BookPDF Available

Pheromones and animal behavior: Chemical signals and signatures, second edition



[A sample from the book is available as a free download from here. It contains the front matter, contents pages, units, preface, Chapter 1, full Index, and full References. Do share with colleagues and students] The book can be purchased from Amazon or your local book seller or from CUP Pheromones and other kinds of chemical communication underlie the behavior of all animals. Building on the strengths of the first edition, widely recognized as the leading text in the subject, this is a comprehensive overview of how pheromones work. Extensively revised and expanded to cover advances made over the last ten years, the book offers a thorough exploration of the evolutionary and behavioral contexts of chemical communication along with a detailed introduction to the molecular and neural basis of signal perception through olfaction. At a time of ever increasing specialization, Wyatt offers a unique synthesis, integrating examples across the animal kingdom. A final chapter critically considers human pheromones and the importance of olfaction to human biology. Its breadth of coverage and readability make the book an unrivaled resource for students and researchers in a range of fields from chemistry, genetics, genomics, molecular biology and neuroscience to ecology, evolution and behavior.
WYATT – 9780521130196 Cover C M Y K
WYATT Pheromones and Animal Behavior
Second edition
Pheromones and other
kinds of chemical
communication underlie
the behavior of all animals.
Building on the strengths
of the first edition, widely
recognized as the leading
text in the subject, this is a
comprehensive overview of
how pheromones work.
Extensively revised and
expanded to cover advances made over the last ten years, the book offers
a thorough exploration of the evolutionary and behavioral contexts
of chemical communication, along with a detailed introduction to the
molecular and neural basis of chemosensory perception. Its breadth of
coverage and readability make the book an unrivaled resource for students
and researchers in a range of fields from chemistry, genetics, genomics,
molecular biology, and neuroscience to ecology, evolution, and behavior.
“Wyatt takes us through a well-judged range of examples of some
amazing chemical communication strategies, from ants right up to
the scent of human attraction. The book is far-reaching, inspiring,
and brilliantly illustrated.”
Patrizia d’Ettorre, University of Paris 13, Sorbonne Paris Cité
“As Wyatt observes, we live in a chemical world, and this lucidly written
and intelligently illustrated book sets the nostrils aquiver as we catch
the scent of a new reality.”
David W. Macdonald, Wildlife Conservation Research Unit,
University of Oxford
“I cannot help thinking that I will cite this book in all research papers
that I will publish in the future.”
Kazushige Touhara, The University of Tokyo
“This thoroughly modern revision of the classic first edition is an
amazing journey through the landscape of pheromones… The book
is a must-read for any undergraduate or graduate student or working
scientist interested in a singular comprehensive resource on this
fascinating topic.”
Leslie B. Vosshall, HHMI-The Rockefeller University
researcher at Oxford University’s
Department of Zoology, and an
Emeritus Fellow of Kellogg
College, Oxford.
Cover illustration
(front): ring-tailed lemur (Lemur
catta), smelling a small bush for
traces of scent from other lemurs,
Madagascar. © Anup Shah/The
Image Bank/Getty Images;
(back): the male silkmoth (Bombyx
mori) has elaborate antennae
covered with sensillae highly
sensitive to the female sex
pheromone, bombykol.
© Walter Leal.
Pheromones and
Animal Behavior
Chemical Signals and Signatures
Second edition
Cover designed by Hart McLeod Ltd
This sample from the book
contains the front matter,
Chapter 1, Index, and full
You may share this freely
with colleagues and
students. (these and the
Appendix introducing
pheromone chemistry for
non-chemists are also
available free from
pheromones ). The book can
be ordered there or from etc.
Pheromones and Animal Behavior
Chemical Signals and Signatures
Pheromones and other kinds of chemical communication underlie the behavior of all animals.
Building on the strengths of the rst edition, widely recognized as the leading text in the
subject, this is a comprehensive overview of how pheromones work.
Extensively revised and expanded to cover advances made over the last ten years, the book
offers a thorough exploration of the evolutionary and behavioral contexts of chemical
communication, along with a detailed introduction to the molecular and neural basis of
chemosensory perception. At a time of ever increasing specialization, Wyatt offers a unique
synthesis, integrating examples across the animal kingdom. A nal chapter critically considers
human pheromones and the importance of olfaction to human biology. Its breadth of coverage
and readability make the book an unrivaled resource for students and researchers in a range of
elds from chemistry, genetics, genomics, molecular biology, and neuroscience to ecology,
evolution, and behavior.
A full list of the references from this book is available for download from www.cambridge.
Tristram D. Wyatt is a researcher at Oxford Universitys Department of Zoology, and an Emeritus
Fellow of Kellogg College, Oxford. He is interested in how pheromones evolve throughout the
animal kingdom, at both molecular and behavioral levels. These broad interests give him a
unique vantage point, enabling him to draw together developments across the subject.
This sample from the book contains the front matter, Chapter 1, Index, and full references.
You may share this freely with colleagues and students. (these and the Appendix introducing
pheromone chemistry for non-chemists are also available free from ). The book can be ordered there or from etc.
Pheromones and
Animal Behavior
Chemical Signals and Signatures
Department of Zoology and Kellogg College,
University of Oxford
University Printing House, Cambridge CB2 8BS, United Kingdom
Published in the United States of America by Cambridge University Press, New York
Cambridge University Press is part of the University of Cambridge.
It furthers the Universitys mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.
Information on this title:
©T.D. Wyatt 2014
First edition ©Cambridge University Press 2003
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2003
Second edition 2014
Printed in the United Kingdom by TJ International Ltd. Padstow Cornwall
A catalog record for this publication is available from the British Library
ISBN 978-0-521-11290-1 Hardback
ISBN 978-0-521-13019-6 Paperback
Additional resources for this publication at
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
To Robert
Preface page xi
Acknowledgments xv
List of SI prexes xvi
List of abbreviations xvii
1 Animals in a chemical world 1
1.1 Intra-specic semiochemicals: pheromones
and signature mixtures 2
1.2 Innatenessof pheromones 16
1.3 How pheromone signals evolve from chemical
cues 18
1.4 Pheromone diversity, specicity, and
speciation 24
1.5 Production of pheromones 31
1.6 Pheromones: signal honesty and costs 32
1.7 Chemical proles from which signature
mixtures are learned for individual and colony
recognition 37
1.8 Differences in response to pheromones 43
1.9 Releaser and primer effects of
pheromones 43
1.10 Multimodal signals 44
1.11 Allohormone pheromones bypassing olfaction
and taste 45
1.12 Pheromones and signature mixtures in
humans? 45
1.13 Pollution disrupts chemical communication in
aquatic organisms 45
Summary 46
Further reading 48
2 Methods for identifying and studying
semiochemicals 49
2.1 Bioassays 49
2.2 Collection and analysis of semiochemicals 55
2.3 Using genetic and other techniques from
molecular biology 59
Summary 63
Further reading 63
3 Pheromones, chemical cues, and sexual
selection 65
3.1 Which sex should advertise? 66
3.2 External fertilization and chemical duets 69
3.3 Scramble competition 69
3.4 Pre-copulatory mate guarding 71
3.5 Contests 72
3.6 Mate choice: overview 73
3.7 Mate choice for good genes, mate quality, and
direct benets 75
3.8 Mate choice for genetic compatibility revealed
by chemical cues 81
3.9 Coolidge effects and rejection of past mates:
been there, done that 86
3.10 Alternative mating strategies 86
3.11 Post-copulatory sexual selection 87
3.12 Sex pheromones and speciation 90
Summary 103
Further r eading 103
4 Coming together and keeping apart:
aggregation pheromones and host-marking
pheromones 105
4.1 Aggregation pheromones and Allee
effects 105
4.2 Host-marking pheromones 110
Summary 112
Further r eading 112
5 Territorial behavior and semiochemicals 113
5.1 Why scent mark on territories? 115
5.2 Scent-fence hypothesis 115
5.3 Scent-matching hypothesis 115
5.4 Border-maintenance hypothesis 120
5.5 Economics of scent-marking patterns in
territories 121
5.6 Dear enemies or nasty neighbors 123
5.7 Counter-marking and over-marking 123
5.8 Scent marking in non-territorial
mammals 124
Summary 125
Further reading 125
6 Semiochemicals and social organization 126
6.1 Colony, kin, family, and individual
recognition 126
6.2 Pheromones and reproduction in social
groups: control or co-operative
signaling? 133
Summary 148
Further reading 148
7 Pheromones and recruitment
communication 150
7.1 Foraging ecology and evolution of recruitment
communication 150
7.2 Social insects as self-organizing systems 160
Summary 164
Further reading 164
8 Fight or flight: alarm pheromones
and cues 165
8.1 Evolution of alarm signals by kin
selection 165
8.2 Subsocial insect families 165
8.3 Alert signals in deer family groups 165
8.4 Clonal sea anemones 167
8.5 Aphids 167
8.6 Social aphids 169
8.7 Termites and social Hymenoptera 169
Summary 172
Further reading 172
9 Perception and response to chemical
communication: from chemosensory receptors
to brains, behavior, and development 173
9.1 How olfaction works: combinatorial
processing of odorants including
pheromones 173
9.2 Evolution of chemoreceptors 188
9.3 The many chemosensory subsystems in
mammals and insects 192
9.4 The overlapping roles and integration of the
accessory and main olfactory systems in
mammals 198
9.5 Pheromones, sex, and brain circuits 201
9.6 Pheromones elicit stereotyped, but modulated,
behavior and/or physiological responses 206
9.7 Pheromone primer effects 209
9.8 Learning of signature mixtures 215
9.9 Interactions between signature mixtures and
pheromones 218
Summary 221
Further reading 222
10 Finding the source: pheromones
and orientation behavior 223
10.1 Investigating orientation behavior
mechanisms 223
10.2 Ranging behavior: search strategies for nding
odor plumes, trails, or gradients 226
10.3 Finding the source: orientation to
pheromones 227
Summary 243
Further reading 243
11 Breaking the code: illicit signalers
and receivers of semiochemicals 244
11.1 Eavesdropping 244
11.2 Chemical communication in mutualisms 249
11.3 Deception by aggressive chemical
mimicry 251
11.4 Social parasites using disguise to escape
detection by social insect hosts 255
Summary 259
Further reading 259
12 Using semiochemicals: applications
of pheromones 260
12.1 Semiochemicals used with benecial and
domestic animals 260
12.2 Pheromones in pest management 263
12.3 Pest resistance to pheromones? 272
12.4 Commercialization: problems and benets of
pheromones 272
Summary 273
Further reading 273
13 On the scent of human attraction: human
pheromones? 275
13.1 Olfactory cues to recognition: signature
mixtures 278
13.2 Choosing mates for genetic compatibility:
avoiding kin and going for optimum
difference? 279
13.3 What molecules do humans give off? 284
13.4 Our sense of smell: perception of odors 291
13.5 Human pheromones? 295
13.6 Where next with human pheromones? 301
Summary:To smell is human 302
Further r eading 303
Appendix An introduction to chemical terms for non-
chemists 304
References 312
List of credits 377
Index 378
This book is designed to bring together people already working on chemical communica-
tion and to encourage others, especially chemists (who have a vital role in this research), to
take up the challenge. My aim has been to make an evolutionary understanding of
chemical communication, including pheromones, accessible to a broad scientic and lay
Pheromone research brings together scientists with many different areas of expertise,
from a rich diversity of chemists to biologists of many kinds. Each area of expertise has its
own jargon and concepts a behavioral ecologist speaks a different language from a
neuroscientist. The book recognizes that every scientist is a novice outside their own
subject, even science close to their own, so I try to explain ideas in terms understandable by
non-specialists while at the same time aiming to be up to date and detailed enough for the
specialist. I also wanted to write a book that could be enjoyed by the majority of the worlds
scientists whose rst language is not English and thus also clearer for everyone.
Pheromones offer exceptional opportunities to study fundamental biological problems.
The rapid progress of the last decades comes from the convergence of powerful techniques
from different areas of science including chemistry and animal behavior, combined with
new techniques in genomics and molecular biology. These allow us to investigate ques-
tions at every level: molecular, neurobiological, hormonal, behavioral, ecological, and
evolutionary. The discoveries from molecular biologists have greatly expanded our
knowledge of the evolutionary biology of olfactory communication. Equally, molecular
biology only makes sense in the context of evolution.
I wrote the rst edition to provide the overview of chemical communication we were
then missing, covering the whole animal kingdom, integrating approaches from ecology to
neurobiology, and all with an evolutionary perspective. I have kept the same overall
structure for the book in the new edition. As before, the book is organized around themes
such as sex, speciation, and social organization, rather than taxonomically. The book also
covers the perception and processing of chemosensory information. In each topic I have
aimed to integrate examples from across the animal kingdom. In the same paragraph you
may read about nematodes, moths, snakes, and mice. I explore the often convergent ways
evolved by different kinds of animals to solve the same communication needs.
All chapters have been comprehensively updated and most chapters have been com-
pletely rewritten. The changes are perhaps most signicant, as you might expect, in those
parts involving molecular biology, especially in the chapter on perception of pheromones.
Recent results include the surprising discovery that insect chemoreceptors have evolved
independently of vertebrate ones. However, there has also been much new to discuss in
evolution and ecology, including results coming from the application of molecular tech-
niques as well as detailed eld work.
Different parts of the book emphasize examples from different taxa. As in the rst
edition, mammals feature more strongly than invertebrates in the sections on individual
variation and hormonal effects of pheromones for example, but invertebrates dominate the
sections covering mechanisms of searching behavior.
Chapter 1 denes pheromones and looks at evolution of pheromones as signals. I raise a
pragmatic distinction between pheromones and the chemistry of individual or colony
odors. I also look at the role pheromones play in speciation. The importance of both
common ancestry and convergence in molding chemical signals is a key theme.
Chapter 2 is about the development of analytical tools and how these are changing the
study of chemical communication, allowing us to identify types of molecules previously
hard to work with. New genomics techniques can be used to identify genes involved in
both production and perception of molecules and not just in model organisms such as
Drosophila. On the behavioral side I emphasize the importance of proper randomization of
treatments and blindingof experimenters wherever possible. Progress will depend on
productive partnerships between chemists and biologists.
The following six chapters cover different aspects of pheromones in the ecology and
behavior of animals. Chapter 3 is about the evolution of pheromones in sexual selection,
drawing out the many parallels between animals in a wide range of taxa. Among the new
material featured in the chapter is work on Drosophila and moths as well as developments
in evolutionary theory.
Chapter 4 covers Allee effects and the roles pheromones have in spacing organisms,
bringing them together, and keeping them apart.
Chapter 5 reviews territorial behavior, largely in terrestrial vertebrates. The discovery of
the male mouse pheromone, darcin, offers fascinating insights into female mouse behavior
(Roberts et al. 2010, 2012). Darcin prompts her to learn the individual odor of the territorial
male and where the scent mark is.
The parallels between complex social behaviors mediated by chemical communication
in social insects and social mammals are explored in Chapter 6. The queen pheromones of
an increasing number of social insects are being identied. It seems, however, that
mammals do not use pheromones to suppressreproduction by subordinate members of
the group.
Recruitment in social insects for foraging and nest building and for defense are covered
by Chapters 7 and 8 respectively. One major change in our understanding is a clearer
distinction between alarm pheromones and cues. The molecules involved in sh alarm are
likely to be cues rather than pheromones.
Chapter 9 explores how olfaction works and how the olfactory receptors themselves
evolve, in enormous variety. Vertebrates and invertebrates are similar in the way they
Preface to the second edition
detect and process chemical cues, by combining inputs from neurons carrying different
olfactory receptors, but they achieve this with quite different receptor families, which
evolved independently.
The mechanisms that animals have evolved to nd an odor source are discussed in
Chapter 10. We understand more about the mechanisms that sh and birds use than when
the rst edition was written. There are some interesting uses of genetically manipulated
Drosophila larvae to explore the ways they orientate in chemical gradients.
Broadcast signals can be eavesdropped. Chapter 11 covers a world of deception and
spying, including new players and a clearer understanding of selection in some classic
Chapter 12 discusses how an understanding of chemical communication can be used for
agriculture and to control disease vectors. Whereas insects formed the main examples of
pheromone control, pheromones are showing promise for the possible control of vertebrate
pest species, notably the sea lamprey.
Chapter 13 covers the roles of chemical communication in human beings. I discuss the
smells we produce and the ones we can perceive. One of the most surprising things that has
emerged from genomics studies on humans is the enormous variation between us as
individuals in what we can smell. Our olfactory receptor repertoires are individually quite
different: it is likely that we each experience unique olfactory worlds. I conclude the
chapter by exploring some of the limitations of current research on human pheromones
and how we could take it forward.
Finally, the appendix explains the common chemical terminology you will come across.
While some of the molecules important in chemical communication are shown in the
gures, there are too many mentioned in the text to illustrate them all. Instead, you can see
them on sites such as, which allows you to search by common name
and shows synonyms as well as the systematic names. Many pheromone molecules,
together with some background, are included on Pherobase
(El-Sayed 2013).
Choice of literature for the second edition
This book necessarily offers a selective distillation of an enormous literature. I have
attempted to reect our consensus understanding of each topic. For reviews, I have
generally used the most recent I could (though I reference earlier reviews if they continue
to be inuential). The papers cited have been chosen to reect both their contributions to
the subject but also because they offer good entry points to the literature (do use
Google Scholar
or Web of Science
to nd papers citing these leads). Sometimes you
will ndareviewandaparticularexperimental paper both referenced, for example
(Cardé & Haynes 2004; Liénard et al. 2010), which will be obvious, I hope, when you look
Preface to the second edition
them up. The references for this edition are also available at
Wherever possible, I have chosen sources that you will be more likely to be able to nd.
Where I have had a choice between equally good papers I have gone for the one in an open
access journal or one that the authors have made available on the web, for example on their
own website. It may be worth searching on an article title to see if it is available. If an
article is not available and you do not have institutional access to the journal, you might
courteously write to the author to see if they have a PDF to send. Most people are pleased to
be asked I know I am.
Have you considered helping edit Wikipedias entries in our subject? It might seem
surprising for a textbook to recommend its readers to consider contributing their expertise
to Wikipedia, the worlds largest online encyclopedia, but this is where the greatest
inuence for our subject will be. As Wikipedia is where most people look rst, Bateman
and Logan (2010) encourage scientists to seize the opportunity to make sure that Wikipedia
articles are understandable, scientically accurate, well sourced, and up to date. Bond
(2011) makes such a call to his fellow ornithologists and presents many advantages of
getting involved. Pheromones and some aspects of chemical communication are briey
covered in Wikipedia but not to the depth and range of many other areas of science. You
might be able to improve this. Logan et al. (2010) give tips for getting started and guidance
on good practice.
If you would like Microsoft PowerPoint
slides of the illustrations in the book for
teaching or talks, do email me,, letting me know which
chaptersgures you would like.
Preface to the second edition
I would particularly like to thank the following for generously reading the whole book in
draft: Bruce Schulte, Jagan Srinivasan, Joan Wyatt, and Vivian Wyatt. I am also grateful to
many other friends and colleagues for help with various chapters and recent writing
projects, which helped me develop ideas explored in the book, including Olle Anderbrant,
Richard Benton, Thomas Breithaupt, Patrizia dEttorre, Monica De Facci, Dick Doty,
Heather Eisthen, Maud Ferrari, Jean-François Ferveur, Kevin Foster, Tom Getty, Stephen
Goodwin, Alan Grafen, Christina Grozinger, Penny Hawken, Matthieu Keller, Jae Kwak,
Jean-Marc Lassance, Darren Logan, Jocelyn Millar, Dan Rittschof, Benoist Schaal, Peter
Sorensen, Számadó Szabolcs, Robert Taylor, Kevin Theis, Martin Thiel,Tobias Uller, Marc
Weissburg, Tom Wenseelers, Danielle Whittaker, Brian Wisenden, and Ben Wyatt.
Any remaining errors are mine of course, and I would welcome comments and sugges-
tions for corrections. You can contact me at
I would like to thank all the scientists in addition to those listed above who advised me
on their areas of expertise and kindly sent reprints and pre-prints of their work. The book
would not have been possible without their help and generosity. In keeping the range of
animal groups represented as wide as possible, I have had to be selective. Inevitably I have
not been able to include many examples that I would have liked to. I apologise to authors
whose research I was not able to describe here despite its high quality.
Many colleagues generously helped me with high-resolution copies of their illustrations.
I would like to give additional thanks to colleagues who produced new or especially
adapted gures for me, including Christina Grozinger, Harland Patch, Troy Shirangi, Jagan
Srinivasan, and John Terschak.
It is a pleasure to thank Martin Grifths, Megan Waddington, Abigail Jones, Kath
Pilgrem, Vania Cunha, and other colleagues at Cambridge University Press for their
encouragement and assistance at all stages of producing the second edition.
I would like to thank the publishers and societies listed at the end of the book for
permission to reproduce gures and tables, particularly those which did not charge fees.
Factor Name Symbol
centi c
milli m
micro µ
nano n
pico p
femto f
2MB2 2-methyl-but-2-enal
AOB accessory olfactory bulb
AOS accessory olfactory system
BNST bed nucleus of the stria terminalis
cAMP cyclic adenosine monophosphate
cGMP cyclic guanosine monophosphate
CHC cuticular hydrocarbon
CNG cyclic nucleotide-gated channel
CNV copy number variant
cVA cis-vaccenyl acetate
ESP1 exocrine gland-secreting peptide 1
GC gas chromatography
FPR formyl peptide receptor
GABA -aminobutyric acid
GPCR G-protein-coupled receptor
GR gustatory receptor (invertebrates)
GSN gustatory sensory neuron (insects)
GUR gustatory receptor (Caenorhabditis
HPLC high-performance liquid chromatography
iGluR ionotropic glutamate receptor
IR ionotropic receptor
JH juvenile hormone
MHC major histocompatibility complex
MGC macroglomerular complex
MOB main olfactory bulb or OB
MOE main olfactory epithelium
MOS main olfactory system
MOT medial olfactory tract
MTMT (methylthio)methanethiol
MUP major urinary protein
MPOA medial pre-optic hypothalamus
OB olfactory bulb (MOB)
OR olfactory receptor
ORCO olfactory receptor coreceptor (insects)
OSN olfactory sensory neuron (also termed
olfactory receptor neuron, ORN)
SEM scanning electron micrograph
SEM standard error of the mean
SNP single nucleotide polymorphism
SPME solid phase micro extraction
T1R taste receptor type 1
T2R taste receptor type 2
TAAR trace amine-associated receptor
TRC taste receptor cell
TRPC2 transient receptor potential channel
2 (= TRP2)
V1R vomeronasal receptor type 1
V2R vomeronasal receptor type 2
VNO vomeronasal organ
VNS vomeronasal system (= accessory olfactory
1Animals in a chemical world
When two dogs meet and sniff, they gain a wealth of
information from each others smells. Each dog will dis-
cover the sex, maturity, and hormonal state of the other;
some of these smells will be species-wide dog phero-
mone signals. Each dog also detects the individual smell
of the other, which it learns as a signature mixtureto
remember in case they meet again.
When two ants meet and sweep antennae over each
other, they have an olfactory exchange of information
similar to that of the dogs, discovering age, sex, ovarian
stage (reproductive or not), and caste (worker, soldier,
queen), all signals from species-wide pheromones. They
also detect the colony odor of the other ant, enabling
them to decide by the signature mixturewhether the
other ant is a nestmate or not.
All animals produce a chemical prole, present on the
body surface, released as volatile molecules, and from
scent marks that they deposit (by dogs on lamp-posts for
example) (Figures 1.1, 1.2, 13.2). As chemical senses are
ancient and widespread, shared by all organisms
including bacteria, animals are pre-adapted to detect
chemical information in the environment (Box 1.1).
Across the animal kingdom, animals of all kinds gain
chemosensory information from other organisms.
Chemical senses are used to locate potential food sour-
ces and detect predators. Chemical senses also mediate
the social interactions that form the focus of this book,
as illustrated by the dogs and ants above. We can prob-
ably say that more organisms use chemosensory com-
munication than any other mode.
A chemical involved in the chemical interaction
between organisms is called a semiochemical (Box 1.2).
Some of the semiochemicals emitted by animals are
pheromones, evolved as signals for communication.
Other semiochemicals, such as the carbon dioxide in
exhaled breath, did not evolve as a signal, but can be
exploited as a cue by blood-sucking mosquitoes as a way
of nding a host. Some of the othermolecules emitted by
animals, such as odors due to infections, may also be
cues. The distinction between signals and cues is explored
further in Section 1.3.
Pheromones and signature mixtures are semiochemi-
cals used within a species. Semiochemicals acting
between individuals from different species are called
allelochemicals and are further divided depending on
the costs and benets to signaler and receiver (Box 1.2)
(Chapter 11) (Nordlund & Lewis 1976; Wyatt
2011). Pheromone signals can be eavesdropped (over-
heard) by unintended recipients: for example, specialist
predatory beetles use the pheromones of their bark bee-
tle prey to locate them. The predators are using the bark
beetle pheromones as kairomones. Animals of one
species can emit fake, counterfeit signals that benet
themselves at the cost of the receiving species. Chemical
signals used in such deceit or propaganda are termed
allomones: for example, bolas spiders synthesize partic-
ular moth pheromones to lure male moths of those spe-
cies. Semiochemicals beneting both signaler and
receiver in mutualisms, such as those between sea ane-
mones and anemone clownsh, are termed synomones.
The multiplicity of terms is only useful as shorthand and
the terms are clearly overlapping, not mutually exclusive
(for example, a molecule used as a pheromone within a
species can be used as a kairomone by its predator).
My aim in this book is to focus on patterns across the
animal kingdom. I have tried to include examples from as
many animal taxa as space allows, but for more detail see
the suggestions in further reading and references in the
text. This chapter introduces the ways in which animals
use semiochemicals and many of the topics are explored
at greater length in later chapters (see Preface for over-
view and rationale).
1.1 Intra-specific semiochemicals:
pheromones and signature mixtures
Modern pheromone research could be said to date from
1959, when the chemist Adolf Butenandt and his team
identied the rst pheromone, the silk moths sex
pheromone bombykol, which prompted the coining of
the word pheromone,from the Greek pherein,to
transfer; hormo¯n, to excite (Butenandt et al. 1959;
Karlson & Lüscher 1959). Butenandts discovery
established that chemical signals between animals
exist and can be identied (Chapter 2). From the start,
Karlson and Lüscher (1959) anticipated pheromones
would be used by every kind of animal, from insects
and crustaceans to sh and mammals. Since then,
pheromones have been found across the animal king-
dom, in every habitat on land and underwater, carry-
ing messages between courting lobsters, alarmed
aphids, suckling rabbit pups, mound-building ter-
mites, and trail-following ants (Wyatt 2009). They are
also used by algae, yeast, ciliates, and bacteria. It is
likely that the majority of species across the animal
kingdom use them for communication of various
kinds. Much is known about the pheromones of
Chemical profile
Made up from many sources, e.g.
small molecule size large
by GC
• secretions
• immune system
• hormones
• bacterial symbionts
• diet
• other conspecifics
• collected from flowers
• infections
Pheromone 1
Pheromone 2
Pheromone 3
Signature mixture A
Signature mixture B
Figure 1.1 Pheromones occur in a background of molecules
that make up the chemical prole consisting of all the mol-
ecules extractable from an individual. The chemical prole
(top) is an imaginary trace from an imaginary column capa-
ble of analyzing all the molecules (at one side is high-
performance liquid chromatography (HPLC) with large pro-
teins, at the other is gas chromatography (GC) with small
volatile molecules). Each peak represents at least one
Much of the chemical prole is highly variable from indi-
vidual to individual. The sources of the molecules in the
chemical prole include the animal itself as well as its envi-
ronment, food, bacteria, and other individuals etc. It is this
complex background that makes identifying pheromones so
challenging in many organisms.
The pheromones could include sex pheromones or ones
related to life stage or caste. The pheromones would be the
same in all individuals of the same type in a species (domi-
nant male, worker ant, forager, etc.); that is, they are anony-
mous, common across the species. As examples, I have
included some possible kinds of pheromones that are
known from organisms (not necessarily in the same species):
aspecic combination of large and small molecules
(Pheromone 1), a combination of small molecules
Figure 1.1 (cont.)
(Pheromone 2), or a particular large molecule by itself
such as a peptide (Pheromone 3).
The signature mixtures (A and B) are subsets of variable
molecules from the chemical prole that are learned as a tem-
plate for distinguishing individuals or colonies. Different
receivers might learn different signature mixtures of the same
individual. For example, a male might learn a different signa-
ture mixture of his mate than the one her offspring might learn.
Hypothetically it is conceivable that the male might learn dif-
texts, say immune-system associated molecules in one context
and more diet inuenced molecules in another. In other words,
Adapted from Wyatt (2010). The layout is inspired by
Figure 1 of Schaal (2009).
Animals in a chemical world
insects, sh, and mammals, but some other taxa have
not been well studied. For example, crabs and other
Crustacea make extensive use of pheromones but
relatively few of these have been chemically identied
(Breithaupt & Thiel 2011). Birds, too, have now been
shown to have a rich olfactory life though we are only
mated alpha
FID (mV)
older sterile
10 20 30 min
35–38 46–48
(c) 9
Figure 1.2 The queenlessant, Dinoponera quadriceps, lives in small groups headed by an alpha female, the only egg-laying
individual in the colony. The hierarchy is maintained by physical aggression. This can include gaster rubbing (a) in which the
alpha female rubs the antenna of the subordinate on the cuticular hydrocarbons, which include the alphas pheromone badge
of dominance, 9-hentriacontene (c, top). This molecule is characteristic of alpha females in all colonies of the species.
(b) If a subordinate female becomes reproductive and starts to produce the molecules characteristic of an alpha female, other
ants in the colony detect this and immobilize her (an example of an honest signal maintained by punishment, Section 1.6).
(c) The colony prole of ants in the colony can be shown in a solid phase micro extraction (SPME) gas chromatographic
analysis of their cuticular hydrocarbons (Monnin et al. 1998) (Chapter 2). As well as the many-peaked hydrocarbon chemical
prole shared by the other ants in the colony, the alpha female also has the additional peak #40 (indicated by the asterisk) which
is the pheromone 9-hentriacontene. Below, her fellow colony members have the same colony prole as her but lack this peak.
(d) Non-destructive SPME sampling allowed changes in the percentage of 9-hentriacontene in the cuticular hydrocarbons of
an individual ant to be followed in the days after she became the alpha female. In a larger sample of ants undergoing the
transition, the signicant difference was between the quantities at 15 and 30 days.
(a) and (b) from Monnin and Peeters (1999), (c) chromatograph from Monnin et al. (1998), (d) from Peeters et al. (1999).
1.1 Intra-specific semiochemicals
Box 1.1 Chemical and other senses compared
Chemical senses are shared by all organisms including bacteria. However, while the general way
that molecules interact with chemosensory receptor proteins in a lock and keymanner is
shared, the chemosensory receptor proteins are highly variable across the animal kingdom and
even within animal taxa. This is because the chemosensory system, like the immune system,
tracks a changing world of molecules generated by other organisms. Over evolutionary time, the
chemosensory systems of organisms co-opt, test, and discard chemosensory receptor genes and
neural coding strategies, leading to great divergences in receptors (Bargmann 2006b;
Bendesky & Bargmann 2011). Chemosensory receptor genes turn over rapidly, in a birth-and-
death process of gene duplication and loss (see Chapter 9). The rapid evolution of chemosensory
receptor proteins, evolved independently in insects and vertebrates, made chemoreception much
harder to investigate than vision (Chapter 9). The key proteins (opsins) for light-detection in eyes
do vary considerably and insect and vertebrate opsins have diverged. However, unlike chemo-
sensory receptor proteins, they form a large monophyletic group within the G-protein-coupled
receptor (GPCR) superfamily (Porter et al. 2012).
At the level of the individual, variation in olfaction is much greater than in the opsin genes.
For humans, mutations in the four genes for opsin receptor proteins sensitive to different
wavelengths of light give us a small number of different kinds of color vision deciency or
color blindness.By contrast, we have more than 400 olfactory receptor genes, each of which
can be mutated, so each of us smells a unique world (Chapter 13) (Olender et al. 2012). For this
reason too, we might each remember different mixtures of molecules as signature mixtures to
recognize the odors of other people.
The chemical senses of olfaction and taste are very different from vision and hearing, which
detect the energy of different wavelengths in the form of light and sound: chemical senses rely on
the physical movement of molecules from the signaler to the sense organ of the receiving animal.
This requires either diffusion, only likely to be important for small organisms at the scale of
millimeters, or ow of currents (Chapter 10). Either way, the time taken for molecules to travel to
the receiver means that chemical signals are rarely instantaneous in the way that visual and
acoustic signals can be.
Challenges remain for studying chemical communication (Chapter 2). We can record and play
back the sound signals of an animal easily enough, but we do not have devices to do the same for
chemical signals. Each molecule needs to be correctly synthesized, in every detail (see
Section 1.4.3 and Appendix), before it can be played backto the animal. This can be
challenging for a team of biologists and makes chemist partners invaluable. For example,
methyl-branched alkanes, important components of ant CHCs, are not commercially available
and synthesizing these is a costly and time-consuming process (van Zweden & dEttorre 2010).
Yet, perhaps more than other modalities such as sound or vision, chemosensory systems are
amenable to molecular manipulation: in model systems we can now study communication at the
Animals in a chemical world
just beginning to discover what molecules their pher-
omones might be (Campagna et al. 2012; Caro &
Balthazart 2010; Hagelin & Jones 2007; Zhang et al.
2010). Research on human semiochemicals is at a
similarly early stage; I review our current state of
knowledge in Chapter 13.
The idea of chemical communication was not new in
1959. The ancient Greeks knew that the secretions of a
female dog attracted males. Charles Butler
(1623) warned in The Feminine Monarchie that if a
beekeeper accidentally crushes a honeybee, the bees
presently nding it by the ranke smell of the poison-
ous humor, will be so angry, that he shall have work
enough to defend himself.In The Descent of Man, and
Selection in Relation to Sex (1871), Charles Darwin
included chemical signals alongside visual and audi-
tory signals as outcomes of sexual selection, describ-
ing the strong smells of breeding males in moths,
pythons, crocodiles, musk ducks, goats, and elephants.
Jean-Henri Fabre (1911), also writing in the 1870s,
described how male great peacock moths, Saturnia
pyri,ocked around a female moth hidden behind
wire-gauze, but ignored visible females sealed under
glass. A female moths smell could be collected on a
cloth and males would ock to that too. Many other
scientists in the nineteenth century and rst half of the
twentieth century, including Niko Tinbergen, had
worked on phenomena we would recognize as being
mediated by pheromones (some are mentioned in
Karlson & Lüscher 1959). However, because the
quantities emitted by an individual animal were so
small, the chemistry of the day could not identify
them, until the inspired idea of using domesticated silk
moths, which could be reared in the hundreds of
thousands necessary to collect enough material for
analysis using the techniques available at that time
(Chapter 2).
The enormous variety of organic molecules identi-
ed as pheromones since the rst, bombykol, in 1959
is as diverse as the animal kingdom, and offers an
ongoing challenge for chemists interested in the
identication, synthesis, and exploration of natural
functions of novel compounds (Cummins & Bowie
2012; El-Sayed 2013; Francke & Schulz 2010). The
likely explanation for the diversity of pheromone
chemistry is that these signals have evolved from
chemical cues naturally released by organisms, facili-
tated by the broad tuning of olfactory receptors
(Chapter 9) (Section 1.3).
Invertebrates and vertebrates, in a wide range of
habitats, use chemical communication in similar ways.
Animals as different as moths and elephants may share
the same molecule(s) as part of their pheromones.
However, there are more fundamental parallels in
sensory processes, even if we are not always sure
whether this has occurred by convergence or via
shared ancestors. The parallels include the combina-
torial way that the sense of smell is organized in the
brain: olfactory sensory neurons with the same olfac-
tory receptor all collect at the same spot (glomerulus)
in the brain; the information from different glomeruli
is combined to identify the molecule (the combinato-
rial mechanism) (Chapter 9).
1.1.1 Pheromones
Pheromones are molecules that have evolved as a
signal between organisms of the same species. The
Box 1.1 (cont.)
level of the genes involved in signal production (e.g., enzyme pathways) and signal reception
(genetics of receptors, brain, and behavior) especially in model animals such as Caenorhabditis
elegans, moths, Drosophila, and the mouse.
1.1 Intra-specific semiochemicals
signal elicits a specic reaction, for example, a
stereotyped behavior (releaser effect) and/or a
developmental process (primer effect) from a conspe-
cic (member of the same species) (Box 1.2)
(Section 1.9) (Wyatt 2010). Many, probably most,
pheromones (including the sex pheromones of most
moths and some mammal pheromones) are not single
compounds, but rather a species-specic combination
of molecules in a precise ratio. This combination is the
pheromone (though sometimes called a multicompo-
nent pheromone or pheromone blend). A pheromone
can elicit a variety of effects, depending on the context
and the receiver (Section 1.8). Responses to phero-
mones usually seem to be innate (though this is not a
part of the denition). In the few instances where
learning is rst required for a pheromone to act, all
animals normally learn the same molecule(s), which is
what denes it as a pheromone (Section 1.2).
Box 1.2 Definitions of chemical mediators
Pheromones are signals. The other categories of semiochemicals in this box are cues that can be
used for information but did not evolve for that function (Section 1.3). Adapted from Wyatt
(2010, 2011) based on Nordlund and Lewis (1976).
See Wyatt (2011) for a discussion of the origins and usage of these terms. I discuss inter-
specic interactions mediated by allelochemicals in Chapter 11. Infochemicalas an alter-
native to semiochemicalwas proposed by Dicke and Sabelis (1988) though its main change
was to replace produced or acquired bywith pertinent to biology ofin each case for
A. Hormone: a chemical agent, produced by tissue or endocrine glands, that controls various
physiological processes within an organism. (Nordlund & Lewis 1976).
B. Semiochemical: a chemical involved in the chemical interaction between organisms.
(Nordlund & Lewis 1976) (from the Greek: semeion, mark or signal).
1. Pheromone: molecules that are evolved signals, in dened ratios in the case of multiple
component pheromones, which are emitted by an individual and received by a second
individual of the same species, in which they cause a specic reaction, for example, a
stereotyped behavior or a developmental process. (Wyatt 2010, modied after Karlson and
Lüscher 1959). (From the Greek: pherein, to carry or transfer, and hormo¯n, to excite or
2. Signature mixture: a variable chemical mixture (a subset of the molecules in an animals
chemical prole) learned by other conspecics and used to recognize an animal as an
individual (e.g., lobsters, mice) or as a member of a particular social group such as a family,
clan, or colony (e.g., ants, bees, badgers). (Wyatt 2010; derived from Johnstonsmosaic
signalsensu 2003, 2005; Hölldobler and Carlins, 1987 ideas; and Wyatts, 2005 sig-
nature odor).
3. Allelochemical: chemical signicant to organisms of a species different from their source,
for reasons other than food as such. (Nordlund & Lewis 1976).
Animals in a chemical world
Karlson and Lüscher (1959) predicted that most
pheromones would act via the conventional senses of
olfaction or taste, but that some pheromones might be
ingested and act directly on the brain or other tissues.
We would call these allohormone pheromones
(Section 1.11). They speculated that royal jelly in hon-
eybees might contain such a pheromone, and indeed an
active molecule (royalactin) has been identied, which
causes larvae receiving it to develop into queens rather
than workers (Chapter 9) (Kamakura 2011).
Pheromones include the familiar sex attractant
pheromones, and numerous others that serve a wide
variety of functions. Some pheromones are specicto
different life stages or castes. One key feature of
pheromones is that they are anonymous,that is, a
given pheromone is the same in all individuals within a
species of the same type (e.g., male or female) or
physiological state, and it conveys a stereotyped mes-
sage that is independent of the individual producing it
(Hölldobler & Carlin 1987).
However, quantities of pheromone can differ
between individuals or in the same individual over
time. Some male mouse pheromones, the farnesenes,
are produced only by dominant male territory holders,
not subordinates (Hurst & Beynon 2004). In the ant
Dinoponera quadriceps, when an ant becomes the top
(alpha) female, she starts to produce the standard
chemical badge of a top femalein her species, 9-
hentriacontene (Figure 1.2) (Peeters et al. 1999).
However, in the male mouse and the top female ant of
these examples, the pheromones are still anonymous
(Hölldobler & Carlin 1987; Hölldobler & Wilson 2009,
p. 270). They indicate the presence of, for example, a
dominant male mouse or an alpha female ant, not a
particular individual.
Some of our expectations of pheromones have been
heavily inuenced by the well studied response of
male moths to the sex attractant pheromones of con-
specic females. For example, the antennae of male
moths have thousands of highly specialized receptors
Box 1.2 (cont.)
species-wide signals
learned by
receiver from
highly variable
chemical profile of
Between members of
same species
Between members of
different species
benefit emitter, of
a different species
benefit receiver, of
a different species
benefit both emitter &
receiver, of different
Diagram showing the relationships between different kinds of semiochemicals. Inspired by Box 7.1 in de Brito-
Sanchez et al. (2008) and other sources.
1.1 Intra-specific semiochemicals
for the pheromone and specicareasofthebrain
dedicated to processing the pheromonal signal.
However, other pheromone processing in insects may
involve less specic receptors, without dedicated
brain areas (glomeruli) (see Chapter 9). Thus, we now
know that narrowly tuned and highly specialized
receptors and dedicated glomeruli are not a prereq-
uisite for pheromone use. For example, honeybee
alarm pheromone components seem to be processed
by receptors and glomeruli that also process other,
non-pheromone molecules (Chapter 9) (Wang et al.
Similarly, male mothsenormously enlarged anten-
nae, covered with thousands of olfactory sensilla that
are tuned specically to the pheromone, reect selec-
tion for extreme sensitivity to low concentrations of
female pheromone, necessitated by the scramble com-
petition to be the rst to reach the female (Chapters 3, 9,
and 10). Based on the great body of work on male
moths, we tend to expect all receivers of pheromones to
be very sensitive to them and to respond at great dis-
tances. However, other animals may not use attractant
pheromones at all, although they may still use sex-
specic contact pheromones for sex and species recog-
nition when in close proximity to each other (for
example the contact sex pheromones used by some
copepods; Snell 2011b). The stimulus concentration on
contact can be high and thus exquisite sensitivity in the
olfactory or gustatory receptors that perceive contact
pheromones is unnecessary. A small number of speci-
alized chemosensory neurons may be sufcient. This
seems to be the case for short range species recognition
mediated by contact chemicals during the courtship of
Drosophila males and females (Chapters 3 and 9).
When the original denition of pheromone was
proposed in 1959, only a single pheromone had been
chemically identied: bombykol of the silk moth
female (Karlson & Lüscher 1959). It is a tribute to
Karlson and Lüscher, and their wide consultation, that
the denition has held up so well (Wyatt 2009). It is not
surprising that the denition has needed to be updated
slightly since then (Box 1.2) (Wyatt 2010). (See Box 1.3
and Box 1.4 for why words matter and how distin-
guishing the concepts can be helpful).
Box 1.3 Pheromones and signature mixtures: why words matter
Denitionsmatter because they can provide useful generalizations and predictions. My purpose in
separating pheromones from signature mixtures is pragmatic and based on the heuristic (rule of
thumb) value of separating these kinds of chemical information. When we say something is a
pheromone,the reader can anticipate that it isa molecule (or a particular combination and ratio of
molecules for a multicomponent pheromone) that will be found, for example, in all sexually
mature females. Quantities of the pheromone may differ between individuals, and this may be
important in mate choice (Chapter 3), but not in ways that allow an individual female to be
recognized as an individual. In Hölldobler and Carlins (1987) terms, the pheromone signal is
anonymous,it could beany female (see also Hölldobler & Wilson 2009, p. 270). (See also Box 1.4
Operational denition of pheromone.)
In contrast, if a phenomenon, such as a male distinguishing his mate from other females, relies
on a learned signature mixture, it would be fruitless to search for a single combination of
molecules eliciting individual mate recognition across the species: it is precisely the great
Animals in a chemical world
Box 1.3 (cont.)
differences between femaleschemical proles that makes learning signature mixtures by males
In the rst edition of this book, I included signature mixtures within the denition of
pheromones(Wyatt 2003, pp. 24). I now think it is more helpful to explicitly separate
signature mixtures as it is emerging that their characteristics are different, in particular the
variability of signature mixtures and the need for learning (Tables 1.1 and 1.2) (Wyatt 2010). It
seems to be a useful distinction, which has helped understand phenomena best explained by
species-specic pheromone molecules appearing on a background of variable chemical proles
from which signature mixtures are learned, in situations as varied as the male effect in sheep
(Hawken & Martin 2012) and trail pheromones in stingless bees (Reichle et al. 2013).
So, to be clear, not all molecules included in this book are pheromones. I will discuss many
molecules that are not pheromones (Section 1.3), including the highly variable signature
mixtures used to avoid mating with kin (Chapter 3) and learned by ants to distinguish nestmates
from non-nestmates (Chapter 6), as well as chemical cues such as barnacle settlement cues
(Chapter 4) and sh alarm cues (Chapter 8).
Box 1.4 Operational definition of pheromone
The formal denition of a pheromone includes both evolved emission and reception of the signal
for that function (Section 1.3) (Table 1.1) (Maynard Smith & Harper 2003, p. 3). However, for
many otherwise respectable pheromones, we do not know enough about the ways in which
production and/or reception may have evolved. So, I propose we formalize an operational
denition of pheromone, which most people already use in practice, as fully identied mole-
cule(s), the same across a species, in all lactating mature females for example, which when
synthesized elicit the same characteristic response in the conspecic receiver as the natural
To legitimately assert that a molecule or specic combination of molecules qualies as a
pheromone for a species (or in a genetically dened subpopulation within a species):
1. The synthesized molecule/combination of molecules (combination) should elicit the same
response as the natural stimulus in the bioassay.
2. It should act in this way at realistic concentrations similar to the natural stimulus.
1.1 Intra-specific semiochemicals
1.1.2 Signature mixtures
Returning to the dogs and ants that opened this chap-
ter, the individually distinctive mixture of molecules
that allows dogs to tell each other apart by smell and
allows ants, at a colony level, to distinguish nestmate
from non-nestmate, are not pheromones and were
not included in the original denition.
We need a different term for the molecules that
animals learn and use to distinguish other individuals
or colonies. I have proposed signature mixture
(Wyatt (2010) inspired by Johnstons (2003, 2005)
mosaic signal,Hölldobler and Carlins (1987) ideas,
and based on Wyatts (2005) signature odor). I think
some of the early doubts about mammal pheromones
(Box 1.5) came from treating signature mixtures as if
they were pheromones. Be aware when reading the
past and current literature that the term pheromone
is still used ambiguously and may be used in contexts
where signature mixtureor chemosensory cues
would be more accurate or helpful.
Signature mixtures are the subsets of variable mol-
ecules from the chemical prole of an individual
(Figure 1.1) that are learned as templates by members
of the same species (conspecics) and used to recog-
nize an organism as an individual or as a member of a
particular social group such as a family, clan, or colony
Box 1.4 (cont.)
3. For multicomponent pheromones, experiments should demonstrate that all compounds in
the combination are necessary and sufcient.
4. Only this molecule or the proposed combination of molecules elicits the effect (and other
similar molecules or combinations that the animal would encounter do not).
5. There should be a credible pathway for the pheromone signal to have evolved by direct or
kin selection.
6. Quantities may vary between individuals (e.g., subordinate and dominant males).
The requirements follow those explored in Chapter 2. They are the equivalent of Kochs
postulatesfor establishing causal relationships for pheromones: initial demonstration of an
effect mediated by a pheromone, then identication and synthesis of the bioactive molecule(s),
followed by bioassay conrmation of activity of the synthesized molecules. It can be equally
important to show that other similar molecules do not have the effect of the proposed
How the response develops (ontogeny) in an individual is a separate question (Section 1.2).
Normally we do not know the details. Fish alarm substances are thought to be cues rather than
pheromones (Chapter 8) as they fail to satisfy criterion #5.
Sadly, the experimental literature on humans, and other mammals, includes many unidenti-
ed extracts or molecules that have never been rigorously demonstrated to be biologically active
by the full bioassay evidence and synthesis process. It is misleading to call them even putative
pheromones(Chapter 13).
Animals in a chemical world
Box 1.5 Mammal pheromones
There is good evidence that mammals have pheromones that t well with the original denition
(Brennan & Zufall 2006; Wyatt 2010), despite doubts from some authors (Beauchamp et al. 1976;
Doty 2010). The many small-molecule mammal pheromones include the rabbit mammary
pheromone 2-methylbut-2-enal (2MB2) (Schaal et al. 2003) (see below), male mouse pheromones
such as (methylthio)methanethiol (MTMT) (Lin et al. 2005), trimethylamine (Li et al. 2013),
dehydro-exo-brevicomin and 2-sec-butyl-4,5-dihydrothiazole (Novotny 2003; Novotny et al.
1985), and Asian elephant pheromones including frontalin (1,5-dimethyl-6,8-dioxabicyclo
[3.2.1]octane) and (Z)-7-dodecen-1-yl acetate (Rasmussen et al. 2003). Many of these molecules
(or ones similar) are also used as pheromones by insects (Section 1.4.1) (Table 1.2). As well as
small-molecule pheromones, mammals also have large-molecule pheromones, such as, in mice,
exocrine gland-secreting peptide 1 (ESP1) (Haga et al. 2010) and the protein pheromone darcin
(Chapter 9) (Roberts et al. 2010, 2012). See Figure 9.7 for some of the large and small molecules
used by mice.
Part of the reason for the earlier doubts was confusion between pheromones and the highly
variable chemical proles of mammals (for example, see Figure 13.2). In addition, over the years,
expectations about pheromones in mammals have built up, perhaps based on misconceptions
about insect pheromones (e.g., Doty 2010): contrary to these expectations, as shown in a variety
of animals including insects, pheromones do not have to be unique species-specic molecules
(Section 1.4.1), animalsresponses to pheromones can vary (Section 1.8), and pheromones can
involve elements of learning (Section 1.2).
A separate set of problems came from some scientists in the 1990s onwards, using mice as
a model system to study mammal pheromones, who seemed to assume, despite earlier
evidence to the contrary, that (a) pheromones would be exclusively detected by the VNO
system, and that (b) all molecules detected by the VNO were pheromones (Baxi et al. 2006;
Wyatt 2009). As explored in Chapter 9, it is now conrmed that (i) pheromones are detected
by both the VNO and the main olfactory system, depending on species and pheromone, (ii)
that the VNO also responds to other odorants, and (iii) that there is extensive integration of
inputs from the two olfactory systems.
An example of a small-molecule pheromone perceived by the main olfactory system is the
rabbit mammary pheromone, 2MB2, which stimulates rabbit pups to suckle (Charra et al.
2012; Schaal et al. 2003). The pups respond to pheromone from their mothers nipple region,
which elicits stereotyped searching, usually successful in just six seconds. The pheromone
also prompts the pups to learn their mothers signature mixture (Coureaud et al. 2010). Newly
born humans use olfactory stimuli, possibly including a pheromone, to nd their mothers
nipple (Chapter 13) (Doucet et al. 2009, 2012).
1.1 Intra-specific semiochemicals
Box 1.5 (cont.)
1,000 1,200 1,400
Retention index
1,600 1,800
Responding pups (%) FID signal (a.u.)
Butanoic acid
(d) 100
Responding pups (%)
Hexan-2-one (40)
Heptan-2-one (15)
Decan-2-one (15)
Dec-2-enal (15)
Undec-2-enal (15)
Butanoic acid (15)
Butyrolactone (20)
Decanal (15)
Benzaldehyde (35)
D,L-Limonene (15)
Cyclopentanone (15)
Cyclohexanone (15)
Oct-1-en-3-one (15)
Pyridine (20)
Octanal (15)
2-Methylbut-2-enal (40)
2-Methylbutan-1-ol (40)
2-Ethylhexan-1-ol (15)
2-Methylpropan-1-ol (40)
-Methylcyclopentan-1-one (15)
Butan-1-ol (40)
Animals in a chemical world
(Box 1.2). Signatureis used as it denotes
The signature mixture is the mixture of molecules
(and likely, their relative ratios) that are learned. The
template is the neural representation of the signature
mixture stored in the memory of the learner (after van
Zweden & dEttorre 2010).
There are two distinguishing characteristics of
signature mixtures: rst, a requirement for learning
and, second, the variability of the cues learned,
allowing other individuals to be distinguished by
their different chemical proles (see Section 1.7 for a
more detailed explanation). Other comparisons
between pheromones and signature mixtures are
explored in Table 1.1.
The signature mixture molecules in the chemical
prole, learned by receivers as the template for
recognition, can be produced by the organism itself,
acquired from the diet, shared local environment, or
other organisms (Section 1.7). In ants, the chemical
prole may have species-characteristic types of
molecules, but each colony produces different com-
binations and ratios of these (van Zweden &
dEttorre 2010). For example, different colonies of
the ant Formica exsecta have different colony-
specic combinations of various (Z)-9-alkenes,
under genetic inuence (Martin & Drijfhout 2009b;
Martin et al. 2008c). What makes this different from
a pheromone is that each colony has a different set
of ratios of these shared molecules it is what
allows the colonies to be distinguished. By contrast
a multicomponent pheromone would be expected to
have a uniform ratio across a population, the same
in each colony.
Different receivers might learn different combina-
tions of molecules from an individuals prole as the
Box 1.5 (cont.)
(Figure facing). The discovery of the rabbit mammary pheromone, 2-methylbut-2-enal (2MB2), used a
linked gas chromatograph-olfactory (GCO) assay, which allowed concurrent detection by neonatal
rabbits and by a ame ionization detector (FID) (Chapter 2). (a) Photographs show the sequence (duration
5 s) of a two-day-old pupssearchinggrasping response directed to the glass funnel of the GC sniff-port.
(b) Typical chromatogram of rabbit milk efuvium (upper panel) and concurrent percentage of pups
responding with searchinggrasping responses (lower panel; inverted scale). The regions of the chromato-
gram eliciting more than 20% of responses (summed across 25 GCO runs) and the compounds eluting in
these regions are shown. (c) Screening, by presenting a glass-rod, of milk volatiles presumed to have
behavioral activity: two-day-old pup at rest (left) and exhibiting grasping (right) to a glass rod carrying
2MB2. (d) Frequency of searching (open bars) and seizing (solid bars) directed to the glass rod carrying one
of the 21 compounds identied in milk. Numbers in parentheses indicate the numbers of pups tested.
(Schaal et al. 2003).
The only signicant mammal group for which chemical communication has not been dem-
onstrated are whales and dolphins (cetaceans). However, it is possible and may be discovered in
future, as baleen (mysticete) whales have a good olfactory system, which they may also use to
detect upwind concentrations of plankton by smell (as albatross do, Chapter 10) (Thewissen et al.
2011). Some baleen whales have a Harderian gland, which in some rodents produces phero-
mones, though it may have other functions (Funasaka et al. 2010).
1.1 Intra-specific semiochemicals
Table 1.1 Contrasting pheromones with signature mixtures. Hölldobler and Carlin (1987) introduced the idea of
anonymous signals (pheromones) contrasted with variable signature mixtures (though their terminology was
different) (see also Hölldobler & Wilson 2009, p. 270). The anonymous pheromone signals are uniform
throughout a category (e.g., species, male, female, and perhaps molt state and dominance status). In contrast,
signature mixtures vary between individuals or colonies and can be used to recognize the organism as an
individual or member of a particular social group such as a family or colony. From Wyatt (2010).
Pheromone Signature mixture
Stimulus A species-wide molecule (or particular
dened combination of molecules).
A combination of molecules, never a sin-
gle molecule. Combination of molecules
varies between individuals or colonies.
Possible receiver side effect: there may
not be one signature mixture for each
individual, as different conspecics
(receivers) may learn different subsets of
molecules in the individuals chemical
prole (Figure 1.2).
Type of
Anonymous (independent of the source
Variable (allows recognition of an indi-
vidual or group such as a colony).
Molecule size Any size or type, depending on habitat,
medium, signal duration, and phylogeny.
Any size or type, depending on habitat,
medium, signal duration, and phylogeny.
Source Make self or acquire/modify. Usually
genetically based.
Make self or acquire/modify. Use chemical
mixtures, genetically based or from the
environment or a combination.
Learning Little requirement for learning of the signal
Innate, stereotyped, or hardwired (with the
caveat of developmental constraints).
Cues learned.
Response Elicits a stereotyped behavior and/or
physiological response. May be context
Learned and can be used to distinguish
individuals or groups (can lead to stereo-
typed response e.g., aggression). May be
context dependent.
Olfactory receptor
Some (e.g., moth sex pheromones) have
high specicity olfactory receptor proteins
(and the labeled linesand dedicated
glomerulithat result). Many other phero-
mones do not.
Low specicity, broadly tuned receptors.
Animals in a chemical world
signature mixture to recognize that individual (see
legend to Figure 1.1). In other words, signature mix-
tures seem to be a receiver sidephenomenon, exist-
ing as a templatein the nervous system of the
receiver. Even if all receivers perceived the world in the
same way, they could each still learn different subsets
of molecules from the chemical prole as the signature
mixture of an individual. A further complication
comes from the way that each receiver smells a slightly
different world, because one of the characteristics of
olfaction is the variability of olfactory receptors
between individuals each of us smells a different
world (Box 1.1) (Chapters 9 and 13). For this reason
too, the learned signature mixtures could differ
between receivers.
With perfect knowledge, one could know the whole
chemical prole of an animal, which molecules from
this prole are learned as the signature mixture by the
receiver, and how the signature mixture is represented
as a template in the nervous system of the receiver. An
outline of what can already be achieved, treating the
system as a black box,is shown by experiments with
the ant Formica japonica, which showed that the
nestmate labelcould be reproduced with synthetic
hydrocarbons matching the colony cuticular hydro-
carbons (CHCs) (Akino et al. 2004).
1.1.3 Which sensory systems, olfaction or
gustation, are used to detect pheromones and
signature mixtures?
In both invertebrates and vertebrates, gustatory (taste)
receptors come from different families of receptors
from olfactory receptors, and link to the brain in dif-
ferent, simpler, ways from the olfactory pathways (in
mammals olfaction includes both the main olfactory
system and vomeronasal organ-accessory olfactory
system; see Chapter 9 for more explanation). Most
pheromones seem to be processed by olfaction.
However, a small but signicant proportion of pher-
omones in invertebrates are processed by taste (gus-
tation) (Chapter 9). These include the response of a
male Drosophila melanogasters gustatory receptors
on its front leg to CHC pheromones important in sex
and species recognition (Chapters 3 and 9). Some
allohormone pheromones in both vertebrates and
invertebrates may act directly on the brain or other
organs (see Section 1.11).
Table 1.1 (cont.)
Pheromone Signature mixture
Processing Mostly combinatorial across glomeruli. Combinatorial across glomeruli.
Detection system:
olfaction or taste
or act directly.
Mostly by glomerularly organized olfac-
tory system(s).
A minority of pheromones by other
chemosensory routes e.g., taste (gestation).
Allohormone pheromones act directly on
tissues or nervous system.
Glomerular olfactory system(s).
In vertebrates with
a vomeronasal
system (VNS).
Detection by the VNS or main olfactory
system or both, depending on pheromone
and species.
Detection by the VNS or main olfactory
system or both, depending on species.
1.1 Intra-specific semiochemicals
All signature mixtures are likely to be processed by
the combinatorial processes of olfaction rather than
taste (gustation), in part because discrimination
learning is likely to be involved (see Box 9.1).
1.2 Innatenessof pheromones
Generally speaking, pheromones do not require
learning: they seem to be innate,”“hardwired,pre-
disposed, or work out of the box(Table 1.2).
However, being innate is not part of the original
pheromone denition (Karlson & Lüscher 1959) or its
updated version (Box 1.2) (see also Sections 1.8 and
9.6) (Wyatt 2010). Instead, the dening point for
pheromones is that they are species-wide rather than
that they are innate (for example, it would be possible
for a pheromone to rely on early learning so long as
typically all individuals imprint on the same molecule
(s) in normal circumstances).
The idea of innatebehavior is itself a long-debated
question in animal behavior. Seemingly innate
behaviors often have developmental and environ-
mental requirements for full expression (Bateson &
Mameli 2007; Mameli & Bateson 2011). Part of the
problem is that the term innatecovers many differ-
ent phenomena (Mameli & Bateson 2011). In practice,
trying to separate the contributions of nature (genes)
and nurture (environment) to the development of a
behavior is like asking whether the area of a rectangle
is due more to its length or width (ascribed to psy-
chologist Donald Hebb, in Meaney 2001).
Geneenvironment interactions on behavior are
explored by Bendesky and Bargmann (2011).
Just as a mammals visual cortex does not form
correctly if the eyes do not receive visual stimuli
during critical periods after birth (Hensch 2004),
olfactory stimuli in amniotic uid before birth can
inuence olfactory bulb development (e.g., Todrank
et al. 2011). Normal responses to pheromones may
not develop unless species-specic conditions are
met. These usually occur as a matter of course in
normal development. Experiments that dissectthe
developmental process can expose the normally hid-
den mechanism(s) by which a response develops. For
example, perhaps surprisingly, early imprinting on
species-specic odors can be important in recogni-
tion of a mate of the correct species when adult
(Chapters 3 and 9) (Doty 2010, p. 39 ff.; Owens et al.
1999). Normally, as the parents are of its own species,
this leads to appropriate courtship choices, but cross-
fostering experiments can demonstrate that these
olfactory choices are learned in some species such as
pygmy mice, house mice, sheep, and deer (Doty 2010,
p. 39 ff.). Cross-fostered young are attracted to the
species of their foster parents.
In the wild, this learning can be important in
sexual selection and speciation in some species
(Verzijden et al. 2012). Cross fostering showed that
early olfactory imprinting by young sh (learning at
a sensitive period; Chapter 9) normally contributes
to sexual isolation in two stickleback species by
inuencing adult mate choices (Kozak et al. 2011).
When adult, learning may also be involved: male
mammals such as rats and mice may need sexual
experience before they can distinguish estrous from
diestrous female odors (Chapter 9) (Swaney &
Keverne 2011).
In some cases, developmental effects have been
shown to act at the periphery of the sensory system: for
example, the behavioral response of young worker
bees to queen mandibular pheromone depends on
exposure to the pheromone soon after pupal emer-
gence, via an effect on dopamine receptor gene
expression in the olfactory sensory neurons (Vergoz
et al. 2009).
Pheromones themselves can prompt learning. While
the response of rabbit pups to the mammary phero-
mone 2-methylbut-2-enal seems hardwired (Box 1.5),
the pheromone stimulates learning of other odors,
which will then also stimulate suckling (Coureaud
et al. 2010; Schaal et al. 2009). Contact with the male
mouse protein pheromone, darcin, prompts a female
mouse to learn both his volatile individual signature
mixture and the location of the scent mark (Chapter 9)
(Roberts et al. 2010, 2012).
Animals in a chemical world
Table 1.2 Biochemical convergence of pheromones among ants, bees, moths and termites, and other animals
including mammals. In some cases, the same or related com pound is used for similar functions in different
species. More commonly, the arbitrary nature of signals is revealed by different uses for same or similar
compound. See other chapters for more details of the functions of these pheromones. After Blum (1982), with
additional information from Kelly (1996), Novotny (2003), Mori (2007) and Breithaupt and Hardege (2012). See
Appendix for notation.
Compound Function Occurrence Animal Genus
Benzaldehyde Trail pheromone
Male sex
Bee, Apidae
Ant, Formicidae
2-Tridecanone Alarm pheromone
Ant, Formicidae
Female sex pheromone,
male sex pheromone
Male and female
Nereid worm
Mouse, Mus
Male sex
Mouse, Mus
Bark beetle, Dendroctonus
Female sex pheromone Mammal
Female Asian elephant
Elephas maximus & 140 species
of moth (as one component
of a multicomponent
Sex pheromone
Bark beetles
Male Asian elephant E. maximus
(both +/enantiomers)
1.2 Innatenessof pheromones
1.3 How pheromone signals evolve from
chemical cues
The ubiquity and chemical diversity of pheromones
can be explained by natural selection and are the
evolutionary consequences of the powerful and exi-
ble way the olfactory system is organized (Chapter 9);
gustation (taste) does not have such exibility. This
may explain why most pheromones are detected by the
olfactory system (in terrestrial vertebrates with both, it
includes the main and accessory olfactory systems).
The olfactory systems of most species have a large
range of relatively non-specicbroadly tuned
olfactory receptors (Chapter 9). This means that almost
any chemical cue in the rich chemical world of animals
will stimulate some olfactory receptors and can
potentially evolve into a pheromone signal.
Pheromones are evolved signals. Signals alter the
behavior of other organisms, have evolved because of
that effect, and work because the receivers response
has also evolved (Maynard Smith & Harper 2003, p. 3;
Seeley 1995, p. 248).
If the signal alters the behavior of the receiver it
must, on average, pay the receiver to respond in this
way, otherwise receivers would evolve not to respond
(Maynard Smith & Harper 2003, p. 3). Signaling is
synonymous with communication as narrowly
dened by Ruxton and Schaefer (2011, p. 2583) in a
helpful discussion of recent debates about animal
communication and information.
In contrast, a cue is any feature of the world, ani-
mate or inanimate, that can be used by an animal as a
guide to future action, but has not evolved for this
purpose (Maynard Smith & Harper 2003, p. 3). For
cues, only the receivers response is evolved. For
example, the CO
released by an animal as it breathes
can be used as a cue by a blood-sucking insect to nd
its host. The mosquitos response is certainly evolved
(and indeed it has highly specialized receptors to detect
), but the release of CO
by the host did not evolve
to have the effect of attracting mosquitoes so it does
not count as a signal. The molecules learned in signa-
ture mixtures of kin or familiar animals are probably
best seen as cues rather than signals, as their produc-
tion has usually not evolved for that purpose (see
discussion in Section 1.7).
Scenarios for how chemical cues become phero-
mone signals fall into two main categories: sender pre-
adaptations, with signals starting from chemical cues
released by the sender, and receiver pre-adaptations
selecting for molecules that match the already existing
sensitivity of the receiver (Bradbury & Vehrencamp
2011, p. 377). If the receiver benets, it may rene its
tuning system. If the sender benets from the response,
the cue will undergo the evolutionary process of ritu-
alization, becoming a pheromone signal (Bradbury &
Vehrencamp 2011, p. 378; Tinbergen 1952). The
changes in ritualization could include increasing its
conspicuousness (e.g., making it more different from
other similar molecules) and reducing its variability
(stereotypy). Thus pheromones evolve from com-
pounds originally having other uses or signicance,
for example from hormones, host plant odors, chem-
icals released on injury, or waste products (Steiger
et al. 2011; Wyatt 2010). There will also be selection
for functional signal features such as longevity and
specicity (Section 1.4). The original functions of the
molecules may or may not be eventually lost.
1.3.1 Pheromone signals derived from
sender precursors
In the sender-precursor model of signal evolution,
pheromones can evolve from any reliable chemical
cue(s) to the senders condition (Figure 1.3) (Bradbury &
Vehrencamp 2011, p. 377 ff.). For example, if there are
molecules leaking from a mature female about to lay
eggs, then mutant males better able to detect them will
nd her rst and gain more matings (Figure 1.3). Over
generations this would result in selection for increas-
ing sensitivity to the females molecules (with multiple
copies of such receptors) and changes in the receptors
for greater specicity. The opportunistic co-option of
chemosensory receptor proteins for detection of new
molecules is discussed in Chapter 9. Although a given
odorant may be unlikely to t any one receptor
Animals in a chemical world
perfectly, it is likely to stimulate some. In turn, if the
sender benets, in the sh example the female would
benet by attracting males to fertilize her eggs, then
production and release of the molecules can evolve
into a signal (pheromone). Molecules become a pher-
omone only if there is positive selection on both sig-
naler and receiver.
Such a scenario has been suggested to explain the
use of body-uid molecules as pheromones by marine
polychaete worms, released with their gametes, which
immediately prompt the other sex to release its game-
tes (Chapter 3) (Figure 3.2) (Breithaupt & Hardege
2012). Similarly, hormones have been co-opted as sex
pheromones in sh, excreted in urine or across per-
meable membranes such as gills (Figure 1.4) (Chung-
Davidson et al. 2011; Stacey & Sorensen 2011). Species
specicity of these multicomponent sh sex phero-
mones comes from other molecules (Section 1.4)
(Levesque et al. 2011; Lim & Sorensen 2012). In ter-
restrial animals such as elephants and mice, many
pheromones are excreted in the urine and though not
necessarily hormones themselves, the quantities and
qualities reect hormonal levels related to body con-
dition (Section 1.6.2).
The aggregation pheromones used by bark beetles
may have evolved from the molecules produced by the
Spying (Change in receiver)
Communication (originator becomes signaler)
No specialized mechanisms for
detection or response
Evolution of detection and response
Receiver benefits
Receiver's response selects
for signal specialization
Chemical released
to the water
Chemical and its
release unchanged
Changes in chemical
and/or its release
Figure 1.3 The sender-precursor model of
signal evolution. Proposed stages in the
evolution of a communication function for
molecules released by an originatorani-
mal (the potential sender). The × in the upper
panel indicates that the receiving individual
has no special adaptations to receive the cue
beyond detecting it. The process starts with
an association between a cue and a condi-
tion of the originator. Receivers must be
able to perceive or evolve receptors for the
cue, and then incorporate the information
into a decision rule and a response. In this
spying phaseonly the receiver benets.
The transition to bilateral benet to both
sender and receiver could occur later if there
is a selective advantage to the sender, lead-
ing to ritualization of the signal to maximize
information transfer.
An original gure by Ivan Hinojosa
( in
Wyatt (2011), inspired by, and with text
adapted from, Stacey and Sorensen (2006)
(with permission from the authors).
Additional text adapted from Bradbury and
Vehrencamp (2011, p. 377).
1.3 How pheromone signals evolve from chemical cues
beetlesdetoxication of the toxic monoterpenes used
by the host trees as a defense against beetle attack
(Chapter 4) (Blomquist et al. 2010). Some of the
detoxication enzymes may have been co-opted as
biosynthetic enzymes for synthesis of pheromones
by the beetle.
Many alarm pheromones in social insects, which
provoke ght or ight in receivers, appear to have
Figure 1.4 Female hormone pheromones co-ordinate reproduction in the goldsh, Carassius auratus, by primer and releaser
effects on the male (Stacey & Sorensen 2009, 2011). It is likely that the hormones evolved into pheromones following the
scenario presented in Figure 1.3.
In the female, the rise and fall of blood concentrations of a succession of hormones (top), from 17β-estradiol (E2) to luteinizing
hormone (LH), steroids, and prostaglandin F
), stimulate release to the water of a succession of hormone pheromones
that reect her hormone levels. First, an unidentied recrudescent pheromone attracts males.
Second, a pre-ovulatory pheromone (the steroid androstenedione (AD), the maturation-inducing steroid 17,20β-P, and its
sulfated metabolite, 17,20β-P-S) are released the night before ovulation. Androstenedione induces agonistic behaviors among
males. As the 17,20β-P:AD ratio increases, males increase their own LH and begin to follow and chase conspecics. Males
exposed to the pre-ovulatory pheromone increase both the quantity and quality of sperm in the milt (semen), increasing the
likelihood of reproductive success.
Third, post-ovulatory pheromone (prostaglandin F
)) and its major metabolite 15K-PGF
stimulate both male
courtship and spawning behaviors and additional male LH increase.
All hormonal cues are released in quantities that range from 10 to >100 ng/h, are detected at concentrations in the picomolar
range, and act in concert to synchronize male behavior and physiology with the female. Figure from and caption after Stacey
and Sorensen (2009).
Animals in a chemical world
evolved from defensive compounds released by ght-
ing or injured conspecics (Chapter 8). There will be a
selective advantage to the potential receivers sensitive
to these compounds and responding appropriately to
protect the colony. Over evolutionary time, defensive
compounds may gain a signal function: for example,
many ant species use the same chemicals for defense
and alarm, to repel enemies and to alert and recruit
nestmates (Hölldobler & Wilson 1990, p. 260). One
example is the use of volatile formic acid for both
functions in Formica ant species (Blum 1996).
1.3.2 Pheromone signals derived from receiver
sensory bias
Any secreted molecule from a sender that overlaps the
receivers pre-existing sensory sensitivities, such as for
food odors, is likely to be selected over others and thus
potentially become a signal (Figure 1.5) (Arnqvist
2006; Bradbury & Vehrencamp 2011, p. 391 ff.;
Endler & Basolo 1998; Ryan 1998). For example, as
female moths use plant odors to nd host plants when
egg laying, their olfactory system is already tuned to
these odors: male moth pheromones appear to have
evolved to exploit this female sensory bias (Chapter 3)
(Figure 1.5) (Birch et al. 1990; Phelan 1997). The male
sex pheromone of the European beewolf, Philanthus
triangulum,(Z)-11-eicosen-1-ol, may exploit a pre-
existing female sensory bias for this molecule as it is a
characteristic volatile molecule given off by its hon-
eybee prey (Chapter 3) (Kroiss et al. 2010; Steiger et al.
2011). In Iberian rock lizards Iberolacerta cyreni (for-
merly Lacerta monticola), a pre-existing sensory bias
in females for a food chemical found in their insect
prey, the lipid cholesta-5,7-dien-3-ol (provitamin D
may have driven selection of this molecule as a com-
ponent of the pheromone secreted by males in their
femoral glands (Martín & López 2008, 2010a) (though
see Font et al. 2012). It is possible that this is also an
honest signal (index) that only high-quality males can
display (Section
As well as the adaptive sensory biases above, there
could also be hidden preferences (receiver
psychology), which are incidental side effects of ho