By Paul Rincon
BBC News Online science staff
The chances that asteroid impacts and huge bouts of volcanism coincide randomly to cause mass extinctions may be greater than previously imagined.
What are the chances of such great events occurring together?
UK researchers conducted statistical tests to determine the probability of such catastrophic events happening at the same time in Earth history.
They found massive releases of lava and space collisions should have overlapped three times in the last 300 million years.
Details will be published in a future issue of the geological journal Lithos.
The work has been done by Dr Rosalind White and Professor Andy Saunders of the University of Leicester.
The probabilities they calculated assumed there was no causal link between the two phenomena - that impacts from space did not set off the volcanism.
Flood basalts, as the term suggests, are formed by massive outpourings of lava from beneath the Earth. Hundreds of thousands of cubic km of material can be spewed on to the surface in short geological timescales.
These eruptions have, like space impacts, been blamed for some of Earth's mass extinctions due to the environmental changes they may trigger.
The Leicester authors contend that because impacts and flood basalts occur more frequently than mass extinctions, it is unlikely the two phenomena bring about mass extinctions on their own.
Over a million cubic km of lava erupted on to surface
Event occurred over several hundred thousand years
However, mass extinctions may be triggered when the two events occur together, they argue.
There is evidence of both phenomena happening at the same time 65 million years ago, when the dinosaurs disappeared from the fossil record. The impact that created the 180km-wide impact crater at Chicxulub in Mexico is generally thought to have played a major part in this extinction.
But some experts think the flood basalts known as the Deccan Traps in India are an alternative "smoking gun". The gases released in this volcanic event would have resulted in major climate changes.
Firstly, the Leicester scientists determined the probability that a flood basalt would coincide with a Chicxulub-sized crater.
They found the probability of this happening at least once over a period of 300 million years was 57%.
Once the researchers reduced the size of the impact slightly, the probabilities increased sharply.
For craters exceeding 100km, the probability of at least three co-occurrences between flood basalts and impacts was 46%. For craters exceeding 60km, the probability of three or more was 97%.
Prof Saunders and Dr White point out that impacts and flood basalt volcanism have been implicated in three major mass extinctions in the past.
Some researchers, such as Professor David Price, of University College London, argue that asteroid impacts can themselves bring on catastrophic volcanism.
Professor Price says there is geochemical evidence to suggest an impactor started the Siberian trap flood basalts, which are associated with the end-Permian mass extinction 251 million years ago.
"I wouldn't be very convinced by the robustness of any statistical analysis when you're dealing with just half a dozen events in the last billion years," Professor Price told BBC News Online.
He also points to evidence uncovered by Professor Mike Coffin, of the University of Texas at Austin, which shows the Ontong Java Plateau - a submarine flood basalt in the western Pacific Ocean thought to be the world's largest - was initiated by the impact of a 20km object from space.
The Leicester researchers say the idea of impacts causing massive volcanism is interesting, but no causal link has been proven.
The co-occurrence of volcanism and impacts may work together to end life
"In the case of the Ontong Java Plateau, we would expect such a large impact to have left lots of signs in the geological record, but none have been found," Dr White said.
Professor Tony Hallam, of the University of Birmingham, commented: "The idea that [flood basalts and impacts] could be purely coincidental is rather a negative one - a null hypothesis - but it strikes me as quite reasonable."
Dr White and Professor Saunders propose in their paper that the "kill mechanisms" associated with flood basalts or impacts by themselves are not sufficiently powerful to cause the worldwide collapse of ecosystems - a point disputed by many other scientists.