Imagine if every animal and plant on the planet collapsed into a single population each, says ecologist Gerardo Ceballos. If lions disappeared except from one small corner of Kenya, the prey they keep in check would run amok everywhere else. If sparrows were no more except in one Dutch forest, the seeds that sparrows disperse would stay in place everywhere else. If honeybees became isolated to one American meadow, the flowers that they pollinate would fail to reproduce everywhere else. None of those species would be extinct per se, “but we’d still be in very bad shape,” says Ceballos.

He uses this thought experiment to show that fixating on the concept of extinction can lead scientists to overestimate the state of the planet’s health. Extinction obviously matters. If a species is completely wiped out, that’s an important and irreversible loss. But that flip from present to absent, extant to extinct, is just the endpoint of a long period of loss. Before a species disappears entirely, it first disappears locally. And each of those local extinctions—or extirpations—also matters.

“If jaguars become extinct in Mexico, it doesn’t matter if there are still jaguars in Brazil for the role that jaguars play in Mexican ecosystems,” says Ceballos. “Or we might able to keep California condors alive forever, but if there are just 10 or 12 individuals, they won’t be able to survive without human intervention. We’re missing the point when we focus just on species extinction.”

He and his colleagues, Paul Ehrlich and Rodolfo Dirzo, have now tried to quantify those local losses. First, they analyzed data for some 27,600 species of land-based vertebrates, and found that a third of these are in decline. That doesn’t mean they are endangered: A third of these declining species are listed as “low concern” by the International Union for Conservation of Nature, meaning that they aren’t in immediate peril. But that, according to Ceballos’s team, provides a false sense of security. Barn swallows, for example, still number in the millions, but those numbers are going down, and the birds are disappearing from many parts of their range. “Even these common species are declining,” says Ceballos. “Eventually, they’ll become endangered, and eventually they’ll be extinct.”

The team also analyzed detailed historical data for 177 species of mammals. In the last century, every one of these species has lost at least 30 percent of its historical range, and almost half have lost more than 80 percent. Consider the lion. If you divide the world’s land into a grid of 22,000 sectors, each containing 10,000 square kilometers, around 2,000 of those would have been home to lions at the start of the 20th century. Now, just 600 of them are. These royal beasts, which once roamed all over Africa and all the way from southern Europe to northern India, are now confined to pockets of sub-Saharan Africa, and a single Indian forest. Their numbers have fallen by 43 percent in the last two decades.

Several other species that were once thought to be safe are also now endangered. Since the 1980s, the giraffe population has fallen by up to 40 percent, from at least 152,000 animals to just 98,000 in 2015. In the last decade, savanna elephant numbers have fallen by 30 percent, and 80 percent of forest elephants were slaughtered in a national park that was one of their last strongholds. Cheetahs are down to their last 7,000 individuals, and orangutans to their last 5,000.

All told, “as much as 50 percent of the number of animal individuals that once shared Earth with us are already gone, as are billions of populations,” Ceballos and his colleagues write. “While the biosphere is undergoing mass species extinction, it is also being ravaged by a much more serious and rapid wave of population declines and extinctions.”

Strong words, but “one has to be incredibly cautious to not be alarmist,” Ceballos tells me.

I point out that the paper in which he describes his new findings begins with the words “biological annihilation,” which sure sounds a little alarmist.

“It would be alarmist if we didn’t have the data,” he counters. “Now it would be irresponsible on our part to not use strong language. I wish we could say we are wrong but unfortunately, this is what is happening.”

These claims feed into a broader debate about whether the Earth is in the middle of a new mass extinction—a crisis in which the majority of living species blink out of existence. There have been five such catastrophes in the planet’s past, the most famous and recent of which wiped out most of the dinosaurs. Scientists like Ceballos, Ehrlich, and Dirzo argue that humans are kicking off the sixth such event.

That seems intuitively right, given what Ceballos and others have found. But other scientists—including many paleontologists who actually study the historical big five—think that the current crisis, though severe, comes nowhere close to those ancient Ragnaroks.

Their argument, as Peter Brannen recently recounted in The Atlantic, is that people underestimate how bad the earlier mass extinctions truly were. The worst of them—the so-called Great Dying at the end of the Permian period—claimed around 97 percent of all species on Earth. And all five mass extinctions killed off not just charismatic megafauna—the equivalents of lions and elephants—but hardy, ubiquitous species like insects and plants. Critically, they finished off widespread species like clams and hard-shelled plankton whose relatively sudden absence is plain to see in the fossil record.

But Ceballos argues that such species are under threat. Although he only looked at mammals and other vertebrates, other researchers have also documented striking declines among insects, snails, and more.

He also says that it’s a question of rates. Extinction may be a natural part of life, but in 2011, one team showed that mammals are going extinct at 3 to 80 times their usual pace. Four years later, Ceballos and his team confirmed that that number of vertebrate species that we’ve lost in the last century would normally have taken between 800 and 10,000 years to disappear. We may not be in a full-blown mass extinction yet, but we’re certainly in the earliest stages. “At the end of this, we’ll probably be able to say, yes, we were in the sixth mass extinction and we lost 75% of everything,” he says. “But we don’t have to wait until the end.”

The problem with this reasoning, says Doug Erwin from the Smithsonian Institution, is that it misleadingly compares what happened in the last century to what happened over the last several million years. “Any comparison of a long-term rate to a short term one almost inevitably comes up with a higher short-term rate,” he says, because the longer intervals average out a lot of fluctuations, while “short-term rates are generally only calculated when they are high.” Erwin isn’t downplaying the magnitude of the current crisis, but he says that Ceballos isn’t comparing like for like.

It’s an odd debate because both sides have a lot of common ground. Erwin says that mass extinctions are huge, runaway, snowballing events; if we were already in one, it’d be too late to do anything about it. Ceballos says that it’s important to recognize that we’re in the early stages, precisely because we still have a small window in which to avert it. They’re both ultimately saying the same thing, even though they disagree about how to label the emergency.

“The real trouble with mass extinctions, from a modern perspective, is that it's really hard to know you're in one before it’s too late,” says Jacquelyn Gill from the University of Maine. “By the time you compile the casualty list, the damage is done. What's really powerful about [Ceballos’s new] study is that it focuses not on the losses, but on the early warning signals. Population declines are a common precursor to extinction, and it's a process we can actually do something about.”

“At the end of the day, protecting biodiversity is the goal,” she adds. “Even if this isn't a mass extinction, we're clearly still losing species that we care about. The loss of the white rhino hurts even if wasn't geologically superlative.”