An extinction event, also called a mass extinction or biotic crisis, is a time when many different types of living things on Earth disappear quickly. Scientists know this has happened when there is a sudden drop in the number and variety of multicellular organisms. This occurs when the speed at which species go extinct becomes faster than the usual rate of extinction and the rate at which new species form.

Experts disagree about how many major mass extinctions have happened in the past 540 million years. Some say there were as few as five, while others believe there were more than twenty. These differences happen because people do not always agree on what makes an extinction event "major," and they use different information to study how life was in the past.

The "Big Five" mass extinction events

In a significant study published in 1982, Jack Sepkoski and David M. Raup discovered five specific time periods in Earth's history with unusually high levels of species loss. These periods were first noticed as unusual compared to the general trend of decreasing extinction rates during the Phanerozoic Eon. However, as more detailed statistical methods were used to analyze the data, scientists confirmed that during the Phanerozoic Eon, multicellular animals have faced at least five major and many smaller mass extinctions. The "Big Five" extinctions are not clearly separated but instead appear to be the most severe events in a continuous pattern of extinction events.

Before the "Big Five" extinctions, there was a much larger mass extinction of microscopic life during the Great Oxidation Event (also called the Oxygen Catastrophe) in the early Proterozoic Eon. Another extinction event, possibly very large, occurred at the end of the Ediacaran Period and just before the Cambrian explosion, which marked the beginning of the Phanerozoic Eon. Some extinction events in the Cambrian and early Ordovician Periods were as severe as or more severe than the "Big Five," but overall life diversity was low until the Great Ordovician Biodiversification Event (GOBE). Sepkoski and Raup (1982) originally measured the actual number of species lost instead of the percentage, so their findings did not include events before the GOBE.

The Late Ordovician mass extinction (LOME) likely happened in two stages, just before and during the transition from the Ordovician to the Silurian Period. The first stage was linked to global cooling and large ice sheets in the south, while the second stage occurred during a warming period. Together, these events caused the extinction of 27% of all families, 57% of all genera, and 85% of all species. Many scientists now consider the LOME the second-largest extinction in Earth's history based on the percentage of genera lost.

In May 2020, studies suggested that the LOME was caused by global warming, volcanic activity, and low oxygen levels in the ocean, not cooling and ice as previously thought. However, this contradicts earlier research that linked the first stage of the LOME to global cooling. More recent findings suggest that volcanic eruptions may have reduced atmospheric carbon dioxide, leading to ice formation and low oxygen levels.

The Late Devonian Period was marked by significant loss of life, with two major extinction events. Scientists believe these events were caused by low oxygen levels in the ocean. These conditions may have resulted from volcanic activity, land plant growth, and climate changes, though the exact causes are still debated.

The larger of these two events was the Kellwasser Event, which occurred at the boundary between the Frasnian and Famennian stages (372 million years ago) in the Late Devonian. This event severely damaged coral reefs and many tropical seabed animals, such as jawless fish, brachiopods, and trilobites. Scientists think this event was caused by nutrients from land being carried into the ocean by rivers. These nutrients led to algal blooms, which used up oxygen in the water, creating low-oxygen conditions that caused mass extinctions.

The other major Late Devonian extinction was the Hangenberg Event, which happened at the boundary between the Devonian and Carboniferous Periods (359 million years ago). This event ended the Devonian Period and nearly wiped out armored fish and newly evolved ammonoids. Together, the Kellwasser and Hangenberg Events caused the extinction of 19% of all families, 50% of all genera, and at least 70% of all species. Sepkoski and Raup (1982) did not initially consider these events statistically significant, but later studies confirmed their major ecological impacts.

The End Permian extinction, also called the "Great Dying," occurred at the boundary between the Permian and Triassic Periods. It was the largest extinction in the Phanerozoic Eon, with 53% of marine families, 84% of marine genera, about 81% of marine species, and an estimated 70% of land vertebrates going extinct. This event also caused the extinction of trilobites and dominant coral groups. While evidence about plants is unclear, new species became dominant after the extinction.

The "Great Dying" had major effects on life on Earth. It reduced many groups of synapsids (mammal relatives) on land, and it took 30 million years for vertebrates to recover. This extinction created opportunities for archosaurs (relatives of dinosaurs and crocodiles) to diversify. In the oceans, the percentage of animals that could not move (sessile animals) dropped from 67% to 50%.

Some scientists now believe the impact on land ecosystems may have been less severe than previously thought. Fossil evidence from China shows that ecosystems recovered quickly, possibly within 75,000 years. Others argue that there may have been no major land extinction at all, supported by findings in plants, insects, and four-legged vertebrates.

The late Permian Period was challenging for marine life even before the End Permian extinction. Recent research suggests that the End-Capitanian extinction event, which occurred before the "Great Dying," may have been a separate event larger than some of the "Big Five" extinctions.

The End Triassic extinction marked the boundary between the Triassic and Jurassic Periods. It caused the extinction of 23% of all families, 48% of all genera, and 70% to 75% of all species. In the oceans, ceratite ammonoids and conodonts disappeared, and reef-building animals were severely affected. Many archosauromorphs, therapsids, and large temnospondyl amphibians were eliminated, leaving dinosaurs with few competitors. Pterosaurs and crocodylomorphs were the only surviving archosaurs, while non-archosaurian diapsids dominated marine environments. Some temnospondyls, like Koolasuchus, survived until the Cretaceous Period in Australia.

The End Cretaceous extinction, also called the K–Pg extinction (formerly K–T extinction), occurred at the boundary between the Cretaceous (Maastrichtian) and Paleogene (Danian) Periods. It caused the extinction of 17% of all families, 50% of all genera, and 75% of all species. In the oceans, ammonites, plesiosaurs, and mosasaurs disappeared, and the percentage of sessile animals dropped to 33%. All non-avian dinosaurs

Sixth mass extinction

Studies done after a key 1982 paper by Sepkoski and Raup show that a sixth mass extinction event, caused by human actions, is happening now. Since 1900, species have been disappearing at a rate more than 1,000 times faster than the normal rate. This extinction is linked to human activities, including population growth, economic expansion, and the overuse of Earth’s natural resources. A 2019 report by IPBES states that about 1 million of the estimated 8 million plant and animal species are at risk of extinction. In late 2021, WWF reported that over a million species could vanish in the next 10 years, calling it the largest mass extinction since the time of the dinosaurs. A 2023 study in PNAS found that at least 73 animal genera have gone extinct since 1500. If humans had never existed, these species would have taken 18,000 years to naturally disappear, according to the report. The term "sixth mass extinction" is sometimes used to describe significant but less well-known prehistoric extinctions, such as the Capitanian mass extinction 260 million years ago, which may have been as severe as the "Big Five" extinctions.

Extinctions by severity

Extinction events can be studied using several methods, such as changes in Earth's layers, effects on ecosystems, comparisons between the number of species that disappear and those that appear, and the loss of variety among groups of organisms. Early research often used families as the main group for study, based on lists of marine animal families created by Sepkoski in 1982 and 1992. Later studies by Sepkoski and others used genera, which are more accurate than families and less likely to be affected by incomplete data or biases in classification. Several important papers have estimated the loss of species or effects on ecosystems from fifteen major extinction events. The methods used in these papers are explained in the next section. The "Big Five" mass extinctions are shown in bold.

Graphed but not discussed by Sepkoski (1996), this event was once thought to be connected to the Late Devonian mass extinction. At the time, it was also considered part of the end-Permian mass extinction. It includes time periods from the late Norian. The loss of species during both major events was calculated together. The events spanned nearby time periods and were calculated separately. This event had significant effects on ecosystems but was not studied directly. It was excluded because scientists disagreed about the timing of events in the Late Triassic.

The study of major extinction events

For much of the 20th century, scientists had trouble studying mass extinctions because they lacked enough information. Although scientists knew mass extinctions happened, they were seen as unusual events that did not fit the idea that life changed slowly over time. In 1980, Luis Alvarez and his team made an important discovery. They found evidence of an asteroid hitting Earth at the end of the Cretaceous period. This idea, called the Alvarez hypothesis, helped scientists and the public pay more attention to mass extinctions and sudden causes for them.

In 1982, David M. Raup and Jack Sepkoski published a study in the journal Science. This work used a detailed list of marine animal families created by Sepkoski. It found five major times when many marine families went extinct, even though extinction rates had generally been decreasing. Four of these times were clearly important: the end of the Ordovician period, the Late Permian period, the end of the Triassic period, and the end of the Cretaceous period. The fifth time was a longer period in the Devonian period, with the most severe extinctions during the Frasnian stage.

Throughout the 1980s, Raup and Sepkoski improved their data on extinction and new species. They created a detailed graph showing changes in biodiversity, called the "Sepkoski curve," and identified groups of life with unique patterns of growth and extinction. Their work became a foundation for future studies. However, in 1984, they suggested a controversial idea: that mass extinctions happened in a cycle every 26 million years. Some astronomers linked this to a possible distant object in the Solar System, called the "Nemesis hypothesis." This idea was later strongly questioned by other scientists.

At the same time, Sepkoski began making a detailed list of marine animal genera. This allowed researchers to study extinctions in more detail. He published early results in 1986, identifying 29 important extinction events. By 1992, he updated his 1982 list of marine families, finding little change in the overall trend despite new data. In 1996, he published a study tracking marine genera extinctions, using three different ways to filter his data: all genera, those found in multiple time periods, and those from well-studied groups. He also revised his family extinction data based on his 1992 update.

New interest in mass extinctions led scientists to study how geological events affected life. In 1995, Michael Benton compared extinction and new species rates in marine and land animals, finding 22 extinction events with no repeating pattern. Books by O.H. Walliser (1996) and A. Hallam and P.B. Wignall (1997) summarized research from the previous two decades. One book listed over 60 possible global extinction events of different sizes. These works helped shape the idea that there were five major mass extinctions, with many smaller ones.

Sepkoski died in 1999, but his marine genera list was published in 2002. This led to new studies on mass extinction patterns. Researchers used his data to compare extinction and new species rates across geological time. A 2006 study by Bambach found 18 distinct extinction events, including four large ones in the Cambrian period. These matched Sepkoski's definition of extinction as short, severe events with high loss of species.

In 2004, Bambach, Knoll, and Wang studied the "Big Five" extinctions. They found each had different relationships between new species and extinction rates. Background extinction rates varied, with some periods having fewer extinctions relative to new species. This showed that the late Ordovician, end-Permian, and end-Cretaceous extinctions were unusually severe, while the late Devonian and end-Triassic extinctions occurred during already stressed times.

In 2005, Foote used computer models to show that sudden extinction events better explained past biodiversity changes than slow, steady rates. This supported the idea that rapid, frequent extinctions were major drivers of change. New species events also occurred, but less often.

Stanley (2007) studied how different groups of life responded to extinction and new species rates. His models showed that biodiversity grew rapidly throughout the Phanerozoic era.

As more data was collected, scientists began checking Sepkoski's work for errors. In 1982, Signor and Lipps noted that the fossil record was incomplete, making extinctions seem less sharp than they were. This idea, called the "Signor-Lipps effect," explains that species might have gone extinct during an event, even if the fossil record shows a slower decline. Foote (2007) found that some geological periods had higher extinction rates due to data from later periods.

Other challenges include studying species that changed quickly or had limited fossil records. Many scientists now use methods like randomized sampling to adjust for these issues. However, these methods can be influenced by how much data is available. A major issue is the "Pull of the Recent," where the fossil record is more complete for recent times, making older periods seem less diverse.

In 2010, Alroy used a method called "shareholder quorum subsampling" (SQS) to reduce errors in diversity estimates. This method samples fossils from a group to compare how many species existed, ensuring fair representation. Every time a new species is found, all its fossils are included in the sample.

Uncertainty in the Proterozoic and earlier eons

Most life on Earth is made up of tiny microbes, which are hard to study using fossils. Because of this, recorded extinction events usually focus on complex organisms that are easier to observe, rather than the full range of life. Well-documented extinction events mostly happened during the Phanerozoic eon, except for the Oxygen Catastrophe in the Proterozoic eon. Before the Phanerozoic, all living things were either microbes or soft-bodied multicellular organisms, which are rarely preserved as fossils. This lack of a clear fossil record for microbes might make it seem like mass extinctions mainly happened during the Phanerozoic, even though extinction rates may have been lower before complex organisms with hard parts appeared.

Extinction happens at different rates. According to the fossil record, about two to five groups of marine animals go extinct every million years.

The Oxygen Catastrophe, which occurred around 2.45 billion years ago during the Paleoproterozoic, is likely the first major extinction event. It may also have been the most severe, but because Earth’s ecosystems before this time are poorly understood and prokaryote genera differ greatly from complex life groups, it is difficult to compare it to later extinction events like the "Big Five," even if more information about Paleoproterozoic life were available.

Evolutionary importance

Mass extinctions have sometimes helped life on Earth evolve faster. When one group of organisms stops being the main group in certain parts of the environment, it is usually not because the new group is "better" than the old one. Instead, it is often because a mass extinction event removes the old group, allowing the new group to take over. This process is called adaptive radiation.

For example, mammaliaformes ("almost mammals") and mammals lived during the time of the dinosaurs, but they could not compete for the large land animal roles that dinosaurs controlled. The end-Cretaceous mass extinction removed the non-avian dinosaurs, allowing mammals to take over these roles. Dinosaurs themselves had benefited from a previous mass extinction, the end-Triassic, which removed most of their main competitors, the crurotarsans. Similarly, within Synapsida, the group of early animals called pelycosaurs (though this group is not fully related) was replaced by therapsids around the Kungurian/Roadian transition, an event sometimes called Olson's extinction. This event may have been a slow change over 20 million years, not a sudden event.

The Escalation hypothesis suggests that species living in environments with a lot of competition between individuals are less likely to survive mass extinctions. This is because traits that help a species grow and stay strong in stable conditions can become a problem when competition decreases during an extinction event.

Many groups that survive mass extinctions do not recover their numbers or variety, and many of these groups slowly decline over time. These groups are often called "Dead Clades Walking." However, groups that survive for a long time after a mass extinction, even if they are reduced to only a few species, may later experience a recovery called the "push of the past."

Darwin strongly believed that interactions between living things, such as competition for food and space—the "struggle for existence"—were more important in driving evolution and extinction than changes in the physical environment. He wrote about this in The Origin of Species.

Patterns in frequency

Some scientists believe that extinction events happen regularly every 26 to 30 million years, or that the variety of life on Earth changes in bursts about every 62 million years. Many ideas, mostly involving space-related causes, try to explain these patterns. These include the possibility of a hidden star that travels with the Sun, movements of Earth through the Milky Way galaxy, or Earth passing through the galaxy’s spiral arms. However, other scientists have found that evidence about marine mass extinctions does not clearly support the idea that extinctions happen in a regular cycle or that ecosystems naturally reach a point where extinction becomes unavoidable. Many of the proposed connections between these events and other factors have been argued to be incorrect or not strong enough statistically. Others claim there is strong evidence showing regular patterns in different records, including changes in nonliving chemical elements like Strontium isotopes, large volcanic eruptions, oxygen-depleted ocean events, mountain-building events, and salt deposits. One possible explanation for this cycle is the movement of carbon between Earth’s atmosphere and its deep layers through oceanic crust.

Mass extinctions may happen when long-term stress is followed by a sudden event. Over the Phanerozoic era, individual groups of organisms seem less likely to go extinct, which might be because food webs became more stable, fewer species were vulnerable to extinction, or other factors like how continents were arranged. However, even after considering how well fossils are preserved, there appears to be a slow decrease in both extinction and new species formation rates over the Phanerozoic. This could mean that groups with faster species turnover are more likely to go extinct by chance, or it might be due to how scientists classify species: larger groups tend to include more species over time, making them less likely to go extinct, and larger taxonomic groups naturally appear earlier in Earth’s history.

Some scientists also suggest that oceans have become more suitable for life over the past 500 million years, making mass extinctions less likely. However, whether species are more or less likely to go extinct at a taxonomic level does not seem to affect the overall chance of mass extinctions happening.

Causes

Scientists are still trying to understand why all mass extinctions happened. Usually, large extinctions occur when a living world that is already under long-term stress experiences a sudden, short-term event. A pattern seems to exist: when there is a high variety of life, the rate of extinction increases over time. When there is a low variety of life, the rate of new species forming increases over time. These patterns, which are likely controlled by the environment, may make small changes, like asteroid impacts, cause larger effects globally.

A good explanation for a specific mass extinction should consider multiple causes. For example, the marine life loss during the end-Cretaceous extinction was likely caused by several overlapping processes that may have had different effects in different areas of the world.

Arens and West (2006) suggested a "press/pulse" model. This model says that mass extinctions usually need two types of causes: long-term stress on ecosystems ("press") and a sudden disaster ("pulse") near the end of the stress period. Their study of marine extinction rates over the Phanerozoic Eon showed that neither long-term stress nor a sudden disaster alone was enough to cause a major increase in extinction rates.

MacLeod (2001) reviewed how mass extinctions relate to events often linked to them, using data from other scientists:

The most common causes of mass extinctions include:

Large igneous provinces formed by flood basalt events could have caused:

Flood basalt events happen in bursts of activity with long periods of inactivity. These events may cause the climate to change between cooling and warming, but overall, they tend to warm the planet because the carbon dioxide they release stays in the atmosphere for hundreds of years.

Flood basalt events are linked to many major extinction events. Scientists think massive volcanism may have caused or contributed to the Kellwasser Event, the End-Guadalupian Extinction Event, the End-Permian Extinction Event, the Smithian-Spathian Extinction, the Triassic-Jurassic Extinction Event, the Toarcian Oceanic Anoxic Event, the Cenomanian-Turonian Oceanic Anoxic Event, the Cretaceous-Palaeogene Extinction Event, and the Palaeocene-Eocene Thermal Maximum. The connection between large volcanic events and mass extinctions has been observed for the last 260 million years. Recently, this link has been found to extend across the entire Phanerozoic Eon.

These events are often marked by layers of sediment showing changes from underwater areas to land, with no signs of geological processes like mountain-building. Sea-level drops could reduce the size of continental shelves (the most productive parts of the oceans), causing marine extinctions, and could disrupt weather patterns, causing land extinctions. However, sea-level drops are likely caused by other events, such as long-term global cooling or the sinking of mid-ocean ridges.

Sea-level drops are linked to most mass extinctions, including all of the "Big Five"—End-Ordovician, Late Devonian, End-Permian, End-Triassic, and End-Cretaceous—and the Capitanian mass extinction, which is as severe as the Big Five.

A 2008 study in Nature found a connection between the speed of mass extinctions and changes in sea level and sediment. The study suggests that changes in ocean environments caused by sea level shifts influence extinction rates and determine the types of life in the oceans.

A large asteroid or comet impact could collapse food chains on land and in the sea by creating dust and aerosols that block sunlight, stopping photosynthesis. Impacts on sulfur-rich rocks could release sulfur oxides, forming toxic acid rain that further harms food chains. Such impacts might also cause megatsunamis or global forest fires.

Most scientists agree that an asteroid hit Earth about 66 million years ago, but there is still debate about whether this impact alone caused the Cretaceous–Paleogene extinction. However, a 2019 study showed that the Chicxulub asteroid impact not only killed non-avian dinosaurs but also rapidly acidified the oceans, leading to ecological collapse and long-term climate changes that were key to the end-Cretaceous mass extinction.

The Permian-Triassic extinction event has been linked to an asteroid impact that formed the Araguainha crater, as the crater's age overlaps with the extinction. However, this idea is widely disputed, and most scientists reject it.

The Shiva hypothesis suggests that Earth experiences more asteroid impacts every 27 million years because the Sun passes through the Milky Way galaxy’s plane, causing extinction events at regular intervals. Some evidence supports this idea in both marine and non-marine settings. Alternatively, the Sun’s movement through the galaxy’s dense spiral arms might cause mass extinctions due to more frequent impacts. However, recent research using maps of the Milky Way’s spiral structure found no clear connection.

A nearby gamma-ray burst (within 6,000 light-years) could destroy Earth’s ozone layer, leaving life vulnerable to harmful sunlight. These bursts are rare, occurring only a few times per million years in a galaxy. Some scientists think a gamma-ray burst caused the End-Ordovician extinction, while a supernova may have caused the Hangenberg event. A supernova within 25 light-years could strip Earth of its atmosphere. Currently, no stars close enough to Earth can produce a dangerous supernova.

Long-term global cooling could kill many polar and temperate species, force others to move toward the equator, reduce tropical habitat space, and make the climate drier by locking water in ice. The cooling cycles of the current ice age had only a small effect on biodiversity, so cooling alone may not cause mass extinctions.

Some scientists think global cooling contributed to the End-Ordovician, Permian–Triassic, and Late Devonian extinctions. Sustained cooling is different from the temporary climate changes caused by flood basalt events or impacts.

Global warming could have the opposite effects: increase tropical habitat space, kill temperate species, force them toward the poles, cause severe polar extinctions, and make the climate wetter by melting ice. It might also cause oxygen-deprived conditions in the oceans.

Effects and recovery

The effects of mass extinction events were different each time. After a major extinction, only species that can live in many different places usually survive. These species are called "weedy species." Over time, new species grow and fill the empty spaces left by extinct species. It generally takes millions of years for life to return to normal after an extinction event. In the most extreme cases, recovery might take 15 to 30 million years.

The worst mass extinction during the Phanerozoic era was the Permian–Triassic extinction, which caused the loss of over 90% of all species. At first, it seemed like life recovered quickly after this event, but this was mostly due to simple, hardy species like the Lystrosaurus. Recent studies show that animals with complex ecosystems, many different species, and intricate food webs took much longer to return. Scientists believe this slow recovery happened because of repeated extinction events and long-lasting environmental problems that lasted into the Early Triassic. Research suggests recovery began around 4 to 6 million years after the extinction, during the mid-Triassic. Some scientists estimate that full recovery did not happen until 30 million years later, during the late Triassic. After the Permian–Triassic extinction, species lived in smaller areas, which may have removed existing species from their habitats and created conditions for new species to later diversify.

The impact of mass extinctions on plants is harder to study because the plant fossil record has many gaps and biases. Some extinctions, like the end-Permian event, were very harmful to plants, while others, such as the end-Devonian event, did not significantly affect plant life.

In media

The term extinction-level event (ELE) has been used in movies and other media. The 1998 film Deep Impact shows a possible comet hitting Earth and calls it an ELE.