If evolution can be viewed as a sort of “arms race,” one often assumes that if an animal manipulates another, the victim will, via random mutation and natural selection, develop counterstrategies to escalate the arms race rather than capitulate. However, Dawkins shows several examples of how this is not necessarily so.
For example, the foster parents of the cuckoo continue to feed their changeling “child” despite the obvious absurdity of a tiny Garden Warbler straining to feed a cuckoo several times its size. Why doesn’t the Warbler recognize this incongruity, when apparently some host parents of cuckoos can indeed recognize flaws in the cuckoo’s egg mimicry? Why does the host parent recognize the parasitic egg but not the more obvious parasitic changeling fledgling bird?
Dawkins’s answer is that natural selection does not act uniformly at all times in any animal’s life. Selective pressure may be stronger at some points in the life cycle than others, or natural selection may have no effect on evolution even if a beneficial mutation were to arise. As an example of the first case, Dawkins points out that recognition of a cuckoo egg in one’s nest gives the host parent the chance to gain an entire breeding cycle, whereas recognition of the incongruous fledgling would buy at most a few days, and that probably too late for the host parent to breed again. Moreover, the actions of the cuckoo, its exaggeratedly gaping mouth, its size, could indeed act as a “drug” on the foster parent, no less than the song of a male nightingale acts as a drug on the female reproductive cycle (and, incidentally, upon the poet’s imagination).
However, a fascinating example of how an animal’s victimhood can be perpetuated is exemplified by slave-making ants. Some species of ants spend a great deal of their time raiding the nests of others and carrying off the larvae and pupae, which subsequently hatch in the new nest and begin to labor for their “masters.” This is a disturbing and puzzling development. Why don’t the enslaved worker ant colonies develop a resistance to the strange environment, filled with others not their genetic sisters—for example, by evolving a genetic disposition to cease work (to go on “strike”) when in a strange queen’s lair?
Remember that worker ants do not reproduce. Therefore, any beneficial mutation that arose in the enslaved ants would not be passed on to the rest of their home nest. At any rate, the raids do not happen often enough to destroy the victims utterly, who are under little selective pressure to evolve complex adaptive countermeasures against slave-making behavior on the part of others that, while aggressive, does not threaten the existence of the nest. An uneven battle ensues, in which the slave-making ants can be said to win the war.
Dawkins indicates that this situation is not unlike the phenomenon of a certain species of hybrid frog, which has one set of chromosomes that is jettisoned in meiosis and one set that is passed on to its offspring. The set of chromosomes (dead-end replicators) that is jettisoned in the hybrid frog is perpetuated in the pure bred species that carries two sets of these chromosomes (which become germ-line, not dead-end, replicators in this species). Thus, any beneficial mutation in the dead-end replicator line will be passed on in the pure bred species that contains two sets of these chromosomes (because in this species these sets of chromosomes are not dead-end replicators), but will not be passed on in the hybrid species. The situation of the enslaved ants is like that of the hybrid species of frog: their genes have phenotypic effects and can even be selected, but they will not be transmitted in the hybrid species, and thus are irrelevant to that particular species' evolution.