Cretaceous-Tertiary extinction event

2007 Schools Wikipedia Selection. Related subjects: Ancient History, Classical History and Mythology; Geology and geophysics

Badlands near Drumheller, Alberta where erosion has exposed the KT boundary.
Badlands near Drumheller, Alberta where erosion has exposed the KT boundary.
KT boundary exposure in Trinidad Lake State Park
KT boundary exposure in Trinidad Lake State Park

The Cretaceous-Tertiary extinction occurred about 65.5 million years ago. It is also known as the K-T extinction event and its geological signature as the K-T boundary ("K" is the traditional abbreviation for the Cretaceous Period, to avoid confusion with the Carboniferous Period, abbreviated as "C"). Since the label "Tertiary" is no longer recognized by most geologists (for example, the International Commission on Stratigraphy) as a geologic 'Period', the K-T demise might also be called the Cretaceous- Paleogene (or K-Pg) extinction event.

Significance of the K-T extinction

The KT extinction event, labeled here as "End K", is shown in comparison to the impact of other events on the extinction intensity for marine fossilerferous genera.
The KT extinction event, labeled here as "End K", is shown in comparison to the impact of other events on the extinction intensity for marine fossilerferous genera.

It was one of the "Big Five" extinctions. Although not the largest, it may be one of the most significant, because:

  • It caused a major change in both marine and land ecosystems. Before the K-T extinction about 50% of known marine species were sessile, and after it only about 33% were sessile. On the land it most famously exterminated the dinosaurs and so made it possible for mammals to become the dominant land vertebrates - and hence paved the way for human evolution.
  • It marks the end of the Cretaceous Period and the beginning of the Tertiary Period. In fact it changed ecosystems so much that it is regarded as the boundary between the Mesozoic and Cenozoic eras.

Casualties and survivors of the K-T extinction


Groups which became totally extinct include:

  • Ammonoids, which are currently thought to have fed on plankton.
  • Rudists, a group of clams which were the major reef-builders of the Cretaceous and also fed on plankton.
  • Inoceramids, giant relatives of modern scallops - they also fed on plankton.
  • Mosasaurs, giant lizards which were the top marine predators.
  • Plesiosaurs, another group of large reptilian marine predators.

Planktonic organisms suffered heavy losses, notably the coccolithophorids (chalk-forming nanoplankton algae which largely gave the Cretaceous period its name).


Groups which became totally extinct include:

  • Non- avian dinosaurs. Note: most paleontologists regard birds as the surviving dinosaurs.
  • Pterosaurs

Groups which suffered heavy losses include:

  • Birds. Some groups became extinct, including Enantiornithes and Hesperornithiformes.
  • Marsupials. The Northern hemisphere family of marsupials became extinct, but those in Australia and South America survived.
  • Freshwater mussels and snails also suffered heavy losses in North America.
  • In North America, as many as 57% of plant species may have become extinct. The Paleocene recovery of plants began with a " fern spike" like that which signals the recovery from natural disasters (e.g. the 1980 Mount St. Helens eruption).

But some other groups were relatively unaffected:

  • Insects. These show no sign of reduced diversity, unlike at the Permo-Triassic extinction.
  • Amphibians
  • Turtles
  • Lepidosaurs. These include tuataras, lizards, snakes and amphisbaenians.
  • Champsosaurs (semi-aquatic archosauromorphs) - they later died out in the early Oligocene.
  • Crocodilians
  • Modern birds ( Aves)
  • Monotremes, egg-laying mammals.
  • Multituberculates, although they later became extinct in the early Oligocene.
  • Placentals, the ancestors of most modern mammals.

Is there a pattern?

Despite its overall severity, the K-T extinction was rather patchy. This raises the question, "Why did some groups die out and others survive?"

There do seem to be some general trends:

  • Organisms which depended on photosynthesis became extinct or suffered heavy losses - from photosynthesing plankton (e.g. coccolithophorids) to land plants. And so did organisms whose food chain depended on photosynthesising organisms, e.g. tyrannosaurs (which ate vegetarian dinosaurs, which ate plants).
  • Organisms which built calcium carbonate shells became extinct or suffered heavy losses (coccolithophorids; many groups of molluscs, including ammonites, rudists, freshwater snails and mussels). And so did organisms whose food chain depended on these calcium carbonate shell builders. For example it is thought that ammonites were the principal food of mosasaurs.
  • Omnivores, insectivores and carrion-eaters appear to have survived quite well. It is worth noting that at the end of the Cretaceous there seem to have been no purely vegetarian or carnivorius mammals. Many mammals, and the birds which survived the extinction, fed on insects, larvae, worms, snails etc., which in turn fed on dead plant matter. So they survived the collapse of plant-based food chains because they lived in " detritus-based" food chains.
  • In stream communities few groups of animals became extinct. Stream communities tend to be less reliant on food from living plants and are more dependent on detritus that washes in from land. The stream communities may also have been buffered from extinction by their reliance on detritus-based food chains. (See Sheehan and Fastovsky, Geology, v. 20, p. 556-560.)
  • Similar, but more complex patterns have been found in the oceans. For example, animals living in the water column are almost entirely dependent on primary production from living phytoplankton. Many animals living on or in the ocean floor feed on detritus, or at least can switch to detritus feeding. Extinction was more severe among those animals living in the water column than among animals living on or in the sea floor.
  • No land animal larger than a cat survived.
  • The largest survivors, crocodilians and champsosaurs, were semi-aquatic. Modern crocodilians can live as scavengers and can survive for as long as a year without a meal.

How long did the K-T extinction take?

This is a controversial issue, because some theories about the extinction's causes require a rapid extinction over a relatively short period (from a few years to a few thousand years). And it is difficult to resolve because of:

  • The Signor-Lipps effect, i.e. the fossil record is so incomplete that most extinct species probably died out a long time after the most recent fossil that has been found.
  • The shortage of continuous beds of fossil-bearing rock which cover a time range from several million years before the K-T extinction to a few million years after it.

At present the best sequence of fossil-bearing rocks known is in Montana, USA (the Hell Creek, Lance Formation and Scollard Formation), running from about 83.5 MYA (million years ago) to 64.9 MYA and covering the Campanian and Maastrichtian ages of the Cretaceous and the beginning of the Paleocene period. They show changes in dinosaur populations over the last 18M years of the Cretaceous:

  • Some groups declined and others grew more diverse.
  • In the middle-late Campanian these formations show a greater diversity of dinosaurs than any other single group of rocks.
  • There is no obvious reduction in dinosaur diversity, not even in the latest part of the Maastrichtian (Fastovsky and Sheehan 1995 and later papers). And the late Maastrictian rocks contain the largest members of almost every major clade: Tyrannosaurus, Ankylosaurus, Pachycephalosaurus, Triceratops and Torosaurus. This suggests food was plentiful not long before the extinction.

In the sediments below the K-T boundary the dominant plant remains are angiosperm pollen grains, but the actual boundary layer contains no pollen and is dominated by fern spores. Normal pollen levels resume immediately above the boundary layer. This is reminiscent of areas blighted by volcanic eruptions, where the recovery is led by ferns which are later replaced by larger angiosperm plants.

Although the Hell Creek, Lance and Scollard formations provide a wealth of information, they cover a relatively small area and it is dangerous to assume that they tell us what happended world-wide.


A good theory of the K-T extinction should:

  • explain all of the losses, not just focus on a few groups such as dinosaurs.
  • explain the selectivity of the extinction, i.e. why particular groups of organisms died out and why others survived.
  • provide killing mechanisms which are strong enough to cause a mass extinction but not a total extinction.
  • be based on events or processes that can be shown to have happened, not just inferred from the extinction.

Alvarez hypothesis

Artistic depiction of asteroidal impact
Artistic depiction of asteroidal impact

In 1980, a team of researchers led by Nobel-prize-winning physicist Luis Alvarez, his son geologist Walter Alvarez and chemists Frank Asaro and Helen Michels discovered that sedimentary layers found all over the world at the Cretaceous-Tertiary boundary contain a concentration of iridium hundreds of times greater than normal. Iridium is extremely rare in the earth's crust because it is very dense, and therefore most of it sank into the earth's core while the earth was still molten. The Alvarez team suggested that an asteroid struck the earth at the time of the K-T boundary. The impact theory can also be traced back to M. W. DeLaubenfels' Dinosaur Extinctions: One More Hypothesis a paper published in the Journal of Paleontology, Vol 30, No 1, p 207-218 January 1956. There is some question as to why the Alvarez's 1980 paper does not give any credit to DeLaubenfels.

The Alvarez impact theory idea is supported by the composition of the K-T boundary layer:

  • chondritic meteorites and asteroids contain a much higher concentration than the earth's crust because they have about the same concentration of iridium as the whole earth.
  • the isotopic composition of iridium in asteroids is similar to that of the K-T boundary layer but differs from that of iridium in the earth's crust.
  • chromium isotopic anomalies found in Cretaceous-Tertiary boundary sediments also strongly support the impact theory and suggest that the impact object must have been an asteroid or a comet composed of material similar to carbonaceous chondrites.
  • shocked quartz granules, glass spherules and tektites are common, especially in deposits from around the Caribbean.
  • all of these constituents are embedded in a layer of clay, which the Alvarez team interpreted as the debris spread all over the world by the impact.

The Alvarez team then estimated:

  • the total amount of iridium in the K-T layer.
  • the size of the asteroid, assuming that it contained the normal percentage of iridium found in chondrites. The answer was about 10km (6 miles) in diameter, about the size of Manhattan. Such a large impact would have had approximately the force of 100,000,000,000,000 tones of TNT, i.e. about 2,000,000 times as great as the most powerful H-bomb ever tested.

The most obvious consequence of such an impact would be a vast dust cloud which would block sunlight and prevent photosynthesis for a few years. This would account for the extinction of plants and phytoplankton and of all organisms dependent on them (including predatory dinosaurs as well as vegetarians). But small creatures whose food chains were based on detritus would have a reasonable chance of survival.

Global firestorms may have resulted as incendiary fragments from the blast fell back to Earth. Analyses of fluid inclusions in ancient amber suggest that the oxygen content of the atmosphere was very high (30-35%) during the late Cretaceous . This high O2 level would have supported intense combustion. The level of atmospheric O2 plummeted in the early Tertiary Period. If widespread fires occurred, they would have increased the CO2 content of the atmosphere and caused a temporary greenhouse effect once the dust cloud settled, and this would have exterminated the most vulnerable survivors of the "long winter".

The impact may also have produced acid rain, depending on what type of rock the asteroid struck. However, recent research suggests this effect was relatively minor. Chemical buffers would have limited the changes, and the survival of animals vulnerable to acid rain effects (such as frogs) indicate this was not a major contributor to extinction (see Kring, D.A. GSA Today v. 10, no.8).

Impact theories can only explain very rapid extinctions, since the dust clouds and possible sulphuric aerosols would wash out of the atmosphere in a fairly short time - possibly under 10 years.

Although further studies of the K-T layer consistently show the excess of iridium, the idea that the dinosaurs were exterminated by an asteroid remained a matter of controversy among geologists and paleontologists for more than a decade.

Chicxulub Crater

Radar topography reveals the 180 kilometre (112 mile) wide ring of the crater
Radar topography reveals the 180 kilometre (112 mile) wide ring of the crater

One problem with the "Alvarez hypothesis" (as it came to be known) was that no documented crater matched the event. This was not a lethal blow to the theory; although the crater resulting from the impact would have been 150 to 200 kilometres in diameter, Earth's geological processes tend to hide or destroy craters over time.

But subsequent research found what many thought was "the smoking gun" - the Chicxulub Crater buried under Chicxulub on the coast of Yucatan This crater is oval, with an average diameter of about 180km, about the size calculated by the Alvarez team. Its shape and location indicate further causes of devastation in addition to the dust cloud:

  • the asteroid landed right on the coast and would have caused gigantic tsunamis, for which evidence has been found all round the coast of the Carribbean and eastern USA - marine sand in locations which were then inland, and vegetation debris and terrestrial rocks in marine sediments dated to the time of the impact.
  • the asteroid landed in a bed of gypsum (calcium sulphate), which would have produced a vast sulphur dioxide aerosol. This would have further reduced the sunlight reaching the earth's surface and then precipitated as acid rain, killing vegetation, plankton and organisms which build shells from calcium carbonate (notably some plankton species and many species of mollusk).
  • the crater's shape suggests that the asteroid landed at an angle of 20° to 30° from horizontal and travelling north-west. This would have concentrated most of the blast and solid debris in the central part of the USA.

Most paleontologists now agree that an asteroid did hit the Earth about 65 million years ago, but many dispute whether the impact was the sole cause of the extinctions.

Gerta Keller suggests that the Chicxulub impact occurred approximately 300,000 years before the K-T boundary. This dating is based on evidence collected in Northeast Mexico, detailing multiple stratigraphic layers containing impact spherules, the earliest of which occurs some 10 metres below the K-T boundary. This chronostratigraphic thickness is thought to represent 300,000 years. This finding supports the theory that one or many impacts were contributary, but not causal, to the K-T boundary mass extinction. However, many scientists reject Keller's analysis, some arguing the 10 metre layer on top of the impact spherules should be attributed to tsunami activity resulting from impact. The Chicxulub crater remains in the centre of a very large controversy.

Deccan Traps

Several scientists think the extensive volcanic activity in India known as the Deccan Traps may have been responsible for, or contributed to, the extinction. A partial reason for the rejection of the impact theory may have been a certain general distrust that a group of physicists was intruding into the paleontologists' domain of expertise.

Before 2000, arguments that the Deccan Traps flood basalts caused the extinction were usually linked to the view that the extinction was gradual, as the flood basalt events were thought to have started around 68MYA and lasted for over 2M years. But Hofman, Féraud and Courtillot (2000) provided evidence that two-thirds of the Deccan Traps were created in 1M years about 65.5.MYA. So these eruptions would have caused a fairly rapid extinction, over a period of thousands of years - but still much slower than one caused entirely by an impact.

The killing mechanisms would have been:

  • dust which blocked sunlight and stopped photosynthesis.
  • sulphur gases which first formed aerosols which also blocked sunlight and then precipitated as acid rain.
  • carbon dioxide emissions which would have increased the greenhouse effect when the dust and aerosols cleared.

In the years when the Deccan Traps theory was linked to a slower extinction, Luis Alvarez (who died in 1988) replied that paleontologists were being misled by sparse data. His assertion did not go over well at first, but later intensive field studies of fossil beds lent weight to his claim. Eventually, most paleontologists began to accept the idea that the mass extinctions at the end of the Cretaceous were largely or at least partly due to a massive Earth impact. However, even Walter Alvarez has acknowledged that there were other major changes on Earth even before the impact, such as a drop in sea level and massive volcanic eruptions in India (Deccan Traps sequence) and these may have contributed to the extinctions.

A very large crater has been recently reported in the sea floor off the west coast of India 2. This, the Shiva crater, 450-600 kilometres in diameter, has also been dated at about 65 million years at the K-T boundary. The researchers suggest that the impact may have been the triggering event for the Deccan Traps. However, this feature has not yet been accepted by the geologic community as an impact crater and may just be a sinkhole depression caused by salt withdrawal. .

Multiple impact event

Several other craters also appear to have been formed at the K-T boundary. This suggests the possibility of near simultaneous multiple impacts, perhaps from a fragmented asteroidal object, similar to the Shoemaker-Levy 9 cometary impact with Jupiter.

  • Boltysh crater (24 km diam., 65.17 ± 0.64 Ma old) in Ukraine
  • Silverpit crater (20 km diam., 60-65 Ma old) in the North Sea
  • Eagle Butte crater (10 km diam., < 65 Ma old) in Alberta, Canada
  • Vista Alegre crater (9.5 km diam., < 65 Ma old) in Paraná State, Brazil

Note: Ma (" mega-annum") means million years.

Maastrichtian Regression

There is clear evidence that sea levels fell in the final stage of the Cretaceous by more than at any other time in the Mesozoic era:

  • in some Maastrichtian rock sequences from various parts of the world the latest rocks are terrestrial, earlier ones represent shorelines and the earliest represent seabeds).
  • these layers do not show the tilting and distortion associated with mountain building, hence by far the likeliest explanation is a regression (drop in sea level).

There is no direct evidence for the cause of the regression, but most likely the mid-ocean ridges became less active and therefore sank under their own weight.

A severe regression would have greatly reduced the continental shelf area, which is the most species-rich part of the sea, and therefore could have been enough to cause a marine mass extinction.

It would also have caused climate changes, partly by disrupting winds and ocean currents and partly by reducing the earth's albedo and therefore increasing global temperatures. These would have caused some extinctions on land, especially among vegetarians because of changes in the vegetation available. But the north American Maastrichtian fossil record for dinosaurs shows:

  • continued high diversity with gains and losses rather than a prolonged mass extinction.
  • a continuing increase in dinosaur sizes, which suggests the total food available was not reduced even if its composition changed.

Supernova Hypothesis

Another proposed cause for the K-T extinction event was cosmic radiation from a relatively nearby supernova explosion. The iridium anomaly at the boundary could support this hypothesis. The fallout from a supernova explosion should contain the plutonium isotope Pu-244, the longest-lived plutonium isotope ( half-life 81 Myr), that is not found in earth rocks. However, analysis of the boundary layer sediments revealed the absence of Pu-244, thus essentially disproving this hypothesis.

Composite theories

The two best-supported theories, based on the Chixculub impact and the Deccan Traps, are not mutually exclusive in the present stage of our knowledge:

  • We only know of one sequence of rocks, the Hell Creek and Lance formations around Montana (USA), which gives a detailed and continuous record of the final stages of the Cretaceous. The evidence of these rocks appears to favour a very quick extinction, most probably caused by the Chixculub impact.
  • We do not know how fast the extinction was in other parts of the world. There is good reason to hope that discoveries in China will add to our knowledge of the K-T extinction. But we have virtually no information about what happened in the southern hemisphere.
  • It is not certain that a catastrophe in the northern hemisphere would have been able to cause a mass extinction in the southern hemisphere - in to-day's earth the two hemispheres share a single ocean current system but have largely separate wind systems, which would have made it difficult for debris from Chixculub to cause a "long winter" in the south. Perhaps the southern extinction was mainly caused by the Deccan Traps at the same time but rather more slowly.

Retrieved from ""