Sexual Reproduction in Flowering Plants

A flower looks decorative, but structurally it is a shoot whose growth has stopped and whose leaves have been modified into four whorls. Working from the outside inwards, these are the calyx made of sepals, the corolla made of petals, the androecium made of stamens, and the gynoecium made of one or more carpels. The first two whorls are accessory: they protect the bud and, in animal-pollinated species, advertise it. The last two are the essential whorls, because they build the male and the female gametophytes. A flower that carries both of them is bisexual; a flower carrying only stamens or only carpels is unisexual.

Each stamen has a slender stalk called the filament and a swollen anther at its tip. A typical angiosperm anther is bilobed and is described as dithecous, and it carries four microsporangia in all, two in each lobe, one at each corner of the anther in cross section. Each microsporangium is a pollen sac and it develops into a pollen chamber packed with pollen grains. Seen in cross section, a young microsporangium is roughly circular and its wall is built of four layers. From outside inwards these are the epidermis, the endothecium, one to three middle layers, and the tapetum. The outer three layers protect the sporangium and help the anther split open when it is ripe; the endothecium develops fibrous thickenings that pull the wall apart as it dries. The tapetum is the innermost layer and its job is purely nutritive. Its cells have dense cytoplasm and usually more than one nucleus, and they feed the developing pollen grains. If the tapetum is damaged, pollen development fails.

The centre of a young microsporangium is packed with compactly arranged sporogenous tissue. Each cell of this tissue can act as a microspore mother cell, also called a pollen mother cell, and it is diploid. When it undergoes meiosis it produces four haploid microspores held together in a cluster called a tetrad. The whole process of forming microspores from a pollen mother cell through meiosis is microsporogenesis. As the anther matures and dehydrates, the tetrads separate and each microspore develops into a pollen grain. The arithmetic here is worth fixing in your head: one mother cell, one meiosis, four pollen grains. Twenty-five mother cells are therefore enough to produce a hundred pollen grains.

A pollen grain is the male gametophyte. It is usually spherical and about 25 to 50 micrometres across, and its wall has two layers. The outer layer, the exine, is made of sporopollenin, one of the most resistant organic materials known. It survives high temperature, strong acids and strong alkalis, and no enzyme that degrades it has been identified so far. That is why pollen walls turn up intact in fossil deposits millions of years old. The exine is not continuous: it is interrupted at germ pores, thin spots through which the pollen tube later emerges. Beneath the exine lies the intine, a thin continuous layer of cellulose and pectin.

Inside its wall, a mature pollen grain is not one naked cell. It contains a large vegetative cell packed with reserve food and carrying a large, irregularly shaped nucleus, plus a small spindle-shaped generative cell with dense cytoplasm that literally floats in the cytoplasm of the vegetative cell. In more than sixty per cent of flowering plants the grain is shed from the anther at this two-celled stage. In the remaining species the generative cell divides mitotically before shedding, so the grain leaves the anther at the three-celled stage carrying a vegetative cell and two male gametes. Pollen viability varies enormously between species. Pollen of cereals such as rice and wheat may lose the ability to germinate within about thirty minutes of release, while pollen of many members of the families Rosaceae, Leguminosae and Solanaceae stays viable for months. Pollen can be kept for years in liquid nitrogen at about minus 196 degrees Celsius, which is how pollen banks are maintained for crop breeding. Pollen also has a public-health side: grains of Parthenium, the carrot grass that came into India as a contaminant of imported wheat, are a well known cause of pollen allergy.

The gynoecium is the female half of the flower. A single unit of it is a carpel, also called a pistil, and it has three regions: a stigma at the top that acts as a landing platform for pollen, an elongated style below it, and a swollen ovary at the base. Inside the ovary is a cushion of tissue called the placenta, and the ovules are attached to it. The number of ovules in an ovary is a species-level character: wheat, paddy and mango have a single ovule per ovary, while papaya, watermelon and orchids have many.

An ovule is the megasporangium. It is attached to the placenta by a stalk called the funicle, and the point where the funicle joins the body of the ovule is the hilum. The body is wrapped in one or two protective coverings, the integuments, which enclose it completely except for a small opening at one end called the micropyle. The opposite end, where the integuments arise, is the chalaza. Inside the integuments is a mass of nutritive parenchyma, the nucellus, and embedded in the nucellus towards the micropylar end is the female gametophyte, better known as the embryo sac.

Megasporogenesis is the formation of megaspores from the megaspore mother cell. A single cell in the micropylar region of the nucellus enlarges, develops dense cytoplasm and a prominent nucleus, and becomes the megaspore mother cell. Being a cell of the nucellus it is diploid. It undergoes meiosis and produces four haploid megaspores. In most flowering plants only one of these four remains functional; the other three degenerate. Because the entire female gametophyte is then built from that one surviving megaspore, this pattern is called monosporic development.

The functional megaspore now divides mitotically, but the divisions are free-nuclear: a nucleus divides and no cell wall is laid down after it. The first division gives two nuclei, which move to opposite ends of the enlarging sac. Two further rounds of division give first four nuclei and then eight, four gathered at each end. Only after the eight-nucleate stage do cell walls appear.

The way those walls appear is worth building up step by step. Think of the mature embryo sac as an oval bag with two ends: a micropylar end, the end nearer the micropyle, and a chalazal end opposite it. Three cells are organised at the micropylar end: one egg cell with a synergid on either side of it. Those three together are the egg apparatus. Each synergid carries special thickenings of its wall at the micropylar tip, and this structure is called the filiform apparatus; it is what guides an arriving pollen tube into the synergid. Three more cells, the antipodals, are organised at the chalazal end. That accounts for six of the eight nuclei. The remaining two, the polar nuclei, are not walled off separately at all. They stay together inside the single huge central cell that fills the middle of the sac. Now count: eight nuclei, but only seven cells, because two of the nuclei share one cell. So a typical angiosperm embryo sac is eight-nucleate and seven-celled; the description eight-celled is not correct, because two of the nuclei never get walls of their own.

One more comparison keeps the two gametophytes straight. The male gametophyte is a pollen grain of two or three cells and it leaves the parent plant. The female gametophyte is a seven-celled sac and it never leaves the ovule; it stays enclosed in the nucellus, inside the integuments, inside the ovary, on the parent plant, and the sperm must be brought to it.

Pollination is the transfer of pollen grains from an anther to the stigma of a pistil. Plants cannot walk to each other, so this transfer has to be arranged, and flowering plants have evolved a remarkable range of arrangements for it. The first way of classifying pollination is by where the pollen comes from.

Autogamy is pollination within the same flower: pollen from an anther reaches the stigma of that very flower. For this to work, the anther and the stigma must lie close together and the pollen must be released exactly when the stigma is receptive. In flowers that open normally, called chasmogamous flowers, both organs are exposed, so autogamy is possible but never guaranteed. Some species also produce cleistogamous flowers, which never open at all. In a cleistogamous flower the anthers dehisce inside the closed bud right beside the stigma, so seed set is assured even if no pollinator ever visits. The cost is that cross-pollination is impossible and the offspring show no new combinations. Viola, the common pansy, along with Oxalis and Commelina, produces both chasmogamous and cleistogamous flowers.

Geitonogamy is the transfer of pollen from an anther of one flower to the stigma of a different flower on the same plant. Read that definition carefully, because it sits exactly between the other two. Functionally it is cross-pollination, since a pollinating agent must carry the pollen from one flower to another. Genetically it is self-pollination, because both the pollen and the ovule belong to the same plant and therefore the same genome. Xenogamy is the transfer of pollen to the stigma of a flower on a different plant. It is the only one of the three that brings genetically different pollen to the stigma, and therefore the only one that produces real recombination.

Pollinating agents fall into two groups. Abiotic agents are wind and water; biotic agents are animals. The great majority of flowering plants use biotic agents. Abiotic pollination is a numbers game: pollen is scattered blindly, most of it never lands anywhere useful, and the plant compensates by producing enormous quantities of it. Both wind and water pollination therefore go with very large pollen output.

Wind-pollinated flowers share a recognisable set of features. The pollen is light, dry and non-sticky. The stamens are well exposed so that moving air can strip the grains away. The stigma is large and often feathery, giving a wide surface to trap drifting grains. Each ovary usually contains a single ovule, and many small flowers are packed into an inflorescence. The corn cob is the standard example: the long threads hanging out of it are the styles and stigmas of the female flowers. Wind pollination is especially common in grasses.

Water pollination is genuinely rare. It is limited to about thirty genera, mostly monocots. In Vallisneria the female flower is carried up to the water surface on a long stalk, while male flowers or free pollen grains are released onto the surface and drift passively with the current until some of them reach a stigma. In sea grasses such as Zostera, the female flowers stay submerged and the pollen grains are long and ribbon-like, carried inside the water itself rather than on its surface. In most water-pollinated species the grains have a mucilaginous covering that protects them from wetting. One misconception is worth correcting here: many aquatic plants, including water hyacinth and water lily, are pollinated not by water but by insects or wind, with their flowers held above the water surface.

Animal pollinators include bees, butterflies, moths, flies, beetles, wasps and ants, along with birds and bats. A few other vertebrates take part too: primates such as lemurs, tree-dwelling rodents, and even some reptiles have been recorded as pollinators. Bees are the dominant group. Animal-pollinated flowers are typically large, coloured, scented and stocked with nectar; where the individual flowers are small, they are grouped into a conspicuous inflorescence so the display is worth a visit. Flowers pollinated by flies and beetles often smell of rotting matter rather than perfume. The usual reward is nectar or surplus pollen, but in some species the reward is a safe place to lay eggs. Amorphophallus, whose flower is taller than an average adult, works this way. The relationship between Yucca and the moth that pollinates it, and between fig plants and the wasp that pollinates them, are obligate mutualisms: neither partner can complete its life cycle without the other. The moth lays its eggs in the locule of the ovary and pollinates the flower while doing so, and its larvae later feed on some of the developing seeds. The fig wasp similarly uses the fig as an egg-laying site and pollinates the enclosed flowers while searching for one.

Continued self-pollination is a poor long-term strategy. It reduces variation and leads to inbreeding depression, in which the progeny become progressively weaker. Many flowering plants have therefore evolved devices that make self-pollination difficult or impossible. Four of these outbreeding devices are worth knowing precisely.

The first device is timing. In many species the pollen is released before the stigma of the same flower becomes receptive, or else the stigma becomes receptive long before the anthers dehisce. Either way the male and female organs of one flower are never ready at the same moment. The second device is position. The anther and the stigma are placed at different heights or angles within the flower so that pollen cannot fall onto the stigma of that same flower. The third device is self-incompatibility. This is a genetic mechanism: the pistil recognises pollen from the same plant and blocks it, either by preventing the pollen from germinating on the stigma or by stopping the growth of the pollen tube in the style. Note that self-incompatibility rejects pollen from the same plant, so it defeats geitonogamy as well as autogamy.

The fourth device is the production of unisexual flowers, and here you have to be careful about exactly what is prevented. If male and female flowers are borne on the same plant, as in castor and maize, then no single flower has both organs, so autogamy is prevented. But geitonogamy is still perfectly possible, because both kinds of flowers belong to one plant and a pollinator can move between them. Only when the male and female flowers are borne on separate plants, as in papaya, are both autogamy and geitonogamy prevented, because the plant that makes the pollen has no ovules at all.

Once pollen lands, the pistil does not accept it automatically. Pollen-pistil interaction covers every event from the moment a grain lands on the stigma to the moment the pollen tube enters the ovule. The stigma acts as a screening surface. Chemical signals pass between the pollen wall and the stigma, and the pistil either accepts a compatible grain and promotes its germination, or rejects an incompatible one and stops it.

A compatible grain absorbs water from the stigma, and the intine bulges out through one of the germ pores to form the pollen tube. The contents of the grain move into this tube. The tube grows down through the tissue of the stigma and then the style and reaches the ovary. If the grain had been shed at the two-celled stage, the generative cell divides mitotically inside the pollen tube during this journey and forms the two male gametes. If the grain had already been shed at the three-celled stage, the two gametes were formed back in the anther and simply travel down the tube. On reaching the ovary, the tube enters the ovule through the micropyle and then pushes into one of the two synergids, guided in by that synergid's filiform apparatus. There it releases both male gametes into the cytoplasm of the synergid. It does not inject them directly into the egg. The synergid that receives them degenerates.

Breeders exploit this whole sequence deliberately in artificial hybridisation, where the aim is to control exactly which pollen reaches which stigma. If the chosen female parent bears bisexual flowers, its anthers are removed from the flower bud with forceps before they dehisce; this step is emasculation. The emasculated bud is then covered with a bag, usually of butter paper, so that no unwanted pollen can reach the stigma; this step is bagging. When the stigma becomes receptive, the bag is opened briefly, mature pollen collected from the anthers of the desired male parent is dusted on the stigma, and the flower is rebagged and left until the fruit develops. If the female parent bears unisexual flowers, emasculation is unnecessary, since the female flowers have no anthers to remove. The female buds are simply bagged before they open, pollinated with the chosen pollen when the stigma is receptive, and rebagged.

Double fertilisation is the event that defines flowering plants, and it is easiest to follow if you first list what has arrived and what is waiting. Arriving, inside the pollen tube and now discharged into the cytoplasm of one synergid, are two male gametes, both haploid. Waiting are a haploid egg cell at the micropylar end of the embryo sac, and two haploid polar nuclei sitting together in the large central cell.

One male gamete moves to the egg cell and fuses with it. The fusion of two gametes is called syngamy, and its product is the zygote. Since a haploid gamete has fused with a haploid egg, the zygote is diploid, written 2n.

The other male gamete moves to the central cell and fuses with the two polar nuclei there. This second fusion involves three haploid nuclei in all, one male gamete plus two polar nuclei, and is therefore called triple fusion. Its product is a single nucleus with three sets of chromosomes, the primary endosperm nucleus, written 3n. The central cell that now contains it is called the primary endosperm cell, and it will go on to form the endosperm. Note the wording carefully: triple fusion is the fusion of three nuclei, not of three gametes. Only one of the three participants is a gamete; the other two are ordinary nuclei of the female gametophyte.

Because two fusions, syngamy and triple fusion, occur inside the same embryo sac, the whole event is called double fertilisation. It is unique to flowering plants and does not occur in any other group of plants.

What makes it distinctive is not just the arithmetic but the logic behind it. In gymnosperms the nutritive tissue that will feed the embryo is haploid female gametophyte tissue, and it is laid down before fertilisation, whether or not fertilisation ever happens. Any ovule that is never fertilised has already cost the plant a store of food. In flowering plants the nutritive tissue is triploid endosperm, and it is created only by the second fusion, which means only after a male gamete has actually arrived. So the plant commits its food reserves to an ovule only once it is certain that an embryo will form there. That efficiency is one reason double fertilisation is regarded as a major advance.

Keep the ploidy of the two products separate and secure. The zygote is 2n and develops into the embryo, which is therefore also 2n. The primary endosperm nucleus is 3n and develops into the endosperm, which is therefore 3n. Both are products of fertilisation, but only one of them is a new individual; the other is a feeding tissue that the new individual will consume.

Once double fertilisation is over, the ovule and the ovary both start changing. The order in which they change matters: endosperm development always begins before embryo development. The advantage is obvious. By the time the zygote starts dividing, there is already a food supply sitting around it, so the young embryo never has to feed itself.

The primary endosperm cell divides repeatedly to form the endosperm tissue, which is filled with reserve food. In the commonest pattern, the primary endosperm nucleus goes through a series of free-nuclear divisions with no wall formation, producing a large number of free nuclei; this stage is called free-nuclear endosperm. Cell walls are laid down later and the endosperm becomes cellular. A tender coconut shows both stages in one fruit: the coconut water is free-nuclear endosperm, containing thousands of nuclei in a common cytoplasm, and the white kernel around it is the cellular endosperm that formed afterwards.

What happens to the endosperm later divides seeds into two groups. In some species the developing embryo eats the endosperm completely before the seed is ripe, storing the food in fleshy cotyledons instead. Such mature seeds have no endosperm and are called non-albuminous or ex-albuminous seeds; pea, groundnut and beans are examples. In other species a good deal of endosperm is left over in the ripe seed, which is then called albuminous or endospermic; castor, coconut, wheat, maize and barley are examples. Occasionally a bit of the nucellus also survives into the mature seed as a separate food store. This persistent nucellus is called perisperm, and it is seen in black pepper and beet. Perisperm is nucellar tissue of the mother plant, so it is diploid, while endosperm is triploid; that ploidy difference is the cleanest way to tell them apart in an answer.

Embryo development, or embryogeny, takes place at the micropylar end of the embryo sac where the zygote lies. Most zygotes wait until a certain amount of endosperm has formed before they begin dividing. The zygote first forms a proembryo, which becomes a globular embryo, then a heart-shaped embryo, and finally a mature embryo.

A typical dicot embryo has an embryonal axis and two cotyledons. The part of the axis above the level at which the cotyledons attach is the epicotyl, and it ends in the plumule, the shoot tip. The cylindrical part below the level of the cotyledons is the hypocotyl, and it ends in the radicle, the root tip, which is covered by a root cap. Remember the pair by their prefixes: epi means above and gives the shoot, hypo means below and gives the root.

A monocot embryo, such as that of a grass, has only one cotyledon, called the scutellum, and it is attached to one side of the embryonal axis rather than terminally. At the lower end of the axis the radicle and root cap are enclosed in an undifferentiated sheath called the coleorrhiza. The portion of the axis above the point where the scutellum attaches is the epicotyl, and it carries a shoot apex with a few leaf primordia, all enclosed in a hollow foliar sheath called the coleoptile. Keep the two sheaths apart by their endings: coleorrhiza wraps the radicle, coleoptile wraps the plumule.

The ovule as a whole matures into the seed. The integuments harden into the tough seed coats, and the micropyle survives in the seed coat as a tiny pore through which water and oxygen enter at germination. As the seed ripens, its water content falls until the seed is relatively dry, holding only about ten to fifteen per cent moisture by mass, and the metabolic activity of the embryo slows down sharply. The seed may then enter a period of inactivity called dormancy, or germinate straight away if conditions are favourable. Dry, dormant seeds can survive astonishing lengths of time: seeds of Lupinus arcticus recovered from Arctic tundra, estimated to be around ten thousand years old, germinated and flowered, and a roughly two thousand year old seed of the date palm Phoenix dactylifera from the Dead Sea region has also been grown successfully.

Meanwhile the ovary matures into the fruit and the ovary wall becomes the fruit wall, or pericarp. The pericarp may be fleshy, as in guava, orange and mango, or dry, as in groundnut and mustard. In most species the fruit develops from the ovary alone and is called a true fruit. In a few species, such as apple, strawberry and cashew, the thalamus also grows and contributes to the fruit, which is then called a false fruit. Some fruits develop from the ovary without fertilisation at all; this is parthenocarpy, and because no ovule was fertilised the fruit is seedless. Banana is the standard example, and parthenocarpy can be induced artificially by applying growth hormones.

It is worth stating plainly why the seed and the fruit matter so much. A seed packages a young plant along with its own food supply inside a protective coat, so the new generation begins life stocked and shielded rather than exposed. Because the seed is dry and metabolically slow, it can wait out a season that would kill a seedling, which is what gives dormancy its value. The seed is also the unit that travels: dispersal by wind, water or animals carries the offspring away from the parent, which reduces competition with the parent plant for light, water and minerals, and lets the species colonise new ground. The fruit is the structure that makes much of that dispersal possible, since a fleshy pericarp attracts animals that carry the seeds away, while dry winged or hooked fruits catch the wind or an animal's coat. Fruits also protect the developing seeds until they are ripe. On top of all this, seeds are the basis of settled agriculture, because a dry seed can be stored from one season to the next and sown when conditions suit. Summarising the fates in one line: ovary becomes fruit, ovary wall becomes pericarp, ovule becomes seed, integuments become seed coat, zygote becomes embryo, primary endosperm nucleus becomes endosperm, and a persistent nucellus becomes perisperm.

The whole chapter so far has described seeds as the product of fertilisation. Some flowering plants break that rule. Apomixis is a form of asexual reproduction that mimics sexual reproduction: a seed is formed, complete with a seed coat and an embryo, but no fertilisation has taken place. It is found in several species of the family Asteraceae and in many grasses.

Apomictic seeds can be produced in more than one way. In some species the egg cell is formed without meiosis, so it is diploid rather than haploid, and this diploid egg then develops directly into an embryo without any male gamete fusing with it. In many varieties of Citrus and mango a different route is taken: cells of the nucellus surrounding the embryo sac start dividing, push into the embryo sac, and develop into embryos there. Since the nucellus is ordinary diploid tissue of the mother plant, every embryo formed this way carries exactly the mother's genotype.

That is the key genetic point about apomixis. Because no gametes fuse and no meiosis contributes, there is no segregation and no recombination. The offspring are genetically identical to the mother plant, so an apomictic seed is effectively a clone packaged as a seed.

Polyembryony is the occurrence of more than one embryo in a single seed. The nucellar route described above is the commonest cause in Citrus and mango: because many nucellar cells can be recruited at once, a single ovule can end up with many embryos of different sizes. If you cut open the seed of an orange carefully you can see several small embryos side by side. Apomixis and polyembryony often appear together, but they are not the same idea. Apomixis is about the absence of fertilisation; polyembryony is about the number of embryos. A seed can have several embryos formed sexually, for instance if more than one embryo sac develops in an ovule, and an apomictic seed may contain just one embryo.

Why breeders care is worth stating clearly. Hybrid varieties give much higher yields, but a hybrid does not breed true: if a farmer sows seed collected from a hybrid crop, the characters segregate in the next generation and the yield advantage is lost. So hybrid seed must be produced afresh every season by controlled crossing, which is expensive, and farmers must buy it again each year. If hybrid varieties could be made apomictic, every seed they set would be a clone of the high-yielding hybrid plant. The characters would not segregate, and a farmer could keep sowing seed saved from the previous crop year after year without any loss of hybrid vigour. This is why the genetics of apomixis is an active area of crop research.