Most of the species that ever existed on the Earth became extinct at some time in the past. It has been estimated that the number of species existing at the present time corresponds to the order of one per mille to one percent of the number of extinct species (Raup 1994). It is rather difficult to perform a more exact quantitative estimate as the probability that a certain species will be found in the paleontological record is directly proportional to the duration of its existence. It is thus apparent that the major part of preserved fossils will correspond to species that survived for a sufficiently long time on the Earth and, to the contrary, we will learn nothing from the paleontological record about the existence of short-lived species. Similarly, it is obvious that a great many of the traits, on the basis of which modern species are distinguished within a certain taxon, are apparent only on the soft parts of their bodies and were thus not preserved in fossils. Nonetheless, the paleontological data clearly demonstrate that species are formed at a certain moment, survive for a longer or shorter time and finally disappear, become extinct. The situation is similar for higher taxa – however, they understandably exist far longer than individual species and become extinct only when their last members become extinct.
At a general level, the most common cause of extinction of a species is “bad luck”, the fact that it was present at the wrong time in the wrong place. Losing the coevolutionary battle with another species apparently occurs far less often. It seems that, at the very least for tetrapod vertebrates, new species mostly expanded into a previously vacated ecological space (Benton 1996). Apparently the vast majority of extinctions occurred in that, at a certain moment, the species was suddenly exposed to conditions that it had not encountered in the past and to which it was thus not adapted. Of course, amongst other things, this means that it is not species and developmental lines that were best able to adapt to the current conditions in their environments that survive for long times, but rather species and evolutionary lines that were fortunately able to survive in situations that they had never encountered before. Thus, paleontologists often point out the contrast between microevolution, based on survival of the fittest, and macroevolution, based rather on survival of the luckiest. Obviously, the degree to which this sharp division corresponds to facts could be a matter for long discussions. In any case, compared to microevolutionary processes, luck certainly plays a much greater role in macroevolution (Raup 1994).
A species become extinct when all its members die. Higher taxa become extinct when all the species that they encompass become extinct. The basic model of extinction assumes that the individual species become extinct independently of one another. If, for example, ten taxa at the highest level are considered, of which each can be divided into ten taxa at the central level and each of these taxa contains ten species then, for example, if, during a certain period in which speciation does not occur, 90% of species die at random, simultaneously 35% of taxa at the central level become extinct and probably only 0.003% of taxa at the highest level (i.e. in this case, probably none). More realistic models, which consider unequal numbers of species and taxa in various taxa at the same level, yield somewhat different numbers, but the results are basically similar.
For example, the number of species that became extinct during the greatest known mass extinction 248 million years ago at the end of the Permian has been estimated by the method of reverse rarefication on the basis of the number of extinct genera and families on the actual distribution of species in the individual taxa as 96% (the newest estimates yield somewhat lower values). The model describing the extinction of higher taxa on the basis of the independent random extinction of their individual species is called the foot soldier in the field model. Even if 90% of foot soldiers were shot and killed, most probably “only” 35% of ten-membered squadrons would be killed, only about 4% of platoons consisting of 3 squadrons and only about 0.008% of troops consisting of 3 platoons, etc. It is apparent that such a model would not apply to a real battle. Either the individual troops and their platoons wouldn’t all attack at once and be mixed up, but in a certain order, or, for example, some of the members of each formation would be left behind somewhere and would most probably survive. In the former case, more of them would be wiped out while, in the second case, fewer higher formations (higher taxa) would disappear than would correspond to the number of dead foot soldiers.
Study of paleontological material demonstrates that, even for the extinction of higher taxa, the foot soldier in the field model does not adequately describe the real situation. For example, it is improbable that all the taxa of the numerous and extremely diversified trilobite taxon could have died in the Paleozoic or that both orders of dinosaurs (Saurischia and Ornithischia) would have become extinct at the end of the Mesozoic. It is apparent that, at least in some periods of the history of life, membership in a taxon affected which species would become extinct and which would survive. However, the foot soldier in the field model is an extremely useful instrument from the methodological point of view as it represents the zero hypothesis against which other more complicated and more realistic models can be tested.