De-extinction: When Woolly Mammoths Return

High in the northern latitudes, grass is growing again. The moss is gone, has given way to something much more productive, where nutrients cycle and carbon is absorbed, where permafrost can penetrate deeper underground. If it was easy to see the ecosystem as lifeless before, that mistake is impossible now: look at the breeze ripple over the grasses, look at the bison roaming, look at the woolly mammoths. 

This is one version of our future, the one championed by biotechnology company Colossal Biosciences and other groups researching de-extinction. For the uninitiated, de-extinction is the emerging science of trying to “bring back” animals that have gone from the Earth. Colossal is leading the charge to genetically engineer a woolly mammoth-like creature, and other labs around the world are trying to pull even more creatures out of the abyss of extinction: Tasmanian tigers, passenger pigeons, and aurochs, the ancestor of today’s cattle. 

First things first: what do we mean when we say de-extinction? You are probably picturing artistic renderings of mammoths: giant shaggy creatures with sloping backs, beady eyes, thick legs, and long, curved tusks. This is, almost certainly, not what we will get.

What we will get is more complicated. For something that has been extinct as long as a mammoth—between roughly 10,000 to 4,000 years—it’s likely impossible to create an exact genetic clone. Instead, Colossal proposes a complex process of genetic engineering to give modern-day elephants certain characteristics of the woolly mammoth. They plan to edit the elephant genome to resemble mammoth genomes pieced together from ancient DNA, creating hybrids that will likely resemble fatter, hairier elephants. 

Of course, we’d be introducing mammoth-like creatures to a world that is much different than the one their ancestors left. Climate change has already increased global temperatures by an average of about 1.1 degrees Celsius (nearly 2 degrees Fahrenheit) since pre-industrial times, and human impact has encroached everywhere. Mammoths lived through many different climate regimes when they were alive, but ecosystems around the world are collapsing. Even if they could withstand temperature changes, could they withstand us? 

The question about what would happen if—and whether we even should—release those new, engineered animals back into the wild is often tucked after long investigations of the how and why of de-extinction. But some groups—like Colossal—forefront ecology as a reason to bring back the mammoth. Reintroducing megafauna will, they claim, slow the melting permafrost, store more carbon in the soil, and aid in elephant conservation. 

What’s more, they believe that engineering a new mammoth is the revival of an entire lost ecosystem: the mammoth steppe. 

About 20,000 years ago, the mammoth steppe was the most extensive biome on Earth, stretching from China to Spain to Canada. But the ecosystem—a cold, dry grassland full of reindeer, bison, horses, and woolly mammoths—disappeared from this Earth along with the mammoth. 

Jacquelyn Gill, a paleoecologist at the University of Maine, has spent much of her career looking at why that happened. Sifting through fossilized feces to find fungal spores, she’s put forth a lot of evidence that those ecosystem changes were the result of mammoth extinction, not the cause of it. 

“Ecosystems collapse because of extinction of large herbivores, not the other way around,” Gill said.

Across the globe, many of Earth’s largest creatures—called megafauna—went extinct between 50,000 and 10,000 years ago, an event often dubbed Late Pleistocene extinctions after the epoch during which they disappeared.

Most mammoths went extinct about 10,000 years ago, likely due to a combination of climate change and human hunting. In two studies focusing on the American Midwest, Gill found that after the extinction of the mammoth, the plant communities completely changed. Comparing that data to other megafaunal extinctions in Australia and Madagascar, she found similar trends. This change in vegetation may also have caused more fires as all that plant matter megaherbivores used to eat could then dry up and serve as fuel. 

There are other, more speculative changes, too: the end-Pleistocene extinctions may have affected ecosystem processes, making ecosystems less productive and limiting the ability of certain plants to spread their seeds. Mammoth extinction may even have warmed the Earth. 

The idea of putting a mammoth-like species back in the Arctic rests on the idea that, in returning megaherbivores, we can reverse the changes that happened 10,000 to 15,000 years ago. If their loss warmed the Earth, could their return help cool it?

The Arctic warms faster than the rest of the globe, and at 3 degrees Celsius (5.5 degrees Fahrenheit) of warming, we’re already seeing the permafrost thaw. The loss of permafrost is a big threat—it stores a lot of carbon, which would be released if it melted. If mammoths returned to the Arctic, the theory is that they could stomp down the winter snow, allowing the cold air to filter underground and cool the permafrost by a few degrees. 

Meanwhile, mammoths would be supporting a healthy grassland ecosystem. Grasslands are lighter in color than forests so will absorb less heat than the small, stunted arctic trees that currently live there. Grasslands also aren’t as susceptible to forest fires, and their extensive root systems store more carbon than current arctic vegetation.

In an experimental stretch of land in the Arctic called Pleistocene Park, a team of scientists are working to bring grasslands back to the Arctic. In the roughly 35,000-acre reserve (whose name echoes the infamous Jurassic Park), scientists are introducing large grazing animals to reverse the ecosystem shift that followed their disappearance. Since Yakutian horses were introduced in 1996, 10 other species have been successfully introduced, some native to the region, some not. They exist now with varying levels of human support—some (like sheep) get extra forage in the winter while others (like moose) live independently from humans. 

One day, maybe the park will be able to add in a mammoth. 

The project was founded in the late ’90s by Russian geophysicist and ecologist Sergey Zimov but is currently helmed by his son, Nikita Zimov, who grew up in the Northeast Science Station in Cherskii, Russia. Nikita likens the loss of the mammoth to losing a right arm: the ecosystem can survive, but everything will be a little bit more difficult. 

A mammoth, too, would have a hard time if you plopped it down in the middle of the Arctic, considering all the changes that have happened since its extinction. The current inhabitants of Pleistocene Park serve to prepare the ecosystem for its arrival, fostering a grassland that a mammoth can feed on. 

To Nikita, it doesn’t really matter whether or not the engineered creature is a full mammoth; it only matters that the creation fills the same role in the ecosystem. His goal with de-extinction is aligned but separate from Colossal’s: not tinkering with a genome but tinkering with an environment. 

“They’re creating the mammoth,” Nikita said. “We’re creating the home for the mammoth.” 

De-extinction entered the public sphere in 2013 after the National Geographic Society convened a group of scientists and ethicists to discuss it. The meeting was the first time such a wide range of people gathered on the topic of de-extinction—from geneticists to conservationists to wildlife biologists—but the idea had been kicking around for a long time. Many conversations around de-extinction end up back at the same source: the 1990 science-fiction novel Jurassic Park. 

No one, at this point, is proposing bringing dinosaurs back, and it’s unlikely we ever could. The dinosaurs went extinct about 65 million years ago, so it’s highly unlikely we’ll find intact cells or enough of their DNA to reconstruct a genome. All the current methods proposed for de-extinction require that the creature went extinct much more recently. 

There are three main kinds of de-extinction: cloning, back-breeding, and genetic editing. The first, cloning, is what most people picture: an exact replica of a long-gone species once again walks the Earth. Cloning requires well-preserved cells, likely from a species that went extinct relatively recently. The second, back-breeding, requires a living descendant of an extinct species that you can carefully breed to resemble its extinct counterpart. In South Africa, researchers are trying this method to create a version of the extinct quagga from the plains zebra. The third, genetic editing, is the method proposed to bring back the mammoth. 

Genetic engineering an animal back into existence requires a sequenced genome of an ancient creature and a living relative of that species. In the case of mammoths, we have the former taken care of. We’ve got the latter, too: elephants. The trouble is: how do you turn an elephant into a mammoth?

Colossal proposes using CRISPR-Cas9 gene editing technology. CRISPR works a bit like genetic scissors, cutting out elephant genes and pasting in mammoth ones. Ideally, you’d edit every single one of the differences to recreate the mammoth genome in full—but that’s a lot of changes. The woolly mammoth and Asian elephant genomes differ by about 1.4 million base pairs, the building blocks of DNA. 

With current technology, we’re limited in how many genes we can edit at once—the most done in peer-reviewed scientific studies is 33 (in research co-led by Colossal cofounder George Church). Editing all 1.4 million base pair differences (which are spread across many different genes) would take an unfeasibly long time in 33-gene increments, so Colossal is working hard to improve gene-editing tech to the point where we can edit hundreds to thousands of genes at once. In the meantime, the group is focusing only on the mammoth traits deemed most important—namely, cold resistance.

Scientists plan to start with an Asian elephant somatic cell—anything that isn’t an egg or sperm. They’ll use the CRISPR tech to edit that to include key mammoth traits (shaggy hair, more fat, cold-adapted blood) and then take that edited nucleus and insert it into an Asian elephant egg cell. Then, they’ll zap the egg to simulate fertilization, watch as an embryo forms, and implant that embryo into an African elephant. (The African elephant is more distantly related to the mammoth but has more stable populations than the Asian elephant, so there is less risk of adding stress to an already-endangered species.) Then, after the 22-month pregnancy, a mammoth-elephant hybrid is born. 

Almost every step of that process presents technical challenges. First, there’s figuring out what genes to edit. Colossal wants to edit over 50 traits to increase cold resistance, but mapping traits to genes is hard work. Even if you can find the genes you think correlate to a specific trait (say, the changes needed to give an elephant shaggy hair), there is no guarantee that those genes don’t interact in surprising ways with other genes, some of which may be far away on the genome from what you’re editing. Then, the process of nuclear transfer is often unsuccessful as it can cause side effects like an abnormal placenta or negative health outcomes once the offspring is born.

After all that, you have to wait nearly two years for a pregnant elephant to give birth to figure out if the process worked. 

Even if you get the elephant as close to a mammoth as possible, it will never be an exact replica: its elephant mother will be raising the hybrid creature with all of her elephant ways. Besides nurture, both the mother and baby’s environment and diet will shape the creature in ways science is still working to understand. By and large, the mother determines the microbiome, and processes like genetic function and activation depend on signals received in the womb and, later, from the environment. 

If you talk to anyone working on de-extinction, they will readily tell you that what we’re creating is never going to be an exact replica of a mammoth. But you can understand the public’s confusion. When you click the “woolly mammoth” tab on Colossal’s website, it proclaims in all caps against a vivid purple background: “WE HAVE THE DNA, THE TECHNOLOGY AND THE LEADING EXPERTS IN THE FIELD. NEXT, WE WILL HAVE THE WOOLLY MAMMOTH. ALIVE AGAIN.” Only through scrolling is it clarified that it will be a “cold-resistant elephant with all of the core biological traits of the Woolly Mammoth.”

Even with great technological advances, it’s hard to count on us “bringing back” a faithful replica of a mammoth. The project, then, rests on the ecological case for these hybrids.

Ben Novak, lead scientist at conservation and biotechnology group Revive & Restore, advocates that we look at “de-extincted” creatures as environmental proxies. The point is not to create an exact replica, but to create something that will fill the same ecological niche as the lost creature. This is not a completely foreign idea in conservation: it’s what’s behind the rewilding movement. 

Novak points to the case of the giant tortoises on Mauritius, which went extinct in the mid-1800s after humans arrived. Over a century and a half later in the late 1990s, a group of scientists began releasing Aldabrachelys gigantea on the islands. The tortoises, which come from the Aldabra Atoll of Seychelles, share a common ancestor with those from Mauritius but developed for millions of years in isolation, becoming completely different species. 

Despite this evolutionary divergence, the role that the Aldabra tortoises fill ecologically is nearly identical. So far, over 20 years out, the experiment seems to be doing well: the Aldabra tortoises are breeding, dispersing seeds, and grazing grasses on the island, slowly restoring some of the functions that were lost with the native giant tortoise. 

Reintroductions (putting species back where they used to be) are controversial, especially when you increase the size of the animal and its proximity to humans (see: gray wolves in Colorado). Introductions (putting in a different species to fill an ecological niche) is even more so and regarded by some as over-intervention, or “playing God.” 

The debates on this subject are intensely political and get into some thorny questions at the heart of conservation, including how exactly you define the boundaries of a species. In the case of putting a de-extincted animal back in the wild, it is often framed as a reintroduction (“bringing back the mammoth”) when it’s more aligned with introduction: inserting a new, similar species with the hopes of revitalizing an ecosystem. 

Adam Searle, a geographer and postdoctoral fellow at the University of Liège in Belgium, spent his Ph.D. investigating the case of the bucardo, the one example we have of de-extinction actually working, if only for a few minutes. In 2003, scientists in Spain successfully created a clone of a bucardo, also known as a Pyrenean ibex, a large wild goat that humans hunted to extinction. Suffering a lung deformity, the creature died seven minutes later. 

Then, in 2014, something else happened that complicated the effort to de-extinct the bucardo. Members of a different subspecies of ibex were introduced to the French Pyrenees. It was thought, for a long time, that no other kind of ibex could survive in the bucardo’s range—it was too high and too cold. But the Iberian ibex are doing quite well. Scampering between the lush green valleys and the naked rock of the mountain tops, these translocated species are venturing down into places that the bucardo abandoned years ago. In this way, they resemble a historical version of the bucardo, one that existed before it came to fear humans. 

Whether you think the Iberian ibex belongs in the French Pyrenees depends on what lens you look through. Genetically, they will never be the same as the bucardo nor will they look quite the same. Though a casual observer will have a hard time telling them apart, the Iberian ibex have slightly smaller horns and a slimmer body. But morphology is mutable, and already the French Pyrenees is leaving its mark on the Iberian ibex: with each generation, they become more adapted to the winter. They are growing fatter and stronger. Ecologically, the two species are essentially identical—they fill the same role.

In the face of this successful translocation, the effort to clone the bucardo has effectively died. Why clone one single female that will likely be raised in captivity when there is an ibex thriving in her absence? This, as with de-extinction as a whole, raises a number of deep, unanswered questions that biologists have been wrestling with for ages. What matters in a species? Is it the precise line-up of genetic information? Or is it the role that it plays in the place where it lives? 

Driving so much of the hype around de-extinction is the idea that something is lost from the world when a species is gone, Searle told me. The question becomes: what can fill that hole?

De-extinction raises a lot of big questions in conservation—from the most basic (how do we define a species?) to the most existential (what role should humans play in the natural world?). For a long time, the biggest question the media have been asking has been: should we be making these creatures?

At this point, that question may not be useful. We are well on our way to creating a proxy mammoth, passenger pigeon, and Tasmanian tiger. Everyone from Peter Thiel to Paris Hilton to the CIA has invested in Colossal Biosciences (as of May 2022, the group’s total funding was a whopping $75 million). Some believe the science will advance genomic technology; others just hope to see a mammoth. 

And who wouldn’t? Mammoths are cool. This is something scientists and ethicists often shared about the de-extinction question. This is one of the clearest reasons to create a mammoth-like creature: it’d be awesome.

All of the ecological benefits of mammoth de-extinction remain theoretical. Backed by science, sure, but they have not yet been tested since we have not had a mammoth to introduce. They also require not just a single, lonely mammoth calf but a whole herd. Besides the fact that we still don’t know much about mammoth ecology—how many there were and how much land they need—how long will it take to create enough mammoths to make a healthy population? Once we do, how long will it take for the climate benefits to materialize? Do we have that kind of time?

There aren’t clear answers to any of these questions. We are still years away from having the first draft of a mammoth, let alone a whole herd. Introducing the mammoth to the Arctic, ensuring that it survives and breeds, protecting it from poachers and climate change—this part of the process looks a lot like conservation. On that, we have a pretty mixed track record. 

Elephants are endangered because of humans. Habitat fragmentation and human development are encroaching on elephant territory, and poaching is a major threat to all three elephant species. The Arctic may be far less populated, but we have already proven that no ecosystem is beyond our reach. If elephant ivory is valuable, how much could a mammoth tusk go for?