Frozen Evolution. Or, that’s not the way it is, Mr. Darwin. A Farewell to Selfish Gene.

In 1964, W.D. Hamilton published the results of his doctoral thesis concerned with some consequences of the existence of sexual reproduction for the progress of microevolution (Hamilton 1964a; Hamilton 1964b).His two articles, together with the ideas of G.C. Williams (Williams 1966), established the basis for a fundamentally new approach to biological evolution in sexually reproducing organisms.This model of biological evolution was popularized in the 1970’s and 1980’s by R. Dawkins as the selfish gene theory.In his best-known book “The Selfish Gene” (Dawkins 1976), he demonstrated, in contrast to Hamilton without using any mathematical models, that the subject of natural selection and thus the actual object of biological evolution cannot be individuals amongst sexually reproducing organisms, whose genome and thus biological properties are not inherited from one generation to the next, and certainly not families, populations or species, but only the various alleles of the individual genes, which are practically always passed down from one generation to the next in unaltered form.Thus, according to Dawkins, biological evolution must be understood as a race between the various alleles of a certain locus for the greatest frequency in the gene pool of the population.The individual alleles of the various genes can variously cooperate together, can conclude various coalitions but, in actual fact, all biological processes are based on a battle amongst the individual selfish genes or, to be more exact, selfish alleles, for the most effective and most frequent replication.

            As an excellent popularizer, Dawkins called his book “The Selfish Gene” and not “The Selfish Allele” and almost always speaks of selfish genes and not alleles, as most lay people have an idea of what a gene is (although mostly erroneously, see II.3.1), while the definition of an allele is not part of the general consciousness.In this conception, organisms are understood to be sort of vehicles, instruments that the genes have created so that they can replicate as fast as possible under the conditions in our biosphere. When an evolutionary biologist studies a certain biological phenomenon, a certain property of living organisms, he should not ask which advantage that property brings its bearer (regardless of whether at the level of the individual, population or species), but how it is advantageous for the allele, the DNA section that codes the given property, how it helps to spread it in the gene pool at the expense of the other alleles of the same gene.

            The selfish gene theory turned out to be an effective instrument for understanding and describing various evolutionary processes.It permits integration of our view of natural selection at all levels.It easily manages to explain the mechanism of the formation of altruistic behaviour and evolutionary processes at the molecular and chromosomal levels.Simultaneously, some biologists do not consider it to be an independent model of evolution, but only an alternative way of describing or viewing nature.Dawkins himself adopted this approach in the first edition of the “Selfish Gene” book.However, an increasing number of evolutionary biologists are coming to the opinion that the selfish gene theory or, to be more exact, the model of intralocus interallele selection, is substantially different from the Neodarwinist model of evolution of adaptive traits.While, according to the classical theory, only an allele that, compared to the other alleles, increases the relative biological fitness of its bearer can spread, it follows from the selfish gene theory that an allele that reduces the relative biological fitness of its bearer can spread in the population, of course, only under the assumption that it will simultaneously increase the probability of its own transfer to the gene pool of the next generation.

Fig. IV.10. The bluebeard model. A normal male has an average of 10 offspring, of which an average of two, a male and female, do not survive to adulthood. A male that has the bluebeard allele on his Y-chromosome kills all the females in his batch of offspring, so that all the males survive to adulthood. From the viewpoint of the classical model of evolution, a bluebeard has lower fitness than a normal male as he has only 5 offspring rather than the usual 8. However, the gene-centred approach has shown that, while a normal allele is passed on to the next generation in only 4 copies (as females do not have a Y-chromosome), the bluebeard allele is passed on in 5 copies. Thus, the content of bluebeard alleles in the population will continuously increase, although the individual biological fitness of their bearers will be reduced.

        An illustrative, although originally purely theoretical example of the spreading of such an allele was described by Maynard Smith (Maynard Smith & Price 1973).This model can be described as the bluebeard modeland its principle is illustrated in Fig. IV.10.Imagine a gene on the Y-chromosome of a male, whose allele M causes that the male will kill all its daughters and feed its sons with their meat.As a consequence, a male with allele M will have almost half-lower biological fitness than a normal male with allele m.It will have only half as many progeny; however, the sons will not have to compete with daughters and will thus be stronger.It would follow from the Neodarwinist model that this bluebeard allele would rapidly disappear from the population.However, it follows from the selfish gene theory that the bluebeard allele will be highly successful and will spread in the population.The success of the allele is caused by the fact that it is transferred to the next generation on the Y-chromosome only to males and thus the fact that the bluebeard does not leave behind and daughters is insignificant.To the contrary, the fact that the sons will have no competition from daughters and will be better fed and thus stronger will contribute favourably to spreading of this gene in the population.

            So far, no case of bluebeard alleles that would spread in the population through the above-described ethological mechanism is known in real nature, but some known ethological mechanisms are quite close to it (Foster, Wenseleers, & Ratnieks 2001).In addition, however, we know a number of alleles that achieve the same effect through other molecular or physiological mechanisms.Probably the best-studied systems consist in the SD-alleles of drosophila and the t-alleles of house mice (Carvalho & Vaz 1999; Ardlie 1998; Vanboven et al. 1996).In both cases, the relevant “bluebeard allele” manages during meiosis to reprogram a gene on the homologous chromosome so that, during spermatogenesis, in which normally the genes of the developing germ cell do not intervene, it actively destroys the developing germinal cell.This means that a heterozygote bearing the bluebeard allele produces substantially fewer sperm than a normal individual (and thus has lower biological fitness), but all its sperm (or most of them) bear the “bluebeard” allele (Fig. IV.11).

Fig. IV.11. Spreading of the t-haplotype in mice. A t-haplotype male produces as few as half as many functional sperm as a male with wild haplotype. However, all his sperm contain the t-haplotype and thus all his progeny will also have this t-haplotype. If a female will always pair with only a single male (a), the reduced amount of ejaculate will not put the male at much of a disadvantage as even half as many sperm is usually sufficient for fertilization of all the eggs in the reproductive tract of the female. However, if the female is more promiscuous and mates with a greater number of males, e.g. two males (b), the bearers of the t-haplotype will be at a disadvantage in competition amongst sperm and will fertilize a smaller number of eggs. From the standpoint of the haplotype, half-as-large male fertility will have no effect as it is transferred to the next generation in the same number of copies as the zero allele. The other genes of a male with t-haplotype, e.g. the gene for black hair colour, which are not genetically bound to the t-haplotype, will, however, be transferred to the next generation in a smaller number of copies. This provides a selection pressure for evolution of modifier genes that are capable of neutralizing the action of t-haplotype alleles.

           Twenty years ago, the gene-centred view of nature seemed rather exotic; however, over time, it has inconspicuously but very successfully penetrated into various areas of biology.The evolution of eusociality in hymenopterous insects will be used here to illustrate the use of the selfish gene theory.