Around 66 million years ago, the sky fell on the dinosaurs’ heads. An asteroid smashed into the Yucatan Peninsula, causing cataclysmic climate changes that marked the end of the Cretaceous period, and killed off some three-quarters of animal species. A small proportion of hardy birds survived, but the other dinosaurs went extinct.

They were, however, already in decline.

Manabu Sakamoto from the University of Reading has shown that dinosaur species were going extinct faster than new ones were appearing, for at least 40 million years before the end of the Cretaceous. The dinosaur opera had already been going through a long diminuendo well before the asteroid ushered in its final coda.

Many other researchers had looked at the fates of the dinosaurs before that infamous extinction event and suggested that they were already declining. But most of these studies had simply tabulated raw numbers of species from different blocks of time. This approach has problems: the rocks from certain time periods may simply be better at preserving fossils, or may have been more intensively scrutinized by fossil-hunters.

After a recent attempt to adjust for these biases, using up-to-date information and better statistical techniques, Stephen Brusatte from the University of Edinburgh found “no evidence for a progressive decline in total dinosaur species richness.” Two groups—the horned ceratopsians and the duck-billed hadrosaurs—were fading in both number of species and variation in body shapes, but the others were not. “Recently, the idea that the dinosaurs were reigning strong has dominated the academic debate,” says Sakamoto.

He begs to differ. Together with Michael Benton and Chris Venditti, he took a recently published family tree, comprising 614 dinosaur species, and modeled the rates at which new species arose and old ones went extinct. “We’re not counting numbers of species throughout the history of dinosaurs, but of speciation events,” he explains. (A speciation event occurs when one species diverges into two.)

The titanic long-necked sauropods went through the biggest downturn. They experienced a burst of speciation in the early Jurassic period that culminated in the late-Jurassic arrival of iconic giants like Brontosaurus and Diplodocus. But by 114 million years ago, they were losing species faster than they could replace them. Even the arrival of the record-breaking titanosaurs couldn’t compensate for this evolutionary recession.

The meat-eating theropods went through a similar, but less pronounced, rise and fall. The Cretaceous period was known for its extremely diverse range of theropods, including sickle-clawed dromaeosaurs like Velociraptor, pot-bellied and scythe-clawed therizinosaurs, ostrich-like ornithomimosaurs, and mighty tyrannosaurs. But many of these lineages originated much earlier in the Jurassic; during their Cretaceous heyday, they were also going extinct faster than they were speciating.

Ironically, the only exceptions to this pattern were the ceratopsians and hadrosaurs—the two groups that Brusatte’s analysis showed were in decline. These plant-eaters radiated into species with very similar builds, but with subtle variations in skulls and teeth that allowed them to exploit different sources of food.

Then again, scientists may have overestimated the number of species in these two groups, says Susannah Maidment from Imperial College London. For example, there’s a fierce ongoing debate about whether Torosaurus was actually an older version of Triceratops. “Over-splitting of this group relative to other dinosaur groups could produce a false picture of high speciation rates at the end of the Cretaceous,” says Maidment. “The method used in [Sakomoto's] study requires us to have an accurate dinosaur evolutionary tree, and although we are very happy with the major branches, the arrangements of the twigs at the end is still debated and constantly undergoing change with each new dinosaur discovered.”

Even if the ceratopsian and hadrosaur trees are right, these groups represented just one seventh of total dinosaur diversity. For the dynasty as a whole, extinction rates were surpassing speciation rates for roughly 48 to 53 million years before the asteroid impact.

The causes of the slowdowns are unclear, says Sakomoto. The Cretaceous was a time of severe environmental change, with intense volcanic activity for tens of millions of years and fluctuating sea levels. “I think that dinosaurs were probably under stress for a very long time,” says Sakomoto. And their prolonged downfall left plenty of space and time for competitors to diversify in their place, including those pesky mammals.

“This isn’t to say that the dinosaurs were going extinct before the asteroid impact, but they were getting more vulnerable and susceptible to mass extinction says Sakomoto. And that’s relevant to us in the modern world. “We are putting a lot of pressure on modern species, and extinctions are happening at an unprecedented rate. If some kind of catastrophe occurs, it might be even more damaging than what we’re observing right now.”

“I love seeing big datasets and new methods thrown at some of these classic mysteries,” says Brusatte. “The result seems very robust, but I question what it means. Does that mean that dinosaurs were doomed to extinction, that they endured some kind of long death march before the asteroid impact finished them off, like a boxer knocking down their opponent with a light punch after several rounds of pummeling? I don’t think so.”

He suspects that the results say more about dinosaur booms than dinosaurs busts. They emerged during the Triassic period and after an extinction event killed off most of their competitors (like the obscure but highly successful crurotarsans), they evolved like mad during the Jurassic. After that burst, a slowdown was inevitable. “That doesn't mean the economy is necessarily doomed; it just means things aren't growing as insanely fast anymore,” says Brusatte. He suspects that you’d see the same pattern in other groups that suddenly rose to power, including mammals like us.

“I think we sometimes have a tendency to overthink the dinosaur extinction, myself included,” Brusatte adds. “The way I see it, it came down to the asteroid. Simple as that. Diversity declines may have made dinosaurs somewhat more susceptible to the asteroid impact, but probably nothing was going to save them.”

The focus on the extinction event also obscures the breathtakingly long nature of the dinosaurs’ reign. It invites us to think that all the dinosaurs we’re familiar with were around at the same time, and then suddenly they weren’t. That’s not true: as Brian Switek says, less time separates us from Tyrannosaurus than separated Tyrannosaurus from Stegosaurus. If anything, Sakomoto’s study, in revealing the dinosaurs’ slow decline, reminds us about just how long they ruled for—a period of 180 million years, during which many species came and went.