Heterosis and Inbreeding Depression

Inbreeding depression is usually defined as the lowered fitness or vigour of inbred individuals compared with their non-inbred counterparts, observed in many (but by no means all) organisms. Its converse is heterosis, the 'hybrid vigour' manifested in increased size, growth rate or other parameters resulting from the increase in heterozygosity in F1 generation crosses between inbred lines.

Problems arise in defining fitness and vigour. So long as we are discussing populations of wild plants, we have the concept of Darwinian fitness: if individuals of one genotype survive to breed more than others, then that genotype confers greater fitness. Fitness is an observed quantity that integrates the effect of all characters that influence the ability of the organism to live and reproduce. As natural selection can only act to increase the frequency of an allele in proportion to the extent to which that allele increases fitness, it was predicted (Falconer, 1989) that the amount of heritable variance for a trait would be inversely proportional to its effect on the organism's fitness. In other words, there would be more inheritable variance in relatively 'neutral' characters than in ones that increased the organism's viability or fecundity. Studies on three very different animal species tended to confirm this prediction regarding individual characters (Kruuk et al., 2000), but the measurement of overall fitness in populations remains problematical.

Vigour is another vague concept: but to most growers it is synonymous with rate of growth or biomass accumulation. In an annual, this is well correlated with production of grain, flowers or fruit since these plants maximise their investment in seed production. But a perennial may increase its growth rate only to reinvest these resources in further vegetative reproduction; examples are bulbil watsonia and crow garlic that have become weeds by the strategy of vegetative reproduction with little or no seeding.

Vigour has sometimes been used as a measure (or even a near-synonym) of fitness in discussions of heterosis. But a plant can be too vigorous for its own good: it can become structurally unstable and vulnerable to destruction by mechanical stress or dependent on a higher and more reliable supply of water and nutrients than the environment can guarantee. And an individual plant whose accelerated growth rate causes it to flower long before the rest of its population may - if it can self-pollinate - have increased its fitness. But its fitness will be reduced to zero if it's a self-incompatible annual.

The relation between fitness and vigour in a wild population is not a straight line, but a curve with a maximum:

In the language of games theory, natural selection might be said to favour strategies that lead to a minimax outcome - the optimum attainable in the real world rather than the theoretical maximum.

Darwinian fitness is not a single character, but the outcome of every trait in an individual that influences the contribution it makes to the next generation. In populations of wild plants, heterosis for a range of quantities presumed to have a strong influence on fitness is more meaningful than heterosis for vigour alone. For example, Buza et al. (2000) measured germination rate, seedling survivorship and vegetative growth rate in their attempt to quantify inbreeding depression in relict Swainsona populations.

Balanced polymorphisms in populations are often due to the heterozygotes having a higher fitness than either homozygote (Ford, 1964).

Possible explanations of this heterosis for overall fitness include:

1. The formation of homozygotes in which alleles with more or less recessive deleterious effects become fully expressed, but had been masked by the dominants in heterozygotes. Recessive and mildly deleterious mutations are common, and if located close to a beneficial allele they will increase in frequency as it does, being "sheltered" from selection in the heterozygotes. By the time a beneficial mutant is common enough for its homozygotes to occur in significant numbers in the population, it may be linked with enough harmful recessives to make these homozygotes less fit than the heterozygote (Fisher, 1927; Ford, 1965). In this case, heterosis is said to be caused by dominance complementation. This may also be described loosely as epistasis of the harmful genes over the beneficial ones - but only if that term is used in the broadest sense.

2. Overdominance - the fitness of the heterozygote per se is higher than either homozygote. This could arise because genes normally have pleiotropic effects, contributing to many measurable traits of the plant. If a mutation is selected for a positive effect on fitness or some desired trait, its other effects are likely to be deleterious. Because a mutant allele at first spreads through a population via heterozygotes, selection will tend to modify and optimise its action by making its beneficial effects dominant and its deleterious effects recessive (Gustafsson, 1950; Sheppard, 1953). This hypothesis has been used to explain balanced polymorphisms, for example in the case of Dactylis glomerata (Apirion & Zohary, 1961).

In both explanations, a beneficial dominant effect is associated with a deleterious recessive effect. Whether these correlated effects are due to the same gene, or to closely linked but separate genes may be a difficult point to decide. Both overdominance and dominance complementation may contribute in varying degrees to observed cases of heterosis. But evidence for actual instances of overdominance remains scarce.

Three predictions of the dominance complementation hypothesis suggest how it might be tested:

1. Inbreeding depression due to genes with deleterious recessive effects that are linked to genes with beneficial dominant effects may be removed by selection; but if it is due to overdominance it cannot.

2. A period of inbreeding can 'purge' deleterious recessive alleles from a population since these will have an increased chance of appearing in homozygotes than can be eliminated by selection (Barrett & Charlesworth, 1991).

3. If two strains of a plant are each selected for heterosis in the F1 produced by crossing them with a third strain, heterosis will also increase in the F1 between these selected strains because they are being purged of deleterious alleles. But if heterosis was due to overdominance, the two strains would converge, since each would be selected for alleles that differed from those of the tester strain at each locus, and their hybrid will show inbreeding depression instead of heterosis.

Wild plants that are normally inbreeders, or have a mixed inbreeder/outbreeder strategy, might be expected to have a genetic architecture that minimises inbreeding depression. In one experiment confirming this, the inbred seeds from cleistogamous flowers of Viola species had only slightly lower fitness than the outcrossed seeds from chasmogamous flowers on the same plants (Berg & Redbo-Torstensson, 1999).

Inbreeding depression is caused mainly by deleterious recessive alleles in both Mimulus guttatus (an outbreeder) and the closely related inbreeder M. micranthus (Dudash & Carr, 1998). Natural selection in experimentally inbred lines of M. guttatus purged the few alleles with major deleterious effects, but numerous other alleles, each with a small effect, accounted for almost all the depression (Willis, 1999).

Cultivated plants

But most of the practical interest in heterosis centres on cultivated plants, and when plant breeders say 'fitness' they really mean 'crop yields'. It is more useful to speak of heterosis for a particular statistic of the crop: for example, leaf area or seed size.

Heterosis is also modified by the interactions between genotype and environment in cultivation. Hybrid sorghum can show heterosis for yield; but this effect varies widely between trials conducted at sites differing in seasonal water supply (Chapman et al., 2000), so that it is more meaningful to characterise a particular hybrid line as showing heterosis for yield at a specific locality or under certain environmental conditions.

In practice, yield and other measurable traits desired by breeders are often correlated with traits that increase the fitness of the plant: increased efficiency of metabolism, photosynthesis etc. As a convenient approximation, we can speak loosely of 'beneficial' and 'deleterious' effects.

Heterosis for size, vigour or yield is most evident in outbreeding crops. The classic case is maize, in which heterosis has long been exploited in the production of uniformly high-yielding F1 seed in commercial quantities.

The definitive experiment by Sprague (1983) showed that this instance of heterosis is due wholly or mainly to dominance complementation. Over many years, two maize populations were each selected for increased yield in the hybrids produced by crossing them with another inbred 'tester' strain. F1 hybrids between the two selected strains showed increased yield, as predicted by the dominance complementation hypothesis. They also produced increased yields in hybrids with other testers.

If dominance complementation is the only factor involved in heterosis for yield, the high productivity of hybrid maize may have been matched by open-pollinated cultivars if enough work had been invested in their development. However, Fu & Dooner (2002) have reported the surprising discovery that maize cultivars differ in the set of genes they carry: some loci in one cultivar lack corresponding alleles in another line. In this case, any component of the heterosis effect due to complementation between different loci could never be fixed in an open-pollinated line, since these loci would remain on separate chromosomes. Maize may be a special case among crop plants - both because its genome has been doubled in size by retrotransposon activity that produced multiple paralogs of some genes, and also because it has been more radically changed by the domestication process than have other crops such as rice.

Xiao et al. (1995) demonstrated that overdominance is not a major cause of heterosis for yield in a cross between the two subspecies of rice, because there was no correlation between most traits and overall genome heterozygosity, heterozygotes were never superior to both homozygotes in analysis of quantitative trait loci, and some F8 inbred lines were actually superior to the F1 for all traits evaluated.

Trials support the hypothesis that heterosis in wheat is due to dominance complementation, with linkage and interaction of alleles (Pickett & Galwey, 1997) - to the loss of those who would like to exploit hybrid wheat as they have done hybrid maize. In some other outbreeding crop plants, such as cucumbers (Cramer & Wehner, 1999), crossing inbred lines rarely leads to heterosis.

The hypothesis of dominance complementation is also supported by evidence that polyploids are to some extent buffered against inbreeding depression because they have additional copies of each gene (Bingham et al., 1994; Husband & Schemske, 1997).

A third explanation of heterosis

An alternative theory was proposed by Milborrow (1998). He suggested that growth of a plant may be limited by the genes that regulate certain metabolic pathways down to lower levels than the maximum possible. Heterozygotes may partially escape this regulation because they have two slightly different alleles for these genes, allowing greater flow on these pathways. This is not overdominance; but, like the overdominance hypothesis, it predicts that heterozygotes have an inherent advantage in vigour that cannot be duplicated by any amount of selection in open-pollinated homozygous lines.

The most interesting part of Milborrow's theory is the implication as to why there are so many "weak" alleles in a population to start with. These are not sublethal mutants that have accumulated under the protection of closely linked genes that are strongly selected, but a necessary part of the genetic adaptation of the population.

Natural selection will maintain weak alleles because an individual with all the strong alleles will be vigorous but of lower fitness. Therefore, selection maintains the average individual in the population at that level of vigour which maximises fitness. This may explain observations that the level of heterosis in wild populations depends on their habitat, sometimes termed habitat-dependent heterosis (Lloyd, 1980). The fitness advantage of heterozygotes is often greater under more severe conditions (eg., Dudash, 1990).

Genetic variation allows more vigorous and less vigorous individuals to exist and take advantage of changing conditions under which they may have the advantage. This is a mixed strategy; and as Von Neumann & Morgenstern (1953) showed, mixed strategies are necessary to achieve minimax outcomes.

But artificial selection of plants in cultivation aims to maximise desired parameters - yield, size etc - within an artificial, optimum environment. It may therefore be possible to purge the alleles that cause inbreeding depression from cultivated lines.

References

Apirion, D. & Zohary, D. (1961) Chlorophyll lethal in natural populations of the orchard grass (Dactylis glomerata L.). A case of balanced polymorphism in plants. Genetics 46: 393-399.

Barrett, S.C.H. & Charlesworth, D. (1991) Effects of a change in the level of inbreeding on genetic load. Nature 352: 522-524.

Berg, H. & Redbo-Torstensson, P. (1999) Offspring performance in three cleistogamous Viola species. Plant Ecology 145: 49-58.

Bingham, E.T.; Groose, R.W.; Woodfield, D.R. & Kidwell, K.K. (1994) Complementary gene interactions in alfalfa are greater in autotetraploids than in diploids. Crop Sci. 34: 832-829.

Buza, L., Young, A. & Thrall, P. (2000) Genetic erosion, inbreeding and reduced fitness in fragmented populations of the endangered tetraploid pea Swainsona recta. Biol. Conservation 93: 177-186.

Chapman, S.C., M. Cooper, M., D. G. Butler, D.G. & Henzell, R.G. (2000) Genotype by environment interactions affecting grain sorghum. I. Characteristics that confound interpretation of hybrid yield. Aust. J. Agric. Res. 51: 197-207

Cramer, C.S. & Wehner, T.C. (1999) Little heterosis for yield and yield components in hybrids of six cucumber inbreds. Euphytica 110: 99-108.

Dudash, M.R. (1990) Relative fitness of selfed and outcrossed progeny in a self-compatible, protandrous species, Sabatia angularis L. (Gentianaceae): a comparison in three environments. Evolution 44: 1129-1139.

Dudash, M.R. & Carr, D.E. (1998) Genetics underlying inbreeding depression in Mimulus with contrasting mating systems. Nature 393: 682-684.

Falconer, D.S. (1989) Introduction to Quantitative Genetics. (Longman: New York).

Fisher, R.A. (1927) On some objections to mimicry theory, statistical and genetic. Trans. Roy. Entom. Soc. London 75: 269-278.

Ford, E.B. (1964) Ecological Genetics. (Methuen: London).

Ford, E.B. (1965) Genetic Polymorphism. (Methuen: London).

Fu, H.H. & Dooner, H.K. (2002) Intraspecific violation of genetic colinearity and its implications in maize. Proc. Nat. Acad. Sci. U.S.A. 99: 9573-9578.

Gustafsson, A. (1950) The cooperation of genotypes in barley. Hereditas 39: 1-18.

Husband, B. & Schemske, D. (1997) The effect of inbreeding in diploid and tetraploid populations of Epilobium angustifolium (Onagraceae): implications for the genetic basis of inbreeding depression. Evolution 51: 737-746.

Kruuk, L.E.B., Clutton-Brock, T.H., Slate, J., Pemberton, J.M., Brotherstone, S. & Guiness, F.E. (2000) Heritability of fitness in a wild mammal population. Proc. Nat. Acad. Sci. 97: 698-703.

Lloyd, D.G. (1980) Demographic factors and mating patterns in angiosperms. In Solbrig, O.T. (ed) Demography and Evolution in Plant Populations. (Blackwell: Oxford).

Milborrow, B. (1998) A biochemical mechanism for hybrid vigour. J. Exp. Bot. 49: 1063-1071.

Pickett, A.A. & Galwey, N.W. (1997) A further evaluation of hybrid wheat. Plant Varieties & Seeds 10: 15-32.

Sheppard, P.M. (1953) Polymorphism and population studies. Symp. Soc. Exp. Biol. 7: 274-289.

Sprague, G.F. (1983) Heterosis in maize: theory and practice. Monogr. Theor. Appl. Genet. 6: 47-70.

Von Neumann, J. & Morgenstern, O. (1953) Theory of Games and Economic Behavior, 3rd ed.

Willis, J.H. (1999) Inbreeding load, average dominance and the mutation rate for mildly deleterious alleles in Mimulus guttatus. Genetics 153: 1885-1898.

Xiao, J.; Li, J.; Yuan, L. & Tanksley, S.D. (1995) Dominance is the major genetic basis of heterosis in rice as revealed by QTL analysis using molecular markers. Genetics 140: 745-754.

Back to Botany page

Home