The story of mammals is one of self-destruction. They first arose roughly 200 million years ago, and after eons spent scurrying in the shadow of the dinosaurs, they finally cut loose and evolved into a breathtaking variety of shapes and sizes, including the largest creatures to ever exist. And after all that, it took barely 100,000 years for one relatively young member of the group—us—to bring everything crashing down.

Throughout our existence, humans and other hominins have hunted other mammals, first for meat, and then for pelts, trophies, trade, and more. Since the last Ice Age, more than 300 species have gone extinct, including mammoths, woolly rhinos, and thylacines. A quarter of the remaining 5,500 species are endangered, thanks to one species: us.

A sobering new study by Matt Davis at Aarhus University throws these losses into stark relief. He estimated how long it would take for mammals to evolve enough new species to replace the ones that we have eradicated. And his most realistic answer is somewhere from 3 million to 7 million years. That’s at least 10 times as long as we have even existed as a species. We have inflicted such grievous wounds on our own family tree that the healing process can’t possibly happen “on any kind of time scale that’s relevant to humans,” Davis says.

Rather than simply counting the numbers of extinct and endangered species, Davis instead worked out how much evolutionary history they represent. This metric, known as phylogenetic diversity, matters because not all species are equal. Some are particularly unusual and irreplaceable.

The pygmy sloth, for example, may be one of the most threatened mammal species, but it’s also one of the youngest, having diverged from its closest relative 9,000 years ago. The aardvark, by contrast, is the last survivor of a once-large group of mammals that split off from the others 75 million years ago. Losing the pygmy sloth would be like snapping off a tiny twig from the mammalian family tree; losing the aardvark would be like sawing down an entire branch.

To work out the extent of these cuts, Davis and his colleagues first built a family tree for all mammals past and present, going back 130,000 years, into the late Pleistocene. By adding up the length of all the missing twigs and branches, they calculated that prehistoric humans robbed mammals of 2 billion years of unique evolutionary history. Since the 16th century, we’ve wiped out an additional 500 million years of evolutionary history, and we stand to lose a further 1.8 billion years within the next five decades. “It is staggering,” Davis says.

Indeed, our actions have been far more destructive than if we’d just killed off species at random. That’s because, as another group showed earlier this year, we have disproportionately targeted the largest species. There used to be giant ground sloths and car-size armadillos; they’re all gone. There used to be six species of elephant-like mammals in North America alone; now there are just three left in the entire world.

And “those big things were also the most evolutionary distinct things,” says Davis. “They were often on their own branches of the tree. We don’t see that pattern in previous mass extinctions.” According to him, humans have pulled off something close to the worst-case scenario for mammalian extinctions. We could barely have destroyed more phylogenetic diversity if we’d planned to.

When the past is this grim, the future is, too. Imagine that we instigate a massive, well-funded, global conservation push that effectively saves all existing mammals from extinction. Imagine also that all the survivors produce new species at twice their highest historical rates, on a par with the African cichlid fish that are textbook exemplars of extremely fast evolution. Even in this implausibly optimistic scenario, it would take half a million years for mammalian diversity to bounce back to its Ice Age zenith.

More realistically, given how fast mammals typically evolve, and given that some living species will inevitably go extinct, the full comeback will likely take 3 million to 7 million years to stage. “That puts us on the same scale as previous mass extinctions,” says Davis. “What we are going through now could have as big an impact as the asteroid” that killed off most of the dinosaurs.

But phylogenetic diversity is just one way to weigh up the loss of life. One could also look at functional diversity, which focuses on what animals do in their environment. Some play crucial roles as seed carriers, pollinators, and nutrient providers. The pygmy sloth might be a young species, but if it “performs a unique function in its ecosystem, its extinction can have dire cascading effects,” says Advait Jukar of the Smithsonian National Museum of Natural History.

These contributions are hard to study and measure, and Davis estimates that it will take even longer to replace them. That is, even after new mammal species evolve, they won’t necessarily step into the ecological vacancies that were created when past ones went extinct. And those vacancies themselves will change as the world warms and the environment shifts.

Regardless of these uncertainties, “it is hard to imagine that a full recovery or either phylogenetic or functional diversity can be achieved within human time-scales,” says Shan Huang of the Senckenberg Biodiversity and Climate Research Center. “But by prioritizing conservation for unique and distinctive lineages, we can at least slow down the losses.”

The Zoological Society of London has been saying as much for 10 years. Its edge of Existence program has been trying to focus attention on evolutionarily distinct species, including those that conservationists rarely pay attention to. Sure, save pandas and tigers, but also think about Attenborough’s long-beaked echidna, the Hispaniolan solenodon, and the Chinese pangolin. “Integrating evolutionary history into conservation planning will be essential in order to avert the loss of many millions of years of unique evolutionary history and all that entails,” says Nisha Owen, who manages the program.

Jukar adds that conservationists could also “map evolutionary distinctiveness onto our current network of protected areas” to see if they’re efficiently defending hotspots of diversity. “The goal for conservation should be to get the most bang for our (very limited) buck.”

“It’ll probably get worse, in all honesty,” says Davis. To avert the worst-case scenarios in his simulations, “we’re talking about a massive, ambitious global-scale project that everyone will need to be involved in. It comes down to whether politicians have the political will to make this happen.”