Coevolution of a predator and its prey, parasite and its host or two species competing for utilization of a certain resource or having a common predator, and even the coevolution of two mutualistic species very frequently has the character of a sort of “arms race”. A certain new evolutionary feature emerges in one of the species, increasing the fitness of its bearers at the expense of members of another species, and thus it gradually spreads in the population of this species (Fig. XVIII.1). Such a new evolutionary feature could, for example, consist in stronger jaw muscles, enabling a predator fish to crack the shell of a certain kind of snail and include it in its diet. This would create a selection pressure on the snail to create a stronger shell, or to synthesize a poisonous or terrible-tasting substance to defend itself against attacks by the fish. The emergence of such an adaptation in the snails would produce a selection pressure on the predatory fish which, sooner or later, would lead to the creation of a suitable counter-strategy, enabling it to somehow overcome the effect of the new evolutionary trait in the snail – for example, through the formation of even stronger jaw muscles or of sharper horny jaws, or the emergence of some suitable pattern of behavior that would enable it to break even stronger shells. A new round in the arms race can be started by either the “attacker” or the “defender” and the participants in this race need not be only a pair of species, but can even consist in large groups of mutually interacting species.
A great many paleontologists assume that some rapid changes in fauna that occurred in certain short periods in the history of the Earth, with the simultaneous participation of an enormous number of species, emerged as a product of just such arms races. A certain new evolutionary feature emerged in one species, creating a selection pressure on a great many species in the environment; they had to somehow react adaptively to this pressure, other species had to react to their reactions, so that the wave of anagenetic changes could finally affect a great portion of the species in the particular ecosystem in a very short period of time (Morris 1995).
Arms races can also occur at an intraspecific level and this can happen on a relatively short time scale in microevolutionary processes. The members of the male or female sex can increase their fitness at the expense of the members of the other sex. Under normal conditions, the speed of evolution of males and females is comparable, so that it cannot be observed that the members of one sex would gain an apparent advantage over the members of the other sex. However, by suitably arranging an experiment, we can ensure, for example, that the females will not be able to respond to the evolutionary drives of the males. The experiment is performed in that the males are allowed to interact ecologically, ethologically and sexually with the females of a certain strain for a number of successive generations, where only the males of the obtained progeny are allowed to survive. The females are replaced in the experimental population in each generation from the stock population (and are thus completely “naive” from the standpoint of interaction with the particular selected line of males). The results of such an experiment performed on drosophila indicate that the males gained a considerable advantage over the females after 41 generations (Rice 1996). They achieved copulation with the females of the particular strain much faster; following copulation, the time before the females began to copulate with some other male was prolonged and, in case of multiple copulation with their own and foreign males, only a small percentage of the eggs were fertilized by the foreign males. The females lost out in the evolutionary battle, not only in the limited ability to chose, from their point of view, the best father for their progeny, but also in shortening of their lifetimes following copulation with the test males (but not with control males) (Fig. XVIII.2).
According to some authors, the coevolutionary battle between the sexes in sexually reproducing species is the main reason for the genetic divergence of allopatric populations and is thus indirectly an important motor for allopatric speciation (Martin & Hosken 2003).