years after Dolly the sheep became the first mammal cloned from an adult cell, the technique is yet to be perfected, but Archer says it’s not quite the horror show people imagine. While the sci-fi nightmare of cloned humans never materialised, livestock can be cloned to preserve breeding lines, and celebrities and millionaires fork out upwards of $40,000 to clone beloved pets.

Some conservationists are also turning to the technique to stop inbreeding in dwindling wildlife populations. In late 2020, Novak teamed up with the US Fish and Wildlife Service to clone a critically endangered black-footed ferret from the frozen cells of a ferret who died in 1988. The clone, named Elizabeth Ann, is now a healthy six-month-old who is fond of tearing apart paper bags, barking at anyone who invades her personal space, and has “three times more genetic variation in her little body than any other [black-footed] ferret on the planet,” Novak says.

Of course, to clone an animal, the cell you are using must still be intact – alive, in a sense. That makes cloning an extinct species almost impossible. But it has happened. In 2003, scientists in Spain cloned an extinct mountain goat known as the Pyrenean ibex from the frozen tissue of the last of its species, Celia.

To clone, you take the egg of a suitably similar host animal, say, a domestic goat, and suck out the DNA-packed nucleus within, replacing it with that of the animal in line for resurrection, in this case, Celia the ibex. Then you hit it with a Frankstein-esque jolt of electricity to fuse the egg and nucleus, and you implant that new egg in a surrogate mother (another goat). If all goes to plan, the DNA will tell the egg to grow an ibex instead. In this case, a baby ibex did arrive but she lived for just 10 minutes, born with a fatal lung defect scientists say can be typical of the species, clone or not.

The experiment was never tried again because by then the Spanish government had released goats into the mountains to replace the ibex, and so the team despaired they were too late. The ibex had lost its habitat. It had effectively gone extinct twice.

Archer hopes there will be a happier ending for the gastric-brooding frog, an extinct Australian species he’s been working to clone since a colleague discovered some intact tissue cells “miraculously still tucked away at the back of an old university freezer”. This frog first caught the eye of medical researchers for its bizarre ability to turn its stomach into a womb and vomit up its babies. “Nothing else in nature can do that,” Archer says. But before it could be studied, in the mid-1980s, it vanished. Then in 2013, Archer’s team had a breakthrough. The extinct frog’s DNA began to replicate when it was implanted in donor frog eggs. Under the microscope, the team watched the embryos start to develop with growing excitement.

“But suddenly it just stopped,” Archer says. The team believes the problem lies not with the DNA, but with their technique for cloning amphibians. “We hit the same wall when we tried a living frog’s DNA. We just need to get one [species] back, one of these [de-extinction] projects over the line, and people will see we’re not making monsters.”

What does bringing back a mammoth have to do with climate change?

OK, so Jurassic Park probably won’t happen but what about a Pleistocene Park for the king of that Ice Age, the woolly mammoth? These towering herbivores were hunted to extinction by early humans some 10,000 years ago, the very last of them surviving on Arctic islands until 4000 years ago. But the mammoth is still the closest genetic relative to the now endangered Asian elephant. “Even closer than the African elephant,” Church says.

He believes resurrecting the mammoth’s ancient genes could stop the Asian elephant from following it into extinction. Splicing in traits that helped the mammoth thrive in the Arctic could open up crucial new habitat, as land-clearing and poaching closer to the equator increasingly whittle down their numbers. Endangered species are already relocated, with varying success, by conservationists, and their genes managed via breeding programs to protect diversity.

Church estimates that editing in about 40 to 100 mammoth genes, chiefly around cold resistance, will be enough to allow Asian elephants to thrive up north. Separate projects have edited about that number in pigs, for different traits, and Church says they are now breeding whole generations of healthy, engineered animals.

In the case of his mammophant, the team would grow the animal in an artificial womb to avoid any risk to the endangered elephant they would otherwise have to use as a surrogate. That means there’s an extra hurdle to scale – growing a mammal artificially, all the way from fertilisation to birth, hasn’t been done before. Church expects to crack the problem in about five years, in mice first, which have a faster gestation period than elephants (20 days versus 22 months). “Then it’ll probably take another five to adapt it to larger animals and then we can see how it scales up for the mammoth.”

If it works, he already has a place to put the herd. Since the ’90s, a group of Russian scientists has been transforming a huge swathe of land in Siberia back to the grasslands that mammoths and other large animals once roamed. It really is called Pleistocene Park and, with the mammoth’s help, some think it could actually slow climate change. Grasslands can absorb more carbon than forests, and mammoths rip down trees and create this tundra as they go. But, more importantly, their heavy feet also trample snow cover, stopping it from acting as insulation and so allowing the permafrost to be chilled by the icy Arctic winds. In theory, Church says this should help slow its thaw, which eventually threatens to release more carbon and methane than the atmosphere holds today.

But Ritchie questions why you would bring back a mammoth, a creature of the Ice Age, to a rapidly warming world? “You’re not going to have herds of thousands of mammoths in time to have a real impact on the permafrost, given how fast it’s melting now with climate change,” he says. “You’ll just end up with an elephant that can’t handle the heat, and probably, a freak show. We have to think very carefully about how the world is going to be when we consider what to bring back.”

What happens when we put an extinct animal back in the wild?

There’s not much point resurrecting a species if it will face the same threat of extinction soon after, like the Pyrenean ibex muscled out of its mountains. And de-extinction proponents stress that animals should fill an empty ecological niche too. When wolves were hunted out of Yellowstone National Park in the United States, elk numbers exploded. With no predator to keep them in check, they tore up the grasses and rivers. Suddenly, the beavers had vanished too. “And when they brought the wolves back, 70 years later, the ecosystem was restored,” Church says.

Of course, for this more classical “rewilding” to work with a resurrected animal, it needs to act the way its ancestor did. But not everything is encoded in genes. How will an engineered mammophant, for example, learn to migrate across the Arctic tundra as mammoths once did if there’s no parent to show it the way? And what if cutting and pasting together species’ genomes, in this case of elephant and mammoth or thylacine and Tasmanian devil, interferes with other natural instincts?

These are problems Archer ponders a lot. “You could even have two sets of instructions [in the DNA] that are contradictory,” he says. “But don’t forget that the mammoth is a kind of specialised elephant, so most of the genome is already the same. Tasmanian devils, too, are close enough to thylacines, even though they’re smaller.” For his part, he believes “we’ll have 99 per cent original behaviours in resurrected species”. Most behaviour really is gene-deep in animals, he says, even the strange case of the gastric-brooding frog swallowing its fertilised eggs. “No frog teaches another frog to do anything, they’re on their own from the moment they’re a tadpole.”

In the case of a resurrected thylacine, there won’t be much to compare it to. There are few records of how the marsupial lived, so some ecologists warn not enough is known to bring it back safely. Archer is quick to point out that the thylacine vanished from Tasmania only 90 years ago, and from the mainland at the same time as the devil (which is itself being considered for reintroduction over the Bass Strait) some 3200 years ago. “We know what it’s going to do. It’s going to become the king of beasts [in Australia] again.”

But consider the case of the passenger pigeon that Novak hopes to bring back to North America. One hundred and fifty years ago, they were the most abundant bird on the planet. And, though they numbered as many as six billion, Novak says there were only three or four flocks flying the world at any one time. When they moved from forest to forest, they came in like a hurricane or a forest fire, breaking branches, destroying canopies and forcing those woods into regeneration cycles.

“No other birds do this,” Novak says. “They were ecosystem engineers. Some of the restoration we thought fire did to the landscape we’ve now shown the birds did.” Novak argues the forest needs them back. He and his team have sequenced the pigeon’s genome and compared it to its closest living relative, the bandtailed pigeon. Of the 25-million-odd genes where they differed, Novak has identified about 30 that could be particularly significant in making a pigeon behave like a passenger pigeon, such as disease resistance and, potentially, extra-social behaviour.

So here’s his plan: Novak imagines a carefully controlled release, first on a netted reserve with nesting baskets packed into dense trees, encouraging the birds to breed in colonies and, to fool them into thinking they are already part of a much bigger flock, with speakers blasting pigeon calls and coos. If the birds gang up as planned, they would be fitted with GPS trackers and set free, by the thousand or so.

With enough funding, which Novak ballparks at about $US25 million, he thinks he could create a live passenger pigeon in the lab within seven years using CRISPR. Parallel work focused on breeding shows it would only take a few more years to build up a healthy sustainable population of 10,000 birds or so. That won’t be enough to make a dent in forests the way the sky-darkening flocks of the last century did. Still, Novak says, it’s a start.

But does that mean that monster pigeon swarms will start descending on cities like New York? Historically, Novak says, the birds stayed clear of urban centres as there was not enough food. “The bigger their flocks get, the more they will stay remote, near tree cover.” And, if things do get out of hand, he says we already know what to do: it was just a few decades of hunting that wiped out those billions of birds in the first place.

What if we crash the ecosystems we’re trying to save?

But suppose passenger pigeon flocks really are too much for American forests already scarred by record wildfires. Or that bizarre little frog becomes the next cane toad. Some have even wondered whether ancient viruses, entangled in the DNA of long-dead species, could be reawakened (cue the buzzing mammoth carcass in the TV sci-fi thriller Fortitude).

Archer, who himself was the first ecologist to sound the alarm on the danger of cane toads in Australia, says the fossil record can offer important clues as to how an ecosystem will fare with a reintroduced species. When he ventured into the Tasmanian bush with one of the last people to see thylacines in the wild, he found their habitat was broadly unchanged since the 1930s. Peter Ward, in his 90s on the hike, had trapped and hunted the tigers as a boy with his father and brother, back when there was a bounty on the marsupials’ head (due to now-debunked fears that thylacines were eating livestock). At the end of the track, Ward’s family hut was still there, just as he’d left it, tins of food still on the shelf. “Tears came into his eyes,” Archer says. “He even remembered what they sounded like. He said they’d make this yip yip yip sound as they circled the hut at night. The forest hasn’t moved on.”

Novak says ecosystems are not a house of cards. “When they have the right pillars – our keystone species like predators and pollinators and herds – then they’re more like a tower that an earthquake wobbles, but it doesn’t collapse.”

Besides, Church adds, the larger the animal, the easier a reintroduction is to reverse. Just as feral goats were removed from the Galapagos Islands, rounding up wayward mammoths wouldn’t be impossible.

But while thylacines could help with Australia’s feral cat problem, as dingoes do on the mainland, Archer says it’s not quite a wolves-in-Yellowstone situation. “They’re not going to be chasing the big animals.” In Tasmania, where there are no dingoes, thylacines would be competing for smaller prey with the Tasmanian devil. On the mainland, it might put pressure on the quoll too. These are both endangered species themselves, and so the thylacine’s impact would have to be closely monitored, Archer says, released as a trial in fenced areas first.

“But 99 per cent of the time, with some careful planning, what happens is what you intend. This is the part of conservation we already know how to do well.”

Indeed, for all the focus on worst-case scenarios, Novak says he could only find one instance of a conservation reintroduction backfiring, after analysing more than a century of US “rewildings”: when moving some endangered water birds into a wetland in 1988 caused others in the area to die off.

In America, regulators have now green-lit the world’s first release of a de-extinct species: a chestnut tree. Once the most abundant on the continent, the towering tree has been genetically engineered to survive the imported fungus that wiped it out eight decades ago. Some Native American tribes have even agreed to replant it on native land. Regulating the chestnut’s return was no easy task and Novak hopes it will now be a guiding light for future de-extinctions, though he admits “mammoths and pigeons are a whole different ball game to trees”.

But, just as gene-editing can bring back life, it can also end it. Gene drives hold awesome power to accelerate evolution and take out feral populations by spreading edits that disadvantage or kill a pest species quickly. Scientists have even proposed such an approach to tackle the mouse plague gripping Australia’s east. Novak says gene drives must be used carefully, but “sometimes the risk of doing nothing, whether that’s gene drives or de-extinction ... is actually a lot worse”.

Will de-extinction help conservation? Or are we playing God?

For scientists eyeing de-extinction projects, there are a lot of vacancies out in the wild in need of filling. Novak says the technologies being developed will benefit existing endangered species too as their gene pools narrow, from the black-footed ferret to the northern white rhino, pictured below. “The passenger pigeons and the mammoths, they’re our moon shots,” he says. “This is never going to replace traditional conservation.”

But others worry that critical funding will be taken away from on-the-ground recovery efforts and shunted into pie-in-the-sky de-extinction projects. Australia has some of the world’s highest extinction rates but spends a tenth of what the US does on conservation efforts. De-extinction, if proven to work, will still carry a higher price tag than traditional conservation. At this late hour, Ritchie says, funnelling more funding into proven methods is a safer bet.

Novak understands the concern but says funding for de-extinction projects so far generally comes from sources not already investing in conservation, such as big tech. “We’ve tried hard [at Revive & Restore] to get money from new places like biotech companies, even Facebook.” Since the not-for-profit was founded by conservationist Stewart Brand in 2012, Novak says about 90 per cent of the funding it’s raised has been spent on genetic rescue, such as their work with ferrets, not de-extinction. With the birth of Elizabeth Ann, he says those projects are advancing enough that they might begin to compete with traditional conservation. “But in conservation, we always fight for funding to re-introduce this species or that one. It’s always triage.”

Archer, who says money for his own frog project comes primarily from people interested in the technology rather than the frog, stresses that forcing a choice between de-extinction and conservation will crush innovation. In Australia, where life has evolved over the past 50 million years cut off from the other continents, he says the case for de-extinction is especially strong. “We have this added responsibility because our animals just don’t exist anywhere else. We’re a whole distinct branch of the global genome.”

Ritchie recalls the story of an Australian naturalist who took a taxidermied northern quoll into parts of the Northern Territory where it had vanished. “A local Aboriginal woman just held it, crying, when she saw it again. It was one of her totemic species, and that pain, that loss, was still so strong.”

Archer and Church say de-extinction could help end the doom and gloom of conservation, turning it around from an unwinnable war into something that could capture the public’s attention (and perhaps real funding). But will the path there be littered with ghastly mistakes, animals trapped in awful lives because of editing blunders?

“Genetic power’s the most awesome force the world has ever seen but you wield it like a kid who’s found his dad’s gun,” Jeff Goldblum’s character Dr Ian Malcolm warns in Jurassic Park.

“How can we stand in the light of discovery and not act?” counters the park’s creator John Hammond, played by Richard Attenborough.

Like Benjamin the thylacine, many of the species we have lost in the past century or two also died of the cold – of our indifference, our cruelty, our thoughtlessness. But does that mean we have a moral obligation to bring them back, as Brand says, to a world that misses them?

Also in this sci-fi explainer series ...

• ‘A numbers game’: Will we ever find aliens (and what are UFOs)?
• Curing cancer, designer babies, supersoldiers: How will gene-editing change us?
• Brain chips, cyborgs and the ‘singularity’: Will artificial intelligence rule the world?
• Beam me up, Scotty: Is teleportation possible? How about intergalactic travel?