Ars Technica has an article on Large Igneous Provinces, which are volcanic eruptions so large (occurring every 15 million years or so on average) they are thought to be the main culprit behind most mass extinctions. Lava and toxic gas can be bad but are not enough to explain mass extinction. They can actually cause large-scale global cooling on a short time scale (a decade or so) due to particulate emissions, but that effect also is not enough to explain mass extinction. The way they are thought to cause global mass extinction this is by burning underground fossil fuels on an enormous scale, causing catastrophic climate change that species cannot adapt to (including global warming, sea level rise, and ocean acidification, which kick in after the shorter-term cooling effect has run its course, and last a lot longer).
I was also wondering about volcanic activity versus comet and asteroid strikes. The article addresses that too:
Debate over what caused these factory resets of life has raged ever since Cuvier’s time. He considered them to be caused by environmental catastrophes that rearranged the oceans and continents. Since then, a host of explanations have been proposed, including diseases, galactic gamma rays, dark matter, and even methane from microbes. But since the 1970s, most scientists have considered the likely root cause to be either asteroid impacts, massive volcanic eruptions, or a combination of both.
Those asteroid (or comet) impacts have captured the public imagination ever since 1980, when Luis and Walter Alvarez found global traces of iridium, which they inferred to be extraterrestrial, at the geological boundary that marked the disappearance of the dinosaurs. The identification of the Chicxulub impact crater in Mexico soon after sealed the deal. Impacts have been proposed to explain other mass extinctions, but there’s very little actual evidence to support those links. In the words of researchers David Bond and Stephen Grasby, who reviewed the evidence in 2016: “Despite much searching, there remains only one confirmed example of a bolide impact coinciding with an extinction event…”
Volcanism, on the other hand, has coincided with most, if not all, mass extinctions—it looks suspiciously like a serial killer, if you like.
According to the article, the rate of fossil fuel burning is similar to what humans are doing now. So it follows that natural ecosystems could collapse this time too. That would be a catastrophe in itself, but how resilient can our species and food production systems be compared to all the others that have come and gone before us?