How many species could go extinct from climate change? It depends on how hot it gets

To consider how climate change could cause some extinctions, imagine a tiny mountain bird that eats the berries of a particular mountain tree.

That tree can only grow at a specific elevation around the mountain, where it's evolved over millennia to thrive in that microclimate. As global temperatures rise, both the tree and the bird will be forced to rise too, tracking their microclimate as it moves uphill. But they can only go so far.

"Eventually, they reach the peak, and then there's nowhere else to go," says Mark Urban, a biologist at the University of Connecticut.

Scientists call this mountain phenomenon the "escalator to extinction" and it's just one way climate change is already squeezing plants and animals from their habitats. Researchers have conducted hundreds of studies projecting how different species might respond to different levels of climate change, finding varied results. In an analysis published Thursday in the journal Science, Urban sought to bring all those studies together.

"I wanted to get a better overall picture, to provide an answer to decision makers who wanted to know exactly how climate change would translate into extinction risk," he says. That picture is worrisome, he found, especially at higher levels of warming.

"Each of these species has encountered and solved life's problems. They're really the great books of knowledge on Earth," says Urban. "We really don't want to burn those books before we get a chance to read them."

If countries meet the shared goal of limiting warming to 1.5 degrees Celsius, 1.8 % of species will be at risk of extinction by the end of the century, Urban reports. But if global warming gets out of hand, warming four or five degrees Celsius, as many as 30% of species could be at risk.

"That 30 percent might be the best case scenario of the worst case scenario," says Cristian Román-Palacios, a biological data scientist at the University of Arizona who wasn't involved in the study.

He points to confounding complexities in how species might respond to such climate extremes that scientists don't yet know. More critters may simply not be able to cope, or ecosystems that lose species after species may collapse altogether. Additionally, many rare species are understudied, or not even discovered, and might be especially vulnerable in ways that don't show up in this analysis.

Still, "we need broad-scale studies like this," says John Weins, an evolutionary biologist at the University of Arizona. "If we want to stop the loss of biodiversity, we need to know what the threats are."

In many instances, stopping biodiversity loss means preserving their natural habitats, through protected areas or national parks. That won't necessarily work for saving species from climate change, says Weins.

"You can protect everything, you can stop all the destruction of rainforests, get everything in preserves, and you could still lose one third of species on Earth," he said. "It requires a more global solution, in terms of stopping carbon emissions."

The globe has already warmed by about 1.3 degrees Celsius from the preindustrial average, which would tip about 1.6 %of species towards extinction by 2100, Urban found. If emissions continue at their current trajectory, about one in twenty could blip out of existence. Beyond that, risk really accelerates, underlining the need for climate action, he says. At a more extreme scenario of 4.3 degrees of warming, nearly 15 % of species would be imperiled. And at 5.4 degrees of warming, 29.7 % of species could die out.

Different species face some different risks. Amphibians, including frogs and salamanders, are more vulnerable, Urban found, perhaps because their habitats are more sensitive to environmental changes. Species that live on islands, mountains and in freshwater could face more challenges, too. Targeted conservation efforts could help slow losses, Urban says, but they're ultimately no substitute for reducing emissions.

Scientists have described roughly 2 million species, but that's likely just a fraction of the biodiversity the planet supports. Estimates vary widely, but converge on something like 10 million species. According to Urban's analysis, that would mean the planet is already on track to lose about 160,000 species, and could see nearly 3 million disappear under the worst-case scenario.

"The study presents a clear choice to decision makers," says Urban. "Will we curb emissions now and work to protect that 2% of species that are currently at risk? Or do we choose one of these other paths that will fundamentally alter the nature of our world?"