Product has been added to the basket

Flood basalts, mantle plumes and mass extinctions

By Steve Self, Open University, Milton Keynes, UK and Mike Rampino, Earth & Environmental Science Program, New York University

Flood Basalts


The great continental flood-basalt eruptions of the geological past are the largest eruptions of lava on Earth, with known volumes of individual lava flows exceeding 2000 cubic kilometres. For comparison, the ongoing eruption of Kilauea volcano on Hawaii has produced just 1.5 cubic kilometres in 16 years! A series of these huge eruptions builds up a thick stack of basalt lava flows (as shown in the photograph) extending over areas of more than a million square kilometres.

Flood basalts are one type of large igneous province (LIP) that characterise the Earth's surface and have been formed at various times in the geological past - some in a submarine environment and some on land (see LIP map).

Notable examples are the Siberian Traps and the Deccan Traps (trap is a Sanskrit word meaning 'step', referring to the step-like topography produced by the stacked layers of lava). The Columbia River province featured in the above photograph is minute in comparison to the size of these enormous outpourings of lava.

back to top

Mantle Plumes

The estimated dates of the younger continental flood basalts compiled from recent sources are shown in Table 1. Several lines of evidence suggest that in most instances the greatest number of individual eruptions and the largest volumes of lava probably occurred within a million years or less. 

Large igneous provinces are thought to be caused by the arrival of a mantle plume in the Earth's outermost layer, the lithosphere. The plumes are proposed to be richer in lighter elements and hotter than the surrounding mantle. As they rise, magma (liquid rock) is generated by partial melting of the plume material. The magma is injected into the lithosphere and erupted onto the Earth's surface to form huge basalt lava flows. The first few million years of a newly arrived mantle plume seem to be the most fertile in terms of magma production and flood basalts are therefore formed in a very short period of geological time.The surface manifestations of mantle plumes are often called hotspots. 


Plumes are thought to originate very deep in the Earth - perhaps at the core-mantle boundary for the larger ones and at a depth of about 600 km deep for the smaller ones - but they seem to be related to the breakup of continents (rifting), so there is some influence from global plate tectonic processes. (A numerical simulation of the development of a mantle plume has been developed by Paul Tackley of UCLA). 

A 'snapshot' view of a theoretical situation involving several plumes in varying stages of development distributed around the globe is shown in the figure (left), kindly provided by Paul Tackley of UCLA. This may represent a typical pattern of plume distribution at any period of geological time.

The estimated dates of the younger continental flood basalts compiled from recent sources are shown in Table 1. Several lines of evidence suggest that in most instances the greatest number of individual eruptions and the largest volumes of lava probably occurred within a million years or less.

Province Age (Myr) Volume (10E6 Km^3) Paleolatitude (degrees) Duration (Myr)
Columbia River  16 ± 1  0.25  45 N  ~ 1 (for 90%) 
Ethiopia  31 ± 1  ~ 1.0  10 N  ~ 1 
North Atlantic  57 ± 1  >1.0  65 N  ~ 1 
Deccan  66 ± 1  >2.0  20 S  ~ 1 
Madagascar  88 ± 1  45 S  ~ 6? 
Rajmahal  116 ± 1  50 S  ~ 2 
Serra Geral/ Etendeka  132 ± 1  >1.0  40 S  ~ 1 or ~ 5? 
Antarctica  176 ± 1  >0.5  50-60 S  ~ 1? 
Karoo  183 ± 1  >2.0  45 S  0.5 - 1 
Newark  201 ± 1  >1.0?  30 N  ~ 0.6 
Siberian  249 ± 1  >2.0  45 N?  ~ 1 
Table 1: Flood Basalt Provinces of the last 250 Myrs

back to top

Flood Basalts & Mass Extinctions

Every now and again in geology, as in any other science, evidence is obtained and presented that cannot easily be explained in terms of familiar processes or accepted ideas. Such a case was continental drift, proposed by Wegener in 1912, which languished as a theory for about 45 years because there was no logical explanation of HOW continents could move. The mechanism of seafloor spreading proposed in the late 1950's led to the development of the modern theory of plate tectonics, which provides an explanation for continental drift.

The time relationship between flood basalt province formation and mass extinctions of organisms is another example of a scientific "hard nut to crack."

Extinction events are increasingly seen as important factors in the history of life on Earth, and recent studies suggest catastrophic causes for at least some mass extinctions. Two catastrophic processes that have been invoked are (1) impacts of asteroids or comets and (2) large volcanic eruptions. The end-Cretaceous (Cretaceous/Tertiary or K/T boundary) mass extinction has been convincingly correlated with the impact of a 10-km diameter asteroid with the Earth about 65 million years ago. Evidence of similar impacts has been found at the times of several other extinction events.

Table 2 compares the LIP ages given in Table 1 with the estimated ages of stratigraphic boundaries involving significant biotic changes, dated according to the most recent geological time scale. In at least three cases (the Deccan, Newark, and Siberian flood basalts), a direct measure of correlation with major extinction events is possible. The probability that three major volcanic events that typically last ~1 Myr should occur within 1 Myr of major extinction events during the last 250 Myr (of which there are ~12) is about 10-4. Thinking about the ways in which these two types of global event might be causally linked is a worthy scientific challenge.
Flood Basalt Episode Age (Myr) Stratigraphic Boundary Age (Myr)
Columbia River  16 ± 1  Early/Mid-Miocene  16.4 
Ethiopia  31 ± 1  Early/Late Oligocene  30 
North Atlantic  57 ± 1  Paleocene/Eocene (Thanetian/Selandian)  54.8 (57.9) 
Deccan  66 ± 1  Cretaceous/Tertiary  65.0 ± 0.1 
Madagascar  88 ± 1  Cenomanian/Turonian (Turonian/Coniacian)  93.5 ± 0.2 (89 ± 0.5) 
Rajmahal  116 ± 1  Aptian/Albian  112.2 ± 1.1 
Serra Geral/Etendeka  132 ± 1  Jurassic/Cretaceous (Hauterivian/Valanginian)  142 ± 2.6 (132 ± 1.9) 
Antarctica  176 ± 1 or 183 ± 1  (Aalenian/Bajocian)  (176.5 ± 4) 
Karoo  183 ± 1  Early/Middle Jurassic  180.1 ± 4 
Newark  201 ± 1  Triassic/Jurassic  205.7 ± 4 
Siberian  249 ± 1  Permian/Triassic  248.2 ± 4.8 
Table 2: Flood Basalt Episodes and Faunal Events

back to top

Environmental Effects of Flood Basalt Eruptions

If there is a causal link between flood basalt events and mass extinctions, it may lie in the environmental impact of the gases released, because basalt eruptions are not particularly explosive. Several kinds of environmental effects have been suggested, including climatic cooling from sulphuric acid aerosols, greenhouse warming from CO2 and SO2 gases, and acid rain. Basaltic magmas are often very rich in dissolved sulphur, and sulphuric acid aerosols formed from sulphur volatiles (largely SO2) are injected into the stratosphere by convective plumes rising above volcanic vents and fissures.

Indirect environmental effects include changes in ocean chemistry, circulation, and oxygenation, especially from basaltic volcanism associated with large submarine oceanic plateaus that may represent flood basalt eruptions in an oceanic environment.

A major uncertainty is the nature and severity of the environmental effects of the eruptions and their potential impact on life. Although the correlation between some flood basalt episodes and extinctions may implicate volcanism in the extinctions, it is also possible that other factors lead to an apparent association. Flood basalt episodes have been attributed to mantle plume activity, and thus may represent one facet of a host of related global geological factors (eg, changes in sea-floor spreading rates, rifting events, increased tectonism and volcanism, sea-level variations) that tend to be correlated, and may be associated with unusual climatic and environmental fluctuations that could lead to significant faunal changes. It has also been suggested that a coincidence of both a large impact and a flood basalt eruption might be necessary in causing severe mass extinctions

back to top

Missing Pieces of the Puzzle

How often do flood basalt lavas occur? It would be desirable to be able to obtain age dates on individual lavas in a flood basalt pile to determine how often these huge eruptions occurred, but the dating methods used do not have sufficient accuracy to pinpoint individual flows in time - virtually a whole flood basalt field fits within the errors on some of the dates.

Were ancient flood basalts associated with major outgassing? Due to their age and altered state, the amount and signature of the minor volatile elements in these old lavas will be difficult to obtain. It is possible to show whether the eruptions degassed vast amounts of sulphur and carbon dioxide if these species can be measured, as they can in some of the younger lavas such as those in the Columbia River province. More accurate modelling of dense atmospheric aerosol clouds, and their effects on atmospheric dynamics and chemistry, is needed before the likely climatic impact of events such as flood basalts can be properly estimated

back to top

Further Reading

  • Coffin, M F & Eldholm, O 1994 Large igneous provinces: Crustal structure, dimensions, and external consequences. Reviews of Geophysics, 32, p. 1-36.
  • Erwin, D H 1994 The Permian-Triassic extinction. Nature, 367, p. 231-236.
  • Mahoney, J J & Coffin, M F (eds.) 1997 Large Igneous Provinces: Continental, Oceanic and, Planetary Flood Volcanism. AGU Geophysical Monograph, 100.Rampino, M R, Self, S & Stothers, R B 1988 Volcanic winters. Annual Review of Earth and Planetary Science, 16, p. 73-99.Rampino, M R & Stothers, R B 1988 Flood basalt volcanism during the past 250 million years. Science, 241, p. 663-668.White, R S & McKenzie, D 1989 Volcanism at rifts. Scientific American, 260, 62-71.White, R S & McKenzie, D 1995 Mantle plumes and flood basalts. Journal of Geophysical Research, 100, p 17543-17585.

back to top