Skip to main contentSkip to navigationSkip to navigation
A grizzly bear and a polar bear
Grizzly bears have begun interbreeding with polar bears after moving north into cooler habitats. Composite: The Observer/Michael Duva/Getty, Peter Hinch/Getty
Grizzly bears have begun interbreeding with polar bears after moving north into cooler habitats. Composite: The Observer/Michael Duva/Getty, Peter Hinch/Getty

Can cross-breeding protect endangered species from the climate emergency?

This article is more than 1 year old

Creating hybrids might make animals more resistant to global heating, but it may mean losing species altogether


The Houston Ship Channel is a rubbish home for a fish. It’s one of the busiest ports in the world and all that traffic has made the water slick with toxic chemicals. Yet the Gulf killifish has found a way: it has evolved pollution resistance by cross-breeding with a different species, the Atlantic killifish, which happened to have a handy mutation.

Cross-breeding, or hybridisation, is more common in nature than we used to think and as global heating makes animals move to areas with lower temperatures, more species may get thrown together. In Alaska and Canada, people have already spotted grolar bears, the result of grizzlies moving up into polar bear territory to escape the heat.

Recently, conservationists have proposed that we could cross-breed animals for their own good. They think that, like the Gulf killifish’s new pollution defence, hybridisation could give vulnerable animals an evolutionary head start in the race to genetically adapt to global heating. Their tolerance of higher temperatures and acidic oceans could improve.

Others are wary of losing millennia-old species as they are mixed with different animals. This debate reveals a chasm within conservation, revolving around the question: can we protect animals while forcing them to change?

A captive-bred grolar bear, or pizzly. Photograph: Reddit

“Lots of things hybridise all the time,” says Michelle Marvier, a conservation biologist at Santa Clara University in California. It happens in plants, fish, amphibians and even some mammals. In fact, many of us carry traces of Neanderthals and Denisovans in our DNA, proof that we mixed with other human species. “It can be something that leads to an evolutionary dead end because the offspring are sterile or it could be halfway to evolutionary adaptation,” she says.

It’s the second aspect that is garnering interest. For species that have a long time between generations, there are fewer opportunities for potentially useful mutations to arise, so if the environment is rapidly changing, normal evolution will probably be too slow for those animals to adapt and survive. Hybridisation could provide a shortcut by quickly bringing in genes from outside the normal gene pool.

‘When you cross-breed different species, the motivation really is to create new gene combinations and increase genetic diversity,” says Madeleine van Oppen, an ecological geneticist at the University of Melbourne. That diversity increases the potential for new adaptations that could save species from extinction caused by global warming, just like the Gulf killifish was saved from pollution by the Atlantic killifish’s mutation.

Van Oppen works on corals, half of which have disappeared over the last 30 years, in large part due to global heating. By crossing corals in the lab, she creates new hybrid offspring and tests them to see how they hold up under warmer conditions. She and her colleagues recently showed that some coral hybrids survived up to 34% better at higher temperature and CO2 pressure than their parents.

But not all species can be bred and tested in the lab. Instead, conservationists could move one species into the habitat of another and hope they breed. While no one has yet tried to make climate-resistant hybrids this way, the approach has been used to combat inbreeding in species with only a few individuals left.

A European wildcat in Glen Feshie in the Scottish Highlands. Photograph: Mark Hicken/Alamy

The Florida panther is a key example. In the mid-1990s, only about 25 remained and scientists thought they would become extinct within two decades. As a last-ditch attempt to save them, conservationists moved eight Texan panthers into their habitat to boost the genetic diversity. “But not without a lot of debate and angst over what that would mean,” says Marvier. Thirty years later, there are still panthers in Florida, but “is it still the Florida panther if you brought in panthers from Texas to interbreed with them?”

The worry with forcing a threatened species to hybridise is that, rather than saving it, you do the opposite: make it go extinct. Its genome no longer survives in its original form. The genes of the newcomers eventually swamp those of the original inhabitants, leaving no trace of what once made them unique.

This is what faces some of Europe’s wildcats. In Scotland, only a few hundred remain and a 2019 report by the International Union for Conservation of Nature found that the main threat to them is cross-breeding with domestic cats. What is now roaming the Scottish Highlands, the report says, is primarily a “hybrid swarm” of wild cats with different degrees of domestic hybridisation.

The situation looks increasingly similar in the Swiss Jura mountains, says conservation biologist Juan Montoya-Burgos of the University of Geneva. Wildcats having a few domestic genes “is not a big problem if they are able to survive and interact and play their ecological role”, he says, but he and his colleague Mathias Currat have shown through modelling that, as it stands, Jura wildcats will become genetically indistinguishable from domestic cats within a century.

How they will fare then is anyone’s guess, but we do know that extensive hybridisation can result in loss of adaptation to the local environment. In the Rocky Mountains of North America, this appears to be happening to the cutthroat trout, an animal so symbolic it is the state fish of seven US states.

“Native trout have adapted to persist in the face of extreme environmental change over time: you know, flooding, wildfires, glaciation,” says aquatic ecologist Clint Muhlfeld from the US Geological Survey. But as they breed with the invasive rainbow trout, released by the millions for fishing stocks during the 20th century, the collection of genes that together are responsible for those adaptations are broken up.

The cutthroat trout of the American Pacific north-west. Photograph: Alamy

This, potentially along with bad new gene combinations, has detrimental effects further down the generational line. In cutthroat trout, hybrids produce fewer and fewer offspring the more rainbow trout ancestry they have. “With as little as 20% hybridisation, you see at least a 50% decline in fitness and we’ve seen this pattern in different populations,” Muhlfeld says.


Being wary of human-facilitated hybridisation, then, is understandable. But Marvier argues that such dire scenarios are rare. “Most introduced species are perfectly benign,” she says. In a recent review, she and her colleagues found that while many studies highlight “invader” hybridisation as a threat to native species, few produced actual evidence of harmful effects, such as poor growth or fertility. The paper argues that if a native species is changed, but in an adaptive way that works within its ecosystem, it should not necessarily be seen as a loss.

Not everyone agrees. “In the field, I marvel at the peculiar and highly idiosyncratic adaptations that [different species] have evolved,” says ecologist Daniel Simberloff at the University of Tennessee. “I just cannot be cheered by the sort of thing that [Marvier and others] are talking about as a replacement.” He is not alone. The last thing we want is to have so much gene flow that we lose that distinctiveness of the species we are trying to protect, says Karin Pfennig, an evolutionary biologist at the University of North Carolina.

From Van Oppen’s perspective, ideally we would maintain all species exactly as they are, but we don’t “have that luxury any more because the environment is changing so rapidly and biodiversity loss is just so fast”. In other words, if we can’t save the ones we have, maybe we can help nature make new ones that are more likely to survive.

Such trade-offs do occur in nature. For example, female Plains spadefoot toads living in the New Mexico desert prefer to mate with a Mexican spadefoot over their own kind, but only if their pond is drying out. Hybrid tadpoles develop faster and have a higher chance of reaching adulthood before the water disappears. It’s a hard choice, though, because only the hybrid daughters are fertile.

Pfennig, who studies spadefoots, suspects hybridisation played a large part in why the Plains spadefoots moved from their ancestral grasslands to the desert, because it “can allow species to move into new habitats that they otherwise wouldn’t be able to live in”. It is not a big leap to think it could do the same for species finding themselves in new or drastically altered environments due to global warming.

Yet adapting genomes using hybridisation “is a really blunt tool”, says Simberloff. “It’s like using a sledgehammer in surgery,” he says, whereas genetic editing would give us much finer control. He highlights a plant case, that of the American chestnut, which was widespread until an accidentally imported blight-causing fungus infected them.

People have tried for a while to hybridise the American chestnut with the resistant Chinese chestnut. “They grow up to be about this high,” he says with a hand held next to his chest, “and then the blight gets them.” But scientists have now managed to fit the American chestnut with a wheat gene that gives them resistance. This, in Simberloff’s surgery analogy, is the scalpel.

A female Plains spadefoot toad mating with a male Mexican spadefoot toad. Photograph: Catherine Chen/davidpfenniglab.com

But evolutionary biologist Andrew Whitehead, from the University of California, Davis, who studies the Gulf killifish, does not see this as a viable option for most species. He says we know too little about how changing or inserting genes will affect animals once they are out in nature. “I think we’re fools to think that we can genetically engineer a future-proof creature.”

Yet evolution will also not do the job for us unaided, he says. People take the story about the Gulf killifish to mean that evolution will provide solutions to pollution and global warming, “but that’s the exact wrong take because it is such an exception.” His own solution is both simpler and, as the UN climate change conferences show, more complicated. “How about burning less fossil fuel? We know that’s gonna work.”

It’s hard to disagree. Protecting all but the most charismatic and important species one by one is impossible, says Montoya-Burgos, regardless of the tools we use. “We need to act at the source of the problem, which is climate change.”

While that is undoubtedly true, some damage has already been done and will be hard to reverse. Just take Florida. Its panther population originally plummeted due to road accidents and disappearing prey, but fixing that might not even matter in the long term. “They’re still not going to make it, probably,” Simberloff says, “because Florida is sinking.” This takes us right back to that difficult question: if we cannot save them as they are, should we try saving them as something else?

Comments (…)

Sign in or create your Guardian account to join the discussion

Most viewed

Most viewed