R.A. Fisher 1930  On speciation

The Genetical Theory of Natural Selection.
Clarendon Press, Oxford.
Chapter VI.

pp. 123-131

The nature of species
From genetic studies in the higher organisms it may be inferred, that whereas genetic diversity may exist, perhaps in hundreds of different loci, yet in the great majority of loci the normal condition is one of genetic uniformity. Unless this were so the concept of the wild type gene would be an indefinite one. Cases are indeed known, as in the agouti locus in mice, in which more than one kind of wild gene have been found, these being both dominant to their other non-lethal allelomorphs; but numerous as are the loci in which such genetic diversity must exist, we have some reason to suppose that they form a very small minority of all the loci, and that the great majority exhibit, within the species, substantially that uniformity, which has been shown to be necessary, if full advantage is to be taken of the chances of favourable mutations. In many loci the whole of the existing genes in the species must be the lineal descendants of a single favourable mutation.

The intimate manner in which the whole body of individuals of a single species are bound together by sexual reproduction has been lost sight of by some writers. Apart from the intervention of geographical barriers so recently that the races separated are not yet regarded as specifically distinct, the ancestry of each single individual, if carried back for only a hundred generations, must embrace practically all of the earlier period who have contributed appreciably to the ancestry of the present population. If we carry the survey back fro 200, 1,000, or 10,000 generations, which are relatively short periods in the history of most species, it is evident that the community of ancestry must be even more complete. The genetical identity in the majority of loci, which underlies the genetic variability presented by most species, seems to supply the systematist with the true basis of his concepts of specific identity or diversity. In his Materials for the Study of Variation, W. Bateson frequently hints at an argument, which evidently influenced him profoundly, to the effect that the discontinuity to be observed between different species must have owed its origin to discontinuities occurring in the evolution of each. His argument, so far as it can be traced from a work, which owed its influence to the acuteness less of its reasoning than of its sarcasm, would seem to be correct for purely asexual organisms, for in these it is possible to regard each individual, and not merely each specific type, as the last member of a series, the continuity or discontinuity of which might be judged by the differences which occur between parent and offspring; and so to argue that these provide an explanation of the diversity of distinct strains. In sexual organisms this argument breaks down, for each individual is not the final member of a single series, but of converging lines of descent which ramify comparatively rapidly throughout the entire specific group. The variations which exist within a species are like the differences in colour between different threads which have crossed and recrossed each other a thousand times in the weaving of a single uniform fabric.

The effective identity of the remote ancestry of all existing members of a single sexual species may be seen in another way, which in particular cases should be capable of some quantitative refinement. Of the heritable variance in any character in each generation a portion is due to the hereditary differences in their parents, while the remainder, including nearly all differences between whole brothers and sisters, is due to genetic segregation. These portions are not very unequal; the correlations observed in human statistics show that segregation must account for a little more than two-fifths, and the hereditary differences of the parents for nearly three-fifths of the whole. These hereditary differences are in their turn, if we go back a second generation, due partly to segregation and partly to hereditary differences in the grandparents. As we look farther and farther back, the proportion of the existing variance ascribable to differences of ancestry becomes rapidly smaller and smaller; taking the fraction due to segregation as only  in each generation, the fraction due to differences of ancestry further back is only about one part in 160, while at 30 generations it is less that one in four millions. It is only the geographical and other barriers to sexual intercourse between different races, factors admittedly similar to those which condition the development of incipient species as geographical races, which prevent the whole of mankind from having had, apart from the last thousand years, a practically identical ancestry. The ancestry of members of the same nation can differ little beyond the last 500 years; at 2,000 years the only differences that would seem to remain would be those between distinct ethnographic races; these, or at least some of the elements of these, may indeed be extremely ancient; but this could only be the case if for long ages the diffusion of blood between the separated groups was almost non-existent.

Fission of species

The close genetic ties which bind species together into single bodies bring into relief the problem of their fission - a problem which involves complexities akin to those that arise in the discussion of the fission of the heavenly bodies, for the attempt to trace the course of events through intermediate stages of instability, seems to require in both cases a more detailed knowledge than does the study of stable states. In many cases without doubt the establishment of complete or almost complete geographical isolation has at once settled the line of fission; the two separated moieties thereafter evolving as separate species, in almost complete independence, in somewhat different habitats, until such time as the morphological differences between them entitle them to 'specific rank'. It would, however, be contrary to the weightiest opinions to postulate that specific differentiation had always been brought about by geographic isolation almost complete in degree. In many cases it may safely be asserted that no geographic isolation at all can be postulated, although this view should not be taken as asserting that the habitat of any species is so uniformly favourable, both to the maintenance of population, and to migration, that no 'lines of weakness' exist, which, if fission is in any case imminent, will determine the most probable geographic lines of division. It is, of course, characteristic of unstable states that minimal causes can at such times produce disproportionate effects; in discussing the possibility of the fission of species without geographic isolation, it will therefore be sufficient if we can give a clear idea of the nature of the causes which condition genetic instability.

Any environmental heterogeneity which requires special adaptations, which are either irreconcileable or difficult to reconcile, will exert upon the cohesive power of the species a certain stress. This stress will be least when closely related individuals are exposed to the environmental differences, and vanishes absolutely if every individual has an equal chance of encountering either of two contrasted environmental situations, or each of a graded series of such situations. It is greatest when associated with circumstances unfavourable to sexual union, of which the most conspicuous is geographic distance, though others, such as earliness or lateness in seasonal reproduction, may in many cases be important. I do not know any such circumstance, which, in the genetical situation produced, differs essentially from geographical distance, in terms of which, therefore, it is convenient to develop the theory.

We may consider the case of a species subjected to different conditions of survival and reproduction at opposite ends of its geographical range. Certain of the genes which exist as alternatives will be favoured at one extreme, and will tend there to increase, while at the other extreme, they will be disadvantageous and tend to diminish in frequency, the intermediate region being divisible into a series of zones in which the advantage increases, from a negative value at one extreme, through zero in a region in which the selective advantage is exactly balanced, to a certain positive advantage at the other extreme. A condition of genetic equilibrium is therefore only established if the increase in frequency in the favourable region and the decrease in frequency in the unfavourable region, not only balance each other quantitatively, but are each equal to the rate at which genes diffuse by migration and sexual union, from the one region to the other. This rate must itself be determined, apart from migratory or sedentary habits of the species, by the density of population along it, and finally by the gradient in the gene frequency ratio between the gene and its allelomorph as we pass across it. So long as a sufficient gradient can be maintained, accompanied by an active diffusion of germinal material, so long as the local varieties, although, possibly, distinct differences between them may be detected, will have no tendency to increase these differences in respect of the frequency of the genes in which they differ, and will be connected by all grades of intermediate types in the population.

The longer such an equilibrium is maintained the more numerous will the genetic differences between the types inhabiting extreme regions tend to become, for the situation allows of the extinction of neither the gene favoured locally nor its allelomorph favoured elsewhere, and all new mutations appearing in the intermediate zone which are advantageous at one extreme but disadvantageous at the other will have a chance of being added to the factors in which they differ. In addition to those genes which are selected differentially by the contrasted environments, we must moreover add those, the selective advantage or disadvantage of which is conditioned by the genotype in which they occur, and which will therefore possess differential survival value, owing not directly to the contrast in different environments, but indirectly to the genotypic contrast which those environments induce. The process so far contains no novel features, it allows of the differentiation of local races under natural selection, and shows that this differentiation must, if the conditions of diffusion are constant, be progressive. It involves no tendency to break the stream of diffusion, or consequently to diminish in degree the unity of ancestry which the species possesses. It is analogous to the stretching of a material body under stress, not to its rupture.

There are, however, some groups of heritable variations which will influence diffusion. In the case we are considering in which the cause of isolation is geographical distance, the instincts governing the movements of migration, or the means adopted for dispersal or fixation, will influence the frequency with which the descendants of an organism, originating in one region, find themselves surrounded by the environment prevailing in another. The constant elimination in each extreme region of the genes which diffuse to it from the other, which are most apt so to diffuse. If it is admitted that an aquatic organism adapted to a low level of salinity will acquire, under Natural Selection, instincts of migration, or means of dispersal, which minimize its chances of being carried out to sea, it will be seen that selection of the same nature must act gradually and progressively to minimize the diffusion of the germ plasm between regions requiring different specialized aptitudes. The effect of such a progressive diminution in the tendency to diffusion will be progressively to steepen the gradient of gene frequency at the places where it is highest, until a line of distinction is produced, across which there is a relatively sharp contrast in the genetic composition of the species. Diffusion across this line is now more than ever disadvantageous, and its progressive diminution, while leaving possibly for long a zone of individuals of intermediate type, will allow the two main bodies of the species to evolve in almost complete independence.

In cases in which the cause of genetic isolation is not merely geographical distance, but a diversity among different members of the species in their habitats or life history, in connexion with which different genetic modifications are advantageous; the isolation will of course not be increased by the differential modification of the instincts of migration, or the means of dispersal; but by whatever type of hereditary modification will minimize the tendency for germinal elements, appropriate to one form of life, to be diffused among individuals living the other form, and among them consequently eliminated.

The power of the means of dispersal alone, without the necessity for selective discrimination in either region, is excellently illustrated by the theory, due to Ray Lankester, which satisfactorily accounts for the diminution or loss of functional eyes by the inhabitants of dark caverns. Ray Lankester pointed out that the possession of the visual apparatus is not merely useless to such animals but, by favouring their migration towards sources of light, will constantly eliminate them from the body of cave inhabitants, equally effectively whether they survive or perish in their new environment. Those which remain therefore to breed in the cavern are liable to selection in each generation for their insensibility to visual stimuli. It should be noted that with such very restricted habitats migrational selection of this sort might attain to very high intensity and in consequence produce correspondingly rapid evolutionary effects.

Sexual preference

A means of genetic isolation which is of general importance in that it is applicable equally to geographical and to other cases is one, which for want of a better term, we may consider under the heading of reproductive or sexual preference.

The success of an organism in leaving a numerous posterity is not measured only by the number of its surviving offspring, but also by the quality or probable success of these offspring. It is therefore a matter of importance which particular individual of those available is to be their other parent. With the higher animals means of discrimination exist in the inspection of the possible mate, for in large groups the sense organs are certainly sufficiently well-developed to discriminate individual differences. It is possible therefore that the emotional reactions aroused by different individuals of the opposite sex will, as in man, be not all alike, and at the least that individuals of either sex will be less easily induced to pair with some partners than with others. With plants an analogous means of discrimination seems to exist in the differential growth rate of different kinds of pollen in penetrating the same style.

An excellent summary of recently established facts in this field has been given by D.F. Jones (Selective Fertilization, University of Chicago, 1928). Cases are known in maize in which discrimination is exercised against pollen bearing certain deleterious mutant factors, and in one case in Oenothera against ovules bearing a certain lethal factor. In these reactions both the genotype of the mother plant and that of the pollen are exposed to selection, and it is this that serves to explain the remarkable fact established by Jones' own observations with maize, that pollen applied in mixtures is on the whole less effective the greater the genetic diversity between the seed parents and the pollen parent. Such a generalized tendency towards homogamy, which is perhaps especially manifest in maize owing to the enormous number of recessive defects, which by continued cross pollination have accumulated in that plant, would, however, be far less effective in promoting the fission of species than would the selection of discriminative tendencies specially directed toward that end, such as must occur, as will be explained more fully below, in a group constantly invaded by the diffusion of unfavourable genes.

In general the conditions upon which discrimination, when possible, can usefully be exercised seem to be (i) that the acceptance of one mate precludes the effective acceptance of alternative mates, and (ii) that the rejection of an offer will be followed by other offers, either certainly, or with such high probability, that the risk of their non-occurrence shall be smaller than the probable advantage to be gained by the choice of a mate. The first condition is satisfied by the females of most species, and in a considerable number of cases by the males also. In other cases, while it would be a serious error for the male to pursue an already fertilized female, it would seem that any opportunity of effective mating could be taken with advantage. The second condition is most evidently satisfied when members of the selected sex are in a considerable majority at the time of mating.

The grossest blunder in sexual preference, which we can conceive of an animal making, would be to mate with a species different from its own and with which the hybrids are either infertile or, through the mixture of instincts and other attributes appropriate to different courses of life, at so serious a disadvantage as to leave no descendants. In the higher animals both sexes seem to be congenitally adapted to avoid this blunder and from the comparative rarity of natural hybridization among plants, save in certain genera where specific distinctness may have broken down through maladaptation in this very respect, we may infer the normal prevalence of mechanisms effective in minimizing the probability of impregnation by foreign pollen. It is therefore to be inferred that in the higher animals the nervous system is so constructed, that the responses normal to an association with a mate of its own species are, in fact, usually inhibited by the differences which it observes in the appearance or behaviour of a member of another species. Exactly what differences in the sensory stimuli determine this difference it is of course impossible to say, but it is no conjecture that a discriminative mechanism exists, variations in which will be capable of giving rise to a similar discrimination within its own species, should such discrimination become at any time advantageous.

A typical situation in which such discrimination will possess a definite advantage to members of both sexes must arise whenever a species occupying a continuous range is in the process of fission into two daughter species, differently adapted to different parts of that range; for in either of the extreme parts certain relatively disadvantageous characters will constantly appear in a certain fixed proportion of the individuals of each generation, by reason of the diffusion of the genes responsible from the other parts of the range. The individuals so characterized will be definitely less well adapted to the situation in which they find themselves than their competitors; and in so far as they are recognizably so, owing, for example, to differences in tint, their presence will give rise to a selective process favouring a sexual preference of the group in which they live. Individuals in each region most readily attracted to or excited by mates of the type were favoured, in contrast to mates of the opposite type, will, in fact, be the better represented in future generations, and both the discrimination and the preference will thereby be enhanced. It appears certainly possible that an evolution of sexual preference due to this cause would establish an effective isolation between two differentiated parts of a species, even when geographical and other factors were least favourable to such separation.

Sexual selection

The theory put forward by Darwin ...