Mass Extinction Events

What counts as a mass extinction

Species go extinct all the time. Background extinction — the slow trickle of lineages dying out as environments shift, parasites evolve, or competitors emerge — runs at roughly one species per million per year over geological time. Mass extinctions are events where that rate spikes by orders of magnitude, with global biodiversity losing roughly three-quarters or more of all species in a few hundred thousand to a few million years. They are visible in the fossil record as sharp drops in species richness across a geological boundary.

Jack Sepkoski and David Raup's 1982 catalogue of marine animal genera through the Phanerozoic identified five spikes that stood far above the noise floor — the Big Five. Smaller events (the end-Capitanian, end-Cenomanian, the Eocene-Oligocene cooling) cluster as second-tier crises but don't reach the same magnitude. The Big Five each have geological-rock and geochemical signatures distinctive enough that they punctuate the textbook chapters on Phanerozoic life: each one ended an era and reset the cast of dominant lineages for the next.

The Big Five compared

	Ordovician	Late Devonian	Permian-Triassic	Triassic-Jurassic	Cretaceous-Paleogene
Date (Mya)	≈ 444	≈ 372 (multi-pulse)	≈ 252	≈ 201	66
Species lost	≈ 85%	≈ 75%	≈ 95% marine	≈ 80%	≈ 75%
Likely trigger	Glaciation, sea-level drop	Anoxia from plant-driven nutrient runoff	Siberian Traps volcanism	Central Atlantic Magmatic Province	Chicxulub asteroid impact
Major casualties	Brachiopods, conodonts, trilobites, graptolites	Reef builders, jawless fish, placoderms	Trilobites (final), most reef life	Conodonts, large amphibians, basal archosaurs	Non-avian dinosaurs, ammonites, mosasaurs
Major beneficiaries	Survivors radiated into reefs	Tetrapods crossed onto land	Archosaurs and synapsids	Dinosaurs took dominance	Mammals, birds, teleost fish
Geochemical signature	Negative δ13C, glacial deposits	Black shales, ocean anoxia	Carbon spike, Siberian sills, sulfides	Carbon excursion, CAMP basalts	Iridium anomaly, shocked quartz, tektites
Recovery time	≈ 5 My	≈ 10-15 My	≈ 10 My (slowest)	≈ 5 My	≈ 5-10 My
Era boundary marked	Within Paleozoic	Within Paleozoic	Paleozoic → Mesozoic	Within Mesozoic	Mesozoic → Cenozoic

Note the diversity of triggers. Three of the five involve large igneous province volcanism (the Siberian Traps and the Central Atlantic Magmatic Province are still visible as continent-scale basalt sheets). One was a bolide impact. One was glaciation-driven sea-level change. The unifying thread is rapid global environmental stress — the climate or chemistry shifts faster than most species can adapt, especially specialists with narrow tolerances.

Worked example: the Permian-Triassic Great Dying

252 million years ago, in roughly 60,000 years — a geological eyeblink — about 95% of marine species and 70% of terrestrial vertebrates went extinct. Reefs disappeared globally and didn't return for 10 million years. Coal forests collapsed. Trilobites, which had survived three previous extinction events, finally vanished. So did most foraminifera, brachiopods, sea urchins, and the dominant land vertebrates of the time. Almost everything we think of as Paleozoic.

The trigger was almost certainly the Siberian Traps — flood basalt eruptions covering an area roughly the size of Western Europe with up to four kilometers of lava. The eruptions intruded into Permian coal and salt deposits, releasing not just CO2 and SO2 but also vast quantities of methane and halogens. Atmospheric CO2 spiked, ocean temperatures rose by 10°C in tropical waters, and the oceans became hypoxic and acidified. Sulfide-producing anaerobic bacteria bloomed in shallow seas, releasing toxic hydrogen sulfide. The combination — heat, acid, anoxia, sulfide — left almost nothing in the photic zone alive.

Recovery took 10 million years, the longest of any of the Big Five. The few surviving lineages — including some early archosaurs — radiated into the empty Triassic world, eventually producing dinosaurs. Mammal ancestors (synapsids), which had dominated the Permian, were reduced to small, mostly burrowing forms. The shape of the rest of the Mesozoic was set by who happened to make it through.

Worked example: the KP impact and the iridium signature

The Cretaceous-Paleogene event 66 million years ago was the most distinctive in geochemical terms. Walter Alvarez, a geologist at Berkeley, and his Nobel-laureate father Luis Alvarez noticed in the late 1970s that a thin clay layer marking the KP boundary in Italy was rich in iridium — an element vanishingly rare in Earth's crust but abundant in chondritic meteorites. Their 1980 paper in Science proposed an asteroid impact, around 10 km in diameter, as the trigger.

For a decade the hypothesis was contested. Then, in 1991, the Chicxulub crater was confirmed off the Yucatán Peninsula — 180 km in diameter, dated by argon-argon to 66.04 Mya, exactly matching the boundary clay. Drilling cores revealed shocked quartz, microtektites (impact melt droplets), and the iridium spike at the same horizon worldwide. Tsunami deposits along the Gulf of Mexico, charcoal layers signaling global wildfires, and a sudden drop in plant pollen all line up. The smoking gun is one of the strongest in geology.

The kill mechanism wasn't the impact itself — that vaporized only the immediate Yucatán region. It was the global ejecta blanket: hot debris re-entering the atmosphere worldwide, igniting wildfires; sulfate aerosols and dust blocking sunlight for months; collapse of photosynthesis; food chains snapping from the bottom up. Anything over roughly 25 kg on land died. So did all non-avian dinosaurs, all pterosaurs, all ammonites, and most marine plankton. Survivors were small, ate detritus or seeds, lived in freshwater, or could go dormant.

Real-world consequences and recovery patterns

• Tropical reefs disappear first. Reef ecosystems collapsed in all five Big Five events and took the longest to rebuild — they need stable temperature, normal pH, and clear water. Today's reefs are showing the same warning signs under rapid warming and acidification.
• Body size matters. Across most events, larger animals die preferentially. Big animals reproduce slowly, eat more, and need more habitat. After the KP event nothing larger than a domestic cat survived on land.
• Specialists die first. Ammonites with narrow shell types went extinct; generalist nautiloids survived. Specialist herbivores tied to one plant family die when that family fails; broad browsers survive.
• Recovery means radiation. Empty niches drive rapid speciation in survivors. The rise of mammals after the KP, the rise of dinosaurs after the Triassic-Jurassic, the recovery of marine reefs after the Permian — each illustrates how surviving lineages diversify into vacated ecological space over 5-10 million years.
• Recovery isn't return. Post-extinction biotas don't recapitulate what came before. After the Permian, brachiopod-dominated benthos was replaced by mollusks; after the KP, dinosaur ecological roles were not restored. Mass extinctions reset the deck.
• The sixth extinction. Current rates of vertebrate species loss are estimated at 100-1,000 times background. Causes include habitat loss (the largest), climate change, invasive species, overexploitation, and pollution. The total magnitude does not yet match the Big Five, but the rate is comparable. Whether it reaches the canonical 75% threshold depends on policy and time.

Common kill mechanisms

• Climate shift. Glaciation in the Ordovician, warming in the Permian and Triassic. Species adapted to one regime cannot relocate fast enough.
• Ocean anoxia. When oceans warm and stratify, oxygen-poor "dead zones" expand. The Devonian and Permian both show widespread black shale deposits — sediments laid down without oxygen, recording mass die-offs of marine life.
• Acidification. CO2 dissolves in seawater as carbonic acid, dropping pH and dissolving carbonate shells. Permian and KP events both show acidification signatures; reef-builders (corals, calcifying plankton) suffer most.
• Volcanism. Large igneous provinces release CO2, SO2, methane, and halogens. The Siberian Traps, the CAMP, and the Deccan Traps (which overlap with the KP boundary) all coincide with major extinctions.
• Impact. The 10-km bolide at Chicxulub is the only confirmed extinction-class impact in the Phanerozoic. Other major events lack iridium spikes or impact craters of comparable scale.
• Sea-level change. Regression (sea drop) destroys shallow-marine habitat where most species live. Several Big Five events coincide with drops of tens to hundreds of meters.
• Trophic cascade. Whatever the initial trigger, food chains snap. Photosynthesizers fail, herbivores starve, predators starve. The longer the food chain, the more vulnerable the apex.

Variants and refinements

• Pulse extinctions. Some events were not single instants. The Late Devonian had at least two pulses (Kellwasser, Hangenberg) spaced about 10 million years apart. Treating these as one event obscures the dynamics.
• Background vs catastrophe. Raup proposed continuum models where mass extinctions are the high-magnitude tail of a single distribution rather than qualitatively different events. The two views have been argued for decades.
• Press vs pulse. Press events (long sustained stress like cooling) and pulse events (sudden shocks like impacts) have different signatures and different survivor profiles. Most real extinctions combine both.
• Selectivity. Some events selected on body size; others on diet; others on habitat. The KP event was biased against large body size and against specialized diets. The Permian was less selective — almost everything died.
• Recovery faunas. The boom of "disaster taxa" — opportunistic generalists like Lystrosaurus after the Permian, fern spikes after the KP — defines the immediate post-extinction interval before normal ecosystems rebuild.
• Defining the sixth. Whether current human-driven decline qualifies depends on how much extinction debt (committed but not yet realized losses) is counted, and how the next centuries unfold. The rate is in the Big Five range; the cumulative magnitude is not yet.

Common pitfalls

• Thinking dinosaurs went extinct first. Dinosaurs were dominant for 165 million years and outlived everything except the impact itself. They were thriving until the day Chicxulub hit.
• Conflating "all dinosaurs" with non-avian dinosaurs. Birds are dinosaurs — modern theropod descendants. Avian dinosaurs survived the KP event and are alive today. Only non-avian dinosaurs went extinct.
• Treating the Big Five as the only extinctions. Smaller events (the End-Capitanian, the End-Eocene) erased many species too. The Big Five are the most severe, not the only crises.
• Assuming impacts caused most extinctions. Only the KP event has a confirmed impact signature. Volcanism is a more common trigger for the others.
• Treating recovery as restoration. Post-extinction biotas are different. After the Permian, mollusks replaced brachiopods; after the KP, mammals replaced large reptiles. Recovery doesn't reset to the prior state.
• Calling current biodiversity loss "the same as the dinosaurs." The Holocene loss is severe and accelerating, but in absolute species count it has not yet reached Big Five levels. The trajectory is concerning; the comparison should be precise.