Were dinosaurs already on their way to extinction before the asteroid impact? This question has sparked debate among scientists for more than 30 years. Now, a research team led by University College London (UCL) suggests that the long-standing belief that dinosaurs were already declining before the asteroid wiped most of them out 66 million years ago may not reflect reality—but rather the limitations of the fossil record from that time period.
The study, published in Current Biology, analyzed fossil data from North America spanning 66 to 84 million years ago, covering the last 18 million years before the asteroid impact at the end of the Cretaceous period. The team focused on four key dinosaur families: the armored, herbivorous Ankylosauridae (like Ankylosaurus), the horned Ceratopsidae (including Triceratops), the duck-billed Hadrosauridae (such as Edmontosaurus), and the carnivorous Tyrannosauridae (like Tyrannosaurus rex).
Initial fossil data suggested that dinosaur diversity peaked around 75 million years ago and declined during the final nine million years leading up to the asteroid impact. However, the researchers found this trend was likely influenced by the fact that fossils from that specific era are significantly harder to uncover—mainly because there are fewer exposed and accessible rock formations from the late Cretaceous period.
This limitation is largely due to geological processes such as mountain uplift and falling sea levels, which have affected how and where fossil-bearing rock layers are preserved and exposed over time.
To better understand the discrepancy, the researchers employed a method known as habitat modeling, typically used in ecology and biodiversity studies. This technique estimates how likely a species was to occupy a given area over time—in this case, during the final 18 million years of the Cretaceous.
“We analyzed the fossil record and found that the quality of the data for four major dinosaur groups declines significantly in the six million years before the asteroid,” explained lead author Chris Dean, a researcher at UCL. “The likelihood of discovering dinosaur fossils decreases, even though the probability that these dinosaurs were actually living in those regions remains constant. That tells us we can't take the fossil record at face value.”
Their findings revealed that, during this timeframe, the proportion of land likely inhabited by these dinosaur families remained largely stable—suggesting that their natural habitats weren’t shrinking, and that their risk of extinction was relatively low. The researchers also assessed how likely it was for each type of dinosaur to be detected in different areas, based on factors like how much land was accessible (e.g., not covered by vegetation), how much relevant rock was exposed, and how often paleontologists had previously searched those areas.
The team concluded that fossil discovery likelihood was low, and the most influential factor was the degree of exposure and accessibility of the appropriate rock formations. Interestingly, unlike the other three families, ceratopsids (e.g., Triceratops) were more likely to appear later in the timeline and were found across a wider area. The team theorized that this might be because ceratopsids preferred grassy plains far from rivers—a type of environment that became dominant during that time due to the retreat of a large inland sea that had once split North America, and the drying up of the river systems feeding it.
