Extinction is the act or process of the dying‐out of a species or evolutionary lineage. It is one of the most common of all ecological/evolutionary processes and represents an inevitable corollary of evolution by natural selection. Information on extinction comes from many sources, including laboratory experiments, field studies and the fossil record. The causes of short‐term extinctions involving few species can, in principle, be well understood. The causes of long‐term extinction events, such as those preserved in the fossil record, remain controversial. Although the contemporary biodiversity crisis is often referred to as a ‘sixth (mass) extinction’, levels of documented losses over the last 400 years have been modest by geological standards. Nevertheless, historical extinction rates are well above palaeontological background rates and are predicted to increase substantially over the next 200 years. Whether the number of species losses over this interval will reach geological magnitudes remains uncertain.

Key Concepts:

  • Extinction is a common and perfectly natural process.

  • The processes that lead to extinction operate at a variety of spatial, ecological and temporal scales.

  • Extinctions can be studied in the laboratory, in the field (modern) and in the fossil record (ancient); however, the ability of researchers to determine extinction patterns accurate and evaluate causal models in detail decreased as the spatial and temporal scales along those spectrum increase and as the biology of the organisms in question grows less familiar.

  • Extinction processes can also be studied via mathematical simulation models. In many (though not all) cases, there is good agreement between the predictions of these mathematical models and field observations.

  • The fossil record of extinction events exhibits two compelling patterns: (1) a historical pattern of long periods of relatively modest (= background) extinction intensities over time, interspersed by short, nonperiodic intervals of elevated extinction intensity and (2) a secular decline in the magnitude of background extinction intensity from the mid‐Palaeozoic to the Recent.

  • Lower‐level peaks in extinction intensity are probably due to the operation of single major or idiosyncratic combinations of moderate intensity process that foster long‐term environmental change (e.g. sea‐level change, continental glaciation, large igneous province volcanism and asteroid/comet impact).

  • Higher‐level peaks in extinction intensity (= mass extinctions) are probably due to idiosyncratic combinations of major process that foster long‐term environmental change (e.g., sea‐level change, continental glaciation, large igneous province volcanism and asteroid/comet impact).

  • Over the last 400 years, extinctions have occurred at a rate that is well above background rates as inferred from the fossil record, but well below the range of even moderate geological extinction events. Models of maximum extinction intensities predicted over the next 200 years are matters for concern, but fall well below the intensities achieved by geological mass extinction events.

Keywords: population; geography; environment; palaeontology; diversity

Figure 1.

Histogram of Phanerozoic extinction intensities for all stages. Red line represents the long‐term linear trend of the data (background extinction‐intensity gradient). Arrows mark stages characterised by local extinction peaks (so‐called mass extinction events). Abbreviations: Sil., Silurian; Carbon., Carboniferous; Perm., Permian. Copyright by Norman MacLeod.

Figure 2.

Genus‐level, extinction‐intensity record for Phanerozoic stages represented as a scatterplot with estimated background extinction gradients shown as linear trend lines with associated 95% confidence bands (dashed) and ‘gradient events’ shown as labelled vertical lines. Solid circles indicate Permian–Recent mass extinctions. The Late Devonian gradient event (Event I, estimated timing 250 Ma) represents the oldest point at which the extinction‐intensity gradient loses coherence. Stages to the left (older) of this line exhibit no statistically significant extinction‐intensity gradient (horizontal line in this segment represents the null model and is within the 95% confidence‐interval bands). Stages to the right (younger) of this line exhibit a highly significant gradient of linearly decreasing extinction intensities over time. A second gradient event (Event II estimated timing 209 Ma) marks the point where inter‐stage variability in extinction‐intensity data collapses. Copyright by Norman MacLeod.



MacLeod N (1998) Impacts and marine invertebrate extinctions. In: Grady MM, Hutchinson R, McCall GJH and Rotherby DA (eds) Meteorites: Flux with Time and Impact Effects, pp. 217–246. London: Geological Society of London.

Raup DM (1991) Extinction: Bad Genes or Bad Luck. New York: W W Norton & Co.

Vermeij GJ (2004) Ecological avalanches and the two kinds of extinction. Evolutionary Ecology Research 6: 315–337.

Further Reading

Courtillot V (1999) Evolutionary Catastrophes: The Science of Mass Extinction. Cambridge: Cambridge University Press.

Eldredge N (1998) Life in the Balance: Humanity and the Biodiversity Crisis. Princeton: Princeton University Press.

Hallam A and Wignall PB (1997) Mass Extinctions and their Aftermath. Oxford: Oxford Science Publications.

Howes C (1997) The Spice of Life: Biodiversity and the Extinction Crisis. London: Blandford.

Lawton JH and May RM (eds) (1995) Extinction Rates. Oxford: Oxford University Press.

MacLeod N (2013) The Great Extinctions: What Causes Them and How They Shape Life. London: The Natural History Museum.

Pearson R (2011) Driven to Extinction: The Impact of Climate Change on Biodiversity. London: The Natural History Museum.

Contact Editor close
Submit a note to the editor about this article by filling in the form below.

* Required Field

How to Cite close
MacLeod, Norman(Oct 2012) Extinction. In: eLS. John Wiley & Sons Ltd, Chichester. http://www.els.net [doi: 10.1002/9780470015902.a0001650.pub3]