Extinction :sx bad genes or bad luck ?sx Many more species have become extinct over the time than survive today .sx Scientists have only just begun to ask why .sx David Raup .sx COUNTLESS species of plants and animals have existed in the history of life on Earth .sx Estimates of the total progeny of evolution range from 5 to 50 billion species .sx Yet , only an estimated 5 to 50 million species are alive today - a rather poor survival record .sx With , at the most , only one in every thousand species surviving , what happened to the others ?sx Even though approximately the same number of species have become extinct as have originated during the progress of evolution , scientists have given their attention almost exclusively to origination .sx Why ?sx It seems as silly as a demographer ignoring mortality rates or a physiographer ignoring erosion .sx But whatever the reason , we still know woefully little about the death of species .sx A species can die outright when all its members die without issue .sx This is called true extinction , and is typified by the disappearance of dinosaurs in the late Cretaceous - whether that extinction was completed in an afternoon or covered several million years .sx But species can also disappear by evolving into something different .sx The ancestral species becomes extinct only in the sense of having been transformed out of existence .sx This is called pseudo-extinction , and is of less concern here .sx The proportions of true and pseudo-extinction in the history of life are not known .sx We can be sure , however , that true extinction has been the fate of large numbers of species .sx Minimum estimates came from groups of organisms that were once highly diverse but died out completely .sx For example , the ammonites ( swimming molluscs of the Mesozoic ) had thousands of co-existing species at the height of their reign .sx Each species constituted a separate lineage - a genome - and the ammonites left no descendants .sx So , regardless of how many transformations and pseudo-extinctions occurred among ammonites , the number of true extinctions cannot be less than the number of separate species living at the time of the ammonites' peak diversity .sx The same reasoning can be applied to the trilobites of the Palaeozoic and to many other large groups documented trough fossils .sx The inescapable conclusion is that true extinction of species has been common in the history of life .sx Yet very little is known about the process .sx Standard text - books on evolutionary biology and palaeobiology hardly mention extinction .sx Much is said about the origin of species and the evolution of species once they are formed , but discussions of extinction are generally limited to casual references to the enigma of the great mass extinctions .sx On causes of extinction , we are apt to read vacuous statements like " Species become extinct when population sizes drop to zero " or " Species die out if they are unable to adapt to changing conditions " .sx The 1987 edition of the Encyclopaedia Britannica says :sx " Extinction occurs when a species can no longer reproduce at replacement levels .sx " These statements are true , of course , but are virtually devoid of content .sx Will Cuppy , in his volume of essays entitled How to Become Extinct , wrote :sx " The Age of Reptiles ended because it had gone on long enough and it was all a mistake in the first place .sx " He followed this by noting :sx " Bats are going to flop , too , and everybody knows it except the bats themselves .sx " It seems impossible to escape the implication that extinction is a mark of failure - failure to compete with better qualified species for resources or failure to adapt to changing physical conditions .sx Extinction-as-failure is deeply embedded in Darwin's writings and in neo-Darwinian theory .sx Although the central argument of Darwin's On the Origin of Species was that of gradual evolution within species by natural selection , he made frequent reference to the importance of extinction and its constructive role in evolution .sx At one point in On the Origins of Species , he wrote :sx " The inhabitants of each successive period in the world's history have beaten their predecessors in the race for life , and are , insofar , higher in the scale of nature .sx " .sx Some observers even include selective extinction of species ( and higher groups ) as part of the natural selection process - although the purist restricts natural selection to changes within species at the level of the breeding population .sx Did the dinosaurs deserve to die ?sx The proposition that extinction results from the failure of a species is appealing and highly reasonable , perhaps so reasonable that testing is not required .sx What could extinction be if not failure ?sx But we still must ask whether the victims of extinction were , in fact , less well adapted than the survivors .sx Could we predict victims and survivors ?sx A case in point is that the extinction of dinosaurs and the survival of mammals at the end of the Cretaceous .sx The two groups had coexisted for something like 140 million years , yet all dinosaur species died out , while enough meal species survived to spawn the great diversification of mammals that led to the 4000 mammal species living today , including Homo sapiens .sx Did the dinosaurs deserve to die ?sx Were they too large , too stupid , or unable to control their body temperatures efficiently enough ?sx There are so many differences between reptiles and mammals that almost any difference can be identified ( and has been ) as the cause of the survival of one but not the other .sx But the sad truth is that there is no hard evidence , other than the fact of the extinctions , for the inferiority of the victims .sx Thus , although extinction-as-failure remains highly plausible , we have no independent confirmation and we are justified in exploring other ways of becoming extinct .sx In these days of heightened concern about endangered species and habitat destruction by humans , the fact that species become extinct may seem unsurprising .sx Extinction , it seems , is 'easy' .sx And if one subscribes to the view that a community of species depends on an intricate and delicately balanced set of interdependencies , then it follows that even minor disturbances will threaten members of the community with extinction .sx The causes of extinction should be as many and varied as there are things that can go wrong in a community , ranging from the purely biological - the invasion of a predatory species from elsewhere , for example - to the purely physical - such as a severe frost or forest fire .sx This general line of thinking has led to the conventional view that the causes of extinction are so complex and so varied that the phenomenon defies scientific inquiry .sx Also , if extinction in the natural world is so common ( and 'easy' ) , why study it ?sx There must be more challenging research topics .sx The conventional view certainly applies in local areas where species have tiny geographic ranges .sx In fact , the classic research on islands by Robert MacArthur of Princeton University and E.O. Wilson of Harvard University ( published in their 1967 book Island Biogeography ) has shown how common extinction is in small island communities .sx Wilson's student Daniel Simberloff followed this up by monitoring changing species on mangrove islands in Florida and found local extinction to be a normal event .sx Few attempts have been made to find out why or how particular species become extinct on small islands , simply because the candidate causes are many and varied .sx But can this view of extinction be applied to the ammonites , dinosaurs , trilobites and other organisms that flourished in the past ?sx I think not .sx About 250 000 fossil species have been found , described , and named , but this is no more than one in every 20 000 species that have lived ( using the low estimate of 5 billion for total progeny) .sx So , the likelihood of any one particular species being preserved in fossil form is negligible .sx It follows that the fossil record must be strongly biased in favour of species that were abundant and geographically widespread .sx They are not the tiny , localised species that go extinct regularly on small islands .sx The average duration of fossil species is about four million years .sx This is undoubtedly much longer than the average existence of all species in the history of life .sx In ecological terms , four million years is an incredibly long time , and suggests that these species were well able to survive the normal vicissitudes of their natural environments .sx Forest fires , unusual frosts or viral epidemics might have killed off some breeding populations , but left enough elsewhere for the species to survive .sx Survival for such a long time suggests either that most successful species are adapted to normal stresses or that they are protected from these stresses by having wide and complex geographical distributions .sx In fact , David Jablonski of the University of Chicago has shown that a large geographic range correlates with the ability to survive for a long time in marine molluscs of the Cretaceous prior to the mass extinction .sx Another important aspect of the record of extinction is that it is highly episodic .sx That is , extinctions are far more clustered in time than would be predicted if each extinction were independent of the others .sx The several large mass extinctions are obvious examples - with two-thirds , or more , of species becoming extinct in a geologically short time - but even the smaller pulses of extinction are more than the chance coincidence of independent events .sx Computer simulations that maintain a constant probability of extinction through time do not yield the concentrated pulses of extinction observed in the fossil record .sx The episodic nature of extinctions divides the geological record into well-defined 'packages' bounded by short intervals when there was a rapid turnover of species .sx It is probably this feature of the record that made it possible for the geologists of the early 19th century to define , in just a few decades , a fossil-based chronology that is recognisable around the world .sx Pulses of extinction typically cut across ecological lines and cover wide geographic areas .sx The mass extinction of the Cretaceous , for example , devastated land vertebrates ( all dinosaurs and a substantial fraction of mammals ) , marine reef communities , pelagic marine animals ( ammonites and swimming reptiles ) , as well as important species of marine plankton .sx Although some groups of organisms survived on both land and sea , the main point is that the extinction pulses recorded by fossils took place across a vastly wider range of habitats than the localised extinctions that have been studied in modern communities such as islands .sx Stresses that are outside normal experience .sx This suggests several things about the extinctions we see in the fossil record .sx First , killing a successful species is not 'easy' , even though it is common on geological time scales .sx Secondly , the stresses causing extinction must be outside the normal experience of the species - the stresses must be so rare as to be beyond the reach of the adaptive power of natural selection .sx And lastly , at least some of the stresses causing extinction must simultaneously affect many habits and modes of life .sx All this makes extinction of the type seen in the geological past very difficult for the biologist to study .sx The present-is-the-key-to-the-past approach that has worked so well for geologists may not apply here - and may possibly be misleading .sx We have been observing nature for a few thousand years , yet complex , multicellular life goes back millions of years .sx It may be that , by pure chance and good fortune , we have not observed the conditions most responsible for the extinction of species .sx In fact , it is rather arrogant for us to assume that our tiny slice of the Earth's history includes a representative sample of that history .sx In colonial America , the heath hen ( Tympanuchus cupido cupido ) was extremely common , with a range that extended from Maine to Virginia .sx It was edible and easy to kill , leading to extensive hunting by the expanding human population .sx The species was so devastated by hunting that by 1870 , the sole survivors were on the small island of Martha's Vineyard , off Cape Cod .sx In 1908 , a 1600-acre refuge was established there as a sanctuary for the remaining fifty birds .sx Protection was successful to the extent that by 1915 , the heath hens occupied the entire island and numbered about two thousand .sx Then , in 1916 , the population suffered the coincidence of several disasters :sx there was a fire spread by strong winds that eliminated much of the breeding ground , an especially hard winter immediately following the fire , an influx of predatory goshawks and , finally , a poultry disease introduced from domestic turkeys .sx