The maintenance of genetic polymorphisms in natural populations is a central problem in genetic evolutionary theory. The paradox is that natural populations exhibit a large degree of variability and are well adapted even in stable environments. The approach to this problem has been theoretical and there are few examples in natural populations that have been thoroughly studied. However, theoretical development of ideas has established a sound basis for future studies of variability in natural populations.

The concept of optimum population fitness in a defined environment has generated many ideas and contradictions in theoretical population biology. Dobzhansky (1955) has considered the origin of the adaptive norm of a population according to two hypotheses. The classical hypothesis maintains that evolutionary changes consist of gradual substitutions and eventual fixation of favourable alleles and chromosome structures. According to this hypothesis, genetic diversity will be neutral, transient or morbid. On the other hand, the balance hypothesis maintains that the adaptive norm is an array of genotypes, heterozygous at numerous gene loci, gene complexes and chromosome structures. Homozygous individuals according to this hypothesis are inferior in fitness to the adaptive norm. The dichotomy presented by the two concepts is more likely a continuum in nature with populations between the two extremes. Dobzhansky (1955) considered the adaptive norm of a population as an array of related genotypes consonant with the environment and which is neither a single genotype nor a single phenotype. In nature there are limitations that prevent populations from existing at either extreme.

The problem of fitness as it pertains to evolutionary changes must be considered both with respect to the population and the individual. Fitness of individuals may be divided into absolute fitness related to the environment in which it lives and the relative fitness of an individual related to the fitness of other individuals in the population. Fitness in the absolute sense is not possible to determine since the environment lies outside the scope of any such definition (Wallace, 1968). The relative fitness of an individual is considered as its contribution to the gene pool of the subsequent generation relative to the contribution of other individuals in the population (Sved et al. 1967, Wallace, 1968). The fitness of each genotype in a population may be considered as the mean relative fitness of each of the individuals of that genotype. The average fitness of a population may be calculated as the sum of the products of fitnesses times the frequencies of individual fitnesses (Wallace, 1968). According to this concept the average fitness of the population is less than that of some of the individuals in the population. The problem with this argument is that the absolute fitness of the population is calculated from the relative fitness of the individuals. It is not entirely valid to state that the fitness of the population is less than that of some if the individuals in the population since relative fitness and absolute fitness are not comparable.

The concept of "genetic load" was first formalized by Crow (1958) who defined it as "the proportion by which the population fitness is decreased in comparison with the optimum genotype." The weakness in this definition is in his use of the term "optimum genotype" which probably does not exist in natural populations. A more suitable definition of genetic load is that it is proportional in natural populations to the extent by which the gene frequency of the breeding individuals deviates from the expected frequencies predicted by Hardy-Weinberg equilibrium due to the forces of selection only.

Heterosis occurs at polymorphic loci when the relative fitness of heterozygotes is greater than that of either of the homozygotes. The conventional model for one heterotic locus with two alleles is as follows (Crow, 1958; Kojima, 1971): where s and u are both positive and less than or equal to 1.

 

Failure to distinguish between heterosis and the phenomena of dominance, over-dominance and luxuriance is a common temptation. However, heterosis refers only to genotypic fitness, whereas the other terms are related to phenotypic expression and are not directly concerned with fitness.

The idea that heterosis is a property of a system of polygenes which are co-adapted by natural selection was submitted by Dobzhansky (1952) with evidence from the ST-CH inversion of Drosophila psuedoobscura in which heterosis occurs only if the constituent chromosomes are derived from the same population or those from nearby localities. Dobzhansky showed further that heterozygotes derived from remote localities did not show increased fitness, suggesting that other loci are involved in the heterosis besides the ST-CH inversion.

The occurrence of heterotic loci has been well documented in a variety of populations. Dobzhansky and Pavlovsky (1955) have described a balanced lethal system in a restricted population of Drosophila. This is the situation where values of both s and u are equal to 1 and results in 50% of the zygotes having no fitness at all (i.e. they make no contribution to the gene pool of the next generation).

The genetic load which results from heterotic loci has been termed the segregation load (Crow, 1958) since random recombination of two alleles in heterozygous individuals will result in one half of the offspring being of the less fit homozygous genotype. Heterosis and panmixia stand in opposition to each other (Dobzhansky, 1955). Only if all heterotic loci are situated on the same chromosome would it be possible to eliminate the homozygotes by non-random mating.

There are a number of ways in which genetic polymorphisms may exist in a population without an overall reduction of population fitness. In natural populations there is a premium on genetic variability which increases the adaptability of populations to spatial or temporal changes in the environment. A population which is completely homozygous may be more fit in a restricted niche but is much more susceptible than a heterozygote population to environmental changes. Through natural selection the fitness of some individuals may be sacrificed for the fitness of the population as a whole (Dobzhansky, 1955).

According to the classical hypothesis of the adaptive norm discussed earlier one would expect that selection should favour the replacement of heterotic alleles by alleles which yield high fitness in the homozygous condition. However, this assumes that the environment is static, which of course it seldom is. Heterozygous loci therefore, have the advantage of greater adaptability which no homozygous locus can achieve.

The idea that the number of genetic polymorphisms simultaneously maintained in a population is restricted by the genetic load imposed, has been a controversial one. This view, which seems sound, is in direct contradiction to the observation of many polymorphisms in nature (Mayr, 1963). The observation that balanced lethal systems of two alleles effectively reduces the fecundity of the zygotes by one half, suggests that zygotic survival could be equally reduced by a series of heterotic loci in which the values of s and u are less than one. Populations with a high reproductive rate such as Drosophila could conceivably maintain more polymorphisms of this type than could populations with a low reproductive rate such as man (Wallace, 1968).

There have been a number of hypotheses that attempt to explain the presence of polymorphisms in the face of genetic load. One obvious way to reduce genetic load in a heterotic system is by developing a series of iso-alleles at each heterotic locus so that the frequency of homozygotes is effectively reduced. Although this appears to have occurred in many natural populations it does not seem to be feasible in many other populations (Mayr, 1963; Kojima, 1971).

I believe that few loci exist in nature that are completely heterotic at all times and in all places. Such loci would place a heavy genetic load on a population. I suspect that heterosis is a local or temporary phenomenon in most cases and that in the long term, polymorphisms exhibiting heterosis are maintained by frequency dependent selection. This would be true most often when the values of s and u are small. When s and u are large or in a balanced lethal system frequency dependent selection is less likely to occur. Kojima (1971) has discussed in more detail the significance of frequency dependent selection and maintains that this is a more important form of selection than heterosis with respect to maintenance of polymorphisms. His major evidence for this is that the genetic load produced by frequency dependent polymorphisms is considerably less for reasonable effective population size (N =200-500).

It is difficult to estimate the evolutionary advantage gained by heterosis. One benefit might be that of regulating population fluctuations under the influence of a changing environment. Lerner (1954) has defined genetic homeostasis as the tendency of a population to equilibrate its genetic composition and to resist sudden changes. He further maintains that Mendelian populations possess self-regulating properties to establish a connection between genetic and developmental homeostasis, and that heterozygosity provides a basis for both phenomena.

I believe that a population with a substantial segregation load and a high reproductive rate will be maintained at low densities by selection against homozygotes. If the selection is temporarily removed as a result of environmental change, the population will increase, but the homozygotes will increase more rapidly than the heterozygotes due to random segregation. Therefore at higher density the fitness of one segment of the population is less than that of others. When selection is resumed it will effect the less fit homozygous segment of the population more than the heterozygous segment. As a result of this the density of the more fit individuals does not fluctuate greatly. Wallace (1968) has attempted to show that the existence of density dependent factors enables a population to meet increased demands of selection by a contraction of population size as well as by an increase in numbers of progeny. He has also published an interesting three dimensional diagram (figure 1) which illustrates the relationships between fitness of a population with respect to two alleles which are both frequency and density dependent. Wallace has stated that simple parabolic relationships between fitness and gene frequency are inadequate models to describe the density dependence of many polymorphisms.

Figure 1. Three dimensional illustration of gene frequency-fitness and population size-fitness dependence relationships (after Wallace, 1968)

Figure 2. Zone of fitness > 1 shaded. This condition would be unlikely in nature.

 

Figure 3. Homozygote "aa" shaded.

 

Figure 4. Heterozygote "Aa" shaded.

 

Figure 5. Homozygote "AA" shaded.

LITERATURE CITED

Crow, J.F. 1958. Some possibilities for measuring selection intensities in man. Hum. Biol. 30(1):1-13.
Dobzhansky, T. 1952. Nature and origin of heterosis. In Heterosis. J.W. Gowen (ed). Iowa State College Press.
Dobzhansky, T. 1955. A review of some fundamental concepts and problems of genetics. Cold Spring Harbour Symposium. 20:1-15.
Dobzhansky, T. and O. Pavlovsky. 1955. An extreme case of heterosis in a central American population of Drosophila tropicalis. Nat. Acad. Sci. Proc. 41:289-295.
Kojima, K. 1971. The distribution and comparison of "genetic loads" under heterotic selection and simple frequency dependent selection in finite populations. Theor. Pop. Biol. 2(2):159-173.
Lerner, M.I. 1954. Genetic Homeostasis. Oliver and Boyd.
Mayr, E. 1963. Animal Species and Evolution. Belknap Press of Harvard Univ. Press, Cambridge, Massachusetts.
Sved, J.A., T.E. Reed and W.F. Bodmer. 1967. The number of balanced polymorphisms that can be maintained in a natural population. Genetics 55:469-481.
Wallace, B. 1968. Polymorphisms, population size and genetic load. In Population Biology and Evolution. R.C. Lewontin (ed).