Pollination and Floral Ecology

By Pat Willmer

Pollination and Floral Ecology is the most comprehensive single-volume reference to all aspects of pollination biology--and the first fully up-to-date resource of its kind to appear in decades. This beautifully illustrated book describes how flowers use colors, shapes, and scents to advertise themselves; how they offer pollen and nectar as rewards; and how they share complex interactions with beetles, birds, bats, bees, and other creatures. The ecology of these interactions is covered in depth, including the timing and patterning of flowering, competition among flowering plants to attract certain visitors and deter others, and the many ways plants and animals can cheat each other.



Pollination and Floral Ecology pays special attention to the prevalence of specialization and generalization in animal-flower interactions, and examines how a lack of distinction between casual visitors and true pollinators can produce misleading conclusions about flower evolution and animal-flower mutualism. This one-of-a-kind reference also gives insights into the vital pollination services that animals provide to crops and native flora, and sets these issues in the context of today's global pollination crisis.

• Provides the most up-to-date resource on pollination and floral ecology


• Describes flower advertising features and rewards, foraging and learning by flower-visiting animals, behaviors of generalist and specialist pollinators--and more


• Examines the ecology and evolution of animal-flower interactions, from the molecular to macroevolutionary scale


• Features hundreds of color and black-and-white illustrations

• Language: English
• Publisher: Princeton University Press
• Release date: Jul 5, 2011
• ISBN: 9781400838943

Pat Willmer

Pat Willmer is professor of zoology at the University of St. Andrews. She has published extensively on pollination biology in leading scientific journals. Her books include Environmental Physiology of Animals.

Book preview

POLLINATION AND FLORAL ECOLOGY

POLLINATION AND FLORAL ECOLOGY

Pat Willmer

PRINCETON UNIVERSITY PRESS

PRINCETON AND OXFORD

Copyright © 2011 by Princeton University Press

Published by Princeton University Press, 41 William Street,

Princeton, New Jersey 08540

In the United Kingdom: Princeton University Press, 6 Oxford Street,

Woodstock, Oxfordshire OX20 1TW

press.princeton.edu

Jacket photograph: The bee Eucera longicornis foraging on the sweet pea

Lathyrus. (C) Andrej Gogala

All Rights Reserved

Library of Congress Cataloging-in-Publication Data

Willmer, Pat, 1953-

   Pollination and floral ecology / Pat Willmer.

          p. cm.

   Includes bibliographical references.

   ISBN 978-0-691-12861-0 (cloth : alk. paper) 1. Pollination. 2. Pollination by insects. 3. Pollination by animals. 4. Plant ecology. I. Title.

QK926.W55 2011

571.8′642—dc22                        2010049061

British Library Cataloging-in-Publication Data is available

This book has been composed in Times

Printed on acid-free paper ∞

Printed in the United States of America

10  9  8  7  6  5  4  3  2  1

CONTENTS

Preface

Acknowledgments

PART I ESSENTIALS OF FLOWER DESIGN AND FUNCTION

Chapter 1         Why Pollination Is Interesting

Chapter 2         Floral Design and Function

Chapter 3         Pollination, Mating, and Reproduction in Plants

Chapter 4         Evolution of Flowers, Pollination, and Plant Diversity

PART II FLORAL ADVERTISEMENTS AND FLORAL REWARDS

Chapter 5         Advertisements 1: Visual Signals and Floral Color

Chapter 6         Advertisements 2: Olfactory Signals

Chapter 7         Rewards 1: the Biology of Pollen

Chapter 8         Rewards 2: the Biology of Nectar

Chapter 9         Other Floral Rewards

Chapter 10       Rewards and Costs: The Environmental Economics of Pollination

PART III POLLINATION SYNDROMES?

Chapter 11       Types of Flower Visitors: Syndromes, Constancy, and Effectiveness

Chapter 12       Generalist Flowers and Generalist Visitors

Chapter 13       Pollination by Flies

Chapter 14       Pollination by Butterflies and Moths

Chapter 15       Pollination by Birds

Chapter 16       Pollination by Bats

Chapter 17       Pollination by Nonflying Vertebrates and Other Oddities

Chapter 18       Pollination by Bees

Chapter 19       Wind and Water: Abiotic Pollination

Chapter 20       Syndromes and Webs: Specialists and Generalists

PART IV FLORAL ECOLOGY

Chapter 21       The Timing and Patterning of Flowering

Chapter 22       Living with Other Flowers: Competition and Pollination Ecology

Chapter 23       Cheating by Flowers: Cheating the Visitors and Cheating Other Flowers

Chapter 24       Flower Visitors as Cheats and the Plants’ Responses

Chapter 25       The Interactions of Pollination and Herbivory

Chapter 26       Pollination Using Florivores: From Brood Site Mutualism to Active Pollination

Chapter 27       Pollination in Different Habitats

Chapter 28       The Pollination of Crops

Chapter 29       The Global Pollination Crisis

Appendix

Glossary

References

Subject Index

Index of Animal Genera

Index of Plant Genera

PREFACE

When I first became interested in pollination biology around thirty years ago, there was essentially just one book to consult for core information, the classic by Faegri and van der Pijl, The Principles of Pollination Ecology, which has been the inspiration to generations, its last edition appearing in 1979 but still structured around the 1966 original. But twenty years on I became frustrated that there had been no single volume since that date which covered the field. This is particularly extraordinary given that the subject itself has expanded and blossomed prodigiously, both in its own right and as a testing ground for wider theories, as well as having assumed greater importance in both scientific circles and as a wider public concern.

It has not been all quiet on the publishing front since the 1970s of course. There have been excellent books with more of a natural history approach (M. Proctor et al. 1996; Buchmann and Nabhan 1996) and botanical books concentrating on floral structures (Barth 1985; Endress 1994) or on floral development (Glover 2007). Over the same period many extremely useful edited volumes have also appeared, each with chapters contributed by experts updating readers on particular themes, such as nectar and nectaries (Bentley and Elias 1983; Nicolson et al. 2007), anthers and pollen (D’Arcy and Keating 1996; Dafni et al. 2000), and floral scents (Dudareva and Pichersky 2006). Several books have compiled useful reviews on practical techniques (Dafni 1992; Kearns and Inouye 1993; Dafni et al. 2005) or on applied aspects of crop pollination (Free 1993), while Chittka and Thomson (2001) took a fresh approach by assembling chapters on aspects of pollinator cognition and learning.

But perhaps most important have been a select few edited volumes with broader pollination coverage and sometimes a more eclectic assemblage of contributions. Here the editors have been very much the leaders in the field: Jones and Little (1983) began the trend with Handbook of Experimental Pollination Ecology, while Lloyd and Barrett (1996) produced Floral Biology: Studies on Floral Evolution in Animal Pollinated Systems, opening up a broader and more ecological and evolutionary perspective on the whole subject. Then two collected contributions came along in 2006, picking up the themes that had been emerging in the preceding decade: Waser and Ollerton in Plant-Pollinator Interactions: From Specialization to Generalization, and Harder and Barrett in Ecology and Evolution of Flowers. These have been seen as especially important in expounding new approaches and exposing older assumptions to a more critical analysis. The debates about generalization and specialization, and the usefulness of the syndrome concept or of alternative modeling approaches, have been particularly thought-provoking, so that anyone starting out as a researcher in pollination ecology will find much to engage them from the authors represented in those two books.

Those readers, however, would not get the background information, the core material of pollination and floral biology. A beginner would still have to explore very widely across the mass of scientific journals, hoping to light upon a reasonably up-to-date review article. And in the last decade the primary literature has undergone an extraordinary explosion in volume, almost more than in any other field, partly because of the practical and applied environmental problems now seen to be emerging with pollination services, and partly because around twenty-five years ago interest began to spread much more widely beyond the European and North American temperate pollination systems, with expertise emerging especially in the New and Old World Tropics, in southern Africa and Australia, and in Japan and China.

For all these reasons, it seemed to me timely to attempt a book that would cover the subject more thoroughly, providing chapters on every key topic, and serve both as an introduction for anyone coming into the field as advanced undergraduate or postgraduate and as a source book for professionals. Begun half way through the decade, it has inevitably burgeoned beyond the original intentions as the expanding literature burst around my ears, to the distress of my publishers. But the resulting book will, I hope, achieve at least part of its aims and will not too greatly annoy those highly respected authors whose work I draw heavily on. Some, seeing pollination biology as pioneering a new era where ecological modelers have more to tell us than old-style field workers, might find my approach too traditional. But at least those coming in to the field hereafter may gain a better understanding of the foundations on which models can be built and so will better see where the key issues may lie.

Pat Willmer, November 2009

ACKNOWLEDGMENTS

It is a pleasure to begin by thanking the many generations of students and postgraduates who have let me know when I taught them in ways that inspired their interest and when I merely confused them. From them, I realized where the real fascinations of pollination lay and where my own interests were too biased or quirky for a wider audience—I have tried, not always successfully, to play down those quirks. Many former postgraduates and postdocs, now running research groups of their own and blossoming into major players in the world pollination and mutualism debates, have been especially important to me, and I thank them all for discussions, arguments, and the chance to read and learn from their own works; in particular I must pay tribute to Graham Stone (GS) and Simon Potts (SP), and more recently to Clive Nuttman (CN), Betsy Vulliamy, Gavin Ballantyne (GB), Caroline King, and Jonathan Pattrick, and to those whose names are followed by initials for use of their photos.

On a wider front, I confess to being a very poor networker, so that I know the pollination biologists of the world far less well on a personal level than I should, which is my loss. My communications with them have nevertheless been wonderfully invigorating, and I thank so many of them for their generosity in exchanging ideas. In particular, I have benefitted at various times from conversations about bees or pollination or wider mutualisms, whether prolonged or brief (and some so brief they will scarcely remember me!), with Jane Memmott, Steve Johnson, Judith Bronstein, Sally Corbet, Florian Schiestl, Dick Southwood, Naomi Pierce, Amots Dafni, Nigel Raine, Ingrid Williams, Cathy Kennedy, Francis Gilbert, Dino Martens, John Free, Jette Knudsen, James Cresswell, Steve Bullock, James Cook, Peter Yeo, Scott Armbruster, Mo Stanton, Spencer Barrett, Dini Eisikowitch, Barbara Gemmill, Peter Gibbs, Aubrey Manning, Gidi Ne’eman, Sue Nicolson, Samy Zalat, Theodora Petanidou, Chris O’Toole, and Rob Raguso.

Several of these colleagues have read and commented wisely on outlines, drafts, or chapters for me, and to them I owe enormous gratitude and an apology where I have insufficiently taken on their criticisms; the misconceptions or infelicities that remain are entirely my fault. Particular thanks to Clive Nuttman on color and scent issues, Peter Gibbs on incompatibility and on general issues of specialization and syndromes, Francis Gilbert (especially for updating my dipteran taxonomy), and Jonathan Pattrick for taking a broad beginner’s view of the book and being even more persnickety about semicolons than I am!

I am deeply grateful to many who have allowed me to use their photographs, which usually put my own efforts to shame. All are acknowledged by initials in the plates, and all of course retain their copyrights in the photos reproduced here. Susie Whiten (SW), Alison Fernie (AF), Ian Johnston (IJ), and Sandy Edwards (SE) were generous with their own photographs from the UK and abroad. Special thanks to Steve Johnson (SJ), Dino Martens (DM), Karen Sarkisyan (KS), and Andrej Gogala (AG) for supplying superb photographs of the more difficult animal visitors. I also owe a massive debt of gratitude to my illustrators, especially Dawn Toshack for most of the pictures of flowers and visitors and their detailed structures, and Caroline King for reproducing most of the graphical figures from other published works, as well as to Dimitri Kar-etnikov for advice on figures in general. I must also thank others at Princeton University Press who have patiently supported my efforts to complete this work; for the initial enthusiasm of Robert Kirk especially, and the subsequent diligence of Stefani Wexler, Beth Clevenger, Gail Schmitt, and Jennifer Slater, as well as the anonymous reviewers who offered comments at an early stage.

Finally a thank-you to many colleagues and friends who have helped to keep me sane through this whole process, especially Clare and Iain and Morven, Sandra who was so often a rock at work, Peter and Elisabeth, and neighbor Pam; all of these have been invaluable.

PART I

ESSENTIALS OF FLOWER

DESIGN AND FUNCTION

Chapter 1

WHY POLLINATION IS INTERESTING

Outline

Which Animals Visit Flowers?

Why do Animals Visit Flowers?

How do Flowers Encourage Animal Visitors?

What Makes a Visitor into a Good Pollinator?

costs, Benefits, and Conflicts in Animal Pollination

Why is Pollination Worth Studying?

The flowering plants (angiosperms) account for about one in six of all described species on earth and provide the most obvious visual feature of life on this planet. In the terrestrial environment, their interactions with other living organisms are dominant factors in community structure and function; they underpin all nutrient and energy cycles by providing food for a vast range of animal herbivores, and the majority of them use animal pollinators to achieve reproduction. Most of the routine work of a plant is carried out by roots and leaves, but it is the flowers that take on the crucial role of reproduction.

A flower is usually hermaphrodite, with both male and female roles. Hence it is essentially a structure that produces and dispenses the male gametophytes (pollen), organizes the receipt of incoming pollen from another plant onto its own receptive surfaces on the stigma, and then appropriately guides the pollen’s genetic material to the female ovules. The flower also protects the delicate male and female tissues (stamens and pistils) and has a role in controlling the balance between inbreeding and outbreeding, hence influencing the genetic structure and ultimately the evolutionary trajectory of the plant. But the plant itself is immobile, so that incoming pollen has to be borne on some motile carrier, sometimes wind or water but much more commonly on a visiting animal. To quote one source (Rothrock 1867), among plants, the nuptials cannot be celebrated without the intervention of a third party to act as a marriage priest! A pictorial overview of the stages is shown in figure 1.1, covering the processes of pollination that are the focus of this book.

A flower also serves to protect the pollen as it germinates and as the male nucleus locates the egg and then to protect the ovules as they are fertilized and begin their development into mature seeds. However, these later events (germination of the pollen and fertilization of the ovule) are technically not part of pollination, and they are covered here only as needed to understand the characteristics and effects of pollen transfer.

Since flowers bring about and control plant reproduction, they are central to much of what goes on in the terrestrial world, and pollination is a key mutualism between two kingdoms of organisms, perhaps the most basic type of exchange of sex for food; the plant gains reproductive success, and the animal—usually— gains a food reward as it visits the plant. But the visitor does not want to be a good pollinator and has to be manipulated by the plant to move on and to carry pollen to another plant. In practice, only about 1% of all pollen successfully reaches a stigma (Harder 2000).

Nevertheless, pollination by animals (biotic pollination) is both more common (Renner 1998) and usually more effective than alternative modes of abiotic pollen movements using wind or water, and animal pollination is usually also associated with more rapid speciation of plants (Dodd et al. 1999; K. Kay et al. 2006). Discussion of animal pollination therefore dominates this book, and around 90% of all flowering plants are animal pollinated (Linder 1998; Renner 1998). Furthermore, plants are, of course, the foundation of all food chains on the planet, and their efficient pollination by animals to generate further generations is vital to ensure food supplies for animals. Natural ecosystems therefore depend on pollinator diversity to maintain overall biodiversity. That dependence naturally extends to humans and their agricultural systems too; about one-third of all the food we eat relies directly on animal pollination of our food crops (and the carnivorous proportion of our diet has some further indirect dependence on animal pollination of forage crops). Pollination and factors that contribute to the maintenance of pollination services are vital components to take into account in terms of the future health of the planet and the food security and sustainability of the human populations it supports.

Figure 1.1 The central processes of pollination in a typical angiosperm flower, with the route taken by pollen from anther to stigma (followed by pollen tube growth into the style) in an animal-pollinated species. (Modified from Barth 1985.)

Beyond its practical significance, the flower-animal mutualism has been a focus of attention for naturalists and ecologists for at least two hundred years and provides almost ideal arenas for understanding some of the fundamental aspects of biology, from evolution and ecology to behavior and reproduction. It is perhaps more amenable than any other area to providing insights into the balance and interaction of ecological and evolutionary effects (Mitchell et al. 2009). Flowers are complex structures, and their complexity admirably reveals the actions, both historical and contemporary, of the selective agents (mainly, but not solely, the pollinators) that we know have shaped them. These factors make floral biology an ideal resource for understanding biological adaptation at all levels, in contrast with many other systems, where there are multiple and often uncertain selective agents.

In this first chapter, some of these central themes are introduced to set the scene for more specialist chapters; it should be apparent from the outset that while each chapter might stand alone for some purposes, it cannot be taken in isolation from this whole picture.

1. Which Animals Visit Flowers?

At least 130,000 species of animal, and probably up to 300,000, are regular flower visitors and potential pollinators (Buchmann and Nabhan 1996; Kearns et al. 1998). There are at least 25,000 species of bees in this total, all of them obligate flower visitors and often the most important pollinators in a given habitat.

There are currently about 260,000 species of angiosperms (P. Soltis and Soltis 2004; former higher estimates were confounded by many duplicated namings), and it has been traditional to link particular kinds of flowers to particular groups of pollinators. About 500 genera contain species that are bird pollinated, about 250 genera contain bat-pollinated species, and about 875 genera predominantly use abiotic pollination; the remainder contain mostly insect-pollinated species, with a very small number of oddities using other kinds of animals (Renner and Ricklefs 1995).

The patterns of animal flower visitors differ regionally. In central Europe, flower visitors over a hundred years ago were recorded as 47% hymenopterans (mainly bees), 26% flies, 15% beetles, and 10% butterflies and moths; only 2% were insects outside these four orders (Knuth 1898). But in tropical Central America, the frequencies would be very different, with bird and bat pollination entering the picture and fewer fly visitors, while in high-latitude habitats the vertebrate pollinators are absent and flies tend to be more dominant. Some of these patterns will be discussed in chapter 27.

2. Why Do Animals Visit Flowers?

The majority of flower visitors go there simply for food, feeding on sugary nectar and sometimes also on the pollen itself. Chapters 7 and 8 will therefore deal in detail with these commodities, and chapter 9 will cover a few more unusual foodstuffs and rewards that can be gathered from flowers; chapter 10 will take an economic view of all these food-related interactions, in terms of costs and benefits to each participant. Major themes in other chapters include the ways that flower feeders can improve their efficiency: learning recognition cues to select between flowers intra- and interspecifically, learning handling procedures, learning to avoid emptied flowers, and avoiding some of the hazards of competing with other visitors.

Flowers are also sometimes visited just as a convenient habitat, often simply because they offer an equable sheltered microclimate for a small animal to rest in, a place that is somewhat protected against bad weather, predators, or parasitoids. Or flowers may offer a reliable meeting site for mates or hosts or prey, or for females an oviposition site providing shelter for eggs and larvae. More rarely they are used as a warming-up site by insects in cold climates, usually because the flowers trap some incoming solar radiation, which enhances their own ovule development, but occasionally because a few flowers can achieve some metabolic thermogenesis that warms their own tissues (chapter 9 will provide more details on this topic).

3. How Do Flowers Encourage Animal Visitors?

Many plant attributes contribute to attraction of visitors: J. Thomson (1983) usefully groups these as plant presentation. Some of these attributes are readily apparent to visitors, and these may be features of individual flowers (e.g., color, shape, scent, reward availability, or time of flowering) or features of whole plants or groups of plants (e.g., flower density, flower number, flower height, or spatial pattern). The more readily apparent plant presentation traits can be divided into attractants (advertising signals), dealt with mainly in chapters 5 and 6, which discuss visual and olfactory signals, and rewards (usually foodstuffs), dealt with in chapters 7–9. Aspects of the timing and spacing of flowers, and how these might be affected by competition between different flowering plants, are given more in-depth treatments in chapters 21 and 22.

Other floral attributes are more cryptic to the visitor and may only determine the reproductive success of the plant in the longer term; these might include pollen amounts, ovule numbers, the genetic structure of the plant population, the presence and type of incompatibility system, etc.

It is generally in the plant’s interest to support and even improve its visitors’ efficiency, encouraging them to go to more flowers of the same species (so ensuring that only conspecific pollen is taken and received) and to go to flowers with fresh pollen available and/or with receptive stigmas for pollen to be deposited upon. Many flowers therefore add signals of status to their repertoire, via color change, odor change, or even shape change. Visitors are thereby directed away from flowers that are too young or too old or already pollinated. Instead they will tend to concentrate their efforts on those (fewer) flowers per plant that are most in need of visitation, thus also being encouraged to move around between separate plants more often and to ensure outcrossing rather than selfing. Reasons for favoring breeding by outcrossing (i.e., with other plants) are covered more fully in chapter 3.

4. What Makes a Visitor into a Good Pollinator?

In many ways this is the crucial theme running through this book. It relates to what is probably the major current debate in pollination ecology, that is, to what extent pollination it is a generalist process and to what extent it is a specialist one. Pollination has in the past nearly always been categorized in terms of syndromes, with particular groups of flowers recognized as having particular sets of characteristics (of color and scent, shape, timing, reward, etc.) that suit them to be visited by particular kinds of animals; and it is implicit in this approach that these suites have often been arrived at and selected for by convergent evolution in plant families that are unrelated. Thus most authors have used terms such as ornithophily to describe the bird pollination syndrome, or psychophily for the butterfly pollination syndrome. Flower characteristics would be listed that fit each syndrome, and an unfamiliar flower’s probable pollinators could therefore be predicted. Flowers in each category were seen as having a degree of specialization that suited them to their particular visitors, with some syndromes being more specialized than others. Nearly all earlier works on pollination were organized around this theme of syndromes, and it served as a useful structure for understanding animal-flower interactions for nearly two centuries. Without knowing this background it would be nearly impossible to follow the current debates that are a major focus for pollination ecologists, and it would also be very difficult to structure the information on flower attractants and flower rewards in chapters 5–9. This book thus retains a syndrome-based approach throughout its early chapters and explicitly considers the evidence in support of a syndrome approach in chapter 11; then it unashamedly covers each of the syndromes in turn in chapters 12–19, providing all the core materials on which later criticisms might be based.

The criticisms and debates focus around the reality of syndromes and how far they have been overplayed in the previous literature, perhaps blinkering or biasing our perceptions. Many authors now regard flower pollination as a much more generalized phenomenon, where most flowers get many different kinds of visitors and have not been and are not being heavily selected to specialize for the needs of one particular best visitor. This approach is specifically addressed in chapter 20. It will be a major argument there that the issue has been clouded by an as yet insufficient distinction between flower visitors and pollinators. So what does make a visitor into a good pollinator?

Physical Factors

Any animal that is to be an effective pollinator must have the ability to passively pick up pollen as its body moves past the anthers of a flower that it visits and carry that pollen to another flower. Normally this will be aided by the animal being a good physical fit in terms of size and shape, so that in alighting on the flower surface, or when inserting its tongue or beak of appropriate length, some specific part of its body touches the anther. Additionally pollen pickup and carriage will be aided if the animal has appropriate surface structures; pollen adheres well to feathers, fur, and hairy or scaly surfaces in insects but does not get transported well on shiny or waxy surfaces (and may even be damaged by certain surface secretions). Hence a small shiny beetle taking some nectar by crawling into the lower surface of a large tubular corolla where the anthers are in the corolla roof may well be a regular visitor to that flower but is unlikely to be an effective mover of its pollen; in effect, it is acting as a cheat as far as the flower is concerned and may be termed an illegitimate visitor.

Behavioral Factors

Different animals land on and forage at flowers in very different fashions. There are many aspects of behavior that will affect whether a given animal is going to be a good pollinator. Pollinators will seldom have a complete perception of all the aspects of plant presentation mentioned above, but they will respond to at least some of them in ways that are useful to the plant:

Their choices of places and times to visit, and exactly which flowers to visit, will be critical. Visits occurring before dehiscence (the splitting of the anthers to reveal the pollen) or after pollen depletion are normally of no value to the flower in ful-filling its male role, and visits before or after the stigma is receptive to incoming pollen are of no value to the flower in its female role.

Their handling of the flowers affect their pollen pickup and deposition characteristics; ideally they should receive pollen at a specific site on their bodies, and one that is also a good site for subsequent deposition of that pollen onto a stigma.

Their handling time per flower affects how many flowers are visited in a given time.

Their speed and directionality of movement between plants affect pollen dispersal.

They should not be too efficient at grooming off the pollen, or indeed at eating it.

Their flower constancy, that is, the likelihood that they will move to another flower of the same species, is perhaps most critical. If they innately or by learning prefer a particular flower phenotype, their high constancy will usually ensure that they move neatly and sequentially among conspecific flowers, not wasting pollen by depositing it in the wrong species. Constancy to a flower (considered in detail in chapter 11) gives economies to the visitor also; it may minimize travel distances, handling times, and learning effort and maximize pollen packing.

Behavioral factors such as these are often the key to being a good pollinator and of course are affected by the animals’ abilities to learn as they become more experienced as foragers. The ability to form a search image or to respond consistently to other cues, associating particular signals with the presence of food, hones their foraging ability and can cement their relationships with particular flowers. Hence later chapters of this book, in considering particular groups of animals that visit flowers, include careful consideration of their sensory and learning capacities.

Physiological Factors

Animals have differing physiological strategies and constraints, and these too can affect their energetic needs and thence their flower-visiting patterns, as will be discussed in chapter 10. Most animals (including nearly all invertebrates, and therefore many of the insect flower visitors) lack elaborate internal physiological regulatory systems, and their thermal balance and water balance are strongly influenced by environmental conditions. They cannot function if they are too hot or too cold or are short of water, and must forage in times and places that provide suitable microclimates, using the sun’s radiation to warm up by basking, or shady places to cool down again, and seeking (usually) sites that are relatively humid. However, birds and bats are physiologically more sophisticated and can regulate their body temperature and water balance more precisely; they generate heat internally through their own metabolism (endothermy) and regulate their own body fluids with efficient skin exchanges, respiratory controls, and excretory organs. They can in principle forage at almost any time and in any habitat, though they may still conserve their own energy by picking more equable sites.

The distinction does not lie exactly between the vulnerable invertebrates and the highly regulated birds and mammals, however. It is now clear that a rather small proportion of insects can also show endothermy, at least some of the time when they need to warm up in the absence of solar inputs (chapter 10); this applies especially to most bees, a few hoverflies, some large moths, and some beetles, occurring more sporadically in other groups. It is perhaps no coincidence that endothermic abilities in insects are most common in the flower-visiting groups, which have access to ready fuel supplies in the form of nectar but which may also need to compete for that nectar in the cool of early mornings or at dusk.

Given the list of factors that can turn a visitor into a good pollinator, two obvious points should emerge:

1. A great many of the animals that go to flowers for a short drink of nectar may be rather poor at pollinating that flower. Those with a poor physical fit, those that cheat, and those with little or no flower constancy are likely to be especially ineffective.

2. Of all the visitors, bees are likely to be especially good as pollinators. They rely solely on flowers for food, both as adults and as larvae, and so must visit more flowers than any other animals. Their sizes, hairy surfaces, and variably long tongues, their excellent learning abilities, communication systems and floral constancy, and their endothermic capacities all equip them to visit flowers efficiently (from their own perspective) and effectively (from the plant’s point of view). Although they are sometimes described as pollen wasters (because they, or rather their offspring back at the nest, eat so much of the pollen that they pick up), they are by far the most important pollinators in most ecosystems, and they do achieve high pollen export from visited flowers (Harder and Wilson 1997); plants have adapted over evolutionary time to make best use of them by providing more than enough pollen to cater for their needs and ensure that sufficient pollen still commonly reaches other flowers.

5. Costs, Benefits, and Conflicts in Animal Pollination

Plants with hermaphrodite flowers benefit greatly because a single animal visit can allow both pickup of pollen and its deposition on a stigma, fulfilling both the male and female functions of that flower at the same time. Animals benefit greatly by finding easily acquired foods, both sugary (nectar) and often also proteinaceous (pollen). Pollination by animals may therefore be a mutualism, of benefit to both participants, but it is not altruistic; for the animals, pollination of the flowers they forage at is almost always just an irrelevant by-product. In fact the plant and animal have a conflict of interest, often with adaptation and counteradaptation going on from both sides through evolutionary time to try and get a bigger share of the benefits. So the situations that we see now are the end products of the long and sometimes quite duplicitous associations of flowers with animals.

The plant ideally wants a visitor that is cheap to feed, alighting only briefly, moving on rapidly to another plant, and being faithful to its chosen plant species; so ideally, the forager should be chronically underfed and continuously on the move. But (again ideally) the animal would prefer to be as well fed and lazy as possible, getting as much food as it can from one flower with minimum energy expenditure, being relatively sedentary, and then moving on to any other nearby flower with copious nectar, whatever its species (although we already noted that some degree of fidelity may improve its foraging efficiency).

Hence, although there are obvious benefits to both partners, there are potentially clear costs as well. The plant has to invest in attractants (its carbon and nitrogen resources are used to make flamboyant petals, pigments, and chemical scents), as well as mere rewards. If the plant reduces its rewards too far, the animal may not get enough food and will give up on that species. The plant generally also has to compete with other plant species for pollination, to obtain a share of the good pollinators, so it cannot afford to skimp on its offerings too much if it is growing within a reasonably diverse plant community. Many plants also have to offset the additional costs of animal exploiters: those visitors who take rewards without pollinating (thereby cheating, chapter 24) and the flower eaters (florivores) or foliage herbivores also attracted by the pollination cues (chapters 25 and 26). For the animals, there may be costs linked to carrying the pollen they have inadvertently picked up (sometimes it is very unwieldy and can interfere with their wings or legs or sense organs), which may favor animals trying to cheat, and there may be costs also from the potential risks of predation or parasitism at flowers, since enemies can use them as a place to find prey or hosts reliably (chapter 24). There are also costs arising from the tendencies of the plants to cheat (chapter 23) by offering no real reward and sometimes by trapping the animal.

6. Why Is Pollination Worth Studying?

Pollination ecology can provide almost unparalleled insights into evolution, ecology, animal learning, and foraging behavior (fig. 1.2). It is perhaps the best of all areas to see and understand some basic biological processes and patterns; studies that deal exclusively with pollination biology have often had major impacts on general ecological and evolutionary theory.

Pollination interactions often show us evolution by natural selection in action almost before our eyes and provide some very clear-cut examples of adaptive radiation and, perhaps, of plant speciation. They are particularly useful for studying coadaptations (coevolution), because such interaction often involve relatively few organisms interacting with relatively high interdependence and incorporating the most fundamental of phenomena (reproduction for the plant, food supply for the animal). There are selection pressures on each side of the partnership, offering hopes that effects at the basic level of male and female gene flow can be quantified, sometimes (in crop pollination especially) on a time scale that can be detected within one scientist’s period of study.

In terms of ecology, the study of pollination sheds light on how different organisms interact and affect each other, especially the competitive effects of plants upon each other, and on the various levels of interactions of plants with pollinators, including resource allocation, competition, exploitation, and simply cheating. In the last two decades there has been an increasing stress on community-level interactions in pollination biology, now seen as an especially useful (because highly quantifiable) arena for more general work on community structures (J. Thomson 1983); so this book inevitably considers the community ecology of pollination, especially in later chapters.

Figure 1.2 Key interactions of major biological topics and themes promoting interest in the study of pollination.

Pollination biology also gives exceptional insight into the ecology of reproductive strategies and the complexities of sex and reproduction. Flowers usually are hermaphrodites, but they have many ways of organizing their sex life sequentially or spatially to maximize their reproductive output and fitness. This book contains rather little coverage of plant reproductive strategies beyond the basics, because the field has now become dominated by theory and modeling, and the topic has also been extensively and recently reviewed in other works (e.g., Harder and Barrett 2006).

In the realm of animal behavior some key influences can be especially easily measured and manipulated with flower visitors, and it is no accident that much of the key work on visual discrimination, learning behaviors, and above all optimal foraging has used pollinators, especially bees. Optimal diet theory can model how animals should behave in an environment offering different proportions of alternative prey as potential food items of differing value (also taking into account factors such as conspicuousness and different variances). The theory predicts that a range of outcomes from complete specialization on one kind of prey item to complete generalization on all possible items is to be expected, even from the same animal, as the prey parameters are varied. Substituting flower for prey item (and with the immense advantage of very easily quantified caloric rewards from nectar), it is not unexpected that pollinators similarly turn out to show almost the full range of possible foraging behaviors. Furthermore, they have proved ideal subjects with which to develop foraging models that can take into account different constraints on the foraging animals, whether from different physiological limits or from different cognitive skills. Learning ability is especially needed where resources are of intermediate predictability (Stephens 1993): that is, too unpredictable over one or a few generations for fixed behavior patterns to be favored, but not so greatly unpredictable that an individual cannot track the changes. This exactly applies to floral resources, so that we should expect flower visitors from any taxon to be good learners. Fortunately this is also readily tested with real or model flowers where just one trait at a time can be varied or associations of traits compared; again, social bees are ideal animals to work with, reliably emerging from their nests and traveling straight to the flowers provided, then displaying clear choices between alternative flowers.

For all these reasons, and with the added concern over human-induced effects on pollinator services in relation to biodiversity and to crops, interest in pollination ecology has burgeoned in the last ten to fifteen years, and the subject is attracting strong interest beyond the traditional academic centers of the developed world, giving us valuable insights into more varied habitats in Asia, Africa, and South America and into a greater diversity of interactions. Increasingly these systems are being modeled, and our preconceptions (of these and of other kinds of mutualisms) are being challenged. But the models are sometimes hampered by reliance on inadequate records, and understanding, of flower-visitor behaviors, and one of the most important issues for the immediate future is ensuring that the new generation of pollination ecologists understand the core subject material of floral biology and can measure and categorize pollination as distinct from mere visitation to feed into their models. We are in need of many and better quantitative studies of the effectiveness of visitors (for example, the average number of conspecific outcrossing pollen grains deposited on a stigma at an appropriate time by a given visitor in a single visit; chapter 11). Then we can properly understand plant and pollinator communities and pollination networks, and the effects of potential extinctions of flower visitors/pollinators on the communities of which they are a part. This book therefore hopes to provide in a single source a useful reference for all the aspects of floral biology and pollination interactions that need to be considered to give a real appreciation of these fascinating mutualisms.

Chapter 2

FLORAL DESIGN AND FUNCTION

Outline

Essential Flower Morphology

The Perianth

The Androecium: Male Structures

The Gynoecium: Female Structures

Flower and Inflorescence Features

Particular Floral Shapes

Flower Size and Size Range

Flower Sex and Flower Design

Overview: Essentials of Floral Design in Relation to Cross-Pollination

Flowers are essentially the containers for a plant’s sex organs, but they must perform several interrelated functions. Most obviously, and taking center stage here, they make and mature the gametes and then dispense the male gametes in such a way that they will be transported to another appropriate flower in the process of pollination, leading to fusion with female gametes. However, the flower must also be structured so that it protects the crucial sex organs through pollination and seed set, both from the environment (excessive rain or drought, freezing, heat load in full sun or during flash fires, physical damage in high winds) and from herbivorous animals (whether these be tiny sucking insects or large browsing and trampling vertebrates). The relative importance of the protective function in flowers varies enormously, and for many plants the issue of protection of the living tissues is perhaps less crucial than for most animals; plants are usually modular, with many sex organs in many flowers, so that they can afford to lose some flowers and regrow a new supply (whereas animals normally are not modular and have a very limited and nonrenewable supply of gonads). Investment in flower protection is therefore often quite limited, but it may become paramount for species with only intermittent flowering, or with just a few large and expensively constructed flowers, or with an extremely limited supply of pollinators; the various ways in which flowers can be protected are considered in more detail in chapter 24. Here we will concentrate primarily on the design features related to sexual function in pollination.

Flowers come in an astonishing variety, whether viewed across the entire spectrum of angiosperm families (appendix) or merely one family at a time. A family such as the Ranunculaceae has flowers ranging from simple bowls (e.g., Ranunculus species such as buttercups) to complex bilateral pendant tubular designs with elaborate internal nectaries accessible only to long-tongued bees (e.g., many Aconitum) and to tiny fluffy flowers where only the anthers are conspicuous (e.g., Thalictrum). Other families show similar variation but often produce the same three broad kinds of flowers, suggestive of convergent evolution.

Not only are individual flowers exceptionally varied in shape, size, and appearance, but the story is further complicated by the tendency of flowers in many families to mass together as inflorescences. Here the overall floral display is determined by the combined appearance of a cluster of small flowers (sometimes then termed florets) into one seemingly whole structure, which may in practice be the unit that is perceived, both by humans and by flower visitors, as the functional flower. We tend to refer to the complex multiple inflorescence of hydrangeas, or of many umbellifers, as their flowers; and we almost always speak of the floral displays in the composites (Asteraceae— daisies, asters, chrysanthemums, and their kin) as flowers, whereas in fact each is made up of many different florets, one kind making up the central disk and another kind forming the outer rays, which appear as petals.

TABLE 2.1

Basic Floral Anatomy

To consider flower design, then, we need to understand both the key components of an individual flower, to analyze their juxtaposition within one flower, and to review how the separate flowers can be assembled together, in many flower families, into complicated inflorescences that become functional units in terms of display and attraction.

1. Essential Flower Morphology

Flowers are complicated assemblages of plant parts, modified originally from leaves. Repetitive patterns of units occur in series, growing centripetally and sitting upon the flower stalk. Table 2.1 shows the general patterns of floral parts, with the four main series of structures from base to tip, or outer to inner: sepals, petals, stamens, carpels. This fundamental structure can be detected in most flowers from the basic arrangement in figure 2.1A, though the details vary enormously; these details can often be conveniently expressed in terms of a simple flower diagram of the type shown in figure 2.1B. In particular, the numbers of each part are highly varied: primitive magnoliids have very large numbers of petals, as do cacti, but in many dicot plants there are just five petals and five sepals, whereas in most monocots these structures come in threes. Even here many exceptions occur, and some important groups, such as the crucifers (Brassicaceae) and the poppies (Papaveraceae), have their parts in multiples of two or four. These petal and sepal numbers may or may not be reflected in the inner series of stamens and carpels: in the geranium family all the parts are in fives, and in the lily and iris families all are in threes, but in some other families this numerical consistency is lost, often with fewer carpels and more stamens than there are petals and sepals.

During flower development each part or series forms in sequence, and is in turn influenced by the initiation and growth of neighboring structures. The original apical meristem that will form the flower is protected in a bract, within which floral organs are initiated; then as floral morphogenesis progresses, the first series of structures, the sepals, grows and surrounds the bud, within which the petals and the androecium (male) and gynoecium (female) form sequentially. To add further variety, the fundamental structures may fuse into a ring as they arise or during ontogeny to form tubular organs. And sometimes parts of different floral organs may fuse, to form organ complexes. However, the basic underlying anatomy can still usually be described by a simple floral diagram as in figure 2.1B.

The primary functions of flowers are always performed by these basic and genuinely floral structures, but as we will see, some of the secondary functions (protecting, guiding, advertising, and rewarding) may be taken over by organs outside the flowers.

2. The Perianth

The outermost structures of a flower collectively form the perianth; this sits upon the receptacle, which is really just the expanded often cone-shaped end of the flower stalk, or pedicel. The perianth itself may be protected by extrafloral bracts, which in some cases are brightly colored and serve both protective and advertising functions.

Figure 2.1 (A) The basic structure of a flower, with both female (gynoecium) and male (androecium) reproductive parts (modified from Faegri and van der Pijl 1979 and later sources). (B) A classical flower diagram showing an idealized structure in transverse section and the relation of the four major whorls of floral parts.

The perianth is made up of rather thin and flat structures that enclose and usually protect the rest of the flower, and its tissues are sterile (nonreproductive). At anthesis (floral opening) the perianth commonly becomes a major part of the floral display, attracting the pollinators.

There may be either one or two series of perianth parts. Where there is only one series (or occasionally two but with no structural differentiation), the component parts are usually called tepals. However, in the majority of plants there are two series of distinctly different structures—the sepals and petals—forming the perianth (as shown in table 2.1).

Calyx and Sepals

The outer series of floral parts is mainly protective at all times in most plants and is termed the calyx, made up of sepals, often the same green color as the foliage but rather more robust. The calyx makes a cup (in which the rest of the flower sits), either by sepals overlapping each other or by their fusion (producing the condition called gamosepaly). Sepals can be formed from a thickened epidermis that is relatively solid and free of airspaces or can be toughened by sclerenchyma to give a semiwoody protective base to the flower; in extreme cases, such as Costus, the sepals form a lignified coating over the whole inflorescence through which small cohorts of individual corollas emerge (plate 34F, G). The external surface of a calyx may be further protected with glandular hairs (for example, in many Solanaceae) or with a waxy layer (for example, in eucalypts). Either of these additions may help to deter possible folivores, and some of the waxy calyces may also serve to reflect solar radiation, keeping the base of the corolla relatively cool.

There are also quite a number of plant genera in which the calyx sepals take on all or part of the role of advertisement and pollinator attraction, both visual and olfactory (with scents from the glandular hairs). Sometimes the sepals form the main ring of colored visual attractants (and tend to be referred to incorrectly as the petals); this is common in the Ranunculaceae, including Helleborus (where petals may become nectaries; chapter 8), and many Anemone and Pulsatilla species.

Alternatively just one or a few of the sepals are enlarged and highly colored, giving the main visual signal in some Rubiaceae and Verbenaceae. Something similar occurs in many euphorbs, such as Dalechampia (plate 11F), although there it is bracts rather than sepals that produce the show (the flowers are in fact compound inflorescences contained within the brightly colored bracts as a group of one female and several male tiny flowers). Or attractive sepals may act in concert with petals, whether being of the same color (e.g., the Salvia in plate 13G) or contrasting, as in some Abutilon (plate 14A) and some Clerodendrum, or enlarged, highly modified, and very differently colored (plate 13F). In yet other cases, the use of sepals as advertising structures occurs in only some flowers of a complex inflorescence; the central flowers are of usual type, whereas outer ones have one or more enlarged colored sepals, increasing the showiness of the whole floral display. Many Hydrangea species and hybrids are familiar temperate examples (plate 11E), as is coriander (plate 11D).

A further specialization occurs with the production of a protective fluid by internal sepal surfaces, producing a water calyx, or water jacket (also achievable by a cup-like calyx or leaf morphology that traps rainwater, as in many bromeliads). This moat of water around the flower base (plate 34D) may help to exclude ants and other small visitors with little role in pollination (e.g., Carlson and Harms 2007; and chapter 24) and may also have a cooling function. Such watery calyces are well known in various tropical plants, including many Bignoniaceae (they were first described in Spathodea from this family). Other genera produce viscous or mucilaginous material; for example, individual Commelina (Commelinaceae) flowers (plate 12A) emerge through a copious layer of thick mucilage held in the calyx, although the function of this substance is rather uncertain.

Finally, there are cases of rewards deriving from sepals; by far the commonest examples are extrafloral nectaries, where glandular areas of the sepals produce nectar that is predominantly collected by ants, in relation to their plant-guarding activities (chapter 25). Other examples are rare, although food bodies arising from sepals and fed on by beetles are not uncommon in the Winteraceae and in some Annonaceae.

Corolla and Petals

The inner series of the perianth structures is the corolla, which is made up of petals and takes on the familiar main display/advertisement function in most angiosperms. The petals tend to be arranged alternately with the sepals (fig. 2.1B and examples in fig. 2.15). As with sepals, the petals may be fused for all or part of their length (gamopetaly) to produce a tubular corolla. The color(s), size, shape, position, or orientation of petals may all contribute to the attractiveness of the display.

The petals are usually colored differently from the foliage and are often more delicate in texture, with a thin epidermis and a mesodermal layer that has plentiful air spaces. However, thickened or fleshy petals are not uncommon, and thick, white or cream-colored corollas are often found in heavily scented dusk- or night-flowering species that are visited by bats or moths using fragrance as an important cue. Petals may release scent through their surfaces or bear specific scent-producing trichomes, or papillae (chapter 6). In some cases, particularly among the Magnoliaceae and Liliaceae, they bear nectar-producing tissues, and occasionally they provide a surface for secondary pollen presentation (chapter 7).

Petals may also present complex microtextured surfaces, which provide easier landing platforms for visitors and may also play a role in close-up flower recognition (Kevan and Lane 1985). Textures involve ridges, overlapping plates, or patterns of pimples, often differing between the upper and lower petal surfaces, or between the upper petals and the lower landing-platform petals in bilateral flowers. For example, through direct tactile perception, bumblebees could discriminate among flowers of Antirrhinum mutants that lacked special conical epidermal cells on petal surfaces (Whitney, Chittka, et al. 2009).

Occasionally the petal tissues, or parts of them, combine with parts of the next inner ring of tissues (the stamens) to form a corona. This appears as an inner corolla in flowers such as daffodils, where it forms the protruding trumpet (plate 6E), and in passion flowers (plates 5A,B and 26G), where the corona is filamentous and becomes an elaborate inner circle of contrasting display.

Both series of perianth structures are usually wilted or shed after anthesis, and in some plants, pollination per se, or the fertilization of the ovules that follows pollination, may act as a trigger for the wilting or senescence of the corolla. However, sepals are sometimes retained to assist in fruit development (for example becoming the wings of the nuts in the large Southeast Asian dipterocarps) or to provide a contrasting background for fruit advertisement. Petals may wilt on the calyx or may be shed by a specific abscission mechanism involving ethylene (a common within-plant signaling system). But even petals are occasionally retained and thickened to form part of the fruiting structure.

3. The Androecium: Male Structures

The male functional organs of a flower, collectively termed the androecium, are a series of stamens. Each is formed from a thin filament, which may be erect, curved or pendant and which supports the terminal anther. The most common type of stamen has an anther with four pollen sacs (or locules), arranged in two pairs, each pair termed a theca. The thecae open at maturity in the process termed dehiscence, so exposing the pollen (dehiscence and its control will be covered in chapter 7).

Relative to other flower parts, stamens are rather constant in size and shape and design (see the review by Bernhardt 1996). In the phylogenetically basal family Magnoliaceae, the distinction between the anther and the filament is poor, but in nearly all other flowers, a two-lobed anther sits on the end of a much thinner and elongate filament. Stamens are rarely more than 20 mm long, although in some plants the filaments are fused basally and can then be strong enough to support lengths of up to 50 mm. This length is often achieved quite late in the floral development process, and elongation can be very fast; in some grasses the filament can grow several millimeters in just a few minutes as dehiscence approaches. On average, filaments are somewhat longer in grasses and other anemophilous (wind-pollinated) flowers than in zoophilous (animal-pollinated) species. However in many of the zoophilous flowers with fused petals, the filaments are inserted onto the inside of the resulting corolla tube (adnate filaments; e.g., fig. 2.2F–H) rather than arising basally, so they are technically short but still function as if they were elongate. Rarely, the filament is elaborated with fine hairs (easily seen in many Verbascum species) or with nodular thickenings (e.g., in Sparmannia), which may enhance their visual signaling effect (fig. 2.2K).

The filament is usually attached to the base of the anther, but sometimes anthers are suspended close to their midpoint by a thin and flexible junction; this is seen in many Liliaceae (fig. 2.2J; plate 12C), where the anthers hang down and move freely in seesaw fashion about this joint.

Within the anthers, the main design variation is in the location and form of the furrows that split apart at dehiscence (fig. 2.3). Most commonly the furrows (or thecal slits) open toward the center of the flower (introrse anthers), so that pollen is presented inward in the direction of the female organs; in flowers visited by larger animals, such as bats or birds, the introrse anthers may open broadly to produce a flat pollen-dispensing surface that can press against fur or feathers, as seen in many passionflowers (plate 5A,B). Externally or laterally opening anthers (extrorse or latrorse, respectively) are also found, but less commonly; they tend to be smaller and are borne on shorter filaments. Occasionally, there is only one furrow and the locules are united internally to open as a single chamber; Pavonia is a good example. At the opposite extreme, the locules may subdivide internally (either transversely or longitudinally) to give a polysporangiate anther. Sometimes dehiscence slits may lose the standard elongate pattern and become curved or almost circular on the tip of a bulbous anther head, giving a valvate stamen (fig. 2.2E).

In some plant groups, especially zoophilous taxa, stamens are modified by becoming united into a more complex structure (fig. 2.4), a phenomenon known as synandry. This is quite common in the large orders

Figure 2.2 Stamens with different filament and anther arrangements: (A–C) simple patterns with dehiscent slits on anthers; (D) elongate with elaborate dehiscence furrow; (E) anther rounded and valvate; (F-H) adnate filaments arising from petals, internal or protruding, in various Boraginaceae and caprifoliaceae; (I) dimorphic stamens set at two heights in Oxalis; (J) seesaw anther head on filament, in Lilium; (K) filament with nodes in Sparmannia; (L) anther with terminal pore in caesalpinoid Fabaceae. Not to scale.

Malvales and Fabales, and it may involve just the filaments or the more distal anthers as well, resulting in a tubular androecium (or even a solid cylindrical structure if there are no female organs in that flower). This greater rigidity in the central floral organs may provide a stronger area to grip or to maneuver around for insect flower visitors. A further possibility is for the anthers to unite at the tips but without the filaments doing so; this is often seen in composite flowers (see Composite Flowers below). It also occurs in many buzz-pollinated plants (chapters 7 and 18), such as Solanum or Cassia, giving a single pollen-dispensing structure that a visiting bee can grasp and vibrate its body against, and in these kinds of plants, it is commonly associated with loss of the anther’s dehiscence furrows and their replacement with one or more terminal or lateral pores through which pollen emerges (poricidal anthers). Single-pored anthers are also found in other situations: in some bellows flowers, such as Cyphomandra, that are squeezed by a bee to release the pollen, and in some of the carrion flowers that attract flies, where the anther is polysporangiate internally but where aggregated pollen emerges from the single pore in a slimy thread or droplet. One other specialization to be noted is the phenomenon of sensitivity in stamens: in a number of plant families the filament is somewhat contractile and may shorten or curve in one direction when stimulated by a visitor (chapter 7 and Compsite Flowers below).

Figure 2.3 Anther heads: (A) transverse section showing internal structure; (B) transverse section of typical tetrasporangiate anther with two locules, closed and then open at dehiscence; (C) longitudinal dehiscence (left) and valvate dehiscence (right) seen in frontal view, x–x marking the dehiscence furrows. (Redrawn from Endress 1994.)

As well as dispensing pollen, whether partly as a reward or solely for reproductive purposes, stamens may serve some other important roles. They are often used in advertisement and attraction, via visual or olfactory signalling. Where they are long and protruding, they markedly increase the apparent size of the flower or inflorescence and may result in a brush blossom (see Brush Blossoms below). Where the rest of the perianth is very small, the anthers then become the main visual signal. While anthers are commonly brown or dull in color when closed and take on the color of the pollen when dehiscing (usually yellow; but see chapter 7), the filaments are often white or almost colorless and so may provide a visual contrast to enhance visibility