Jump to ContentJump to Main Navigation
Understanding Flowers and Flowering$

Beverley Glover

Print publication date: 2007

Print ISBN-13: 9780198565970

Published to Oxford Scholarship Online: January 2008

DOI: 10.1093/acprof:oso/9780198565970.001.0001

Subscriber Login

Forgotten your password?


Why Are Flowers Different? Pollination Syndromes—The Theory

(p.127) CHAPTER 13 Why Are Flowers Different? Pollination Syndromes—The Theory
Understanding Flowers and Flowering

Beverley J. Glover

Oxford University Press

Abstract and Keywords

It is clear from a merely cursory glance around any garden in the summer months that flowers come in an enormous variety of sizes, shapes, colours, and scents. The book now focusses on the differences between flowers, as opposed to the molecular similarities that unite them. This chapter begins by considering the different ways that flowers can be pollinated. It is a basic premise underlying much of floral biology that differences in pollination system explain many of the differences in floral form. The evidence to support this premise is not as compelling as we might like to think, as discussed in later chapters. However, to set the stage for those discussions, this chapter looks at the historical concept of the pollination syndrome and the predictions it makes about floral morphology. The chapter considers the roles different animal pollinators may play in influencing floral evolution.

Keywords:   animal pollinators, floral evolution, floral form, floral morphology

Until this point, we have largely ignored the concept that the flowers of different species may be very different from one another. However, it is clear from a merely cursory glance around any garden in the summer months that flowers come in an enormous variety of sizes, shapes, colours, and scents. In Section III of this book we focus on the differences between flowers, as opposed to the molecular similarities that unite them. In this chapter we begin by considering the different ways that flowers can be pollinated. It is a basic premise underlying much of floral biology that differences in pollination system explain many of the differences in floral form. The evidence to support this premise is not as compelling as we might like to think, as will be discussed in later chapters. However, to set the stage for those discussions, in this chapter we look at the historical concept of the pollination syndrome and the predictions it makes about floral morphology.

13.1 Cross pollination

If cross pollination is to occur within a population of flowering plants, it is necessary for pollen to be transported from the anthers in a flower on one individual to the stigma in a flower of a second individual. The alternative is self-pollination, when pollen is transferred from the anther of a single flower onto the stigma of the same flower (autogamy), or, confusingly, is transferred from the anther of one flower to the stigma of a second flower on the same plant (geitonogamy). Self-pollination is thought of as an adaptation to unreliable pollen vectors, particularly seen in short-lived annual plants and those species which regularly invade new habitats. However, self-pollination is not optimal for genetic recombination and so in the long term may compromise the genetic diversity and evolutionary lability of a population (see discussion in Chapter 12). Therefore a great many flowering plants have adaptations which maximize their potential for cross pollination. Since plants are immobile, this transport of pollen from plant to plant must be accomplished with the aid of a pollen vector. The pollen vector may be abiotic or biotic.

13.2 Abiotic pollen vectors

13.2.1 Wind pollination

Abiotic pollen vectors consist primarily of the wind and water. Of these, wind pollination (anemophily) is by far the more common, being found in 18% of angiosperm families (including the ecologically dominant Poaceae, the grasses) as well as in the conifers (Ackerman 2000). Wind pollination is a secondarily derived state in angiosperm flowers, as the earliest flowers are believed to have been pollinated by beetles (Thien et al. 2000). Anemophily is thought to evolve in response to changes in the environment which decrease the efficiency of biotic pollination while enhancing the success of wind pollination. Examples of environmental changes which might cause a shift towards wind pollination include variations in pollinator abundance, migration into an area with a very dry climate, or the arrival of another plant species which might compete for pollinator attention (Culley et al. 2002). Many plant species have been shown to be both wind- and animal-pollinated, either simultaneously or at different (p.128) times within the same season (Cox 1991). This combination of biotic and abiotic pollination systems is known as ambophily, and may be the transitional route through which full wind pollination most commonly evolves.

Angiosperm species that are wind-pollinated typically show a particular set of characteristics. Most of these characteristics have been shown to enhance the success of wind pollination, so can be thought of as adaptations to a wind-pollinated reproductive system (Culley et al. 2002). As wind pollination is a passive process, many of these adaptations are to do with the environments and habitats in which anemophilous plants live. To begin with, wind-pollinated species typically inhabit environments with stronger winds but lower humidity than those of biotically pollinated species. Wind-pollinated plants are found in habitats with low rainfall, ensuring pollen is not washed away. Wind-pollinated species are usually found in moderately high densities within an area, but with low densities of other species surrounding them. This combination of features maximizes the chances of released pollen landing on another individual of the same species.

Besides common features of environment and habitat, wind-pollinated species also share a typical set of floral characteristics. Flowers and inflorescences are commonly held above or away from vegetation, which maximizes access by the wind to reproductive organs. Very often the inflorescences are pendulous, like those of willow trees. The flowers within these inflorescences are usually unisexual. It would be hard to prevent self pollen clogging the stigmas of flowers exposed to the wind, if both male and female reproductive organs were present in a small area. The outer two whorls of floral organs, the petals and sepals, are usually much reduced or absent in wind-pollinated flowers, and, where present, do not usually produce much scent or pigmentation. The loss of a brightly coloured and strongly scented corolla is unsurprising in view of the unconscious nature of the wind as a pollen vector. The stamen filaments of wind-pollinated flowers are usually long, exposing the locules to the wind, and the pollen grains they produce are smooth-surfaced, small in size, and enormous in quantity. Small size and a smooth surface have been shown to enhance the aerodynamic properties of pollen travelling in wind (Niklas, 1985), while large quantities are necessary to increase the chances of wind-blown pollen reaching a stigma. The stigmas themselves are usually long and feathery, and protrude a considerable way

                         Why Are Flowers Different? Pollination Syndromes—The Theory

Figure 13.1 The long tasselled flowers of a wind-pollinated grass hang far from the main body of the plant.

(p.129) beyond the outer edges of the flower, maximizing surface area available for pollen capture (Culley et al. 2002). Wind-pollinated grass flowers are shown in Fig. 13.1.

13.2.2 Water pollination

Water pollination (hydrophily) is a relatively rare system of gamete transfer, but it is employed by a few species of grasses and waterweeds, occurring in only 31 genera of 11 different families, 9 of the families in the Monocots (Cox 1988). In most hydrophilous species, pollen is released below the water surface and carried passively by currents to female reproductive structures. This mechanism of water pollination is used by many marine plants, such as the Caribbean turtle grass, Thalassia testudinum, which releases pollen grains underwater, bound up in strands of mucilage. These are carried by currents below the water surface, and the mucilage encourages them to stick to the female flowers, which are also underwater.

However, in some species pollen is released onto the surface of the water and is carried on the surface to female reproductive structures. The pollen of such plants, for example Elodea canadensis (the Canadian waterweed), is ornamented with many tiny spikes, which trap pockets of air and ensure the pollen floats. Anther dehiscence is often explosive, scattering the pollen grains widely across the water surface. The female reproductive structures may always be held at surface level in some hydrophilous angiosperms, but in others the female flower closes around the pollen and is then withdrawn below the water surface after pollination. A variation on the theme of releasing pollen onto the water surface is shown by species of tape-grass, Vallisneria, which release entire male flowers onto the water, with the pollen never leaving the anthers until it arrives at a female flower. The female flowers of many hydrophilous species create depressions in the surface tension of the water, encouraging passing pollen grains to slide down towards them. While hydrophilous flowers do produce relatively large quantities of pollen, this use of water surface tension dynamics to facilitate pollination does mean that less pollen is required than might be predicted (Cox 1988).

13.3 Biotic pollen vectors

Biotic pollen vectors provide some or all of the pollination service for the majority of flowering plants. The frequency with which we observe this interaction leads to a common assumption that plants and animals are cooperating in a mutualistic association. However, animals rarely deliberately pollinate flowers, but instead view flowers as food sources. The reward offered is usually a mixture of nectar, a sugar solution (which may contain various concentrations of several different sugars, along with other nutrients), and pollen itself, which is very rich in amino acids. In collecting this food, animals inadvertently brush against the reproductive organs of the flower, causing pollen to be transferred from stamens to their bodies, and from their bodies onto stigmatic surfaces. Many flowers show structural features which enhance this inadvertent contact between pollinator body and floral reproductive organs, such as narrow corolla tubes or dorsally positioned anthers which deposit pollen on the back or head of the pollinator.

13.4 Principles underlying the pollination syndrome concept

Biotic pollen vectors consist of a range of animals— most commonly insects, as well as birds, bats, and a small number of other vertebrates. In a seminal book published in 1966, Faegri and van der Pijl set out to ‘formulate the general principles of pollination ecology, applicable anywhere’. A major device used by Faegri and van der Pijl was the pollination syndrome, a suite of floral traits associated with the attraction of a particular group of pollinators. The assumption behind the pollination syndrome concept is that coevolution between plant and animal has led to the acquisition by plants of floral features which maximize their chances of attracting and being pollinated by a particular animal or group of animals.

13.4.1 Do flowers act as specialist or general advertisements?

The pollination syndrome concept explains the differences between different flowers as consequences (p.130) of the types of animals that they attract. This concept relies on the principle that characteristics such as particular colours, scents, shapes, and markings are specific to flowers pollinated by specific types of animals. The different floral characteristics can be viewed as animal-specific advertisements, adapted to maximize the chances of attracting a particular type of honeybee or butterfly or beetle.

However, there is an alternative way of explaining the many differences between different flowers. It was Darwin himself who first articulated the reason flowers are brightly coloured:

Flowers rank amongst the most beautiful productions of nature; but they have been rendered conspicuous on contrast with the green leaves, and in consequence at the same time beautiful, so that they may be easily observed by insects. (Darwin 1859)

On first sight this might be read as consistent with the pollination syndrome concept, in saying that different colours attract different animals. In fact all that this quote says is that colour (and by extension shape and scent) simply provides contrast between flowers and the green vegetation around them. In this case any feature that makes a flower stand out even more against the green backdrop is likely to enhance the success of that flower in attracting pollinators. In this argument, all the different floral characteristics are simply a number of different ways that plants have hit upon to make their flowers more attractive to a broad spectrum of pollinators. Instead of providing species-specific advertisements, the different floral characteristics might simply represent an almost infinite number of different advertisements, all of which serve to attract animals in general. While it is true that each possibility is likely to be correct in the case of certain plant species, it should be borne in mind that the pollination syndrome concept rests on the idea of specialization, and may not be compatible with a view of flowers as non-specific advertisers of rewards to generalist pollinators.

13.4.2 Plants and pollinators want different things, making species-specific interactions unlikely

Plants and their pollinators have different requirements of the interaction, and these differences may reduce the likelihood of coevolution occurring between individual species. A plant will experience maximum reproductive fitness if it maximizes pollen dispersal between individuals of the same species, without its stigmas becoming clogged with pollen of other species and without wasting much energy on nectar production. Therefore its optimal pollen vector will be an animal which alights only briefly on each flower, moves rapidly between individuals, is faithful to only one plant species, and does not eat much.

The animal will conserve energy if there is no need to forage, and so should prefer flowers with a large reward, where it can stay a long time. If the reward is insufficient in one species it should forage on a range of plant species.

However, very specific interactions between plant and pollinator carry their own risks. A one-on-one relationship between flower and pollinator should mean that little pollen is wasted, and so its production can be less prolific, saving energy. Female fertility should also be enhanced as foreign pollen will not be applied to the stigma, where it can block acceptable pollen from germinating. However, such an extreme specialist can expect to set no seed at all if the pollinator is absent in any one season, through disease or climatic changes, a situation which would be catastrophic for an annual plant and may also be very disadvantageous for a perennial.

It is generally argued that for these reasons the evolution of plants and their pollinators has followed a middle road, with groups of pollinators coevolving with groups of flowers, rather than one-on-one species-specific interactions. It is the broad association between a group of flowers and the group of animals that pollinates them that has come to be called a pollination syndrome (Faegri and van der Pijl 1966).

13.5 The pollination syndromes

A pollination syndrome classically describes a suite of adaptations shown by a plant to a taxonomic order of animals, and by those animals to a particular group of flowers, which may not be phylogenetically related to each other. The adaptations shown by the animals may be behavioural or morphological, while the plants can only show (p.131) morphological adaptations, which may include flower size, structure, colour, reward, and timing of floral induction and opening. A table summarizing the key floral traits traditionally associated with different pollinators is shown in Fig. 13.2.

13.5.1 Beetle pollination (Cantherophily)

Beetle pollination is widely believed to have been the first pollination syndrome, the one used by the first angiosperms, as the Coleoptera, the beetles, constitute one of the oldest groups of insects and were already numerous at the time that the angiosperms came into existence, having themselves arisen around 260–280 mya (Ponomarenko 1995). Fossilized pollen found in the digestive tracts of beetles has shown that they had already acquired the habit of grazing on the pollen of cycads, conifers, and other gymnosperms, before the angiosperms appeared (Labandeira 1997; Thien et al. 2000). When considering a pollination syndrome, it is useful to begin with the features of the animal's biology relevant to its role as an agent of pollen transfer. Beetles have mouth parts positioned parallel to the axis of the body, which limits

                         Why Are Flowers Different? Pollination Syndromes—The Theory

Figure 13.2 Table to show the general features of flowers involved in different pollination syndromes.

their ability to manipulate food sources, particularly those with depth. They are also quite big animals, with a moderately high demand for protein as well as carbohydrate. Beetles do not have good colour vision, but do have a strong sense of smell and a particular attraction to fruity smells. Flowers that have become adapted for pollination by beetles might be expected to have characteristics which make them both attractive to beetles and easy to obtain food from once the animal has landed. These features would include the provision of a floral reward consisting of both nectar (for sugar) and pollen (for protein) in a flat structure from which the reward can be lapped. Beetle-pollinated flowers are usually saucer or bowl-shaped, with nectar secreted into the shallow bowl, and often have excess anthers to produce extra pollen as a reward. Since colour vision is unimportant to beetles, bee-tle-pollinated flowers might be expected to waste little energy on the production and modification of pigments, but instead to produce a strong scent. Many beetle-pollinated flowers do produce a fruity fragrance, and they are often greenish or off-white in colour. The classic example of a beetle-pollinated flower is magnolia, essentially unchanged for (p.132) 100 million years and still pollinated by the same sorts of animals (Fig. 13.3a). Lilies, wild roses, and some poppies are also often beetle-pollinated.

13.5.2 Fly pollination (Myiophily)

The Diptera, the fly order, show the greatest variation in methods and habits of pollination of any

                         Why Are Flowers Different? Pollination Syndromes—The Theory

Figure 13.3 Insect-pollinated flowers. (a) Magnolia flowers are beetle-pollinated. (b) The fly-pollinated flowers of Fatsia japonica. (c) Bumblebee (Bombus terrestris) entering a Hebe flower. (d) Many daisies are butterfly-pollinated. (e) The flowers of Angraecum sesquipedale have very long nectar spurs and are pollinated by extremely long-tongued moths. Photographs (a), (d), and (e) kindly supplied by Cambridge University Botanic Garden and H. Rice. See also Plate 10.

group of insects. This makes it hard to define the morphological traits that we would expect to see in a fly-pollinated flower. Indeed, some flies behave very like beetles, and pollinate flowers morphologically very similar to beetle-pollinated flowers (Thien et al. 2000). Other flies may have more in common with wasps, and pollinate flowers with very different structural features. One feature of flies which has been utilized by flowering plants is their persistence throughout all seasons. Flies are one of the few groups of insects that are not strictly periodic, and, as such, many plants flowering under adverse conditions or at odd times of the year can be entirely dependent on flies for their pollination. Another feature common to all flies is that they do not feed their offspring, and are usually lighter-bodied than many other insects. In consequence they do not need much food, and fly-pollinated flowers usually supply only a small quantity of nectar. Although there is much variation within the order, in general flies are more visually acute animals than beetles, and have a positive preference for pale and yellow colours, and for nectar guides. Fly-pollinated flowers are often coloured cream or yellow, and do not usually have much scent. Classic examples of fly-pollinated flowers include species such as carrot and other members of the Umbelliferae, as well as species such as groundsel and other Asteraceae. A fly-pollinated Fatsia japonica flower is shown in Fig. 13.3b.

13.5.3 Bee pollination (Melittophily)

Of the Hymenoptera, the bees have specialized most towards a diet of nectar and pollen. Wasps will take nectar to meet their sugar requirements, but do not actively collect either nectar or pollen. There are a few specialized instances of ants feeding on nectar, often from extra-floral nectaries, but for the most part it is the bees that are the great pollinators of the Hymenopteran order. Bees are large animals, and have a substantial energy requirement. They usually forage for nectar for themselves, and pollen to feed to larvae back in their hive or brood chamber. The pollen is carried on their bodies, in specialized structures that range from simply having hairy feet to having pollen baskets on the hind legs, and is then groomed off back at the hive. In response to these (p.133) features, bee-pollinated flowers are usually quite large to bear the weight of the animal, and often have a clear landing platform. They are also often closed, with petals or other structures which must be pushed aside by the bee to access the nectar. This sort of mechanism prevents nectar robbing by other animals, and helps to keep bees constant, as the supply of nectar is likely to be good. Inside these large flowers is usually a reasonable volume of nectar, and a small amount of excess pollen. Bees have the ability to perceive depth and many species have long tongues. The nectar is often secreted at the base of deep corolla tubes or nectar spurs, again denying access to species without the size and tongue length to access it. Since pollen found on the bee's body is groomed off and fed to the larvae, many bee-pollinated flowers have dorsal anthers which deposit the pollen on the back of the bee's neck, a position from which it cannot easily be groomed. Bees have good colour vision and can see in ultraviolet, blue, and yellow, but do not have receptors to perceive red which they see only as a weak green signal. Bee-pollinated flowers are usually brightly coloured, with yellow and blue being considered classic bee colours. However, many bee-pollinated flowers are red or pink. It has been shown that in many of these cases the presence of UV absorbing pigments modifies the red to a colour more highly visible to the bee (see Chapter 18; Chittka and Waser 1997). As quite intelligent animals, bees are sensitive to nectar guides, which enable them to handle the flowers more quickly by directing them straight to the nectar. Many bee-pollinated flowers have nectar guides visible either in the range of the spectrum that we can see or else in the ultraviolet part. Classic examples of bee-pollinated flowers include the garden snapdragon (Antirrhinum majus), nettles, and foxgloves. A bumblebee visiting a Hebe inflorescence is shown in Fig. 13.3c.

13.5.4 Butterfly pollination (Psychophily)

In contrast to bees, the other most prominent group of pollinating invertebrates in temperate climates does not feed its offspring and has relatively low energy requirements. Butterflies are relatively light in weight, and usually alight on flowers, which reduces their energy expenditure. They also have long tongues, often 1 to 2 cm in depth. Flowers pollinated by butterflies usually combine a flat structure to alight on with deep tubes in which the nectar is secreted. This can be achieved either by presenting a tube-shaped flower with a landing rim around it, like a buddleia, or else by clustering a number of small tubes together to produce a larger flat surface, the system used by members of the Asteraceae, such as Michaelmas daisies (Fig. 13.3d). Butterflies are not known to have much sense of smell at all, and butterfly-pollinated flowers do not usually produce much scent. However, butterflies do have good colour vision, and can see red. Butterfly-pollinated flowers are usually brightly coloured, with reds and yellows often predominant.

13.5.5 Moth pollination (Phalaenophily or Sphingophily)

Although moths and butterflies are members of the same insect order, the Lepidoptera, their methods of pollination and the structures of the flowers they visit are completely different. The major reason for this is the difference in their behaviour—most moths are nocturnal while butterflies are diurnal. Moths also prefer not to land on flowers, collecting their nectar while hovering. Many moths are also much heavier in the body than butterflies. The combination of a heavy body and hovering flight means that moths have very high energy requirements indeed, and moth-pollinated flowers usually produce more nectar than butterfly- or bee-pollinated flowers. To facilitate nectar collection while hovering, moth-pollinated flowers are usually bilaterally symmetrical, and presented to the pollinator with the petal lobes bent backwards and the corolla tube open for easy access. To prevent theft of this large supply of easily accessible nectar by other animals, it is common for moth-pollinated flowers to close up or fold over during the day, and only open at night. Alternatively, the nectar may be presented in a tube so long that it is inaccessible to other animals. Moth tongues are longer than those of any other insect, and moth-pollinated flowers often have much longer nectar spurs than any other flower type. The extreme example of this (p.134) relationship is the story of the Madagascan orchid, Angraecum sesquipedale, which has a nectar spur up to 30 cm in length (Fig. 13.3e). In 1862, on seeing one of these plants in flower at the Royal Botanic Gardens, Kew, Darwin predicted that there must exist a moth with a proboscis long enough to reach the nectar at the bottom of this spur. He was proved correct in 1903 with the identification of the hawkmoth Xanthopan morganii ssp. praedicta, which has a tongue of a similar length to the nectar spur, and has been shown to pollinate the flower in a caged environment. Colour vision is irrelevant to moths, which fly at night, and moth-pollinated flowers are often white or cream. This allows them to stand out against the dark vegetation at night. However, moths do have a good sense of smell, and it is usual for moth-pollinated flowers to release a strong scent, often in the evening. Classic examples of moth-pollinated flowers include gardenia and some honeysuckles.

13.5.6 Bird pollination (Ornithophily)

Since the first pollinating animals were insects, the later emergence of pollinating vertebrates meant that these animals had access to flowers already adapted for pollination by another animal. There are sufficient similarities between butterfly- and bird-pollinated flowers to suggest that the early pollinating birds fed from flowers which had coevolved with butterflies. There is some speculation as to how bird pollination evolved, with some authors suggesting that birds hunting nectar-gath-ering insects in flowers accidentally discovered nectar. It is certainly true that even hummingbirds, which take almost all of their energy from nectar, still eat the occasional insect to meet their protein requirements. The classic pollinating bird is the hummingbird, which is found only in the Americas, but a variety of other groups also feed on nectar, including the African sunbirds, Australian lorikeets, and American honey-creepers. Different types of birds have different feeding strategies. For example, hummingbirds hover beneath or in front of flowers while feeding, so hummingbird-pollinated flowers tend to be either pendant (like a fuchsia, Fig. 13.4a) or stand out with free space in front. Sunbirds, on the other hand, perch while feeding, so sunbird-pollinated flowers have a perch with the nectary facing towards it (Fig. 13.4b). Bird-pollinated flowers are usually either brush or tube-shaped, and the nectar is secreted into spurs, which are usually shorter and wider than those on butterfly flowers. They also must be quite tough, as a beak is both stronger and harder than a butterfly's tongue. One of the key features of a bird-pollinated flower is the quantity of nectar secreted. Birds are larger animals than the invertebrates discussed previously, and their energy requirements can be very high indeed. They feed from flowers which produce very large quantities of very concentrated nectar, so much so that it will actually drip from the flowers at certain times of the year. Birds have good colour vision (see Chapter 18) and the flowers they pollinate are usually red, often with contrasting yellow marks to act as nectar guides. However, scent is not important in bird pollination. Classic examples of bird-pollinated flowers include red columbine, poinsettia, eucalyptus, hibiscus, and passion flower.

13.5.7 Bat pollination (Chiropterophily)

A quarter of all bat species use flowers for food to some extent, and a few species rely on flowers for all of their nutritional requirements. Bats are large, heavy animals which sometimes land on the flowers they feed from. Those flowers tend to be large and robust, and usually saucer-shaped for ease of lapping nectar. Because bats must find the flowers at night they are not usually within the foliage, but either hang below it for easy access (Fig. 13.4c) or actually develop on the trunk of the plant itself. Bat–pollinated flowers produce the most nectar of any flower type, with as much as 15 ml sometimes recorded from single flowers. Those bat species which have become entirely dependent on flowers eat pollen as their only protein source. The flowers they visit usually have greatly enlarged anthers, and sometimes a very great number of them. The flowers of some bat-pollinated species commonly develop over 2000 anthers per flower. Those anthers may only open at night, to prevent pollen robbing by beetles. Since bats are nocturnal, the flowers themselves may sometimes only open at night and may only last one night before senescing. (p.135)

                         Why Are Flowers Different? Pollination Syndromes—The Theory

Figure 13.4 Vertebrate-pollinated flowers. (a) The pendant form of Fuchsia flowers is ideal for hovering hummingbirds. (b) Bird of Paradise (Strelitzia regina) flowers provide a sturdy landing platform for non-hovering birds. Photograph kindly supplied by Cambridge University Botanic Garden. (c) The flowers of Strongylodon macrobotrys, the jade vine, hang far below the foliage, making them readily accessible to bats. See also Plate 11.

Bats are colour-blind, so flower colour is irrelevant in attracting them. Bat-pollinated flowers are usually white to cream or sometimes a greenish pink colour. The main attractant used by flowers to attract bats is scent. Bat-pollinated flowers generate a very strong scent, often containing butyric acid. Bats which are specialized to feed on nectar often have a nasal cavity larger than that of insectivorous bats, suggesting that scent is more important to flower bats than to insectivorous bats. In contrast the sonar apparatus is sometimes reduced, as little time is spent hunting insects. Flower bats have longer tongues and narrower snouts than insectivores, and the tongue often has papillae on the end for lapping up the nectar. Classic examples of batpollinated flowers include many cacti, as well as members of the Bignoniaceae (trumpet creepers) and the Bombaceae.

(p.136) 13.5.8 Deceit pollination

The term deceit pollination covers such an enormous range of relationships that it would be impossible to describe them all in this section. Instead we will consider three examples of deceit pollination, each involving increasingly specialized flowers which appear to have coevolved with particular pollinating animals. Although the relationships considered in this section are more complex than the basic pollination syndromes described earlier, they still represent recognizable syndromes with clearly apparent matches between flower morphology and animal behaviour.

The simplest form of deceit pollination is a version of Batesian mimicry, a phenomenon which occurs when an organism mimics another and has particular features of the model attributed to it. For example, the yellow and black banding of hoverflies is a form of Batesian mimicry, causing predatory birds to avoid the hoverfly in case it has the same poisonous sting as the wasps and bees it mimics. A similar situation exists where species of plants mimic the flowers of other species, but do not provide a reward. This sort of pollination system requires the mimic to develop the appropriate colour, shape, and scent to match the model, but to fail to produce nectar. The match does not have to be perfect, as the visual acuity of many pollinators is insufficient to distinguish between generally similar flowers. For instance, the rewardless orchid Orchis israelitica has been shown to function as a mimic of the nectar-supplying Bellevalia flexuosa, even though the similarity between them is only at the very general level of inflorescence architecture, approximate floral shape, and white colour (Galizia et al. 2005). Another orchid species, Dactylorhiza sambucina, produces no reward and mimics a range of rewarding species. Within Europe there are two colour morphs of this species, yellow and magenta, and recent reports indicate that negative frequency-dependent selection maintains this colour polymorphism, as naïve bees preferentially visit the rarer morph, having not yet learned to associate it with a lack of reward (Gigord et al. 2001). This form of mimicry occurs where a large number of flowering plants of different species grow in a defined area and flower at similar times. The flowers mimicked may be pollinated by any of the common pollinating animals, so the mimic may develop all the features (except the reward) of any of the standard pollination systems. This sort of pollination by deceit saves the plant energy, but it can only be successful when the mimic is at low frequencies in the community; otherwise the pollinators will abandon both the mimic and the genuine flowers.

A less diffuse system of deceit pollination is called sapromyiophily. Sapromyiophilous flowers attract carrion and dung flies and carrion and dung beetles, by mimicking the appearance and scent of a piece of rotting flesh or of carnivore or herbivore faeces. The insects are attracted to the flower to feed or lay their eggs, and transfer pollen in the process. The flower provides no reward to the insect, and may in fact reduce its fitness by causing eggs to be laid on a structure which will not provide sustenance to any maggots that hatch. A range of different colours and scents are used by sapromyiophilous flowers, but they are commonly dark red, brown, or blue-tinged. They may have a surface covering of fine hairs, believed to enhance the visual mimicry of dead flesh. Most flowers that attract egg-laying insects also release a strong scent. A recent report of the components of the scents of a range of sapromyiophilous flowers in the family Apocynaceae observed that different species released different mixtures of hexanoic acid, carboxylic acids, pyrazines, heptanal, octanal, dimethyl oligosulphides, indole, and cresol, and that the mix of compounds released gave specific corpse, urine, or dung odors (Juergens et al. 2006). To facilitate the dispersal of scent, some flowers are thermogenic, generating heat through a salicylic acid-activated signal transduction cascade. There are many examples of sapromyiophilous flowers, particularly in the orchid family, but the most famous example is the world's largest inflorescence, the titan arum (Fig. 13.5a).

The most specialized pollination system by deceit is that used by flowers which mimic the female of a species of insect, and invite the male to attempt to mate with them. The colours and scents produced by these mimicking flowers are highly species specific, and may include the release of compounds which mimic the insect's pheromones. (p.137)

                         Why Are Flowers Different? Pollination Syndromes—The Theory

Figure 13.5 Floral mimicry. (a) The Titan arum (Amorphophallus titanium) attracts pollinators by releasing a strong scent reminiscent of rotting flesh. Image kindly provided by Cambridge University Botanic Garden. (b) Ophrys episcopalis, which mimics female insects to achieve pollination through pseudocopulation. Image kindly provided by Richard Bateman (Royal Botanic Garden, Kew). (c) The composite inflorescence of Gorteria diffusa mimics its pollinating flies. See also Plate 12.

Colours usually involve yellows, purples, and browns, and may be distributed on the flower to mimic the appearance of the wing cases of the female. No reward is provided to the pollinating animal, but, because this kind of mimicry releases instinctive behaviour, the animal cannot learn to avoid these flowers, and so they are not as fre-quency-limited as food deceivers. A classic example of a flower pollinated in this way is the fly orchid, Ophrys insectifera, which mimics the female of the scoliid wasp Campsoscolia ciliate. Although the visual similarity between the flower and the wasp is very striking, it is the scents produced by the flower which are most astonishing. The orchid releases a range of very specific volatiles, which closely mimic the sex pheremones of the female wasp. Some of these, such as (omega-1)-hydroxy acids and (omega-1)-oxo acids, are not found elsewhere in the plant kingdom (Ayasse et al. 2003). A number of related species produce similarly impressive mimics (Fig. 13.5b). A less extreme, but nonetheless remarkable, example is the South African ‘beetle daisy’, Gorteria diffusa. The composite inflorescence of these daisies contains bright orange ray florets, several of which develop large black spots (Fig. 13.5c). The spots are composed of both intense pigmentation and specialized cells which add three-dimensionality and sheen. They attract male bee flies, Megapalpus nitidus, by mimicking the female flies which often spend the night in the closed inflorescences (Johnson and Midgely 1997). These are perhaps some of the most extreme examples of coevolution between plant and pollinator, and certainly give much credit to the idea that floral features have evolved in response to pollinator preferences.

The different pollination syndromes discussed in this chapter vary from the extremely diffuse (fly pollination) to the very specialized (some of the deceit pollination examples). While there is little doubt that some specialized flowers have evolved traits which make them more attractive to particular animals, there is a great deal of debate in the literature as to whether the majority of ‘ordinary’ flowers can really be said to show pollination syndromes. In the next few chapters we will discuss the mechanisms by which the great diversity of shapes and colours is generated, before returning to consider the adaptive significance of these features in Section IIIB.