Bats

This page intentionally left blank

Bats
From Evolution to
Conservation
Second Edition

John D. Altringham
Professor of Animal Ecology and Conservation,
University of Leeds

DRAWINGS BY

Tom McOwat and Lucy Hammond

Myotis myotis (drawing by Lucy Hammond)

1
3
Great Clarendon Street, Oxford ox2 6dp
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide in
Oxford New York
Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam
Oxford is a registered trade mark of Oxford University Press
in the UK and in certain other countries
Published in the United States
by Oxford University Press Inc., New York
# John D. Altringham, 2011
All drawings # Tom McOwat unless otherwise credited.
The moral rights of the author have been asserted
Database right Oxford University Press (maker)
First published 1996
Second edition published 2011
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, or under terms agreed with the appropriate
reprographics rights organization. Enquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above
You must not circulate this book in any other binding or cover
and you must impose the same condition on any acquirer
British Library Cataloguing in Publication Data
Data available
Library of Congress Cataloging in Publication Data
Data available
Typeset by SPI Publisher Services, Pondicherry, India
Printed in Great Britain
on acid-free paper by
CPI Antony Rowe, Chippenham, Wiltshire

ISBN 978–0–19–920711–4 (Hbk.)

978–0–19–920712–1 (Pbk.)

10 9 8 7 6 5 4 3 2 1
Table of contents

Preface vii
Acknowledgements ix
Introduction—a biology lesson from the bats xi

1. Evolution and diversity 1

Evolution ....................................................................................................................................1
Classification and the distribution and diversity of bats ................................................ 18
Yinpterochiroptera ............................................................................................................... 20
Yangochiroptera ..................................................................................................................... 26
Adaptive radiation—why are the Phyllostomidae so numerous and so diverse? ...... 35

2. Flight 37
Some basic aerodynamics: aerofoils and flapping flight ................................................ 37
The evolution of flight ......................................................................................................... 43
The advantages of flight and the demands on the flyer: physiological
and biomechanical aspects of flight .................................................................................... 46
The wing and what it does in flight .................................................................................... 50
Ecological aspects of flight ................................................................................................... 51
3. Echolocation and other senses 61
What is echolocation? ............................................................................................................ 61
Sound generation and perception in echolocating bats ................................................... 63
Echolocation calls .................................................................................................................. 64
How do bats avoid being confused by other bats’ sonar? ............................................. 80
Auditory adaptations and the neural basis of echolocation .......................................... 80
Echolocation and behavioural studies in the wild ............................................................ 82
The cost of echolocation and its consequences .................................................................. 86
Other senses ........................................................................................................................... 88
The ecology of echolocation: interactions between flight, food, and foraging
habits ....................................................................................................................................... 90
Evolution of echolocation...................................................................................................... 92
4. Torpor and hibernation 97
Energy balance........................................................................................................................ 97
Concepts: ectothermy, endothermy, and heterothermy ................................................... 97
Concepts: torpor and hibernation ....................................................................................... 99
The cost of staying warm—the importance of body size ................................................ 99
The physiology of torpor ................................................................................................... 102
vi TABLE OF CONTENTS

The ecology of torpor and hibernation ........................................................................... 104

Biological clocks .................................................................................................................. 109
Evolution of torpor and hibernation .............................................................................. 111
5. Reproduction and development 113
Reproductive cycles ............................................................................................................ 113
Gestation .............................................................................................................................. 118
Birth, development, and survival .................................................................................... 119
Infant–mother communication and maternal care ......................................................... 124
Roost selection and maternal foraging patterns ............................................................ 125
Mating behaviour ............................................................................................................... 126

6. Roosting and feeding ecology 137

Roosting ecology ............................................................................................................... 137
Feeding ecology .................................................................................................................. 154
Information transfer ............................................................................................................ 163
Interactions between habitat, food, foraging, social structure, and behaviour ......... 165

7. Migration, social structure, and population structure 175

Migration .............................................................................................................................. 175
Navigation ........................................................................................................................... 179
Social and population structure ....................................................................................... 180
8. Biogeography, macroecology, community ecology, and the interactions
between bats and other organisms 195
Biogeography and macroecology .................................................................................... 195
Community ecology............................................................................................................ 211
Predators and prey ............................................................................................................ 220
9. Conservation 239
Why conserve bats? ............................................................................................................ 239
An overview of current status .......................................................................................... 239
Identifying and predicting conservation needs ............................................................ 241
Threats to bats ..................................................................................................................... 243
From the negative to the positive: how to conserve bats ............................................. 266
Bat conservation organizations ....................................................................................... 284

References 285
Index 319
 Preface

Preface to the first edition advances in almost all areas of bat biology and
most chapters have required a major rewrite and
This book is aimed primarily at undergraduate and expansion. In some areas developments have been
graduate students wishing to learn about bats, but so significant I have created new chapters. I have
also aims to show how a study of one group of also included a major chapter on conservation. My
animals can contribute to a wider understanding of interest in bats began in conservation and it has also
the processes that shape the natural world. It there- been responsible for the shift in my own research
fore has two main objectives. The first is to give an from biomechanics to bat ecology. Bats are under
account of the biology of the world's bats, emphasiz- threat from habitat destruction, degradation and
ing those aspects that are unique or highly adapted, fragmentation, from climate change, and from the
notably flight and echolocation. The second objec- many other pressures humans bring to bear on the
tive is to illustrate processes and concepts of broad planet. They therefore need our help. As a large,
biological relevance, many of which are major diverse, and widespread group of animals, their
themes in current research. The coverage is broad, value as indicators of environmental health has
but by no means comprehensive. I have tried to long been evident to biologists and has recently
make the book accessible to the growing number been more widely recognized. Although the book
of bat enthusiasts in all walks of life, by giving is broader in its scope than its parent it is now even
some relevant background to what I imagine are more difficult to be comprehensive, so I apologize if
the more difficult sections and by explaining termi- your favourite subject gets too little attention! It is
nology and principles that may be unfamiliar. increasingly difficult to do justice to the technical
complexity of many areas of study in a book of
such breadth. I have tried to walk that fine line:
Preface to the second edition
giving you sufficient detail to properly inform with-
In this updated and expanded version my aim has out getting you mired in the detail. In bringing you
not changed, but the state of bat biology has! Scien- up to date with the key advances in our knowledge
tific and public interest in bats has grown consider- and understanding of bats, I hope I have also illu-
ably in recent years and the scientific literature on strated some of the major changes in the way we
bats has exploded. There have been significant think about and study the natural world.
This page intentionally left blank
 Acknowledgements

Most of the drawings of bats are by Tom McOwat, a I am very grateful to Roger Butlin for introducing
few by Lucy Hammond—all are superb and make me to the power of molecular genetics and to Man-
the book a treat for the eye. The colour photographs uel Ruedi for casting his expert eye over the genetics
were all taken by Merlin Tuttle and I am grateful to sections of the book.
Bat Conservation International for making these My wife Kate and our daughter Anne did most of
available. Many of the black and white photographs the hard work of compiling and checking the refer-
are my own, and where they are not, the photogra- ences—a task I was very grateful to pass on! Again,
phers are gratefully acknowledged in the legends. if mistakes remain they are mine.
Chloe Bellamy, Anita Glover, Elena Papadatou, Helen Eaton at OUP has been a model of patience,
Kirsty Park, Chris Scott, and Paula Senior read efficiency, and helpfulness and it has been a plea-
many or all of the chapters—many thanks for your sure to work with her. Ian Sherman at OUP has also
comments, corrections and encouraging words. Any been very patient—Ian, I hope this long overdue
remaining failings are of course entirely mine! book is worth the wait.
Thanks to all of them and to other past and present Most of all, thanks to Kate for her unwavering
members of my research group for the hard work support!
and fun we’ve had while studying bats.
This page intentionally left blank
 Introduction—a biology lesson from
the bats

Bats are one of the most successful mammalian or- about (yes, there’s still uncertainty) 29 mammalian
ders, and probably the most diverse. The 1,110-plus orders, including animals as varied as primates,
species provide an unparalleled exhibition of varia- carnivores, cetaceans, rodents, insectivores, ungu-
tions on the mammalian theme, and a broad lesson lates, seals, sloths, and marsupials. Bats are
in biology. In the bats we see excellent illustrations distributed all around the world: over 200 species
of adaptive radiation, optimal foraging, coevolu- are found in Africa and Madagascar, over 300 in
tion, convergent evolution, reciprocal altruism, the South and Central America and the Caribbean, and
arms race between predator and prey, the complex a similar number in South East Asia and Australa-
interactions between behaviour and population sia. They are also well represented in higher lati-
structure and key macroecological and bioge- tudes: about 40 species are resident in both North
ographical principles, to name just a few examples. America and western and central Europe. Several
This is a book about bats, but it will also use them as vespertilionid bats (for example Eptesicus, Lasiurus,
a vehicle to show how the natural world is shaped. and Myotis species) spend the summer north of
Rodents are the only mammalian order to out- the Arctic Circle. Other members of this very large
number bats with, the last time I looked, 2,227 spe- family (for example Dasypterus species) forage in the
cies, but they are arguably less diverse in their chill and windy regions of southern Patagonia.
biology. Over 20 per cent of all mammals are Bats have found their way to most islands, however
bats—surprising when you think that there are remote, where they may be the only native

Brown long-eared bat, Plecotus auritus

xi
xii INTRODUCTION

mammals. New Zealand has only two species of evidence for independent origin, with subsequent
land mammal, both bats: one species from the convergent evolution, was published during the
endemic family Mystacinidae, and a vespertilionid. 1980s. The traditional viewpoint seemed to be shak-
The nearest relatives of Mystacina tuberculata are the ing on its apparently weak, and largely anatomical,
fisherman bats (Noctilio) of South America—one of foundations and controversy raged. But new evi-
several possible examples of related families sepa- dence for the common origin of bats emerged from
rated by the breaking up of Gondwanaland in the the labs of molecular biologists and anatomists. In the
late Cretaceous. This is one of several bits of circum- end, the evidence for a common ancestor became
stantial evidence for the very early origin of bats, overwhelming and the controversy died, but not
now supported by firmer evidence from genetic without generating some fascinating results. In
studies. Bats probably witnessed the demise of the resolving the debate, new and surprising discoveries
dinosaurs in the mass extinction at the end of have been made about the evolution of bats and the
the Cretaceous. origins of flight and echolocation. Ironically and sur-
Bats range in size from one of the smallest mam- prisingly (to me at least), the closest relatives to the
mals (the bumblebee bat, Craseonycteris thonglongyai, non-echolocating, Old World fruit bats appear to be
1.5–2 g) to 1 kg flying foxes (Pteropus species) with the horseshoe bats—the most advanced and sophis-
wingspans of over 1.5 m. They also come in a wide ticated echolocators!
range of shapes and colours. Most bats are admit- The niche that many bats exploit as aerial, noctur-
tedly rather drab, but there are exceptions, like the nal hunters is a demanding one. Flight places major
painted bats (Kerivoula), whose bright and cryptic anatomical and physiological restrictions on bats,
patterning may camouflage them in their exposed but the rewards, evident in their success, are great.
tree roosts—some look like flowers and fruit. The A better understanding of the aerodynamics of flap-
tube-nosed fruit bat (Nyctimene major) has wing pat- ping flight stimulated studies of the relationship
terning to match the tree trunks to which it clings. between wing morphology, flight characteristics,
The wonderfully grotesque hammer-headed bat and feeding ecology. Add to flight the ability to
(Hypsignathus monstrosus) has a nose of immense locate and catch prey in the dark using sound, and
proportions. Males hang in the trees along rivers you have in bats a highly adapted product of evolu-
and call to passing females, who select the best (the tion, with many interesting biological stories to tell.
most impressive callers?) for mating. This is one of We are only just beginning to understand the com-
the best documented examples of lekking in mam- plexity, subtlety, and remarkable perceptual abil-
mals, but more are now being found among bats. ities of bat echolocation. Advances in technology
Then there’s the striking crested free-tailed bat and some ingenious experimentation have uncov-
(Chaerephon chapini) which distributes pheromones ered dazzling feats—such as the ability of greater
from the erectile crest of hair on its head. The nose- horseshoe bats (Rhinolophus ferrumequinum), at least
leaves and varied facial protuberances of many bat under laboratory conditions, to distinguish different
families are often useful identification features. Most prey species from the modulated echoes returned by
have a functional role in echolocation, for example flying insects. Like flight characteristics, the type of
those of the horseshoe bats (Rhinolophidae), but the echolocation used is determined by environment,
function of others has yet to be determined, if indeed foraging style, and the prey sought. Studies of bats
they have one! in the wild have shown how whiskered bats, Myotis
A complex and exciting story has unfolded around mystacinus, subtly alter the structure of their echolo-
the very origins of bats. Until recently two sub-orders cation calls in flight to compensate for errors in
were described, the Megachiroptera, the Old World target ranging due to Doppler shift and changing
fruit bats or flying foxes, and the more widespread, position—mechanisms worthy of the most sophisti-
more numerous and more diverse Microchiroptera. cated modern military technology.
The traditional and widely-accepted view was that The success of bats in high latitudes is due in no
these two sub-orders arose from a common ancestor. small measure to their ability to reduce body tem-
However, a substantial and broad-based body of perature and save energy when insect availability is
 INTRODUCTION xiii

low. Torpor is the ability to reset body temperature amphibians, reptiles, birds, mammals (including
to a level well below that required for normal activ- other bats), fruit, nectar, pollen, occasional leaves
ity, to actively regulate it within narrow limits, and and seeds, and of course blood. Many are highly
to actively return to full operating temperature: few specialized in their diets, but others, including
mammals perform this task as well as bats. We many spear-nosed bats (Phyllostomidae) of South
commonly think of torpor in the context of the and Central America, are omnivorous and will
long winter hibernation, but its use is an important take insects, vertebrates, and fruit. The very diverse
part of a flexible, day to day, energy-saving strategy feeding ecologies of bats have been sources for
among many temperate bats. many interesting and informative investigations.
Few potential roost sites have been overlooked The neotropical vine Mucuna holtonii bears a mod-
by bats. Brazilian free-tailed bats (Tadarida brasilien- ified petal or vexillum that acts as a powerful acous-
sis) in Central America and the southern United tic mirror, reflecting most of a bat’s echolocation call
States form cave dwelling colonies that number back towards the bat over a wide range of angles of
millions of individuals. The hoary bat (Lasiurus incidence. Wild bats show a very strong preference
cinereus) is a solitary tree dweller, hanging (with for flowers with an intact vexillum since it indicates
its young) from high branches in the boreal forests. a rich nectar source. Bats don’t only use echolocation
Incidentally, this species ranges all the way to to find food—passive sound, smell, vision, and even
South America, and is the only species found in heat sensors are also important. Many bat-polli-
Hawaii. A few species live in underground bur- nated flowers are violet and reflect ultra-violet
rows, and in the case of the African slit-faced bats light, and some bats can perceive this. The nectar-
(Nycteris species), those of the aardvark! The short- feeding Glossophaga soricina was found to be sensi-
tailed bat (Mystacina tuberculata) of New Zealand tive to wavelengths down to 310 nm, well into the
frequently forages on the ground, and burrows into UV. But it does not end there. Even more recently it
fallen and decaying kauri trees, where they roost has been shown that G. soricina and Carollia perspi-
like peas in a pod. Adaptations for this unusual cillata have cone cells for colour vision, with two
way of life include tough wings that can be tucked types of light-sensitive opsin protein for short
away in pouches on the body, and strong talons on (blue/UV) and long (green/red) wavelengths—
the thumb and toes. Several species (for example these bats have all that is necessary for daylight
Artibeus, Ectophylla, and Uroderma) bite through the vision, dichromatic colour vision, and UV vision!
main supporting ribs of palm and Heliconia leaves Nectar and fruit eating bats also provide the biol-
to collapse them into tents and, in Gabon, Myotis ogist with fascinating examples of coevolution. The
bocagei roosts inside the flowers of the water arum. long tongues of nectar-feeding glossophagine bats
Tylonycteris bats roost inside bamboo shoots, gain- are well known. The record is held by the recently
ing entry through the internodal emergence holes discovered Anoura fistulata at 85 mm, 150 per cent of
of a chrysomelid beetle. They have fleshy pads or its body length—second only to chameleons within
‘suckers’ on their wrists and ankles to grip the the vertebrates. When not in use, this remarkable
inside of the culm. Suckers are also present on tongue is retracted into the thoracic cavity with its
bats of two other families that roost in furled base between the heart and the sternum, a structure
leaves, the Myzopoda of Madagascar and the Thyr- similar to that independently evolved by ant-eating
optera of Central and South America—a good pangolins. This enormous tongue enables the bat to
example of convergent evolution. feed from the slender, 80–90 mm long corolla of the
Bats feed on a wider variety of food than any flowers of Centropogon nigricans. No other animal
other mammalian order. Most feed on insects and has been recorded visiting the plant, suggesting
other arthropods—the Pallid bat (Antrozous pallidus) that it is probably the only pollinator.
of the south-western United States, and some Afri- Interactions between predator and prey have led
can slit-faced bats, have a liking for scorpions! One to a continual ‘arms race’, with the prey evolving
population of the fisherman bat (Noctilio leporinus) better means of escape, and the predator, of neces-
eats lots of fiddler crabs. Others feed on fish, sity, overcoming them in the fight for survival. This
xiv INTRODUCTION

The flat headed bat, Sauromys petrophilus

arms race is nowhere better illustrated than in the and three species of vampire bat. This diversity of
relationship between bats and their insect prey. feeding habits is paralleled by a fine display of var-
Noctuid moths, for example, have evolved ‘ears’ to iations in form, physiology, and ecology—the long
detect approaching bats and have stereotyped and bristly tongues of nectar feeders, the white tent-
avoidance mechanisms hardwired into their ner- makers, the record-breaking kidneys of vampires,
vous system. Some bats have evolved echolocation and the chin-flap-cum-night-cap of the wrinkle-
calls that are less audible to moths, or catch them faced bat.
without using echolocation. New and increasingly Over the last 20 years several areas of research
subtle adaptations are uncovered almost every year. have exploded into activity with the rapid develop-
There is also evidence for arms races between bats ment of molecular genetics and easy access to user-
and their predators. friendly and powerful computing. They have given
Few cited examples of reciprocal altruism are us powerful tools to investigate, among other
clear cut and beyond explanation in terms of kin things, the evolutionary origins of bats and their
selection. One exception appears to be blood sharing ability to fly and echolocate, historical distribution
in the common vampire bat (Desmodus rotundus). patterns, and the complex social lives of bats.
Vampires must have their 25 ml blood meals on a For example, the consequences of the most recent
regular basis to survive, and in close-knit groups glaciations and the role of mountains as barriers to
within a colony bats will regurgitate some of their post-glacial recolonization are being revealed by
last meal to a ‘buddy’ who has been unable to feed. an analysis of current population genetic structure.
From studies of the behaviour of wild bats in The subtle adaptive value of the complex social life
the roost, and of captive bats whose degree of relat- of the greater horseshoe bat is being unravelled
edness was known, it appears that this behaviour by genetic analysis over many generations.
of vampire bats is truly altruistic and cannot always The abundance and diversity of bats makes them
be explained on the basis of kin selection. The sys- good models in the fields of biogeography and
tem works because the donor will only give if macroecology and this is another area that has
the benefit to the recipient is far greater than its seen a rapid expansion in recent years. Studies
own loss, and because the favour is returned at a of bats are helping us understand the rules that
later date. govern the distribution of life on the Earth, such as
The Phyllostomidae (New World leaf- or spear- species–area relationships and the ways in which
nosed bats) are the family to go to for a lesson in latitude and elevation determine the richness and
adaptive radiation. The family contains about 160 abundance of life.
species in 55 genera. They are fewer and far less Sadly, much of this fascinating diversity is
widespread than the Vespertilionidae or evening under threat. Bats are subject to the same pres-
bats, but are unmatched in the range of food they sures as the rest of life on the planet—relentless
eat. Many species are insectivorous, like all in the human pressure leading to habitat degradation,
ancestral family, but there are now large numbers of fragmentation, and loss. They are also subject to
fruit, flower, nectar and pollen-eaters, carnivores, some more or less unique problems because of
 INTRODUCTION xv

their life history strategies. Their tendency to form fascinating biology and their ecological roles as
large aggregations for breeding and hibernating, pollinators, seed dispersers, and pest control-
often in close proximity to humans, make them lers—and we are taking the messages learned to
particularly vulnerable. Their low reproductive a wider audience; the public image of bats is also
rates make them slow to recover from population improving. This is an important part of the conser-
decline. The driving force behind much current vation effort. A growing understanding and
research is conservation: a desire to understand appreciation of the ecological role of bats, and a
bats so that we are better able to protect them. fascination for their unique biology, must make
We are learning more and more about their conservation easier.
This page intentionally left blank
 CHAPTER 1

Evolution and diversity

Fossils and what they tell us about the origin of bats. Molecular genetics tells us more about the evolutionary
history of bats. Why did bats evolve? The origins of flight and echolocation. Past controversies, the current
concensus and the implications to bat evolution. A brief classification of modern bats. Brief descriptions at
family level. The Phyllostomidae—an example of adaptive radiation.

Evolution unquestioned and, controversially, it had even

been suggested that they might have different an-
Fossils and the origins of bats
cestors. The resolution of this controversy has led to
Bats are the second largest mammalian order. There some surprising results, which will be discussed in
are an estimated 1,116 species and the number detail later in the chapter. For the moment you will
grows annually as new species are named, primarily have to take the phylogenetic tree in Fig. 1.1 on trust.
through the splitting of existing species based on It shows the known fossil bats in relation to extant
genetic evidence. They have long been divided into bats, and a number of other mammalian orders: note
two sub-orders, the Mega- and Microchiroptera, that the traditional sub-orders have disappeared.
commonly referred to as the megabats and micro- It is generally said that bats are not well repre-
bats. The justification for these two sub-orders, as sented in the fossil record. There is no shortage of
we will see, has recently been undermined, and the species, just a shortage of anything more than jaws
terms now have to be interpreted differently. I will for most of them, and the ratio of known extinct to
define them and continue to use them where appro- living species is low in relation to other mammals.
priate in this chapter, since they have historical and Nevertheless, fossils of almost 200 species have been
biological significance and you will find them found. There are fossil representatives of all modern
widely used in the literature. However, as these families except Rhinopomatidae and Craseonycter-
two terms are now obsolete I will endeavour not to idae, in addition to fossils of species from a number
use them after this chapter! The megabats are the of long lost families. Most fossils are too recent to
Old World fruit bats: often large (for bats that is, shed much light on the origins of bats, but several do
at 20–1,500 g), exclusively plant-eating (they eat date back to the Eocene.
mainly fruit, flowers, nectar, and pollen), and con- What do the earliest fossils tell us about bat evo-
fined to Africa, tropical Asia, and Indo-Australasia. lution? The answer is surprisingly little. The oldest
As I write there are 186 known, living species, all fossil bat, Icaronycteris index, was found in the Green
belonging to one family, the Pteropodidae. Micro- River, Polecat Bench formation of Wyoming, not far
bats on the other hand are found on every continent from Yellowstone Park in the United States, and has
except Antarctica, and, like the megabats, on many been dated to the early Eocene, 50 million years ago
isolated islands. They are generally smaller than (Jepson 1966, 1970). Icaronycteris looks remarkably
megabats (1.5–150 g), and they eat all sorts of things, like a modern insectivorous bat. The best European
although the ancestral microbat almost certainly ate specimens were found in the famous oil-shale
insects and other arthropods. The 930-plus species pits at Messel, near Darmstadt in Germany: Archae-
are distributed among 17 or 18 families. Until quite onycteris, Palaeochiropteryx (Fig. 1.2), and Hassianyc-
recently this division into sub-orders was largely teris (Smith and Storch 1981). Preservation is

1
2 BATS

Eulipotyphla
Cetartiodactyla
Perissodactyla
Carnivora
Pholidota
Icaronycteris
Archaeonycteris
Hassianycteris
Paleochiropteryx
Pteropodidae
Rhinolophidae
Hipposideridae
Rhinopomatidae
Craseonycteridae
Megadermatidae
Emballonuridae
Nycteridae
Natalidae
Molossidae
Miniopteridae
Vespertilionidae
Myzopodidae
Mystacinidae
Thyropteridae
Furipteridae
Noctilionidae
Mormoopidae
Phyllostomidae
Figure 1.2 Palaeochiropteryx tupaiodon, an Eocene fossil bat from
Figure 1.1 An evolutionary tree of modern and fossil bats. Adapted from Messel in Germany.
Teeling et al. (2005), with permission from AAAS. At the top (dashed lines)
are the relationships to a number of other mammalian orders, the
Eulipotyphla (insectivores), Cetartiodactyla (whales and relatives), million years ago. If so, they shared the world with
Perissodactyla (odd-toed ungulates—horses, etc.), Carnivora (carnivores), the dinosaurs, and watched their extinction at the
and Pholidota (pangolins). The dotted lines denote extinct bats. The
end of the Cretaceous.
Miniopteridae have until recently been included in the Vespertilionidae.
Fossils are lacking only from the Craseonycteridae and Rhinopomatidae.
What evidence can we cite in support of this time
scale? Until recently there was little evidence of real
substance, but some of a persuasive, if circumstan-
so good that recognizable insect remains can be seen tial, nature. As we’ll see a little later, bats show no
in the gut of some specimens. On the fossilized close affinities to any other mammalian order: ac-
wingscales of moths eaten by the bats, identifiable cording to the latest analyses (Murphy et al. 2001
pollen grains can be seen—evidence of where the and Fig. 1.1), their nearest, but still distant, relatives
moth took its last meal—ecology captured in stone! include the pangolin and the horse! These are
All of these bats date back about 45 million years, unlikely ancestors for the bats, and they probably
and all resemble modern bats. A lavishly illustrated evolved from something resembling modern tree
book has been published on the Messel beds, with shrews. This early inability to link bats to any
superb photographs of these bats (Schaal and Zeig- known mammalian group in itself suggested a
ler 1992). It is now sadly out of print, but well worth very early origin and prompted a more oblique
finding in a library. No fossil bats were known that search for clues. Some moths, mantids, lacewings,
were in any way intermediate in form between a and other insects have ‘ears’ whose main function
modern bat and some early, tree-living ancestor appears to be to detect the echolocation calls of bats
which might have got around by jumping or glid- and trigger escape responses (Fullard 1987, Bailey
ing. However, if we allow time for the evolution of 1991). Gall and Tiffney (1983) discovered the fossi-
these sophisticated aerial insectivores, with an lized egg of a noctuid moth in deposits at Martha’s
apparently advanced echolocating capability, then Vineyard in Massachusetts which date back to about
bats probably made their appearance over 65 75 million years ago. All known living and extinct
 EVOLUTION AND DIVERSITY 3

noctuids are tympanate, and their ears are tuned to landmass. The break up started with the separation
the ultrasonic frequency range used by many echo- of Antarctica/Australasia from South America, and
locating bats. If the moth that laid this eggs was ended with the separation of Australia from Antarc-
tympanate, then echolocating bats may already tica about 50 million years ago (Fig. 1.3).
have been around 75 million years ago. Finally, New Zealand has just two endemic mammalian
there are a number of cases where closely related species, both bats. The closest living relatives of the
bat species live on now distant fragments of Gond- short-tailed bat (Mystacina tuberculata) appear to be
wana. Before the break up of this southern super- the fisherman bats (Noctilio) of South America (Pier-
continent these bats presumably shared the same son et al. 1986, Teeling et al. 2003). The more recent

0Ê 0Ê
0Ê 30 0Ê 30Ê
–3 Ê –3
0Ê

0Ê
60

60
–6

–6
Ê

Ê
–90Ê

–90Ê
90Ê

90Ê
0Ê

0Ê
–1

–1
12

12
20

20
Ê

–1 0Ê –1 0Ê
50 15 50 15
Ê Ê
180Ê 180Ê
100 Ma Reconstruction 75 Ma Reconstruction
0Ê
0Ê
Ê 30
– 30 Ê
0Ê 30
–3 Ê
0Ê

60
0Ê

–6

Ê
60
–6

–90Ê

90Ê
–90Ê

90Ê

New
Zealand
0Ê
–1
0Ê

12
20
–1

12

Ê
0Ê2

–1 0Ê
–1 50 15
50 0Ê Ê
Ê 15
180Ê
180Ê
65.0 Ma Reconstruction 47 Ma Reconstruction

Figure 1.3 The break-up of Gondwana. Plate fragments 100, 75, 65, and 47 million years ago are shown as solid blocks outlined in black. Current day
shorelines are shown largely within the outlines of the old land masses. Small islands between Antarctica and New Zealand may have aided the movement of the
ancestors of Mystacina. Adapted from maps available at www.odsn.de.
4 BATS

work of Teeling and colleagues is based on an exten- complex data they produce, have led to a revolution
sive genetic comparison of Mystacina, Noctilio, repre- in our understanding of how and when bats
sentatives of twelve other bat families, and six evolved. Figure 1.4 shows a basic molecular phylog-
mammalian outgroups. The ancestors of these two eny of the bats (from Teeling et al. 2005) and where
bats appeared to separate from other bats about in the world the major groups are thought to have
67 million years ago, probably in South America. evolved. Figure 1.5 shows the estimated dates of the
Mystacina separated from Noctilio about 47 million major radiations (Teeling et al. 2005). These results
years ago and probably dispersed to what was to be are based on over 13 kb of sequence data (13,000
Antarctica and Australia. New Zealand had already base pairs of DNA sequence) from 18 nuclear genes
started to drift away from the Antarctic–Australian and species from all known bat families were
Plate about 75 million years ago (Griffiths and Varne included in the study. Recently it has been sug-
1972), but had perhaps not travelled too far by gested that all 19 species in the genus Miniopterus
47 million years ago to prevent this migration, per- should be elevated to family status, the Miniopter-
haps by island-hopping. idae (for example Hoofer and van den Bussche
That was the story when I thought I had finished 2004), rather than be included within the Vesperti-
writing this chapter and moved on to the next. But, lionidae. A re-analysis of the morphological data of
not for the first time in the writing of this book, I was Simmons and Geisler (1998) by Teeling et al. (2005)
forced to go back and describe new developments. suggested that the four extinct Eocene families were
Simmons et al. (2008) have recently described a new closely related and sister taxa to all living bats.
and important fossil species, again from the Green This tree is supported by that shown in Fig. 1.6
River Formation in Wyoming. Onychonycteris finneyi constructed by Eick et al. (2005) and based on a
is similar in age to Icaronycteris, but appears more 4 kb sequence from four intron markers. Introns
primitive in several ways. The cochlea is proportion- are non-coding sections of DNA situated between
ally smaller than in other Eocene bats, suggesting those coding for proteins. The analysis by Eick and
that echolocation was either poor or absent. The colleagues included 58 species from all families
fingers in its short, broad wings are unique amongst except the Craseonycteridae. Eick et al. (2005) also
bats in that all have retained their claws, and attempted to look at the geographical origins of bats
although it has an anatomy consistent with powered and, in contrast to Teeling et al. (2005), concluded
flight, it does appear to be primitive. Its limb pro- that modern families had a southern hemisphere
portions are intermediate between those of non- origin, probably in Africa. The molecular trees con-
volant mammals and all known bats, including firm an early origin for bats, with the last common
other Eocene species. This does not necessitate ancestor about 64 million years ago and possibly
bringing the evolution of bats forward, but it does earlier. The bats underwent a major diversification
shed light on a number of questions. It shows, for about 50 to 52 million years ago when all four major
example, that the claws on the third to fifth fingers microbat lineages appeared. According to Teeling
were lost after the evolution of flight and that both et al. (2005) they appear to have evolved in the
the arm and hand bones continued to elongate, ancient northern landmass of Laurasia, in what is
leading to more aerodynamically efficient wings. now North America, before migrating and evolving
Onychonycteris may also shed light on one of the across Laurasia and Gondwana to the south (but
more debated questions about bat evolution: which see Eick et al. (2005) for an alternative view). Their
evolved first, echolocation or flight, or did they radiation arose from two newly identified sub-
evolve together? I’ll leave that question until later. orders, the Yinpterochiroptera and the Yangochir-
optera. The Yinpterochiroptera evolved in Asia and
The current picture–molecular genetics meets have an exclusively Old World distribution. The
origins of the Yangochiroptera are less clear, but
morphology
they too probably arose in Laurasia, in either Asia
The appearance of sophisticated molecular techni- or Europe, but now have a global distribution which
ques, and advanced methods for interpreting the raises some interesting questions about dispersal.
 EVOLUTION AND DIVERSITY 5

Pteropodidae
Asia
Rhinolophidae
Hipposideridae
Rhinopomatidae
Craseonycteridae
Europe/Asia Yinpterochiroptera Megadermatidae
Emballonuridae
Nycteridae
Natalidae
Europe/Asia? Molossidae
Miniopteridae
LAURASIA Vespertilionidae
Myzopodidae
Mystacinidae
Thyropteridae
Furipteridae
Central & South Noctilionidae
America Mormoopidae Figure 1.4 A molecular phylogeny of the bats showing
Phyllostomidae where in the world the major groups were thought to
have evolved. (Adapted from Teeling et al. (2005) with
GONDWANA Yangochiroptera permission from AAAS.)

GENUS FAMILY SUPERFAMILY

Pteropus
24 23

Yinpterochiroptera
Rousettus Pteropodidae
Cynopyterus
22 Nyctimene
58 39 Rhinolophus Rhinolophidae
Hipposideros Hipposideridae
Megaderma Megadermatidae Rhinolophoidea
52 16
43 Macroderma
49 Craseonycteris Craseonycteridae
Rhinopoma Rhinopomatidae
Nycteris Nycteridae
64 Emballonura
52 30 Emballonuridae Emballonuroidea
42 Rhynchonycteris
Taphozous
22 Tonatia
Artibeus
19 Phyllostomidae
26 Anoura
Yangochiroptera

36 Desmodus
Noctilionoidea
56 Pteronotus Mormoopidae
42 Noctilio Noctilionidae
36 Furipterus Furipteridae
46 40
Thyroptera Thyropteridae
52 Mystacina Mystacinidae
Myzopoda Myzopodidae
Antrozous
54 10 Rhogeessa Vespertilionidae
20
Myotis
47 Tadarida
Vespertilionoidea
22 Molossidae
50 Eumops
Natalus Natalidae

70 60 50 40 30 20 10 0

K–T boundary Cenozoic

Figure 1.5 A proposed time scale for the evolution of the representatives of all but one (Miniopteridae) of the bat families. (Adapted from Teeling et al.
(2005) with permission from AAAS.) All numbers refer to millions of years before the present. K–T boundary = Cretaceous–Tertiary boundary.
6 BATS

Peropteryx kappleri
Rhynconyceris naso
Saccopteryx bilineata
Taphozous mauritianus EMBALLONURIDAE
Nycteris grandis
Nycteris javanicus
Nycteris macrotis
Nycteris thebaica NYCTERIDAE
Furipterus horrens FURIPTERIDAE
Noctilio albiventris
Noctilio leporinus NOCTILIONIDAE
Mormoops megalophylla
Pteronotus parnellii
Pteronotus personatus MORMOOPIDAE
Artibeus jamaicensis
Glossophaga soricina
Trachops cirrhosus
Desmodus rotundus PHYLLOSTOMIDAE
Mystacina tuberculata MYSTACINIDAE
Thyroptera lavali
Thyroptera tricolor THYROPTERIDAE
Miniopterus australis
Miniopterus macrocneme
Miniopterus fraterculis
Miniopterus inflatus
Miniopterus natalensis MINIOPTERIDAE
Eptesicus hottentotus
Nycteceinops schlieffeni
Scotophilus dinganii
Myotis tricolor
Cistugo lesueuri
Cistugo seabrai VESPERTILIONIDAE
Chaerephon ansorgei
Otomops martiensseni
Mormopterus petrophilus
Tadarida aegyptiaca MOLOSSIDAE
Natalus major
Natalus micropus NATALIDAE
Myzopoda aurita MYZOPODIDAE
Anthops ornatus
Hipposideros caffer
Hipposideros commersoni
Cleotis percivali HIPPOSIDERIDAE
Rhinolophus capensis
Rhinolophus darlingi
Rhinolophus fumigatus
Rhinolophus hildebrandtii RHINOLOPHIDAE
Cardioderma cor
Megaderma spasma
Megaderma lyra MEGADERMATIDAE
Rhinopoma hardwickei RHINOPOMATIDAE
Cynopterus sphinx
Epomophorus minor
Epomophorus wahlbergii
Rousettus aegyptiacus PTEROPODIDAE
Genet
Pangolin
Horse
0.05 substitutions/site

Figure 1.6 A phylogenetic tree derived from sequences of four intron markers by Eick et al. (2005). Note that the Craseonycteridae were not included
in this study.

For example, did the New World emballonurids, Wyss 1998)? Teeling et al. (2005) suggest that the
which separated from their African cousins about noctilionoids arose in Gondwana, perhaps South
30 million years ago, cross the Atlantic from Africa America. Among them, the phyllostomids, mor-
on stepping stones or vegetation rafts as has been moopids, noctilionids, furipterids, and thyropter-
suggested for new world monkeys (Flynn and ids are largely confined to the neotropics, but the
 EVOLUTION AND DIVERSITY 7

two mystacinid species (one of them now extinct) Gliding and flying
are found only in New Zealand and the one my- What follows is largely informed speculation, but
zopodid is an endemic of Madagascar! The two I think few biologists would question it. Over
speciose families that make up most of the vesper- thousands of years of jumping around after insects,
tilionoids, the Vespertilionidae and the Molossi- from branch to branch, and tree to tree, the ancestors
dae, both have global distributions and their of microbats probably evolved gliding membranes
place of origin is uncertain. similar to those of modern mammals like flying
squirrels and sugar gliders (Fig. 1.7). We should
not be surprised that this may have happened—
Why did bats evolve?
gliding has evolved independently many times in
The evolution of new species is the product of the the vertebrates, with living examples among the
spontaneous generation of random genetic mutations fish, amphibia, reptiles, marsupials, and eutherian
and the forces of natural selection—often a response mammals (see for example Rayner 1981).
to a changing environment. Change means a new Before going any further, we ought to ask the
physical environment, new sources of food, new ha- question: why did flight evolve? Two very powerful
bitats, new competitors, and new predators. At the reasons come quickly to mind. Less energy is ex-
time bats are thought to have been evolving, the pended gliding from tree to tree than running
flowering plants were in the first stages of their mas- down the trunk, running across the ground, and
sive diversification. Müller (1981) conducted a review running up the trunk of the next tree. Furthermore,
of the pollen record and demonstrated a proliferation if the animal doesn’t come down to the ground, it
of angiosperms at all taxonomic levels. They became doesn’t have to face terrestrial predators either. Con-
dominant over more primitive plants in the Cenoma- trolled, flapping flight brings other advantages, dis-
nian period (100–95 million years ago), and modern cussed in Chapter 2.
families appeared in great numbers from the begin- Let’s go back to the evolution of gliding. A nar-
ning of the Maastrichtian (69 million years ago). By row extension of the skin between front and hind
the end of the Cretaceous the insects supported by legs probably became more extensive, and exten-
these plants were abundant and insectivorous and ded to the spaces between hind legs and tail. With
frugivorous mammals were becoming well estab- the appearance of webbing between the fingers
lished (Lillegraven 1974). Teeling et al. (2005) cite evi- and toes, the fingers could elongate, carrying the
dence for a 7  C rise in mean temperature, a webbing with them, dramatically increasing the
significant increase in plant diversity and a peak in wing area. A study by Sears et al. (2006) suggests
Tertiary insect diversity that coincided with the that increased local expression of a single protein,
appearance of the four major microbat groups. This Bmp2, which causes proliferation of bone forming
progressive increase in insect diversity provided an chondrocytes, may have been a critical step in the
abundant food source for bats and other mammals, evolution of the bat wing. It appears to have been a
but they didn’t have them all to themselves. During rapid evolutionary step, since the lengths of the
the day, they would have had to compete with birds third to fifth fingers have remained very similar
and other insect eaters. Archaeopteryx, the first bird, in length (relative to body size) over the last 50
dates back to the early Cretaceous (135 million years million years (with the notable exception of Ony-
ago), and birds were abundant by the time bats ap- chonycteris, Simmons et al. 2008). By having its
peared on the scene. Birds were likely to be significant fingers within its wings the ancestral bat gained
competitors and predators of small mammals by the greater control over wing shape, giving it more
late Cretaceous (65 million years ago). For these rea- aerial control. Ultimately wings used for gliding
sons many early mammals (like their modern coun- must eventually have begun to be used for active
terparts) were nocturnal, and it is presumed that bats flapping flight, as the necessary neuromuscular,
evolved from one of these small, nocturnal, and arbo- circulatory, and respiratory adaptations evolved
real (tree-dwelling) mammals. (see Chapter 2).
8 BATS

Malayan colugo
Cynocephalus variegatus

African flying squirrel

Anomalurus sp.

southern flying squirrel

Glaucomys volans

Figure 1.7 Modern gliding mammals.

 EVOLUTION AND DIVERSITY 9

Echolocation tupaidon both fell in the middle of a cluster of data

As early bats became more agile, perhaps even points for modern echolocating bats. For their size,
before they became bats, they would have had to their basal turns appeared to be as well developed as
improve their orientation skills to be successful those of most modern bats. However, a later analy-
night flyers. Echolocation, orientation by analysis sis of a more extensive data set tells a different story
of the echoes from emitted sound pulses, probably (Habersetzer and Storch 1992). Figure 1.8 shows
increased in sophistication alongside flight. It is dif- basicranial (skull) width (skull length, as used by
ficult to imagine how the two could have reached Novacek (1985), may depend on feeding strategy)
their present level of sophistication unless they coe- plotted against cochlea diameter for the Eocene
volved, at least in the absence of good night vision. It Messel bats. They have cochleas similar to, or smal-
is likely that the ancestors of bats, like some modern ler than, the smallest of the Vespertilionidae, bridg-
insectivores, emitted ultrasonic sounds and perhaps ing the gap between echolocating insectivorous bats
had a simple form of echolocation, which became and the non-echolocating Old World fruit bats.
increasingly sophisticated as bats became more A number of the extant echolocating bats studied
agile flyers. Novacek (1985) produced evidence to fall in the same area: all of them have foraging
suggest that the very early fossil bats, Icaronycteris strategies which make use of vision, olfaction, or
index and Palaeochiropteryx tupaidon, had a well prey-generated sound, in addition to echolocation.
developed echolocation system—perhaps better In other words, they do not depend entirely on
developed than that of some modern bats. This evi- echolocation. This suggests that the cochlear sys-
dence came from a study of the internal structure of tem of the Eocene bats was not as advanced as that
the fossil skulls. The single most important feature is of modern insectivorous bats, and that they too did
the large size of the basal turn of the cochlea—the not rely exclusively on echolocation. The recently
structure in the inner ear which sorts and processes discovered Eocene bat Onychonycteris (Simmons
sounds by frequency. The basal turn is receptive to et al. 2008) not only has a small cochlea, but lacks
the echoes of very high frequency echolocation calls. other features, present in Icaronycteris and Palaeo-
The size of this basal turn will obviously depend chiropteryx, that suggested to the authors that it
upon the size of the bat, but when these two para- may not even have been able to echolocate. More
meters were plotted against each other, so that the recent work by Veselka et al. (2010) on the structure
size of the basal turn was shown relative to the size of the larynx suggests otherwise—this is an area of
of the bat, Icaronycteris index and Palaeochiropteryx intense research that I will return to in Chapter 3.

Figure 1.8 Cochlear size and echolocation.

The relationship between the width of the
cochlea and skull width for a sample of 6
Eocene and 286 extant species. (Based on
Habersetzer and Storch 1992.)
10 BATS

Note in Fig. 1.8 that there is little overlap between our hypothesized ancestral bat, since they have an
the megabats and microbats. All microbats use extensive gliding membrane and webbed fingers.
echolocation at least some of the time, even those They are not lemurs, but dermopterans, and
that feed on fruit and nectar. Amongst megabats, they do not fly, but glide. The two living species
only a few species from the genus Rousettus echo- are the only dermopterans. Colugos can glide well
locate, using a very different and probably less in excess of a 100 m, and move their limbs around
capable method. Most bats generate their echoloca- to turn and change altitude. In one recorded glide
tion calls in the larynx, but Rousettus do it by click- of 136 m, an individual lost only 11 m in altitude.
ing their tongues. Why the difference? If the oldest They are herbivores the size of a small cat and
fossil bats are echolocators, why don’t the mega- their adaptations to gliding have left them clumsy
bats use echolocation? It seems remarkable that not on the ground. For a long time they were assumed
only may they have lost the ability to echolocate, to provide a useful illustration of how bats might
but that they also appear to have lost the anatomi- have evolved. Someone then suggested that they
cal and physiological adaptations which made it were not just a useful illustration, but perhaps real
possible. In the 1980s this observation led scientists ancestors to the megabats.
to ask the question: did all bats really have a com- The earliest megabat is Archaeopteropus transiens,
mon ancestor? The question had been asked before, which dates back 35 million years to the Oligocene
but only at this point did the tools exist to address it (36–25 million years ago) and was found in Venetia,
in depth. Italy (Dal Piaz 1937). Until the 1980s it was widely
accepted that all bats had a common ancestor. How-
The evolution of bats: a recent debate ever, it was then suggested that the megabats
evolved independently, following a quite different
and a new consensus
evolutionary line to that of the insectivorous micro-
The last 20 years have seen the birth and death of a bats. A strong body of evidence in support of this
major controversy in bat evolution and phylogeny. idea was put forward, triggering a very vigorous
It would be simpler and easier to pass over this and controversy and new research. It is worth discussing
describe the current state of affairs, but we would in some depth for several reasons. First, to dispel the
miss out on an interesting and informative debate, common idea that taxonomy plays little part in
so we’ll start at the beginning and see how the modern biology. Second, to show that answers to
investigations unfolded. The modern colugos, or many important biological questions only come
flying lemurs (Fig. 1.9), resemble in some respects with input from a wide variety of fields–biology is

Figure 1.9 A gliding dermopteran (flying

lemur).
 EVOLUTION AND DIVERSITY 11

a truly multidisciplinary subject. Finally, because of Pettigrew (1986) found it in megabats and Dermop-
the implications of independent origins for micro- tera. He argued that it was highly unlikely that
bats and megabats—principally that flight, the most either of these two patterns could have evolved
anatomically and physiologically specialized and from the other, suggesting that microbats and mega-
demanding mode of locomotion, evolved twice in bats are not at all closely related, but that megabats
the mammals, and that the striking similarities were in fact ‘flying primates’. It is interesting to note
between microbats and megabats are the result of at this point that in 1758, after studying megabats,
convergent evolution. We have already noted the Linnaeus originally classified bats as primates.
absence of echolocation and the small cochlea in Subsequent evidence which showed that microbats
megabats. What other evidence was put forward in were not related to primates resulted in all bats
support of diphyly (independent origins) in bats, being reclassified.
and what evidence is there for monophyly (common Pettigrew et al. (1989) went on to show, through a
origins)? cladistic analysis of some 24 different characteristics
This debate had in fact been around for some time of the nervous systems of 14 mammalian species,
(see for example Jones and Genoways 1970, Smith that microbats appeared very early in mammalian
and Madkour 1980) but it gained momentum with evolution, and that their nearest relative among the
the publication of a paper (Pettigrew 1986) which mammals studied was the tree-sloth Bradypus! Mega-
suggested that the pattern of neural connections bats appeared to have evolved much later, from an
between the mid-brain and the retinal cells of the early primate branch, around the same time as the
eyes were very different in microbats and megabats. Dermoptera. The cladogram is shown in Fig. 1.11.
All neurones in the right superior colliculus (s.c.) A cladogram is a form of evolutionary tree which
of the mid-brain appeared to project to the retinal links species according to shared, derived character-
cells of the left eye of a microbat and those of the left istics. That is, those characteristics that are not
s.c. to the right eye (Fig. 1.10). This pattern has been ancestral to all of the species in the group being
found in all mammals except primates, and is studied, but evolved after their divergence from a
believed to be the ancestral mammalian pattern. In common ancestor. The more characteristics shared
contrast, neurones from the right s.c. of a megabat between two species, the more closely related they
project to both eyes, but only to the left half of the
visual field. Neurones from the left s.c. projected to
the right half of the visual field of both eyes. This
pattern was thought to be unique to primates, but

visual field

retina

optic chiasma

superior
colliculus

left right left right

Figure 1.11 Cladogram based on the analysis of 24 neural characters. The
Figure 1.10 Simplified diagram of the primitive/microbat (left) and megabats Pteropus and Rousettus appear to be closely related to the primates
primate-like/megabat (right) connections between the mid-brain (superior and dermopterans and are separated from the microbats, Macroderma and
colliculi) and the eyes (Based on Pettigrew et al. 1989.) Mormopterus, by several other taxa. (Based on Pettigrew et al. 1989.)
12 BATS

Box 1.1 Cladistics

The philosophical framework of modern taxonomy is based fossils. The technique therefore has limited practical value
on the concepts of evolutionary theory. Cladistics was in many cases.
developed in an attempt to introduce a set of more objective 3. Embryological evidence. It is assumed that in the embry-
and rigidly applied rules than those used in traditional ological development of a group of species, the general,
evolutionary taxonomy (Hennig 1966). In cladistic analysis, ancestral characters appear before the more specialized,
given a group of animals to classify, the first and most crucial derived characters. This assumption is certainly not valid
task is to sort out which forms of a particular character are all of the time. The big debate concerns the frequency
ancestral and which are derived. There are three lines of with which it is valid. It is perhaps safest to say that any
evidence, but it is rare to be able to use all three in any evidence drawn from embryology should be used with
particular case. some caution.
1. Outgroup analysis. A comparison of characters with those Given the difficulties, it is clear that cladistics must use all
of another species, or group of species, which is known of the techniques at its disposal and cross-reference
not to belong to those under study, but which is relatively wherever possible. The use of unrooted trees can help
closely related. Any shared characteristics are by definition greatly. An unrooted evolutionary tree indicates the
ancestral. It sounds fine in principal, but it requires prior relationships within a group of animals, but does not indicate
knowledge of the relations between groups. This has led the order of their evolution. Unrooted trees can be
some people to criticize the technique as circular. This is constructed first on morphological or biochemical evidence,
an unfair simplification, and it would be more realistic to and rooted later, if key evidence can be found. Once a root
view it as part of an iterative process—constant refine- has been found, the direction of the tree’s ‘growth’ is known,
ment of the model to arrive at an answer that best fits the and the evolutionary relationships between the species can
known facts—a perfectly valid scientific technique. be resolved.
Another criticism is that the method assumes that there is Finally, in constructing cladograms, the various methods
no evolutionary convergence, or at least that it is rare. The and computer programs used are generally based on the
problem is one of separating homologous structures, that is concept of parsimony. The tree that uses the least number of
those with a common origin, from analogous structures, evolutionary steps and the smallest number of assumptions
which are derived from different parts of the body but serve to explain the data is sought. This is not necessarily the way
a similar function. Good examples of convergent evolution, nature works, it is simply a practical scientific approach—
and analogous structures, are the wings of birds and that of Occam’s razor. Other approaches are constantly being
insects and the eyes of vertebrates and cephalopods. developed, such as maximum likelihood methods, and these
Convergence probably is rare, but when it does occur it will may lead to different conclusions. Good basic accounts of
lead to errors in cladistic analyses: but then it can upset cladistic theory and methods, and the debates surrounding
traditional methods too. them, can be found in Ridley (1986) and Patterson (1987)
2. Palaeontological evidence. A good fossil record will pro- and for a more recent practical guide you might look at Forey
vide good evidence of which characters are ancestral and et al. (1998). Whatever you read, it will be out of date unless
which are derived. However, fossil records are usually full you go to the primary literature, since this field changes
of gaps, and many characters are simply not preserved in constantly as new methods and programs are developed.

are. A cladogram is constructed by computing the evolutionary biology by the traditional morpholo-
evolutionary tree which uses the least number of gist and, as we will see below, the molecular
evolutionary steps to explain the different degrees taxonomist.
of relatedness between the species. Cladograms can The argument that primates, dermopterans, and
be very persuasive, but have their problems. For megabats are closely related, and that microbats
those interested, Box 1.1 takes a closer look at cla- are unrelated to megabats, was strengthened by
distic techniques, since they are now widely used in additional lines of evidence, from factors as simple
 EVOLUTION AND DIVERSITY 13

as a consideration of body size ranges, to the anal- drive the last nails into the coffin of the ‘flying
ysis of the amino-acid sequence of haemoglobin. primate’ hypothesis. I will come to that later, but I
Table 1.1 lists some of the differences between first want to show how we got to that point. In the
megabats and microbats not discussed in the text early stages of the debate a wide range of ap-
(Pettigrew et al. 1989). At the time this was coun- proaches were used to provide evidence for both
tered by the evidence in Table 1.2, which lists sides of the argument. Some of this is presented
features common to mega- and microbats, but below, to show how the controversy progressed
not found in other mammals. It was these known and how persuasive both sides could be. In the
differences that led to uncertainty, and started the end, most of the evidence in favour of diphyly was
controversy. undermined, but some of the issues raised are still
The research that was prompted by this work debated.
goes on still, but in the eyes of most scientists One feature that was used persistently to argue
the debate is over. It is now widely accepted that for a close relationship between megabats and mi-
bats are monophyletic and researchers are concerned crobats was the similarity of their wings. The first
with other evolutionary questions. The primary tool part of Table 1.2 lists just a few of those cited by
is molecular genetics, and one of its first uses was to Baker et al. (1991): the wings of megabats and

Table 1.1. Some of the differences between megabats and microbats not covered in detail in the text. Discussed by Pettigrew et al. (1989).

MICROBATS MEGABATS

distribution worldwide palaeotropical (Old World tropics)

orientation primarily by echolocation, all species generate sonar pulses in the primarily visual, tongue clicking orientation sounds produced
larynx by a few species only
diet ancestral insectivores, a small minority have evolved to feed on fruit, nectar, and pollen
fruit, nectar and pollen, vertebrates, and blood
teeth W-shaped cusps, or evidence for past possession of such teeth simple, no evidence of W-shaped cusps (i.e. of insectivorous
ancestry)
eyes simple retinal blood circulation; tapetum lucidum (reflective layer complex retinal blood circulation; tapetum lucidum often
behind receptor cells) rarely present; ganglion cell streak below present; ganglion cell streak above optic disk; eyes open
optic disk; eyes open after birth before birth
ears pinna (external ear) often complex, margin incomplete; tragus pinna simple, margin complete to form a tube; tragus never
(cartilaginous projection) often present inside pinna; Paaw’s present; Paaw’s cartilage absent; size of cochlea closely
cartilage in middle ear cochlea (sound reception and processing related to size of bat; extra turn absent; cochlea in contact
apparatus in inner ear) variable in size and often large; cochlea with skull
has large, extra basal turn for high frequency sound reception;
cochlea acoustically isolated from skull
limbs metacarpals (palm bones) long in relation to first phalanges (finger metacarpals and phalanges similar in length; opposable
bones); thumb and forefinger have minimal independent thumb and mobile forefinger; forelimbs move together,
mobility; limbs move independently, many species very agile on movement is slow and clumsy
the ground
skin hair erector muscles are striated (like skeletal muscle) hair erector muscles are smooth (like those of internal organs)
penis corpus spongiosum not enlarged to form glans penis corpus spongiosum enlarged to form glans penis
torpor widespread in two families and highly developed poorly developed, and only found in nectar feeders
roosting posture neck extended (head bent towards back)—neck vertebrae specially neck flexed (bent towards chest)
adapted
threat behaviour primarily acoustic, wing spreading not seen often involves wing spreading and other visual threats
central nervous inferior colliculus (auditory centre) larger than superior colliculus superior colliculus larger than inferior colliculus; primate-like
system (visual centre); primitive pathway between eye and brain; pathway between eye and brain; forebrain well developed
forebrain usually less well developed than hindbrain; low as cerebral cortex; low frequencies map at front of auditory
frequency sounds map at rear of auditory cortex; motor cortex cortex; motor cortex shows primate-like arrangement of
shows primitive arrangement of cortico-spinal areas; hindlimb is cortico-spinal areas; hindlimb is represented by a large area
represented by a small area of somatosensory cortex of somatosensory cortex
14 BATS

Table 1.2. Some of the characteristics shared by microbats and megabats. is under very different functional constraints. The
For detailed coverage, see Baker et al. (1991). dermopteran Cynocephalus, and the oldest megabat
Anatomical features common to microbats and megabats fossil, Archaeopteropus, fall among the megabats, as
do primates. The ratio for the forelimb is very vari-
Occipito-pollicalis muscle along the leading edge of the wing able in microbats, but shows little variation in mega-
Fingers 2–5 of forelimb greatly enlarged
bats: but then all of the latter belong to a single
Claws restricted to digits 1 or 1 and 2
Hindlimbs rotated 90  outward, i.e. knee directed to the side family.
Calcar present on foot By the same argument against functionally impor-
Head of the femur aligned almost parallel to the shaft tant morphological characters, the muscles of the
Premaxilla greatly reduced wing should also be excluded from the list of sup-
Jugal greatly reduced
porting evidence for monophyly. The specialized
Several anatomical features of the middle ear
Anatomy of preplacenta and placenta occipito-pollicalis muscle complex (which controls
Somatosensory map of forelimb reversed relative to other mammals the shape of the leading edge of the wing) was cited
as evidence for monophyly (Wible and Novacek
1988), but it is present not only in microbats, mega-
microbats certainly do look very similar. However, bats, and dermopterans, but also in the clearly unre-
this is to be expected since there is considerable lated flying squirrels (Johnson-Murray 1977) and
similarity in mode of flight, and therefore in the birds (Raikow 1985). Thewisen and Babcock (1991)
evolutionary pressures moulding wing shape. Fly- presented us with an interesting twist to this story.
ing is not easy. There are few ways in which the They looked not at the muscle complex, but at its
vertebrate body plan can be adapted to meet the pattern of innervation. If the occipito-pollicalis mus-
intolerant energetic, mechanical, and aerodynamic cles are truly homologous rather than analogous
demands of flight. Pettigrew and his colleagues they should have similar innervation patterns due
argued that wing morphology in fact yields evi- to the close link between the development of a mus-
dence for the independent evolution of megabats cle and its nerve supply. (Homology: derived from
and microbats, and for the close relationship the same ancestral muscles; analogy: convergent
between megabats, primates, and dermopterans. evolution leading to similar structures derived
The hypothesis put forward was that the relative from unrelated muscles). Thewisen and Babcock
lengths of the bones of the third and fourth fingers found that features of the innervation pattern were
are not going to be important in determining flight quite unique among mammals, and common to mi-
performance, and are therefore not subject to the crobats, megabats, and dermopterans: good news
evolutionary pressures of flight. (Finger 5 is known for the proponents of monophyly.
to play an important role in altering wing camber
and was excluded from the analysis.) We might
Molecular studies
expect these fingers to be similar in all bats, or at
least for no clear pattern to emerge between groups, Early molecular evidence in taxonomy came from
if all bats are closely related. immunological studies, to be followed by protein
This was not found to be the case. In Fig. 1.12 the sequencing, and finally the sequencing of DNA
ratios between the length of the metacarpals and the itself. Like the morphological evidence, it has been
first phalanges of fingers 3 and 4 for a large number controversial. Several early studies (for example
of megabats and microbats have been plotted Ammerman and Hillis 1992, Bailey et al. 1992)
against the forearm length of each species. There is came down firmly on the side of monophyly, but
no overlap between the data for microbats and Pettigrew (1994, 1995) questioned the validity of
megabats: the microbats have proportionally longer some of the base alignments which lead to this con-
metacarpals. Can it be argued that the wings of clusion, and raised another complication. He argued
megabats and microbats are so very similar? The that megabat DNA was rich in the nucleotide bases
relationship holds true for the hindlimb too, which A-T at the expense of G-C. It was suggested that an
 EVOLUTION AND DIVERSITY 15

(a)
digit 1
1
metacarpel
metacarpel

2 2
phalanx 1
phalanx 1
5
5 4
3
4
3
megabat microbat

(b)
12

10
metacarpophalangeral index

MICROBATS
Icaronycteris
8

6 Archaeopteropus
Cynocephalus
4

2
MEGABATS

0 60 120 180
forearm length (mm)

Figure 1.12 Forelimb morphology in bats. (A) The forelimb skeletons of megabats and microbats. (B) The ratio of metacarpal length/first phalanx length
(for fingers 3 and 4), plotted against forearm length for a large sample of microbat and megabat species. A fossil microbat, Icaronycteris, and megabat,
Archaeopteropus, and the dermopteran Cynocephalus are also included. (Based on Pettigrew et al. 1989.)

A-T content of more than 70 per cent, as was found based on four nuclear and three mitochondrial
in megabats, raised doubts about the validity of genes (a total of over 8,000 base pairs). Sixteen bat
sequencing evidence, which assumes A-T and G-C species were used, with representatives from all
are equally abundant. More recent studies did not microbat superfamilies, megabats, and four out-
find a high A-T bias in the genes sequenced and groups (human, flying lemur, dog, and mouse).
used methods that took into account the moderate After rigorous analysis using the latest approaches
heterogeneity in base composition that was observed. and tools (see Box 1.2) they were able to reject the
A major turning point came with the first of a series ‘flying primate’ hypothesis and found no evidence
of papers by Teeling and colleagues (Teeling et al. to suggest that the Dermoptera were a sister group
2001). They carried out a phylogenetic analysis to the bats. If this was not significant enough, they
16 BATS

came up with an even more startling result. The know more and includes some useful references.
more observant reader will have spotted it already Some of the difficulties revolve around the cladistic
in earlier figures: the bats in the superfamily Rhino- problems discussed earlier, but there are others
lophoidea are more closely related to the megabats unique to the molecular approach. As laboratory and
than they are to other microbats. The implications of analytical methods improve, molecular taxonomy be-
this are as profound as those of the ‘flying primate’ comes an ever more powerful approach and the rate at
hypothesis. If true, then either the echolocation sys- which new data are published continues to accelerate.
tems of the rhinolophoids evolved quite indepen- So, monophyly wins the day. At the height of the
dently of those in other microbats, or they were lost monophly–diphly debate Pettigrew et al. (1989) pre-
in the evolution of the megabats. There are plausible sented four scenarios for the relationships between
explanations for both. Let’s start with the possibility microbats and megabats (the fallen angel, the deaf
that microbats evolved echolocation more than fruit bat, the blind cave bat, and the flying primate).
once. Echolocation is not unique to microbats: ceta- For each, they suggested ways in which the situa-
ceans, some cave-nesting birds, and megabats in the tion might have arisen, and discussed the implica-
genus Rousettus use it. Within the microbats, there is tions and difficulties of each. Only two of the four
considerable variation in the anatomical and physi- scenarios justify further discussion.
ological adaptations for echolocation and in call
structure. Some of this variation is specific to partic- The deaf fruit bat
ular groups, and may therefore be the result of Megabats arose from microbats, losing the capacity for
independent origins. Similarities in echolocation echolocation, and the associated anatomical and physio-
systems need not imply monophyly within the logical features, and acquiring a primate-like brain in the
microbats, but convergence due to the functional process. Recent work has questioned the evidence for
constraints of echolocation. Alternatively, could a primate-like brain (for example Ichida et al. 2000)
echolocation have evolved only once in ancestral and I have already suggested some reasons why the
bats and the megabats have subsequently lost the loss of echolocation may have occurred. An addi-
ability? The evolution of exceptional night vision in tional reason may be that since echolocation re-
megabats alongside their shift to a vegetarian diet quires an extensive processing centre in the cortex
may have reduced the need for echolocation. Fur- of the brain, perhaps there is not the space for this
thermore, echolocation may limit the maximum size alongside a large visual centre. However, some of
of bats, since the emission of echolocation calls is the differences listed in Table 1.1 remain and require
coupled to the respiratory and wingbeat cycles to explanation: I suspect few have been fully investi-
reduce energy costs. As bats get bigger and wing- gated. For example, if the differences in metacarpal/
beat frequency falls then call emission frequency phalanges indices are representative of all bats,
also falls, reducing the rate at which the bat receives what underlies them?
information. If megabats are going to be big, then
maybe they can’t echolocate. Rousettus is an inter- The flying primate
esting exception, since it echolocates by clicking Microbats evolve. Megabats evolve independently on an
its tongue and is probably not constrained by the early branch of the primate line. Is it likely that flight
wingbeat in the same way. I will come back to the could have evolved twice in the mammals, and that
evolution of echolocation in Chapter 3. Clues to megabats and microbats could have undergone
this new-found relationship between megabats such striking convergent evolution? As pointed out
and rhinolophoids were there to be seen—for earlier, gliding has evolved many times in the verte-
example, it is consistent with the exclusively Old brates and three times in the marsupials alone
World distribution of both groups. (Archer 1984). Powered flight almost certainly
I have glossed over the methods that underpin this evolved from gliding in all living and extinct ani-
new view of bats. Box 1.2 says a little about the com- mals (see Chapter 2), so its independent origin in
plexities of phylogenetics for those who would like to microbats and megabats is a reasonable possibility.
 EVOLUTION AND DIVERSITY 17

Box 1.2 Phylogenetics

The principles: the research community and improved, and new methods
and programs are made available at frequent intervals. Any
1. To identify homologous DNA sequences in a group of
attempt to explain the theory behind even one or two of
animals. That is, identify the DNA sequence or gene
them would require a large box, so if you want to know more
responsible for producing a particular protein, which has
try the following references: Avise (2004, 2006), Felsenstein
essentially the same function in all of the organisms under
(2004), and Page and Holmes (1998).
study. More recently, analyses have been carried out
Molecular taxonomy has a major advantage over many
based on non-coding homologous sections of DNA.
morphological methods: the degree of difference between
2. Determine the nucleotide sequences of these
homologues can be quantified in terms of the differences in
homologues. The greater the number of differences
their nucleotide sequences. Morphology is a complex and
between sequences (due to base substitution, insertion,
poorly understood expression of these molecular differences
or deletion), the more distantly related the organisms
that cannot be readily quantified. As is usually the case in
will be.
biology, there are a number of flies in the DNA soup. Some of
3. Determine which is the ancestral form, and the paths by
the most important difficulties relate to the identification of
which each derived form evolved. In other words, construct
homologous DNA sequences—the crucial first step in the
an evolutionary tree, usually by cladistic analysis.
process. Fortunately, pitfalls are increasingly easy to avoid as
The most common analysis methods involve maximum the genomes of a growing number of organisms are
parsimony and maximum likelihood approaches and these progressively mapped and made public.
and a variety of other methods (for example distance matrix, Different genes appear to undergo mutations at different,
Quartet, Bayesian inference) can be accessed with varying but remarkably constant, rates. Functionally important DNA
degrees of user-friendliness in a range of software packages, changes slowly: any deleterious mutations, which impair the
many freely downloadable from the Internet. Joe Felsenstein function of its protein products, will be eliminated by natural
at the University of Washington, Seattle, a research leader in selection. Introns mutate more rapidly, and in homologous
this field, maintains a comprehensive and up-to-date website DNA sequences show greater differences between species
of available programs (http://evolution.genetics.washington. than their functional exons. This may complicate things for
edu/phylip/software.html). It is common practice to use the evolutionary biologist, but it also provides a powerful
several approaches and several programs in a single study, tool—molecular clocks to measure the time course of
with additional programs often being used in data evolution itself—assuming each clock can be calibrated. So,
preparation, assumption testing, and so on. The ready it is not only possible to identify evolutionary branch points,
availability of these programs has been a major factor in the but also the lengths of individual branches, as shown for
growth of this field. Programs are constantly evaluated by example in Fig. 1.5.

What about the similarity of form? Given a five- Many of the assumptions underlying Pettigrew’s
fingered mammalian forelimb, just how many hypotheses have been questioned from the start
ways could it develop into an aerodynamically func- (Baker et al. 1991, Simmons et al. 1991), but the
tional wing? Probably not many, and there are a debate has been interesting and exciting. Even
number of well-known and striking examples of though the flying primate hypothesis has proved
convergent evolution to lend credibility to the idea: to be wrong, as Pettigrew (1991b) himself said ‘it
the eyes of vertebrates and cephalopods, the gills of will still have been a most fruitful, wrong hypothe-
fish and cephalopods, and the hydrodynamically sis’. It has raised interesting questions, prompted
efficient body form of fast pelagic swimmers new research, unlikely collaborations, the applica-
among cetacea, teleost fish, and sharks. However, tion of new techniques to the study of bats, and
an overwhelming body of molecular and morpho- made us question long-held beliefs. That is what
logical data rejects this hypothesis. keeps research buzzing.
18 BATS

Classification and the distribution and Before taking a more detailed look at bat phylog-
diversity of bats eny, let’s look at the global distribution of bats.
Among mammals, only humans (and some of the
A brief tour of modern bat families mammals closely associated with humans) are more
Bats are the most widely distributed and (by spe- widely distributed than bats. Bats are found every-
cies) the second most numerous group of mammals, where except the highest latitudes, the most inhos-
outnumbered only by the rodents (almost 2,300 spe- pitable deserts, and the most remote islands. In
cies). Precisely how many species of bats there are in common with all other forms of life, the number of
the world is uncertain. The uncertainty is due in part species declines away from the equator, although
to the difficulties of defining a species: when does a the pattern is disturbed by geographical features
sub-species become a new species, and so on? What such as the Sahara Desert. The approximate num-
is certain is that the number of species is increasing, bers of species in different regions of the globe are
through the recognition of cryptic species within shown in Fig. 1.13. The neotropics of South and
known species using molecular approaches, and Central America is the richest area, with over 200
through the discovery of distinctly new bats. When species, followed by the palaeotropical regions of
writing the earlier version of this book 15 years ago Asia and then Africa. Temperate regions are
all sources suggested there were fewer than 1,000 impoverished by comparison, with about 40 species
species (for example Hill and Smith (1984) sug- in North America (nearctic) and 85 or so in northern
gested 966 and Findley (1993) opted for 963 species). Eurasia (palaearctic). A number of species of the
The current tally is 1,116 (Wilson and Reeder 2005), family Vespertilionidae spend the summer in high
but that does not include new species that I may latitudes in both hemispheres, and a few are found
well mention before the end of the book. Bats there- north of the Arctic Circle. The Old World fruit bats
fore account for 20 per cent of the approximately have island-hopped halfway across the Pacific from
5,420 species of mammals. the western rim, and on some islands rare, endemic

0 0

20

20
40
40 20
60
60
80
100 80
120
100 100
80 60
60 40
20
40
60
20 20
40

Figure 1.13 The number of bat species in 500 km2 quadrats in different parts of the world. (Based on Findley 1993.)
 EVOLUTION AND DIVERSITY 19

species have evolved. The vespertilionid hoary bat, we accept the Miniopteridae, 18 families across
Lasiurus cinereus, has colonized the Galapagos and these lineages.
Hawaiian Islands. New Zealand’s only native mam- Figure 1.14 summarizes the chiropteran family
mals are the single remaining species of the endemic tree down to the sub-family level, with the number
family Mystacinidae, and another member of the of genera and species in each. It will undoubtedly
well-travelled vesper bats. Bats probably evolved change before this book is even published! The clas-
in the tropics, and almost 80 per cent of all species sification used is primarily that given by Teeling et
are tropical. Most families are found only in the al. (2005). The numbers of sub-families, genera, and
tropics: just two, the Vespertilionidae and Rhinolo- species are taken from Wilson and Reeder (2005).
phidae, contain truly temperate species. Species are not evenly distributed across families,
All bats are included in the order Chiroptera with four (Pteropodidae, Molossidae, Vespertilioni-
(meaning hand-wing), but new molecular phyloge- dae and Phyllostomidae) containing 834 of the
nies (for example Teeling et al. 2005, Eick et al. 2005) 1,116. Distribution is also very uneven within fa-
do not support the traditional division into the sub- milies and even genera. For example the Vesperti-
orders Megachiroptera and Microchiroptera. The lionidae are dominated by the Vespertilioninae and
megabats all belong to the same family, the Ptero- Myotinae and the genus Myotis. What follows is a
podidae, the Old World fruit bats or flying foxes, thumbnail sketch of each family. The major external
and now sit within one of the four main microbat features of bats used in the descriptions are shown
lineages. The microbats are distributed in 17 or, if in Fig. 1.15. Photographs of representives of all but

SUPERFAMILY FAMILY SUBFAMILY (genera, species) Genera, species

Pteropodidae 42, 186
Yinpterochiroptera Rhinolophidae 1, 77
Hipposideridae 8, 81
Rhinopomatidae 1, 4
RHINOLOPHOIDEA Craseonycteridae 1, 1
Megadermatidae 4, 5
EMBALLONUROIDEA Emballonuridae Taphozoinae 2, 18 13, 51
Emballonurinae 11, 33
Yangochiroptera Nycteridae 1, 16
Natalidae 3, 8
Molossidae Molossinae 15, 99 16, 100
Tomopeatinae 1, 1
Miniopteridae 1, 19

Vespertilionidae Vespertilioninae 38, 238 47, 388

VESPERTILIONOIDEA Antrozoinae 2, 2
Myotinae 3, 106
Murininae 2, 19
NOCTILIONOIDEA Kerivoulinae 2, 23
Myzopodidae 1, 2
Mystacinidae 1, 1
Thyropteridae 1, 4
Furipteridae 2, 2
Noctilionidae 1, 2
Mormoopidae 2, 10
Phyllostomidae Phyllostominae 16, 42 55, 160
Stenodermatinae 18, 67
Desmondontinae 3, 3
Brachyphyllinae 1, 2
Phyllonycterinae 2, 5
Glossophaginae 13, 32
Carolliinae 2, 9

Figure 1.14 An evolutionary tree for the bats to sub-family level based on Teeling et al. (2005).
20 BATS

one of the families (Furipteridae) are in the colour consistent within a species, and can be easily
plate section. measured to the necessary degree of accuracy in a
live bat. As a general indication of overall size, the
body length of a bat (not including the head) is
Yinpterochiroptera
similar to its forearm length. Megabats weigh any-
Family: Pteropodidae, the ‘megabats’ (186 species in thing from 20 g to 1.5 kg, with wingspans approach-
42 genera). Plates 1–3 ing 2 m. With the exception of several species of the
The Old World fruit bats, or flying foxes, are con- genus Rousettus (Fig. 1.16), megabats do not echolo-
fined to the Old World tropics and feed almost cate, but rely on vision (and smell) for night orienta-
exclusively on fruit, flowers, nectar, and pollen. In tion and the location of food. They generally have
the past several sub-families were recognized but large eyes, simple ears, and simple muzzles (Fig.
they are now generally viewed as a single group. 1.15). The second and third fingers are largely inde-
As a group, they are larger than other bats, but show pendent of each other, and the second has a claw
considerable variation, with forearm lengths of (Fig. 1.15). Tail and tail membranes are typically
40–220 mm. The length of the forearm is relatively small or non-existent. Although strong flyers, they

tragus

Lissonycteris angolensis Otonycteris hemprichi

thumb

forearm Pteropus poliocephalus

thumb
Myotis daubentonii

Craseonycteris Sturnira sp.

thonglongyi

tail membrane calcar

Figure 1.15 Some important features of the external
Notopteris macdonaldi Natalus sp. anatomy of bats.
 EVOLUTION AND DIVERSITY 21

Figure 1.17 Tube-nosed bat, Nyctimene species.

Figure 1.16 Egyptian fruitbat, Rousettus aegyptiacus.

lack the wide variation in wing form and flight style

seen in other families, and they lack their complex
shoulder anatomy. Most have rather dog-like faces,
hence the name flying foxes. Skull and teeth are
adapted to their particular feeding habits. Skull
and jaws are typically large and strong with large
muscles to deal with tough-skinned fruit. The shape
of the head often reflects the need for large areas for
muscle attachment and large biting forces: this often
means short, heavy jaws, and a large braincase.
Molar and premolar teeth have large, and relatively
flat, grinding/crushing surfaces, and the canines are
large to grasp and penetrate tough fruit. The incisors
are usually small. Many species crush fruit to extract
the juice and spit out the pulp. Nectar and pollen
feeders have lighter skulls and their jaws are often Figure 1.18 Masked fruit bat, Pteropus personatus.
elongated and narrow with correspondingly smaller
teeth. This allows them to probe deeply into flowers head and shoulders, often found with tufts of long
and, with the aid of long tongues, extract nectar. The hair and associated scent glands—features related to
male hammer-headed bat, Hypsignathus monstrosus, courtship behaviour. The face itself may have white
has a huge larynx and snout, used to generate loud or coloured markings (Fig. 1.18). A number of spe-
and spectacular hoots, as part of a display to attract cies have wing membranes mottled with white, yel-
passing females to a lek (group of displaying males). low, or shades of red.
The tube-nosed bats (Nyctimene, Fig. 1.17) have lat- Most species roost in trees, in large and occasion-
erally directed, tube-like nostrils, of uncertain func- ally noisy colonies. Tree colonies, or camps, may
tion. They are said to stretch and vibrate when the number several thousand individuals. Many spe-
bat makes audible calls in flight. Similar but less cies, often the smaller megabats, roost in small
pronounced nostrils are found in Murina species groups or even singly and these are the most likely
(Vespertilionidae). to have cryptic colouration or patterning. A signifi-
Although many are unexciting shades of brown, cant minority of species roost in caves, usually
some are patterned and/or brightly coloured. Males close to entrances. Cave roosts of echolocating Rou-
of some species have bright colouration around the settus bats can contain thousands of individuals.
22 BATS

Pteropodids will often forage in large groups, hav- Rhinolophidae and Hipposideridae are used to
ing located a flowering or fruiting tree or stand of detect fluttering prey.
trees. Roost sites may be 50 km or more from a
particular feeding site, and the colony will take Family: Rhinolophidae (77 species in 1 genus). Plate 4
flight before dusk to travel to it. Some species Widespread, insectivorous Old World bats rang-
undergo seasonal migrations, following the fruiting ing from Britain and the whole of Africa in the
season of particular species of tree. west, to Japan and eastern Australia in the east.
All remaining bats fall into the old sub-order Micro- Characterized by the most complex noseleaves
chiroptera. Although now dissolved, it still makes some and large, mobile ears (Figs 1.19, 1.20). The nose-
sense to provide an overall description of this large and leaf is very variable in form and complexity (spe-
very diverse group before looking at the different lineages. cies can often be identified by it), and it serves as
Forearm length ranges from 22 to 110 mm. All species an acoustic lens, focusing the nasally emitted
have the ability to orientate by echolocation, and most echolocation pulses. Rhinolophids have evolved
make full use of it. The ears are often large and complex constant frequency, flutter detecting, echolocation
and may contain a tragus (Fig. 1.15), and many species pulses, and use Doppler shift compensation. They
have noseleaves: all features associated with echolocation. are broad winged bats that commonly forage in
They generally have small eyes, but their visual capability cluttered environments, feeding by gleaning (tak-
is often found to be very good—none are blind. Some ing prey off foliage and the ground), aerial hawk-
species often locate their prey without echolocation, by ing close to the ground, and ‘flycatching’—
listening for prey-generated sounds, and/or by using making sorties from a perch after passing insects.
vision. These species have very large eyes and excellent The majority are tropical, but some can tolerate
vision. The eyes are often forwardly directed, and proba- low temperatures by becoming torpid, a feature
bly give good binocular vision. Wing and tail membranes rare in tropical bats. About 15 species have colo-
vary enormously in shape and size, and with echolocation nized temperate regions of Europe and Asia and,
characteristics that are related to prey type, foraging like many vespertilionids, are truly heterothermic
strategy and habitat, although it should be said that (that is they are true physiological hibernators).
most species have flexible feeding habits. They probably Rhinolophidae are predominantly cave roosting,
evolved from an early insectivore and their teeth reflect colonial or solitary, often in association with Hip-
this ancestry, despite their diversification into a wide posideros species.
range of feeding habits. The chewing teeth have W-shaped
cusps for cutting and crushing food. Skull and jaw shape Family: Hipposideridae (81 species in 8 genera). Plate 5
is very variable and determined primarily by the food The leaf-nosed bats have sometimes been included
taken, but in some cases by roosting habit. Their enor- in the Rhinolophidae, and there is no doubt that
mous diversity can really only be dealt with in a family by they are closely related. They are found throughout
family account. the Old World tropics and subtropics. They have
elaborate noseleaves but generally not as complex
as those of the Rhinolophidae. They are primarily
Superfamily: Rhinolophoidea
colonial or solitary cave roosters and have similar
Five Old World families of insectivorous or carnivo- feeding habits to the Rhinolophidae. Figure 1.21
rous bats with complex noseleaves make up this shows Hipposideros caffer, some populations of
group: the Rhinolophidae, Hipposideridae, Rhinopo- which have rich, golden fur.
matidae, Craseonycteridae, and Megadermatidae.
Most have short, broad wings for slow manoeuvrable Family: Rhinopomatidae (4 species in 1 genus). Plate 6
flight, and many often forage by hovering, gleaning, The small, insectivorous mouse-tailed bats (with
or by hawking from perches (flycatching). Wing their distinctive, long free tails, Figs. 1.22, 1.23) are
loading (body mass/wing area) is low in most, found in the arid and semi-arid regions of the Old
enabling them to carry heavy prey. The long, con- World, from North Africa east to Thailand and Su-
stant frequency echolocation pulses emitted by many matra. Possible adaptations to their desert life are
 EVOLUTION AND DIVERSITY 23

Figure 1.19 Lesser horseshoe bat, Rhinolophus hipposideros.

Figure 1.20 Lesser horseshoe bat, Rhinolophus hipposideros.

the valved nostrils, which may prevent the entry of name suggests, they have very long tails, which
dust, and kidneys modified for the production of project well beyond the end of the tail membrane.
concentrated urine to limit water loss. They are colo-
nial roosters in caves and man-made structures. In Family: Craseonycteridae (1 species in 1 genus). Plate 7
the cooler months they may remain dormant for Discovered in 1973 in Thailand, the bumble-bee bat,
long periods, living off accumulated fat. As the Craseonycteris thonglongyi, is the smallest known bat
24 BATS

Figure 1.21 Sundevall’s leaf-nosed bat, Hipposideros caffer.

Figure 1.23 Lesser mouse-tailed bat, Rhinopoma hardwickei (drawing

by Lucy Hammond).

1993). It is rare and under some threat due to defor-

estation of its habitat in Thailand, but its recent
discovery in Myanmar is encouraging.

Family: Megadermatidae (5 species in 4 genera).

Plates 8 and 9
The Old World false vampire bats (so-called
Figure 1.22 Lesser mouse-tailed bat, Rhinopoma hardwickei. because, like many bats, they were once thought
to feed on blood) are large and spectacular. Macro-
derma gigas, the Australian ghost bat (Figs. 1.24,
(1.5 g, forearm length 22–26 mm), and arguably the 1.25), has a forearm length of 105 mm, a wingspan
smallest known mammal (Hill 1974). It roosts high of 0.6 m, and weighs up to 150 g. Megaderma lyra,
in the ceilings of limestone caves. It feeds by aerial the Asian greater false vampire has attractive blue-
hawking in open spaces close to its roosts, using grey fur, Macroderma gigas is more or less white,
calls of high frequency and intensity (Surlykke et al. and Lavia frons from Africa, has blue-grey fur in
 EVOLUTION AND DIVERSITY 25

addition to its yellow wings. Lavia frons frequently

flies in daylight. Most are partly carnivorous, feed-
ing on a range of small vertebrates, including fish,
frogs, reptiles, birds, rodents, and other bats, as
well as large invertebrates. These bats have a for-
midable array of senses for detecting prey, which
may be responsible for their very adaptable hunt-
ing strategy. Their echolocation pulses are short,
low in intensity, and broadband (sweeping a
broad range of frequencies)—effective in cluttered
environments, but not easily detectable by prey.
The large ears are sensitive not only to the weak
echoes returning from these pulses, but also to
prey-generated sounds. The bats need not use
echolocation under all circumstances, but can sim-
ply listen. Finally, they all have well-developed,
forward-pointing eyes, and can locate prey visu-
ally. They hunt both by hawking from perches,
and by flying low over the ground. Prey are usu-
ally taken to a favourite perch, and the discarded
remains beneath these perches reveal a lot about
the bats’ diet. Macroderma gigas has been known to
take prey more than 60 per cent its own body
Figure 1.24 Australian ghost bat, Macroderma gigas. weight. They roost in small groups in caves, rock
crevices, and hollow trees.

Figure 1.25 Australian ghost bat, Macroderma gigas.

26 BATS

Yangochiroptera
Superfamily: Emballonuroidea
Composed of two families of aerial insectivores: the
Emballonuridae and the Nycteridae.

Family: Emballonuridae (51 species in 13 genera).

Plates 10 and 11
These are the sheath-tailed bats of the Old and
New World tropics. They are currently divided
into two sub-families: the Emballonurinae (11
genera, 33 species) and the Taphozoinae (2 genera,
18 species, Fig. 1.26). Emballonurids have long nar-
row wings, typical of fast flying, aerial insectivores.
The wings are so long that at rest they have one
extra fold relative to other bats. The tail is loosely
bound to the tail membrane, and curls over the top
surface when the bat is roosting. They are some-
times called sac-winged bats, because of the phero-
mone-producing glands present near the shoulder
in some New World species. These sac-like glands
are usually larger in the male. The four Diclidurus
species are white or grey/white, with pink wings,
ears, and face, and are known as ghost bats. Sac- Figure 1.27 Large slit-faced bat, Nycteris grandis.
copteryx species have two thin white stripes run-
ning the full length of the back, but are otherwise
brown, like most other species. Stripes are found Some species have pale grey and yellow pattern-
on bats of several unrelated groups (for example ing, to camouflage them against lichen covered
the Phyllostomidae), and are thought to be cryptic, rocks and branches. Emballonurids use a wide vari-
breaking up the outline of the bat in the roost. ety of roosts: caves, tree holes, and man-made struc-
tures. They are very agile, and often remain alert
when roosting.

Family: Nycteridae (16 species in 1 genus). Plate 12

The Old World slit-faced bats are found in the
African tropics, and the arid regions of the Near-
East. Oddly, one species is isolated from the rest,
in Indonesia. The feature which gives them their
name is a prominent slit running back along the
muzzle from the nostrils, bordered by fleshy pro-
tuberances, which is thought to be important in
sound emission for echolocation. They have large
ears and short, broad wings. They are primarily
insectivorous, taking prey in the air and from the
ground. Nycteris grandis (Fig. 1.27, forearm length
up to 66 mm) also takes scorpions and small
Figure 1.26 Black-bearded tomb bat, Taphozous melanopogon vertebrates, including frogs, birds, and bats. Like
(drawing by Lucy Hammond). the megadermatids, they take their prey back to
 EVOLUTION AND DIVERSITY 27

favourite feeding perches. They roost in small Sub-family Phyllostominae

groups or singly, in caves, hollow trees, and
The earliest fossil phyllostomid, Notonycteris magda-
man-made structures. Some have been known to
lensis, dates back to the Miocene (13–25 million years
roost in the burrows of aardvarks and porcupines.
ago) (Savage 1951), and has been assigned to the
sub-family Phyllostominae, generally regarded as
Superfamily: Noctilionoidea the oldest group of Phyllostomidae. There are 42
This is a large tropical superfamily, very diverse in species of Phyllostominae, in 16 genera. Most spe-
its biology, which probably originated in the neo- cies are omnivorous, taking insects and fruit. There
tropics, colonized subtropical regions, and has even is a trend towards carnivory in some of the larger
penetrated some temperate areas. There are seven species. Chrotopterus auritus, Trachops cirrhosus (Fig.
families, the Phyllostomidae, Mormoopidae, Nocti- 1.28), and Vampyrum spectrum all feed extensively
lionidae, Furipteridae, Thyropteridae, Mystacini- on small vertebrates. Phyllostomus hastatus also takes
dae, and Myzopodidae. vertebrates, but to a lesser extent than the others.
However, they all take insects and fruit. Both car-
Family: Phyllostomidae (160 species in 55 genera). nivory and frugivory appear to be an extension of
Plates 13–20 insectivory in these bats. The teeth of even those
The 160 species of New World leaf-nosed or spear- Phyllostominae which feed extensively on fruit are
nosed bats exhibit greater diversity than any other arguably less well developed to that end than those
family, including the Pteropodidae and Vesperti- of other members of the family: one of the reasons
lionidae (families with more species, but fewer for suggesting that they are the most primitive (in
genera). They almost certainly evolved from an the sense that they evolved first) of the Phyllostomi-
insectivorous ancestor, but now feed on a wide dae. Lonchorhina aurita (Fig. 1.29) has arguably the
variety of food. Seven sub-families are currently most impressive noseleaf of all bats.
described, although some authorities have laid
claim to as many as nine. The Phyllostominae eat
insects and other arthropods if they are small, and
include vertebrates in their diet if they are large.
Some larger species eat fruit. The Stenodermatinae
are specialized fruit eaters, and the Glossophagi-
nae and Phyllonycterinae feed mainly on nectar
and pollen. The Brachyphyllinae eat fruit, nectar,
and pollen. The Carolliinae show a marked prefer-
ence for the fruit of pepper (Piper species). Finally,
the Desmodontinae are the vampires. Most Phyl-
lostomidae (vampires are a notable exception) have
a relatively simple, vertical, blade-like noseleaf,
and the ears are simple but often long. In almost
all other respects, the family is very variable, a
reflection of their diverse foraging behaviour and
diet. They are found in the tropics and subtropics
of the New World, but in habitats as varied as hot
and humid lowland forests, cool montane forests,
and semi-arid deserts. They exploit virtually all
possible roosts, and several species of Stenoderma-
tinae build foliage tents. Two Cynopterus species
(Pteropodidae) also make tents. Figure 1.28 Fringe-lipped bat, Trachops cirrhosus.
28 BATS

Figure 1.31 Honduran white bat, Ectophylla alba.

Sub-family Stenodermatinae
A large sub-family of 67 species in 18 genera. These
bats have a short muzzle, broad flat molars, and
often a cutting edge to some of the teeth: adaptations
Figure 1.29 Tome’s long-eared bat, Lonchorhina aurita.
for dealing with courser fruits than those eaten by
the Glossophaginae (see below). Some species also
take insects at certain times of the year. The group
includes some fascinating bats: the tent-making Ur-
oderma bilobatum (Fig. 1.30) and the all white Ecto-
phylla alba (Fig. 1.31), another tent-maker. The most
bizarre is Centurio senex (Fig. 1.32), the wrinkle-faced
bat, with its face full of fleshy folds, the transparent
chin flap which covers its face when roosting, and its
partially transparent wings.
Sub-family Brachyphyllinae
Two species of relatively specialized fruit, nectar,
and pollen feeders from the Antilles in the Carib-
bean. They have a long muzzle, the teeth are small
and simple, and the tongue is long and extensible.
They are closely related to the more specialized
Glossophaginae.
Sub-family Phyllonycterinae
A small group (5 species in 2 genera) with similar
habits to the Brachyphyllinae, they are found on the
Caribbean islands of Cuba, Jamaica, Caymans, and
Figure 1.30 Peter’s tent-making bat, Uroderma bilobatum.
Caicos.
 EVOLUTION AND DIVERSITY 29

Figure 1.33 Lesser long-nosed bat, Leptoncteris yerbabuenae.

Piper, pepper plants. Although related to the Glos-

Figure 1.32 Wrinkle-faced bat, Centurio senex. sophaginae, they lack their specializations, and
some certainly take insects.
Sub-family Desmodontinae
Sub-family Glossophaginae
The three members of this family all feed exclusively
The 32 species of flower bats (in 13 genera) have on blood, and have a number of specific adaptations
progressed further along the road to specialization for this. They have heat sensors in the noseleaf for
in nectar, pollen, and soft fruit. They typically have locating capillary-rich areas of skin, modified ca-
long, often very long, muzzles and greatly reduced nines for fur clipping, long, sharp incisors for pain-
teeth (Fig. 1.33). The extensible tongue frequently lessly opening a wound, anticoagulants to prevent
ends in a brush-like collection of papillae, which clotting, and a grooved tongue to help move blood
increase the quantity of nectar or pollen that can be rapidly to the mouth. A specialized stomach and
carried on it. Many of the smaller species can hover kidneys rapidly remove the blood plasma, which
in front of flowers. It is possible that this feeding the bats often begin to excrete before they have
strategy evolved more than once in the Glossopha- finished their meal. Since a blood meal may be up
ginae, as it may have done in their Old World to 60 per cent of the bat’s body weight, and only the
equivalents amongst the smaller megabats. The red blood cells are of value, removal of the plasma is
recently discovered Anoura fistulata has a tongue essential: the return flight to the roost will be impos-
that can reach 85 mm into the corolla tube of the sible if the bat cannot reduce its weight. Does the
flowers it feeds from (Muchhala 2006). This name of the common vampire (Fig. 1.34), Desmodus
uniquely long tongue, more than twice the length rotundus, come from its after-dinner appearance?
found in other nectar bats, is stored in an equally These 30–40 g bats are remarkably agile on the
unique tube behind the sternum. Facial structure in ground, and will often land before approaching
bats may be determined by factors other than feed- their prey. Desmodus now feeds primarily on domes-
ing behaviour. tic mammals, but will take blood from native mam-
mals and birds, which were presumably its natural
Sub-family Carolliinae
prey before humans provided an abundant alterna-
A small but very abundant group of bats (9 species tive. Diaemus youngii, the white-winged vampire,
in 2 genera) specializing in the fruit of the genus also feeds on cattle in some parts of its range, and
30 BATS

Figure 1.35 Sooty moustached bat, Pteronotus quadridens.

Figure 1.34 Common vampire bat, Desmodus rotundus.

swallows. They typically roost in hot humid caves,

on birds in others: the hairy-legged vampire, Di- often in large colonies, in forest and arid areas.
phylla ecaudata feeds mainly on birds. Both species
are much less common than Desmodus. All three Family: Noctilionidae (2 species in 1 genus). Plate 22
species can roost in caves or tree cavities. There are two fisherman bats of the New World
tropics and subtropics. Both have large droopy
Family: Mormoopidae (10 species in 2 genera). Plate 21 lips, and are sometimes referred to as the bulldog
The Mormoopidae are found in tropical and sub- bats. The larger species Noctilio leporinus (forearm
tropical areas of the New World. They are known as length up to 88 mm, weight to 60 g) trawls for fish
the moustached or naked-backed bats (Pteronotus up to 100 mm in length, and the bat will adjust its
spp., Fig. 1.35) and ghost-faced bats (Mormoops flight path to intercept a moving fish. With long legs,
spp.). The features responsible for the first two big feet, and long claws for gaffing the fish, it is well-
names are a fringe of long hairs around the mouth, adapted to its foraging style. The tail membrane is
and wing membranes which arise close to the mid- kept out of the water by long calcars—cartilaginous
line of the back—but underneath the wing the back projections from the feet towards the tail, which
is still furred. The hairs round the mouth, together support the trailing edge of the tail membrane in
with flaps on the lower lip, may help to focus the many bats (Fig. 1.15). Fish are carried to a perch to
echolocation pulses and/or funnel insects into the be eaten. The diet also includes aquatic insects and
mouth. At least one species (Pteronotus parnellii) has small crustacea. Like its smaller, mainly insectivo-
independently evolved a flutter-sensitive echoloca- rous, relative N. albiventris (Fig. 1.36), it feeds on
tion system, analogous to that used by the Old coastal waters as well as rivers and lakes. In com-
World Rhinolophidae. The unusual wing mem- mon with many bats, particularly those that feed
brane attachment may serve to increase wing area over water, they can swim and take off easily from
without increasing span, but there have been no the water surface. Both species roost primarily in
attempts to test this idea. Mormoops have the long, caves and hollow trees, as clusters of non-breeding
narrow wings characteristic of fast flying aerial in- males, or harem groups—a pattern common to
sectivores, and they feed close to the ground like many bats.
 EVOLUTION AND DIVERSITY 31

Family: Furipteridae (2 species in 2 genera)

The smoky bats of the South American tropics
resemble the Thyropteridae, and the unrelated Na-
talidae, with a high, domed skull, and funnel shaped
ears (Fig. 1.46). A characteristic feature is the tiny
thumb. The single species of Amorphochilus is one of
the few bats known to live in the dry, low country
west of the Andes.

Family: Thyropteridae (4 species in 1 genus). Plate 23

The disk-winged bats of the neotropics have tiny
suction cups on their wrists and feet (Fig. 1.37),
used to grip the smooth walls of their roost: the
furled leaves of Heliconia and related banana-like
plants. As the roost leaves grow and unfurl, the
small colonies must constantly move on to younger
leaves. They have the domed skull and funnel
shaped ears of the Natalidae and Furipteridae.
Figure 1.36 Small fisherman bat, Noctilio albiventris.

Figure 1.37 Disk-winged bat, Thyroptera

tricolor, with disks on wrists and feet.
32 BATS

Family: Mystacinidae (2 species, 1 now extinct, in 1 Dactylanthus taylori. An equally unique and mono-
genus). Plate 24 typic family of bat-fly feeds on the fungi and yeast
The short-tailed bats are a rather unique and enig- which grow on its guano.
matic family, endemic to New Zealand. There were
two species, but there have been no confirmed speci- Family: Myzopodidae (2 species in 1 genus). Plate 25
mens caught or seen of the greater short-tailed bat Myzopoda aurita, the Old World sucker-footed bat, is
since 1965, and it is probably extinct. New Zealand confined to Madagascar. It resembles the New
has only one other native terrestrial mammal, a World Thyroptera in several respects: the high
vespertilionid bat. Morphological and molecular domed skull, large ears, small thumb, toes with
evidence suggests close affinities with the Noctilio- only two phalanges, fused toes, and suckers on
nidae. Mystacina tuberculata (Fig. 1.38) is a small bat hands and feet. However, the ears are not funnel
with remarkable adaptations to a partly terrestrial shaped, and the thumb sucker is not stalked. A new
way of life. The wing membranes are very tough species has recently been described, M. schliemanni.
close to the body, and the delicate distal regions can
be folded away and tucked into skin pouches on the
Superfamily: Vespertilionoidea
flanks. The legs are short and stout, and the toes and
thumb have talons near the base. It spends much of The largest superfamily, dominated by the Vesper-
its time on the ground, will burrow for food, and tilionidae, and to a lesser extent the Molossidae, it
search for insects in the burrows of seabirds. It even also includes the recently created family the Mini-
burrows into fallen and rotten trees, particularly the opteridae and is completed by the little known
kauri, to roost. They forage in the air, in trees, and on Natalidae. All are insectivores, but several species
the ground, on a wide variety of food. Although also take fish.
primarily insectivorous, they will also eat fruit, nec-
tar, and pollen and are the primary pollinator of the Family: Vespertilionidae (388 species in 47 genera).
rare and unusual parasitic plant: the woodrose, Plates 26–29
The vesper or evening bats are the second largest
mammalian family after the Muridae (Old World
rats and mice), and one of the most widespread
mammalian groups. They are absent only from the
Antarctic, the most northern areas of North America
and Eurasia, and some isolated oceanic islands.
They thrive even in desert environments, and
many species survive temperate winters by hiber-
nating. Some species undertake substantial migra-
tions to avoid the worst of the temperate winter and
to find suitable hibernation sites. Most species have
simple faces with no noseleaf, but their ears are very
variable (Figs. 1.39, 1.40). Although typically small
with a simple tragus (a cartilaginous projection
inside the external ear or pinna, Fig. 1.15), some
ears are as long as the bat that wears them (Fig.
1.41). These long ears are often furled around the
head at rest, and held erect when the bat becomes
active. Large ears are sometimes partially fused
along their inner edges. The tail membrane is usu-
ally large and, with the wings, is used to catch
insects. Vesper bats are almost exclusively insectiv-
Figure 1.38 Short-tailed bat, Mystacina tuberculata. orous, feeding by aerial hawking, gleaning, and
 EVOLUTION AND DIVERSITY 33

Figure 1.39 Noctule, Nyctalus noctula (drawing by Lucy Hammond).

Figure 1.41 Brown long-eared bat, Plecotus auritus (drawing by

Lucy Hammond).

colour extends along the fingers, in stunning con-

trast to its black wings. Bright and cryptic coloura-
tion is most common in those species roosting in
trees, often in the open (Figs. 1.42, 1.43).
Vesper bats exploit virtually all possible roost
sites: trees, tree holes, under bark, in flowers, inside
bamboo stalks, old birds’ nests, caves, mines, tun-
nels, houses, and so on. Members of several genera
have suction pads on their thumbs and feet, similar
to those seen in the Thyropteridae and Myzopodi-
dae, an adaptation for roosting inside the smooth,
Figure 1.40 Australian big-eared bat, Nyctophilus timoriensis.
furled leaves of plants like banana.
Vesper bats are divided into five sub-families,
flycatching. The term insectivorous includes the but there is considerable uncertainty about the
consumption of other arthropods. The Mexican classification of these bats, and there are frequent
fish-eating bat, Myotis vivesi, and several other spe- revisions. The largest sub-family is the Vespertilio-
cies of Myotis (for example Myotis adversus) take ninae, with 238 species in 38 genera. In the Myo-
small fish, and have the large feet and sharp claws tinae, 103 of the 106 species belong to the genus
of other fish-eating bats. Myotis (arguably the most widespread genus on
Most vesper bats are shades of brown, grey, and the planet after Homo). Other sub-families are the
black, but there are some colourful exceptions, Murinae (tube-nosed bats, 19 species in 2 genera),
including the painted bats of the genus Kerivoula. Kerivoulinae (23 species in 2 genera), and the An-
Kerivoula picta has bright red-orange fur, and the trozoinae with just two species.
34 BATS

Figure 1.42 Butterfly bat, Glauconycteris superba.

Figure 1.43 Welwitsch’s hairy bat,

Myotis welwitschii.

Family: Miniopteridae (19 species in 1 genus). Plate 30 Family: Molossidae (100 species in 16 genera).
The Miniopterinae are the long-fingered bats, all of Plates 31–33
the genus Miniopterus. It has only recently been All but one species belong to the sub-family Molos-
suggested on molecular grounds that this family be sinae. The exception is Tomopeas ravus (Tomopeati-
separated from the Vespertilionidae (Hoofer and nae) from Peru. The free-tailed bats are found in
van den Bussche 2004). They are all long-winged, the Old and New Worlds, including warmer tem-
fast flying species, many of them migratory. As in perate regions. They are adapted for rapid aerial
the Emballonuridae, their long wings make an extra pursuit of insects in the open, with long, narrow
fold when they are tucked away at rest. wings. Wing membranes are generally very tough.
 EVOLUTION AND DIVERSITY 35

Figure 1.44 Egyptian free-tailed bat, Tadarida aegyptiaca.

Figure 1.46 Cuban funnel-eared bat, Natalus micropus

brasiliensis, house several million bats. Bats may fly

over 50 km a night in search of food. In temperate
zones T. brasiliensis and several other species are
migratory, and it has been suggested that some
may even undergo short periods of torpor, but this
does not appear to be as deep as in true hibernators.

Family: Natalidae (8 species in 3 genera). Plate 34

The funnel-eared bats of the New World tropics are
small and delicate, with the broad wings of a slow-
flying, manoeuvrable gleaner of small insects. The
ears are large, forward pointing funnels. They roost
in large colonies in humid caves.

Adaptive radiation—why are the

Figure 1.45 Greater naked bat, Cheiromeles torquatus. Phyllostomidae so numerous and so
diverse?
The tail always projects beyond the end of the tail Many different factors have played a part in the
membrane. The ears are frequently stiff, tilted well evolution of the phyllostomids, some of which we
forward, and joined along part of their length (Fig. can only make informed guesses about. Some influ-
1.44). The skull is flattened in some crevice roosting ences are of general biological relevance, others are
species. The two Cheiromeles species (Fig. 1.45) are specific to the bats themselves. First of all, the phyl-
naked but most molossids have a thick covering of lostomids evolved in the neotropics, which provide
short fur. three important influences: age, climate, and size.
They use a wide variety of roosts, and some cave The neotropical forests have been around since the
roosts of the Brazilian free-tailed bat, Tadarida end of the Cretaceous, 65 million years ago, giving
36 BATS

lots of time for the bats to evolve, and the first known strategy, and most importantly, with large, sharp
phyllostomid is probably about 25 million years old incisors to puncture skin. The large animal-eating
(Savage 1951). In both animals and plants, species bats of the Old World lack this last important
diversity typically decreases with distance from the requisite.
equator: few people are unaware of the tremendous Whether hawking, flycatching, or gleaning, a bat
biodiversity of tropical rainforests relative to that of will encounter vertebrates, and to a large bat, a small
northern boreal forests, for example. More species vertebrate, like a large arthropod, is food. Mammals
can be supported by large areas: this is one of the few (including other bats), birds, reptiles, and amphibia
general statements that most ecologists would agree are all taken by some of the larger Phyllostominae,
on (for example Schoener 1986) and the neotropical such as Chrotopterus auritus and Vampyrum spectrum
forests are by far the largest in the world. (and in the Old World, by the Megadermatidae and
Back in the Miocene, and probably earlier, the first Nycteridae). None appear to be exclusively carnivo-
insectivorous phyllostomids were evolving in the rous: vertebrates are just a greater or lesser part of
vast neotropical rain forests. Most modern bats their varied diet, made possible by their flexible
have pretty flexible foraging strategies, and there is foraging strategies.
no reason to believe their ancestors were any differ- The smaller fisherman bat, Noctilio albiventris (a
ent. As well as taking insects in the air during con- relative of the phyllostomids) feeds by trawling for
tinuous flight or by flycatching from perches, early insects over water. Its larger cousin, N. leporinus, is
bats probably gleaned prey from a range of surfaces, large enough to take fish, and has become a fish
including fruit and flowers. Bats must have specialist, an obvious transition. Fish are part of
exploited this source and frequently taken a mouth- the diet of several trawling Vespertilionidae (for
ful of fruit, got a snoutful of pollen, or a taste of example Myotis vivesi).
nectar. It is easy to see how insectivores could Flexible foraging involves flexible prey detec-
become omnivores, as many modern phyllostomids tion. All phyllostomids echolocate, and more of
are, and eventually fruit or nectar/pollen specialists. that later, but many use prey-generated sounds
The absence of competition from the Old World fruit to catch their food—as do some bats in other
bats may also be an important factor in the evolution parts of the world. One phyllostomid, Trachops
of the fruit eating phyllostomids. cirrhosus, listens to the calls of male frogs, which
Insects also gather on larger animals, often can make up a substantial part of its diet at some
around wounds, and Fenton (1992) has suggested locations (Tuttle and Ryan, 1981). Tonatia sylvicola,
that the vampires may have evolved from bats a related species, is attracted to singing katydids
feeding on insects and their larvae on the wounds (Tuttle et al. 1985).
of large mammals and even birds. He provides The conditions under which the Phyllostomidae
persuasive arguments as to why vampires have evolved uniquely favoured their astonishing adap-
only evolved in the New World. Only here were tive radiation, but many of the conditions applied to
the conditions right: lots of large mammals and other bat families. Together with the mobility and
large birds with the requisite wounds and insects ecological flexibility conferred by flight, they have
(mammals frequently bear scars from predators or enabled the bats to become arguably the most
inter-specific conflict), bats with a flexible foraging diverse mammalian order.
 CHAPTER 2

Flight

Basic aerodynamics—aerofoils, flight theory, and the mechanics of flapping flight. The evolution of flight.
Why flight evolves, and why it has not evolved more often—advantages conferred, anatomical/physiological
adaptations and constraints. Ecology of flight—power v speed curves. Maximum range and minimum
power speeds. Wing morphology and flight performance in relation to ecology—aspect ratio, wing loading,
and wing shape. Scaling effects.

Some basic aerodynamics: aerofoils overcome the effect of gravity, and thrust to over-
come the drag experienced by the wings and body of
and flapping flight
the bat that resists forward movement. Drag is usu-
Aerofoils ally minimized by having a streamlined shape, but
is always significant, at around 10–20 per cent of the
There are several excellent accounts of aerodynam-
lift. Since lift acts at right angles to the movement of
ics in relation to animal flight, and most will be
the wings, the wings must move forwards and
mentioned somewhere in the chapter, but for a
downwards relative to the airflow. A gliding bat
wide ranging book on all aspects of vertebrate flight
achieves the right angle of incidence (Fig. 2.1c) to
you should go to Norberg (1990). Pennycuick (1972)
overcome drag by losing height. A flapping bat
has written a short and very readable book on ani-
must use appropriate wing movements (kinematics)
mal flight, and more recently Tennekes (2009) has
during the downstroke and the upstroke to over-
published a second, extended edition of his superb
come drag.
little book called The Simple Science of Flight: From
Insects to Jumbo Jets which covers all aspects of flight
Blade-element and momentum jet theories
in an entertaining and accessible way. A real under-
standing of flight can only be gained if you are The wings of birds and bats have been dealt with
familiar with a little of the terminology and some in the same way as fixed wings in many theoreti-
of the concepts of aerodynamics, so that is what we cal analyses of flight. The blade-element and
will start with. momentum jet theories are applied when it is
The shape of an idealized aerofoil, and that of a thought that steady-state or quasi-steady-state
bat’s wing, are shown in Fig. 2.1a. The tapered pro- conditions are operating. In the blade-element
file and convex upper surface are responsible for the approach, the velocities and forces operating on
forces generated by an aerofoil that keep the bat in different sections along the wing are calculated
the air and move it forward. Because of this shape, for each phase of the wingbeat, and summed for
air flowing over the top of the wing must move the whole wing. The momentum jet theory calcu-
faster than that flowing below (Fig. 2.1a). This res- lates the air velocities in the momentum jet of air
ults in an area of low pressure above the wing, and behind the wing. The aerofoil action of the wing,
high pressure below, and therefore a net aerody- with its upward and forward force, induces an
namic force (NAF) that raises the wing (Fig. 2.1b). equal and opposite jet of air, backwards and
If the bat is to go forward, then the NAF must have downwards. These two approaches are often
two components, a vertical, lift component to applied together.

37
38 BATS

(a) a b

high speed
low pressure

airflow
section through a
low speed section through b
high pressure
wing membrane
wing skeleton

(b) lift
NAF

airflow

drag
direction and magnitude
of air pressure around wing

(c)
NAF lift NAF lift
Figure 2.1 Aerofoils and the generation of lift
thrust thrust and thrust. a) Transverse section through an
aerofoil in an airstream. A bat’s wing (right) has a
lo w low similar section, and also acts as an aerofoil. b) The
airf fb at airf nt
e nt o veme bat pressure differences around the wing (left), and the
vem mo Net Aerodynamic Force (right) which results from
mo net them. The NAF is conventionally resolved into lift
wing and drag components. c) How gliding and flapping
bats gain lift and forward thrust (based in part on
gliding flapping flight Rayner 1986).

In a steady- or quasi-steady-state analysis, the lift beyond the scope of this book, but details can be
and drag coefficients of the wing must be calculated. found in Norberg (1990). The relevance of these
Drag and lift are defined by the following equations: calculations will become apparent later in the chap-
ter when we look at the power requirements of
Drag; D ¼ 0:5rV2 SCD flight.
Lift; L ¼ 0:5rV2 SCL
Vortex wake theory and wing kinematics
where æ = air density, V = forward velocity of the
wing, S = wing area, and CD, CL are the coefficients Although still widely applied, and very useful,
of drag and lift respectively. The last two indicate blade-element theory underestimates the NAF gen-
the capacity of the wing, at a given angle of inci- erated under certain conditions, and a number of
dence, to generate drag and lift. As I said above, a people have developed a new approach which
good aerofoil has a lift:drag ratio of about 10:1. Total often gives a better match between theory and
drag on the flying bat has three components, all of experiment—this is vortex wake theory. The camber
which must be calculated. Induced drag is due to of a wing, in addition to increasing the velocity of air
perturbations in the airflow round the wing, profile flow over it, also deflects the air downwards, and
drag is the friction and pressure drag on the wings, induces rotational flow around the wing, and
and parasite drag is the friction and pressure drag on behind it: these are the bound and trailing vortices
the body. A discussion of how these are calculated is shown in Fig. 2.2. If Newton’s third law is to be
 FLIGHT 39

(a)

bound vortex
CIRCULATION
starting vortex
(b)

trailing vortex

rotation in
vortex

(c)
lift NAF

Figure 2.2 Vortex wake theory. a) The motion of the

airflow drag wing at the start of the downstroke sets up the starting
vortex at the trailing edge, and induces an equal and
opposite rotation of air around the wing—the bound
vortex, or circulation. b) As the wing goes through the
bound vortex downstroke, a trailing vortex forms at each wing tip, as air
downwash is shed and rotated. c) A schematic of an active aerofoil
showing: the downward deflection of the air (downwash),
starting vortex
the starting and trailing vortices and the bound circulation
trailing vortex responsible for the NAF (based in part on Rayner 1986).

obeyed (action and reaction are equal and opposite), A bat must therefore change its wing shape and/
then the downward momentum of the trailing vor- or the angle of incidence during a wingbeat if it is
tices in the wake of the bat (action) is balanced by to maintain forward flight. Lift acts at right angles to
the upward NAF acting on the wings (reaction). the direction of wing movement. It is proportional
A study of the vortices in the wake of a flying bat to the wingspan, the speed of movement, and to the
(the momentum jet of the previous section) will circulation, or strength, of the bound vortex. The
therefore tell us a lot about how the wings generate circulation is determined primarily by the angle of
lift and thrust. incidence, increasing with increasing angle until the
Detailed descriptions of wing kinematics and vor- airflow becomes turbulent, when the wing stalls and
tex wake theory in relation to bat flight can be found lift is lost. An active downstroke will give weight
in Rayner (1986, 1987): what follows is a simplified support and forward thrust, and will be the most
account. To recap, in flapping flight a bat must use important phase of the wingbeat. To get a mean
its wings to support its weight and overcome drag if forward thrust over the wingbeat, the upstroke
it is to move forward. It cannot do this by simply must therefore be changed. Circulation can be
flapping its wings with fixed wing geometry and reduced by decreasing the angle of incidence (Fig.
angle of incidence with the airflow. This would 2.3a) or changing aerofoil section. Alternatively (or
keep it airborne, but would generate equal amounts in addition), the wingspan can be reduced by flexing
of forward and backward thrust in the downstroke the wing, or sweeping back the tip (Fig. 2.3b).
and upstroke respectively. This is shown in Fig. 2.3a. Whether or not a bat generates lift on the upstroke
40 BATS

(a)

wing s
ect
io net weight support
n no thrust angle of incidence

m
forward motion

ot
constant

ion
of bat
airflow

airflow
do weight support
w
ns
tro thrust inactive upstroke,

ke
ke

tro
no lift, reduced

s
up
circulation

no lift

(b)

reduce circulation by........

wing flexion or.....

sweeping back
wing tip

Figure 2.3 How to generate lift and net forward thrust. a) Top—with a constant angle of incidence to the airflow, the wing supports the weight of the
bat, but there is no forward thrust. The lower diagram shows one method of overcoming the problem. The wing is ‘feathered’, and the upstroke is
aerodynamically inactive. The upstroke can remain active by flexing the wing, or sweeping back the wing tip (b). Circulation is reduced but forward thrust
is maintained, at the cost of reduced lift during the upstroke relative to the downstroke (based in part on Rayner 1986).

will depend upon the relative size of the lift great deal about how bats fly. By flying bats through
and drag components. If drag is low, the bat neutrally buoyant, helium-filled soap bubbles, Ray-
may be in a position to generate lift on the upstroke, ner et al. (1986) were the first to visualize the trailing
but when drag is high, the upstroke is likely to be vortices. This was done by taking stereo photographs
passive, to avoid negative thrust. Drag is deter- of the bubbles at 8 ms intervals, and plotting their
mined by flight speed and wing morphology, so paths to reveal the direction and velocity of air flow.
we can expect these to have a major impact on As we will see, the picture painted by these bubbles
wing kinematics. has recently been shown to lack the detail revealed by
This brings us back to vortices. In combination more recent work, but an account of this study will
with analyses of wing kinematics they can tell us a help us understand the complexities that follow.
 FLIGHT 41

The noctule, Nyctalus noctula, is a moderately because drag is low at these speeds. As we will see
large, fast-flying bat, with long, narrow wings. below, drag is high only at very low and very high
Figure 2.4 shows idealized trailing vortices behind speeds.
a noctule flying at 7.5 ms1. The bat trails a continu- At low speeds, wing kinematics and trailing vor-
ous vortex from the tip of each wing, indicating that tices are very different. The wake of a noctule flying
the bound circulation remains intact throughout the at 3 ms1 is made up of a series of discrete, roughly
wingbeat. Both the downstroke and the upstroke are circular ring vortices, shed during the downstroke.
active and generate lift. This is achieved because the (Fig. 2.5).
wing remains flat, and the wing has a positive angle Trailing vortices are only shed when the wing is
of incidence, over most of the wingbeat. The armw- generating lift, that is during the downstroke. The
ing (that part between the body and wrist) is merely absence of vortices during the upstroke indicate that
flexed a little during the upstroke, reducing the it is aerodynamically inactive. When the circulation
wingspan (and hence lift) and ensuring that there of the bound vortex changes, a transverse vortex,
is net forward thrust. The upstroke is active, sup- equal in magnitude to the change in circulation, is
porting the bat’s weight over the whole wingbeat, at shed from the wing’s trailing edge. In this case it
the cost of some negative thrust during the occurs when the circulation is created at the begin-
upstroke. The negative thrust is not very important ning of the downstroke, and destroyed at the end.

Figure 2.4 Fast flight. Vortices behind a fast flying noctule, Nyctalus noctula.

Figure 2.5 Slow flight. Vortices behind a slow flying noctule, Nyctalus noctula.
42 BATS

These transverse vortices break the trailing vortices explain why this species can use a gait at high flight
into rings, which are opened at the beginning of the speeds that has a negative thrust component in
downstroke, and closed at the end. The wing beat the upstroke. The brown long-eared bat (Plecotus
frequency is faster than at 7 ms1, and the upstroke auritus) has only been seen to use the passive
is made inactive by flexing the wings to reduce the upstroke gait. The high drag of their low-aspect
span, and by rotation to reduce the angle of inci- ratio, rounded wings may prevent them from
dence to zero or negative values. using an active upstroke. I will come back to wing
Figure 2.6 shows the wing kinematics of a greater shape and the different components of drag later in
horseshoe bat during slow flight, with an inactive the chapter.
upstroke. At the top of the wingbeat, the wing is, so Twenty one years later, technological advances
to speak, palm down, and the back of the wing tilted enabled Hedenstrom et al. (2007) to show that the
slightly forward. The aerodynamic forces acting on helium bubbles did not reveal all. The nectar-feeding
the wing as it is swept down in this position will bat Glossophaga soricina was flown in a wind tunnel
move the bat forwards and upwards. As the wing filled with a fine fog (1 μm diameter particles) and
reaches the bottom of its stroke it is quickly flipped by providing bats with nectar from a feeder, their
over, so that the palm faces forward, and may tilt position in the tunnel could be kept constant.
slightly backwards. It then sweeps up in this posi- Digital Particle Image Velocimetry (DPIV) cameras
tion, and folds as it goes. captured the movement of the fog particles in the
The noctule therefore has two gaits: at high flying wake of the bat by illuminating them with a power-
speeds the wings are aerodynamically active ful sheet of pulsed laser light in two planes, parallel
throughout the wingbeat, and at lower speeds, and perpendicular to the air flow. At the same time,
when drag is higher, the upstroke becomes inactive. wing kinematics were filmed with high speed cam-
The noctule has high aspect ratio wings (that is long eras, again from two planes. The three dimensional
and narrow, see below), with pointed tips. Wings of wakes still change with flight speed, but they are
this shape have low induced drag (drag due to the more complex and generate lift in different ways. At
movement of the wings themselves), which may slow speed, as the wing turns into the upstroke it is

Figure 2.6 Wing movements during slow flight of a greater horseshoe bat, Rhinolophus ferrumequinum (from Aldridge 1986).
 FLIGHT 43

flipped upside down, reversing the circulation, and fly slowly when foraging—to hover or turn rapidly.
travels back faster than the bat is flying forward. If the angle of attack is too great, the vortex bound to
Because the circulation is reversed, the wing con- the wing breaks away, and there is turbulence and
tinues to generate lift and forward thrust, so it is loss of lift. Bats can flex their fingers, and the wing
active in both downstoke and upstroke. At higher membrane itself has thin sheets of muscle in it to
flight speeds this reversal does not occur, but the keep it taut. The wing also attaches to the hind limb,
pattern is far from simple, with two opposing vorti- and bats can move this too. They therefore have
ces: a strong wing tip vortex and a second, of about considerable control over their wing shape and posi-
half the strength, rotating in the opposite direction tion: perhaps more control than birds. Flexing the
at the base of the wing (Fig. 2.7). wing increases camber (increased curvature, convex
It was calculated that some as yet unknown non- up), just like lowering the flaps on an aeroplane, and
steady state mechanisms must be operating to gen- this reduces the stalling speed by allowing a higher
erate lift at low speeds. Not only is lift generation angle of attack before breakaway. This makes bats
more complex than previously thought, but experi- particularly good at slow speed, highly manoeuvr-
ments carried out by Hedenstrom and his colleagues able flight. This is just what is needed to chase flying
show that bats and birds appear to do things quite insects, which are also often highly manoeuvrable.
differently. More work must be done before we can
say how general these differences are, but if you are
The evolution of flight
interested in following up this story you should start
with Johansson et al. (2008) and Muijres et al. (2008) Vertebrates took to the air a long time ago. The first
for recent developments and some impressively to do so were the reptilian weigeltisaurs of the Per-
detailed flow visualizations of bat wakes. mian period, which died out at the time of the great
I noted earlier that increasing the angle of attack end-Permian mass extinction, about 240 million
of the wing increases lift. As a bat slows down, it years ago (myr BP). The weigeltisaurs, like the later
needs to do this to keep airborne. Many bats need to keuhneosaurs of the Triassic (245–208 myr BP), had

Figure 2.7 The complex vortices around the wings during the downstroke in Glossophaga soricina. The strength and direction of the airflow are shown
by the direction and size of the arrows. The strong wing-tip vortices are accompanied by weaker but distinct opposing vortices at the base of the wings
(from Hedenstrom et al. 2007, with permission from AAAS).
44 BATS

gliding membranes supported by extended ribs. birds) has concentrated on a study of living animals,
Longisquama, another Triassic glider, had a spectac- often from a biomechanical viewpoint. Let’s take a
ular double row of long scales on its back, which look at the idea that bats evolved from a gliding
probably dropped into a horizontal position for ancestor, and see why it is the preferred option. A
gliding. At around the time Longisquama was glid- very concise and readable account has been given by
ing, the first vertebrates to use powered flight were Rayner (1992).
evolving—the pterosaurs. Among the first to First of all, if the ancestral bat is going to glide, it
appear, 230 myr BP, were Eudimorphodon and Dim- has to be able to climb, to gain height for its jump.
orphodon, and the last to die, 165 million years later When an animal climbs a tree its muscles do work
in the end-Cretaceous extinction, included Pterano- and use energy. Some of this muscular energy is
don and Quetzalcoatlus. In the midst of all these converted into potential energy (potential energy =
gliding and flying pterosaurs, the first known bird, the mass of the animal X its height above the ground
Archaeopteryx, emerged 135 million years ago, and at X acceleration due to gravity). The higher it climbs,
about the same time mammals were experimenting the more energy it expends, but the more potential
with gliding (Meng et al. 2006). Perhaps the energy it gains. This potential energy is used by the
unknown first bat was flying towards the end of animal to glide. When the protobat jumps (Fig. 2.8),
the pterosaurs’ reign. gravity pulls it to the ground, and it loses potential
energy.
Gliding or jumping? Gliding is powered by potential energy: when the
protobat opens its wings as it falls, they generate a
Two theories have been put forward to explain the force which counteracts the force of gravity acting
evolution of flight (flight refers to powered, flapping on its body, holding the protobat up. In other words
flight, as distinct from gliding): the arboreal and the the wings act as an aerofoil. However, as the aerofoil
cursorial. The arboreal theory for the evolution of moves through the air, there is a backwardly
bat flight was probably put forward first by Darwin directed force (drag) as well as the upward force
(1859), and has been elaborated many times since. (Fig. 2.8 and see previous section). Drag resists the
Put simply, tree or cliff dwelling ancestors evolved forward motion of the animal, and slows it down.
flight through an intermediate gliding stage. The A gliding animal maintains speed because it is fall-
cursorial theory is more recent, and has not had a ing. The energy needed to overcome drag and main-
wide following, but Caple et al. (1983) formalized it, tain gliding speed comes from the loss of potential
and raised numerous objections to the arboreal the- energy as the animal approaches the ground. Good
ory. These objections have been effectively coun- gliders have good aerofoils, with little drag. They
tered by a number of authors (for example can glide a long way because only a little potential
Norberg 1985, Rayner 1992), and the arboreal or energy (height loss) is needed to overcome drag, so
gliding theory was widely favoured, but recent less height is lost over a given horizontal distance:
bird fossils have revived interest in the cursorial their glide angle is small (Fig. 2.8).
theory. Hedenstrom et al. (2009) provide an up-to- So, we have established that a glider and future
date review comparing flight in bats and birds, flyer needs to be able to climb. We do not need to
including evolutionary aspects. In birds, there are look very far to find evidence which suggests that
reasons for considering the cursorial theory (Rayner our protobat, a small mammal, could climb: the
1986, Hedenstrom 2002), but it can be discounted world is full of very agile mammalian climbers of
when considering bats. The cursorial theory de- all sizes. What does it need to be able to glide?
mands that the animal runs and leaps, gaining suffi- A gliding surface or aerofoil, and the strength to
cient speed to get lift from its outstretched wings to hold the aerofoil open. Rayner (1992) has calculated
glide, and then fly. There is little doubt that our that the muscular strength required to hold the aero-
protobat was unable to run at any speed! In the foil open is much less than that needed for climbing,
absence of an abundance of evidence from the fossil so a climber can glide if it has an aerofoil. Aerofoils
record, work on the evolution of flight in bats (and and gliding have evolved at least once in every
 FLIGHT 45

NAF
lift
good glider
high lift:drag ratio
M
drag
PE = Mgh low glide angle
h Mg

NAF lift
poor glider Figure 2.8 Gliding. A gliding mammal climbs a tree to gain
drag low lift:drag ratio potential energy PE (PE = Mgh; M = body mass, g = acceleration
due to gravity, h = height above ground). Its gliding membrane
generates lift to keep it airborne, and the loss of height (and
Mg therefore loss of PE), provides the energy to overcome drag and
high glide angle maintain glide speed. The higher the lift to drag ratio, the
shallower the glide.

vertebrate group, and several times in most. There membrane can be increased, there is little scope for
are gliding fish, frogs, lizards, snakes, and mam- increasing aspect ratio without significant modifica-
mals, and some of their aerofoils (pectoral fins, tion of the limb skeleton. If the gliding membrane
webbed feet, extended ribs, flattened bodies, flaps extended to between the fingers, then simple elon-
of skin between the fore and hindlimbs) differ little gation of the fingers could increase aspect ratio, and
from the standard anatomical plan of their non- this was probably a very important step in the evo-
gliding cousins. The protobat would have little dif- lution of powered flight. Not only would it allow an
ficulty in becoming a bad glider as the skin flap increase in aspect ratio, but adjustments to the flight
between fore and hindlimbs developed, and from path could also be made by movement of the fingers
there, a better one. to change the shape of the aerofoil. Modern-day
How was the aerofoil’s performance improved? flying lemurs (dermopterans) have webbed hands
First of all, the amount of lift generated increases as with short fingers—a step away from a long-
the area of the aerofoil increases. More lift means fingered flying bat. Sears et al. (2006) have shown
slower, safer speeds, and longer glides, so an evolu- that the lengths (relative to body size) of the third,
tionary selective pressure for increased wing area is fourth, and fifth fingers in bats have remained simi-
understandable. Another factor is wing shape. Mod- lar in length for the last 50 million years. They show
ern mammalian gliders such as flying squirrels and that the critical step, occurring at an early stage in
marsupial sugar gliders have aerofoils of low aspect the evolution of bats, was probably an increased
ratio (Fig. 2.9). local expression of a single protein, Bmp2, which
Aspect ratio is a measure of wing shape, calcu- causes the proliferation of bone-forming chondro-
lated as the length from tip to tip (span), divided by cytes and an elongation of the finger bones.
the width, front to back (chord). For the oddly So, there’s no reason why a gliding protobat
shaped wings of animals, a better measurement is should not have evolved, and there are several
wing span squared divided by area. Low aspect good reasons why it should. A single tree would
ratio wings generate lots of drag, the lift to drag have been unlikely to provide our protobat with all
ratio is high, the glide angle is steep, and the glide of its food, or cater for other needs, such as a safe
therefore short (Norberg 1985). The length of the shelter and mates, and like many modern mammals
glide can be increased by lengthening the wing, it would have moved from tree to tree. It had two
thus increasing aspect ratio, which increases the lift choices, jump from one tree to the next, or climb all
to drag ratio. Although the area of the gliding of the way down, cross the ground, and climb back
46 BATS

cost—its life—if caught by a ground predator.

A jump, or better still a glide from tree to tree, is a
good way to avoid a more agile predator. Lots of
animals have learnt to move through trees by jump-
ing, and many spread their limbs for stability and
for landing. It is a small step from there to gliding.
The next step is to active, flapping flight. Rayner
(1986) has calculated that flapping the wings over
an amplitude of 25º or more, with flexing on the
upstroke, is sufficient to generate useful thrust to
assist gliding, and that a climbing animal could
supply the appropriate energy: the way to flapping
flight is therefore open. A detailed aerodynamic
model for this transition has been developed by
Norberg (1985, 1990).

The advantages of flight and the

demands on the flyer: physiological
and biomechanical aspects of flight
What are the advantages conferred on our protobat
by the evolution of flight, and the obstacles, physio-
logical and anatomical, that stand in its way?

The advantages: why fly?

I have already noted one major advantage, the cost
of transport. Flight may demand twice the energy
per second as running, but a flying animal travels
very much faster, and the cost of transport may
therefore be five times lower than that for a flightless
counterpart of similar size. In terms of energy ex-
Figure 2.9 The low aspect ratio gliding membranes of the sugar
pended per unit distance moved, flying is cheap.
glider, Petaurus breviceps. Bats can therefore forage over wide areas, exploiting
food supplies that are ephemeral or show seasonal
fluctuations. They are able to commute over consid-
up again. The latter is an expensive option: climbing erable distances between roost site and foraging
down can use as much energy as climbing up. Glid- area and, in so doing, exploit otherwise unavailable
ing on the other hand is cheap. A good glider, like a resources. Flight gives them access to new food
colugo, can travel over a 100 m, lose only 10 m sources, such as airborne insects, the fruit and flow-
height, and expend almost no energy. Gliding is ers of many trees, and vertebrates such as fish and
also fast: five or more times faster than running on birds. A wide range of roost sites can be used, many
the ground (Rayner 1986, Scholey 1986). Although safe from all, or almost all, predators. Flight itself
the cost of climbing is high, the high gliding speeds also reduces predation risk (Pomeroy 1990). Migra-
mean that the overall cost of transport (cost per unit tion is an option open to flying animals, allowing
distance travelled, or unit time travelling) is low. A different, distant environments to be exploited at
small animal on the ground could pay the ultimate different times of the year. Flight enables bats to
 FLIGHT 47

cross major barriers such as deserts, seas, and moun- fatigue. If a bat needs to supply twice as much
tains. The advantages conferred by flight may have energy to its muscles as a mammal on the ground,
major consequences. One feature common to the it must have some special physiological adaptations.
three living groups of flying animals—bats, birds, When a bat takes off, its breathing rate rapidly
and insects—is the relatively large number of spe- increases 4–6 fold, to match its wingbeat frequency
cies in each group. The mobility given to them by (around 10 Hz for an average-sized bat). Not only is
flight has probably been at least partly responsible there one breath to one wingbeat, but the two pro-
for their diversification. cesses are closely coupled in the bat species studied
(for example Carpenter 1986). Bats breathe out late
The obstacles: why is flying difficult? in the upstroke when the forces pulling the wing
down are greatest, and these forces also compress
If flight can give an animal enormous advantages in the thorax. The movements of the thorax during
life, why hasn’t it evolved more often? One answer flight may assist ventilation and thus save energy,
is that flying is difficult. If you are going to fly well, as has been suggested for some terrestrial mammals
you have to be built for the job, and built to a high (Alexander and Young 1992, Young et al. 1992). Not
specification plan. The mechanical, physiological, surprisingly, bats often emit their echolocation
and aerodynamic conditions that must be fulfilled pulses late in the upstroke too, and echolocation,
make it unlikely that flight will evolve very often. expensive in a stationary bat, comes at little or no
I have already said that flight is expensive in terms extra cost to a flying bat (Speakman and Racey
of the rate at which energy must be supplied to the 1991). When a terrestrial mammal starts running,
flight muscles. This is shown in Fig. 2.10: the meta- the changes in respiration rate are smaller than in
bolic cost of transport decreases with increasing bats, but not always that much smaller. The extra
body size, but it is greater in bats than non-flying oxygen supplied to the flight muscles of a bat cannot
mammals, and comparable to that of birds. be attributed to differences in the breathing rate. The
To meet the sustained high energy demands of Brazilian free-tailed bat, Tadarida brasiliensis, has
flight, oxygen must be taken in by a capable respira- been shown to have a lower resting oxygen con-
tory system, an effective heart and circulatory sys- sumption rate than rodents of similar size (Canals
tem must deliver it to the flight muscles, and the et al. 2005) but a higher maximum rate and aerobic
flight muscles must work for long periods without scope. In combination with thinner blood–gas bar-
riers and relatively larger lungs (Canals et al. 2005,
Maina 2000), bats have an oxygen diffusion capacity
up to three times higher than that of the rodents
oxygen consumption (ml/kg/min)

700 studied (Maina 2000). Physical differences in the

geometry of the bronchioles leading to the lungs
500 flying birds may also make airflow more efficient. What about
the circulatory system? Heart rates of resting bats
300
flying bats are comparable to those of other mammals of similar
size (for example Studier and Howell 1969, Thomas
running and Suthers 1970, Carpenter 1985), but are higher in
mammals flight than those of running mammals, and compa-
100 rable, at up to six times the resting rate, to those of
0.02 0.1 1.0 flying birds (Berger and Hart 1974, Carpenter 1986).
body mass (kg) A bat’s heart, like a bird’s, is as much as three times
the size of that of a terrestrial mammal of the same
Figure 2.10 The cost of locomotion in relation to body size. As animals
body size. This means that in a single heartbeat, it
get bigger, the amount of energy consumed per kg of body weight (in this
case oxygen consumption) declines—locomotion becomes more efficient. may be able to pump around twice as much blood to
Flight in both birds and bats is about twice as expensive as running in the flight muscles—data are still a little sparse
mammals (based on Thomas 1987). (Thomas and Suthers 1972). Oxygen from the lungs
48 BATS

is carried in the blood bound to haemoglobin in the auritus, should need 1.83 W to fly. Using her blade-
red blood cells. Bats have more red blood cells than element theory, the calculated mechanical power
birds and terrestrial mammals and more haemoglo- input was 0.36 W (Norberg, 1990). Assuming an
bin in them, and can therefore carry more oxygen efficiency of 0.25 for the conversion of chemical
(for example Wolk and Bogdanowicz 1987). To sum energy to mechanical work, the metabolic input =
up, there are no startling physiological adaptations 0.36/0.25 = 1.44 W, add 0.11 W for resting metabo-
as seen in birds: the high energy demand of bat lism, and this results in a total metabolic input of
flight muscle is apparently met by small, cumulative 1.55 W, in good agreement with the physiological
changes in several physiological parameters, but the estimate. Metabolic rate during flight would there-
picture is far from clear in many of the species fore be 14 times resting metabolism. In bats, the
studied and more work needs to be done. More flight muscles are typically 10 per cent of the body
detailed accounts of these physiological aspects can weight; this will be 0.9 g in a brown long-eared bat,
be found in Thomas (1987), Norberg (1990), and and they are predominantly fast contracting, oxida-
Maina (2000). Although bats appear to match the tive fibres (Altenbach and Hermanson 1987). They
performance of birds in many respects, they are will produce their maximum power output at the
probably still constrained by having tidally venti- wingbeat frequency of 10–14 Hz (see Altringham
lated lungs. This may be one reason why the largest and Young 1991). Measurements of mammalian
bats are just one tenth the weight of the largest flying skeletal muscle power output under conditions
birds. simulating those operating during locomotion (for
The flight muscles of bats are typical of those of example James et al. 1995) suggest that these fast
most mammals. They must be capable of sustained contracting muscles should be capable of generating
activity, and are therefore highly aerobic. They have up to 150 Wkg1. The total mechanical power avail-
a rich capillary blood supply, the fibres themselves able to Plecotus is thus 0.14 W: lower than the 0.36 W
are rich in myoglobin to facilitate oxygen transfer calculated by Norberg for Plecotus. However, it is
from the blood to the muscle, and they have numer- identical to the 0.14 W calculated by Norberg et al.
ous mitochondria to synthesize ATP aerobically. (1993) for a 12 g Glossophaga soricina in forward
Most mammals, including many if not most bats, flight. Total power required was 1.63 W, or 12
fuel activity by metabolizing lipid and glycogen times resting metabolic rate. Helverson and Winter
stores. However, if the energy demands of flight (2003) summarise work they and colleagues have
are particularly high, this may not be the most effi- carried out on flight in glossophagine fruit bats.
cient source. In common with hummingbirds, the They come to one surprising conclusion—hovering
small, nectarivarous bat Glossophaga soricina fuels flight is cheaper than in both hummingbirds and
hovering flight directly with the sugars present in hawkmoths.
ingested nectar (Voigt and Speakman 2007), which The flight musculature of bats is complex relative
can be rapidly metabolized after ingestion. This re- to that of birds, since the muscles have a role to play
duces the need for large fat stores, which would in both flight and terrestrial locomotion. There are
increase the cost of flight by increasing the weight five major downstroke muscles and two major
of the bat, and it saves the energy which would be upstroke muscles: in birds there is one major down-
consumed in converting the sugars to lipids—and stroke muscle, assisted by one other (reviewed in
back again! Hill and Smith 1984, Norberg, 1990). Altenbach and
The increase in metabolic rate during the transi- Hermanson (1987) have carried out the most
tion from rest to flight is enormous. Measurements detailed experiments on bats, which show that the
of oxygen consumption during flight (Thomas 1975, downstroke muscles begin to contract late in the
Carpenter 1986, Speakman and Racey 1991) show upstroke, and similarly the upstroke muscles are
that there is a tight relationship between flight cost active before the downstroke is complete. This
and body size which can be used to predict oxygen early onset of activity presumably acts to brake the
consumption for bats of known weight. Norberg wings at the top and bottom of the strokes and
calculated that a 9 g brown long-eared bat, Plecotus initiate changes in wing shape. The initial stretching
 FLIGHT 49

of the active muscle will also increase muscle power in the planes of maximum bending force (Norberg
output (for example James et al. 1995). Few muscles 1970). We will look at the wing in more detail in the
are active in the middle of the strokes. next section: first, a brief look at other skeletal adap-
The discussion so far has concentrated on physio- tations, summarized in Fig. 2.11.
logical adaptations for flight: there are also bio- In common with birds and pterosaurs, bats have
mechanical constraints. Flying animals are best short, streamlined bodies and the vertebral column
made light so that less energy is needed in flying, has regions of restricted movement, and even rigid-
and the forces on the body are smaller, placing ity, due to the fusion of bones. The large ridge on the
smaller demands on the musculo-skeletal system. sternum seen in birds, for the attachment of the
Bats show a number of adaptations that reduce major flight muscle (pectoralis), is often absent in
weight, notably in the bones and membranes of the bats. However, it may be replaced by a ligamentous
wings (Norberg 1970). However, these structures sheet which serves the same function. The shoulder
must also be strong, to stand up to the large forces girdle is attached to the sternum by the clavicles, to
that act on them during flight. The wing bones of form the strong support needed for the wings.
bats are unusually rigid, a property not achieved by Flight demands adaptation in many other areas of
internal struts as in birds, but by simply having the body. For example, in both insectivorous and
slender bones with the greatest diameters oriented fruit-eating bats, the gastrointestinal tract has an

short, rigid vertebral column

clavicle

view from anterior

clavicle

sternum

keel Figure 2.11 Adaptations of the skeleton

bird view from left bat of bats for flight. See text for details.
50 BATS

absorptive surface area exceeding that of other by a short ligament. The increased rigidity resulting
mammals, due to increased epithelial folding and from this resists the large strains present at the lead-
microvilli formation (Maina 2000) to help fuel flight. ing edge during flight. The metacarpals are much
Even the brain is under the influence of flight: bats longer than the phalanges. The thumb has a free
that fly in more complex environments and need to claw, used in climbing and grooming. The humerus
process more sensory information have larger infe- interlocks with the scapula at the top of the
rior colliculi to deal with the auditory demands of upstroke, so that during the downstroke, the ful-
echolocation and a larger hippocampus, important crum of the wing is at the joint between the clavicle
in spatial memory (Safi and Dechmann 2005). and the scapula. This increases the mechanical effi-
ciency of the downstroke (Altenbach and Herman-
son 1987). Aerodynamic forces bulge the wing
The wing and what it does in flight
during flight, and shape is maintained by thin mus-
All the major features of a typical bat’s wing can be cles in the membrane, in addition to a network of
seen in Fig. 2.12. elastin and collagen fibres. The hindlimb is also
The wing membrane, or patagium, is supported long, and supports the posterior part of the flight
primarily by the elongated forelimb. This is made membrane, which ends somewhere along the tibia/
up of the humerus, the long radius (the ulna has fibula or the proximal part of the foot. The leg has
regressed), and the elongated metacarpals and pha- rotated during the evolution of bats, and the foot
langes of fingers 2–5. Fingers 2 and 3 are usually and claw point backwards. Bats are generally less
joined at the distal section of the wing’s leading edge agile on the ground than other mammals (but not

Figure 2.12 Skeleton of a greater horseshoe bat,

Rhinolophus ferrumequinum, to show wing structure.
 FLIGHT 51

universally so!), but they can use their legs to alter boundary layer, reducing the velocity gradient.
wing shape, and hang in very awkward sites. The wing bones project above the upper surface of
A special locking mechanism in the tendons of the the wing and, together with hairs on the leading
foot enables bats to hang upside down without edge of the membrane, may introduce this micro-
effort (Schaffer 1905, Bennett 1992). A cartilaginous turbulence into the boundary layer (Pennycuick
projection from the ankle, the calcar, extends 1971, Norberg 1972). Fast flying bats have wing
towards the tail and supports the trailing edge of bones that are flattened in the plane of the wing to
the flight membrane. A small flap of skin, the post- give a smooth profile. Slow flyers, that would
calcarial lobe, may project beyond the calcar. benefit from turbulence generation, have rounded
Finally, the long tail (if present) is typically bones that project above the wing surface (Vaughan
incorporated into the flight membrane but, at the and Bateman 1980).
other extreme, is completely free in some species. The wing section between the body and the fifth
A number of small muscles extend from the limbs finger is the plagiopatagium (armwing), which gen-
into the flight membrane, and some are found erally provides most of the lift or weight support
entirely within the membrane. They are used to during flight. The dactylopatagium (handwing),
tension parts of the wing during flight manoeuvres, between fingers 2 and 5, generates most of the for-
and aid wing folding at rest. ward thrust. The uropatagium (tail membrane) pro-
The wing membrane can be divided into several vides some lift, and in fact moves up and down
functional parts. A section of the membrane, the synchronously with the wings. It may be used as a
propatagium, lies anterior to the humerus and brake, and (with the wings) can be used to catch
radius. With the small membrane between fingers insects in flight. Of course bats are constantly chang-
2 and 3 (dactylopatagium minus), it acts as a leading ing direction and this is achieved by subtle move-
edge wing flap that can be lowered by the thumb ments of the wings and tail membrane. The wings
and finger 2. These flaps can delay stalling at high can be moved independently and, by folding one
angles of attack by maintaining laminar flow over wing, rapid tight turns can be made (Fig. 2.13).
the wings, preserving lift at low flight speeds. The Analyses of how bats perform some of these man-
propatagium is largest in slow flying, manoeuvrable oeuvres have been made by Norberg (1976, 1990).
bats. The lowering of these flaps increases the cam- Because of their hindlimb structure, take-off is a
ber of the wing (the degree of upward curvature of tricky business for many bats, and although most
the wing in cross-section), which can also be are capable of taking off from level ground, some are
achieved by flexing finger 5. Increased camber in- not very good at it. They get round the problem by
creases lift, at the expense of increased drag. This launching themselves from their roost site: after a
compromise is worthwhile in slow flying bats, when short free-fall, they are going fast enough to open
drag is low. Bats with the broadest wings, for exam- their wings and gain sufficient lift to stay in the air.
ple Old World fruit bats, in addition to a large Some of the fastest flying bats with long narrow
propatagium, have short metacarpals, enhancing wings fall 5–6 m before unfurling them.
their ability to camber the wing.
Bats have other mechanisms for preventing stal-
ling at high angles of attack—for example the wing Ecological aspects of flight
bones may act as turbulence generators. Stalling The power required for flight and its relation
occurs when the laminar airstream over the wing
to flight speed
separates from the upper surface, creating major
turbulence and loss of lift. Frictional forces slow Figure 2.14 shows the various components that con-
the air down close to the wing surface, and so tribute to the aerodynamic power required for flight
there is a gradient of increasing air velocity from in relation to airspeed, determined using blade-
the surface out. If this gradient is too steep, separa- element theory. Let’s see how this graph was
tion occurs. Separation can be prevented by intro- arrived at, and what it means in ecological terms.
ducing controlled microturbulence into the The total power required for flight is the sum of the
52 BATS

Figure 2.13 A turning brown long-eared bat, Plecotus auritus.

to be more than a minor component of total flight

Psum power (Psum).
power for flight (W)

0.2 Psum ¼ Paer þ Pine

The aerodynamic power has three components: par-
asite power (Ppar), profile power (Ppro), and induced
0.1 Pine Ppro
power (Pind)
Ppar Paer Paer ¼ Ppar þ Ppro þ Pind
Pind
0 Parasite power is needed to overcome the drag on
RMR the body. It is proportional to the frontal area of the
2 4 6 8 10
body, and increases dramatically with flight speed
Vmp Vmr (Æ speed3). Streamlining is therefore important, par-
air speed (ms–1) ticularly to fast flying bats, if parasite power is to be
minimized. Profile power overcomes the drag on the
Figure 2.14 The power required for flight, and its dependence on flight wings and increases with speed in a similar fashion
speed for a small (9 g) bat. Psum = total aerodynamic power required in watts to parasite power. It also increases with wing area.
(W). Ppro, Ppar and Pind = profile, parasite, and induced power respectively.
The induced power is that needed for lift and thrust
Pine = inertial power. Vmp = minimum power speed, the speed at which
power consumed per unit time is lowest. Maximum range speed (Vmr) is the
generation and decreases with flight speed. It is the
speed which will take the bat furthest on a given amount of fuel. The major power component of slow flight and hover-
mechanical equivalent of resting metabolic rate (RMR) is also indicated, ing. Induced power also increases with increasing
with RMR set at zero (based on Norberg 1987, 1990). body weight and decreases as wingspan increases.
Power requirements will therefore be minimal in a
streamlined bat (low Ppar), with low body weight
aerodynamic power and the inertial power. The (low Pind), long wings (low Pind), and high aspect
aerodynamic power (Paer) is needed to generate the ratio (low area, low Ppro).
lift and thrust that keep the bat in the air, and move The sum of all the power components, Psum, has a
it forward. The inertial power (Pine) is needed to distinct minimum at a speed characteristic of each
move the wings up and down, and is negligible at species. What does all this mean to bat behaviour?
all but the slowest speeds since wing inertia is Let’s look at some examples. If a bat is on migration,
turned into useful aerodynamic work at the end of with no real opportunity to feed, it wants to go as
the downstroke (Norberg 1987). It may be signifi- far as possible for the minimum cost. In other words,
cant during slow flight and hovering, but is unlikely it must maximize airspeed and minimize power
 FLIGHT 53

requirements. This is its maximum range speed (Myotis daubentonii) flies between Vmp and Vmr when
(Vmr) and can be found by drawing a tangent from foraging (Jones and Rayner 1988), but the similar
the origin of the graph to the Psum line (Fig. 2.14). If M. adversus foraged at speeds slightly above Vmr
food is plentiful, and the bat is in the air for a long (Jones and Rayner 1991). The nectarivore Leptonyc-
time but not in a hurry, then it may want to fly at a teris yerbabuenae commutes at speeds between Vmp
speed which keeps the cost per unit time as low as and Vmr (Sahaly et al. 1993). There are complications
possible: that is maintain the minimum cost of trans- however: for example, the speeds of M. daubentonii
port. This is the lowest point on the Psum line, the and M. adversus are influenced by aerodynamic
minimum power speed, Vmp. A foraging bat wants ‘ground effects’ because they fly close to water. Fur-
to keep transport costs down, but it may need to fly thermore, different aerodynamic models can give
faster than Vmp to cover sufficient ground to catch different results. Grodzinski et al. (2009) measured
sufficient food. This will be a delicate equation for commuting and foraging flight speeds in Pipistrellus
the bat to balance, but it should fly somewhere kuhlii using three-dimensional acoustic tracking to
between Vmp and Vmr. When a bat leaves the roost measure flight speed more accurately than could be
to fly to its foraging site, it may be important to get achieved in previous studies. They then compared
there fast—to maximize foraging time when the measured speeds with predicted optima, based
insect populations peak at dusk, and also to avoid on two published aerodynamic models. Only one
predators. It might then exceed Vmr. Is there any model yielded a good fit to theory, with foraging
evidence that bats actually follow these rules? The flight speed being faster than Vmp and commuting
answer is yes. Work on pipistrelle bats, Pipistrellus speed a little slower than Vmr, as predicted. I will not
pipistrellus, (Fig 2.15) by Jones and Rayner (1989) get into a discussion of the merits of different mod-
shows that individuals commuting to and from the els, but leave you to read Grodzinski et al. (2009),
roost fly significantly faster than Vmr, and foraging who also provide a good review of this topic.
bats fly close to Vmr, as predicted. Daubenton’s bat
Wing morphology and foraging strategy
Vmp Vmr
Vcommuting
Bats have wings of different shapes and sizes and a
12
few examples are shown in Fig. 2.16. The differences
10
8
6
4
2 Otomops
number of flights

martiensseni
0

Vforaging
Plecotus
auritus
12
0.2m
10
8
6
4
2
Rousettus
0 aegyptiacus
0.5 2.5 4.5 6.5 8.5 10.5
flight speed (m/s)
Mimetillus
Figure 2.15 Observed commuting and foraging flight speeds of the moloneyi

pipistrelle bat in relation to calculated minimum power (Vmp) and

maximum range (Vmr) speeds (based on Jones and Rayner 1989). Figure 2.16 Variation in the wing shapes of bats.
54 BATS

Hypsignathus
monstrosus

Mimetillus
moloneyi

Eidolon
helvum

Nycteris
thebaica

Taphozous
peli

Cardioderma
cor

Tadarida
midas

Figure 2.16 (Continued)

are largely a reflection of the foraging strategy of the ratio (AR = span2/area), or long and narrow, that
bat—where they feed, how they feed, and what they is high AR.
feed on. There are two main ways in which wings Other structural differences in wing form are also
can vary. First, wing area can be large or small important. The relative lengths and areas of the arm
relative to the size of the bat. We describe this and handwing vary considerably. The shape of the
using a measurement called ‘wing loading’. This is wing tip is also variable—it may be broad and
the weight of the bat, divided by the total area of its gently rounded, or narrow and pointed. All of
flight membrane. A high wing loading means a these measurements tell us about a bat’s flight
large bat with relatively small wings. Secondly, style, and can be related to its foraging strategy.
wings can be short and broad, that is low aspect Evolution is working on bats to ensure that each
 FLIGHT 55

species is evolving the best wings for the job. The flexible feeders, and can show considerable plastic-
biggest problem we have in explaining wing design ity in their foraging strategies.
is that we do not always know exactly what that job
is. There may also be considerations beyond the Wing loading and aspect ratio
purely aerodynamic: the best wings for a particular Let’s start with the two major components of wing
flight style may not be suited to another aspect of the morphology: wing loading (WL) and aspect ratio
bat’s lifestyle. However, a general picture has (AR). With two variables, we can have four possible
emerged, which we’ll look at after a word of cau- combinations, represented by the four quadrants of
tion. For the sake of analysis and illustration, bats Fig. 2.17. In this figure, species whose flight/beha-
are put into particular foraging categories. It should viour patterns are well documented are placed
always be remembered that most bat species are in their appropriate positions on a scatter plot of

low WL/high AR high WL/high AR

3 fast, open-air
slow, open-air hawking, short
flight, longdistance Myotis Noctilio range migration
migration bocagei leporinus migrators

T. longimanus
T. nudiventris
Myotis Taphozous melanopogan
vivesi Emballonura Miniopterus inflatus
increasing aspect ratio

monticola
Tadarida brasiliensis
trawlers E. semicaudata Lasiurus cinereus

Myotis daubentonii
Nyctalus noctula
–3 3
Megaderma
lyra Nyctalus leisleri
Anoura geoffreyi
ground gleaners Lasiurus borealis
Plecotus townsendi
P. auritus Myotis yumanensis
Nycteris hispida Choeronycteris mexicana
Nycteris macrotis hoverers
Leptonycteris hoverers/gleaners
yerbabuenae

high WL/low AR
low WL/low AR –3 fast flight in cluttered
slow, manoeuvrable habitats
flight in cluttered
habitats

increasing wing loading

open symbols = insectivores or short range migrators
closed symbols = also take vertebrates or long range migrators

Figure 2.17 Wing loading and aspect ratio combination in bats. A principal component analysis (Norberg and Rayner 1987).
56 BATS

a principal component analysis by Norberg and Taphozous species from the emballonurids and the
Rayner (1987). Species referred to in the text are free-tailed bats (Figs. 2.17, 2.18). They feed in the
named. The analysis compensates for the effects of open, catching insects on the wing. Their fast flight
size, an important point since mass increases in means that they cannot turn tight circles. They
proportion to length3, but wing area in proportion would not be good at hunting among trees or at
to length2, so that bigger bats tend to have higher hovering to pick insects off foliage.
wing loadings if they remain geometrically similar These and other bats may fly long distances dur-
(that is, all parts of the body increase in the same ing migration. Energy efficient, high AR wings are
proportions). We will come back to matters of size a therefore useful, and so too is a high WL, if high
little later. The equations in Box 2.1 should help you speed and a short migration time are important
in the discussion that follows. (such as noctules, Fig. 2.17). If the bat is not in a
The higher the WL the faster the bat must fly if its hurry, or migration distances are short, its wing area
small wings are to generate the lift needed to stay does not need to be so small in relation to its body
airborne. However, profile drag on the small wings weight. In this case very long, high AR wings are
is low (eqn. 4), so less energy is wasted and flight best, to reduce induced power to a minimum (eqn. 2),
efficiency is high. High WL is often found in combi- and WL may be quite low (for example the embal-
nation with a high AR: narrow wings are aerody- lonurids Emballonura monticolla and E. semicaudata,
namically more efficient than broad wings of the Fig. 2.17). Long-distance, fast-flying migrants should
same total area, since induced power is low (eqn. 2). have the longest, high AR wings, with a high WL (for
So, if a bat regularly flies long distances to feed, it example Tadarida brasiliensis, Fig. 2.17, 2.18). These
needs to be fast to get to the feeding site quickly species all need to fly in the open, since their long
(high WL), and it needs to waste as little energy as wings would be a hindrance in vegetation.
possible getting there (high AR and WL). These bats High AR and low WL are also found in fish-eating
are also streamlined to keep down parasite power bats such as Noctilio leporinus and Myotis vivesi (Figs.
(eqn. 5). Bats that fall into this category include 2.17, 2.19). Flying in the open over water, with no
Nyctalus, Lasiurus, and Miniopterus species among need to make tight turns, they can have long, effi-
the vespertilionids and Tadarida species from the cient wings. WL is low so that they can carry heavy
molossids. The best examples are perhaps some loads—the fish they feed on. The insectivorous
trawlers tend to have average AR and WL (for
example Myotis daubentonii and M. yumanensis),
although Myotis bocagei is an interesting and
Box 2.1 Power for flight extreme exception (Fig. 2.17).
Low WL in combination with low AR is found in
It is worth putting down a few equations which describe many bats that feed among vegetation. This combi-
the factors determining flight power. They will help you nation allows them to fly slowly without stalling,
understand the discussion of wing morphology in relation make tight turns, and even hover: at low speed
to lifestyle. profile power is low even with a large wing area
(eqn. 4). All of the bats on this part of the plot (Fig.
(1) Pind Æ Mg3/2/b in hovering flight
(2) Pind Æ Mg2/b2V in forward flight 2.17) are gleaners and hoverers. The low WL also
(3) Ppro Æ b3S/T3 in hovering flight (T is approx. Æ b, thus enables them to carry heavy prey and take off easily,
Ppro Æ S) even if they are carrying prey. Short wings are also
(4) Ppro Æ SV3 in forward flight useful when taking off from the ground and moving
(5) Ppar Æ AV3 in forward flight in cluttered environments, in addition to conferring
high manoeuvrability. The slit-faced bats (Nycteridae)
M = mass of bat, g = acceleration due to gravity,
b = wingspan, V = forward velocity of bat, are typically ground gleaners, and have the lowest
S = wing area, T = duration of a wingbeat, A = frontal AR/WL combination (Fig. 2.17). Next to them on
area of bat’s body. Fig. 2.17 are the long-eared bats (Plecotus species,
Fig. 2.20), representatives of a group of gleaners
 FLIGHT 57

Figure 2.18 Free-tailed bat, Otomops martiensseni, a fast, efficient flyer.

Figure 2.19 Noctilio leporinus, the fisherman bat.

that take food from the ground and from foliage and environment and limit manoeuvrability. These bats
that frequently hover. are principally nectar and pollen feeders and are
The specialist hoverers fall into the low AR/high found among the Phyllostomidae (for example
WL quadrant. Long wings would be aerodynami- Anoura, Leptonycteris, Choeronycteris Figs. 2.17, 2.21)
cally more efficient for hovering, since induced and the smaller Old World fruit bats (for example
power decreases with increasing wing span (eqn. Macroglossus species). The high WL gives them high
1), but they are a hindrance in a cluttered flight speeds, perhaps an important factor when
58 BATS

food supply is patchy and commuting time between

patches must be minimized—some of these bats
dart from flower to flower like hummingbirds. The
long wings that are ideal for hovering may have
been compromised by the need for speed and possi-
bly access to flowers.
Carnivorous species are essentially big-time glea-
ners and have low AR and low WL for the same
reasons as their insectivorous cousins. A good exam-
ple is Megaderma lyra (Figs. 2.17, 2.22). Many of these
bats have large ears for detecting prey without echo-
location. Big ears would produce a lot of drag, so
they are generally only found on slow flying bats.

Arm- and handwings, wing tip shape and camber

There are more variables to consider than WL and
AR, as I said at the beginning of this section: for
example relative lengths and areas of the arm- and
handwings, wing camber, and wing tip shape.
Baage (1987) and Norberg and Rayner (1987)
have carried out the most extensive analyses of
Figure 2.20 Brown long-eared bat, Plecotus auritus, a gleaner. these characteristics. Norberg and Rayner (1987)
have shown that short, rounded wing tips are
found on slow flying, manoeuvrable bats, in associ-
ation with low AR and low WL. To be manoeuvrable,
that is to have a small turning circle, a bat must fly
slowly. The size of the turning circle increases with
flight speed, hence the association with low WL.

Figure 2.21 A hovering nectar feeder, the long-tongued bat, Glossophaga soricina.
 FLIGHT 59

Figure 2.22 Megaderma lyra, a carnivorous ground

gleaner.

Bats that need to be manoeuvrable are usually those To initiate a roll º should be large and J should be
that forage in a cluttered environment, hence the small. The wings perhaps play a more dominant
low AR. These bats also tend to have high camber role in determining J than the body (Norberg and
wings, and have the ability to control camber by Rayner 1987). Wing inertia can be minimized by
flexing digit five and/or lowering the hindlimbs. having pointed wings, that is by having the
The long-eared bats of the genus Plecotus fit into ‘heavy’ carpal joint shifted towards the body. This
this category. seems to be the important factor in fast flying bats
The fast, efficient flyers hawking on insects in the such as the noctule, Nyctalus noctula. The pointed
open, the high WL high AR species, are not very wing tip also increases aerodynamic efficiency (by
manoeuvrable. However they often have pointed reducing wing tip vortices), reducing flight costs, so
wing tips, which increases their agility. Agile bats it is a feature seen in migrating species.
have the ability to rapidly initiate a roll, altering It is also possible to have good low speed agility.
their flight path. The relationship between wing mor- Horseshoe bats, Rhinolophus species, have high roll
phology and agility is complex, but some important rates, yet they have small, rounded wing tips that
factors can be identified. The maximum angular increase inertia. This is compensated for by the high
acceleration, Æmax, (that is the fastest roll) is given by: camber and low WL, that give them extra lift on the
handwing, and therefore a high º. This ability to roll
amax ¼ t=J rapidly at low speeds is used by horseshoe bats, and
where  = rolling torque, J = total roll moment of those of similar morphology, such as Myotis nattereri,
inertia (the inertia resisting the roll). Barbastella barbastellus, and Cardioderma cor to roll
into their roosting positions with remarkable agility.
t ¼ Mgbl
Baage (1987) categorized the 14 species of bat
where M = mass of bat, g = acceleration due to found in Scandinavia on the basis of five morpho-
gravity, b = wingspan, and º = moment of distribu- logical measurements, roughly speaking: wing cam-
tion of lift across the wingspan. Thus: berability, size and movement of propatagium,
mass, AR, and wing-tip shape. The bats included
amax ¼ Mgbl=J fast aerial hawkers like the noctule (Nyctalus
60 BATS

noctula), the gleaning long-eared bat (Plecotus auri- in birds: 1.5 g to 15 kg. Why are there no really big
tus), the trawling/gaffing Daubenton’s bat (Myotis bats? The answer lies primarily in the effects of size
daubentonii), and small, slow hawkers such as the on flight performance (Norberg and Rayner 1987).
whiskered bat (Myotis mystacinus) and the pipistrelle On the positive side, bigger bats fly faster, and the
(Pipistrellus pipistrellus). Predictions about their cost of transport decreases with increasing size.
flight performance and foraging habits were made However, the mechanical power required for flight
on the basis of these measurements and compared increases more rapidly with size than the power
to field observations of each species in ‘search phase’ available from the muscle, so there has to be a size
flight. Correlations between observed and predicted beyond which flight is impossible. This is around 15
behaviour were generally good. Studies of this kg, and explains the upper limit for flying birds, but
nature are of value in attempts to look at the struc- not for bats. Wing inertia increases with size, so
ture of bat communities. large bats are not very agile. Manoeuvrability also
declines with increasing size—one good reason why
Gliding bats that have to chase insects are small. Since prey
Bats are not generally thought of as either gliders are detected at short-range by echolocation, high
or soarers. However, although none make use of manoeuvrability enables bats to turn and capture
long glides, some small bats (for example Pipistrellus prey detected at the short ranges over which echolo-
pipistrellus) are known to glide for very short periods cation operates (Barclay and Brigham 1991). Since
(<1 s) and some large pteropodids are well-known most bats link breathing and echolocation pulse
flap-gliders (for example Pteropus mariannus). Day- emission with the wingbeat, very large bats, with
flying Pteropus samoensis from American Samoa low wingbeat frequencies, would have low pulse
have long been known to soar, making use of both emission rates: perhaps too low for effective prey
thermals and the updraft on slopes. Norberg et al. capture and orientation (Jones 1994). Holderied and
(2000) compared the flight morphology and gliding Helversen (2003) suggest that bats match detection
performance of soaring and non-soaring Pteropus distance to wingbeat period by having appropriate
species from Pacific and Indian Ocean islands. The echolocation calls (discussed in the next chapter).
soarers tended to have longer wings and lower wing Increased manoeuvrability also accounts for the
loadings (after correction for body size) than the small size of nectarivores such as the glossophagine
non-soaring species, although the differences were phyllostomids that hover in front of flowers. The
not statistically significant, perhaps due to the small macroglossine bats are larger, and perch to feed
sample sizes. There appear to be no aerodynamic from flowers. Large bats tend to take large food
constraints preventing the non-soarers from doing items, bigger fruit eaters take bigger fruits, and the
so. The key factor seems to be whether or not the larger ‘insectivorous’ bats take large invertebrate or
species has become a diurnal forager, since it is only vertebrate prey.
during the day that they can make use of thermals. Aerodynamic factors do not appear to set a lower
Despite the small morphological differences, the limit on bat size, but there may be mechanical con-
soarers were better performers, with lower mini- straints in skeletal strength. One important factor is
mum sinking speeds and smaller turning radii. the increase in surface area to volume ratio as size
decreases. The large surface area of a small endo-
Body mass therm loses heat quickly, and the small volume can-
There is one morphological feature we have yet to not generate sufficient heat to maintain a high body
cover, the size of the bat. So far we been concerned temperature. Small mammals have to eat constantly
with largely size-independent variables, such as to meet the high metabolic demand and below a
WL, AR, and wing-tip shape. Although bats range certain size they simply do not have sufficient for-
in size from 1.5 g to 1.5 kg, the range is much greater aging time (McNab 1982).
 CHAPTER 3

Echolocation and other senses

What is echolocation? Sound generation and perception. Echolocation call characteristics: FM calls—basic
properties. Autocorrelation function and range acuity. Spectral changes in echoes and target fine structure.
Spectral change as a detector of movement. Target location in 3-D space. Can bats intercept prey? CF
calls—basic properties. Doppler shift and velocity determination in CF bats. Prey detection and identifica-
tion using CF calls. How do bats avoid being confused by other bats’ sonar? Auditory adaptations and the
neural basis of echolocation. Paired sounds—enhanced neural response to the second sound. Time windows
and target ranging. Target ranging by pulse-echo delay. Dealing with sound intensity variation. Echoloca-
tion and behavioural studies in the wild. How good is echolocation in the real world? Do bats select their
prey? Echolocation and navigation. The cost of echolocation and its consequences. Echolocation in neotropi-
cal fruit-eaters. Other senses. Hunting with other senses. Old World fruit bats. Navigational senses. The
ecology of echolocation: interactions between flight, food, and foraging habits. Evolution of echolocation and
echolocation calls.

What is echolocation? High frequency sound is not essential for echolo-

cation although, as we will see shortly, it does
Echolocation is a complex and highly evolved pro- confer some major advantages. A number of ani-
cess that has given bats the ability to exploit an mals echolocate with lower frequency sounds. The
ecological niche closed to all but a few animal best known examples are the cave dwelling birds
groups—the night sky. Echolocation is not unique (Fig. 3.1): the South American oilbird or guacharo
to bats, but it has arguably reached its evolutionary (Steatornis caripensis) and the South-east Asian, Indo-
peak in these mammals. What is echolocation? Australian cave swiftlets (Collocalia species, Fig. 3.1),
A simple definition is that it is the analysis by an use 2–10 kHz sounds of 1 ms duration, at repetition
animal of the echoes of its own emitted sounds, by rates of up to 20 Hz. The birds use these loud,
which it builds a sound-picture of its immediate audible sounds to locate their nests or roosting
environment. A number of animals use sound in sites in dark caves, but switch off when there is
this way, and many use high frequency sound or enough light to see by (Sales and Pye 1974, Fullard
ultrasound, beyond the range of human hearing. et al. 1993, Price et al. 2004).
These include, as well as the bats, the Cetacea Old World fruit bats of the genus Rousettus (Fig.
(whales and dolphins: see for example Au 1993, 3.1) echolocate within their cave roosts, generating
Thomas et al. 2004) and the insectivores (shrews sound by clicking the tongue. The pulses are short
and tenrecs). It is not clear how sophisticated ultra- (less than 1 to 5 ms), high frequency (10–60 kHz) and
sonic echolocation is in the insectivores but there is have low energy content: an order of magnitude
no doubt about its value in cetaceans. Ultrasonic lower than those of bats producing ultrasound in
sounds are also emitted by some rodents and a the larynx (Holland et al. 2004). Their ability to
number of marsupials, but we do not know if they negotiate obstacles using echolocation has been
are used for echolocation or just for communication studied in the lab, and although it does not match
(Sales and Pye 1974). that of other bats, it is well developed, which is

61
62 BATS

Figure 3.1 The Egyptian fruit bat, Rousettus aegyptiacus, an echolocating, Old World fruit bat, and two echolocating birds, Steatornis caripensis,
the oilbird or guacharo, and Collocalia fuciphaga, a cave swiflet. All use echolocation to find their way to roosts/nests in caves.
 ECHOLOCATION AND OTHER SENSES 63

(Continued)

something of a puzzle given some of the limitations colleague. Griffin went on to do much of the pioneer-
inherent in the system (Holland et al. 2004). I am not ing work on echolocation and a fascinating account
going to say any more about Rousettus—much more can be read in his book Listening in the Dark (Griffin
is known about other bats. Throughout history peo- 1958, reprinted 1986).
ple have wondered just how bats navigate in the
dark, but real scientific enquiry started in the 1790s, Sound generation and perception
when Spallanzani performed some meticulous exp-
in echolocating bats
eriments, clearly indicating that hearing was crucial
to orientation, and sight largely unimportant. No one First, we need to look at how bats generate their
knew about ultrasound, so a plausible explanation sonar calls and how they detect the echoes. In com-
wasn’t found until this century. In the late 1930s mon with other mammals, sounds are generated in
Griffin cracked the code with the first generation of the larynx. The larynx in echolocating bats is pro-
ultrasonic detectors, borrowed from a physicist portionally larger than in most other mammals, but
64 BATS

the mechanism of action is the same: air passing Recorded from 10 cm, a pipistrelle’s echolocation
over the vocal chords makes them vibrate. The mus- pulses may be up to 130 dB: loud enough to be
cles of the larynx are used to adjust the tension in the uncomfortable, if we could hear it. All sounds
vocal chords and change the frequency of the sound. decrease in amplitude or loudness as they travel
Bats emit their calls through the mouth or the nos- away from a point source, due to spherical spread-
trils. Those using the nostrils often have complex ing. High frequency sounds decrease in amplitude
noseleaves: folds of skin and cartilage of varying even more rapidly as they travel through the air,
complexity that may extend over a large part of due to attenuation. This presents bats with a chal-
their face (see Fig. 3.2). In many species the noseleaf lenge. They need very sensitive hearing to detect
appears to act as an acoustic lens, focusing the the very weak echoes returning from their insect
sound into a narrow beam in front of the bat. prey, and so their hearing is very acute. But, how
Bats have to hear well, so they generally have do they avoid deafening themselves by the very
large external ears or pinnae. They also have a tra- loud sounds they emit? Bats do not emit echoloca-
gus (Fig. 3.2), a cartilaginous projection from the tion sounds continuously: they come in discrete
base of the ear, that sits inside the pinna. The ear/ pulses. During the pulse some species make them-
tragus combination may be designed to limit the selves temporarily deaf, and between pulses
major receptive field to an area 30–40 either side hearing is at its most sensitive. There are two tiny
of the midline, and improve the directionality or muscles in the ear and one of them, the stapedius,
sensitivity to incoming echoes. attaches to the stapes (Fig. 3.3). Just before a bat
Figure 3.3 shows a cross-section of the ear struc- emits an echolocation pulse, this muscle contracts
tures of an echolocating bat, but the basic layout is and pulls the stapes away from the oval window
common to all mammals. Sound passes along the (Henson 1965, Suga and Jen 1975). The sound can-
canal from the external ear, and vibrates the ear- not now be passed on to the cochlea, a very deli-
drum or tympanic membrane. This is only 2–11 cate structure that can be damaged by loud
μm thick in bats. The vibrations are passed along sounds. As soon as the pulse has been emitted,
the malleus, incus, and stapes, the three ear ossicles, the bat relaxes its stapedius and its hearing is
to the oval window. The higher the frequency of the fully restored. In the final stages of taking an insect,
echolocation pulses used by a bat, the smaller the a bat’s sonar pulse repetition rate may exceed 200
area of the eardrum, and the smaller and lighter Hz (Box 1), and its stapedius muscle can (but does
the ossicles—because they need to be able to vibrate not always) operate at the same frequency—one of
more rapidly. The external and middle ear cavities the highest rates recorded in vertebrate muscle.
are air filled, the inner ear is fluid filled. There are other sound dampening specializations
Vibrations of the oval window are transmitted associated with the cochlea—instead of being fused
along the spiral canal of the cochlea. The cochlea to the skull as is common in mammals, it is loosely
does the work of sifting and categorizing the suspended in its cavity, and surrounded by blood
sound by frequency before the brain interprets it. It sinuses and fatty deposits.
consists of three canals spiralling in parallel, and
between them is the organ of Corti. From the base
Echolocation calls
of the spiral to its apex, cells in the organ of Corti
respond to sounds of decreasing frequency. Echolo- Bats are clearly very good at echolocation. They
cating bats have 2.5–3.5 complete turns, relative to have had a long time to perfect it, and we can
the 1.75 typical of Old World fruit bats (and many probably assume that evolution has ensured that
other mammals), and the basal turns are particularly the characteristics of a bat’s echolocation signals
large. Sound waves vibrate the tectorial membrane are close to the optimum for the job, although we
and trigger the tiny cilia on the sensory cells. Electri- must always remember that there may be con-
cal impulses from these cells pass along the auditory straints we are unaware of. A good review of the
nerve to the mid-brain for initial signal processing, general principles of echolocation has been written
before final processing in the auditory cortex. by Simmons et al. (1975). What follows is a basic
 Brachyphylla
cavernarum
Megaderma Diaemus
spasma youngi Nycteris
hispida
Erophylla
bombifrons

Desmodus
Lavia frons rotundus
Pharotis
imogene Rhinolophus
ruwenzorii

Rhinopoma
hardwickei

Anthops
ornatus Rhinonycteris
aurantius
Macrophyllum
macrophyllum

Asellisus Triaenops
sp. persicus
Hipposideros
ridleyi
Carollia
perspicillata

Sturnira
lilium
Cleotis
persivali
Chiroderma
salvini

Figure 3.2 Noseleaves and ears

66 BATS

Taphozous
Rousettus
wildegardae
aegyptiacus

Taphozous
perforatus

Taphozous
Rhinopoma Stenonycteris
peli
sp. lanosus

Nycteris
sp.

Epomophorus
labiatus
Rhinolophus
landeri

Lavia
frons

Pteronotus
macleayii Myotis
bocagei

Hipposideros
bicolor

Laephotis
wintoni

Mormoops
megalophylla Desmodus
rotundus

Tonatia
sp.

Tadarida
sp. Macrotus
californicus

Figure 3.2 (Continued)

 ECHOLOCATION AND OTHER SENSES 67

tensor tympani muscle

organ of corti
middle ear

cochlea
external ear

eardrum

malleus
incus
inner ear stapes

stapedius muscle
auditory
nerve Figure 3.3 Auditory apparatus (based on Hill and Smith
oval window 1984).

Box 3.1 Sound

For those lacking a background in physics, this is a good Sound is normally represented by a wave, as shown in Fig.
place for a brief description of the physics of sound, and for 3.4. You can think of it as a vibration travelling over a water
defining some of the important terminology. Sound is a wave surface. The frequency is simply the number of waves which
of vibration passing through the air. The source of the sound, pass a given point in space in one second. The wave is
for example the vibrating larynx of a bat, sends out a wave of passing through the air at a speed of 340 m s-1, the speed of
compression of the air at each vibration, packing the sound. As frequency increases, and since sound always
molecules of the air closer together (Fig. 3.4). The number of travels at the same speed, more waves pass a given point in
times a wave of compression is sent out every second is the one second. This means that as frequency goes up,
frequency of the sound, and is measured in Hertz (Hz). The wavelength goes down. The wavelength is the distance
wavelength of the sound, the distance between between two corresponding points on a wave, say from one
each compression, is the reciprocal of the frequency or peak to the next. Low frequency sounds are low pitched, like
1/frequency. a double bass. High frequency sounds are high pitched, like a
violin. Human hearing ranges from around 40 Hz up to
20 kHz. Echolocation calls of most bat species fall
somewhere in the range 20 kHz to 120 kHz. The other main
feature of a sound wave is its height or amplitude, that’s the
distance between peak and trough. The amplitude is
determined by the amount of energy put into the sound. High
energy sounds have high amplitudes, and are loud. Sound
intensity is measured in decibels (dB).
amplitude
The waveform shown in Fig. 3.4 is the most simple, a sine
wave. It is a pure tone, the sound you get from a tuning fork.
Most sounds are more complicated, as shown in Fig. 3.5.
wavelength (1/frequency)

Figure 3.4 Sound. compression of molecules in the air in sound

generation, and its graphical representation as a sine wave, having a
frequency and amplitude.

continues
68 BATS

Box 3.1 (continued)

more complex and are determined by the structures that
make them, the human larynx or a grasshopper rubbing its
legs together, for example.
Each sound will have a fundamental frequency, that is the
lowest frequency. Let’s say it is 1,000 Hz. All the other
components of this complex sound are multiples of this
10 ms
fundamental frequency, 2,000, 3,000, 4,000 Hz, and so on.
Figure 3.5 A natural sound—human voice
These are called the harmonics: 2,000 Hz is the second
harmonic, 3,000 the third, and so on. Each can have a
different amplitude or loudness, and this can complicate the
Why are they so complicated? Each is made up of sounds
final waveform further, when they are all added together to
of more than one wavelength of more than one amplitude.
give waveforms like those shown in Fig. 3.5.
Pure tones are rare in nature; most biological sounds are

description of the properties of echolocation pulses by changing their sonar pulses when over 2 m away,
and their functions. and to 0.18 mm wires from 0.9 m (Grinnell and
Let’s start with frequency and wavelength. Why Griffin 1958). Individual fruit flies elicit a response
are high frequency calls used? There are several when up to 1 m away (Griffin et al. 1960). The bat
answers: let’s quickly cover the minor ones first. To may detect the target earlier, but not respond with
begin with, relatively few other natural sounds are an altered sonar signal, so these are minimum esti-
so high, so interference from other natural sources is mates of detection distances. In more subtle tests,
minimal. Furthermore, high frequency sounds do Kick (1982) showed that the big brown bat, Eptesicus
not travel very far in air before becoming very fuscus, detected 19 mm diameter spheres at 5.1 m,
quiet. Energy absorption increases exponentially and 4.8 mm spheres at 2.9 m. Detection of the echoes
with frequency, and a bat echolocating at 30 kHz is occurred close to the threshold of hearing, and pos-
unlikely to have a range exceeding 30 m, decreasing sibly involved integration of information from sev-
to 10 m at 100 kHz, and 4 m at 200 kHz (Lawrence eral pulses. Ambient noise and clutter in the field
and Simmons 1982a). The use of high frequencies almost certainly leads to poorer performance, but as
may be one way in which bats avoid interfering we will see, there are compensatory factors.
with each other’s echolocation. They may also Bats emit echolocation sounds in pulses. These
enable bats to get very close to prey that can hear. pulses are usually described as being frequency
High frequency sound is also less likely to alert most modulated (FM) or constant frequency (CF), but
predators. However, by far the most important many, perhaps most, species of bat use combina-
answer lies in target discrimination. This is technical tions of the two. Few bats use pure CF calls, and
jargon for how small an insect the bat can detect. As the terms broadband (for FM) and narrowband (for
we saw earlier, high frequency sounds have very most CF) are often more appropriate. Various
short wavelengths. Sounds of 100, 50, and 10 kHz authors have categorized pulse patterns in various
have wavelengths of 3.4, 6.8, and 34 mm respec- ways, and these classifications have some functional
tively. The best sound for detecting an object is one value, but it should be remembered that simple
with a wavelength similar in length to that of the labels are often used for complex calls. I will start
object. Bats catch small insects and therefore need with generalized descriptions of basic FM and CF
short wavelength, high frequency sounds. Beha- echolocation: other patterns will come up as specific,
vioural experiments show that little brown bats, real examples throughout the chapter. A sample of
Myotis lucifugus, respond to 3 mm diameter wires echolocation calls is shown in Fig. 3.6.
 ECHOLOCATION AND OTHER SENSES 69

110
90
frequency (kHz)

70
Myotis dasycneme Myotis mystacinus
50
30 Myotis bechsteinii
(two calls)
10
50ms 10ms 50ms

110
90
frequency (kHz)

70 Pipistrellus nathusii
50
Nyctalus noctula
30 Tadarida teniotis

10
10ms 50ms 20ms

110
Rhinolophus
90 euryale
frequency (kHz)

Barbastella
R. ferrumequinum barbastellus
70 Eptesicus
serotinus
50
30
10
Figure 3.6 Bat sonograms. Time versus
50ms 50ms 50ms frequency plots of echolocation calls.

Frequency modulated calls

Since sound travels at 340 ms–1, the echo from an
Basic properties object 1 m away returns 5.9 ms after the emitted
In its simplest form a broadband FM pulse is char- pulse. If the pulse was more than 5.9 ms long, then
acterized by a short, steep sweep down the frequen- the bat would be listening to the echo before it had
cies, from above 60 to about 30 kHz in a few ms in finished emitting the pulse. In most bats that use FM
many temperate vespertilionids, for example. Fre- sonar, neural mechanisms of echo interpretation
quency modulation or FM simply means that the demand that pulse/echo overlap is avoided. Having
frequency of the pulse is not constant, but altered detected its prey from the echo a bat wants to know
or modulated. The frequency versus time curves, or how far away it is. Since sound waves always travel
sonograms, of several species are shown in Fig. 3.6. at the same speed through air, by measuring the
Although calls in the vast majority of species sweep pulse-echo delay the bat can calculate the distance
down the frequency range, upwards sweeps have to the insect (Simmons 1971). So, a bat emits a short
been recorded in some species (for example Molos- pulse, waits for the echo, calculates the distance to
sops temminckii, Guillen-Servent and Ibáñez 2007). the insect, emits another short pulse to update its
Broadband FM pulses are short: typically 2–5 ms information, and so on.
long and sometimes less than 0.2 ms. Why are short The sequence of events leading to a successful
pulses used? Bats detect objects in their surround- aerial catch is shown in Fig. 3.7. Both the duration
ings by listening to the echo of each emitted pulse. and the bandwidth, or range of frequencies swept in
70 BATS

(a)
100
frequency (kHz)

search phase
approach
terminal

capture

0
0 0.2 0.4 0.6 0.8 1.0

(b)
search phase
approach
100 terminal
frequency (kHz)

capture

0 Figure 3.7 FM echolocation call changes from

0 0.2 0.4 0.6 0.8 1.0
search to capture: a) fundamental frequency only
time (s) b) with harmonics

the pulse, change. During the search phase, as a bat This type of pulse is sometimes used by bats, but in
hunts for its prey but before it detects it, it emits many cases it is more complex. Figure 3.7 shows one
pulses at a frequency of about 10 Hz (the frequency example of these more complex patterns. FM signals
of the wingbeat and respiration, which are coupled). potentially provide much more detailed 3-D pictures
When it detects its prey the bat enters the approach of the environment than constant frequency signals—
phase: as it gets closer there is less time between they are better at discerning target distance, and
pulse and echo, so the pulses have to get shorter to potentially (this will be discussed later) details of
avoid overlap. The bat also updates its information target structure. Bats therefore increase the band-
about the whereabouts of the insect more fre- width of their pulses when they detect and attack
quently, so its pulse emission rate gets progressively an insect. They do this in two ways: by increasing
faster. In the terminal phase, pulses may be emitted the range of frequencies swept over by the funda-
at frequencies of more than 200 Hz, and each pulse mental, and by introducing harmonics. The lower
may be only a fraction of a ms long. There is consid- lines in Fig. 3.7b are the FM sweeps of the fundamen-
erably more subtlety in this than I am revealing. To tal frequency. The lines above them are components
get a greater insight into how bats manipulate pulse of the second and third harmonics (not all frequencies
emission I suggest reading Moss et al. (2006): pulse are of the same intensity, so parts of the harmonics
emission patterns can be more complex and depen- may not be detected). Other changes in pulse pattern
dent upon the difficulties of the task before them. may be seen during approach and capture. Since low
However, this basic account explains the need for frequency sounds attenuate less in air than higher
short duration pulses. What about the frequency frequencies, some species will put more energy into
modulation? The explanation for that goes hand in the lower harmonics during the search phase, but
hand with another feature. enhance the higher harmonics as they approach their
The pattern I have shown in Fig. 3.7a is simplistic— prey for finer detail and target ranging (Simmons
a sweep involving a single fundamental frequency. 1971, 1979). Why are large bandwidth FM calls
 ECHOLOCATION AND OTHER SENSES 71

(a) The ACR also defines t, the width of the envelope at

t
half its maximum height. This is approximately the
T reciprocal of the signal’s bandwidth. The introduction
of FM higher harmonics decreases the height of the
side peaks, reducing t. We can think of the bat’s
perceptual acuity as its ability to determine unambig-
uously the timing (that is position) of the peak of the
ACR. The narrower the central peak, and the lower
the side peaks, the more acute the bat’s target ranging.
(b) From what has been said, it is clear that this will be
t achieved if the bandwidth is large (that is t is small),
the average fundamental frequency is high (T is also
T small), and FM harmonics are present (the side peaks
are of low amplitude). This is only part of the story—
that for a stationary target. Simmons and Stein (1980)
show how this call structure also minimizes errors
–400 0 400 due to moving targets.
time lag (ms) In Fig. 3.9a, the relationship between call band-
width and target-ranging acuity is shown for five
Figure 3.8 Autocorrelation functions: a) narrowband, pure tone b) species of bat. The horseshoe bat has a ‘constant
broadband with harmonics. T = average period, t  1/bandwidth (see
frequency’ call with only a short, narrow bandwidth
text for further details). (From Simmons and Stein 1980, with permission
from Springer Science + Business Media).
FM tail, and can only distinguish between two ob-
jects if one is more than 12 mm closer (or further
away) from the bat than the other (Fig. 3.9b). The
with multiple harmonics so well suited to fine phyllostomid, Phyllostomus hastatus, has a broad
discrimination? bandwidth call, and can resolve distances down to
almost 4 mm. This means being able to differentiate
The autocorrelation function between echoes arriving as little as 20 μs apart.
A bat’s ability to resolve small differences in the Early experiments involved training bats on a plat-
distances to two objects is related to the properties form to fly to the nearest target for a reward (for
of its echolocation call, in particular the bandwidth example Simmons and Vernon 1971), or by simulat-
and the average wavelength (or period) of the com- ing echoes with variable time delays (Simmons
ponent frequencies. These two properties can be 1973). Whilst it is doing these tests, the bat is moving
described by the autocorrelation function (ACR) of its head to investigate the targets, and this will
the echolocation pulse (Simmons 1971). I am not reduce its sensitivity since target distance changes
going to try to explain this function in detail; for as its head moves. To find out just what can be
those interested Simmons (1971) and Simmons and achieved in ideal conditions, Simmons (1979)
Stein (1980) give excellent accounts. The bottom line trained bats to discriminate between a simulated
is as follows. The bat behaves as if it stores the echo at constant delay of 3 ms, and one made to
emitted pulse (in the inferior colliculus of the mid- jitter, or oscillate, back and forth by a tiny distance,
brain), and cross-correlates it with the returning at a mean delay of 3 ms. Under these conditions the
echo. Figure 3.8 shows ACRs for two pulse types, a big brown bat, Eptesicus fuscus, could identify a
1 ms duration 5 kHz (25–20 kHz) bandwidth pulse target jittering by just 1.3 μs. This means that the
in (a), and a 1 ms, 30 kHz (50–20 kHz) bandwidth bat had to measure a delay of several ms with an
pulse with its second harmonic (100–40 kHz) in (b). accuracy of 1 μs, and then store that information
The ACR describes the average period T of the until it could compare it with the next pulse. In the
signal (the reciprocal of the average fundamental fre- more complex conditions encountered in the field
quency), which decreases with increasing frequency. the bat could not perform as well as this, but it gives
72 BATS

(a)
80
range difference (mm)
12 Rhinolophus ferrumequinum

1/B
time difference (ms)

60

8 Pteronotus
suapurensis
40
Eptesicus fuscus Phyllostomus
hastatus
4 Eptesicus fuscus
20

0 0
0 10 20 30 40
bandwidth (kHz)
(b)

range = x mm
Figure 3.9 Relationship between call bandwidth and target ranging
targets acuity. Target ranging resolution should theoretically be equal to the
range = y m
m reciprocal of the bandwidth, the dotted line in the figure. The four
species studied by Simmons (1973) have target ranging abilities very
close to those expected. The lower diagram shows what is meant by
range difference = y–x mm resolved difference distance.

you an idea of the sophistication of the system, and surface (Simmons et al. 1974). The differences may
this remarkable sensitivity almost certainly has an be as little as 0.6 mm, and Simmons et al. originally
important role in other aspects of echolocation, such suggested that the bats were performing this task on
as determining target fine structure. the basis of spectral differences in the echoes—you
could think of it as a change in the colour composi-
Spectral changes in FM echoes—determining target tion. The remarkable resolution of pulse-echo delay
fine structure timing, suggested by the jitter experiments, later led
As I said earlier, FM pulses are, in theory, good at Simmons (1979) to suggest that discrimination may
determining the fine structure of a bat’s surround- be based on temporal cues. The results of some of
ings—a communications expert will tell you that these jitter experiments have been hotly debated
more frequencies carry more information. FM sig- (Pollak 1993, Simmons 1993). Schmidt (1988) has
nals carry this information back to the bat because shown that the Indian false vampire, Megaderma
the echo is substantially changed by the target. FM lyra, can learn frequency spectra from objects with
bats gather information from spectral changes— different textures. When given two simulated ech-
modulation of the frequency components, and oes which differed from the learnt spectrum, the
amplitude changes—modulation of the strength of bats selected the pattern closest to the real one. The
the echo. As yet, we have a poor understanding of broadband FM call (100–20 kHz) used by M. lyra
how this is done, but we do know that vesper bats enables it to analyse textural depths (that is surface
of the genera Myotis and Eptesicus can discriminate roughness) in the range 0.9–4.2 mm. The narrow-
between objects of different size, shape, and even band FM tail to the call of greater horseshoe bats will
texture. Differences in size appear to be determined be limited to 1.1–1.5 mm.
from amplitude differences in the echo, shape and
textural differences must involve more subtle mec- Spectral change as a detector of movement
hanisms. Eptesicus fuscus can discriminate between It may be that spectral changes between FM calls
flat plates that have holes of different depths on their and their echoes are used to detect movement in
 ECHOLOCATION AND OTHER SENSES 73

cluttered environments. The complex spectrum re- behavioural experiments (Fig. 3.10) in which bats
turning from a leafy branch will carry no specific were trained to discriminate between horizontal or
cues about potential prey. However, even the slight- vertical arrays of rods which subtended different
est movement by an insect, either flying past, or angles at the eye.
moving on a leaf, will change the echo spectrum The big brown bat, Eptesicus fuscus, was capable
‘colour’ in a characteristic way, different to the of resolving vertical angles down to 3.5 , and hori-
changes due to movement of the branch by the zontal angles to 1.5 . Horizontal position may be
wind. In theory, a bat echolocating at around 20 determined by differences in the time of arrival
kHz will detect textural changes (movement) of and/or the intensity of echoes between the two
around 200 μm, and at 150 kHz resolution would ears (Fig. 3.11). Analysis by differences in the time
be a remarkable 5 μm, for a frequency shift of just 1 of arrival would have to work on delays of 1–2 ms
per cent (Neuweiler 1990). The movement of the bat between the ears. The exact mechanism is still
itself introduces spectral changes which could com-
plicate analysis. Is this why bats such as M. lyra and
the long-eared bat, Plecotus auritus, often hover
subtended angles
motionless when gleaning? Bats that feed over
water may also detect prey by changes in echo spec-
tra from the textured water surface, since all use
brief broadband FM signals (for example Myotis
daubentonii and M. adversus, Jones and Rayner
1988, 1991; Noctilio leporinus, Wenstrup and Suthers
1984; Myotis vivesi, Suthers 1967). Interestingly, Figure 3.10 Locating objects in 3-D space. Experimental setup for
determining bats’ ability to resolve horizontal (or vertical) angles. Bats are
Megaderma lyra, which as we’ve noted uses broad-
trained to identify the larger (or smaller) angles subtended by vertical (or
band calls, also fishes (Suthers 1967). horizontal) arrays of rods, and signal their choice by moving onto one of the
Neuweiler (1990) points out that there has been forward platforms.
very little work on the auditory analysis of changing
spectral patterns, and suggests that rich rewards
await those entering this field. I confess to not pur-
suing this in updating the book and leave it to
interested readers.
Before moving on, a note of caution: what bats can
do in theory may be quite different from what they
do in practice. The utilization of FM echolocation,
including spectral cues, appears to be severely com-
promised in cluttered environments and several
species appear to locate prey by listening for sounds
generated by their movement, rather than by det-
ecting movement using echolocation (for example
Arlettaz et al. 2001).

Target location in three-dimensional space

So far the FM bat has determined the distance to the
object, its size and shape, and something about its
texture. It needs several more important bits of
information: an object’s location in the horizontal
Figure 3.11 How bats resolve horizontal components of position in
and vertical planes, and its relative velocity and space. Horizontal displacement is probably determined by differences in
direction of movement. Lawrence and Simmons the intensity and/or time of arrival of echoes at the two ears, due to
(1982b) and Simmons et al. (1983) carried out differences in echo path length (dotted lines) between object and ears.
74 BATS

uncertain, and the basic arguments are discussed by suggested that they do, but the patterns of head
Suthers and Wenstrup (1987). When the tragus is movements of bats during pursuit, although not
temporarily displaced (by the researcher) vertical conclusive, suggested that they do not (Masters
acuity is significantly reduced and bats are less suc- et al. 1985). More recent work (Ghose et al. 2006)
cessful at capturing tethered prey (Chiu and Moss, not only shows that they can, but also suggests
2007), but the success rate is still better than 5 per that they do it in such a way as to minimize the
cent and the bats quickly learn to compensate. The time-to-capture of fast-flying and erratic prey,
tragus is known to alter the spectral composition using a strategy previously unknown in the animal
and direction of sounds entering the ear and cues kingdom. Many previous studies have shown that
from these changes may help Eptesicus to resolve an approach known as the constant bearing strategy
vertical angles in ways that remain to be fully is widespread in animals chasing prey, and even in
understood. humans, for example when catching a ball. The
External ears accentuate direction-specific differ- animal holds constant the angle between its heading
ences and, together with the tragus, improve bats’ and the target as it homes in on the target and, at the
ability to determine direction (see for example same time, tries to maintain a straight line course.
Aytekin et al. 2004). Can this combination also help However, this only works when pursuer and pur-
a bat recognize an object and determine its orienta- sued are moving at constant speeds, or at least when
tion? Glossophaga soricina is known to use echoloca- the speed of the pursued does not change too often
tion when seeking and feeding at flowers of a or too quickly, so that the purser has time to com-
tropical vine, Mucuna holtonii (Helversen and pensate. It does not work well for targets that
Helversen 1999) and in the laboratory the same spe- change speed and direction rapidly. Ghose and col-
cies has an ability to sort objects by shape indepen- leagues have shown that bats make use of a constant
dent of size. How do they do this? Using an artificial absolute target direction (CATD) strategy. The
bat head of an ultrasound loudspeaker and two tiny speeds of pursuer and pursued can be resolved
microphones, Holderied and Helversen (2006) were into two components, one parallel to an imaginary
able to show that differences in the echoes received line linking them and the other perpendicular to this
from objects, including the flowers of a bat-polli- line, the transverse component. In the CATD strat-
nated plant, carry information about both shape egy, the transverse velocities of the bat and insect
and orientation. However, a bat faces substantial are matched, keeping constant the absolute direc-
difficulties if it is to be able to utilize this informa- tion to the target, however erratic the insect’s flight
tion. For example, a bat can only make good use of is. Although not previously described in animals,
this information if it can separate cues related only the strategy is known in military circles as parallel
to the shape of the object from those due to the navigation and is used in guided missiles, which
direction of its approach. Possible solutions to this have onboard sensors and computers to enable
and other tasks have been suggested, but remain them to perform this complex task. But how do the
theoretical. bats do it? When a bat is chasing a target, its head
locks on to the target. When the bat is using a CATD
Can bats intercept their prey? strategy, the absolute direction of its head is held
An FM bat appears to determine prey velocity pri- constant, despite the fact that its body may change
marily by measuring the change in distance between orientation and speed. The bat could maintain
it and its prey from one echo to the next. This facility CATD by manoeuvring to cancel any changes in
is clearly important in the approach and capture head direction that its vestibular system senses. Tar-
phases of the chase, and could be used by bats to get ranging through its echolocation system will tell
plot prey trajectory, and hence fly on an interception it if it is approaching its target as it maintains CATD.
course. There has been some debate in the past as to Other mechanisms are suggested by Ghose and col-
whether bats can and do fly on interception paths. A leagues, but this strikes me as the most plausible and
study of pursuit flight sequences by Webster (1967) elegant solution.
 ECHOLOCATION AND OTHER SENSES 75

Automatic gain control described above, are to give the bat more informa-
As a bat approaches its prey and the echoes get tion about the position and nature of the target. Why
louder, its hearing becomes less sensitive (Kick and is the CF component there at all? CF echolocation
Simmons 1984) due to the action of the middle ear did not evolve without reason, so how do bats uti-
muscles (Suga and Jen 1975). In combination with lize these pulses? There are several possible func-
accurate aiming of the head and adjustments to the tions for CF calls, and over the last 30 years evidence
intensity of the emitted pulses (for example Kobler has been gathered in support of some subtle and
et al. 1985; Hiryu et al. 2007) this acts as an automatic remarkable uses. Note also that although pulse
gain control, keeping the neural stimulation due to duration decreases during a feeding buzz, many
the echoes at a relatively constant level. Boonman species with long CF components to their calls
and Jones (2002) have shown that in Myotis dauben- use an auditory processing system that is tolerant
tonii at least, the initial intensity of the call and of pulse-echo overlap. This is possible because
intensity compensation during approach is stereo- bats in the families Rhinolophidae and Hipposider-
typical and not dependent on auditory feedback: it idae (and the New World moustached bat, Pterono-
is independent of target size and therefore echo tus) have an acoustic fovea: a region of the cochlea
strength. that is extremely sensitive to the echo frequency
of their calls. Pulse and echo are separated in the
Constant frequency calls frequency—rather than the time-domain (Neuweiler
1990).
Basic properties I have already noted that high frequency sounds
At the other end of the range from short, broadband are attenuated rapidly in air. Two ways to make the
FM sweeps are the long, constant frequency (CF) calls pulse carry further are to concentrate all, or most, of
used by many bats. They are typically 10–50 ms in the energy into one frequency, rather than spread it
duration, and are rarely entirely CF, since they often over a broad FM pulse, and to make this component
have brief, narrowband FM components at one or relatively low frequency. Many bats have a promi-
both ends (Fig. 3.6). CF pulses show a lot of interspe- nent CF or shallow FM component to their search
cific variation and are more accurately referred to as phase call, even species we think of as FM bats (for
CF/FM or even FM/CF/FM pulses. example Pipistrellus species). When a CF or shallow
CF/FM pulses commonly undergo similar chan- FM component is present, it is often as a low fre-
ges to FM pulses as the bat homes in on its target. The quency tail to the FM sweep. Some bats alternate CF
pulse gets shorter, by a reduction in the central CF or narrowband FM pulses with broader FM pulses
component, and the FM component increases in (for example Nyctalus species). Examples of these
bandwidth (Fig. 3.12). Additional harmonics may calls are shown in Fig. 3.6. In general this trick is
also be introduced. used by bats that hunt in relatively open environ-
The pulses get shorter to prevent pulse-echo over- ments, and the low frequency CF or narrowband
lap, and the changes in the FM component, as call may help them to detect prey simply by

search phase approach

100 terminal
frequency (kHz)

capture

0
0 0.2 0.4 0.6 0.8 1.0 Figure 3.12 Call structure during prey capture in a CF
time (s) bat.
76 BATS

maintaining the intensity of the pulse, and hence 84

that of the echo, over greater distances. This is not predicted Doppler-

frequency (kHz)
the whole answer, since many bats using long CF shifted echo
pulses (for example Rhinopoma hardwickei) include 83
several energy-draining higher harmonics in the
call. Another suggestion, discussed below, is that real echo
the CF pulses are designed to detect fluttering in-
82
sects. This may be plausible for some species, but in
0 1 2 3
many the percentage of total time occupied by calls
time (s)
is so small that the detection of fluttering wings is
unlikely (Neuweiler 1990). Figure 3.14 Doppler shift compensation in the greater horseshoe bat.
The dotted line shows the predicted shift in the echoes returning to the bat
Doppler shift due to the movement of a simulated target. The solid line shows the
A phenomenon called Doppler shift appears to be actual echoes. As the target moves and the echo frequency changes, the
bat shifts its emitted frequency to keep the echo frequency constant (based
exploited by some bats that use CF pulses. Everyone
on Simmons 1974).
is familiar with the way in which the sound of a
siren changes as a police car or ambulance passes in
the street. As it approaches it is high pitched, and the
sound suddenly becomes low pitched as it passes A greater horseshoe bat emits its FM/CF/FM
and recedes. This is due to Doppler shift (Fig. 3.13). pulses at a frequency of about 82 kHz. Let’s say a
As the vehicle approaches, the sound waves from pulse hits a target towards which the bat is moving
the siren are pushed against your ears due to the at 5 ms1. The pulse will be compressed and the
motion of the vehicle. The waves are compressed returning echo will be logged at 84 kHz by the bat.
like a spring, decreasing the wavelength, increasing It will shift its emitted pulse down to 80 kHz, so that
the frequency, and therefore increasing the pitch. As the returning echo is logged at 82 kHz again. Figure
the vehicle passes, the sound waves are suddenly 3.14 shows an example of a bat responding to a
stretched by the source of the sound as it goes away simulated target that suddenly moves towards it at
from you, the wavelength increases, the frequency 2 ms1 for 2 s, and then stops again (Simmons 1974,
decreases, and the pitch goes down. The change in Simmons et al. 1975).
pitch is directly related to the speed at which the The bat alters its emission frequency to keep the
source approaches or recedes. If an echolocating bat echo at 82 kHz. It does this because its ears are tuned
is approaching an insect, the frequency of the echoes to be particularly sensitive to sounds of 82–83 kHz
will be determined by the combined speeds of the (Neuweiler 1970) and a large part of the cochlea is
bat and insect. Echoes received by a bat approaching given over to this frequency range (Bruns 1970). The
a stationary insect will be increased in frequency, shift in emitted frequency required to keep the echo
and increased further if the insect is approaching the at 82 kHz tells the bat how fast it is moving, and in
bat. Some bats make use of Doppler shift to calculate what direction, relative to the object. A similar
their speed relative to objects around them. mechanism operates in the moustached bat,

Figure 3.13 How Doppler shift works.

 ECHOLOCATION AND OTHER SENSES 77

Pteronotus parnellii, and probably in many other CF Fenton 1984, Link et al. 1986). The echolocation sys-
bats (see Neuweiler 1990 for a detailed review). This tem of such bats is extremely sensitive to minute
should be an ideal mechanism to compute the speed frequency modulations, suggesting that they are of
of insect prey, but Doppler shift compensation is functional importance (Vater 1982). Recent work
used by horseshoe bats to relate their own motion also shows that bats may learn rather more from
to the environment (Trappe and Schnitzler 1982): the echoes than simply that an insect is present.
the movement of prey is probably determined by Some of the most elegant studies have been done
pulse-echo delay measurements. on the greater horseshoe bat, Rhinolophus ferrumequi-
Noctilio leporinus uses CF/FM pulses to locate fish num. With a 50 ms CF call, it could monitor a com-
at the surface of the water. It is tempting to suggest plete wingbeat cycle of any insect beating its wings
that Doppler shift of the CF component is used for at a frequency greater than 20 Hz.
prey movement detection. Wenstrup and Suthers Figure 3.15 shows the changes in echo intensity,
(1984) trained bats to approach moving targets in or amplitude modulation, of the CF pulse during
an indoor pool. The bats could discriminate between two wingbeat cycles of a moth flying directly
two targets with a velocity difference of 35–45 cm s1. towards the bat and perpendicular to the bat’s
Were they using Doppler shift, or simply the rate at path (Schnitzler et al. 1983, Schnitzler 1987). The
which target distance changed from pulse to pulse bat could get quite a lot from this information.
to detect prey movement? When bats were trained First of all, it could determine the wingbeat fre-
to respond to a simulated Doppler shift in the ech- quency from the period of the repeating pattern. It
oes, the best they could do was to detect an apparent could also work out the orientation of the insect
velocity difference of 170 cm s1. They therefore from the rather different patterns produced. You
appear to calculate velocity by using updated dis- can see that, when flying across the bat’s path, the
tance information from the FM component of their large wing area of the moth is presented face on to
pulses. The Doppler shift mechanism is clearly the bat and produces high intensity echoes or glints.
important in determining prey movement, and To these can be added the FM modulation due to
even velocity in some species, but care must be wing movement. This FM and AM modulation of
taken not to overstate its importance. the echo relative to the emitted pulse is species-
specific, as well as being determined by prey
Prey detection and identification using CF calls orientation.
Doppler shift may have another important role to We can see this more clearly in the time-averaged
play in prey detection and identification. You can spectra shown in Fig. 3.16 (Schnitzler et al. 1983).
think of an insect as an acoustic mirror, reflecting These show the relative intensities of the different
sound back towards the bat. When it presents the frequency components in the echoes. There is a
flat face of its wings to the bat it acts as a better strong component around 80 kHz, close to the emit-
mirror than when it presents its wings edge on: in ted frequency, in the echoes of all species, but more
other words the echo will be more intense. This striking are the many differences related to the size
means that the strength of the echo will change as and shape of the insect and its wings, the wingbeat
the insect beats its wings up and down, and if the CF frequency, and the patterns of wing movement.
pulse is long relative to the insect’s wingbeat period, Bats could in theory make use of all this
the bat may be able to detect and time these ampli- information—do they? The answer appears to be
tude modulations. In addition to amplitude modu- yes: trained bats in the laboratory are capable of
lation, the moving wings will also lead to frequency some astoundingly subtle discriminatory feats.
modulation of the echo, through Doppler shift, as Greater horseshoe bats have been trained to make
the wings move towards and away from the bat. Do a choice between pairs of real, simulated, or mod-
bats take advantage of these effects? It is certainly ified echoes from insects (von der Emde and Schnit-
the case that bats respond better to moving targets, zler 1990). The echoes were manipulated so that the
in particular fluttering insects. Several CF bats show ‘phantom’ targets were all fluttering at exactly
a strong preference for fluttering targets (Bell and 50 Hz, so that the bats could not use frequency as a
78 BATS

(a)

(b)
0° 90°

downstroke

upstroke

downstroke

upstroke
10ms

amplitude

Figure 3.15 Amplitude and frequency modulation of a CF echolocation call by fluttering insects. a) The movement of the insect’s wings leads to
frequency modulation of the echo by Doppler shift (represented by changes in line spacing on the echo). Changes in wing profile, and hence the intensity
of the echo, lead to amplitude modulation (represented by changes in line thickness). b) Amplitude modulation of the CF call of a greater horseshoe bat
over two wingbeat cycles of a moth, flying towards and at right angles to the bat (based on Schnitzler et al. 1983).

guide to identification. The echoes of four different head-on and back view cranefly echoes—echoes
insects were used, a sphingid moth, a noctuid moth, they had not so far experienced. Despite this they
a scarabid beetle, and a cranefly, from three different still went for the cranefly. Whatever they had learnt
orientations: head-on, side view, and back view. In about side-on cranefly echoes, they could appar-
the first experiment, the bats were given a reward if ently see the same characteristics in very different
they picked out a side-on cranefly echo in preference echoes from craneflies oriented at other angles. The
to any other echo from the other insects. This they experimenters themselves could identify no com-
did consistently after a short period of training. mon characteristics between echoes from a study
Without any further training they were given a of the sonograms and oscillograms. Interestingly,
choice between echoes from the other species and two of the four bats were rather better at this test
 ECHOLOCATION AND OTHER SENSES 79

50 50
Ichneumonidae Tachinidae
sound pressure (dB)

0 0
78 80 82 78 80 82

50 50
Campaea margaritata Dolichovespula saxonica

Figure 3.16 Time-averaged echospectra

from four flying insects. Each species has a
0 0
spectrum with a characteristic frequency
78 80 82 78 80 82 range of different intensities (based on
frequency (kHz) Schnitzler et al. 1983).

Figure 3.17 Leisler’s bats, Nyctalus leisleri, feeding around a streetlamp.

80 BATS

than the other two. When the echoes were manipu- components of their own CF/FM pulses. However,
lated, to remove either the large amplitude glints or if simulations of the entire CF/FM pulse were used,
the lower amplitude events between glints, bats the bats performed the discrimination tests badly.
behaved differently. One bat could identify the cra- The researchers went on to perform experiments
nefly only on the basis of the information between that suggested N. albiventris could only receive and
glints: with glints only it performed the test badly. process the FM component, to determine target
However, another bat could use either: different range, if it was preceded by the CF component
bats must learn different techniques. If the echoes with the appropriate temporal spacing. The CF com-
were manipulated to remove either FM or AM com- ponent of the CF/FM pulse is presumed to open
ponents the bats could not identify the cranefly a neural gate of short duration, enabling the bat to
echoes. process the echoes from its own pulses. The CF/FM
pulses or echoes of other bats are not likely to be
heard during this brief window, and are therefore
How do bats avoid being confused by
ignored. Finally, it is also becoming clear that in
other bats’ sonar? many species each bat has its own ‘personal’ call
In real life, there are all sorts of extraneous sounds, frequency—I’ll give some examples later in the
including the sonar of other bats, to interfere with a book.
bat’s own sonar processing (Fig. 3.17). There is no
doubt that such factors do influence a bat’s perfor- Auditory adaptations and the neural basis
mance level, but laboratory studies have shown that
of echolocation
echolocation has evolved to overcome many of the
problems. This is a large and complex field, but I’ll We saw earlier that the auditory system of bats has
give you an example or two of work in this area. a number of adaptations for picking up high fre-
The moustached bat (Pteronotus parnellii) over- quency sounds, notably the large basal turn of the
comes the interference caused by other bats’ calls by cochlea. Few features of the cochlea are unique,
suppressing the first harmonic in its sonar pulse to since many small mammals have good high fre-
around 1 per cent of the total energy of the call. It is quency hearing. Many of the important adapta-
then so weak that other bats are very unlikely to hear tions which enable bats to use ultrasound for
it. However, the bat hears its own first harmonic echolocation are in the auditory centres of the
directly through the tissues between vocal chords brain. Some of the more important adaptations
and cochlea. The first harmonic is used to open a will be described in this section. Neuweiler (1990)
timed neural gate that enables the bat’s auditory suggests that the key to understanding neural pro-
system to receive and process the echo from that cessing in echolocation lies in the hypothesis that
call. The bat does not hear and respond to the weak the emission of the echolocation pulses triggers
first harmonics of other bats, so they do not therefore specific mechanisms in the brain that enable bats
open the neural gate that initiates processing: it is not to perform their remarkable sonar feats. In the
therefore confused by the presence of other echolo- previous section I described two mechanisms by
cating bats. which bats minimize interference due to the
A similar mechanism is used by the small fisher- sounds of other bats. Both mechanisms depend
man bat, Noctilio albiventris. Roverud and Grinnell upon the bats’ own echolocation pulses triggering
(1985) trained bats to discriminate a target range neural gates that briefly open windows enabling
difference of 5 cm and then attempted to confuse the auditory system to receive and interpret echoes.
the bats with recorded sounds. These bats emit There is evidence to suggest that mechanisms simi-
paired pulses at 7–10 Hz: an 8 ms CF pulse followed lar to those used against interference are important
by a CF/FM (6 ms/2 ms) pulse. The bats were found in target ranging, and researchers are looking to see
to be insensitive to a wide range of potential jam- if time windows are important in other echoloca-
ming sounds, including simulations of CF or FM tion tasks. Considerable effort is also being put in
 ECHOLOCATION AND OTHER SENSES 81

to uncovering other auditory adaptations. This with maximum sensitivity between 0.7 and 2.4 m.
research has been reviewed by O’Neill (1987), The bat can therefore attend selectively to its near
Roverud (1987), Pollak and Casseday (1989), Neu- environment, and ignore echoes from more distant
weiler (1990), Suga (1990), Moss and Sinha 2003, objects. Roverud reports that a similar mechanism
and Neuweiler 2003: I can only give you a flavour operates in the horseshoe bat Rhinolophus rouxi,
of what this research is telling us about the way a which uses long FM–CF–FM pulses, and the big
bat’s brain has evolved to meet the stringent re- brown bat, Eptesicus fuscus, which emits short FM
quirements of echolocation. However, we must pulses only. The time window in all three species
not forget that the brain of a bat is fundamentally was open for around 30 ms, allowing the bats to
like that of other mammals, and adaptations for target range over distances up to 5 m, more than
echolocation are frequently exaggerations or refine- sufficient distance for detecting and catching prey
ments of existing mammalian structures and pro- on the wing.
cesses that make them more accessible to the
scientist. The lessons being learnt from bats there-
Target ranging by pulse-echo delay
fore often have much wider relevance, increasing
our understanding of how the mammalian brain is The ability of bats to measure the distance to an object
organized and functions. from the pulse-echo delay has been described above.
Two types of mechanism appear to exist in the brain
Paired sounds—enhanced neural response of a bat to process target-range information. In the
auditory centre of the mid-brain of FM bats (in a
to the second sound
region called the inferior colliculus) up to 50 per
In most mammals, the neural response to a sound cent of the cells sampled are of a type called ‘phasic
(for example the firing of specific cells in the brain constant latency responders’ (pCLRs). This simply
that trigger target-ranging computation) is usually means that they respond to impulses from the ear
suppressed by a simultaneous or slightly earlier (irrespective of their intensity) by firing rapidly and
sound. This is also true for bats, with the exception briefly (Suga 1970, Pollak et al. 1977a,b). They can
of horseshoe bats, when the sounds are around their encode very precisely the timing of pulse-echo pairs
best frequencies—that is, the frequencies of their CF for processing higher in the brain. These temporal
calls and echoes (Möller 1978). A tone around 81.5 pattern-encoding neurons are uncommon in CF
kHz dramatically increases the response to a second bats, but another type is found in the mid-brain and
tone of 82 kHz. The first can be thought of as the auditory cortex of both CF bats (Suga et al. 1978) and
emitted pulse, the second the weak echo of slightly FM bats (Feng et al. 1978): the range-tuned neurons
higher frequency due to Doppler shift (see CF calls (see O’Neill 1987, for references and review). These
earlier in chapter). This mechanism is almost cer- cells only fire if they receive a pair of impulses (from
tainly one that has evolved to selectively enhance the sonar pulse and its echo) with a specific delay
the brain’s response to the weak echoes of the bat’s between them. Every cell is programmed to respond
CF calls. The enhanced response would work for to a certain delay, and the higher centres of the brain
echoes returning from up to 9 m away. will determine what the pulse-echo delay is by which
cell fires.
In the moustached bat, Pteronotus parnelli, these
Time windows and target ranging
cells make up a two-dimensional map on the surface
The neural gate used by Noctilio albiventris (Roverud of the auditory cortex (Suga and O’Neill 1979,
1987) not only acts to filter out extraneous sounds, it O’Neill and Suga 1982), and are divided into three
also acts as a distance filter for the bat’s own sounds. types, each sensitive to the delay between the fun-
The window is open for 21 ms after the bat’s own damental frequency and one of the three higher
FM component (27 ms after the CF gate opener) that harmonics present in the FM component of the call
is, to echoes coming from targets up to 3.6 m away, of the bat (Fig. 3.18). Delays of 2–8 ms take up most
82 BATS

5 Dealing with sound intensity variation

6
8 A bat must be able to process its own pulses at
10
FM1–FM3 14
16
80–130 dB, followed by the echo, typically up to
4 FM1–FM4 only 40 dB (and remember that dB are on a logarith-
18ms
2
1
mic scale). We saw earlier how the stapedius muscle
comes into play here, but there are other mechan-
FM1–FM2
isms. The pulse itself may increase the response of the
auditory system to the echo (Grinnell 1963, Möller
1978). The grey bat, Myotis grisescens, attenuates its
Figure 3.18 A neural map of range-tuned neurons in the auditory response to its own calls by 25 dB relative to record-
cortex of the moustached bat (based on O’Neill and Suga 1982). ings of its call (Suga and Schlegel 1972). The inten-
sity of the echo itself will increase as the target is
approached. Bats can smooth out variation in the
of the map, corresponding to target ranges of response to the echo by reducing the intensity of the
0.35–1.4 m. These cells are not too sensitive to the call as they approach the target (Kobler et al. 1985),
FM frequency of the echo, so are not upset by Dopp- or by increasing the threshold for echo detection
ler shifts, and they are either on or off. In other (Kick and Simmons 1984). Many neural responses
words, they are frequency and level tolerant, but (at all levels of the system) have upper thresholds,
not delay tolerant. This is just what is required of and do not respond differently to sounds over a
cells whose function is to encode target ranging, wide range of intensities (for example Suga 1977,
pulse-echo delays. Although level tolerant, most Vater 1982), a very simple way of dealing with
neurons respond to echoes within a particular changing intensity signals.
range of amplitudes—that is from a target of a par-
ticular size. Why the delay between fundamental
and harmonics is so important is not fully under- Echolocation and behavioural studies
stood, but the same phenomenon is seen in Rhinolo- in the wild
phus rouxi (O’Neill et al. 1985). One function is in the
How good is echolocation in the real world?
anti-interference strategy of the moustached bat
described in the last section. The FM little brown Is there any evidence that bats in the field can take
bat, Myotis lucifugus, has range-tuned neurons advantage of some of the subtle abilities they show
responsive to FM pulses with identical spectra (Sul- in the laboratory? Until recently it has been very
livan 1982). difficult to study echolocation in the field, but tech-
Going one step further into the system, researchers nical advances have now made this possible. The
are asking what processes lie behind range-tuned results of such studies suggest we have only just
neurons. One much studied mechanism involves begun to understand the degree of sophistication
delay lines in neural networks (for example Olsen of bat echolocation! I will illustrate this with a recent
and Suga 1986, Suga 1990). The network delays the example: how do bats navigate accurately around
response of a cell in the mid-brain to the pulse, so that objects despite the difficulties introduced by their
it arrives at the range-tuned neuron in the cortex at own rapid movement? This is important because
the same time as the response to the echo, which gets many bats fly close to objects, either to catch prey
there unimpeded. The delay can be introduced by a or to reduce the risk of being eaten themselves—but
series of interneurons, each with an axon and a syn- the risk of collision, injury, or even death is poten-
apse with finite transmission times. The echo response tially high. Holderied et al. (2006) used an array of
is carried more or less directly to the range-tuned microphones to track the flight paths of individual
neuron. By altering the length of the delay line, or by whiskered bats, Myotis mystacinus, as they flew
direct neural inhibition, range-tuned neurons sensitive along a hedge. The position of a bat at the time of
to different delays can be created. echolocation call emission could be calculated from
 ECHOLOCATION AND OTHER SENSES 83

the different arrival times of the call at each of the

microphones. A bat’s ability to accurately determine distance of focus
its distance from the hedge is impaired because it
changes position between call and echo and because
the echoes themselves are modified by Doppler shift.
Since the bat moves closer to the target between hedge
pulse and echo, the distance to the target calculated
from pulse-echo delay is underestimated. The Dopp-
ler shift-induced increase in echo frequency will lead
to an overestimate of target range, since it shifts the 1m
position of the cross-correlation function peak 2m
(described earlier in the chapter). Holderied and his
colleagues were able to show that the bats changed signal overlap zone
call design during flight such that these two sources
of ranging error cancelled each other out. But
the complexity does not end there. The Doppler
shift overestimation is dependent on call design, hedge
but independent of distance to target. The under-
estimation due to the change in position between
pulse and echo increases with target distance. So,
for a particular call design, the ranging errors will
only be cancelled for targets at a particular distance
from the bat: the Distance of Focus (DOF). The bats
altered call design during flight so that the DOF
extended only to the nearest point of the hedge Figure 3.19 Vertical sections through a hedge and flight paths
through adjacent airspace. Each circle represents the distance of focus
(Fig. 3.19). In this way, target ranging was most
(top) or signal overlap zone (bottom) of individual bats (at the centre of
accurate for those objects of greatest importance to each circle) as they passed through the plane of measurement. See text for
the bats. details (adapted with permission from Holderied et al. 2006).
How is call design manipulated? Doppler shift
ranging errors are lowest in short, broadband calls,
Do bats select their prey?
so bats can alter independently call duration and/or
bandwidth to change the magnitude of the ranging We can put the question we asked at the beginning
error. Furthermore, hyperbolically modulated calls of this section in a less direct, but perhaps more
are Doppler-tolerant, in that Doppler shift does not functionally relevant way: does echolocation allow
alter ranging acuity (precision), although they are bats to select their prey, or are they taken at random,
still subject to ranging error. Since bats do not in proportion to their abundance in the environ-
make use of hyperbolic calls at all times, this raises ment? If they are able to select, is it done on crude
the possibility that bats are also making use of criteria like size, or are they more sophisticated in
changes in Doppler tolerance. their prey selection? Do they select insects that rep-
There is a final twist of sophistication. The edge of resent the best energetic returns for the effort put in
the Signal Overlap Zone (SOZ) tracked the DOF. to catch them? In other words, are they foraging
The SOZ is the space around the bat (determined by optimally? This is an ambitious question of funda-
pulse duration and distance to target) where pulse mental ecological importance, and one I’ll come
and echo will overlap, impairing echolocation per- back to later in the book. For the moment, in the
formance. The first objects outside the signal overlap context of their sensory abilities, all we need to ask
zone were those at the DOF and therefore most is: can bats use echolocation to select prey in the
accurately localized. wild?
84 BATS

Figure 3.20 shows data on feeding preferences of The answer in the case of Scottish pipistrelles is
the little brown bat, Myotis lucifugus, collected by yes, some of the time, for some insects. Figure 3.21
Anthony and Kunz (1977) in New Hampshire. The shows the percentage of insects in droppings plotted
upper histogram shows the percentage of insects against those found in suction traps, for three differ-
available to the bats, divided into classes of different ent insect groups (Swift et al. 1985). Each dot repre-
sizes (that is, the insects caught in suction traps). The sents one night’s data. If the dot falls above the line,
lower shows those eaten by the bats (that is, the then that group of insects are over-represented in
insects found in droppings). There are many more the diet, assuming the traps themselves are not
small insects available to the bats than those in selective (incidentally, a point of continued debate
larger size classes, but they eat mainly those falling whatever the trapping method used). Mayflies
into the middle size classes. Although there are very appear to be actively selected, but moths are
few large insects available, they make up a signifi- under-represented. Midges are taken in proportion
cant part of the diet, so the bats must be selectively to their availability. As we will discover later, some
feeding on them. One factor limiting the upper size moths can hear bats and take evasive action, and
limit may be the physical difficulty involved in some will simply be too large for the small pipis-
catching and eating large prey. Even if the bats can trelle to handle.
do it, it may be less economical in energetic terms Selection is also seen in the hoary bat, Lasiurus
than handling smaller prey. Small prey may also be cinereus. The insects eaten are shown in Fig. 3.22 as
uneconomical, and detected but avoided. The very a percentage of the total diet, together with the
smallest prey may not even be detected, so the ques- insects available, assessed by sticky and Malaise
tion of selection doesn’t come into it! So, some bats traps (Barclay 1985, 1986). Moths made up only 10
appear to select by size. Can they also select by any per cent of the insects caught in traps, but consti-
other criteria? tuted 40 per cent of the bats’ diet, and a similar
pattern was seen for dragonflies and beetles. Dra-
gonflies don’t fly at night—so how was the bat
80
finding and taking these? The bats appear almost
to ignore mosquitoes, flies, and caddis flies.
% available by number

60
100
40 midges

20 11
% by number in faeces

0 mayflies
0 100
60
% in diet by number

5 moths
40

20
0 5
0 0 4
0–2 2–4 4–6 6–8 8–10 >10 % by number in traps
size class of prey (mm body length)
Figure 3.21 The diet of the pipistrelle, P. pipistrellus. Each graph shows
Figure 3.20 Prey selection by size. Insect size classes available to, and the number of insects available (i.e. caught in suction trap), as a percentage
eaten by, the little brown bat, Myotis lucifugus (based on Anthony of the total, plotted against the number eaten (determined from faecal
and Kunz 1977). analysis) (based on Swift et al. 1985).
 ECHOLOCATION AND OTHER SENSES 85

60

50
available

eaten
percentage by weight

40

30

20

10

0
others moths dragonflies beetles mosquitoes muscoid caddis flies Figure 3.22 The diet of the hoary bat, Lasiurus
flies cinereus (based on Barclay 1986).

Insect availability of course changes with the sea- They should certainly be able to separate these from
son, and this is reflected in the diet of some species. smaller insects with faster wingbeat frequencies.
Figure 3.23 shows the composition of the diet of the There must be more to it than wingbeat frequency,
greater horseshoe bat, Rhinolophus ferrumequinum, however, since one of the greater horseshoe bat’s
determined by Jones (1990): it includes some of the favourite items, the cockchafer, has a wingbeat fre-
largest flying insects found in the U.K. In early quency similar to several abundant but neglected
April, 50 per cent of the diet was dor beetles, Geo- species: perhaps the experiments of von der Emde
trupes, and over 30 per cent ichneumon flies. Later in and Schnitzler (1990) described earlier are relevant
the month, craneflies, Tipula, dominated, constitut- here?
ing up to 60 per cent of the diet. Cockchafers, Melo- How do these field results relate to laboratory
lontha, are eaten in May, but from mid-May to early experiments? It is not too difficult to believe that
July moths are the big thing, being slowly replaced bats in the field can select by size much as they do
by dung beetles, Aphodius. A casual study might in controlled experiments, but what about the com-
suggest opportunistic feeding: the bats simply eat plex task of learning frequency spectra carried out
what is most abundant. However, when the data by Megaderma lyra, or the interpretation of AM/FM
were analysed it was shown that the bats were modulated CF carrier waves by greater horseshoe
actively selecting certain prey, taking them in bats? It is unlikely that these bats have to rely on this
greater numbers than predicted on the basis of sort of learning procedure in the wild, but they may
their abundance in the environment. The long CF have a head full of inherited spectra that tell them
sonar pulse of the horseshoe bat could allow it to what to feed on. Alternatively, spectral analysis may
determine the wingbeat frequencies of its major be used only rarely to identify specific prey in the
prey items, which fall mainly within the range wild—most studies suggest feeding is opportunistic
50–80 Hz. In the laboratory, these bats sit up and (but not random) in most bats (Fenton 1990). This
take notice, by increasing the length of their CF does not diminish the relevance of all this work—it
pulse, when presented with fluttering targets oscil- is a study of the exceptions, and of the controversial,
lating at 30–100 Hz (Schnitzler and Ostwald 1983). that leads to some of biology’s greatest insights.
86 BATS

10 others
0
10 small Diptera
0

50
Tipulids

0
30
Ichneumonids
percentage of diet by volume

Lepidoptera
50

0
30
Melolontha
0

50
Aphodius

0
50

Geotrupes Figure 3.23 Seasonal variation in the

diet of the greater horseshoe bat,
0
Rhinolophus ferrumequinum (based on
April May June July August September October Jones 1990).

The cost of echolocation and its down, but these events occupy a small proportion of a
consequences bat’s flying time.
Because bats link breathing and echolocation
Vision is a low-cost sense, the light used to perceive pulse emission with the wingbeat, large bats with
objects comes from the sun at no cost to the animal. low wingbeat frequencies have low pulse emission
Echolocation on the other hand can be expensive—the rates. This places an upper size limit on bats, since
emission of high intensity sounds at a frequency of 10 very low pulse rates will be too low for effective
Hz costs a resting pipistrelle 10 times its resting meta- prey capture and even orientation (Jones 1994,
bolic rate at 25  C. By comparison, flight costs a pipis- 1999). Holderied and Helversen (2003) suggest that
trelle around 15–20 times its resting metabolic rate. bats match detection distance to wingbeat period by
However, a flying bat gets echolocation for free having appropriate echolocation calls. A bat can
(Speakman and Racey 1991). In flight, a bat breathes make greatest use of the echo from one of its calls
at the same frequency as it flaps its wings, breathing if it arrives before the next call is emitted, since this
out on the upstroke and in on the downstroke. The avoids both pulse-echo overlap and misassignment
movement of the wings is driven by the flight muscles of an echo to the pulse that generated it. The longer
on the thorax, and these also drive the movements of the wingbeat period, the more distant the object that
the thorax that are responsible for breathing. Since the can be detected in this ‘time window’. It makes
bat is expelling air from the lungs during the sense for larger bats with longer wingbeat periods
upstroke, it costs next to nothing to also emit its to use calls of greater intensity and lower frequency.
echolocation pulse at the end of the upstroke/expira- In a study of 11 European bat species they were able
tion. During a feeding buzz this coupling must break to show that echolocation calls in these bats appear
 ECHOLOCATION AND OTHER SENSES 87

100
132
10% maximum Sls (dB peSPL)

relative echo energy (%)

130 80
30º
128 60 0º

126 40
124
20
122
0
120 50 40 30 20 10 0 –10 –20 –30 –40 –50
70 angle between microphone and loudspeaker (º)
call end frequency (kHz)

60

50

40

30

20

10
180
mean wingbeat period (ms)

Figure 3.25 Sound energy reflected by the vexillum of M. holtonii.

160
Frequency-modulated sweeps were broadcast directly at the vexillum and
140 with the vexillum rotated by 30  in the horizontal plane (based on von
Helversen and Helversen 1999, with permission from Nature Publishing Group).
120

100 spikes that stand erect from the plant itself (Hessel
and Schmidt 1994, Thies et al. 1998). In contrast,
80
Phyllostomus hastatus use only echolocation to find
60 the dangling fruit of Gurania spinulosum (Kalko and
40 Condon 1998). However, bats often need to look for
5 6 7 8 10 15 20 30 40 50 60 fruit and flowers amongst the leaves and branches
body mass (g) of their food plant, conditions that do not favour the
use of echolocation. Nevertheless, echolocation still
Figure 3.24 Changes in echolocation call intensity, call frequency, and
plays an important role. Korine and Kalko (2005)
wingbeat period in relation to body size in European bats (based on
Holderied and Helversen 2003, with permission from the Royal Society).
studied two small fruit-eaters, Artibeus watsoni and
Vampyressa pusilla, that feed on small fruit hidden
to have evolved to make full use of the available amongst leaves. Both species use smell to locate ripe
time window (Fig. 3.24). figs amongst leaves, but echolocate throughout,
showing characteristic changes in echolocation in
the transition from search to approach, but without
Echolocation in neotropical fruit-eaters
the terminal buzz seen in insectivorous species. The
The fruit-eaters of the neotropics, all members of the bats are probably using smell for initial detection
Phyllostomidae, evolved from insect-eating, echolo- and subsequent classification (they are able to select
cating ancestors and many still make use of echolo- ripe, pest-free fruit) and echolocation to approach
cation when finding food. Corollia perspicillata and their target and identify fruit at close range.
C. castanea locate stands of Piper primarily by smell, Nectar feeders also make use of echolocation, and
but use echolocation to find the individual fruit some flowers have evolved to make themselves easier
88 BATS

to locate by echolocating bats. The neotropical vine prey-generated sounds, for example the pallid bat,
Mucuna holtonii bears a structure called a vexillum, a Antrozous pallidus (Bell 1982, Fuzessery et al. 1993),
modified petal that stands erect, facing away from the the long-eared bat, Plecotus auritus (Coles et al. 1989,
flower, when the flower opens (Fig. 3.25). The vexil- Anderson and Racey 1991), and the sibling mouse-
lum in M. holtonii is concave and acts as a powerful eared bats Myotis myotis and M. blythii (Arlettaz et al.
acoustic mirror, reflecting most of a bat’s echolocation 2001). Large carnivores such as the Australian ghost
call back towards the bat over a wide range of angles bat, Macroderma gigas (Guppy and Coles 1988), the
of incidence (Helversen and Helversen 1999). Wild Indian false vampire, Megaderma lyra (Marimuthu
bats showed a very strong preference for flowers and Neuweiler 1987), and Cardioderma cor (Ryan
with an intact vexillum: removal of the vexillum, or and Tuttle 1987) also appear to make full use of
the insertion of a pad of cotton wool into its concave prey-generated sounds. Audiograms for three of
face, destroyed the flower’s reflective properties and these species (Fig. 3.26) show that all have extremely
led to a huge reduction in the proportion of flowers sensitive hearing in the range 10–20 kHz, and they
visited by bats. are the most sensitive mammalian ears yet docu-
mented. This is partly due to their large external
ears, which give greater directionality and increase
Other senses
sound pressure at the eardrum by up to 30 dB.
Ever since echolocation was first described by Donald Obrist et al. (1993) have shown that the acoustic
Griffin in the 1930s it has taken centre stage in properties of the external ear are very well matched
research into bat sensory biology. In a review of to the particular hunting strategy of the species. The
bat sensory ecology, Altringham and Fenton (2003) common vampire, Desmodus rotundus, is sensitive to
noted that during the 1980s and 1990s research pa- breathing noises at 20–35 kHz and may also make
pers on echolocation outnumbered those on other use of prey-generated sounds to find its prey
senses almost tenfold. In recent years, the balance (Schmidt et al. 1991, Obrist et al. 1993).
has begun to be redressed, as it has become better A number of bats appear to rely to a large extent
appreciated that even echolocating bats make use of on vision. The California leaf-nosed bat, Macrotus
other senses when foraging, and that other senses californicus, is another gleaner that hunts using
play an important role in the lives of bats. In this prey-generated sound as well as low-intensity echo-
chapter I will confine myself to the use of other location. Under laboratory conditions, when illumi-
senses in navigation, orientation, and foraging. nation is equivalent to that of a brightly moonlit
Other aspects of sensory biology will be discussed, night, it locates prey by echolocation only one third
where appropriate, in other chapters. of the time, relying on vision at other times, in the
absence of prey movement and sound (Bell 1985).
Vision is also utilized at much lower light levels. The
Hunting with other senses
brown long-eared bat, Plecotus auritus, is also able to
Virtually all bats have good auditory sensitivity in hunt visually (Eklof and Jones 2003), as are some
the frequency range 12–25 kHz, as well as over the Nyctophilus species in Australia (Grant 1991). The
particular range used in echolocation (Neuweiler eyesight of many bats appears to be as good (Al-
1990). They therefore have the ability to listen to tringham and Fenton 2003) as that of Macrotus and
the sounds made by moving prey, and many bats other small, nocturnal mammals: vision may play a
do indeed locate and track their prey by this means. greater role than we imagine in the foraging of many
This may be a legacy from their insectivorous ances- other bat species.
tors, or a later adaptation for prey detection in the The observation that many bat-pollinated flo-
dark. The big brown bat, Eptesicus fuscus, can detect wers are violet and reflect ultra-violet (UV) light
katydids from over 100 m (Buchler and Childs 1981) prompted Winter et al. (2003) to ask if nectar-feeding
and Trachops cirrhosus identifies edible frogs from bats had UV vision to help them locate flowers at
their calls (Ryan and Tuttle 1983). Many insectivo- dusk or in moonlight. The visible spectrum extends
rous bats, particularly gleaners, may rely largely on from a wavelength of 700 nm (red) to 400 nm
 ECHOLOCATION AND OTHER SENSES 89

100

80 Plecotus auritus

Macroderma
60 gigas
threshold (dB SPL)

40 Megaderma
lyra

20

-20
Figure 3.26 Audiograms of three species of gleaning bat.
These bats take advantage of prey-generated sounds when
hunting, and their ears are most sensitive to sounds of
1 2 4 6 8 10 20 40 60 100 150
around 10–20 kHz—those of animals rustling in the
frequency (kHz) undergrowth (see text for references).

(violet). Most mammals, including bats, have little since atmospheric scattering at dusk shifts the light
or no colour perception and no sensitivity to light to shorter wavelengths (Smith 1982).
below 400 nm. However, in a series of behavioural I contemplated rewriting the above section in the
choice experiments the small, nectar-feeding phyl- light of a recently published paper, but I thought it
lostomid, Glossophaga soricina, was found to be sen- would illustrate the pace of scientific developments
sitive to wavelengths down to 310 nm, and showed to simply report the new findings. Muller et al. (2009)
two peaks of sensitivity, at 365 nm in the UV and have recently shown that two phyllostomids, Glosso-
510 nm in the green part of the visual spectrum. UV phaga soricina and Carollia perspicillata, have cone cells.
vision has been reported in a number of rodents and These have two types of cone opsin, short (blue/UV)
marsupials, conferred by a specific UV-sensitive and long (green/red) wavelength. I will leave you to
pigment. Glossophaga, however, is not dichromatic, read the details, but the authors conclude that these
but monochromatic: a single pigment appears to be bats have all that is necessary for daylight vision,
responsible for sensitivity to both visible and UV dichromatic colour vision, and UV vision!
light. Although this is the first monochromatic, UV The choice of which combination of senses to use
sensitive, mammalian eye to be reported, it is per- when foraging is often dependent upon context. For
haps not as surprising a finding as it first appears. example, Ratcliffe et al. (2005) have shown that
The rods in the human eye have a very similar although Megaderma lyra will make use of vision
spectral sensitivity, but the lens absorbs the UV and spatial memory in familiar environments,
light, so we are not able to see it. UV vision may when in unfamiliar surroundings echolocation was
only be possible in small mammals, since they have consistently used to find prey, irrespective of light
eyes with a thin lens and a short path through the levels. This situation-dependent variation in beha-
cornea and aqueous and vitreous humours to the viour may explain some of the conflicting results
retina: sufficiently short to allow UV light to pass. reported in some studies: many experiments are
UV sensitivity may be of more general utility in bats, performed in laboratory conditions that will become
90 BATS

familiar to the bats under study. Ratcliffe and col- is likely to be more important over long journeys.
leagues suggest that in unfamiliar surroundings Very recently, evidence has been found for a mag-
bats may have to make use of echolocation, but netic navigational sense in bats (Holland et al. 2006).
when familiarity, moonlight, or noisy prey allow, Big brown bats, Eptesicus fuscus, carrying radio-
echolocation may be switched off. transmitters were displaced 5 km from their roost
Suprisingly, in the few species studied, bats will and their return paths tracked from a light aircraft.
preferentially use visual cues over echolocation in an Control bats flew straight home, but individuals
escape response, and of course many species rely on subjected to rotated magnetic fields prior to release
visual cues to time emergence. These and related were initially confused. If the magnetic field was
issues are reviewed by Altringham and Fenton (2003). rotated anticlockwise, bats initially flew west of
Smell has been shown to be important in finding the true course and clockwise rotation caused bats
fruit in both Old and New World fruit-eaters, gener- to fly on a more easterly track. The bats were only
ally in conjunction with vision and echolocation subject to the rotated fields for 45 min before and
(see above and Chapter 8). after sunset, suggesting that their magnetic sense
Finally, the common vampire, Desmodus rotundus, may be calibrated against sunset. This has been
(and the two other species of vampire?) has heat- confirmed by recent work on the greater mouse-
sensing organs in pits around its noseleaf that are eared bat, Myotis myotis, by Holland et al. (2010),
insulated from the surrounding tissue and main- who showed that timing was critical—calibration
tained at a temperature 9  C lower (Kürten and only worked at sunset. It has been known for some
Schmidt 1982). These sensors may help the bat time that birds use a map and compass navigation
choose an appropriate bite site, where surface that makes use of a magnetic sense, sometimes with
blood vessels raise the local skin temperature. the aid of vision and olfaction, but there is still
considerable controversy surrounding bird naviga-
tion (Bingman and Cheng 2005, Walcott 2005). Per-
Old World fruit bats
haps the study of bats can shed light on common
The Old World fruit bats (with the exception of Rou- mechanisms.
settus) do not echolocate and rely primarily on good
night vision and a good sense of smell to find food.
Pteropodids alone amongst bats have a tapetum lu- The ecology of echolocation: interactions
cidum: a reflective layer at the back of the retina that
between flight, food, and foraging habits
reflects light back through the photoreceptor cells to
increase the probability that they will be triggered In the last chapter we looked at relationships between
under low light conditions. There is also some evi- wing morphology/flight performance and habitat/
dence that these bats have cone cells in their retina for food/foraging strategy. In this chapter, I want to add
limited colour vision (Suthers 1970). echolocation characteristics to the story. You will
have gained some appreciation of these interactions
already: I want to pull the strands together and com-
Navigational senses
plete the picture. A detailed review of the subject has
Do bats use echolocation when making long flights been written by Norberg and Rayner (1987), and you
between roost and foraging site, or when migrating? can read it in a more condensed form in Norberg’s
As we have seen, many insectivorous bats have book (1990). Other excellent reviews include Fenton
good low-light eyesight and since echolocation (1990) and Neuweiler (1990).
works only over short ranges, sight is likely to be Fenton (1990) has rightly emphasized the flexible
important in navigation. Homing studies carried out nature of foraging styles in bats, and the danger
in the 1950s and 1960s (reviewed by Holland 2007) of too rigidly placing species into particular ecologi-
yielded confusing and often contradictory results. cal niches. However, a number of feeding strategies
However, on balance, they suggested that echoloca- can be clearly defined, and they are usually asso-
tion may have a role to play, but suggest that vision ciated with particular forms of echolocation. Among
 ECHOLOCATION AND OTHER SENSES 91

animal-eating bats, six strategies are readily identifi- Jung et al. (2007). Figure 3.27 shows the calls of 10
able. Let’s go through them in turn. emballonurid species recorded in Central America,
illustrating the use of calls at two frequencies and
the trend towards lower frequency in less cluttered
Open-space hawking above vegetation environments (Jung et al. 2007). Inter-pulse interval
Fast-flying hawkers, with high aspect ratio wings increased with decreasing clutter and call frequency
and high wing loading (see Chapter 2), typically use decreased with increasing size, although across all
low frequency, narrowband FM, or CF search calls bats there were exceptions to this relationship.
up to 60 ms in duration. When there are no harmo-
nics, which is often the case, all of the energy is Open-space hawking between vegetation
concentrated in these 12–30 kHz pulses that can
Bats feeding in this habitat are similar to those
detect prey (aided by flutter detection) over long
described in the previous section, and there is con-
distances. Early detection is needed if these bats,
siderable overlap in flight and echolocation char-
which fly rapidly and have large turning circles, are
acteristics. There is a tendency towards higher
to catch their prey. In such an uncluttered environ-
frequency calls, both the CF search and the FM
ment, detailed spectral analysis using FM pulses is
capture components. This perhaps reflects the
not needed. When prey are detected, some species
slower flight, and the shorter distances over which
switch to short (about 5 ms) broadband FM pulses.
prey are detected. Vespertilionids such as Lasiurus
These will enhance target ranging, and provide
and Pipistrellus species forage in this way. Some
spectral cues that give details of target structure.
species in this group, for example Pipistrelles,
Calls can be of very high intensity. Many emballo-
have very flexible call structure, increasing call
nurids, molossids, and vespertilionids fall into this
bandwidth in clutter for example (Kalko and
category. Some species alternate narrow and not so
Schnitzler 1993).
narrow broadband FM pulses in their search phase
flight, with the second call typically being higher in
Foraging in and around vegetation
frequency. This may be a mechanism for increasing
detection range or the ability to pick up a wider Bats that feed by slow hawking in vegetation have
range of prey sizes. However, a number of other broad wings and low wing loading, giving them
functions have been suggested for this behaviour. If excellent low speed manoeuvrability. They detect
you want a full discussion of the merit of these and take insects at close range. In this cluttered
various hypotheses, see Kingston et al. (2003) and environment, the bat has to deal with complex
so
na
ris
te
yc
on

120
ch

isl
yn

ra
nt
Rh

ra
ce

100
tu

ta
ris

a
lep

ne
frequency (kHz)

ta
te

ica
li
c

yx
ny

bi

tis
ris
er

pl

80
ro

yx

i
ro
pt

er
st

yx
ta
nt

ac
er

pl
co

ro
ec

er
Ce

p
c

vi
op

al

ba
pt

ka
Sa

yx
re
cc

ps

io

al
yx
ab

60
er
Sa

nt
ro

us
er
pt
ur

la
tta

ur
pt
ro
Ba
rm
Cy

id
ro
Pe
Co

icl
Pe

40
D

20
0
decreasing clutter 10 ms

Figure 3.27 Echolocation call structure of 10 emballonurid species from Central America (from Jung et al. 2007, with permission from John Wiley and Sons).
92 BATS

echoes from the vegetation. Two strategies are open readily from their webs. This led to one of my favour-
to bats, both designed to reveal moving insects ite bat film sequences!
against background clutter, and both usually make
use of high frequency calls (>50 kHz). Rhinolo-
phids, hipposiderids, and some emballonurids use Trawling and gaffing
CF calls to detect fluttering insects (for example Bats that take prey by speculative trawling or tar-
Siemers et al. 2005, Zhao et al. 2003, Jung et al. geted gaffing of prey from water (noctilionids, at
2007). Most other bats in this group, including least one phyllostomid, and some vespertilionids,
many vespertilionids, use broadband FM echoloca- Fig. 3.28) can have long wings, since they feed over
tion, often with several harmonics. Spectral analysis clutter-free open water, and (in the case of fish-
of the FM pulses may be used as a movement detec- eaters) low wing loadings to carry heavy prey.
tor, and the FM pulse gives accurate range finding. Their echolocation calls are broadband FM, possibly
for spectral analysis of water texture and to detect
moving prey. All species studied show a strong
Flycatching or perch-hunting preference for hunting over smooth water: the bats
These bats feed by hanging from a perch, intercept- are able to detect insects more easily against a water
ing passing prey. They use echolocation techniques surface that reflects clutter-free echoes, whether the
similar to those bats feeding in vegetation, but have insect is flying above the surface, or floating on it
variable aspect ratio and wing loading since flight (for example Siemers et al. 2001). Some have a CF
performance may be dictated by factors other than component to their call. Since many also hawk for
foraging. Many of the larger Rhinolophidae hunt insects, this may be used for flutter detection (for
from perches. example Siemers et al. 2001, Weinbeer and Kalko
2007).
Another important adaptation, to be expected
Gleaning and hovering from the above discussion, is shown in Fig. 3.29.
Gleaning bats that often hover over prey generally The frequency range to which a species’ ears are
have low aspect ratio, low wing loading, and best tuned is either close to that it uses in echoloca-
rounded wingtips for slow manoeuvrable flight. tion or, in the case of gleaners, in the range of the
These characteristics also facilitate take-off from the sounds generated by their prey.
ground, possibly with heavy prey. They use short
(less than 2 ms) FM echolocation pulses, of low inten-
Evolution of echolocation
sity, designed to discriminate fine texture and possi-
bly detect target movement by spectral analysis over Echolocation, like flight, is of such fundamental
short distances. Many gleaners (insectivores and car- importance to the way of life of bats that it must
nivores) rely heavily on prey-generated sounds, and have played a vital role in their evolution. Flight and
even vision, alongside or as an alternative to echolo- echolocation almost certainly increased in sophisti-
cation. Examples of this category are found in several cation together as the protobats took to the night
families, including the Phyllostomidae, Megaderma- sky, although which came first has been a matter of
tidae, Nycteridae, and the Vespertilionidae (for considerable speculation and investigation. Our
example Arlettaz et al. 2001; Ratcliffe et al. 2005). understanding of the evolution of echolocation is
Natterer’s bat, Myotis nattereri, has calls of unusually making major advances as the technology for study-
high bandwidth, 135 kHz, and is able to take prey in ing echolocation improves (for example Jones and
flight within 5 cm of vegetation (Siemers and Schnit- Holderied 2007) alongside our understanding of the
zler 2000). At the time this work was published I was evolution of bats themselves (for example Teeling et
advising the BBC on bat behaviours suitable for film- al. 2000, 2005, Springer et al. 2001, Eick et al. 2005, see
ing for the Life of Mammals TV series and suggested Chapter 1). The publication of complete genomes of
that Natterer’s bats would probably take spiders bats, beginning with Myotis lucifugus, Pteropus
 ECHOLOCATION AND OTHER SENSES 93

Figure 3.28 Daubenton’s bats, Myotis daubentonii, gaffing prey from the surface of a river.

search capture

Tadarida aegyptiaca

10 Taphozous kachhensis
foraging height above ground (m)

Taphozous melanopogon
8

Rhinopoma hardwickei

6 Plecotus
auritus
Pipistrellus dormeri

4 Hipposideros speoris
Pipistrellus Hipposideros bicolor
mimus
capture
2 search
Megaderma
lyra

0
0 10 17 24 26 35 50 55 137 151
best frequency of audition (kHz)

Figure 3.29 Foraging strategy in relation to echolocation calls and auditory characteristics. Foraging height is plotted against the best frequency of audition.
Bats are loosely divided into gleaners (ground and foliage), above canopy hawkers, low level open-air hawkers, and hawkers in cluttered habitats,
and the characteristic sonagrams of each group shown (based on Neuweiler 1990).
94 BATS

vampyrus, and Rhinolophus ferrumequinum, is being vision. When Rousettus evolved echolocation by
undertaken by the National Human Genome clicking its tongue it avoided the size constraint
Research Institute (www.genome.gov) and this is imposed by laryngeal echolocation. If both his-
enabling researchers to study the evolution of echo- tories are plausible, which one is the more
location at the level of the genes themselves. My probable?
aims here are modest: to give a brief overview of A re-evaluation of morphological data by Teel-
our current understanding of how and when echo- ing et al. (2005), constraining their analysis to what
location evolved and how it reached its current level was possible based on their own molecular phy-
of sophistication and diversity. Several recent and logeny, put Icaronycteris index firmly at the base of
complementary reviews deal with these questions in the tree of all bats. Since I. index is morphologically
considerable detail, each from a different angle with similar to all living echolocating bats, this strongly
a different emphasis: Neuweiler (2003), Schnitzler suggested that echolocation evolved only once in
et al. (2003), Jones and Teeling (2006), Jones and bats, and the Old World fruit bats must therefore
Holderied (2007), and Teeling (2009). However, as have lost this facility. Eick et al. (2005) favoured the
we’ll see it is a rapidly developing field, so not all independent evolution of echolocation in the Rhi-
views expressed will fit the current concensus! nolophoidea, but this was based on the presence of
Perhaps a good place to start is with a question: a unique, calcified rib cartilage, possibly part of a
is the ability to echolocate ancestral to all bats, or mechanism to reduce the cost of echolocation
did it evolve more than once? The question relates when stationary. However, the evidence for a sin-
only to laryngeal echolocation, we’ll leave the odd gle evolutionary appearance of pharyngeal echolo-
case of echolocation in Rousettus, which generates cation was at the time stronger. Furthermore, there
its echolocation calls by tongue-clicking, until later was some tentative evidence for the loss of echolo-
in the discussion. This question arises because of cation in pteropodids, such as the use of multi-
the recent finding (for example Teeling et al. 2000, harmonic, broadband communication calls and a
2005, Springer et al. 2001, Eick et al. 2005) that the large cochlea (Springer et al. 2001). A resolution
horseshoe bats and their relatives in the Rhinolo- seemed possible with the sequencing of two
phoidae share a common ancestor with the non- genes involved in hearing and probably in echolo-
echolocating Old World fruit bats (see Chapter 1). cation in both echolocating and non-echolocating
This raises two possible histories: echolocation bats (Li et al. 2007, 2008): FOXP2 has a role in
evolved in a common ancestor to all bats and the sound generation and interpretation and Prestin is
Old World fruit bats lost the ability, or echoloca- central to determining cochlear sensitivity. The re-
tion evolved independently in the Rhinolophoidae sults were complex (read the papers!), but the
and the Yangochiroptera (Emballonuroidea, Nocti- work on Prestin in particular did suggest that echo-
lionoidea, and Vespertilionoidea). I gave reasons in location evolved more than once in bats, with
Chapter 1 why both of these are plausible options, subsequent convergence. However, the authors
but it is worth recapping. Echolocation has evolved do stress the need for more data—the evolution
independently in cetaceans, two groups of birds of echolocation involves more than a single gene.
(cave-swiflets and the guacharo), bats of the In the latest development, Liu et al. (2010) have
genus Rousettus, and possibly in shrews. It is there- demonstrated convergence in the Prestin gene
fore not implausible that it could have arisen more between bats and dolphins, driven by natural
than once in the rest of the bats. Equally plausible selection, an exciting discovery!
is its loss in the Pteropodidae. Since echolocation
call emission is coupled to the respiratory and
Evolution of echolocation calls
wingbeat cycles to reduce energy costs, a progres-
sive increase in the size of the Old World fruit bats Echolocation calls in living bats show considerable
would make it less and less effective. The progres- diversity. We might expect this to be determined by
sive loss of echolocation may have been compen- phylogeny to some degree, but since call shape is so
sated for by the evolution of exceptional night clearly allied to function, we should not be surprised
 ECHOLOCATION AND OTHER SENSES 95

Pteropodidae multiharmonic calls. Since they are found in the

Rhinolophidae widely separated Rhinopomatidae/Craseonycteri-
Hipposideridae dae, Emballonuridae and Thyropteridae, and some
Rhinopomatidae
Craseonycteridae
Mormoopidae, this call type has probably evolved
Megadermatidae several times. Short broadband calls are used by
Emballonuridae many bats that forage in clutter, to prevent pulse-
Nycteridae echo overlap and to aid target localization. Short
Natalidae
broadband calls with a dominant first (fundamen-
Miniopteridae
Molossidae tal) harmonic are used exclusively by the Vesperti-
Vespertilionidae lionidae, notably Myotis species. Similar calls, but
Myzopodidae with more harmonics are much more widely used
Mystacinidae (Megadermatidae, Nycteridae, Phyllostomidae,
Thyropteridae
Mystacinidae, Vespertilionidae, and Natalidae) and
Furipteridae
Noctilionidae probably evolved several times. Only one species,
Mormoopidae the endemic sucker-footed bat from Madagascar,
Phyllostomidae Myzopoda aurita, (Myzopodidae) uses long duration
broadband calls. The functional value of such a
Narrowband, dominated by fundamental harmonic design is not known. Finally, there are those bats
Narrowband, multiharmonic
that make use of a long, pure, constant frequency
Short, broadband, multiharmonic
Long, broadband, multiharmonic component in their calls. As discussed earlier in the
Constant frequency chapter this is used for target detection, as a flutter-
Polymorphic
detector, for target identification, and in Doppler
Not known
shift compensation. Most bats incorporate terminal
broadband components for improved target locali-
Figure 3.30 Echolocation call type mapped onto the bat molecular
zation. These highly sophisticated calls, and the
phylogeny. Several calls types are widespread across the tree, showing
extensive convergent evolution (Jones and Teeling 2006, with permission auditory and neural adaptations associated with
from Elsevier). them, have evolved independently in both the Yinp-
terochiroptera (Rhinolophidae and Hipposideridae)
to see in echolocation call design excellent illus- and Yangochiroptera (Mormoopidae (Pteronotus
trations of adaptive radiation and convergent parnelli) and Noctilionidae). Despite many similari-
evolution. Jones and Teeling (2006) categorized ties, the rhinolophids and Pteronotus have solved the
search-phase call design on the basis of bandwidth, many problems of echolocation in slightly different
harmonic content, and duration. Figure 3.30 shows ways.
how these superimpose on the molecular phyloge- Can we say what form the ancestral calls took?
netic tree. There are some clues, but few certainties. The most
The very short, broadband clicks used by Rouset- widespread calls are multiharmonic, and other fea-
tus almost certainly evolved after the Pteropodidae tures that may have been common to the calls of
had lost laryngeal echolocation. They are probably early bats are short duration, low intensity, and high
more sophisticated than was once thought (Holland bandwidth (Schnitzler et al. 2004). The ability of
et al. 2004), but I will not discuss them here. As many Vespertilionidae and Molossidae to simplify
discussed earlier in the chapter, species that forage the call to a pure fundamental appears to have
in the open typically have long, narrowband calls, evolved later. Many of these species do emit multi-
dominated by the fundamental harmonic, to detect harmonic calls during prey capture. Finally, the
distant targets. A high proportion of the Molossidae sophisticated CF calls are probably a recent
and Miniopteridae, as well as many Vespertilioni- innovation.
dae, use such calls. From their position on the tree it What did echolocation evolve for? We primarily
is possible that this call structure is ancestral to these study it and discuss it in terms of prey capture,
bats. Some open-air species use narrowband, although frequently in the context of how bats
96 BATS

perform this task in complex environments. Schnit- alternative ways of categorizing functional groups
zler et al. (2003) argued that it evolved first for spa- of echolocating bats.
tial orientation and navigation and only later for Echolocation research is a complex and rapidly
prey capture, replacing the use of passive listening growing field and one I find particularly fascinating.
in many species. Others have suggested that vision It is difficult to know where to stop, but I will stop
continued to be used for orientation as echolocation here and move on to other things. Suffice to say,
evolved for prey-capture. Schnitzler et al. (2003) dis- many new and interesting studies will be published
cuss this in some detail and also present some by the time this book is released!
 CHAPTER 4

Torpor and hibernation

Energy balance. Ectothermy, endothermy, and heterothermy. Torpor and hibernation. The cost of staying
warm—the importance of body size. The physiology of torpor. The ecology of torpor and hibernation,
comparative aspects. Hibernacula. Biological clocks. Evolution of hibernation and torpor.

Energy balance territorial defence, mate attraction, mating, and a

range of other social needs, all demand energy at
All animals have a daily cycle of activity, carrying different times. One factor is always a major variable
out the tasks essential to their individual survival in the energy balance equation: thermal homeosta-
and those required to pass on their genes to the next sis, the maintenance of a stable body temperature
generation. All of these tasks consume energy above that of the environment. It is particularly
derived from metabolizing the food they eat. Life important in temperate, insectivorous bats since
is a continuous attempt to balance the input:output they face the challenge of an ephemeral food supply
equation: energy in—food, must equal energy out— in an often harsh climate, but it is relevant to many
the physiological and behavioural costs of life. To be others too. In this chapter we will look at how bats
successful, that is to survive and pass on its genes, manipulate their body temperature as part of their
an animal must operate efficiently, and evolution energy budgeting process.
drives the animal towards optimal solutions for all
its tasks. If one animal feeds more efficiently than
another, it can spend more time looking for a mate Concepts: ectothermy, endothermy,
or a safe roost, and might therefore have a competi-
and heterothermy
tive edge in the fight for survival. Optimal solutions
may never be achieved in a changing world, but this First of all, a definition or two. Bats, like all mammals,
drive towards optimality is strong, and the solutions are endotherms. That is they maintain high body
we see today are often marvellously close to the temperatures by metabolizing their food to generate
optimal. So, life is governed by the energy balance heat internally. To do this they must have high meta-
equation—what does it mean to bats? Bats are most bolic rates and high food intakes. Many reptiles can
active between dawn and dusk, and spend much of maintain high body temperatures during the day, but
the day asleep, or indulging in various social activ- they achieve this by basking in the sun and they cool
ities. On average, a bat must catch enough food each rapidly at night. Reptiles are ectotherms: they have
night to provide the energy needed to get it through low metabolic rates and cannot generate sufficient
to the next night. Energy stored primarily in the metabolic heat to compensate for heat loss across
form of fats will smooth out daily fluctuations in the body surface. Endothermy, as practised by mam-
supply and demand. The greatest energy demand mals and birds, has big advantages—the animal can
on a bat is collecting this food in the first place. The go about the business of survival independently of
importance of an optimal foraging strategy is there- the vagaries of the climate. If the sun goes in, a lizard
fore obvious. Although each species is adapted to a cools and all activity slows or even stops, since the
particular foraging strategy, the most critical vari- chemical reactions that keep its body on the move
able in the optimal foraging equation is often the slow as temperature falls. However, endothermy also
energetic cost of flight. Grooming, care of young, has potential drawbacks. The high metabolic rates

97
98 BATS

needed for effective endothermy demand high food 20

O2 consumption (cm3/g/h)
intake. This can create problems in two ways.
thermal
1) The food supply may be seasonal—if so, the animal 15 neutral
zone
needs to hibernate, switch to an alternative and
more plentiful food supply, store food in times 10
homeothermic
of plenty, or migrate to a place where there is food. response

2) If the animal is very small, it needs to eat a lot of 5

food relative to its body size. This is due to a
simple physical law: as the size of an animal heterothermic response

decreases it loses proportionally more of its 4 8 12 16 20 24 28 32 36 40

body heat to the surrounding air, because it has ambient temperature (ºC)
a proportionately greater surface area to lose it
from, relative to the volume generating it. This is Figure 4.1 Energy expenditure during homeothermy and heterothermy
why small mammals like shrews have such a in the western pipistrelle, P. hesperus (from Hill and Smith 1984).
prodigious appetite: they have to keep eating to
provide the energy that keeps them warm. In
active bat is minimal) is around 30–36  C—a typical
other words, their mass-specific metabolic rate
warm summer evening over much of this bat’s
(energy produced per kg of tissue) is high (dis-
range. As temperature increases beyond the thermal
cussed below), although their absolute rate is
neutral zone, energy must be expended to keep
low. At the other end of the scale, many dino-
cool—through sweating and wing-fanning for
saurs were probably inertial homeotherms: their
example. Below the thermal neutral zone, more
large size gave them very low surface area:vol-
and more metabolic energy must be diverted into
ume ratios, minimizing heat loss, and enabling
heat production. In heterotherms such as temperate
them to maintain elevated body temperatures
bats, when the bat is at rest in the roost, the body
with low metabolic rates.
temperature thermostat can be set to a new low
Many animals cope with both small size and a sea- temperature when the ambient temperature falls
sonal food supply. Some brave it out, and remain below a critical level, making enormous energy sav-
homeothermic—that is they maintain a high body ings. This critical temperature may vary, depending
temperature all the time. They can migrate (many upon species, physiological state, the status of
birds), lay in food stores (many small rodents), or stored energy reserves, and the behavioural/ecolog-
change diet (small birds and rodents). Others hiber- ical demands on an individual bat—more on this
nate, or become heterothermic: they actively regu- later. When the thermostat is reset, the bat enters
late their body temperature so as to minimize their torpor, but even in this state energy must be ex-
energy requirements. The body temperature main- pended to prevent death if the temperature falls
tained is related to ambient temperature, the energy below a critical level. The bat may increase its meta-
reserves of the animal and the energetic and beha- bolic rate to prevent excessive cooling, or come out
vioural demands on it. of torpor to move to a warmer hibernation site.
Figure 4.1 shows how the western pipistrelle Heterothermy is a specialized form of home-
(Pipistrellus hesperus), responds to temperature othermy, not a primitive one. It is found mainly in
change (Hill and Smith 1984). It is a small (3–5 g) small mammals in temperate regions and in hum-
species from the south-western deserts of the United mingbirds, which evolved in the high, and often
States, where the annual temperature range is 10 cool, forests of the Americas. The changes in body
to 45  C. The upper and lower thermal lethal limits temperature are under the control of the animal, and
for this species are 1 and 44  C: in other words they it can arouse from torpor using its own energy,
die if the ambient temperature remains outside these without an increase in air temperature. Hetero-
limits for long. The thermal neutral zone (in which thermy is therefore a facultative, not an obligate,
energy expenditure on thermoregulation by an process in bats (for example Audet and Fenton
 TORPOR AND HIBERNATION 99

1988). This is perhaps a good place to point out that elephant

most bats are homeothermic most of the time: only 1000

resting metabolic rate (w)

temperate members of the Vespertilionidae and Rhi-
nolophidae (and the recently separated Miniopteri- 100 human
dae) make extensive use of deep facultative
heterothermy. However, as we will see, many 10
other bats use torpor to a limited extent.
1 rat
Concepts: torpor and hibernation
mouse
What exactly do we mean by hibernation and tor- 0.1
por? Definitions of hibernation are often vague, but
we can be more specific about the term torpor. In
torpor an animal allows its body temperature to fall 0.01 0.1 1 10 100 1000 10000
below its active, homeothermic level. The fall in body mass (kg)
body temperature is controlled: it does not fluctuate
Figure 4.2 The relationship between resting metabolic rate
freely with air temperature, and is maintained within
and body mass in birds and mammals (based on Schmidt-Nielsen 1984).
narrow limits at this lower level. If the air gets too
cold, the torpid bat will burn off some of its stored
fuel to provide the energy needed to keep itself from metabolic rate (BMR, which is very close to RMR,
becoming too cold. If that is torpor, what is hiberna- the metabolic rate in a resting animal) for mammals
tion? Simply put, hibernation is an extended form is shown in Fig. 4.2, as a log-log plot. The use of such
of torpor, lasting for days, weeks, or months, occur- graphs is common practise among biologists inter-
ring on a seasonal basis, in response to a prolonged ested in the effects of size, since a remarkable num-
fall in ambient temperature or reduction in food ber of biological variables can be related to body
supply. Bats do not usually spend their time in mass (M) by the simple equation:
continuous hibernation, and in a maritime climate
y ¼ a:Mb
bats may be more or less active right through a mild
winter. Ransome (1990) has suggested that the dif- By plotting the log of the variable y against log M,
ference between ‘summer’ and hibernation torpor the exponent b, can be calculated from the slope of
hinges on the arousal process that starts at dusk in the line (a is the intercept of the line with the y-axis).
response to the peak in the bats’ daily temperature If y increases in direct proportion to M, then the
cycle. In summer, this always leads to arousal, but in slope b = 1. If y increases more slowly than M,
winter arousal is suppressed for many days at a b < 1, and b > 1 if y increases faster than M. If y
time. Others like to draw a clear line between torpor decreases as M increases, then b is negative. Body
and hibernation. Rather than wrangle over defini- size has an enormous influence on all aspects of
tions, it is perhaps better to pose a question: how biology, and for a fascinating and readable introduc-
does torpor fit into the life history strategy of a bat? tion to the subject see Schmidt-Nielsen’s book
Before answering this question, we need to look at (1984). In the case in question, resting metabolic
the cost of endothermy. rate increases with increasing body mass, but it in-
creases more slowly than if it was directly propor-
The cost of staying warm—the tional to M: b = 0.75: larger animals have relatively
low metabolic rates. Kleiber (1932) was the first to
importance of body size describe this important relation and more recently it
Let’s start by looking at the costs of homeothermy was determined for bats alone by Speakman and
and heterothermy. The most important single factor Thomas (2003). Basing their calculation on all the
influencing the cost of homeothermy is body size. data available, 84 bat species from 11 of the 19
The relationship between body mass and basal families, they found the same exponent.
100 BATS

This is seen best if we plot the metabolic rate per gets smaller its energy stores decrease faster (Æ M1)
kg of tissue (mass-specific rate) against body mass than the rate at which it uses energy (Æ M0.75), so it
on a similar log-log plot, as shown in Fig. 4.3: b = has to feed more often (McNab 1982). Starvation is
0.25: small mammals have disproportionately never far away for a small mammal, particularly in a
high mass-specific metabolic rates. This is not cold environment. The problem is too extreme to be
surprising: look at the even steeper relationship solved by extra insulation: the little brown bat
between minimal thermal conductance, the rate at would need to be at the centre of a furry football. It
which energy is lost from the body, and body mass therefore makes sense that if a small mammal has no
in Fig. 4.4a. (McNab 1969, 1982). This is a very reason to remain homeothermic all the time, then a
predictable consequence of the increased surface heterothermic option may evolve. The two extreme
area:volume ratio with decreasing size. courses of action shown in Fig. 4.1 are not the only
Figure 4.4b shows that, with the exception of the options: the thermostat could be set to any tempera-
biggest pteropodids, the bats studied tended to have ture between ambient and full homeothermy and
higher thermal conductance than other mammals. the temperature chosen depends upon the ambient
McNab (1969) carried out his experiments on tropi- temperature, the energy reserves of the bat, and its
cal bats, and the higher conductance may be due to physiological and ecological requirements at the
the fact that tropical mammals have poor insulation time. As a general rule, however, the set tempera-
relative to temperate mammals of the same size ture is usually ambient or a little above it, at tem-
(Scholander et al. 1950). It has been suggested that peratures above 1  C.
higher conductance may be due to heat loss from the A 9 g fringed myotis (Myotis thysanodes) has a
wings, but bats can keep their wings folded when at thermal neutral zone of 30–38  C. Flight and night
rest and control blood flow through them. They may roost maintenance use up 1.55 kJ g1 day1. Day
even help to retain heat when they are wrapped roost maintenance at 20  C ambient temperature for
around the body, by trapping a layer of still air a regulating bat costs an additional 0.67 kJ g1
(Bartholomew et al. 1964). Thermal images of flying day1. However, if the bat becomes heterothermic,
bats in fact show that most of the heat is lost from only an additional 0.34 kJ g1 day1 are required
the body. But, to return to the main story, if a bat (O’Farrell and Studier 1970). This is a 15 per cent
needs to remain homeothermic, the smaller the bat, daily energy saving. On a minute by minute basis
the higher the mass-specific metabolic rate, and the the regulating 3–5 g western pipistrelle discussed
more food it needs per gram of body mass to main- earlier uses 4–5 times as much oxygen at an ambient
tain its temperature. The 8 g little brown bat, Myotis temperature of 20  C, than when it is in its thermal
lucifugus, needs food to generate 3.7 kJ g1 day1, neutral zone, and 8–9 times that of a non-regulating
a 180 g false vampire, Vampyrum spectrum, only bat at 20  C. Any opportunity to down-regulate body
1.1 kJ g1 day1. As a bat, or any other homeotherm, temperature can mean big energy savings. The very
smallest species of bat may be obligatory hetero-
therms some of the time—that is they may have no
option. However, it is clear that the rest are faculta-
log specific metabolic rate

tive heterotherms—they can choose when to be het-

erothermic, and how much to down-regulate.
slope = –0.25
As with many things in biology, the situation is
not quite that simple. Above (Fig. 4.2), I described
the relationship between body mass and resting
metabolic rate (RMR) for homeothermic mammals,
saying that RMR increased in proportion to M0.75.
Two questions have arisen about this relation. Is the
log body mass
slope really 0.75, and is this the slope for all home-
Figure 4.3 The relationship between mass-specific resting metabolic otherms? It is important to clear up these questions,
rate and body mass in birds and mammals (from Schmidt-Nielsen 1984). since they are relevant to bats. First, if RMR was
 TORPOR AND HIBERNATION 101

0.5
min.thermal conductance
(cm3 oxygen/g/h/ºC)

b = –0.51

0.1

0.05

0.01 0.1 1
body mass (kg)
thermal conductance (cm3 O2/g/h/ºC)

0.5
diet of bats
0.4 fruit
nectar
meat
insects
0.3 blood

0.2

temperate mammals
0.1
other than bats
Figure 4.4 The relationship between minimal thermal
conductance and body mass in bats. (a) Log-log and (b)
25 50 75 100 150 360 linear plots (from McNab 1969, 1982, with permission
body mass (g) from Elsevier).

simply proportional to body surface area, then the advantage. It turns out that bats also deviate from
slope should be 0.67. Despite much debate, dis- the curve, and this deserves a closer look (Fig. 4.5).
cussed by Schmidt-Nielsen (1984), the slope is 0.75, McNab (1969, 1982) found that, as a general rule,
not 0.67. One of the arguments against the 0.75 value insectivorous bats have BMRs below the predicted
hinges on the deviation from the line of some mam- line, but all other bats (carnivorous and fruit/nectar
malian groups. Variation is widespread: across all eating bats) fall on, or even a little above, the line.
mammals 23 per cemt of RMR is not explained by Species with mixed diets, such as Phyllostomus has-
body mass. Amongst individual orders, 10–45 per tatus and Noctilio leporinus, have intermediate BMRs
cent is unexplained. Thus, depending upon which (McNab 1982). These differences appear to be truly
mammals are included in the overall calculation, the related to diet rather than to phylogenetic origin
slope of the relationship varies. Shrews have higher (that is taxonomic relations are unimportant). The
BMRs than those predicted from the allometric low BMR of insectivores may reflect the low energy
equation, as do seals and whales, but desert rodents content and relatively unreliable nature of their food
have lower rates. If all homeotherms are included, supply. The debate about the relative importance of
however, the slope is 0.75. These deviations from the ecology and phylogeny in biological scaling is long
expected curve often convey an important message. standing and wide ranging. The controversy has
Marine mammals have the problem of maintaining included studies of bats (see McNab 1992, 2003)
high body temperatures in cold water, which con- and several studies have questioned McNab’s find-
ducts heat away from the body rapidly—hence the ings. Speakman and Thomas (2003) are the latest to
high metabolic rate. Desert mammals, from rats to review this and much else in bat energetics. They
camels, often have to contend with a poor food point to problems on both sides of the debate,
supply: a low RMR is therefore an adaptive related to estimates of metabolic rate and to a poor
102 BATS
mass specific resting metabolic rate (cm3 O2/g/h)

3.0
diet
fruit
nectar
meat
2.0 insects
blood

best fit line

for terrestrial
mammals
1.0

Phyllostomus
hastatus

0.5 Noctilio leporinus

Figure 4.5 The relationship between mass-
5 10 50 100 500
specific resting metabolic rate and body mass in
body mass (g) bats with different diets (from McNab 1982).

knowledge of bat phylogeny. With some caveats, and nectarivore groups are tropical Pteropodidae or
they initially came out in support of McNab: insec- Phyllostomidae. When this is taken into account,
tivorous bats have lower BMRs than frugivores, as a Speakman and Thomas conclude that there is little
function of diet, not phylogeny. But is it that simple? evidence linking BMR to diet or phylogeny.
Speakman and Thomas (2003) also found that
although the relationship between body mass and
The physiology of torpor
BMR had the same exponent as other mammals
(0.75), the average BMR of bats is lower than that Can we define torpor in physiological terms? We
of rodents and higher than that of marsupials. Does can certainly list some features characteristic of
phylogeny underlie the differences, as some would torpor:
suggest? Speakman and Thomas took care to com-
1. A controlled reduction of body temperature to a
pare bats with taxonomically homogeneous groups
varying extent, but in deep torpor to as little as
in the same size range. However, most of the bats in
1–2  C above ambient.
the sample are tropical species, and the rodents are
2. A corresponding fall in O2 consumption, breath-
largely temperate. Animals that need to generate a
ing rate, heart rate, and metabolic rate.
lot of heat to cope with low ambient temperatures,
3. Peripheral vasoconstriction, and in extreme cases
in this case the rodents, may require a higher BMR.
restricted blood flow to only a few vital organs
BMR is in fact correlated with latitude in birds.
for long periods.
Speakman and Thomas did not find such a relation-
4. An ability to arouse spontaneously and indepen-
ship in bats and suggest it may be because temper-
dently of ambient temperature.
ate bats spend the day in sheltered roosts, buffered
from their environment, and do not therefore need a Let’s follow a bat into hibernation and look at the
high BMR. When they are out of the roost they are in physiological changes that occur. Our bat will pre-
almost perpetual flight—and generating huge quan- pare for hibernation by depositing large reserves of
tities of heat in their flight muscles. The generally fat. To do this many bats appear to rely more on
low BMR in bats may be a reflection of this. But, if regular post-feeding torpor than an increase in feed-
roost and ambient temperature are so important, ing activity. Some are known to forage close to the
don’t they complicate the phylogeny versus diet sites used at least in the early stages of hibernation,
debate? Indeed they do, since the major frugivore and are thus able to enter torpor immediately after
 TORPOR AND HIBERNATION 103

feeding (Twente 1955, Krzanowski 1961). The 9 g

fringed myotis (Myotis thysanodes) can store almost
0.2 g fat per day by this method, adding over 2 g to
its weight in 11 days (Ewing et al. 1970). A bat will
typically enter hibernation with fat reserves of 20–30
per cent body weight (Ewing et al. 1970). Day length
is probably the cue for laying down fat prior to
entering hibernation. The most important stimulus
for the onset of hibernation itself appears to be
ambient temperature, in contrast to rodents, where
food shortage and photoperiod are the main factors.
Hormonal and other physiological factors are
undoubtedly important, but I leave you to explore
them. Once hibernation does begin the changes are
profound.
Heart rates measured during hibernation range Figure 4.6 The location of brown adipose tissue in a bat.
from 10–16 beats min1 in the red bat, Lasiurus
borealis (Reite and Davis 1966), to 42–62 beats
min1 in the big brown bat, Eptesicus fuscus (Kallen increased blood flow is sent in part to the brown
1977). Contrast this with 250–450 beats min1 at rest, adipose tissue (Fig. 4.6, abbreviated to BAT!)—a
and 800 beats min1 during flight (Studier and deposit of fat on the back, which has cells speci-
Howell 1969, Studier and O’Farrell 1976). The alized for heat production. The fat cells in the BAT
blood supply to the limbs is shut down and the of the bat have an unusually large number of mito-
excess red blood cells are stored in the spleen. Only chondria, the organelles responsible for turning fat
the vital organs like the brain and heart retain a into ATP, the high-energy molecule that powers
normal, regular blood supply. The bat breathes most cellular processes. In this case, the energy is
slowly and erratically. It may go for 60–90 min released directly as heat. BAT is found in many
without a breath. The oxygen consumption rate of small mammals and in the young of larger mam-
a hibernating little brown bat (Myotis lucifugus) at mals, including humans.
2  C is 140 times slower than that of a fully homeo- The blood is warmed as it passes through the BAT
thermic individual (Thomas et al. 1990). This uses and thus warms the rest of the body. The bat may
just 4 mg of fat per day, from the 2.5 g available to a begin to shiver, once the muscles are warm enough
well-fed bat. In theory, this would keep it going for to do so, and the heat generated by this muscular
almost two years! However, a small rise in tempera- activity accelerates the warming up process. Muscle
ture, or a brief arousal, increases this rate enor- is only 20–40 per cent efficient at producing mechan-
mously. Thomas (1995) calculated that an arousal ical power—the rest of the energy is lost as heat. In
of several hours duration uses about 100 mg of fat, 10–30 minutes the bat can be fully active (Fig. 4.7).
equivalent to 65 days in deep torpor. If a bat aroused Small mammals have high mass-specific rates of
every 14 days, it would account for 85 per cent of the heat loss, and during arousal we would expect them
fat metabolized over the winter. How long can a bat to lose heat more rapidly to the environment than
remain in torpor? Little brown bats have remained larger heterotherms, and therefore warm up more
torpid for up to 140 days under laboratory condi- slowly. However, Stone and Purvis (1992) found
tions (Menaker 1964). In the wild, hibernation is that mean warm up rate increased with decreasing
almost invariably interrupted by frequent arousal. body size in heterothermic mammals, a result that
We’ll look at why they undertake these expensive runs counter to predictions based on scaling effects.
arousals later in the chapter. It seems that small mammals have evolved mechan-
How do bats arouse themselves? It starts with an isms to speed up arousal, possibly by having more
increase in heart rate and breathing rate. The brown adipose tissue. Furthermore, the warm up
104 BATS

ambient = 17ºC Its fuel tanks, in the form of fat deposits, are full
after foraging on warm, insect-rich evenings. The
600
capricious temperate spring brings in cold, wet
weather, and the bat can’t forage, so it stays in the
500 roost. It’s a pregnant female, with a developing
heart rate (beats/min)

heart foetus, so it maintains a high body temperature

400 rate 36 throughout most of the day and night to speed the

body temperature (ºC)

growth of its baby, using up its stored energy. The
arousal 32
300 bad weather doesn’t last, so it is soon out foraging
stimulus
28 again. A couple of weeks later, there is a longer cold
body snap. The bat’s reserves of fat are reduced to a
200 temp.
24 critical level, and it is forced to become heterother-
mic to save energy. This slows foetal development,
100 20 but there is no choice. In a bad year, this may lead to
abortion. However, this year the bat is soon feeding
0 10 20 30 again, and spends most of its time in a homeother-
time (min) mic state through gestation, lactation, and into
weaning. The evidence for the use of torpor by
Figure 4.7 Heart rate and body temperature changes during arousal pregnant and lactating females is contradictory.
from torpor in the mouse-eared bat, Myotis myotis (from Kulzer 1967).
Data gathered on Myotis lucifugus by Kurta and
Kunz (1987) and Kurta and Kunz (1988) under con-
ditions closely simulating those in a natural roost
rate increased as the temperature range over which suggest that females will maintain high body tem-
the animal arouses increased, and surprisingly, peratures at all costs. This contrasts with earlier
warm up rate was inversely related to resting meta- work (for example Studier and O’Farrell 1976), that
bolic rate (although positively related to peak meta- showed that a high proportion of females under
bolic rate). The mechanisms underlying these laboratory conditions reduced body temperature
interesting relationships are not yet understood, when faced with declining ambient temperature.
but they all make sound functional sense. Arousal The important message here seems to be that what
is an expensive process and mechanisms that mini- bats may do under laboratory conditions is not nec-
mize the time of arousal reduce the energy wasted. essarily what they choose to do in the wild, when
The low RMR reduces the cost of maintaining a high conditions may be subtly different.
body temperature and we might expect this trait to During this time a nearby male has a different
be selected for during the evolution of small hetero- strategy. He hasn’t got the burden of a developing
therms. The data are too few to interpret them with foetus, he is probably not defending a territory or
confidence in terms of natural torpor patterns, but attracting mates at this time of year, so he doesn’t
they are certainly thought-provoking. need to maintain a high body temperature. He can
afford to forage only on those nights best suited for
foraging, and spends a lot of time in torpor to save
The ecology of torpor and hibernation
energy. An opposing view, based on laboratory ex-
Torpor, used on a daily basis for energy budgeting periments, is that he does need energy to produce
or for long periods of hibernation, is an important sperm, and he may not have it quite so easy as
and integral component of the life history strategy of previously supposed. Hamilton and Barclay (1994)
bats in both temperate and tropical regions. To give and Grinevitch et al. (1995) monitored big brown
a general picture of how torpor can be used, let’s bats, Eptesicus fuscus, in the wild with temperature-
briefly look at a year in the life of a temperate insec- sensitive radio transmitters. Both sexes made use of
tivore, before broadening the discussion to cover all torpor throughout the summer, but males did
bats. We’ll start in spring, with a well-fed bat. indeed use it more frequently, and their torpor was
 TORPOR AND HIBERNATION 105

deeper. Again, what bats can do and what they fat reserves substantially during the winter months
choose to do in the wild may be different. (Ransome 1990). These bats can clearly top up their
Let’s jump now to late summer and autumn. Food food stores by winter foraging. Park et al. (1999,
is plentiful, and the bats are making the most of it. 2000) have shown that arousal is synchronized
This is the mating season for bats, so they will around dusk in most greater horseshoe bats
remain homeothermic for at least some of the time (Fig. 4.8) and this synchrony was most evident in
they spend in the roost. However, despite substan- individuals with low body mass. Bats remained
tial reserves of fat, the females will spend a lot of the active for 2.5 to 9 hours on average with the longer
daylight hours in torpor, saving energy. This is the periods being recorded at ambient temperatures
turn of the males to exert themselves, defending above 10  C.
territories, attracting mates, and mating. This beha- In captive pipistrelles, Pipistrellus pipistrellus,
viour will continue well into autumn. mass loss during torpor is independent of original
It is difficult to define when daily torpor turns into mass (Speakman and Racey 1989), and although
hibernation, and it is probably uninformative and winter feeding occurs, they do not appear to regu-
misleading to try. It all depends on the temperature, late their body mass in the same way as horseshoe
the food supply, and other demands on the bat’s bats—a light bat in a group is always a light bat
time. However, the bats will go into torpor more (Avery 1985). Speakman and Racey (1989) suggest
often, and for longer periods as the winter ap- that they may not need to regulate mass: by regulat-
proaches. Young bats are still growing, and may ing fat reserves over the winter they ensure that
not be able to afford the luxury of torpor quite so even late in hibernation they are well above levels
often, so they may enter hibernation later than that could lead to mortality. Arousal to drink may
adults and at a lower weight (for example Rhinolo- be at least as important as arousal to feed since there
phus ferrumequinum, Ransome 1968, 1990; three Myo- is considerable evaporative water loss during hiber-
tis species, Ewing et al. 1970). Males often remain nation (Thomas and Cloutier 1992). Laboratory
active for longer than females, and may also enter studies on Pipistrellus pipistrellus (Speakman and
hibernation at a lower weight (Ransome 1971). Even Racey 1989) and Plecotus auritus (Hays et al. 1992)
in the depths of the harshest winter, bats do not also suggest that the primary function of arousal is
remain in continuous torpor. They will arouse at
intervals of as little as a day and rarely remain torpid
18 Jan
for more than two weeks. Given the enormous cost
sunrise

sunset

of arousal, why do they do it? Many suggestions 01 Feb

have been put forward, most have been tested: there 15 Feb
is evidence to suggest that bats wake up to eat,
drink, move to another hibernation site, or even 01 Mar
mate. However, not all bats appear to arouse for 15 Mar
the same reasons and in some instances arousal 29 Mar
remains unexplained. Males are more active than
females in some species (for example Avery 1985), 12 Apr
possibly because they need to feed to compensate 26 Apr
for a lower body weight at the onset of hibernation
10 May
(Ransome 1990). In periods of warm weather, com-
mon in temperate winters in maritime regions, 24 May
many bats come out and feed. In mild years bats 00:00 04:00 08:00 12:00 16:00 20:00 24:00
may forage right through the winter (Ransome 1971, time (24 hour clock)
Avery 1985, Brigham 1987). In greater horseshoe
Figure 4.8 Arousal from torpor (horizontal bars) during hibernation is
bats, Rhinolophus ferrumequinum, there is no correla- synchronized with dusk in Rhinolophus ferrumequinum, based on
tion between mass on entering torpor in winter and recordings of wild bats made with temperature sensitive radio transmitters
that at the end of hibernation, and they can increase (from Park et al. 2000, with permission from John Wiley and Sons).
106 BATS

to drink. Thomas and Cloutier (1992) showed that the western mastiff bat, Eumops perotis, in the United
more than 99 per cent of water loss was by evapora- States and the greater mouse-tailed bat, Rhinopoma
tion across the skin, and suggested that the cluster- microphyllum, in India. Daily torpor is used to a
ing of bats frequently seen in hibernacula may be an limited extent by many tropical, insectivorous bats
adaptation to reduce water loss. whose food supply is unreliable and often low in
energy (McNab 1969, Studier and Wilson 1970,
McNab 1982).
Comparative aspects The fruit-eating pteropodids are almost exclu-
sively homeothermic, due in part to their large
Daily, controlled torpor is used by most (perhaps size, but also to the relative abundance and reliabil-
all) species in the large, temperate branches of the ity of their food supply—they can afford the expense
Rhinolophidae and Vespertilionidae, including the of continuous homeothermy. Large Pteropus and
recently separated Miniopteridae. Most (all?) tem- Rousettus species wrap their wings around their
perate, and a few subtropical, species from these body and tuck in their heads when temperature is
families are hibernators. In these bats, torpor is an very low, creating large, still air spaces to reduce
integral part of their energy budgeting strategy. In heat-loss. Conversely, wing flapping, panting, and
other bats, the situation is far less clear, but as a other behavioural methods are used to keep cool
general rule torpor appears only to be used in emer- when temperatures are high. Some small pteropo-
gencies, in the face of major energy shortfalls or dids have been shown to make use of daily torpor
extreme weather conditions. Torpor is more charac- when ambient temperature is very low. The tube-
teristic of insectivorous bats, and this is related not nosed bats Nyctimene albiventer (30 g) and Paranycti-
only to their small size, but to the ephemeral nature mene raptor (20 g) from New Guinea can slip into
of their food supply. Having made these bold state- torpor and reduce oxygen consumption fourfold at
ments, let’s look at some of the detail. 25  C (Bartholomew et al. 1970). The African long-
Some members of the Molossidae, Hipposideri- tongued fruit bat, Megaloglossus woermanni, also en-
dae, Rhinopomatidae, and Mystacinidae have also ters daily torpor, reducing body temperature to as
been reported to use torpor, but less frequently than little as 26  C at an ambient temperature of 23  C
in vesper and horseshoe bats and body temperature (Kulzer and Storf 1980).
reductions are rarely as extreme. In species that The diverse Phyllostomidae of South and Central
bridge the temperate–subtropical divide, the situa- America includes both frugivores and insectivores.
tion can be more complex than in those confined However, the over-riding factor determining the
to temperate regions. The Brazilian free-tailed bat, thermoregulatory strategy is size. The insectivores
Tadarida brasiliensis, ranges from northern California may resort to torpor as a last resort. When the fru-
south to Mexico. Populations in northern California givores in this family were first studied some
do not migrate, are perhaps capable of using torpor, conflicting results were obtained, one investigation
and have been found ‘hibernating’ in cold sites typi- suggesting homeothermy in the fruit-eaters (McNab
cal of those used by temperate bats (Pagels 1975). 1969), the other heterothermy (Studier and Wilson
They do not, however, allow their temperature to 1970). The problem was resolved when it was rea-
drop close to ambient, but maintain a differential of lized just how important the nutritional state of the
24–31  C between body and ambient temperature bats was in determining their response to low ambi-
(Pagels 1975). T. brasiliensis in Arizona also do not ent temperatures. A well-fed bat can afford the lux-
migrate, but they appear not to use torpor, spending ury of homeothermy, a hungry bat may not (Studier
the winter roosting in warm chimneys, caves, and and Wilson 1979). The importance of nutritional
tunnels. Individuals that spend the summer in other state and body mass was demonstrated in con-
south-western states migrate to Mexico. trolled experiments on Carollia perspicillata and Stur-
Some species on the temperate–subtropical divide nira lilium by Audet and Thomas (1997). It seems
use daily torpor only in the winter months, and likely that in these bats too torpor is a last resort, not
seem unable to utilize it at other times, for example a frequently used strategy.
 TORPOR AND HIBERNATION 107

The relatively large size and high energy diet of

carnivorous bats should incline them towards
homeothermy. Data are sparse and frequently anec-
dotal, so should be viewed with caution. The 150 g
Australian ghost bat, Macroderma gigas, appears to
be homeothermic over a wide temperature range,
and so too does the phyllostomid Chrotopterus aur-
itus (100 g) (McManus 1976). The larger fisherman
bat, Noctilio leporinus (60 g) does appear to have the
capacity for heterothermy, due to its size and per-
haps its mixed diet—it probably relies on insects to a
large extent in some circumstances (McManus 1976).
Nycteris grandis (30 g), which takes large numbers of
vertebrates, appears not to use torpor. The common
vampire, Desmodus rotundus, (35 g) also appears to
be homeothermic down to very low temperatures in
the wild, although capable of using torpor. The two
other vampire species, Diphylla ecaudata and Diae-
mus youngi may resort to torpor more often. Perhaps
the avian blood they feed on is harder to obtain than
the blood of the largely domestic prey of Desmodus.

Hibernacula
Figure 4.9 A hibernating whiskered bat, Myotis mystacinus, covered
This section concerns only temperate, insectivorous in condensed water droplets.
bats. Different species have different microclimatic
requirements for successful hibernation, and the
success or failure of a bat to survive the winter is humidity, since it reduces the amount of water they
determined to a large extent by finding the right lose to the air and they do not need to drink so often.
conditions—and some species have different re- The whiskered bat’s glistening coat is due to a cov-
quirements at different times. Figure 4.9 shows the ering of condensed water droplets.
common perception of a hibernating bat—a whis- Many species habitually select open sites and often
kered bat, Myotis mystacinus, a small vespertilionid, hang free of the rock, for example horseshoe bats,
in deep torpor in a cave. Caves selected by bats have which characteristically wrap their wings around
several features that make them ideal sites. They are their body (Fig. 4.10). Many Myotis species, for exam-
cool in winter, but not too cold: away from the ple mouse-eared bats, Myotis myotis, in Europe and
entrance of a cave with limited airflow temperatures the little brown and grey bats, M. lucifugus and
are typically 2–10  C throughout even the harshest M. grisescens, in North America also tend to roost in
temperate winter. The temperature may be very the open. Others, for example Natterer’s bat, Myotis
stable in any given location in the cave, although nattereri, commonly crawl into cracks to hibernate,
different locations may be quite different in temper- although many more, such as Daubenton’s bat,
ature. For example it is cold on the floor, but depres- M. daubentonii, are found in both situations. The rub-
sions in the ceiling retain rising warm air and can be ble on cave floors and even gaps between wall and
considerably warmer than other parts of the cave. clay infill are also used by many species. A single
Bats roosting in such a depression can raise the cave may be used by many species of bat, and it is
temperature further. Different species can therefore not uncommon to find clusters containing several
select the location which suits them best. Caves are species. Throughout the temperate world bats have
typically humid and most species benefit from high adapted to artificial hibernacula such as mines,
108 BATS

and moved position in 44 days, but at 20  C half

moved in just 1 day (McNab 1974b). In contrast,
half of the 25 g greater horseshoe bats, Rhinolophus
ferrumequinum, moved in 1 day at 10  C. To compen-
sate for this effect, larger species will often select
cooler locations in caves. As well as the difference
in size, roosting habit is important, and probably
accounts for some of the difference in behaviour
between Pipistrellus subflavus, a solitary hibernator,
and Rhinolophus ferrumequinum, which often hiber-
nates in clusters. The importance of clustering, inde-
pendent of size, can be seen by comparing Myotis
lucifugus, a cluster former, and Myotis emarginatus,
frequently a solitary hibernator. Both weigh about
9 g, but M. lucifugus arouses far more frequently
than M. emarginatus (McNab 1974a). Clustering
bats tend to maintain higher body temperatures,
relative to ambient, than solitary bats and will select
cooler sites in caves (Beer and Richards, 1956). In
warm caves (Twente 1955, Hall 1962) and in mild
weather (Hooper and Hooper 1956), bats form smal-
ler clusters than in cool caves and cold weather.
Assuming bats have selected the most appropri-
ate natural hibernation sites, how often do they
Figure 4.10 Hibernating greater horseshoe bats, Rhinolophus
arouse? The short answer, as discussed earlier, is
ferrumequinum.
that once every week or two is a rough average,
but the period between arousals can vary from a
disused railway and canal tunnels, cellars, and aban- day to many weeks. A number of studies have
doned kilns that offer similar microclimates. shown that within a species torpor bout duration is
The literature on cave-hibernating bats is quite inversely related to ambient temperature (for exam-
extensive and a number of general statements can ple Brack and Twente 1985, Dunbar and Tomasi
be made about their behaviour. However, as with 2006). A large number of studies of vesper and
studies on the use of torpor during the summer horseshoe bats (for example Daan 1973, Ransome
months, some must be interpreted cautiously, since 1971, Brack and Twente 1985) all agree on the
what bats do under laboratory conditions is not point that arousal is frequent and bats leave the
always what they do in the wild and some of the hibernaculum (Daan 1973) to drink and feed, and
conclusions drawn about the use of hibernacula are to change hibernation site (for example Park et al.
based on laboratory studies. Bats very rarely spend 1999, 2000).
the entire winter in torpor: as discussed above, Work by Ransome (1971) on greater horseshoe
arousal occurs at frequent intervals, dependent bats shows that they select sites within a hibernacu-
upon ambient temperature, the size of the bat, and lum on the basis of ambient temperature and season.
its roosting habits (for example Daan 1973, McNab Warmer sites were chosen after warm days, possibly
1974b, Ransome 1990). A number of studies have to trigger arousal so that the bats could feed during
shown that, within a species, the interval between these warm spells. At the same ambient tempera-
arousals decreases with increasing temperature (for ture, they choose cooler sites in spring than in winter
example Ransome 1971, Twente et al. 1985). Small (Fig. 4.11)—to compensate for an observed increase
bats also tend to arouse less frequently. At 10  C, in arousal frequency at a given temperature, from
half of the 6 g Pipistrellus subflavus in a cave aroused winter to spring.
 TORPOR AND HIBERNATION 109

been carried out on tree hibernators. The red bat,

mean site temperature selected by bats (ºC)

11 Lasiurus borealis, has been found in deep hibernation

torpor hanging from tree branches, but this and
many other tree species typically hibernate in
9
winter tree holes. One of the most extensive studies was
carried out on the noctule, Nyctalus noctula, a 25 g
7 spring bat found throughout Europe (Sluiter et al. 1973).
Bats were kept singly or in clusters of 50 in sections
5
of tree trunk, either in the open or in temperature-
controlled rooms. Single or clustered bats survived
without food for the duration of the experiment
3 (almost 4 months), and clusters of bats could tolerate
ambient temperatures of 16  C by maintaining sur-
1 face body temperature at 9  C. Between 9  C and
0 4 8 12 16 20 10  C, bats aroused every 4–8 days. Temperatures
max. external temperature on day of arousal (ºC) above and below this range often led to daily
arousal. Noctules are migratory in many parts of
Figure 4.11 Hibernation temperature selection in greater horseshoe continental Europe (Strelkov 1969): even holes in
bats, Rhinolophus ferrumequinum, in relation to external temperature and
the thickest tree trunks may not provide sufficient
season (from Ransome 1990).
insulation against the extreme temperatures of a
continental climate, so they may be forced to seek
Boyles et al. (2007) remind us that saving energy is less extreme climates in the south and west.
not the only concern in hibernation. There are eco- Some bats use the most unusual and often
logical and physiological costs to hibernation, such puzzling sites for hibernation. Barbastelle bats, Bar-
as increased predation risk, the inability to maintain bastella barbastellus, have occasionally been found in
physiological and immunological competence when the leaf litter among the roots of large trees. Others
torpid (for example Humphries et al. 2003, Prender- species have been found in small cavities in the soil,
gast et al. 2002), and even sleep deprivation (Daan often gaining entry through a bank side. Some spe-
et al. 1991, Trachsel et al. 1991). Although these have cies, such as the pipistrelle, Pipistrellus pipistrellus,
been studied primarily in rodents, the results are appear to tolerate daily temperature fluctuations of
relevant to bats. It therefore makes sense to balance over 20  C in some of their hibernation sites, and just
costs and benefits. Boyles et al. were able to show how they deal with this in energetic terms remains a
that little brown bats, Myotis lucifugus, with low mystery.
energy reserves selected lower temperatures for
hibernation to minimize energy consumption, rela-
Biological clocks
tive to bats with larger reserves, perhaps incurring
greater costs in an effort to save energy. Under The movement of the Earth brings about environ-
laboratory conditions, blocking fatty acid availabil- mental changes that influence the behaviour of vir-
ity caused big brown bats, Eptesicus fuscus, to select tually all organisms. At the most fundamental level,
lower torpor temperatures. This trade-off is proba- light intensity, temperature, and humidity change
bly important whenever bats use torpor, not just in on a daily (circadian) and an annual (circannual)
hibernation (Wojciechowski et al. 2007). cycle. These changes cause complex, but to some
Some bats, such as the brown long-eared bat (Ple- degree predictable, changes in the biotic and abiotic
cotus auritus) in Europe, are less commonly found environment. Animals have adapted to these cycles
deep in caves, but often spend part of the winter by evolving their own internal clocks, that allow
near the entrance. For much of the winter, long- them to synchronize their activity patterns with the
eared bats hibernate in more exposed sites, often in environmental cycles, and even prepare for changes
buildings or in trees. Relatively few studies have in advance. There are three main types of internal
110 BATS

clock: circadian, circannual, and tidal. I do not think

140
the latter is of any major importance to bats, October
120
although vampires will delay evening emergence
100
to avoid moonlight, and the activity of insectivorous
80
species can be influenced by moonlight (for example
60
Bay 1978, Jung and Kalko 2010). The circadian and mean
40
circannual clocks of bats play a crucial role in torpor peak
20
and hibernation.

60 November
Circadian clocks
40

number of flights within cave

In the absence of external cues, few biological clocks 20
are accurate—for example circadian rhythms are not
set to exactly 24 hours, but can be several hours December
20
longer or, more frequently, shorter. In normal life,
they are kept in synchrony with the environmental January
20
cycles by external cues that reset the biological clock
each day. The most common circadian cue is photo-
February
period—the most invariant of the environmental 20
cues. The importance of a good circadian clock to a
bat is obvious. With insect abundance peaking at
dusk and dawn it’s no good if the bat sleeps in and 60 March
misses them. On the other hand it does not want to 40
arouse too early and waste energy. Some of the best 20
evidence for circadian clocks in bats comes from
studies of hibernating bats.
40 April
We have already seen one example of arousal
20
synchronizing with dusk in Rhinolophus ferrume-
quinum (Fig. 4.8). Figure 4.12 shows the frequency 0 6 12 18 24 6 12 18 24
of flights within a cave over a 48 h period, from
time of day (h)
October to April, of three Myotis species (Daan
1973). The frequency of flights is a good indicator Figure 4.12 The frequency of flights of Myotis bats within a hibernation
of arousal from torpor. In October, the bats arouse in cave (from Daan 1973).
large numbers, with peaks in the evening. The bats
are still leaving the cave to feed outside, so the
biological clock is therefore entrained with the envi-
ronment. As the winter progresses there is much less 2  C, clock period varied from 13–21 h (Twente and
activity, the bats remain torpid for up to 20 days at a Twente 1987), but most bats kept good time and
time, and the peaks are much less evident, since the aroused within two or three hours of dusk. Thomas
clocks of most bats are free-running and no longer (1993, 1995) used automatically-logged echolocation
entrained. In spring, as the bats come out of hiber- calls and temperature-sensitive radio transmitters to
nation and feed, the clocks are again entrained. study hibernating little brown bats in the wild. In
Because it sums the activity patterns of numerous contrast to the laboratory studies, he found no evi-
bats, Fig. 4.12 does not show that in individual bats dence for an endogenous circadian rhythm tightly
the circadian clock can be very stable, even in bats in regulating the timing of arousal. Differences in
deep torpor, although clock period can vary consid- arousal patterns between laboratory and field stud-
erably between individuals. For example, in captive ies are common, and are reviewed by Thomas
big brown bats, Eptesicus fuscus, held in the dark at (1995). In many hibernacula, temperature may
 TORPOR AND HIBERNATION 111

provide no clue to the weather outside, so an ability

to maintain a reasonably accurate circadian rhythm 50
through torpor is important if bats are to take

body weight (g)

advantage of warm nights when feeding might be 40
possible. Failure to arouse means a missed opportu-
nity, and arousal during the day is wasteful of 30
energy. Bats with an intrinsically more stable and
reliable circadian clock should have a selective advan- 20
tage over others. However, as shown in Fig. 4.11,
in the absence of external cues the clock free runs, Aug Aug Aug
1979 1980 1981
and synchrony between the bat and its environment
is lost. Thomas suggests that synchrony of arousal Figure 4.13 Changes in body mass over three years, of a pallid bat
in the laboratory is an artefact, and that wild popu- (Antrozous pallidus), kept under constant environmental conditions (from
lations quickly desynchronize. There are, however, Beasley et al. 1984a).
significant inter-specific differences: synchrony
was the norm in the study of greater horseshoe
bats, Rhinolophus ferrumequinum, by Park et al.
(1999, 2000). 1986), pallid bats were again kept under constant
photoperiod (14 h light) and temperature (23  C),
but the food supply was unrestricted. The bats
Circannual clocks
showed patterns of weight change, food intake,
Another way of keeping ahead in the survival game and reproductive state comparable to those of a
is to be prepared for the future. In bats, this could wild bat. The value of this circannual clock should
mean preparation for the onset of winter. A circann- be as obvious as that of the circadian. The bats will
ual clock, free at least to some degree of the vagaries need to put down fat reserves in readiness for hiber-
of the temperate climate, could be useful to a tem- nation. They must do this in the autumn in the face
perate bat in triggering preparation for hibernation. of fluctuating weather and food supplies. At other
Do circannual clocks exist in bats? times of the year, excess weight increases the cost of
Figure 4.13 shows the body weight changes in a locomotion and impairs flight performance: the
male pallid bat, Antrozous pallidus, over three years amount of stored fat must reflect the needs of the
in a study by Beasley et al. (1984a). The bat was kept bat at that time and not food availability, since this
under constant conditions of photoperiod (14 h may be unpredictable. The bats, in essence, have a
light), temperature (23  C) and food supply. Despite continuously regulated food reserve. The ‘correct’
the lack of external cues, it still put on weight from amount varies with the season, and is regulated by
September to December in preparation for a winter the circannual clock. This clock can free run but is
that did not come. A similar pattern was seen in normally entrained by environmental factors, prin-
many other bats in the study. The clock period cipally photoperiod, the most predictable indicator
varied from bat to bat, and small differences were of the season. On top of this the bat has a fine
noted between the sexes, but the mean for all bats control, determined by a more variable set of physi-
was 300 days, significantly less than one year. Food ological and ecological factors linked to the life his-
intake and reproductive cycles also showed similar tory cycle.
behaviour. The changes in reproductive state and
body weight normally seen in male bats in autumn
Evolution of torpor and hibernation
could be induced by the appropriate change in pho-
toperiod, or by subcutaneous implants of the hor- It is now generally accepted that torpor evolved in
mone melatonin, which is normally secreted by the the tropics, probably in small insectivorous bats.
pineal gland in response to photoperiod changes We have seen in this chapter how tropical ins-
(Beasley et al. 1984b). In another experiment (Beasley ectivores have lower BMRs than frugivores or
112 BATS

omnivores of a similar size, and how they are more Hibernation occurs in members of the Vespertilio-
likely to resort to facultative torpor due to fluctua- nidae and Rhinolophidae, and may have evolved
tions in their food supply and their small size. independently on several occasions. Interestingly,
From their origins in the tropics, bats invaded the some temperate bats may have invaded the neo-
subtropical and then the temperate zones, and tropics, which may explain why tropical horseshoe
many were pre-adapted to some extent to a hetero- and vesper bats do not always maintain high body
thermic existence. As they colonized higher lati- temperatures like other tropical insectivorous bats:
tudes, facultative torpor evolved further, enabling they may not have been there long enough to
bats to exploit some of the most extreme temper- evolve a ‘tropical’ thermoregulatory strategy. For
ate climates by hibernating for extended periods. a more detailed discussion, see McNab (1969).
 CHAPTER 5

Reproduction and development

Reproductive cycles, timing the onset of gestation—different strategies. Monoestrous and polyoestrous
cycles and the environment. Gestation, birth, development, and survival, energetic considerations. Life
expectancy. Infant–mother communication and maternal care. Roost selection and maternal feeding pat-
terns. Mating patterns, sperm competition, sexual dimorphism.

I’ll look at reproduction predominantly from an about the physiological mechanisms that have
ecological and behavioural perspective, concentrat- evolved to enable bats to adjust the timing of the
ing on how different strategies have evolved to meet key stage in reproduction.
different ecological and environmental demands.
The detailed physiology of reproduction in bats is Starting gestation at the right time
varied, but follows the essential mammalian plan in
outline, and I will not cover it in any detail. Hill and Bats have evolved a number of mechanisms to
Smith (1984) give a good account of the basics and ensure that birth is timed to give the greatest chance
Crichton and Krutzsch (2000) have edited a multi- of survival to both the female and its young. This
author volume that brings most aspects up to date. may mean that timing has to be very precise. For
Excellent reviews on various aspects of bat repro- example, in the case of bats at the highest latitudes,
duction can be found in this book and have also the summer is short and the young must be born as
been written by Racey (1982), Tuttle and Stevenson early in the year as is feasible to maximize the time
(1982), Kunz (1987), Kurta and Kunz (1987), Ransome for full development before the onset of winter. The
(1990), and Barclay and Harder (2003). females need a good food supply through gestation
and lactation, the young must be rapidly weaned,
and there must still be time for both mothers and
Reproductive cycles
young to fatten themselves up for hibernation. Just
Bats are placental mammals. After mating and how do they achieve this? The temperate summer is
fertilization, the egg is implanted in the wall of barely long enough for mating, gestation, lactation,
the uterus and the foetus completes its develop- weaning, and pre-hibernation feeding. The last four
ment over a gestation period characteristic of the factors must follow each other in fairly rapid succes-
species. All nourishment comes from the placenta, sion. Once the foetus has begun to develop,
which also meets the embryo’s other physiological although bats have some scope for slowing devel-
requirements—delivery of oxygen, removal of waste opment, as we will see, there is little scope for speed-
products, and immunological defence. After birth ing it up, so a timely start is important. Bats have
the young are protected and given shelter, fed on evolved ways of isolating the first step, mating, from
the mother’s milk, and possibly instructed for a the last four steps of the reproductive cycle, so that
period, before weaning and eventual independence. gestation can start as soon as the conditions are
Within that framework bats show considerable var- right.
iation, and this can generally be linked to climate One mechanism was thought to be unique to bats,
and feeding habits. Before looking at variations in but it has now been observed in a small number of
the reproductive cycle, I need to say something rodents and lagomorphs and is probably more

113
114 BATS

widespread still. This is delayed ovulation and fer- bats, and is found in most pteropodids (for exa-
tilization. It is the most common method used by mlple Pteropus poliocephalus, Nelson 1965) and the
temperate species, primarily vesper and horseshoe majority of species in other families, for example the
bats. Sperm production reaches a peak in late sum- megadermatids (Megaderma lyra, Brosset 1962),
mer/early autumn when the females are in oestrus emballonurids (Saccopteryx bilineata, Bradbury and
(for example Papadatou et al. 2008). The bats will Vehrencamp 1977b), phyllostomids (Phyllostomus
mate any time during this period, and mating can hastatus, McCracken and Bradbury 1981, Brachy-
continue right through the winter in some species, phylla cavernarum, Nellis and Ehle 1977), and vesper-
during periods of arousal from hibernation. How- tilionids (Tylonycteris pachypus and T. robustula,
ever, fertilization does not occur, because the egg Medway 1972). The timing of the cycle is presum-
has yet to be released into the oviduct. The sperm ably governed by a circannual clock, entrained by
are stored and possibly nourished in the oviduct or photoperiod and modulated by environmental fac-
the uterus right through the winter, and ovulation tors. The reproductive cycle is precisely timed, and
and fertilization occur in late winter/early spring, pregnancy can be delayed after copulation in many
often when the female is arousing from hibernation. species, to synchronize birth with the abundant
This ensures that as soon as the climate is suitable food available during the rainy seasons (for exam-
the bat can leave hibernation and the foetus can ple Artibeus jamaicensis). The gestation period is
begin to develop. This is a remarkable process, variable and sometimes surprisingly long: 12–13
since it means that foreign cells, that would nor- weeks in the very small Tylonycteris pachypus (4 g)
mally be attacked and destroyed by the female and T. robustula (6 g) (Medway 1972), which may
immune system, are not only tolerated but perhaps be an adaptation to the fluctuating insect supply.
even nourished for several months. In those species Monoestry is the rule in temperate bats. Mating
that can mate throughout the winter, sperm produc- usually occurs in the autumn and oestrus is
tion normally ceases in September–November (in ‘extended’ over the winter by one of the strategies
the northern hemisphere) and the males may store described in the previous section, so that pregnancy
viable sperm through the winter. begins in the spring. Because of the short summer,
Another strategy is delayed implantation, a method the breeding cycle is typically highly synchronous,
common to many mammals. In this case, ovulation and births occur over a period of as little as two
occurs about the time of mating and fertilization weeks in a given species at a particular location
occurs in the oviduct. Ovulation is spontaneous (Racey 1982). The females of most temperate species
in most bats, but in some it is triggered by copula- form nursery colonies, and the clusters share the
tion. The fertilized egg then undergoes its first few cost of thermoregulation during pregnancy and lac-
cell divisions. What happens next depends upon tation (Dwyer 1971). The energy savings are very
geographical location and climate (see below). considerable and foetal and infant growth are accel-
These strategies are not unique to temperate bats: erated (Herreid 1963, 1967, Tuttle 1975). Synchrony
many phyllostomids use them to synchronize birth is best seen in small species that use cool roost sites,
with the rainy seasons, as we will see below. The last since they make the biggest gains from clustering
strategy, embryonic diapause, appears to be unique (Dwyer 1971). For example, the 7–9 g grey bat,
to a few species of phyllostomid bats of Central and Myotis grisescens, occupies the coldest reported sum-
South America, such as the Jamaican fruit bat, Arti- mer roosts in North America (down to 13  C), and
beus jamaicensis. In this case the embryo implants in birth is highly synchronous (Tuttle 1975). The little
the uterus, but becomes dormant for up to 10 weeks. brown bat, Myotis lucifugus, is the same size, but
roosts at much higher temperatures in the same
areas and shows a lower degree of synchrony than
Monoestry
any bat of similar size and distribution (O’Farrell
In the simplest reproductive patterns, there is a sin- and Studier 1973). The largest North American spe-
gle cycle of oestrus, pregnancy, and lactation each cies, the western mastiff bat, Eumops perotis, selects
year (Fig. 5.1). This pattern is common in tropical warm roosts and has the least synchronous birth
 REPRODUCTION AND DEVELOPMENT 115

ember Januar
ember Januar
Dec y
Fe Dec y
Fe
r br r
be be br
tation
m m trous

ua
oes

ua
e

lac

e
ov

ry

ov

ry
N

N
Mar
er October

anoestrous

Mar
er October

pre
ch
oestrous

ch
gnancy
Phyllostomus
Pteropus hastatus

an oe
April
temn

April
temn
poliocephalus
Sep

st r
Sep
us

n
o
p re o
ati
M
st

M
gn a nc y l act
st
ay
gu gu
ay
Au Au
June June
July July

ember Janua
Dec ry
Fe
r br
be
strous
m
ua
e

oe pr
ov

ry

e
N

gn

Mar
er October

ancy

ch

Tyloncteris
pachypus
April
temn

n
an

ti o
Sep

oe

st
ta

ro c
us la M
st

ay
gu
Au
June
July
Figure 5.1 Monoestrous reproductive patterns.

pattern (Barbour and Davis 1969). O’Farrell and Polyoestry

Studier (1973) studied two related species in the
In the generally more equable but still variable cli-
same building. In Myotis thysanodes, a bat that
mates of the tropics, bats exhibit a very wide range
often uses cooler roosts, births were highly synchro-
of reproductive patterns (see Wilson 1979, Racey
nous, but they were asynchronous in M. lucifugus.
1982, Tuttle and Stevenson 1982, Hill and Smith
Bats with a wide geographic distribution show
1984, Crichton and Krutzsch 2000, for reviews
the most varied patterns. The bent-winged bat Mini-
from different perspectives). Most tropical species
opterus schreibersii (Fig. 5.2), once thought to be a
have two or even three annual oestrus cycles. These
single, widespread species is in fact an assemblage
may be continuous (for example Rousettus aegyptiacus
of three groups of similar species, ranging from
(Fig. 5.3) and Eonycteris spaelaea) or interrupted by
southern Europe down to Australasia (Appleton et
anoestrous (Glossophaga soricina, Fig. 5.3), delayed
al. 2004), but is nevertheless a good illustration (Ri-
implantation (Miniopterus schreibersii, Fig. 5.2), or
chardson 1977). In the tropics foetal development in
embryonic diapause (Artibeus jamaicensis, Fig. 5.3).
Miniopterus begins immediately after mating: in
Individual bats may produce young two or even
March in the north and in October in the south. In
three times a year.
north and south temperate regions embryonic
In the cases investigated, multiple cycles are gen-
development is arrested in its early stages and the
erally linked to seasonal trends in food supply,
embryo is not implanted in the uterus. After a delay
which in turn are determined by environmental fac-
of up to five months, implantation occurs at the
tors such as rainfall and temperature. Underlying
beginning of the summer and gestation continues.
all this is the photoperiod-driven circannual clock.
A similar pattern is seen in the moustached bat,
The importance of the clock itself is uncertain and
Pteronotus parnellii, in the New World.
116 BATS

ember January ember January

Dec Fe Dec Fe
r br r br
be be
mp
ed i lantat lactation
m m

ua
ua

e
e

lay

ov
ov

ry
io

ry
e n

temn October N
temn October N

Mar
Mar

p re

o e st ro u s
ch
ch

gn an c y
North South
o e s t ro

temperate

er
er

temperate
n cy

April
April
na

Sep
Sep
us

eg

el

d
lac pr M ay
ed i t io
n M
m p l a nt a

st
st

gu tatio n ay gu ay
Au Au
June June
July July

ember January ember January

Dec Fe Dec Fe
r br r br
be be cy
m oest m an
gn lac
ua

ua
ro
e

re ta
ov

ov
ry

ry
temn October N

temn October N

p
us

tio
Mar

Mar
anoestrous

n
ch

ch
North South
o es tro
an cy

tropical tropical
er

er
April

April
s
gn

ou
us
Sep

Sep
re

tr

p s
M
oe M
st

st

gu l a c t a ti o n ay
gu an ay
Au Au
July
June
July
June Figure 5.2 Reproductive patterns in the Miniopterus
schreibersii species complex.

probably varies, being more important at higher two and occasionally three pups in separate litters
latitudes where changes in photoperiod are (see also Thomas and Marshall 1984).
significant. One of the most detailed investigations was that
Species whose food is not subject to seasonal fluc- by Bradbury and Vehrencamp (1976a, b; 1977a, b) of
tuations tend to show aseasonal, asynchronous emballonurids in Costa Rica. The proboscis bat,
polyoestry. The common vampire, Desmodus rotun- Rhynchonycteris naso, and the lesser white-lined bat,
dus, breeds continuously and asynchronously, but Saccopteryx leptura, forage in habitats with little sea-
each bat has just two litters of a single pup each year sonal change in their availability of insect prey, and
(Fleming et al. 1972). The fruit eating Artibeus litera- typically breed asynchronously more than once per
tus (Phyllostomidae), and the insectivorous Tapho- year. The white-lined bat, Saccopteryx bilineata, and
zous longimanus (Emballonuridae) and Myotis the sac-winged bat, Balantiopteryx plicata, feed more
mystacinus (Vespertilionidae), appear to breed asea- on seasonally abundant insects and have a single,
sonally and asychronously over much of their range. synchronous birth peak each year. This stresses the
Pipistrellus dormeri fits the same pattern, but from importance of food supply and the flexibility of
November to February the majority of births are reproductive patterns, even between closely related
twins, with a single pup predominant for the rest species. This flexibility can be seen within species
of the year (Madhaven 1978). Patterns are discern- too: the ‘big fruit-eating bat’, Artibeus lituratus, is
ible in species that at first appear to be asynchro- aseasonal and asynchronous near the equator, syn-
nous. The mastiff bat, Molossus sinaloae, although chronously bimodal in Central America, and in
asynchronous, has a tendency to show birth peaks Mexico produces just one young, asynchronously,
twice a year (LaVal and Fitch 1977). Myotis albescens in the first half of the year (Wilson 1979). This and
in Paraguay (Myers 1977) and M. adversus in south- other geographical trends are also presumably due
ern Queensland, Australia (Dwyer 1970a) breed to changes in food supply, but most lack detailed
asynchronously for only half of the year, producing investigation. Many bats have birth peaks that
 REPRODUCTION AND DEVELOPMENT 117

ber Janua ber Janua

ecem ry ecem ry
rD Fe rD Fe
be br be br
m pregn a nc
m oest.

ua
e
ua
e

ov
ov

y us r

ry
ry

r N
r N

eg
str

Mar
lactati s

Mar
on

er Octobe
rou

nan
er Octobe

anoe

ch
ch
oest

Rousettus Glossophaga

cy
o e s a ti o n
t ro u s
aegyptiacus soricina

s
o es t n .
tro u
temn
temn

April
April

la c
t
lac

Sep
Sep

pr ct

la
nc

y
eg ati
nan o n p re g n a

st
M
st

M
Au
gu cy ay
Au
gu ay

June June
July July
r
embe January
r Dec Fe
be br
pre u
m
slow gna
nc
e

ar
ov

y
N

Mar
er October

d ia p

ch

Artibeus
ti o n
ause

ou s

jamaicensis
la c ta

April
o e s tr
temn

la
Sep

o e ct at y
st r i o n nc
ou na
t

p re g
s
gu s ay
Au
June
July
Figure 5.3 Polyoestrous reproductive patterns.

coincide with known or predicted peaks in food suppressed early in the long dry season, presumably
abundance, which makes good sense, since the most to save energy in those individuals not physiologi-
energetically demanding period for the females is cally fit for the second pregnancy. Also coincident
lactation (see Racey 1982). with the rainy seasons were peaks in male body
Figure 5.4 summarizes studies by McWilliam mass and testicular activity. Ovulation and mating
(1987) on Coleura afra roosting in coral caves near occurred at the end of each rainy season. Foetal
Mombassa in Kenya, where there is a long rainy growth occurred at similar rates in the two cycles
season (May–June inclusive) and a short rainy sea- but the poorer condition of the females led to an
son (November and December) each year. The insect early delay in the short rains pregnancy and a longer
groups most commonly taken by Coleura all showed gestation period. Early growth rates of the young
peaks roughly coincident with the rainy seasons, as bats were significantly slower after the short rain
did ‘feeding success’, as measured by terminal feed- pregnancy.
ing buzzes on an ultrasound detector. The first preg- Given the apparent difficulties of the second preg-
nancy was coincident with the short dry season nancy, why do the bats persist with it? The answer
(January and February), and nearly all females in lies in survival. The smaller number of young fe-
the colony gave birth at the onset of the long rainy males from the short rains pregnancy survived the
season. However, the second pregnancy over-ran short dry season better than the long rains young
the long dry season (July–September inclusive) survived the long dry season. Females become
with birth occurring in the middle of the short mature in their first year, and can breed in the sea-
rainy season, when many females failed to give son in which they were born. All long rains females
birth. Their lower body mass at ovulation and bred in their first season, but the majority of short
early in pregnancy, relative to that after the long rains females, like older females, did not breed over
rainy season, suggests this may be due to poor the first long dry season, a tactic that may increase
condition. Reproduction in some of the bats is their overall reproductive success.
118 BATS

3000
(a)

number of insects trapped/night

400 rainfall 2000

300 insects
rainfall (mm)
200 1000

100

(b)
rainy seasons

50
mean no. feeding buzzes
40
every 15 min
30
20
10

(c)
mating

maternity colonies
ovulation
pregnancy
lactation

M AM J J A S O N D J F M A M J J A S O ND
1978 1979

Figure 5.4 The reproductive cycle of Coleura afra in Kenya (from McWilliam 1987).

The themes and variations within tropical bats are to birth. As we have seen, bats can make use of
endless and many are still poorly understood. For a sperm storage, delayed implantation, and embryonic
further glimpse at this complexity, read Bonaccorso’s diapause. The California leaf-nosed bat, Macrotus
(1979) description of Panamanian bats, Fleming et al. californicus, can slow foetal development right down
(1972) on some Central American communities, to achieve the same ends. The gestation period is thus
Wilson (1979) on the Phyllostomidae, and Thomas very variable in bats, both within and between spe-
and Marshall (1984) on three West African fruit bats. cies. As a general rule the larger the bat the longer the
gestation period, from about 40 days to 5–6 months.
In vampires it is as long as 8 months. The 40–50 days
Gestation
typical of a 5–8 g bat is relatively long: in a similar
Determining when gestation actually begins can be a sized rodent or shrew gestation is 20–30 days. Foetal
difficult task and so, therefore, is determining the growth rates in bats are among the slowest in mam-
length of gestation—the period from implantation mals, and comparable with primates (Racey 1973).
 REPRODUCTION AND DEVELOPMENT 119

We saw in the last chapter that heterothermic bats the mother’s weight, but in rare examples can be up
have generally low metabolic rates, and this may in to 40 per cent (for comparison, those of most other
part explain the low rate of growth. mammals, including humans, are 5–10 per cent of
Variation within species is largely due to environ- adult weight). In a survey of the literature by Kurta
mental factors affecting food supply. European pi- and Kunz (1987) the average was 22 per cent of
pistrelles, Pipistrellus pipistrellus, kept at 11–14  C adult weight, compared to an average of 8 per cent
with no food entered torpor, and gestation period for other mammals of a similar size range. However,
increased by 14 days in these bats relative to fed total litter weights were 26 and 25 per cent respec-
individuals within their natural temperature range tively: bats have bigger babies because they have
(18–26  C). However, given sufficient food bats rem- fewer of them. Small bats with low wing loadings
ained homeothermic and gestation period was inde- tend to have relatively larger babies/litters, but
pendent of temperature between 10–25  C (Racey there is considerable variation among the smaller
1969, 1973). At 5  C gestation period increased by species: similar to the situation in non-flying mam-
5 days, and at 30–35  C it was reduced by 2–3 days. mals. Size and degree of development at birth are
In the wild, in two successive years, pipistrelle ges- quite variable even within a species and depend
tation period (from the end of hibernation to birth) upon maternal size, nutritional and hormonal
was 41 and 51 days, the variation being due to poor state, and the size of the litter. The combined weight
weather and fewer flying insects in the second year of twins is generally significantly larger than that of
(Racey and Swift 1981). Gestation period in greater a single baby from the same or a related species
horseshoe bats, Rhinolophus ferrumequinum, in the (Kurta and Kunz 1987), but the degree of develop-
wild can be lengthened when bad weather prevents ment may be similar (Ransome 1973). At birth, pter-
them feeding and forces them into torpor (Ransome opodids have a good covering of fur, their eyes are
1973). Low temperature acts in two ways: by reduc- open, and they are alert. Most other bats are propor-
ing insect availability and by increasing the cost of tionally larger, but are naked with closed eyes (Orr
homeothermy. During periods of poor weather, and 1970). However, their skin rapidly pigments and
thus insect availability, delayed birth may be a cho- grows fur, and the eyes open within a few days.
sen option, not simply obligatory. Bent-winged bats, The phyllostomids are unusual in giving birth to
Miniopterus schreibersii, that typically form large co- advanced young. All bats are born with milk teeth
lonies, have been seen to disperse into smaller, and quickly find the nipple and take a firm grasp.
cooler roosts late in pregnancy, possibly to delay They have the ability to climb and cling to their
birth (Dwyer and Harris 1972). Similar behaviour mother, but are otherwise relatively helpless. Why
has been observed in the grey bat, Myotis grisescens, do bats have such large young? Kurta and Kunz
by Tuttle (in Tuttle and Stevenson 1982). (1987) put forward several ideas: a longer period of
High temperature may also influence gestation. protection in the uterus, better thermoregulation,
Aestivation has been put forward as a possible reduced mortality, a more advanced developmental
cause of delayed birth in the tomb bat, Taphozous state, and the simple fact that female bats have
longimanus (Krishna 1978). Gestation in this 5 g bat unfused pelvic girdles and are therefore able to
may last for up to 4 months. give birth to bigger young. The relative importance
of these factors will be very variable, and is still
largely unknown.
Birth, development, and survival Most bats produce just one pup per litter (which
makes it easier to carry them, if necessary, Fig. 5.5),
Early development
and only in the Vespertilionidae are multiple births
Bats will give birth in the head-down roosting posi- at all common. Most vesper bats produce a single
tion, but the majority of those studied use a head-up pup, a significant minority can produce twins and a
or cradle (horizontal) position, often using the wings few up to three. Four species of Lasiurus give birth
and tail membrane to cradle the baby during birth. to three, four, and sometimes five pups, although
At birth, bats are typically around 20–30 per cent of only the red bat, L. borealis, produces more than
120 BATS

Figure 5.5 Mother and young of the

mouse-tailed bat, Rhinopoma hardwickei.

two with any frequency (see Tuttle and Stevenson their wings must achieve close to adult dimensions,
1982 for references). Older more experienced fe- but this can take as little as 2–3 weeks (Kunz 1987).
males tend to have the larger litters. Some species The greater horseshoe bat, Rhinolophus ferrumequi-
that produce only one pup in part of their range num, (25–30 g) is typical of smaller bats and several
may typically produce twins elsewhere. For exam- colonies have been intensively studied by Ransome
ple the European pipistrelle, Pipistrellus pipistrellus, for over 40 years in the UK (Ransome 1973, 1990,
and noctule, Nyctalus noctula, usually have a single Hughes et al. 1989). In this species, young are capa-
pup in Britain but twins in other parts of Europe. ble of flight in as little as 15 days and are competent
Eptesicus fuscus has one pup in western North flyers at 24 days. Over this brief period, rapid
America, and often twins in the east. Season and growth of the wing reduces wing loading by 30
variation in climate from year to year also play a per cent, and it continues to decrease to day 40, by
role. Myers (1977) found that Eptesicus furinalis which time the young are becoming independent.
generally has twins in its first annual cycle in Para- They are generally weaned between days 45 and 62.
guay, but the second litter of the year, when food Echolocation calls increase in frequency and length
supplies are diminishing, invariably contains just to reach adult values around day 30. This rapid
one pup (see also Madhaven 1978). development is essential if the bats are to meet the
Growth rates are lower in pteropodids than in aerodynamic constraints of flight at an early stage
most other bats, but relative to other mammals and give themselves time to become fully indepen-
they are all fast, and bats achieve independence dent. It also gives their mothers time to recover and
rapidly, aided also by their high birth weight. Before mate prior to hibernation. In the greater horseshoe
they can fly bats generally need to reach 90–95 per bat, large females tend to give birth to larger young.
cent of adult skeletal size (70 per cent mass) and Pups that are born large and grow quickly into large
 REPRODUCTION AND DEVELOPMENT 121

bats tend to have the greatest chance of survival. the wild females do make use of torpor during repro-
Interestingly, reproductive success increases in duction, but to a lesser extent than males (for example
females up to the age of 8 years, and can continue Hamilton and Barclay 1994), and considerable inter-
to be high even at 20 years (Ransome 1990). The specific variation is seen in its use. In the greater
cause of this success in older bats is not yet known, horseshoe bat, Rhinolophus ferrumequinum (Ransome,
but large babies that develop quickly are produced 1973), and the little brown bat, Myotis lucifugus
by well-nourished mothers. (Studier and O’Farrell 1976), females are heterothe-
Development after birth depends primarily on rmic during early pregnancy, but homeothermic
two external factors: the microclimate of the roost, in mid-late pregnancy, before reverting to hetero-
and food availability and the environmental factors thermy just prior to birth and during lactation. A
that influence it. The most important aspect of similar pattern was observed in free-living brown
microclimate for most species, particularly those long-eared bats, Plecotus auritus (Speakman and
from higher latitudes, is temperature. At low tem- Racey 1987), although they remained homeothermic
peratures the metabolic cost of thermoregulation is right up to birth. Audet and Fenton (1988), using
high and less energy can be put into growth. Many temperature-sensitive radio transmitters attached to
species choose warm roosts and/or form large co- big brown bats, Eptesicus fuscus, found that females
lonies, reducing the cost of thermoregulation by were more likely to be heterothermic during preg-
clustering (Dwyer 1963, Kunz 1973). Tuttle (1975) nancy than during lactation. Speakman and Racey
has shown that the early growth rate of the grey bat, (1987) also measured energy expenditure and com-
Myotis grisescens, increases with increasing tempera- pared it to that predicted on the basis of continuous
ture, and at a constant roost temperature it increases homeothermy. Energy costs were significantly
as colony size increases. Very young bats do not lower than predicted overall, and from a low level
thermoregulate and all of the energy goes into during early pregnancy they rose rapidly to exceed
growth in the first few days, hence the importance the predicted level at the time of birth, before falling
of roost temperature (Kunz 1987). By choosing to lower than predicted levels during lactation. Why
roosts that receive radiant heat from the sun, or don’t the bats remain heterothermic and save
that are warmed by the bats themselves, high tem- energy throughout the cycle? Presumably, the foe-
peratures can be maintained (Kunz 1982, 1987, see tus cannot develop sufficiently rapidly under these
below). Rapid growth requires regular feeding for circumstances. Lactation however does not seem to
the mother for milk production. Strong winds, be compromised by heterothermy. Perhaps the
heavy rain, fog, drought, and low temperature all female remains homeothermic long enough to pro-
reduce insect availability or foraging success (for duce milk and feed its offspring, before entering
example Tuttle and Stevenson 1982, Kunz 1987). torpor until its next foraging trip.

Balancing the energy budget during Is energy intake by the female the factor
reproduction which limits litter size and development?
Bats go into torpor when adverse conditions prevent Barclay (1995) made the interesting suggestion that
feeding and this option is open to bats during preg- calcium requirement, not energy intake, limits litter
nancy and lactation too: do they take it? The energy size and growth. Since the overall litter weight is the
demands of reproduction are high, and a pregnant same for bats and other mammals of a similar size
bat could respond by providing this energy, by (Kurta and Kunz 1987), prenatal constraints do not
eating more, or by utilizing stored fat. Alternatively, appear to be the crucial factor—bats put their effort
it could reduce its overall energy requirements by into a single pup, rather than several small ones.
becoming heterothermic. However, heterothermy Young bats and birds are both large when they
may have the disadvantage of slowing down some achieve independence, and this appears to be a con-
stages of reproduction. Several studies suggest that in straint imposed by flight. The forces acting on the
122 BATS

wing bones of bats in flight are very high (Swartz et al. Later development and survival
1992), and fully calcified bones may be required
Survival rates vary with the stage in development.
before young bats can fly: hence the production of a
During preflight development, survival rates appear
single pup that must grow to adult size before inde-
to be high: Tuttle and Stevenson’s (1982) review of
pendence. Until the youngster is flying, the female
the literature yielded a figure of 0–0.12 mortality,
must supply all of its nutritional needs, and the diets
with a mean of about 0.03, and Ransome (1990)
typical of almost all bats (insects, fruit, pollen, and
arrives at a similarly low figure. However, in excep-
nectar) are low in calcium. Birds and terrestrial mam-
tionally poor conditions, such as prolonged bad
mals get calcium from various organic and inorganic
weather during early pregnancy, the females may
sources. Bats can usually meet their energy demands
abandon the roost to ensure their own survival (to
during lactation, but not their calcium requirements,
breed the following year), and mortality may be 1
so the female must put calcium into the milk from her
(for example Nellis and Ehle 1977, Stebbings 1988).
own reserves. A number of testable predictions arise
Typical causes of mortality include maternal stress,
from this hypothesis, but I am not aware that they
predation, and falls from the roost (although fe-
have been investigated. If flight does demand large
males will often retrieve young). Frick et al. (2010a)
‘fledgling’ size, then this has important consequences
using more sophisticated models have recently esti-
to the life-history strategy of bats: the short life, large
mated that first year survival of little brown bats,
litter pattern seen in other small mammals is not an
Myotis lucifugus, was only 0.23–0.46, lower than es-
option for them.
timates in other recently published studies, where
Barclay (1995) raises another interesting point.
values are much closer to those of adults. Young
There is a big debate as to how important size
born late in the summer were much less likely to
alone is in determining variation in mammalian
survive and return to the natal roost and breed the
life-history patterns (see for example Promislow
following year than bats born early in the summer.
and Harvey 1990). Big mammals tend to live for a
The time to first flight shows enormous variation
long time and produce few, slow growing, slow
within and between species, and is dependent on a
developing young. In contrast, small mammals live
wide range of external factors, reviewed by Tuttle
life at a fast pace, producing lots of rapidly develop-
and Stevenson (1982). For example, first flight times
ing youngsters, before dying young. How much of
of Myotis lucifugus in different studies ranged from
this variation is due to size alone, rather than to
14–15 days to 21–30 days. Differences in colony size
phylogenetic, ecological, or other factors? Some biol-
and temperature led to first flight times of 24 and 33
ogists argue size is everything (or almost every-
days in Myotis grisescens.
thing), but in most studies bats (the second largest
Large pteropodids fly at 9–12 weeks of age, with
order of mammals) are under-represented or even
weaning at 15–20 weeks. In most small bats the
neglected altogether. Bats are an effective argument
figures are 2–6 and 5–10 weeks, respectively. Early
against the over-riding importance of size.
flight and weaning are associated with weight loss
Size does have a profound effect on growth rate.
in the young (Tuttle and Stevenson 1982, Kunz
In a review of the available data, Kunz and Allgaier
1987). Stored fat is mobilized to carry the bat over
(1995) found a strong negative relationship: big bats
this difficult period, and in bats with a short wean-
grow slowly. After removing the effects of size no
ing period (for example in Myotis species) fat depos-
apparent effects on growth could be attributed to
its are greater than in those with long weaning
diet, phylogeny (although based on the old megabat
periods (for example Eptesicus and Antrozous)
and microbat suborders), or resting metabolic rate.
(Kunz 1987). During this time the young must per-
Temperate insectivorous bats did, however, grow
fect their flight and echolocation skills, learn to catch
faster than their tropical counterparts. Intraspecific
prey, and familiarize themselves with foraging sites.
differences in growth rate could be attributed to
Mortality over this period is hard to assess, and
gender, litter and colony size, climate, latitude, and
there are few published values. It appears to be
food abundance in certain instances.
comparable to preflight mortality in some cases,
 REPRODUCTION AND DEVELOPMENT 123

but is more variable and can be high (10 to more summer due to predation or the stress of reproduc-
than 50 per cent) (Tuttle and Stevenson 1982, tion. However, Papadatou et al. (2009) found that
Ransome 1990). Young, unskilled flyers are particu- survival rates over winter were comparable to those
larly susceptible to predation and accidents. Other over the summer, as did Sendor and Simon (2003) in
factors influencing mortality include weather, prox- Pipistrellus pipistrellus. With data collected over 16
imity of foraging sites to the roost, the level of years, Frick et al. (2010a) were able to investigate
maternal care, and pesticide poisoning. differences due to weather, showing that female
In all bats, sexual maturity is normally reached in survival was highest in wet summers, probably
1–2 years. In some insectivorous bats, females may due to the increased availability of insect prey. An
be mature in only 3 months. In half of the species even longer data set was used by Schorcht et al.
studied females mature before males, but males are (2009), who analysed a 20-year run of recapture
only rarely ahead of females. Early maturity is most data for Leisler’s bat, Nyctalus leisleri, using bat
common in polyoestrous species. boxes in a forest in southern Germany. Survival
was higher after the first year and in females,
although lower apparent survival in males may be
Life expectancy
due at least in part to permanent emigration—a
How long do bats live? This question was first ad- common problem with all such studies. Bats also
dressed in some detail by Tuttle and Stevenson survived the winter better if it was warm and wet,
(1982), but more recent improvements in methodol- in contrast to findings from other studies (for exam-
ogy have led to a progressive lowering of estimated ple Sendor and Simon 2003). The reasons for the
mortality rates in the literature. Data on mortality differences are not clear, but may be related to
rates after birth are variable, and in many cases choice of hibernation site or whether bats are seden-
unreliable, but the general trend is always the tary or migratory. An important take-home mes-
same; mortality is high in the first year and de- sage, that may be more widely applicable in bats,
creases over the next few years. A good start in life is that adult survival appears to be a major driver
is important. An early birth, a roost with the right of population dynamics, since temporal variation
microclimate, and good foraging close to the roost in adult survival was surprisingly high. This is an
have been shown to be critical factors (Tuttle 1976a, unexpected finding and needs further study.
Ransome 1990, Frick et al. 2010a). I have already A recent study by Papadatou et al. (2011) found
mentioned that young bats enter hibernation late, significant inter-colony variation in survival in the
since they are still feeding and growing as winter isabelline serotine, Eptesicus isabellinus.
approaches. Bats are rather long-lived for their size. Bats, on average, live 3.5 times longer than non-
If it gets through its first year, a typical small bat has flying mammals of comparable size (Wilkinson and
a good chance of surviving to 7–8 years, and there South 2002). The difference is likely to be greater
are many instances, from a wide range of species, of still, since a very much higher proportion of the
wild individuals living for 10–30-plus years. The data from non-flying mammals in this study was
development of increasingly sophisticated capture- from captive animals that usually live longer.
mark-recapture models has led to a re-evaluation of It was long ago suggested that bats were long-
mortality rates in recent years. Annual adult sur- lived because they spent a large part of their lives in
vival rates in papers published up until the mid- torpor, in line with the idea that animals that live life
1990s vary between 0.28 and 0.86, but since 2000, in the fast lane (expending energy and accumulating
published rates have been consistently higher, often tissue damage), die earlier. However, this does not
above 0.80 and as high as 0.94. For example, Papa- explain why non-hibernating bats appeared to live
datou et al. (2009) found that survival rates of adult almost as long as their hibernating cousins. An alter-
male and female long-fingered bats, Myotis capacci- native explanation is that the ability to fly reduces
nii, were comparable at 0.89–0.94. It has been vari- death by predation and this in turn increases lon-
ously suggested that mortality could be higher in gevity. In species prone to predation, disease, or
winter due to starvation during hibernation, or in accidental death, there is no natural selection against
124 BATS

late-developing, deleterious mutations that will the mothers respond to the calls of their own young
shorten life. So, even if they escape the extrinsic (Balcombe 1990). Scherrer and Wilkinson’s (1993)
cause of death, death will nevertheless catch up study of evening bats, Nycticeius humeralis, has
with them early. Bats (and other flying and gliding shown that the calls are heritable, and carry infor-
vertebrates), in cheating predation and living lon- mation about family identity. The young do not
ger, may allow natural selection to weed out these learn the calls from their mothers (although this
early, intrinsic killer mutations, further increasing has been suggested for other species), but use them
their longevity. Wilkinson and South (2002) set out within minutes of birth. This is important if proper
to explain variation in the longevity of bats on the mother–offspring bonding is to occur, since the
basis of differences in their ecology and life history mothers may leave the roost to forage within hours
strategies with interesting results. Reproductive of giving birth, and will need to find their own
rate (pups/year) was the biggest predictor of lon- youngsters on their return. Isolation calls are not
gevity: fewer pups lead to a longer life. Hibernating the only means of identification. Visual and olfac-
species lived, on average, six years longer than non- tory cues are also important (Fenton 1985, Gustin
hibernators, those that made use of cave roosts lived and McCracken 1987, Altringham and Fenton 2003).
an average of five years longer than those that did In most bats, but by no means all, the young are
not, and bigger bats lived longer than small ones. left in the roost when the females leave to forage.
Other factors, such as colony size and diet, did not Foraging insectivorous bats would in general pay
make significant contributions to the variation in too high a penalty in terms of loss of flight perfor-
longevity. It is clear from these results that both mance and cost of transport (Hughes and Rayner
intrinsic and extrinsic factors contribute to longevity 1993). The young of many species, for example
and some of the results are not easily explained, but Tadarida brasiliensis, Macrotus californicus, and Mini-
Wilkinson and South (2002) draw some fascinating opterus schreibersii cluster when left in the roost,
conclusions. reducing the costs of homeothermy. Very young
bats are poor thermoregulators and may not be
Infant–mother communication able to maintain high body temperatures without
clustering. The females of these species roost sepa-
and maternal care rately, unless they are suckling young. Female
The rearing of young is carried out exclusively by fringed myotis, Myotis thysanodes, and little brown
the females. The bond between females and young bat, M. lucifugus, are said to baby-sit young while
is established at birth and vocalization plays a major the rest of the colony is foraging.
role in this bonding. Fenton (1985) and Altringham Communal behaviour does not appear to extend
and Fenton (2003) have reviewed communication to suckling young as frequently as was once
between mother and offspring and other aspects of thought. For example, it was widely accepted to be
bat communication. On returning to the roost after the case in the Mexican free-tailed bat, Tadarida bra-
foraging, the females of many species emit special siliensis, but in a study by McCracken (1984) 83 per
search calls, to which the young have specific cent of sampled females were feeding their own
replies. In echolocating bats, these calls are often young. Kin selection or reciprocal altruism are
ultrasonic and resemble segments of the foraging unlikely to be operating in the 17 per cent of cases
calls: they become increasingly complex as the where bats are feeding young other than their own,
young develop (Jones et al. 1991). There is now since the youngsters move appreciable distances
considerable evidence that calls emitted by an indi- between feeding bouts, and individuals within a
vidual bat are sufficiently reproducible, and calls given area are unlikely to constitute a kin or recipro-
between bats sufficiently varied, to enable mother cation group. The still significant degree of indis-
and youngster to identify each other (for example criminate suckling observed may simply be due to
Balcombe and McCracken 1992, Scherrer and Wilk- the enormous size of maternity roosts, and the diffi-
inson 1993), and behavioural studies confirm that culties mothers have in finding their own offspring.
 REPRODUCTION AND DEVELOPMENT 125

In T. brasiliensis, maternity roosts can contain Vaughan 1976). Young vampires will feed with
millions of bats. Wilkinson (1992a) studied the even- adults, and the young of some insectivorous bats
ing bat, Nycticeius humeralis, with colonies of only (for example lesser white-lined bats, Saccopteryx lep-
15–300 females. Evening bats are weaned around 40 tura) will shadow their parents, mimicking their
days after birth. Communal nursing was never seen twists and turns, and copying echolocation calls to
before day 8, but increased slowly to a peak over the the terminal feeding buzz (Bradbury and Emmons
last two weeks, when the youngsters were begin- 1974). Observations of the pallid bat, Antrozous pal-
ning to forage for themselves. During this period lidus, suggest that young may learn to forage from
females suckled young other than their own 18 per their mothers (O’Shea and Vaughan 1977). This is not
cent of the time. As in the case of the Mexican free- the case, however, in the little brown bat (Buchler
tailed bat, kin selection and reciprocal altruism are 1980), and radiotracking of greater horseshoe bats,
not operating. Interestingly, the females tended to Rhinolophus ferrumequinum, has shown that mother
accept non-descendent female young, but rejected and offspring leave the roost separately and forage
males. Wilkinson makes the intriguing suggestion apart (Jones et al. 1995).
that the females may feed non-descendent young- Many maternity colonies include males as well as
sters to dump excess milk and reduce wing loading females, but the males do not appear to play any
prior to foraging. Why should they restrict feeding part in rearing the young. Male Phyllostomus disco-
to female youngsters? Since females generally return lour are often found with youngsters sleeping on
year after year to the roost in which they were born, their backs (Bradbury 1977b), but to what extent
this strategy may help to maintain colony size, and there is bonding is unknown.
hence information transfer between females about
foraging and roosting sites. Males on the other hand
are likely to disperse. In a review of the literature Roost selection and maternal foraging
Fenton (1985) found that the vast majority of species
patterns
studied appeared to nurse only their own young
most of the time. This topic is really part of the next chapter, but one
In the later stages of lactation maternal care in or two points are perhaps most relevant here. First
large insectivorous and carnivorous species may of all, roost selection is critical to reproductive suc-
extend to food provisioning. For example, captive cess. As we have seen above, roost microclimate is
Indian false vampires, Megaderma lyra, will supply one of the most important factors influencing
young with whole frogs and dismember larger frogs growth and development. Suitable roosts may not
(Raghuram and Marimuthu 2007). In the wild, the be available close to the best foraging sites, and
pallid bat, Antrozous pallidus, is known to feed its commuting distance may influence the foraging suc-
young on arthropods, and African sheath-tailed cess of the adults. Tuttle (1976a) showed that as
bats, Coleura afra, return to the roost with cheek commuting distance increased, the mortality of
pouches full of insects. Other species believed to young grey bats, Myotis grisescens, making their
take food back to the roost include the small Indian first flights also increased.
false vampire, Megaderma spasma, and the Austra- Weather permitting, the females will typically
lian ghost bat Macroderma gigas. Captive Molossus forage every night during pregnancy and lactation,
ater feed their young on insects, as do Noctilio albi- to feed their growing young. Few data are available
ventris. This practise may be more widespread than from individual bats, but studies of whole colonies
is generally thought, although it has not been can, if interpreted with caution, give some insights
observed in many extensively studied species. In into behaviour.
captivity the common vampire will regurgitate Figure 5.6 shows the percentage of female pipis-
blood for offspring (see Racey 1982 and Fenton trelles out of the roost at different times during the
1985 for references). Once they are flying, young- night, over a Scottish summer (Swift 1980). Early in
sters may accompany mothers on foraging flights pregnancy, the females do not need to forage for
(Myotis adversus, Dwyer 1970b, Cardioderma cor, long. They all leave at sunset, but soon start to
126 BATS

early pregnancy late pregnancy lactation weaning post-weaning

100
percentage
of colony
outside
roost
0
20.00 04.00
time

Figure 5.6 Activity patterns of the pipistrelle, P. pipistrellus, during pregnancy and lactation (from Swift, 1980, with permission from John Wiley and Sons).

come back, and few bats are out in the early morn- informative ways of categorizing bat mating systems.
ing. As pregnancy progresses more bats stay out McCracken and Wilkinson (2000) organized them on
late. During lactation and weaning, when the de- the basis of roosting and mating associations and
mands on the females by their growing young are their stability and seasonality. In perhaps the most
greatest, they go out to feed twice, at dawn and widely cited general review of mammalian mating
dusk. Once the youngsters start to fend for them- systems, Clutton-Brock (1989) identified the most
selves, the females need to feed less, and the activity important factors influencing mating strategy, and
curve reverts to its unimodal form. This pattern has related them to the patterns observed in nature.
been observed in a number of species, including I thought it would be interesting to see to what
some where individual bats were tracked, and illus- extent bats could be described in this way. Figure
trates how reproduction can influence foraging 5.7 summarizes mating systems and the ecological/
activity, but there is considerable variation. Once behavioural characteristics of females normally as-
the youngsters are weaned, the colony may begin sociated with each system. It is based on the scheme
to break up at any time. From now until hibernation, set out by Clutton-Brock, with some modification to
the adult bats are pre-occupied initially with mating fit the observed patterns in bats. It is widely recog-
and later with preparation for hibernation. The nized that it is the social and ecological characteris-
young must become competent foragers, familiarize tics of the females that are largely responsible for the
themselves with roosting sites, complete their mating strategy that has evolved in the males. A
growth, and prepare for hibernation. Those that common strategy in bats is that known as ‘resource
mature early may also mate. defence polygyny’, in which males defend a resource
against other males, enabling them to copulate with a
Mating behaviour large number of females. This resource may be a
foraging area, a roost site, or the females themselves.
Bats probably show more forms of mating beha- Although a widely recognized strategy, few studies
viour than any other mammalian order. The bewil- actually identify the resource being defended. In Fig.
dering range of patterns seen must have its origin in 5.7 the different resource defence strategies are listed
natural and sexual selection. If we can understand separately. I will briefly describe the patterns first,
how the different mating strategies work, why a and then look at examples.
particular species has evolved to use a particular
strategy, and the selective pressures operating on
Mating patterns
each system, then some form of order should come
out of the confusion. Bradbury (1977a) has described Obligate monogamy, males mating with a single
mating systems in some detail, and Bradbury and female in a given season, commonly arises when
Vehrencamp (1977a) developed a general model of the young can be reared successfully only with the
mammalian mating systems based on their studies of help of the male, or when male help significantly
emballonurid bats. There are several sensible and increases breeding success. This situation has not yet
 REPRODUCTION AND DEVELOPMENT 127

rear young only rears young

range defensible range indefensible

foraging site
defended groups stable groups unstable

females females
defended defended

roost
variable
defended
strategy
resource defence polygny

solitary groups groups groups groups s in group many s in common s solitary

small large large small home range, migration, or widely
range hiberntion distributed

obligate facultative uni multi multi uni mating mating aggregations/swarming roving s
monogamy monogamy/ groups groups groups groups territories leks
polygyny (harems) (harems) temporary harems/territories

Nycteris spp? Saccopteryx Saccopteryx Rhynchonycteris Phyllostomus hastatus Pipistrellus nanus Aggregations/swarming:
Lavia frons? leptura bilineata naso [stable] P. pipistrellus Balantiopteryx plicata
Taphozous peli? Peropteryx Pteropus giganteus Pteropus poliocephalus Nyctalus noctula Miniopterus schreibersii
[stable] Myotis adversus Myotis lucifugus
Rhinolophus luctus? keppleri
Artibeus jamaicensis Artibeus jamaicensis Myotis sodalis
Vampyrum spectrum? Hipposideros - in caves [unstable] - in trees Leks: Hypsognathus monstrosus
beatus Miniopterus australis Ectophylla alba? Harems: Tylonycteris spp.
Rhinolophus [unstable]
sedulus?

Figure 5.7 Mating patterns in bats (adapted from Clutton-Brock 1989).

been shown convincingly in bats. Males do not territories at sites frequented by females. As female
appear even to participate in rearing—they are cer- density increases, male territories become less dis-
tainly not essential in any species of bat studied. persed, ending in leks at the highest densities of
All other patterns arise when the male is dispens- females: small, adjacent male territories (or even an
able as far as rearing is concerned. He is thus more of absence of territories) from which males display to
a free agent, and able to pursue other females. The compete for passing females. Some mammals mate
main factors become the range and social behaviour on migration or at the onset of hibernation, and
of the females. If the home range of the females can behaviour is flexible: defence of individuals, (tran-
be defended by a male, then three main systems sient) harems, mating territories, and swarming are
may arise, depending upon the size of the female all common. Again, all are seen in bats. Finally, if
groups: facultative monogamy/polygyny, unimale females are thinly or unpredictably distributed
(harem) groups, or multimale groups. All involve males rove in search of individual females. I have
males holding a territory and defending the female not seen this documented for bats, but I suspect it
(s) within it. If the female group is large enough, does occur.
then more than one male may be involved. All three
systems are seen in bats, particularly in the tropics.
Some examples
If the females’ range is too large or too mobile to
be defended, then female group stability may be the Too few bat species have been studied in sufficient
next determinant. With stable groups, harem and detail to give many examples of each pattern, or to
multimale groups may form, based on defence of place them in a category with any certainty. In fact
the group, rather the territory. If the female groups the various categories are not mutually exclusive,
are unstable, males typically defend small mating and many bats show elements of several strategies.
128 BATS

I have categorized them, sometimes tentatively, on against other males in the colony, and competition
the basis of the most prevalent strategy. Obligate for females is intense. However, males do not
monogamy has yet to be verified in any bat, but a attempt to retain females that leave their territory.
number of species are routinely found in pairs or Mating occurs at the day roost, so males can only
family groups, and possible candidates include the benefit if the females that roost with them are the
African slit-faced bats (Nycteris species), the neo- only bats allowed into their feeding territories.
tropical phyllostomid Vampyrum spectrum, and the Recent work by Behr et al. (2006) has revealed
Old World bats Taphozous peli and Rhinolophus luctus more detail of this species’ behaviour. Males main-
(see Bradbury 1977a for references), but all could be tain their territories by counter-singing whenever
facultative. Monogamy has also been reported in the they leave or return to the roost. The frequency
yellow-winged bat, Lavia frons (Vaughan and with which they call and the spectral characteristics
Vaughan 1986, 1987), with a pair sharing a territory, of the calls themselves both influence mating suc-
defended by the male, for about three months. Their cess (parental assignment was determined by micro-
harsh, scrubby habitat may make cooperation essen- satellite analysis). Territorial males father more
tial to the rearing of young, and several behavioural offspring than non-territorial males and those terri-
traits are described that suggest some degree of help torial males that call most frequently, using the low-
from the male. Monogamy may also occur in Hippo- est frequencies in a characteristic ‘long buzz’, were
sideros beatus and Rhinolophus sedulus (see Heller the most successful.
et al. 1993), but further study is needed. Why should A fourth emballonurid, Rhynchonycteris naso,
it occur if the male does not have to participate in feeds along rivers in the same part of the world.
rearing? Foraging area defence by single females Group size is similar to that of S. bilineata, but stable.
and large home ranges used by females could both However, bats frequently change roost. Female
lead to the evolution of monogamy, since it would rather than resource defence would therefore be
be energetically costly for males to chase several the predicted strategy, and that is what is seen
females. Both could be factors for carnivorous bats (Bradbury and Vehrencamp 1977a). Multi-male
such as Vampyrum spectrum. groups form, and males only defend territory
The neotropical emballonurids Saccopteryx leptura against males from other colonies. Within the col-
and Peropteryx keppleri, two of five species studied ony, individual males will follow foraging females if
by Bradbury and Vehrencamp (1977a), defend feed- they move feeding site.
ing territories used by females and typically form The fifth species, Balantiopteryx plicata, forms very
monogamous pairs. Foraging areas are small, fe- large colonies of up to 10,000 bats. Group stability is
males roost singly or in small, stable groups and low, and roosting and foraging sites change fre-
they are thinly scattered through their wet forest quently. Defence of foraging site, roosting site, or
habitat. It is therefore feasible for males to defend females by the males would be impractical and
feeding territories, but in so doing they will defend costly. Although it was difficult to study this species
only one or perhaps a few females. The main in detail, the evidence suggests male mating aggre-
resource appears to be the foraging area, which is gations are visited in their cave roosts by females
defended by males and females, but the females during the short copulation season. This pattern is
themselves may also be defended against other observed in some temperate species that mate prior
males. to hibernation, for example Miniopterus species in
In contrast, Saccopteryx bilineata roosts in large Australia (Dwyer 1966) and probably in northern
but unstable groups (of up to 40 bats) and forms temperate regions too.
harems of one male and up to eight females within Myotis, Eptesicus, Plecotus, and Barbastella species
the group (Bradbury and Vehrencamp 1977a). swarm in late summer and autumn at underground
Although the individual bats present in a group hibernation sites. Swarming has been studied in
turnover quite frequently, the roost site is used for North America (for example Barbour and Davis
long periods. Each male defends part of the colony’s 1969, Fenton 1969, Thomas et al. 1979), Europe
feeding territory, and the tree buttress roost site, (Parsons et al. 2003, Rivers et al. 2006, Glover and
 REPRODUCTION AND DEVELOPMENT 129

Altringham 2008, Gottfried 2009) and probably oc- (Fig. 5.8c, Glover and Altringham 2008, Kerth et al.
curs in other temperate parts of the world too. Pri- 2003), behaviour characterized by intense chasing
marily during August and September (Fig. 5.8a), (for example Rivers et al. 2006), and some evidence
large, transient populations visit caves and mines, of skewed mating success (Watt and Fenton 1995).
often from considerable distances. Swarming sites A fascinating observation is the arrival of cohesive
may attract hundreds or thousands of bats in a male groups at swarming sites over several seasons.
single night, of more than ten species, with many This is particularly unexpected, given the very high
thousands visiting in a single season. The mating turnover of bats and the fact that males frequently
system involved in swarming has not yet been roost alone. An analysis of the recapture patterns of
worked out, but several features suggest a form of ringed bats showed that groups of up to five Myotis
lek (see below): a marked male bias (70–95 per cent, nattereri, with near consecutive ring numbers (they
Fig. 5b), that is less marked at the peak of the season were ringed together on a previous occasion), were

(a)
80

70
mean number of bats caught

60

50

40

30

20

10

0
09/08 16/08 23/08 30/08 06/09 13/09 20/09 27/09 04/10 11/10 18/10 25/10
Date (week ending day/month)

(b)
(c)
80 30
% females

60 20
% catch

40 10
M

20 0
early mid late end
F swarming
0
s
tii

us

itu
i

i
er

i
on

nd

cin
r

r
te

au
nt

ra

ta
at

be

.b

P.
ys
.n

au

.m
M

.d

M
M

Figure 5.8 Swarming activity of Myotis and Plecotus species peaks in August and September (data from four caves in the Yorkshire Dales, UK) (a) and there is
a marked male bias in all species (b), that is less evident at the peak of the season (c). Adapted from Glover and Altringham (2008).
130 BATS

captured far more often than could occur by chance, Tropical species also make use of mating terri-
up to four years apart (Rivers 2005). This has subse- tories, for example Pipistrellus nanus in Africa
quently been shown to be true for Myotis daubentonii (O’Shea 1980), Myotis adversus in Australia (Dwyer
and Plecotus auritus (unpublished) and presumably 1970b), and Artibeus jamaicensis in the neotropics
holds for other swarming species too. Are these (Morrison 1979). When A. jamaicensis roosts in
males roosting together at distant ‘summer’ roosts trees, single males defend holes in the trunk or
and making the long journey to their swarming site branches against other males (Morrison 1979).
as a group? These groups appeared to be no more Groups of females (4–11) roost in these holes, but
related to each other than a random sample of other group composition is very variable. Their food
males caught on the same night, but the methods source (Ficus species) is dispersed but abundant,
used may not have had sufficient resolving power. trees fruiting asynchronously for 1–2 weeks each
Individual bats are typically faithful to a single year. Females disperse to feed, preventing their de-
swarming site. In an area where numerous swarm- fence by the male. Instead he spends 90 per cent of
ing sites are in close proximity, less than 1 per cent of the night flying within 50–100 m of the roost, chas-
bats were recaptured at more than one site (Glover ing males and escorting returning females back to
and Altringham 2008). the hole. This flight is fuelled by occasional foraging
The timing of swarming is correlated with cli- trips for fruit, which is carried as much as 1.5 km
mate, occurring earlier at higher latitudes and alti- back to the roost for consumption (Morrison and
tudes: at 1,800 m, in the Tatra Mountains of Poland, Hagen-Morrison 1981). Females feed close to the
it peaks in July and August (Piksa 2008), but in more fig tree providing the fruit. The male may expend
equitable climates it may peak a month or more later 30 per cent of his daily energy budget on roost
(for example Rivers et al. 2006). Some of the varia- defence.
tion in the timing of swarming may be due to differ- As the density of females increases, male terri-
ences in species composition, since the timing of tories may become smaller and more closely packed.
peak activity varies between species (for example Ultimately this grades into a lek: females visit
Glover and Altringham 2008). Swarming sites are groups of males occupying small adjacent territories
frequently also hibernation sites and an additional and the males display to the females, who choose a
function may be to locate and assess hibernacula. mate for copulation. No resource is defended, but
Early in the season there is no net influx into the site some males are more successful than others at at-
at the end of the night, but late in the season bats tracting females due to some quality of their display.
stay in increasing numbers to hibernate (for example The African epomophorine bat, Hypsignathus mon-
Rivers et al. 2006). strosus (hammer-headed bat) is undoubtedly the
Lek formation may evolve in the face of large best known example (Bradbury 1977b). The males
numbers of mobile females within a common have an enormous larynx (20 per cent of the body
home range. This could be thought of as an exten- cavity) which enables them to emit loud, low fre-
sion of the mating territory strategy to high densi- quency honks. As many as 90 males line up along
ties, so I will deal with the two together. If female river banks, one every 50 m, and honk continuously
groups are moderately large and unstable, but fe- to passing females. The bats honk at a rate of 50–120
males forage within a well-defined home range, per min and beat their wings. When females
males may set up mating roosts within foraging approach, honk rate and wing flapping increase.
areas or on commuting routes and attract passing The specific cues that win female attention are not
females, defending the roost and a small airspace known, but the call must have an important role.
around it. This is common among European temper- The behaviour of Artibeus jamaicensis in tree roosts
ate bats: males occupy tree roosts and call from their was described above. This species is interesting in
roost (Nyctalus noctula, Sluiter and van Heerdt 1966), that it frequently roosts in caves, where its beha-
or use a songflight (Pipistrellus pipistrellus, Gerell viour is different. Kunz et al. (1983) found that the
and Lundberg 1985) to advertise their presence favoured sites were solution cavities in the cave
(Fig. 5.9). roof. Unlike the tree holes used by the bats in the
 REPRODUCTION AND DEVELOPMENT 131

Figure 5.9 Male pipistrelle, P. pipistrellus, in songflight.

Morrisons’ studies, these cavities were not in limited stable: turnover is low and comparable to the annual
supply: only 13 per cent were used at any one time, mortality rate. This stability is maintained even
and the females frequently changed roosts. The cav- through changes of harem male.
ities are clearly not a defensible resource: are the Dechmann et al. (2005) describe an interesting case
males defending the females instead? If so, they of resource-defence polygyny. Lophostoma silvicolum
should move from cavity to cavity with the females, roosts in active termite nests by excavating a small
but fidelity of males to particular female groups has chamber on the underside. Single males carry out
yet to be demonstrated, and there is switching of this energy-consuming activity and attract a small
females between groups. Whatever the strategy, it is harem of up to six females. Males gain from the high
not that used by the tree roosting members of the reproductive success associated with harem-keeping.
species studied. A similar system seems to be The females appear to benefit from a warm, dry,
operating in Miniopterus australis in Borneo (Med- robust roost and probably the high fitness of the
way 1971). Males are found with harems of up to six males capable of excavating roosts in the very hard
females, but roost fidelity and group stability are nests.
both low. Roost defence is unlikely, but female de- Defence of one or more females is also typical of
fence would involve frequent changes of the females many Old World fruit bats. In Pteropus poliocephalus
defended. in Australia, single males select one or more females
In the neotropical Phyllostomus hastatus, the case for and the group as a whole defends its small territory
female defence is clear (Fig. 5.10). Single males defend on the camp tree (Nelson 1965). Elements of both
harems of 7–25 females (18 on average) in their cave roost and female defence are therefore present,
roosts, in some cases for several years (McCracken although the latter is perhaps the most important.
and Bradbury 1977, 1981). Female groups are very Male Pteropus giganteus in India are very possessive
132 BATS

Figure 5.10 Phyllostomus hastatus

harem.

of females in the vicinity of their defended tree roost, Helversen 2003) who suggest that female choice is
but bonding does not occur and females move an important component of this harem system.
around the camp tree to some extent. Sometimes a particular behaviour offers only long
As suggested by the work on Artibeus jamaicensis, term benefits. Artibeus jamaicencis have a complex
a species may use more than one mating strategy. system with dominant, subordinate, and satellite
Senior et al. (2005) have shown that in Daubenton’s males. Dominant males tolerate subordinate males
bat, Myotis daubentonii, the mating strategy used is around the harem, since it allows them to defend
intimately related to habitat quality and female be- larger harems. The subordinates gain little immedi-
haviour, and that mating can occur in summer ate benefit, fathering only 9 per cent of offspring
roosts as well as at swarming sites in the autumn relative to 69 per cent by the dominant males. Satel-
(see also Encarnação et al. 2007). Mating strategy in lite males, that sneak into the harem, fathered 22 per
Saccopteryx bilineata is to some degree age-depen- cent. However, subordinates are more likely to fill
dent. Older males are more likely to be territorial positions left by missing dominant males than satel-
and to reproduce than younger males (Heckel and lites. In large harems, subordinates appeared to be
Helversen 2002). However, peripheral males did sons of the dominant males and together they
reproduce and some individuals were as successful fathered all offspring in these groups (Ortega and
as some territory holders. The social and mating Arita 2002, Ortega et al. 2003). I could go on describ-
system of this bat is complex and was explored in ing the apparently infinite variety of mating sys-
more detail by the same authors (Heckel and tems, but time to move on.
 REPRODUCTION AND DEVELOPMENT 133

Sperm competition
Given the long periods over which bats can store
sperm, and the wide variety of social and mating
systems they exhibit, they are an ideal group for the
study of sperm competition (Hosken 1997, 1998).
Not surprisingly, testes mass varies from 0.12 to
8.4 per cent body mass (Wilkinson and McCracken
2003) and is strongly correlated with social group
size and mating system, in particular female pro-
miscuity (Hosken 1997, 1998, Wilkinson and
McCracken 2003, Pitnick et al. 2006). As the likeli-
hood of sperm competition increases amongst spe-
cies, more resources are put into sperm production.
About half of the variation in relative testes size
could be explained by group size in Old World
fruit bats (Hosken 1998) with the mating system
presumably explaining much of the remainder.
Vonhof et al. (2006) found that almost half of the 26
pairs of twins found in colonies of the big brown bat,
Eptesicus fuscus, had different fathers, and similar
findings were made by Mayer (1995) in the noctule,
Nyctalus noctula. Females of a number of other spe-
cies are known to solicit copulation from more than
one male (for example the greater horseshoe bat,
Rhinolophus ferrumequinum, Rossiter et al. 2000a). It
is of course possible that there is a significant ele-
ment of choice on the part of the females.
Figure 5.11 Male Hypsignathus monstrosus.

Sexual dimorphism and sexual signals

insectivorous bats too: females are commonly larger
I have mentioned songflights and lekking behaviour than males. A very striking example is the large crest
already. The hammer-headed bat, Hypsignathus of the African crested free-tailed bat, Chaerephon
monstrosus, shows an extreme form of sexual dimor- chapini (Fig. 5.12). A more detailed look at this sub-
phism: the enlargement of the larynx and other ject, and other areas of communication, can be found
structures for honking gives the males a grotesque in Fenton (1985).
and distinctive head (Fig. 5.11). I’ll finish this section with a very brief look at
Males of many epomophorine species are larger vocalization in courtship. A study by Behr and Hel-
than females and have display markings, typically versen (2004) demonstrates the level of complexity
on the head and shoulders. They often go in for seen in some species. Male sac-winged bats, Saccop-
wing shaking to attract the females, and mutual teryx bilineata, have different calls for territorial de-
grooming. The shoulders of these and other bats fence and courtship. Courtship songs were long,
have glandular sacs that secrete an odour to attract complex, and variable (Fig. 5.13), being made up of
females (for example Sac-winged bats, Saccopteryx, calls that fell into four main groups. Different males
Voigt and Helversen 1999). Quite a few bats have had different repertoires, giving females the poten-
courtship songs, or rather courtship screams, grunts, tial to choose mates based on the richness and com-
and bellows. Sexual dimorphism is seen in many plexity of their calls. If you want to know more, Behr
134 BATS

Figure 5.12 Male Chaerephon chapini.

trills
frequency (kHz)

noise burst
80

0 Figure 5.13 Mating call components of male

50 ms
sac-winged bats, Saccopteryx bilineata. These
short tonal
quasi-CF components are combined to make complex
songs (from Behr and Helversen 2004, with
permission from Springer Science+Business
Media).

and Helversen’s (2004) paper is a good starting What about genes?

point, with Davidson and Wilkinson (2004), also
Finally, mating is all about passing on genes, and
on Saccopteryx bilineata, and a more recent paper by
mating systems have a profound effect on the genetic
Bohn et al. (2009) on Brazilian free-tailed bats, Tadar-
structure of populations—yet I have said very little
ida brasiliensis.
about genetics. I will cover this when I discuss social
 REPRODUCTION AND DEVELOPMENT 135

structure, dispersal, and population structure in elevated inbreeding. Greater horseshoe bats, Rhinolo-
Chapter 7. I will finish this chapter with one phus ferrumequinum, achieve this because females in
fascinating study that does rely on genetics. Rossiter the same maternal line share males and mate with
et al. (2005) addressed the interesting question of how specific males across the years. This strengthens
a mating strategy can lead to increased kinship genetic ties between females, improving group cohe-
within social groups of females without the risk of sion whilst minimizing the genetic costs.
This page intentionally left blank
 CHAPTER 6

Roosting and feeding ecology

Roosting ecology—the importance of finding the right roost. Advantages and disadvantages of communal
roosting. Where do bats roost: caves and crevices, trees, foliage, tent-making bats, man-made roosts, etc.
Behavioural ecology of roosting, sexual segregation. Feeding ecology: the diversity of feeding habits and
adaptations, with diversions into reciprocal altruism, optimal foraging, and convergent evolution. Informa-
tion transfer. Interactions between habitat, food, foraging, and social structure and behaviour, sexual
segregation again.

Roosting ecology · competition avoidance—few other vertebrates

can make use of most bat roosts.
The importance of finding the right roost
The processes involved in roost selection are com-
Bats spend most of their lives in their roosts. Roost- plex, with many mutually dependent interactions.
ing habits influence local and global distribution, The availability of suitable roosts, for example, will
densities, foraging and mating strategy, social influence foraging behaviour, but roosting beha-
structure and seasonal movements, and even the viour itself may be influenced by the abundance
morphology and physiology of bats. An under- and dispersion (in time and space) of food, and
standing of their roosting ecology is therefore fun- both have an important role in determining the
damental to an understanding of bats. The most social structure of intra- and interspecific bat
wide ranging and thorough reviews are by Kunz communities.
(1982) and Kunz and Lumsden (2003). I have Before looking at some of these interactions I want
already talked a little about roosting ecology to review the enormous diversity of roosting beha-
when we covered hibernation and reproduction. viour. There are some general trends, and there are
In this chapter I will take a broader view, and many exceptions to all of them, but here are the
look at the great diversity of roosting adaptations broadest. Bats in the tropics are known to use exposed
and show how they relate to other aspects of bat roosts but in the harsher temperate climate exposed
biology. By choosing a suitable roost, bats can gain roosts are rare. Bats that roost in the open tend also
many advantages. Here are the most important to be large. Bats in high latitudes tend to roost in
potential benefits: larger groups. Opportunistic (adaptable, generalist)
·· protection from the weather;
protection from predators;
roosters, such as the long-eared bats shown in
Fig. 6.1, tend to be more widespread than specialists.
· cheaper thermoregulation—energetic savings du-
ring roosting;
Bats utilizing stable roosts, such as caves, are fre-
quently very faithful to these sites, over many years
·· reduced commuting costs to foraging sites;
improved mating opportunities;
and many generations. Ephemeral roosts, such as
those in foliage, may increase local movements,
·· improved maternal care;
information transfer—exchanging knowledge of
but bats may still be faithful to particular locati-
ons. Within this broad framework is a wealth of
foraging and roosting sites; fascinating detail.

137
138 BATS

Figure 6.1 Brown long-eared bats,

Plecotus auritus, roosting in the roof of a
building.

Where do bats roost? weak solubility in rainwater. Even when present,

few caves may be suitable for bats, and of those
Caves and rock crevices
that are, only part of each cave may offer the appro-
Caves
priate microclimate. It may be these factors, together
Let’s start with caves, the most stable and persistent with the obvious thermoregulatory advantages of
roosts (Fig. 6.2). Bats are the only vertebrates to communal living, which have led to the formation
really make use of caves for permanent shelter, of vast colonies by some species. As an order, bats
although there are a few notable exceptions among are undoubtedly the most gregarious and social of
fish and amphibians. We have already looked at the mammals. Six colonies of the Brazilian free-tailed
advantages offered by caves to hibernating bats, and bat, Tadarida brasiliensis, in south-eastern United
it may have been an extension of their use of warmer States, were said to contain 54 million bats (Davis
caves in the tropics that led bats, as they spread to et al. 1962). Recent, more accurate, estimates (Betke
higher latitudes, to use caves as hibernacula. In sub- et al. 2008) suggest only 4 million, but this is never-
tropical and warm-temperate zones caves can make theless a lot of bats! It is also possible that much of
suitable maternity, mating, and hibernation roosts. the difference reflects a real decline in the bat popu-
At higher latitudes caves are rarely used by mater- lation. Several species across the world assemble in
nity colonies, although small numbers of bats, par- roosts of 100,000 to over 500,000 (for example Hip-
ticularly males, may be present all year round. In the posideros caffer in Africa, Brosset 1966; the bent-
tropics caves are widely used throughout the year winged bat, Miniopterus schreibersii, in Australia,
by many species. Caves are a scarce commodity in Dwyer and Hamilton-Smith 1965; the grey bat, Myo-
most landscapes, since they are common only in tis grisescens, in the United States, Tuttle 1976a), and
certain rock types, primarily limestone, due to its populations in some remote caves have yet to be
 ROOSTING AND FEEDING ECOLOGY 139

Figure 6.2 Cave bats. Summer roost of

Schreiber’s bent-winged bat, Miniopterus
schreibersii.

estimated. Many caves harbour a number of differ- Few non-echolocating fruit bats venture far into
ent species: Bateman and Vaughan (1974) estimated caves, but echolocating members of the genus Rou-
that four species of mormoopid in a Mexican cave settus will penetrate the darker regions and roost in
totalled 800,000 bats. At the other extreme, caves are the roof void. Emballonurids are also found in rela-
home to many species that roost in small groups and tively well-lit areas, and vesper bats will use crevices
even individually. close to the entrance. Vesper bats roost in the dark
The varied formations and microclimates found regions of the cave, and hipposiderids and rhinolo-
within a single cave can encourage colonization by phids are typical of the deepest regions. Tropical
several species, each with their own particular re- caves show only minor variations in microclimate
quirements. Temperature, humidity, and cavity size and few are uninhabited by bats. In temperate re-
are the most important factors, but precise require- gions microclimate is strongly influenced by cave
ments have been worked out for few species. A large structure. Cold air falls into descending caves, and
and structurally complex cave in the tropics will in colder regions these are rarely used by bats at any
frequently be home to more than ten species. Figure time of year. Caves with little or no air flow are also
6.3 shows the distribution of bats in a cave in Papua frequently too cold for summer use. The most com-
New Guinea, from Hill and Smith (1984). monly used caves have several entrances, significant
The cave dwelling Old World fruit bat, Dobsonia movement of air, and a variety of different tempera-
moluccensis, roosts in the large and well-lit entrance. ture regimes.
140 BATS

Hipposideros cervinus
Hipposideros calcaratus Dobsonia moluccensis
Rhinolophus euryotis
Pipistrellus angulatus
Hipposideros diadema
Emballonura nigrescens
Rhinolophus megaphyllus Aselliscus
tricuspidatus Rousettus amplexicaudatus
Miniopterus tritis
Hipposideros cervinus Miniopterus schreibersii

Hipposideros calcaratus

Figure 6.3 Distribution of bats in a cave in Papua New Guinea (from Hill and Smith 1984).

Bats can alter roost microclimate in a cave may be reduced by the presence of large
numbers of some species: few bats can tolerate the
Bats that form large maternity colonies can have a
high ammonia levels generated by a large colony of
significant impact on cave microclimate, raising the
the Brazilian free-tailed bat, Tadarida brasiliensis. The
temperature in roof cavities by as much as 10  C.
bats, and their piles of guano, have evolved their
Some species, such as the grey bat, Myotis grisescens,
own unique ecosystems, including the parasites and
in the United States (Tuttle 1976a) successfully rear
diseases of the bats themselves. The increased inci-
young in cool caves. This said, only 2.4 per cent of
dence of pests and disease is one of the two major
the 1,635 known caves in the cave-rich state of Ala-
disadvantages of community life: the other is the
bama were used by the grey bat in summer (Tuttle
possible increased competition for food in the vicin-
1979). The increased temperatures generated by one
ity of the roost. A large colony of Brazilian free-
species can benefit another. Miniopterus australis is
tailed bats must forage over a very wide area.
always found in association with larger colonies of
Cave dwelling bats are frequently very faithful to
its more common relative, Miniopterus schreibersii, at
their roost sites, occupying one cave for life and
the southern limit of its distribution in New South
sometimes for many hundreds of generations,
Wales, Australia (Dwyer 1968). In the warmer north,
whether the roost is a nursery, mating, or hiberna-
they are not usually found together. A similar
tion site. Alternatively a small number of sites may
advantage may be gained by the grey bat in the
be used by a colony, and one or more sites will be in
United Sates, where it associates with Myotis aus-
use at any one time, depending on factors such as
troriparius. Postnatal growth rates, and post-flight
cave microclimate, food availability, physiological
survival are low if the grey bat colony is too small
requirements, social interaction, and parasite
(Tuttle 1976a), so reproductive success may be
avoidance.
greater in mixed roosts.
In large numbers, bats alter their cave habitat in
Rock crevices
other ways. A large colony produces a lot of urine
and guano, leading to massive accumulations of From crevices in caves, it is a short flight to crevices
nitrogenous waste on the cave floor and a high in rock outside the cave. Rock crevices rarely offer
ammonia concentration in the air. Species diversity the same thermal stability as caves, or the same
 ROOSTING AND FEEDING ECOLOGY 141

degree of protection, but they are more common. 14–18  C were not used by the bats at this time of
Availability facilitates, and unstable microclimate year. However, groups of up to 176 bats did occupy
often necessitates, mobility in crevice-dwelling these crevices in the hot summers, when crevice
bats. Although many are faithful to particular temperatures were a stable 28–33  C: the thermal
roosts, many are not. Because of this, and the fact neutral zone of the bats. Outside the roost the diur-
that they roost in small groups, they are difficult to nal, summer temperature varied between 18  C at
find, and their roosting ecology is not well under- night and 38  C in the daytime shade (50  C on rock
stood. All manner of crevices are used in (to hu- faces). In addition to moving around within the
mans) the most inaccessible places: climbers system of crevices, bats would cluster when neces-
frequently comment (for example Kirkpatrick 2008) sary to reduce heat loss and conserve metabolic
on bats roosting in cracks on the most impressive energy. Brazilian free-tailed bats, Tadarida brasilien-
rock faces on Earth, such as the 1,000 m wall of El sis, and Yuma myotis, Myotis yumanensis, occasion-
Capitan in Yosemite, California, where the easily ally shared these roosts. Bats did not use the sites in
audible echolocation calls suggest the mastiff bat, winter, when the mean daily ambient temperature
Eumops perotis. The general trend is for bats to was as low as 6  C. Many bats make use of rock
move around within and among cracks, choosing crevice roosts in the more arid regions of the United
the temperature appropriate to their immediate States, from the 70-plus g western mastiff bat, Eu-
needs. Vaughan and O’Shea (1976) studied the mops perotis (Vaughan 1959), to the 6 g western
roosting ecology of pallid bats, Antrozous pallidus, pipistrelle, Pipistrellus hesperus (Hayward and
in the semi-arid desert of Arizona, a habitat typical Cross 1979).
of many crevice roosting bats. In the spring and In their study of pallid bats, Vaughan and O’Shea
autumn groups of up to about 20 bats occupied (1976) described a pattern of behaviour that is typi-
vertical crevices that showed large diurnal varia- cal of many species. More time was spent in the day
tions in temperature (Fig. 6.4). roost than in foraging and night roost occupation
For most of the day the temperature was below combined—about 13–16 hours in each 24 hour
20  C, and the torpid bats conserved energy. The period. The bats emerged earlier, and returned
afternoon sun raised the temperature in these cre- later, on summer nights. On their return large
vices to 32  C, more than 10  C above outside, ambi- groups of bats circled the roost entrance for up to
ent temperature, warming the bats prior to their 45 min before entering. During this time they would
evening emergence. Deep, horizontal crevices at make frequent brief landings close to the entrance

6–7 July

36 ambient
12-13 April
32
32
28 vertical crevice
deep horizontal
temperature (°C)

28 cracks
24 deep horizontal cracks
20
16 24

12
8 ambient 20

11 15 19 23 03 07 11 08 12 16 20 24 04
time (h) time (h)

Figure 6.4 Temperature variation in rock crevices used by the pallid bat, Antrozous pallidus (from Vaughan and O’Shea 1976).
142 BATS

crack, and emit regular ‘directive’ calls that attracted

other bats. This rallying behaviour has been widely
reported, although often without the directive calls,
but its function remains uncertain. Vaughan and
O’Shea suggest that the scent cues on landings
allow the bat to confirm that it is at the right
entrance. The rallying itself may enable the bats to
refresh their knowledge of the detailed topography
of the roost. Rallying has also been described in
many other species, including the common Euro-
pean pipistrelle, Pipistrellus pipistrellus (Swift 1980),
the South-east Asian bamboo bats Tylonycteris
pachypus and T. robustula, (Medway and Marshall
1972), and the New World Tadarida macrotis
(Vaughan 1959). Anecdotal observations from my
own studies suggest that it may be the time when
bats make a decision to switch roost, since radio-
tracked bats may return at dawn to the roost used
the previous day, before moving on to a different
day roost.
Crevice dwelling bats often show anatomical
adaptations to this way of life. Bats tend to be dor-
soventrally flattened for aerodynamic reasons, but
some crevice dwellers show a more extreme flatten-
ing, notably of the skull in certain molossids of the
genera Platymops and Neoplatymops (Peterson 1965).
Platymops setiger is probably the most extreme exam-
ple of this trend.

Tree roosting bats

A very large proportion of the world’s bats roost in
trees and you find them everywhere: from hanging
Figure 6.5 Daubenton’s bat, Myotis daubentonii, in a tree crevice.
between the leaves to crawling into deep crevices
(Fig. 6.5). A few species even modify their roosts,
making tents or excavating cavities. Trees with et al. (2007) is primarily concerned with bats in
hollow trunks are common in nutrient poor soils North American forests, but much of it is applica-
in the tropics, and Janzen (1976) has suggested that ble to bats in forests in other parts of the world.
this may be an adaptation to encourage use by bats
Bark roosts
and other animals, providing the trees with valu-
able nitrogen and minerals from the faeces. If Jan- Tree-bark cracks and the often large but narrow
zen is correct then bats may have an additional, spaces under exfoliating bark, on both live and
important role in tropical forest nutrient cycles. dead trees, are used by many species in all parts of
Before going into detail it is worth noting two the world, but until recently there were few detailed
references. Kunz and Lumsden (2003) have written accounts. Cervený and Bürger (1989) found small
a very readable and comprehensive review on the numbers of male Daubenton’s bat, Myotis daubento-
ecology of tree and foliage roosting in bats. If nii, and Bechstein’s bat, M. bechsteinii, under the
I don’t provide enough information on a particular bark of alder trees in spring. Small nursery colonies
topic, they probably do. A recent book by Lacki of the Indiana myotis, Myotis sodalis, roost under the
 ROOSTING AND FEEDING ECOLOGY 143

bark of the bitternut hickory, and in the winter the A very common roost site in the tropics is the large
thicker bark of the shagbark hickory is preferred cavity that forms inside the bole of many trees as the
(Humphrey et al. 1977). The tiny African vespertilio- inside rots out, assisted by fire, fungal attack, and
nid, Mimetillus moloneyi, and the larger Laephotis invertebrates. The temperature and humidity inside
wintoni, also roost behind tree bark (Kingdon these cavities can be remarkably stable. Smaller cav-
1974). The common name of the former, Moloney’s ities form where branches have fallen or where
flat-headed bat, suggests an adaptation to this way storms and age have cracked the trunk or limbs
of life. Kunz and Lumsden (2003) provide a long list (again assisted by various biological agents from
of more recent references on tree-bark roosts in vari- fungi to woodpeckers) and these are common in
ous parts of the world and discuss the nevertheless trees throughout the world. Cavities in a large tree
still limited information on roost selection, competi- can persist for hundreds of years and even long after
tion, and predation. the tree has died.
The baobab tree in Africa is a first class bat roost
Tree cavities
(Fig. 6.6), and home to some of the big carnivorous
Cavities in the trunk and branches of trees offer megadermatids such as the heart-nosed bat, Cardi-
roosts that are both more durable and thermally oderma cor, in groups of up to 80 individuals
more stable than bark crevices, and they are widely (Vaughan 1976). The large New World carnivore
used in both temperate and tropical regions (see Vampyrum spectrum also roosts in tree cavities
Bradbury 1977a, Kunz 1982 p. 7, Kunz and Lumsden (Vehrencamp et al. 1977). Many other phyllostomids
2003). Cavities vary enormously in size and position use tree cavities (Tuttle 1976c, noted 28 species), as
and in the size and location of the entrance holes. do molossids, vesper bats, and noctilionids. It is also

Figure 6.6 Bat roost in a baobab tree,

Adansonia digitata.
144 BATS

common to find more than one species in a tree high humidity and temperature. Where predation
roost: Saccopteryx bilineata with Vampyrum spectrum risk is high bats may prefer cavities with holes no
(Vehrencamp et al. 1977), and the common vampire, bigger than themselves, to exclude predators. Sev-
Desmodus rotundus, with phyllostomids such as the eral neotropical species that roost in accessible cav-
fringe-lipped bat, Trachops cirrhosus and Carollia per- ities, such as Saccopteryx bilineata, remain alert when
spicillata (Tuttle 1976c). These cavities are used as roosting, presumably to reduce predation. At the
maternity roosts, mating roosts, and night feeding level of the individual tree or stand, many bats
roosts. Even some of the smaller Old World fruit prefer large diameter trees that are taller than adja-
bats make use of tree cavities (references in Kunz cent trees and free of clutter. Larger trees are more
and Lumsden 2003). likely to have suitable cavities and the uncluttered
In temperate regions, tree cavities are widely used approach may not only facilitate access, but also
by vesper bats. They are found in trees throughout solar heating. This may also underlie the preference
the year, using cavities as nursery, mating and night of some species for trees at the forest edge. Selection
feeding roosts, and as hibernacula. Despite centuries for particular tree species, reported in a number of
of forest clearance and intensive, insensitive man- studies, is most likely a reflection of their ability to
agement, tree cavities are still important bat roosts provide suitable roosts, due to their tendency to
in Europe and under-valued until recently. Cervený form crevices and their insulating properties. Spe-
and Bürger (1989) found five species using tree holes cies with thicker bark are better insulators (Nicolai
in a small study site in Czechoslovakia: Myotis nat- 1986). Similarly, the preference for dead trees seen in
tereri, M. bechsteinii, M. daubentonii, Nyctalus noctula, some studies may also reflect the greater probability
and N. leisleri. The two Nyctalus species were found of them having suitable cavities and of being
roosting together, and both were also found with warmed by the sun. The composition and structure
M. daubentonii. Colonies of the last species num- of surrounding trees have been shown to be impor-
bered over 85 individuals, and Gaisler et al. (1979) tant in several studies, but few broad patterns
found colonies of up to 70 N. noctula. The depen- emerge, so I will leave you to consult Kunz and
dence of many European species on buildings is Lumsden (2003) for species-specific detail. A num-
probably a reflection of the long history of defores- ber of other important questions are raised by Kunz
tation. Human impact on North American and Aus- and Lumsden, such as: are tree cavities a limiting
tralian forests is more recent and less intense and it resource and is there significant competition
is likely that a larger proportion of species still rely between bats, and indeed between bats and other
heavily on trees (Kunz and Lumsden 2003). animals, for these roosts? These questions remain
Considerable recent research has focused on roost unanswered.
selection in tree-roosting bats and it is clear that The most unusual tree hole bat is the endemic
some species show strong preferences for tree spe- short-tailed bat of New Zealand, Mystacina tubercu-
cies and cavity type. For example, 85 per cent of lata. In addition to using natural holes, it is said to
Nyctalus noctula roosts in a study in The Netherlands burrow into dead and rotting kauri trees, Agthis
were in woodpecker holes (Boonman 2000). Kunz australis, (Daniel 1990). Groups of up to 200 bats
and Lumsden (2003) summarized results from 26 line up like peas in pods in these burrows, where
studies in an attempt to draw some general conclu- they can raise the temperature to 39 ºC and the
sions on roost selection at different levels from cav- humidity to 100%.
ity, through tree, to the landscape scale. Cavity size Some species roost in the open against the trunk
and shape, in addition to affecting microclimate, or branches of trees. The proboscis bat Rhynchonyc-
will certainly determine the maximum number of teris naso, a small (4–5 g) neotropical emballonurid,
roosting bats and may also influence social struc- is typical. It is frequently found on branches that
ture. Few studies compared the cavities that bats overhang water, and has a speckled yellow-grey
used with those available, but from the few that colouration that resembles lichen, making it difficult
did it can be tentatively concluded that bats show to see (Bradbury and Vehrencamp 1976a, Bradbury
a preference for dry, well-insulated spaces that have 1977a). Small colonies of 5–11 bats are typical and
 ROOSTING AND FEEDING ECOLOGY 145

they often space themselves in a vertical line 2–4 cm They roost inside the stem (culm) of the large
apart (Fig. 6.7). bamboo, Gigantochloa scortechinii (Medway and
Despite the open nature of the roost, colony size Marshall 1970, 1972), entering through the interno-
can be as great as 45, and in addition to their cryptic dal emergence holes of a chrysomelid beetle, Lasio-
colouration the bats remain quiet and still (but alert) chila goryi, that has pupated in the culm. Both
to avoid detection. Other bats known to roost on species have fleshy pads on their wrists and ankles
exposed tree trunks or branches include the related to assist the claws in gripping the inside of the culm
Saccopteryx leptura (Bradbury and Vehrencamp (Fig. 6.8). They roost in the typical upside-down
1976a) and the African butterfly bats of the genus posture at the top of the internodal cavity. Both
Glauconycteris (Kingdon 1974). Glauconycteris varie- also have flattened skulls, to allow them to enter
gata has a beautiful reticulated pattern on its wings the narrow hole left by the beetle. Roosting groups
that strongly resembles leaves. Some of the smaller are typically small (3–5) but can be as large as 20–30.
Old World fruit bats also roost on tree trunks and The beetle is (or was at that time) sufficiently abun-
have cryptic colouration,for example Nyctimene dant that there was no apparent shortage of roost
major in South-east Asia (Walker 1975). cavities and the bats moved on an almost daily
Although bamboos are not trees, they can be large basis. The other two species that habitually roost in
and relatively long-lasting, so they are best dealt bamboo, Eudiscopus denticulus in Thailand and Glis-
with in this section. Four species appear to make chropus tylopus in Borneo, have similar pads on their
regular use of bamboo, and at least four more hands and feet.
make at least occasional use. Tylonycteris pachypus
and T. robustula are the small bamboo bats of India Foliage and other temporary roosts
and South-east Asia (Fig. 6.8). Foliage roosts are usually freely available, but they
are also very short-lived. Many bats use foliage
roosts: most are solitary or roost in relatively small
groups, and most are tropical. The exceptions to the
first rule are the larger Old World fruit bats. They
have few natural predators and will roost in large
‘camps’, frequently of several thousand individuals,
in the canopies of large trees (for example Nelson
1965, Kingdon 1974). Camps are frequently noisy,
smelly, and very conspicuous, and can be made
more conspicuous by the defoliation often caused
by these large colonies. Defoliation may or may not
be incidental, but can have consequences: it exposes
the bats to sun, wind, and rain, but may allow them
to spot predators more easily. Camps can be surpris-
ingly persistent: some have been in use for almost
100 years. Roost selection is poorly understood, but
important criteria include proximity to foraging
sites, high canopy, the presence of emergent trees,
and in dry regions, the presence of water (for exam-
ple Tidemann et al. 1999), but general patterns are
hard to find. One reason for this may be that human
disturbance and destruction of habitat has disrupted
normal roost selection behaviour.
Most foliage roosting bats are less conspicuous,
particularly smaller species or those roosting in
Figure 6.7 Rhynchonycteris naso, the proboscis bat, on a tree bole. smaller groups. The variety of roosting locations is
146 BATS

Figure 6.8 Tylonycteris pachypus, the bamboo bat,

roosting in a bamboo culm.

enormously varied, with bats taking advantage of Exceptions to the tropical rule for foliage roosting
the rich diversity of form of both living and dead bats are members of the genus Lasiurus. In North
plants. Many foliage roosters have cryptic coloura- America, the hoary bat, Lasiurus cinereus, roosts in
tion (Fig. 6.9) and adopt distinct postures to avoid trees and has a long, thick coat of grizzly fur (that is
detection. frosted, like the grizzly bear), which is clearly cryp-
Camouflage is achieved by several means. Multi- tic against many backgrounds (Fig. 6.11). Lasiurus
coloured fur, for example in the vespertilionid genus species use both conifers and hardwoods, typically
Kerivoula (Walker 1975), can resemble moss, lichen, roosting under denser leaf clumps. Some tropical
dead foliage, and other backgrounds. Specific pat- foliage roosters also have long, thick fur, presum-
terns may match specific substrates or mimic leaves ably for insulation, for example Lavia frons, the yel-
or fruit. It has been suggested that the light coloured low-winged bat of Africa. Externally roosting bats
heads and shoulders of pteropodids act as counter- also tend to have good vision, to keep an eye open
shading. The white lines common on the body and for predators, although in some cases it is also for
head of many foliage roosting phyllostomids may visual hunting. This is seen particularly well in the
disrupt the outline of the bats against a background phyllostomids. Some phyllostomids have transpar-
of leaf veins and shafts of light penetrating the ent windows in the handwing, so that they can see
foliage (Fig. 6.10). The white coat of the white bat, what is going on even when the wings are wrapped
Ectophylla alba, takes on a green hue when lit around the head and body (Vaughan 1970). One
through foliage. species, the wrinkle-faced bat, Centurio senex, also
 ROOSTING AND FEEDING ECOLOGY 147

Figure 6.9 The hoary bat, Lasiurus cinereus, has fur of

several colours and a grizzled (frosted) appearance, which
may help to camouflage it when roosting.

Figure 6.10 The phyllostomid Vampyrodes caraccioli.

The white lines may serve to break up the bats’ outline and
help to camouflage them.

has a translucent chinfold that can be lowered over and Wilson 1974). Two recently discovered species
its face (Goodwin and Greenhall 1961). (T. devivoi and T. lavali) have yet to be studied. One
In the Thyropteridae, the two well-known species of the two species of sucker-footed bat (Myzopoda
of neotropical disk-winged bats (Thyoptera tricolour aurita, family: Myzopodidae) of Madagascar has
and T. discifera), roost inside furled leaves of Helico- been found in rolled leaves of the palm Ravenala.
nia and similar plants in Central America (Findley All species have very well developed adhesive
148 BATS

Figure 6.11 The hoary bat, Lasiurus cinereus, frequently roosts in the
open and has thick insulating fur which covers the tail membrane and parts
of the wings.

disks on their wrists and feet that can grip the

smooth inside surface of the leaves (Fig. 6.12). This
adaptation may be so specialized that it restricts the
bats to this sort of roost, and Findley and Wilson
suggest that competition for unfurled leaves may
limit the distribution and numbers of Thyroptera
tricolor. These roosts are abandoned as the leaves
begin to unfurl, so bats move frequently. Similar,
but usually less well-developed pads are found on
at least 11 species of vespertilionid bat (see Kunz
1982 for references). One species in Gabon, Myotis
bocagei, has been found roosting inside the flowers of
the water arum, but it is more commonly found in
banana leaves (Brosset 1976).

Tent-making bats
Figure 6.12 (a) The New World disk-winged bat, Thyroptera tricolour,
The final word in foliage roosting behaviour is roosting in a rolled leaf, and (b) the African banana bat, Pipistrellus
found in those species that make their own roost nanus, which lacks disks, in a tightly rolled leaf.
 ROOSTING AND FEEDING ECOLOGY 149

sites. Tent-making, or at least tent-use, has evolved protection from the weather and predators. Fig.
independently in three families of bats, in the neo- 6.13 shows the phyllostomid Ectophylla alba in a tent.
tropics, India/Indonesia, and the Philippines. At Palm and Heliconia are the most commonly used
least fifteen phyllostomids, three pteropodids, and plants, but there are some unusual variations: for
one vespertilionid are believed to modify the leaves example in India the pteropodid Cynopterus sphinx,
under which they roost to make tents (for example in addition to making more typical tents, makes a
Kunz 1982, Rickart et al. 1989, Brooke 1990, Foster structure resembling an upside down bird’s nest
1992, Charles-Dominique 1993, Balasingh et al. 1995), from the stems of the creeper Vernonia scandens
although the evidence is often circumstantial. Many (Balasingh et al. 1995). The extent to which tent-
use the same basic technique: the veins or midrib of making is innate or learned is unknown, and proba-
large, broadleaved, and often pinnate leaves are par- bly varies from species to species. In a study of the
tially chewed through, causing the leaf to collapse. white bat, Ectophylla alba, in Costa Rica (Brooke
The partially enclosed tent thus formed gives 1990) 57 per cent of all tents were found to be

Figure 6.13 The white bat, Ectophylla

alba and a Heliconia tent.
150 BATS

made from two species of Heliconia, but at least 26 base for entry and exit. Commonly used nests
other plants species were used for the 361 tents include those of weaverbirds, wrens, and sunbirds
located. Despite this diversity, all showed the same (Skinner and Smithers 1990). Cup- and bottle-shaped
pattern of perforations on similar leaves, suggesting nests of swallows and swifts are used by bats in
that the construction technique is largely innate. the Old and New Worlds (references in Kunz and
Tent-making in Uroderma bilobatum is more flexible Lumsden 2003).
and it uses several plants of several leaf forms. Most
tents made by this species use single leaves, but Roosts in arboreal nests of ants and termites
Choe (1994) has found U. bilobatum making tents At least four phyllostomids, one vespertilionid, and
from up to 14 leaves of young, unbranched Coccoloba one pteropodid have been found roosting, in small
manzanillensis. Bats start by collapsing the bottom groups, in excavated ant or termite nests (Kunz and
leaf and work up the plant over several days. Kunz Lumsden 2003). It has yet to be confirmed that the
and Lumsden (2003) summarize the different archi- bats excavated the holes, since several birds are
tectural styles of bat tents. Not all modify leaves: known to excavate in old nests. Both old and active
Balionycteris maculata excavates the root masses of nests are used.
epiphytic plants.
Tent construction has to be time consuming and Man-made roosts
energetically expensive, so the cost must be offset by By now it will be clear that many bats are very
some clear adaptive advantage. Can a single bat adaptable and opportunistic roosters. This adapt-
make a tent in one night, or is construction a group ability has enabled bats to exploit a wide range
venture? Only Cynopterus sphinx has been observed of man-made structures throughout the world
making a tent, and in this case the harem male did (Fig. 6.14). Many are obvious analogues of natural
the work. A tent may be used for as little as 1–2 sites: mines, tunnels, tombs, and similar structures
consecutive nights, but most tent bats make several are used by cave bats; roof cavities in wooden
tents, and will utilize many of them for several buildings substitute for tree cavities, as do the
weeks or months. They are used as day roosts by numerous smaller cavities found in buildings.
nursery groups, harems and solitary bats, and at Crevice dwellers occupy cracks in the stonework
night to consume fruit in safety (Brooke 1990, of buildings and bridges, or the expansion joints of
Charles-Dominique 1993). Roosting groups range more modern concrete and steel buildings and
from 1–80 depending upon species, and there is evi- bridges. Some species have become so well-adapted
dence for considerable year to year fidelity to specific to man-made sites that they are now rarely found in
locations, despite the short life of the tents them- natural roosts: the serotine, Eptesicus serotinus, and
selves. Tent bats are often highly localized, being pipistrelles (Pipistrellus pipistrellus and P. pygmaeus)
dependent upon the number and distribution of in Britain, the big (Eptesicus fuscus) and little (Myotis
suitable plants. The dependence of foraging beha- lucifugus) brown bats in North America (Barbour
viour, social patterns, and mating systems of tent and Davis 1969), Molossus molossus and Myotis ni-
bats on the distribution of tent plants and food plants gricans in the neotropics (Greenhall and Stell 1960,
is an interesting topic beyond the scope of this book, Wilson 1971), and Taphozous melanopogan and Mega-
but covered by Kunz and Lumsden (2003). derma lyra in India (Brosset 1962).

Roosts in birds’ nests Behavioural ecology of roosting

Several species are known to make use of old bird
nests. Kerivoula and Murina species in the Paleaeo- I have looked at where bats roost but have so far
tropics roost in suspended bird nests. Most are oc- said relatively little about the behaviour of roosting
cupied by single bats, but although up to eight bats bats beyond the specifics of roost selection. It’s time
have been found, there was no evidence of use as a to look at behaviour and the evolutionary driving
nursery (Schultz 1997, 2000). Some appear to have forces behind it. As already stated, a characteristic
been modified, with the bats creating a hole in the common to most bats is their extreme sociability.
 ROOSTING AND FEEDING ECOLOGY 151

Figure 6.14 Natterer’s bat, Myotis

nattereri, roosting in the roof of a building.

Although colony size typically varies from a few to a may be to avoid those predators that have located
few hundred bats, colonies of a few species form the their roost (Fenton et al. 1994). Emergence time from
largest known mammalian aggregations. Why are the roost is variable, from before sunset to several
bats colonial? There are three main reasons, dis- hours after, and is related to factors that are likely
cussed briefly below. Many of the topics are covered to determine vulnerability to predation, such as
to varying degrees and in different contexts in other flight performance, foraging habitat, and foraging
parts of the book, but it is useful to bring them strategy (Jones and Rydell 1994). For example, fast-
together here under the umbrella of roosting flying aerial hawkers emerge much earlier than
behaviour. slow-flying and gleaning species. Emergence time
Reduced predation risk. By living in groups bats also varies with colony size within a species (Fenton
gain protection from predators, both in the roost and et al. 1994). Large colonies emerge earlier than small
on emergence. The individual risk of being taken by colonies, probably because of the lower risk to in-
a predator decreases as colony size increases. In a dividuals in large colonies from diurnal and crepus-
colony of thousands, the risk becomes vanishingly cular predators.
small. Bats adopt a number of strategies that further Reduced thermoregulation costs. Colony forma-
reduce the risk of predation on emergence. For tion also leads to thermoregulatory benefits, because
example, many species emerge from the roost in bats that cluster together reduce their exposed sur-
clusters. This temporal clustering probably makes face area and therefore heat loss (Speakman and
it difficult for predators to fix on a single bat, enhanc- Thomas 2003). This is particularly important to
ing the dilution effect (for example Speakman et al. bats since they have large surface area:volume
1999). One reason bats frequently change roost ratios. Furthermore, living together gives bats the
152 BATS

ability to alter their own environmental conditions spatially and temporally (Dwyer 1966). Nursery co-
by increasing the ambient temperature of their roost lonies were predominantly all female, and what
space. Clustering can also limit water loss (Thomas Dwyer labelled ‘adult’ colonies comprised both
and Cloutier 1992). sexes, some of which were mating roosts where
Social benefits of cooperation and information transient females visited resident males. Juveniles
transfer. Colonial behaviour can facilitate information and yearlings also formed colonies, separate from
transfer about good foraging (for example Wilkinson adults. All colony types varied seasonally in size
1992b) and roost (Kerth and Reckardt 2003) sites. Sev- and composition in relation to the reproductive
eral studies provide good evidence for information cycle. European colonies of Schreiber’s bat show
transfer and these are discussed later. Coloniality will similar patterns, widespread in other insectivorous
also facilitate the evolution of group foraging strate- families such as Rhinolophidae, Hipposideridae and
gies that may increase an individual’s foraging Emballonuridae, and in the Old World fruit bats
efficiency. Again, these will be discussed below. Famil- (Pteropodidae).
iarity with particular roosts, habitats, and roost inha- Despite persecution and habitat loss, some Old
bitants can bring benefits through improved foraging World fruit bat camps can still number tens of
efficiency and reduced predation. thousands and sometimes over 100,000 individuals
Some of the benefits of colonial living depend (Kunz and Lumsden 2003). There is structure in the
upon colony stability. Transient groupings can gain colonies of several of the species studied, for exam-
from reduced thermoregulatory costs or reduced ple in the grey-headed flying fox, Pteropus polioce-
predation risk, but are less likely to benefit from phalus (Nelson 1965). This species often forms
cooperative strategies or an improved knowledge sexually segregated camps prior to parturition,
of the roost and its surroundings. It is becoming with males and females occupying different trees
clear that although many bats gain from quite ran- or gathering at different levels in the canopy of a
dom associations, structure and social interaction single tree. As the young are weaned males and
play a major role in many species and can give females mix, the males form small territories within
additional benefits. the camps (maintained by aggression and scent
marking) and attract one or more females. After
Bats that form large colonies mating the females leave to form large single sex
groups and the males quickly establish male-only
Species that form very large colonies are found pri- groups. In winter large camps may break up into
marily among cave-roosting, insectivorous species smaller groups. Similar patterns have been observed
and tree-roosting Old World fruit bats. The largest in other species.
roosts are those of the Brazilian free-tailed bat,
Tadarida brasiliensis, in North, South, and Central
America. Nursery colonies in caves in Texas, as we
Bats that form smaller colonies
noted earlier, can have bats counted in millions. The vast majority of species form colonies of a few
Many of the males do not migrate north with the individuals to a few hundred. In most temperate
females, but stay in the southern part of their range species, small colonies are made up largely or exclu-
until the females return, when mixed colonies form sively of females, the males living apart, singly or in
for mating. In common with most bats, there small groups. One of the best studied examples is
appears to be no structure or hierarchy within the the greater horseshoe bat, Rhinolophus ferrumequi-
roost. num (Ransome 1990). These colonies often comprise
Many other insectivorous bats form large colonies only a small number of matrilines. The brown long-
in caves. The Schreiber’s bent-winged bat (Miniopterus eared bat, Plecotus auritus, forms small mixed-sex
schreibersii) species complex is widely distributed in colonies, both sexes showing marked natal philopa-
both northern and southern hemispheres. Colonies try (Burland et al. 1999, Entwistle et al. 2000, Burland
can contain tens of thousands of individuals of et al. 2001). Similarly, in Bechstein’s bat, Myotis bech-
both sexes, with colony composition varying both steinii, nursery colonies of 20–40 females are closed,
 ROOSTING AND FEEDING ECOLOGY 153

with just one or two matrilines (Kerth et al. 2000). In constructs tents in the fruit/flower clusters of the
contrast to long-eared bats, male Bechstein’s bats kitul palm tree. They form harems, comprising a
roost singly, away from the females. Colonies of single male with 1–37 females and dependent
females frequently fragment, with considerable mix- young. Average group size rises from 6 in the wet
ing in the sub-groups, but associations are strong season to 14 in the dry season and although the
between lactating females, possibly for improved overall harem structure is maintained throughout
social thermoregulation. Colony fragmentation and the year, individuals and groups of bats switch
fusion are common to many species, and will be roosts regularly (Storz et al. 2000). The yellow-
discussed in detail later in the book. Fragmentation eared bat, Uroderma bilobatum, is another tent-
and roost switching may be to reduce predation, as building bat. A study in Costa Rica focused on a
described above, to reduce parasite burden (Lewis colony of about 30 bats, which moved into a coconut
1996), to find a roost with a better microclimate grove prior to parturition. They divided themselves
(Kerth et al. 2001b), or to move nearer to better amongst an average of nine tents. Two of these tents
foraging sites. Roost switching is not confined to were nursery roosts containing the majority of the
bats using small or semi-permanent roosts, since females and occasionally males, but most of the
many cave-roosting bats are also very mobile. In males were usually found roosting alone or in
northern Greece, female long-fingered bats, Myotis small groups. There were regular movements of
capaccinii, regularly fly between nursery roosts in bats both into and out of the colony during the
disused mines and caves that are 30 km apart in a study (Lewis 1992). Ectophylla alba is a tent-making
straight line, and up to 50 km via the rivers along bat that roosts in Heliconia leaves (Fig. 6.13). Post-
which they travel (Papadatou et al. 2008). parturition group composition changes from mixed
In many tropical and subtropical species, for groups of four to eight bats to harems and male only
example in many bats of the family Phyllostomidae, roosts (Brooke 1990). Even temperate species may
a single male roosts with a harem of females. In the roost in very small groups or singly. The females of
Jamaican fruit-eating bat, Artibeus jamaicencis, colo- Lasiurus species in North America frequently roost
nies of over 200 bats roost in caves. A dominant alone or with their offspring, hanging from the
male defends a group of between 4 and 18 females. foliage of broad-leaved and coniferous trees (Kunz
In groups with over 14 females there may also be a and Lumsden 2003). Carnivorous species roost in
subordinate male, who, although not appearing to small colonies or family groups. The Australian
actively defend the harem, seems to deter take-over ghost bat, Macroderma gigas, is a large insectivore
bids from other males and will take priority as the and carnivore that can form roosts of over 400
next dominant male (Ortega and Arita 2000, 2002). bats, but rather less than 100 is more typical. The
This species will also roost in tree cavities, when a carnivorous Vampyrum spectrum, on the other hand,
single male will defend a small group of females. A typically roosts in pairs. It is clear from this that
study of a colony of greater sac-winged bat, Saccop- colony size and composition are often influenced
teryx bilineata, in Costa Rica showed that single by mating strategy, and this is discussed further in
males defended a particular territory in the roof of Chapter 5.
a building even if there were no females present.
Harem size was between one and five females and
Sexual segregation in the roost
non-territory holding males roosted close to the har-
ems. In this species, juvenile females disperse and Segregation between the sexes occurs at various
the male juveniles remain with their natal colony spatial levels, from the roost through home range
(Voigt and Streich 2003). to a broader geographical scale. Let’s start at the
Bats that use ephemeral roosts typically form the smallest scale and ask: why do males and females
smallest and most nomadic colonies. Good exam- so often roost apart during the nursery season?
ples include the neotropical disk-winged bats Surprisingly, few studies address this question
and palaeotropical sucker-footed bats described directly and there is little direct evidence in support
above. Cynopterus sphinx, the short-nosed fruit bat, of most of the explanations put forward. In
154 BATS

temperate bats, perhaps the most plausible reason for competition or exclusion at foraging sites.
and that supported by the greatest body of evidence, Assuming exclusion at roost and/or foraging sites
relates to physiological and energetic differences does occur in some cases, how do the females exert
between the sexes. Males and females of temperate their apparent dominance? Differences in foraging
species often require a different thermal environ- behaviour may also be related to sex-based differ-
ment, forcing them to roost separately. Although ences in foraging strategy or diet (discussed later in
there is some variation between species and during the chapter), but this topic has not been explored in
the stages of pregnancy and lactation, females gen- any detail.
erally remain homeothermic most of the time to Segregation may also have evolved to avoid
maximize foetal growth rate and possibly also milk aggression, either between males or between males
production. Males, free of these burdens, can make and the females and their offspring, but again direct
use of torpor more often than females in the sum- evidence is lacking and there is little evidence for
mer, and to maximize energy savings may choose aggression amongst individuals in bat roosts.
cool roosts (for example Hamilton and Barclay 1994, Finally, limitations in the size and number of suit-
Lausen and Barclay 2002, 2003). able roosts, particularly for nurseries, may drive
Another reason for segregation may be to reduce segregation in some species, but this too requires
competition for food in the vicinity of the roost study. I will come back to segregation on a land-
(Kunz 1973). A large nursery colony consumes a scape scale later in the chapter, after reviewing feed-
considerable amount of food each night and females ing ecology, since the two cannot be discussed in
may be better able to provide for their young with- isolation.
out competition from males. Foraging is generally
more efficient and infant mortality lower if the fe-
Feeding ecology
males can feed close to the roost (Tuttle 1979). This
explanation assumes that if males and females share The feeding habits of bats are almost as varied as
the same roost they will share foraging areas. Male those of the mammals as a whole, and this dietary
and female Myotis daubentonii that shared the same diversity is responsible for much of the morphologi-
roost were found to forage in the same home range cal, physiological, and ecological diversity seen in
(Senior et al. 2005), but non-colony males fed in less bats. Bats feed on insects and other arthropods,
favourable habitat. However, they may not be ac- including freshwater and marine species, mammals,
tively excluded from the colony home range, but birds, reptiles, amphibians, fish, blood, carrion,
may choose to forage elsewhere to avoid competi- fruit, flowers, nectar, pollen, and to a limited extent
tion. Plecotus auritus also forms mixed sex roosts and seeds and foliage.
the males forage on the periphery of the home range A comprehensive review of the feeding ecology of
(Entwistle et al. 1996). Wilkinson and Barclay (1997) bats would need a large book in itself. I am going to
showed that male big brown bats, Eptesicus fuscus, go through the list, briefly cover the basics, and
foraged over larger areas than females, spending highlight some features I find particularly interest-
more time away from the main feeding area. ing or informative.
Again, in both cases, the males may, despite the
longer commuting distances, get more food by feed-
Insectivory
ing away from the core area where competition with
the females would be more intense. In contrast, male Although insects are the dominant component of
and female Myotis bechsteinii that do not share roosts most bats’ diet, spiders, scorpions, crustacea, and
do have overlapping foraging ranges (Kerth and other arthropods are eaten (Fig. 6.15). For simplicity
Morf 2004). Males commuted shorter distances and I use the term insectivorous to describe all of those
occupied smaller foraging areas than females, as species that feed on insects and other arthropods.
seen in male and female Myotis daubentonii that do Around 70 per cent of all bats are insectivorous:
share a roost (Senior et al. 2005). Despite widespread arthropods are an abundant and widespread food,
segregation in the roost, there is no strong evidence and it’s likely that their diversification was a
 ROOSTING AND FEEDING ECOLOGY 155

Figure 6.15 Natterer’s bat, Myotis nattereri, taking a spider from its web.

primary driving force behind the evolution and (Barclay and Brigham 1991). Size is important in
diversification of bats. The vast majority of insectiv- other ways: it is a generally accepted ecological
orous bats are small, and catch their food on the rule that the size of the predator is related to the
wing. Their small size gives them the manoeuvrabil- size of its prey. Optimal foraging theory predicts
ity and agility necessary to catch flying insects de- that predators, and in particular carnivores (includ-
tected by their short-range echolocation system ing insectivores), should, under certain conditions,
156 BATS

select prey that are most profitable in terms of necessarily optimal foraging—and opportunistic
energy intake per unit handling time. An insect feeding can be the optimal strategy. A selective bat
that is too big may take a disproportionately long (a specialist) will in principle take only the most
time to catch, subdue, and eat, and this may incur a profitable prey, but it may spend a lot of time and
considerable energy cost in itself. If the insects are energy looking for them. An opportunist (generalist)
too small, it may not be possible to catch and eat will spend little time and energy looking for partic-
sufficient to meet daily energy requirements. The ular prey, and many may give it little energy in
number and size of teeth, the size of the jaw, the return, but its rate of food intake may still be high.
size of the cranial crest where the chewing muscles An optimal forager will adjust the time it spends
are attached, and other morphological features, can using each strategy to maximize its net energy
all be related to diet (for example Freeman 1981, intake (MacArthur and Pianka 1966). Let’s assume
Findley and Black 1983). Bats that eat hard-cased the bat is feeding and taking profitable prey, when it
insects like beetles have strong jawbones, large jaw encounters a slightly less profitable insect. Should it
muscles, and a few, large teeth. expand its diet and eat it? The answer is yes, if it
Several species eat scorpions: members of the increases its overall net energy intake rate—more
Nycteridae—the slit-faced bats—in the Old World specifically, if the profitability of that item, (the ith)
tropics, and the pallid bat, Antrozous pallidus, in is greater than the expected rate of energy intake if it
North and Central America. The fish-eating bats, ignores this item and continues to search for more
Noctilio leporinus and Myotis vivesi take crustacea; profitable prey. The profitability of its new item is:
N. leporinus in Puerto Rico are known to eat fiddler
Ei
crabs (Brooke 1994) and in this population insects
hi
were more important in the wet season, fish and
crustacea in the dry season. where Ei is its energy content, and hi its handling
time. The rate of energy intake on the present diet is:
Optimal foraging E
In Chapter 3 I discussed the subject of prey selection
ðs þ hÞ
in relation to the perceptual abilities of echolocation.
I want to raise it again, in the context of optimal where E is the average energy content of the present
foraging. There has been a long standing debate diet, h the average handling time, and s the average
among bat biologists as to whether bats can indeed search time. The bat should therefore take this ith
select their prey and forage optimally: studies can be item when:
cited in support of both selective and opportunistic
Ei E
foraging, but the case for optimal foraging is harder 
hi ðs þ hÞ
to establish. Fenton (1990) and Jones (1990) provide
interesting insights into the debate. The problem,
Searching and handling—generalist or specialist?
like many in ecology, is a complex one, with many
factors likely to influence the foraging strategy There are two major components to any foraging
adopted by a bat. Let’s review some of the more strategy: searching and handling. If search time
important, and then see if the observed behaviour is long in relation to handling (s>h), then the bat
of bats can be explained with reference to them. should be a generalist or opportunist, because
Unfortunately, few studies really address the ques- Ei/hi will be large for a wide range of insects, and
tion of optimal foraging in bats. Far more questions E=ðs þ hÞ small. Once it has found a prey item, it
are asked than answered and there has been surpris- might as well eat it. All bats that feed on the wing
ingly little recent research, probably not unrelated to are relatively small, and encounter prey quite fre-
the difficulties involved in such work. quently (1–20 each minute, with a capture success
Optimal foraging can involve both selective and rate of about 40 per cent, reviewed by Fenton 1990).
opportunistic foraging, a point sometimes missed in Prey are predominantly small, and handling times
the literature on bats—selective feeding is not will be short, so theory predicts that these bats
 ROOSTING AND FEEDING ECOLOGY 157

will be generalists. Published evidence is contradic- moving, stinging scorpion that must be gleaned
tory (see Jones 1990, Fenton 1990), with studies sug- from a cluttered ground habitat.
gesting both generalist and specialist strategies,
Foraging style
even for the same species (for example Myotis luci-
fugus, Anthony and Kunz 1977, Belwood and Fen- Several styles have been identified in bats, but for
ton 1976). However, the generalist strategy does the purpose of this discussion, I’ll consider two
appear to be that most commonly adopted. extreme categories: aerial hawking and flycatching
If handing time is long in relation to search time, (perch-hunting). In the former bats detect, pursue,
then it may be energetically more advantageous to and eat their prey on the wing; in the latter prey may
reject some prey items and search again. This is be detected from a perch, pursued, caught, and then
because when s is small, the bat must effectively eaten back at the perch. Jones and Rayner (1989)
maximize E=h: that is keep eating only the most compared the feeding behaviour of the closely
profitable prey. It has been suggested that this spe- related and morphologically similar 7 g lesser horse-
cialist strategy is used by the larger aerial insecti- shoe bat (Rhinolophus hipposideros), and 25 g greater
vores. For example, the hoary bat, Lasiurus cinereus, horseshoe bat (R. ferrumequinum). Although both
takes only large prey, when small insects are known had very flexible foraging styles, the lesser horse-
to be available (Barclay 1985). This has been noted shoe bat was predominantly an aerial hawker, and
for a number of species (see Barclay and Brigham was never observed flycatching. Greater horseshoe
1991), although again exceptions have been noted bats often spent much of their time flycatching.
(Rydell 1986). It may be true that this is selection for This is a common trend among larger bats—large
profitable prey, but Barclay and Brigham (1991) nycterids (slit-faced bats) and megadermatids (for
present an alternative and persuasive hypothesis. example the yellow-winged bat, Lavia frons,
They note that insectivorous bats are small, even Vaughan and Vaughan 1986) are typically flycatch-
by comparison with insectivorous birds, suggesting ers. As I said above, large animals take large prey,
that aerodynamic constraints alone (small size con- and it may be the difficulty of handling large prey
fers increased manoeuvrability and agility) do not in flight that makes these bats return to a perch.
explain their small size. They argue that because Why should they hunt from a perch in the first
bats use echolocation, a short-range orientation sys- place? The cost of flight is around 20 times the
tem, prey are detected only when the bats are very resting metabolic rate for an insectivorous bat, the
close, giving them little time to change flight path. cost of flycatching may be only 3 times. Large prey
This makes adaptations for manoeuvrability and items generally occur at lower densities, so search
agility even more important: hence the strong selec- times and flight costs will be high. Flycatching—
tive pressure for small size. Bigger bats are less sitting and waiting for large prey, is likely to be a
manoeuvrable and less agile, since they have higher cost-effective strategy. Foraging in clutter, as many
wing loadings, and hence faster flight speeds. They of these bats do, demands low aspect ratio wings
compensate for this by having lower frequency and low wing loading for slow manoeuvrable
echolocation pulses that are attenuated less in air, flight—characteristics that make flight even more
and can thus detect prey at greater distances. How- expensive. This large difference in energy cost
ever, the penalty for this is a decrease in resolution; between the two strategies must be a major factor
they cannot detect small insects, and this may be the in determining which style is adopted, as the density
reason why many large bats (but by no means all) and type of prey being sought changes through the
only take large prey—selection may not enter into night and with the seasons. There are a number of
the equation. descriptive accounts of bats changing foraging style
The relative lengths of searching and handling during the night.
times will depend upon the habitat, the behaviour
Prey density
of the prey, and the nature of the prey itself. Com-
pare for example a slow-flying, small, soft moth in Search time, s, will increase as prey density decreases,
an uncluttered airspace with a large, hard, fast- and this should lead to the adoption of a generalist
158 BATS

strategy. Some of the apparent contradictions seen in just how important birds are in their diet. The very
the literature may be resolved if future studies take high wing loading and aspect ratio suggest birds are
this into account. taken in flight, which would indeed make it almost,
but not quite, unique: the great evening bat, with the
Perception and pursuit of prey
wonderful scientific name Ia io, probably also
As we saw in Chapter 2, bats differ in their flight catches birds in flight (Thabah et al. 2007). Vam-
performance, and in Chapter 3 we saw that they pyrum spectrum and Macroderma gigas are the biggest
differed in their perceptual abilities in terms of the non-pteropodid bats, with a wingspan of almost 1
size of prey they can detect, their ability to function m, and a mass of 130–180 g (Fig. 6.16). A feature
in cluttered environments and the detailed informa- common to all except N. lasiopterus is the use of low
tion that echolocation gives them about their prey. intensity, broadband echolocation calls, and/or
All of these factors must be taken into account when prey-generated sound cues when hunting. None
trying to explain foraging strategy. appear to be exclusive carnivores, and a broad diet
characterizes most.
The predator as prey
Megadermatids will eat arthropods, other bats,
Even if all of the above could be quantified, a bat small rodents, birds, lizards, and frogs. Vampyrum
may not conform to the rules of optimal foraging— spectrum, Trachops cirrhosus, and Chrotopterus auritus
its behaviour may be influenced by other factors, have similar diets. The foraging behaviour of Vam-
such as avoiding being eaten itself. pyrum spectrum, studied by Vehrencamp et al. (1977)
in Costa Rica, is probably typical. They roosted in
small groups in tree holes, and left and returned
Carnivory
singly during the night. They probably therefore
A carnivorous bat is generally thought of as one that foraged alone. Foraging time was very variable,
takes small vertebrates (excluding fish) as a signifi- and on some nights they did not forage at all. This
cant component of its diet. Only about 11 species are is consistent with the high energy content of the
confirmed carnivores: four species of false vampire predominantly avian prey of this bat—one 75 g
(Megadermatidae)—Megaderma lyra and M. spasma bird every second or third night may meet its energy
(India and South-east Asia), Cardioderma cor (East
Africa,) and Macroderma gigas (Australia); four phyl-
lostomids—Trachops cirrhosis, Vampyrum spectrum,
Phyllostomus hastatus, and Chrotopterus auritus (Cen-
tral and South America); one nycterid—Nycteris
grandis (Africa); and two vespertilionids—Antrozous
pallidus (southern United States and Central Amer-
ica) and Nyctalus lasiopterus (Europe). None are
small bats (all weigh more than 20 g) and all but
two have low aspect ratio wings and low wing
loadings—useful attributes for catching and carry-
ing large (often ground dwelling) prey. The excep-
tions are Phyllostomus hastatus and Nyctalus
lasiopterus. Although P. hastatus does take verte-
brates, its higher aspect ratio and wing loading
have led Norberg and Fenton (1988) to predict that
they do not make up a significant part of this spe-
cies’ diet. Similarly, N. lasiopterus has been shown to
eat passerine birds, particularly during nocturnal
migration (Ibáñez et al. 2001, Popa-Lisseanu et al.
2007) and there was some initial controversy over Figure 6.16 Linnaeus’ false vampire, Vampyrum spectrum.
 ROOSTING AND FEEDING ECOLOGY 159

requirements. Food was sometimes taken back to insects. Two species are confirmed fish-eaters—Nocti-
the roost, possibly for consumption, for a mate lio leporinus (60 g, Noctilionidae) and Myotis vivesi (25
guarding its pup, or even for feeding the pup. The g, Vespertilionidae), from the tropics and subtropics
only clue is the observation that a captive male of the Americas. This is another good example of
provided food for a female. The bats specialized in convergent evolution. Both species have long legs
birds that roosted communally and/or had strong and huge feet for effective fishing and long, sharp
body odours. Vehrencamp et al. (1977) make the claws for gaffing fish. They have high aspect ratio
interesting comment that these two characteristics wings for efficient flight over water: flight is free
have been put forward as anti-predation mechan- from clutter, making long wings practical, and lift is
isms. They clearly do not work against the bats, but gained by flying close to the surface. Finally, both
may be more effective against other predators. have a low wing loading, an adaptation for slow
Other studies suggest a more varied diet, and a flight, and for carrying large prey (Fig. 6.17). The
very broad diet has been reported for Nycteris smaller noctilionid, N. albiventris (30 g), although pre-
grandis (see Norberg and Fenton 1988 for references) dominantly insectivorous, also takes fish. Other insec-
and Megaderma lyra (Audet et al. 1991). tivorous trawlers/gaffers are known to take fish to
Carnivory almost certainly evolved from insectiv- varying extents, for example Myotis capaccinii (10 g) in
ory and most carnivorous bats still take large num- Europe, as do the large gleaners Megaderma lyra and
bers of arthropods: little modification of the teeth Nycteris grandis.
was necessary to deal with the new prey. Relatively N. leporinus has a strong, long calcar that can be
large size may be a prerequisite of carnivory, and folded forward to lift the tail membrane clear of the
also a consequence permitted by the rich diet. water as the bat swoops to gaff fish. They search for
fish breaking the water surface using characteristic
echolocation calls in high (20–50 cm above water) or
Piscivory
low (4–10 cm) search flights, dipping the feet in only
Piscivory is really a specialized form of carnivory, and in an attempt to catch located fish. Alternatively
probably evolved in trawling/gaffing insectivorous they can trawl randomly in areas of high fish activ-
bats. Like carnivores, fish-eating species do still take ity, dragging their feet through the water in up to

Figure 6.17 Noctilio leporinus, the fisherman bat.

160 BATS

10 m straight lines without targeting a specific fish

(Schnitzler et al. 1994).

Sanguivory
This is the habit that gave bats their undeserved bad
image. Ironically, only three species feed on blood.
They all belong to the Desmodontinae a sub-family
of the large and diverse Phyllostomidae. By far the
most abundant is the common vampire, Desmodus
rotundus.
It is widespread in the tropical and subtropical
areas of the New World, and extends just into the
southern United States. The other two, the white
winged vampire (Diaemus youngi) and the hairy-
legged vampire (Diphylla ecaudata, Fig. 6.18), are
much less common and much less widespread.
Diphylla ecaudata apparently has a strong preference
for bird blood. Desmodus on the other hand has a
preference for large mammals, and the introduction
of domestic horses, cattle, and pigs, has probably led
to a dramatic increase in its numbers over the last
300 years. The booming human population also
provides an occasional food source in some areas.
Vampires roost in colonies of up to 100 or so
individuals in caves or hollow trees. They locate
their prey by a combination of smell, passive
hearing, echolocation, and possibly heat. They are
remarkably agile on the ground, and will often land
near their prey, run up to it and jump onto the foot Figure 6.18 Diphylla ecaudata, the hairy-legged vampire.
or leg. They have heat-sensitive cells in the nose to
help them locate capillary rich areas of skin, which
they nick with their razor sharp, blade-like, upper nitrogenous waste, and the kidney switches from a
incisors. The shallow wound is kept open by an water expelling to a water conserving mode.
anticoagulant in the saliva, and the bats lap up the
blood as it flows—they do not suck. It takes them Reciprocal altruism
20–30 min to take in around 25 ml of blood—a large Another fascinating thing about vampires is their
quantity for a bat that weighs only 30–35 g. They can social behaviour, studied to great effect by Wilkin-
be so heavy that they are unable to take off: their son (1984, 1990). Examples of reciprocal altruism, in
specific name, rotundus, is therefore appropriate to which animals return favours to their mutual
their after-dinner appearance! Fortunately for the benefit, have been put forward for many years.
bats, since this situation leaves them open to preda- However, most turn out to be either false, or exam-
tion, they have very efficient kidneys: if the meal is a ples of kin-selection, and the individuals are really
long one, they begin to urinate excess plasma from protecting their own genetic interests. In vampires
the ingested blood before they finish feeding. In the we find one of the few cases of reciprocal altruism
subsequent digestion of this protein rich diet, they that stands up to scrutiny. Within a colony of fe-
have to cope with very high concentrations of males (males roost separately, defending territories),
 ROOSTING AND FEEDING ECOLOGY 161

there are groupings of 8–12 individuals that roost starvation point. However the recipient, if it is rela-
close together on a regular basis over several years. tively close to starvation, gains 18 h. Computer si-
If bats do not get their 25 ml of blood on a regular mulations showed that the buddy system will work
basis they rapidly deteriorate: 2–3 nights without a well only if donors give to their buddies when the
meal takes a bat close to starvation. However, recipients have 24 h or less to live, and this is just
within these social groups, back at the roost, bats what happened in the laboratory experiments.
that did get a meal on a particular night will regur- Wilkinson (1987) has studied other phyllostomids,
gitate blood for those that did not, and this beha- concluding that both kin-selection and reciprocal
viour is reciprocated on later nights. Genetic altruism are plausible mechanisms governing aspects
analysis of blood samples from bats within a group of the social behaviour in those bats with stable group
showed that although there are sibling and parent– composition and/or a high degree of relatedness
offspring groupings, many of the bats within groups among individuals in the group. Research on the
are not related. Study of the blood sharing beha- behaviour of other species with varied social systems
viour showed that bats regurgitated blood for is proving to be very informative in this context—no
related and unrelated individuals in the group, and other group of mammals can boast such a diverse
that sharing occurred between unrelated bats only if social organization—and some of this work will be
they roosted together regularly—in other words covered before the end of the book.
with those able and likely to do the same for them
on a later occasion. Vampires therefore provide us
Bats that eat plant products
with an example of both kin-selection and reciprocal
altruism. When the field studies were followed up The Old World fruit bats feed almost exclusively
by laboratory experiments, it was shown that the on fruits, flowers, pollen, and nectar. In the New
bats in fact tended to set up buddy systems, with World, plant-eating appears to have evolved inde-
pairs of bats forming the tightest blood-sharing pendently on several occasions within the large and
associations. diverse Phyllostomidae. It is assumed that in taking
Figure 6.19 shows weight loss with time for a insects from fruit and flowers they developed a taste
vampire after returning to the roost with a good for the vegetation underneath. Many of the phyllos-
meal inside it. Weight loss is initially rapid, slowing tomid species are probably less than committed
gradually to a starvation point after about 60 hours. vegetarians, taking insects to varying extents. Phyl-
In the example shown, if a donor gives just 5 per lostomus hastatus, although it eats predominantly
cent of its body weight in blood to a buddy, then it fruit, takes the occasional vertebrate in addition to
loses the equivalent of 7 h of its time to the insects: perhaps protein-limited lactating females.

120

donor gives 5% body weight in blood

110
loses ~7h
% pre-fed weight

100
recipient receives 5% body weight
gains ~18h
90

80

70
Figure 6.19 Weight loss curve after feeding in Desmodus
60 50 40 30 20 10 0
rotundus, the common vampire. A cost-benefit analysis of blood-
time to starvation (h) sharing (adapted from Wilkinson 1984).
162 BATS

Conversely, several species thought of as insectivores may be retained to puncture nectaries. However,
eat nectar and pollen on a seasonal basis. Many fruit- the Old and New World environments differ in
eating bats may need to supplement their diet to many respects, and the plant-eaters in the two
meet their protein requirements (for example Courts worlds evolved in different ways to meet the chal-
1998), but the need will depend upon physiological lenges posed by their environments. Fleming (1986,
requirements and the fruit eaten. Wendeln et al. 1993, and Fleming et al. 1987), argues that the key
(2000) suggest that a mixed diet of several fig spe- lies in the diversity and ecology of the food plants,
cies, for example, can meet nutritional needs. which have led to two fundamentally different
Pollen and nectar provide a rich diet. In addition strategies. Both families are large and evolution-
to carbohydrate, bat nectar is often rich in lipids, and arily successful (Pteropodidae—42 genera, 186 spe-
pollen in protein. Around 70–75 per cent of the diet cies; Phyllostomidae—55 genera, 160 species) and
of Leptonycteris yerbabuenae (L. sanborni) is nectar, Fleming studied their adaptive radiation by exam-
and a smaller amount of pollen provides all the ining a number of morphological traits at the
necessary protein. The hair on the head and generic level of each family, in relation to those of
shoulders of some bats is rough and scaly to collect the most primitive genus in each (Rousettus and
pollen, which is groomed off and eaten back at the Macrotus). The results are summarized in Fig. 6.20.
roost. The long tongue is useful for pollen grooming Average size has not changed over the course of
as well as for nectar extraction. The latter is made the evolution of both families, but the Pteropodidae
more efficient by the evolution in some species of are bigger. Since the mass-specific cost of flight de-
brush like papillae on the tip of the tongue. creases with increasing body size, Old World fruit
Some Old World fruit bats eat leaves (for example bats fly more efficiently. As they evolved, the wings
Kunz and Ingalls 1994, Kunz and Diaz 1995), per- of pteropodids increased in aspect ratio—long nar-
haps to provide some essential nutrients not present row wings are aerodynamically more efficient and,
in the rest of their food. One reason why no bats use with a high wing loading, lead to fast, efficient
leaves as a primary energy source is that the size and flight. Phyllostomids, in contrast, are relatively
weight of the specialized teeth and digestive system small bats, with low aspect ratios and wing loadings
required would place serious constraints on flight for slow, manoeuvrable flight. The New World tro-
performance—not to mention the huge quantities pics have a greater diversity of food plants, and
of matter that need to be ingested and digested! show less spatial and temporal patchiness, than
A number of species have been shown to eat seeds those of the Old World. In the Old World, a smaller
(for example Nogueira and Peracchi 2003) which number of food plant species are generally more
may be another important source of protein as well scattered, and a greater proportion go in for ‘big
as other nutrients. bang’ flowering or fruiting, that is they flower or
fruit explosively over just a few days. Pteropodids
Convergent evolution again? may commute 50 km or more in their search for
We might expect to see a classic case of convergent fruiting trees, hence their adaptations for rapid, effi-
evolution between the Old World fruit bats and the cient flight. Phyllostomids, with a richer and more
phyllostomids, and at first glance we do. For exam- predictable source of food, rarely range more than
ple, species that feed almost exclusively on nectar 10 km from the roost, and frequently much less. The
and pollen are found in both families, and they jaw structure of pteropodids appears to have
exhibit many similarities. The Macroglossinae (Pte- diverged little from that of their ancestors, since
ropodidae), like their counterparts in the Phyllos- they must remain generalist feeders in order to
tomidae (such as Glossophaginae) are small, and exploit any flowering or fruiting tree they encounter
some hover like hummingbirds in front of flowers. in their long foraging trips. The richer habitat of the
Members of both groups have long muzzles for phyllostomids has enabled them to specialise, and
probing flowers, and long tongues to extract nec- many are specialist frugivores or nectarivores, with
tar. Since they have little use for teeth, they are jaw structure to match. The divergence resulting
often reduced, although a sharp pair of incisors from these two different strategies: the large, wide
 ROOSTING AND FEEDING ECOLOGY 163

2.0
1.8

relative 1.6
forearm Phyllostomi-
1.4 dae Pteropodidae
length
1.2
1.0

0.8
0.6
2.6
2.2
1.8
1.4
relative jaw width 1.0
to length ratio 0.6
0.2 1.2 1.4
0.8 1.0

relative aspect ratio

Figure 6.20 Adaptive radiation in plant-eating bats (from Fleming 1993). Changes in forearm length, wing aspect ratio, and jaw shape relative to bats of the
most primitive pteropodid (Rousettus) and phyllostomid (Macrotus) genera.

ranging generalist versus the small, home range suggests that the bats behave as if following a
specialist, can be seen in Fig. 6.20. simple rule of thumb: leave the panicle when a
total of eight flowers have been found empty.
Optimal foraging again This particular strategy was thought to lead to
Studies of Leptonycteris yerbabuenae (Fig. 6.21) by more efficient foraging than other possible ‘rules’
Howell (1979) and Howell and Hartl (1980) re- and in fact was close to the best strategy possible.
vealed an interesting example of optimal foraging. For a cost of 0.003 kJ in circling the panicle (6.1 m),
This bat spends much of its time feeding on nectar a bat gets 0.84 kJ of energy—sufficient to power a
from the large flower panicles of the agave, Agave flight of 0.25–0.5 km—not a bad return. There are
palmeri, a desert succulent of Mexico and the other important points to note: by foraging in
south-western United States. It is a medium-sized flocks the bats enhance their chance of finding
(17 g) bat which roosts in groups of 20–100 indivi- the next panicle unexploited, and hence with full
duals, and after dusk they forage as a group. flowers—essential for the successful operation of
When one bat finds a panicle with flowers full of their foraging strategy. The bats can thus switch
nectar, all of the other bats join it and circle round with minimal risk of making a bad decision. Note
the plant. They then leave the circle at random and that in this example, the bat is optimizing energy
take turns to dip into the flowers for nectar. At input in a 24 hour period: this may not always be
some point, one of the bats leaves this panicle, and the most important consideration (see Heithaus
flies in search of another. Apparently, all of the 1982, p. 355).
remaining bats follow and make no further
attempt to feed on that panicle. The first bat to
Information transfer
leave can be any one of the group, and there is
no evidence of a social hierarchy within the flock. Given the gregarious nature of bats, we might
Why do the bats feed in flocks, and what makes expect them to have well-developed communication
them suddenly leave? A combination of field- skills, and to intentionally and unintentionally share
based experiments and mathematical modelling information about roost sites and foraging areas.
164 BATS

Figure 6.21 Leptonycteris yerbabuenae

at an agave panicle.

Although communication in bats has received con- for a brief discussion and references). Insectivorous
siderable attention (see Fenton 1985, Altringham bats quickly gather round patches of prey, presum-
and Fenton 2003), information transfer in particular ably as the echolocation calls of feeding bats attract
has received relatively little. A review by Wilkinson other bats in the vicinity. Barclay (1982) showed that
(1995) provides a good summary and points the little brown bats (Myotis lucifugus) flew to loudspea-
way to later research. Wilkinson describes four kers broadcasting calls of both little (Myotis lucifu-
basic mechanisms of information transfer: local gus) and big (Eptesicus fuscus) brown bats: good
enhancement, social facilitation, imitative learning, evidence for eavesdropping behaviour. Bats may
and intentional signalling. also find good foraging sites by following other
Local enhancement involves unintentionally di- bats. Wilkinson (1992b) has shown that the foraging
recting another organism to a particular part of the success of evening bats, Nycticeius humeralis, is
habitat. Although many species of insectivorous bat increased by following behaviour. These bats
appear to forage singly, group foraging is seen in exploit insect patches that persist for several days.
some species some of the time, in particular among It appears that bats that have failed to find these
aerial hawkers, most likely when emphemeral insect patches will follow at random another member of
prey occur in dense patches (see Kerth et al. 2001a, the colony on its next flight from the roost, and in
 ROOSTING AND FEEDING ECOLOGY 165

doing so reduce their chance of returning without a females approach the calling males, the intentional
meal. This system will be most beneficial to those nature of this signal is clear. Mating calls are also
bats making frequent foraging trips from the roost made many other bats, for example the European
in a given night—notably females during lactation. pipistrelle, Pipistrellus pipistrellus (Lundberg and
Social facilitation is an increase in individual Gerell 1986) and the white-lined bat, Saccopteryx
foraging success brought about by group foraging bilineata (for example Bradbury and Emmons 1974,
behaviour. Some degree of coordinated behaviour Behr and Helversen 2004, Bohn et al. 2009). Territo-
has been reported for a number of group-foraging rial calls, to defend foraging sites or mates, are
bats, for example proboscis bats, Rhynchonycteris widely described for temperate and tropical insecti-
naso, follow each other when feeding on insects vores, for example the European pipistrelle (Racey
(Bradbury and Vehrencamp 1976a), and Phyllosto- and Swift 1985) and the white-lined bat (Bradbury
mus discolor visit flowers in turn (Heithaus et al. and Vehrencamp 1976a). The Australian ghost bat,
1974), but whether this improves foraging success, Macroderma gigas, (Tidemann et al. 1985) calls at
and by what mechanisms, is largely unknown. feeding sites in the absence of territorial behaviour,
Wilkinson (1995) suggests that the decrease in fre- suggesting advertisement of food. A long-term
quency of echolocation pulses in the terminal feed- study of the spear-nosed bat, Phyllostomus hastatus,
ing buzz observed in some species may serve to (see Boughman (2006) for the latest stages and a
advertise prey and increase group size. summary of past work) has shown that stable,
largely unrelated groups of females use these calls
to assist the formation of foraging groups as they
Imitative learning
leave their cave roosts. Group foraging reduces
Whilst a number of studies have looked for evi- search time and may enable the bats to defend a
dence that bats can learn foraging techniques from food patch against conspecifics. It also appears to
other bats (notably their mothers) most are far from lead to improved pup condition at weaning. Group
conclusive. The most persuasive case can be made foraging, as expected, is most common during per-
for the common vampire, Desmodus rotundus, in iods when food is particularly patchily distributed.
which feeding ranges of mother and yearling- Studies of distress calls by Russ et al. (2004) indi-
daughter overlap significantly more than expected cate that they work across species, genera, and even
(Wilkinson 1985). The long period of association families, suggesting convergence in call structure,
between mother and young perhaps facilitates which makes sense if mobbing can deter preda-
this. There is good evidence to show that the social tors—the more the better and the species of the
and echolocation calls of young bats can closely mobber is largely immaterial.
resemble those of their mothers (for example Brazi-
lian free-tailed bats, Tadarida brasiliensis, Balcombe
and McCracken 1992; greater horseshoe bats, Rhi- Interactions between habitat, food,
nolophus ferrumequinum, Jones and Ransome 1993).
foraging, social structure, and behaviour
Although it has been shown that isolation calls are
heritable and very specific even in bats just 2–5 Over the last two chapters we have looked at mating
days old (evening bats, Nycticeius humeralis, Scher- systems, roosting ecology, feeding ecology, and
rer and Wilkinson 1993), Jones and Ransome (1993) information transfer. Can we pull these aspects
provide some evidence to suggest that echolocation together and say something about the way they
call structure in greater horseshoe bats does have a interact to give a bat a particular life history? This
learnt component. is an essential task if we are to really understand the
Intentional signalling can take many forms: mat- biology of bats, but it is also difficult and few studies
ing calls, territorial calls, alarm calls, and food calls have set out to do this explicitly.
have all been reported in bats. The most spectacular Several questions are of fundamental importance.
mating calls are those of the epomophorine bats What roosting sites are available? What is the life
(Bradbury 1977a, b). Since call rate increases when span of these roosts? How abundant is the food
166 BATS

supply? Is it evenly or patchily distributed? How foraging and commuting, or separate these tasks?
does food supply vary in time? Where are the roosts It would make sense to combine them if food was
in relation to foraging sites? What are the risks of abundant and uniformly distributed. Conversely, in
predation? When we have answers to these ques- getting to a known, rich patch of food in an unpre-
tions we can begin to piece together the story. dictable environment, commuting by the shortest
The importance of food distribution can be best route without foraging is likely to be the optimal
illustrated with some extreme examples. If the food strategy. Predation risk may also play an important
supply of a bat is widely and densely distributed, role, but it is difficult to generalize, since the risk will
and does not vary with the seasons, the bat need be determined not only by the foraging style of the
only travel a short distance from its roost to find bat, but also by the strategy of the predator. As an
food, and it can return again and again to those sites example, long flights combining commuting and
it knows well. Such sites may offer other benefits in foraging will make bats particularly vulnerable to
addition to low commuting cost and time, for exam- aerial predators: perhaps other bats and owls. The
ple a reduced risk of predation. Such a bat may proximity and availability of roosts will also influ-
forage most effectively if it forages alone. An abun- ence strategy: cave roosts are relatively scarce, and
dant food supply need not be defended: it may even may necessitate long commuting flights irrespective
be energetically unfavourable to expend time and of the spatial and temporal distribution of food. Tree
energy on defence rather than on feeding. An abun- and foliage roosts are generally more abundant, and
dant and invariant food supply may also enable the will enable bats to roost close to their food source.
bat to specialize in a small number of food types. The use of special night roosts, or the ability to
At the opposite extreme, we have a species that return to the day roost between foraging excursions
relies on a patchy source of food, and this food will influence foraging style, and the level of infor-
undergoes large seasonal, or even unpredictable, mation transfer between group members. Night
changes in abundance and distribution. This bat roosts used for resting or food consumption may
may need to travel considerable distances in search also reduce the risk of predation. The size of the
of food, and it will have to change its foraging sites bat species may be important, since small bats
to locate new sources. Foraging in this species may have high energy requirements, which may necessi-
be more effective in groups, since a group can search tate a high energy content diet and a specific forag-
a wider area, and when one individual finds food, ing strategy. It is clear that all of these factors, and
others may be attracted to it intentionally or unin- many others, determine the foraging strategy of an
tentionally. Whether or not a patch of food is individual bat, whether or not the bat will forage as
defended by the group, or small patches by indivi- part of a group, the compositional stability of a
duals, will depend upon how rich the food supply is group, its home range or territory, and its fidelity
and how defensible the patch. For example, a very to particular roosts sites. As we have seen in the last
rich but difficult to defend patch is not worth chapter, it is these factors that in turn determine
defending—best to just tuck in and eat. A patchy mating strategy. In a species that forms stable
and ephemeral food supply may encourage a gener- female groups, with a high fidelity to a particular
alist feeding strategy, with bats taking a wide range roost site, a male will find it easy to defend a harem
of transiently available food. against other males. If the females forage as a group,
This is just the simple tip of a complex iceberg: the male or males may forage with them, and the
there’s much more under the surface. If the patchy group may even defend a territory. When female
food of our group forager undergoes rapid changes group composition varies, and roost sites change
in abundance and distribution, then any group of frequently, other strategies must be adopted, such
bats is likely to be as successful as another. If patches as mating roost defence or lekking. Finally, the
persist for days or weeks, then stable groups, within social status of the bat is important. A dominant
which information about patch locality and abun- male will spend time in defence of a roost or
dance can be shared, may be more successful for- harem or in advertisement, and this may necessitate
agers. Should a bat or group of bats combine changes in foraging strategy. I could go on, but the
 ROOSTING AND FEEDING ECOLOGY 167

point has been made, and I think it is time to look at minimize the size of their foraging territory, such
some specific examples and see if there is any evi- that it contains only one active feeding site. The
dence to suggest that bats conform to any of these smaller the territory, the better the bats’ knowledge
rules. of the site and the food patches within it. On this
basis, the wet habitat bats should be high density
overall with a fine-grained dispersion, that is, many
Insectivores
small colonies with small territories, to exploit the
Bradbury and Vehrencamp (1976a, b, 1977a, b) stud- many small insect patches in close proximity to each
ied five species of emballonurid in Costa Rica. The other. In dry habitats, large seasonally rich insect
food supply of all species was spatially and tempo- patches will support large colonies, but these will
rally variable, so we might expect group foraging be well separated (coarse-grained) and overall pop-
and possibly stable social groups. Of course life is ulation density will be low.
rarely that simple, and this study is a good lesson in So much for the theory, what about the practice?
how the complexity of real situations makes it diffi- The fit is surprisingly good. The wet habitat bats,
cult to uncover the basic rules governing behaviour. Rhynchonycteris naso and Saccopteryx leptura, do
Important aspects of the biology of the five species indeed live in small colonies, forage in small terri-
are summarized in Table 6.1. tories, and have high population densities. In con-
There was no obvious relationship between social trast, the dry habitat Balantiopteryx plicata lives in
structure (colony size and stability) and group for- large colonies and forages several kilometres from
aging. The three smaller species all foraged fre- the roost. Unfortunately, population density is
quently in groups, the larger two were always unknown. Saccopteryx bilineata moves between the
solitary feeders. Clearly, we need to look deeper. wet and dry forests over the seasons, exploiting
What about the habitats, and the way the bats larger, more dispersed insects when in the wet for-
used them? est—hence the coarse-grained, low density disper-
There are two dominant habitats, ‘wet’ and ‘dry’. sion. Peropteryx kappleri is something of a puzzle—a
Wet habitats are characterized by high plant species wet forest bat with a fine-grained dispersion, but
diversity, a heterogeneous mix of plants, and high low population density.
plant activity (that is a high turnover of foliage, Are any other features of Table 6.1 worth closer
flowers, and fruit), with minimal seasonal variation. study? We can divide the five species into two
Since there is little synchrony of activity among groups: R. naso, S. leptura, and P. kappleri—the wet
plants, mean plant activity within a patch is rela- forest species; and B. plicata and S. bilineata—the dry
tively low. Since insect abundance is related to plant habitat species. The wet forest bats are all specialist
activity and dispersion, these habitats are character- foragers, adapted to feeding in specific habitats.
ized by small, low density patches of insects around They utilize several roost sites within their small
active plants, and these small patches are densely territories, group composition is stable, and there is
packed: a ‘fine-grained’ insect dispersion. In con- a tendency towards monogamy. Dry habitat bats are
trast, dry habitats have lower plant species diversity opportunistic foragers, utilizing a wide range of
and a tendency towards monospecific stands of low habitats in their large territories. They use a single
activity. Plants often re-leaf or flower in synchrony, roost site, group composition is unstable, and mat-
so within a patch, plant activity levels can be sea- ing is by ‘swarming’ (not to be confused with that
sonally very high. Dry habitats are therefore char- discussed in Chapter 5) or harem formation. These
acterized by large, well-separated patches of insects patterns can’t be explained with any confidence, but
that can be of very high density: a coarse-grained here is one possible scenario for debate. In the wet
dispersion. How does this influence the dispersion forest species, the small territory size gives bats an
of the bats? First of all, Bradbury and Vehrencamp intimate knowledge of their feeding sites. In combi-
(1976b) persuasively argue that food supply is the nation with the small, low density food patches, it is
major determinant of population density in these therefore more profitable to forage singly or in small
insectivorous bats, and that the bats should groups. The stability of the wet forest environment
Table 6.1. Aspects of the biology of five species of emballonurids in Central America (based on Bradbury and Vehrencamp 1976a, 1976b, 1977a).

Species weight (g) colony (group) group mating system foraging dispersion, territory foraging strategy/ roost site
size mean, range compositional defence, territory size, mean habitat habitat
stability density of bats

Rhynchonycteris 4.5 7.8, 5–11 high Multi-male groups, female f: group beats, at territory centre m: specialist: over water, tree boles near water,
naso defence by males solitary, at territory periphery colony rarely pasture, wet move around 3–6 sites
territory defended, 1.1 ha, 8 bats/ha forest, river
Saccopteryx 5.0 3.6, 1–9 high Monogamous pairs (?), solitary and group beats common colony specialist: below mature or tree boles in forest, move
leptura resource/female territory defended, 0.14 ha, 17.6 riparian forest canopy at around several sites
defence by male bats/ha (Trinidad), 1.4 ha, 2.5 bats/ dusk, moving above
ha (Costa Rica) canopy with insects
later in the night, wet
forest
Balantiopteryx 6.1 ?, 50–2,000 low swarming (?) male strategy solitary and group beats, defended, opportunistic: pasture, caves and crevices, no
plicata unknown unknown but large, unknown swamps, woodland in movement
highly seasonal areas,
dry seasonal savannah
Saccopteryx 7.7 6.8, 2–32 low temporary harems, solitary beats, clustered to form colony opportunistic: all habitats, tree buttress and trunk
bilineata resource defence by territory, females of harem forage on as insects peak cavities, no movement
male adjacent patches, and harem territory seasonally, wet/dry
defended by male (against males) forests
and females (against females), 7 ha,
0.7 bats/ha
Peropteryx 9.0 4.3, 1–7 high Monogamous pairs (?), solitary beats, unknown, 0.6 bats/ha specialist: sub-canopy of fallen logs, tree buttress
kappleri resource defence by mature forest, wet cavities, some
male forest movement between
roosts
 ROOSTING AND FEEDING ECOLOGY 169

allows bats to specialize at foraging in particular constant throughout the night (Fig. 6.22b, c).
habitats. Since females forage singly or in small M. evotis forages by hawking and probably also by
groups, males can only defend small groups, hence gleaning: it has the low intensity echolocation calls
the tendency towards monogamy and stable group and low wing loading characteristic of confirmed
composition. Small colony size enables them to gleaners, and captive individuals used prey-gener-
exploit a wide range of roost sites, and they may ated sounds to take stationary moths. In the wild it
occupy the site closest to active food patches. Dry feeds primarily on moths (Fig. 6.22d), despite their
habitat bats must be opportunistic foragers to con- low abundance, but by doing so is exploiting a
tend with the marked spatial and temporal variation temporally more reliable food source. Barclay sug-
in food supply. The large colonies of B. plicata limit gests that with their more flexible foraging style and
choice of roost, and bats must disperse over a wide more reliable food source, female M. evotis can meet
area to find sufficient food. Female defence is per- the high energy demands of reproduction, and can
haps not possible in the large cave aggregations, therefore breed successfully in the region. Female
hence the ‘swarming’ mating strategy and the unsta- M. lucifugus would need to double their nightly
ble groups. It is more difficult to suggest reasons for foraging time to meet the high cost of reproduction,
the use of a single roost site for S. bilineata. Although a strategy unlikely to pay in the cool mountain
females forage singly, they do form groups and habitat studied. They must therefore move to habi-
individual members forage on adjacent beats. tats with a higher insect abundance, the warmer
Males can thus defend a number of females, to prairies to the east, where nursery colonies are
form harems. found in large numbers. Males of both species
have low energy demands and make extensive use
Sexual segregation on a landscape scale of torpor when prey availability is low (for example
I know of no other comparative studies of this Hamilton and Barclay 1994) and so are able to
nature, and we must piece together examples occupy habitat at all altitudes.
mainly from studies of just one or two species. Segregation by altitude has been observed in
Barclay (1991) gives a very interesting example of other species too. For example, in the Black Hills of
how food availability and foraging style influence South Dakota USA, females, and particularly repro-
the distribution of related species and lead to sexual ductive females, of all 11 species of bat captured,
segregation on a landscape scale. Little brown bats, declined in abundance with increasing altitude rela-
Myotis lucifugus, and long-eared myotis, M. evotis, tive to males of the same species (Cryan et al. 2000).
are the two most common species on the eastern Other North American studies show similar trends
slopes of the Rocky Mountains in Canada. Both are (for example Grindal et al. 1999, Storz and Williams
small (5–8 g), hibernate in caves and mines, and 1996, and see references in Cryan et al. 2000). In
form nursery colonies in trees or buildings. How- Europe, Daubenton’s bat, Myotis daubentonii, also
ever, Barclay noted that over 90 per cent of segregates in the Jura mountains (Leuzinger and
M. lucifugus caught in harp traps and mist nets in Brossard 1994), Italy (Russo 2002), and in the U.K.
the summer were males, but the sex ratio in M. evotis (Senior et al. 2005). Studies of an Old World fruit bat,
was 1:1. He argued that differences in foraging style, Otopteropus cartilagonodus, on Luzon in the Philip-
in a habitat of cool nights and relatively low insect pines (Ruedas et al. 1994) showed marked sexual
abundance, are responsible for the very different sex segregation along an altitudinal gradient and dis-
ratios. tinct differences in size and skull/jaw structure. It
M. lucifugus typically forages opportunistically was suggested that this would reduce inter-sexual
over water, feeding on the wing on small prey— competition for food, but no behavioural or ecologi-
notably chironomids (Fig. 6.22d)—which are very cal studies have been carried out.
abundant, but only for the first couple of hours We have investigated the mechanisms underlying
after sunset (Fig. 6.22a). M. evotis forages along altitudinal segregation of one species in some detail
paths and within forests, where insect abundance and looked at the consequences to mating patterns
is lower than over water but remains relatively (Senior et al. 2005, R.L. Angell, R.K. Butlin, and
170 BATS

(a)
(b)
aerial biomass (mg per trap)

120
100 100
80 80
60 60
40 40
20 20
0 0
early middle late water path forest

(c)
(d)
stationary biomass (mg)

300 M. evotis
60
percentage by volume

M. lucufugus

200 Figure 6.22 Temporal and spatial

40 patterns in food supply, and the diets of
two vespertilionid bats. Biomass and spatial
100 distribution of aerial prey available to M.
20
lucifugus and M. evotis in eastern Canada
(a,b). Biomass of stationary prey on
vegetation (c) and the diet (% by volume of
early middle late Lepidoptera Hymenoptera Neuroptera prey) of the two species (d) (from Barclay
Chironomidae Homoptera 1991).

J.D. Altringham. in prep.). Daubenton’s bat, Myotis forearm length ratio). This suggests that they may
daubentonii, is similar in size and behaviour to be excluded from better habitat lower down the
M. lucifugus in N. America, roosting close to rivers valley by more dominant males or by females. This
and feeding by aerial hawking and by gaffing in- view is supported by the finding that the males that
sects off the water surface. At high elevations in roost at intermediate elevations with the nursing
river valleys only males have been found roosting females have a much higher probability of fathering
and feeding, over many years. At lower elevations their offspring (based on microsatellite genotyping
both males and females are found, but nursery col- of females, their young and males). This species is
onies are almost exclusively comprised of females. typical of its genus in that its primary mating system
In the best-studied valley, a mixed-sex nursery col- appears to be autumn swarming. It appears that
ony roosts at an intermediate elevation. The males at dominant males may in this case monopolize
high, intermediate and low elevations are discrete females prior to autumn swarming. Males at the
populations, each bat occupying the same region lowest elevations do not share nursery roosts with
over many summer seasons, showing high fidelity females, and the offspring from these roosts are
to both roosts and foraging sites. Habitat at high more likely to be fathered by males at swarming
elevations is less than optimal, with cooler and sites. Given the widespread nature of sexual segre-
more variable night temperatures and a river gation in summer roosts of Myotis species, why is
which is frequently narrow and turbulent. The the roost at intermediate elevations mixed? One
males occupying this region feed for significantly possibility is that females tolerate males in the
longer each night and travel further to their foraging roost to increase roost temperature in what may be
sites than males at intermediate elevations, yet they marginal habitat for a nursery colony. The males
have a lower body condition index (body mass: benefit from the observed increase in mating
 ROOSTING AND FEEDING ECOLOGY 171

success. Whatever the reason, foraging patterns, and M. nattereri enables males to defend females,
social structure, and mating patterns are complex and relatively high fidelity (within and between
and clearly determined by changing habitat quality years) between males and female groups was
along the valley. found. Unfortunately, we have no idea if the two
Park et al. (1998) noted different patterns in the gleaning species do make use of information
social structure and mating systems of three species exchange or group foraging strategies, leaving a
of bat with different foraging strategies. The com- large gap in this potentially elegant explanation.
mon pipistrelle, Pipistrellus pipistrellus, brown long- Studies of Bechstein’s bat, Myotis bechsteinii, (Kerth
eared bat, Plecotus auritus, and Natterer’s bat, Myotis et al. 2001a) suggest that we ought to be cautious.
nattereri, used artificial roost boxes in the same for- Despite strong social cohesion in female nursery
est. P. pipistrellus is an opportunistic hawker of a roosts, there appears to be little information transfer
wide range of small insects around forest rides, about feeding sites: individual females were faithful
edges, and clearings. In contrast, P. auritus and to their own foraging sites within and between
M. nattereri are predominantly gleaners within the years. We must look elsewhere for an explanation
forest, P. auritus taking a disproportionately large of their sociality. Perhaps, as Kerth et al. (2001a)
numbers of moths and M. nattereri a large propor- suggest, philopatry gives them the intimate knowl-
tion of non-volant insects. The two gleaning species edge of their small feeding sites that is essential for
formed maternity colonies in the roost boxes, but efficient foraging. Social cohesion may bring other
P. pipistrellus did not. Roost box use by P. pipistrellus benefits too, such as cooperative breeding. These
was predominantly by solitary males during spring alternative explanations may underlie the strong
and summer, and in the autumn these males de- philopatry and social cohesion observed in many
fended territories and attracted small groups of up other species not expected to benefit greatly from
to 10 transient females, presumably from nearby information transfer and group foraging. For exam-
nursery roosts. Plecotus auritus and M. nattereri ple, Daubenton’s bats, M. daubentonii, are also faith-
formed mixed-sex roosts throughout the summer ful to just one or two short stretches of river, within
and autumn, with females outnumbering males and between years (Senior et al. 2005).
2–3 fold. As suggested by Barclay (1991), the glea- It is difficult to know where to draw the line on
ners may have a more reliable food supply and a this discussion, but I do not want to stray into the
more flexible foraging style, enabling them to complexities of social structure just yet. Many recent
exploit the woodland more efficiently. Although studies have investigated patterns in colony and
these differences may be related to foraging strat- population structure, how they change in time and
egy, roosting requirements may also be important. space, and the ecological and behavioural mechan-
Maternity colonies of P. pipistrellus are usually larger isms responsible for them. This topic will be covered
than those of the other two species and almost exclu- in the next chapter.
sively in buildings: are roost size and microclimate Brigham (1991) drew attention to the importance
more critical? of geographical variation within a species. The big
Plecotus auritus and M. nattereri, with their empha- brown bat, Eptesicus fuscus, formed maternity colo-
sis on relatively large prey, probably have to deal nies in tree hollows in the semi-arid Okanagan val-
with a more widely distributed and patchy food ley of British Columbia, and bats consistently
supply than P. pipistrellus. Female groups in both foraged over a short stretch of river up to 4 km
species were more stable, which makes sense from their day roosts, where swarms of Trichoptera
because they may have more to gain from informa- provided more than half of their diet. In contrast, the
tion exchange and group foraging. Low stability of same species studied near Ottawa in Ontario used
female P. pipistrellus groups means low defensibil- buildings for maternity roosts. Since these were
ity, hence the defence of a mating roost by males, close to the centre of their foraging ranges bats
rather than defence of the females themselves. Indi- flew less than 1 km to feed, but foraging sites varied
vidual males defended the same boxes for up to five nightly. The use of buildings in Ontario is probably
successive years. High group stability in P. auritus due to the scarcity of natural sites in an area that
172 BATS

retains little of its native woodland. The different Phyllostomus hastatus has a more varied and flexi-
foraging pattern presumably reflects insect distribu- ble diet, opportunistically feeding on fruit, pollen,
tion in this more temperate region. and insects (McCracken and Bradbury 1981, Willig
et al. 1993). Some particularly profitable food items
may be thinly and unpredictably distributed.
Plant-eaters
McCracken and Bradbury (1981) have shown that
Plant-eating bats eat a wide array of wild and com- this bat forms very stable groups of unrelated
mercial fruits including figs, kapok, pepper, females. Males form harems by defending a female
mangoes, bananas, peaches, and even young coco- group, whose composition may remain unchanged
nuts and cocoa pods. At least 188 species from 64 for several years. McCracken and Bradbury consid-
genera are exploited by pteropodids alone. In the ered a wide range of possible benefits that might
tropics, trees come into fruit throughout the year, arise from group formation, ruling out all but coop-
and are well distributed spatially, giving a constant eration in foraging and foraging site defence.
supply of food. Although the food supply is avail- Although bats from a particular group left and ret-
able all year, it can be very patchy, both temporally urned to the roost alone, each group had its own
and spatially, and the foraging habits of the bats are feeding territory, within which each bat had its
moulded by the plants they feed on. As with many own beat. They also noted that all of the bats
other aspects of bat behavioural ecology, we know would occasionally leave their own beats to forage
only a little about only a few plant-eating bats. Early together, and that they were very vocal when forag-
work has been summarized and interpreted by ing. Group foraging was observed around flowering
Fleming (1982), and can be supplemented by several Hymenaea trees and fruiting Lecythis trees. Bloedel
more recent studies. The phyllostomid Artibeus ja- (1955) reported group feeding on termites. Although
maicensis is a widespread and numerous species some crucial evidence was lacking, McCracken and
with a very flexible frugivorous diet. However, in Bradbury concluded that the driving force for the
the moist tropical forest of Barro Colorado Island in establishment of these stable groups was recipro-
Panama, it relies primarily on two abundant species cated group foraging, with successful bats alerting
of fig which produce heavy crops of fruit (Ficus
insipida produces 40,000 fruits over a period of one
week, once or twice annually). With this rich and
predictable food supply, individuals forage singly,
amplitude

usually within 1 km of the roost, and make frequent

forays between fig trees and their night feeding
roosts, typically within 200 m (Morrison 1979, Mor-
rison and Hagen-Morrison 1981). Females only 400 ms
return to the roost to suckle young. Male defence
of feeding sites is impractical, given their ephemeral
and rich nature, as is the defence of females, since
12 kHz
they disperse to forage and roosting groups are
unstable. The males in fact defend tree hole roosts,
which are in short supply. Males have a very differ-
frequency

ent foraging strategy: they spend 90 per cent of the

night in roost defence, feeding only when the
females are absent on foraging trips. When they do
feed, they bring their food back to the roost to eat it,
necessitating long commuting trips. Defence flights
0 kHz
and long commuting flights dramatically increase
the energy demands of the males, covered by eating Figure 6.23 Social screech calls of Phyllostomus hastatus used to attract
more fruit (Morrison and Hagen-Morrison 1981). co-members of social groups (from Wilkinson 1995).
 ROOSTING AND FEEDING ECOLOGY 173

group members to rich patches of food. Wilkinson expected by chance alone. Again, harem males
and Boughman (1998) have shown that females use behaved differently, spending most of their time in
low frequency calls to attract group members as or close to the roost, and foraging for as little as one
they emerge from caves to forage and when feeding hour each night. Bachelor males also form groups,
(Fig. 6.23). but they are unstable, foraging sites are scattered,
Bats screech when leaving the cave, but almost and they forage for significantly longer than either
never on returning. Screeching bats are accompa- harem males or females, suggesting a rather differ-
nied by other group members more often than ent foraging strategy.
This page intentionally left blank
 CHAPTER 7

Migration, social structure,

and population structure

Migration, sex differences, time, and cost. Navigation. Social and population structure. Roosts and colonies.
Population structure on large and small scales. The effects of history on population structure.

Migration higher 13C:12C ratio. Migratory L. yerbabuenae had

13
C:12C ratios that were related to the season and
I’ll begin where I ended in the last chapter, with two latitude at which they were caught (Fig. 7.1a), indicat-
more phyllostomids, but look at a different aspect of ing a slow switch from C3 to CAM plants during the
the link between food and behaviour, as a way of spring migration north, and a return to C3 plants dur-
introducing the fascinating topic of migration. Most ing the autumn migration south. A study of the flow-
nectarivorous bats live in the tropics and subtropics, ering times of columnar cacti and agave broadly fits
where food can be found all year round, but a num- the pattern of bat migration (Fig. 7.1b). Four species of
ber of phyllostomids are migratory, spending the cactus and one agave flower roughly sequentially
summer in the deserts of the south-western United along the spring migration corridor, and as many as
States, and the winters in Mexico, feeding on route nine agaves are known to flower later in the year,
during their unhurried migration. Fleming et al. probably coincident with the autumn migration.
(1993) used a novel approach to discover which I’ve included migration in this chapter since it is
plants the migratory Leptonycteris yerbabuenae fed an aspect of behaviour that is often linked to social
at over the year, and compared its diet to non- structure and mating system and thus has a signifi-
migratory populations in Baja California, and the cant influence on population structure, which will
non-migratory Glossophaga soricina in southern be discussed in the rest of the chapter. Before going
Mexico. Plants utilize carbon dioxide in a number further, perhaps I should define migration: it can be
of different ways during photosynthesis, leading to defined as the twice-annual, seasonal movements of
characteristic differences in the ratio of the two sta- animals in response to changes in climate and food
ble isotopes of carbon, 13C and 12C, in the tissues of abundance. A phylogenetic analysis by Bisson et al.
C3 and CAM plants. By analysing the 13C:12C ratio (2009) suggests that migration has evolved indepen-
in the tissues of the bats, it was possible to determine dently in bats on several occasions, that strategies of
which plants they fed at. The rate at which the short- and long-distance migration evolved sepa-
carbon is taken up is dependent upon the rate of rately, and that the reasons for its evolution proba-
turnover of the tissue in question. By studying 3 mg bly differ between groups. Relative to birds, few
skin and muscle samples a snapshot of each bat’s bats migrate: between 30 and 40 out of over 1,100
diet over the previous two months was obtained. species. This number does depend upon how long
G. soricina fed mainly on C3 plants throughout the or short a journey is defined as migration and our
year: members of the Bignoniaceae, Bombacaceae, knowledge, or lack of it, about the movements of
and Leguminosae in particular (Fig.7.1a). many species. Bats migrate from cold to warmer
L. yerbabuenae, resident in Baja California, speci- climates so that they can continue to feed or to find
alized on CAM plants, the desert succulents of the a more suitable climate for hibernation. Bats also
cactus and agave families with their characteristically migrate to follow fruiting and flowering plants.

175
176 BATS

(a) excessive disturbance it caused and more sensitive

L. curasoae approaches are now used. Nevertheless, these early
resident
carbon isotope ratio (0/00)

–12 CAM studies provided valuable information. Interest in

migration has recently grown, driven by conserva-
–16
L. curasoae
tion concerns and the link between bats and some
migratory emergent diseases (see Chapter 9). It is likely to
–20
receive further impetus as new techniques open up
–24 G. soricina exciting new possibilities, for example the progres-
C3
sive miniaturization of satellite tracking technology
J F M A M J J A S O N D and the use of stable isotope analysis (for example
Month Britzke et al. 2009). The driving force for migration
(b) in the temperate zone is particularly strong, since
at a time when low environmental temperatures
late summer/autumn flowering
increase the energy demands of homeothermy,
agaves
food becomes scarce. Some bats hibernate locally,
but many migrate, either to find suitable hibernation
sites, or to find new food sources and remain active.
Migration is, as expected, most common in temper-
ate bats, and in long distance migrants usually in-
volves movements to higher latitudes in spring and
a return to lower latitudes in autumn. Migrants
spring flowering cacti and agave usually show a high degree of fidelity to both their
summer and winter habitats and even to specific
roosts. Although the majority of migratory bats are
Figure 7.1 The food plants of migratory and non-migratory nectarivores. temperate, some tropical and subtropical species,
(a) The 13C:12C ratio of tissue samples from two nectar feeding such as Leptonycteris yerbabuenae described above
phyllostomids during the year. High ratios indicate dependence on CAM and Pteropus poliocephalus in Australia (Eby 1991)
plants (desert cacti and agaves) low ratios on C3 plants. (b) The flowering
do migrate, the primary stimulus being variations
patterns of CAM plants on the Pacific coast of Mexico and the south-western
United States, which provide a food corridor for migrating bats (from Fleming in food supply due to seasonal variation in rainfall.
et al. 1993, with permission from Springer Science + Business Media). Migration has been studied most extensively in
temperate species in North America and Europe
(e.g. Findley and Jones 1964, Griffin 1970, Strelkov
1969, Tuttle 1976b, Cryan et al. 2004, Britzke et al.
Bisson et al. (2009) suggest several ways in which 2009). (Note that the sensory aspects of migration
migration may have evolved, none mutually exclu- will be covered later in the chapter). Many species
sive, and the paper is worth further exploration. in the genera Nyctalus, Lasiurus, Pipistrellus, and
Popa-Lisseanu and Voigt (2009) have written the Tadarida, to name a few, undergo seasonal migra-
most recent review of bat migration and the same tions of up to 1,700 km, in a generally north–south
journal volume has papers from a symposium they direction. Most of these migrations are between
organized on the subject. A little earlier, Fleming summer roosts and winter hibernation sites. Many
and Eby (2003) reviewed the global literature on of these bats hibernate in tree cavities and these long
migration, and a summary of European studies has journeys may be necessary to escape the longer and
also been published (Hutterer et al. 2005). more extreme winters at high latitudes. Other bats,
Migratory movements were studied intensively such as temperate members of the widespread
for several decades in the mid-twentieth century, genus Myotis, usually migrate over shorter dis-
due to the ease with which large numbers of bats tances, probably because ‘local’ caves offer suitable
could be ringed (banded) in their hibernation sites. hibernation conditions: caves are better buffered
This practice quickly fell out of favour due to the than tree-cavities against environmental extremes.
 MIGRATION, SOCIAL STRUCTURE, AND POPULATION STRUCTURE 177

The movements of short-range migrants are not roosting hoary bat, Lasiurus cinereus (Findley and
necessarily north–south, but are still between sum- Jones 1964). Like the noctule, it is a large bat for
mer roost and winter hibernaculum. temperate regions and a strong flyer, and has been
Although I refer to short and long distance reported to migrate in substantial waves, sometimes
migration, migratory distance is probably some- accompanying migrating birds. Populations trave-
thing of a continuum, with considerable variation lling the shortest distances do so to hibernate,
even within species. However, there is a general whereas those migrating to the southern end of
trend for long distance migrants to show adapta- their range will have a short or non-existent hiber-
tions for efficient and fast flight: high aspect ratio nation period. This pattern is common to several
wings, pointed wing tips and high wing loading other Lasiurus species, and bats of other genera.
(see Chapter 2). Long distance migrants are there- We saw in Chapter 4 how migration and hiberna-
fore generally found amongst species that forage tion patterns of the Brazilian free-tailed bat, Tadarida
by aerial hawking in open spaces. Short-range and brasiliensis, are dependent upon the geographical
sedentary bats are more likely to have the short, location of a particular population.
broad wings associated with gleaning. However, The grey bat, Myotis grisescens, has been studied
other factors are also important and provide ex- intensively and it typifies seasonal migrations in
ceptions to this general rule. many North American and European house and
As already stated, long range migration is most cave bats.
common in tree roosting bats, since tree holes are About 90 per cent of the grey bats in the south-
poor hibernation sites in the coldest climates. The 30 eastern United States hibernate in just three caves
g European noctule, Nyctalus noctula, migrates to the (Fig. 7.4): 375,000 in two caves in Tennessee, and
south or south-west in autumn, away from the cold, 1,500,000 in a single cave in northern Alabama (Tut-
continental interior, to warmer regions under the tle 1976b). The bats show almost perfect fidelity to
moderating influence of the Atlantic Ocean or the these caves. Of 3,110 bats ringed (banded) in one
Black Sea (Fig. 7.2). Not all bats in a particular pop-
ulation migrate, and they often disperse to different
locations, from 80–1,600 km away. Nathusius’ pip-
istrelle, P. nathusii, although small (6 g), undertakes
similar migrations in the same geographical region.
A similar picture is seen in North America. Figure
y
ul

7.3 shows the seasonal distribution of the tree e- J

Jun
Oc
to
be

May
r

autumn migrations
spring migrations
De
ce a ry

Jan
u be
m

r
Fe
-

b rua
ry

Figure 7.2 Migration patterns of the noctule, Nyctalus noctula, Figure 7.3 Migration of the hoary bat, Lasiurus cinereus, in
in Europe (based on Strelkov 1969). North America (from Findley and Jones 1964).
178 BATS

availability of these structures led to changes in

migratory behaviour, with fewer bats making long
migrations?
Tennessee

Sex differences
In most species, females are more likely to be migra-
spring ringing
tory than males. Where both sexes migrate females
sites typically migrate further and often start their migra-
tion earlier (Fleming and Eby 2003). This differential
hibernation migration probably occurs for two reasons. Spring
sites migration to a resource-rich summer habitat must be
Alabama
weighed against the high costs (time, energy, and
predation risk) of migration itself. The high energy
demands of pregnancy and lactation in females tip
the balance in favour of migration. This applies to
both insectivorous bats and the small number of
New World nectar feeders that have been studied
(see Fleming and Eby 2003 for examples). Females
Florida
100 km may also migrate to find nursery roosts with more
favourable microclimates: usually warmer sites that
reduce thermoregulatory costs. Males can remain in,
Figure 7.4 Migration patterns of the grey bat, Myotis grisescens, or migrate to, less favoured habitats since they
in the southern United States (adapted from Tuttle 1976b). require less food and cooler roosts due to their
more frequent use of torpor during the summer
months. For example, male hoary bats, Lasiurus
cave in winter, none were found in either of the cinereus, spend their summers in the mountains of
other caves during a 14-year study, despite high the western states of the USA and Canada. Females
recovery rates. Of 6,486 bats ringed during hiberna- tend to form their nursery colonies further east
tion only one was recovered at a different site. In (Findley and Jones 1964, Cryan 2003). Such differ-
spring they disperse to summer roosts up to 525 km ences can occur on even smaller geographical scales.
away, perhaps covering the ground in just a few Ibáñez et al. (2009) studied three species of Nyctalus
days. Fidelity to their summer roosts is equally on the Iberian peninsula and found that in all three
impressive. only males were present year-round. Females were
Most species that hibernate in caves are either only present during the late summer and autumn
sedentary or short-range migrants (up to about 500 mating season. Males of N. noctula and N. leisleri
km), although it is hard to draw a line between the appear to set up mating roosts on the migration
two conditions since there is as much variation routes of the females. N. lasiopterus is largely absent
within species as between them. Migration distance from northern Europe and is not therefore likely to
and direction may depend to a large extent on the migrate through northern Spain, so some other mech-
availability of suitable hibernation sites. Forming anism must explain the absence of females at other
primarily in limestone, natural caves are a scarce times, perhaps segregation by altitude, with females
resource in many parts of the world, so species forming nursery colonies only at low elevations.
that live in regions with few or no caves may have Local segregation and movement will be covered
to travel long distances to hibernacula. These species later in this chapter. The males of other temperate
now often use artificial underground sites, such as species (for example Pipistrellus nathusii) also set up
mines and disused railway tunnels, which are wide- mating roosts or territories on the autumn migration
spread and not so reliant on local geology. Has the routes of females.
 MIGRATION, SOCIAL STRUCTURE, AND POPULATION STRUCTURE 179

Although poorly documented, it is likely that only bats are nevertheless capable of sustained, all-night
a very small proportion of tropical species migrate. flights over water. There is some evidence that, like
For those that do, the reasons for migration are birds, bats may migrate in large groups. In species
again related to variations in food supply, and that habitually roost in large numbers, tens of
most often relate to seasonal patterns in rainfall thousands may arrive at a migration roost over-
(Fleming and Eby 2003). The few known, short- night. On completing migration, temperate bats
range migrants include several east African insectiv- will need to lay down fat again for hibernation, so
orous species described by O’Shea and Vaughan for many species there may be several periods of fat
(1980) and Vaughan (1977) that move to coastal deposition over the autumn.
habitats when their inland habitats become too
dry. Long distance migrants include the nectarivor-
ous phyllostomids Leptonycteris yerbabuenae and
Navigation
Choeronycteris mexicana (Fleming and Eby 2003) Long distance migratory flights consume a lot of
and the larger African pteropodids Eidolon helvum, energy so navigation must be accurate to avoid
Myonycteris torquata, and Nanonycteris veldkampi excessive energy expenditure. There is also no selec-
(Thomas 1983). During the wet season the African tive advantage in arriving in the wrong location at
species studied by Thomas migrated north from the the end of migration! So, how do bats navigate? We
forest zone to the drier savannah, and E. helvum know that they use echolocation for orientation and
covered as much as 1,500 km. Although food was short-range navigation between roosts and foraging
scarcer in the north, there appeared to be fewer bats sites, and that some make use of vision. Echoloca-
and therefore less competition. Richter and Cum- tion provides little information beyond 10 m in most
ming (2006) studied the same species in Zambia, species and perhaps 50 m at best, yet some bats are
where the arrival of 5–10 million bats in Kasanka known to migrate at much greater heights above the
National Park coincided with rising food availabil- ground. It was also frequently supposed that bats’
ity. After breeding they departed as the food supply vision was not up to the task. However, they clearly
diminished. In Australia, migration of the large pter- migrate with skill, and homing experiments have
opodids is often less predictable, due to the unpre- shown that bats will return to a roost even if persis-
dictability of the flowering of eucalyptus and other tently removed, often to locations well outside their
trees that provide their food. These trees flower and home range (for example Davis and Cockrum 1962).
fruit asynchronously and not every year, but pro- So what senses do they use? Navigation requires
duce huge quantities of nectar and fruit when they both a compass and a map. Birds are known to
do, so the bats must follow them around. How they make use of stars, sunset, and magnetic compasses.
locate these ephemeral riches is not known. What they use for maps is more controversial, but
the Earth’s magnetic field (by a different mechanism
Migration: how long does it take and what to the compass) and olfactory cues have been pro-
posed. Do bats use the same mechanisms? The
are the costs?
release of bats inside a planetarium with an artificial
Bats, like birds, must lay down fat to fuel migration. post-sunset glow shows that this can be used as a
The fat reserves carried by most bats are sufficient compass (Buchler and Childs 1982). Many species
for flights of several hundred kilometres at most and are also able to detect light sources equivalent to
flight speeds are relatively slow. Although there stars (for example Childs and Buchler 1981), but
have been few direct observations to support the we do not yet know if they can navigate by the
statement (see Fleming and Eby 2003 for a review stars. Holland et al. (2006, 2010, see also Chapter 3)
of what is known) it is likely that most bats take have recently shown that bats can make use of a
many nights, perhaps several weeks, to complete magnetic compass alongside these other cues. We
their migration. They will often stop over to refuel, know absolutely nothing about the maps used by
perhaps using traditional transitional roosts. Some bats. Although we know very much less about
180 BATS

navigation in bats relative to birds, interesting dif- their home. A colony refers to a group of bats, be-
ferences have already been noted. For example, the haviourally and/or genetically related to each other,
magnetic compass in birds is based on the angle of that share one or more roosting sites over a long
inclination: that of all mammals studied to date period. Colony composition, as we will see, varies
(including bats?) on polarity. Holland (2007) gives considerably from small, cohesive groups to large
an up-to-date review of what we do and do not and fluid assemblages.
know about bat navigation. In discussing reproduction, mating systems,
Bats almost certainly follow traditional migration roosting behaviour, and even foraging behaviour it
routes, but these are not at all well-studied. Furman- became clear that a roost is not a random collection
kiewicz and Kucharska (2009) have shown that the of bats. Sex and age ratios, group size, and composi-
Oder River valley in Poland is a spring and autumn tion at the roost level all vary. So far I have discussed
migration corridor for several migratory species and roosting group structure in specific contexts where
cite a number of earlier studies that suggest river the focus was on some other topic. I want now to
corridors and the coast are used in many parts of look at studies in which social structure was the
Europe. Ahlén et al. (2009) reported migrating bats focus of attention. We are sometimes given the
up to 14 km from shore in the Baltic Sea. impression that roosts and colonies are stable enti-
ties, but this is often far from true.
Many bats change roost site frequently and roost
Social and population structure
switching is most commonly seen in tree-roosting
In the last chapter we looked at roosting ecology and bats (for example Myotis bechsteinii (Kerth et al.
asked: why are bats social, what factors determine 2006), Eptesicus fuscus (Willis and Brigham 2004),
colony size and composition, and why is sexual seg- and Nyctalus lasiopterus (Popa-Lisseanu et al. 2008))
regation so often seen in the roost? In this section we although it is probably not uncommon in cave-
will probe a little deeper and from different angles to roosting bats that form larger roosting groups (for
look at behaviour and its consequences across all example Myotis capaccinii, Papadatou et al. 2008,
geographical scales, from social structure and the 2009). Frequent roost changing in Bechstein’s bat,
relationships between individual bats, to large scale Myotis bechsteinii, is driven in part by changes in
population structure and the historical background roost microclimate and the needs of the females,
to current patterns. This is a very active research area, since they prefer warmer roosts when bringing up
making effective use of rapidly developing techni- pups (Kerth et al. 2001b). Roost-switching may also
ques in molecular genetics alongside traditional and be driven by the need to avoid ectoparasites (Reck-
newly developed field methods. The topic is also ardt and Kerth 2007). Other reasons for changing
closely linked to dispersal patterns and mating sys- roost were discussed in the last chapter. As bats
tems. It is difficult to talk about any one of these move from roost to roost, roosting groups fre-
subjects in isolation and we get greater insights by quently fragment and reform many times over the
bringing them all together—but can I do it success- course of a season. This tendency towards changing
fully? I approach this section with some trepidation! association patterns in roosting bats has become
known as fission-fusion behaviour.
Roosts and colonies: who lives with who In many cases, this frequent break-up and reform-
ing of roosting groups as they move around a set of
and why?
shared roost sites may involve no more than ran-
Before I start, I need to define some terms to avoid dom associations of bats (for example in the two
confusion. In the last chapter a roost was a physical bamboo bats, Tylonycteris spp., Medway and Mar-
structure or location containing a group of roosting shall 1972), but we know that many non-random
bats. The word ‘roost’ is often used to describe the associations exist, since I have already described
bats themselves and I will use it frequently in this more than a few in previous chapters. Associations
way in the following pages. It should be clear from may be between related individuals, driven by kin
the context whether I am referring to the bats or selection, or between particular individuals that
 MIGRATION, SOCIAL STRUCTURE, AND POPULATION STRUCTURE 181

cooperate in some way, such as reciprocal altruism roosting associations did not serve as a means of
in blood-sharing vampires (Wilkinson 1985). More information transfer about foraging sites. Individual
simply, non-random associations may result from females showed high fidelity to small foraging
common needs, for example nursing females may areas, even between years, with little or no overlap
select a warm roost or one with suitable locations in between the areas used by different colony mem-
which to leave their young when they leave to bers. There was a negative correlation between
forage. roosting association and the distance between forag-
The most comprehensive studies in this area have ing sites: bats that roost together have foraging sites
been performed by Kerth and colleagues on Bech- that are close together. This may result from the
stein’s bats, Myotis bechsteinii, so I will use these as tendency of bats to minimize commuting distance
the basis of the discussion. Bechstein’s bat is a small between roost and foraging site. To benefit from
(7–14 g) gleaner that roosts primarily in tree holes both communal roosting and short commuting dis-
and forages within mostly deciduous woodland tances, bats are most likely to roost with colony
across much of Europe. Nursery colonies are typi- members using neighbouring foraging sites.
cally made up of 20–40 females and the males live a How do bats exchange information and maintain
more solitary life. Nursery colonies are closed, cohesive groups in fission-fusion societies? Kerth
matrilineal societies due to the extreme natal philo- and Reckardt (2003) introduced 20 pairs of suitable
patry of the females (Kerth et al. 2000). Kerth and and unsuitable (that is with a blocked inner
König (1999) studied a colony of female Bechstein’s entrance) bat box roosts into their study forest.
bats in a Bavarian forest that had adopted bat boxes PIT-tagged Bechstein’s bats discovered these boxes
erected on trees. PIT tags (passive induction trans- after an average of three months and attempted to
ponders) under the skin allowed individual bats to enter suitable and unsuitable roosts at random.
be identified without disturbance using magnetic However, once a pair of boxes had been discovered,
readers outside the boxes, as they moved around naïve bats were much more likely to visit the suit-
the forest. Up to 50 roosts were used each year and able box, in the company of ‘experienced’ females,
individuals roosted with most other colony mem- than visit the unsuitable box. Bats probably gath-
bers as often as expected by chance (69–85 per cent ered at their day roost towards the end of the night
over three years). However, almost all females had and were then led to the new roost by one or more
significant positive and negative associations with females that had already visited this new roost. This
15–30 per cent of other members of the colony. may be one reason why bats returning to a day roost
Based on analysis of both autosomal and mitochon- habitually fly in groups around the entrance rather
drial microsatellite markers, they concluded that than entering directly. There was no evidence for
there was no evidence for a significant relationship the reciprocal sharing of information through
between relatedness and either positive or negative exchanged roles: some individuals were frequent
association. In contrast, lactating females were much recruiters, others typically the recruited. There was
more likely to be found roosting together than exp- also no evidence for kin selection, in that recruiter
ected by chance and this appeared to be the driving and recruited were no more closely related than
force behind the associations. Females made regular random pairs of bats. The behaviour is probably
face–to-face contact in the roost and groomed each stable because all bats benefit. Recruited bats gain
other, behaviour that is likely to aid recognition. For from an increased knowledge about suitable roosts,
example, Safi and Kerth (2003) have shown that the recruiters presumably benefit from the larger
facial secretions are characteristic enough to identify group size through increased roost temperatures.
colony members. Kerth and König (1999) suggested Kerth et al. (2006) went on to show that decisions
that lactating females might benefit from mutual about roost switching are made collectively, since
warming, defence of pups, and even communal decisions reflected the information available to the
nursing, although we saw in Chapter 5 that the majority of roosting bats. However, not all bats
latter probably only occurs under rare circum- reach the same decision, perhaps due to conflicting
stances. Kerth et al. (2001a) went on to show that needs, hence the tendency for fission and fusion.
182 BATS

The addition of genetics to behavioural studies benefits such as thermoregulation, as seen in many
gives us a powerful research tool, and over the last previous studies.
decade or so our understanding of bat social sys- I have focused very much on temperate bats, since
tems has improved enormously. This discussion of this is where most research is concentrated, but
the mechanisms underlying group cohesion is a there have been studies of tropical species and
good example, and is nicely illustrated by work on these have revealed some novel results. Vonhof
greater horseshoe bats by Rossiter et al. (2002). They et al. (2004) looked at roosting associations in Spix’s
argued that the strong natal philopatry seen in disk-winked bat, Thyroptera tricolor, in Costa Rica.
females would lead to groups of highly related indi- These little bats lived in the rolled leaves of Heliconia
viduals and favour the evolution of kin-selected or Calathea plants in mixed-sex groups of 4–14 indi-
behaviours. With the benefit of the most compre- viduals. Close associations were seen between mem-
hensive long-term ringing data available for any bers of these small groups that persisted for more
species they were able to show very high degrees than a year in many cases. However, despite major
of relatedness in 15 matrilines, set against low back- spatial overlap of the small home ranges of numer-
ground relatedness in the colony. Radiotracking ous roosting groups (an average of only 0.2 ha with
over a three year period showed that females and centres typically only 100 m apart), individuals from
their adult daughters shared both roosts and forag- different groups almost never roosted together. The
ing areas, but were separated temporally: they gen- authors suggest a number of reasons why this inter-
erally foraged alone and rarely occupied a roost esting social structure may exist, but as yet there is
together. The bats do not appear to benefit from no evidence in support of any of them. It does how-
information exchange about good roosts and rich ever appear that the bonds must be quite close and
food sources, nor from shared territorial defence. driven by kinship and/or some form of cooperative
Females may simply benefit from the shared inheri- behaviour.
tance of their mother’s roosting and feeding sites, Another analytical tool has recently been applied
together with the knowledge of how best to exploit to bats. Social network analysis, originally devel-
them. An interesting twist to this story is that the oped for studying human social interactions, is
offspring only really begin to associate with their now being used with increasing frequency on wild
mothers in their second year. In their first summer animals, including bats, since it allows us to quan-
their ranges expand slowly as their bones and flight tify the strength of the associations between indivi-
skills mature, and the mothers have already begun duals in complex social groups. Patriquin et al.
to disperse to mating sites by the time they have (2010) looked at a nursery colony of northern long-
reached their maxima. eared bats, Myotis septentrionalis, and were able to
Methany et al. (2008a, b) investigated the interest- show that during gestation, females formed fewer
ing case of the long-term relocation of part of a colony and closer associations with other females than dur-
of a fission-fusion species. A significant proportion of ing lactation. (This appears to contradict earlier find-
a colony of big brown bats, Eptesicus fuscus, moved ings by Willis and Brigham (2004) and Garroway
from one locality to another 7 km away. Almost all of and Broders (2007), but this is a reflection of differ-
the females in a matriline with high average related- ences in the analytical approaches that I won’t go
ness made the move to the new location. Females into here!). This difference in behaviour between
from other matrilines with lower average relatedness gestation and lactation is not easily explained, but
showed a more even split between the old and new may be related to the need for greater social thermo-
roosting localities, suggesting that kinship was at regulation when energy demands are high during
least in part driving the decision-making process. lactation. Bats may be less choosy about who they
The mechanism is unknown, but may have involved roost with when they need to form large groups
communication at night roosts that were subse- during lactation. Younger females formed very
quently used as day roosts. In contrast, day to day much weaker associations and were more likely to
associations within a given roosting location were not form associations with a greater number of other
driven by kinship but more likely by short-term bats, but the results are consistent with associations
 MIGRATION, SOCIAL STRUCTURE, AND POPULATION STRUCTURE 183

mixed nursery

nursery

nursery male roost

Figure 7.5 Social network for Daubenton’s

swarming bat, Myotis daubentonii. Links indicate
association indices >0.25. Clusters coincide well
site
with known social groups: nursery colonies,
swarming summer roosts of males and swarming groups.
site Grey circles = males, white circles = females,
white squares = cutpoint males. When the
swarming site network shows only links with association
male roost
indices >0.75, cutpoint males are the only links
between clusters/groups.

based on matrilines, as seen in the genetics studies of spread of information and disease among colonies
Kerth and colleagues described earlier. Network of the greater noctule, Nyctalus lasiopterus.
analysis is likely to reveal more about associations
at the individual level and how they are influenced
Population structure: the small scale picture
by factors such as sex, age, relatedness, and repro-
ductive condition as well as ecological factors. This In this and previous chapters I have frequently
will not only enlighten us about the subtle complex- made statements that imply genetic knowledge of
ities of bat society, but will have value in studies of the bats being studied: for example, I’ve noted that
information exchange, gene flow, and disease trans- bats were related or unrelated, or that particular
mission. Recent work of our own (Rigby 2010, E.L. males were more likely to father offspring than
Rigby, J. Aegerter, and J.D. Altringham, in prepara- others. In a world driven by natural selection,
tion), on social structure in Daubenton’s bats, Myotis genetic relationships are all important and therefore
daubentonii, has revealed that in contrast to high an integral part of modern behavioural and ecologi-
levels of associations among females, males typi- cal studies. I have deliberately avoided going into
cally form only weak associations; but a small num- the technical detail but it has now become essential.
ber of males (‘cutpoints’) are responsible for the few Burland and Worthington Wilmer (2001) provide a
strong associations between clusters of bats with review of early work in the field. The last decade
high levels of association. Clusters typically match has seen an explosion in the quality and quantity
closely to roosts: single sex or mixed summer roosts of genetic information that can be obtained from
and swarming sites (Fig. 7.5). These few males may the smallest tissue samples—with a corresponding
have a disproportionate influence on information increase in the number of important papers making
exchange, gene flow, and the transmission of dis- use of this information to explore bat ecology and
ease. Fortuna et al. (2009) have shown, using net- behaviour. Before going on it would therefore
work analysis, how colony structure may control the be useful to provide a brief summary of the
184 BATS

methodology involved and what the principle social structure. Kerth et al. (2002) constructed gene-
genetic markers can tell us (Box 7.1). This informa- alogical trees for females from 4 nursery colonies of
tion will help you to understand the rest of this 10–35 adults, based on data from one mitochondrial
chapter if you are not already familiar with the and 11 nuclear microsatellite markers. Across all
terminology and concepts. four colonies, 75 per cent of adult females had one
Let’s return to bats and specifically to Bechstein’s or more close relatives in their colony (coefficient of
bats, because they perhaps represent one extreme of relatedness r  0.25, r = 1 for identical twins, 0.5 for

Box 7.1 Genetic markers in ecological and behavioural studies: from roost mates to population
structure

This is a very superficial look at the subject, but hopefully questions being posed, but 5–30 is ideal for most studies.
provides sufficient detail to understand what follows for Again, depending upon the questions to be asked,
those unfamiliar with the subject. A book I have found material with an appropriate intrinsic mutation rate is
invaluable as a relative newcomer to the field is Ecological chosen, rapid if you are interested in present day, fine-
Genetics by Lowe et al. (2004). scale population structure, slow if you are more concerned
with historic events. Microsatellites are the most widely
used markers in bat research. They are tandem repeats of
The tools:
short (1–4 base) nucleotide sequences (for example
We can now extract useful amounts of DNA from the GACGACGACGAC . . . n), typically having 10–50 repeats.
smallest of tissue samples. A 3 mm diameter disk of wing They are common and randomly distributed in nuclear
membrane, currently the most widely used tissue from bats, DNA. Mutation rates are high giving rise to numerous
will yield enough DNA for almost all purposes. Sampling is alleles with slightly different numbers of repeats, making
sterile, rapid, and painless. Since wing membrane is a them ideal for studying structure within populations.
metabolically active tissue the hole is repaired in as little as Because they are widespread it is possible to develop
10 days and the effects on the bat are negligible. The tissue primers for large numbers of microsatellites. They are
can be stored indefinitely in alcohol before use. Techniques widely used to investigate gene diversity, population
have been developed to extract DNA from faeces and may structure, and gene flow (see below).
well replace invasive techniques in the near future. The small It is also common to sequence specific regions of DNA, most
amounts of DNA extracted can be safely stored at 80  C. commonly mitochondrial DNA (mtDNA), for direct comparison.
Nanogram quantities are then used in PCR (polymerase chain The sequencing of mtDNA is also used in the investigation of
reaction) amplification. This reaction uses enzymes derived gene diversity, population structure, and gene flow, but the
from hot-springs bacteria to replicate specific sections of the most valuable application has been in phylogenetic and
DNA millions of times at high reaction temperatures. This phylogeographic studies (see Chapters 1 and 8).
provides sufficient DNA for further analysis. Specific regions Nuclear DNA and mtDNA are often used in parallel. The
are replicated with the aid of ‘primers’: short sections of DNA advantage of mtDNA is that it is passed down the female
that are complementary in sequence to the tissue DNA of line, is not influenced by male mating patterns and dispersal,
interest. After an initial denaturation step to separate the and does not undergo recombination. It is therefore more
two strands of DNA, the primers bind to their complementary likely to reveal population structure in migratory species and
sites and serve as starting points for replication of the chosen preserve historic patterns of gene flow, since females are
section of DNA. Repeated cycles of replication and denaturation typically highly philopatric. For example, the distribution of
produce the millions of copies needed for analysis. mitochondrial haplotypes (since the mitochondrial DNA is a
But which bits of DNA do we want and why? If we are single, circular molecule, and is haploid—not diploid as most
interested in the relationships between individuals and autosomal genes—one halotype corresponds to one allele
populations, we usually want to compare from several to present in a given individual) has been used to determine
many, single-locus, polymorphic markers. A polymorphic past patterns of post-glacial recolonization (for example
marker is one present as multiple alleles in the study Ruedi et al. 2008). Nuclear DNA is inherited by both sexes
population. The ideal number of alleles varies with the and is most useful, in the form of microsatellites, in
 MIGRATION, SOCIAL STRUCTURE, AND POPULATION STRUCTURE 185

addressing questions related to current gene flow. Both are proportions in different study populations (HS). The fixation
used to investigate mating systems and paternity. Sex-linked index (FST) is the reduction in heterozygosity within
markers, present on either X or Y chromosome,s are also of populations due to selection or drift. It is calculated directly
value, but the development of Y chromosome markers is still from HT and HS. This reduction should be small (low FST) in
very much in the early stages. large, well-mixed populations and greatest (high FST) in
small, isolated populations. If there is significant gene flow
between large populations alleles are less likely to be lost.
When considering the significance of an FST value it is
Things we may want to measure
important to take into account other factors, such as the
Genetic variation can be measured in several ways, all spatial scale over which it is measured in relation to the
potentially informative. We can measure the total amount biology of the species. Perhaps the most widely used
of genetic variation in a population (genetic diversity), the method for estimating FST and its related parameters is that
distribution of variation among populations (genetic of Weir and Cockerham (1984). This and many other
differentiation), and the amount of variation between two methods, such as Analysis of Molecular Variance (AMOVA)
particular populations (genetic distance). A wide range of can be implemented using programs such as Arlequin
techniques are available for estimating these and applying (Excoffier et al. (2005). AMOVA uses a genetic distance
appropriate statistical analysis. Fortunately, as the number matrix to examine patterns of differentiation within and
of analytical methods grows, so does the number of between populations. Genetic distance is a measure of the
relatively user-friendly (and usually free) software programs degree of similarity between two individuals, populations,
that enable you to make the most of the increasingly or even species. Arlequin is perhaps the single most useful
complex data sets. We can also estimate gene flow and versatile population genetics program, but there are
between populations, relatedness between individuals and many more and most studies will make use of several, at
populations, and we can assess paternity, or the least one of which will probably be new! Before leaving this
probability of paternity, when studying mating systems. topic I should point out that there are many variations
Underlying most of these methods are some critical on the F statistic, all with their pros and cons in the complex
assumptions, the most important being Hardy–Weinberg world of molecular and population genetics.
Equilibrium (HWE): in a large population, and in the Back to things we want to measure. Spatial effects can
absence of mutation, genetic drift, gene flow, non-random be taken into account using approaches such as Isolation
mating, and natural selection, gene frequencies and By Distance (IBD). Different methods are used to
genotype ratios remain constant from generation to investigate the relationship between linear distance
generation. Of course in real life one or more of these and genetic distance in discrete or continuously distributed
factors is pushing the population out of HWE. HWE individuals or populations. Differences in this
provides a theoretical basis for investigating these effects. relationship reflect differences in intrinsic mobility or
Here are some of the more widely used concepts and behaviour that influence gene flow. More complex spatial
measurements. Let’s start with genetic diversity. Mean analyses are too complex to discuss here, but you might
observed heterozygosity (HO) is the mean number of like to investigate programs like STRUCTURE, BAPS,
heterozygotes at a locus divided by the total number of and GENELAND!
individuals sampled. At the risk of getting too simplistic, a Gene flow generally has to be estimated indirectly and is
heterozygous individual has different alleles at a particular often derived from FST as the number of migrants per
locus (that is a nuclear, autosomal locus) on the two generation Nm. This is independent of population size since
homologous chromosomes inherited from its parents. In a drift is slow in large populations so one migrant can
large population with high gene flow HO should be high. effectively counteract it. However, in small populations with
The mean expected heterozygosity (HE), often referred to high drift rates the same single migrant is now a significant
as gene diversity, is an estimate of the proportion of addition to the population.
all individuals in the population that would be The loss of genetic variation due to a population crash is
heterozygous, based on allele frequencies from the sampled known as a bottleneck. Population size may recover quickly,
population. If observed and expected heterozygosity are but the recovery of genetic variation will be much slower.
different, something interesting may be happening in your Bottlenecks are an extinction threat due to increased
population dynamics. We can investigate population inbreeding, alongside the purely demographic threat.
differentiation by estimating total gene diversity (HT) and its Bottlenecks may also be interesting and illuminating
continues
186 BATS

Box 7.1 (continued)

episodes in the history of a species. A number of methods are Of course, this works on sequence data, and is most easily
available to detect past bottlenecks through decreases in and commonly applied to mtDNA sequence data. One
allelic richness and heterozygosity. approach mentioned later in the chapter is Parsimony
Relatedness and maternity/paternity can be confirmed or Analysis, which builds a phylogenetic tree that requires the
at least estimated with stated degrees of probability by fewest mutations. Another common approach is the
comparing genotypes. maximum-likelihood method that maximizes the probability
of observing the data, given a model of DNA evolution. As
described in Chapter 1, if we know mutation rates, a
timescale can be put on these trees too. It is therefore
Historical processes
possible to learn a great deal about the events behind current
The phylogenetic approaches described in Chapter 1 can population structure, which makes it that much
be combined with population genetics techniques to yield easier to interpret data in terms of current behaviour and
a very powerful tool for investigating historical processes. ecology.

Long Newton

Ellerburn
Ryedale Windy Pits
Wintringham
70 Harpham
Summerbridge
Wintringham Kirkburn
Low
Catton Kilnwick
assignment to Windy Pits (%)

Woodfield Woodfield
Ellerburn 20 km South Cave
60
Low Catton
Harpham South Cave
Long Newton

Kilnwick
50
Summerbridge
Kirkburn

40
20 30 40 50 60 70
distance from Windy Pits (km)

Figure 7.6 Percentage of individual Natterer’s bats, Myotis nattereri, from different summer nursery roosts, assigning to the Ryedale swarming sites.
Arrows on the map indicate known movements of ringed bats. From Rivers et al. (2005), with permission from John Wiley and Sons.
 MIGRATION, SOCIAL STRUCTURE, AND POPULATION STRUCTURE 187

mother and daughter and for siblings, 0.25 for half- daughters may inherit foraging areas from their
siblings), with numerous mother:daughter and sis- mothers.
ter pairs. Not a single female from another colony Few other species have been studied in such
appeared to have joined any of the colonies over a detail, but female natal philopatry has been well
five-year study period. In a separate study, Kerth documented in the evening bat, Nycticeus humeralis
et al. (2000) showed that most of the 10 colonies (Wilkinson & Chapman 1991), the brown long-eared
studied had only 1–2 matrilines and estimated that bat, Plecotus auritus (Burland et al. 1999), the com-
only 1 female would switch colony every 9–14 years mon vampire bat, Desmodus rotundus (Wilkinson
(mtDNA FST = 0.68, an exceptionally high value). 1987), and Daubenton’s bat, Myotis daubentonii
Strong female philopatry is common to many bats (Angell, 2008). In most species studied, although
(for example Myotis myotis, Castella et al. 2001, philopatry is common, group composition is more
Plecotus auritus, Burland et al. 1999, Rhinolophus fer- flexible and genetic structure is weak or absent (for
rumequinum, Rossiter et al. 2000b), but this degree of reviews see McCracken 1987, Wilkinson 1987).
isolation had not been reported before. It is difficult What are the males up to? To go back to Bech-
to say how widespread it may be since few bat stein’s bat, Kerth and Morf (2004) showed that males
populations have been studied in such detail. How- occupying the same forests as the females were
ever, even in the absence of intensively collected PIT almost exclusively solitary roosters. Adult males
tag data, other studies have been able to show inter- were never found roosting together, and on only
colony movement (for example Plecotus auritus, 37 out of 515 census days was a male found roosting
Entwistle et al. 2000, Rhinolophus ferrumequinum, with females. Males were most frequently found
Rossiter et al. 2000b), suggesting that such a high with females (32/37) during the mating season
degree of isolation may be unusual. (August and September) and never during lactation
Despite this, the average relatedness r within a ( June and July). However, these resident males ap-
colony of Myotis bechsteinii was only 0.02. Average peared to father few offspring.
relatedness is typically low in philopatric, temperate Safi and Kerth (2007) investigated the causes of
bat species (for example Plecotus auritus, Burland sociality in males of different species. Due to a lack
et al. 2001, Rhinolophus ferrumequinum, Rossiter et al. of data, their analysis was restricted to 45 northern
2000a, Myotis nattereri, Rivers et al. 2005). Although hemisphere temperate species, but the results may be
this result may seem surprising at first, it simply typical of all bats. They concluded that solitary roost-
shows that few females within a colony share a ing was the ancestral condition but that sociality had
father: low levels of polygyny and reproductive evolved in several lineages, primarily due to the
skew are characteristic of the mating systems of benefits of information transfer about food. Social
these bats. For example, genetic data show that the species were typically those that fed on ephemeral
primary mating system in Plecotus and Myotis spe- insects in the open air. Information transfer is likely
cies is swarming, where males and females from to be of greatest benefit to these bats and their low
large catchment areas come together to mate prior frequency, high intensity echolocation calls are those
to hibernation, a system that promotes gene flow most likely to be heard by conspecifics foraging
(for example Kerth et al. 2003, Veith et al. 2004, nearby. These species also have the low flight costs
Rivers et al. 2005, 2006, Furmankiewicz and Altring- associated with high aspect ratio wings and low
ham 2007). Although many sister pairs were found wing loading and are thus able, when necessary, to
by Kerth et al. (2002), most were half-sisters, as forage over wider areas to reduce competition.
expected.
What are the likely benefits of living in such a
closed society? We have already seen that there is
Population structure: a bigger picture
no evidence for kin selection, but the bats groom
each other, they share information about new roosts, Of course roosting behaviour and social structure
they help each other to keep the roost warm, and are not necessarily fixed, but may be influenced by
188 BATS

environment and history. For example, Senior et al. sites (for example Furmankiewicz and Altringham
(2005) described major differences in social structure 2007). Since there is now strong evidence that hiber-
and behaviour over very modest distances in Myotis nation sites are frequently also swarming sites
daubentonii. Kerth et al. (2008) therefore asked the where bats mate, this presents a significant oppor-
ambitious question: is the social structure of Bech- tunity for outbreeding that is likely to have a major
stein’s bat the same across its entire geographical influence on population structure.
range? By sequencing two regions of mitochondrial The higher mitochondrial genetic diversity at
DNA (mtDNA) they were able to show how Bech- swarming sites compared with summer sites re-
stein’s bat recolonized Europe from a refugium in ported for M. bechsteinii (Kerth et al. 2003) has also
the Balkans after the last ice age. They then used been found in Plecotus auritus (Veith et al. 2004),
data from seven nuclear (autosomal, that is not on supporting the view that bats gather at swarming
the X or Y chromosomes) and two mitochondrial sites from many summer colonies. The high levels of
microsatellite loci to compare population structures nuclear gene diversity found within summer colo-
across the range. Both Balkan and European popu- nies (Kerth et al. 2003, Veith et al. 2004, Rivers et al.
lations showed greater differentiation in mtDNA 2005) strongly suggests high levels of gene flow
relative to nuclear DNA right across their range, between these apparently isolated colonies. To
suggesting strong female philopatry and male dis- date, in only one study (Rivers et al. 2005) has
persal with outbreeding, presumably at swarming nuclear gene diversity been found to be higher at
sites. Mitochondrial genetic diversity was greater swarming sites relative to summer sites and the
and relatedness lower at swarming sites relative to difference was small. However, both mitochondrial
nursery colonies, supporting this conclusion. and nuclear gene diversities are nevertheless high at
We have just introduced two additional elements all sites. High diversity could be due to mating at
that must be considered in any detailed discussion swarming sites, or because males move between
of population structure: mating systems and post- summer colonies. Paternity studies have been car-
natal dispersal patterns. The emphasis of this section ried out on juveniles at summer colonies of several
has so far been largely on females and we have seen species. Low numbers (3–15 per cent) of within-col-
that in most species they are highly philopatric. In ony fathers were found in Myotis myotis, M. bechstei-
common with most mammals (Greenwood 1980) it nii, M. daubentonii, and P. auritus, suggesting that
is the males that disperse from their natal colonies, extra-colony mating at swarming sites is more typi-
but how far they go varies. In the case of Bechstein’s cal (respectively, Petri et al. 1997, Kerth and Morf
bat, 45 per cent of the solitary males captured in the 2004, Angell 2008, Burland et al. 2001). The case of
vicinity of a nursery colony were offspring from M. daubentonii is more complex (see Senior et al. 2005
other colonies (identified by their mtDNA haplo- and Chapter 5), but again the norm appears to be
types), but most of these were from neighbouring extra-colony mating.
colonies 1–7 km away (Kerth et al. 2000). Rivers et al. Rivers et al. (2005, 2006) have so far conducted the
(2006) show that male M. nattereri also do not stray only detailed study that combined genetics and
far from their natal area, even if they do disperse mark-recapture on bats moving between known
from their natal roosts. Male M. daubentonii disperse swarming and summer sites. They found that the
along the rivers on which they roost and feed, but overall FST among summer colonies was very low
dispersal distances appear short, rarely exceeding a (less than 2 per cent). FST is a measure of the reduc-
few tens of kilometres (Angell 2008). However, tion in heterozygosity within populations due to
although the males of many species do not appear selection and/or genetic drift (see Box 7.1). This
to establish their summer homes very far from their reduction should therefore be greatest in small,
natal area, they may nevertheless undertake much isolated populations. The low FST therefore indicates
larger scale seasonal migrations to their hibernation high levels of gene flow despite high levels of phi-
sites as discussed earlier. Even highly sedentary lopatry. Mating within small independent popula-
species, such as P. auritus have been shown to tions would lead to a greater FST. However, the
undertake journeys of up to 88 km to hibernation story does not end there. The overall FST, although
 MIGRATION, SOCIAL STRUCTURE, AND POPULATION STRUCTURE 189

small, was significantly different from zero and pair- mark-recapture evidence that females are philopa-
wise comparison across 11 summer colonies re- tric. Mean within-colony relatedness values in this
vealed significant differences, despite the fact that and other studies of swarming species are all low (2–
the colonies were only 6–130 km apart and bats were 6 per cent, Kerth et al. 2002, Burland et al. 2001, Veith
known to travel at least 60 km (and probably much et al. 2004). However, as discussed above, if colony
further) between summer roost and swarming site. members are not paternally related, the number of
As we will see a little later, most studies of bat popu- pairwise relatedness values of approximately zero
lations have revealed little structure at this small would make the overall mean relatedness low, even
scale, with migratory species in particular showing if there are close matrilineal relatives in the colony
near panmictic populations (that is populations in and high mitochondrial DNA structure (for example
which mating appears to be random across all indi- M. bechsteinii, Kerth et al. 2000). Even related females
viduals). But let’s not jump ahead: for the moment arriving at a swarming site on the same night are
we’ll stay with those species that do not undertake unlikely to mate with the same male, since hundreds
large migrations. A significant FST implies some pop- are present each night (for example Rivers et al. 2006,
ulation structure. What form does this structure take? Glover and Altringham 2008). Male-mediated gene
In the case of M. nattereri there is a strong correlation flow through mating at swarming sites, combined
between proximity to a swarming site and the pro- with female philopatry, can explain all or most of
portion of females from a nursery colony that can be the observed genetic structure in swarming species.
genetically assigned to the swarming site (Fig. 7.6). As in most other mammals, male-mediated gene
The closer a nursery colony is to a swarming site, the flow is likely to be one of the key factors determin-
more likely that the bats will visit that as opposed to ing population structure in bats whatever the mat-
some other swarming site. ing system. The noctule, Nyctalus noctula, is a
Over this relatively small spatial scale isolation by migratory species and mating takes place when
distance (IBD) was only marginally significant in M. highly philopatric females migrate through the ter-
nattereri, but much more significant when measured ritories of males on their way to hibernation sites.
across the whole of England (Glover et al. 2011). IBD, The males occupy mating roosts in tree holes and
as its name implies, is a simple measure of the call to attract the females. Individuals of both sexes,
degree to which individuals of a species become from different colonies, hibernate together (Petit and
more genetically isolated as the distance between Mayer 2000). Petit and Mayer (1999) found that
sampling locations increases. IBD is more marked over 20 per cent of the variation in mtDNA was
in more sedentary species such as P. auritus (Burland distributed between the 13 nursery colonies sam-
et al. 1999) and less so in the more mobile M. dau- pled across Europe from western France to Kazakh-
bentonii (Angell 2008, Frith 2010). In migratory spe- stan, but between-colony variation only accounted
cies IBD may be negligible over very large scales. for 0.5 per cent of the nuclear microsatellite markers,
Linear distance is potentially a crude measure of reflecting the very marked contrast between female
separation, since other factors such as behaviour philopatry and male dispersal—in this case dispersal
and geography are likely to have an important influ- over considerable distances. Nevertheless, the pop-
ence. Glover et al. (2011) found structure amongst ulation was not panmictic, although the structure
swarming sites within the approximately 40 km could not be explained either by non-random dis-
diameter of the Yorkshire Dales National Park. persion along migration routes or by distance. There
This genetic structure could not be explained on was some evidence to suggest that the Alps may act
the basis of IBD, but appeared to be related to the as a barrier to migration and that the ‘panmictic
summer catchment areas of the swarming sites and unit’ may be over 3,000 km wide: an astonishingly
migration routes along river valleys. large area for such a small mammal. Petit et al. (2001)
Mean within-colony relatedness among females were able to show that this huge unit could be
in M. nattereri (Rivers et al. 2005) was significantly explained by random male dispersal alone,
higher than between-colony relatedness, supporting although other explanations were possible.
190 BATS

Female philopatry and male dispersal (and there- and Altringham 2007), is probably very common in
fore male-mediated gene flow) may be the norm, but temperate species. It can easily lead to the same
there are exceptions. Indeed, because most studies population structure seen with permanent male dis-
have been of temperate bat species that rarely have persal. Even in the absence of direct evidence for
year-round roost sites, alternative strategies may be such movements in other swarming species, this
more widespread than the literature suggests. Per- pattern is the most likely. For example in M. bech-
haps the most important reason for males to dis- steinii, males disperse only over short distances and
perse is to reduce the competition for mates. most of the paternity appears to come from males
Dispersal in females is more likely to have evolved encountered at swarming sites, rather than local
to avoid inbreeding (Clutton-Brock 1989, Clobert males (Kerth & Morf 2004).
et al. 2001). If males are long term residents in a Although I have so far discussed species that
roost then females may need to disperse to avoid show marked population structure, most studies
inbreeding. Dechmann et al. (2007) describe an inter- of bat populations have uncovered little structure,
esting example in the neotropical bat Lophostoma particularly at small scales, with migratory species
silvicolum. Males defend groups of females in cav- showing near panmictic populations. However,
ities they excavate in termite mounds, an apparently beyond the general conclusion that many popula-
typical example of resource defence polygyny. Per- tions approach panmixia lie some interesting results
haps because excavating these roosts is a major task that are well worth investigating. McCracken and
for a male, once installed they may stay in residence Gassell (1997) studied Tadarida brasiliensis in Califor-
for two to three years. Young females reach sexual nia, Texas, Arkansas, and Florida and found little or
maturity in as little as six months, so dispersal to no differentiation between migratory and non-
avoid mating with their fathers makes sense. migratory populations of T. b. cynocephala and even
Nuclear microsatellite and mtDNA analysis showed between the two subspecies T. b. cynocephala and
that within these small harems (four to eight adult T. b. mexicana. Sinclair et al. (1996) studied the little
females, zero to five juveniles) both mitochondrial red flying-fox, Pteropus scapulatus, a species with a
and nuclear gene diversity was high, indicating that distribution across most of eastern and northern
the probability that any two bats were related was Australia. Genetic differentiation was minimal
very small: young females do leave the natal roost. between six sub-populations sampled across its
This is in complete contrast to the highly philopatric entire range and the entire population was close to
groups of related females common to the species we panmictic. Similar results were reported for two
have discussed so far. Parentage analysis revealed other Australian flying foxes, the black (P. alecto)
probable parent–offspring pairs for about half of the and grey-headed (P. poliocephalus) by Webb and
160 bats tested, but only 8 of these pairs were ever Tidemann (1996). In complete contrast, the nine po-
found using the same roost—and only 5 simulta- pulations of the Australian ghost bat, Macroderma
neously. Clearly in this species, at least in this gigas, studied by Worthington-Wilmer et al. (1999)
instance, both sexes disperse. In tropical species are highly structured. This large, endangered carni-
that do not benefit from some form of cooperative vore was once widespread in Australia but is now
behaviour, female dispersal may be widespread. found only in isolated populations across the north-
It is implicit in many studies that males disperse ern fringe of the continent, after a range contraction
permanently to give rise to the genetic patterns we that appears to have progressed very rapidly in the
have discussed. There is certainly evidence for per- last century. The contraction is thought to have been
manent dispersal in some species such as N. noctula due to the increasingly arid climate in central Aus-
(Petit et al. 2001, Petit and Mayer 1999), but dispersal tralia. Each of the nine populations studied was
patterns are not known in most other species. Tem- made up of several colonies with estimated popula-
porary dispersal of both sexes for mating and for tion sizes of 60–1,200 individuals. There is major
hibernation, and the return of both sexes to their genetic differentiation between populations and evi-
natal area as seen, for example, in M. nattereri (Riv- dence for reduced gene flow within populations,
ers et al. 2005, 2006) and P. auritus (Furmankiewicz particularly those on the contracting southern edge
 MIGRATION, SOCIAL STRUCTURE, AND POPULATION STRUCTURE 191

of the distribution. These southern populations also Differentiation was also found between popula-
show less genetic diversity than those located fur- tions of the migratory Miniopterus schreibersii nata-
ther north and there is less gene flow between them. lensis across South Africa (Miller-Butterworth et al.
Movement may be restricted because of the scarcity 2003). A surprising 56 per cent of the variation in
of underground roost sites with an appropriate mtDNA occurred between three populations in the
microclimate, particularly at the southern limits of west, south, and east of the country (Fig. 7.7). More
the range. A comparison of mitochondrial and surprisingly, the microsatellite markers revealed the
nuclear DNA suggested that gene flow, where it same three populations, suggesting that both males
does occur, is strongly male-biased, as we might and females are highly philopatric. Across their
expect. broad geographic range Miniopterus species migrate
Wilkinson & Fleming (1996) sequenced an to hibernate in caves, and in common with many
approximately 300 base-pair region of mtDNA in other species they show high individual fidelity to
Leptonycteris yerbabuenae to investigate population hibernacula. It appears that within the three South
structure of this migratory bat from the south-west- African populations males and females migrate to
ern USA down to Mexico. Their results lend support the same hibernacula. Mating is believed to take
to the stable isotope studies described at the begin- place in the hibernacula prior to hibernation, giving
ning of the chapter in suggesting that there are the opportunity for mating and gene flow between
coastal and inland migration routes. There was bats from different colonies within, but crucially not
greater genetic differentiation between bats sampled between, populations. Mating behaviour therefore
on the two routes, typically only 50–200 km apart, has some similarity to swarming. As would be pre-
than between bats sampled along the full 3,000 km dicted from this complex behaviour there was no
length of the migration corridors. Bats appear to fall simple or strong relationship between genetic differ-
largely into two populations when moving north, entiation and geographical distance either across the
one that migrates along the coast, following the country or within the large eastern population. Evi-
spring flowering cacti, and another that follows dence was also found for morphological differentia-
the summer-flowering agaves along the foothills tion, with bats from the north and east that may
of the Sierra Madre. The southern migration of undertake migrations in excess of 500 km, having
both groups appears to follow the inland corridor higher aspect ratio wings, an adaptation for more
(Fleming et al. 1993). Significant isolation by distance efficient flight. Finally, there was some correspon-
was also detected, showing that movement and in- dence between population structure and the princi-
terbreeding were to some degree restricted across pal biomes of the region. The western, southern,
the species’ range. and eastern populations occupy Nama/Succulent

savanna

grassland
western population

karoo
eastern population
Figure 7.7 Three genetically distinct populations (based on
mitochondrial and nuclear DNA) of Schreiber’s bat,
fynbos Miniopterus schreibersii, in southern Africa. Sampling
southern population locations shown by white, grey, and black circles, occupying
different biomes. From Miller-Butterworth et al. (2003), with
permission from Nature Publishing Group.
192 BATS

Karoo, Fynbos, and Savanna habitats respectively, prompted Petit et al. (1999) to investigate the popu-
habitats that have remained largely unchanged for lation structure of the European noctule, Nyctalus
at least a million years and perhaps very much noctula. This is a relatively large, migratory species
longer. Karoo is semi-arid shrubland, Fynbos is an that lives in forests throughout Europe. Based on the
evergreen shrubland with winter rains, and analysis of polymorphisms in two mtDNA regions,
Savanna is a summer rainfall habitat of woodland, they concluded that noctules had recolonized Eur-
shrubland, and grassland. As bats adapted to local ope from several refugia, but that there was no
habitat structure and patterns of insect availability, evidence for a population bottleneck. Genetic varia-
selection would favour philopatry and progressive bility was generally high and, surprisingly, the least
differentiation in both sexes. This brief historical genetically diverse populations were those on the
perspective serves as an appropriate introduction southern boundary of the species’ range, close to the
to the next section, the influence of history on cur- presumed refugia. The high variability appears to be
rent population structure. due to rapid evolution of the mitochondrial region
studied, as the population expanded in size and
shifted its range north with changing climate and
Population structure: historical effects
forest expansion. The southern and western popula-
Depending upon the questions that interest us most, tions on the fringes of the range may have colonized
we can view historical effects on population struc- these areas not directly from refugia, but at a later
ture as a fascinating insight into the past, or as date from the north and east. If southern and west-
complicating factors that make interpretation in ern regions offer suboptimal habitat, the low genetic
terms of recent ecology and behaviour more difficult diversity may be a consequence of high mortality in
and uncertain. Either way, they must be studied! the colonists and small populations resulting from
Perhaps the most studied historical event, and one low reproductive success.
that is a good illustration of the importance of his- Castella et al. (2001) investigated patterns in both
tory, is the recolonization of Europe after the most mitochondrial and nuclear DNA in Myotis myotis
recent glaciations. As the ice spread south it pushed across the Alps. An analysis based on 15 microsatel-
all flora and fauna, bats included, before it. At the lite loci revealed little structure (FST = 0.02), pointing
Last Glacial Maximum 18,000 years ago bats were to the now familiar pattern of male-mediated gene
confined to ‘glacial refugia’ in the Iberian Peninsula, flow. However, mtDNA patterns pointed to three
Italy, and the Balkans. Range contraction and dis- distinct lineages that probably arose in separate gla-
placement, possibly accompanied by a dramatic cial refugia. After the retreat of the ice these bats
decline in numbers, was later followed by recoloni- appear to have met in a contact zone in the Alps.
zation and perhaps expansion of both range and Because of the high degree of natal philopatry
population size. It would be surprising if these dra- among the females, this historical pattern is pre-
matic events did not have profound effects on pop- served in the mtDNA. Parsimony analysis put the
ulation structure: effects that may have persisited to 16 haplotypes into 3 clades. North and west of the
the present time and which must be unravelled from Alps all of the bats in eight colonies fell into a single
current processes that determine population struc- clade (A), a further four colonies within the Alpine
ture. Predictions include a loss of genetic variability region were composed of bats from two to three
due to expansion from small, remnant populations clades (A, B, and C). The single colony sampled
(population bottlenecks) and the expectation that south of the Alps in northern Italy had bats only
genetic diversity would decrease with increasing from clades B and C. The most divergent colonies
distance from refugia due to the founder effect (He- (in both mtDNA and nuclear microsatellites) were
witt 1996, 2000). The founder effect is a reduction in those separated by the Alps. This and other evi-
genetic diversity due to expansion and colonization dence strongly supports the view that three geneti-
by small numbers of individuals from a parent pop- cally distinct populations evolved in ice age refugia
ulation that may already have a limited gene pool. and have since met in this region as their ranges
Evidence for such bottlenecks in other animals expanded, with the Alps acting as a significant
 MIGRATION, SOCIAL STRUCTURE, AND POPULATION STRUCTURE 193

barrier to dispersal. Ruedi and colleagues have between northern and western colonies, which are
continued to explore the history of M. myotis in made up primarily of bats of a single clade (A), and
more recent papers. Ruedi and Castella (2003) were those in the Alps, Italy, and Greece, which typically
able to show that the vast majority of colonies sam- have bats from two or three clades. Italian colonies
pled right across Europe, from Spain to Poland, also show a high degree of mtDNA differentiation.
belonged to clade A from a presumed refugium in A detailed analysis of the data led the authors to a
Spain (Fig. 7.8). number of conclusions. The Italian population has a
Ruedi et al. (2008) went on to build a more com- complex history, with endemic populations that sur-
plete picture by analysing both mitochondrial and vived the ice age and mixed with post-glacial im-
nuclear data obtained from almost 600 bats across 29 migrants from Greece and the Balkans. These bats
colonies, most notably sampling in Italy (Fig. 7.8). have remained highly philopatric, not only to Italy,
Seven major clades were found across Europe, five but to their locale. However, their population struc-
of which were found in Italy. One of the five was ture cannot be explained by isolation by distance
unique to southern Italy, the others are also found in and there is no evidence for changing diversity
the Alpine region or beyond. Two clades were with latitude. The Alps appear to have been a
unique to Greece. There is a dramatic contrast major barrier to movement, but other potential

A
A+C
A+B+C

A+B
B+C
B+F

CladeA B+G

B
A+D

Clades E+D
B+C
F
CladeA

Clades D+E

CladeF

Figure 7.8 Distribution of mtDNA haplotypes of the greater mouse-eared bat, Myotis myotis, in Europe with the probable post-glacial recolonization pattern
from southern refugia. Adapted from Ruedi et al. (2008) and Ruedi and Castella (2003), with permission from John Wiley and Sons.
194 BATS

barriers appear to have been less important. For the Pyrenees from Spain to colonize NW Europe.
example, bats in the colony on the island of Mallorca, Both mitochondrial and nuclear data support these
300 km off the Spanish coast, have near identical conclusions, although as expected the effects are
alleles to those of mainland Spanish colonies (Ruedi more marked in mtDNA patterns.
and Castella 2003). This is in sharp contrast to the 14 There are many other studies in which genetics
km Strait of Gibraltar that neatly separates M. myotis have been used to probe historical and geographical
in Europe from M. punicus in North Africa (Castella et questions, but I want to leave them for the following
al. 2001). Similarly, the Alps act as an effective barrier chapter and discuss them in the wider context of
to clades B and C, yet clade A appears to have crossed biogeography.
 CHAPTER 8

Biogeography, macroecology,
community ecology, and the
interactions between bats and
other organisms

Biogeography and macroecology. Species–area relationships. Molecular genetics and patterns of speciation
and colonization. Latitude, elevation, and species richness. Community ecology, niche partitioning,
modelling community structure. The interface between macroecology and community ecology. Bats as
predators and as prey: arms races. Bats and plants—co evolution and mutualism, New World versus Old.
Pollination and seed dispersal.

spatial scales and investigates patterns of abun-

Biogeography and macroecology
dance, distribution, and diversity. The two disci-
Patterns in the distribution of animals and plants plines therefore overlap considerably and have
have been recognized for a very long time, and many aims and methods in common. They are the
research aimed at explaining such patterns plays a final step in the expansion of both temporal and
major role in modern ecology. This research has a spatial scales of measurement as we move from
new and urgent goal—that of charting, understand- studying individual behaviour through population
ing, and predicting the influence of climate change ecology to evolution. The boundaries are of course
and other human impacts on our fauna and flora there for convenience and it is all a continuum. You
(see for example Whittaker et al. 2005, Whittaker and will find overlapping material in the previous chap-
Fernández-Palacios 2007). The richness and diversity ter, where the emphasis is on population structure
of bats make them an ideal group for biogeographical of individual species in relation to behaviour.
and macroecological studies. However, they have Although the study of bats has lagged behind that
been somewhat neglected until quite recently, per- of many other taxa, significant advances have been
haps due to the difficulties inherent in quantifying made and we can still use bats to illustrate some of
population and community structure in small, noc- the key principles, methods, and findings.
turnal, and highly mobile mammals. A short book on
bat biogeography by Findley (1993) and the growth
Species–area relationships and related topics
of the field of macroecology (for example Brown
1995, Gaston and Blackburn 2000, Blackburn and The relationship we have to look at first, since it
Gaston 2003, Gaston et al. 2008) have probably underlies many other patterns, is the species–area
helped to stimulate research, since work in this relationship. We have known for a very long time
area has accelerated over the last decade or so. Bio- that we tend to find fewer species on small islands
geography has been defined as the study of biodi- than larger ones, and that both tend to have fewer
versity in space and in time. Macroecology is species than comparable areas of mainland. It is also
typically described as the study of relationships apparent that this principle applies not only to true
between organisms and their environment at large islands, those surrounded by water, but also to

195
196 BATS

bodies of water surrounded by land, and to moun- widely used in theoretical and practical ecology.
tain tops, isolated forests, and so on. We can go Much of the recent debate and complexity in the
further and say that even within the same essentially field has revolved around the possible over-simplis-
homogeneous region the number of species in- tic nature of this model. It works well at intermedi-
creases with area. Understanding this relationship ate scales, but does it explain the whole picture? Is a
is central to investigations in macroecology: species– more complex model needed to explain deviations
area relations (SARs) vary with organism, ecosys- from a power law at large and small scales? At the
tem, geographical region, and the scale of investiga- very largest scales it has been suggested that
tion. The many theories that aim to explain variation the curve should approach a plateau as all the avail-
in the distribution and abundance of species, for able species are recruited. At the other end of the
example in relation to latitude and altitude, make scale must the model explain the small island effect:
specific predictions about the shape of the SAR. the observation that species richness appears to
SARs are therefore central to models that aim to be independent of size on the smallest islands?
explain the ecological, evolutionary, geographic, Lomolino (2000) discusses these possibilities in
and climatic mechanisms that determine macroeco- some depth, but Dengler (2009) argues strongly in
logical patterns. The same concepts are put to wide favour of retaining the simple power law. Many
practical use in conservation, for example for pro- others have entered the debate, but this is all getting
ducing biodiversity maps, identifying hotspots, and a little technical for current purposes and beyond
predicting the consequences of human impact on what can be discussed using data from bats. Let’s
the environment. Because of its importance and its get back to basics.
complexity, this field has been both active and Three types of primary mechanism are likely to
something of a minefield: in the last decade a con- explain species–area relations. All will probably
stant stream of papers has been published, propos- play a role in most situations, but their relative impor-
ing (to the non-specialist at least) a bewildering tance will vary with ecosystem, taxon, geography,
number of methods, models, and rationales for and so on, hence the observed widespread variation
studying and explaining SARs (and the related Spe- in the constants C and z in the power function.
cies Sampling Relationships, SSRs). For those inter-
1. Habitat diversity: bigger areas contain more
ested in pursuing this area further I suggest starting
habitats and can therefore support more species.
with a characteristically readable and insightful
2. The number of species will also be determined
review by Lawton (1999), who asks the question
by the balance between the rates of immigration
‘Are there general laws in ecology?’. It’s worth
reading in its entirety, since the earlier sections put
those on macroecology nicely into context. Rosenz-
weig (1995), Lomolino (2000), and Dengler (2009)
S = number of species
and the references within them will give you some
A= area
idea of the breadth and depth of the topic. I will stick
log number of species

to a basic account!
Arrhenius (1921) described species–area relation-
ships using a power function: slope

S ¼ CAz
S = CAz
Where S = species richness, A is area, and C and z are
constants (Fig. 8.1). C and z vary with taxa and
intercept
ecosystem in often poorly understood ways. The
exponent z (the slope of the line on a log–log plot)
usually lies between 0.1 and 0.5 but C (the intercept) log area
can span orders of magnitude. Although simple, this
model has survived the test of time and is still Figure 8.1 The species–area relationship.
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 197

and extinction. Small islands are small targets for species richness against area for bats across tropical
immigrant species, and small populations on and temperate regions and from tropical forest
small islands will be more vulnerable to extinc- alone—not surprisingly bats obey the species–area
tion than the larger populations on larger islands law. Because the relationship is best described by a
(equilibrium theory, MacArthur and Wilson power function it is usual to plot species–area rela-
1967). Increasing remoteness from the source of tions on log scales to yield a linear relationship.
new species will also decrease immigration rate. A number of studies have described species–area
3. Speciation rates increase with island size and on relationships for bats on islands, such as Wright
large islands speciation may make a greater con- (1981), Ahlén (1983), Lomolino (1984), Lawlor
tribution than immigration to species richness (1986), Johansson and de Jong (1996), Ricklefs and
(Losos and Schluter 2000). Finally there may be Lovette (1999), and Carvajal and Adler (2005). In a
an unwanted, artifactual contribution due to wide-ranging survey, Lomolino (1984) used data
sampling, since sampling larger areas (or sam- from 19 archipelagos across the world to test the
pling for longer) will sample more individuals applicability of the equilibrium theory to mammals,
and therefore an increased number of less abun- including bats. The archipelagos included inland,
dant or elusive species. coastal, and oceanic islands and mountaintops. He
confirmed that species richness was positively cor-
Can studies of the distribution of bats illustrate or related with island area and negatively correlated
test any of these concepts? Figure 8.2 shows plots of with isolation. As predicted, z values increased with

1000

tropics
No. of species

100

temperate
10

1 10 100 1000 10000

area (km ´ 10 )
2 –3

1000

Neotropica
No. of species

S.E. Asia
Borneo
100
Jawa Sulawesi Africa
Mindanao New Guinea
Sumatra
Luzon
Negros

Figure 8.2 The number of species plotted

10 against area for (a) temperate and tropical
10 100 1000 10000
regions and (b) tropical forests (from Findley
rainforest area (km2 ´ 10–3) 1993).
198 BATS

isolation and decreased with the vagility (mobility) influenced by the key geographic feature, area, but
of the mammal order, with bats having by far the their different responses to habitat diversity reflect
lowest z value. This was done by using data for the influence of differences in biology. Bats, and to a
Malaysia, with vagility expressed as the percentage lesser extent birds, show relatively little habitat spe-
of genera in each order found on both sides of the cialization and have lower population densities rel-
Wallace Line. The Wallace Line (named after Alfred ative to the other groups, hence the dominant effect
Russel Wallace) is a zoogeographic boundary of area on species richness. Larger islands support
between species of Indo-Malaysian and Australa- larger populations that are probably better able to
sian origin that passes between Bali and Lombok withstand environmental challenges and popula-
and Borneo and Sulawesi: the boundary between tion fluctuations. Butterflies and amphibians/rep-
two tectonic plates that brought once distant islands tiles, with high population densities, are less prone
together. A similar study was performed by Lawlor to extinction, so species richness is less dependent
(1986), who probed a little deeper into the differ- on island area and more dependent on habitat diver-
ences between bats and non-volant mammals. He sity, since many are specialists. There are some
was primarily interested in the non-volant mam- sweeping statements in this brief summary, but the
mals, but in making comparisons with bats provides evidence presented in the paper is persuasive. The
us with useful information. In a nutshell, the species strong dependence on area seen with butterflies
richness of bats conforms to equilibrium theory with using simple correlation is probably due to the effect
immigration and extinction being important deter- of area on habitat diversity, which makes these two
minants. In contrast, patterns in non-volant mam- factors difficult to separate fully in this and most
mals appear to have arisen from the fragmentation other studies.
and partial extinction of ancient continental faunas, Presley and Willig (2008) and Willig et al. (2009)
against a background of low immigration. Bats and focused their efforts on the bats of the Bahamas and
non-volant mammals therefore show quite different Greater and Lesser Antilles in the Caribbean and
z value patterns: z values for bats are higher on assessed the relative importance of island area,
oceanic islands than on land-bridge islands, reflect- maximum elevation, inter-island distance, and hur-
ing the greater isolation of oceanic islands. Because ricane-induced disturbance not only on species rich-
non-volant mammals on oceanic islands are relict ness, but also on community composition. A greater
fauna, z values are lower than on land-bridge is- elevational range usually means greater habitat
lands, since immigration rates are low. Comparing diversity, so elevation is a useful, simple surrogate.
bats directly with non-volant mammals, bats have There is a considerable amount of information in
higher z values on oceanic islands, but lower values these two papers, covering many complex interac-
on land-bridge islands. tions: I will extract only some key points. Species
Ricklefs and Lovette (1999) were interested in richness increased with island area in all three island
separating the influences of geographic and groups, with the Bahamas showing the weakest
biological factors on the species–area relationship. relationship. Species richness also increased with
To do this they compared bats, birds, butterflies, elevation in the Greater and Lesser Antilles, but
and amphibians/reptiles on 19 islands in the Lesser not in the Bahamas, since the islands are all low.
Antilles in the Caribbean. Island area varied over Similar patterns were found for guild richness. In
two orders of magnitude, from 13–1500 km2. Using the Greater and Lesser Antilles compositional simi-
simple correlation, all animal groups showed a sig- larity declined with increasing inter-island distance
nificant species–area relationship (Fig. 8.3), but there and increasing size difference. In the Bahamas, only
the similarity ended. inter-island distance could explain variation in com-
In multiple regression that included area and hab- position. These results held for all bats combined
itat diversity, area was a significant factor for species and when they were divided into broad carnivore/
richness of bats and birds, and habitat diversity for herbivore guilds. Overall, inter-island distance was
birds, butterflies, and amphibians/reptiles. What the most important determinant of compositional
does all this mean? All groups are strongly similarity. Species composition was found to be
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 199

The role of speciation in species–area relationships

was investigated by Carvajal and Adler (2005), who
studied species richness of the native mammals of
30 tropical Pacific archipelagos (Fig. 8.4).
Lesser
Antilles The basic statistics are interesting in themselves.
Caribbean Sea
Of the 106 species present, 83 were bats, 18 rodents,
and 5 marsupials. They were categorized as conti-
100 nental (species also found on continents or land-
bridge islands: 4 marsupials, 4 rodents, and 37
bats), Pacific (species found on more than one archi-
pelago: 7 bats) and endemic (species restricted to a
butterflies single archipelago: 1 marsupial, 14 rodents, and 42
birds bats). Endemics made up 20, 78, and 51 per cent of
species richness

marsupials, rodents, and bats respectively. The Bis-

reptiles and
10 marck and Solomon archipelagos stand out, with 52
amphibians
and 47 native mammal species respectively, all
others having 12 species or less. Fourteen of the
bats
archipelagos had no records of native mammals.
In five, a single bat species was the only native
mammal. Native marsupials and rodents were con-
fined to the Bismarck and Solomon archipelagos.
1 Continental bats were found only on archipelagos
close to the Australasian continental landmasses
10 100 1000 and land-bridges: Bismarcks, Solomons, New Cale-
area (km2) donia, Vanuatu, and Santa Cruz. The one exception
was Hawaii, which has a sub-species of Lasiurus
Figure 8.3 Species richness in relation to island area of the Lesser Antilles cinereus from the Americas. Pteropus tonganus and
in the Caribbean Sea (adapted from Ricklefs and Lovette 1999, with Emballonura semicaudata were the most widespread
permission from John Wiley and Sons).
species, found on ten and eight archipelagos respec-
tively. Species richness in the four groups (all spe-
nested in all island groups, but nestedness was cies, continental, Pacific, and endemic) was related
weakest in the Bahamas due to the apparently ran- to ten geographical variables. All species, continen-
dom distribution of a number of species. Nestedness tal and endemic groups, showed very strong rela-
measures the extent to which species composition tionships (simple Poisson regression) with all
across an archipelago departs from random, with geographic variables. As expected, these were posi-
small assemblages being hierarchical subsets of tive for area and elevation (remember this correlates
larger assemblages. For example, the communities with habitat diversity) and negative for isolation.
on small islands at increasing distances from a large Pacific species were strongly influenced by area,
source island may be progressively smaller sub-sets but not by isolation.
of the assemblage on that large island, due perhaps Bats have colonized Pacific archipelagos from
to different dispersal abilities of the bats or to greater Australasia and the Philippines, but only as far
habitat similarity between neighbouring islands. In as the Cook Islands. However, with the excep-
this study, both of these factors were thought to tions of the remote Marquesas and the Society
contribute to the pattern observed. Nestedness is a Islands (Fig. 8.4), those archipelagos without bats
complex measurement: methods and software are are made up of small, low, sandy islands with
still evolving. Whittaker and Fernandez-Palacios poor vegetation. Bats have almost certainly
(2007, p. 126) give a simple introduction to the reached many of these but have simply failed
subject. to colonize. The high species richness on the
200 BATS

1000 km
Hawaii
Wake

Marianas

Marshalls

Carolines
Belau
Nauru Tungaru Lines
Bismarcks
Banaba
New Phoenix
Solomons Marquesas
Guinea Tuvalu

Santa Cruz Wallis Samoa

Vanuatu Rotuma Societies Tuamotus
Cooks

Niue
Australia New Caledonia Fiji Tubuai
Tonga Pitcairn

Figure 8.4 Pacific island archipelagos studied by Carvajal and Adler (2005), with permission from John Wiley and Sons.

Solomons and the Bismarcks is probably due to a birds, and skinks all have similar rates of ende-
number of factors: mism in the tropical Pacific. It seems that despite
differences in vagility and hence colonization, spe-
(1) proximity to continental source populations will
ciation rates are great enough to ensure that ende-
have led to multiple colonizations;
mism is similar in all groups. Birds have colonized
(2) large area favours large populations and low
all 30 archipelagos, skinks 27, yet bats are found on
extinction rates; and
only 16. It is suggested that bats need large land
(3) high habitat complexity and within- and
masses to maintain persistent populations. The low
between-island barriers will promote spe-
reproductive rates of bats may be a contributory
ciation.
factor.
It is usually suggested that endemism will be I’ll finish this section on a study with a different
greater on remote islands, but it appears that the approach and focus. Frick et al. (2008) looked at bats
barriers between these two archipelagos and the on the small islands in the Gulf of California, Mex-
source populations (and the inter-island barriers) ico. They investigated the influence of area, isola-
are sufficiently small to promote regular coloniza- tion, and habitat on species richness. The latter was
tion, but great enough to limit gene flow and pro- made particularly convenient since the islands fell
mote speciation. The endemics on smaller and more into two groups approximately 500 km apart and
distant island groups are probably due to inter- the northerly group were significantly drier and
rather than intra-archipelago speciation. Speciation less vegetated in summer than those in the south
has been more common in the pteropodids than in (Fig. 8.5). The larger islands were more topographi-
other bats in the Pacific. More recent colonization cally diverse, which may give them greater habitat
and/or slower evolution are put forward as possible diversity, but diversity was not measured.
mechanisms, but evidence is poor as yet. Fifteen species were recorded on the Baja penin-
Carvajal and Adler (2005) also compared their sula (using echolocation recording and mist-
results with published data on other taxa. They netting), but only eleven of these were detected on
came up with the perhaps unexpected result that the islands and several of these were missing from
taxa as different as bats, non-volant mammals, the northern sites. The results are summarized in
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 201

roosts on the mainland (or indeed both) is not yet

known.

Mexico
Molecular genetics: a deeper look into
the past
Until recently, patterns of colonization and specia-
Ba
ha

tion had to be based on indirect and often scant

pe

N
n

evidence. Molecular genetics, as in many other

in
su

fields, has led to major advances. By constructing

l

300 km
a

phylogenetic and phylogeographic trees, using new

analytical tools, and with input from other fields
such as palaeoclimatology, we can map and explain
historical processes in increasing detail and with
greater confidence. Hewitt (2004) has written a
short and readable review, describing the techni-
12
ques and summarizing the broad global picture
(for animals and some plants) over the last few
10
million years. The techniques are being widely
species richness

8 applied to bats and what follows is far from com-

southern islands prehensive, but gives a taste of what is being
6 discovered.
O’Brien et al. (2009) constructed a phylogenetic
4 tree (based on mitochondrial DNA sequencing) of
northern islands
17 species and sub-species of fruit bat from the
2 Indian Ocean and beyond to test the hypothesis
that the current distribution of species in the Indian
0 Ocean can be explained by a single colonization
1 10 100 1000 10000 from Asia. The results turned out to be more com-
area (ha) plex. At the base of their tree is Pteropus scapulatus,
of Australian origin and perhaps one of the oldest
Figure 8.5 Species–area relationships for two groups of small islands in
species in the genus. The other bats fall into two
the north and south of the Gulf of California. Adapted from Frick et al.
(2008), with permission from John Wiley and Sons.
clades. The first includes P. poliocephalus and P. con-
spicillatus from Australia and P. hypomelanus, from
both Australasia and South-east Asia. All other spe-
cies fall into an Indian Ocean clade, but they do not
have a recent common ancestor and the pattern of
Fig. 8.5. Species richness increased with area in both colonization that gave rise to them appears to have
archipelagos and was significantly higher on the involved three distinct events (Fig. 8.6).
southern, more heavily vegetated islands. Semi-log The first islands to be colonized were the most
models were explored to explain the data and the distant from Asia: the Comoros (P. livingstonii and
best fit was obtained using a model that P. seychellensis comorensis), Pemba Island off the
incorporated both island area and isolation (shortest coast of Africa (P. seychellensis comorensis and the
flight path), despite the proximity of the islands to endemic P. voeltzkowi), and Mafia between them
the mainland (0.4–25 km). Whether the observed (P. seychellensis comorensis). There is no evidence for
effect of isolation reflects differences in immigration island-hopping, so it is presumed their ancestors
and extinction rates of resident populations, or for- arrived directly from India and Pakistan on
aging patterns of bats visiting the islands from monsoon winds. It is interesting that they failed to
202 BATS

India
P. giganteus
P. hypomelanus

P. giganteus arie
3 Maldives P. hypomelanus maris (extinct)

2
Pemba P. seychellensis seychellensis
P. voeltzkowi Seychelles
P. aldabrensis
Mafia
P. s. comorensis Aldabra
Chagos
Comoros
P. s. comorensis
P. livingstoni

P. rufus Figure 8.6 Current distribution of Pteropus

Rodrigues P. rodricensis
Mauritius P. niger bats on Indian Ocean islands and the three
Reunion P. subniger (extinct) presumed colonization routes, based on
Madagascar mtDNA sequencing and analysis. Adapted
Indian Ocean from O’Brien et al. (2009), with permission
from Elsevier.

colonize Africa, since they were very likely to have Pacific islands, their movements hindered or
been blown there! The second event was the coloni- assisted by sea-level changes during glacial periods.
zation of Rodrigues by the ancestor of P. rodricensis. When sea level fell, land area increased and land-
The remaining species arrived in a third wave: P. bridges joined islands, leading to changes in habitat
aldabrensis on Aldabra, P. seychellensis seychellensis distribution. As sea level rose again, land area
on the Seychelles, P. rufus on Madagascar, and P. decreased and islands became more isolated.
niger (and presumably the now extinct P. subniger) Campbell et al. (2006) asked: do sympatric, related
on Mauritius. This colonization may have occurred species respond to such change in ways that can be
as little as 125,000 years ago, although the dating predicted by differences in their ecology? They
evidence is weak as yet. Sub-species of P. giganteus studied four broadly sympatric species of small
and P. hypomelanus (now extinct) from mainland Cynopterus fruit bats in the Sunda region of South-
Asia were both found on the Maldives, not far off east Asia. Two species are forest specialists
the southern tip of India. The study also revealed (C. horsfieldii and C. brachyotis Forest), the other
that the current taxonomy of Indian Ocean fruit two (C. sphinx and C. brachyotis Sunda) are open-
bats, based on morphology, is in need of revision. habitat generalists. All are non-migratory, philopa-
The minimal differences in mtDNA sequences tric species with similar mating systems, so nothing
observed suggest that P. aldabrensis, P. niger, P. in their behaviour should lead to major differences
rufus, and P. seychellensis may not merit species sta- in population structure. Figure 8.7 shows the
tus, but are races of P. giganteus. Both my account current geography of the region, the sampling
and the map are simplifications of the complex and areas, and the extent of the land-bridges during a
tentative picture drawn in the paper. typical major glacial period. Palaeoecological and
At the same time as bats were colonizing the palaeoclimatic data indicate that the land that
Indian Ocean, others were moving around the emerged from the sea became savanna and scrub,
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 203

The Pleistocene glaciations also altered island

Sampling sites geography in the Philippines and Heaney et al.
C. brachyotisSunda
C. brachyotisForest (2005) studied genetic variation (in protein allo-
C. sphinx
C. horsfieldi
zymes, now a rarely used technique) of six fruit bat
species (and one rodent) across eleven islands cov-
ering most of the major sites of endemism in the
archipelago.
Current patterns in genetic variability were
Malaysian broadly (but not entirely) consistent with hypoth-
Peninsula
eses based on Pleistocene sea levels and the ecology
and vagility of the bats (and the rat!). The islands
and study species are summarized in Fig. 8.8. Three
bats are widespread species frequently associated
Borneo
with disturbed habitats (Cynopterus brachyotis,
Sumatra Macroglossus minimus, and Rousettus amplexicauda-
tus), two others (Haplonycteris fischeri and Pteno-
chirus jagori) and the rat (Rattus everreti) are
Philippine endemics, but widely distributed in the
Java islands and more generally (but not exclusively)
associated with undisturbed forest. The remaining
bat species, Ptenochirus minor, is a forest species
Figure 8.7 Sampling sites of four species of Cynopterus fruit bat on the
endemic to Greater Mindanao. What did they find?
Malaysian peninsula, used to test the hypothesis that the genetic population
structure of generalist species would show evidence of range expansion
Firstly, genetic differentiation in the rat was consis-
during past glacial periods. The white line encloses land exposed during a tent with the degree of isolation of the source islands
typical glacial maximum. Adapted from Campbell et al. (2006), with during the late Pleistocene glacial maximum. Pat-
permission from John Wiley and Sons. terns for the bats were less clear, but the populations
of the four small, historically isolated islands (Sibu-
yan, Dalupiri, Fuga, and Barit) were typically the
most differentiated. Those on Biliran and Leyte,
but tropical forest remained much the same in once part of Greater Mindanao and now only
location and extent. isolated by narrow channels, are less well differen-
It was hypothesized that the open-habitat gener- tiated. Genetic variation within populations ap-
alists would expand their range into new habitat peared to be independent of both island area and
whilst the tropical forest specialists would be con- isolation from source populations, but was signifi-
fined to their forest patches and that these differ- cantly greater in the widespread species than in the
ences would be reflected in current population endemic forest species. It appears that even on these
structure, which was examined using mtDNA and small islands populations are still large enough to
microsatellite markers (see Chapter 7). Two species prevent loss of genetic variability. In none of the
did indeed appear to have had long and stable species was a relationship found between present
demographic histories (C. brachyotis Forest and geographical distance between islands and genetic
C. sphinx) and two showed signs of past expansions distance. However, some evidence was found for a
(C. horsfieldii and C. brachyotis Sunda): one species of positive relationship between late Pleistocene inter-
each ecological type fell into each category! Clearly island distances and genetic difference. There was
the underlying mechanisms were more complex also a clear trend towards increased gene flow in the
than permitted by the simple hypothesis and diffi- widespread, adaptable species, relative to the forest-
cult to untangle without further study. As in the dwelling endemics. The authors go on to address a
previous study, the taxonomy of the species under number of other questions, but I will wind up with
investigation was to some extent rewritten. a summary: geological history and ecology are
204 BATS

high
medium genetic
differentiation
Cb low Dulapiri
Hf
high Mm
Pj Ra minimal Fuga & Barit
isolation

Ra
med. Hf Cb
Pj Mm
Ra
low Pj Cb
Hf Pm Mm
low med. high
colonising ability Luzon
Wide spread forest endemics
Hf Haplonycteris fischeri
Pj Ptenochirus jagori
Greater Mindanao forest endemic
Pm Ptenochirus minor
Widespread generalists
Cb Cynopterus brachyotis
Mm Macroglossus minimus
Ra Rousettus amplexicaudatus Catanduanes
sampling sites
Mindora
Sibuyan

Biliran
Panay
Leyte
Palawan

Negros

Mindanao
Figure 8.8 The Philippine archipelago, showing the
coastline (white lines) during the Pleistocene ice ages, when
many of the islands were joined by land-bridges. Heaney et al.
GREATER (2005) sampled six species of bat to investigate the relationship
BORNEO MINDANAO between historical processes and ecology in determining
200 km population structure. Figure adapted from Heaney et al.
(2005), with permission from John Wiley and Sons.

both important in determining present population perhaps single, founding events and that movement
structure within and between species—and the two between islands would be rare. The entire cyto-
are not always easy to separate. There are numerous chrome b gene was sequenced from wing mem-
other studies related to Pacific Ocean bats (for exam- brane samples to look at within-species population
ple Schmitt et al. 1995, Maharadatunkamsi et al. structure. Bats were sampled on the small islands of
2000), but I think it is time to look at studies in other the northern Lesser Antilles, Montserrat, Nevis, St
parts of the world addressing different questions. Kitts, St Eustatius, Saba, and St Maarten, since they
The islands of the Caribbean form a long chain are far enough apart (30–100 km) to make migration
between the Yucatan peninsula in the west and difficult.
Venezuela in the east (Fig. 8.9). Carstens et al. Artibeus jamaicensis is a widespread and common
(2004) studied three species of phyllostomid bat in species found right across the Caribbean, Brachy-
the Lesser Antilles to determine how they were phylla cavernarum is restricted to Puerto Rico and
likely to have reached the islands. Formed over 20 the Lesser Antilles, and Ardops nichollsi to the north-
million years ago, these volcanic islands have never ern Lesser Antilles. All three are medium-sized
been connected. It has therefore been proposed that omnivores, A. jamaicensis and A. nichollsi are
colonization was likely to have been through rare, tree-roosters, but B. cavernarum roosts in caves,
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 205

500 km

Artibeus jamaicencis

Haiti & Lesser

Cuba Dom. Rep. Antilles
Yucatan

Jamaica Puerto Rico

Ardops nichollsi
Brachyphylla cavernarum

Figure 8.9 Ranges of three Caribbean bat

species investigated by Carstens et al. (2004)
to study patterns of colonization. Figure
adapted from Carstens et al. (2004), with
permission from John Wiley and Sons.

which may make it less vulnerable to the frequent mitochondrial cytochrome b of five species of fruit-
hurricanes of the area. A. jamaicensis produces up to eating Carollia from sites across Central America and
three pups per year, the other two species a single tropical South America, demonstrating the impor-
pup (or more rarely, twins), perhaps giving A. jamai- tance of the uplift of the Andes and the formation of
censis a greater ability to recover when populations the Panamanian land-bridge in determining current
are reduced by hurricanes or volcanic activity. The population structure. In Europe, Juste et al. (2004),
evidence suggests that each species has a unique again using cytochrome b, described two clades of
colonization history. A. nichollsi appears to be mono- Plecotus that arose over 5 million years ago, their
phyletic, indicating that the species arose from a subsequent colonization of Madeira and the Canary
single founding event. The population structure of Islands, and the importance of the Strait of Gibraltar
B. cavernarum suggests that there was a single as a biogeographic barrier. Pestano et al. (2003a, b)
founding event, but that genetic diversity arose by and Salgueiro et al. (2007) investigated the origins
relatively recent and incomplete lineage sorting of and current population structures of bats on the
the founding population from the western Carib- Canary Islands and the Azores respectively.
bean. Lineage sorting is the loss of some lineages On the other side of the world, Lloyd (2003) con-
due to genetic drift or natural selection. In contrast, ducted a detailed study of the endemic short-tailed
A. jamaicensis appears to have colonized the Lesser bat, Mystacina tuberculata, in New Zealand. Multiple
Antilles several times, perhaps by island-hopping mitochondrial gene sequences were used to deter-
from both the west and the south, leading to the mine the population structure and demographic
extreme genetic variation seen on the islands of history of six phylogroups of this endangered bat,
Montserrat, Nevis, and St Kitts. Inter-island move- revealing considerable detail (Fig. 8.10). All three
ment of A. jamaicensis appears to be quite common, phylogroups on the North Island expanded during
perhaps due to a combination of abundance, tree- interstadials (when forests were most extensive)
roosting habits, and an obvious ability to survive before the last glacial maximum. A central North
being blown off-course by the wind. Island hybrid zone may result from range expansion
I’ll finish this section with a few very brief sum- as the forest recovered from large scale volcanic
maries to illustrate the range of recent work in this eruptions. On the South Island it appears that bats
field. Hoffman and Baker (2003) sequenced spread south from refugia in the north of the
206 BATS

known populations sampled

hypothesized post-glacial West

range expansion

Central Volcanic
Region

South East

South Island 1

South Island 2

Figure 8.10 The range expansion of the short-

tailed bat, Mystacina tuberculata, after recent
glacial periods, following the reforestation of New
200 km
Zealand. Adapted from Lloyd (2003), with
permission from John Wiley and Sons.

island as the forests followed the retreat of the conservation concern. Evidence for cryptic taxo-
glaciers 9–10,000 years ago. nomic divisions was also found between China
Finally, a study of a single species, the greater and Europe and within China itself. A follow up
horseshoe bat, Rhinolophus ferrumequinum, across study (Flanders et al. 2009) emphasized the complex-
its entire range: Rossiter et al. (2007) analysed micro- ity of this type of work and the benefits of using
satellite data from 56 localities between the UK and more than one genetic marker. They re-analysed
Japan. Data are consistent with a north-westward microsatellite data and combined it with sequencing
expansion across Europe from west Asia in the dis- of a segment of the mitochondrial ND2 gene. Mito-
tant past, with a more recent expansion from south- chondrial DNA alone suggested that Europe was
ern European refugia after the last glacial maximum. recolonized just once by a single expansion event.
Central European bats originated in the Balkans, Combined data points to two expansion events, one
those in Western Europe from Iberia and/or Italy. before and one after the Last Glacial Maximum and
It is interesting to note that there is greater genetic shows that Asia Minor was an important refugium
sub-division within the south-western UK than for the first of these expansions.
across 4,000 km of continental Eurasia, probably This is a rapidly expanding field and it has only
due to population fragmentation–a major been possible to give a brief overview. The language
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 207

120
100
No. of species

80
60
40
20

10 20 30 40 50 60 70
latitude

1000
No. of species

100 New World

10 Old World

Figure 8.11 Number of bat species in relation to latitude.

10 20 30 40 50 60 70
(a) In the New World, and (b) in the world. From Willig and
latitude Selcer (1989) and Findley (1993).

and techniques used present significant barriers to the number of species in 500  500 km quadrats at
the uninitiated and it would be difficult to go more each latitude in the New World (Willig and Selcer,
deeply into it without the need for substantial back- 1989). The pattern is also obvious for the whole
ground explanation. However, anyone interested in world in Fig. 8.11b (Findley, 1993) and Fig. 1.13.
this aspect of bat biology would benefit from the This increase in species richness with proximity to
overview given in the relevant chapters of Lowe the equator is seen in a wide range of flora and
et al. (2004). fauna, and a recent meta-analysis of almost 600 studies
Given the numerous studies carried out on is- by Hillebrand (2004) shows that the relationship is
lands, this is an appropriate point to mention the remarkably consistent across taxa, regions, habitats,
book, Island Bats, edited by Fleming and Racey and scales.
(2009) which covers many of the topics I’ve dis- Latitudinal gradients in species-richness are seen
cussed in greater detail, in addition to much else. in all studies of bats, but the precise pattern is
In summary, the study of bats, using a wider dependent upon the particular taxa studied, the
range of traditional and emerging techniques, region of study, the methods employed, and the
shows that SARs result from processes in all three scale at which measurements are taken (for example
categories: habitat diversity, equilibrium theory, Willig et al. 2003b, Lyons and Willig 2002). Most
and speciation, and that there are many interesting work has been done in the New World: two large,
historical diversions. inter-connected continents spanning the equatorial
zone in an essentially north–south direction, lend
themselves to such studies. Figure 8.12 shows spe-
Latitude
cies richness across the entire length of the New
What other large-scale patterns do we see in the World in relation to the expected distribution. The
distribution of bats? The most obvious is the latter assumes bats are distributed randomly in the
decrease in species richness with increasing latitude, absence of environmental gradients (Willig and
clearly seen in Fig. 8.11a which shows the decline in Lyons 1998), but takes into account shape and area
208 BATS

actual number of species

150
predicted
number of species
species richness

100

Equator
50

Figure 8.12 Real and predicted number of bat

0 species in relation to latitude in the New World
–60 –40 –20 0 20 40 60
(from Willig et al. (2003b)), with permission from
latitude (º) University of Chicago Press.

of the land. Since a strong latitudinal gradient is and colleagues took what they believed to be the
predicted by even a random distribution, it is impor- most plausible and well-supported hypotheses and
tant to relate real patterns to this null model. It can discussed them in turn. These hypotheses are sum-
be seen that there are more species in the tropics marized in Table 8.1.
than the null model predicts and fewer at higher Mittelbach and colleagues concluded that few of
latitudes. The departures from the model are also the many hypotheses can be rejected and that many
greater in the northern hemisphere than the south. are probably important—latitudinal gradients are
The next step is to explain these differences—but complex and there is no unifying theory. The tropics
that work has only just begun. appear to be an important source of new species and
The Old World is much more geographically com- these species tend to persist for longer. Tropical
plex and shows more complex latitudinal patterns in environments have been more extensive and more
species richness that have yet to be examined in any stable over time and there is evidence for a tropical
detail. The variation in latitudinal diversity patterns origin for many of life’s lineages. Palaeontological
has led to considerable debate over the mechanisms and phylogenetic evidence suggest that tropical re-
underlying the gradient and it is clear that method gions have had higher rates of diversification.
alone can have a profound effect, with considerable Higher speciation rates have contributed to this,
scope for introducing bias. In a recent review Mit- but in the absence of data on extinction rates little
telbach et al. (2007) put the many proposed mechan- can be said about their relative contributions. Can
isms into three broad categories: we explain the higher rates of diversification in the
tropics? Not yet—there are few data to test most of
(1) ecological hypotheses that focus on mechanisms
the hypotheses in Table 8.1, despite the fact that
of species coexistence;
many have been around for several decades and
(2) evolutionary hypotheses related to speciation
two for well over 100 years (Mittelbach et al. 2007,
rates; and
Table 1). A good general account of the complexity
(3) historical hypotheses relating to the extent and
of this and related issues can be found in Begon,
stability of tropical environments.
Harper and Townsend (2006). An important take-
These are obviously similar to those discussed in the home message, stated with clarity by Lawton (1996),
previous section on species–area relationships. This is that many ecological patterns have several causes
is a good, readable review for those wanting an and an obsession with finding a single explanation
introduction to this complex topic. Willig et al. is ‘unwise and unhelpful’, a view that has gained
(2003a) take a similar approach to the subject if wider acceptance after a period when one explana-
you would like a second viewpoint. Mittelbach tion was often set against another.
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 209

Table 8.1 Major hypotheses that could explain the latitudinal gradient in species richness (from Mittelbach et al. 2007).

Diversification rates are Diversification rates are higher in the tropics than in temperate regions
independent of latitude, but
there has been more time for
diversification in the tropics

Higher speciation rates Lower extinction rates

Tropical environments are old and Genetic drift in small populations accelerates evolution Stability of tropical climates reduces
many extant taxa evolved in them extinction rates
Dispersal of clades out of the tropics is Climatic variation results in higher speciation in the tropics Larger tropical areas have larger
limited and recent populations, larger species
ranges, and lower extinction rates
Higher likelihood of parapatric and sympatric speciation
Larger tropical areas provide greater opportunity for isolation
Narrower physiological tolerances in tropical organisms reduce dispersal
across unfavourable environments
Higher temperatures result in increased evolutionary speed
Stronger biotic interactions lead to greater specialization and faster
speciation

Altitude
considerations are likely to be just as important to
Another widely observed gradient is a decline in bats. Like many areas of macroecology this is a
species richness with increasing altitude. As with complex subject, but a very readable account of the
latitudinal gradients, it is a complex and controver- main issues has been written by Grytnes and
sial area and the underlying mechanisms will only McCain (2007).
be properly understood if all key factors are consid- One of the most methodical studies of bats is that
ered. For example it is important to control for by Patterson et al. (1998) who pooled data collected
changes in area, but until recently few studies had over numerous surveys into 250 m altitudinal bands
done so and even fewer involving bats. As you go between 340 and 3,450 m in the Manu Biosphere
up a mountain the area within a given altitudinal Reserve in the Peruvian Andes. The gradient passed
range usually declines (but see later!), so you would through five major zones: lowland rainforest, mon-
predict a decline in species richness on this basis tane forest, cloud forest, elfin forest, and puna. With
alone. Rahbek (1997), discussed this in some detail, increasing altitude species richness showed a steady
based on his analysis of bird data from across tropi- decline from over 100 species to less than 20
cal South America. He found a monotonic decline in (Fig. 8.13a). A similar pattern was seen in birds.
species richness with increasing altitude, but when When species were ranked by the midpoint of
change in area with altitude was taken into account their altitudinal range a distinctive pattern was rev-
species richness peaked towards the lower end of ealed: bats with higher altitudinal midpoints were
the altitudinal range. Rahbeck examines the validity found over increasingly large altitudinal ranges
of some of the basic assumptions that led to the (Fig. 8.13b). As altitude increases in mountains, the
expectation of a monotonic decline—such as a range of climatic conditions experienced by an ani-
monotonic decline in productivity. He also points mal over its lifetime will increase, so bats living at
to the severe geometric constraints of living on a higher altitudes need to be adapted to cope with a
mountain. The habitat width of birds in the Peru- wider range of climatic conditions than those living
vian Andes for a given altitudinal unit is only 4–20 lower down the mountain—hence the increasing
km. This is likely to have a major effect on viable range amplitude? Of 129 species, 111 were found
population sizes, extinction rates, and other ecologi- in the lowlands and only 5 were restricted to ranges
cal processes—which in turn will have a strong above 1,500 m. The pattern observed by Iniguez
influence on the species richness gradient. These Davalos (1993) in the Sierra de Manantlan of Mexico
210 BATS

(a)

100
species richness

80

60

40

20

1000 2000 3000

elevation (m)
(b)

3000
elevational range (m)

2000

Figure 8.13 (a) Species richness in relation to elevation.

1000
(b) Elevational range of 129 species ranked by mid-
elevational range (centre line). The range of elevations
(top and bottom lines) at which a species is found
increases steadily with elevation at all but the very highest
30 60 90 120
sites. From Patterson et al. (1998), with permission from
elevational rank of 129 species John Wiley and Sons.

was rather different. The gradient extended from influence latitudinal gradients and because there
400 m to 2,980 m, passing through savanna, tropical are only two independent gradients that can be
deciduous and semi-deciduous forests, gallery, oak, studied! There are numerous independent altitudi-
pine–oak, and pine forest and finally montane meso- nal gradients in mountain ranges across the world
phytic forest. The distribution of 27 of the 55 species and they typically occur over a small enough spatial
known to occur in the region is summarized in scale to eliminate latitudinal effects. She studied
Fig. 8.14. small non-volant mammals (26 gradients) and bats
Both species richness (Fig. 8.14a) and range ampli- (8 gradients) across 34 altitudinal gradients with the
tude (Fig. 8.14b) were higher at intermediate alti- aim of testing the importance of both area and spa-
tudes. Fewer than half of the species had ranges tial constraints. Rosenzweig (1995) argues that local
that extended into the lowlands. This pattern may species–area relations will be driven by habitat
be explained by the meeting of discrete lowland and diversity and the ties to habitat of individual spe-
montane communities that overlap on the slopes cies, but that on large scales extinction and specia-
of the mountains. Two studies, two results, two tion rates will be more important, and in all cases
explanations. species richness should decline as area declines.
Finally, let’s look at a study on bats that explicitly McCain suggests that patterns on mountains are
considers area in investigating species richness gra- likely to fall somewhere in the middle of this contin-
dients. McCain (2007a) chose to study altitude uum. Spatial constraints are usually described in
because too many poorly understood factors terms of the mid-domain effect (MDE): spatial
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 211

not a consistent global driver of altitudinal diversity

patterns. On average only 30 per cent of the varia-
20
bility in the data could be explained by an MDE
species richness

model and results were very variable. Spatial con-

15
straints, like area, appear not to be a major factor
either. Elimination of these leaves climatic and evo-
10
lutionary mechanisms. Evolutionary history may
play a role: for example on large mountains changes
5
in climate may have significantly different effects on
speciation and extinction rates at different altitudes.
1000 2000 3000 Colonization rates may also vary with altitude.
elevation (m) Unfortunately there are no data available with
which to investigate these possibilities. We do have
some information on the effects of climate and alti-
tude from another study by McCain (2007b), con-
ducted exclusively on bats. An analysis of data from
20 published gradients revealed that species rich-
2000 ness was greatest when temperatures were high
elevational range (m)

and water was abundant. On mountains with arid

bases species richness was highest at intermediate
altitudes, but it declined monotonically with
increasing altitude on mountains with warm, wet
1000 bases. These results make intuitive sense: thermo-
regulatory costs will decline and the abundance of
food will increase with increasing temperature. Sim-
ilarly, both the bats and their food (be it insects,
plant products, or even other vertebrates) need
water and arid ecosystems typically have lower
5 10 15 20 25 productivity than wet ones. There was also a posi-
elevational rank tive relationship between species richness and abun-
dance at most sites, underlining the conclusion that
Figure 8.14 (a) Species richness in relation to elevation. (b) elevational
range ranked by mid-elevational range. The range of elevations at which
productivity, driven by climate is a major factor in
a species is found is greatest for mid-elevation species. From Iniguez determining altitudinal species richness gradients.
Davalos (1993).

Community ecology
boundaries, in this case the tops and bottom of a In recent decades, community ecology has gone
mountain, will lead to the overlap of more species from being a largely qualitative science to one of
ranges in the middle of an area—the slope of the the more conceptually complex and mathematical,
mountain—leading to more species somewhere on as the emphasis has moved from describing to
the slope of the mountain. What did McCain find? understanding what we see. Complexity is inevita-
Bats and small mammals yielded broadly similar ble, since community ecology explores the many
results. To cut a long but interesting story short, interactions between species in order to explain
only 13 of the 34 gradients showed a strong diver- observed patterns of species distribution, abun-
sity response to area—those with a clear monotonic dance, and diversity. However, community ecol-
decline in diversity with altitude. The remaining ogy is not just about current patterns, but seeks
sites, with a mid-altitude peak in diversity, were to explain how these patterns arose through com-
not significantly influenced by area. Area is clearly munity assembly and dynamics. Many factors
212 BATS

are involved: competition, predation, mutualism,

environmental variability (biotic and abiotic), and
the diversity, productivity, and distribution (in community
time and space) of resources. In essence, what are guild
the rules that govern the coexistence of organisms?

taxa
This deceptively simple question has led to the ensemble
development of a number of theoretical approaches,
in particular competitive trade-off models, neutral
assemblage
models, and most recently, stochastic niche theory,
which incorporates elements of the first two and
appears to overcome the shortcomings of both (Till-
ecological function
man 2004). Competitive trade-off models work
from the assumption that a species cannot do Figure 8.15 Schematic definitions of the terms community, guild,
everything well. To be good at one ecological func- assemblage, and ensemble. From Patterson et al. (2003) after Fauth et al.
tion a species must be less good at another, so that (1996), used with permission from University of Chicago Press.
species coexist because each has a unique suite of
strengths and weaknesses that defines its niche.
Although they are supported by numerous field following terms are widely accepted, but they have
studies, the primary failing of basic trade-off mod- sometimes been used in a rather carefree manner in
els is that they don’t provide a general mechanism the literature. It can therefore be confusing relating
to explain differences in species abundance. Neutral studies on specific taxa to conceptual papers that are
models assume that there is no need for trade-offs more precise in their terminology. Fauth et al. (1996)
and that coexistence results from the balance provide a brief rationale for the scheme (illustrated
between speciation and random extinction as pop- in Fig. 8.15) that follows.
ulation sizes fluctuate. They too can explain many The species coexisting at a given site, in most cases
field observations, but they also predict that there one defined by the ecologist who wishes to study it,
will be no relationship between species traits are defined as a community: for example the species
and their abundances. This, as we will see in the living in a particular habitat within a larger and
next chapter, is certainly not the case: particular more diverse landscape, or the species on an island,
traits can make species rare, limited in distribution, in a particular valley, or at a particular altitudinal
and prone to extinction. Tillman (2004) took a com- range on a mountain or even in a random square.
petitive trade-off model and added stochastic ef- The species under study may be further constrained
fects that govern the likelihood that a new species to a particular guild or guilds—defined as local
will establish itself in a habitat, to produce the guilds. A guild is a group of species that exploit a
stochastic niche theory, overcoming these major similar resource in a similar way: for example herbi-
problems. vores, predators, parasites, saprophytes, and so on.
The abundance, diversity, and ecological impor- Each of these may be subdivided into other guilds,
tance of bats make it important to understand their for example predators could be carnivores or insec-
roles in communities, but it is difficult to test the- tivores. The community may alternatively be con-
ories in the large and complex communities they strained to certain taxonomic groups, in which case
occupy. Bat research in this field is largely one of they become an assemblage. If your geographically
observation, description, and comparison—manip- restricted community is restricted to a guild (or
ulative experiments are the preserve of other biolo- guilds) within a limited taxonomic group, then you
gists working on more amenable systems. However, are studying an ensemble. A niche is the spatial,
basic predictions can be tested and the models can temporal, behavioural (etc!) ‘space’ occupied by a
give valuable focus and insight in bat research. species, often described as its functional role in a
Before discussing studies on bats we should first community—a description of how it makes its liv-
be clear about some of the terminology—the ing. So, having got that straight, what are the big
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 213

questions in community ecology? The list is a long is discussed in detail in Chapter 2: briefly they deter-
one, but here’s a few to give you a feel for them. How mine flight speed, efficiency, and manoeuvrability.
do species in a guild coexist—do they all occupy In addition, foraging bats were placed in one or
distinct niches or is there some shared niche space? more of five microhabitat categories (by direct
Is a niche determined in part by the presence or observation or mist netting), in relation to the man-
absence of other guild members? Are there rules grove stands.
that govern interactions within and between guilds, A plot of WL against AR for all species in the area
leading to non-random community composition? shows considerable overlap in morphological space
Do different rules become important at different (Fig. 8.16a). However, within the ensemble observed
scales? Community structure on a local scale must at each of the six sites studied, little or no overlap
be determined in part by the pool of species present was seen (two examples are shown in Fig. 8.16b).
at higher spatial scales, so the boundaries between A comparison with randomly generated ensembles
community ecology and macroecology are necessar- strongly supported the view that the natural ensem-
ily fuzzy. Macroecological rules dictate the species bles had not arisen by chance. Attempts to add other
pool upon which community ecology rules act to species known to occur in the area to ensembles
determine community structure. Since the rules at almost invariably led to morphological overlap.
all levels are complex and controversial, there are WL and AR were good predictors of microhabitat
some major challenges! niche (Fig. 8.16b). We can tentatively conclude that
we have evidence for a non-random, structured bat
community, and in this instance one with little mor-
Niche partitioning
phological, and apparently niche, overlap. This clear
In earlier chapters we saw how wing morphology, separation of members of an ensemble solely on the
echolocation call structure, and body size influenced basis of flight morphology has not been seen in other
foraging style and food preferences of particular studies, but it is clearly one of the dominant factors
species. Bats therefore have potential as a model in niche separation as we shall see.
for community ecology studies, since many of the In several studies, echolocation call structure data
key parameters can be measured with relative ease have been included alongside those on flight mor-
and accuracy. Early attempts to study the structure phology and foraging behaviour. Aldridge and Rau-
of insectivorous bat assemblages or ensembles were tenbach (1987), in a study of 26 insectivorous species
based around differences in wing morphology and in the Kruger National Park, South Africa, measured
its presumed, or observed, effect on flight character- three flight morphology indices (WL, AR, and wing
istics and, in turn, the habitat used. Later studies tip shape index) and five call parameters (highest and
often included other variables, including body size lowest frequencies, frequency of maximum intensity,
and echolocation call parameters. Analytical meth- search phase call shape, and call duration). Captured
ods have become more sophisticated, the questions bats were also flown through an obstacle course to
more ambitious, and a wide range of other defining provide an index of manoeuvrability. The foraging
characters are now being used, as we will see. So behaviour of light tagged individuals was observed
let’s begin the questions. Is there evidence of com- in different habitats, and these were ranked in
petition for resources, and if so, might this competi- order of increasing clutter. Finally, faecal pellets
tion contribute to niche separation? Are the niches from captured bats were analysed for prey size.
occupied by different bats discrete, or is there over- Four groups of species were identified on the
lap between niches? basis of WL, AR, and mass (Fig. 8.17). Group one
McKenzie and Rolfe (1986) studied 15 insectivo- species had low WL, AR, and mass, and were
rous species feeding in the mangrove forests of calculated to have low flight speeds and high
northern Western Australia. Flight morphology manoeuvrability. They also had clutter-resistant
was expressed in terms of wing loading (WL = echolocation calls, and should thus be capable of
body mass/wing area) and aspect ratio (AR = wing- feeding in clutter (for example Nycticeius, Eptesicus,
span2/wing area). The importance of these variables Pipistrellus, and Hipposideros species). Group two,
214 BATS

(b)

9 Taphozous
flaviventris

8 IN THE OPEN
Chaerophon jobensis
OVER CANOPY
(a) Chalinolobus Mormopterus loriae
7 nigrogriseus

Nycticeius greyi
FOREST EDGE
6 Nyctophilus arnhemensis

aspect ratio
INSIDE FOREST

5 7 9 11 13 15 17
9
FOREST EDGE Taphozous georgianus
Pipistrellus
8 tenuis IN THE OPEN
Rhinonycteris
aurantius Miniopterus schreibersii
7 OPEN SPACES BESIDE AND ABOVE CANOPY

Nycticeius “sanborni” INSIDE FOREST

Myotis adversus
6 FOREST EDGE
Nyctophilus arnhemensis INSIDE FOREST

5 7 9 11 13 15 17
wing loading (Nm–2)

Figure 8.16 Wing loading plotted against aspect ratio for an insectivorous bat community in an Australian mangrove forest. (a) All species in the region,
(b) examples of two of the six localities studied, with the microhabitats in which each species foraged. Adapted from McKenzie and Rolfe (1986).

16 1 2 3 clutter > woodland edge

14
12
aspect ratio

4 open
10
8
6
4
2 mass (g)
–2 )
m 20 40 50 100
( N 4
i ng
o ad 8 Figure 8.17 Three-dimensional plot of mass in relation to
gl aspect ratio and wing loading for a South African
in 12
w
insectivorous bat community. The different symbols refer to
16
the four classes discussed in the text. Adapted from Aldridge
20 and Rautenbach (1987).

with intermediate WL and mass and slightly higher species. Group four included five Tadarida species
AR, were calculated to be a little less manoeuvrable and one Taphozous, all characterized by high WL,
and, in conjunction with their echolocation calls, high AR, and pointed wing tips. All six have low
were predicted to feed in both clutter (Rhinolophus frequency, narrowband FM calls. These are bats with
hildebrandtii) and intermediate clutter/woodland fast, agile, and efficient flight, and an ability to detect
edge habitats (for example Scotophilus species). prey at some distance, and are therefore likely to
Group three comprised only one species, Hipposi- forage in the open. Do these predictions of foraging
deros commersoni, because of its high body mass, but niche match the observed foraging patterns, and is
it was predicted to be another intermediate clutter there evidence for niche separation?
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 215

canopy specialists
gap incorporators
Mormopterus
20 gap specialists beccarii
wing loading (Nm–2)

Scotorepens balstoni
Chalinolobus Chaerephon
nigrogiseus jobensis
Hipposideros
Rhinolophus diadema
megaphyllus Mormopterus
10 loriae
Eptesicus
pumilus
Hipposideros unidentified
ater Eptesicus emballonurid
sagittula
Nyctophilus
bifax
Figure 8.18 Wing loading in relation to aspect
0 ratio for another Australian insectivorous bat
5 6 7 8 9 community. See text for details. Adapted from
aspect ratio Crome and Richards (1988).

Four habitat-use strategies were identified: open three foraged in both (Fig. 8.18). Again, flight mor-
foraging (by for example Tadarida condylura and phology and echolocation call structure were good
T. pumila), woodland-edge foraging (by for example predictors of habitat use: for example three of the
Nycticeius schlieffini, Eptesicus capensis, and Pipistrel- five gap species were fast flying molossids, typified
lus nanus), intermediate-clutter foraging (by for by long, narrow, wings and long, narrowband FM
example Scotophilus borbonicus and S. dingani), and calls. All five gap species also foraged in the neigh-
clutter foraging (by for example Hipposideros caffer bouring eucalypt woodland.
and Rhinolophus hildebrandtii). Habitat use therefore Fullard et al. (1991) found that seven species of
correlated well with manoeuvrability index and insectivorous bats in the Perup forest of south-west-
with flight morphology. Although each species ap- ern Australia formed loose clusters of clutter, edge,
peared to have flight and echolocation characteris- and open habitat foragers, with substantial inter-
tics well suited to its foraging style, most were specific overlap (Fig. 8.19).
flexible, using several of the seven habitat categories WL, AR, and call structure correlated well with
defined. Each niche was also occupied by several foraging zone classification. They also noted the
species. Furthermore, although larger bats tended flexibility shown by most species—six of the seven
to eat larger insects, larger species also took a larger foraged in a wide range of microhabitats, and a wide
size-range of prey, and there was little evidence for variety of insects was taken by most species. The
partitioning on the basis of prey. In contrast to the study was carried out in the dry season, when insects
mangrove ensemble studied by McKenzie and Rolfe are less abundant. The authors suggest that some
(1986), we have here an ensemble with considerable species could be more selective in the wet season,
overlap of prey and microhabitat between species. but this appears not to have been followed up.
McKenzie and Rolfe (1986) argue that the clear cut Kalko (1995) observed foraging emballonurids in
patterns they observe may be due to the particular Panama with night vision goggles and recorded
stability of the mangrove habitat: in other habitats, echolocation call structure. The results are summar-
the influx/efflux of species due to seasonal and long ized in Fig. 8.20, which shows foraging microhabi-
term habitat changes are likely to blur boundaries in tats, typical flight paths, and search phase
ensemble structure. echolocation call structure. The two species feeding
Crome and Richards (1988) looked at partitioning in the open had low frequency, quasi-CF search
between rainforest canopy and artificial gaps phase pulses around 24–32 kHz. At the other
(0.03–0.07 ha) created by logging in Queensland, extreme, the tiny Rhynchonycteris naso, which foraged
Australia. Four species were confined to the closed over small streams in the vegetation, had a high
canopy, five more to the gaps, and the remaining frequency call (100–105 kHz), with a steep FM tail to
216 BATS

OPEN too with broadly similar results, with the addition

Tadarida australis of parameters such as cranial shape and bite force.
EDGE
Aguirre et al. (2002), Hodgkison et al. (2004), and
Mormopterus planiceps
Campbell et al. (2007) are recent studies that will
Chalinolobus gouldii give you an entry into this field.
The consensus from these studies is thus for
Chalinolobus morio
CLOSED some partitioning of resources, at least in terms of
Nyctophilus gouldi foraging habitat, but with appreciable overlap
between species within ensembles, and considerable
Eptesicus regulus
flexibility of foraging style in many species. However,
Nyctophilus geoffroyi this may be too simplistic: there are many factors that
have yet to be taken into account. We need more
information about prey preferences, microhabitat,
Figure 8.19 Overlap of foraging niches in another Australian bat
capture methods, seasonal variation in prey, and tem-
community. From Fullard et al. (1991).
poral foraging patterns, to name just a few. Arlettaz
(1999) has shown that the two sympatric, sibling spe-
most of its short CF pulses. The three species feeding cies Myotis myotis and Myotis blythii, despite their
in gaps in the forest understorey had intermediate genetic and morphological similarity, occupy differ-
frequency (42–56 kHz) pulses, with FM sweeps either ent niches (Fig. 8.21). M. myotis fed on ground dwell-
side of a quasi-CF pulse. The two Saccopteryx species ing insects, primarily carabid beetles, captured in
also fed close to the ground, and often emitted pulses habitats that made this kind of prey more readily
in pairs. Evidence for microhabitat partitioning is available: freshly cut meadows and mown orchards,
clear, but there is overlap in both apparent niche forest with little undergrowth or tall groundcover.
and call structure, particularly in the three species M. blythii preferred species typical of uncut grassland,
feeding in the understorey. mainly bush crickets.
I have concentrated on insectivorous bats, but Siemers and Swift (2006) looked at two
studies have been carried out on fruit-eating bats more sympatric Myotis species, M. bechsteinii and

120
freq. (kHz) Diclidurus albus
0
0 100 200 300
time (ms)

Peropteryx kappleri?

Cormura brevirostris

Saccopteryx bilineata

Saccopteryx leptura
Rhynchonycteris naso

Figure 8.20 Foraging niches and echolocation call structures of six emballonurid bats in Panama. Based on Kalko (1995).
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 217

mean percentage frequency

50 Myotis myotis
Myotis blythii
40
*** P<0.001
30 ** P<0.01

20

10
y

le

er

pe

rd

st
ur
ck

ba

re
ab

do

do
at

a
ep

st

ch
ro

fo
ur

w
ar

ea

ea
st

pa

or
m

m
*** **

n
** ***

ns

ow
de

m
*** **

Figure 8.21 Foraging niche partitioning in two sympatric, sibling species, Myotis myotis and Myotis blythii. Adapted from Arlettaz (1999), with permission
from John Wiley and Sons.

M. nattereri, roosting in the same orchard and forag- Heller and Helversen (1989) studied 12 syntopic
ing in the same forests and orchards. M. bechsteinii (same habitat and range) Rhinolophus and Hipposi-
frequently captured prey using only prey-generated deros species in the Krau Game Reserve, Malaysia.
sound, something M. nattereri never did, and this The frequencies of their narrowband echolocation
was reflected in their significantly different diets. calls were significantly more evenly distributed
Although morphologically similar, M. bechsteinii than predicted by chance, a result not due to differ-
has significantly larger ears and evidence was ences in bat size. There was also evidence for a
found for sensitivity to the sounds made by poten- minimum frequency separation between species.
tial prey. Evidence of segregation through habitat These results could be interpreted as evidence
use has also been reported in comparisons of pairs of for partitioning of prey on the basis of size: higher
Myotis species in Canada (Herd and Fenton 1983, frequencies are better at detecting smaller prey, but
Saunders and Barclay 1992). over shorter distances (see Chapter 3). However, call
Other studies have looked at diet. For example, separation of this nature could be for communi-
Findley and Black (1983) examined the morphology cation, but the two functions are not mutually exclu-
and stomach contents of nine insectivorous species sive. As an interesting aside, call separation may
captured outside a cave in Zambia. Morphology and be a significant factor in the evolution of new spe-
diet were closely related. On the basis of cies, with either social (for example Kingston et al.
multivariate analysis they proposed an ensemble 2001) or foraging (Kingston and Rossiter 2004) dri-
consisting of a cluster of invariant specialists and vers in operation. Siemers and Schnitzler (2004)
a smaller number of species with more variable were able to show that differences in bandwidth in
morphology and diet. A novel approach was taken five European Myotis species were correlated with
by Tschapka (2004), who measured the energy avail- differences in their ability to capture prey close to
able from flowers (in terms of energy density, kJ ha1 clutter (Fig. 8.22), suggesting that small differences
day1) to nectarivorous bats in a Costa Rican rain- in sensory ability may be important in niche
forest. The very large differences in energy density, differentiation.
spatial distribution, and phenology of the plants are An interesting example of possible resource parti-
probably a significant factor in niche partitioning— tioning comes from a study in the south-western
I’ll say more about this study later in the chapter. deserts of the United States and northern Mexico
218 BATS

(a) Testing models: do ecological processes

shape the structure of bat ensembles and
100
assemblages?
80 M. nattereri
Over the last 20 years or so a number of papers (for
capture success (%)

M. emarginatus
example Willig and Moulton 1989, Arita 1997) have
60
addressed the question of whether local bat ensem-
40
M. mystacinus bles and assemblages are more than just random
M. daubentonii
sub-sets of the species in a regional pool, reflecting
20 the importance of ecological processes. Some have
M. dasycneme
failed to find evidence for a significant effect of
0
ecological processes, but in an accessible and useful
review Patterson et al. (2003) raise important points
0 10 20 30 40 50 not always considered in earlier work: if studies are
distance of prey from clutter screen (cm)
done over large heterogeneous areas (and thus mea-
(b) sure regional or gamma diversity) they are unlikely
to be studying the true interactions within local en-
16
sembles (alpha diversity): some of the animals
minimal capture distance (cm)

M. dasycneme under study may rarely or never interact. Equally

12 important, species or whole elements of the local
M. daubentonii fauna that do interact with the study organisms
may be excluded. A recent study by Moreno et al.
8 M. mystacinus (2006) explicitly addressed this issue. They studied
local ensembles and assemblages in a landscape
M. emarginatus
with different levels of species richness. The
4
M. nattereri 42 km2 study area in central Veracruz in Mexico
had 14 sampling sites across 7 habitats. The study
0 was confined to Phyllostomidae and Mormoopidae
20 40 60 80 100 120 140 since it was believed these could be sampled effec-
bandwidth of search calls (kHz) tively. Ecological space was defined in terms of
wing loading, aspect ratio, and wing tip index.
Figure 8.22 In European bats of the genus Myotis, bandwidth of search
They looked at the average morphological distances
phase echolocation calls is correlated with ability to take prey close to
clutter. Adapted from Siemers and Schnitzler (2004), with permission from
between species in real ensembles and assemblages
Nature Publishing Group. and compared them with randomly generated units.
Because assemblages include species that rarely or
never interact, since they are in different guilds, they
predicted that ecological processes would be weak,
by Herrera et al. (1993). Carbon stable-isotope ratios but that these processes would be strong and detect-
(see Chapter 6) suggest that although the pallid bat, able at the ensemble level. These predictions were
Antrozous pallidus, and the California leaf-nosed bat, supported by the data: the structure of assemblages
Macrotus californicus, both feed on large arthropods, of neotropical bats appear to be determined by
primarily by gleaning, only the pallid bat regularly neutral processes, but ecology, in particular compe-
feeds at cactus and agave flowers. Incidentally, tition, is a significant determinant of ensemble
although it is unlikely that the bat is feeding on structure.
nectar, it does appear to be an effective pollinator: Schoeman and Jacobs (2008) asked the question:
are we looking at an early stage in the evolution of is ensemble composition non-random in ways pre-
nectarivory? dicted by theory? The chosen ensemble was the
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 219

insectivorous bats of southern Africa and they remember is that similarity of form, and hence for-
looked for evidence that patterns in phenotypic aging style and niche occupation, may arise through
structure (body size, wing morphology, echoloca- phylogenetic constraints: related bats are usually
tion call structure) were influenced by competition morphologically similar. The part phylogeny plays
and prey-defence mechanisms (generically referred has been considered in few studies. We should also
to as deterministic filters). The authors make the remember that other resources may be important
point that bats are ideal in this respect since these in determining community structure, such as roost-
mechanisms are most likely to be important in stable ing requirements. For example, for species using
systems, and bats are characterized by long life, low bamboo culms, or furled leaves, roosting require-
fecundity, low predation pressure, and stable popu- ments are so specific that they may determine the
lations. One important step was to try and include presence or absence of the species in a community.
all species in the analysis, since missing species They may also be important determinants of a bat’s
could have a significant influence on the patterns. morphology. Humphrey (1975) found that
Of 46 known species, all but 4 rare species were increased topographic diversity in the environment
included in the study. They used principle compo- (primarily trees and buildings) led to an increase in
nent analysis to measure phenotypic distances the species richness and diversity of colony forming
between species in ensembles, predicting that in bats.
ensembles structured by competition these should
be greater and less variable than those in randomly
selected ensembles: that is phenotypes should be The interface between macroecology
well and evenly spaced. This would apply to all and community ecology
phenotypic characters. Conversely if structure is
It will by now be obvious that there is a very fuzzy
determined by adaptations to overcome the anti-
line between macroecological and community eco-
predation mechanisms of their prey, then pheno-
logical processes. Scale is a pervasive factor in both
typic distances related to echolocation call structure
and the boundary shifts depending on location, time,
should be smaller than those in randomly chosen
community, assemblage, ensemble, and so on.
ensembles. These contrasting predictions were
I thought I would finish this section with a descrip-
tested on local and regional assemblages, and in
tion of some of the work that focuses on this inter-
three functional groups within them: open air, clut-
face. Useful reviews of theoretical frameworks for
ter-edge, and clutter foragers. The take-home mes-
this multi-scale approach to community ecology
sage was that some ensembles and functional
include Kneitel and Chase (2004) and Leibold et al.
groups showed non-random patterns that pointed
(2004). Leibold and colleagues describe a metacom-
to competition or prey-defence mechanisms as
munity model in which a set of communities
important deterministic filters, however, the major-
is linked by species that can disperse from each
ity appeared random. Those that did appear to be
and are capable of colonizing other communities.
influenced by competition were the more species-
Four types of model are reviewed that emphasize
rich ensembles and the functional groups in which
different processes, but which are likely to merge
we might expect competition to be more intense.
into a single, more broadly applicable model:
The small number of non-random echolocation pat-
terns observed may reflect counter-adaptations to  patch-dynamics—identical patches, subject to
prey in a coevolutionary arms race. However, they random and ecologically-driven extinctions, in
may simply reflect the fact that echolocation gener- which extinction is influenced by interaction
ally works best over a narrow range of frequencies. between the species and dispersal to new patches
Even if these results are taken as unequivocal evi-  species-sorting—patches are heterogeneous and
dence for the influence of competition and predator– species interactions are dependent on the abiotic
prey coevolution, much remains unexplained. environment
There are a number of issues I have not addressed,  mass effects—in which dispersal becomes a
but which should not be forgotten. One thing to major factor in determining community structure
220 BATS

 neutral—in which species are essentially equiva- model may be widely applicable to bats, given their
lent and dynamics are very slow—a null model in mobility and, as we will see in the next chapter, may
which the trade-offs inherent in the other models be an important factor in conservation strategies.
are absent. For yet another new approach to metacommunity
analysis in bats I suggest you look at Presley et al.
These concepts, at least in the form of detailed,
(2009). It’s time we moved on to new ground. In fact,
integrated models, are quite new. Studies on bats,
we are going to make a big step—after talking about
and indeed studies in general, are only now catch-
little else but bats, we are going to look at some of
ing up with the development of these predictive
the animals and plants that bats interact with.
models, but the foundations are being laid by
describing and comparing communities and explor-
ing the possible mechanisms underlying differences. Predators and prey
Stevens and Willig (2002) took a very broad
Bats as predators
approach and looked at the scale-dependence of
latitudinal gradients in richness in New World Bats feed on a wide range of animals (Fig. 8.23), but
bats. They concluded that the large scale gradient I know of few published studies that investigate the
(measuring gamma diversity) was generated by effects bats have on community dynamics. The few
both an increase in species richness at the local species feeding on vertebrates are unlikely to have a
level (alpha diversity) and an increase in species major effect on the community of which they are a
turnover amongst communities (beta diversity). part—there are relatively few species and they are
Most studies have taken a more spatially limited not abundant. On the other hand, the enormous
approach, such as that by Stevens et al. (2004), who concentrations of insectivorous bats found in many
compared two forest reserves in Paraguay. Differ- parts of the world must have a significant impact on
ences in species richness, composition, and rank insect population dynamics, and in turn are likely to
abundance patterns were explained on the basis of influence other parts of their ecosystem. Estimated
differences in habitat heterogeneity between the two from the mean increase in bat weight after foraging,
sites, which facilitate coexistence. Stevens et al. a colony of one million Brazilian free-tailed bats,
(2007) were among the first to explicitly apply the Tadarida brasiliensis, will eat 500–1,000 tonnes of in-
new metacommunity models to large scale empiri- sects in a single summer. ‘Control’ of fruit-eating
cal data. They collected data on species composition bats and vampires in several countries has been
from 26 sites in Paraguay. They estimated the vari- crude and indiscriminate, killing insectivorous bats
ance in species composition across sites that could too, and increases in local insect pests have been
be explained by spatial characteristics (geographical reported as a consequence, but the reports lack sci-
coordinates) and local environment: elevation, pre- entific corroboration. Until very recently, there had
cipitation, temperature, and evapotranspiration been no published attempts to quantify the effects
(which proved to have minimal influence) and bats may have on insect populations, but this is
percentage cover of 17 different vegetation types beginning to change. Cleveland et al. (2006) esti-
(which were important). Both could alone explain mated that T. brasiliensis, by eating crop pests, was
a significant component of the variation in species worth around 0.75 million dollars to the annual 5.5
composition, but spatial factors were more impor- million-dollar cotton industry in one region of Texas
tant. This suggests that a mass effects model—one in alone. I will discuss this work in more detail in the
which dispersal is the most significant factor—de- next chapter, but a bat that has effects of commercial
scribes the observed metacommunity structure bet- significance must be having a significant impact on
ter than species-sorting or neutral models. In a community ecology. Several studies published in
species-sorting metacommunity environmental fac- respected journals have investigated the effects of
tors should be dominant, and in a neutral model birds on the insects on tropical trees by placing
neither suite of factors should be significant predic- exclosures over the trees. However, the exclosures
tors of metacommunity structure. The mass-effects were left on day and night and therefore actually
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 221

in Mexico. Kalka and colleagues reported similar

effects of birds and bats on the number of insects
found on plants. However, as the number of insects
increased with the exclusion of both birds and bats,
there was a near proportional increase in herbivory
(Fig. 8.24b). The exclosures in this study covered
native, tropical understorey plants. The effects are
likely to be even more significant in the canopy,
where productivity is higher and where there are
frequently more bats. Such studies are enormously
labour-intensive and difficult and until they are
extended we can only guess at the effects insectivo-
rous bats have on community structure. However,
we can say rather more about the interaction
between bats and their insect prey at the individual
level.

The arms race: bats versus insects

The interaction between bats and their prey has
received considerable attention from researchers,
since it has many of the features of an evolutionary
arms race and is in many respects easy to study.
Numerous reviews have been written: a short but
Figure 8.23 Pallid bat, Antrozous pallidus, taking a scorpion wide-ranging summary by Jones and Rydell (2003)
would be my choice of a place to start any further
looked at the effects of both birds and bats. It is a reading, since it puts the arms race into a wider
sign of how little attention is given to bats by many context, bringing together aspects of bat–insect in-
biologists that this fundamental flaw in experimen- teractions that are of necessity dealt with in different
tal design went apparently unnoticed by authors, chapters of this book. It has the added interest of a
referees, and journal editors! In 2008 two studies fine poem on the subject by David Pye, one of the
appeared back-to -back in the same issue of Science pioneers in the field. Most insectivorous bats locate
that went some way to rectifying the situation. Wil- and capture their prey using echolocation. How-
liams-Guillén et al. (2008) and Kalka et al. (2008) ever, many insects can hear ultrasound: species
carried out their work in Mexico and Panama from at least six orders (Coleoptera, Dictyoptera,
respectively. They also used exclosures but designed Diptera, Lepidoptera, Neuroptera, and Orthoptera)
their studies so as to assess the importance of both possess auditory organs capable of detecting sounds
bats and birds. Williams-Guillén and colleagues at frequencies of more than 20 kHz (for example
found that the exclusion of both birds and bats led Michelson and Larsen 1985, Yager and Hoy 1986,
to significant increases in the number of arthropods Spangler 1988, Robert et al. 1992). These ears (many
on coffee plants (Fig. 8.24a). tympanic or drum-like) have evolved from a variety
The effect of bats was small but significant during of organs, and are located on the abdomen, thorax,
the dry season but increased markedly during the head, legs, or even the wings, depending upon the
wet season and was then greater than that of birds: insect. The ears of moths, the best studied group, are
bat exclusion led to an 84 per cent increase in the tuned to the frequencies (20–60 kHz) emitted by the
number of arthropods. There was no detectable bats that are known, or are most likely, to prey upon
effect on plant herbivory in this study, but coffee is them (Fullard 1987) and from 50 to over 90 per cent
a non-native crop plant and may have few predators of moth species in a given locality may be able to
222 BATS

(a)

arthropods/100 leaves
arthropods/100 leaves

20 12
both both
15 birds bats
bats 8
10 birds
control 4
5 control

0
22 Jan 5 Feb 19 Feb 27 Feb 9 July 23 July 6 Aug 22 Aug

(b)
**
**
15 16 **
*

12
mean arthropods-2m (s.e.m

mean percent herbivory (sem)

)

12

*
9 **

4
3
birds bats birds bats
control control
excluded excluded excluded excluded

Figure 8.24 The effects of excluding bats (and birds) on the number of arthropods found on plants and in (b) on plant herbivory. From (a) Williams-Guillén et al.
(2008) and (b) Kalka et al. (2008) with permission from AAAS. See text for details.

hear (Fenton and Fullard 1979). It is now widely quickly take the moth out of a bat’s flight path (for
accepted, at least in the case of moths, that hearing example Roeder 1964, Rydell et al. 1997). There is
evolved as a defence against predation by bats (for evidence to suggest that moths and mantids gain
example Roeder 1965, Miller 1975, Yager et al. 1990, information about the proximity of bats from the
Fullard and Yack 1993). In others, notably Orthop- increased pulse repetition rates of the feeding buzz
tera (crickets, locusts, and so on), a bat-evasion func- (Fullard 1992, Yager and May 1990). Roeder and
tion may have evolved in ears used primarily for Treat (1961) estimated that these strategies reduced
intra-specific communication (Fullard and Yack the chance of capture in the field by about 40 per
1993). Yager et al. (1990) suggest that hearing has cent. Later field studies suggest that insects with
evolved independently at least eight times in the ears have a significantly greater chance of survival
insects, three times in the Lepidoptera alone. (for example Acharya 1992, Rydell 1992). In many
Sound reception generally produces a two-stage species males have to seek out stationary (and in
evasive reaction. Low intensity ultrasound causes some cases flightless) females. It is perhaps not
moths (and some other insects studied) to fly away surprising that males, at greater risk from bats,
from the sound. High intensities induce complex often have much better developed hearing systems
spirals, loops, and dives or cessation of flight, to than females (for example Rydell et al. 1997). Some
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 223

Figure 8.25 Distasteful Arctiidae moths emit ultrasonic clicks in response to bat echolocation calls. These clicks may be the equivalent of aposematic
colouration and/or ‘jam’ bat echolocation.

species of Arctiidae (Fig. 8.25) emit loud clicks in and Rydell (2003) review the evidence for other
response to very loud bat ultrasound calls, and these potentially adaptive changes in insects, such as the
can impair the performance of attacking bats (Miller observation that earless moths tend to be very small
1991). (to avoid detection), or very large (too large for many
Fullard et al. (1994) suggest that the late emission bats to capture), and/or fast (to evade capture). In-
of the clicks, during the bat’s terminal feeding buzz, sects may also have evolved to fly outside the major
will maximize this jamming/startle effect. The click flight times of bats. The ears of some autumn-flying
may also be a warning to predators that the moth is noctuids have degenerated, suggesting reduced pre-
unpalatable (for example Acharya and Fenton 1992). dation pressure—and those of diurnal moths have
Arctiidae are typically unpalatable and have bright degenerated still further. Hofstede et al. (2008) have
aposematic colouration to deter diurnal predators. recently shown that there is a strong correlation
A number of studies (for example Bates and Fenton between nocturnal flight time and auditory sensitivity
1990, Barber et al. 2009) have found that naive big in noctuid moths (Fig. 8.26). More striking still, in the
brown bats (Eptesicus fuscus) are startled by clicks endemic noctuid moths of Tahiti, where there are no
made by arctiids, but individuals learn to associate bats, the hearing system is vestigial and dysfunctional
the clicks with unpalatability. Palatable species will (Fullard et al. 2004).
mimic unpalatable species, but this is not always a There is evidence that some species make up for
successful strategy (Barber and Conner 2007). Rat- the lack of ears by flying close to the ground, con-
cliffe and Nydam (2008) have shown that this dual cealing themselves in clutter. One reason crickets,
predation pressure, from visual and auditory hun- grasshoppers, and katydids call from within vegeta-
ters, has exerted strong selective pressure on arc- tion may be to reduce the risk of being eaten by a
tiids. In spring, when birds but few bats were active, bat. Many insects emerge synchronously around
arctiids did not produce clicks, but those flying later dusk in large swarms. These swarms attract bats in
in the season did. Diurnal species in spring and sum- large numbers, but may effectively swamp the bats
mer were visually conspicuous, nocturnal species and then decline before bat activity peaks.
were visually cryptic. Some hawkmoths and beetles, The evolution of ears by moths, together with
unpalatable in some cases, also emit sounds. Jones adaptive behavioural responses to bat echolocation
224 BATS

1.6
log % nocturnal flight time

1.4

1.2

1.0

0.8

0.6
Figure 8.26 Nocturnal flight time in noctuid moths
is correlated with overall auditory sensitivity to bat
1.70 1.72 1.74 1.76 1.78 echolocation calls. From Hofstede et al. (2008), with
auditory sensitivity permission from the Royal Society.

calls, can be thought of as the second step in an high frequency CF bat species at tympanic prepara-
evolutionary arms race: a response to the first step, tions of two species of noctuid moth. The high fre-
predation by bats using ultrasound. The term ‘arms quency calls of Rhinolophus hipposideros (peak
race’ is very graphic, but it has been criticized on the frequency 110 kHz) were significantly less apparent
grounds that there must be evidence of response to the moths than those of the other species with peak
and counter-response if it is to be appropriate. An frequencies of 49–83 kHz. Bats could of course use
arms race does appear to be occurring between bats calls below the best hearing range of their prey: Ta-
and moths, since there is evidence that bats have darida teniotis, the European free-tailed bat, uses a
taken step three, and evolved ways of avoiding narrowband call of 11–12 kHz, and large moths and
detection by moths with ears. Neuroptera make up 90 per cent of its diet (Rydell
There is widespread evidence that moths with and Arlettaz 1994). Another large molossid, the Afri-
ears make up a significant part of the diet of many can Otomops martiensseni, also uses low frequency
bats, despite the moths’ defences, and in some cases calls to feed on tympanate moths (Rydell and Yalden
this is due to one of several apparently adaptive 1997) as (perhaps more surprisingly) does the spotted
responses by the bats. The first response has been bat, Euderma maculatum, a vespertilionid with enor-
the evolution of echolocation calls of very high fre- mous ears (Fullard and Dawson 1997).
quency: above the 20–60 kHz to which moths’ ears An alternative response is to use very low inten-
are sensitive. High frequency calls are used by a sity, short duration calls, for example Myotis evotis
number of bats known to feed largely on moths and M. septentrionalis (Faure et al. 1990, Faure et al.
(see Fullard 1987 for review). The highest frequency 1993). The most convincing study is a recent one by
calls recorded were by the bat Cloeotis percivali, at Goerlitz et al. (2010) of the barbastelle, Barbastella
212 kHz (Fenton and Bell 1981): several years earlier barbastellus, which emits calls up to 100 times lower
Whittaker and Black (1976) had noted that it fed in amplitude than other aerial hawkers studied.
exclusively on moths. Fullard (1987) suggested that They show that only bats emitting such calls will
such a strategy could only persist if the bats using detect moths before they themselves are detected.
high frequency ultrasound were a small component Plecotus auritus uses similar calls in Europe, but
of the bat fauna, otherwise moths would ultimately Waters and Jones (1995) found its calls to be just as
evolve higher frequency hearing, a view supported audible to moths as those of FM bats with more
by more recent research (for example Schoeman and intense calls. Anderson and Racey (1991) have
Jacobs 2003). Waters and Jones (1995, 1996) broad- shown that in captivity this bat can use prey-gener-
cast real ultrasound recordings of three FM and two ated sound to locate its prey. This is a strategy used
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 225

Figure 8.27 Frog being preyed upon by the fringe-lipped bat, Trachops cirrhosus.

by a number of other species, including Trachops a range of small vertebrates, the population studied
cirrhosus, Tonatia sylvicola (Tuttle et al. 1985), Myotis by Tuttle and Ryan (1981) on Barro Colorado Island
evotis (Faure and Barclay 1992), and M. myotis and in Panama took large numbers of frogs (Fig. 8.27).
M. blythii (Arlettaz et al. 2001). Other species may Trachops preyed on males of Physalaemus pustulosus
achieve essentially the same effect during the feed- at breeding ponds, and capture rates were signifi-
ing buzz, by reducing call intensity and duration. cantly higher when male frogs were calling to attract
Finally, some insectivorous species are known to use females. By playing recordings of frog calls to bats in
vision to locate prey when sufficient light is avail- enclosures and in the field, Tuttle and Ryan were
able, for example the California leaf-nosed bat, able to show that the bats preferred the calls of P.
Macrotus californicus (Bell 1985) and the long-eared pustulosus to those of a local poisonous species and
bat, Plecotus auritus (Eklof and Jones 2003). So, is this another species too large for the bats to handle. The
a coevolutionary arms race? The evidence is persua- frogs alter the complexity of their call, a whine fol-
sive, but not watertight (depending upon how lowed by up to six chucks, by increasing the number
restrictive your definition is), since alternative rea- of chucks in response to the calls of other males. The
sons can be put forward for some of the apparent chucks carry information on the size of the male,
counter-adaptations of bats. For example, low fre- and females prefer the more complex calls. Why
quency calls may have evolved to facilitate long- don’t the males use complex calls all the time? Pos-
range prey detection and high frequency calls to sibly because it increases the predation risk—Tra-
detect small prey. The fact that they enable bats to chops has been shown to prefer complex calls too
catch eared moths may be a beneficial side effect. (Ryan et al. 1982). However, they may use simple
calls when they can because complex calls are also
Another arms race? Bats versus frogs energetically costly, so the evidence to suggest this
Evidence also exists for an arms race between the is a response to the bats is equivocal. Bats are typi-
phyllostomid bat, Trachops cirrhosus, and its prey. cally most sensitive to the range of sound frequen-
Although this species commonly takes insects and cies used in their echolocation calls, and are
226 BATS

rarely sensitive to low frequencies (see Chapter 3).

Trachops’ sensitivity to sound is high at frequencies
greater than 15 kHz and falls in the range 15–5 kHz
as expected, but then increases again from 5–0.2
kHz: the dominant frequency range in the frogs’
mating calls (Ryan et al. 1983). Page and Ryan
(2005) have revealed a new level of sophistication
by testing the ability of the bats to adapt to changing
cues. After providing a reward when bats flew to a
loudspeaker broadcasting the call of P. pustulosus,
they then provided a reward only with the call of a
poisonous toad. Bats adapted with remarkable
speed and after a few trials switched their prefer-
ence—an important ability in a flexible predator that
is faced with frequent changes in the type of prey
that are abundant.

Carnivores—hunting by sight, sound, and echolocation

Those bats that feed on other vertebrates also have
the problem of alerting prey with their echolocation
calls. To overcome this, some appear to have a very
flexible foraging strategy, combining echolocation
with the use of prey-generated sound and vision.
Echolocation calls are typically broadband FM, and
suited to the cluttered environment in which the
bats hunt, but they are also low intensity, possibly
to reduce the likelihood of detection by prey (Nor-
berg and Fenton 1988). Several species are known
to make use of prey-generated sound, for example
the Indian false vampire, Megaderma lyra (Fiedler
1979, Marimuthu et al. 2002), and the large ears
characteristic of insectivorous bats which make use
of this strategy are found in most of the known
carnivores. Although it has not been investigated,
I suspect the large eyes of many carnivorous bats are
used to detect prey—some insectivores certainly use
vision (for example Bell 1985, Eklof and Jones 2003).
The large, forward pointing eyes of M. lyra (Fig. 8.28)
and the Australian ghost bat, Macroderma gigas, can
appear almost owl-like (Figs. 1.24 and 1.25).
Figure 8.28 Indian false vampire, Megaderma lyra, with rodent.

Bats as prey
cats, bats themselves, raptors, snakes, frogs, fish,
Evidence for predators (other than humans) having and a few large arthropods, mainly spiders, but
a significant effect on bat populations is sparse. Few also giant centipedes. Only birds and snakes appear
predators specialize on bats, but a largely anecdotal to be regular predators, and likely to have an effect
literature includes many occasional bat-eaters: on bat populations. At least five snake species in
monkeys, a loris, mustelids, racoons, opossums, Central and South America and Africa are known
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 227

Figure 8.29 (a) Bat falcon, Falco rufigularis and (b) African harrier hawk, Polyboroides typus.

to take bats roosting in caves and trees (Hill and mortality. This figure could be higher in mainland
Smith 1984), and sometimes in flight as they leave Europe. Individual owls occasionally take large
caves. The brown tree snake, Boiga irregularis, intro- numbers of bats, and could have a significant local
duced to Guam in the late 1940s is having a very effect on populations (Julian and Altringham 1994),
significant and worrying effect on the population of but in general impact is probably low: Lee and Kuo
the rare Pteropus mariannus by taking young (see (2001) estimated that red-tailed hawks and pere-
Mickleburgh et al. 1992). Owls, hawks, and falcons grine falcons took only a fraction of a per cent of
take bats throughout the world (Gillette and Kim- free-tailed bats from a large cave roost in Texas.
bourgh 1970), but I know of only one attempt to
estimate their impact on bats on a regional scale. Raptors versus bats
Speakman (1991) calculated that although bats The benefits of roosting in colonies were discussed
made up a very small component of the diets of in Chapter 6, and I also mentioned some of the
owls, the only important avian predators in Britain, potential costs, one of which was predation
they could account for 10 per cent of the annual bat (Fig. 8.29). Bats are particularly vulnerable to aerial
228 BATS

predation as they leave the roost: they often emerge Bats and plants
in large numbers, over a relatively short period, and
Over 250 species of bat eat one plant product or
at a time that can be predicted by predators. Several
another, ranging from obligate frugivores, nectari-
raptors are known to feed on emerging bats, and
vores (50 at least are flower specialists, Fig. 8.30),
large numbers hunt at some of the vast colonies of
through versatile omnivores, to insectivores that
Brazilian free-tailed bats, Tadarida brasiliensis, in the
take a little fruit when gleaning insects. Pteropodids
southern United States. Fenton et al. (1994) studied
are known to feed on plants from almost 200 genera,
the behaviour of both bats and birds at roosts in
phyllostomids from about 150 genera (Marshall
South Africa. Wahlberg’s eagle, Aquila wahlbergi,
1983, Heithaus 1982), and a more recent tally
and hobby, Falco subbuteo, stooped on free-tailed
would certainly increase considerably the 750 spe-
bats (Tadarida pumila and T. condylura) emerging
cies known to be visited by bats at the time. The
from roosts in buildings and bridges. The African
transition to nectarivory/frugivory in one form or
goshawk, Accipiter tachiro, pursued bats in flight.
another has been a major factor in the adaptive
The raptors appeared at dusk, and either flew close
radiation of the phyllostomids. For example, several
to the roosts or perched nearby until the bats began
of the sub-families of the Phyllostomidae appear to
to emerge. Some birds were also present for the
have undergone their major radiations following
dawn return of the bats. Half of the attempted at-
their independent evolution of nectarivory/frugiv-
tacks were successful, and handling time (0.5–5 min)
ory (van der Pijl 1972). In turn, the bats appear to
was the major factor influencing overall predation
have had their effect on the adaptive radiation of
rate, given the short emergence time (10–30 min).
the plants. Vogel (1969) suggested that bat pollina-
Bats were usually taken to a perch to be eaten, but
tion has arisen independently in at least 27 New
some birds reduced their handling time by eating in
World plant families, affecting over 500 species.
flight. A successful bird could meet more than 50 per
Similarly, the use of bats as major seed dispersers
cent of its daily food requirements. Bats appeared to
also appears to have arisen independently on many
use three strategies to minimize risk: late departure
occasions—but I can’t find supporting evidence for
from the roost at small colonies, emergence in clus-
this! Bat–plant interactions therefore provide enor-
ters, and roost switching. Large colonies emerged
mous potential for the study of a wide range of
earlier than small ones, perhaps gaining from the
ecological and evolutionary principles and this
safety in numbers factor. Since insect populations
peak in the early evening, early emergence is desir-
able but increases the risk of predation by diurnal
predators: the actual emergence time is probably
a compromise between these conflicting factors
(Jones and Rydell 1994). Serotines, Eptesicus seroti-
nus, increased clustering when a tame barn owl
(Tyto alba) was placed outside the roost (Petrzelkova
and Zukal 2003), since it probably reduces the abil-
ity of a predator to fix on a single target, as well as
simply reducing the probability that a particular bat
will be taken. Clustering may have other functions,
particularly where aerial predators are less com-
mon. In some cases, clustering behaviour may be
related to foraging strategy. Evening bats (Nycticeius
humeralis) leave the roost within 10 s of each other
far more often than predicted by chance: females
unsuccessful at foraging earlier that evening follow
other bats to increase their chances of feeding (Wilk-
inson 1992b). Figure 8.30 Macroglossus woermanni, African long-tongued fruit bat.
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 229

potential has certainly been exploited by research- pollinated only by a single bat species, Helversen
ers. Wide-ranging reviews have been written by and Winter (2003) estimated that 60 per cent of all
Fleming (1982), Heithaus (1982), and Marshall species visited by glossophagine bats were polli-
(1983) covering the early literature of New and Old nated almost exclusively by them and only 30 per
World bat–plant interactions. Research has gained cent were generalists. Although pollination by sin-
momentum in recent years, with the growth of fields gle species is rare, pollination by one or a few func-
such as community ecology, ecosystem services, tional groups appears to be the norm (Fenster et al.
and conservation biology, and much of the recent 2004).
work has also now been reviewed by Helversen and The Old World fruit bats have been frugivores
Winter (2003), Dumont (2003), Muscarella and throughout their known history and evidence for
Fleming (2007), and Fleming and Muchhala (2008), coevolution, true or diffuse, is best sought in the
each focusing on a particular topic. I will cover the plants. The first insectivorous bats to evolve frugiv-
field in two broad areas that could be thought of as orous habits, the phyllostomids of the neotropics,
the evolutionary and the ecological: (1) what the probably did so in association with plants that
interaction can tell us about coevolution and mutu- were preadapted to bats, either through visitations
alism in the organisms themselves; and (2) the eco- from other neotropical, nocturnal animals (Sussman
logical aspects of pollination and seed dispersal. and Raven 1978), or through evolution from ptero-
podid-pollinated ancestors in the Old World (Baker
Coevolution? 1973). The first steps into coevolution in the New
The mutual dependence that exists between some World were probably made by the bats. Baker (1973)
bats and plants has led to the suggestion that coevo- drew attention to the danger of too readily using
lution has occurred frequently. Strictly speaking, the coevolution as an explanation for many bat–plant
term should be used to describe only the process by relations. Ceiba and Parkia species, bat-pollinated
which one species is the source of the selective forces plants that have been used to define bat flowers,
which determine the evolution of another, with the must have existed in both the Old and the New
changes (morphological, physiological, or beha- Worlds before the evolution of either pteropodids
vioural) in one inducing changes in the other. In or phyllostomids. To become the principle pollina-
time, the association may become so specialized tors of these plants, bats in both Worlds must have
that the two species are mutually and totally depen- displaced other pollinators. Adaptation in the plants
dent upon each other for their existence. True coevo- to facilitate the switch presumably offered signifi-
lution is probably much rarer and less important as cant benefits, such as longer range pollination and
an evolutionary force than once believed (for exam- dispersal relative to smaller or non-volant pollina-
ple Nitecki 1983, Futuyma and Slatkin 1983) and tors/dispersers. The small glossophagine bats
many examples may now be explained in terms of appear to have displaced, or in many cases now
diffuse coevolution or preadaptation. Diffuse coevo- share their plants, with birds and hawkmoths
lution occurs when more than one species, and often (Sphingidae), an obvious step—all are small and
many, are involved on both sides of the interaction. capable of hovering flight. Adaptation to pollination
The degree of specificity and mutual dependence in by glossophagines may have evolved indepen-
the interaction between any two species is therefore dently more than 100 times in neotropical plants.
likely to be low. Preadaptation is when one species Of course there are also examples in which the
is able to interact with another simply because inde- plant gains no benefit and the bats are robbers:
pendent evolutionary pressures have, by chance, some glossophagines take nectar by inserting their
given it the attributes which enable it to do so. The tongues between petals from the side of the flower.
evolution of bat–plant interactions appears to have
New World versus Old World
been driven primarily by diffuse coevolution, as
appears to be the case for the majority of pollination We saw in Chapter 6 that a comparison of palaeo-
systems (Waser et al. 1996, Fenster et al. 2004). How- tropic and neotropic frugivores revealed morpho-
ever, although few (perhaps no) plant species are logical differences reflecting the environments in
230 BATS

which they evolved. Fleming and Muchhala (2008) variety of chemicals common to a wide range of
extended this approach by comparing Old and New non-bat flowers. Only the baobab, Adansonia digi-
World nectarivorous bats and birds, and the rela- tata, had significant quantities of sulphur com-
tionships they have with their food plants. There pounds in its flower scent. More interesting still
was a consistent trend towards greater diversity in was the finding that the kapok, Ceiba pentandra,
the New World. Both bats and birds showed greater which is native to both Africa and South America,
species richness, greater morphological diversity in has sulphur compounds only in the New World. It is
their more specialized taxa, and their food plants pollinated by bats on both sides of the Atlantic, but
also showed greater taxonomic and ecological diver- the phyllostomids and pteropodids clearly exert
sity and again, greater morphological specialization. very different selective pressures.
Two factors appear to explain the richness of the Bat-pollinated flowers are typically either dull
neotropics relative to the palaeotropics: the greater and visually inconspicuous or white. Colour ap-
diversity and greater predictability of plant re- peared to be largely unimportant to the nectarivor-
sources in the New World favoured radiation into ous phyllostomids since, until very recently, all bats
numerous specialized niches. The comparison of were believed to be colour-blind. Recent evidence
birds and bats is interesting too, but I have no shows that Glossophaga soricina and Carollia perspicil-
space for it here! lata (and probably many other species) have the
Let’s start our more detailed study by looking at necessary eye structure for dichromatic colour
adaptations (real or apparent!) in bat-pollinated vision and ultraviolet vision (Muller et al. 2009).
plants, move on to the bats and then the interactions. Winter et al. (2003) having noted previous studies
showing that some bat-pollinated plants have flow-
Bat plants: flowers
ers that reflect UV light strongly, were prompted to
The flowers of plants have evolved mechanisms that conduct experiments that showed that some species
attract bats and ensure that pollen is deposited on of bat are indeed sensitive to UV light. It now seems
the bat for transfer to the next flower. However, a that they may also take some notice of colour—data
complex cost-benefit mechanism is in operation. The from the same group show that Carollia (but not
flower must provide enough nectar to attract bats, Glossophaga) can differentiate between red and
but little enough that bats make repeated visits and green in behavioural experiments. However, Glosso-
transfer pollen efficiently. Sufficient nectar must also phaga (and probably other species too) can locate
be produced to effectively compete with other flowers with other cues. Mucuna holtonii have flowers
plants of the same and other species. One way to that, due to their particular shape, are very conspicu-
limit the cost of nectar production is to make it ous to echolocating bats (Helversen and Helversen
profitable for bats to visit in ways that are less costly 1999, see Chapter 3) making them easy to locate in a
to the plant, for example by reducing search time, cluttered environment. Bat-pollinated plants have
handling time, and predation risk (see Chapter 6). It evolved to advertise themselves in many ways!
therefore pays to have conspicuous, easy to find Bat flowers may be large and bell-like, with large
flowers from which nectar can be obtained with nectaries deep in the bowl, and others take the form
minimal effort and away from predators. of ‘shaving brushes’ with exposed nectaries (Fig.
Bat-pollinated flowers frequently have a strong 8.31). Some flowers have landing platforms for
odour that is almost invariably unpleasant, to humans bats. Nectaries are often large and may hold as
at least. A surprising number of unrelated plants in much as 5–10 ml of nectar when full. The stamens
the New World produce odours using the same fam- are placed to leave pollen on the bat—from a heavy
ily of sulphur-containing compounds that suggests covering over much of the body, to a strategic dab
convergent evolution amongst bat-pollinated plants on the top of the head. Mucuna holtonii is unusual, at
(Pettersson et al. 2004). Just why phyllostomid least among bat-pollinated plants, in releasing pol-
bats find them so enticing is not known, but field len explosively when first visited (Helversen and
trials show that they do. In contrast, eight species of Winter 2003). Attracted by the vexilla, a sound-re-
tree visited by pteropodid bats in Africa had a flecting petal bowl, the first bat pushes its head into
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 231

the flower, gets its reward of about 100 μl of nectar some extent these patterns must reflect coevolution
and simultaneously bursts the keel to catapult pol- between a plant and its pollinators.
len over its rump. Subsequent visitors get only 20 μl
Bat plants: fruits
from the same flower for their pollinating duty, but
are encouraged to continue visiting the plant as the Dumont (2003) has written a very thorough review
inflorescence continues to open new flowers until that brings all of the important elements together so
about midnight. Helversen and Helversen (1999) I will cover only the essential points. First, a number
and Helversen and Winter (2003) provide some of general considerations. Bats eat ripe fruit most of
fascinating detail. For example, bird-pollinated and the time and seeds pass through the gut with the
palaeotropic species of Mucuna do not have the erect, probability of germination unchanged or even
concave vexilla that attracts echolocating bats in the enhanced (for example Shilton et al. 1999, Heer
neotropics, and they suggest that the soft hairy ce- et al. 2010). Due at least in part to the weight con-
phalium that surrounds the flowers of many colum- straints of flight, many bats process fruit before
nar cacti may serve to enhance the acoustic contrast eating it, dropping pulp and seeds varying distances
between it and the flowers within (Fig. 8.31). from the parent plant. As discussed in Chapter 6,
Bat plants are frequently trees and vines, and pteropodids are typically generalists due to their
several other characteristics make it easier for bats dependence on a wide range of plant species,
to find and get to the flowers. For example, flowers many of them thinly scattered big-bang strategists
are often pendulous, they are typically held free of (for example many Ficus species). The smaller phyl-
obstruction, for example on the periphery of the lostomids of the New World are more varied spe-
tree, and the tree may even be leafless during flow- cialists, due to the greater abundance and diversity
ering (Vogel 1969). They frequently open only at of food plants, a high proportion of which are
night (and perhaps only for one night), and some- steady-state producers (for example Cecropia, Piper,
times only for a limited time after dusk. Temporal and Solanum species). What all of these bats have in
differences in flowering and fruiting patterns are common is that they are looking for fruit. What cues
also important and can be characterized by the two are they seeking? In the neotropics, bat fruit is gen-
extremes seen (Heinrich and Raven 1972, Gentry erally a dull brown or green, suggesting that vision
1974, Heithaus 1982). Some plants exhibit ‘big- is not very important. This is supported by the work
bang’ reproduction, producing vast numbers of of Kalko and colleagues, described in Chapter 3,
flowers over just a few days. They attract many which indicates that olfaction and echolocation are
opportunistic foragers including bats, other mam- the primary senses, with the dominant sense vary-
mals, insects, and birds. Flower shape and nectar ing between species. Many figs hide their fruit
production are such that few potential pollinators amongst foliage, which also suggests vision is not
are excluded. At the other extreme are the ‘steady important. Lacking echolocation, Old World fruit
state’ plants that produce just a few flowers each bats rely on sight and smell and the majority of bat
night over a long period: from one month to as fruits are either pale or orange to red and most are
much as ten months, perhaps with only one flower displayed in the open. Those that are not presum-
opening per night in some species. Pollinators have ably rely on smell, but the study of this is not as
time to memorize the location of flowering trees and advanced as that of flower odour.
visit repeatedly, often by trap-lining, but because What is there to attract bats to particular fruits
there are few flowers per tree they must visit several once they have found them? The nutritional value
trees, improving pollination success. Bats have of bat fruits varies considerably (Dumont 2003). Figs
excellent spatial memory and are able to relocate tend to be high in indigestible fibre and low in
plants and even individual flowers with consider- proteins and lipids, perhaps surprising given the
able precision. Flower shape may restrict access to number of both Old and New World species that
these regular visitors and discourage opportunistic eat them. However, small phyllostomids that rely
foragers. Between the two extremes there are many heavily on figs appear to be very selective (Wendeln
other patterns, including the ‘multiple-bang’. To et al. 2000, see Chapter 6). Steady-state producers in
232 BATS

Tricanthera sp.
Durio zibenthus
(durian)

Adansonia digitata
(baobab)

Freycinetia reinechei (liana)

Musa sp.
(banana)

Angophora coscata
(apple gum tree)

Carnegia gigantea
(saguaro cactus)

Agave palmeri

Figure 8.31 Bat plants.

 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 233

the understorey, like Piper and Solanum, are usually tubes. The range was 55–77 mm compared to 11–24
those favoured by smaller bats and have lower fibre mm in the non-glossophagine, opportunistic nectar
and higher protein and lipid content. Figs do appear feeding Carollia perspicillata. The record is held by
to be an important source of calcium. Many phyllos- the recently discovered Anoura fistulata at 85 mm,
tomids thought of as frugivores certainly supple- 150 per cent of its body length (Muchhala 2006)—
ment their diet with insects. second only to chameleons within the vertebrates.
Bat fruit also vary in handling time, through vari- When not in use, this remarkable tongue is retracted
ation in size, shape, and hardness and the way they into the thoracic cavity with its base between the
are presented on the plant. Fruit size is sometimes heart and the sternum! This glossal tube is similar to
correlated with the size of the bats that eat them, but that independently evolved by ant-eating pangolins.
there are many exceptions: small bats will eat large The tongues of some nectar feeders are feathered
fruit in situ and some large bats will eat small fruit. to improve nectar retention. Glossophagines also
Hardness too varies, and some fruit can only be often have long slender muzzles and sometimes a
eaten by bats with the necessary jaw strength and reduced dentition. In stark contrast, the robust
structure. There is as yet little evidence to suggest skulls and jaws of fruit-eaters are clearly adapted
that any plants have evolved fruit that are more to their food (Dumont 2003). They presumably have
accessible to bats other than in presenting them for the musculature to match, but a simple analysis of
easy removal or to aid landing. However, differ- bite force (Aguirre et al. 2002) found no difference
ences in hardness and size probably contribute to between insectivores, omnivores, and frugivores a
resource partitioning in fruit-eating bats. sample of 20 species from a tropical savanna, after
Only those plants that have evolved close and correction for size. However, differences in biting/
specific relations with particular bats are likely to chewing behaviour can have a major effect on force
have fine-tuned their bat-attracting qualities but, as and efficiency (Dumont 2003).
discussed, plants rarely depend entirely on one spe-
cies of bat or even on bats alone. Similarly, the bat Bat–plant interactions: coevolution and mutualism
may feed on many species of plant. Some bats even The relationships that have evolved are too numer-
eat many of the flowers they are pollinating, groom ous and complex to describe here, and the reader
the pollen from their fur before it can be transferred should go to the references I gave at the beginning
to another plant, or steal nectar without pollinating of this section for information on particular bats and
the flower. Similarly, the benefits to the plant of plants. I want to look at a few examples, starting with
having its fruit eaten by bats may not always be one that perhaps is an example of true coevolution.
greater than those of other dispersers. The partners The record-breaking tongue extension of Anoura
in a mutualism are both in it for what they can get— fistulata (Fig. 8.32) enables it to feed from the 80–90
they are exploiting each other. mm long, slender corolla of Centropogon nigricans. No
other animal has been recorded visiting the plant,
Adaptations in the bats
suggesting that it is probably the only pollinator.
All bats have adaptations to suit their foraging style Is coevolution also evident in the relationship
and diet. My aim here is to highlight briefly the between Oroxylum iridicum and the small pteropo-
most striking and interesting (bias acknowledged!) did, Eonycteris spelaea? Gould (1978) found that
that have not already been discussed. If you want it roosted in caves and flew long distances to its
morphological or physiological detail look at the feeding grounds, arriving well after dark. It fed on
reviews listed at the beginning of the section. As- the nectar produced by the flowers of the tree Orox-
pects of this topic are also covered in Chapters 2 ylum iridicum, which did not open until about two
(flight morphology and performance), 3 (echoloca- and a half hours after dusk, and dropped before
tion and other senses), and 6 (feeding ecology). dawn. This suited the late arrival of E. spelaea and
The long tongues of glossophagine bats are well prevented earlier feeders from getting the nectar.
known. Winter and Helversen (2003) measured the The flower and bat fit together snugly and the
operational length in nine species feeding from glass bat’s tongue is just long enough to reach the
234 BATS

Figure 8.32 Anoura fistulata taking nectar from the flowers of Centropogon nigricans.

nectaries, but only if the bat pushes far enough in to Agave palmeri, in the Arizona summer. Flowers of
ensure a good covering of pollen on head and the agave produce most of their nectar early in the
shoulders. The flower has to be tipped on its rigid evening, with minimal production in the second half
stem by a bat the size of E. spelaea before the nec- of the night. Flocks of L. yerbabuenae must feed at the
taries charge the tuft of hairs that the bat feeds from. flowers of several plants in one night for foraging to
Since only a little nectar is given up each time, the be efficient. The optimal foraging strategy involves
bat is forced, as are most pollinators, to move repeat- leaving a plant before exhausting the nectaries and
edly from flower to flower—something larger pter- moving on to a previously unvisited plant, rather
opodids do less frequently. The bottom line is that than returning to those visited earlier in the night.
other bat species did not appear to feed at this tree. The pattern of nectar production ensures cross-pol-
However, although the plant is highly dependent on lination, since this plant is not intra-fertile. The
the bat, the bat feeds on many other tree species, and heads and shoulders of the bats are covered in pol-
there is no direct evidence that E. spelaea has made len after feeding, the pollen clinging to specially
specific evolutionary responses to Oroxylum—it adapted hairs. Much of it is groomed off and eaten
simply has the long muzzle and tongue, and other and its amino acid composition appears to be
characteristics, common to many nectar-feeding matched to the physiological needs of the bat (Ho-
bats. well 1974). Again, however, the dependence is not
Leptonycteris yerbabuenae (previously L. sanborni) is absolute for either partner. The agave is also polli-
a nectar-eating phyllostomid that feeds on a broad nated to a limited extent by sphinx moths and car-
range of plants in Central America and the south- penter bees, as are other bat-pollinated succulents
western United States, but it may depend primarily like the saguaro cactus. Seed set losses are respec-
on a single species at certain times of the year. Ho- tively only 16 and 28 per cent lower when the
well (1979) studied its interaction with the agave, saguaro is pollinated by moths and bees rather
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 235

Figure 8.33 A New Zealand short-tailed bat, Mystacina tuberculata, feeding on the nectar of Dactylanthus. See text for details.

than by L. yerbabuenae (McGregor et al. 1962). As In this relationship we have one of the world’s most
already stated, L. yerbabuenae is a generalist, feeding unusual and endangered bats in a mutualistic rela-
on a wide range of plants, presumably specializing on tionship with an unusual and endangered plant. The
A. palmeri only when it is energetically advantageous. study of this interaction may be important to the
Close, mutual dependence between just two spe- survival of both species.
cies is an unstable situation, since a decline in one Finally, a few words on competition and ensemble
species (even temporarily) will inevitably lead to a structure. Helversen and Winter (2003) provide an
decline in both, possibly to extinction (May 1981). interesting analysis of glossophagine bats based on
This may be why such systems are hard to find in the size-dependence of daily energy requirements.
nature and why mutualisms are generally more dif- Glossophagine bats, like hummingbirds, have high
fuse: dependence between species is not absolute, daily energy requirements, in fact the highest so far
and more than two species are usually involved. recorded in mammals. With some simple and rea-
Begon et al. (2006) provide an excellent summary sonable assumptions, backed by much detailed
of mutualistic relationships, pointing out that they work of their own and analysis of that of others,
make up most of the world’s biomass! Ecologists they calculated the minimum nectar requirements
and ecology textbooks for a long time concentrated per flower for bats of 9, 11, and 18 g (Fig. 8.34) for
on competition and predator–prey relations, but different inter-flower distances. Larger bats need
the importance of mutualism is now universally more food each day and must be able to gather
recognized. that food efficiently. Plants whose flowers yield
The endemic Dactylanthus taylorii is New Zeal- only small amounts of nectar per visit are unlikely
and’s only completely parasitic plant. It has large, to be exploited by larger bats, but small bats can
dull flowers on unisexual plants that produce lots of profitably visit all flowers. A 9 g bat can make a
nectar and a strong odour. The only real candidate living from flowers that deliver about half as much
for its principle pollinator is another endemic spe- nectar as that needed to sustain an 18 g bat with the
cies, the short-tailed bat, Mystacina tuberculata. It has same average inter-flower flight distance.
been seen to make up to 40 visits per night to take Since territoriality appears rare or absent in glos-
nectar (Fig. 8.33) over a four week period (Ecroyd sophagines and there is no evidence to suggest that
1993). Mystacina is the only insectivorous bat outside any flowers favour visitation by larger bats, this
the phyllostomids known to feed on plant products. simple energetic model may underlie ensemble
236 BATS

and appear to be wide-ranging opportunists. At

per flower (ml 17% sugar)

L. robusta
minimum nectar volume

30 G. soricina only 7 g, L. obscura should not be a specialist, so its

18 g 11 g position in the ensemble is a mystery, as is that of
G. soricina.
20
G. commissarisi Pollination and seed dispersal
9g
10 For many years the importance of bats as pollinators
and seed dispersers was greatly underestimated and
0 only now are we beginning to appreciate both the
0 50 100 magnitude of their role and the unique aspects of
mean inter-flower distance (m)
this relationship. Since so many plant species are
adapted to exploit bats as pollinators and seed dis-
Figure 8.34 Larger bats require more nectar per flower than smaller bats
at any given inter-flower distance. From Helversen and Winter (2003), used persers, bats influence their density and distribution
with permission from University of Chicago Press. through the roles they play in key ecological pro-
cesses such as colonization and succession. Ecosys-
tems are in constant change as organisms invade
structure. In a habitat where nectar supply is limited new habitat and compete with residents and other
larger species would lose the competitive exclusion colonists to make a living in that habitat. Bats can
battle with smaller species that can exploit the wider assist plants since they contribute to two critical
range of nectar volumes supplied by different processes: more efficient pollination can increase
plants. These predictions were borne out by experi- the number and quality of propagules produced
ments on captive bats: when nectar supplies to arti- and more effective dispersal increases the chances
ficial feeders were limited the 10 g Glossophaga of a species maintaining its place in an existing
soricina out-competed the 17 g Choeronycteris mexi- habitat and colonizing new habitats. Bats can poten-
cana. The larger bats have at least one card to play. tially transfer pollen over considerable distances, an
Flight may be more expensive for a large bat, but it is important factor for those plants that are randomly
cheaper per gram of body mass and bigger bats fly spaced at low density, and also important in main-
faster. A bigger bat also carries more body fat. taining a large breeding population (Heithaus 1982).
Larger species can use their stored energy reserves Different bats may respond to the same plant in
not only to survive periods of poor foraging, but different ways. The large Phyllostomus discolor and
also to forage over wider areas in search of more the small Glossophaga soricina both feed on the nectar
ephemeral, but perhaps more rewarding, flowers. of Bauhinia pauletia. P. discolor empties the flower in
This is an option not open to smaller species that a single visit and forages in flocks, possibly to
are constrained to operate in smaller home ranges reduce the chance of an individual bat visiting a
and cannot survive long periods of starvation. Coex- flower that has already been emptied. G. soricina
istence is maintained through competition of species forages along traplines, taking only a little nectar at
with fundamentally different approaches: specia- each flower, and it can return to a single flower
lists versus generalists. How do these predictions many times (Heithaus et al. 1974). In terms of the
fit observations in the real world? A study of the quantity of nectar provided, the distribution of
bats of La Selva biological station in Costa Rica food is perceived differently by the two bats, and
(Tschapka 2004) provides an answer. Lowland for- they have evolved different strategies to exploit it.
est was dominated by resident populations of two The plant, faced with this diversity, must evolve
small species, Hylonycteris underwoodi (7.6 g) and a nectar producing strategy which is a compromise
Glossophaga commissarisi (8.8 g). Two larger species, between the needs of these and perhaps other
Lonchophylla robusta (15.7 g) and G. soricina (11.5 g), pollinators.
and the small Lichonycteris obscura (7 g) are seasonal By evolving mechanisms to disperse its seeds, a
visitors. L. robusta and L. obscura move into the area plant can gain in several ways, through increased
during the nectar-rich months (October–February) gene exchange, enhanced colonization of new sites,
 BIOGEOGRAPHY, MACROECOLOGY, COMMUNITY ECOLOGY 237

reduced competition, and reduced seed predation. late successional canopy trees. A major exception to
Those bats that transport large numbers of seeds this rule is the bat–fig relationship. Both Old and
long distances without damaging them are thus New World bats have a major role in the early succes-
likely to be beneficial to plants. Many pteropodids sion of figs (Ficus). Whenever figs play an important
and phyllostomids fall into this category. Morrison role in succession (which is often), then bats are
(1978) calculated that Artibeus jamaicensis disperses important too. Artibeus, Corollia, Sturnira, and the pre-
at least 7 per cent of the annual fig crop on Barro dominantly nectarivorous Glossophaga are common
Colorado Island in Panama, and pteropodids con- phyllostomid genera that specialize on early succes-
sume large numbers of fruits before commuting up sion plants. Their palaeotropical counterparts include
to 50 km to their day roosts. Bonaccorso et al. (1980) Cynopterus, Nyctimene, Epomops, and the nectarivor-
found that of the 16 species of mammal (including 4 ous Syconycteris. Although larger than the 10–20 g
bats) that ate the fruits of Dipteryx on Barro Color- phyllostomids, they are usually less than 60 g. Even
ado Island, only the phyllostomid Artibeus lituratus within the phyllostomids there are behavioural differ-
was a ‘high-quality’ disperser. One of the other spe- ences that will influence the way they disperse seeds.
cies that feeds on Dipteryx fruits, Carollia perspicillata, For example Corollia and Glossophaga typically feed
is too small to disperse the seeds, but is the principle quickly, swallowing soft fruit and their seeds, which
disperser for several Piper species and Solanum haye- can pass through the gut in 30 min or less. Artibeus
sii (Heithaus and Fleming 1978). If a plant depends and Sturnira more often feed more slowly on harder
on several pollinators, pressures on the evolution of fruit, chewing more, spitting out the pulp and swal-
fruit size and shape may be conflicting: large bats lowing fewer seeds. With the exception of Syconycteris
can carry fruit away from the parent tree, promoting the pteropodids are slow feeders that produce spats
seed dispersal, but a small bat may need to eat the as they eat, which are typically dropped under the
fruit at the parent tree, and uneaten seeds will fall parent tree. Slow eaters are therefore less likely to
below it. If gut transit times are short, as is the case disperse seeds in quantity over long distances. This
for a large proportion of the seeds ingested, few is not to say that they are not capable of being long-
seeds will be dispersed over significant distances in distance dispersers: Shilton et al. (1999) have shown
the faeces. However, the damage done to seed cases that Cynopterus can retain viable seeds in the gut for
in their passage through the bat, and the nutrient more than 12 hours and thus have the potential
content of the faeces in which they are deposited, for spreading seeds, even between islands. The recol-
may improve germination. A recent paper by Heer onization of Krakatau after the eruption of 1883
et al. (2010) demonstrates very clearly the benefits has been studied in some detail. Early colonization
many fig seeds gain from passage through Artibeus by figs could have been either bird- or bat-assisted,
jamaicensis. but by 1919 Cynopterus sphinx had colonized the is-
Muscarella and Fleming (2007) have recently rev- lands, Rousettus amplexicaudatus was found in 1933,
iewed the role of frugivorous bats in tropical forest and Pteropus vamprus was established by 1985 (Whit-
succession in both Old and New Worlds. Given the taker and Jones 1994). By 1992, over 40 of the 124 plant
ecological and evolutionary differences in bat–plant species were bat/bird dispersed and 30 more were
interactions in these two worlds, we will not be brought by birds but subsequently dispersed around
surprised to see that there are substantial differences the islands by bats. The role of bats in succession is
here too. The majority of the New World phyllosto- clearly important, but apparently not as dominant as
mids are small, manoeuvrable understorey specia- in the New World, and this appears to hold true for
lists and play a critical role in colonization and early secondary succession in other Old World studies (for
succession, because of their long evolutionary asso- example Ingle 2003 and see Muscarella and Fleming
ciation with the numerous shrubs and other early 2007 for additional references).
succession plants that produce small fruits with Bats have a role beyond early succession. On the
abundant small seeds. Pteropodids, large and fre- Rio Manu in the Peruvian Andes, bats dispersed 25
quently confined to the canopy where they generally per cent of the plants that had become established on
feed on large fruit, more commonly disperse larger, newly exposed river sandbanks (Foster et al. 1986)
238 BATS

and two bat-dispersed species dominate plant bio- different—they are far more likely, for example, to
mass. Even 500 years post-disturbance, two of the drop seeds in forest clearings and other open, dis-
four dominant canopy trees are bat-dispersed, the turbed areas.
other two wind-dispersed. Muscarella and Fleming
Pteropodids and Pacific island floras
(2007) give a number of other neotropical examples,
but pteropodid bats are less well studied. Hodgkison The depauperate faunas of many Pacific islands
et al. (2003) found that dietary overlap with birds was include few pollinators of the native flora: Ptero-
minimal, but significant between bats and other podids are the only indigenous, frugivorous mam-
mammals. However, other mammals were fre- mals on most islands, and many islands have few
quently seed predators that did not contribute to pollinating birds, or even insects, relative to conti-
dispersal. They determined that around 15 per cent nental areas. The role bats play in maintaining the
of tree species in a 1 ha plot of old growth forest in diversity of the native flora, and in turn the fauna,
Malaysia relied on bats for pollination/dispersal. may be crucial (for example Cox et al. 1991). Many
Banack (1998), in a dietary study of Pteropus samoen- Pacific island bats are endemic to single or small
sis and P. tonganus on the Samoan archipelago, found groups of islands, and are under threat
a marked preference for the fruits of primary over from hunting and habitat destruction. At least one
secondary forest tree species. species has become extinct in recent years, and
Due to differences in food preferences, food populations of several other species are probably
handling, and other aspects of behaviour, bats too low for them to be effective pollinators and
often play a role that is distinct from that of birds. seed dispersers (for example McConkey and Drake
Because bats more often carry food away from the 2006), which introduces the next and final chapter,
parent tree than birds, and because they more often Conservation, in which I will come back to this
defecate in flight, patterns of seed dispersal are often topic.
 CHAPTER 9

Conservation

Why conserve bats? An overview of current status. Identifying and predicting conservation needs.
Population decline, range contraction, and extinction. Predicting extinction risk. Threats to bats—habitat
loss, alien species, hunting, climate change, development, and other factors. From negative to positive—how
to conserve bats. From reserves to landscapes, agroforestry, logging regimes, agriculture, evidence-based
conservation and mitigation, education. Bats as bioindicators.

Although bats have some more or less unique con- Whether we study bats as professional biologists
servation problems, they are basically subject to the or as amateur naturalists, or simply watch them in
same biological processes, and therefore conserva- our gardens and on television wildlife programmes,
tion concerns, as other animals. Much of the rich and they enrich our lives. They are another facet of the
complex conservation literature is therefore relevant natural world to marvel at and to try to understand.
to bats. I do not want to discuss conservation theory The study of the natural world, and the communi-
in any detail, but I do want the reader to see the links cation of what we learn to a wider public, is an end
between conservation theory, broad global con- in itself: a part of our culture (Fig. 9.1). As I hope this
cerns, and the specifics of bat conservation. I will book shows, when we study bats we also learn a
therefore outline some very general (and sometimes great deal about the ways in which the natural
self-evident) principles as and when they seem world works. The study of bats has had frequently
appropriate, but for more detailed information I unsuspected medical and economic benefits, as
suggest looking at one of the many excellent books basic research has led on to strategic research.
on conservation biology. A look at my shelf suggests Despite this, society continues to place too little
Pullin (2002, a good general text), Whittaker and value on the natural environment and the study of
Fernandez-Palacios (2007, because islands are so how it works.
important in conservation and bat conservation
in particular), and Sutherland (2000, as a bridge
An overview of current status
between conservation theory and practice).
Approximately 25 per cent of the 1,100 or so bat
species are globally threatened (Mickleburgh et al.
Why conserve bats?
2002), primarily by habitat destruction, degradation,
Why conserve bats? First and foremost, for a reason and fragmentation, misguided persecution, and
we do not see stated as often as we should in con- exploitation for food. Bat conservation has made
servation arguments: simply because they are, like enormous strides over the last 20–30 years but, like
all other organisms, part of our astonishingly rich conservation in general, continues to be hampered
biosphere. Bats are part of the global ecosystem, by a lack of funding, rooted in a widespread reluc-
with a part to play in its continuing evolution. tance in politicians and big business to take conser-
Must we justify their existence only in terms of vation seriously. We lack the resources needed to
what they can do for humans? Sadly, all too often gather the information that is essential for effective
the answer appears to be ‘yes’. conservation planning and the resources to fully

239
240 BATS

Figure 9.1 Taking bats to a wider public—working with the BBC Natural History Unit on a live broadcast from a cave hibernaculum.

implement the plans that are developed. Perhaps bats have in many parts of the world can make
the most hard-pressed area is that of monitoring conservation an uphill struggle. However, over the
(short- and long-term), which severely compromises last decade or two, there have been highly successful
our ability to practice evidence-based conservation: education programmes in many parts of the world
the value of a great deal of what we do in the name that have dispelled damaging myths about bats and
of conservation remains untested. In many parts of stressed the important roles bats have in natural
the world we have yet to gather even basic data on ecosystems. Attempts to give bats adequate legal
the number, distribution, and status of species pres- protection have not always been so successful and
ent. Surveys repeatedly result in major revisions of they have little or no legislation to protect them in
regional bat faunas (for example Francis et al. 1999, many parts of the world: in some countries they are
Bates et al. 2000, Pottie et al. 2005). A high proportion still classed as vermin! The IUCN Chiroptera Spe-
of the species at greatest risk are island endemics cialist Group produced action plans for Old World
that face multiple threats: habitat loss, hunting, and fruit bats (Mickleburgh et al. 1992) and Microchir-
intense tropical storms. A disproportionate number optera (Hutson et al. 2001) which have been the
are pteropodids (Mickleburgh et al. 2002) and many focus of a large number of important initiatives.
are keystone species—important pollinators and Bat populations in many parts of the world are
seed-dispersers in depauperate ecosystems. The loss clearly in decline and the ranges of many species
of some of these bats could therefore be catastrophic have contracted. The evidence for this is over-
to their island ecosystems. Members of the smaller whelming, albeit scattered and anecdotal for the
families are also among the more threatened, such as great majority of species (Mickleburgh et al. 1992,
the Craseonycteridae, Myzopodidae, and Furipteri- Hutson et al. 2001, IUCN 2009). Eight species have
dae. Some countries, such as Madagascar and Japan, not been recorded for fifty years or more, and a
have a high proportion of endemic species and a bat further four have almost certainly been driven to
fauna that is under more widespread threat. extinction more recently (IUCN 2009). The list in-
Good conservation is generally underpinned by cludes large Old World fruit bats and small insectiv-
good education and legislation. The negative image orous bats, but a few are so poorly documented that
 CONSERVATION 241

their distribution and true status as species are to avoid them. Applying the ‘precautionary princi-
uncertain and will remain so: ple’ may be necessary before population data are
available. We can also plan at a more strategic
 Pteropus pilosus: Palau, Micronesia, probably lost
level: some ecological traits or suites of traits, wide-
in the late-nineteenth century.
spread within a particular genus or even family,
 Nyctimene santacrucis: Solomon Islands, last rec-
may make them unusually vulnerable (for example
orded in 1907.
those of slow-flying, gleaning specialists, as we’ll see
 Acerodon lucifer: Philippines, notified as extinct in
later). This knowledge can be used for longer term,
1996.
more global planning, as well as in local decision
 Pteropus tokudae: Guam, last recorded in 1967.
making. Let’s look at these issues in a little more
 Pteropus brunneus: Percy Island, Australia, last
detail.
recorded in 1878.
 Pteropus loochoensis: Japan, questions have been
raised about its very existence. Population decline, range contraction,
 Pteropus subniger: Mauritius and Réunion, proba- and extinction
bly lost in the nineteenth century.
Some of the easiest questions to ask a biologist are
 Phyllonycteris major: Puerto Rico, extinction date
among the most difficult to answer—and they are
unknown.
often central to conservation. When giving public
 Pipistrellus sturdeei: Japan, notified extinct in 2000.
lectures I am frequently asked apparently simple
 Mystacina robusta: New Zealand, last seen in 1967.
questions that I am expected to know the answers
 Nyctophilus howensis: Lord Howe Island, known
to. How many bats are there of a species? Is the
from a single twentieth century skull.
species in decline? Is it seriously threatened? At
It is worth noting that many were island species, this point I often have to say that I don’t know,
that many were lost relatively recently, and that, because it is difficult to estimate population sizes
understandably, information on them is sparse. and trends. As a scientist it is important that I try
The list is still growing—two more extinct pteropo- to give objective answers, since I frequently stress
did species have just been described from Samoa the need for an objective, evidence-based approach
(Helgen et al. 2009): both were identified from speci- to the study and conservation of bats. Estimating
mens collected in the mid-nineteenth century and population size with any accuracy requires good
neither has been recorded since, although anecdotal field data, which usually takes considerable effort.
evidence suggests one of them may have been seen The best estimates come from intensive studies
as recently as the 1980s. over a limited area and may not be representative
across a species’ range. In practice, we tend to count
Identifying and predicting conservation the number of bats in nursery colonies or hiberna-
cula, or the activity levels of flying bats using bat
needs
detectors, as indicators of relative population size
Perhaps the first step in conservation is identifying (for example the UK National Bat Monitoring
that there is, or will be, a problem. The decline or Programme, BCT 2009). The methods are often in-
disappearance of a species from a locality clearly herently biased or flawed, but may be the best avail-
indicates that there is cause for concern on a local able. Nursery colonies often fragment and move
scale and suggests we should look to see if the within and between seasons, making interpretation
decline is more widespread. We must then identify of any observed change difficult. Changes in num-
the cause of this decline and take appropriate action. bers in hibernacula are also difficult to interpret if
Even without evidence of decline, if we understand not all sites are known and monitored, particularly
a species’ particular ecological needs, such as a given that in many sites only a small proportion of
dependence on a scarce and vulnerable type of the bats present may be visible at any one time. For
roost or food, we can anticipate problems and try single sites, or more discrete populations, we may
242 BATS

use mark–recapture methods alongside other meth- effort, since bats can be difficult to sample (for exam-
ods (for example Sendor and Simon 2003, Rivers ple Moreno and Halffter 2000) and the methods
et al. 2006, Papadatou et al. 2009). These methods used each have their biases and limitations (for
are becoming increasingly refined (for example example Flaquer et al. 2007, MacSwiney et al. 2008).
Papadatou et al. 2009, Schorcht et al. 2009), but
still require considerable effort. Novel approaches
Predicting extinction risk
to monitoring single colonies or roosts are being
developed, such as the genetic analysis of faecal Of course, if we understand a species’ biology or
samples to count the number of bats occupying a study patterns in life history cycles of entire taxa,
roost over a specific interval (Peuchmaille and Petit we can make predictions from past trends and antic-
2007), or the use of thermal imaging cameras to ipate future problems. Over recent years a number
count emerging bats (Betke et al. 2008). Revealing of papers have been published that have examined
trends requires that this effort is repeated over many global issues in conservation and extinction risk
years, particularly for long-lived animals like bats, in mammals and other vertebrates (for example
but again, the results may not be representative of Grenyer et al. 2006, Davies et al. 2008, Schipper
what is happening across a species’ range. Intensive, et al. 2008, Cardillo et al. 2008). Their broad aims
long-term monitoring of this nature is generally have been to examine patterns of distribution (of
low on the list of research priorities of those organi- common and rare species), phylogenetic history,
zations capable of funding and/or doing the work. and life history traits in different taxa to draw
In the absence of good population trend data, what broad conclusions about extinction risk and global
do we look for to tell us bats are in trouble? The conservation policy and practice. One important
simple answer is that we look for major local popu- conclusion was the need for studies at narrow taxo-
lation decline, local extinction, and therefore range nomic levels if models were to be precise enough to
contraction: the first steps to more widespread be of practical value. Similar approaches have been
extinction. In other words, a species may have to used to study bats specifically (for example Jones et
be in very real trouble before we are even aware that al. 2003, Safi and Kerth 2004), to see if there are non-
there is a significant problem. random patterns of threat and extinction risk that
Although, as we have seen, there is often good can be explained by life history traits and used to
reason to focus on particular species that are faced build predictive models. Jones et al. (2003) looked at
with unique or at least clearly definable threats, in all bats and found that 48 per cent of the variation in
many cases it is more appropriate to look at an the IUCN assessment of threat could be explained
assemblage of bats or a particular habitat. This by just two factors: restricted geographic range and
may be as part of a more general biodiversity assess- low aspect ratio wings (low efficiency, high man-
ment or an effort to monitor the consequences of oeuvrability) increased extinction risk indepen-
habitat change: either the possibly damaging conse- dently. Low aspect ratio wings are typically found
quences of ‘development’ or attempts to manage in species that forage in cluttered woodland habi-
habitat for the benefit of wildlife. In this case we tats, have small home ranges, small colonies, and
may be more interested in measures of diversity tend to be specialist foragers: all factors consistent
rather than precise estimates of population size: for with increased extinction risk. In Old World fruit
example species richness, diversity as measured by bats, which were found to be at greater risk than
a variety of indices (for example Simpson’s, Shan- other bats, low reproductive rate was more impor-
non–Wiener, Margalef’s) or the use of indicator spe- tant than wing shape in predicting high risk.
cies or taxa. We will also be interested in the scale Another important finding was that closely related
over which this information is gathered: alpha species were more likely to have similar levels of
(within a habitat), beta (across an ecotone or envi- risk than expected by chance. This study, in com-
ronmental gradient), or gamma (regional) diversity. mon with many others, used the same index of risk
We must also be careful to gather accurate data assessment, the IUCN scale of extinction risk: least
through the use of appropriate techniques and concern (0), near-threatened (1), vulnerable (2),
 CONSERVATION 243

endangered (3), critically endangered (4), and this means food and shelter and the ability to repro-
extinct (5). It is worth noting that some of the factors duce. This species has evolved to function in a par-
explored in these studies, such as geographic range ticular habitat and this habitat can be measured in
or small population size, are also used in assigning two fundamental ways: size and quality—how large
species to the IUCN scale in the absence of reliable is it and how close is it, in biological terms, to its
evidence for population decline, so there is a risk of ‘original’ state? By original I mean prior to rapid
circularity in the method. Jones et al. (2003) acknowl- modification by humans. Habitats evolve, since
edge that 95 of the 239 threatened species in their they are made up of species evolving in a changing
study were assigned IUCN threat levels based abiotic/biotic environment, so they are dynamic,
entirely on these two criteria. They therefore not static. However, with infrequent but often spec-
repeated their analysis using only species classified tacular natural exceptions, change—before humans
on the basis of population declines and obtained came onto the stage—was generally at a pace that
similar results. As discussed above, reliable data allowed slow and subtle changes in ecosystems. We
on current and past population sizes are rarely as changed all that. The most visible and easily under-
good as we would like and there are continuous stood way in which we made life difficult for other
efforts to improve the IUCN criteria themselves species was by reducing the size and quality of their
and the data behind them (see for example Rodri- habitat. Cut down forest trees and you deprive bats
gues et al. 2006). In a study of 35 temperate insectiv- of food and shelter. Reduce the area of a forest and
orous bats, Safi and Kerth (2004) also concluded that what’s left will almost invariably support a smaller
bats with low aspect ratio wings were more likely to bat population. And the relationship can be unpre-
have a high extinction risk. Interestingly, dietary dictable and harsh when the losses are great—you
breadth was not a predictor of extinction risk, sug- may lose most of the bats, species and individuals,
gesting that the critical correlations may be those when the forest shrinks below a critical size. This
between wing shape and habitat specialization, relationship is much discussed amongst ecologists,
range size, and migratory ability, but not that as we saw in the last chapter, not least because it is of
between wing shape and diet. For another way of fundamental concern to conservationists and deter-
looking at this issue read Meyer et al. (2008)—I’d like mines the effectiveness of nature reserves locked in
to move on and discuss the ways in which humans ‘alien’ landscapes. It’s complex, of course, because
are putting bats under so much pressure. the many organisms within the forest depend upon
each other to varying extents and in varying ways,
each responding to a reduction in habitat area in a
Threats to bats different way.
In common with all threatened animals and plants, We don’t have to cut the forest down to cause
the pressures faced by bats are numerous, and problems, we can simply fragment it. If we build
underlying them all is our reckless exploitation of a motorway through the middle of a forest then a
the Earth. We can identify several overlapping and species that can’t cross the road is now living in
interlocking mechanisms of loss that apply not just a smaller forest. Divide the forest with lots more
to bats, but to biodiversity in general. I will discuss roads and many of the smaller fragments may be
them in broad, conceptual terms and provide refer- unable to support viable populations of some spe-
ences for specific examples. The subject is huge and cies. Turn the land adjacent to the roads into towns
I will leave you to study some of these examples in and farms and you reduce forest area and increase
detail by reading the original papers. However, the isolation of the forest fragments. As populations
some will be discussed a little as they come up. in the forest undergo ‘natural’ fluctuations in size,
those in the smaller, more isolated fragments are
more likely to die out and we see the first signs of
Habitat loss, fragmentation, and degradation
a species in decline: falling numbers and range con-
To survive, a population of a species requires suffi- traction. Throughout this process, the patches of
cient habitat to meet its needs. At the simplest level forest that remain are probably also changing, and
244 BATS

perhaps not for the better—habitat loss and frag- we unwittingly provided, for example in build-
mentation are usually accompanied by degradation ings, bridges, tunnels, and mines. In temperate
of what is left. Some change is an inevitable conse- countries large numbers may gather in mines to
quence of area reduction and fragmentation, since hibernate (Fig. 9.2), and buildings often house the
both generally increase the amount of ‘edge’ habitat, most important nursery roosts. However, destruc-
favouring some species but reducing habitat quality tion of these artificial roosts has now become a major
for others. Edges can bring other complications, problem, through the disturbance, redevelopment,
such as the influx of generalist competitors and pre- or destruction of old mines, and the demolition or
dators and greater exposure to extreme weather renovation of buildings. Although bats may cause
events. Further degradation may result from man- minor problems in occupied buildings these can gen-
agement/exploitation of the forest and from the erally be overcome, but urban humans have often
pollution and introduced alien species that fre- lost the willingness to live in close proximity to other
quently accompany human occupation of a land- animals.
scape. Although I have used forests as an obvious Many studies have investigated the effects of for-
example, these changes will naturally apply to other est clearance, fragmentation, and degradation on
habitats. bats in both temperate and tropical regions. Major
Bats can be very adaptable, and many species fragmentation and degradation are undoubtedly
have learned to exploit the human environment, bad for the bats, but the results for more modest
which can sometimes provide better roosting and fragmentation are mixed, with some studies
even foraging than degraded rural landscapes. showing little or no effect or even some ‘benefits’
However, the built environment is often the for bats, depending upon the indicators used. Too
most rapidly changing of all, presenting new much has been done for me to attempt to cover this
threats as well as opportunities. As we destroyed topic comprehensively—I will use a few selected
bats’ natural roosts, some species adapted to those papers to illustrate some of the key points.

Figure 9.2 Brandt’s bats, Myotis brandtii, hibernating in an abandoned nineteenth-century lead mine in northern England. Photo by Anita Glover.
 CONSERVATION 245

Many studies have focused on the neotropics, flooding and continued for four years. Fragmenta-
where bats are abundant and diverse, but where tion led to a rapid fall in bat abundance, species
forest clearance is occurring at an alarming rate. richness, and ensemble structure. The effects were
One of the earliest studies (Fenton et al. 1992), in seen more rapidly on small islands, but were com-
south-east Mexico, reported significantly higher parable on all islands after year three, suggesting
diversity (Shannon–Weiner), but not species rich- that larger fragments (in this case still only 40 ha)
ness, in undisturbed relative to disturbed forest. did not protect the bats (see also Estrada et al. 1993).
This difference was due in large measure to the Cosson et al. (1999) make the important point that
significantly greater presence of members of the they do not believe that their small forest islands are
insectivorous (and one carnivorous) Phyllostominae supporting separate populations. They believe the
in undisturbed forest. Fenton et al. (1992) sampled bats are visiting them from nearby larger fragments
bats using mist nets and ultrasound detectors, but or continuous forest to feed, but when fragments are
the latter were found to be ineffective: phyllosto- too small they are no longer worth visiting due to
mids have very low intensity echolocation calls habitat changes (for example through edge effects)
that are hard to detect and it was also impossible or because the bats are reluctant to cross open
to identify most species from their calls. Most spaces. It is worth noting that species that exploit
subsequent studies have used only mist nets. In a the understorey were more sensitive to forest frag-
more detailed study, also in Mexico, Medellin et al. mentation than canopy feeders, as were smaller spe-
(2000) confirmed that the presence of Phyllostomi- cies (which typically have smaller home ranges)
nae species is a good indicator of low levels of relative to larger bats.
disturbance. They also found that phyllostomid Over a 10 year period, and further downstream of
richness, diversity, and the number of rare species the same flooded river used by Cosson et al. (1999),
increased as disturbance level declined along a gra- Henry et al. (2007) studied a single, abundant under-
dient extending through cornfields, young and old storey frugivore, Rhinophylla pumilio, following the
abandoned fields, cacao plantation, and undis- foraging movements of bats (using radiotelemetry)
turbed forest (Fig. 9.3). whilst estimating food availability across islands
Cosson et al. (1999) followed the response of bats and mainland. A (rather complex!) measure of bat-
to the creation of islands by the damming of a river specific landscape connectivity was developed,
in Guyane (French Guiana). The study of nectar- based on average forest cover in neighbouring land-
and fruit-eating species started one year before scape units, weighted by the probability that bats

35

30 Total

25 Forest
bat species

Cacao
20
Oldfield
15 Cornfield

10

0 Figure 9.3 Species accumulation curves in undisturbed

0 1 2 3 4 5 6 7 8 9 10
and disturbed habitats in Chiapas, Mexico. From Medellin
month et al. (2000), with permission from John Wiley and Sons.
246 BATS

Gatún Lake

2 km
Barro Colorado
Island

Figure 9.4 Gatún Lake and Barro Colorado Island (simplified!), Panama showing the eleven islands and six mainland sites chosen for study
by Meyer and Kalko (2008), with permission from John Wiley and Sons.

would use the units, estimated from the frequency you to explore. Mist netting over two years was
distribution of flight distances. The abundance of used to sample bats systematically. In common
Rhinophylla pumilio increased strongly with connec- with results from most neotropical studies, virtually
tivity, but was not related to local food availability. all the bats caught were phyllostomids (98 per cent).
Despite its apparent mobility, this understorey spe- Bat assemblages on islands were shown to be less
cies was deterred by open water. The choice of an diverse and structurally simpler than those on main-
understorey species for this study was based on the land sites, due to the loss of gleaning (insect- and
knowledge gained from their earlier study (see pre- vertebrate-eating) species, omnivores, and vampires
vious paragraph), that it would be more sensitive to (Fig. 9.5). Among the frugivores, canopy specialists
fragmentation than canopy species. were significantly less affected by fragmentation
Another study to take advantage of artificial is- and isolation than understorey species.
lands was that by Meyer and Kalko (2008). In this Species richness decreased with distance to main-
case the islands were in Gatún Lake in Barro Color- land and increased with forest cover, but relative
ado Nature Monument in Panama, formed in 1914 importance depended on scale, with distance being
by the damming of the Chagres River during the important at small scales, area at larger scales. This
construction of the Panama Canal. The lake has over suggests that many species use more than one
200 islands varying in size from less than 1 ha to island. On the mainland itself, forest edge sites
1,560 ha. Meyer and Kalko (2008) chose 11 islands were less rich than forest interiors, although species
and 6 mainland sites for their study (Fig. 9.4). composition did not differ significantly. One of the
The islands fell into two groups relative to the take-home messages was that, after isolation, habi-
mainland: near (less than 500 m) and far (more tat loss was the next most important factor deter-
than 1.5 km) and there were other important ele- mining the response of the bats: in minimizing
ments to the experimental design that I will leave to fragmentation, conservation measures should also
 CONSERVATION 247

1200
1000 islands
far
% change in relative abundance

200

islands
150
near

100

50 mainland
edge
0
s s
ore ore
giv riv
–50 fru cta
ne
g s Figure 9.5 Differences in the abundance of
–100 in ore re
s es
an aliv vo r bats on islands in Gatún Lake relative to
l e
g im ni u ivo
an om mainland sites. Adapted from Meyer and
ng
sa Kalko (2008).

attempt to minimize habitat loss. The importance of species evenness was highest in largely undisturbed
habitat area was also highlighted by Struebig et al. forest. The one member of the Phyllostominae re-
(2008), who found it to be more important than ported, the insectivorous/carnivorous Chrotopterus
isolation in a study of Malaysian forest. They sug- auritus, was abundant only in undisturbed forest, add-
gest that fragments greater than 300 ha contribute ing further weight to the view that these bats are good
significantly to landscape-scale bat diversity, but indicators of low levels of disturbance (Fenton et al.
that even small fragments will have value. Meyer 1992, Medellin et al. 2000).
and Kalko’s (2008) work also highlights the impor- The important role of nectar-eating bats as polli-
tance of the ‘fragment-matrix’ contrast, that is the nators was discussed in the previous chapter. Ques-
magnitude of the habitat differences between forest ada et al. (2003) reported that two of the three
fragment and the surrounding area. When the effective pollinators of the tree Ceiba grandiflora, an
matrix is water, as in this study, the bats’ response important component of the dry forest along the
to fragmentation and isolation is greater than when Pacific coast of Mexico, visited the tree far less
the matrix is less ‘hostile’ to the bats. The matrix often in disturbed relative to undisturbed forest.
generally offers less food, fewer roosts, less protec- The third species was not significantly affected.
tion from weather and predators, and is energeti- This led to lower levels of pollen deposition on
cally costly to cross. Water offers nothing! stigmas and reduced fruit set (Fig. 9.6).
The rapid loss of frugivorous phyllostomids As we would predict from studies described ear-
immediately after flooding in the study of Cosson lier, it was the two smaller species, Musonycteris
et al. (1999) contrasts with the dominance of frugi- harrisoni and Glossophaga soricina (8–13 g) that were
vores on islands found by Meyer and Kalko (2008), sensitive to disturbance, not the larger Leptonycteris
90 years after the formation of Gatún Lake. This yerbabuenae (previously L. curasoae) (20–30 g).
suggests that some species may have some capacity Not all of the work in this area has been done in
to recover from major habitat change. the neotropics. Law et al. (1999), working in New
Gorresen and Willig (2004) studied a bat assemblage South Wales, Australia, used bat detectors to iden-
composed primarily of fruit- and nectar-eating phyl- tify species and assess the activity of insectivorous
lostomids in Paraguay. Species richness was highest in bats across habitats. They found that forest area,
areas that had undergone modest fragmentation, but habitat diversity, and structural complexity were
248 BATS

(a)
14

12
visits (no. flowers/night)

10

8
undisturbed
6

4 disturbed

0
Leptonycteris Glossophaga Musonycteris
curasoae soricina harrisoni

(b) (c)
6
fruit set (no. fruits/flower)
pollen grains (no./stigma)

5
400
4
300
3
200 2
Figure 9.6 Flower visitation rates (a) by three species of
nectar bat in undisturbed and disturbed forest in Mexico and
100 1 the consequences to pollination (b) and fruit set (c). From
0 Quesada et al. (2003), with permission from Springer Science
undisturbed disturbed undisturbed disturbed +Business Media.

the best predictors of activity for a large proportion area of 20,000 km2 using automated bat detectors.
of the bat ensemble occupying the fragmented Not surprisingly, species diversity was positively
landscape. The sensitive bats were the slow-flying, related to forest cover and negatively related to
manoeuvrable species best adapted to the forest urban development, although some species appear
habitat. Fast, open-air foragers were more tolerant able to exploit urban landscapes (for example Gehrt
of fragmentation and insensitive to isolation effects and Chelsvig 2003, 2004). These results broadly par-
on the scale studied. Although flight activity in frag- allel several previous studies and show that land-
ments and corridors (narrow strips of forest scapes modified by humans rarely benefit bats (for
between fragments) was not significantly different example Walsh and Harris 1996, Jaberg and Guisan
from that in continuous forest, feeding activity was 2001, Russ and Montgomery 2002). Many of these
significantly higher in the latter. On an individual studies support the analyses of Jones et al. (2003) and
species level, activity in corridors was no higher Safi and Kerth (2004) in showing that bats adapted
than in the adjacent open landscape, suggesting for life in cluttered environments, with low aspect
that the bats were not using the corridors for com- ratio wings and slow flight, are largely excluded
muting. Corridors are widely seen as a possible from urban and intensely agricultural environ-
solution to forest fragmentation but this is not the ments. Forest/woodland cover is critical to many
first study to suggest that we should look more of the rarer bat species and this woodland should
closely at this issue (see for example Davies and be mature and structurally diverse.
Pullin 2007). Duchamp and Swihart (2008) con- Ecologists have long recognized the pervading
ducted a large scale study across an agricultural importance of scale in many of the processes they
landscape in Indiana in the temperate United States: study (see Chapter 8) but until recently most stud-
27 4 km X 4 km squares were sampled over a total ies, often for very practical reasons, have been
 CONSERVATION 249

conducted on a single scale. However, new analy- Underground roosts

tical methods and the use of Geographical Infor-
In warmer climates caves (and old mines) hold the
mation Systems have led to the incorporation of
largest colonies of bats on the planet, and these sites
scale-dependence in some recent studies and scale-
may be occupied all year round. A single cave may
dependence is likely to be an important component
be home to many species, with diverse social struc-
of future work. Gorresen et al. (2005) studied 14 sites
ture and feeding ecology (for example Arita 1996,
10 km apart in subtropical, semi-deciduous forest in
Furman and Ozgul 2002, 2004, Niu et al. 2007). In the
Paraguay. Each site comprised nested concentric cir-
temperate world the use of underground sites is
cles of 1, 3, and 5 km radius. Bats were sampled by
highly seasonal but just as important: hibernacula
intensive mist netting and abundance, richness, and
frequently hold tens or hundreds of thousands of
evenness were correlated with landscape variables
bats, and occasionally over one million. Cave roosts
(forest cover, patch size, patch density, edge density,
in some parts of the world are under threat from
nearest neighbour distance, patch proximity, and
quarrying. Abandoned mines have the additional
patch shape) at these three scales. Significant rela-
threats of total closure for reasons of public safety,
tionships were much more likely to be found at
collapse as old timber props decay, or of being re-
larger scales. At the 5 km scale 12 significant re-
opened as new and more efficient extraction pro-
sponses were found involving all 8 species exam-
cesses make them financially viable. The Australian
ined, only 5 responses were found at the 3 km scale
ghost bat, Macroderma gigas, is very reliant on old
and only 2 at the 1 km scale. At both 5 km and 3 km
mines, with one site alone holding 1,500 of the esti-
scales regression identified three significant relation-
mated total population of 10,000 (Phillips 1990).
ships, associated with three landscape variables,
Several of its major roost sites have been threatened
between habitat and species richness and evenness.
with large-scale quarrying and others are in danger
At the 1 km scale a single significant relationship was
of collapse (McKenzie and Hall 2008). The cancella-
associated with only one landscape variable. The
tion of limestone quarrying licences gave a reprieve
details of the many relationships uncovered are too
to several major bat roosts in southern India (Mur-
complex to go into here, but it is worth giving a few
phy 1987) and some large hibernacula have been
examples. Patch size could account for 44 and 42 per
saved in the USA by preventing the sealing of
cent of the variability in species richness at the 5 km
mines (Tuttle and Taylor 1998). Sadly many sites
and 3 km scales, but had no significant effect at the 1
have been lost with little or no publicity or attempts
km scale. Patch density and patch size together ac-
to save them. The nests of cave swiflets are highly
counted for 62 per cent of the variation in species
prized by the Chinese and in 2001 had a street value
richness at the 5 km scale, but at the 1 km scale no
of $4,000 per kilogram (Sankaran 2001). Nest collect-
landscape variables were able to explain species rich-
ing not only threatens the swiftlets, but also the bats
ness. Four of the eight species and species richness
that roost in the same caves due to excessive distur-
(but not evenness) demonstrated significant scale-
bance. The large deposits of guano that build up
dependence, with bats responding positively to
under insectivorous bat colonies have been exp-
some variables and negatively to others.
loited for centuries, and in many tropical countries
There are important messages to take from this:
it is still an economically important agricultural fer-
we must study bat:habitat relationships at the
tilizer. Kingston (2010), who lists the conservation of
appropriate scale and we must manage habitat
cave-dependent bats as one of the four primary
for bats at the appropriate scale. One difficulty is
conservation priorities for South-east Asia, notes
establishing this scale. Many of the bats studied by
that over 250 tonnes of guano are extracted from
Gorresen et al. (2005) may roost and forage in the
caves each year in Myanmar and collection from a
same area all year round, yet a complex situation
single cave in Thailand generates an annual income
was revealed. Where bats undertake significant sea-
of $135,000 for local villagers. However, guano col-
sonal movements it could get more complex and the
lection can also disturb cave bats unless carefully
scale very large.
250 BATS

regulated. More general disturbance is a factor in

many regions (for example Petit et al. 2006).
Cavers can be very disruptive to bats, and an
increasing awareness of this has led to improved
procedures in some countries, with the development
of voluntary codes of practice for cavers and legal
protection for some of the caves (for example Hut-
son et al. 1995). Survey work is often done with the
cooperation and assistance of cavers (for example
Glover and Altringham 2008) and the majority of
cavers and caving organizations are keen conserva-
tionists. Bats are now an integral part of many cave
conservation documents and caving organizations
regularly participate in research and conservation
expeditions. Cavers have played a significant part
in trying to control the worrying emergence of
White Nose Syndrome (WNS) in the north-eastern
United States (Turner and Reeder 2009, Frick et al.
2010b). The National Speleological Society has es-
tablished a WNS fund and is taking practical steps
to try and limit the possible spread of the disease by
cavers through new codes of practice, cave closure,
and education. WNS is discussed in more detail
below.
The growth of tourism and eco-tourism has also Figure 9.7 Carlsbad Cavern, home to a large colony of Brazilian free-
had its effects. Many caves opened up for tourism tailed bats, where photography by tourists is now controlled.
have had their entrances and even interiors modi-
fied, sometimes destructively; walkways, railing,
and lighting are often installed and the caves are persecuted. Even when the bats, not the cave, are the
exposed to a constant stream of people. All clearly focus of attention, poor management of the crowds
constitute major disturbance if bats are present. A can lead to significant population declines. The
number of major hibernacula have been lost or dam- use of cameras has been banned at the evening
aged by the development of show caves, and some emergence of Brazilian free-tailed bats at Carlsbad
are still under threat from tourism. Furman and Cavern (Fig. 9.7) in New Mexico due to fears that
Ozgul (2002, 2004) have drawn attention to the bio- intensive use of flashguns was disturbing the bats.
logically rich caves of Turkey which are used as both
hibernacula and summer roosts. None of the caves
Pollution: chemical, noise, light
have adequate protection and some are under seri-
ous threat: two have recently had tourist walkways There are many forms of pollution, some of which
built directly under hibernation and nursing sites. can affect bats directly, while others contribute to
This threat is widespread, for example China has habitat degradation in more subtle ways. As DDT,
extensive karst landscapes with numerous caves prior to its belated widespread ban in the 1970s and
occupied by bats (for example Niu et al. 2007) and 1980s, inflicted well-documented devastation on
with the increase in both domestic and foreign tour- more iconic animals like the bald eagle and the
ism many of these caves are being opened up to peregrine falcon, it was undoubtedly having similar
sightseers. It is possible in some cases for bats to and largely unrecorded effects on bats. In the United
live alongside tourism, but it is frequently the case States DDT was actually used to control bat ‘infes-
that bats are neglected in development plans or even tations’ in homes! DDT was just the potent and
 CONSERVATION 251

well-publicized tip of the pollution iceberg. Many that effluent levels do alter the habitat for bats, but
damaging chemicals have been banned and others not necessarily in predictable ways. However, over-
are now used more sparingly and more carefully, all bat activity and feeding activity were not signifi-
but we continue to pour chemicals into the environ- cantly different. Much therefore depends upon what
ment, particularly in the developing world (for you look for. Park and Cristinacce (2006) compared
example Fenton and Rautenbach 1998, McGinn insect and bat activity over percolated filter beds, a
2001). The most direct chemical pollutants kill traditional method still used for small communities
or weaken bats, or impair reproduction, as they in the UK, and the activated sludge plants that have
accumulate in their bodies. Pesticides kill the inver- replaced them at high throughput works. Filter beds
tebrates that bats feed on and herbicides kill the supported large insect populations and large num-
plants that these invertebrates rely on for their exis- bers of foraging pipistrelles (Pipistrellus pipistrellus
tence. Unless the use of pesticides and herbicides and P. pygmaeus), at levels comparable to nearby
is carefully targeted and controlled there is an inevi- ‘natural’ habitats, whereas insect and bat activity
table loss of biodiversity and disruption to ecosys- was low at activated sludge plants. There is no
tems. And you don’t have to try to kill things. simple take-home message when it comes to effluent
A significant cause of biodiversity loss is enhanced treatment, but clearly, as far as bats at least are
nutrient levels resulting from the use of fertilizers in concerned, it is not necessarily bad.
intensive agriculture. Fertilizers are washed beyond There are other forms of pollution that are per-
the confines of the fields to which they are applied, haps less immediately obvious, but may be no less
increasing nutrient levels in the surrounding coun- important. Humans rarely confine their activities to
tryside via streams and rivers. the hours of daylight, but flood the night with
Sewage or wastewater treatment works can have light—our villages, towns and cities, transport sys-
a major influence on aquatic environments and a tems, industrial areas, sports stadiums, even some of
number of studies have looked at insect and bat our agricultural land, are lit to allow us to work and
communities in relation to sewage plants. The re- play into the night. Despite the financial and envi-
sults are very much dependent on the type of works ronmental costs (in terms of energy wastage alone)
being studied, but data are rarely given on effluent we also feel the need to light up public buildings
nature and quantity. The results from early studies and private homes purely for visual effect and
are sometimes conflicting, showing significant (for aggrandizement. Artificial light has significant ef-
example Vaughan et al. 1996) or non-significant fects on many animals (for example Longcore and
(Racey et al. 1998) effects of river eutrophication. Rich 2004, Rich and Longcore 2006) and bats are
When an effect was observed, the response was clearly going to be among the worst affected. Light
species-dependent, probably due to differences in avoidance has the potential to exclude bats from
prey preference: some may benefit from an increase otherwise good roosting and foraging habitat and
in the abundance of pollution tolerant insect species, it probably prevents them from travelling freely
others may depend on insects intolerant of pollu- about their home range. Stone et al. (2009) have
tion. Kalcounis-Rueppell et al. (2007) looked at insect shown that illuminating hedgerows with ‘street
and bat communities up- and downstream of the lights’ leads to a major reduction in the activity of
discharge pipes (‘point-source nitrogen effluent’) the lesser horseshoe bats (Rhinolophus hipposideros)
into streams in North Carolina, USA. Nocturnal that habitually commute along them. The conse-
insect abundance was higher upstream than down- quences of light avoidance in this case are not
stream of discharges in both study streams. Diptera, clear, but are likely to include longer commuting
Coleoptera, and Lepidoptera were more abundant journeys that waste time and energy, and perhaps
upstream, Odonata were more abundant down- loss of suitable foraging habitat and therefore a
stream. The generalist big brown bat, Eptesicus fus- reduction in foraging success. One reason for avoid-
cus, was more abundant upstream, and Perimyotis ing lights is an increased risk of predation, so indi-
subflavus (the eastern ‘pipistrelle’), an aquatic spe- viduals that are unable to avoid them may suffer
cialist, was more abundant downstream, suggesting increased predation. Artificial light also has the
252 BATS

potential to interfere with the navigation of migrat- mouse-eared bats, Myotis myotis, experienced a 25-
ing bats, since there is evidence that vision is used in fold reduction in foraging efficiency. A significant
navigation (see Chapters 3 and 7). Some bat species effect was found as far away as 50 m, when search
are known to feed on the insects attracted by artifi- time was 150 per cent of controls. The effects of these
cial light that emits wavelengths in the ultraviolet single vehicles would therefore extend to 60 m and,
(Rydell 1992, Blake et al. 1994). However, these spe- as traffic densities increased, well beyond that. Less
cies appear to be in the minority and most probably direct evidence suggests that aerial hawking species
avoid lights. The concentration of insects around are also likely to be affected by traffic and other
lights may well benefit a few of the fast flying, aerial artificial noise (for example Mackey and Barclay
hawking species, but at a significant cost to the 1989, Szewczak and Arnett 2006).
others, in that lights may reduce the abundance of
these insect species in areas away from lights. For Alien species: predators, competitors,
example, Arlettaz et al. (2000) suggest that increased
and diseases
competition due to the population expansion of the
more light tolerant pipistrelle, Pipistrellus pipistrellus, Humans have become increasingly mobile, travelling
may have contributed to the decline of the lesser with increasing frequency, over greater distances,
horseshoe bat, Rhinolophus hipposideros. Avila-Flores more quickly, and in greater numbers. This increased
and Fenton (2005) found that bat activity in Mexico mobility, like many other aspects of human activity,
City was higher than in nearby forest, but that this has been particularly marked in the last century. As
was due to the high activity of a small number of we travel we take many organisms with us: diseases
species exploiting insects in lit areas. Natural forest in our bodies, pests on them, pets that escape, agri-
had lower activity, but more species. cultural animals, agricultural plants, and many ani-
We surround ourselves not just with light but mals and plants that just sneak a ride. In the past,
with noise, and this too is known to have wide- settlers in new lands deliberately introduced animals
spread consequences for many animals, such as and plants to create a home from home. The traffic is
birds and cetaceans, impairing communication, nav- of course two-way: when travellers return home they
igation, and foraging activity (Barber et al. 2010). bring back other diseases and pests, exotic animals
Only recently have we begun to consider its effects that escape, plants for gardens that also escape. Many
on bats. Gillam and McCracken (2007) have shown natural habitats have been replaced by non-native
that Brazilian free-tailed bats (Tadarida brasiliensis) crop plants. We introduce disease or predators to
will increase the frequency of their echolocation calls control our pests, sometimes pests that are them-
in the presence of high background intensity insect selves recent introductions. For thousands of years
calls. They found a strong positive correlation we have introduced organisms into environments
between the frequency of broadcast insect calls and they could never have reached on their own. The
that of echolocation calls. Schaub et al. (2008) con- vast majority of these organisms are benign. They
cluded from lab experiments that traffic noise up to often fail to thrive in their new environment and
50 m from a road is likely to deter species that forage many are sink populations, only capable of surviving
by listening passively for prey, although it is worth by sustained reintroduction. A small proportion of
noting that the noise generated by the movement of introduced species will thrive and establish large,
vegetation could be equally or even more disturb- widespread, and healthy populations, but have little
ing. Artificially generated broadband noise, at simi- or no discernable impact on the environment. A very
lar intensities, was the most effective deterrent of small proportion not only become established and
foraging bats, probably due to the higher frequen- thrive but have a profound effect on their new
cies present. In a more recent study, Siemers and home, driving other species to extinction through
Schaub (2010) played back noise under controlled predation, competition, or by more complex and
lab conditions equivalent to the average generated indirect mechanisms (Blackburn et al. 2004, Clavero
by single cars and trucks passing at 80 km h1. At a and Garcia-Berthou 2005). Problems are particularly
distance of 7.5 m from the ‘road’ edge greater acute on islands, where natural populations are often
 CONSERVATION 253

small and species have frequently evolved in a extraordinarily high: the UK had over 9 million pet
depauperate ecosystem, lacking specific predators, cats in 2003 (data from the UK pet food industry
diseases, or competitors. A classic example is that of cited in Beckerman et al. 2007) and a less certain
the introduction of the Asian and Australasian brown estimate of two million feral cats. Numbers are
tree snake, Boiga irregularis, onto the Pacific islands rising at an average of 13 per cent per year. The
(Fritts and Rodda 1998). The snake was introduced domestic cat population in the USA has been esti-
onto Guam around 1950 and by 1990 bats, birds, and mated to be over 80 million, the feral population
reptiles had been largely eliminated from much of the 30–70 million. Worldwide there may be over 500
island by this introduced predator—most forested million, but estimates are often little more than
areas retained little more than three small lizard spe- guesses. They are widely acknowledged to be a
cies. On Guam and other islands in Micronesia and serious conservation concern in urban and ‘wild’
the Northern Marianas the snake has had a devastat- landscapes and on both continents and islands.
ing effect on the Marianas fruit bat, Pteropus marian- They underlie the extinction of many island popula-
nus, by eating young before they are able to fly. The tions. In a recent review, Nogales et al. (2004)
population on Guam is expected to become extinct describe how cats have been removed from at least
within 10 years (Allison et al. 2008). Introduced pre- 48 islands across the world in an attempt to reduce
dators are not its only problem: habitat loss, hunting their impact on rare wildlife. Most of these islands
for food (discussed in the next section), volcanic activ- are small (less than 5 km2) since the larger the island,
ity, and typhoons are all major factors. Some of these the more difficult eradication is, but the benefits are
threats compound each other: for example, after a now becoming evident and eradication pro-
typhoon, hunting activity often increases (Esselstyn grammes more ambitious. There is good evidence
et al. 2006). that this predation extends to bats (for example
Introduced predators are an important factor in Woods et al. 2003) and that the effects can be signifi-
the decline of the endemic New Zealand long-tailed cant, particularly on rarer species. I used the data
bat, Chalinolobus tuberculatus (Pryde et al. 2005). In from Woods et al. (who did not single bats out for
warm years, with plenty of food from the Nothofagus particular attention) to make some conservative cal-
beech forests, populations of introduced mice and culations on total kills (Altringham 2003). These
rats were high, supporting large populations of their data indicated that cats were killing at least 250,000
introduced predator, the stoat, Mustela erminea. The bats per annum: more than the total UK population
ship rat, Rattus rattus, and the stoat in turn prey on of all but our two most common species, and a very
bats in the roost. Extinction of the bats in the study significant proportion of these too. The effects of cats
valley was predicted to occur within 50 years, and it and other introduced predators on wildlife may go
was suggested that if climate change brings warmer beyond the immediate kill, as illustrated by Becker-
summers the current problem will worsen as rat and man et al. (2007). They show that even when preda-
stoat populations rise. Efforts have begun to control tor mortality is low, changes in behaviour induced
these predators and appear to be having some suc- by predators (such as in foraging patterns or habitat
cess. Rats and stoats are having a similar effect on choice) can lower fecundity and greatly reduce pop-
New Zealand’s other endemic species, the short- ulation size. Beckerman et al. (2007) provide a good
tailed bat, Mystacina tuberculata (O’Donnell 2008a), overview of the debate between the ‘doomed sur-
and the introduced Pacific rat, Rattus exulans, ap- plus’ hypothesis, which proposes that alien preda-
pears to have been responsible for the almost certain tors simply mop up those animals that would die
extinction of the greater short-tailed bat, Mystacina anyway, and that of the ‘hapless survivor’, which
robusta, last seen on Big South Cape Island in 1967 proposes that natural mortality and the effects of
(O’Donnell 2008b). alien predators are additive.
Cats are among the most numerous introduced In some situations top predator eradication can be
predators in many countries, as pets and as sub- counter-productive, by relaxing population control
stantial feral populations. Numbers can be on other introduced, but smaller ‘mesopredators’, as
254 BATS

observed by Rayner et al. (2007). Cat eradication on

Little Barrier Island in New Zealand led to an
increased number of introduced Pacific rats and a
decline in the breeding success of the endangered
Cook’s petrels, until a rat eradication programme
got underway.
I am not aware of any examples of introduced
competitors to bats, and only one possibly intro-
duced disease. I will come back to disease a little
later in another context, but this is a good point to
introduce the one disease of very real conservation
concern: White Nose Syndrome (WNS), which
emerged in the north-eastern United States in 2006
(for example Turner and Reeder 2009). WNS is
apparently caused by a cold-loving fungus, Geo-
myces destructans, and it has been responsible for
the death of over one million bats in the US since
2006. The fungus forms a conspicuous growth on
the face, ears, and wings of six species of cave-
hibernating bats (Eptesicus, Myotis, and Perimyotis)
and infiltrates the underlying tissues (Fig. 9.8).
The disease is not yet well understood. It is typi-
cally found in bats with minimal fat reserves at
a time when these should be extensive enough to
get them through the winter (Blehert et al. 2009). It is
not yet clear if the fungus is the primary cause of Figure 9.8 Little brown bat, Myotis lucifugus, with white-nose
syndrome. Photo by Marvin Moriarty, USFWS.
death or a secondary infection, but after extensive
study, the absence of other vectors suggests it is the
cause. Until very recently WNS was confined to the
north-eastern United States, but it has now been
discovered in five species of cave-hibernating Myotis such resistance? The disease first came to light in a
species in six continental European countries show cave in New York State, and may have got
(France, Germany, Switzerland, Hungary, Czech there through a human carrier from Europe, or some
Republic, and Slovakia, Puechmaille et al. 2010, Wib- other part of the world where it has yet to be identi-
belt et al. 2010, Martinkova et al. 2010), and may fied. Scientists and conservationists are lobbying
prove to be more widespread in Europe (and per- hard for funds to investigate the disease, fearing
haps beyond). In Europe G. destructans infection that it could be a very significant threat to a large
does not appear to be associated with significant number of species. In a recent paper Frick et al.
mortality, but may have been present since 1995 or (2010b) calculate that at current mortality rates the
earlier. One ringed (banded) and infected individual little brown bat, Myotis lucifugus, a widespread and
was recaptured on two subsequent occasions (5 and common species, may be regionally extinct in less
11 months later) and showed no signs of infection. than 20 years. Even under more optimistic model
In North America mortality can exceed 90 per cent, a parameters the future for this species looks bleak.
level that would not go unnoticed in the closely More worryingly, at least five more species may face
monitored European sites. One possibility is that a similar future and the disease is spreading south
the fungus has coevolved with bats in Europe and and west (Fig. 9.9). I will leave you to pursue this
bats are therefore resistant to it. Is it a recent intro- fast-moving topic, since anything I write will be
duction to North America, where the bats have no rapidly out of date.
 CONSERVATION 255

400 km
In the 1980s approximately 13,000 bats were im-
ported each year (Wiles 1992), resulting in signifi-
cant declines in bat populations in the exporting
countries (Rainey 1998). Some of the imported spe-
cies were endemics, or species with limited distribu-
tions, and several are now threatened by this trade.
Import/export control under the CITES Agreement
(Convention on International Trade in Endangered
2006 Species of Wild Fauna and Flora) has been larg-
2007
2008
ely successful, but some illegal trade continues
2009 (Worthington and Taisacan 1996). A recent review
2010
(Mickleburgh et al. 2009) suggests that consumption
of bats by humans continues to be a significant
concern in Asia, Pacific and Indian Ocean islands,
and western Africa, since ‘off-take’ rates are unlikely
to be sustainable. There are examples of successful
regulation of the hunting of island populations, but
these are rare and further action is clearly necessary.
A detailed study by Epstein et al. (2009) of Pteropus
vampyrus illustrates the difficulties. Since 1996 hunt-
ing has more than doubled, with 90,000 bats taken
between 2002 and 2005, at a rate that is almost
certainly unsustainable. This does not include illegal
hunting or ‘pest’-control activity. This species was
Figure 9.9 County by county spread of white-nose syndrome across the also shown to travel hundreds of kilometres
north-eastern USA and Canada. Adapted from Frick et al. (2010b) with between roosts in a given year, and their journeys
permission from AAAS. take them across several international boundaries.
Heavy losses due to hunting in Malaysia therefore
threaten the population over a large region.
Hunting, persecution, disease—and what Although fruit bats are the most common target,
large numbers of insectivorous bats are eaten in
bats can do for humans
some countries. Francis et al. (1999) reported that
Many animals are hunted for food and others killed smaller species were eaten right across Laos and
because they are real or perceived pests or threats. that where consumption was commercial, or in-
Some are killed through little more than ignorant volved rare species, it was probably of conservation
prejudice. The massive increase in the human popu- concern.
lation and the ease with which we can now kill wild There is a persistent and widely held, homocentric
animals has put enormous pressure on bats. Two view of the world, that if an animal has no apparent
aspects of their behaviour make them particularly economic value then it is not worth preserving, and
vulnerable: the formation of large, vulnerable colo- if it is perceived to be a pest then it can be eradicated.
nies and their use of buildings as roosts, which A more enlightened view is becoming more preva-
brings them into very close proximity to humans. lent, as is an appreciation of the importance of intact,
Bats are eaten in several parts of the world, but functional ecosystems to human survival. Sadly,
until recently we had little evidence to indicate how despite widespread acceptance of the concepts of
much of a threat this was. In Guam, fruit bats have ecosystem services and the economic value of natu-
traditionally been eaten in large numbers, and hav- ral habitats, this homocentric view of the world is
ing eaten their way through their own bats, the still widespread and bats, like many other animals,
islanders imported bats from other Pacific islands. continue to be persecuted.
256 BATS

Most of the claims that bats are vermin, as carriers Bats, notably pteropodids, have also been shown
of disease or as destroyers of buildings and crops, are to be hosts to a number of emerging viral diseases
either false or exaggerated: bats are rarely a signifi- that have led to human fatalities in South-east Asia
cant problem to humans. They are involved in the and significant economic costs (Halpin et al. 2007).
transmission of a small number of diseases, most The emergence of these new viruses has been linked
seriously rabies, the rabies-like lissaviruses, and his- to ecological pressures on the bats as their habitat
toplasmosis. However, their impact on humans is has been fragmented and destroyed and the bats
very small and does not justify ill-conceived persecu- themselves exploited. Humans are being exposed
tion (see Daszak 2010 for a recent perspective). to these diseases, at least in part, through agricul-
Histoplasmosis is a respiratory disease caught by tural expansion into virgin forest and trade in wild-
inhaling a species of fungus (Histoplasma capsulatum) life. The hope is that the lesson to be learned is not
found in soil rich in organic matter—for example that bats are pests to be exterminated, but that we
under large roosts of birds or bats. Infections are must make more responsible use of the environ-
uncommon and rarely go beyond flu-like symp- ment. A recent study by Streicker et al. (2010), who
toms, but in a small percentage of cases the fungus analysed rabies virus sequences from 23 North
is disseminated and the resulting inflammation can American bat species suggest that even within
be fatal if untreated. The fungus is found in the bats, the probability of cross-species transmission
southern United States, South America, and Africa, and host-shifting diminishes rapidly with increasing
with isolated reports from some other parts of the phylogenetic distance. Their data suggest that if this
world. Most recorded infections with links to bats is a common feature of these viruses, the threat to
are in people who enter tropical caves with large bat humans of viral diseases emerging in wild animals
populations. is perhaps not as great as many have feared.
Rabies is a widespread and serious disease. Only Deliberate attempts to control vampires in caves
one species, the common vampire, is a major vector, (Pint 1994) or bats believed to be crop pests (Makin
and has been estimated to be responsible for the and Mendelssohn 1987, Hall and Richards 2000) not
death of approximately 100,000 domestic cattle per only kill the bats in question, but frequently many
year in South and Central America at a cost of $30 other species that share their roosts. Control mea-
million (Brass 1994). Vaccination programmes have sures, even if species-specific, can be extreme: in the
decreased the incidence of rabies in domestic live- Maldives 75 per cent of the fruit bats were culled
stock, but it still remains a problem. Globally, bat- every 3 or 4 years, ostensibly to maintain a stable
related human deaths due to rabies were believed to population (Dolbeer et al. 1988). The economic sense
be less than 10 per year between 1929 and 1990 as well as the justification for this action (in terms of
(Brass 1994). Rabies is no more prevalent in bats science and conservation) must be questionable.
than in other animal-eating mammals, and since Many of these species are now known to benefit
people do not encounter bats very often, they are humans. Insectivorous bats are important as a natu-
far more likely to catch rabies from domestic or wild ral control of insects, including many medical and
carnivores. In Europe, bats are host to the rabies-like agricultural pests. For example, Cleveland et al.
European bat lissaviruses (EBLVs). There are two (2006) have conservatively estimated that Brazilian
strains, prevalent in just a few species (Harris et al. free-tailed bats (Tadarida brasiliensis) contribute an
2006). EBLV-1 is found almost exclusively in Eptesi- average $0.75 million per year to the $4.6–6.4 million
cus serotinus and EBLV-2 in Myotis daubentonii south-central Texas cotton harvest by eating insect
and Myotis dasycneme. Active surveillance in the pests, in particular the cotton bollworm moth (Heli-
UK found antibody prevalence of EBLV-2 in Myotis coverpa zea). The bats also feed on fall armyworm
daubentonii to be at 1–5 per cent. Transmission to (Spodoptera frugiperda), cabbage looper (Trichoplusia
humans is very rare: an unvaccinated bat rehab- ni), and tobacco bud-worm (Heliothis virescens)
ilitator died in 2002, presumably after being bitten. moths. The names of the second two speak for
It is not considered a significant threat to the general themselves, the fall armyworm caterpillar is a major
public. pest that feeds on a wide range of crops including
 CONSERVATION 257

sweetcorn, alfalfa, barley, cotton, clover, oats, pea- consequent increase in the deposition of seeds of
nuts, rice, ryegrass, sugarbeet, soybean, sugarcane, colonizer plant species.
wheat, and in some years a variety of fruit trees. Roosting bats can occasionally cause a nuisance
Given the abundance of this species (albeit a fraction when they form large colonies in buildings, particu-
of historic numbers), its contribution to the US econ- larly in the tropics. The biggest problems are noise
omy must be very significant. This is being increas- and odour due to the accumulation of droppings.
ingly recognized—many farmers are replacing However, these rarely constitute a significant health
chemical insecticides with bat houses (Fig. 9.10)! hazard. In temperate regions, problems are far less
See also Federico et al. (2008). frequent, and can generally be overcome without
Insectivorous bats are of course important consu- cost to humans or bats if common sense and a bal-
mers of non-pest insects and presumably therefore anced approach is taken—sadly this is frequently
have an impact on natural systems. Two recent papers not the case.
(Kalka et al. 2008, Williams-Guillen et al. 2008) have
shown that excluding bats from native and non-
Climate change
native plant communities increased the number of
insects found on plants and Kalka et al. (2008) also Predicting the consequences of climate change to
detected an increase in invertebrate herbivory. Wil- our fauna and flora is fraught with difficulty (for
liams-Guillen et al. (2008) found that the number of example Araújo and Rahbek 2006) but the difficul-
arthropods on coffee plants also declined, but did ties only begin there. Conservation policy and prac-
not detect a corresponding increase in herbivory tice already tread an extraordinarily complex path
(see Chapter 8 for more detail of these two studies). through theoretical, practical, political, and eco-
Unfortunately, as natural habitat is destroyed, nomic minefields. The debate on how to incorporate
and replaced with commercial fruit crops, bats are climate change into practical conservation is there-
forced to turn to these crops for food and this prob- fore intense (see for example Heller and Zavaleta
ably underlies most of the observed conflict between 2009), not least because it must be done on a land-
farmers and bats. Damage to fruit crops by bats can scape scale that will inevitably cross political bound-
occasionally be severe, but even this is generally aries. The effects of climate change on bats are likely
overestimated—much of the fruit eaten is too ripe to be as marked as on many other animals, and
to be of commercial value. When costs have been equally unpredictable on a fine scale. As local tem-
assessed objectively they are frequently low, and peratures rise bats should, on average, be able to
must be offset against the positive gains of expand their ranges to higher latitudes and alti-
improved pollination success. The durian fruit tudes, but may in turn be forced out of habitat at
alone, from a bat-pollinated tree, adds well over lower latitudes and altitudes. Events will of course
$100 million to local economies in South-east Asia be more complex and other consequences of climate
(for review see Fujita and Tuttle 2001). Bats may also change will distort this simple scenario. Changes in
be important in maintaining populations, through temperature are likely to be accompanied by
pollination and seed dispersal, of wild fruit plants changes in the timing and abundance of rainfall
that are not cultivated but are important to indige- and the frequency and intensity of extreme weather
nous people. Wild strains of many crop plants are events. High intensity summer rainfall, flooding,
important in plant breeding, to ensure a constant and more frequent strong winds have been common
supply of healthy and disease-resistant commercial at my own study sites in the UK in the last few years.
strains. Fruit-eating bats in the New World have If this develops into a persistent pattern it is likely to
been identified as a vital factor in forest regeneration be a significant threat in itself. More extreme
after timber operations, since they are the most weather events in many parts of the world are a
important dispersers of the seeds of ‘colonizer’ tree predicted consequence of most climate change mod-
species (see Chapter 8). Kelm et al. (2008) found that els (IPCC 2010) and tropical storms can have a major
artificial roosts in degraded forest areas were impact on bats, particularly on islands, through
quickly occupied by frugivorous bats with a direct habitat damage and subsequent predation
258 BATS

Figure 9.10 Bat houses are being introduced onto farms in the USA as part of pest control programmes, such as these in (a) California
(photo by Mylea Bayless, BCI). The second photo (b) shows little brown bats, Myotis lucifugus, in a bat house on an Oregon farm (photo
by Merlin D. Tuttle, BCI).
 CONSERVATION 259

and hunting (Dashback 1990, Pierson and Rainey bats were more tolerant, but outcomes were very
1992, Esselstyn et al. 2006.). The mobility conferred dependent upon the climate change model used.
by flight gives bats the ability to move in the face of Under many models, some species showed initial
a changing environment, but their success depends range increases, although they were often followed
upon the integrity of the ecosystem in which they by major declines. The general pattern, whatever the
live and many of the component species may not climate change model, was a decline in species rich-
have the same ability to move and adapt if the ness. Figure 9.11 shows some of the data for temper-
pace of climate change is too great or potential ate bats: current predicted distributions and the
new habitats are fragmented. Equally important, projections for 2050 and 2090 based on different
key resources such as hibernacula may not become climate change models. The A1F1 model predicts
available, or may not be found, as existing ones are massive declines by 2090, whereas B2 predicts
lost. The changes may be so rapid in the future that a maintenance of diversity in Europe, but with a
the behaviour of these long-lived mammals may major shift north.
even be a constraint. As the authors themselves note, interpretation
There is strong evidence to show that other flying should be cautious, since the models took no
animals, such as birds (Huntley et al. 2007, 2008, Jetz account of habitat availability or ecological interac-
et al. 2007) and butterflies (Thomas et al. 2004, For- tions. They are therefore likely to represent the most
ister et al. 2010) are changing their distributions in optimistic outcomes, since, as discussed above, the
response to climate change. Both birds and butter- ecology and behaviour of the bats themselves, and
flies are diurnal, easily identified in flight, and stud- the responses of their preferred habitats, are likely to
ied by a large number of amateur and professional limit the ability of bats to adapt to a changing
naturalists and scientists. The cryptic and nocturnal climate.
bats will not be so easy to study in this context: key Adams and Hayes (2008) took a novel approach
studies on birds and butterflies have relied on good to studying the consequences of reduced water
past and present distribution data, and even these availability as a consequence of climate change in
are lacking for bats. What evidence do we have for the arid regions of the western USA. Using passive
the effects of climate change on bat distribution? induction (or integrated) transponders implanted in
LaVal (2004) analysed bat capture data from 1973– female fringed myotis, Myotis thysanodes, and a
1999 in the Monteverde cloud forest of Costa Rica. submerged receiver, they were able to count drink-
Capture rates did not change, but 24 new, predomi- ing visits made to an isolated artificial water source
nantly lowland species were recorded, suggesting close to a maternity roost. Lactating females had
temperature-induced movements to higher alti- very high water demands, 13 times higher than
tudes, as found by Pounds et al. (1999) for other non-reproductive females. Modelling the climate
vertebrates in the forest. However, the effects of change-induced fall in water storage capacity of
concomitant reforestation could not easily be sepa- the study area showed that without conservation
rated from those of climate change. measures, drinking water was likely to become a
Rebelo et al. (2010) started with the known distri- limiting resource, leading to major declines in repro-
butions of 28 European bat species, dividing them ductive success.
up into three biogeographic groups: boreal, temper- What actually constrains an animal to a specific
ate, and Mediterranean. Using these ‘presence-only’ climatic range? What are the links between the cli-
data and current climatic data (temperature, precip- mate variables so widely used in biogeography and
itation, and humidity), they modelled future distri- climate change research and the animals them-
butions under a number of IPCC climate-change selves? There is a huge literature on physiological
projections, over the next century. Boreal species and energetic constraints and adaptations asso-
were, as expected, predicted to be intolerant of ciated with climate (see for example Johnston and
climate change and many projections led to species Bennett 1996), but relatively few attempts have been
being pushed north, with subsequent extinction pre- made to model distributions on a physiological
dicted before 2100. Temperate and Mediterranean basis. However, Humphries et al. (2002) have done
260 BATS

species was consistent with the model, and by in-

number of species
0-2 corporating climate change predictions they were
3-5 able to show the predicted northward shift in distri-
6-8 bution (Fig. 9.12). It is worth noting that this north-
9-10
ward shift may not be an expansion of the bat’s
range but a shift—more southerly hibernacula may
become too warm. This study, like others predicting
range shifts, makes the assumption that the environ-
ment the bats move into fulfils their other require-
ments: that is it has suitable hibernacula and
suitable foraging habitat in which to feed and
build up fat reserves. This is a big assumption.
Finally, Parris and Hazell (2005) report an inter-
2050 - A1F1 2090 - A1F1 esting aside to the climate change theme: since the
1950s, a localized increase in temperature, a reduc-
tion in the number of frosts, and extensive artificial
watering of parks and gardens (the equivalent of 590
mm rain per annum) has drawn a population of
1000 km grey-headed flying foxes, Pteropus poliocephalus,
2050 - B2 2090 - B2 into the Royal Botanic Gardens of Melbourne. With-
out this very local anthropogenic climatic change,
Melbourne would be well outside the climatic range
for the species.

Infrastructure—wind turbines, roads,

and other built structures
Figure 9.11 Top, current predicted distributions of a ‘temperate’ group
of European bats and below, projections for 2050 and 2090 based on As we’ve seen, some species of bat are adaptable
two different climate change models. Adapted from Rebello et al. (2010), and can exploit built structures as roosts and artifi-
with permission from John Wiley and Sons. cial lights for food. However, there is little doubt
that the balance between the detrimental and the
beneficial effects of human infrastructure falls firmly
just that, by developing an energetic model for on the detrimental side. The physical structures we
hibernation in the little brown bat, Myotis lucifugus, place on the planet take many forms, but to illustrate
in North America. Hibernation temperature will some of the issues I will concentrate on two exam-
have a profound effect on the energy cost of hiber- ples, roads and wind turbines.
nation and a rise in temperature will in most cases
elevate the temperature of underground hibernation Roads
sites. Cave temperatures typically vary very little or As with many other human developments, there
not at all through the year and reflect the mean appears to be an insatiable demand for new roads:
annual temperature above ground. Bats that are across the world, longer, wider, faster, more com-
forced to move to higher latitudes by climate change plex road systems are being built at a pace that
will also be subject to longer winters. Humphries shows no sign of slowing, despite the acknowledged
et al. therefore built a model that took into account conservation concerns and the need to make better
the maximum size of the bat’s energy stores as it use of public transport and reduce fossil fuel use.
entered hibernation, the temperature dependent And, alarmingly, the traffic on these roads continues
rate of energy use, and winter length. The current to increase and congestion continues to be a prob-
limit of distribution of known hibernacula for the lem. A few statistics say it all: only 2 per cent of
 CONSERVATION 261

2080
Present limit
limit
known northern
hibernacula

Figure 9.12 Predicted shift in the northern limit of hibernacula of the little brown bat, Myotis lucifugus, in North America, due to climate change. The dark
grey area represents the current limit, a broad area of uncertainty due to variation in cave microclimate and the amount of fat deposited by bats. Below the lower
boundary of this zone virtually all bats can hibernate successfully, above the upper boundary, almost none. The light grey area shows the predicted shift
northward by 2080. Adapted from Humphries et al. (2002), with permission with Nature Publishing Group.

Germany is made up of landscape fragments greater to traffic when either foraging along roads or when
than 100 km2 (Jaeger et al. 2007) and only 17 per cent attempting to cross roads on commuting flights.
of the US landscape is more than 1 km from a road Being small, bats can probably be pulled easily into
(Riiters and Wickham 2003). Surprisingly, despite the slipstream of passing vehicles. Russell et al.
the many ways in which roads can impact on bats, (2009) watched over 26,000 bat crossings (primarily
rigorous, peer-reviewed studies have only begun to little brown bats, Myotis lucifugus) on a highway in
be published in the last few years. Because the the USA. Bats approached the road using tree can-
impact of roads can take many forms, it’s convenient opy cover and fewer bats were recorded crossing
to deal with them one at a time, but the various where cover was absent. The lower the cover, the
impacts are likely to be cumulative. lower the bats crossed the road. Where bats were
forced to cross an open field in leaving the roost,
Roads: collision mortality
most did so at a height of less than 2 m. Collection of
Although agile and manoeuvrable in flight, most bat road-kill corpses led to a conservative estimate of an
species fly at low speeds (less than 20 km/h) and annual mortality of 5 per cent of the bats in local
many fly close to the ground (0–4 m), particularly roosts. I arrived at a similar, conservative estimate
when crossing open spaces. In contrast to the major- for a road in the UK crossed by bats from a large
ity of birds, most bats also spend most of the time roost (Altringham 2008). Lesinski (2007) recorded
they are out of the roost in flight. They also make bat casualties on an 8 km section of two-lane high-
extensive use of linear landscape features, such as way by weekly searches for corpses over four sum-
woodland edges and hedgerows along roads, for mers. Casualties ranged from 0.3 bats/km/year in
foraging and as navigational aids. In combination, built-up areas to 6.8 bats/km/year where roads
these behavioural traits make bats highly vulnerable were bordered by trees. However, a study by Slater
262 BATS

(2002) suggests, based on the rate of removal of populations is slowed, further increasing the likeli-
‘corpses’ (small pieces of chicken!) by scavengers, hood of local extinction. These factors are likely
that a census of this kind may underestimate wild- to be most relevant to rare species that already
life road kills as much as 12–16 fold. At dawn, small have small and fragmented populations. Indeed,
corpses were, on average, removed by scavengers their very rarity is likely to be due in large measure
within 30 min. The difficulty of spotting small bat to their susceptibility to these and other anthropo-
corpses and the fact that many will be thrown clear genic pressures.
of the road or carried some distance on the vehicle, Kerth and Melber (2009) have recently shown
suggests underestimates will be even greater. Arnett that a major road through a forest in Germany
(2006) found that humans (in the absence of scaven- does indeed have a major barrier effect. Only 3 of
gers) were able to find only 14 and 42 per cent of bat 34 radiotracked, nursing female Bechstein’s bats,
carcasses placed at two wind farm sites. Forman et Myotis bechsteinii, crossed a major road (4–5 lanes)
al. (2003, pp. 120–22) show that wildlife collisions when foraging. All three used an underpass that
(there are no data on bats) increase with increases in was fortuitously built very close to their roost.
vehicle speed and traffic volume and with proximity Other bats from the same colony, and another five
to wildlife habitat and wildlife movement corridors. nearby colonies, did not cross the road. Females
Theoretical studies (for example Lande 1987, With had smaller foraging areas where one edge of
and King 1999, Carr and Fahrig 2001) show that their colony range was the road, relative to those
populations of animal species with low reproduc- from colonies bounding ‘natural’ forest edges and,
tive rates and high intrinsic mobility (such as bats) most significantly, these females had lower repro-
are more susceptible to decline and ultimately ductive success. In the same study, six barbastelle
extinction by the additional mortality caused by bats, Barbastella barbastellus, were also tracked and
roads. five repeatedly flew over the road itself, highlight-
ing the fact that the effects of roads are species-
Roads as barriers
specific.
Roads are potential barriers to flight between roosts In the absence of other studies on bats, it is worth
and foraging sites, between different foraging sites, looking at studies of other animals. Many of these
and between summer, mating, and winter roosts. are summarized by Forman et al. (2003) who show
They can therefore prevent bats from accessing that roads act as significant barriers to a variety of
parts of their habitat or force them to make longer mammals from voles to grizzly bears. Primary
journeys that consume time and energy. Roads may roads are significantly more effective barriers than
act as barriers because they break existing flight secondary roads, and the barrier effect increases
lines, because some species are reluctant to cross with increasing traffic volume. The effects in some
open ground, because some species avoid lit areas cases are severe. Gerlach and Musolf (2000) have
(road and vehicle lights) and, at least initially, shown that populations of bank vole are genetically
because they represent sudden change in the bats’ different either side of a busy four-lane highway,
familiar landscape. Barriers therefore fragment hab- but not either side of a two-lane country road or a
itat, decreasing its area and quality. If habitat area railway. Genetic isolation such as this can only
and quality are major determinants of population occur with very low levels of migration. Highways
size, which they usually are, then habitat fragmen- have been shown to be major genetic barriers even
tation will lower the sustainable population size. to large and mobile animals such as coyotes and
Barriers such as roads may also limit the flow of lynx (Riley et al. 2006). Corlatti et al. (2009) asked the
individuals between populations. This has two con- question: do overpasses connect landscape on
sequences: (a) it reduces gene flow and increases either side of roads and prevent genetic isolation?
inbreeding, leading to a greater risk of local extinc- The short answer was that a lack of effective moni-
tion (b) it slows recovery from local population toring and poor experimental design means that the
crashes because recruitment from neighbouring question cannot even be addressed effectively—a
 CONSERVATION 263

conclusion I came to when specifically studying Roads: time lag and cumulative effects
bats (Altringham 2008). Balkenhol and Waits
Most of the factors discussed above will be cumula-
(2009) review the potential of genetic approaches
tive. The effects of each need not therefore be great to
to the study of road ecology.
have a profound effect on a bat population. Further-
Studies of birds are likely to be particularly rele-
more, there will be a lag (known as the extinction
vant to bats. Develey and Stouffer (2001) and Laur-
debt) between cause and effect (Tilman et al. 1994,
ance et al. (2004) have shown that even narrow,
Loehle and Li 1996). This is illustrated in Fig. 9.13.
unpaved forest roads can act as a barrier to forest
The effects of habitat loss and reduced habitat
birds, supporting the view that woodland bats are
quality may be seen quickly, but collision mortality,
likely to be most affected by roads.
unless very high, may not have a significant effect
Roads: loss of habitat for several generations and the barrier effect may
take several more generations to show itself. A
Road development frequently involves the removal
study of the effects of roads on wetland biodiversity
of trees and buildings that hold potential or actual
(birds, mammals, reptiles, amphibians, and plants)
bat roosts. The removal of trees, hedges, scrub,
suggests that the full effects may not be seen for
water bodies, and unimproved (‘natural’) grassland
several decades (Findlay and Bourdages 2000).
also reduces available foraging habitat. The road
A very recent study by Berthinussen and Altring-
surface alone removes significant habitat: 7 ha for
ham (2011) has shown that bat species richness and
every 10 km of 7 m wide, two-lane road. Associated
abundance were strongly and negatively correlated
verges and structures remove yet more habitat. Sev-
with proximity to a major road. Activity 1.6 km from
eral studies (for example Rydell 1992, Blake et al.
the road was three times greater than at 100 m. An
1994, Stone et al. 2009) have shown that road
effect of this magnitude over such a distance sug-
lighting deters many bat species from approaching
gests much more than simple, local avoidance.
not just the road, but the habitat for some distance
I haven’t even discussed pollution from roads:
on either side of it. However, light can also attract
transport is the fastest growing source of green-
some species. This may be a mixed blessing: certain
house gases. In the USA, over 50 per cent of domes-
types of lighting attract insect prey, increasing for-
tic CO2 emissions come from cars.
aging efficiency, but bats may be at greater risk of
collision with traffic.
Again, with so few studies on bats, it is helpful to Wind turbines
look at relevant studies on other animals. A series of It is ironic that when there is a desperate need to
studies on birds (Foppen and Reijnen 1994, Reijnen develop sustainable, environmentally-friendly forms
and Foppen 1994, 1995, Reijnen et al. 1995, 1996) has of energy generation that electricity generation with
shown that traffic disturbance reduces both species wind turbines, a technology with real potential, has
richness and population density on either side of become so highly controversial. There are several
major highways, in both grassland and woodland. reasons why wind turbines are unpopular, two
The distance over which these effects were observed being of particular relevance here: they kill bats
varied with species and traffic volume, but signifi- (and birds) and they have an impact on terrestrial
cant effects were detected up to 2.8 km from roads. habitats due to the infrastructure associated with
The magnitude of the effect was also substantial: 26 them: all need roads, maintenance facilities, and a
of 43 species showed reduced densities, and within means of taking the power to the grid. In addition,
100 m of the road this reduction was as high as 98 many people feel very strongly that they have an
per cent. Traffic noise was the most important factor unwanted visual impact on the landscape: they spoil
in these studies (see above). the view from their windows and the sweeping
Bats will roost in road bridges (for example Kee- panoramas of their recreational walks. There is a
ley and Tuttle 1999), but not over busy roads. How- complex spectrum of arguments from the purely
ever, bats roosting and foraging close to roads, aesthetic, through intrusion into the wilderness, to
perhaps also attracted by new planting, may do so the ‘not in my backyard’. They are not relevant here
at the expense of greater risk of collision. other than because they are sometimes the most
264 BATS

natural,
pre-construction
fluctuations

habitat loss

low
population size

extinction risk reduced habitat quality

collision mortality large fluctuations

and a high,
post-construction
barrier effect extinction risk

time lag (extinction debt) and

the cumulative consequences
of several ecological effects

road construction time

Figure 9.13 Hypothesized extinction debt due to road-building. Based on Forman et al. (2003), used with permission from Island Press, Washington, DC.

powerful voices in the public debate—and they do ones being commissioned are now reaching 5 MW,
not always lend it clarity. but there is likely to be a poor correlation between
Wind turbines are large structures, typically power and kills: location and height are probably
60–100 m high with blades up to 50 m long, and the key factors. Published figures on kills, particu-
they are getting larger. In 2009 they had the capacity larly early ones, are sometimes uncorrected for car-
to produce 160 gigawatts or 2 per cent of the world’s cass loss due to scavengers or poor detection (see the
energy use, and this figure is set to grow rapidly section above on roads and collision deaths) and so
(WWEA 2010). There has been concern about the many are very likely to be underestimates. The pro-
impact of wind turbines on birds for many years, portion of bats killed that are not recovered in
but their potential to kill bats has only recently been subsequent searches can be large and very variable,
addressed. Research is still in its early stages and depending on the terrain under the turbine, the
much of the work published to date is concerned methods, timing, and intensity of searches, and the
with devising and evaluating methods for assessing abundance of predators and scavengers. I don’t
the scale and nature of the impact and methods to want to get embroiled in global numbers, since this
mitigate against this impact. The most detailed of is probably not the best way to view the problem.
these is Kunz et al. (2007), and a good review of the In North America about 75 per cent of fatalities
literature has been written by Cryan and Barclay are from a small group of migratory, tree-roosting
(2009). Despite the short time since research began, species, in particular three members of the Lasiurus
some conclusions can already be drawn. Wind genus. In Europe, ecologically similar Nyctalus, Pi-
farms on forested ridge tops in North America are pistrellus, Eptesicus, and Vespertilio species are most
killing up to 50 bats per megawatt (MW) per year at affected (for example Dürr and Bach 2004, Rodri-
some sites. At many other sites fatalities are more gues et al. 2008, Rydell et al. 2010). Most bats appear
usually less than 1 to 10 bats/MW/year. A typical to be killed at low winds speeds (less than 6 m/s)
commercial turbine is rated at 0.5–3.0 MW and new (Horn et al. 2008) and where deaths are greatest
 CONSERVATION 265

turbines are primarily killing these species during little firm evidence in support of any theory, but it is
their autumn migration. However, bats were not all important that the most plausible are tested if we are
simply ‘passing through’, but also fed around tur- to make sensible decisions about how to reduce the
bines and appeared to investigate the blades (Horn impact of wind turbines on bats. A recent paper by
et al. 2008, Cryan and Barclay 2009, Rydell et al. Long et al. (2010) suggests that turbines of certain
2010). Cryan and Barclay (2009) reviewed the possi- colours may be attractive to insects and in turn
ble reasons that bring bats and turbines together and attract bats.
the causes of death: why are bats there and how are Sovacool (2009) looked at the problem from a
they killed? The evidence suggests that bats, unlike very different perspective and highlights two sig-
birds, do not generally collide with stationary tur- nificant issues in conservation—the need to make
bines or turbine pillars. Death appears to be caused appropriate comparisons and to look at the big
by direct collision with moving blades or by baro- picture. The paper has some serious flaws which
trauma: air pressure changes around the blades were quickly pointed by Willis et al. (2010). Despite
cause rapid expansion of the air in the lungs that its flaws, which I will come back to, there are some
leads to internal injuries. Collisions may be random issues worthy of discussion, a point acknowledged
events, they may be unfortunate coincidences due, by Willis and colleagues. The critical questions
for example, to the fact that turbines are sited on Sovacool asked were: (i) are all wind turbines
migration or other flight lines, or it may be that bats equally ‘bad’, (ii) how bad are they compared to
are actually attracted to wind turbines—or indeed a conventional sources of energy, and (iii) how do
combination of factors. Rydell et al. (2010) have they compare to other anthropogenic causes of
recently explored the very plausible hypothesis bird and bat mortality? It is clear from the above
that bats may be attracted by migrating insects that summary that the answer to the first question is no,
accumulate around turbines. Since turbines clearly some wind turbines/sites are clearly much worse
kill certain species at certain times, most collisions than others. I don’t want to spend more time on
are not random. Migration itself, or behavioural this particular issue, other than to stress that we do
traits characteristic of migratory species, (for exam- need to identify which are good and which are bad
ple high, aerial hawking) are clearly important and and why. The other questions take us into a mine-
there is evidence that fatalities are concentrated in field of contentious complications. Sovacool cites a
migration corridors (for example Baerwald and Bar- number of studies that compare mortality due to
clay 2009). In addition to using migration corridors wind turbines, primarily on birds, with that from
it may be that migratory bats, by flying high and by anthropogenic factors that rarely receive the atten-
exploiting high-altitude insect patches, make them- tion of environmentalists: impact with tall buildings,
selves more vulnerable. However, there are still windows, road traffic and power lines, predation by
some puzzles associated with migration, such as cats, poisoning by pesticides, habitat loss through
why there appears to be no spring mortality peak. agriculture, and others. A report by The National
Are spring and autumn migrations (of bats and in- Academy of Sciences (NAS 2007) concluded that
sects) different? The study of migration itself in bats wind turbines accounted for less than 0.003 per
is also a relatively new science (see Chapter 7) and cent of anthropogenic bird deaths. However you
we need to understand migratory behaviour better manipulate the assumptions and variables in the
before we can properly assess its role in wind tur- analysis it is hard to avoid the conclusion that
bine mortality. Cryan and Barclay (2009) make wind turbines are responsible for relatively few
detailed and persuasive arguments in favour of the bird deaths. With some critically important caveats,
careful testing of a wide range of ‘coincidence hy- I suspect the situation for most bats may not be so
potheses’ that they lay out in some detail. They also very different. But, as Willis et al. (2010) point out,
discuss the possibility that turbines attract bats. The bats are not birds, an oversimplification that Sova-
more plausible ideas collated and explored fall into cool seems quite happy with! The ‘rules’ that deter-
two areas: attraction for roosting, feeding, or mating mine kill rates in birds and bats are probably very
or curiosity and/or misperception. As yet, there is different and we cannot simply extrapolate from
266 BATS

one group to the other. We also know that wind making them more efficient at generating energy—
turbines target particular and potentially rare spe- we also need to make them less efficient at killing
cies, so that the total number of bats (or for that wildlife. More importantly, we need to be less prof-
matter birds) killed is misleading. The life history ligate in our use of energy. If we used energy more
strategies of birds and bats can also be very differ- efficiently we would not need to generate so much
ent, so the consequences of increased mortality are of it.
not comparable. However, we are still left with the The worthy causes of conservation are frequently
question: why are wind turbines being given so hijacked by those who wish to prevent develop-
much attention, when death by other means receives ment for reasons other than conservation, and
so little? Is it because it is a new threat? Because we wind turbines are a good example of this. Many
feel that there is little we can do with the long- people do not think that turbines have a place in
standing problems that are part of the fabric of ‘wild’ places or ‘unspoilt’ landscapes (even if some
life? I’m not suggesting less attention be given to of the latter are in fact agricultural deserts) or sim-
wind turbines, but more to these other issues. ply in the view from their window. This is not the
One way to increase the objectivity of an analysis place for a debate about the ruination of the aes-
is to compare deaths per gigawatt hour (GWh) for thetics of wilderness or the view from your home,
different power generation methods. The calcula- but conservationists should question the motives of
tions are necessarily underpinned by assumptions detractors with a hidden agenda who become tem-
and simplifications, but Sovacool’s calculations ap- porary conservationists. I receive frequent requests
peared to be supported by published data, which I for help in blocking developments where the not-
will leave you to look at. However, Willis et al. in-my-backyard motive is not even concealed: the
(2010) point to some serious flaws in the data, bats are merely a means of achieving sometimes
which you may also care to examine. Having done less laudable aims.
so myself, what conclusions can I draw? Sovacool’s
calculations for bird and bat deaths per GWh for
wind, nuclear power, and coal-fired electricity gen- From the negative to the positive:
eration are indeed flawed, and any conclusions
how to conserve bats
based on their absolute values and their values rela-
tive to each other are suspect, so I will not repeat That’s enough about our unwanted impact on bats,
them here. However, the principles behind the cal- let’s look at what we are doing, or can do, to con-
culations remain sound. For example, they include serve bat populations. We have covered at least one
the associated running costs, not just the direct costs: important task already. It’s what this book has been
those on the effects of coal-fired power generation all about so far: understanding bats and understand-
include the effects of coal mining, plant operation, ing our impact on them. Without this information
acid rain, mercury, and climate change. Sovacool’s we cannot practice effective conservation. Conser-
analysis may be flawed, and his neglect of important vation practice should be based on the best available
biological detail unintentionally provocative, but his scientific knowledge and its effectiveness or other-
approach is surely the right one: make appropriate wise should be assessed using the most appropriate
comparisons and look at the big picture. Newman and scientifically rigorous methods. I could be
and Zillioux (2009) made a similar study to that of accused of stating the obvious, but regrettably con-
Sovacool and came to broadly similar conclusions. servation frequently does not follow this simple
We clearly need to design and run wind turbines to rule. It is broken for many reasons: practicality and
minimize their impact, but do we need to stop build- pragmatism, constraints on time and resources (that
ing them as some would advocate? Whether they is money), political or economic barriers, entrenched
replace or supplement existing energy generation attitudes, poor communication between researchers
plants, with care, they may prove to be the lesser and practitioners, and so on. Conservation is almost
of the energy generation evils. Ongoing research is always a compromise.
 CONSERVATION 267

As ever, I make no attempt to be comprehensive biodiversity in the long term? The total reserve area
in my coverage, but I hope what follows covers the was over 600,000 km2, but 35 per cent of this was in
major topics and highlights some of the more impor- Alaska, as far away from the major biodiversity hot-
tant developments, such as the shift to landscape- spots as one can go. Of 23 ‘protected’ 250  250 km
scale conservation and the increased reliance on an quadrats (those with more than 10 per cent of the
evidence-based approach. Inevitably, subjects over- land protected), 15 had no threatened or small-range
lap, and it is not always clear where a subject fits bat species. Over 80 per cent of threatened and/or
into the grand scheme, so what you hope to find small-range bat species were not adequately pro-
may not always be where you expect to find it. tected by reserves. This mismatch across the whole
of the western hemisphere is clearly evident from
From conserving the ‘best’ to improving Fig. 9.14, which also shows that modelling reserve
distribution to better protect bats yields very differ-
the whole ent maps.
Ever since the conservation movement started, Ceballos and Ehrlich (2002) have shown, perhaps
effort and resources have gone largely into protect- not surprisingly, that mammalian extinctions are
ing those areas that retain unspoilt habitat, have concentrated in areas of high human density,
high biodiversity, rare ecosystems, or endangered where intensive agriculture and hunting take their
(and often charismatic) species. This approach has toll. Andelman and Willig (2003) also found that for
undoubtedly been hugely successful, but in the face bats, regions low in reserves and high in threatened
of climate change and the ever increasing human species also had large and expanding human
pressure on the rest of our environment (for exam- populations.
ple Sala et al. 2000) we must do more to protect The obvious need to take a more structured and
biodiversity outside reserves. We must also facilitate analytical approach to the creation of reserves, to
the movement of organisms between suitable habi- maximize their effectiveness, has led to the develop-
tat patches inside and outside protected areas as ment of a variety of methods to assess just how well
climate change and land-use change drive organ- they do their job. These methods are based on mea-
isms out of once suitable areas. Issues such as scale- sures such as rates of deforestation and hunting or
dependence, connectivity, landscape structure and the diversity of indicator groups, and often involve
heterogeneity, and a more integrated approach to comparisons between the reserve and the surround-
reserve selection (and conservation in general) have ing land (for example Chape et al. 2005, Balmford
therefore come to the fore. This topic is the most et al. 2005). These methods are becoming increas-
‘theoretical’ in the section, partly because the science ingly sophisticated as the quality of data and ana-
and the practice are so new but mainly because so lytical methods improve and new problems are
little has been done in relation to bats. I’ll start identified and taken into account. One such problem
with one of the few studies that have looked at bats. is ‘leakage’, discussed by Ewers and Rodrigues
This large scale perspective was taken by Andel- (2008). When activities, such as forestry, are stopped
man and Willig (2003), who highlighted the pro- in reserves, there may be a tendency for this activity
blems of conserving bats through a reliance on to be displaced into the area surrounding the
existing protected areas that were created over reserve, exaggerating the differences and thus lead-
many years, often with very different goals. They ing to undue optimism about a reserve’s effective-
compared distributions of 249 of the 255 New World ness. The false picture created could have a major
species with the distribution of 1,413 reserves of influence on policy and be damaging to conserva-
IUCN Category I and II status. A little over 80 per tion effort. Leakage could also accelerate the rate at
cent of reserves had areas of less than 100 km2, which the reserve becomes isolated from the rest of
almost 60 per cent had areas of less than 10 km2, the landscape, since the greatest damage is likely to
and the median size was less than 5 km2. How many be done around the reserve boundaries—and this
of these are big enough to retain all or most of their brings me on to the next topic.
268 BATS

Species richness
1–13
14–26
27–38
Total area in reserves (sq km) 39–51
52–64
1–5 65–76
6–14 77–89
15–32 90–101
33–59 102–114
60–117 115–127
118–239
240–496
497–1500
1501–3510
3511–3674

Richness of small-range
species
250 km × 250 km quadrats
1
with >10% of their area 2
in reserves 3
4
5
6
7
8
9

Richness of threatened
species
1
2
3
4
5
6
7
8
9

Figure 9.14 The maps on the left show the distribution of reserves in 250 x 250 km quadrats in the New World, expressed as total area and as
quadrats with more than 10% of their area within reserves. On the right three different ways of selecting areas that need protection: species richness, richness
of small-range species, and richness of threatened species. From Andelman and Willig (2003), with permission from John Wiley and Sons.
 CONSERVATION 269

In addition to creating single reserves in the best powerful applications that can be more widely used
place and in the best way, we need to think about in conservation (Guisan and Thuiller 2005). One of
how reserves interact with each other and with the the very significant benefits of the GIS-based
landscape around them. We need to blur the bound- approach to conservation practitioners is that the
aries between reserves and the ‘working’ or human results of complex studies can often be presented
landscape around them to enable bats and other in a series of easily interpreted maps.
organisms to move as freely as possible. This has Perhaps the first GIS-based attempt to better
frequently happened, by accident or by design, and understand the relationship between bats and their
to varying extents: a reserve boundary may vary landscape was made by Jaberg and Guisan (2001).
from little more than an imaginary line present They took bat distribution data for western Switzer-
only on a map, to a dramatic change in habitat that land compiled using a wide range of methods to
can be seen from space. These issues are particularly identify roosting, foraging, or hibernating bats of
important to bats, because of their mobility and their 20 species. Abundance was based on a semi-quanti-
use of resources that can be scattered over a wide tative index of ‘regularity’—the number of years a
area, but also because the structure of the landscape species was recorded in a particular location over
can create barriers as well as movement corridors, as the 14 years of data collection. Seventeen GIS-based
discussed earlier in this chapter. Debate about the environmental predictors, at 25 and/or 100 m reso-
size, shape, number, and interconnectedness of re- lution, included a range of topographical and land-
serves has gone on for many years and is covered in use variables. Relationships were then explored
the books I listed at the beginning of this chapter. between the environment and the ensemble and a
A very recent study investigates the importance of number of selected species, using both univariate
some of these factors to bats. Lookingbill et al. (2010) and multivariate statistical models. Elevation and
assessed the importance of wetland area, connectiv- natural vegetation structure were the most impor-
ity, and landscape to bats in five small national tant predictors of the distribution of most species.
parks in the eastern United States with varying de- Anthropogenic landscapes, whether urban or agri-
grees of urbanization of the surrounding landscape. cultural, were generally poor predictors. Ensemble
Wetland area was a good predictor of activity for composition showed marked gradients from lake
some species, urban area for other species. The addi- shore to the Jura mountain tops and from urban
tion of a measure of connectivity between wetlands areas to forest. The results were used to draw habitat
significantly improved correlations. suitability maps, with prediction successes between
The logical next step of this approach is to con- 71 and 87 per cent. This was surprisingly good
sider managing the entire landscape for bats. To do considering the deficiencies in a data set that was
this we must look at bat activity in relation to a not compiled specifically for the task and was cer-
much larger number of environmental variables, tainly not complete, but the authors do draw atten-
and consider the important question of scale, as tion to some possible biases. Nevertheless, it was
part of the overall management process (see Chapter clear that the approach was useful in the study of
8 and Wiens 1989, Willis and Whittaker 2002). This bats and it is now being used with increasing fre-
is becoming increasingly feasible with the growing quency, often using data collected specifically for
sophistication and power of geographical informa- the task.
tion systems (GIS) and statistical tools, and the ad- Greaves et al. (2006) combined generalized linear
vances made in integrating them. For example, Store modelling (GLM) and GIS to develop predictive
and Jokimaki (2003) pioneered a versatile approach maps of the distribution of the rare long-tailed bat,
that calculates and maps habitat suitability indices Chalinolobus tuberculatus, in New Zealand, based
and can handle multiple species and scales—one of on records collected over a nine-year period. The
many such developments. This is a rapidly growing value of these maps was then tested against obser-
field, with the publication of new methods almost vations at over 150 field sites, an important step not
matching the pace of their application. Perhaps taken by many previous studies, most of which
the next phase will be one of standardization into relied on statistical resampling techniques. Fourteen
270 BATS

landscape and climatic variables were used in the and abundance it was likely to support and the
model, six of which were shown to be good predic- structural complexity it would present to the bats.
tors of presence. This species is found on forest Bat activity was most clearly associated with fine
boundaries of low slope, where winter temperatures scale habitat structure and became less influenced
are not too low and there’s plenty of sunshine—and by vegetation as spatial scale increased. A similar
it helps if southern beech, Nothofagus species, makes result has been reported by others studying forest
up a significant proportion of the trees. Historical habitats in the Pacific North West (for example
surveys that relied on anecdotal and subjective Erickson and West 2003) using different approaches.
information to target effort had a detection fre- It will be interesting to see if this result holds for
quency of just 12 per cent. The predictive value of different and perhaps more varied habitats. Having
the map was modest, but it did increase detection lived in the Pacific North West I know that the
frequency to 45 per cent, a significant practical forests are rich and diverse—but they are neverthe-
advantage. The addition of as yet unavailable infor- less all forests!
mation, for example on forest composition, struc- The ecological and statistical methods underlying
ture, and maturity, is likely to lead to significant the mapping of species distributions, habitat suit-
improvements in the models and maps. ability, and so on are complex and there are still
Milne et al. (2006) took a similar approach to the many pitfalls and problems to overcome (for exam-
modelling of habitat suitability maps for 25 of the 28 ple Dormann et al. 2007). However, GIS-based
insectivorous species at the ‘Top End’ of Australia’s mapping is becoming an increasingly valuable con-
Northern Territory. Data were collected from histor- servation tool due to its enormous potential in
ical sources and from a series of acoustic surveys. power, accuracy, and ease of use. Maps can distill
Over half of the maps were robust, and annual a mass of information into a form that can be readily
rainfall and habitat complexity were significant in understood, and are often aesthetically pleasing.
most models, but the maps showed considerable Figure 9.16, a work in progress of Chloe Bellamy, a
inter-specific variation. By combining the maps a member of my research group, summarizes the
species richness map was derived, showing clear principles, if not the pitfalls and problems, which
hotspots in the north-west and north-central zone are too complex to discuss here. Bat distribution
(Fig. 9.15), highlighting the potential value of this data for the Lake District National Park in the UK
approach in identifying localities for the use of lim- are combined with other GIS layers that detail dif-
ited conservation resources. Most of one of these ferent environmental data, such as elevation, aspect,
hotspots falls within the Kakadu National Park. Is slope, habitat, water, urban development, and so on.
this due to bias from the intensive survey within the Once collated models are built at different spatial
park, or good judgement when the park was scales incorporating different variables, and the pre-
formed? The complex river systems and geology of dictions tested, ideally against independent data
the park provide the habitat complexity associated from outside the immediate study area. The final
with high richness, suggesting good judgement. In model, with the greatest accuracy, may include
favour of the sampling bias hypothesis is the fact data at several spatially explicit scales.
that the other hotspot is the relatively accessible
land around the city of Darwin.
Agroforestry and timber-extraction
Ober and Hayes (2008) incorporated scale-depen-
dence into their study of bats foraging along streams Trees are the most important habitat feature in the
in the hills of the Oregon coast of the USA. They lives of most bat species. How we manage the
used three ascending scales, defined as stream reach world’s forests is therefore vitally important to
(30 m of stream), nightly activity area (1.5 km bats. Is there any room for optimism in the face of
radius), and watershed. Habitat was characterized an ever more rapid rise in the rate of deforestation in
on the basis of richness, composition, and cover in the world’s biodiversity hotspots? Clearly our most
terms that could be related to the insect diversity important task is to try to stop this madness, but
 CONSERVATION 271

Kakadu National Park

Relative species richness light: low, dark: high

Hipposideros diadema

Figure 9.15 The large map shows predicted

relative species richness of the ‘Top End’ of
the Northern Territory of Australia. The small
maps show species records (left) maps and
Taphozous kapalgensis predicted probability of occurrence (right) for
two representative species. From Milne et al.
Records probability of occurrence (20-100%) (2006), with permission from Elsevier.

many of the best mechanisms for this lie more in the many parts of the world there is no longer an option
realms of education, economics, and politics, which of leaving forests ‘untouched’, since we have
I will come to later. already been exploiting them for hundreds if not
If we accept that trees are a resource that will thousands of years. One possible way is to adopt
inevitably be exploited by humans, and that some more sensitive harvesting regimes than the large-
forests simply can’t be left as nature reserves, are scale clear-felling that has been a common practice
there ways in which we can harvest them and mini- throughout the world. Another approach is to inte-
mize the impact on bats, or even benefit bats? In grate forestry with other forms of agriculture: a
272 BATS

model inputs and bird ensembles were richer in cabruca than in

the adjacent forest, but where forest was rare, cab-
ruca had poorer bird and bat ensembles and lacked
some of the species characteristic of the forest habi-
tats, such as the gleaning insectivores as discussed
earlier in the chapter (for example Medellin et al.
2000). Cabruca can clearly be valuable habitat, but
only if there is an abundance of natural forest
species data environmental data around it. It is not a substitute for natural forest
model building and a careful balance must be struck between the
two if this form of agroforestry is to be an effective
conservation measure. Bats were found to be roost-
model output ing in the cabruca, not only foraging in it or simply
habitat suitability index: flying through. Some caution should be exercised
suitable however: bats were sampled using mist nets and it
is likely that the simpler structure of the cabruca
not suitable may have made it easier to catch bats than in
the forest. Capture rates were certainly higher, but
whether this influenced species composition is
not known.
Harvey and Villalobos (2007) also looked at both
bats and birds in similarly structured cacao and
banana plantations in Costa Rica, comparing them
Figure 9.16 Illustration of how habitat suitability maps are built. GIS with surrounding forest and monocultured plan-
layers, for the Lake District National Park in the UK, of bat occurrence and tain. Agroforestry plantations retained rich bat en-
habitat data, are combined in models that predict habitat suitability across sembles, including three species of conservation
large areas. See text for more details.
concern and the same dominant species. Plantations
also had more nectarivorous species than surround-
practice that goes under the name of agroforestry. ing forests, since bananas, present in both systems,
There is an extensive literature assessing the benefits provide more nectar. Again, however, agroforestry
or otherwise of these approaches (for example Bhag- was not a substitute for forest. It is interesting to
wat et al. 2008, Garcia et al. 2010) and they have note that in both studies bats appeared able to adapt
potential in both tropical and temperate regions. to agroforestry. As expected, there was minimal use
Agroforestry will probably have its greatest impact of the monocultural plantain fields. A number of
in the tropics, where pressure on forests and the other studies support these conclusions, including
potential gains for humans and biodiversity are Estrada and Coates-Estrada (2001) who also looked
greatest. It has the potential to protect reserves by at shade coffee plantations. To sum up, agroforestry
making other land more productive, it can provide can be good for bat conservation if it has good
buffer zones around reserves, and it can improve spatial and species complexity and sufficient natural
connectivity between forest patches. There is much forest is retained. This has to be achieved under
discussion about what it has done and can achieve circumstances that encourage farmers to plant
for biodiversity, which I leave you to explore, and trees and retain forest, ideally through a direct
I will stick primarily to bats. Faria et al. (2006) looked reward system, for example by producing certifi-
at bat and bird diversity in ‘cabruca’ or shade cacao cated environmentally-friendly products for a guar-
plantation in Bahia, Brazil. The understorey in cab- anteed market. This would ideally be supported by
ruca is replaced by cacao, and the native canopy education and training programmes. There are good
species are thinned or left unmanaged. Where natu- examples of such systems in many parts of the
ral forest made up over 90 per cent of the land, bat world and we must work to promote more.
 CONSERVATION 273

A final point on this issue, raised most recently by Extraction techniques will also determine roost
Garcia et al. (2010). Integrating agricultural land- availability: clear cuts don’t have roosts, other than
scapes and forest in nature conservation, in a land- around the edges! Selective thinning on the other
scape scale approach, must not only work for bats hand can leave suitable roost trees standing. There
and other wildlife—it must work for the people too. is an extensive literature on the use of temperate
They suggest that many of the current approaches forest by bats which, although relevant, is periph-
have yet to get the balance right, with too little con- eral to my main aim in this section. If you want to
cern for the aspirations of farmers. Success requires know more, an up-to-date summary can be found in
an equal partnership between conservation and Lacki et al. (2007).
farming—with the benefits to both being obvious. The merits of reduced impact logging systems
In both tropical and temperate forests the trees have been explored primarily in the tropics. Clarke
themselves are usually the crop and only a very et al. (2005a) compared the effects of open range and
small proportion of the world’s forests are pro- periodic block harvesting on bat ensembles in forest
tected from this massive agricultural industry. The in Trinidad. Open range logging criteria are based
most widespread timber extraction method in most on minimum girths for extraction for a limited num-
parts of the world has been large-scale clear-felling, ber of species and show minimal concern for conser-
a highly destructive process that forests recover vation. Periodic block criteria, on the other hand,
from slowly, if at all. There has been a shift towards have been designed to help maintain biodiversity.
selective logging, driven primarily by economics, Logging is carried out over one to two years and
but with increasing consideration given to conser- plots are then left for thirty years. Selection is based
vation in some countries (for example Dickinson on factors such as canopy closure, the maturity of
et al. 1996, Grieser John 1997, Dawkins and Philip trees, their value to wildlife, and rarity. Only 4–8
1998), although there is considerable debate on the trees/ha are extracted, rather fewer than under open
best methods to use (for example Chazdon 1998). range management. Species richness was similar
Natural forests are typically a mosaic of stands of in primary forest and in forest managed on both
different densities and age structures, with inter- regimes, and species abundance distributions were
vening open areas, due to the underlying terrain also similar. However, diversity was lower under
and the ravages of disease, weather, and natural both logging regimes, with a reduction in the num-
succession. Logging on a small scale can mimic this ber of gleaning insectivores and carnivores and an
process, but large-scale clear-felling comes nearer increase in the number of frugivores. This is a recur-
to mimicking, on a too frequent and regular basis, ring theme in disturbed forests, described earlier in
the catastrophic environmental events that occur the chapter. The increase in frugivore abundance is
very rarely in nature. probably due to an increase in the pioneer plant
I’ll start with work in the boreal forests of Canada species that they feed on and disperse. The bat
by Patriquin and Barclay (2003) who compared bat ensemble structure under periodic block manage-
activity in cleared, thinned, and unharvested stands. ment did show signs of being more similar to that
As might be expected, the results were species- of primary forest, although the differences were
dependent, with the large, fast-flying, open-air for- small. Studies in the same forest by Clarke et al.
ager, the silver-haired bat, Lasionycteris noctivagans, (2005b) demonstrated that the bat ensemble has
foraging only in clear cuts. The gleaning northern some capacity to recover over the 30 years between
long-eared bat, Myotis septentrionalis, preferred logging periods. Gleaning insectivore and carnivore
unharvested forest, and the little brown bat, Myotis numbers increased and the number of frugivores
lucifugus, a small aerial insectivore, fed primarily decreased, but recovery to pre-logging composition
along the edge of clear cuts. The message to take was not complete. These studies do suggest that
from this is that management affects different spe- periodic block, and perhaps other conservation-
cies in different ways and we must understand the minded logging systems, can help to maintain bats,
needs of different species to formulate sensible but the logged forests will nevertheless have bat
plans. This study looked only at foraging bats. ensembles that differ from those of primary forests.
274 BATS

Castro-Arellano et al. (2007) assessed the effects of Bats in agricultural landscapes

reduced impact logging (RIL, extracting 18 m3 of
For many thousands of years humans have been
timber/ha) on phyllostomids in Brazilian Amazo-
replacing forest with agriculture and bats have had
nia, 20–42 months after harvesting. Although 46
to adapt to the agricultural landscape or perish.
species from five families were caught, phyllosto-
Some forms of agriculture can to some extent
mids accounted for 91 per cent of the species and
mimic forest, such as shade plantations of cacao
95 per cent of the individuals, leaving too few data
and coffee (Fig. 9.17a), as discussed in the previous
for detailed analysis of many species. Of the 15
section. This section is concerned primarily with
species studied in detail, 10 showed no change in
landscapes in which little is natural and trees are
abundance, three increased and two decreased, with
generally a minor component (Fig. 9.17b). But I’ll
a resultant increased diversity and reduced domi-
start with an exception that bridges the gap, the
nance in harvested forest relative to controls, but
olive groves of the Mediterranean region. These are
changes were relatively small. As in other studies,
part of a huge industry and the groves are a major
several species were recorded only in control forest,
component of the agricultural landscape. Bat activ-
including two gleaning animalivores. As in the pre-
ity in olive groves can be high (Russo and Jones
vious study, logging led to an increase in the species
2003, Davy et al. 2007) and they are used by at least
feeding on and dispersing early successional plants.
one rare species, the Mediterranean horseshoe bat,
The conclusion was that RIL, which appears to have
Rhinolophus euryale (Russo et al. 2002). This is per-
strong similarities to periodic block, is probably a
haps not too surprising since the tree is native to the
viable conservation management logging regime.
region, has been managed for thousands of years,
The authors also summarize some useful guidelines
and some groves are still under low input tradi-
to help protect rare and keystone species: the main-
tional management. However, intensification has
tenance of high densities of older seed trees, the
reduced the value of olive groves to wildlife in
protection of riparian areas, the establishment of
recent decades and caused considerable environ-
no-logging zones within harvested stands, and
mental degradation through soil erosion and pollu-
reduced disturbance of standing and fallen dead
tion. Changes in European environmental law,
trees. However, the same group, analysing data
agricultural funding, and initiatives such as the
gathered in the same study (Presley et al. 2008),
EU-LIFE ‘Life among the Olives’ programme
draw some more disturbing conclusions in a
(Camarsa et al. 2010) hopefully reflect a change for
subsequent paper: RIL led to reduced abundance
the better. I’m not aware of any work on bats in
or loss of 16 species, most of them rare species,
other silvicultural systems, but the potential for bat
including 6 gleaning animalivores. Control forest
conservation is certainly worth exploring. The cork
was home to 7–17 more species than logged forest.
oak woodlands of Spain and Portugal are the target
This is not very surprising—many of the studies
of campaigns to maintain traditional management
already discussed noted the loss of certain species,
practices because of their value to wildlife, and they
in particular the rare gleaning species. Whether the
are likely to be important to bats.
loss is due to the disturbance of their closed canopy
This move towards more sustainable agriculture
habitat, loss of food and/or roost sites, or some
across many parts of the world has been welcomed,
other factor is not yet known. The message is clear:
promoted, and supported by conservationists, but
whether a logging regime is ‘good’ or ‘bad’ for bats
the welcome is often tainted by scepticism over the
is determined by what you measure and what you
pace of progress and the evident lack of conviction
look for. The authors suggested that by arranging
amongst many politicians and agricultural indus-
logged sites within a matrix of undisturbed forest,
trialists. The debate has also become increasingly
source-sink dynamics may mitigate against the
complex, with genetically manipulated crops, food
detrimental effects of logging. The fine details of
security, climate change, and organic farming added
many of the papers published in this field, taken
to the mix, together with no small number of red
together, suggest that it is important that this idea
herrings, white elephants, and other diversions. The
is put to the test.
 CONSERVATION 275

Figure 9.17 (a) Honduran cloud forest plantations, outside Cusuco National Park; (b) tea plantation in the Western Ghats, India.
276 BATS

challenges are huge—feeding an ever growing studied, the scale, and the landscape. However, re-
human population without destroying what little sults were highly variable and it was concluded that
biodiversity is left in many of our agricultural land- the benefits were likely to be significant in inten-
scapes. There is no doubt that many of the principles sively managed landscapes, but perhaps not in
of organic farming make sense: we are well aware of small scale landscapes with varied land use, where
the damage done by pesticides, herbicides, excessive habitat structure may already favour biodiversity.
fertilizer, the pollution from industrial scale animal Recent work by Gabriel et al. (2010) stresses the need
farms, the removal of hedges, the scaling up of for management on multiple scales, including very
operations, and the homogenization of the land- large ones, if biodiversity is to benefit. This will
scape. We are also well aware of the waste involved require policy changes, legislation, and incentives
in some farming methods and the continued depen- that will facilitate change on appropriate scales—
dence on rapidly diminishing oil reserves. There decisions cannot always be left to individual farm-
have been some significant improvements in con- ers. A complicating factor in many of these studies is
ventional farming too, such as a decrease in the the difficulty of separating the effects of habitat
quantities of chemicals applied to the land and structure and heterogeneity from those of farming
their persistence in the environment. It is important practice—organic farms tend to be found in land-
to remember the gains too: for example a four-fold scapes that are likely to have the heterogeneous,
increase in yield on UK farms since 1945. However small-scale habitat structure that favours biodiver-
this has come at the cost of a decrease in habitat sity (Benton et al. 2003).
diversity and quality and consequently biodiversity Structure was addressed in some studies and
(for example Robinson and Sutherland 2002, Green clearly is important, for example Wickramasinghe
et al. 2005). Organic farming has been shown to lead et al. (2003) found a strong correlation between for-
to an increase in farmland biodiversity, including aging activity and hedgerow height (Fig. 9.18).
bats (for example Wickramasinghe et al. 2003, 2004, Hedgerow width, length, and composition will
Fuller et al. 2005). Since pesticides will reduce insect probably be important too, as will features of farm
abundance and diversity and herbicides the abun- woodland and wetland. The importance of habitat
dance and diversity of the plants the insects rely on, structure (for example Firbank et al. 2008) and land-
this ought to have consequences for the bats. Wick- scape scale factors such as the aggregation of
ramasinghe et al. (2003, 2004) studied 24 matched organic farms (Gabriel et al. 2009) should certainly
pairs of conventional and organic farms in the UK not be underestimated. Views often become very
and found that organic farms had higher insect spe- polarized, entrenched, and simplified on this issue.
cies richness and abundance, higher bat activity (up If we are to find solutions the underlying complexity
61 per cent), higher bat foraging activity (up 84 per must be understood and accepted by scientists, pol-
cent), and higher bat species richness. Insect abun- icy makers, and farmers.
dance was also higher in woodland and water
habitats on organic farms relative to those on con-
Migratory species
ventional farms. Notably, moth diversity was also
higher and moths are an important food source The conservation of migratory species, in addition to
for some rarer bat species (in this case greater and many of the concerns already discussed, has the
lesser horseshoe bats, Rhinolophus ferrumequinum added issues of large scale or multiple habitat use
and R. hipposideros) found only on the organic and the need for international cooperation. A good
farms. The activity of these moth-eating bats was illustration of these problems is the plight of the
correlated with moth abundance. Insects of five of Malayan flying fox, Pteropus vampyrus, discussed
the 18 families important in the diet of bats were earlier in the chapter. Epstein et al. (2009) have
more abundant on organic farms. shown that unsustainable hunting in Malaysia
A recent review of organic farming (Bengtsson threatens bats that spend part of their year in Indo-
et al. 2005) suggested that the average biodiversity nesia and Thailand. Conservation will only be effec-
gain was 30–50 per cent, depending upon the taxa tive with international cooperation. Other studies
 CONSERVATION 277

Figure 9.18 Small-scale agricultural landscape in England preferred by bats, with a high density of species-rich hedgerows with many mature trees.

have also stressed the importance of international compromise: switch the turbines off at low winds
cooperation, for example between countries that speeds and reduce bat mortality for relatively little
share species that migrate between summer habitat cost in lost energy (Arnett et al. 2010). There is
and winter hibernacula (for example Papadatou some evidence to suggest that this strategy works
et al. 2009). Migratory bats may also face new chal- (Behr and Helversen 2005). Another strategy is to
lenges from climate change and the growing num- design turbines that don’t attract bats (assuming
ber of wind turbines, my next topic. that they do) or that are more easily perceived, and
perceived as a danger, by bats. We also need to
Wind turbines design turbines that are less likely to injure or kill
It is unlikely that we will stop building wind tur- bats by barotrauma. We can explore possible de-
bines. Wind-energy generation is a massive and terrents, such as radar (Nicholls and Racey 2007).
growing industry with considerable political and All of these are under investigation and there is no
economic momentum. It also has enormous poten- reason why many of them should not be success-
tial as a sustainable energy source, whatever some ful. Turbines are becoming increasingly efficient, so
of its critics say. Rather than expend effort trying to that more energy is produced by fewer turbines,
ban them should it not be spent trying to make which will have some effect on the rate at which
them safer? As Baerwald and Barclay (2009) stress, new turbines are built. However, the benefits of
we need first to understand why bats and turbines this alone may be small if, as expected, large num-
come together and why bats are killed—we can bers will be built anyway as we attempt to derive
then seek evidence-based solutions to the pro- more and more of our energy from renewable
blems. Given the high mortality among migratory sources and continue to do far too little to use
bats, more informed site selection is clearly a prior- energy more efficiently.
ity: learn more about bat migration routes and Smaller, domestic microturbines may be a greater
avoid them. Research may identify other vulnera- threat to the majority of bat species than large com-
ble species and habitats to avoid. Since bats are mercial turbines, since they are closer to the ground
killed primarily at low wind speeds, when turbines and may be located close to roosts, commuting routes
are least efficient, there is clearly the potential for or foraging sites. Some of the problems to be solved
278 BATS

may be common to both types of turbine, but the whether or not bats use these crossing points in
effects of scale also need to be considered. Long significant numbers and even less that they are
et al. (2009) after studying the interaction between effective in conserving bat populations by increas-
synthesized echolocation pulses and microturbine ing permeability or reducing mortality (Altringham
blades suggested more blades, of greater width, 2008). In the UK millions of pounds are being
would probably make them more apparent to bats. spent on mitigation-related construction that has
The controversy surrounding wind turbines will no objective support for its efficacy, due to a lack
mean that research into their effects, and the effec- of funding, forethought, planning, an apparent lack
tiveness of mitigation measures, will need to be of of commitment to conservation by many develo-
the highest standard if it is to have any impact on the pers, and poorly designed and implemented survey
debate. The issue of assessing mitigation success and monitoring. The situation appears to be similar
will be considered in the next section. I have not at best in some other countries and worse in the
discussed the effects of disturbance during develop- majority.
ment, or the long term loss of habitat, since these As with most other forms of development, the
apply to all forms of development: even wind tur- momentum behind road building is apparently
bines occupy land, lead to new roads, alter site unstoppable. Road construction is an ever growing,
hydrology, and so on. worldwide industry that has a major impact on the
environment, directly and through the increased
Roads volume and speed of traffic that follows. Since this
In contrast to the debate surrounding wind turbines, is not likely to change, at least in the foreseeable
we continue to build roads at an astonishing rate future, all we can do is promote a more rigorous
with hardly a murmur about their effects on bats, or and scientific approach to making roads safer for
indeed most other wildlife, and a general acceptance bats and other wildlife. This requires significant
that mitigation practices are effective. Read almost changes to common practice. Conservation needs
any of the literature produced by those building to be considered as early as possible in the planning
roads, whether government or private sector and process so that adequate pre-construction survey
you will get an optimistic vision of the light and and monitoring can be carried out. Without this,
friendly footprints that modern roads make across assessing the success of mitigation is difficult or
the landscape. Some only just stop short of telling us even impossible. This also allows the best possible
that roads are good for wildlife! This is at best inno- solutions (in conservation, construction, and eco-
cent self-delusion, since there is little or no evidence nomic terms) to be built into the construction
to support most of the claims and increasing evi- programme from the beginning. Experience tells us
dence to the contrary (for example Kerth and Melber that last minute design changes or retrofitting of
2009, Berthinussen and Altringham 2011). We need crossing points are expensive and more likely to be
to assess the real impact of roads from collision ineffective since they usually involve compromises.
mortality, pollution, barrier effects, and habitat loss More rigorous standards in survey and monitoring
and decide which can be mitigated against most must be implemented. If these are not met the statu-
effectively and economically. We also need to exam- tory conservation agencies must be allowed to exer-
ine more rigorously current mitigation practices, cise their power at any stage from planning to
such as the use of lights and barriers to deter bats reporting. Survey and monitoring plans are typi-
from crossing dangerous roads, and the building of cally not designed well enough to address the
crossing points to make roads more permeable and key questions, programmes rarely run for long
safer to cross. Several plausible solutions have been enough to gather sufficient data, and analysis and
proposed and implemented to help bats cross roads, interpretation are unsupported even by basic statis-
with various designs of underpass, overpass, and tics. The reports are consequently of little value.
‘bat gantry’ or ‘bat guidance systems’ (Fig. 9.19). A high proportion of these flawed plans are
Despite their expense and the considerable faith not fully implemented, devaluing them still further.
placed in them, there is little or no evidence to show I have laboured these points because they apply
 CONSERVATION 279

Figure 9.19 Bat guidance system on a UK road, presumed, without evidence, to guide bats across roads, safely above the traffic.

equally well to many other types of development, Walsh and Harris 1996, Vaughan et al. 1997, Gehrt
such as the wind turbines discussed in the previous and Chelsvig 2003, 2004, Avila-Flores and Fenton
section, yet they are critical to the development of 2005). However, some generalist species appear to
evidence-based conservation. Why does practice fall have adapted quite well to urban environments and
so short of the ideal? Lack of funding is invariably can be abundant, particularly in cities that retain
cited as a reason for not pursuing conservation ‘best good tree cover. To look at just one example,
practice’ yet the costs are negligible relative to the Avila-Flores and Fenton (2005) found that natural
overall cost of a typical project—and low even rela- forest fragments outside Mexico City had more bat
tive to the cost of many of the unproven mitigation species than parks in the city, but large parks had
features: a single ‘bat gantry’ can cost over £100,000. higher overall activity. These two habitats contained
It is hard not to conclude that the problems are the highest insect densities and vespertilionid spe-
a lack of commitment to conservation in the devel- cies were largely confined to these areas. However,
opers, interference when political, economic, and several species, notably fast flying molossids, made
conservation goals appear to conflict, and poor com- extensive use of more urban habitats. Low-flying
munication between conservation scientists, devel- vesper bats are probably more affected by noise
opers, and ecological/environmental consultants. and light pollution and the exposed nature of city
There is some hope for the future. The messages of environments, which make them more dangerous
conservation science are being made more accessible and difficult places in which to forage, as well as
to those who need to put them into practice and there having less food. Most other studies report similar
is increasing cross-party collaboration, but there is results.
still a long way to go. On the other hand, there is no doubt that, relative
to other habitats, such as intensively farmed land,
some of our cities have the potential to be havens for
The urban environment
wildlife. If we are to make the most of our cities and
There can be little doubt that urbanization is a bad towns as bat habitat we need to understand how
thing for bats, reducing species diversity and abun- bats use the urban environment. This means ques-
dance relative to ‘natural’ habitats (for example tioning some of our basic assumptions and ‘best
280 BATS

guesses’ to gather evidence about what really does to forest trees in very large numbers for several
benefit bats. For example, Oprea et al. (2009) inves- decades by the Forestry Commission in the UK,
tigated the potential value of tree-lined streets as and similar schemes operate in many other parts
corridors between wooded parks in Vitória, Brazil, of the world. Bat Conservation International has
but found no evidence to suggest that they were run a successful bat house scheme for many years
any more valuable than streets without trees. This and published a regular newsletter on bat houses
finding may not be applicable to other locations and for over 10 years.
urban environments, but does highlight the need for Despite the large number of bat box schemes, few
conservationists to gather evidence rather than bat box studies have reached the scientific literature.
make assumptions, even when these appear quite A recent scheme that has been reported was in-
reasonable and rational. Rhodes et al. (2006) demon- itiated by Flaquer et al. (2006) in the wetlands of
strated the need for managers to look closely at the Ebro in northern Spain. This is a rich habitat,
individual tree roosts in urban Brisbane, Australia, but one with few potential natural roosts, so 69
since not all roosts are equivalent. A large colony of boxes of two designs were erected on trees, build-
white-striped freetail bats, Tadarida australis, were ings and posts. The pipistrelle, Pipistrellus pygmaeus,
shown by social network analysis to be highly colonized the boxes, and during the breeding season
dependent on a single ‘hub’ tree roost among the showed a preference for those on posts and build-
many used. This was the only roost used by all ings, but avoided those on trees. The mean occu-
members of the colony and therefore important for pancy rate was higher than in any previously
maintaining social integrity. reported study (but as the authors note, few studies
give convincing data) and was of course seasonal,
Artificial roosts being higher in summer (93 per cent) than in winter
One resource that urban environments do provide (40 per cent). In the third year of study over half of
in abundance is potential roost sites. Bats have made the boxes were occupied by females with pups dur-
extensive use of a wide range of built structures in ing the summer months. A single box could be
and out of our towns and cities. This has led many occupied by more than 100 bats. Clearly in this
people to experiment with purpose built roosts in an particular habitat boxes are a potentially valuable
attempt to compensate for the loss of natural roosts conservation tool.
or even to encourage bats in their roles as pest A number of studies have looked into the micro-
control agents or pollinators and dispersers. Artifi- climate within boxes in an attempt to improve de-
cial roosts may be no more than wooden boards signs of boxes used for mitigation after roost loss
attached to walls or corrugated metal sheets (for example Lourenço and Palmeirim 2004). Bat
wrapped around trees. At the other end of the boxes have also been used in research to facilitate
scale they can be large and elaborate bat houses the study of various aspects of bat behaviour (for
and artificial hibernacula (Fig. 9.20). example Park et al. 1996, 1998 and the many papers
There is no doubt that many of these are used by by Kerth and colleagues discussed in previous chap-
bats, and some have been spectacularly successful, ters), but there is clearly much more that could be
even attracting large nursery colonies. At the other done.
end of the scale bat boxes are often part of ill- One study has shown how important artificial
considered and misguided ‘mitigation’ plans to roosts can be, not only to bats, but to the habitats
replace roosts lost to development. Their success in which they live. Kelm et al. (2008) used artificial
as conservation tools is extremely variable and has roosts to attract 10 species of frugivorous and nec-
often been hard to quantify due to a lack of moni- tarivorous bats into a forest pasture mosaic in Costa
toring. However, even those that may be of little Rica. Five of these species colonized the roosts on a
direct value to bats can have very positive effects long term basis in both the primary forest and the
because they are part of education programmes deforested areas, leading to a significant increase in
and are a good way to get the public involved. seed rain around the roosts in deforested areas, of
A variety of small ‘bat boxes’ have been attached seed from 69 different plants, primarily early
 CONSERVATION 281

Figure 9.20 Artificial bat roosts. (a) The London Bat House at the London Wetlands Centre, winner of a competition to design a striking but functional
roost. Photo: Teri Pengilley. (b) An early twentieth-century municipal bat roost built in San Antonio Texas (photo from the George Grantham Bain
Collection in the US Library of Congress). The hope was that it would serve the dual purposes of mosquito control and guano production. (c) An artificial
hibernaculum built by the Forestry Commission in the UK. Photo: Charles Critchley. (d) A small wall bat house, a solitary bat choosing to roost outside
it! Photo: Corrie Bruemmer.
Figure 9.20 (Continued )
 CONSERVATION 283

successional species. Large sawdust and concrete published by various organizations. Gates do not
roosts were used, 0.5 m square and almost 2 m necessarily prevent public access, but may help to
high, to simulate hollow tree roosts, with an regulate access, whether for scientific study, explo-
entrance near the base. All were colonized, some ration, recreation, or profit. In many caves bats and
after just three days, but on average in three people are compatible, since bat use may be sea-
weeks. This rapid colonization, even within the for- sonal, or restricted to certain locations within a
est, suggests natural roosts may be limited. Roost cave. Access restrictions may be voluntary or legally
size and composition were comparable to that in enforced and if cave users understand and appreci-
natural roosts. Most of the colonizing bats were ate the need to protect bats, education can take the
of relatively common species and all were generalist place of steel bars.
foragers, taking fruit, pollen, and nectar. Low
seed rain is known to be one of the primary causes
Education
of poor recovery of disturbed forest, so artificial
roosts could prove to be a valuable conservation Education can be the most effective tool in conser-
tool. vation, but the effort needed should not be under-
estimated. To be effective education needs to be well
Cave roosts organized, balanced, and continuous. Audiences
Bats use caves year round in many parts of the have to be persuaded that what you have to say is
world, often forming huge aggregations to breed interesting and important, people forget and new
or to hibernate. Bats are particularly vulnerable dur- audiences are born and grow—so the task is never
ing hibernation: they are torpid and therefore help- complete. Educators also have to keep up with our
less until fully aroused, they often gather in ever expanding knowledge, not just of bats, but also
hibernacula in very large numbers, and they have of many important areas on the edge of bat conser-
little or no opportunity to replenish food reserves if vation. Education also needs to be tailored to partic-
disturbed. Disturbance at hibernacula has led to ular audiences. If we take a broad definition of
some well documented population declines (for education, our audiences will include children,
example Rabinowitz and Tuttle 1980, Wegiel and their teachers and parents, politicians, legislators,
Wegiel 1998). Protecting existing hibernacula is developers, planners, architects, builders, farmers,
therefore of primary importance and considerable indeed everyone, including other conservationists!
effort has gone into this in many countries. Large Each needs a unique approach, but if you can con-
nursery colonies in caves are also vulnerable. Failure vey your enthusiasm and fascination for bats to
to reproduce due to disturbance or roost loss will others the battle is at least half won—and it is a
have severe consequences since bats have low repro- poor biologist who cannot make bats fascinating!
ductive rates. For some audiences we need go no further, but for
One effective way to protect roosts has been to fit others it is just the beginning, and in some cases we
gates at the entrances to caves, abandoned mines, may need to work against prejudice, ignorance, and
and other underground sites (Tuttle 1977, 1979, Tut- real or apparent conflicts of interest. It is essential
tle and Taylor 1998, Sherwin et al. 2002, 2009). How- that we work to overcome these barriers: conserva-
ever, there has been some concern over the effects of tion works best if we work with people rather than
the gates themselves, if not designed carefully, since against them. Conservation by consent, and if nec-
they may alter microclimate (Tuttle and Stevenson essary some compromise, is better than conserva-
1978, Richter et al. 1993) and be unwelcome obsta- tion by conflict.
cles to bats (Pugh and Altringham 2005, Spanjer and
Fenton 2005, Slade and Law 2008). Some early gat- Bats as bioindicators
ing projects therefore led to significant declines of In monitoring the environmental changes that result
the bats they were meant to protect (Tuttle 1977, from our misuse of the planet, we use a number of
MacGregor 1993), but gating has become a reliable taxa as bioindicators—surrogates for all the rest of
conservation tool and designs and guidance are life on the planet. Traditionally insects and birds
284 BATS

have been favoured, since they represent a large participants in their work, and are typically an
proportion of invertebrate and vertebrate diversity enthusiastic mixture of professionals and amateurs.
respectively and can be relatively easy to identify They support a wide range of activities including
and survey. In the case of insects it is often necessary education, practical conservation, survey work, and
to select particular, narrow taxonomic groups for research. Some have been responsible for major leg-
ease of study. Bats are obvious candidates for selec- islation and conservation projects, and the success-
tion as mammalian bioindicators. There are lots of ful promotion of a more informed and publicly
species, they are globally widespread and their eco- acceptable image for bats. Many produce newslet-
logical diversity is high: bats fill many trophic niches ters in addition to booklets and leaflets, reports on
and have diverse habitat requirements and life his- specific projects, conservation plans, guides to sur-
tory patterns. Bats are high in trophic webs and vey techniques, manuals for bat house construction
therefore sensitive to changes in their environment and siting, and so on. They may run a wide range
through changes in their prey. They are also impor- of public and professional courses and educational
tant in ecosystem services: if bats decline we can and professional resources and guides to bats are
expect to see consequences in the environment due also produced.
to their roles as pollinators, seed dispersers, and Those I have direct knowledge of include:
insect predators. Finally, bats are in some respects
 Bat Conservation Trust
easy to survey and identify and their taxonomy is
www.bats.org.uk
relatively stable: all important practical considera-
 Bat Conservation International
tions. They have therefore been used as biodiversity
www.batcon.org
indicators, implicitly and explicitly, in many papers
 Lubee Bat Conservancy
over recent years and there has been some formal
www.batconservancy.org/
recognition of this role by conservation organizations
and governments. Promoting bats as bioindicators is
an excellent way to promote bat conservation. By Other organizations
using bats to gauge the health of the biosphere we IUCN
also learn more about bats and their ecological roles, The World Conservation Union’s Species Survival
helping to reinforce the message that they are worth Commission has a Chiroptera Specialist Group, a
conserving. Jones et al. (2009) have recently made a network of about 100 authorities on the world’s
detailed, formalized and persuasive case for bats, bats. The group acts on particular conservation is-
arguing that they can serve as effective biodiversity, sues and prepares broad policies and strategies for
ecological, and environmental indicators. Their paper bat conservation.
makes any further comment from me unnecessary
and I suggest you read it! Eurobats
The Eurobats secretariat was set up to help Euro-
pean governments fulfil their obligations under The
Bat conservation organizations
Agreement on the Conservation of Populations of Euro-
Organizations have been set up in many countries pean Bats. It promotes information exchange, coordi-
specifically to promote bat conservation. Their activ- nates research and monitoring initiatives, puts
ities vary considerably, but many seek the active forward its own initiatives, and stimulates public
involvement of the public as members and awareness.
 References

When there are large numbers of authors, only the first Altringham, J.D. (2003) British Bats. New Naturalist series,
author’s name is given. pp 218. Harper Collins, London.
Adams, R.A. and Hayes, M.A. (2008) Water availability Altringham, J. D. (2008) Bat ecology and mitigation: Report
and successful lactation by bats as related to climate to the public enquiry on the A350 Westbury Bypass.
change in arid regions of western North America. White Horse Alliance.
J. Anim. Ecol. 77, 1115–1121. Altringham, J.D. and Fenton, M.B. (2003) Sensory ecology
Acharya, L. (1992) Are ears valuable to moths flying and communication. In: Bat Ecology, edited by T. H. Kunz
around lights? Bat Res. News 33, 47. and M.B. Fenton, pp 90–127. University of Chicago Press,
Acharya, L. and Fenton, M.B. (1992) Echolocation Chicago.
behaviour of vespertilionid bats (Lasiurus cinereus and Altringham, J.D. and Young, I.S. (1991) Power output and
Lasiurus borealis) attacking airborne targets, including the frequency of oscillatory work in mammalian dia-
arctiid moths. Can. J. Zool. 70, 1292–1298. phragm muscle: the effects of animal size. J. Exp. Biol.
Aguirre, L.F., Herrel, A., van Damme, R., and Matthysen, 157, 381–389.
E. (2002) Ecomorphological analysis of trophic niche Ammerman, L.K. and Hillis, D.M. (1992) A molecular test
partitioning in a tropical savannah bat community. of bat relationships: monophyly or diphyly? Syst. Biol.
Proc. R. Soc. Lond. B. 269, 1271–1278. 41, 222–232.
Ahlen, I. (1983) The bat fauna of some isolated islands in Andelman, S. J. and Willig, M. R. (2003) Present patterns
Scandinavia. Oikos 41, 352–358. and future prospects for biodiversity in the Western
Ahlén, I., Baage, H.J. and Bach, L. (2009) Behavior of Hemisphere. Ecol. Lett. 6, 818–824.
Scandinavian bats during migration and foraging at Anderson, M.E. and Racey, P.A. (1991) Feeding behaviour
sea. J. Mammal. 90, 1318–1323. of captive brown long-eared bats, Plecotus auritus. Anim.
Aldridge, H.D.J.N. (1986) The flight kinematics of the Behav. 42, 489–493.
greater horseshoe bat, Rhinolophus ferrrumequinum. In: Angell, R.L. (2008) Population ecology and mating systems
Biona Rep. 5, Bat flight-Fledermausflug, edited by W. Nach- of Daubenton’s bat, Myotis daubentonii. PhD thesis, Uni-
tigall, pp 127–138. Gustav Fischer Verlag, Stuttgart. versity of Leeds, UK.
Aldridge, H.D.J.N. and Rautenbach, I.L. (1987) Morpho- Anthony, E.L.P. and Kunz, T.H. (1977) Feeding strategies
logy, echolocation and resource partitioning in insectiv- of the little brown bat, Myotis lucifugus, in southern New
orous bats. J. Anim. Ecol. 56, 763–778. Hampshire. Ecology 58, 775–786.
Alexander, R.McN. and Young, I.S. (1992) Dynamic Appleton, B.R., McKenzie, J.A., and Christidis, L. (2004)
models of breathing. In: S.E.B. Seminar Series 51: Oxygen Molecular systematics and biogeography of the bent-wing
transport in biological systems, edited by S. Egginton and bat complex Miniopterus schreibersii (Kuhl, 1817) (Chirop-
H.F. Ross. Cambridge University Press, Cambridge. tera: Vespertilionidae). Mol. Phylogenet. Evol. 31, 431–439.
Allison, A., Bonaccorso, F., Helgen, K., and James, R. (2008) Araújo, M.B. and Rahbek, C. (2006) How does climate
Pteropus mariannus. In: IUCN 2009. IUCN Red List of change affect biodiversity? Science 313, 1396–1397.
Threatened Species. Version 2009.2. www.iucnredlist.org. Archer, M. (1984) The Australian marsupial radiation. In:
Altenbach, J.S. and Hermanson, J.W. (1987) Bat flight mus- Vertebrate Zoogeography and Evolution in Australasia, edi-
cle function and the scapulohumeral lock. In: Recent ted by M. Archer and G. Clayton, pp 633–808. Hesperian
Advances in the Study of Bats, edited by M.B. Fenton, Press, Carlisle, W. Australia.
P.A. Racey, and J.M.V. Rayner, pp 100–118. Cambridge Arita, H.T. (1996) The conservation of cave-roosting bats in
University Press, Cambridge. Yucatan, Mexico. Biol. Conserv. 76, 177–185.

285
286 REFERENCES

Arita, H.T. (1997) Species composition and morphological cues for sound localization in azimuth and elevation.
structure of the bat fauna of Yucatan, Mexico. J. Anim. J. Acoust. Soc. Am. 116, 3594–3605.
Ecol. 66, 83–97. Baage, H.J. (1987) The Scandinavian bat fauna: adaptive
Arlettaz, R. (1999) Habitat selection as a major resource wing morphology and free flight in the field. In: Recent
partitioning mechanism between the two sympatric sib- Advances in the Study of Bats, edited by M.B. Fenton,
ling bat species Myotis myotis and Myotis blythii. J. Anim. P.A. Racey, and J.M.V. Rayner, pp 57–74. Cambridge
Ecol. 68, 460–471. University Press, Cambridge.
Arlettaz, R., Godat, S., and Meyer, H. (2000) Competition Baerwald, E. F. and Barclay, R. M. R. (2009) Geographic
for food by expanding pipistrelle bat populations variation in activity and fatality of migratory bats at
(Pipistrellus pipistrellus) might contribute to the decline wind energy facilities. J. Mammal. 90, 1341–1349.
of lesser horseshoe bats (Rhinolophus hipposideros). Biol. Bailey, W.B. (1991) Avoiding predation. In: Acoustic
Conserv. 93, 55–60. Behaviour of Insects: An Evolutionary Perspective, Chapter
Arlettaz, R., Jones, G., and Racey, P.A. (2001) Effect of 10. Chapman and Hall, London.
acoustic clutter on prey detection by bats. Nature 414, Bailey, W.J., Slightom, J.L. and Goodman, M. (1992) Rejec-
742–745. tion of the ‘flying primate hypothesis’ by phylogenetic
Arnett, E. B. (2006) A preliminary evaluation on the use of evidence from the -globin gene. Science 256, 86–89.
dogs to recover bat fatalities at wind energy facilities. Baker, H.G. (1973) Evolutionary relationships between
Wildlife Soc. B. 34, 1440–1445. flowering plants and animals in American and African
Arnett, E.B., Huso, M.M.P., Hayes, J.P., and Schirmacher, forests. In: Tropical Forest Ecosystems in Africa and South
M. (2010) Effectiveness of changing wind turbine cut-in speed America: A Comparative Review, edited by B.J. Meggers,
to reduce bat fatalities at wind facilities. A final report E.S. Ayensu, and W.D. Duckworth, pp 145–159. Smith-
submitted to the Bats and Wind Energy Cooperative. sonian Institute Press, Washington.
Bat Conservation International. Austin, Texas, USA. Baker, R.J., Novacek, M.J., and Simmons, N.B. (1991) On
Arrhenius, O. (1921) Species and area. J. Ecol. 9, 95–99. the monophyly of bats. Syst. Zool. 40, 216–231.
Au, W. (1993) The Sonar of Dolphins. Springer-Verlag, Balasingh, J., Koilraj, J., and Kunz, T.H. (1995) Tent
Berlin. construction by the short-nosed fruit bat, Cynopterus
Audet, D. and Fenton, M.B. (1988) Heterothermy and the sphinx (Chiroptera: Pteropodidae), in southern India.
use of torpor by the bat Eptesicus fuscus (Chiroptera: Ethology 100, 210–219.
Vespertilionidae): a field study. Physiol. Zool. 61, Balcombe, J.P. (1990) Vocal recognition of pups by mother
197–204. Mexican free-tailed bats, Tadarida brasiliensis mexicana.
Audet, D., Krull, D., Marimuthu, G., Sumithran, S. and Anim. Behav. 39, 960–966.
Bala Singh, J. (1991) Foraging behaviour of the Indian Balcombe, J.P. and McCracken, G.F. (1992) Vocal recogni-
False Vampire Bat, Megaderma lyra (Chiroptera: Mega- tion in Mexican free-tailed bats: do pups recognise
dermatidae). Biotropica 23, 63–67. mothers? Anim. Behav. 43, 79–88.
Audet, D. and Thomas, D.W. (1997) Facultative hypother- Balkenhol, N. and Waits, L.P. (2009) Molecular road ecol-
mia as a thermoregulatory strategy in the phyllostomid ogy: exploring the potential of genetics for investigating
bats Carollia perspicillata and Sturnira lilium. J. Comp. transportation impacts on wildlife. Mol. Ecol. 18,
Physiol. B 167, 146–152. 4151–4164.
Avery, M.I. (1985) Winter activity of pipistrelle bats. Balmford, A. et al. (2005) The Convention on Biological
J. Anim. Ecol. 54, 721–738. Diversity’s 2010 target. Science 307, 212–213.
Ávila-Flores, R. and Fenton, M.B. (2005) Use of spatial Banack, S.A. (1998) Diet selection and resource use by
features by foraging insectivorous bats in a large urban flying foxes. Ecology 79, 1949–1967.
landscape. J. Mammal. 86, 1193–1204. Barber, J.R., Chadwell, B.A., Garrett, N., Schmidt-French,
Avise, J.C. (2004) Molecular Markers, Natural History B., and Conner, W.E. (2009) Naïve bats discriminate
and Evolution. Sinauer Associates, Sunderland, arctiid moth warning sounds but generalize their apose-
Massachusetts. matic meaning. J. Exp. Biol. 212, 2141–2148.
Avise, J.C. (2006) Evolutionary Pathways in Nature: A Phylo- Barber, J.R. and Conner, W.E. (2007) Acoustic mimicry in a
genetic Approach. Cambridge University Press, predator–prey interaction. PNAS 104, 9331–9334.
Cambridge. Barber, J.R., Crooks, K.R., and Fristrup, K.M. (2010) The
Aytekin, M., Grassi, E., Sahota, M. and Moss, C.F. (2004) costs of chronic noise exposure for terrestrial organisms.
The bat head-related transfer function reveals binaural TREE 25, 180–189.
 REFERENCES 287

Barbour, R.W. and Davis, W.H. (1969) Bats of America. Beasley, L.J., Pelz, K.M., and Zucker, I. (1984a) Circannual
University Press of Kentucky, Lexington. rhythms of body weight in pallid bats. Am. J. Physiol. 246,
Barclay, R.M.R. (1982) Interindividual use of echolocation R955–R958.
calls: eavesdropping by bats. Behav. Ecol. Sociobiol. 10, Beasley, L.J., Smale, L., and Smith, E.R. (1984b) Melatonin
271–275. influences the reproductive physiology of male pallid
Barclay, R.M.R. (1985) Long- versus short-range foraging bats. Biol. Reprod. 30, 300–305.
strategies of hoary (Lasiurus cinereus) and silver-haired Beckerman, A.P., Boots, M., and Gaston, K.J. (2007) Urban
(Lasionycteris noctivagans) bats and the consequences for bird declines and the fear of cats. Anim. Conserv. 320,
prey selection. Can. J. Zool. 63, 2507–2515. 320–325.
Barclay, R.M.R. (1986) Foraging strategies of silver haired Beer, J.R. and Richards, A.G. (1956) Hibernation of the big
(Lasionycteris noctivagans) and hoary (Lasiurus cinereus) brown bat. J. Mammal. 37, 31–41.
bats. Myotis 23–24, 161–166. Begon, M., Harper, J.L,. and Townsend, C.R. (2006)
Barclay, R.M.R. (1991) Population structure of temperate Ecology: Individuals, Populations and Communities. 4th
zone insectivorous bats in relation to foraging behaviour edition. Blackwell, Oxford.
and energy demand. J. Anim. Ecol. 60, 165–178. Behr, O. and von Helversen, O. (2005) Gutachten zur
Barclay, R.M.R. (1995) Constraints on reproduction by Beeinträchtigung im freien Luftraum jagender und
bats: energy or calcium? In: Ecology, Evolution and ziehender Fledermäuse durch bestehende Windkraftan-
Behaviour of Bats, edited by P.A. Racey and S.M. Swift, lagen. Wirkungskontrolle zum Windpark ‘Roßkopf’
pp 245–258. Symp. Zool. Soc. Lond. 67. Oxford University (Freiburg i. Br.). Friedrich -Alexander-Universität Erlan-
Press, Oxford. gen-Nürnberg, Institut für Zoologie II: 1–42.
Barclay, R.M.R. and Brigham, R.M. (1991) Prey detection, Behr, O. and von Helversen, O. (2004) Bat serenades—
dietary niche breadth and body size in bats: why are aerial complex courtship songs of the sac-winged bat (Saccop-
insectivorous bats so small? Amer. Nat. 137, 693–703. teryx bilineata). Behav. Ecol. Sociobiol. 56, 106–115.
Barclay, R.M.R. and Harder, L.D. (2003) Life histories Behr, O., Helversen, O., Heckel, G., Nagy, M., Voigt, C.C.,
of bats: life in the slow lane. In: Bat Ecology, edited by and Mayer, F. (2006) Territorial songs indicate
T.H. Kunz and M.B. Fenton, pp 209–253. University of male quality in the sac-winged bat Saccopteryx bilineata
Chicago Press, Chicago. (Chiroptera: Emballonuridae). Behav. Ecol. 17, 810–817.
Bartholomew, G.A., Leitner, P., and Nelson, J.E. (1964) Bell, G.P. (1982) Behavioural and ecological aspects of
Body temperature, oxygen consumption and heart rate gleaning by a desert insectivorous bat, Antrozous pallidus
in three species of flying foxes. Physiol. Zool. 37, 179–198. (Chiroptera: Vespertilionidae). Behav. Ecol. Sociobiol. 10,
Bartholomew, G.A., Dawson, W.R., and Lasiewski, 217–223.
R.C. (1970) Thermoregulation and heterothermy in Bell, G.P. (1985) The sensory basis of prey location by
some of the smaller flying foxes (Megachiroptera) of the California leaf-nosed bat, Macrotus californicus
New Guinea. Zeitsch. fur Vergl. Physiol. 70, 196–207. (Chiroptera: Phyllostomidae). Behav. Ecol. Sociobiol. 16,
Bates, D.L. and Fenton, M.B. (1990) Aposematism or 343–347.
startle? Predators learn their responses to prey. Can. Bell, G.P. and Fenton, M.B. (1984) The use of Doppler-
J. Zool. Lond. 68, 49–52. shifted echoes as a flutter detection and clutter rejection
Bates, P.J.J., Nwe, T., Pearch, M.J., Swe, K.M., Hla Bu, S.S., system: the echolocation and feeding behaviour of Hip-
and Tun, T. (2000) A review of bat research in Myanmar posideros ruber. Behav. Ecol. Sociobiol. 15, 109–114.
(Burma) and results of a recent survey. Acta Chiropt. 2, Belwood, J.J. and Fenton, M.B. (1976) Variation in the diet
53–82. of Myotis lucifugus (Chiroptera: Vespertilionidae). Can.
Bateman, G.C. and Vaughan, T.A. (1974) Night activities of J. Zool. Lond. 54, 1674–1678.
mormoopid bats. J. Mammal. 55, 45–65. Bengtsson, J., Ahnström, J., and Weibull, A-C. (2005)
Bay, F.A. (1978) Light control of the circadian activity The effects of organic agriculture on biodiversity and
rhythm in mouse-eared bats (Myotis myotis). J. Interdis- abundance: a meta-analysis. J. Appl. Ecol. 42, 261–269.
cipl. Cycle Res. 9, 195–209. Bennett, M.B. (1992) Structural modifications involved
Bat Conservation Trust. (2009) National Bat Monitoring in the fore- and hind limb grip of some flying foxes
Programme. www.bats.org.uk/pages/nbmp.html (Chiroptera: Pteropodidae) J. Zool. Lond. 229, 237–248.
[accessed 28 February 2011]. Benton, T.G., Vickery, J.A., and Wilson, J.D. (2003) Farm-
Beasley, L.J. (1986) Seasonal cycles of pallid bats (Antrozous land biodiversity: is habitat heterogeneity the key? TREE
pallidus): proximate factors. Myotis 23–24, 115–123. 18, 182–188.
288 REFERENCES

Berger, M. and Hart, J.S. (1974) Physiology and energetics Boughman, J.W. (2006) Selection on social traits in greater
of flight. In: Avian Biology, Vol. 4, edited by D.S. Farner spear-nosed bats, Phyllostomus hastatus. Behav. Ecol.
and J.R. King. Academic Press, London. Sociobiol. 60, 766–777.
Berthinussen, A. and Altringham, J.D. (2011) The effects of Boyles, J.G., Dunbar, M.B., Storm, J.J., and Brack, V. (2007)
major roads on bat activity and diversity. J. Appl. Ecol. Energy availability influences microclimate selection of
In press. hibernating bats. J. Exp. Biol. 210, 4345–4350.
Betke, M., Hirsh, D.E., Makris, N.C., McCracken, G.F., Brack, V. and Twente, J.W. (1985) The duration of the
Procopio, M., Hristov, N.I., Tang, S., Bagchi, A., period of hibernation of three species of vespertilionid
Reichard, J.D., Horn, J.W., Crampton, S., Cleveland, bats. I. Field studies. Can J. Zool. 63, 2952–2954.
C.J., and Kunz, T.H. (2008) Thermal imaging reveals Bradbury, J.W. (1977a) Social organisation and communi-
significantly smaller Brazilian free-tailed bat colonies cation. In: Biology of Bats, Vol. 3, edited by W.A. Wimsatt,
than previously estimated. J. Mammal. 89, 18–24. pp 1–72. Academic Press, New York.
Bhagwat, S.A., Willis, K.J., Birks, H.J.B., and Whittaker, R.J. Bradbury, J.W. (1977b) Lek mating behaviour in the
(2008) Agroforestry: a refuge for tropical biodiversity? hammer-headed bat. Z. Tierpsychol. 45, 225–255.
TREE 23, 261–267. Bradbury, J.W. and Emmons, L.H. (1974) Social organi-
Bingman, V.P. and Cheng, K. (2005) Mechanisms of sation of some Trinidad bats. I. Emballonuridae. Z. Tier-
animal global navigation: comparative perspectives pyschol. 36, 137–183.
and enduring challenges. Ethol. Ecol. Evol. 17, 295–318. Bradbury, J.W. and Vehrencamp, S.L. (1976a) Social
Bisson, I-A., Safi, K., and Holland, R.A. (2009) Evidence for organisation and foraging in emballonurid bats. I. Field
repeated independent evolution of migration in the studies. Behav. Ecol. Sociobiol. 1, 337–381.
largest family of bats. PloS One. 4, e7504. Bradbury, J.W. and Vehrencamp, S.L. (1976b) Social
Blackburn, T.M., Cassey, P., Duncan, R.P., Evans, K.L., and organisation and foraging in emballonurid bats. II.
Gaston, K.J. (2004) Avian extinction and mammalian A model for the determination of group size. Behav.
introductions on oceanic islands. Science 305, 1955–1958. Ecol. Sociobiol. 1, 383–404.
Blackburn, T.M. and Gaston, K.J. (2003) Macroecology: Bradbury, J.W. and Vehrencamp, S.L. (1977a) Social
Concepts and Consequences. Symp. Brit. Ecol. Soc. 43. organisation and foraging in emballonurid bats. III.
Cambridge University Press, Cambridge. Mating systems. Behav. Ecol. Sociobiol. 2, 1–17.
Blake, D., Hutson, A.M., Racey, P.A., Rydell, J., and Bradbury, J.W. and Vehrencamp, S.L. (1977b) Social organi-
Speakman, J.R. (1994) Use of lamplit roads by foraging sation and foraging in emballonurid bats. IV. Parental
bats in southern England. J. Zool. Lond. 234, 453–62. investment patterns. Behav. Ecol. Sociobiol. 2, 19–29.
Blehert, D.S., et al. (2009) Bat white-nose syndrome: an Brass, D.A. (1994) Rabies in Bats, Natural History, and Public
emerging fungal pathogen? Science 323, 227–227. Health Implications. Livia Press, Connecticut.
Bloedel, P. (1955) Observations on the life histories of Brigham, R.M. (1987) The significance of winter activity by
Panama bats. J. Mammal. 36, 232–235. the big brown bat (Eptesicus fuscus): the influence of
Bohn, K.M., Schmidt-French, B., Schwartz, C., Smotherman, energy reserves. Can. J. Zool. 65, 1240–1242.
M., and Pollak, G.D. (2009) Versatility and stereotypy Brigham, R.M. (1991) Flexibility in foraging and roosting
of free-tailed bat songs. PLoS One 4, e6746. behaviour by the big brown bat (Eptesicus fuscus) Can.
Bonaccorso, F.J. (1979) Foraging and reproductive ecology J. Zool. Lond. 69, 117–121.
in a Panamanian bat community. Bull. Fla. State Mus. Britzke, E.R., Loeb, S.C., Hobson, K.A., Romanek, C.S., and
Biol. Ser. 24, 359–408. Vonhof, M.J. (2009) Using hydrogen isotopes to assign
Bonaccorso, F.J., Glanz, W.E., and Sandford, C.M. (1980) origins of bats in the eastern United States. J. Mammal. 90,
Feeding assemblages of mammals at fruiting Dipteryx 743–751.
panamensis (Papilionaceae) trees in Panama: seed predation, Brooke, A.P. (1990) Tent selection, roosting ecology and
dispersal, and parasitism. Rev. Biol. Trop. 28, 61–72. social organisation of the tent-making bat, Ectophylla
Boonman, M. (2000) Roost selection by noctules (Nyctalus alba, in Costa Rica. J. Zool. Lond. 221, 11–19.
noctula) and Daubenton’s bats (Myotis daubentonii). Brooke, A.P. (1994) Diet of the fishing bat, Noctilio leporinus
J. Zool. Lond. 251, 385–389. (Chiroptera: Noctilionidae). J. Mammal. 75, 212–218.
Boonman, A. and Jones, G. (2002) Intensity control during Brosset, A. (1962) The bats of central and western India.
target approach in echolocating bats; stereotypical Part I. J. Bombay Nat. Hist. Soc. 59, 707–746.
sensori-motor behaviour in Daubenton’s bats, Myotis Brosset, A. (1966) La biologie des chiroptères. Masson, Paris.
daubentonii. J. Exp. Biol. 205, 2865–2874. Brosset, A. (1976) Social organisation of the African bat,
Myotis bocagei. Z. Tierpsychol. 42, 50–56.
 REFERENCES 289

Brown, J.H. (1995) Macroecology. University of Chicago Carpenter, R.E. (1986) Flight physiology of intermediate
Press, Chicago. sized fruit bats (Pteropodidae). J. Exp. Biol. 120, 79–103.
Bruns, V. (1970) Peripheral auditory tuning for fine fre- Carr, L.W. and Fahrig, L. (2001) Impact of road traffic on
quency analysis by the CF-FM bat, Rhinolophus ferrume- two amphibian species of different vagility. Conserv. Biol.
quinum. II. Frequency mapping in the cochlea. J. Comp. 15, 1071–1078.
Physiol. A. 106, 87–97. Carstens, B.C., Sullivan, J., Davalos, L.M., Larsen, P.A.,
Buchler, E.R. (1980) The development of flight, foraging and Pedersen, S.C. (2004) Exploring population genetic
and echolocation in the little brown bat (Myotis lucifugus). structure in three species of Lesser Antillean bats. Mol.
Behav. Ecol. Sociobiol. 6, 211–218. Ecol. 13, 2557–2566.
Buchler, E.R. and Childs, S.B. (1981) Orientation to distant Carvajal, A. and Adler, G.H. (2005) Biogeography of
sounds by foraging big brown bats (Eptesicus fuscus). mammals on tropical Pacific islands. J. Biogeogr. 32,
Anim. Behav. 29, 428–432. 1561–1569.
Buchler, E.R. and Childs, S.B. (1982) Use of post-sunset Castella, V., Ruedi, M., and Excoffier, L. (2001) Contrasted
glow as an orientation cue by the big brown bat (Eptesicus patterns of mitochondrial and nuclear structure among
fuscus). J. Mammal. 63, 243–247. nursery colonies of the bat Myotis myotis. J. Evol. Biol. 14,
Burland, T.M., Barratt, E.M., Beaumont, M.A., and Racey, 708–720.
P.A. (1999) Population genetic structure and gene flow in Castro-Arellano, I., Presley, S.J., Saldanha, L.N., Willig, M.G.,
a gleaning bat, Plecotus auritus. Proc. R. Soc. Lond. B. 266, and Wunderle, J.M. Jr. (2007) Effects of reduced-impact
975–980. logging on bat biodiversity in terra firme forest of lowland
Burland, T.M. and Worthington Wilmer, J. (2001) Seeing Amazonia. Biol. Conserv. 138, 269–285.
in the dark: molecular approaches to the study of bat Ceballos, G. and Ehrlich, P.R. (2002) Mammal population
populations. Biol. Rev. 76, 389–409. losses and the extinction crisis. Science 296, 904–907.
Burland, T.M., Barratt, E.M., Nichols, R.A., and Racey, P.A. Cervený, J. and Bürger, P. (1989) Density and structure of
(2001) Mating patterns, relatedness and the basis of natal the bat community occupying an old park at Zihobce
philopatry in the brown long-eared bat, Plecotus auritus. (Czechoslovakia). In: European Bat Research 1987, edited
Mol. Ecol. 10, 1309–1321. by V. Hanák, I. Horácek, and J. Gaisler, pp 475–488.
Camarsa, G., Gardner, S., Jones, W., Eldridge, J., Hudson, Charles University Press, Praha.
T., Thorpe, E., and O’Hara, E. (2010) Life among the Olives. Chape, S., Harrison, J., Spalding, M., and Lysenko, I. (2005)
European Commission, Luxembourg. Measuring the extent and effectiveness of protected
Campbell, P., Schneider, C.J., Adnan, A.M., Zubaid, A., areas as an indicator for meeting global biodiversity
and Kunz, T.H. (2006) Comparative population targets. Philos. Trans. R. Soc. Lond. B Biol. Sci. 360,
structure of Cynopterus fruit bats in peninsular Malaysia 443–455.
and southern Thailand. Mol. Ecol. 15, 29–47. Charles-Dominique, P. (1993) Tent-use by the bat
Campbell, P., Schneider, C.J., Zubaid, A., Adnan, A.M., Rhinophylla pumilio (Phyllostomidae: Carolliinae) in
and Kunz, T.H. (2007) Morphological and ecological cor- French Guiana. Biotropica 25, 111–116.
relates of coexistence in malaysian fruit bats (Chiroptera: Chazdon, R. L. (1998) Tropical Forests—Log ‘em or leave
Pteropodidae). J. Mammal. 88, 105–118. ‘em? Science 281, 1295–1296.
Canals, M., Atala, C., Olivares, R., Guajardo, F., Figueroa, Childs, S.B. and Buchler, E.R. (1981) Perception of simulated
D.P., Sabat, P., and Rosenmann, M. (2005) Functional stars by Eptesicus fuscus (Vespertilionidae) - a potential
and structural optimization of the respiratory system of navigational mechanism. Anim. Behav. 29, 1028–1035.
the bat Tadarida brasiliensis (Chiroptera, Molossidae): Chiu, C. and Moss, C.F. (2007) The role of the external
does airway geometry matter? J. Exp. Biol. 208, ear in vertical sound localization in the free flying bat,
3987–3995. Eptesicus fuscus. J. Acoust. Soc. Am. 121, 2227–2235.
Caple, G., Balda, R.P., and Willis, W.R. (1983) The physics Choe, J.C. (1994) Ingenious design of tent roosts by Peter’s
of leaping animals and the evolution of preflight. Am. tent-making bat, Uroderma bilobatum (Chiroptera:
Nat. 121, 455–467. Phyllostomidae). J. Nat. Hist. 28, 731–737.
Cardillo, M., Mace, G.M., Gittleman, J.L, Jones, K.E., Clarke, F.M., Pio, D.V., and Racey, P.A. (2005a). A compar-
Bielby, J., and Purvis, A. (2008) The predictability of ison of logging systems and bat diversity in the neotrop-
extinction: biological and external correlates of decline ics. Conserv. Biol. 19, 1194–1204.
in mammals. Proc. R. Soc. B. 275, 1441–1448. Clarke, F.M., Rostant, L.V., and Racey, P.A. (2005b) Life
Carpenter, R.E. (1985) Flight Physiology of flying foxes. after logging: post-logging recovery of a neotropical bat
J. Exp. Biol. 114, 619–647. community. J. Appl. Ecol. 42, 409–420.
290 REFERENCES

Clavero, M., and Garcia-Berthou, E. (2005) Invasive species Dal Piaz, G. (1937) I mammiferi dell’Oligocene Veneto. Arch-
are a leading cause of animal extinctions. TREE 20, 110–110. aeopteropus transiens. Mem 1st Geol. Univ. Padova 11, 1–8.
Cleveland, C.J., Betke, M., Federico, P., Frank, J.D., Hallam, Daniel, M.J. (1990) Order Chiroptera. In: The Handbook of
T.G., Horn, J., López Jr, J.D., McCracken, G.F., Medellin, New Zealand Mammals, edited by C.M. King, pp 114–137.
R.A., Moreno-Valdez, A., Sansone, C.G., Westbrook, Oxford University Press, Auckland.
J.K., and Kunz, T.H. (2006) Economic value of the pest Darwin, C. (1859) On the Origin of Species. John Murray,
control service provided by Brazilian free-tailed bats in London.
south-central Texas. Front. Ecol. Environ. 4, 238–243. Dashback, N. (1990) After the hurricane. Bats 8, 14–15.
Clobert, J., Danchin, E., Dhondt, A.A., and Nichols, J.D. Daszak, P. (2010) Bats, in black and white. Science 329,
(2001) Dispersal. Oxford University Press, Oxford. 634–635.
Clutton-Brock, T.H. (1989) Mammalian mating systems. Davies, T.J., Fritz, S.A., Grenyer, R., Orme, C.D.L., Bielby,
Proc. R. Soc. Lond. B. 236, 339–372. J., Bininda-Emonds, O.R.P, Cardillo, M., Jones, K.E.,
Coles, R.B., Guppy, A., Anderson, M.E., and Schlegel, P. Gittleman, J.L., Mace, G.M., and Purvis, A. (2008) Phylo-
(1989) Frequency sensitivity and directional hearing in genetic trees and the future of mammalian biodiversity.
the gleaning bat, Plecotus auritus (Linnaeus 1758). PNAS 105, 11556–11563.
J. Comp. Physiol. A. 165, 269–280. Davies, Z.G. and Pullin, A.S. (2007) Are hedgerows
Corlatti, L., Hackländer, K., and Frey-Roos, F. (2009) effective corridors between fragments of woodland
Ability of wildlife overpasses to provide connectivity habitat? An evidence-based approach. Landscape Ecol.
and prevent genetic isolation. Conserv. Biol. 23, 548–556. 22, 333–351.
Cosson, J.F., Pons, J.M., and Masson, D. (1999) Effects of Davidson, S.M. and Wilkinson, G.S. (2004) Function of
forest fragmentation on frugivorous and nectarivorous male song in the greater white-lined bat, Saccopteryx
bats in French Guiana. J. Trop. Ecol. 15, 515–534. bilineata. Anim. Behav. 67, 883–891.
Courts, S.E. (1998) Dietary strategies of Old World fruit Davis, R.B. and Cockrum, E.L. (1962) Repeated homing
bats (Megachiroptera, Pteropodidae): how do they exhibited by a female pallid bat. Science 137, 341–342.
obtain sufficient protein? Mammal Rev. 28, 185–194. Davis, R.B., Herreid, C.F., and Short, H.L. (1962) Mexican
Cox, P.A., Elmqvist, T., Pierson, E.D., and Rainey, W.E. free-tailed bats in Texas. Ecol. Monogr. 32, 311–346.
(1991) Flying foxes as strong interactors in south Pacific Davy, C.M., Russo, D., and Fenton, M.B. (2007) Use of
island ecosystems: a conservation hypothesis. Conserv. native woodlands and traditional olive groves by forag-
Biol. 5, 448–454. ing bats on a Mediterranean island: consequences for
Crichton, E.G. and Krutzsch, P.H. (2000) Reproductive conservation. J. Zool. Lond. 273, 397–405.
Biology of Bats. Academic Press, London. Dawkins, H. C. and Philip, M. (1998) Tropical Moist Forest
Crome, F.H.J. and Richards, G.C. (1988) Bats and gaps: Silviculture and Management: A History of Success and
microchiropteran community structure in a Queensland Failure. CAB International, Oxfordshire, United
rain forest. Ecology 69, 1960–1969. Kingdom.
Cryan, P.M. (2003) Seasonal distribution of migratory Dechmann, D.K.N., Kalko, E.K.V., and Kerth, G. (2007) All-
tree bats (Lasiurus and Lasionycteris) in North America. offspring dispersal in a tropical mammal with resource
J. Mammal. 84, 579–593. defense polygyny. Behav. Ecol. Sociobiol. 61, 1219–1228.
Cryan, P.M. and Barclay, R.M.R. (2009) Causes of bat Dechmann, D.K.N., Kalko, E.K.V., König, B., and Kerth, G.
fatalities at wind turbines: hypotheses and predictions. (2005) Mating system of a Neotropical roost-making bat:
J. Mammal. 90, 1330–1340. the white-throated, round-eared bat, Lophostoma silvico-
Cryan, P.M., Bogan, M.A., and Altenbach, J.S. (2000) Effect lum (Chiroptera: Phyllostomidae) Behav. Ecol. Sociobiol.
of elevation on distribution of female bats in the Black 58, 316–325.
Hills, South Dakota. J. Mammal. 81, 719–725. Dengler, J. (2009) Which function describes the species-
Cryan, P.M., Bogan, M.A., Rye, R.O., Landis, G.P., and area relationship best? A review and empirical evalua-
Kester, C.L. (2004) Stable hydrogen isotope analysis of tion. J. Biogeogr. 36, 728–744.
bat hair as evidence for seasonal molt and long-distance Develey, P.F. and Stouffer, P.C. (2001) Effects of roads on
migration. J. Mammal. 85, 995–1001. movements of understorey birds in mixed-species flocks
Daan, S. (1973) Activity during natural hibernation in three in central Amazonian Brazil. Conserv. Biol. 15, 1416–1422.
species of vespertilionid bats. Neth. J. Zool. 23, 1–71. Dickinson, M.B., Dickinson, J.C., and Putz, F.E. (1996) Nat-
Daan, S., Barnes, B.M., and Strijkstra, A.M. (1991) Warming ural forest management as a conservation tool in the
up for sleep – ground squirrels sleep during arousals tropics: divergent views on possibilities and alternatives.
from hibernation. Neurosci. Lett. 128, 265–268. Commonw. Forest. Rev. 75, 309–315.
 REFERENCES 291

Dolbeer, R.R., Fielder, L.R., and Rasheed, H. (1988) Man- Eick, G.N., Jacobs, D.S., and Matthee, C.A. (2005) A nuclear
agement of fruit bat and rat populations in the Maldive DNA phylogenetic perspective on the evolution of echo-
Islands, Indian Ocean. In: Proceedings of the Thirteenth location and historical biogeography of extant bats
Vertebrate Pest Conference, edited by A.C. Crabb and (Chiroptera). Mol. Biol. Evol. 22, 1869–1886.
R.E. Marsh, pp 112–118. University of California, Davis. Eklof, J. and Jones, G. (2003) Use of vision in prey detection
Dormann, C.F., McPherson, J.M., Araújo, M.B., Bivand, R., by brown long-eared bats Plecotus auritus. Anim. Behav.
Bolliger, J., Carl, G., Davies, R.G., Hirzel, A., Jetz, W., 66, 949–953.
Kissling, W.D., Kühn, I., Ohlemüller, R., Peres-Neto, Emde, G.V.D. and Schnitzler, H-U. (1990) Classification of
P.R., Reineking, B., Schröder, B., Schurr, F.M., and insects by echolocating greater horseshoe bats. J. Comp.
Wilson, R. (2007) Methods to account for spatial auto- Physiol. A. 167, 423–430.
correlation in the analysis of species distributional data: Encarnação, J.A., Kierdorf, U., and Wolters, V. (2007) Do
a review. Ecography 30, 609–628. mating roosts of Daubenton’s bats (Myotis daubentonii)
Duchamp, J.E. and Swihart, R.K. (2008) Shifts in bat com- exist at summer sites? Myotis 43, 31–39.
munity structure related to evolved traits and features of Entwistle, A.C., Racey, P.A., and Speakman, J.R. (1996)
human-altered landscapes. Landscape Ecol. 23, 849–860. Habitat exploitation by a gleaning bat, Plecotus auritus.
Dumont, E.R. (2003) Bats and fruit: an ecomorphological Phil. Trans. Roy. Soc. Lond. B. 351, 921–931.
approach. In: Bat Ecology, edited by T.H. Kunz and M.B. Entwistle, A.C., Racey, P.A., and Speakman, J.R. (2000)
Fenton, pp 398–429. University of Chicago Press, Chicago. Social and population structure of a gleaning bat, Pleco-
Dunbar, M.B. and Tomasi, T.E. (2006) Arousal patterns, tus auritus. J. Zool. Lond. 252, 11–17.
metabolic rate, and an energy budget for eastern red Epstein, J.H., Oliva, K.J., Pulliam, J.R.C., Smith, C., Wes-
bats (Lasiurus borealis) in winter. J. Mammal. 87, 1096–1102. trum, J., Hughes, T., Dobson, A.P., Zubaid, A., Rahman,
Dürr, T. and Bach, L. (2004) Bat deaths and wind turbines: S.A., Basir, M.M., Field, H.E., and Daszak, P. (2009)
a review of current knowledge, and of the information Pteropus vampyrus, a hunted migratory species with
available in the database for Germany. Bremer Beiträge a multinational home-range and a need for regional
für Naturkunde und Naturschutz. 7, 253–264. management. J. Appl. Ecol. 46, 991–1002.
Dwyer, P.D. (1963) The breeding biology of Miniopteris schrei- Erickson, J. L. and West, S.D. (2003) Associations of bats
bersii blepotis (Temminck) (Chiroptera) in north-eastern with local structure and landscape features of forested
New South Wales. Aust. J. Zool. 11, 219–240. stands in western Oregon and Washington. Biol. Conserv.
Dwyer, P.D. (1966) The population pattern of Miniopterus 109, 95–102.
schreibersii (Chiroptera) in north-eastern New South Estrada, A. and Coates-Estrada, R. (2001) Species com-
Wales. Aust. J. Zool. 14, 1073–1137. position and reproductive phenology of bats in a tropical
Dwyer, P.D. (1968) The little bent-winged bat - evolution in landscape at Los Tuxtlas, Mexico. J. Trop. Ecol. 17,
progress. Aust. Nat. Hist. 1968, 55–58. 627–646.
Dwyer, P.D. (1970a) Latitude and breeding season in a Estrada, A., Coates-Estrada, R., and Meritt, D. (1993) Bat
polyestrous species of Myotis. J. Mammal. 51, 405–410. species richness and abundance in tropical rain forest
Dwyer, P.D. (1970b) Social organisation in the bat Myotis fragments and in agricultural habitats at Los Tuxtlas,
adversus. Science 168, 1006–1008. Mexico. Ecography 16, 309–318.
Dwyer, P.D. (1971) Temperature regulation and cave- Esselstyn, J.A., Amar, A., and Janeke, D. (2006) Impact of
dwelling in bats: an evolutionary perspective. Mammalia post-typhoon hunting on Mariana fruit bats (Pteropus
35, 424–453. mariannus). Pac. Sci. 60, 531–539.
Dwyer, P.D. and Hamilton-Smith, E. (1965) Breeding caves Ewers, R.M. and Rodrigues, A.S.L. (2008) Estimates of
and maternity colonies of the bent-winged bat in south- reserve effectiveness are confounded by leakage. TREE
eastern Australia. Helictite 4, 3–21. 23, 113–116.
Dwyer, P.D. and Harris, J.A. (1972) Behavioural acclimati- Ewing, W.G., Studier, E.H., and O’Farrell, M.J. (1970)
sation to temperature by pregnant Miniopterus (Chirop- Autumn fat deposition and gross body composition in
tera) Physiol. Zool. 45, 14–21. three species of Myotis. Comp. Biochem. Physiol. 36,
Eby, P. (1991) Seasonal movements of grey-headed flying 119–129.
foxed, Pteropus poliocephalus, from two maternity camps in Excoffier, L., Laval, G., and Schneider, S. (2005) Arlequin
northern New South Wales. Wildlife Research 18, 547–559. ver. 3.0: An integrated software package for population
Ecroyd, C. (1993) In search of the wood rose. Forest and Bird genetics data analysis. Evolutionary Bioinformatics Online
Feb, 24–28. 1, 47–50.
292 REFERENCES

Faria, D., Laps, R.R., Baumgarten, J., and Cetra, M. (2006) the conservation of bats in African woodlands. In: Bat
Bat and bird assemblages from forests and shade cacao Biology and Conservation, edited by T.H. Kunz and
plantations in two contrasting landscapes in the Atlantic P.A. Racey, pp 261–270. Smithsonian Institution Press,
Forest of southern Bahia, Brazil. Biodivers. Conserv. 15, Washington DC.
587–612 Fenton, M.B., Rautenbach, I.L., Smith, S.E., Swanepoel, C.
Faure, P.A. and Barclay, R.M.R. (1992) The sensory basis of M., Grosell, J., and van Jaarsveld, J. (1994) Raptors
prey detection by the long-eared bat, Myotis evotis, and the and bats: threats and opportunities. Anim. Behav. 48,
consequences for prey selection. Anim. Behav. 44, 31–39. 9–18.
Faure, P.A., Fullard, J.H., and Barclay, R.M.R. (1990) The Fiedler, J. (1979) Prey catching with and without echoloca-
response of tympanate moths to the echolocation calls of tion in the Indian false vampire bat (Megaderma lyra).
a substrate gleaning bat, Myotis evotis. J. Comp. Physiol. A. Behav. Ecol. Sociobiol. 6, 155–160.
166, 843–849. Findlay, C.S. and Bourdages, J. (2000) Response time of
Faure, P.A., Fullard, J.H., and Dawson, J.W. (1993) The wetland biodiversity to road construction on adjacent
gleaning attacks of the northern long-eared bat, Myotis lands. Conserv. Biol. 14, 86–94.
septentrionalis, are relatively inaudible to moths. J. Exp. Findley, J.S. (1993) Bats: A Community Perspective. Cam-
Biol. 178, 173–189. bridge University Press, Cambridge.
Fauth, J.E., Bernardo, J., Camara, M., Resetarits Jr., W.J., Findley, J.S. and Black, H.L. (1983) Morphological
Van Buskirk, J. and McCollum, S.A. (1996) Simplifying and dietary structuring of a Zambian insectivorous bat
the jargon of community ecology: a conceptual community. Ecology 64, 625–630.
approach. Am. Nat. 147, 282–286. Findley, J.S. and Jones, C. (1964) Seasonal distribution of
Federico, P., et al. (2008) Brazilian free-tailed bats as insect the hoary bat. J. Mammal. 45, 461–470.
pest regulators in transgenic and conventional cotton Findley, J.S. and Wilson, D.E. (1974) Observations on the
crops. Ecol Appl. 18, 826–837. neotropical disk-winged bat, Thyroptera tricolor. J. Mammal.
Felsenstein, J. (2004) Inferring Phylogenies. Sinauer Asso- 55, 562–571.
ciates, Sunderland, Mass. Firbank, L.G., Petit, S., Smart, S., Blain, A., and Fuller, R.J.
Feng, A.S., Simmons, J.A., and Kick, S.A. (1978) Echo (2008) Assessing the impacts of agricultural intensifica-
detection and target-ranging neurons in the auditory tion on biodiversity: a British perspective. Phil. Trans. R.
system of the bat Eptesicus fuscus. Science 202, 645–648. Soc. B. 363, 777–787.
Fenster, C.B., Armbruster, W.S, Wilson, P., Dudash, M.R., Flanders, J., Jones, G., Benda, P., Dietz, C., Zhang, S., Li, G.,
and Thomson, J.D. (2004) Pollination syndromes and Sharifi, M., and Rossiter, S.J. (2009) Phylogeography of
floral specialization. Annu. Rev. Ecol. Syst. 35, 375–403. the greater horseshoe bat, Rhinolophus ferrumequinum:
Fenton, M.B. (1969) Summer activity of Myotis lucifugus contrasting results from mitochondrial and microsatel-
(Chiroptera: Vespertilionidae) at hibernacula in Ontario lite data. Mol. Ecol. 18, 306–318.
and Quebec. Can. J. Zool. 47, 597–602. Flaquer, C., Torre, I., and Ruiz-Jarillo, R. (2006) The value
Fenton, M.B. (1985) Communication in the Chiroptera. Indi- of bat-boxes in the conservation of Pipistrellus pygmaeus
ana University Press, Bloomington. in wetland rice paddies. Biol. Conserv. 128, 223–230.
Fenton, M.B. (1990) The foraging behaviour and ecology of Flaquer, C., Torre, I., and Arrizabalaga, A. (2007) Compar-
animal-eating bats. Can. J. Zool. 68, 411–422. ison of sampling methods for inventory of bat commu-
Fenton, M.B. (1992) Wounds and the origin of blood-feed- nities. J. Mammal. 88, 526–533.
ing in bats. Biol. J. Linn. Soc. 47, 161–171. Fleming, T.H. (1982) Foraging strategies of plant-visiting
Fenton, M. B., Acharya, L., Audet, D., Hickey, M.B.C., Merri- bats. In: The Ecology of Bats, edited by T.H. Kunz,
man, C., Obrist, M.K., and Syme, D.M. (1992) Phyllostomid pp 287–325. Plenum Press, New York.
bats (Chiroptera: Phyllostomidae) as indicators of habitat Fleming, T.H. (1986) Opportunism vs. specialization:
disruption in the Neotropics. Biotropica 24, 440–446. the evolution of feeding strategies in frugivorous
Fenton, M.B. and Bell, G.P. (1981) Recognition of species of bats. In: Frugivores and Seed Dispersal, edited by A.
insectivorous bats by their echolocation calls. J. Mammal. Estrada and T.H. Fleming, pp 105–118. W. Junk, Dor-
62, 233–243. drecht, Netherlands.
Fenton, M.B. and Fullard, J.H. (1979) The influence of bat Fleming, T.H. (1993) Plant-visiting bats. Amer. Sci. 81,
hearing on bat echolocation strategies. J. Comp. Physiol. 460–467.
132, 77–86. Fleming, T.H., Breitwisch, R., and Whitesides, G.H. (1987)
Fenton, M.B. and Rautenbach, I.L. (1998) Impacts of Patterns of tropical vertebrate frugivore diversity. Ann.
ignorance and human and elephant populations on Rev. Ecol. Syst. 18, 91–109.
 REFERENCES 293

Fleming, T.H. and Eby, P. (2003) Ecology of bat migration. Centre for Protected Areas and Watershed Management,
In: Bat Ecology, edited by T.H. Kunz and M.B. Fenton, Lao PDR.
pp156–208. University of Chicago Press, Chicago. Freeman, P.W. (1981) Correspondence of food habits
Fleming, T.H., Hooper, E.T. and Wilson, D.E. (1972) Three and morphology in insectivorous bats. J. Mammal. 62,
Central American bat communities: Structure, repro- 166–173.
ductive cycles and movement patterns. Ecology 53, Frick, W.F., Hayes, J.P., and Heady, P.A. (2008) Island
555–569. biogeography of bats in Baja California, Mexico: patterns
Fleming, T.H. and Muchhala, N. (2008) Nectar-feeding bird of bat species richness in a near-shore archipelago.
and bat niches in two worlds: pantropical comparisons J. Biogeogr. 35, 353–364.
of vertebrate pollination systems. J. Biogeogr. 35, 764–780. Frick, W.F., Reynolds, D.S., and Kunz, T.H. (2010a) Influ-
Fleming, T.H., Nunez, R.A., and Sternberg, L. da S. L. ence of climate and reproductive timing on demography
(1993) Seasonal changes in the diets of migrant and of little brown myotis Myotis lucifugus. J. Anim. Ecol. 79,
non-migrant nectarivorous bats as revealed by carbon 128–136.
stable isotope analysis. Oecologia 94, 72–75. Frick, W.F., Pollock, J.F., Hicks, A.C., Langwig, K.E., Rey-
Fleming, T.H. and Racey, P.A. (2009) Island Bats: Evolution, nolds, D.S., Turner, G.G., Butchkoski, C.M., and Kunz, T.
Ecology, and Conservation. University of Chicago Press, H. (2010b) An emerging disease causes regional popula-
Chicago. tion collapse of a common North American bat species.
Flynn, J.J. and Wyss, A.R. (1998) Recent advances in South Science 329, 679–682.
American mammalian paleontology. TREE 13, 449–454. Frith K.E. (2010) Investigating the population structure of
Foppen, R. and Reijnen, R. (1994) The effect of car traffic on Daubenton's bat in Europe: a multiple marker approach.
breeding bird populations in woodland. II. Breeding PhD Thesis. University of Sheffield, UK.
dispersal of male willow warblers (Phylloscopus trochilus) Fritts, T.H. and Rodda, G.H. (1998) The role of introduced
in relation to the proximity of a highway. J. Appl. Ecol. 31, species in the degradation of island ecosystems: A case
95–101. history of Guam. Annu. Rev. Ecol. Syst. 29, 113–40.
Forey, P.L., Kitching, I.J., and Williams, D. (1998) Cladistics: Fujita, M.S. and Tuttle, M.D. (2001) Flying foxes (Chirop-
Theory and Practice. 2nd revised edition. Oxford University tera, Pteropodidae): threatened animals of key ecological
Press, Oxford. and economic importance. Conserv. Biol. 5, 455–463.
Forister, M.L., McCall, A.C., Sanders, N.J., Fordyce, J.A., Fullard, J.H. (1987) Sensory ecology and neuroethology
Thorne, J.H., O’Brien, J., David P., Waetjen, D.P., and of moths and bats: interactions in a global perspective.
Shapiro, A.M. (2010) Compounded effects of climate In: Recent Advances in the Study of Bats, edited by M.B.
change and habitat alteration shift patterns of butterfly Fenton, P.A. Racey, and J.M.V. Rayner, pp 244–272.
diversity. PNAS 107, 2088–2092. Cambridge University Press, Cambridge.
Forman, R. T. T., Sperling, D., Bissonette, J. A., Clevenger, Fullard, J.H. (1992) The neuroethology of sound production
A. P., Cutshall, C. D., Dale, V. H., Fahrig, L., France, R., in tiger moths (Lepidoptera, Arctiidae) I. Rhythmicity
Goldman, C. R., Heanue, K., Jones, J. A., Swanson, F. J., and central control. J. Comp. Physiol. A. 170, 575–588.
Turrentine, T., and Winter T. C. (2003) Road Ecology: Fullard, J.H., Barclay, R.M.R., and Thomas, D.W. (1993)
Science and Solutions, pp 481. Island Press, Washington. Echolocation in free-flying atiu swiftlets (Aerodramus
Fortuna, M.A., Popa-Lisseanu, A.G., Ibáñez, C., and sawtelli). Biotropica 25, 334–339.
Bascompte, J. (2009) The roosting spatial network of a Fullard, J.H. and Dawson, J.W. (1997) Why do diurnal
bird-predator bat. Ecology 90, 934–944. moths have ears? Naturwissenschaften 86, 276–279.
Foster, M.S. (1992) Tent roosts of Macconnell’s bat Fullard, J.H., Koehler, C., Surlykke, A., and McKenzie, N.L.
(Vampyressa macconnelli). Biotropica 24, 447–454. (1991) Echolocation ecology and flight morphology of
Foster, R.B., Arce, J., and Wachter, T.S. (1986) Dispersal insectivorous bats (Chiroptera) in south western Austra-
and the sequential plant communities in Amazonian lia. Aust. J. Zool. 39, 427–438.
Peru floodplain. In: Frugivores and Seed Dispersal, edited Fullard, J.H., Ratcliffe, J.M., and Soutar, A.R. (2004) Extinc-
by A. Estrada and T.H. Fleming, pp 357–370. W. Junk, tion of the acoustic startle response in moths endemic to
Dordrecht, Netherlands. a bat-free habitat. J. Evol. Biol. 17, 856–861.
Francis, C.M., Guillen, A., and Robinson, M.F. (1999) Order Fullard, J.H., Simmons, J.A., and Saillant, P.A. (1994)
Chiroptera: Bats. In: Wildlife in Lao PDR. 1999 Status Jamming bat echolocation: the dogbane tiger moth
Report, compiled by J.W. Duckworth, R.E. Salter, and Cynia tenera times its clicks to the terminal attack calls
K. Khounboline, pp. 225–235. IUCN—The World of the big brown bat Eptesicus fuscus. J. Exp. Biol. 194,
Conservation Union, Wildlife Conservation Society and 285–298.
294 REFERENCES

Fullard, J.H. and Yack, J.E. (1993) The evolutionary biology Gaston, K.J., Chown, S.L., and Evans, K.L. (2008) Ecogeo-
of insect hearing. TREE 8, 248–252. graphical rules: elements of a synthesis. J. Biogeogr. 35,
Fuller, R.J., Norton, L.R., Feber, R.E., Johnson, P.J., Cham- 483–500.
berlain, D.E., Joys, A.C., Mathews, F., Stuart, R.C., Town- Gehrt, S.D. and Chelsvig, J.F. (2003) Bat activity in an urban
send, M.C., Manley, W.J., Wolfe, M.S., Macdonald, D.W., landscape: patterns at the landscape and microhabitat
and Firbank, L.G. (2005) Benefits of organic farming to scale. Ecol. Appl. 13, 939–950.
biodiversity vary among taxa. Biol. Lett. 1, 431–434. Gehrt, S.D. and Chelsvig, J.F. (2004) Species-specific
Furman, A. and Ozgul, A. (2002) The distribution of cave- patterns of bat activity in an urban landscape. Ecol.
dwelling bats and conservation status of underground Appl. 14, 625–635.
habitats in Istanbul area. Ecol. Res. 17, 69–77. Gentry, A.H. (1974) Flowering phenology and diversity in
Furman, A. and Ozgul, A. (2004) The distribution of cave- tropical Bignoniaceae. Biotropica 6, 64–68.
dwelling bats and conservation status of underground Gerell, R. and Lundberg, K. (1985) Social organisation in
habitats in Northwestern Turkey. Biol. Conserv. 120, the bat Pipistrellus pipistrellus. Behav. Ecol. Sociobiol. 16,
243–248. 177–184.
Furmankiewicz, J. and Altringham, J.D. (2007) Genetic Gerlach, G. and Musolf, K. (2000) Fragmentation of
structure in a swarming brown long-eared bat (Plecotus landscape as a cause for genetic subdivision in bank
auritus) population: evidence for mating at swarming voles. Conserv. Biol. 14, 1066–1074.
sites. Conserv. Genet. 8, 913–923. Ghose, K., Horiuchi, T.K., Krishnaprasad, P.S., and Moss,
Furmankiewicz, J. and Kucharska, M. (2009) Migration C.F. (2006) Echolocating bats use a nearly time-optimal
of bats along a large river valley in southwestern Poland. strategy to intercept prey. PLoS Biol 4, e108.
J. Mammal. 90, 1310–1317. Gillam, E.H. and McCracken, G.F. (2007) Variability in the
Futuyma, D.J. and Slatkin, M. (1983) Coevolution. Sinauer echolocation of Tadarida brasiliensis: effects of geography
Associates, Sunderland, Massacheusetts. and local acoustic environment. Anim. Behav. 74, 277–286.
Fuzessery, Z.M., Buttenhof, B., Andrews, B., and Kennedy, Gillette, D.D. and Kimbourgh, J.D. (1970) Chiropteran
J.M. (1993) Passive sound localisation of prey by the pallid mortality. In: About Bats, edited by B.H. Slaughter and
bat (Antrozous pallidus). J. Comp. Physiol. A. 171, 767–777. D.W. Waiton, pp 262–281. Dallas Southern Methodist
Gabriel, D., Carver, S.J., Durham, H., Kunin, W.E., Palmer, University Press, Dallas.
R.C., Sait, S.M., Stagl, S,. and Benton, T.G. (2009) The Glover, A.M. and Altringham, J.D. (2008) Cave selection
spatial aggregation of organic farming in England and and use by swarming bat species. Biol. Conserv. 141,
its underlying environmental correlates. J. Appl. Ecol. 46, 1493–1504.
323–333. Glover, A.M., Butlin, R.K., and Altringham, J.D. (2011)
Gabriel, D., Sait, S.M., Hodgson, J.A., Schmutz, U., Kunin, Population structure of the swarming Natterer’s bat,
W.E., and Benton, T.G. (2010) Scale matters: the impact Myotis nattereri, at multiple spatial scales. Under review.
of organic farming on biodiversity at different spatial Goerlitz, H.R., ter Hofstede, H.M., Zeale, M.R.K., Jones, G
scales. Ecol. Lett. 13, 858–869. and Holderied, M.W. (2010) An aerial-hawking bat uses
Gaisler, J., Hanák, V., and Dungel, J. (1979) A contribution stealth echolocation to counter moth hearing. Curr. Biol.
to the population ecology of Nyctalus noctula (Mamma- 20, 1568–1572.
lia, Chiroptera). Acta Sci. Nat. Brno. 13, 1–38. Goodwin, G.G. and Greenhall, A.M. (1961) A review of the
Gall, L.F. and Tiffney, B.H. (1983) A fossil noctuid moth bats of Trinidad and Tobago. Bull. Am. Mus. Nat. Hist.
egg from the late Cretaceous of eastern North America. 122, 187–302.
Science 219, 507–509. Gorresen, P.M. and Willig, M.R. (2004) Landscape
Garcia, C.A., Bhagwat, S.A., Ghazoul, J., Nath, C.D., responses of bats to habitat fragmentation in atlantic
Nanaya, K.M., Kushalappa, C.G., Raghuramulu, Y., forest of Paraguay. J. Mammal. 85, 688–697.
Nasi, R., and Vaast, P. (2010) Biodiversity conservation Gorresen, P.M., Willig, M.R., and Strauss, R.E. (2005)
in agricultural landscapes: challenges and opportunities Multivariate analysis of scale-dependent associations
of coffee agroforests in the Western Ghats, India. between bats and landscape structure. Ecol. Appl. 15,
Conserv. Biol. 24, 479–488. 2126–2136.
Garroway, C.J. and Broders, H.G. (2007) Nonrandom asso- Gottfried, I. (2009) Use of underground hibernacula by the
ciation patterns at northern long-eared bat maternity barbastelle (Barbastella barbastellus) outside the hiberna-
roosts. Can. J. Zool./Rev. Can. Zool. 85, 956–964. tion season. Acta Chiropt. 11, 363–373.
Gaston, K.J. and Blackburn, T.M. (2000) Pattern and Process Gould, E. (1978) Foraging behaviour of Malaysian nectar-
in Macroecology. Blackwell Science, Oxford. feeding bats. Biotropica 10, 184–193.
 REFERENCES 295

Grant, J.D.A. (1991) Prey location by two Australian long- Guillen-Servent, A. and Ibáñez, C. (2007) Unusual echolo-
eared bats, Nyctophilus gouldi and N. geoffroyi. Aust. cation behavior in a small molossid bat, Molossops tem-
J. Zool. Lond. 39, 45–56. minckii, that forages near background clutter. Behav. Ecol.
Greaves, G.J., Mathieu, R., and Seddona, P.J. (2006) Predic- Sociobiol. 61, 1599–1613.
tive modelling and ground validation of the spatial Guisan, A. and Thuiller, W. (2005) Predicting species dis-
distribution of the New Zealand long-tailed bat tribution: offering more than simple habitat models. Ecol.
(Chalinolobus tuberculatus). Biol. Conserv. 132, 211–221. Lett. 8, 993–1009.
Green, R.E., Cornell, S.J., Scharlemann, J.P.W., and Balm- Guppy, A. and Coles, R.B. (1988) Acoustical and neural asp-
ford, A. (2005) Farming and the fate of wild nature. ects of hearing in the Australian gleaning bats Macroderma
Science. 307, 550–555. gigas and Nyctophilus gouldi. J. Comp. Physiol. 162, 653–668.
Greenhall, A.M. and Stell, G. (1960) Bionomics and Gustin, K. and McCracken, G.F. (1987) Scent recognition in
chemical control of free-tailed house bats (Molossus) in the Mexican free-tailed bat, Tadarida brasiliensis mexicana.
Trinidad. Spec. Sci. Rep. Wildl. Serv. Washington D.C. Anim. Behav. 35, 13–19.
Greenwood, P.J. (1980) Mating systems, philopatry and dis- Habersetzer, J. and Storch, G. (1992) Cochlea size in extant
persal in birds and mammals. Anim. Behav. 28, 1140–1162. Chiroptera and middle Eocene Microchiropterans from
Grenyer, R. et al. (2006) Global distribution and conserva- Messel. Naturwiss. 79, 462–466.
tion of rare and threatened vertebrates. Nature. 444, Hall, J.S. (1962) A life history and taxonomic study of the
93–96. Indiana bat, Myotis sodalis. Sci. Pub. Reading Pub. Mus.
Grieser Johns, A. (1997) Timber Production and Biodiversity Art Gall. 12, 1–68.
Conservation in Tropical Rain Forests. Cambridge Univer- Hall, L. and Richards, G. (2000) Flying Foxes. Fruit and
sity Press, Cambridge, UK. Blossom Bats of Australia. University of New South
Griffin, D.R. (1958) Listening in the Dark. Reprinted 1986. Wales Press, Sydney.
Yale University Press, New Haven, Conn. Halpin, K., Hyatt, A.D., Plowright, R.K., Epstein, J.H.,
Griffin, D.R. (1970) Migration and homing of bats. In: Daszak, P., Field, H.E., Wang, L., and Daniels, P.W.
Biology of Bats, Vol. I, edited by W.A. Wimsatt, (2007) Emerging viruses: coming in on a wrinkled wing
pp 233–246. Academic Press, New York. and a prayer. Emerg. Infect. 44, 711–717.
Griffin, D.R, Webster, F.A. and Michael, C.R. (1960) The Hamilton, I.A. and Barclay, R.M.R. (1994) Patterns of daily
echolocation of flying insects by bats. Anim. Behav. 8, torpor and day-roost selection by male and female big
141–154. brown bats (Eptesicus fuscus). Can. J. Zool. 72, 744–749.
Griffiths, J.R. and Varne, R. (1972) Evolution of the Tasman Harris, S.L., Brookes, S.M., Jones, G., Hutson, A.M., Racey,
Sea, Macquarie Ridge and Alpine Fault. Nature Phys. Sci. P.A., Aegerter, J., Smith, G.C., McElhinney, L.M., and
235, 83–86. Fooks, A.R. (2006) European bat lyssaviruses: Distribu-
Grindal, S.D., Morissette, J.L., and Brigham, R.M. (1999) tion, prevalence and implications for conservation. Biol.
Concentration of bat activity in riparian habitats over an Conserv. 131, 193–210.
elevational gradient. Can. J. Zool. 77, 972–977. Harvey, C.A. and Villalobos, J.A.G. (2007) Agroforestry
Grinevitch, L., Holroyd, S.L., and Barclay, R.M.R. (1995) systems conserve species-rich but modified assemblages
Sex-differences in the use of daily torpor and foraging of tropical birds and bats. Biodivers. Conserv. 16,
time by big brown bats (Eptesicus fuscus) during the 2257–2292.
reproductive season. J. Zool. Lond. 235, 301–309. Hays, G.C., Speakman, J.R., and Webb, P.I. (1992) Why do
Grinnell, A.D. (1963) The neurophysiology of audition in brown long-eared bats (Plecotus auritus) fly in winter?
bats: intensity and frequency parameters. J. Physiol. 167, Physiol. Zool. 65, 554–567.
38–66. Hayward, B.J. and Cross, S.P. (1979) The natural history of
Grinnell, A.D. and Griffin, D.R. (1958) The sensitivity of Pipistrellus hesperus (Chiroptera: Vespertilionidae). Office
echolocation in bats. Biol. Bull. 114, 10–22. Res. West. N.M. 3, 1–36.
Grodzinski, U., Spiegel, O., Korine, C., and Holderied, M.W. Heaney, L.R., Walsh, J.S., and Peterson A.T. (2005) The
(2009) Context-dependent flight speed: evidence for roles of geological history and colonization ability in
energetically optimal flight speed in the bat Pipistrellus genetic differentiation between mammalian populations
kuhlii? J. Anim. Ecol. 78, 540–548. in the Philippine archipelago. J. Biogeogr. 32, 229–247.
Grytnes, J-A. and McCain, C.M. (2007) Elevational trends Heckel, G. and von Helversen, O. (2002) Male tactics
in biodiversity. In: Encyclopaedia of Biodiversity, edited by and reproductive success in the harem polygynous bat
S. Levin. Elsevier, Amsterdam. Saccopteryx bilineata. Behav. Ecol. 13, 750–756.
296 REFERENCES

Heckel, G. and Helversen, O. (2003) Genetic mating system Herd, R.M. and Fenton, M.B. (1983) An electrophoretic,
and the significance of harem associations in the bat morphological, and ecological investigation of a putative
Saccopteryx bilineata. Mol. Ecol. 12, 219–227. hybrid zone between Myotis lucifugus and Myotis yuma-
Hedenström, A. (2002) Aerodynamics, evolution and nensis (Chiroptera: Vespertilionidae) Can. J. Zool. 61,
ecology of bird flight TREE 17, 415–22. 2029–2050.
Hedenström, A., Johansson, L.C., Wolf, M., von Busse, R., Herreid, C.F. (1963) Temperature regulation and metabo-
Winter, Y., and Spedding, G.R. (2007) Bat flight gener- lism in Mexican free-tailed bats. Science 142, 1573–1574.
ates complex aerodynamic tracks. Science 316, 894–897. Herreid, C.F. (1967) Temperature regulation, temperature
Hedenström, A., Johansson, L.C. and Spedding G.R. (2009) preference and tolerance, and metabolism of young and
Bird or bat: comparing airframe design and performance. adult free-tailed bats. Physiol. Zool. 40, 1–22.
Bioinsp. Biomim. 4, 015001. Herrera, L.G., Fleming, T.H., and Findley, J.S. (1993) Geo-
Heer, K., Albrecht, L., and Kalko, E.K.V. (2010). Effects of graphic variation in carbon composition of the pallid bat,
ingestion by neotropical bats on germination parameters Antrozous pallidus, and its dietary implications. J. Mam-
of native free-standing and strangler figs (Ficus sp., mal. 74, 601–606.
Moraceae). Oecologia 163, 425–435. Hessel, K. and Schmidt, U. (1994) Multimodal orientation
Heinrich, B. and Raven P.H. (1972) Energetics and pollina- in Carollia perspicillata (Phyllostomidae). Folia Zool. 43,
tion. Science 176, 597–602. 339–346.
Heithaus, E.R. (1982) Coevolution between bats and plants. Hewitt, G.M. (1996) Some genetic consequences of ice ages,
In: The Ecology of Bats, edited by T.H. Kunz, pp 327–367. and their role in divergence and speciation. Biol. J. Linn.
Plenum Press, New York. Soc. 58, 247–276.
Heithaus, E.R. and Fleming, T.H. (1978) Foraging move- Hewitt, G.M. (2000) The genetic legacy of the Quaternary
ments of a frugivorous bat, Carollia perspicillata (Phyllos- ice ages. Nature 405, 907–913.
tomidae). Ecol. Monog. 48, 127–143. Hewitt, G.M. (2004) The structure of biodiversity – insights
Heithaus, E.R., Opler, P.A., and Baker, H.G. (1974) from molecular phylogeography. Front. Zool. 1, 16pp.
Bat activity and pollination of Bauhinia pauletia: plant- Hill, J.E. (1974) A new family, genus and species of bat
pollinator coevolution. Ecology 55, 412–419. (Mammalia: Chiroptera) from Thailand. Bull. Brit. Mus.
Helgen, K.M., Helgen, L.E., and Wilson, D.E. (2009) Pacific (Nat. Hist.). Zoology 32, 29–43.
Flying Foxes (Mammalia: Chiroptera): Two New Species Hill, J.E. and Smith, J.D. (1984) Bats: A Natural History.
of Pteropus from Samoa, Probably Extinct. Amer. Mus. British Museum (Natural History). London.
Novit. 3646, 1–37. Hillebrand, H. (2004) On the generality of the latitudinal
Heller, K.-G., Achmann, R. and Witt, K. (1993) Monogamy diversity gradient. Am. Nat. 163, 192–211.
in the bat Rhinolophus sedulus. Z. Saug. 58, 376–378. Hiryu, S., Hagino, T., Riquimaroux, H., and Watanabe, Y.
Heller, K.-G. and Helversen, O. (1989) Resource partition- (2007) Echo-intensity compensation in echolocating bats
ing of sonar frequency bands in rhinolophid bats. (Pipistrellus abramus) during flight measured by a teleme-
Oecologia 80, 178–186. try microphone. J. Acoust. Soc. Am. 121, 1749–1757.
Heller, N.E., Zavaleta, E.S. (2009) Biodiversity manage- Hodgkison, R., Balding, S.T., Zubaid, A., and Kunz, T.H.
ment in the face of climate change: A review of 22 (2003) Fruit bats (Chiroptera: Pteropodidae) as seed dis-
years of recommendations. Biol. Conserv. 142, 14–32. persers and pollinators in a lowland Malaysian rain
Helversen, D. and Helversen, O. (1999) Acoustic guide in forest. Biotropica 35, 491–502.
bat-pollinated flowers. Nature 398, 759–760. Hodgkison, R., Balding, S.T., Zubaid, A., and Kunz, T.H.
Helversen, O. and Winter, Y. (2003) Glossophagine bats and (2004) Habitat structure, wing morphology, and the ver-
their flowers: costs and benefits for plants and pollinators. tical stratification of Malaysian fruit bats (Megachirop-
In: Bat Ecology edited by T.H. Kunz and M.B. Fenton, tera: Pteropodidae). J. Trop. Ecol. 20, 667–673.
pp 346–397. University of Chicago Press, Chicago. Hoffman, F.G. and Baker, R.J. (2003) Comparative phylo-
Hennig, W. (1966) Phylogenetic Systematics. University of geography of short-tailed bats (Carollia: Phyllostomidae).
Illinois Press, Urbana. Mol. Ecol. 12, 3403–3414.
Henry, M., Pons, J-M., and Cosson, J-F. (2007) Foraging Hofstedeter, H.M., Ratcliffe, J.M., and Fullard, J.H. (2008)
behaviour of a frugivorous bat helps bridge landscape Nocturnal activity positively correlated with auditory
connectivity and ecological processes in a fragmented sensitivity in noctuoid moths. Biol. Lett. 4, 262–265.
rainforest. J. Anim. Ecol. 76, 801–813. Holderied, M.W. and Helversen, O. (2003) Echolocation
Henson, O.W. (1965) The activity and function of the middle range and wingbeat period match in aerial-hawking
ear muscles in echolocating bats. J. Physiol. 180, 871–887. bats. Proc. R. Soc. Lond. B. 270, 2293–2299.
 REFERENCES 297

Holderied, M.W. and Helversen, O. (2006) ‘Binaural echo Hughes, P.M. and Rayner, J.M.V. (1993) The flight of pip-
disparity’ as a potential indicator of object orientation istrelle bats Pipistrellus pipistrellus during pregnancy and
and cue for object recognition in echolocating nectar- lactation. J. Zool. Lond. 230, 541–555.
feeding bats. J. Exp. Biol. 209, 3457–3468. Humphries, M.M., Thomas, D.W., and Kramer, D.L. (2003)
Holderied, M.W., Jones, G., and Helversen, O. (2006) Flight The role of energy availability in mammalian hiberna-
and echolocation behaviour of whiskered bats commut- tion: a cost-benefit approach. Physiol. Biochem. Zool. 76,
ing along a hedgerow: range-dependent sonar signal 165–179.
design, Doppler tolerance and evidence for ‘acoustic Humphries, M.M., Thomas, D.W., and Speakman, J.R.
focussing’. J. Exp. Biol. 209, 1816–1826. (2002) Climate-mediated energetic constraints on the
Holland, R.A. (2007) Orientation and navigation in bats: distribution of hibernating mammals. Nature 418,
known unknowns or unknown unknowns? Behav. Ecol. 313–316.
Sociobiol. 61, 653–660. Humphrey, S.R. (1975) Nursery roosts and community
Holland, R.A., Borissov, I., and Siemers, B.M. (2010) A diversity in nearctic bats. J. Mammal. 56, 321–346.
nocturnal mammal, the greater mouse-eared bat, cali- Humphrey, S.R., Richter, A.R., and Cope, J.B. (1977) Sum-
brates a magnetic compass by the sun. PNAS 107, mer habitat and ecology of the endangered Indiana bat,
6941–6945. Myotis sodalis. J. Mammal. 58, 334–346.
Holland, R.A., Thorup, K., Vonhof, M.J., Cochran, W.W., Huntley, B., Collingham, Y.C., Willis, S.G., and Green, R.E.
and Wikelski, M. (2006) Navigation—Bat orientation (2008) Potential impacts of climatic change on European
using Earth’s magnetic field. Nature 444, 702. breeding birds. PLoS One 3, e1439.
Holland, R.A., Waters, D.A., and Rayner, R.M.V. (2004) Huntley, B., Green, R., Collingham, Y.C., and Willis, S.G.
Echolocation signal structure in the megachiropteran (2007) A Climatic Atlas of European Breeding Birds. Lynx
bat Rousettus aegyptiacus Geoffroy 1810. J. Exp. Biol. 207, Edicions, Barcelona.
4361–4369. Hutson, A.M., Mickleburgh, S. and Mitchell-Jones, A.J.
Hoofer, S. R., and van den Bussche, R.A. (2003) Molecular (1995) Bats Underground: A Conservation Code, 2nd edi-
phylogenetics of the chiropteran family Vespertilioni- tion. The Bat Conservation Trust, London.
dae. Acta Chiropt. 5, 1–63. Hutson, A.M., Mickleburgh, S.P., and Racey, P.A. (2001)
Hooper, J.H.D. and Hooper, W.H. (1956) Habits and move- Microchiropteran Bats: Global Status Survey and Conserva-
ments of cave-dwelling bats in Devonshire. Proc. Zool. tion Action Plan. IUCN/SSC Chiroptera Specialist
Soc. Lond. 127, 1–26. Group. IUCN, Gland, Switzerland and Cambridge, UK.
Horn, J.W., Arnett, E.B., and Kunz, T.H. (2008) Behavioral Hutterer, R., Ivanova, T., Meyer-Cords, C., and Rodrigues,
responses of bats to operating wind turbines. J. Wildlife L. (2005) Bat Migrations in Europe. A Review of Banding
Manage. 72, 123–132. Data and Literature. Naturschutz und Biologische Vielfalt
Hosken, D. J. (1997) Sperm competition in bats. Proc. R. Soc. Heft, 28. Federal Agency for Nature Conservation, Bonn.
Lond. B. 264, 385–392. Ibáñez, C., Guillén, A., Agirre-Mendi, P.T., Juste, J.,
Hosken, D. J. (1998) Testes mass in megachiropteran bats Schreur, G., Cordero, A.I. and Popa-Lisseanu, A.G.
varies in accordance with sperm competition theory. (2009) Sexual segregation in Iberian noctule bats. J. Mam-
Behav. Ecol. Sociobiol. 44, 169–177. mal. 90, 235–243.
Howell, D.J. (1974) Bats and pollen: physiological aspects Ibáñez, C., Juste, J., Garda-Mudarra, J.L., and Girre-Mendl,
of the syndrome of Chiropterophily. Comp. Biochem. P.T. (2001) Bat predation on nocturnally migrating birds.
Physiol. 48, 263–276. PNAS 98, 9700–9702.
Howell, D.J. (1979) Flock foraging in nectar-feeding bats: Ichida, J.M., Rosa, M.G.P., and Casagrande, V.A. (2000)
advantages to the bats and to the host plants. Amer. Nat. Does the visual system of the flying fox resemble that
114, 23–49. of primates? The distribution of calcium-binding pro-
Howell, D.J. and Hartl, D.L. (1980) Optimal foraging in teins in the primary visual pathway of Pteropus polioce-
Glossophagine bats: when to give up. Am. Nat. 115, phalus. J. Comp. Neurol. 417, 73–87.
696–704. Ingle, N.R. (2003) Seed dispersal by wind, birds, and bats
Hughes, P.M., Ransome, R.D., and Jones, G. (1989) Aero- between Philippine montane rainforest and successional
dynamic constraints on flight ontogeny in free-living vegetation. Oecologia 134, 251–261.
greater horseshoe bats, Rhinolophus ferrumequinum. In: Iniguez-Davalos, L.I. (1993) Patrones ecologicos en la
European Bat Research 1987, edited by V. Hanák, I. Hor- comunidad de murcielagos de La Sierra de Manantlan.
ácek, and J. Gaisler, pp 255–262. Charles University In: Avances en el estudio de los mamiferos de Mexico,
Press, Praha. edited by R.A. Medellin and G. Ceballos, pp 355–370.
298 REFERENCES

Publicaciones Especiales, 1, Asociacion Mexicana de Jones, G., Duverge, L., and Ransome, R.D. (1995) Conser-
Mastozoologia, A.C., Distrito Federal, Mexico. vation biology of an endangered species: field studies
IPCC (2010) Intergovernmental Panel on Climate Change. of greater horseshoe bats. In: Ecology, Evolution and
www.ipcc.ch [accessed 28 February 2011]. Behaviour of Bats, edited by P.A. Racey and S.M. Swift,
IUCN (2009) Red List of Threatened Species, IUCN-SSC. Symp. Zool. Soc. Lond. 67, pp 309–324. Oxford University
www.iucnredlist.org [accessed 28 February 2011]. Press, Oxford.
Jaberg, C. and Guisan, A. (2001) Modelling the distribution Jones, G. and Holderied, M.W. (2007) Bat echolocation
of bats in relation to landscape structure in a temperate calls: adaptation and convergent evolution. Proc. R.Soc.
mountain environment. J. Appl. Ecol. 38, 1169–1181. Lond. B. 274, 905–912.
Jaeger, J.A.G., Scharz-von Raumer, H-G., Esswein, H., Jones, G., Hughes, P.M., and Rayner, J.M.V. (1991) The
Müller, M., and Schmidt-Lüttmann, M. (2007) Time development of vocalizations in Pipistrellus pipistrellus
series of landscape fragmentation caused by transporta- (Chiroptera: Vespertilionidae) during post-natal growth
tion infrastructure and urban development: a case and the maintenance of individual vocal signatures.
study from Baden-Württemberg (Germany). Ecology J. Zool. Lond. 225, 71–84.
and Society 12, 22. www.ecologyandsociety.org/vol12/ Jones, G., Jacobs, D.S., Kunz, T.H., Willig, M.R., and Racey,
iss1/art22/ [accessed 28 February 2011]. P.A. (2009) Carpe noctem: the importance of bats as
James, R.S., Altringham, J.D., and Goldspink D.F. (1995) Bioindicators. Endangered Species Research 8, 93–115.
The mechanical properties of fast and slow skeletal mus- Jones, G. and Ransome, R.D. (1993) Echolocation calls of
cles of the mouse in relation to their locomotory function. bats are influenced by maternal effects and change over a
J. Exp. Biol. 198, 491–502. lifetime. Proc. Roy. Soc. Lond. B. 252, 125–128.
Janzen, D.H. (1976) Why tropical trees have rotten cores. Jones, G. and Rayner, J.M.V. (1988) Flight performance,
Biotropica 8, 110. foraging tactics and echolocation in free-living Dauben-
Jepson, G.L. (1966) Early Eocene bat from Wyoming. ton’s bats Myotis daubentonii (Chiroptera: Vespertilioni-
Science 154, 1333–1339. dae) J. Zool. Lond. 215, 113–132.
Jepson, G.L. (1970) Bat origins and evolution. In: Biology of Jones, G. and Rayner, J.M.V. (1989) Optimal flight speed in
Bats, Vol. I, edited by W.A. Wimsatt, pp 1–64. Academic pipistrelle bats, Pipistrellus pipistrellus. Eur. Bat Res. 1987,
Press, New York. 247–253.
Jetz, W., Wilcove, D.S., and Dobson, A.P. (2007) Projected Jones, G. and Rayner, J.M.V. (1991) Flight performance,
impacts of climate and land-use change on the global foraging tactics and echolocation in the trawling insec-
diversity of birds. PLoS Biology 5, e157. tivorous bat Myotis adversus (Chiroptera: Vespertilioni-
Johansson, L.C., Wolf, M., von Busse, R., Winter, Y., dae) J. Zool. Lond. 225, 393–412.
Spedding G.R., and Hedenstrom, A. (2008) The near Jones, G. and Rydell, J. (1994) Foraging strategy and pre-
and far wake of Pallas’ long tongued bat (Glossophaga dation risk as factors influencing emergence time in
soricina). J. Exp. Biol. 211, 2909–2918. echolocating bats. Phil. Trans. Roy. Soc. B. 346, 445–455.
Johansson, M. and de Jong, J. (1996) Bat species diversity in Jones, G. and Rydell, J. (2003) Attack and defence: interac-
a lake archipelago in central Sweden. Biodivers. Conserv. tions between echolocating bats and their insect prey. In:
5, 1221–1229. Bat Ecology, edited by T.H. Kunz and M.B. Fenton,
Johnson-Murray, J.L. (1977) Myology of the gliding mem- pp 301–345. University of Chicago Press, Chicago.
branes of some petauristine rodents (genera: Glaucomys, Jones, G. and Teeling, E.C. (2006) The evolution of echolo-
Pteromys, Petinomys and Petaurista). J. Mammal. 59, cation in bats. TREE 21, 149–156.
374–384. Jones, K.E., Purvis, A., and Gittleman, J.L. (2003) Biological
Johnston, I.A. and Bennett, A.F. (1996) Animals and Temper- correlates of extinction risk in bats. Am. Nat. 161, 601–614.
ature: Phenotypic and Evolutionary Adaptation. Cambridge Jones, K.J. and Genoways, H.H. (1970) Chiropteran sys-
Univ. Press, Cambridge. tematics. In: About Bats: A Chiropteran Symposium, edited
Jones, G. (1990) Prey selection by the greater horseshoe by R.H. Slaughter and D.W. Walton, pp 3–21. Southern
bat (Rhinolophus ferrumequinum): optimal foraging by Methodist University Press, Dallas.
echolocation? J. Anim. Ecol. 59, 587–602. Julian, S. and Altringham, J.D. (1994) Bat predation by a
Jones, G. (1994) Scaling of wingbeat and echolocation pulse tawny owl. Naturalist 119, 49–56.
emission rates in bats: why are aerial insectivores so Jung, K. and Kalko, E.K.V. (2010) Where forest meets
small? Funct. Ecol. 8, 450–457. urbanization: foraging plasticity of aerial insectivorous
Jones, G. (1999). Scaling of echolocation call parameters in bats in an anthropogenically altered environment.
bats. J. Exp. Biol. 202, 3359–3367. J. Mammal. 91, 144–153.
 REFERENCES 299

Jung, K., Kalko, E.K.V., and Helversen, O. (2007) Echoloca- Kerth, G., Mayer, F., and Konig, B. (2000) Mitochondrial
tion calls in Central American emballonurid bats: signal DNA (mtDNA) reveals that female Bechstein’s bats live
design and call frequency alternation. J. Zool. 272, 125–137. in closed societies. Mol. Ecol. 9, 793–800.
Juste, J., Ibáñez, C., Muñoz, J., Trujillo, D., Benda, P., Kar- Kerth, G., Mayer, F., and Petit, E. (2002) Extreme sex-biased
ataş, A., and Ruedi, M. (2004) Mitochondrial phylogeo- dispersal in the communally breeding, nonmigratory
graphy of the long-eared bats (Plecotus) in the Bechstein’s bat (Myotis bechsteinii). Mol. Ecol. 11,
Mediterranean Palaearctic and Atlantic Islands. Mol. 1491–1498.
Phylogenet. Evol. 31, 1114–1126. Kerth, G. and Melber, M. (2009) Species-specific barrier
Kalcounis-Rueppell, M.C., Payne, V.H., Huff, S.R., and effects of a motorway on the habitat use of two
Boyko, A.L. (2007) Effects of wastewater treatment threatened forest-living bat species. Biol. Conserv. 142,
plant effluent on bat foraging ecology in an urban stream 270–279.
system. Biol. Conserv. 138, 120–130. Kerth, G. and Morf, L. (2004) Behavioural and genetic
Kalka, M.B., Smith, A.R., and Kalko, E.K.V. (2008) Bats data suggest that Bechstein’s bats predominantly mate
limit arthropods and herbivory in a tropical forest. outside the breeding habitat. Ethology 110, 987–999.
Science 320, 71. Kerth, G., Petrov, B., Conti, A., Anastasov, D., Weishaar,
Kalko, E.K.V. (1995) Echolocation signal design, foraging M., Gazaryan, S., Jaquiéry, J., König, B., Perrin, N.,
habitats and guild structure in six neotropical sheath- and Bruyndonckx, N. (2008) Communally breeding
tailed bats (Emballonuridae). In: Ecology, Evolution and Bechstein’s bats have a stable social system that is inde-
Behaviour of Bats, edited by P.A. Racey and S. Swift, pendent from the postglacial history and location of the
Symp. Zool. Soc. Lond. 67, pp 259–273. Oxford University populations. Mol. Ecol. 17, 2368–2381.
Press, Oxford. Kerth, G. and Reckardt, K. (2003) Information transfer
Kalko, E.K.V. and Condon, M. (1998) Echolocation, olfac- about roosts in female Bechstein’s bats: an experimental
tion, and fruit display: how bats find fruit of flagellichor- field study. Proc. R. Soc. Lond. B. 270, 511–515.
ous cucurbits. Funct. Ecol. 12, 364–372. Kerth, G., Wagner, M., and Konig, B. (2001a) Roosting
Kalko, E.K.V. and Schnitzler, H.-U. (1993) Plasticity in together, foraging apart: information transfer about
echolocation signals of European pipistrelle bats in food is unlikely to explain sociality in female Bechstein’s
search flight: implications for habitat use and prey detec- bats (Myotis bechsteinii). Behav. Ecol. Sociobiol. 50,
tion. Behav. Ecol. Sociobiol. 33, 415–428. 283–291.
Kallen, F.C. (1977) The cardiovascular system of bats: Kerth, G., Weissman, K., and König, B. (2001b) Day roost
structure and function. In: The Biology of Bats, Vol. III, selection in female Bechstein’s bats (Myotis bechsteinii): a
edited by W.A. Wimsatt, pp 289–483. Academic Press, field experiment to determine the influence of roost tem-
New York. perature. Oecologia 126, 1–9.
Keeley, B. and Tuttle, M. (1999) Bats in American bridges. Kick, S.A. (1982) Target detection by the echolocating bat,
In Proceedings of the Third International Conference on Eptesicus fuscus. J. Comp. Physiol. 145, 431–435.
Wildlife Ecology and Transportation, edited by GL Evink Kick, S.A. and Simmons, J.A. (1984) Acoustic gain control
et al., 167–172. FL-ER-73–99. Tallahassee, Florida in the bat’s sonar receiver and the neuroethology of
Department of Transportation. echolocation. J. Neurosci. 4, 2725–2737.
Kelm, D.H., Wiesner, K.R,. and Helversen, O. (2008) Effects Kingdon, J. (1974) East African Mammals, Vol. IIA (Insecti-
of artificial roosts for frugivorous bats on seed dispersal vores and bats). Academic Press, London.
in a neotropical forest pasture mosaic. Conserv. Biol. 22, Kingston, T. (2010) Research priorities for bat conservation
733–741. in Southeast Asia: a consensus approach. Biodivers. Con-
Kerth, G., Ebert, C., and Schmidtke, C. (2006) Group deci- serv. 19, 471–484.
sion making in fission–fusion societies: evidence from Kingston, T., Jones, G., Akbar, Z., and Kunz, T.H. (2003)
two-field experiments in Bechstein’s bats. Proc. R. Soc. Alternation of echolocation calls in five species of aerial-
Lond. B. 273, 2785–2790. feeding insectivorous bats from Malaysia. J.Mammal. 84,
Kerth, G., Kiefer, A., Trappmann, C,. and Weishaar, M. 205–215.
(2003) High gene diversity at swarming sites suggest Kingston, T., Lara, M.C., Jones, G., Akbar, Z., Kunz, T.H.,
hot spots for gene flow in the endangered Bechstein’s and Schneider, C.J. (2001) Acoustic divergence in two
bat. Conserv. Genet. 4, 491–499. cryptic Hipposideros species: a role for social selection?
Kerth, G. and König, B. (1999) Fission, fusion and nonran- Proc. R. Soc. Lond. B. 268, 1381–1386.
dom associations in female Bechstein’s bats (Myotis bech- Kingston, T. and Rossiter, S.J. (2004) Harmonic-hopping in
steinii). Behaviour 136, 1187–1202. Wallacea’s bats. Nature 429, 654–657.
300 REFERENCES

Kirkpatrick A. (2008) Psychovertical. Hutchinson, London. Kurta, A. and Kunz, T.H. (1987) Size of bats at birth and
Kleiber, M. (1932) Body size and metabolism. Hilgardia 6, maternal investment during pregnancy. Symp. Zool. Soc.
315–353. Lond. 57, 79–106.
Kneitel, J.M. and Chase, J.M. (2004) Trade-offs in commu- Kurta, A., and T.H. Kunz. (1988) Roosting metabolic rate
nity ecology: linking spatial scales and species coexis- and body temperature of male little brown bats (Myotis
tence. Ecol. Lett. 7, 69–80. lucifugus) in summer. J. Mammal. 69, 645–651.
Kobler, J.B., Wilson, B.S., Henson, O.W., and Bishop, A.L. Kürten, L. and Schmidt, U. (1982) Thermoperception in
(1985) Echo intensity compensation by echolocating the common vampire bat (Desmodus rotundus). J. Comp.
bats. Hearing. Res. 20, 99–108. Physiol. A 146, 223–228.
Korine, C. and Kalko, E.K.V. (2005) Fruit detection and LaVal, R.K. and Fitch, H.S. (1977) Structure, movement and
discrimination by small fruit-eating bats (Phyllostomi- reproduction in three Costa Rican bat communities.
dae): echolocation call design and olfaction. Behav. Ecol. Occas. Pap. Mus. Nat. Hist. Univ. Kans. 69, 1–28.
Sociobiol. 59, 12–23. Lacki, M.J., Hayes, J.P., and Kurta, A. (2007) Bats in Forests:
Krishna, A. (1978) Aspects of reproduction in some Indian Conservation and Management. John Hopkins University
bats. Ph.D. thesis, Banaras Hindu University, Varanasi. Press, Baltimore.
Krzanowski, A. (1961) Weight dynamics of bats wintering Lande, R. (1987) Extinction thresholds in demographic
in the cave at Pulawy (Poland) Acta Theriol. 4, 249–264. models of territorial populations. Am. Nat. 130, 624–635.
Kulzer, E. (1967) Die Hertztätigkeit bei lethargischen Laurance, S.G.W., Stouffer, P.C., and Laurance, W.F. (2004)
und winterschlafenden fledermäusen. Zeitschr. f. vergl. Effects of road clearings on movement patterns of under-
Physiol. 56, 63–94. story rainforest birds in central Amazonia. Conserv. Biol.
Kulzer, E. and Storf, R. (1980) Schlaf-lethargie bei 18, 1099–1109.
dem afrikanischen langzungenflughund Megaloglossus Lausen, C.L. and Barclay, R.M.R. (2002) Roosting beha-
woermanni Pagenstecher. Z. Saugetierk. 45, 23–29. viour and roost selection of female big brown bats (Epte-
Kunz, T.H. (1973) Population studies of the cave bat sicus fuscus) roosting in rock crevices in south-eastern
(Myotis velifer): Reproduction, growth and development. Alberta. Can. J. Zool. 80, 1069–1076.
Occas. Pap. Mus. Nat. Hist. Univ. Kans. 15, 1–43. Lausen, C.L. and Barclay, R.M.R. (2003) Thermoregulation
Kunz, T.H. (1982) Roosting ecology. In: The Ecology of Bats, and roost selection by reproductive female big brown
edited by T.H. Kunz, pp 1–55. Plenum Press, New York. bats (Eptesicus fuscus) roosting in rock crevices. J. Zool.
Kunz, T.H. (1987) Post-natal growth and energetics of Lond. 260, 235–244.
suckling bats. In: Recent Advances in the Study of Bats, Laval, R.K. (2004) Impact of global warming and
edited by M.B. Fenton, P.A. Racey, and J.M.V. Rayner, locally changing climate on tropical cloud forest bats.
pp 395–420. Cambridge University Press, Cambridge. J. Mammal. 85, 237–244
Kunz, T.H., Arnett, E.B., Cooper, B.M., Erickson, W.P., Law, B.S., Anderson, J., and Chidel, M. (1999) Bat commu-
Larkin, R.P., Mabee, T., Morrison, M.L., Strickland, M.D., nities in a fragmented forest landscape on the south-west
and Szewczak, J.M. (2007) Assessing impacts of wind- slopes of New South Wales, Australia. Biol. Conserv. 88,
energy development on nocturnally active birds and 333–345.
bats: a guidance document. J. Wildlife Manage. 71, Lawlor, T.E. (1986) Comparative biogeography of mam-
2449–2486. mals on islands. Biol. J. Linn. Soc. 28, 99–125.
Kunz, T.H., August, P.V., and Burnett, C.D. (1983) Harem Lawrence, B.D. and Simmons, J.A. (1982a) Measurement of
social organisation in cave roosting Artibeus jamaicensis atmospheric attenuation at ultrasonic frequencies and
(Chiroptera: Phyllostomidae) Biotropica 15, 133–138. the significance for echolocation by bats. J. Acoust. Soc.
Kunz, T.H., and Diaz, C.A. (1995) Folivory in fruit-eating Am. 71, 585–590.
bats, with new evidence from Artibeus jamaicensis Lawrence, B.D. and Simmons, J.A. (1982b) Echolocation
(Chiroptera: Phyllostomidae). Biotropica 27, 106–120. in bats: the external ear and perception of the vertical
Kunz, T.H. and Ingalls, K.A. (1994) Folivory in bats: an position of targets. Science 218, 481–483.
adaptation derived from frugivory. Funct. Ecol. 8, Lawton, J.H. (1996) Patterns in ecology. Oikos 75, 145–147.
665–668. Lawton, J.H. (1999) Are there general laws in ecology?
Kunz, T.H. and Lumsden L.F. (2003) Ecology of cavity and Oikos 84, 177–192.
foliage roosting bats. In: Bat ecology, edited by T.H. Kunz Lee, Y.F. and Kuo, Y.M. (2001) Predation on Mexican
and M.B. Fenton, pp 3–89. University of Chicago Press, free-tailed bats by peregrine falcons and red-tailed
Chicago. hawks. J. Raptor Res. 35, 115–123.
 REFERENCES 301

Leibold, M.A., Holyoak, M., Mouquet, N., Amarasekare, Long, C.V., Flint, J.A., and Lepper, P.A. (2010) Insect attrac-
P., Chase, J.M., Hoopes, M.F., Holt, R.D., Shurin, J.B., tion to wind turbines: does colour play a role? Eur.
Law, R., Tilman, D., Loreau, M., and Gonzalez, A. (2004) J. Wildl. Res. DOI 10.1007/s10344-010-0432-7
The metacommunity concept: a framework for multi- Lookingbill, T.R., Elmore, A.J., Engelhardt, K.A.M.,
scale community ecology. Ecol.Lett. 7, 601–613. Churchill, J.B., Gates, J.E., and Johnson, J.B. (2010) Influ-
Lesinski, G. (2007) Bat road casualties and factors ence of wetland networks on bat activity in mixed-use
determining their number. Mammalia 71, 138–142. landscapes. Biol. Conserv. 143, 974–983.
Leuzinger, Y. and Brossard, C. (1994) Répartition de M. Losos, J.B. and Schluter, D. (2000) Analysis of an evolution-
daubentonii en fonction du sexes et de la période de ary species-area relationship. Nature 408, 847–850.
l’année dans le Jura Bernois. Résultats preliminaries. Lourenço, S.I. and Palmeirim, J.M. (2004) Influence of
Mitt. Natf. Ges. Schaffhausen 39, 135–143. temperature in roost selection by Pipistrellus pygmaeus
Lewis, S.E. (1992) Behaviour of Peter’s tent-making bat, (Chiroptera): relevance for design of bat-boxes. Biol.
Uroderma bilobatum, at maternity roosts in Costa Rica. Conserv. 119, 237–243.
J. Mammal. 73, 541–546. Lowe, A., Harris, S., and Ashton, P. (2004) Ecological Genetics.
Lewis, S.E. (1996) Low roost-site fidelity in pallid bats: Blackwell, Oxford.
associated factors and effect on group stability. Behav. Lundberg, K. and Gerell, R. (1986) Territorial advertisement
Ecol. Sociobiol. 39, 335–344. and mate attraction in the bat Pipistrellus pipistrellus.
Li, G., Wang, J.H., Rossiter, S.J., Jones, G., and Zhang, S.Y. Ethology 71, 115–124.
(2007) Accelerated FoxP2 evolution in echolocating bats. Lyons, S.K. and Willig, M.R. (2002) Species richness,
PLoS ONE 2, e900. latitude, and scale-sensitivity. Ecology, 83, 47–58.
Li, G., Wang, J.H., Rossiter, S.J., Jones, G., Cotton, J.A., and MacArthur, R.H. and Pianka, E.R. (1966) On optimal use of
Zhang, S.Y. (2008) The hearing gene Prestin reunites a patchy environment. Amer. Nat. 100, 603–609.
echolocating bats. PNAS 105, 13959–13964. MacArthur, R.H. and Wilson, E.O. (1967) The Theory of
Lillegraven, J.A. (1974) Biogeographical considerations of Island Biogeography. Princeton University Press, Prince-
the marsupial-placental dichotomy. Ann. Rev. Ecol. Syst. ton, N.J.
5, 263–283. MacGregor, J. (1993) Responses of winter populations
Link, A., Marimuthu, G., and Neuweiler, G. (1986) Move- of the federal endangered Indiana bat (Myotis sodalis) to
ment as a specific stimulus for prey catching behaviour cave gating in Kentucky. In 1991 National Cave Manage-
in rhinolophid and hipposiderid bats. J. Comp. Physiol. A. ment Symposium Proceedings, edited by D.L. Foster,
159, 403–413. pp. 364–370. American Cave Conservation Association,
Liu, Y., Cotton, J.A., Shen, B., Han, X., Rossiter, S.J., and Horse Cave, Kentucky, pp. 405.
Zhang, S. (2010) Convergent sequence evolution Mackey, R.L. and Barclay, R.M.R. (1989) The influence of
between echolocating bats and dolphins. Curr. Biol. 20, physical clutter and noise on the activity of bats over
R53–R54. water. Can. J. Zool. 67, 1167–1170.
Lloyd, B.D. (2003) The demographic history of the New MacSwiney, M.C., Frank, G., Clarke, M., and Racey, P.A.
Zealand short-tailed bat Mystacina tuberculata inferred (2008) What you see is not what you get: the role of
from modified control region sequences. Mol. Ecol. 12, ultrasonic detectors in increasing inventory complete-
1895–1911. ness in Neotropical bat assemblages. J. Appl. Ecol. 45,
Loehle, C. and Li, B.L. (1996) Habitat destruction and the 1364–1371.
extinction debt revisited. Ecol. Interact. 6, 784–789. Madhaven, A. (1978) Breeding habits and associated
Lomolino, M.V. (1984) Mammalian island biogeography: phenomena in some Indian bats. Part V, Pipistrellus
effects of area, isolation and vagility. Oecologia 61, dormeri (Dobson), Vespertilionidae. J. Bombay. Nat. Hist.
376–382. Soc. 75, 426–433.
Lomolino, M.V. (2000) Ecology’s most general, yet protean Maharadatunkamsi, Hisheh, S., Kitchener, D.J., and
pattern: the species-area relationship. J. Biogeogr. 27, Schmitt, L.H. (2000) Genetic and morphometric diversity
17–26. in Wallacea: geographical patterning in the horseshoe
Longcore, T. and Rich, C. (2004). Ecological light pollution. bat, Rhinolophus affinis. J. Biogeogr. 27, 193–201.
Front. Ecol. Envir. 2, 191–198. Maina, J.N. (2000) What it takes to fly: the structural
Long, C.V., Flint, J.A., Lepper, P.A., and Dible, S.A. (2009) and functional respiratory refinements in birds and
Wind turbines and bat mortality: interactions of bat bats. J. Exp. Biol. 203, 3045–3064.
echolocation pulses with moving turbine rotor blades. Makin, D. and Mendelssohn, H. (1987) The fruit-bat issue
Proc. Inst. Acous. 131, 185–192. in Israel. Bat News 11, 4–5.
302 REFERENCES

Marimuthu, G. and Neuweiler, G. (1987) The use of acous- McGinn, A.P. (2001) Malaria’s lethal grip tightens. In Vital
tical cues for prey detection by the Indian false vampire Signs 2001–2002. The Trends that are Shaping our Future,
bat, Megaderma lyra. J. Comp. Physiol. 160, 509–515. pp. 134–135. Earthscan Publications Ltd., London.
Marimuthu, G., Rajan, K.E., Kandula, S., Parsons, S., McGregor, S.E., Alcorn, S.M., and Olin, G. (1962) Pollination
and Jones, G. (2002) Effects of different surfaces on the and pollinating agents of the saguaro. Ecology 43, 259–267.
perception of prey-generated noise by the Indian false McKenzie, N. and Hall, L. (2008) Macroderma gigas. In:
vampire bat Megaderma lyra. Acta Chiropt. 4, 25–32. IUCN 2009. IUCN Red List of Threatened Species. Ver-
Marshall, A.G. (1983) Bats, flowers and fruit: evolutionary sion 2009.2. www.iucnredlist.org.
relationships in the Old World. Biol. J. Linn. Soc. 20, McKenzie, N.L. and Rolfe, J.K. (1986) Structure of bat
115–135. guilds in the Kimberley mangroves, Australia. J. Anim.
Martinkova, N., et al. (2010) Increasing incidence of Ecol. 55, 401–420.
Geomyces destructans fungus in bats from the Czech McManus, J.J. (1976) Thermoregulation. In: Biology of Bats
Republic and Slovakia. PLoS ONE 5: e13853. of the New World Family Phyllostomidae. Part I, edited by
Masters, W.M., Moffat, A.J.M., and Simmons, J.A. R.J. Baker, J.K. Jones, and D.C. Carter. Spec. Pubs. Mus.
(1985) Sonar tracking of horizontally moving targets Texas Tech. Univ. 13, 281–292. Texas Technical Press,
by the big brown bat, Eptesicus fuscus. Science 228, Texas.
1331–1333. McNab, B.K. (1969) The economics of temperature regulation
May, R.M. (1981) Models for two interacting populations. in Neotropical bats. Comp. Biochem. Physiol. 31, 227–268.
In: Theoretical Ecology: Principles and Applications. 2nd McNab, B.K. (1974a) The behaviour of temperate cave bats
edition, edited by R.M. May, pp 78–104. Blackwell in a subtropical environment. Ecology 55, 943–958.
Scientific Publications, Oxford. McNab, B.K. (1974b) The energetics of endotherms. Ohio J.
Mayer, F. (1995) Genetic population structure of the Sci. 74, 370–380.
noctule bat, Nyctalus noctula: a molecular approach and McNab, B.K. (1982) Evolutionary alternatives in the physi-
first results. Symposia of the Zoological Society of London 67, ological ecology of bats. In: The Ecology of Bats, edited by
387–396. T.H. Kunz, pp 151–200. Plenum Press, New York.
McCain, C.M. (2007a) Area and mammalian elevational McNab, B.K. (1992) A statistical analysis of mammalian
diversity. Ecology 88, 76–86. rates of metabolism. Funct. Ecol. 6, 672–679.
McCain, C.M. (2007b) Could temperature and water McNab, B.K. (2003) Standard energetics of phyllostomid
availability drive elevational species richness patterns? bats: the inadequacies of phylogenetic-contrast analyses.
A global case study for bats. Global Ecol. Biogeogr. 16, 1–13. Comp. Biochem. Physiol. A. 135, 357–368.
McConkey, K. R. and Drake, D. R. (2006) Flying foxes cease McWilliam, A.N. (1987) The reproductive and social biol-
to function as seed dispersers long before they become ogy of Coleura afra in a seasonal environment. In: Recent
rare. Ecology 87, 271–276. Advances in the Study of Bats, edited by M.B. Fenton,
McCracken, G.F. (1984) Communal nursing in Mexican P.A. Racey, and J.M.V. Rayner, pp 324–350. Cambridge
free-tailed bat maternity colonies. Science 223, 1090–1091. University Press, Cambridge.
McCracken, G.F. (1987) Genetic structure of bat social Medellín, R.A., Equihua, M., and Amin, M.A. (2000) Bat
groups. In: Recent Advances in the Study of Bats, edited diversity and abundance as indicators of disturbance in
by M.B. Fenton, P.A. Racey, and J.M.V. Rayner, Neotropical rainforests. Conserv. Biol. 14, 1666–1675.
pp. 281–298. Cambridge University Press, Cambridge. Medway, Lord. (1971) Observations of social and
McCracken, G.F. and Bradbury, J.W. (1977) Paternity and reproductive biology of the bent-winged bat Miniopterus
genetic heterogeneity in the polygynous bat, Phyllosto- australis in northern Borneo. J. Zool. Lond. 165, 261–273.
mus hastatus. Science 198, 303–306. Medway, Lord (1972) Reproductive cycles of the flat-
McCracken, G.F. and Bradbury, J.W. (1981) Social organi- headed bats Tylonycteris pachypus and T. robustula
sation and kinship in the polygynous bat, Phyllostomus (Chiroptera: Vespertilionidae) in a humid equatorial
hastatus. Behav. Ecol. Sociobiol. 8, 11–34. environment. J. Linn. Soc. Lond. Zool. 51, 33–61.
McCracken, G.F. and Gassell, M.F. (1997) Genetic structure Medway, Lord and Marshall, A.G. (1970) Roost-site
in migratory and non-migratory populations of Brazilian selection among flat-headed bats (Tylonycteris spp).
free-tailed bats. J. Mammal. 78, 348–357. J. Zool. Lond. 161, 237–245.
McCracken, G.F. and Wilkinson, GS. (2000) Bat mating Medway, Lord and Marshall, A.G. (1972) Roosting associ-
systems. In: Reproductive Biology of Bats edited by ation of flat-headed bats, Tylonycteris species (Chirop-
E.G. Crichton and P.H. Krutszch, pp 321–362. Academic tera: Vespertilionidae) in Malaysia. J. Zool. Lond. 168,
Press, New York. 463–482.
 REFERENCES 303

Menaker, M. (1964) Frequency of spontaneous arousal Möller, J. (1978) Response characteristics of inferior
from hibernation in bats. Nature 203, 540–541. colliculus neurons of the awake CF-FM bat Rhinolophus
Meng, J., Hu, Y., Wang, Y., Wang, X., and Li, C. (2006) ferrumequinum. II. Two tone stimulation. J. Comp. Physiol.
A Mesozoic gliding mammal from northeastern China. A. 125, 227–236.
Nature 444, 889–893. Moreno, C.E., Arita, H.T., and Solis, L. (2006) Morphological
Metheny, J.D., Kalcounis-Rueppell, M.C., Bondo, K.J., and assembly mechanisms in Neotropical bat assemblages
Brigham, R.M. (2008a) A genetic analysis of group and ensembles within a landscape. Oecologia 149, 133–140.
movement in an isolated population of tree-roosting Moreno, C.E. and Halffter, G. (2000) Assessing the com-
bats. Proc. R. Soc. B. 275, 2265–2272. pleteness of bat biodiversity inventories using species
Metheny, J.D., Kalcounis-Rueppell, M.C., Willis, C.K.R., accumulation curves. J. Appl. Ecol. 37, 149–158.
Kolar, K.A., and Brigham, R.M. (2008b) Genetic relation- Morrison, D.W. (1978) Foraging ecology and energetics of
ships between roost-mates in a fission–fusion society of the frugivorous bat, Artibeus jamaicensis. Ecology 59,
tree-roosting big brown bats (Eptesicus fuscus). Behav Ecol 716–723.
Sociobiol 62, 1043–1051. Morrison, D.W (1979) Apparent male defense of tree
Meyer, C.F.J., Fründ, J., Pineda Lizano, W., and Kalko, hollows in the fruit bat, Artibeus jamaicensis. J. Mammal.
E.K.V. (2008) Ecological correlates of vulnerability to 60, 11–15.
fragmentation in Neotropical bats. J. Appl. Ecol. 45, Morrison, D.W. and Hagen-Morrison, S. (1981) Economics
381–391. of harem maintenance by a neotropical bat. Ecology 62,
Meyer, C.F.J. and Kalko, E.K.V. (2008) Assemblage-level 864–866.
responses of phyllostomid bats to tropical forest frag- Moss, C.F., Bohn, K., Gilkenson, H., and Surlykke, A.
mentation: land-bridge islands as a model system. (2006) Active listening for spatial orientation in a com-
J. Biogeogr. 35, 1711–1726. plex auditory scene. PloS Biol 4, e79.
Michelson, A. and Larsen, O.N. (1985) Hearing and sound. Moss, C.F. and Sinha, S.R. (2003) Neurobiology of echolo-
In: Comprehensive Insect Physiology, Biochemistry and Phar- cation in bats. Curr. Opin. Neurobiol. 13, 751–758.
macology, Vol. 9, edited by G.A. Kerkut and L.I. Gilbert, Muchhala, N. (2006) Nectar bat stows huge tongue in its rib
pp 495–556. Pergamon Press, Oxford. cage. Nature 444, 701–702.
Mickleburgh, S.P., Hutson, A.M., and Racey, P.A. (1992) Muijres, F.T., Johansson, L.C., Barfield, R., Wolf, M., Sped-
Old World Fruit Bats: An Action Plan for their Conservation. ding, G.R., and Hedenstrom, A. (2008) Leading-edge
I.U.C.N./S.S.C., Gland, Switzerland. vortex improves lift in slow-flying bats. Science 319,
Mickleburgh, S.P., Hutson, A.M., and Racey, P.A. (2002) 1250–1253.
A review of the global conservation status of bats. Oryx Müller, B., Glösmann, M., Peichl, L., Knop, G.C., Hage-
36, 18–34. mann, C., and Ammermüller, J. (2009) Bat eyes have
Mickleburgh, S., Waylen, K., and Racey, P.A. (2009) Bats as ultraviolet-sensitive cone photoreceptors. PLoS ONE 4,
bushmeat: a global review. Oryx 43, 217–234. e6390.
Miller, L.A. (1975) The behaviour of flying green lacewings Müller, J. (1981) Fossil pollen records of extant angios-
in the presence of ultrasound. J. Insect Physiol. 121, perms. Bot. Rev. 47, 1–142.
205–219. Murphy, M. (1987) India’s Samanar Hill caves saved. Bats
Miller, L.A. (1991) Arctiid moth clicks can degrade the 5, 6.
accuracy of range difference discrimination of echolocat- Murphy, W.J., Eizirik, E., O’Brien, S.J., Madsen, O., Scally,
ing brown bats, Eptesicus fuscus. J. Comp. Physiol. A. 199, M., Douady, C.J., Teeling, E., Ryder, O.A., Stanhope, M.J.,
571–579. de Jong, W.W., and Springer, M.S. (2001) Resolution of
Miller-Butterworth, C.M., Jacobs, D.S., and Harley, E.H. the early placental mammal radiation using Bayesian
(2003) Strong population substructure is correlated phylogenetics. Science 294, 2348–2351.
with morphology and ecology in a migratory bat. Nature Muscarella, R. and Fleming, T.H. (2007) The role of frugiv-
424, 187–191. orous bats in tropical forest succession. Biol. Rev. 82,
Milne, D.J., Fisher, A., and Pavey, C.R. (2006) Models of the 573–590.
habitat associations and distributions of insectivorous Myers, P. (1977) Patterns of reproduction of four species of
bats of the Top End of the Northern Territory, Australia. vespertilionid bats in Paraguay. Univ. Calif. Publ. Zool.
Biol. Conserv. 130, 370–385. 107, 1–41.
Mittelbach, G.G., et al. (2007) Evolution and the latitudinal National Academy of Sciences. (2007) Environmental Im-
diversity gradient: speciation, extinction and biogeogra- pacts of Wind-Energy Projects. Washington, DC: National
phy. Ecol. Lett. 10, 315–331. Research Council.
304 REFERENCES

Nellis, D.W. and Ehle, C.P. (1977) Observations on the (flying foxes, Pteropus spp.): wing morphology and flight
behaviour of Brachyphylla cavernarum (Chiroptera) in performance. J. Exp. Biol. 203, 651–664.
the Virgin Islands. Mammalia, 41, 403–409. Norberg, U.M. and Fenton, M.B. (1988) Carnivorous bats?
Nelson, J.E. (1965) Behaviour of Australian Pteropodidae Biol. J. Linn. Soc. 33, 383–394.
(Megachiroptera). Anim. Behav. 13, 544–557. Norberg, U.M., Kunz, T.H., Steffensen, J.F., Winter, Y., and
Neuweiler, G. (1970) Neurophysiologische untersuchun- Helversen, O. (1993) The cost of hovering and forward
gen zum echoortungssystem der grossen hufeisennase, flight in a nectar-feeding bat, Glossophaga soricina, esti-
Rhinolophus ferrumequinum. Z. Zgl. Physiol. 67, 273–306. mated from aerodynamic theory. J. Exp. Biol. 182,
Neuweiler, G. (1990) Auditory adaptations for prey 207–227.
capture in echolocating bats. Physiol Rev. 70, 615–641. Norberg, U.M. and Rayner, J.M.V. (1987) Ecological mor-
Neuweiler, G. (2003) Evolutionary aspects of bat echoloca- phology and flight in bats (Mammalia; Chiroptera): wing
tion. J. Comp. Physiol. A. 189, 245–256. adaptations, flight performance, foraging strategy and
Newman, J. and Zillioux, E. (2009) Comparison of reported echolocation. Phil. Trans. Roy. Soc. Lond. B. 316, 335–427.
effects and risks to vertebrate wildlife from six electricity Novacek, M.J. (1985) Evidence for echolocation in the
generation types in the New York/New England region. oldest known bats. Nature 315, 140–141.
New York State Energy Research and Development Ober, H.K. and Hayes, J.P. (2008) Influence of vegetation
Authority, New York. on bat use of riparian areas at multiple spatial scales.
Nicholls, B. and Racey, P.A. (2007) Bats avoid radar J. Wildlife Manage. 72, 396–404.
installations: could electromagnetic fields deter bats O’Brien, J., Mariani, C., Olson, L., Russell, A.L., Say,
from colliding with wind turbines? PLoS One 2, e297. L., Yoder, A.D. and Hayden, T.J. (2009) Multiple coloni-
Nicolai, V. (1986) The bark of trees: thermal properties, sations of the western Indian Ocean by Pteropus fruit bats
microclimate and fauna. Oecologia 69, 148–160. (Megachiroptera: Pteropodidae): the furthest islands
Nitecki, M.H. (1983) Coevolution. University of Chicago were colonised first. Mol. Phylogenet. Evol. 51 294–303.
Press, Chicago. O’Donnell, C. (2008a) Mystacina tuberculata. In: IUCN 2009.
Niu, H., Wang, N., Zhoa, L., and Liu, J. (2007) Distribution IUCN Red List of Threatened Species. Version 2009.2.
and underground habitats of cave-dwelling bats in www.iucnredlist.org.
China. Anim. Conserv. 10, 470–477. O’Donnell, C. (2008b) Mystacina robusta. In: IUCN 2009.
Nogales, M., Martín, A., Tershy, B.A., Donlan, C.J., Veitch, D., IUCN Red List of Threatened Species. Version 2009.2.
Puerta, N., Wood, B., and Alonso, J. (2004) A review of www.iucnredlist.org.
feral cat eradication on islands. Conserv. Biol. 18, 310–319. O’Farrell, M.J. and Studier, E.H. (1970) Fall metabolism in
Nogueira, M.R. and Peracchi, A.L. (2003) Fig-seed preda- relation to ambient temperatures in three species of Myo-
tion by two species of Chiroderma: discovery of a new tis. Comp. Biochem. Physiol. 35, 697–703.
feeding strategy in bats. J. Mammal. 84, 225–233. O’Farrell, M.J. and Studier, E.H. (1973) Reproduction,
Norberg, U.M. (1970) Functional osteology and myology of growth and development in Myotis thysanodes and
the wing of Plecotus auritus Linnaeus (Chiroptera). Arkiv M. lucifugus (Chiroptera: Vespertilionidae). Ecology 54,
for Zoologi 22, 483–543. 18–30.
Norberg, U.M. (1972) Bat wing structures important O’Neill, W.E. (1987) The processing of temporal informa-
for aerodynamics and rigidity (Mammalia, Chiroptera). tion in the auditory systems of echolocating bats. In:
Z. Morphol. Tiere 73, 45–61. Recent Advances in the Study of Bats, edited by M.B. Fenton,
Norberg, U.M. (1976) Some advanced flight manoeuvres of P.A. Racey, and J.M.V. Rayner, pp 171–199. Cambridge
bats. J. Exp. Biol. 64, 489–495. University Press, Cambridge.
Norberg, U.M. (1985) Evolution of vertebrate flight: an O’Neill, W.E., Schuller, G., and Padke-Schuller, S. (1985)
aerodynamic model for the transition from gliding to Functional and anatomical similarities in the auditory
active flight. Am. Nat. 126, 303–327. cortices of the Old World horseshoe bat and neotropical
Norberg, U.M. (1987) Wing form and flight mode in bats. moustached bat for processing of similar biosonar
In: Recent Advances in the Study of Bats, edited by signals. Assoc. Res. Otolaryn. 193, 148.
M.B. Fenton, P.A. Racey, and J.M.V. Rayner, pp 43–56. O’Neill, W.E. and Suga, N. (1982) Encoding of target range
Cambridge University Press, Cambridge. and its representation in the auditory cortex of the mous-
Norberg, U.M. (1990) Vertebrate Flight. Springer-Verlag, tached bat. J. Neurosci. 2, 17–31.
Berlin. Oprea, M., Mendes, P., Vieira, T.B., and Ditchfield, A.D.
Norberg, U.M, Brooke, A.P., and Trewhella, W.J. (2000) (2009) Do wooded streets provide connectivity for bats
Soaring and non-soaring bats of the family Pteropodidae in an urban landscape? Biodivers. Conserv. 18, 2361–2371.
 REFERENCES 305

O’Shea, T.J. (1980) Roosting, social organisation and the Park, K.J., Altringham, J.D., and Jones, G. (1996) Assortative
annual cycle in a Kenya population of the bat Pipistrellus mating in two phonic types of the bat Pipistrellus
nanus. Z. Tierpsychol. 53, 171–195. pipistrellus: further evidence for two cryptic species.
O’Shea, T.J. and Vaughan, T.A. (1977) Nocturnal and Proc. Roy. Soc. B. 263, 1495–1499.
seasonal activities of the pallid bat, Antrozous pallidus. Park, K.J., and Cristinacce, A. (2006) Use of sewage treat-
J. Mammal. 58, 269–284. ment works as foraging sites by insectivorous bats.
O’Shea, T.J. and Vaughan, T.A. (1980) Ecological observa- Anim. Conserv. 9, 259–268.
tions on an East African bat community. Mammalia 44, Park, K.J., Jones, G., and Ransome, R.D. (1999) Winter
485–496. activity of a population of greater horseshoe bats
Obrist, M.K., Fenton, M.B., Eger, J.L., and Schlegel, P.A. (Rhinolophus ferrumequinum) J. Zool. Lond. 248, 419–427.
(1993) What ears do for bats: a comparative study Park, K.J., Jones, G., and Ransome, R.D. (2000) Torpor,
of pinna sound pressure transformation in Chiroptera. arousal and activity of hibernating Greater Horseshoe
J. Exp. Biol. 180, 119–152. Bats (Rhinolophus ferrumequinum). Funct. Ecol. 14,
Olsen, J.F. and Suga, N. (1986) The auditory thalamus of 580–588.
the moustached bat: convergent input and coincidence Park, K.J, Masters, E., and Altringham, J.D. (1998) Social
of excitation from orientation sound and echo. Neurosci. structure and mating patterns of three sympatric species
Abstr. 12, 1272. of vesper bats. J. Zool. Lond. 244, 379–389.
Orr, R.T. (1970) Development: prenatal and postnatal. Parris, K.M., and Hazell, D.L. (2005) Biotic effects of
In: Biology of Bats Vol. I, edited by W.A. Wimsatt, climate change in urban environments: The case of
pp 217–231. Academic Press, New York. the grey-headed flying-fox (Pteropus poliocephalus) in
Ortega, J. and Arita, H.T. (2000) Defence of females by Melbourne, Australia. Biol. Conserv. 124, 267–276.
dominant males of Artibeus jamaicensis (Chiroptera: Parsons, K.N., Jones, G., Davidson-Watts, I., and Green-
Phyllostomidae). Ethology 106, 395–407. away, F. (2003) Swarming of bats at underground sites in
Ortega, J. and Arita, H.T. (2002) Subordinate males in Britain—implications for conservation. Biol. Conserv.
harem groups of Jamaican fruit-eating bats (Artibeus 111, 63–70.
jamaicensis): satellites or sneaks? Ethology 108, 1077–1091. Patriquin, K.J. and Barclay, R.M.R. (2003) Foraging by
Ortega, J., Maldonado, J.E., Wilkinson, G.S., Arita, H.T., bats in cleared, thinned and unharvested boreal forest.
and Fleischer, R.C. (2003) Male dominance, paternity, J. Appl. Ecol. 40, 646–657.
and relatedness in the Jamaican fruit-eating bat (Artibeus Patriquin, K.J., Leonard, M.L., Broders, H.G., and
jamaicensis). Mol. Ecol. 12, 2409–2415. Garroway, C.J. (2010) Do social networks of female
Page, R.A. and Ryan, M.J. (2005) Flexibility in assessment northern long-eared bats vary with reproductive
of prey cues: frog-eating bats and frog calls. Proc. R. Soc. period and age? Behav. Ecol. Sociobiol. 64, 899–913.
Lond. B. 272, 841–847. Patterson, B.D., Stotz, D.F., Solari, S., Fitzpatrick, J.W., and
Page, R.D.M. and Holmes, E.C. (1998) Molecular Evolution: Pacheco, V. (1998) Contrasting patterns of elevational
A Phylogenetic Approach. Blackwell, Oxford. zonation for birds and mammals in the Andes of south-
Pagels, J.F. (1975) Temperature regulation, body weight eastern Peru. J. Biogeogr. 25, 593–607.
and changes in total body fat of the free-tailed bat, Patterson, B.D., Willig, M.R., and Stevens, R.D. (2003)
Tadarida brasiliensis cynocephala (LeConte). Comp. Bio- Trophic strategies, niche partitioning and patterns of
chem. Physiol. 50, 237–246. ecological organisation. In: Bat Ecology, edited by
Papadatou, E., Butlin, R.K., and Altringham, J.D. (2008) T.H. Kunz and M.B. Fenton, pp 536–579. University of
Seasonal roosting habits and population structure of the Chicago Press, Chicago.
long-fingered bat Myotis capaccinii in Greece. J. Mammal. Patterson, C. (1987) Introduction. In: Molecules and
89, 503–512. Morphology in Evolution: Conflict or Compromise?, edited
Papadatou, E., Butlin, R.K., Pradel, R., and Altringham, J. by C. Patterson, pp 1–22. Cambridge University Press,
D. (2009) Sex-specific roost movements and population Cambridge.
dynamics of the vulnerable long-fingered bat, Myotis Pierson, E.D. and Rainey, W.E. (1992) The biology of flying
capaccinii. Biol. Conserv. 142, 280–289. foxes of the genus Pteropus: a review. In Pacific Island
Papadatou, E., Ibáñez, C., Pradel, R., Juste, J., and Gimenez, Flying Foxes: Proceedings of an International Conservation
O. (2011) Assessing survival in a multi-population Conference, eds. D.E. Wilson and G.L. Graham. pp. 1–17.
system: a case study on bat populations. Oecologia, 165, US Fish and Wildlife Service Biological Report 90(23). US
925–933. Department of the Interior, Fish and Wildlife Service,
Washington DC.
306 REFERENCES

Pierson, E.D., Sarich, V.M., Lownstein, J.M., Daniel, M.J., Pettigrew, J.D. (1991) A fruitful, wrong hypothesis?
and Rainey, W.E. (1986) A molecular link between the Response to Baker, Novacek and Simmons. Syst. Zool.
bats of New Zealand and South America. Nature 323, 40, 231–239.
60–63. Pettigrew, J.D. (1994) Flying DNA. Curr. Biol. 4, 277–280.
Pennycuick, C.J. (1971) Gliding flight of the dog-faced bat, Pettigrew, J.D. (1995) Flying primates: crashed or crashed
Rousettus aegyptiacus, observed in a wind tunnel. J. Exp. through? In: Ecology, Evolution and Behaviour of Bats,
Biol. 55, 833–845. edited by P.A. Racey, and S.W. Swift, Symp. Zool. Soc.
Pennycuick, C.J. (1972) Animal Flight. Edward Arnold, Lond. 67, pp 3–26. Oxford University Press, Oxford.
London. Pettigrew, J.D., Jamieson, B.G.M., Robson, S.K., Hall, L.S.,
Pestano, J., Brown, R.P., Suárez, N.M., Benzal, J., and McAnally, K.I., and Cooper, H.M. (1989) Phylogenetic
Fajardo, S. (2003a) Intraspecific evolution of Canary relations between microbats, megabats and primates
Island Plecotine bats, based on mtDNA sequences. (Mammalia: Chiroptera and Primates). Phil. Trans. Roy.
Heredity 90, 302–307. Soc. Lond. B. 325, 489–559.
Pestano, J., Brown, R.P., Suárez, N.M., and Fajardo, S. Phillips, W.R. (1990) Priorities for Ghost Bat (Macroderma
(2003b) Phylogeography of pipistrelle-like bats within gigas) Conservation and Management. Australian National
the Canary Islands, based on mtDNA sequences. Mol. Parks and Wildlife Service, Canberra.
Phylogenet. Evol. 26, 56–63. Pierson, E.D. and Rainey, W.E. (1992) The biology of flying
Peterson, R.L. (1965) A review of the flat-headed bats of the foxes of the genus Pteropus: a review. In Pacific Island
family Molossidae from South America and Africa. Life Flying Foxes: Proceedings of an International Conservation
Sci. Contr. R. Ont. Mus. 64, 3–32. Conference, eds. D.E. Wilson and G.L. Graham. pp. 1–17.
Petit, E., Balloux, F., and Goudet, J. (2001) Sex-biased US Fish and Wildlife Service Biological Report 90(23). US
dispersal in a migratory bat: a characterization using sex- Department of the Interior, Fish and Wildlife Service,
specific demographic parameters. Evolution 55, 635–640. Washington DC.
Petit, E., Excoffier, L., and Mayer, F. (1999) No evidence of Pierson, E.D., Sarich, V.M., Lownstein, J.M., Daniel, M.J., and
bottleneck in the post-glacial recolonization of Europe Rainey, W.E. (1986) A molecular link between the bats of
by the noctule bat (Nyctalus noctula). Evolution 53, New Zealand and South America. Nature 323, 60–63.
1247–1258. Piksa, K. (2008) Swarming of Myotis mystacinus and other
Petit, E. and Mayer, F. (1999) Male dispersal in the noctule bat species at high elevation in the Tatra Mountains,
bat (Nyctalus noctula): where are the limits? Proc. R. Soc. southern Poland. Acta Chiropt. 10, 69–79.
Lond. B. 266, 1717–1722. Pint, J.J. (1994) Who cares about Mexican bats? National
Petit, E. and Mayer, F. (2000) A population genetic analysis Speleological Society News 52, 94.
of migration: the case of the noctule bat (Nyctalus noc- Pitnick, S., Jones, K.E., and Wilkinson, G.S. (2006) Mating
tula). Mol. Ecol. 9, 683–690. systems and brain size in bats. Proc. R. Soc. Lond. B. 273,
Petit, S., Rojer, A., and Pors, L. (2006) Surveying bats 719–724.
for conservation: the status of cave-dwelling bats on Pollak, G.D. (1993) Some comments on the proposed per-
Curaçao from 1993 to 2003. Anim. Conserv. 9, 207–217. ception of phase and nanosecond time disparities by
Petri, B., Paabo, S., Haeseler, A., and Tautz, D. (1997) echolocating bats. J. Comp. Physiol. A. 172, 523–531.
Paternity assessment and population subdivision in a Pollak, G.D., Bodenhamer, R., Marsh, D.S., and Souther, A.
natural population of the larger mouse-eared bat Myotis (1977a) Recovery cycles of single neurons in the inferior
myotis. Mol. Ecol. 6, 235–242. colliculus of unanaesthetised bats obtained with
Petrzelkova, K.J. and Zukal, J. (2003) Does a live barn owl frequency-modulated and constant frequency sounds,
(Tyto alba) affect emergence behaviour of serotine bats J. Comp. Physiol. B. 120, 215–50.
(Eptesicus serotinus)? Acta Chiropt. 5, 177–184. Pollak, G.D. and Casseday, J.H. (1989) The Neural Basis of
Pettersson, S., Ervik, F., and Knudsen, J.T. (2004) Floral Echolocation in Bats. Zoophysiology, Vol. 25. Springer-
scent of bat-pollinated species: West Africa vs. the New Verlag, Berlin.
World. Biol. J. Linn. Soc. 82, 161–168. Pollak, G.D., Marsh, D.S., Bodenhamer, R., and Souther, A.
Pettigrew, J.D. (1995) Echolocation. In: A Dictionary of Birds, (1977b) Characteristics of phasic on neurons in inferior
edited by B. Campbell and E. Lack T. and A.D. Poyser, colliculus of unanaesthetised bats with observations
Calton. relating to mechanisms for echo ranging. J. Neurophys.
Pettigrew, J.D. (1986) Flying primates? Megabats have the 45, 208–26.
advanced pathway from eye to midbrain. Science 231, Pomeroy, D. (1990) Why fly? The possible benefits for
1306–1306. lower mortality. Biol. J. Linn. Soc. 40, 53–65.
 REFERENCES 307

Popa-Lisseanu, A.G., Bontadina, F., Mora, O., and Ibáñez, C. Quesada, M., Stoner, K.E., Rosas-Guerrero, V., Palacios-
(2008) Highly structured fission-fusion societies in an Guevara, C., and Lobo, J.A. (2003) Effects of habitat dis-
aerial-hawking, carnivorous bat. Anim. Behav. 75, 471–482. ruption on the activity of nectarivorous bats (Chiroptera:
Popa-Lisseanu, A.G., Delgado-Huertas, A., Forero, M.G., Phyllostomidae) in a dry tropical forest: implications for
Rodríguez, A., Arlettaz, R,. and Ibáñez, C. (2008) Bats’ the reproductive success of the neotropical tree Ceiba
conquest of a formidable foraging niche: the myriads of grandiflora. Oecologia 135, 400–406.
nocturnally migrating songbirds. PLoS One 2, e205. Rabinowitz, A. and Tuttle, M.D. (1980) Status of summer
Popa-Lisseanu, A.G. and Voigt, C.C. (2009) Bats on the colonies of the endangered gray bat in Kentucky. J. Wildlife
move. J. Mammal. 90, 1283–1289. Manage. 44, 955–960.
Pottie, S.A., Lane, D.J.W., Kingston, T., and Lee, B.P.Y.H. Racey, P.A. (1969) Diagnosis of pregnancy and experimen-
(2005) The microchiropteran bat fauna of Singapore. Acta tal extension of gestation in the pipistrelle bat, Pipistrellus
Chiropt. 7, 237–247. pipistrellus. J. Reprod. Fertil. 19, 465–474.
Pounds, J. A., Fogden, M. P. L., and Campbell, J. H. (1999) Racey, P.A. (1973) Environmental factors affecting the
Biological response to climate change on a tropical length of gestation in heterothermic bats. J. Reprod. Fertil.
mountain. Nature 398, 611–615. Suppl. 19, 175–189.
Prendergast, B.J., Freeman, D.A., Zucker, I., and Nelson, R.J. Racey, P.A. (1982) Ecology of bat reproduction. In: The
(2002) Periodic arousal from hibernation is necessary Ecology of Bats, edited by T.H. Kunz, pp 57–104. Plenum
for initiation of immune responses in ground squirrels. Press, New York.
Am. J. Physiol. 282, R1054–R1062. Racey, P.A. and Swift, S.M. (1981) Variations in gestation
Presley, S.J., Higgins, C.L., López-González, C., and length in a colony of pipistrelle bats (Pipistrellus pipistrel-
Stevens, R.D. (2009) Elements of metacommunity struc- lus) from year to year. J. Reprod. Fertil. 61, 123–129.
ture of Paraguayan bats: multiple gradients require anal- Racey, P.A. and Swift, S.M. (1985) Feeding ecology of
ysis of multiple ordination axes. Oecologia 160, 781–793. Pipistrellus pipistrellus (Chiroptera: Vespertilionidae)
Presley, S.J., and Willig, M.R. (2008) Composition and during pregnancy and lactation. I. Foraging behaviour.
structure of Caribbean bat (Chiroptera) assemblages: J. Anim. Ecol. 54, 205–215.
effects of inter-island distance, area, elevation and Racey, P.A., Swift, S.M., Rydell, J., and Brodie, L. (1998)
hurricane-induced disturbance. Global Ecol. Biogeogr. 17, Bats and insects over two Scottish rivers with contrasting
747–757. nitrate status. Anim. Conserv. 1, 195–202.
Presley, S.J., Willig, M.R., Wunderle, J.M., and Saldanha, L.N. Raghuram, H. and Marimuthu, G. (2007) Maternal feeding
(2008) Effects of reduced-impact logging and forest of offspring with vertebrate prey in captive Indian false
physiognomy on bat populations of lowland Amazo- vampire bat, Megaderma lyra. Acta Chiropt. 9, 437–443.
nian forest. J. Appl. Ecol. 45, 14–25. Rahbek, C. (1997) The relationship among area, elevation
Price, J.J., Johnson, K.P., and Clayton, D.H. (2004) The and regional species richness in tropical birds. Am. Nat.
evolution of echolocation in swiftlets. J. Avian Biol. 35, 149, 875–902.
135–143. Raikow, R.J. (1985) Locomotor system. In: Form and Func-
Promislow, D.E.L. and Harvey, P.H. (1990) Living fast tion in Birds, edited by A.S. King and J. McLelland.
and dying young: a comparative analysis of life history pp 57–147. Academic Press, London.
variation among mammals. J. Zool. Lond. 220, 417–437. Rainey, W.E. (1998) Conservation of bats on remote Indo-
Pryde, M.A., O’Donnell, C.F.J., and Barker, R.J. (2005) Fac- Pacific islands. In Bat Biology and Conservation, edited by
tors influencing survival and long-term population via- T.H. Kunz and P.A. Racey, pp. 326–341. Smithsonian
bility of New Zealand long-tailed bats (Chalinolobus Insitution Press, Washington DC.
tuberculatus): Implications for conservation. Biol. Conserv. Ransome, R.D. (1968) The distribution of the greater horse-
126, 175–185. shoe bat, Rhinolophus ferrumequinum, during hibernation,
Puechmaille, S.J., Verdeyroux, P., Fuller, H., Ar Gouilh, M., in relation to environmental factors. J. Zool. Lond. 154,
Bekaert, M., and Teeling, E.C. (2010) White-nose syn- 77–112.
drome fungus (Geomyces destructans) in bat, France. Ransome, R.D. (1971) The effect of ambient temperature on
Emerg. Infect. Dis. 16, 290–293. the arousal frequency of the greater horseshoe bat,
Pugh, M. and Altringham, J.D. (2005) The effects of gates Rhinolophus ferrumequinum, in relation to site selection
on cave entry by swarming bats. Acta Chiropt. 7, 293–300. and the hibernation state. J. Zool. Lond. 164, 353–371.
Pullin, A.S. (2002) Conservation Biology. Cambridge Univer- Ransome, R.D. (1973) Factors affecting the timing of births
sity Press, Cambridge, UK. of the greater horseshoe bat (Rhinolophus ferrumequinum)
Period. Biol. 75, 169–175.
308 REFERENCES

Ransome, R.D. (1990) The Natural History of Hibernating woodland. III. Reduction of density in relation to the
Bats. Christopher Helm, London. proximity of main roads. J. Appl. Ecol. 32, 187–202.
Ratcliffe, J.M. and Nydam, M.L. (2008) Multimodal Reijnen, R., Foppen, R., and Meeuwsen, H. (1996) The
warning signals for a multiple predator World. Nature effects of car traffic on the density of breeding birds in
455, 96–100. Dutch agricultural grasslands. Biol. Conserv. 75, 255–260.
Ratcliffe, J.M., Raghuram, H., Marimuthu, G., Fullard, J.H,. Reite, O.B. and Davis, W.H. (1966) Thermoregulation in
and Fenton, M.B. (2005) Hunting in unfamiliar space: bats exposed to low ambient temperatures. Proc. Soc.
echolocation in the Indian false vampire bat, Megaderma Exp. Biol. Med. 121, 1212–1215.
lyra, when gleaning prey. Behav. Ecol. Sociobiol. 58, Rhodes, M., Wardell-Johnson, G.W., Rhodes, M.P., and
157–164. Raymond, B. (2006) Applying network analysis to the
Rayner, J.M.V. (1981) Flight adaptations in vertebrates. In: conservation of habitat trees in urban environments: a
Vertebrate Locomotion, edited by M.H. Day, pp 137–171. case study from Brisbane, Australia. Conserv. Biol. 20,
Academic Press, London 861–870.
Rayner, J.M.V. (1986) Vertebrate flapping flight mechanics Rich, C. and Longcore, T. (2006) Ecological Consequences of
and aerodynamics, and the evolution of flight in bats. Artificial Night Lighting. Island Press, Washington, DC.
In: Biona Rep. 5, Bat flight-Fledermausflug, edited by Richardson, E.G. (1977) The biology and evolution of
W. Nachtigall, pp 27–74. Gustav Fischer Verlag, Stuttgart. the reproductive cycle in Miniopterus schreibersii and
Rayner, J.M.V. (1987) The mechanics of flapping flight in M. australis (Chiroptera: Vespertiliondae). J. Zool. Lond.
bats. In: Recent Advances in the Study of Bats, edited by 183, 353–375.
M.B. Fenton, P.A. Racey, and J.M.V. Rayner, pp 23–42. Richter, A.R., Humphrey, S.R., Cope, J.B., and Brack, V.
Cambridge University Press, Cambridge. (1993) Modified cave entrances: thermal effects of body
Rayner, J.M.V. (1992) Avian flight evolution and the prob- mass and resulting decline of endangered Indiana bats
lem of Archaeopteryx. In: Biomechanics and Evolution, (Myotis sodalis). Conserv. Biol. 7, 407–415.
edited by J.M.V. Rayner and R. Wootton, pp 183–212. Richter, H.V. and Cumming, G.S. (2006) Food availability
S.E.B. Symposium. Cambridge University Press, and annual migration of the straw-colored fruit bat
Cambridge. (Eidolon helvum). J. Zool. Lond. 268, 35–44.
Rayner, J.M.V., Jones, G., and Thomas, A. (1986) Vortex Rickart, E.A., Heideman, P.D., and Utzurrum, R.C.B. (1989)
flow visualisations reveal change in upstroke function Tent-roosting by Scotophilus kuhlii. J. trop. Ecol. 3, 433–436.
with flight speed in bats. Nature 321, 162–164. Ricklefs, R.E. and Lovette, I.J. (1999) The roles of island
Rayner, M.J., Hauber, M.E., Imber, M.J., Stamp, R.K., and area per se and habitat diversity in the species–area
Clout, M.N. (2007) Spatial heterogeneity of mesopreda- relationships of four Lesser Antillean fauna groups.
tor release within an oceanic island system. PNAS 104, J. Anim. Ecol. 68, 1142–1160.
20862–20865. Ridley, M. (1986) Evolution and Classification: The Reforma-
Rebelo, H., Tarroso, P., and Jones, G. (2010) Predicted tion of Cladism. Longman, London.
impact of climate change on European bats in relation Rigby, E.L. (2010) Population ecology of Daubenton’s bats.
to their biogeographic patterns. Glob. Change Biol. 16, PhD thesis, University of Leeds, UK.
561–576. Riley, S.P.D., Pollinger, J.P., Sauvajot, R.M., York, E.C.,
Reckardt, K. and Kerth, G. (2007) Roost selection and roost Bronley, C., Fuller T.K., and Wayne, R.K. (2006) A south-
switching of female Bechstein’s bats (Myotis bechsteinii) ern California freeway is a physical and social barrier to
as a strategy of parasite avoidance. Oecologia 154, gene flow in carnivores. Mol. Ecol. 15, 1733–1741.
581–588. Riiters, K.H. and Wickham, J.D. (2003) How far to the
Reijnen, R. and Foppen, R. (1994) The effects of car traffic nearest road? Frontiers in Ecology and the Environment 1,
on breeding bird populations in woodland. I. Evidence 125–129.
of reduced habitat quality for willow warblers (Phyllos- Rivers, N.M. (2005) Seasonal changes in population
copus trochilus) breeding close to a highway. J. Appl. Ecol. structure and behaviour of the Natterer’s bat (Myotis
31, 85–94. nattereri). PhD thesis. University of Leeds, UK.
Reijnen, R. and Foppen, R. (1995) The effects of car traffic Rivers, N.M., Butlin, R.K., and Altringham, J.D. (2005)
on breeding bird populations in woodland. IV. Influence Genetic population structure of Natterer’s bats explained
of population size on the reduction of density close to a by mating at swarming sites and philopatry. Mol. Ecol.
highway. J. Appl. Ecol. 32, 481–491. 14, 4299–4312.
Reijnen, R., Foppen, R., ter Braak, C., and Thissen, J. (1995) Rivers, N.M., Butlin, R.K,. and Altringham, J.D. (2006)
The effects of car traffic on breeding bird populations in Autumn swarming behaviour of Natterer’s bats in the
 REFERENCES 309

UK: population size, catchment area and dispersal. Biol. Ruedas, L.A., Demboski, J.R., and Sison, R.V. (1994)
Conserv. 27, 215–226. Morphological and ecological variation in Otopteropus
Robert, D., Amoroso, J., and Hoy, R.R. (1992) The cartilagonodus (Koch, 1969) (Mammalia: Chiroptera:
evolutionary convergence of hearing in a parasitoid fly Pteropodidae) from Luzon, Philippines. Proc. Biol. Soc.
and its cricket host. Science 258, 1135–1137. Wash. 107, 1–16.
Robinson, R.A., and Sutherland, W.J. (2002) Post-war Ruedi, M. and Castella, V. (2003) Genetic consequences of
changes in arable farming and biodiversity in Great the ice ages on nurseries of the bat Myotis myotis:
Britain. J. Appl. Ecol. 39,157–176. a mitochondrial and nuclear survey. Mol. Ecol. 12,
Rodrigues, A.S.L., Pilgrim, J.D., Lamoreux, J.F., Hoffmann, 1527–1540.
M., and Brooks, T.M. (2006) The value of the IUCN Red Ruedi, M., Walter, S., Fischer, M.C., Scaravelli, D.,
List for Conservation. TREE 21, 71–76. Excoffier, L., and Heckel, G. (2008) Italy as a major
Rodrigues, L., Bach, L., Dubourg-Savage, M. J., Goodwin, Ice Age refuge area for the bat Myotis myotis (Chirop-
J., and Harbusch, C. (2008) Guidelines for consideration of tera: Vespertilionidae) in Europe. Mol. Ecol. 17,
bats in wind farm projects. UNEP/EUROBATS Secretariat, 1801–1814.
Bonn, Germany. Russ J.M., Jones, G., Mackie, I.J., and Racey, P.A. (2004)
Roeder, K.D. (1964) Aspects of the tympanic nerve Interspecific responses to distress calls in bats (Chirop-
response having significance for the avoidance of bats. tera: Vespertilionidae): a function for convergence in call
J. Insect. Physiol. 10, 529–556. design? Anim. Behav. 67, 1005–1014.
Roeder, K.D. (1965) Moths and ultrasound. Sci. Am. 212, Russ, J.M. and Montgomery, W.I. (2002) Habitat associa-
94–102. tions of bats in Northern Ireland: implications for
Rosenzweig, M. L. (1995) Species Diversity in Space and conservation. Biol Conserv. 108, 49–58.
Time. Cambridge University Press, Cambridge. Russell, A.L., Butchkoski, C.M., Saidak, L., and McCracken,
Rossiter, S.J., Benda, P., Dietz, C., Zhang, S., and Jones, G. G.F. (2009) Road-killed bats, highway design, and the
(2007) Rangewide phylogeography in the greater horse- commuting ecology of bats. Endangered Species Research
shoe bat inferred from microsatellites: implications for 8, 49–60.
population history, taxonomy and conservation. Mol. Russo, D. (2002) Elevation affects the distribution of the
Ecol. 16, 4699–4714. two sexes in Daubenton’s bats Myotis daubentonii (Chir-
Rossiter, S.J., Jones, G., Ransome, R.D., and Barratt, E.M. optera: Vespertilionidae) from Italy. Mammalia 66,
(2000a) Parentage, reproductive success and breeding 543–551.
behaviour in the greater horseshoe bat (Rhinolophus fer- Russo, D. and Jones, G. (2003) Use of foraging habitats by
rumequinum). Proc. R. Soc. Lond. B. 267, 545–551. bats in a Mediterranean area determined by acoustic
Rossiter, S.J., Jones, G., Ransome, R.D., and Barratt, E.M. surveys: conservation implications. Ecography 26, 197–209.
(2000b) Genetic variation and population structure in the Russo, D., Jones, G., and Migliozzi, A. (2002) Habitat selec-
endangered greater horseshoe bat Rhinolophus ferrume- tion by the Mediterranean horseshoe bat, Rhinolophus
quinum. Mol. Ecol. 9, 1131–1035. euryale (Chiroptera: Rhinolophidae) in a rural area of
Rossiter, S.J., Jones, G., Ransome, R.D., and Barratt, E.M. southern Italy and implications for conservation. Biol.
(2002) Relatedness structure and kin-biased foraging in Conserv. 107, 71–81.
the greater horseshoe bat (Rhinolophus ferrumequinum). Ryan, M.J. and Tuttle, M.D. (1983) The ability of the frog-
Behav. Ecol. Sociobiol. 51, 510–518. eating bat to discriminate among novel and potentially
Rossiter, S.J., Ransome, R.D., Faulkes, C.G., Le Comber, S.C., poisonous frog species using acoustic cues. Anim. Behav.
and Jones, G. (2005) Mate fidelity and intra-lineage polyg- 31, 827–833.
yny in greater horseshoe bats. Nature 437, 408–411. Ryan, M.J. and Tuttle, M.D. (1987) The role of prey-generated
Roverud, R.C. (1987) The processing of echolocation sound sounds, vision and echolocation in prey localisation by
elements in bats: a behavioural approach. In: Recent the African bat, Cardioderma cor. J. Comp. Physiol. A. 161,
Advances in the Study of Bats, edited by M.B. Fenton, 59–66.
P.A. Racey, and J.M.V. Rayner, pp 152–170. Cambridge Ryan, M.J., Tuttle, M.D., and Barclay, R.M.R. (1983)
University Press, Cambridge. Behavioural responses of the frog-eating bat, Trachops
Roverud, R.C. and Grinnell, A.D. (1985) Echolocation cirrhosus, to sonic frequencies. J. Comp. Physiol. A. 150,
sound features processed to provide distance informa- 413–418.
tion in the CF/FM bat Noctilio albiventris: evidence for Ryan, M.J., Tuttle, M.D., and Rand, A.S. (1982) Bat preda-
a gated time window utilising both CF and FM compo- tion and sexual advertisement in a neotropical frog. Am.
nents. J. Comp. Physiol. A. 156, 457–469. Nat. 119, 136–139.
310 REFERENCES

Rydell, J. (1986) Foraging and diet of the northern bat Schaal, S. and Ziegler, W. (1992) Editors. Messel: An Insight
Eptesicus nilssoni in Sweden. Holar. Ecol. 9, 272–276. into the History of Life and of the Earth. Clarendon Press,
Rydell, J. (1992) Exploitation of insects around street lamps Oxford.
by bats in Sweden. Funct. Ecol. 6, 744–750. Schaffer, J. (1905) Anatomisch-histologische untersuchun-
Rydell, J. and Arlettaz, R. (1994) Low-frequency echoloca- gen uber den bau der zehen bei fledermausen und eini-
tion enables the bat Tadarida teniotis to feed on tympanate gen kletternden saugeteiren. Zeitschr. Wissensch. Zool. 83,
insects. Proc. Roy. Soc. Lond. B. 257, 175–178. 231–284.
Rydell, J., Skals, N., Surlykke, A,. and Svensson, M.G.E. Schaub, A., Ostwald, J., and Siemers, B.N. (2008) Foraging
(1997) Hearing and bat defence in geometrid winter bats avoid noise. J. Exp. Biol. 211, 3174–3180.
moths. Proc. R. Soc. Lond. B. 264, 83–88. Scherrer, J.A. and Wilkinson, G.S. (1993) Evening bat isola-
Rydell, J. and Yalden, D.W. (1997) The diets of two tion calls provide evidence for heritable signatures.
high-flying bats from Africa. J. Zool. Lond. 242, 69–76. Anim. Behav. 46, 847–860.
Rydell, J., Bach, L., Dubourg-Savage, M-J., Green, M., Schipper, J. et al. (2008) The status of the world’s land
Rodrigues, L., and Hedenstrom, A. (2010) Mortality of and marine mammals: diversity, threat, and knowledge.
bats at wind turbines links to insect migration? Eur. Science 322, 225–230.
J. Wildl. Res. DOI 10.1007/s10344-010-0444-3 Schmidt, S. (1988) Evidence for a spectral basis of texture
Safi, K. and Dechmann, D.K.N. (2005) Adaptation of perception in bat sonar. Nature 331, 617–619.
brain regions to habitat complexity: a comparative Schmidt, U., Schlegel, P., Schweizer, H., and Neuweiler, G.
analysis in bats (Chiroptera). Proc. R. Soc. Lond. B. 272, (1991) Audition in vampire bats (Desmodus rotundus).
179–186. J. Comp. Physiol. A. 168, 45–51.
Safi, K. and Kerth, G. (2003) Secretions of the interaural gland Schmidt-Nielsen, K. (1984) Scaling: Why is Animal Size so
contain information about individuality and colony mem- Important? Cambridge University Press, Cambridge.
bership in the Bechstein’s bat. Anim. Behav. 65, 363–369. Schmitt, L.H., Kitchener, D.J., and How, R.A. (1995)
Safi, K. and Kerth, G. (2004) A comparative analysis of A genetic perspective of mammalian variation and evo-
specialization and extinction risk in temperate-zone lution in the Indonesian Archipelago: biogeographic
bats. Conserv. Biol. 18, 1293–1303. correlates in the fruit bat genus Cynopterus. Evolution
Safi, K. and Kerth, G. (2007) Comparative analyses 49, 399–412.
suggest that information transfer promoted sociality Schnitzler, H.-U. (1987) Echoes of fluttering insects: infor-
in male bats in the temperate zone. Am. Nat. 170, mation for echolocating bats. In: Recent advances in the
466–472. study of bats, edited by M.B. Fenton, P.A. Racey, and
Sahaly, C.T., Horner, M.A., and Fleming, T.H. (1993) Flight J.M.V. Rayner, pp 226–243. Cambridge University
speeds and mechanical power outputs of the nectar- Press, Cambridge.
feeding bat, Leptonycteris curasoae (Phyllostomidae: Glos- Schnitzler, H.-U. and Ostwald, J. (1983) Adaptations for the
sophaginae). J. Mammal. 74, 594–600. detection of fluttering insects by echolocation in Horse-
Sala, O.E., et al. (2000) Biodiversity – global biodiversity shoe bats. In: Advances in Neuroethology, edited by
scenarios for the year 2100. Science 287, 1770–1774. J.P. Ewart, R.R. Capranica, and D.J. Ingle, pp 801–827.
Sales, G. and Pye, D. (1974) Ultrasonic Communication by Plenum Press, New York.
Animals. Chapman and Hall, London. Schnitzler, H.-U., Kalko, E.K.V., and Denzinger, A. (2004)
Salgueiro, P., Ruedi, M., Coelho, M.M., and Palmeirim, J.M. Evolution of echolocation and foraging behavior in
(2007) Genetic divergence and phylogeography in the bats. In: Echolocation in Bats and Dolphins, edited by
genus Nyctalus (Mammalia, Chiroptera): implications J. Thomas, pp. 331–339. University of Chicago Press,
for population history of the insular bat Nyctalus azoreum. Chicago.
Genetica 130, 169–181. Schnitzler, H.-U., Moss, C.F., and Denzinger, A. (2003)
Sankaran, R. (2001) The status and conservation of the From spatial orientation to food acquisition in echolocat-
Edible- nest swiftlet (Collocalia fuciphaga) in the Andaman ing bats. TREE 18, 386–394.
and Nicobar Islands. Biol. Conserv. 97, 283–294. Schnitzler, H.-U., Kalko, E.K.V., Kaipf, I., and Grinnell, A.D.
Saunders, M.B. and Barclay, R.M.R. (1992) Ecomorphology (1994) Hunting and echolocation behaviour of the fisher-
of insectivorous bats: a test of predictions using two man bat, Noctilio leporinus, in the field. Behav. Ecol. Socio-
morphologically similar species. Ecology 73, 1335–1345. biol. 35, 327–345.
Savage, D.E. (1951) A Miocene Phyllostomatid bat from Schnitzler, H.-U., Menne, D., Kober, R., and Heblich, K.
Colombia, South America. Univ. Calif. Pubs in Geol. 28, (1983) The acoustical image of fluttering insects in echo-
357–366. locating bats. In: Neuroethology and Behavioural
 REFERENCES 311

Physiology. Roots and Growing Points, edited by F. Huber vegetation using echolocation signals of very broad
and H. Markl, pp 235–250. Springer-Verlag, Heidelberg. bandwidth. Behav. Ecol. Sociobiol. 47, 400–412.
Schoeman, M.C. and Jacobs, D.S. (2003) Support for the Siemers, B.M. and Schnitzler, H.-U. (2004) Echolocation
allotonic frequency hypothesis in an insectivorous bat signals reflect niche differentiation in five sympatric
community. Oecologia 134, 154–162. congeneric bat species. Nature 429, 657–661.
Schoeman, M.C. and Jacobs, D.S. (2008) The relative influ- Siemers, B.M., Stilz, P., and Schnitzler, H.-U. (2001)
ence of competition and prey defenses on the phenotypic The acoustic advantage of hunting at low heights
structure of insectivorous bat ensembles in Southern above water: behavioural experiments on the European
Africa. PLoS ONE 3, e3715. ‘trawling’ bats Myotis capaccinii, M. dasycneme and
Schoener, T.W. (1986) Overview: kinds of ecological M. daubentonii. J. Exp. Biol. 204, 3843–3854.
communities - ecology becomes pluralistic. In: Commu- Siemers, B.M., Baur, E., and Schnitzler, H.-U. (2005) Acoustic
nity Ecology, edited by J. Diamond and T.J. Case, pp mirror effect increases prey detection distance in trawling
467–479. Harper and Row, New York. bats. Naturwissenschaften 92, 272–276.
Scholander, P.F., Walters, V., Hock, R., and Irving, L. (1950) Siemers, B.M., and Swift, S.M. (2006) Differences in sensory
Body insulation of some arctic and tropical mammals ecology contribute to resource partitioning in the
and birds. Biol. Bull. 99, 225–236. bats Myotis bechsteinii and Myotis nattereri (Chiroptera:
Scholey, K.D. (1986) The evolution of flight in bats. Vespertilionidae). Behav. Ecol. Sociobiol. 59, 373–380.
In: Biona Rep. 5, Bat Flight-Fledermausflug, edited by Siemers, B.M, and Schaub, A. (2011) Hunting at the high-
W. Nachtigall, pp 1–12. Gustav Fischer Verlag, Stuttgart. way: traffic noise reduces foraging efficiency in acoustic
Schorcht, W., Bontadina, F., and Schaub, M. (2009) Varia- predators. Proc. Roy. Soc. B. Lond. In press.
tion of adult survival drives population dynamics in a Siemers, B.M., Beedholm, K., Dietz, C., Dietz, I., and
migrating forest bat. J. Anim. Ecol. 78, 1182–1190. Ivanova, T. (2005) Is species identity, sex, age or indi-
Schultz, M. (1997) Bats in bird nests in Australia: a review. vidual quality conveyed by echolocation call frequency
Mammal Rev. 27, 69–76. in European horseshoe bats? Acta Chiropt. 7, 59–274.
Schultz, M. (2000) Roosts used by the golden-tipped bat Simmons, J.A. (1971) Echolocation in bats: Signal proces-
Kerivoula papuensis (Chiroptera: Vespertilionidae). sing of echoes for target range. Science 171, 925–928.
J. Zool. Lond. 250, 467–478. Simmons, J.A. (1973) The resolution of target range by
Sears, K.E., Behringer, R.R., Rasweiler IV, J.J., and Niswander, echolocating bats. J. Acous. Soc. Amer. 54, 157–173.
L.A. (2006) Development of bat flight: morphologic and Simmons, J.A. (1974) Response of the Doppler echolocation
molecular evolution of bat wing digits. PNAS 103, system in the bat Rhinolophus ferrumequinum. J. Acous.
6581–6586. Soc. Amer. 56, 672–682.
Sendor, T. and Simon, M. (2003) Population dynamics of Simmons, J.A. (1979) Perception of echo phase information
the pipistrelle bat: effects of sex, age and winter weather in bat sonar. Science 204, 1336–1338.
on seasonal survival. J. Anim. Ecol. 72, 308–320. Simmons, J.A. (1993) Evidence for perception of fine echo
Senior, P., Butlin, R.K., and Altringham, J.D. (2005) Sex and delay and phase by the FM bat, Eptesicus fuscus. J. Comp.
segregation in temperate bats. Proc. R. Soc. Lond. B. 272, Physiol. A. 172, 533–547.
2467–2473. Simmons, J.A. and Stein, R.A. (1980) Acoustic imaging in
Sherwin, R. E., Altenbach, J. S., and Haymond, S. (2002) bat sonar: echolocation signals and the evolution of
Response of Bats to Gates. Bat Gate Design: A Technical echolocation. J. Comp. Physiol. 135, 61–84.
Interactive Forum, pp. 8. The Fish and Wildlife Service, Simmons, J.A. and Vernon, J.A. (1971) Echolocation:
Office of Surface Mining, and Bat Conservation, Austin, discrimination of targets by the bat Eptesicus fuscus.
Texas. J. Exp. Zool. 176, 315–328.
Sherwin, R.E., Altenbach, J.S., and Waldien, D.L. (2009) Simmons, J.A., Howell, D.J., and Suga, N. (1975) Informa-
Managing Abandoned Mines for Bats, pp. 103. Bat Conser- tion content of bat sonar echoes. Amer. Sci. 63, 204–215.
vation International, Austin, Texas. Simmons, J.A., Kick, S.A., Lawrence, B.D., Hale, C., Bard, C.,
Shilton, L.A., Altringham, J.D., Compton, S.G., and and Escudie, B. (1983) Acuity of horizontal angle discrim-
Whittaker, R.J. (1999) Old World fruit bats can be ination by the echolocating bat, Eptesicus fuscus. J. Comp.
long-distance seed dispersers through extended retention Physiol. A. 153, 321–330.
of viable seeds in the gut. Proc. R. Soc. Lond. B. 266, 219–223. Simmons, J.A., Lavender, W.A., Lavender, B.A., Dor-
Siemers, B.M. and Schnitzler, H.-U. (2000) Natterer’s bat oshow, C.A., Kiefer, S.W., Livingston, R., Scallet, A.C.,
(Myotis nattereri Kuhl, 1818) hawks for prey close to and Crowley, D.E. (1974) Target structure and echo
312 REFERENCES

spectral discrimination by echolocating bats. Science 186, Speakman, J.R. (1991) The impact of predation by birds on bat
1130–1132. populations in the British Isles. Mammal. Rev. 21, 132–142.
Simmons, N.B., and Geisler, J.H. (1998) Phylogenetic rela- Speakman, J.R. and Racey, P.A. (1987) The energetics
tionships of Icaronycteris, Archaeonycteris, Hassianycteris, of pregancy and lactation in the brown long-eared bat,
and Palaeochiropteryx to extant bat lineages, with com- Plecotus auritus. In: Recent Advances in the Study of
ments on the evolution of echolocation and foraging Bats, edited by M.B. Fenton, P.A. Racey, and J.M.V.
strategies in Microchiroptera. Bull. Am. Mus. Nat. Hist. Rayner, pp 367–393. Cambridge University Press,
235, 1–182. Cambridge.
Simmons, N.B., Novacek, M.J., and Baker, R.J. (1991) Speakman, J.R. and Racey, P.A. (1989) Hibernal ecology of
Approaches, methods, and the future of the chiropteran the pipistrelle bat: Energy expenditure, water require-
monophyly controversy: a reply to J.D. Pettigrew. Syst. ments and mass loss, implications for survival and the
Zool. 40, 239–243. function of winter emergence flights. J. Anim. Ecol. 58,
Simmons, N.B., Seymour, K.L., Habersetzer, J., and 797–813.
Gunnell, G.F. (2008) Primitive Early Eocene bat from Speakman, J.R. and Racey, P.A. (1991) No cost of echoloca-
Wyoming and the evolution of flight and echolocation. tion for bats in flight. Nature 350, 421–423.
Nature 451, 818–822. Speakman, J.R., Irwin, N., Tallach, N., and Stone, R. (1999)
Sinclair, E.A., Webb, N.J., Marchant, A.D., and Tidemann, Effect of roost size on the emergence behaviour of
C.R. (1996) Genetic variation in the little red flying-fox pipistrelle bats. Anim. Behav. 58, 787–795.
Pteropus scapulatus (Chiroptera: Pteropodidae): implica- Speakman, J.R. and Thomas, D.W. (2003) Physiological
tions for management. Biol. Conserv. 76, 45–50. ecology and energetics of bats. In: Bat Ecology, edited
Skinner, J.D. and Smithers, R.H.N. (1990) The Mammals of by T.H. Kunz and M.B. Fenton, pp 430–490. University
the Southern African Subregion. University of Pretoria, of Chicago Press, Chicago.
Pretoria. Springer, M.S., Teeling, E.C., Madsen, O., Stanhope, M.J.,
Slade, C.P. and Law, B.S. (2008) An experimental test of and de Jong, W.W. (2001) Integrated fossil and molecu-
gating derelict mines to conserve bat roost habitat in lar data reconstruct bat echolocation. Proc. Natl. Acad.
southeastern Australia. Acta Chiropt. 10, 367–376. Sci. USA 98, 6241–6246.
Slater, F.M. (2002) An assessment of wildlife road Stebbings, R.E. (1988) Conservation of European Bats.
casualties – the potential discrepancy between numbers Christopher Helm, London.
counted and numbers killed. Web Ecology 3, 33–42. Stevens, R.D. and Willig, M.R. (2002) Geographical ecology
Sluiter, J.W. and van Heerdt, P.F. (1966) Seasonal habits of at the community level: perspectives on the diversity of
the noctule bat (Nyctalus noctula). Arch. Neerl. Zool. 16, new world bats. Ecology 83, 545–560.
423–439. Stevens, R.D., Willig, M.R., and Gamarra de Fox, I. (2004)
Sluiter, J.W., Voute, A.M., and van Heerdt, P.F. (1973) Hiber- Comparative community ecology of bats from Eastern
nation of Nyctalus noctula. Period. Biol. 75, 181–188. paraguay: taxonomic, ecological, and biogeographic
Smith, H. (1982) Light quality, photoreception and plant perspectives. J. Mammal. 85, 698–707.
strategy. Ann. Rev. Plant Physiol. 33, 481–518. Stevens, R.D., López-González, C., and Presley, S.J. (2007)
Smith, J.D. and Madkour, G. (1980) Penial morphology and Geographical ecology of Paraguayan bats: spatial
the question of chiropteran phylogeny. In: Proc. Fifth Int. integration and metacommunity structure of interacting
Bat Res. Conf., edited by D.E. Wilson and A.L. Gardner, assemblages. J. Anim. Ecol. 76, 1086–1093.
pp 347–365. Texas Technical Press, Lubbock. Stone, E.L., Jones, G., and Harris, S. (2009) Street lighting
Smith, J.D. and Storch, G. (1981) New middle Eocene bats disturbs commuting bats. Curr. Biol. 19, doi:10.1016/j.
from Grube Messel near Darmstadt, W. Germany. cub.2009.05.058
Senckenberg. Biol. 61, 153–167. Stone, G.N. and Purvis, A. (1992) Warm-up rates during
Sovacool, B.K. (2009) Contextualizing avian mortality: arousal from torpor in heterothermic mammals: physio-
A preliminary appraisal of bird and bat fatalities from logical correlates and a comparison with heterothermic
wind, fossil-fuel, and nuclear electricity. Energ. Policy 37, insects. J. Comp. Physiol. B. 162, 284–295.
2241–2248. Store, R. and Jokimäki, J. (2003) A GIS-based multi-scale
Spangler, H.G. (1988) Moth hearing, defence and com- approach to habitat suitability modelling. Ecol. Model.
munication. Ann. Rev. Entomol. 33, 59–82. 169, 1–15.
Spanjer, G.R., and Fenton, M.B. (2005) Behavioral Storz, J.F., Bhat, H.R., and Kunz, T.H. (2000) Social struc-
responses of bats to gates at caves and mines. Wildlife ture of a polygynous tent-making bat, Cynopterous sphinx
Society Bulletin, 33, 1101–1112. (Megachiroptera). J. Zool. Lond. 251, 151–165.
 REFERENCES 313

Storz, J.F. and Williams, C.F. (1996) Summer population Craseonycteris thonglongyai and Myotis siligorensis. Behav.
structure of sub-alpine bats in Colorado. Southwest Nat. Ecol. Sociobiol. 33, 1–12.
41, 322–324. Sussman, R.W. and Raven, P.H. (1978) Pollination by
Streicker, D.G., Turmelle, A.S., Vonhof, M.J., Kuzmin, I.V., lemurs and marsupials: an archaic coevolutionary sys-
McCracken, G.F., and Rupprecht, C.E. (2010) Host phy- tem. Science 197, 885–886.
logeny constrains cross-species emergence and Sutherland, W.J. (2000) The Conservation Handbook.
establishment of rabies virus in bats. Science 329, 676–679. Research, Management and Policy. Blackwell, Oxford.
Strelkov, P.P. (1969) Migratory and stationary bats Suthers, R.A. (1967) Comparative echolocation by fishing
(Chiroptera) of the European part of the Soviet Union. bats. J. Mammal. 48, 79–87.
Acta Zool. Cracov. 14, 393–439. Suthers, R.A. (1970) Vision, olfaction and taste. In: The
Struebig, M.J., Kingston, T., Zubaid, A., Mohd-Adnan, A., Biology of Bats edited by W.A. Wimsatt, pp 265–309.
and Rossiter, S.J. (2008) Conservation value of forest frag- Academic Press, New York.
ments to Palaeotropical bats. Biol. Conserv. 141, 2112–2126. Suthers, R.A. and Wenstrup, J.J. (1987) Behavioural
Studier, E.H. and Howell, D.J. (1969) Heart rate of female discrimination studies involving prey capture by echo-
big brown bats in flight. J. Mammal. 50, 842–845. locating bats. In: Recent Advances in the Study of Bats,
Studier, E.H. and O’Farrell, M.J. (1976) Biology of Myotis edited by M.B. Fenton, P.A. Racey, and J.M.V. Rayner,
thysanodes and M. lucifugus (Chiroptera: Vespertilioni- pp 122–151. Cambridge University Press, Cambridge.
dae). III. Metabolism, heart rate, breathing rate and Swartz, S.M., Bennett, M.B., and Carrier, D.R. (1992) Wing
general energetics. Comp. Biochem. Physiol. 54, 423–432. bone stresses in free flying bats and the evolution of
Studier, E.H. and Wilson, D.E. (1970) Thermoregulation skeletal design for flight. Nature 359, 726–729.
in some Neotropical bats. Comp. Biochem. Physiol. 34, Swift, S.M. (1980) Activity pattern of pipistrelle bats (Pipis-
251–262. trellus pipistrellus) in north-east Scotland. J. Zool. Lond.
Studier, E.H. and Wilson, D.E. (1979) Effects of captivity on 190, 285–295.
thermoregulation and metabolism in Artibeus jamaicensis Swift, S.M., Racey, P.A., and Avery, M.I. (1985) Feeding
(Chiroptera: Phyllostomidae). Comp. Biochem. Physiol. 62, ecology of Pipistrellus pipistrellus (Chiroptera: Vesperti-
347–350. lionidae) during pregnancy and lactation II Diet. J. Anim.
Suga, N. (1970) Echo-ranging neurons in the inferior Ecol. 54, 217–225.
colliculus of bats. Science 170, 449–452. Szewczak, J.M. and Arnett, E.B. (2006) An acoustic deter-
Suga, N. (1977) Amplitude spectrum representation in rent with the potential to reduce bat mortality from wind
the Doppler shifted CF processing area of the auditory turbines. Bat Res. News 47, 151–152.
cortex of the moustached bat. Science 196, 64–67. Teeling, E.C. (2009) Hear, hear: the convergent evolution of
Suga, N. (1990) Biosonar and neural computation in bats. echolocation in bats? TREE 24, 351–354.
Sci. Amer. 262, 34–41. Teeling, E.C., Scally, M., Kao, D.J., Romagnoli, M.L.,
Suga, N. and Jen, P.H.S. (1975) Peripheral control of Springer, M.S., and Stanhope, M.J. (2000) Molecular evi-
acoustic signals in the auditory system of echolocating dence regarding the origin of echolocation and flight in
bats. J. Exp. Biol. 62, 277–311. bats. Nature 403, 188–192.
Suga, N. and O’Neill, W.E. (1979) Neural axis representing Teeling, E.C., Madsen, O., Murphy, W.J., Springer, M.S.,
target range in the auditory cortex of the moustached and O’Brien, S.J. (2003) Nuclear gene sequences confirm
bat. Science 206, 351–353. an ancient link between New Zealand’s short-tailed bat
Suga, N. and Schlegel, P. (1972) Neural attenuation of and South American noctilionoid bats. Mol. Phylogenet.
responses to emitted sounds in echolocating bats. Science Evol. 28, 308–319.
177, 82–84. Teeling, E.C., Springer, M.S., Madsen, O., Bates, P.,
Suga, N., O’Neill, W.E., and Manabe, T. (1978) Cortical O’Brien, S.J., and Murphy, W.J. (2005) A molecular phy-
neurons sensitive to combinations of information- logeny for bats illuminates biogeography and the fossil
bearing elements of bio-sonar signals in the moustached record. Science 307, 580–584.
bat. Science 200, 778–781. Tennekes H. (2009) The Simple Science of Flight. MIT Press,
Sullivan, W.E. (1982) Neural representation of target Cambridge, MA.
distance in auditory cortex of the echolocating bat Myotis Thabah, A., Li, G., Wang, Y., Liang, B., Hu, K., Zhang, S.,
lucifugus. J. Neurophys. 48, 1011–1032. and Jones, G. (2007) Diet, echolocation calls, and phylo-
Surlykke, A., Miller, L.A., Mohl, B., Andersen, B.B., genetic affinities of the great evening bat (Ia Io; Vesperti-
Christensen-Dalsgaard, J., and Jorgensen, M.B. (1993) lionidae): another carnivorous bat. J. Mammal. 88,
Echolocation in two very small bats from Thailand: 728–735.
314 REFERENCES

Thewisen, J.G.M. and Babcock, S.K. (1991) Distinctive Tidemann, C.R., Vardon, M.J., Loughland, R.A., and
cranial and cervical innervation of wing muscles: new Brocklehurst, P.J. (1999) Dry season camps of flying
evidence for bat monophyly. Science 251, 934–936. foxes (Pteropus spp) in Kakadu World Heritage Area,
Thies, W., Kalko, E.K.V., and Schnitzler, H.-U. (1998) The North Australia. J. Zool. Lond. 247, 155–163.
roles of echolocation and olfaction in two Neotropical Tillman, D. (2004) Niche tradeoffs, neutrality, and commu-
fruit-eating bats, Carollia perspicillata and C. castanea, nity structure: a stochastic theory of resource competi-
feeding on Piper. Behav. Ecol. Sociobiol. 42, 397–409. tion, invasion, and community assembly. PNAS 101,
Thomas, C.D., et al. (2004) Extinction risk from climate 10854–10861.
change. Nature 427, 145–148. Tilman, D., May, R.M., Lehman, C.L., and Nowak, M.A.
Thomas, D.W. (1983) The annual migration of three species (1994) Habitat destruction and the extinction debt.
of West African fruit bats (Chiroptera: Pteropodidae). Nature 371, 65–66.
Can. J. Zool. 61, 2266–2272. Trachsel, L., Edgar, D.M., and Heller, H.C. (1991) Are
Thomas, D.W. (1993) Lack of evidence for a biological ground-squirrels sleep-deprived during hibernation?
alarm clock in bats (Myotis spp.) hibernating under Am. J. Physiol. 260, R1123–R1129.
natural conditions. Can. J. Zool. 71, 1–3. Trappe, M. and Schnitzler, H.-U. (1982) Doppler-shift com-
Thomas, D.W. (1995) The physiological ecology of pensation in insect-catching horseshoe bats. Naturwissen.
hibernation in vespertilionid bats. In: Ecology, Evolution 69, 193–194.
and Behaviour of Bats, edited by P.A. Racey, and Tschapka, M. (2004) Energy density patterns of nectar re-
S.M. Swift, Symp. Zool. Soc. Lond. 67, pp 233–244. Oxford sources permit coexistence within a guild of Neotropical
University Press, Oxford. flower-visiting bats. J. Zool., Lond. 263, 7–21.
Thomas, D.W. and Cloutier, D. (1992) Evaporative water Turner, G.G. and Reeder, D.M. (2009) Update of white nose
loss by hibernating little brown bats, Myotis lucifugus. syndrome in bats, September 2009. Bat Res. News 50,
Physiol. Zool. 65, 443–456. 47–53.
Thomas, D.W., Cloutier, D., and Gagne, D. (1990) Arryth- Tuttle, M.D. (1975) Population ecology of the grey bat
mic breathing, apnea and non-steady-state oxygen (Myotis grisescens): Factors influencing early growth
uptake in hibernating little brown bats (Myotis lucifugus). and development. Occas. Pap. Mus. Nat. Hist. Univ.
J. Exp. Biol. 149, 395–406. Kans. 36, 1–24.
Thomas, D.W., Fenton, M.B., and Barclay, R.M.R. (1979) Tuttle, M.D. (1976a) Population ecology of the grey bat
Social behaviour of the little brown bat, Myotis lucifugus. (Myotis grisescens): factors influencing growth and
I. Mating behaviour. Behav. Ecol. Sociobiol. 6, 129–136. survival of newly volant young. Ecology 57, 587–595.
Thomas, D.W. and Marshall, A.G. (1984) Reproduction Tuttle, M.D. (1976b) Population ecology of the grey bat
and growth in three species of West African fruit bats. (Myotis grisescens): philopatry, timing and patterns of
J. Zool. Lond. 202, 265–281. movement, weight loss during migration, and seasonal
Thomas, J.A., Moss, C.F., and Vater, M. (2004) Echolocation adaptive strategies. Occ. pap. Mus. Nat. Hist. Univ. Kans.
in Bats and Dolphins, pp 631. University of Chicago Press, 54, 1–38.
Chicago. Tuttle, M.D. (1976c) Collecting techniques. In: Biology
Thomas, S.P. (1975) Metabolism during flight in two of Bats of the New World Phyllostomidae. Part 1, edited by
species of bats, Phyllostomus hastatus and Pteropus gouldii. R.J. Baker, J.K. Jones, and D.C. Carter, pp 71–88. Spec.
J. Exp. Biol. 63, 273–293. Publ. Mus. Texas Tech. Univ. Lubbock 10, 1–218.
Thomas, S.P. (1987) The physiology of bat flight. In: Recent Tuttle, M. D. (1977) Gating as a means of protecting cave
Advances in the Study of Bats, edited by M.B. Fenton, dwelling bats. pp. 77–82. In: 1976 National Cave Manage-
P.A. Racey, and J.M.V. Rayner, pp 75–99. Cambridge ment Symposium Proceedings, edited by T. Aley and
University Press, Cambridge. D. Rhodes, pp. 106. Speleobooks, Albuquerque.
Thomas, S.P. and Suthers, R.A. (1970) Oxygen con- Tuttle, M. D. (1979) Status, causes of decline, and manage-
sumption and physiological responses during flight in ment of endangered gray bats. J. Wildlife Manage. 43, 1–17.
an echolocating bat. Fedn. Proc. 29, 265. Tuttle, M.D. and Ryan, M.J. (1981) Bat predation and the
Thomas, S.P. and Suthers, R.A. (1972) The physiology and evolution of frog vocalisations in the neotropics. Science
energetics of bat flight. J. Exp. Biol. 57, 317–335. 214, 677–678.
Tidemann, C.R., Priddel, D.M., Nelson, J.E., and Pettigrew, Tuttle, M.D. and Stevenson, D.E. (1978) Variation in
J.D. (1985) Foraging behaviour of the Australian ghost bat, the cave environment and its biological implications.
Macroderma gigas (Microchiroptera: Megadermatidae). pp. 108–121. In: 1977 National Cave Management Sym-
Aust. J. Zool. 33, 705–713. posium Proceedings, edited by R. Zuber, J. Chester,
 REFERENCES 315

S. Gilbert, and D. Rhodes, pp. 140. Adobe Press, Vaughan, T.A. and Vaughan, R.P. (1987) Parental behavior
Albuquerque. in the African yellow-winged bat (Lavia frons). J. Mammal.
Tuttle, M.D. and Stevenson, D. (1982) Growth and survival 68, 217–223.
of bats. In: The Ecology of Bats, edited by T.H. Kunz, Vehrencamp, S.L., Stiles, F.G., and Bradbury, J.W. (1977)
pp 105–150. Plenum Press, New York. Observations on the foraging behaviour and avian prey
Tuttle, M.D., Ryan, M.J., and Belwood, J.J. (1985) Acoustic of the Neotropical carnivorous bat, Vampyrum spectrum.
resource partitioning by two species of phyllostomid J. Mammal. 58, 469–478.
bats (Trachops cirrhosus and Tonatia sylvicola). Anim. Veith, M., Beer, N., Kiefer, A., Johannesen, J., and Seitz, A.
Behav. 33, 1369–1370. (2004) The role of swarming sites for maintaining
Tuttle, M.D. and Taylor, D.A.R. (1998) Bats and Mines. gene flow in the brown long-eared bat Plecotus auritus.
Revised edition. Resource Publication Number 3. Bat Heredity 93, 342–349.
Conservation International, Austin. Veselka, N., McErlain, D.D., Holdsworth, D.W., Eger, J.L.,
Twente, J.W. (1955) Some aspects of habitat selection and Chhem, R.K., Mason, M.J., Brain, K.L., Faure, P.A., and
other behaviour of cavern-dwelling bats. Ecology 36, Fenton, M.B. (2010) A bony connection signals laryngeal
706–732. echolocation in bats. Nature 463, 939–942.
Twente, J.W. and Twente, J. (1987) Biological alarm clock Vogel, S. (1969) Chiropoterophilie in der neotrophischen
arouses hibernating big brown bats, Eptesicus fuscus. flora. Neue Mitt. II, III. Flora Abt. B. 158, 185–222, 289–323.
Can. J. Zool. 65, 1668–1674. Voigt, C.C. and von Helversen, O. (1999) Storage and
Twente, J.W., Twente, J., and Brack, V. (1985) The duration display of odour by male Saccopteryx bilineata (Chirop-
of the period of hibernation of three species of vesperti- tera, Emballonuridae). Behav. Ecol. Sociobiol. 47, 29–40.
lionid bats. II. Laboratory studies. Can J. Zool. 63, Voigt, C.C. and Speakman, J.R. (2007) Nectar-feeding bats
2955–2961. fuel their high metabolism directly with exogenous car-
van der Pijl, L. (1972) Principles of Dispersal in Higher Plants. bohydrates. Funct. Ecol. 21, 913–921.
Springer-Verlag, New York. Voigt, C.C. and Streich, W.J. (2003) Queuing for harem
Vater, M. (1982) Single unit responses in the cochlear access in colonies of the greater sac-winged bat. Anim.
nucleus of horseshoe bats to sinusoidal frequency and Behav. 65, 149–156.
amplitude modulated signals. J. Comp. Physiol. A. 149, Vonhof, M.J., Whitehead, H., and Fenton, M.B. (2004)
369–388. Analysis of Spix’s disc-winged bat association patterns
Vaughan, N., Jones, G., and Harris, S. (1996) Effects of and roosting home ranges reveal a novel social structure
sewage effluent on the activity of bats (Chiroptera: Ves- among bats. Anim. Behav. 68, 507–521.
pertilionidae) foraging along rivers. Biol. Conserv. 78, Vonhof, M.J., Barber, D., Fenton, M.B., and Strobeck, C.
337–343. (2006) A tale of two siblings: multiple paternity in big
Vaughan, N., Jones, G., and Harris, S. (1997) Habitat use by brown bats (Eptesicus fuscus) demonstrated using micro-
bats (Chiroptera) assessed by means of a broadband satellite markers. Mol. Ecol. 15, 241–247.
acoustic method. J Appl Ecol. 34, 716–730. Walcott, C. (2005) Multi-modal orientation cues in homing
Vaughan, T.A. (1959) Functional morphology of three bats: pigeons. Int. Comp. Biol. 45, 574–581.
Eumops, Myotis and Macrotus. Publ. Mus. Nat. Hist. Univ. Walker, E.P. (1975) Mammals of the World, Vol. 1. John
Kans. 12, 1–153. Hopkins University Press, Baltimore.
Vaughan, T.A. (1970) The transparent dactylopatagium Walsh, A.L. and Harris, S. (1996) Factors determining the
minus in phyllostomid bats. J. Mammal. 51, 142–145. abundance of vespertilionid bats in Britain: geographi-
Vaughan, T.A. (1976) Nocturnal behaviour of the African cal, land class and local habitat relationships. J. Appl.
false vampire bat (Cardioderma cor). J. Mammal. 57, 227–248. Ecol. 33, 519–529.
Vaughan, T.A. (1977) Foraging behaviour of the giant Waser, N.M., Chittka, L., Price, M.V., Williams, N.M., and
leaf-nosed bat Hipposideros commersoni. E. Afr. Wildlife J. Ollerton, J. (1996) Generalisation in pollination systems,
15, 237–250. and why it matters. Ecology 77, 1043–1060.
Vaughan, T.A. and Bateman, M.M. (1980) The molossid Waters, D.A. and Jones, G. (1995) Echolocation call
wing: some adaptations for rapid flight. Proc. 5th Int. structure and intensity in five species of insectivorous
Bat Res. Conf., pp 69–78. Texas Tech. Press, Lubboch. bat. J. Exp. Biol. 198, 475–489.
Vaughan, T.A. and O’Shea, T.J. (1976) Roosting ecology of Waters, D.A. and Jones, G. (1996) The peripheral auditory
the pallid bat, Antrozous pallidus. J. Mammal. 57, 19–42. characteristics of noctuid moths: responses to the search-
Vaughan, T.A. and Vaughan, R.P. (1986) Seasonality phase echolocation calls of bats. J. Exp. Biol. 199, 857–868.
and the behavior of the African yellow-winged bat. Watt, E.M. and Fenton, M.B. (1995) DNA-fingerprinting
J. Mammal. 67, 91–102. provides evidence of discriminate suckling and
316 REFERENCES

nonrandom mating in little brown bats Myotis lucifu- agricultural intensification on bat foraging. Conserv.
gus. Mol. Ecol 4, 261–264. Biol. 18, 1–10.
Webb, N.J. and Tidemann, C.R. (1996) Mobility of Austra- Wiles, G.J. (1992) Recent trends in the fruit bat trade on
lian flying-foxes, Pteropus species (Megachiroptera): evi- Guam. In: Pacific Island Flying Foxes: Proceedings of an Inter-
dence from genetic variation. Proc. R. Soc. Lond. B. 263, national Conservation Conference, edited by D.E. Wilson and
497–502. G.L. Graham, pp. 53–60. US Fish and Wildlife Service
Webster, F.A. (1967) Interception performance of echolo- Biological Report 90(23). US Department of the Interior,
cating bats in the presence of interference. In: Animal Fish and Wildlife Service, Washington DC.
sonar systems, Vol. 1, edited by R.G. Busnel, pp 673–713. Wilkinson, G.S. (1984) Reciprocal food sharing in the
Laboratoire de Physiologie Acoustique, Jouy-en-Josas. vampire bat. Nature 308, 181–184.
Wegiel, A. and Wegiel, J. (1998) Bat protection in caves in Wilkinson, G.S. (1985) The social organisation of the
Poland. Myotis 36, 63–69. common vampire bat. I. Pattern and cause of association.
Weinbeer, M. and Kalko, E.K.V. (2007) Ecological niche Behav. Ecol. Sociobiol. 17, 111–121.
and phylogeny: the highly complex echolocation behav- Wilkinson, G.S. (1987) Altruism and co-operation in bats.
ior of the trawling long-legged bat, Macrophyllum macro- In: Recent Advances in the Study of Bats, edited by
phyllum. Behav. Ecol. Sociobiol. 61, 1337–1348. M.B. Fenton, P.A. Racey, and J.M.V. Rayner, pp 299–323.
Weir, B.S. and Cockerham, C.C. (1984) Estimating F-statistics Cambridge University Press, Cambridge.
for the analysis of population structure. Evolution 38, Wilkinson, G.S. (1990) Food sharing in vampire bats. Sci.
1358–1370. Am. 262, 64–70.
Wendeln, M. C., Runkle, J.R., and Kalko, E.K.V. (2000) Wilkinson, G.S. (1992a) Communal nursing in the evening
Nutritional values of 14 fig species and bat feeding bat, Nycticeius humeralis. Behav. Ecol. Sociobiol. 31,
preferences in Panama. Biotropica 32, 489–501. 225–235.
Wenstrup, J.J. and Suthers, R.A. (1984) Echolocation of Wilkinson, G.S. (1992b) Information transfer at evening bat
moving targets by the fish-catching bat, Noctilio lepori- colonies. Anim. Behav. 44, 501–518.
nus. J. Comp. Physiol. A. 155, 75–89. Wilkinson, G.S. (1995) Information transfer in bats. In:
Whittaker, J.O. and Black H.L. (1976) Food habits of cave Ecology, Evolution and Behaviour of Bats, edited by
bats from Zambia, Africa. J. Mammal. 57, 199–204. P.A. Racey, and S.M. Swift, Symp. Zool. Soc. Lond. 67,
Whittaker, R.J., Araújo, M.B., Jepson, P., Ladle, R.J., pp 345–360. Oxford University Press, Oxford.
Watson, J.E.M., and Willis, K.J. (2005) Conservation Wilkinson, G.S., Chapman, A.M. (1991) Length and
biogeography: assessment and prospect. Diversity and sequence variation in evening bat d-loop mtDNA.
Distributions 11, 3–23. Genetics, 128, 607–617.
Whittaker, R.J. and Fernández-Palacios, J.M. (2007) Island Wilkinson, G.S. and Boughman, J.W. (1998) Social calls
Biogeography: Ecology, Evolution, and Conservation, 2nd coordinate foraging in greater spear-nosed bats. Anim.
edition. Oxford University Press, Oxford. Behav. 55, 337–350.
Whittaker, R.J. and Jones, S.H. (1994) The role of frugivo- Wilkinson, G.S. and Fleming, T.H. (1996) Migration and
rous bats and birds in the rebuilding of a tropical forest evolution of lesser long-nosed bats Leptonycteris curasoae,
ecosystem, Krakatau, Indonesia. J. Biogeogr. 21, 245–258. inferred from mitochondrial DNA. Mol. Ecol. 5, 329–339.
Wibbelt, G. et al. (2010) White-Nose Syndrome Fungus Wilkinson, G. S. and McCracken, G. F. (2003) Bats and
(Geomyces destructans) in bats, Europe. Emerg. Infect. balls: sexual selection and sperm competition in the
Dis. 16, 1237–1243. Chiroptera. In: Bat Ecology, edited by T. H. Kunz and
Wible, J.R. and Novacek, M.J. (1988) Cranial evidence for the M. B. Fenton, pp. 128–155. University of Chicago Press,
monophyletic origin of bats. Am. Mus. Novit. 2911, 1–19. Chicago.
Wiens, J.A. (1989) Spatial Scaling in Ecology. Funct. Ecol. 3, Wilkinson, G. S. and South, J. M. (2002) Life history,
385–397. ecology and longevity in bats. Aging Cell 1, 124–131.
Wickramasinghe, L.P., Harris, S., Jones, G., and Vaughan, N. Wilkinson, L.C. and Barclay, R.M.R. (1997) Differences in
(2003) Bat activity and species richness on organic and the foraging behaviour of male and female big brown
conventional farms: impact of agricultural intensifica- bats (Eptesicus fuscus) during the reproductive period.
tion. J. Appl. Ecol. 40, 984–993. Ecoscience 4, 279–285.
Wickramasinghe, L., Harris, S., Jones, G., and Vaughan, N. Williams-Guillén, K., Perfecto, I., and Vandermeer, J. (2008)
(2004) Abundance and species richness of nocturnal in- Bats limit insects in a neotropical agroforestry system.
sects on organic and conventional farms: effects of Science 320, 70.
 REFERENCES 317

Willig, M. and Moulton, M.P. (1989) The role of stochastic Winter, Y., López, J. and Helversen, O. (2003) Ultraviolet
and deterministic processes in structuring Neotropical vision in a bat. Nature 425, 612–614.
bat communities. J. Mammal. 70, 323–329. Winter, Y. and Helversen, O. (2003) Operational tongue
Willig, M.R. and Selcer, K.W. (1989) Bat species density length in phyllostomid nectar-feeding bats. J . Mammal.
gradients in the New World: a statistical assessment. 84, 886–896.
J. Biogeog. 16, 189–195. With, K.A. and King, A.W. (1999) Extinction thresholds for
Willig, M.R., Camillo, G.R., and Noble, S.J. (1993) Dietary species in fractal landscapes. Conserv. Biol. 13, 314–326.
overlap in frugivorous and insectivorous bats from Wojciechowski, M.S., Jefimow, M., and Tefi gowska, E.
edaphic cerrado habitats of Brazil. J. Mammal. 74, (2007) Environmental conditions, rather than season,
117–128. determine torpor use and temperature selection in large
Willig, M.R. and Lyons, K.S. (1998) An analytical model mouse-eared bats (Myotis myotis). Comp. Biochem. Phys. A.
of latitudinal gradients of species richness with an 147, 828–840.
empirical test for marsupials and bats in the New Wolk, E. and W. Bogdanowicz (1987) Hematology of the
World. Oikos 81, 93–98. hibernating bat, Myotis daubentonii. Comp. Biochem.
Willig, M.R., Kaufman, D.M., and Stevens, R.D. (2003a) Physiol. 88, 637–39.
Latitudinal gradients of biodiversity: pattern, process, Woods, M., McDonald, R.A., and Harris, S. (2003) Preda-
scale, and synthesis. Annu. Rev. Ecol. Evol. Syst. 34, tion of wildlife by domestic cats Felis catus in Great
273–309. Britain. Mammal Rev. 33, 174–188.
Willig, M.R., Patterson, B.D., and Stevens, R.D. (2003b) Worthington, D.J. and Taisacan, E.M. (1996) Fruit Bat
Patterns of range size, richness and body size in Research: 1995 Annual Report. Division of Fish and Wild-
Chiroptera. In: Bat Ecology, edited by T. H. Kunz and life, Saipan, Commonwealth of the Northern Mariana
M. B. Fenton, pp. 580–621. University of Chicago Press, Islands.
Chicago. Worthington, W.J., Hall, L., Barratt, E., and Moritz, C.
Willig, M.R., Presley, S.J., Bloch, C.P., and Genoways, H.H. (1999) Genetic structure and male-mediated gene flow
(2009) Macroecology of Caribbean bats: effects of area, in the ghost bat (Macroderma gigas). Evolution 53,
elevation, latitude, and hurricane-induced disturbance. 1582–1591.
In: Island Bats: Evolution, Ecology, and Conservation, edited Wright, S.J. (1981) Intra-archipelago vertebrate distribu-
by T.H. Fleming and P.A. Racey, pp 216–264. University tions: the slope of the species-area relation. Am. Nat.
of Chicago Press, Chicago. 118, 726–748.
Willis, C. K. R., and Brigham, R.M. (2004) Roost switching, World Wind Energy Association (2010) World Wind
roost sharing and social cohesion: forest-dwelling big Energy Report 2009. www.wwindea.org [accessed 28
brown bats, Eptesicus fuscus, conform to the fission– February 2011].
fusion model. Anim. Behav. 68, 495–505. Yager, D.D. and Hoy, R.R. (1986) The cyclopian ear: a new
Willis, C.K.R., Barclay, R.M.R., Boyles, J.G., Brigham, R.M., sense for the praying mantis. Science 231, 727–729.
Brack, V.R., Waldien, D.L., and Reichard J. (2010) Bats Yager, D.D., and May, M.L. (1990) Ultrasound-triggered,
are not birds and other problems with Sovacool’s (2009) flight-gated evasive manouevres in the praying mantis
analysis of animal fatalities due to electricity generation. Parasphendale agrionina. II. Tethered flight. J. Exp. Biol.
Energy Policy 38, 2067–2069. 152, 41–58.
Willis, K.J. and Whittaker, R.J. (2002) Species Diversity— Yager, D.D., May, M.L., and Fenton, M.B. (1990) Ultrasound-
Scale Matters. Science 295, 1245–1248. triggered, flight-gated evasive escape manouevres in
Wilson, D.E. (1971) Ecology of Myotis nigricans (Mamma- the praying mantis, Paraspenandale agrionina. I. Free flight.
lia: Chiroptera) on Barro Colorado Island, Panama Canal J. Exp. Biol. 152, 17–39.
Zone. J. Zool. Lond. 163, 1–13. Young, I.S., Warren, R.D., and Altringham, J.D. (1992)
Wilson, D.E. (1979) Reproductive patterns. In: Biology of Some properties of the respiratory and locomotory sys-
Bats of the New World Family Phyllostomidae. Part III, tems of mammals in relation to body size, J. Exp. Biol.
edited by R.J. Baker, J.K. Jones, and D.C. Carter, pp 164, 283–29.
317–378. Spec. Publ. Mus. Texas Tech. Univ. 16. Lubbock. Zhao, H., Shuyi, Z., Mingxue, Z., and Jiang, Z. (2003)
Wilson, D.E. and Reeder, D.M. (2005) Mammal Species of Correlations between call frequency and ear length
the World: A Taxanomic and Geographic Reference. John in bats belonging to the families Rhinolophidae and
Hopkins University Press, Baltimore. Hipposideridae. J. Zool. Lond. 259, 189–195.
This page intentionally left blank
 Index

Illustrations of bats are in bold type. Common names of bats are not included.

Accipiter tachiro African goshawk 228 arousal from torpor 99, 103–6, Cecropia 231
Acerodon lucifer 241 108–11, 114 Ceiba kapok 229
Adansonia digitata baobab 143, 230, 232 aspect ratio 42, 45–6, 52, 54–5, 90, 92, Ceiba grandiflora 247
adaptive radiation 35–6 157–9, 162–3, 177, 188, 192, 213–5, Ceiba pentandra 230
aerofoils 37–39 218, 242–3, 248 Centronycteris centralis 91
Agave palmeri 163, 232, 234, Plate 19 audiogram 88–89 Centropogon nigricans xiii, 233, 234
agility (flight) 58–60 auditory apparatus 63–4, 67 Centurio senex 28, 29, 146, Plate 15
agricultural landscapes, bats in auditory cortex 81–2 Chaerephon
274–6 autocorrelation function 71–2 Chaerephon ansorgei 6
agroforestry 270–3 Chaerephon chapini xii, 133, 134
Agthis australis 144 Balantiopteryx Chalinolobus
alien species 252–4 Balantiopteryx plicata 91, 116, Chalinolobus gouldii 216
altitudinal effects on diversity 209–11 127–128, 167–168 Chalinolobus morio 216
altruism, reciprocal 124, 160–61 Barbastella, 128 Chalinolobus nigrogriseus 214–5
Amorphochilus 31 Barbastella barbastellus 59, 69, 109, Chalinolobus tuberculatus 253, 269
Angophora coscata apple gum tree 232 224, 262 Cheiromeles 35
Anomalurus, flying squirrel 8 Bauhinia pauletia 236 Cheiromeles torquatus 35
Anoura 5, 57 biogeography 195–211, 259 Chiroderma salvini 65
Anoura fistulata xiii, 29, 233, 234 bioindicators, bats as 283–4 Choeronycteris 57
Anoura geoffreyi 55 biological clocks 109–111 Choeronycteris mexicana 55, 179,
Anthops ornatus 6, 65 birth 113, 114, 115 116, 117, 118, 119–121 236
Antrozous 5, 122 blade-element theory 37–8 Chrotopterus
Antrozous pallidus xiii, 88, 111, 125, body mass Chrotopterus auritus 27, 36, 107,
141, 156, 158, 218, 221, Plate 29 and cost of locomotion 47 158, 247
Aphodius dung beetle 85–86 and flight 60 circadian clock 110–11
Aquila wahlbergi Wahlberg’s eagle 228 and metabolism, torpor 99–102 circannual clock 111
Archaeonycteris 1–2 Boiga irregularis brown tree circulatory system 47
Archaeonycteris pollex 9 snake 227, 253 Cistugo
Archaeonycteris trigonodon 9 Brachyphylla Cistugo lesueuri 6
Archaeopteropus 14–15 Brachyphylla cavernarum 65, 114, 205 Cistugo seabrai 6
Archaeopteropus transiens 10 breathing rate 47, 60, 86, 102, 103 cladistics 11–2
Archaeopteryx 7, 44 brown adipose tissue 103 Cleotis percivali 6, 65
Artibeus xiii, 5, 237 climate change 257–60
Artibeus jamaicensis 6, 117, 127, 130, Calathea 182 Coccoloba manzanillensis 150
132, 153, 172, 204–5, 237 camber (wing) 14, 38,43, 51, 58–60 coevolution 219, 225, 229–36
Artibeus literatus 116, 237 camouflage 26, 146–7 Coleura 116
Artibeus watsoni 87, 114–5 Cardioderma cor 6, 54, 59, 88, 125, 143, 158 Coleura afra 117–8, 125
artificial roosts 280–3 Carnegia gigantea saguaro cactus 232 Collocalia 61
Aselliscus 65 carnivory 158–159, 225–6 Collocalia fuciphaga cave swiftlet 62
Aselliscus tricuspidatus 140 Carollia 205, 230 communication
arctiid moths 223 Carollia perspicillata xiii, 65, 89, 107, mother-infant 124–5
arms races 219, 221–5 144, 230, 233, 237 sexual 133–4

319
320 INDEX

communal nursing 125, 181 Calls CF 75–80 Falco subbuteo hobby 228, 227
community ecology 211–20 cost of 86 feeding 154 et seq.
conservation 176, 196, 206, 220, 239 ecology of 90–4 Ficus 130, 231, 237
et seq. evolution of 9–10, 92–6 Ficus insipida 172
landscape scale 267–270 phyllostomids, in 87–8 flight 37 et seq.
organisations 284 neural basis 80–2 anatomy 49–51
urban environment 248, 269–70, prey selection by 83–5 cost of (energetics) 47–9
279–280 Ectophylla xiii ecological aspects 51–60
convergent evolution 11, 12, 14, 16, Ectophylla alba 28, 28, 127, 146, 149, evolution of 7, 43–6
17, 94, 95, 159, 162, 230 149, 153, Plate 16 power requirements and speed
Craseonycteridae 5–6, 19, 23–4 ectothermy 97 51–8
Cormura brevirostris 91, 216 education 283 flutter detection 22, 30, 76–8, 85, 91,
Corollia 237 Eidolon helvum 54, 179, Plate 2 92, 95
Corollia perspicillata 87 elevational effects on diversity flycatching 22, 33, 36, 92, 157
Corollia castanea 87 209–11 foraging strategies and styles,
Craseonycteris 5 Emballonuridae 5–6, 19, 26 insectivorous bats 90–2, 154–61
Craseonycteris thonglongyai xii, 20, Emballonura 5 forestry, impacts, methods 270–4
23, Plate 7 Emballonura monticola 55–6 fossil bats 1–6
Cynocephalus, dermoptera, Emballonura nigrescens 140 Freycinetia reinechei liana 232, Plate 1
colugo 14–5 Emballonura semicaudata 56, 199 frugivores 161–3, 172–3, 228–38
Cynocephalus variegates 8 endothermy 97–9 Furipteridae 5–6, 19, 31
Cynopyterus 5, 27, 202–3, 237 energy balance 97 Furipterus 5
Cynopterus brachyotis 203–4 Epomophorus Furipterus horrens 6
Cynopterus horsfieldii 203–4 Epomophorus labiatus 66
Cynopterus sphinx 6, 149–50, 153, 237 Epomophorus minor 6 Geomyces destructans White Nose
Cyttarops alecto 91 Epomophorus wahlbergii 6 Syndrome fungus 254
Epomops 237 Geotrupes dung beetles 85–86
Dactylanthus 235 Eptesicus xi, 122, 128, 213, 255, 264 gestation 118–9
Dactylanthus taylori 32, 235 Eptesicus capensis 215 Gigantochloa scortechinii bamboo 145
Dasypterus xi Eptesicus isabellinus 122 Glaucomys volans flying squirrel 8
Desmodus 5, 29–30, 107, 160 Eptesicus furinalis 120 Glauconycteris 145
Desmodus rotundus xiv, 6, 29, 30, Eptesicus fuscus 68, 69, 71–74, 81, Glauconycteris superba 34
65–6, 88, 90, 107, 116, 144, 160–1, 88, 90, 103, 104, 109, 110, 120, 121, Glauconycteris variegata 145
165, 184 133, 154, 164, 171, 180, 182, 223, gleaning 22, 32, 35, 36, 55, 56, 58, 59,
development 119–23 251 60, 73, 88, 98, 90, 92, 93, 151, 157,
Diaemus youngi 65, 107, 160, 29 Eptesicus hottentotus 6 159, 169, 171, 177, 181, 218, 228,
Diclidurus 26 Eptesicus pumillus 215 241, 246, 247, 272, 273, 274
Diclidurus alba 91, 216 Eptesicus regulus 216 gliding 7, 44–5, 60
Diphylla ecaudata 30, 107, 160, 160, Eptesicus sagittula 215 Glischropus tylopus 145
Plate 20 Eptesicus serotinus 150, 228, 256 Glossophaga, 89, 230, 237
Dimorphodon 44 Eonycteris spelaea 115, 233–4 Glossophaga commissarisi 236,
disease 123–124, 140, 176, 183, 250, Erophylla bombifrons 65 Plate 18
252, 254, 255–6 Euderma maculatum 224, Plate 28 Glossophaga soricina xiii, 6, 42–3, 48,
dispersal Eudimorphodon 44 58, 74, 88–9, 115, 117, 175, 230,
ability of bats 199, 209, Eudiscopus denticulus 145 236, 247–8
219–20 Eumops 5 Gurania spinulosum 87
post-natal 180, 184, 188, 190 Eumops perotis 106, 114, 141
seed 229, 236–8, 257 Eumops underwoodi Plate 33 habitat degradation, destruction,
Dobsonia European bat lissavirus 256 fragmentation 238, 239, 240,
Dobsonia moluccensis 139–40 evolution 1 et seq. 243–9, 250–1, 253, 256, 257, 259,
Doppler shift 22, 76–8, 81–2, 83 Convergent 11, 12, 14, 15, 17, 94–5, 262–4, 265, 274, 276, 278
Durio zibenthus durian 232 159, 163, 230, predictive mapping 259–60, 269–72
echolocation 9–10, 92–6 Haplonycteris fischeri 203–4
ecosystem services 220, 221–2, flight 7, 43–6 Hassianycteris 1–2
255, 284 extinction (recent) 32, 197, 198, 200, Hassianycteris revilliodi 9
echolocation 61 et seq. 201, 202, 208–12, 219, 235, 238, Hassianycteris messelensis 9
Calls, FM 64, 68 et seq. 240–3, 252–3, 254, 259, 262–4, 267 heart and heart rate 47, 102–4
 INDEX 321

Heliconia xiii, 31, 147, 149, 149–50, latitudinal effects on diversity 207–9 Miniopterus australis 6, 127, 131, 140
153, 182 Lavia Miniopterus fraterculis 6
Helicoverpa zea cotton bollworm Lavia frons 24–5, 65–6, 127–8, 146, Miniopterus inflatus 6, 55
moth 256 157, Plate 9 Miniopterus macrocneme 6
Heliothis virescens tobacco budworm Lecythis 172 Miniopterus natalensis 6
moth 256 lek 21, 127, 129, 130, 133, 166, Miniopterus schreibersii 115–16, 119,
heterothermy 97–9 Leptonycteris 57 124, 127, 138, 139, 140, 152, 191,
hibernacula 107–9, 249 Leptonycteris curasoae 248 214, Plate 30
hibernation 99–111 Leptonycteris yerbabuenae 29, 53, 55, Miniopterus schreibersii
ecology of 104–106 162–163, 164, 175–6, 179, 191, 234, natalensis 191
evolution of 111–2 247, Plate 19 Miniopterus tritis 140
migration for 176–8 Lichonycteris obscura 236 molecular clock 17
Hipposideridae 5–6, 19, 22 life expectancy 123 Molossidae 5–6, 19, 33–34
Hipposideros 5, 22, 213, 217 lift 37–45, 51, 52, 56, 59, 159 Molossops temminckii 69
Hipposideros ater 215 Lissonycteris angolensis 20 Molossus
Hipposideros beatus 127–8 Lonchophylla robusta 236 Molossus ater 125
Hipposideros bicolor 66, 93 Lonchorhina aurita 27, 28 Molossus molossus 150
Hipposideros caffer 6, 22, 24, 138, 215 Longisquama 44 Molossus sinaloae 116
Hipposideros calcaratus 140 Lophostoma silvicolum 131, 191 momentum jet theory 37–8
Hipposideros cervinus 140 monoestry 114–15
Hipposideros commersoni 6, 214 Macroderma 5, 11 monogamy 126, 127, 128, 167, 168, 169
Hipposideros diadema 140, 215, 271 Macroderma gigas 24–25, 25, 88–89, Mormoopidae 5–6, 19, 30
Hipposideros ridleyi 65 107, 125, 153, 158, 165, 191, 226, Mormoops 30
Hipposideros ruber Plate 5 249, Plate 8 Mormoops megalophylla 6, 66,
Hipposideros speoris 93 macroecology 195–211 Plate 21
Histoplasma capsulatum 256 Macroglossus 58 Mormopterus 11
Homo 33 Macroglossus minimus 203–4 Mormopterus beccarii 215
hovering 22, 29, 43, 48, 52, 55, 56–8, Macroglossus woermanni 228 Mormopterus loriae 214–5
60, 73, 92, 162, 229 Macrophyllum macrophyllum 65 Mormopterus petrophilus 6
hunting and persecution 255–7 Macrotus 88, 162–163 Mormopterus planiceps 216
Hylonycteris underwoodi 236 Macrotus californicus 66, 88, 118, Mucuna 231
Hypsignathus monstrosus xii, 21, 54, 124, 218, 225, Plate 17 Mucuna holtonii xiii, 74, 87, 230
130, 133, 133 magnetic compass 90, 179–81 Murina 21, 150
maternal care 123, 124–5, 137 Musa banana 232
Icaronycteris 1–2, 4, 9, 15 mating patterns 126–32 Musonycteris
Icaronycteris index 1, 9, 94 megabats (Megachiroptera) xii, 1, Musonycteris harrisoni 247–248
information transfer 125, 137, 152, 10–6, 19, 20 Mustela erminea stoat 253
163–6, 171–2, 181–3, 184–5 Megadermatidae 5–6, 19, 24–5 mutualism 212, 229, 233, 235
insectivory 154–8, 167–72 Megaderma 5 Myotis xi, 5, 19, 33, 72, 95, 105, 107,
isolation calls 124, 165 Megaderma lyra 6, 24, 55, 58, 59, 110, 122, 128, 129, 170, 176, 217,
72–73, 85, 88–89, 93, 114, 125, 150, 218, 254
Kerivoula xii, 146, 150 158–9, 226, 226 Myotis adversus 33, 73, 116, 125,
Kerivoula picta 33 Megaderma spasma 6, 65, 125, 158 127, 130, 214
kin selection 124–5, 161, 180, 181, 184 Melolontha may bug 85–86 Myotis albescens 116
metabolic rate 47–8, 52, 60, 86, Myotis austroriparius 140
lactation 104, 113–8, 121–2, 125–6, 97–104, 119, 121–2, 141, 157 Myotis bechsteinii 68, 142, 144, 152,
153, 154, 161, 165, 178, 181, 182, microbats (Microchiroptera) xii, 1, 154, 171, 180, 181, 183, 187–90,
187, 259 10–16, 19 216, 217, 262
Laephotis wintoni 66, 143 migration 22, 32, 46, 52, 55–6, 175–80, Myotis blythii 216, 217, 225
Lasiochila goryi chrysomelid 184, 188–92, 204, 265, 277 Myotis bocagei xiii, 55, 56, 66, 148
beetle 145 conservation of migratory Myotis brandtii 224
Lasiurus xi, 56, 91, 119, 146, 153, 176, species 276–7 Myotis capaccinii 123, 153, 159, 180
177 cost of 179 Myotis dasycneme 68, 218, 256
Lasiurus borealis 55, 102, 109, 119, sex differences 178 Myotis daubentonii 20, 53, 55, 56, 60,
264 Mimetillus moloneyi 53–4, 143 73, 74, 93, 107, 130, 132, 142, 144,
Lasiurus cinereus xiii, 19, 55, 84, 85, Miniopteridae 5–6, 19, 34 154, 169, 170, 183, 187, 188, 189,
146, 147, 148, 157, 177, 178, 199 Miniopterus 4, 34, 56, 115, 128, 191 218, 256
322 INDEX

Myotis (cont.) Notopteris macdonaldi 20 philopatry 152, 171, 181–187, 188–94

Myotis emarginatus 108, 218 Nyctalus 56, 75, 144, 177–8, 264 Phyllostomidae 5–6, 19, 27–30
Myotis evotis 169, 170, 224, 225 Nyctalus lasiopterus 158, 178, 180, Phyllostomus
Myotis grisescens 82, 107, 114, 119, 183 Phyllostomus discolor 125, 165, 236
121, 122, 125, 138, 140, 177, 178 Nyctalus leisleri 55, 79, 123, 178 Phyllostomus hastatus 27, 71–2, 87,
Myotis lucifugus 68, 82, 84, 92, 100, Nyctalus noctula 33, 41, 41, 55, 59, 101–102, 114–115, 127, 132, 132,
103, 104, 107, 108, 109, 114, 121, 69, 109, 120, 127, 130, 133, 144, 158, 161, 165, 172
122, 124, 127, 150, 157, 164, 169, 177, 189, 192 phylogeny, molecular 4–7, 14–17
170, 254, 258, 260, 261, 273 Nycteceinops schlieffeni 6 Physalaemus pustulosus Tungara
Myotis myotis 88, 90, 104, 107, 187, Nycteridae 5–6, 19, 26–7 frog 225
188, 192, 193, 194, 216, 217, 225, Nycteris xiii, 5, 66, 127–128 Pipistrellus 75, 91, 176, 215, 264
252 Nycteris grandis 6, 26, 26, 107, 158–9 Pipistrellus angulatus 140
Myotis mystacinus xii, 60, 68, 82, Nycteris hispida 55, 65 Pipistrellus dormeri 93, 116
107, 116, 218 Nycteris javanicus 6 Pipistrellus hesperus 98, 141
Myotis nattereri 59, 92, 107, 129, Nycteris macrotis 6, 55, Plate 12 Pipistrellus kuhlii 53
144, 151, 155, 171, 186, 187, Nycteris thebaica 6, 54 Pipistrellus mimus 93
188, 218 Nycticeius 215 Pipistrellus nanus 127, 130, 149, 215
Myotis nigricans 150 Nycticeius greyi 214 Pipistrellus nathusii 69, 178
Myotis punicus 194 Nycticeius humeralis 124–5, 164–5, Pipistrellus pipistrellus 53, 60, 84,
Myotis septentrionalis 182, 224, 273 228 105, 109, 119–20, 123, 126–127,
Myotis sodalis 127, 142 Nycticeius sanborni 214 130, 142, 150, 165, 171, 251–2
Myotis thysanodes 100, 103, 115, Nycticeius schlieffini 215 Pipistrellus pygmaeus 150, 280
124, 259 Nyctimene 5, 21, 21, 237 Pipistrellus sturdeei 241
Myotis tricolor 6 Nyctimene albiventer 106 Pipistrellus subflavus 108
Myotis vivesi 33, 36, 55, 56, 73, Nyctimene major xii, 145 Pipistrellus tenuis 214
156, 159 Nyctimene santacrucis, 241 piscivory 159–60
Myotis welwitschii 34 Nyctophilus 88 Platymops 142
Myotis yumanensis 55, 141 Nyctophilus arnhemensis 214 Platymops setiger 142
Mystacinidae 5–6, 19, 32 Nyctophilus bifax 215 Plecotus 48, 56, 59, 128–9, 187, 205
Mystacina 3, 4–5, 235 Nyctophilus geoffroyi 216 Plecotus auritus 33, 42, 48, 52, 53, 55,
Mystacina robusta 241, 253 Nyctophilus gouldi 216 58, 60, 73, 88–89, 93, 105, 109, 121,
Mystacina tuberculata xii, 3–6, 32, Nyctophilus howensis 241 130, 138, 152, 154, 171, 187–8,
144, 205–6, 235, 253, Plate 24 Nyctophilus timoriensis 33 224–5
Myzopodidae 5–6, 19 Plecotus townsendi 55
Myzopoda xiii, 5 Onychonycteris 4, 7, 9 pollution 250–252
Myzopoda aurita 6, 32, 95, 147, Onychonycteris finneyi 4 polygyny 126–7, 131, 187, 190
Plate 25 Optimal foraging 83–85, 97, 156–8, polyoestry 115–8
Myzopoda schliemanni 32 163, 234 population structure 188–92, 202–5
Oroxylum iridicum 233–4 historical effects 192–4
Natalidae 5–6, 19, 35 Otomops martiensseni 6, 55, 57, 224 preadaptation 229
Natalus 5, 20 Otonycteris hemprichi 20 predators
Natalus major 6 Otopteropus cartilagonodus 169 bats as 220–6
Natalus micropus 6, 35 ovulation 114, 117–8 of bats 226–8
Natalus stramineus Plate 34 pregnancy 104, 114–15, 117–19,
Neoplatymops 142 Palaeochiropteryx 1–2, 9 121–2, 125, 126, 154, 178
niche partitioning 213–9 Palaeochiropteryx tupaidon 9 prey selection 83–5, 156–7
navigation 90, 179–80, 252 Palaeochiropteryx spiegeli 9 Ptenochirus
Noctilionidae 5–6, 19, 30 Paranyctimene raptor 106 Ptenochirus jagori 203–4
Noctilio 3–5 Parkia 229 Ptenochirus minor 204
Noctilio albiventris 6, 30, 31, 36, Perimyotis subflavus 251, 254 Pteranodon 44
80–1, 125, 159 Peropteryx Pteronotus 5, 30, 75, 95
Noctilio leporinus xiii, 6, 30, 36, 55–6, Peropteryx kappleri 6, 91, 127–8, Pteronotus macleayii 66
57, 73, 77, 101–2, 107, 156, 159, 167–8, 216 Pteronotus parnellii 6, 30, 77, 80–1,
159, Plate 22 Peropteryx macrotis 91 95, 115
noctuid moths 2–3, 223–4 Petaurus breviceps 46 Pteronotus personatus 6
Nothofagus southern beech 253, 270 Piper 27, 29, 87, 231, 233, 237 Pteronotus quadridens 30
Notonycteris magdalensis 27 Pharotis imogene 65 Pteronotus suapurensis 72
 INDEX 323

Pteropodidae (Old World fruit bats, Rhinolophus megaphyllus 140, 215 245–9, 259, 263, 268, 270–1, 273,
flying foxes) 1, 5–6, 19–22, 179 Rhinolophus rouxi 81–82 276
Pteropus xii, 5, 11, 60, 106 Rhinolophus ruwenzorii 65 sperm competition 133, 133
Pteropus aldabrensis 202 Rhinolophus sedulous 127–8 sperm production 114
Pteropus alecto 190 Rhinophylla pumilio 245–6 Spodoptera frugiperda fall armyworm
Pteropus brunneus 241 Rhinopomatidae 5–6, 19, 22–3 moth 256
Pteropus conspicillatus 201 Rhinopoma 5, 66 stable isotopes 176, 191, 218
Pteropus giganteus 127, 131, 202 Rhinopoma hardwickei 6, 24, 65, 76, Steatornis caripensis oilbird 61, 62
Pteropus hypomelanus 201–2 93, 120, Plate 6 Stenonycteris lanosus 66
Pteropus livingstonii 201 Rhinopoma microphyllum 106 Sturnira 20, 237
Pteropus loochoensis 241 Rhogeessa 5 Sturnira lilium 65, 107
Pteropus mariannus 60, 227, 253, Rhynchonycteris 5 survival (life expectancy) 122–4
Plate 1 Rhynchonycteris naso 91, 116, swarming 127, 128–30, 132, 169–70,
Pteropus niger 202 127–128, 144, 145, 165, 183–7, 189, 190, 192
Pteropus personatus 21 167–8, 215–6, Plate 10 Syconycteris 237
Pteropus pilosus 241 roads, effects on bats 260–4, 278–9
Pteropus poliocephalus 20, 114–15, roosting 137 et seq. Tadarida 5, 56, 66, 176, 214
127, 131, 152, 176, 190, 201, 260 ecology of 150 et seq. Tadarida aegyptiaca 6, 35, 93
Pteropus rodricensis 202 fidelity to roosts 131, 151, 166, 170, Tadarida australis 216, 280
Pteropus rufus 202 171, 176, 177, 178, 181, 191; see Tadarida brasiliensis xiii, 35, 47,
Pteropus samoensis 60, 238 also philopatry 55–56, 106, 124, 134, 138–41, 152,
Pteropus scapulatus 190, 201 roosts 138–150 165, 177, 190, 220, 228, 252, 256,
Pteropus seychellensis 202 Rousettus 5, 10–11, 16, 20, 21, 61–3, 90, Plate 31, Plate 32
Pteropus seychellensis comorensis 201 94, 106, 139, 162–163 Tadarida brasiliensis cynocephali 190
Pteropus seychellensis Rousettus aegyptiacus 6, 21, 53, 62, Tadarida brasiliensis mexicana 190
seychellensis 202 66, 115, 117 Tadarida condylura 215, 228
Pteropus subniger 202, 241 Rousettus amplexicaudatus 140, Tadarida macrotis 142
Pteropus tokudae 241 203–4, 237, Plate 3 Tadarida midas 54
Pteropus tonganus 199, 238 Tadarida pumila 215, 228
Pteropus vampyrus 92, 237, 255, 276 Saccopteryx 26, 133, 216 Tadarida teniotis 69, 224
Pteropus voeltzkowi 201 Saccopteryx bilineata 6, 91, 114, 116, Taphozous 5, 56, 214
127–8, 132–4, 144, 155, 165, 167–8, Taphozous flaviventris 214
Quetzalcoatlus 44 216 Taphozous georgianus 214
Saccopteryx leptura 91, 116, 125, Taphozous kachhensis 93
rabies 256 127–8, 145, 167–8, 216 Taphozous kapalgensis 271
Rattus sanguivory 160–1 Taphozous longimanus 116, 119
Rattus everreti 203 Scotophilus 214 Taphozous mauritianus 6, Plate 11
Rattus exulans 253 Scotophilus borbonicus 215 Taphozous melanopogon 28, 55, 93,
Rattus rattus 253 Scotophilus dinganii 6, 215 150
reciprocal altruism 160–1 Scotorepens balstoni 215 Taphozous peli 54, 66, 127–8
reproductive cycles 113–18 senses other than echolocation Taphozous perforates 66
Rhinolophidae 5–6, 19, 22 88–90 Taphozous wildegardae 66
Rhinolophus 5, 59, 217 sexual dimorphism 133 territoriality 105, 127–128, 130–1, 133,
Rhinolophus capensis 6 sexual segregation 152–153, 165, 166, 167–8, 171, 172,
Rhinolophus darling 6 in the roost 153–154 178, 182, 189, 235
Rhinolophus euryale 69, 274 in the landscape 169–172 thermal neutral zone 98
Rhinolophus euryotis 140 social structure 127, 132–3, 135, Thyropteridae 5–6, 19, 31
Rhinolophus ferrumequinum xii, 42, 135–40, 152–3, 160–1, 163–73, Thyroptera xiii, 5, 32
50, 69, 72, 77, 85–86, 92, 105, 108, 180–8 Thyroptera lavali 6
108–11, 119–21, 125, 133, 135, 152, Solanum 231, 233 Thyroptera tricolor 6, 31, 148, 148,
157, 165, 187, 206, 276, Plate 4 Solanum hayesii 237 182, Plate 23
Rhinolophus fumigatus 6 sound generation and perception Tipula crane fly 85
Rhinolophus hildebrandtii 6, 214–5 63–4 Tomopeas ravus 34
Rhinolophus hipposideros 23, 157, species-area relationships 195–201, Tonatia 5, 66
224, 251–2, 276 208, 210 Tonatia sylvicola 36, 225, Plate 14
Rhinolophus landeri 66 species richness (diversity) 196–201, torpor 99, 102–11
Rhinolophus luctus 127–8 207–11, 218, 219, 220, 230, 242, ecology of 104–6
324 INDEX

torpor (cont.) Uroderma xiii white nose syndrome 250,

evolution of 111–2 Uroderma bilobatum 28, 28, 150, 153 254–5
Trachops, 225–6 wind turbines, effects on bats 263–6,
Trachops cirrhosus 6, 27, 27, 36, 88, Vampyressa pusilla 87 277–8
144, 158, 225, 225 Vampyrum spectrum 27, 36, 100, 127–8, wing
Triaenops persicus 65 143–4, 153, 158, 158, Plate 13 loading 22, 54–6, 60, 90–2, 119, 125,
Tricanthera 232 Vernonia scandens 149 157–9, 162, 169, 177, 188, 213–5,
Trichoplusia ni cabbage looper Vespertilionidae 5–6, 19, 32–4 218
moth 256 Vespertilio 264 kinematics 38–43
Tylonycteris xiii, 127, 180 vortex wake theory 38–43 morphology 53–60
Tylonycteris pachypus 114, 142, 145,
146 weaning 113, 120, 122, 125–6, Yangochiroptera 26–35
Tylonycteris robustula 114, 142, 145 152, 165 Yinpterochiroptera 20–5
 Plate 2. A tightly-packed cluster of roosting straw-coloured fruit bats,
Eidolon helvum. Pteropodidae.

Plate 1. Marianas flying fox or fruit bat, Pteropus mariannus,

feeding on the nectar of Freycinetia flowers. Old World
Pteropodidae.

Plate 3. Part of a large colony of Geoffroy’s rousette, Rousettus Plate 4. Greater horseshoe bat, Rhinolophus ferrumequinum, a member
amplexicaudatus, roosting in a cave. Pteropodidae. of the widespread Old World Rhinolophidae.

All photographs in the plate section by Merlin Tuttle © Bat Conservation International
Plate 5. Noack’s roundleaf bat, Hipposideros ruber. The Plate 6. Lesser mouse-tailed bat, Rhinopoma hardwickei,
Hipposideridae are another successful Old World family, one of just four species of Rhinopomatidae, all found in the
closely related to the Rhinolophidae. more arid regions of the Old World tropics.

Plate 7. The tiny (1.5-2 g) Bumblebee bat, Craseonycteris Plate 8. The large, carnivorous Australian ghost bat, Macro-
thonglongyi, solitary member of the Craseonycteridae from SE derma gigas. The large eyes and ears enable it to hunt using sight
Asia. and passive sound as well as echolocation. Megadermatidae.
 Plate 10. Proboscis bat, Rhynchonycteris naso, a diminutive New World
member of the tropical Emballonuridae.

Plate 9. African yellow bat, Lavia frons. An African member

of the Old World Megadermatidae.

Plate 12. Dobson’s slit-faced bat, Nycteris macrotus.

Female nursing a pup. The Megadermatidae are Old World
insectivores, the larger species perhaps being occasional
carnivores.

Plate 11. Mauritian tomb bat, Taphozous mauritianus, an

Old World member of the Emballonuridae.
Plate 13. The carnivorous spectral or false-vampire bat, Vampyrum Plate 14. Round-eared bat, Tonatia sylvicola, taking a kaydid to a perch to
spectrum, one of the largest species of the New World Phyllostomidae, eat it. Phyllostomidae.
easily the most ecologically diverse of all bat families.

Plate 15. The small but bizarre wrinkle-faced bat, Centurio senex. In this Plate 16. A typically small social group of the frugivorous Honduran or
photo its facial wrinkles are partially covered by the loose flap of skin over its white tent bat, Ectophylla alba, in the tent they have constructed from a leaf.
chin. A fruit-eating member of the Phyllostomidae. Phyllostomidae.

Plate 17. The California leaf-nosed bat, Macrotus californicus, with the Plate 18. Two Commissaris’ nectar bats, Glossophaga commissarisi,
characteristically large, forward-pointing eyes and ears of a predator that feeding at (and pollinating) the flowers of the tropical climber, Macuna.
need not always use echolocation to capture prey. Phyllostomidae. Phyllostomidae.
 Plate 19. A long nosed-bat, Leptonycteris yerbabuenae, taking nectar Plate 20. Hairy-legged vampire, Diphylla ecaudata. One of three species
from a saguaro cactus. This species is an important pollinator of many desert of vampire bat. Phyllostomidae.
plants in south-western North America, including Agave palmeri from which
tequila is made. Phyllostomidae.

Plate 21. Ghost-faced bat, Mormoops megalophylla, fast-flying aerial Plate 22. The fisherman or bulldog bat, Noctilio leporinus, takes insects,
insectivore from the New World Mormoopidae. fish and even crustacea from the water with its long toes and claws. One of
just two species of Noctilionidae.

Plate 23. Spix’s disk-winged bat, Thyroptera tricolor, uses tiny cups on wrists Plate 24. Short-tailed bat, Mystacina tuberculata, one of two remaining bat
and ankles to stick to the smooth internal surfaces of furled leaves. Another species native to New Zealand and the only remaining member of the
small family, the New World Thyropteridae has only four species. Mystacinidae. Like many of the country’s animals it has an unusual natural history.
 Plate 26. South-eastern myotis, Myotis austroriparius. A North American
member of the largest and most widespread genus of the Vespertilionidae.
Like many warm-temperate bats it forms large nursery roosts in caves.

Plate 25. Like the disk-winged bats of the Americas, the

sucker-footed bat, Myzopoda aurita, from Madagascar has
also evolved cups to enable it to cling to smooth leaves.

Plate 27. The dense fur and striking colouration of the hoary bat, Lasiurus Plate 28. Spotted bat, Euderma maculatum. Another eye-catching vesper
cinereus make it particularly attractive. A relatively large, fast-flying aerial bat with, in relation to its size, the largest ears of any bat.
insectivore from the Vespertilionidae.
Plate 29. The pallid bat, Antrozous pallidus, another vespertilionid, takes Plate 30. Schreiber’s bent-winged bat, Miniopterus schreibersii, is a
much of its prey from the ground, including large and potentially dangerous member of a recently proposed new family, the Miniopteridae: one of many
prey like scorpions and centipedes. taxonomic revisions arising from genetic studies.

Plate 31. The Brazilian free-tailed bat, Tadarida brasiliensis (Molossidae).

This aerial insectivore has been shown to make a substantial economic
contribution to the cotton industry of the southern United States by eating
insects that are major crop pests.

Plate 32. Some of the four million Brazilian free-tailed bats, Tadarida
brasiliensis, emerging from their roost in Bracken Cave, Texas: the largest
known bat colony on Earth.
Plate 33. Underwood’s mastiff or bonneted bat, Eumops underwoodi. One of 100 members of the widespread Molossidae—all
fast-flying insectivores.

Plate 34. The Mexican funnel-eared bat, Natalus stramineus, one of eight species of New World Natalidae. All are small, gleaning,
cave-roosting bats.