7.2.4 Sources of genetic variation and restrictions on breeding
(a) Self- versus cross-pollination
The extent of genetic variation in a population often relates to its breeding system. Plants are either self- or cross-pollinated, or a mixture of the two mechanisms may operate in a single plant or species. Self-pollination, also known as autogamy, is the transfer of pollen from anther to stigma of the same flower or another flower on the same plant by a pollinating agent. In extreme examples self-pollination is automatic, when the anthers contact the stigma of the same flower either in the open flower or in the unopened bud. Mechanisms include homogamy, which is simultaneous maturation of the male and female organs, and cleistogamy, in which pollen is shed and the stigma is receptive before anthesis. Cleistogamy operates in peas and coffee, and assists plant breeders to generate true-breeding (homozygous) plant lines through simple repeated selﬁng. Cross-pollination, also known as allogamy, involves the transfer of pollen from an anther on one plant to a stigma of a flower on another plant. Plants with high levels of cross-pollination show greater genetic variation than those which are self-pollinated, as there is more opportunity for gene recombination and gene flow within a population.
Plants have evolved a wide range of outbreeding mechanisms which place restrictions on self-fertilisation and have the overall effect of increasing cross-fertilisation and hence genetic diversity. This maintains a high level of heterozygosity within a popu-lation and avoids a phenomenon called inbreeding depression, which relates to accumulation of alleles which are deleterious as homozygotes. The main outbreeding mechanisms are spatial or temporal separation of sexes, sexual incompatibility and male or female sterility.
(b) Spatial separation of sexes
Unlike the typical hermaphrodite ‘perfect’ flower, in about 10% of species not all flowers or individual plants are sexually identical (Irish and Nelson 1989). The most common type of spatial separation is monoecy, with female and male repro-ductive organs borne on unisexual flowers within the same plant. It is characteristic of maize and the family Cucurbitaceae. There are many variations on this theme, including andromonoecy with male and hermaphrodite flowers on the same plant, gynomonoecy with female and hermaphrodite flowers on the same plant, and androgynomonoecy (trimonoecy) with female, male and hermaphrodite flowers on the same plant (Table 7.5). The effect on pollination probabilities is to enhance, but rarely to guarantee, outbreeding.
Species producing unisexual flowers on different plants are termed dioecious. Plants with female flowers are gynoecious, and male-bearing ones are androecious. Clearly, these single-sex plants do not have the capacity to self-fertilise, so they are obligate outbreeders. This is not the case with some of the variants, including androdioecy with male and hermaphrodite flowers on different plants, gynodioecy with female and hermaphrodite flowers on different plants, and androgynodioecy (trioecy) with female, male and hermaphrodite flowers on different plants (Table 7.5). Pistachio (Pistacia vera), kiwifruit (Actnidia chinensis) and Casuarina (Figure 7.25a, b) are dioecious as are most pawpaw (Carica papaya) genotypes.
(c) Sex expression
Sex expression in plants is both genetically and developmentally determined. Heteromorphic sex chromosomes are recognised in asparagus and hop, but in many plants sex development is influenced by plant hormones. This may relate to expression of alternate sets of sex genes rather than presence or absence of female and male chromosomes. Applied auxins or ethylene promote femaleness, as in cucumber, pine, papaya and date. Gibberellins tend to promote maleness in cucumber, mulberry and oil palm, and cytokinins will induce hermaphrodites in male grapes. There are, however, many exceptions. Auxins promote male cone buds in some gymnosperms such as Pseudotsuga, and ethylene promotes maleness in Chinese chestnut. Gibberellins are present in higher concentrations in female than in male inflorescences of carob and date palm, and applied gibberellins promote femaleness in maize and Chinese chestnut. Some of this apparent confusion may relate to mismatch between physiological hormone levels and usually very high concentrations of applied plant growth regulators, and to differences in the forms active in sex expression. In cucumber, ethylene may be the primary hormone determining overall plant sex expression (Yin and Quinn 1995), but we cannot yet explain the subtle mechanisms that control patterns of male and female flowers within individual monoecious plants. In this species there are strong positional effects with exclusively male flowers on basal nodes, but a relatively high proportion of females further up the stem and on lateral branches. Environmental factors affect this pattern: low light intensity or low nutritional status often favours maleness, whereas femaleness is promoted by high light intensity and good nutrition. Temperature and photoperiod can also be influential, with short days and low temperatures enhancing femaleness in cucumber. Some Australian plants show variable sex expression, and in Leptospermum pistil abortion can occur at any stage of floral development.
(d) Temporal separation of sexes
In this mechanism, termed dichogamy, the female and male organs of the flower mature at different times, thus reducing the probability of fertilisation of a pistil with pollen from the anther of the same flower. In protogynous flowers, for example ﬁg and avocado (Persea americana), the female matures before the male, whereas in protandrous species, for example the Australian genera Banksia and Leptospermum (Figure 7.25c), the opposite occurs. In monoecious lychee (Litchi chinensis), spatial and temporal separation are combined, with a male–female–male sequence within each inflorescence. Dichogamy is most effective at preventing selﬁng within a single flower, but in most plants, especially trees with many thousands of flowers opening over a period of several weeks, there is a continuum of male and female function and consequently many chances of selﬁng.
Many plants have evolved sophisticated genetic mechanisms which prevent mating by self or related individuals by disrupting molecular interactions during pistil–pollen recognition. This is self-incompatibility, of which there are three major types: gametophytic, homomorphic sporophytic and heteromorphic sporophytic. In all cases, prevention of self-fertilisation is due to expressed alleles common to both parents. Gametophytic incompatibility refers to pollen–pistil interactions genetically controlled by the haploid (hence gametophytic) genome of the pollen grain and the diploid genome of the pistil tissue (Figure 7.26). This version occurs in Prunus, Lycopersicon and Nicotiana and is attributed to one or more multiallelic S loci. In sporophytic homomorphic incompatibility, S-genes are again involved but the pollen–pistil interaction is genetically controlled by the diploid (hence sporophytic) genome of the parent plant in which the pollen developed, and the diploid genotype of the pistil tissue, as in the genus Brassica. The pollen parent effect probably relates to deposition of tapetum (i.e. parental) proteins into the pollen coat. These proteins are recognised by the female parent at the stigma surface. Sporophytic heteromorphic incompatibility is found in Averrhoa(carambola, starfruit), Primula and Linum (flax). In Primula, heteromorphic refers to the two floral forms called pin and thrum. In the former, the style is long so that the pistil has the appearance of a pin but the stamen ﬁlaments of the pin flower are short. In the latter, the style is short and the stamen ﬁlaments are long, giving the flower the appearance of a thrum, a fringe of threads. The mechanism is controlled by a single locus with two alleles, S and s, with S dominant to s. The pin plant is homozygous recessive ss, whereas the thrum plant is heterozygous Ss. Sporophytic pollen control maintains the population as ss or Ss, because SS progeny do not survive.
Gametophytic and sporophytic self-incompatibility (GSI and SSI) systems differ in a number of ways. As already stated, the principal difference is that GSI is governed by the haploid genotype of the pollen, as opposed to the diploid genotype of the pollen parent in SSI. Thus SSI prevents sibling mating as well as self-mating and is a more exclusive system. In addition, GSI alleles act independently, whereas dominance relationships in pollen and pistil are common in SSI. The pollen of most plants with GSI is binucleate on release from the anther and germinates well in vitro, whereas that of plants with SSI tends to be trinucleate and is more difﬁcult to grow in vitro. In GSI, pollen tubes are generally inhibited in the upper third of the style, whereas in SSI they are inhibited on the stigma. Pollen tube inhibition is controlled by S-glycoprotein which is the product of the S-gene. Callose is deposited in the stigma papillae adjacent to inhibited SSI pollen grains, but this is not part of the GSI mechanism (Figure 7.27c, d). At present, detailed genetic and molecular studies of self-incompatibility have been restricted mainly to Brassica, Nicotiana and Prunus. Many other species do not conform exactly, and a generic model will require wider-ranging research.
(f) Late acting self-incompatibility
Even in the absence of the hurdles of SSI and GSI, there are further potential barriers to fertilisation, collectively known as late-acting self-incompatibility. This can occur at various locations in the ovary and at different stages of the life cycle. For example, pollen tube inhibition occurs at the placenta in some eucalypts, or at the nucellus in Acacia retinodes. Embryo sac inhibition is seen in chestnut and cocoa, and in the latter species is reportedly controlled both gametophytically and sporophytically. In other instances, the mechanisms operate post-zygotically, with early embryo abortion following selﬁng reported in Rhododendron, Pinus, Liquidambar, Pseudotsuga, Olea, Picea and Persea. Possibly this is an early expression of inbreeding depression, caused by accumulation of deleterious alleles following selﬁng. At present, however, there is little direct evidence to support this or any alternative theory. Particularly in the tropics, many of these woody species have not been extensively bred in cultivation and are probably highly heterozygous. In addition, there are many examples of plants which yield poorly following selﬁng where post-zygotic embryo abortion is implicated, but the controlling mechanism has not been elucidated. These include Erythrina, Eucalyptus, Camellia, Vaccinium, Ziziphus, Carya, Anacardium and Hevea.
One intriguing element of partial self-incompatibility is that seeds with different male parents can exist on a single plant, even within a single inflorescence. Many studies have compared growth and survival rates of self fruits and cross-fruits (Denney 1992), and in most cases cross-pollinated fruits are larger (known as the xenia effect) and tend to show less premature fruit abscission. This occurs in many nut crops (macadamia, pecan, hazelnut, almond) where the economically important product is the seed itself, but is also in fruits such as lychee. This may relate partly to increased vigour of hetero-zygous individuals (heterosis). However, usually the seed and the fruit tissues are larger, yet the latter are derived entirely from maternal tissue, and therefore fruit growth must be stimulated indirectly as a result of the seed’s genotype. The commercial implications are diverse, with nut growers wanting maximum seed yield (Table 7.6) with minimised shell and fruit, whereas best-quality fruit are seedless or small seeded. Inappropriate genotype mixtures in citrus orchards can lead to very seedy although larger fruit (Table 7.7).
(g) Overcoming self-incompatibility
The ability to manipulate expression of incompatibility has become a vital tool in plant breeding, allowing hybridisation of otherwise incompatible parents. Pre-anthesis bud pollination is effective for species such as Brassica, as expression of S-glycoprotein (the product of the S-gene) is minimal until the flower is ready to open. Polyploidy can also be used, as self-incompatible diploid plants may become self-compatible when tetraploid, as in Leucaena. Low temperature in the range of 10–25°C is effective in almond, cherry and apple because the selﬁng optimum is lower than the crossing optimum, whereas temperatures above 32°C reduce S-glycoprotein activity in apple and pear. Old and end of season flowers have weaker self-incompatibility than young early flowers. Alternatively, removal of the stigma and top part of the style followed by pollen application to the cut surface can be effective, for example in cherry, because of removing the sites where the incompatibility reactions normally occur. Gamma irradiation of the style operates in much the same way. Finally, mentor pollen (that is, dead compatible pollen mixed with live incompatible pollen; see Feature essay 7.1) may allow self seed to be produced in otherwise self-incompatible genotypes of apple, pear, citrus and cherry.
(h) Interspeciﬁc incompatibility and incongruity
Interspeciﬁc hybridisation can be of enormous beneﬁt to plant breeders attempting to generate new genotypes. Although often hard to predict, success is frequently achieved in genera such as Citrus and Prunus, but rarely in others including Populus or between subgenera of Eucalyptus. Interspeciﬁc incompatibility may relate to taxonomic distance between species, as in Eucalyptus (Figure 7.27a, b), or to simple physical differences such as style length, which prevents pollen tubes of short-styled species reaching the ovary of long-styled species in Rhododendron, Prunus and Eucalyptus (Gore et al. 1990). In some cases it is related to self-incompatibility, for example the cross between the self-incompatible almond and the self-compatible peach is incompatible, whereas the reverse cross is compatible. We deduce that recognition is involved in interspeciﬁc incompatibility, because mentor pollen can assist interspeciﬁc fertility in apple and poplar. Live interspeciﬁc pollen plus dead compatible pollen or live interspeciﬁc pollen plus compatible pollen wall proteins are effective. Similarly, recognition by the stigma has been demonstrated by removal of stigma secretion with solvents which aids interspeciﬁc fertility in Eucalyptus and Populus. However, not all cases of interspeciﬁc sterility are related to pistil–pollen incompatibility, because some may be due to species incongruity. This is the failure of seed production due to non-relatedness, and thus to non-recognition at one or more stages of the pollination and seed development processes. At some stage of taxonomic divergence, interspeciﬁc sterility is attributable to incongruity rather than to interspeciﬁc incompatibility.
(i) Selection for self-compatibility
In commercial crops, self-incompatibility can drastically reduce yields, particularly in plantings consisting of limited numbers of genotypes. Selection for self-fertility genes in a self-incompatible population is a common aim of plant breeding programs. This may occur through natural hybridisation with a self-compatible relative, as in the cross between almond (Prunus dulcis) and P. webbii. A similar outcome can be achieved by controlled hybridisation with a self-compatible relative, for example crossing peach (P. persica) with almond (P. dulcis). Alternatively, mutation breeding has generated self-fertile genotypes of apple, cherry and almond. Although the mechanism is not clear, polyploidy can also be effective: tetraploid blueberries are self-compatible, unlike their diploid progenitors.
(j) Male and female sterility
Most botanists agree that ancestral flowering plants were hermaphrodite, and the subsequent evolution of dioecy required male or female sterility. Triploidy is an alternative mechanism which promotes both, as it invariably results in faulty meiosis and hence sterile gametes. For plant breeders, male sterility is extremely valuable in preventing self-fertilisation of otherwise self-compatible individuals.
Pollen production can fail due to mutation of one or more genes of which there are three major classes. Genic male sterility, also known as chromosomal or Mendelian sterility, operates via chromosomal genes with Mendelian inheritance. Cytoplasmic male sterility is also known as maternal, mito-chondrial or plastid sterility because it operates via the genomes of cytoplasmic organelles which are inherited only through the female parent. Gene–cytoplasmic male sterility is a com-bination of the two. The most obvious change is modiﬁcation of structural differentiation of the stamen resulting in absent or highly reduced anthers, as in male sterile cultivars of cucumber and tomato. Faulty differentiation of the anther can result in feminisation, as in sl (stamenless) mutant tomatoes (Table 7.8),or functional male sterility can result simply from the failure of pollen release from the anther, as in tomato, eggplant and grape. Breakdown in microsporogenesis at meiosis can result in loss of contact with the tapetum, as in tomato and squash, and abnormal tapetal development in onion and carrot can lead to abortion of the microgametophyte at the post-meiotic stage.
Female sterility is less well understood, but it can also manifest in a number of ways. Female flower abscission occurs in walnut, and absent, incomplete or retarded development of the embryo sac has been observed in plum and lychee. In mango and pistachio, the embryo sac can degenerate. Adverse environmental conditions can cause female sterility, such as spring frost effects on apple flowers, or incomplete style development due to cold weather during mango flower development.
Plants have evolved many ways to restrict inbreeding and promote outbreeding. Indeed, a single genus or species may exhibit multiple mechanisms, the relative importance of which may vary with habitat, environment and genotype. For example, walnut (Juglans regia) and lychee are both monoecious and dichogamous. Within the Prunus genus, almond (P. dulcis) is self-incompatible whereas peach (P. persica) is self-compatible. Understanding the ecological and agronomic implications of these diverse mechanisms can assist species conservation in the wild and exploitation in cultivation.