Life on planet Earth is a fragile thing. All it takes is one wayward asteroid, and bam, there goes the most dominant group of land animals on our planet.

If it wasn't for the 10-kilometre-wide meteorite that hurtled into Earth some 66 million years ago, dinosaurs might have continued to dominate the land, a new paper suggests. As for mammals like ourselves, we may never have gotten the chance to rise.

The results are the latest in a long and drawn-out debate on the extinction of non-avian dinosaurs.

While the Chicxulub asteroid and the consequences of its impact - which blocked the Sun's rays and triggered global climate cooling - are generally considered the prime candidates for the massive Cretaceous extinction event, some recent evidence suggests certain dinosaur species were already in decline tens of millions of years before that.

The authors of these findings have claimed "overwhelming support for a long-term decline across all dinosaurs and within all three major dinosaur groups", but among palaeontologists, support for that idea is hardly overwhelming. In fact, it's extremely contentious.

In the years since this idea was first put forward, several other lines of research have disagreed with its conclusions - not necessarily with the data itself, but with the interpretations drawn.

Gaps in dinosaur fossils and sampling biases mean we might be under sampling certain Cretaceous dinosaurs, while overestimating the presence of others.

"Previous studies done by others have used various methods to draw the conclusion that dinosaurs would have died out anyway, as they were in decline towards the end of the Cretaceous period," explains paleontologist Joe Bonsor from the University of Bath.

"However, we show that if you expand the dataset to include more recent dinosaur family trees and a broader set of dinosaur types, the results don't actually all point to this conclusion - in fact only about half of them do."

Rather than simply counting the number of dinosaur species present at the time using fossil records, the team used statistical methods to look at the rate of speciation within dinosaur families.

Analysing thousands of family tree combinations in 12 dinosaur families, researchers tested whether species diversification was slowing down, staying the same or speeding up before the asteroid's impact - an indication that could tell us how quickly extinct dinosaurs are replaced with new ones.

Of all 2,727 speciation models, only 518 (less than 20 percent) unambiguously showed terminal decline before the asteroid impact.

Even when the authors considered those on the border of terminal decline and no decline, that's just over half the models that would support a more drawn out extinction.

As such, the team says they are skeptical of the terminal extinction theory, instead suggesting that dinosaur diversity would have remained high throughout the Late Creataceous, even as species richness varied between branches.

"The main point of our paper is that it isn't as simple as looking at a few trees and making a decision. The large unavoidable biases in the fossil record and lack of data can often show a decline in species, but this may not be a reflection of the reality at the time," says Bonsor.

"Our data don't currently show they were in decline, in fact some groups such as hadrosaurs and ceratopsians were thriving and there's no evidence to suggest they would have died out 66 million years ago had the extinction event not happened."

The results are supported by another recent study, which found habitats for North American dinosaurs did not decline during the Late Cretaceous.

Fossil outcrops for this region are small, which means we're probably under sampling the area and underestimating its species richness.

Systemic sampling errors like this are rife in palaeontology and, to a certain extent, unavoidable. But relying on species richness as opposed to total species number also comes with its limitations.

For starters, there may very well be no correlation between speciation rates in dinosaur evolution and extinction. That's an assumption we're making in hindsight.

"We may never know the true levels of speciation and extinction of Mesozoic dinosaurs," the authors of the new paper admit, "but an increased focus on filling gaps in the fossil record will be the primary way in which palaeontologists will continue to build a more accurate picture of past dinosaur diversity."

The new paper once again emphasises the numerous gaps and biases in our knowledge, and the authors are calling for continued research that is detailed, regionally controlled and well distributed in time to recreate the most holistic history possible.

The study was published in the Royal Society Open Science.