Around 252 million years ago, volcanoes across what is now Siberia erupted repeatedly for more than a million years, releasing perhaps 100,000 billion tonnes of carbon dioxide and helping wipe out roughly 90 per cent of marine species in the deadliest mass extinction Earth has ever known

The end-Permian extinction was not a single volcano exploding on one bad day. It was a long volcanic province, a changing climate system, and a marine biosphere pushed past several limits at once.

Researchers usually place the crisis close to 252 million years ago, at the boundary between the Permian and Triassic periods. A 2015 Science Advances paper by Seth Burgess and Samuel Bowring used high-precision uranium-lead dating to connect the Siberian Traps large igneous province with Earth’s most severe known extinction, while showing that the magmatism began before the main biological collapse and continued after it.

The simple version is that huge eruptions released carbon dioxide and other gases. The more useful version is that the most damaging gases may not have come only from lava reaching the surface.

The timing problem

Large igneous provinces last too long to explain a geologically abrupt mass extinction by duration alone. The Siberian Traps were active over hundreds of thousands to more than a million years, depending on which part of the province is counted, while the sharpest extinction interval appears much shorter.

That mismatch is why the timing matters. In a 2017 Nature Communications paper, Burgess, James Muirhead, and Bowring argued that the key interval was not simply the earlier flood of surface lavas. They identified an abrupt shift from mostly flood lavas to widespread sill intrusions as the likely trigger interval for the end-Permian collapse.

A sill is magma that spreads sideways through rock underground rather than rising straight to the surface. That distinction sounds technical, but it changes the environmental story.

Why underground magma may have mattered more than lava

The Tunguska basin, beneath the Siberian Traps, contained carbonate, evaporite, clastic, and hydrocarbon-bearing rocks. When magma intruded through those sediments, it could bake them, drive chemical reactions, and release greenhouse and toxic gases at a scale ordinary surface degassing alone may not explain.

The 2017 Nature Communications study reported that the sill complex covered more than 1.5 million square kilometres and argued that heat from those sills likely liberated the large greenhouse gas volumes needed to drive the extinction. This is the reason estimates sometimes reach the scale of perhaps 100,000 billion tonnes of carbon dioxide, though the exact number depends on the model, the assumed source rocks, and whether the estimate is expressed as carbon, carbon dioxide, methane, or combined thermogenic gases.

That uncertainty should not be smoothed away. A 2011 Nature paper led by Stephan Sobolev offered a different but related mechanism, arguing from petrological evidence that recycled oceanic crust in the plume head could have allowed massive degassing of carbon dioxide and hydrochloric acid. The paper’s model suggested such degassing could itself trigger a mass extinction, but it is still one model in a broader debate over how much came from mantle sources, heated sediments, coal, carbonates, and hydrocarbons.

What the oceans record

The biological scale of the event is not in much doubt. A 2012 review by Jonathan Payne and Matthew Clapham in Annual Review of Earth and Planetary Sciences describes the end-Permian marine extinction as the largest in the Phanerozoic record, with losses often discussed at roughly 80 to 90 per cent of marine species depending on the taxonomic method and dataset.

The headline figure of roughly 90 per cent is therefore a rounded expression of a severe loss, not a census number from a complete fossil inventory. Fossil records are uneven. Some older estimates folded in nearby late-Permian losses, while newer analyses often separate the end-Capitanian and end-Permian crises more carefully.

Still, the main point survives that caution. Marine ecosystems were hit harder than in any other known extinction interval, and the selectivity of the losses points toward environmental stress rather than a random pruning of life.

Heat, low oxygen, acid, and toxic metals

Carbon dioxide matters in this story because it warms air and ocean, changes ocean chemistry, and makes it harder for some marine animals to maintain normal physiology. Warming also reduces oxygen solubility in seawater and can help expand low-oxygen zones.

That does not mean there was one kill mechanism. The end-Permian oceans appear to have faced a combination of heating, deoxygenation, acidification, and chemical stress. A 2015 Science paper led by Matthew Clarkson used boron isotopes to argue for a pulse of ocean acidification during the extinction, consistent with rapid carbon input into the ocean-atmosphere system.

The 2017 Nature Communications paper also points to a roughly 10 degree Celsius rise in global sea surface temperature as one of the lines of evidence for massive greenhouse forcing. A warmer ocean can become a hostile ocean quickly, especially for organisms that need stable carbonate chemistry to build shells and skeletons.

What this does not prove

The Siberian Traps are the favoured explanation for the end-Permian mass extinction, but the details are not settled in the way a simple cause-and-effect diagram suggests. Different studies place different weight on plume degassing, sediment heating, coal combustion, methane release, sulphur gases, halogens, mercury, acid rain, anoxia, and ocean acidification.

What the stronger papers have in common is not a single neat culprit. They show a tight timing relationship between Siberian magmatism, carbon-cycle disruption, warming, and biological collapse. They also show why the deadliest interval may have depended on where the magma went, not only on how much lava erupted.

The lesson from the rocks is therefore narrower and more precise than the usual phrase “volcanoes caused an extinction”. A vast volcanic system intruded and erupted through the wrong rocks at the wrong moment, releasing gases fast enough to overwhelm parts of Earth’s climate and ocean chemistry.

That is enough to make the end-Permian extinction less like an ancient curiosity and more like a record of how a planet responds when carbon enters the atmosphere faster than its stabilising systems can remove it.