Convergent Evolution Limited Forms Most Beautiful Ebook3000 PDF

www.ebook3000.
com
Convergent Evolution
www.ebook3000.com
Vienna Series in Theoretical Biology
Gerd B. Müller, Günter P. Wagner, and Werner Callebaut, editors
The Evolution of Cognition, edited by Cecilia Heyes and Ludwig Huber,

2000
Origination of Organismal Form: Beyond the Gene in Development and
Evolutionary Biology, edited by Gerd B. Müller and Stuart A. Newman,
2003
Environment, Development, and Evolution: Toward a Synthesis, edited
by Brian K. Hall, Roy D. Pearson, and Gerd B. Müller, 2004
Evolution of Communication Systems: A Comparative Approach, edited
by D. Kimbrough Oller and Ulrike Greibel, 2004
Modularity: Understanding the Development and Evolution of Natural
Complex Systems, edited by Werner Callebaut and Diego Rasskin-
Gutman, 2005
Compositional Evolution: The Impact of Sex, Symbiosis, and Modularity
on the Gradualist Framework of Evolution, by Richard A. Watson, 2006
Biological Emergences: Evolution by Natural Experiment, by Robert
G. B. Reid, 2007
Modeling Biology: Structure, Behaviors, Evolution, edited by Manfred D.
Laubichler and Gerd B. Müller, 2007
Evolution of Communicative Flexibility: Complexity, Creativity,
and Adaptability in Human and Animal Communication, edited by
Kimbrough D. Oller and Ulrike Greibel, 2008
Functions in Biological and Artificial Worlds: Comparative Philosophical
Perspectives, edited by Ulrich Krohs and Peter Kroes, 2009
Innovation in Cultural Systems: Contributions from Evolutionary Anthro-
pology, edited by Michael J. O’Brien and Stephen J. Shennan, 2009
The Major Transitions in Evolution Revisited, edited by Brett Calcott and
Kim Sterelny, 2011
Transformations of Lamarckism: From Subtle Fluids to Molecular
Biology, edited by Snait B. Gissis and Eva Jablonka, 2011
Convergent Evolution: Limited Forms Most Beautiful, by George R.
McGhee Jr., 2011
www.ebook3000.com
Convergent Evolution
Limited Forms Most Beautiful
George R. McGhee Jr.
The MIT Press

Cambridge, Massachusetts
London, England
www.ebook3000.com
© 2011 Massachusetts Institute of Technology
All rights reserved. No part of this book may be reproduced in any form by any electronic
or mechanical means (including photocopying, recording, or information storage and
retrieval) without permission in writing from the publisher.
For information about special quantity discounts, please email special_sales@mitpress

.mit.edu
This book was set in Syntax and Times Roman by Toppan Best-set Premedia Limited.
Printed and bound in the United States of America.
Library of Congress Cataloging-in-Publication Data
McGhee, George R.
Convergent evolution : limited forms most beautiful / George R. McGhee Jr.
p. cm.
Includes bibliographical references and index.
ISBN 978-0-262-01642-1 (hardcover : alk. paper)
1. Convergence (Biology) I. Title.
QH373.M34 2011
576.8—dc22
2011007840
10 9 8 7 6 5 4 3 2 1
www.ebook3000.com
For Marae
’S òg a thug mi gràdh dhuit.
www.ebook3000.com
www.ebook3000.com
Contents
Series Foreword ix
Preface: Limited Forms Most Beautiful xi
1 What Is Convergent Evolution? 1

Recognizing Convergent Evolution 1
Convergent Evolution versus Parallel Evolution? 5
Evolutionary Constraint and Convergent Evolution 7
Mimicry, Camouflage, and Convergent Evolution 8
On the Organization of This Book 9
2 Convergent Animals 13
Locomotion 13
Carnivores: Prey Detection 30
Carnivores: Prey Capture 42
Herbivores: Plant Processing 50
Defense: Antipredator Adaptations 60
Organ Systems 66
Reproduction 81
3 Convergent Plants 93
Arborescence 93
Photoautotrophs: Light Capture 97
Photoautotrophs: Carbon Processing 101
Defense: Antiherbivore Adaptations 105
Defense: Antidehydration Adaptations 112
Reproduction 115
4 Convergent Ecosystems 135

One Ecological Role, Many Convergent Players 135
One Ecosystem Play, Many Convergent Casts of Actors 155
Isochronous and Heterochronous Ecosystem Convergence 163
A Periodic Table of Niches? 168
www.ebook3000.com
viii Contents
5 Convergent Molecules 177

Convergent Molecules? 177
DNA 178
Proteins and Protein Functions 189
Enzymes and Enzyme Functions 199
Convergent Evolution of Cellular and Tissue Structures? 205
6 Convergent Minds 209

Convergent Minds? 209
Problem-Solving Behavior 209
Group Behavior 228
Convergent Mentalities 238
7 Functional and Developmental Constraint in Convergent Evolution 245

Convergent Evolution in Theoretical Morphospace 246
Life as We Know It 252
Life as We Do Not Know It 253
A Periodic Table of Life? 260
8 Philosophical Implications of Convergent Evolution 265

The Argument for Unpredictability: Creative Freedom or Chaotic
Randomness? 266
The Argument for Predictability: Comforting Certainty or Depressing
Inevitability? 269
Judging the Arguments: Evolutionary Views on Trial 271
The Question of Extraterrestrial Life 276
In Conclusion: A Rewrite of Darwin’s View of Life 276
Appendix: A Phylogenetic Classification of Life 279

References 287
Index of Common Names 303
Index of Species 309
Index of Topics 319
www.ebook3000.com
Series Foreword
Biology is becoming the leading science in this century. As in all other

sciences, progress in biology depends on interactions between empirical
research, theory building, and modeling. But whereas the techniques and
methods of descriptive and experimental biology have evolved dramati-
cally in recent years, generating a flood of highly detailed empirical data,
the integration of these results into useful theoretical frameworks has
lagged behind. Driven largely by pragmatic and technical considerations,
research in biology continues to be less guided by theory than seems
indicated. By promoting the formulation and discussion of new theoreti-
cal concepts in the bio-sciences, this series intends to help fill the gaps in
our understanding of some of the major open questions of biology, such
as the origin and organization of organismal form, the relationship
between development and evolution, and the biological bases of cogni-
tion and mind.
Theoretical biology has important roots in the experimental biology
movement of early-twentieth-century Vienna. Paul Weiss and Ludwig
von Bertalanffy were among the first to use the term theoretical biology
in a modern scientific context. In their understanding the subject was not
limited to mathematical formalization, as is often the case today, but
extended to the conceptual problems and foundations of biology. It is
this commitment to a comprehensive, cross-disciplinary integration of
theoretical concepts that the present series intends to emphasize. Today
theoretical biology has genetic, developmental, and evolutionary com-
ponents, the central connective themes in modern biology, but also
includes relevant aspects of computational biology, semiotics, and cogni-
tion research, and extends to the naturalistic philosophy of sciences.
The “Vienna Series” grew out of theory-oriented workshops organized
by the Konrad Lorenz Institute for Evolution and Cognition Research
(KLI), an international center for advanced study closely associated with
x Series Foreword
the University of Vienna. The KLI fosters research projects, workshops,

archives, book projects, and the journal Biological Theory, all devoted to
aspects of theoretical biology, with an emphasis on integrating the devel-
opmental, evolutionary, and cognitive sciences. The series editors
welcome suggestions for book projects in these fields.
Gerd B. Müller, University of Vienna and KLI

Günter P. Wagner, Yale University and KLI
Werner Callebaut, Hasselt University and KLI
Preface: Limited Forms Most Beautiful
In 1859, Charles Darwin concluded his momentous book On the Origin

of Species by musing on the grandeur of the evolutionary view of life:
“from so simple a beginning endless forms most beautiful . . . have been,
and are being, evolved” (Darwin 1859, 490). A century and a half after
Darwin published those words, many evolutionary biologists are ques-
tioning whether the evolution of life actually does produce “endless
forms.” In fact, in many cases we see that evolution has produced the
same form—or a very similar one—over and over again in many inde-
pendent species lineages, repeatedly, on timescales of hundreds of mil-
lions of years.
The phenomenon of convergent evolution has led many of us to
suggest that that famous sentence should perhaps be rewritten to read,
“from so simple a beginning limited forms most beautiful . . . have been,
and are being, evolved.” The purpose of this book is to reveal to the
reader how ubiquitous the phenomenon of convergent evolution is in
life and that it occurs on all levels of evolution, from tiny organic mol-
ecules to entire ecosystems of species, and even in the way in which we
think.
Why does evolution produce limited forms most beautiful? The phe-
nomenon of convergent evolution leads directly to the concept of evo-
lutionary constraint—that is, the number of evolutionary pathways
available to life is in fact not endless, but is quite restricted. In the final
two chapters of the book, the causes of convergent evolution and the
implications of evolutionary constraint will be examined in detail. Those
implications are not only of scientific but also of philosophical interest.
And those implications all concern, in one way or another, an essential
question: How predictable is the evolutionary process?
My fascination with convergent evolution dates back to my first visit,
as a small boy, to a natural history museum and zoological garden. I was
xii Preface
astonished to see that the eyes of the deadly pit viper, gazing coldly
at me through the glass enclosure of its pen, were identical to the eyes
of my beloved black cat back at home. Both animals had eyes with very
odd-looking vertical-slit pupils and irises of a beautiful gold color. How
could it be that this deadly snake, a cold-blooded scaly reptile so vastly
different from a warm, furry mammal, had eyes like a cat? In my later
university studies I began to realize that the number of evolutionary
possibilities was not endless, and that there was a very good reason
why the eyes of a predatory snake and a predatory mammal are so
similar. By the end of this book, I hope the reader will also share my
fascination with an evolutionary process that produces limited forms
most beautiful.
I thank the Konrad Lorenz Institute for Evolution and Cognition
Research, Altenberg, Austria, for the fellowship that enabled me to work
at the institute in 2010 and to bring this book to completion. I thank
Gerd Müller, director of the KLI, for all of his help in preparing this
book for the Vienna Series in Theoretical Biology. I thank all my col-
leagues (in particular Karl Niklas and Doug Erwin) who have discussed
the scientific implications of convergent evolution with me over the
years. I thank Simon Conway Morris, Celia Deane-Drummond, and
Michael Ruse for our discussions of the philosophical implications of
convergent evolution while we attended a conference on the subject at
the Vatican Observatory, Castel Gandolfo, Italy (and I thank Simon for
inviting me to the conference). Needless to say, philosophical opinions
expressed in this book are my own. Finally, I thank my wife, Marae, for
her patient love.
1 What Is Convergent Evolution?
The question “How common is convergence?” remains unanswered and may be

unanswerable. Our examples indicate that even the minimum detectable levels
of convergence are often high, and we conclude that at all levels convergence
has been greatly underestimated.
—Moore and Willmer (1997, 1)
Recognizing Convergent Evolution
A porpoise looks like a fish. It has a fusiform, streamlined body like that
of a swordfish or a tuna. It has four fins on the ventral side of its body,
instead of four legs. It has a large fin at its posterior end, instead of a tail.
And it even has a vertical fin centered on its back, so it looks very much
like a shark when it is swimming through the water toward you.
Astonishingly, all appearances to the contrary, a porpoise is not a fish;
it is a mammal. It possesses all the distinctive combination of mammalian
traits: a porpoise is placental, gives live birth, nurses its young with spe-
cialized mammary glands, has an endothermic metabolism, has three
bones in its inner ears, has mammalian milk teeth, reproduces via inter-
nal fertilization, and so on. But it has lost other traits that are found in
most mammals: it has no legs, no tail, no fur, and has instead evolved fins
like those of a fish in place of legs and a tail.
The porpoise inherits its mammalian traits from its mammalian
ancestors, and the possession of these traits indicates to us that the por-
poise belongs to the evolutionary lineage, or clade, of the Mammalia.
These traits are synapomorphies, derived traits that are inherited from
a common ancestor and that define membership within a particular
clade.
The fins of the porpoise are not directly inherited from fish ancestors;
they are independently derived convergent traits. That is, porpoises
have evolved fins that have converged on the morphology of the fins that
2 Chapter 1
are seen in the fish, and, even though the fins of the porpoise look very
much like those of a fish, they are in fact not fish fins but rather are
mammal fins.
Distinguishing synapomorphic traits from convergent traits is critical
in the recognition of convergent evolution. In the literature (particularly
the older literature), synapomorphic traits are sometimes called second-
ary homologous traits, and convergent traits are called homoplastic traits
(Lecointre and Le Guyader 2006). Thus, the critical determination, for
the purpose of recognizing convergent evolution, is sometimes referred
to as the determination of secondary homology versus homoplasy (a
term that I have always found to be most uneuphonious—it sounds like
some type of disease).
There are actually three ways in which homoplastic traits—traits that
look similar but that are not inherited from a common ancestor—might
arise in evolution: convergence, parallelism, and reversion. Let us con-
sider a hypothetical example in which we know the evolutionary rela-
tionships between six species, illustrated by the cladogram given in figure
1.1. Species 1, 2, and 3 all possess synapomorphy S, and thus belong to
the monophyletic clade S. Species 4, 5, and 6 all possess synapomorphy
T, and thus belong to the monophyletic clade T. Although clade S and
clade T have diverged in their evolution, they nevertheless evolved from
a common ancestor that evolved the derived trait R, a trait that all six
species still possess by inheritance; thus, trait R is a synapomorphy
for the larger monophyletic clade R, which contains both clade S and
clade T.
So far, so good. Now let us suppose that in clade S, the new trait Z
arises by evolutionary modification of the preexisting trait A, as seen in
species 3 (figure 1.2). However, in clade T a new trait very similar to trait
Figure 1.1
Cladogram of evolutionary relationships between six hypothetical species.
What Is Convergent Evolution? 3
Figure 1.2
Convergent evolution of trait Z in species 3 and 6.
Z also arises, but by evolutionary modification of the preexisting trait B,

as seen in species 6 (figure 1.2). Trait Z is thus a convergent trait, a
homoplastic trait, having evolved independently in species 3 and species
6. If we were to mistakenly think that trait Z was a synapomorphic trait,
we would mistakenly think that species 3 and species 6 were closely
related sister species, since they both possess trait Z, and we would thus
mistakenly include species 3 and species 6 in the new clade Z. The erro-
neous clade Z would thus be a polyphyletic clade. If, for example,
we were to include swordfish and porpoises together in the clade
“fins” because we mistakenly believed that fins were a synapomorphic
trait, we would have created an erroneous polyphyletic clade because
the fins present in these two species are convergent traits, not
synapomorphies.
Parallel evolution is a particular type of convergent evolution, where
parallelism is “a similarity that has appeared independently in different
closely related taxa,” closely related in that the same trait has indepen-
dently evolved “from the same ancestral character in different taxa. A
parallelism is a special case of convergence” (Lecointre and Le Guyader
2006, 541). A hypothetical case of parallel evolution is illustrated in figure
1.3. All six species possess the same ancestral character R in the original
cladogram given in figure 1.1. In figure 1.3, however, trait R has been
modified into trait Z independently, in parallel, in both species 3 and
species 6.
The conceptual difference between the evolutionary scenarios illus-
trated in figures 1.2 and 1.3 lies in the traits that have been modified in
producing the convergent trait Z. In figure 1.2, two separate traits, A and
B, have been modified in two separate lineages to produce the same
convergent trait Z. In figure 1.3, only one trait, the ancestral trait R, has
4 Chapter 1
Figure 1.3
Parallel evolution of trait Z in species 3 and 6.
Figure 1.4
Reverse evolution of trait R in species 6 within clade T, whereas all the species of clade S
simply inherit trait R from a common ancestor.
been modified in two separate lineages to produce the parallel conver-

gent trait Z.
Last, a homoplastic trait can also be produced by the process of reverse
evolution. In figure 1.4, trait R has been modified into trait B, a new
derived trait possessed by species 4 and species 5. In species 6, however,
trait B has been modified back into trait R—an evolutionary reversion.
Species 1, 2, and 3 all possess trait R by inheritance from a common
ancestor. Since species 6 did not inherit trait R directly from that ances-
tor, trait R is a homoplastic reversion in this species. Consequently, it
would be erroneous to conclude that species 6 was more closely related
to species 1, 2, and 3 than it is to species 4 and 5 simply because it also
possesses trait R. We would create an erroneous polyphyletic clade if we
were to mistakenly include species 6 in clade S along with species 1, 2,
and 3. Instead, species 6 possesses synapomorphy T (figure 1.1) and
belongs in clade T along with species 4 and 5, even though it does not
possess trait B as species 4 and 5 do (figure 1.4). At present, the impor-

tance and amount of homoplasy in evolution that is produced by rever-
sion is uncertain (Spinney 2007).
It is easy for us to see the conceptual differences between convergent,
parallel, and reverse evolution in figures 1.2–1.4 because we already
knew the true evolutionary relationships of the six hypothetical species
(figure 1.1). In real life, however, evolutionary relationships are not
known beforehand, and we must carefully analyze all traits that we see
are shared by different species. Do these species share these traits simply
because they all inherit them from a common ancestor? Or are these
traits convergent—have they independently arisen in species that belong
to separate evolutionary lineages, in separate clades? A porpoise is not
a fish, even though it looks like one.
Convergent Evolution versus Parallel Evolution?
In modern phylogenetic systematics, we see that parallel evolution is

simply a special case of convergent evolution (Lecointre and Le Guyader
2006). In the literature (particularly the older literature), not only is
parallel evolution sometimes erroneously considered to be different
from convergent evolution, but in some cases it is considered to be the
actual disproof of convergent evolution. For example, in the next chapter
we will see that eyes have convergently evolved in animals in some 49
independent lineages (table 2.15). It was long thought that each separate
animal lineage convergently evolved its type of eye from separate origi-
nal traits, in a case of classic convergent evolution as illustrated in figure
1.2. However, we now know that all animal eyes are produced by modify-
ing the same conserved regulatory gene present in the animal genome
(the Pax-6 gene), and thus the evolution of eyes is a case of parallel
convergent evolution, as illustrated in figure 1.3. The recent molecular
discovery that the same ancestral character was independently, in paral-
lel, modified in the evolution of eyes was mistakenly taken by Rey et al.
(1998) to be a disproof of convergent evolution: “An oft invoked example
of convergent evolution has been compound eyes of insects versus sin-
gular eyes of vertebrates. This proved to be a wrong example, however,
because development of all metazoan eyes recently has been shown to
be under control of the same regulatory gene that encodes Pax-6 protein”
(Rey et al. 1998, 6212).
All life on Earth evolved from a single common ancestor. We know
this because all life on Earth uses the exact same coding molecule, DNA,
6 Chapter 1
and the exact same molecular construction: DNA codes for RNA, and
RNA codes for the same 20 amino acids that are assembled to produce
proteins, the building blocks of life. This fact leads us to the concept of
“deep homology” and parallelism in evolution: “One of the most impor-
tant, and entirely unanticipated, insights in the past 15 years was the
recognition of an ancient similarity in patterning mechanisms in diverse
organisms, often among structures not thought to be homologous on
morphological or phylogenetic grounds. In 1997, prompted by the
remarkable extent of similarities in genetic regulation between organs
as different as fly wings and tetrapod limbs, we suggested the term
‘deep homology’ to describe the sharing of the genetic regulatory appa-
ratus that is used to build morphologically and phylogenetically dispa-
rate animal features” (Shubin et al. 2009, 818). However, rather than
viewing deep homology as a disproof of convergent evolution, Shubin
et al. (2009) recognize that it is proof of parallel convergent evolution:
“The deep homology of generative processes and cell-type specification
mechanisms in animal development has provided the foundation for
the independent evolution of a great variety of structures . . . common
genetic mechanisms are used to generate diverse adaptations and can
lead to the parallel evolution of novelties” (Shubin et al. 2009, 818; italics
mine).
In actual practice it is sometimes quite difficult to differentiate con-
vergent and parallel evolution. Rokas and Carroll (2008) give unambigu-
ous examples of both types of evolution at the molecular level: both
dolphins and rhinos have independently evolved the amino acid valine
at the same amino acid site. The dolphins, however, modified alanine to
valine at that site, whereas the rhinos modified tyrosine to alanine. The
same amino acid, valine, was independently evolved from two different
amino acid precursors in a process of convergent evolution (the process
illustrated in figure 1.2). In contrast, the filamentous ascomycetes Asper-
gillus clavatus and Aspergillus oryzae have independently evolved glu-
tamic acid at the same amino acid site in parallel, in that both modified
their ancestral amino acid asparagine to glutamic acid at that site (the
process illustrated in figure 1.3).
In other cases the determination is not so easy. The wings of birds and
bats are overwhelmingly used in the literature as examples of convergent
evolution. And indeed these are two very distantly related animals—the
bats belong to the clade Mammalia and the birds to the clade Dinosauria.
We have to go back in time to the Carboniferous, some 340 million years
ago, to find a common ancestor for these two animals. Both have wings,
but the birds modified dinosaurian forelimbs to wings whereas bats

modified mammalian forelimbs to wings in what was a process of con-
vergent evolution (the process illustrated in figure 1.2). Or perhaps it was
not? Even though dinosaurian forelimbs are very different from mam-
malian forelimbs, they are both forelimbs. And both dinosaurs and
mammals have inherited (and modified) their forelimbs from their
ancient amphibian ancestors. One could thus just as well argue that the
evolution of wings in birds and bats is a case of parallel evolution: the
same ancestral character (forelimbs) was modified in parallel to produce
wings in birds and in bats (the process illustrated in figure 1.3). Argu-
ments of deep homology would push that parallelism even further back
in time, in that the genetic regulatory mechanisms used to produce fore-
limbs were evolved before the Cambrian diversification of animal life
over 540 million years ago.
In the chapters that follow, I will not draw a sharp distinction between
convergent evolution and parallel evolution. Again, from the perspective
of modern phylogenetic systematics, parallel evolution is simply a type
of convergent evolution (Lecointre and Le Guyader 2006). Where I think
the distinction is useful, and reasonably well demonstrated, I will point
out examples of these two evolutionary processes in operation.
Evolutionary Constraint and Convergent Evolution
Convergent evolution was once thought to be almost exclusively pro-

duced by functional constraint, the old maxim that “form follows func-
tion.” Given the same function, evolution via natural selection will
produce the same form to serve that function. However, modern studies
in evolutionary development (“evo-devo” to its practitioners) have
revealed that a great deal of convergent evolution of life on Earth is in
fact due to developmental constraint: the phenomenon of “deep homol-
ogy” that we considered in the previous section of the chapter is just one
example. Ancient homologous regulatory genes have been indepen-
dently switched on and off many times in the evolution of life to produce
convergent forms, such as the many types of eyes seen in animals (which
we will consider in the next chapter). The same forms have been pro-
duced by the repeated channeling of evolution along the same develop-
mental trajectory. Natural selection has a limited repertoire of potential
forms from which to choose, and convergent evolution is the result.
In order to understand the phenomenon of convergent evolution,
we now need to carefully consider the action of both functional and
8 Chapter 1
developmental constraints in the evolutionary process: “Patterns of par-

allel evolution can provide even stronger illustrations of the need to
distinguish explanations based on the similarity of natural selection from
those involving developmental bias or genetic channeling” (Brakefield
2006, 364). Throughout this book I will point out examples of functional
and developmental constraint in operation. The theoretical analysis of
evolutionary constraint in producing convergent evolution will be con-
sidered in detail in chapter 7.
Mimicry, Camouflage, and Convergent Evolution
Mimicry is a form of convergent evolution in which one species indepen-

dently evolves a morphology very similar to that of another species
simply in order to fool a third species. In Batesian mimicry, a harmless
mimic converges on the morphology of a harmful model species in a case
of false warning—potential predators will avoid the harmless mimic,
thinking it is the harmful model. An example is the harmless New Mexico
milk snake (Lampropeltis triangulum celaenops), which looks almost
identical to the deadly Arizona coral snake (Micruroides euryxanthus).
In Müllerian mimicry, two harmful species converge on a common mor-
phology in a case of real warning—potential predators easily recognize
the harmful species because the warning morphology is twice as abun-
dant in nature as it would be if it was characteristic of only a single
species. An example is the yellow-and-black striping seen in both bees
and wasps.
Mimicry is similar to camouflage, where species evolve morphologies
that converge on the form of either a living or a nonliving model in order
to blend into the surroundings so that the camouflaged species cannot
be seen. An example of the former is the giant walkingstick (Mega-
phasma dentricus), an insect that so closely resembles the leafless twig
of a plant that it is unnoticed by predators. An example of the latter is
the white fur color that is convergently evolved by numerous animal
species that live in snowy habitats, such as the Asian snow leopard (Uncia
uncia) and the Arctic polar bear (Thalarctos maritimus).
All the myriad forms of mimicry and camouflage convergence serve a
single function: deception. Predator species repeatedly and indepen-
dently evolve forms that deceive prey animals into not noticing them
(until it is too late), and prey animals independently evolve forms that
deceive predators so that the prey either are not noticed or are noticed
but avoided. Although the convergence in form that is seen in mimicry
and camouflage in nature is sometimes spectacular, the fundamental

function of that convergence is rather simple, in that it is all deception.
Entire books have been written on the evolution of mimicry and cam-
ouflage, and I will not consider these forms of convergence in more detail
in this book, with the exception of the carrion-mimic flowers, discussed
in chapter 3.
On the Organization of This Book
This book is not intended to be an “Encyclopedia of Convergent Evolu-

tion” (an effort that would run to many volumes). Instead, I have con-
centrated on listing a single species example from as many separate and
disparate phylogenetic lineages that I am aware of in which a trait has
originated independently by convergent evolution.
A phylogenetic classification of life is absolutely essential in determin-
ing whether a similar trait found in separate species is shared by those
species simply because they inherited it from a common ancestor—that
is, a synapomorphic trait—or whether that trait has arisen independently
in each separate species lineage—and is thus a convergent trait. The
phylogenetic classification that I use throughout this book is given in
abbreviated form in table 1.1 and in detail in the appendix. This classifi-
cation is modified chiefly from the phylogenetic classification of living
life forms by Lecointre and Le Guyader (2006) and the APG II (2003);
additional information on extinct life forms is taken from Benton (2005),
Donoghue (2005), and Niklas (1997).
For each species example given in the chapters that follow, I have
provided an abbreviated list of the major nodes in its phylogenetic clas-
sification, which will allow the reader to quickly see the close or distant
relationship of a group of species that have independently evolved a
similar trait by convergent evolution. I have not repeated the entire
sequence of nodes, as that would be redundant with the appendix and
would burden the lists of species with information overload. The major
clades I have used in the text of this book are given in table 1.1 and are
set in boldface type in the appendix, allowing the reader to quickly
assemble the entire evolutionary sequence of nodes for any species
lineage in which he or she is particularly interested.
For each individual convergent structure or trait that is discussed in
the chapters that follow, I will list the phylogenetic lineages of species
that have independently evolved that structure or trait in a “step down”
fashion from least derived to most derived; that is, I will list lineages that
10 Chapter 1
Table 1.1
The major lineages of life on Earth
Eubacteria
Archaea
Eukarya
– Bikonta
– – Chlorobiota
– – – Embryophyta
– – – – Tracheophyta
– – – – – Lycophyta
– – – – – Euphyllophyta
– – – – – – Moniliformopses
– – – – – – Lignophyta
– – – – – – – Spermatophyta
– – – – – – – – Angiospermae
– – – – – – – – – Euangiosperms
– Unikonta
– – Metazoa
– – – Cnidaria
– – – Bilateria
– – – – Protostomia
– – – – – Lophotrochozoa
– – – – – Ecdysozoa
– – – – Deuterostomia
– – – – – Echinodermata
– – – – – Chordata
– – – – – – Vertebrata
– – – – – – – Chondrichthyes
– – – – – – – Osteichthyes
– – – – – – – – Actinopterygii
– – – – – – – – Sarcopterygii
– – – – – – – – – Tetrapoda
– – – – – – – – – – Batrachomorpha
– – – – – – – – – – Reptiliomorpha
– – – – – – – – – – – Amniota
– – – – – – – – – – – – Sauropsida
– – – – – – – – – – – – – Diapsida
– – – – – – – – – – – – – – Lepidosauromorpha
– – – – – – – – – – – – – – Archosauromorpha
– – – – – – – – – – – – Synapsida
– – – – – – – – – – – – – Therapsida
– – – – – – – – – – – – – – Mammalia
Note: The complete phylogenetic classification of life used in this book is given in the
appendix.
occur at the top of table 1.1 first and arrange subsequent lineages in their
sequence of evolution as outlined in table 1.1. Thus, in the discussion of
the convergent evolution of eyespots in animals given in chapter 2, the
nonbilaterian jellyfish species Leuckartiara octona (lineage Metazoa:
Cnidaria in table 1.1) is listed before the more derived bilaterian rotifer
species Asplanchna brightwelli (lineage Metazoa: Bilateria: Protostomia:
Lophotrochozoa in table 1.1), which itself is listed before the deuteros-
tomous sea cucumber species Synaptula lamperti (lineage Metazoa: Bila-
teria: Deuterostomia: Echinodermata in table 1.1).
I will take the list of nodes for each pair of convergent lineages back
to the point where the lineages diverged in evolution so the reader can
see how distantly related species are that have independently acquired
a convergent trait. Thus, in the previous example, the living sea cucumber
Synaptula lamperti and the rotifer Asplanchna brightwelli both have
eyespots, yet you have to go all the way back to the evolutionary split
between the protostomous and deuterostomous lineages of bilaterian
animals to find a common ancestor for the two species (Bilateria: Proto-
stomia versus Bilateria: Deuterostomia). In order to find a common
ancestor for these bilaterian animals and the cnidarian jellyfish Leuck-
artiara octona, you have to go all the way back to the evolution of
the metazoans themselves, over 600 million years ago (i.e., Metazoa:
Cnidaria versus Metazoa: Bilateria).
Published sources for all the convergent lineages discussed in this
book are given in the references. These have partially come from my own
lists of convergences, amassed over many years, and all the many text-
books and encyclopedia articles on evolution I have read, all of which
contain examples of convergent evolution (in many cases, the same
ones). In addition, three compendia that are exclusively devoted to the
subject of convergent evolution, and that have been particularly helpful
in my analyses, are Conway Morris’s Life’s Solution (2003), Barlow’s Let
There Be Sight! A Celebration of Convergent Evolution (2005), and
Berg’s Nomogenesis, or Evolution Determined by Law (1922), and I
recommend these to the reader as they also contain many additional
examples not included in this volume.
The reader can quickly grasp the structure of my analysis of the
phenomenon of convergent evolution by reading the section headings
within each of the individual chapters given in the table of contents of
the book. Throughout the book I will discuss the functional and devel-
opmental constraints that have resulted in convergent evolution at every
level of life, from the external forms of living organisms down to the very
12 Chapter 1
molecules from which they are constructed, from their ecological roles
in nature to the way in which their minds function. I will also use the
analytical techniques of theoretical morphology in that discussion
(McGhee 1999, 2001, 2007). In theoretical morphology, the consideration
of nonexistent biological form is just as important as that of existent
biological form, and can give us valuable insights into the phenomenon
of convergent evolution.
2 Convergent Animals
In all cases of two very distinct species furnished with apparently the same
anomalous organ, it should be observed that, although the general appearance
and function of the organ may be the same, yet some fundamental difference can
generally be detected. I am inclined to believe that in nearly the same way as
two men have sometimes independently hit on the very same invention, so
natural selection, working for the good of each being and taking advantage of
analogous variations, has sometimes modified in very nearly the same manner
two parts in two organic beings, which owe but little of their structure in common
to inheritance from the same ancestor.
—Darwin (1859, 193–194)
Locomotion
Some of the most spectacular examples of convergent evolution are

clearly due to the functional constraints of locomotion. Consider one
of the most frequently cited cases of convergent evolution: the aston-
ishing morphological similarity between the extinct Mesozoic marine
reptile Ichthyosaurus platyodon and the living marine mammal Pho-
caena phocaena, the harbor porpoise, or Tursiops truncatus, the bottle-
nose dolphin. Not only do they look amazingly similar to one another,
but they all look amazingly similar to large, fast-swimming fishes like
Xiphias gladius, the swordfish, or Carcharodon carcharias, the great
white shark. The cartilaginous fishes and the bony fishes both solved the
physics of swimming in the dense medium of water back in the Silurian
by evolving streamlined, fusiform morphologies (table 2.1). Some 230
million years later, a group of land-dwelling reptiles rediscovered this
same morphology in their evolutionary return to the sea. And around
175 million years after that, a group of land-dwelling mammals also
rediscovered this same morphology in their own evolutionary return to
the sea.
14 Chapter 2
Table 2.1
Convergent evolution of animal swimming morphologies
1 Convergent structure and function: FUSIFORM BODY (missile-shaped form for drag
reduction)
Convergent lineages:
1.1 Great white shark (Vertebrata: Chondrichthyes: Elasmobranchii: Lamnidae;
Carcharodon carcharias)
1.2 Swordfish (Vertebrata: Osteichthyes: Actinopterygii: Teleostei: Xiphiidae; Xiphias
gladius)
1.3 Ichthyosaur (Osteichthyes: Saurcopterygii: Reptiliomorpha: Amniota: Sauropsida:
Diapsida: Ichthyosauria: Ichthyosauridae; Icthyosaurus platyodon †Jurassic)
1.4 Harbor porpoise (Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
Laurasiatheria: Cetartiodactyla: Cetacea: Odontoceti: Phocaenidae; Phocaena
phocaena)
2 Convergent structure and function: EEL-SHAPED BODY (elongated, cylindrical form
for drag reduction)
2.1 Reed fish (Osteichthyes: Actinopterygii: Cladista: Polypteridae; Erpetoichthyes
calabaricus)
2.2 European eel (Actinopterygii: Teleostei: Elopomorpha: Anguillidae; Anguilla
anguilla)
2.3 South American lungfish (Osteichthyes: Sarcopterygii: Dipnoi: Lepidosirenidae;
Lepidosiren paradoxa)
2.4 Mosasaur (Sarcopterygii: Reptiliomopha: Amniota: Sauropsida: Diapsida:
Lepidosauromorpha: Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Mosasauridae; Platecarpus ictericus †Cretaceous)
2.5 Geosaur (Diapsida: Archosauromorpha: Crurotarsi: Crocodilia: Metriorhynchidae;
Geosaurus giganteus †Jurassic)
3 Convergent structure and function: PADDLE-FORM APPENDAGES (walking
appendages modified to paddle shapes for paddling/rowing through water)
3.1 Sea scorpions (Bilateria: Protostomia: Ecdysozoa: Arthropoda: Cheliceriformes:
Merostomata: Eurypterida: Eurypteridae; Eurypterus remipes †Silurian)
3.2 Marbled diving beetle (Arthropoda: Mandibulata: Hexapoda: Dytiscidae;
Thermonectes marmoratus)
3.3 Green sea turtle (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Reptiliomorpha: Amniota: Sauropsida: Anapsida: Testudines: Chelonioidae; Chelonia
mydas)
3.4 Plesiosaur (Sauropsida: Diapsida: Lepidosauromorpha: Sauropterygia:
Elasmosauridae; Muraenosaurus leedsii †Jurassic)
3.5 King penguin (Diapsida: Archosauromorpha: Ornithodira: Dinosauria: Saurischia:
Theropoda: Maniraptora: Aves: Sphenisciformes: Spheniscidae; Aptenodytes patagonica)
3.6 African manatee (Amniota: Synapsida: Therapsida: Mammalia: Eutheria: Afrotheria:
Sirenia: Trichechidae; Trichechus senegalensis)
3.7 California sea lion (Eutheria: Laurasiatheria: Carnivora: Caniformia: Otariidae;
Zalophus californianus)
Note: The geological age of extinct species is marked with a †.

Convergent Animals 15
The evolution of an ichthyosaur or a porpoise morphology is not

trivial. It can be correctly described as nothing less than astonishing that
a group of land-dwelling tetrapods, complete with four legs and a tail,
could devolve their appendages and tails back into fins like those of a
fish, while also reevolving a dorsal fin. Highly unlikely, if not impossible?
Yet it has happened twice, convergently in the reptiles (sauropsid amni-
otes) and the mammals (synapsid amniotes), two groups of animals that
are not closely related. We have to go back in time as far as the Carbon-
iferous to find a common ancestor for the mammals and the reptiles; thus,
their genetic legacies are very different. Nonetheless, the ichthyosaur and
the porpoise both independently reevolved fins.
An extremely ancient form of swimming morphology can be seen in
some of the first chordate animals to evolve back in the Cambrian, not
long after the evolution of animals themselves some 600 million years
ago. These animals have elongated cylindrical bodies, usually laterally
compressed, with caudal fins located along the dorsal and ventral sur-
faces of the posterior one-third or so of the body. Superficially, these
animals look somewhat like arrows, but rather than moving rigidly like
an arrow, these animals undulate through the water, with a sinusoidal
motion of the body, propelled by the caudal fins at the rear. Such eel-like
morphologies can be seen in the extinct conodonts, such as Eoconodon-
tus notchpeakensis (the “dawn conodont” from Notchpeak) from the
Late Cambrian, which were very ancient craniate chordates possessing
a skull but no jaws or vertebrae. The same plesiomorphic, eel-like mor-
phology can be seen in the modern myxinoid hagfish Myxine glutinosa,
which also have no jaws or vertebrae, and in the modern sea lamprey
Petromyzon marinus, which does possess vertebrae but still has no jaws.
Surprisingly advanced, derived animals have repeatedly converged on
this ancient swimming morphology—including the modern eels them-
selves, such as the European eel Anguilla anguilla (table 2.1). Modern
eels are highly modified teleost fish, the Elopomorpha, and the teleost
fishes themselves are highly derived actinopterygian fish. Yet a modern
eel looks very much like an ancient Cambrian craniate, and swims in the
same undulatory fashion, propelled by caudal fins. The teleost fishes were
not the first to evolve eel-forms among the ray-fin fishes, however, as this
same form was convergently evolved within the plesiomorphic cladistian
fishes, a basal group of actinopterygian fishes, as can be seen in the
modern reed fish.
This same morphology has been convergently evolved in the second
major extant branch of the bony fishes, the lobe-finned sarcopterygians,
16 Chapter 2
as can be seen in the South American lungfish Lepidosiren paradoxa,

which is virtually identical to an eel. And, as one might expect from the
previous example of the convergent evolution of fusiform swimming
morphologies, these convergences are not confined to disparate types of
fish (table 2.1). The Cretaceous mosasaurs were eel-like marine reptiles,
closely resembling the mythical sea serpent, and evolved from land-
dwelling lizards. The Jurassic geosaurs were eel-like marine archosaurs,
a more derived form of reptile than the lizards, yet they too evolved from
land-dwelling tetrapods that devolved the tetrapods back into fins and
returned to the sea. Both the mosasaurs and geosaurs developed caudal
fins by dorsoventrally elongating, and laterally compressing, the caudal
vertebrae themselves.
In analyzing evolution from the perspective of theoretical morphology,
it is important to consider both nonexistent convergent form and existent
convergent form (McGhee 2001, 2007). There is one major group of land
animals that has not yet convergently evolved an eel-like swimming
morphology—the mammals (see table 2.1). Mammalian “fish-forms”
have evolved in the Cenozoic (porpoises, etc.), but no mammalian “eel-
forms” have evolved as yet. Why not? At present, the sea otters (Enhydra
lutris), with their webbed feet, elongated bodies, and long flattened tails
used in swimming, might be the best example of a precursor to the con-
vergent evolution of the eel-form in mammals. Careful consideration of
nonexistent convergent form may give us valuable insights into the phe-
nomenon of convergent evolution, and we shall return to this topic
throughout this book.
Last, extensive convergence in form is seen in animals that have sec-
ondarily evolved the capability to paddle or row through water with their
former walking appendages (table 2.1). Among the arthropods, the
extinct sea scorpions (ancestors of modern land scorpions) modified a
pair of their fore appendages into two paddles back in the Silurian, and
420 million years later the modern diving beetles have modified a pair
of their hind appendages into paddles. All four of the tetrapods of
modern sea turtles have been modified into paddles, and appear very
similar in form to the four paddles of the ancient plesiosaurs, reptiles
that underwent this same modification some 170 million years before.
Among the advanced dinosaurian reptiles, the modern-day penguins
have modified their wings—forelimb structures that were themselves
modified from walking appendages back in the Jurassic—into paddles
for use in swimming rather than in flight. Two separate groups of modern
mammals have convergently modified their forelimbs into paddles: the
manatees and their kin, and the sea lions and their kin (table 2.1).
Although a manatee and a walrus look very similar, they are very differ-
ent types of mammals. The afrotherian manatee is more closely related
to an elephant, and the laurasiatherian walrus is more closely related to
a wolf, than they are to each other. Their apparent similarity in form is
entirely due to convergent evolution.
The second most commonly cited example of amazing feats in conver-
gent evolution is the evolution of flying in land animals. No fewer than
three separate groups of tetrapods have modified their forelimbs into
wings: the extinct pterosaurs and the living birds and bats (table 2.2). At
first glance, it would seem to be extremely unlikely that an originally
quadrupedal animal, using all four legs for walking, would somehow be
able to evolve into an animal in which the forelimb is now used as a
flying structure. Yet not only has this occurred in the evolution of life, it
has occurred three times independently! The wings of birds, bats, and
pterosaurs have an amazingly similar appearance, which is clearly due to
the functional constraints of locomotion via flying in the thin gaseous
medium of the Earth’s atmosphere. This similarity is deceptive, however,
for closer examination reveals that the wings of the three groups of
animals are constructed differently. The wing of a bat largely consists of
its hand, in which the fingers are vastly elongated and between which is
stretched a webbed membrane of skin. The wing of a pterosaur is similar,
but consists of a single finger (the fourth digit) from which a membrane
of skin is stretched all the way to the side of the animal. The wing of a
bird is even more different, in that the surface of the wing is not con-
structed of skin membranes, but consists of elongated flight feathers
attached to its arm. Still, wings in all three groups are convergent modi-
fications of an original walking structure, the forelimb.
The wings of insects, the fourth major group of animals to evolve
powered flight (table 2.2), are radically different from those of the tet-
rapods. The insects did not modify their forelimbs in the evolution of
wings. Rather, insect wings consist of modified gill branches originally
present only in the larval stages. Originally present on multiple segments
in the abdomens of larvae, these gill structures were expanded in some
body segments while being developmentally suppressed in others, leading
to the four-wing condition seen in primitive adult insects such as mayflies
(Carroll 2006).
In examining the convergent evolution of wings (table 2.2), we begin
to see the effect of a different constraint on the evolution of form in
nature, in addition to that of functional constraint. All three tetrapod
18 Chapter 2
Table 2.2
Convergent evolution of animal flying morphologies
1 Convergent structure and function: WINGS (flat to curved planar structures for
generating lift in active flying)
1.1 Green darner dragonfly (Bilateria: Protostomia: Ecdysozoa: Panarthropoda:
Arthropoda: Mandibulata: Hexapoda: Odonata: Aeschnidae; Anax junius)
1.2 Hairy devil pterosaur (Bilateria: Deuterostomia: Chordata: Osteichthyes:
Sarcopterygii: Reptiliomorpha: Amniota: Sauropsida: Diapsida: Archosauromorpha:
Ornithodira: Pterosauria: Rhamphorhynchidae; Sordes pilosus †Jurassic)
1.3 Great blue heron (Archosauromorpha: Ornithodira: Dinosauria: Saurischia:
Theropoda: Maniraptora: Aves: Neognathae: Neoaves: Ciconiiformes: Ardeidae; Ardea
herodias)
1.4 Mouse-eared bat (Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
Laurasiatheria: Chiroptera: Microchiroptera: Vespertilionidae; Myotis myotis)
2 Convergent structure and function: GLIDER MEMBRANES (expanded skin
membranes stretched between fore and hind limbs for retarding falling speed in passive
flying)
2.1 Flying Mesozoic mammal (Synapsida: Therapsida: Mammalia: Volaticotheria:
Volaticotheridae; Volaticotherium antiquum †Cretaceous)
2.2 Flying possum or sugar glider (Mammalia: Marsupialia: Diprotodontia: Petauridae;
Petaurus breviceps)
2.3 Feathertail glider (Marsupialia: Diprotodontia: Acrobatidae; Acrobatus pygmaeus)
2.4 Greater glider (Marsupialia: Diprotodontia: Pseudocheiridae; Petauroides volans)
2.5 True flying squirrel (Mammalia: Eutheria: Euarchontoglires: Rodentia: Sciuridae;
Glaucomys volans)
2.6 Scaly-tailed flying squirrel (Eutheria: Euarchontoglires: Rodentia: Anomaluridae;
Anomalurus derbianus)
2.7 Flying lemur or colugo (Eutheria: Euarchontoglires: Dermoptera: Cynocephalidae;
Cynocephalus volans)
3 Convergent structure and function: GLIDER BODIES (dorsoventrally flattened,
laterally expanded body for retarding falling speed in passive flying)
3.1 Flying diapsid reptile (Sauropsida: Diapsida: Weigeltisauridae; Coelurosauravus
jaekeli †Permian)
3.2 European flying Mesozoic lizard (Diapsida: Lepidosauromorpha: Kuehneosauridae;
Kuehneosaurus latus †Triassic)
3.3 North American flying Mesozoic lizard (Lepidosauromorpha: Kuehneosauridae;
Icarosaurus siefkeri †Triassic)
3.4 Black-bearded flying lizard (Lepidosauromorpha: Lepidosauria: Squamata: Iguania:
Agamidae; Draco melanopogon)
3.5 Flying paradise snake (Lepidosauria: Squamata: Scleroglossa: Autarchoglossa:
Anguimorpha: Serpentes: Colubridae; Chrysopelea paradisi)
4 Convergent structure and function: GLIDER FEET (expanded skin membranes
stretched between elongated toes in feet for retarding falling speed in passive flying)
4.1 Wallace’s flying tree frog (Tetrapoda: Batrachomorpha: Lissamphibia: Batrachia:
Anura: Rhacophoridae; Rhacophorus nigropalmatus)
4.2 Costa Rican flying tree frog (Lissamphibia: Batrachia: Anura: Hylidae; Agalychnis
spurrelli)
Table 2.2
(continued)
4.3 Kuhl’s flying gecko (Tetrapoda: Reptiliomorpha: Amniota: Sauropsida: Diapsida:

Lepidosauromorpha: Squamata: Scleroglossa: Gekkonta: Gekkonidae; Ptychozoon
kuhli)
5 Convergent structure and function: GLIDER FINS (greatly elongated, horizontally
oriented pectoral fins for retarding falling speed in passive flying)
5.1 Flying fish (Osteichthyes: Actinopterygii: Exocoetidae; Exocetus volitans)
5.2 Flying gurnard fish (Actinopterygii: Dactylopteridae; Dactylopterus volitans)
6 Convergent structure and function: GLIDER APPENDAGES (elongated, flattened
appendages and wide, flanged heads for retarding falling speed in passive flying)
6.1 Giant gliding ant (Arthropoda: Mandibulata: Hexapoda: Hymenoptera: Myrmicinae:
Cephalotini; Cephalotes atratus)
6.2 Mexican gliding ant (Arthropoda: Mandibulata: Hexapoda: Hymenoptera:
Pseudomyrmecinae; Pseudomymex gracilis)
6.3 South American gliding ant (Arthropoda: Mandibulata: Hexapoda: Hymenoptera:
Formicinae; Camponotus heathi)
Note: The geological age of extinct species is marked with a †. For data sources, see text.
groups modified their forelimbs in the evolution of wings because only

those structures were present for potential modification: unlike the
ancient insects, the tetrapods had no gill branches to modify. The triple
modification of forelimbs to wings in the tetrapods is an example of
developmental constraint, a constraint that did not apply to the insects
in this instance. Hypothetically, a group of insects could have modified
their forelimbs to wings, like the tetrapods, but they were not confined
to that developmental pathway and, as the geological record demon-
strates, they evolved along another developmental pathway not open to
the tetrapods.
One of the analytical techniques of theoretical morphology involves
the generation of nonexistent biological form, usually by mathematical
modeling and computer simulation (McGhee 2007). As a thought experi-
ment in theoretical morphology, we can consider the nonexistent mythi-
cal animal the dragon, a reptile-like animal with functional fore and hind
limbs and functional wings as well. From our previous discussion of the
convergent evolution of wings in tetrapods and insects, we can easily see
why dragons are nonexistent. First, a group of tetrapods would have to
develop an additional pair of forelimbs (the serial replication of a pair
of limbs is not developmentally impossible, but extremely rare in the
vertebrates) and, secondly, it would have to develop wings from the new
20 Chapter 2
pair of forelimbs. For the insects, such a morphological transition would

have been relatively simple, as they already possessed three pairs of
walking appendages, and it is interesting to note that they never took
this potential developmental pathway. Instead, they modified a different
structure, their larval gill branches, into flying structures. Later in this
chapter we shall consider another group of insects who in fact have
convergently modified their first pair of walking appendages into struc-
tures no longer used in locomotion.
Considerably more animals have convergently evolved gliding mor-
phologies than powered-flight morphologies (table 2.2). At least seven
separate groups of mammals have evolved the capability to glide by
stretching skin membranes between their fore and hind limbs, beginning
with the volaticoterian mammals in the Cretaceous. These ancient gliders
looked amazingly similar to modern-day southern flying squirrels, Glau-
comys volans, in North America. The true flying squirrels are widely
distributed around the planet, except in Africa. Thus, it is fascinating to
note that in Africa a separate group of rodents, the scaly-tailed flying
“squirrels” (which are not squirrels at all), have convergently evolved
the same morphology. In Southeast Asia, another group of mammals, the
flying lemurs or colugos, have converged on this morphology, and these
animals are not any type of rodent at all but dermopterans, more closely
related to primates than to rodents. Last, this same convergent morphol-
ogy was evolved by very distantly related marsupial mammals, and is
found in three different groups of flying possums and marsupial gliders
in Australia and parts of Southeast Asia (table 2.2).
Glider morphologies have repeatedly been evolved in the diapsid
reptiles, but in a form very different from that found in the mammals.
Reptilian gliders have bodies that are dorsoventrally flattened and later-
ally widened. Within their bodies, the ribs are flattened and greatly elon-
gated laterally, away from the anteroposterior axis of the vertebral
column. Species of the modern flying lizard genus Draco have bodies
that resemble a discus in flight—with a pair of forelimbs and a head
attached at one side and a pair of hind limbs and a tail attached to the
opposite side. This identical morphology was evolved more than a quarter
of a billion years ago by ancient diapsid weigeltisaurid reptiles in the
Permian (table 2.2). A very interesting example of parallel evolution of
this morphology occurred in the Triassic, where the lepidosauromorph
Kuehneosaurus latus evolved in Europe and a second group of lepido-
sauromorphs separately evolved the same form in North America, as
seen in Icarosaurus siefkeri (Kuhn-Schnyder and Rieber 1986). Last, a
modification of the same glider morphology has been evolved by the

flying paradise snakes. Rather than flattening and flaring the ribs in the
midsection of the body, the entire snake is flattened like a ribbon, and it
flares its ribs laterally outward as it glides through the air.
Both groups of animals—the mammals and the reptiles—have conver-
gently evolved glider morphologies with expanded surface-area struc-
tures. The construction of those glider structures is very different in the
two groups of animals, however: the mammals repeatedly evolve skin
membranes stretched between their fore and hind limbs, and the reptiles
repeatedly evolve flattened bodies with laterally flared rib cages (table
2.2). Apparently no reptiles ever evolved skin-membrane gliding mor-
phologies, and no mammals ever evolved glider bodies with flattened,
flared rib cages. The two groups of animals appear to be developmentally
constrained to a particular convergent form for each group.
Skin-membrane gliding structures are also found in another group of
animals, but with quite a different construction from those produced by
the mammals. These are the flying tree frogs, which possess skin mem-
branes stretched between the elongated toes of their feet. The flying tree
frogs also present a modern example of parallel evolution. The Old
World frogs, such as Wallace’s flying tree frog, belong to an entirely dif-
ferent family from the New World frogs, such as the Costa Rican flying
tree frog; we can see that the same glider morphology has independently
been evolved by the frogs in two separate regions of the Earth. Lizards
have also convergently evolved webbed feet for gliding, as in Kuhl’s
flying gecko and related species (table 2.2). Interestingly, these geckos
also have added a narrow strip of skin running along the sides of their
head, body, and tail to provide extra gliding surface area, as their webbed
feet are not nearly as large in proportion to their bodies as those found
in the flying tree frogs. If any of the lizards are to converge in the future
on the glider-membrane morphology found in the mammals, these gecko
lizards are perhaps the most likely candidates.
Convergent evolution of gliding morphologies is not confined to land
animals. At least two separate groups of fishes have evolved gliding
morphologies by expanding and elongating their pectoral fins, which are
held horizontally as wing surfaces while the fish leap and then glide
through the air (table 2.2). Again, this convergent morphology appears
to be confined to actinopterygian fishes, but it is interestingly similar to
the modification of forelimbs to wings in the descendants of another
group of fish, the sarcopterygians (these descendants being the ptero-
saurs, birds, and bats; see table 2.2).
22 Chapter 2
Last, convergent gliding morphologies have evolved in at least three

separate groups of ants (table 2.2). Curiously, members of two of these
gliding ant groups (the cephalotini and pseudomyrmecinae) glide back-
ward, abdomen first (Yanoviak et al., 2005). What is perhaps more odd
is the evolution of gliding morphology at all in the ants, given that so
many species of these animals have wings and powered flight.
Most all discussions of convergent evolution begin with the spectacu-
lar examples of convergence of form in fast swimming animals (fusiform
bodies, table 2.1) and flying animals (wings, table 2.2). They often end
there as well. Convergence of form due to locomotory functional con-
straints does not stop with swimming and flying morphologies, however,
so let us consider further convergent consequences of these constraints.
The act of walking in the gravitational field of the Earth requires that an
animal not only resist the pull of gravity in standing up, but also remain
standing and balanced in the progressive falling-forward motion used in
walking locomotion. Walking requires legs, which are lever assemblages
composed of rigid structural elements with connecting tissues and
powered by muscle contractions.
Two major groups of animals have convergently evolved legs: the
arthropods and the tetrapods (table 2.3). As in the convergent evolution
of wings (table 2.2), close examination of the legs of arthropods and
tetrapods reveals interesting differences to solving the same functional
problem, that of walking via lever assemblages. In tetrapod legs, the rigid
structural elements (bones) are located inside the leg, and the connecting
tendons and muscles are located on the outside of the leg (and covered
with a thin layer of skin tissue). Arthropod legs are exactly the reverse:
the rigid structural elements (composed of chitin) are located on the
outside of the leg, and the connecting tendons and muscles are located
inside the leg. A disadvantage of tetrapod endoskeletal legs is that the
soft-tissue muscles and tendons are located on the outside of the leg, and
thus are easily damaged (as we all painfully discover at one time or
another in our lives). On the other hand, the soft-tissue muscles and
tendons of the arthropod exoskeletal leg are protected by the enclosing
chitin shell of the leg. A disadvantage of arthropod legs, however, is
growth of the leg itself. Arthropods must periodically split and shed the
outside rigid covering of the exoskeletal leg (molting), a process not
necessary in growth of the endoskeletal legs of the tetrapod.
Walking locomotion has been hugely successful in the evolution of
animal life on Earth: there are more species of arthropods on Earth than
all other animals species combined, and tetrapod species of vertebrates
Table 2.3
Convergent evolution of animal walking morphologies
1 Convergent structure and function: LEGS (articulated, muscle-controlled lever

assemblages for standing and walking in a gravitational field)
1.1 Giant walkingstick arthropod (Bilateria: Protostomia: Ecdysozoa: Panarthropoda:
Arthropoda: Mandibulata: Hexapoda: Phasmatodea: Phasmidae; Megaphasma
dentricus)
1.2 Giraffe tetrapod (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Tetrapoda: Reptiliomorpha: Synapsida:Therapsida: Mammalia: Eutheria: Laurasiatheria:
Cetartiodactyla: Giraffidae; Giraffa camelopardalis)
2 Convergent structure and function: BIPEDALISM (adaptation for [1] maximum oxygen
uptake during sustained running or hopping, [2] predator detection at a distance, [3] freeing
the forelimbs for functions other than locomotion)
2.1 Bolosaurs (Amniota: Sauropsida: Anapsida: Bolosauridae: Eudibamus cursoris
†Permian)
2.2 Ornithosuchids (Diapsida: Archosauromorpha: Crurotarsi: Ornithosuchidae;
Ornithosuchus longidens †Triassic)
2.3 Dinosauromorphs (Archosauromorpha: Archosauria: Ornithodira: Dinosauro-
morpha: Lagosuchidae; Marasuchus illoensis †Triassic)
2.4 Dinosaurs (Archosauria: Ornithodira: Dinosauria: Herrerasauridae; Herrerasaurus
ischigualastensis †Triassic)
2.5 Red kangaroo (Amniota: Synapsida: Therapsida: Mammalia: Marsupialia:
Diprotodontia: Macropodidae; Macropus rufus)
2.6 Fawn hopping mouse (Mammalia: Eutheria: Euarchontoglires: Rodentia: Muridae;
Notomys cervinus)
2.7 California kangaroo rat (Eutheria: Euarchontoglires: Rodentia: Heteromyidae;
Dipodomys californicus)
2.8 Springhaas (Eutheria: Euarchontoglires: Rodentia: Pedetidae; Pedetes capensis)
2.9 Humans (Eutheria: Euarchontoglires: Primates: Catarrhini: Hominidae: Hominini;
Ardipithecus ramidus †Pliocene)
are much more diverse than nontetrapod. However, one particular form
of tetrapod walking locomotion is curiously quite restricted: that of
bipedalism. The first tetrapods evolved in the Late Devonian, and were
quadrupedal amphibians (nonamniote tetrapods). The plesiomorphic
quadrupedal condition is the norm for almost all living tetrapods; very
few have made the transition to standing up and walking solely on their
hind limbs.
The adaptive significance of bipedalism is still open to debate. Stand-
ing erect on the hind limbs clearly gives an animal the ability to see much
farther in the distance, thus giving it an early-warning capability in preda-
tor detection. Walking only on the hind limbs also gives an animal the
capability to use the forelimbs for functions other than locomotion: both
theropod maniraptoran dinosaurs and primate mammals convergently
24 Chapter 2
evolved prehensile hands in their forelimbs. Another possible bipedal

adaptation is allowing greater expansion of the lungs and rib cage during
fast running or hopping, for greater oxygen uptake. Many species of
quadrupedal lizard will stand up on their hind limbs in order to rapidly
run away from danger. This behavior is most notorious in the brown
basilisk lizard, Basiliscus basiliscus, of Central and South America, where
it is known as the Lagartija Jesucristo (Jesus Christ lizard) because of its
ability to run so fast on its hind limbs that it can actually run across the
surface of water without sinking.
Bipedal locomotion has convergently arisen multiple times in both
reptiles and mammals (table 2.3). The oldest currently known bipedal
reptile is the Late Paleozoic anapsid species Eudibamus cursoris (Berman
et al. 2000). Although it was a very ancient type of reptile, it was not only
bipedal but digitigrade, standing on its toes. Following the end-Permian
mass extinction, bipedal locomotion reevolved independently in two
major groups of the archosaurs, the advanced reptiles, in the Triassic. The
ornithosuchids, or “birdlike crocodiles,” were bipedal predators much
more closely related to living crocodiles than other archosaurians
(Benton 2005). Much more significantly, in terms of subsequent evolu-
tion, is the convergent appearance of bipedalism in the advanced orni-
thodires and the earliest dinosaurs in the Triassic. Marasuchus was a
bipedal dinosauromorph predator—not quite a dinosaur—and the earli-
est dinosaurs Herrerasaurus and Eoraptor were bipedal predators
(Benton 2005). In further dinosaurian evolution, the earliest of the other
major branch of the dinosaurs, the ornithischians, were bipedal fabrosau-
rids, such as Lesothosaurus (Benton 2005). Thus, bipedalism may well be
a synapomorphy for the entire Dinosauria, with subsequent quadrupe-
dalism in dinosaurs evolving as a secondary trait. All of the living dino-
saurs, the birds, are bipedal.
Bipedalism has convergently arisen in two major groups of mammals:
the marsupials and the placentals (table 2.3). The fascinating bipedal
marsupials of Australia are well known, particularly the largest, the red
kangaroo. Within the placentals, bipedalism has evolved independently
several times in the rodents, and notably in the primates. The fawn
hopping mouse is typical of Australian hopping-mice placental species
that coexist with marsupial bipedal species. Independently, in North
America, the Californian kangaroo rat (which is not a rat at all) and
related species have evolved bipedal locomotion. And in South Africa,
yet another independent group of rodents has evolved bipedalism, the
Springhaas Pedetes capensis, which has been described as the “kangaroo
rabbit” although it is neither a rabbit nor a kangaroo. Last, bipedalism

arose independently in a group of primates in Africa, the Hominini, and
is a synapomorphy for all of the various human species.
Why is bipedalism so rare in the mammals in contrast to the dino-
saurs? The dinosaurian ecosystem was enormously successful; it per-
sisted for some 150 million years, and it contained numerous bipedal
animal groups. All of the theropods, the predators of the dinosaurian
ecosystem, were bipedal. Numerous herbivore groups, both saurischian
prosauropods and ornithischian ornithopods, were bipedal, and many of
these were quite large animals, such as the hadrosaurs. Why are preda-
tory bipedal lions or tigers nonexistent in mammalian ecosystems? Why
are there no large, herbivorous bipedal horses or buffalos?
The rarity of bipedal mammalian animals is probably a function of
developmental constraint, as opposed to functional constraint, for the
earliest mammals and their ancestors were quadrupedal. The persistence
of this symplesiomorphic trait in mammalian evolution may be due more
to “developmental inertia,” in that the quadrupedal condition is the
inherited norm, than to any possible nonfunctionality of bipedal mam-
malian forms. In contrast, bipedalism is a synapomorphy for the Dino-
sauria—their earliest forms, both carnivores and herbivores, were bipedal.
Only when they evolved such large and heavy animals as the gigantic
sauropods, the massively armored ankylosaurs, and ceratopsids with
enormously large skulls did dinosaurs depart from this developmental
norm because bipedalism was no longer functional; at that point, they
secondarily reevolved quadrupedal forms.
To conclude this section on walking-locomotion morphologies, I would
like to take another thought-experiment excursion into theoretical mor-
phology. The mythical centaur is described as a vertebrate animal with
four walking legs present on a horselike posterior, and two arms present
on a humanlike anterior. In considering this nonexistent form, Maclaurin
and Sterelny (2008, 104) ask: “No six-limbed vertebrates have ever
evolved from four-limbed ancestors. Is this evidence of the developmen-
tal impossibility of centaurs? How can we tell from the fact that the
elements in a trait cluster did not diverge independently of one another,
that they could not evolve independently of one another?” This is an
interesting question, and it illustrates the difficulty of proving develop-
mental constraint. But, as in our previous thought experiment on the
nonexistence of dragons, we can argue for developmental constraint on
the grounds of improbability. In order to evolve a centaur-form, a group
of tetrapods would have to first develop an additional pair of legs (not
26 Chapter 2
impossible, but highly improbable for vertebrates) and, second, develop

arms from the new pair of legs. It is interesting to note here that the
insects have in fact developed a centaur-form: the praying mantis, Mantis
religiosa. Evolved from ancestors for whom the six-legged condition is
the developmental norm, the mantid has four walking appendages on its
posterior and two arms on the upright thorax below the head, like a
centaur. And not only have the insects evolved centaurs, but they did it
twice! Centaur morphologies very similar to true mantids (Hexapoda:
Mantodea: Mantidae) have been independently evolved by the “false
mantid” mantidflies (Hexapoda: Neuroptera: Mantispidae), a group of
insects more related to ant lions than mantids. The green mantidfly of
southeastern North American, Zeugomantispa minuta, looks virtually
identical in form to Mantis religiosa but is smaller, being only about one-
fifth the length of the mantid. Thus, even centaurs have been conver-
gently evolved, but only by insects.
So far we have considered animal locomotion under water (swim-
ming), aloft in the air (flying), and on the surface of the Earth (walking).
Now let us consider animal locomotion under the surface of the Earth:
burrowing. The number of independent convergences on burrowing
morphologies is amazing (table 2.4), particularly when it would seem that
these morphologies are related to a rather restrictive and highly special-
ized way of life.
The evolution of the four walking limbs of the earliest amphibians
from their fish ancestors’ lobed fins is a complicated process spanning
Table 2.4
Convergent evolution of animal burrowing morphologies
1 Convergent structure and function: SNAKE-SHAPED BODIES (penetrator adaptations:

streamlined, pointed heads for soil penetration; elongated, thin cylindrical bodies to
minimize drag resistance in burrows; loss of fore and hind limbs no longer used in
locomotion)
1.1 Aistopods (Tetrapoda: Lepospondyli: Aistopoda: Phlegethontiidae; Phlegethonia
linearis †Carboniferous)
1.2 Greater yellow-banded caecilian (Tetrapoda: Batrachomorpha: Lissamphibia:
Gymnophiona: Cecelidae; Ichthyophis glutinosus)
1.3 Florida worm lizard (Tetrapoda: Reptiliomorpha: Amniota: Sauropsida: Diapsida:
Lepidosauromorpha: Lepidosauria: Squamata: Scleroglossa: Amphisbaenea:
Amphisbaenidae; Rhineura floridana)
1.4 Burton’s legless lizard (Lepidosauria: Squamata: Scleroglossa: Gekkonta:
Pygopodidiae; Lialis burtonis)
1.5 Brazilian bachian lizard (Squamata: Scleroglossa: Autarchoglossa: Scincomorpha:
Gymnophthalmidae; Bachia oxyrhinus)
Table 2.4
(continued)
1.6 California legless lizard (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:

Anniellidae; Anniella pulchra)
1.7 Eastern glass lizard (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Anguidae; Ophisaurus ventralis)
1.8 Scarlet snake (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha: Serpentes:
Colubridae; Cemophora coccinea)
2 Convergent structure and function: MOLE-SHAPED BODIES (excavator adaptations:
compact cylindrical bodies to fit within burrows, powerful forelimbs or hind limbs with
enlarged claws or shovel structures for digging, rudimentary eyes)
2.1 Mole cricket (Bilateria: Protostomia: Ecdysozoa: Arthropoda: Mandibulata:
Hexapoda: Orthoptera: Gryllotalpidae; Gryllotalpa gryllotalpa)
2.2 Mexican burrowing toad (Bilateria: Deuterostomia: Chordata: Osteichthyes:
Sarcopterygii: Tetrapoda: Batrachomorpha: Lissamphibia: Batrachia: Anura:
Rhinophrynidae; Rhinophrynus dorsalis)
2.3 Couch’s spadefoot toad (Batrachomorpha: Lissamphibia: Batrachia: Anura:
Pelobatidae; Scaphiopus couchii)
2.4 Guinea shovel-snout frog (Lissamphibia: Batrachia: Anura: Hemisotidae; Hemisus
guineensis)
2.5 Mole salamander (Lissamphibia: Batrachia: Urodela: Ambystomidae; Ambystoma
talpoideum)
2.6 Southern marsupial mole (Tetrapoda: Reptiliomorpha: Amniota: Synapsida:
Therapsida: Mammalia: Marsupialia: Notoryctemorpha: Notoryctidae; Notoryctes
typhlops)
2.7 Pink fairy armadillo (Mammalia: Eutheria: Xenarthra: Dasypodidae;
Chlamydophorus truncatus)
2.8 Cape golden mole (Mammalia: Eutheria: Afrotheria: Afrosoricida: Chrysochloridae;
Chrysochloris asiatica)
2.9 European mole (Mammalia: Eutheria: Laurasiatheria: Eulipotyphles: Talpidae;
Talpa europaea)
2.10 Ancient palaeanodonts (Eutheria: Laurasiatheria: Pholidota: Palaeanodonta:
Epoicotheridae; Epoicotherium unicum, Xenocranium pileorivale †Oligocene)
2.11 Common vole (Eutheria: Euarchontoglires: Rodentia: Myomorpha: Arvicolidae;
Microtus arvalis)
2.12 Lesser mole rat (Eutheria: Euarchontoglires: Rodentia: Myomorpha: Spalacidae;
Spalax leucodon)
2.13 Siberian zokor (Eutheria: Euarchontoglires: Rodentia: Myomorpha: Cricertidae:
Mylospalacinae; Myospalax myospalax)
2.14 Asian bamboo rat (Eutheria: Euarchontoglires: Rodentia: Myomorpha:
Rhizomyidae; Rhizomys sumatraensis)
2.15 Chilean coruro (Eutheria: Euarchontoglires: Rodentia: Hystricomorpha:
Octodontidae; Spalacopus cyanus)
2.16 Rio Negro tuco-tuco (Eutheria: Euarchontoglires: Rodentia: Hystricomorpha:
Ctenomyidae; Ctenomys rionegrensis)
2.17 African mole rat (Eutheria: Euarchontoglires: Rodentia: Hystricomorpha:
Bathyergidae; Cryptomys hottentotus)
2.18 Plains pocket gopher (Eutheria: Euarchontoglires: Rodentia: Sciuromorpha:
Geomyidae; Geomys bursarius)
28 Chapter 2
millions of years in the Late Devonian and Early Carboniferous. So it is

astonishing to note that some ancient amphibians promptly devolved
these same hard-earned appendages and became limbless in the later
Carboniferous. These are the aistopods, like Phlegethonia linearis, which
has a morphology virtually identical to a modern snake (Benton 2005)
but is not a snake. Amphibian convergence on penetrator-form burrow-
ing morphologies did not end in the Carboniferous, for an entire group
of legless amphibians exist today, the Gymnophiona, whose caecilian
members strongly resemble earthworms. In addition, they have second-
arily evolved a hydrostatic system for burrowing and locomotion that is
convergent on the ancient plesiomorphic hydrostatic system seen in true
earthworms (O’Reilly et al. 1997).
These amphibians are not alone, as the amphisbaenid Florida worm
lizard is also identical in overall appearance to an earthworm, even to
its pinkish-purple color, yet it is a lepidosaurian reptile. Lee and Shine
(1998) have demonstrated that the amphisbaenid worm lizards are more
closely related to gekko lizards than to the snakes. But other groups of
lizards have evolved snakelike penetrator morphologies, like the ancient
amphibian aistopods. These modern reptiles include the California
legless lizard of western North America, the legless glass lizards of
eastern North America, the pygopodid legless lizards of Australia, and
the recently discovered gymnophthalmid legless lizard Bachia oxyrhinus
of Brazil, South America. The true penetrator-form burrowing snakes,
like the very pretty scarlet snake of southeastern North America, are
more closely related to modern varanid lizards, such as the very impres-
sive predator Varanus komodensis, the Komodo dragon (Lee and Shine
1998), and more distantly related to the ancient marine snakelike mosa-
saurs (see eel-form swimmers in table 2.1).
When burrowing animals are mentioned, most people think of the
mole, with its short, compacted, cylindrical body and powerful digging
forelimbs—typical excavator-form burrowers. However, this same form
can be found in the insects, as seen in the mole crickets. And, just as the
mammalian mole is considered by many to be a pest, the mole cricket is
injurious to garden plants and crops. Excavator-form burrowers are also
found in the amphibians. For example, ambystomid mole salamanders
have short, compacted, cylindrical bodies and blunt heads for burrowing.
The amphibian toads have convergently evolved excavator-forms twice,
but in a backward fashion: both the Mexican burrowing toads and the
spadefoot toads burrow with their powerful hind limbs, not their fore-
limbs, and in both groups the hind feet of the animals have been enlarged
and stiffened into spadelike digging structures. Last, excavator forms are
also found in the frogs, as in the Guinea shovel-snout frog.
In modern placental mammals, the mole-form has independently
evolved in three very distantly related groups: the laurasiatherian true
talpid moles, so common in Europe and North America; the afrotherian
golden chrysochlorid moles; and the xenarthran armadillos (table 2.4).
In the Oligocene, some 30 million years ago, another group of placental
mammals independently evolved the mole-form, the ancient epoicother-
ids (Rose and Emry 2005). An even more distantly related mole-form
has been convergently evolved by the marsupial mammals, as is seen in
the modern southern marsupial mole.
Last, multiple convergent evolutions of excavator-form morphologies
have taken place in another group of placental mammals, the rodents, as
documented by Nevo (1999). Mole-forms have been convergently
evolved in four groups of myomorph rodents around the world: the voles,
the lesser mole rats, the Siberian zokors, and the Asian bamboo rats. In
South America, two groups of hystricomorph rodents have evolved
mole-forms: the coruros, endemic to Chile, and the tuco-tucos, found far
to the southeast in Uruguay. In Africa, another group of hystricomorph
rodents have evolved mole-forms: the African mole rats. And last, in
North America, the sciuromorph rodents have produced the pocket
gophers.
From the perspective of theoretical morphology, it is interesting to
note that burrowing mammals have never produced penetrator-forms—
there exist no mammalian snake-forms. Yet it is easy to envision
an elongated, furry, snakelike mammal, one that has secondarily lost its
legs. Of all the mammals, the weasels and ferrets—with their elongated
bodies and small legs—are perhaps the best candidates for producing
a hypothetical snake-form mammal. In contrast, the reptiles have never
produced excavator mole-forms; many reptiles excavate (turtles come
to mind) but usually only sporadically and shallowly, as when they
are preparing to hibernate. The burrowing reptiles’ preference for
penetrator-forms and the burrowing mammals’ preference for excavator-
forms again raise the question of the role of functional versus
developmental constraint in explaining the absence of theoretically
possible morphologies in evolution.
In the animals, mode of locomotion is usually independent of how the
animal feeds; that is, whether it is a carnivore or an herbivore. Some
burrowing mammals are carnivores, like talpid moles, while others are
herbivores, like tuco-tucos (Nevo 1999). Some flying birds are carnivores,
30 Chapter 2
like hawks; others are herbivores, like parrots. Similary, some flying bats
are carnivores, like the insectivorous microbats, while others are herbi-
vores, like the frugivorous megabats. In addition to locomotion, the
particular mode of feeding of an animal imposes additional functional
constraints, which again are reflected in subsequent convergent evolu-
tion, as we shall see in the next section of the chapter.
Carnivores: Prey Detection
As heterotrophic organisms, animals are incapable of synthesizing their

own food, unlike plants, which we shall consider in the next chapter.
Because all animals need food, it is no surprise that many different
animal groups have repeatedly, convergently, evolved the same success-
ful forms and structures used in food acquisition.
In order to survive, carnivores must be able to detect, capture, and kill
prey animals. One very obvious way to detect prey is to be able to see
the prey animal, that is, to have eyes. The convergent evolution of eyes
themselves will be considered later in this chapter; of interest here are
the modifications of eyes found in the carnivores. An oft-repeated rule
for animals is that carnivorous animals have binocular vision—their eyes
are located forward enough on their skulls to give the animal overlap-
ping fields of vision, which gives the animal three-dimensional depth
perception and thus enables it to precisely locate a prey animal in space.
In contrast, herbivorous animals usually have eyes located on either side
of their heads, giving them an almost 360-degree field of vision for the
detection of a predator in front of them, to either side, or even behind
them. The disadvantage of herbivorous-animal vision is that the animal
has either no binocular vision at all or only very restricted binocular
vision directly in front of it. In some carnivores, highly developed binocu-
lar vision has the disadvantage that their field of vision is restricted to
the front of the animal, so that they must turn their heads in order to see
to their sides or to look behind themselves.
Exceptions to this rule are numerous. For example, the primates have
highly developed binocular vision that is, in general, unrelated to a car-
nivorous mode of life. Instead, many primates need binocular vision
because they are arboreal—they need precise depth perception in order
to jump from tree branch to tree branch without missing the branch, and
perhaps falling to their deaths to the forest floor below.
Binocular vision is a very ancient condition for chordate animals. Even
though their eyes are placed on either side of their laterally flattened
bodies, many fish still possess a limited degree of binocular vision

directly in front of their bodies, where their visual fields briefly overlap.
For this reason, the rainbow trout, Oncorhynchus mykiss, can accurately
stalk and precisely bite a floating insect resting on the water surface
above it.
Thus, in many cases the possession of binocular vision in a carnivore
is simply a plesiomorphic trait that the animal inherited from a distant
ancestor, and not a newly derived or convergent trait. What is often
convergent, however, is the degree of binocular vision—the degree to
which the eyes have been moved forward on the skull. Consider the owl
and the cat. These are very different animals—one is an avian dinosaur
and the other is a placental mammal, and they represent two branches
of the evolutionary tree that diverged over 340 million years ago (the
sauropsids and the synapsids). Yet both animals have eyes rotated so far
anterior on their skulls that they face directly forward, a convergent
condition also found in humans. These animals have very highly devel-
oped three-dimensional depth perception, as we do.
However, if we closely examine the eyes of the owl, cat, and human,
we immediately spot a major difference—the cat has eyes with vertical-
slit pupils. The vertical-slit pupil has been repeatedly and independently
evolved many times throughout the evolutionary history of animals
(table 2.5). In the seas, three separate groups of modern cartilaginous
fishes have convergently evolved eyes with vertical-slit pupils. The angel
sharks (species of Squatina) are bottom dwellers with flattened bodies,
and lie in wait for passing prey that they then ambush. They often
operate in dim light or at night, where it is difficult for prey animals to
detect them lying on the sea bottom. The carcharhinid sharks, such as
the whitetip shark, have eyes with vertical-slit pupils and gold irises that
are astonishingly similar to those often found in black cats. Last, the
myliobatid rays, such as the beautiful spotted eagle ray, have indepen-
dently evolved the same type of pupil.
On land, the vertical-slit pupil has repeatedly evolved independently
in the eyes of amphibians, reptiles, birds, and mammals (table 2.5). In
the amphibians, the great majority of frogs and toads have eyes with
horizontal-slit pupils (a trait we shall consider in more detail later).
However, some tree frogs have evolved eyes with vertical-slit pupils, a
pupil type that has also been convergently evolved in the peculiar tailed
frogs, which are aquatic carnivores that inhabit streams and creeks. Ver-
tical-slit pupils have evolved independently in two families of fossorial
toads: the rhinophrynid burrowing toads and the pelobatid spadefoot
32 Chapter 2
Table 2.5
Convergent evolution of predator eye structures and vision systems
1 Convergent structure and function: VERTICAL-SLIT PUPILS IN EYES (allows [1] full
usage of the diameter of the lens of the eye in bright light as well as in very low light
intensities, with well-focused images in all light intensities; [2] particular detection of
motion in the horizontal plane)
1.1 Pacific angel shark (Vertebrata: Chondrichthyes: Elasmobranchii: Squatinidae;
Squatina californica)
1.2 Whitetip shark (Chondrichthyes: Elasmobranchii: Carcharhinidae; Charcharhinus
longimanus)
1.3 Spotted eagle ray (Chondrichthyes: Elasmobranchii: Myliobatidae; Aetobatus
narinari)
1.4 Brownbelly leaf frog (Vertebrata: Osteichthyes: Sarcopterygii: Tetrapoda:
Batrachomorpha: Lissamphibia: Batrachia: Anura: Hylidae; Phyllomedusa tarsius)
1.5 Tailed frog (Batrachomorpha: Lissamphibia: Batrachia: Anura: Ascaphidae;
Ascaphus truei)
1.6 Mexican burrowing toad (Lissamphibia: Batrachia: Anura: Rhinophrynidae;
Rhinophrynus dorsalis)
1.7 Couch’s spadefoot toad (Lissamphibia: Batrachia: Anura: Pelobatidae; Scaphiopus
couchii)
1.8 Helmeted gecko (Tetrapoda: Reptiliomorpha: Amniota: Sauropsida: Diapsida:
Lepidosauromorpha: Lepidosauria: Squamata: Scleroglossa: Gekkonta: Gekkonidae;
Tarentola chazaliae)
1.9 Burton’s legless lizard (Lepidosauria: Squamata: Gekkonta: Pygopodidae; Lialis
burtonis)
1.10 Granite night lizard (Lepidosauria: Squamata: Scleroglossa: Autarchoglossa:
Scincomorpha: Xantusidae; Xantusia henshawi)
1.11 Indian python (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha: Serpentes:
Boidae; Python molurus)
1.12 Timber rattlesnake (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Serpentes: Viperidae: Crotalinae; Crotalus horridus)
1.13 Nile crocodile (Diapsida: Archosauromorpha: Archosauria: Crurotarsi: Crocodylia:
Crocodylidae; Crocodylus niloticus)
1.14 Black skimmer (Archosauria: Ornithodira: Dinosauria: Saurischia: Theropoda:
Maniraptora: Aves: Charadriiformes: Rhynchopidae; Rhynchops niger)
1.15 Small cats (Amniota: Synapsida: Therapsida: Mammalia: Eutheria: Laurasiatheria:
Carnivora: Feliformia: Felidae; Felis sylvestris catus)
1.16 Red fox (Mammalia: Eutheria: Laurasiatheria: Carnivora: Caniformia: Canidae;
Vulpes vulpes)
1.17 Harp seal (Eutheria: Laurasiatheria: Carnivora: Caniformia: Phocidae; Pagophilus
groenlandicus)
1.18 Slow loris (Eutheria: Euarchontoglires: Primates: Lorisiformes; Nycticebus
coucang)
2 Convergent structure and function: HORIZONTAL-SLIT PUPILS IN EYES (allows
[1] full usage of the diameter of the lens of the eye in bright light as well as in very low
light intensities, with well-focused images in all light intensities; [2] particular detection of
motion in the vertical plane)
2.1 Common octopus (Bilateria: Protostomia: Lophotrochozoa: Mollusca: Cephalopoda:
Coleoidea: Octopodidae; Octopus vulgaris)
Table 2.5
(continued)
2.2 Common toad (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:

Tetrapoda: Batrachomorpha: Lissamphibia: Batrachia: Anura: Bufonidae; Bufo bufo)
2.3 Bullfrog (Batrachomorpha: Lissamphibia: Batrachia: Anura: Ranidae; Rana
catesbeiana)
2.4 Green tree frog (Lissamphibia: Batrachia: Anura: Hylidae; Hyla cinerea)
2.5 Asian palm civet (Tetrapoda: Reptiliomorpha: Amniota: Synapsida: Therapsida:
Mammalia: Eutheria: Laurasiatheria: Carnivora:Viveridae; Paradoxurus hermaphroditus)
2.6 Bottlenose dolphin (Mammalia: Eutheria: Laurasiatheria: Cetartiodactyla: Cetacea:
Odontoceti: Delphinidae; Tursiops truncatus)
3 Convergent structure and function: ENLARGED EYES (increases retinal surface area
for low-light-intensity vision in nocturnal predators)
3.2 Squirrelfish (Bilateria: Deuterostomia: Chordata: Osteichthyes: Actinopterygii:
Beryciformes: Holocentridae; Holocentrus adscensionis)
3.3 Great horned owl (Osteichthyes: Sarcopterygii: Reptiliomorpha: Amniota:
Sauropsida: Diapsida: Archosauromorpha: Ornithodira: Dinosauria: Saurischia:
Theropoda: Maniraptora: Aves: Strigiformes: Strigidae; Bubo virginianus)
3.4 Tasmanian devil (Amniota: Synapsida: Therapsida: Mammalia: Marsupialia:
Dasyuromorpha: Dasyuridae; Sarcophilus harrisii)
3.5 Small cats (Mammalia: Eutheria: Laurasiatheria: Carnivora: Feliformia: Felidae;
Felis sylvestris catus)
3.6 Lesser bushbaby (Mammalia: Eutheria: Euarchontoglires: Primates: Lorisiformes:
Galagidae; Galago senegalensis)
3.7 Philippine tarsier (Eutheria: Euarchontoglires: Primates: Tarsiiformes: Tarsiidae;
Tarsius syrichta)
4 Convergent structure and function: INFRARED “VISION” ([1] pit organs in snakes
that detect infrared light, used for detecting heat from warm-blooded, endothermic prey
animals even in total darkness; [2] infrared receptors in insects, used to detect forest fires)
4.1 Pyrophyllic beetle (Bilateria: Protostomia: Ecdysozoa: Arthropoda: Mandibulata:
Hexapoda: Coleptera: Buprestidae: Buprestinae; Melanophila acuminata)
4.2 Australian fire beetle (Arthropoda: Mandibulata: Hexapoda: Coleptera: Buprestidae:
Chrysobothrinae; Merimna atrata)
4.3 Australian flat bug (Arthropoda: Mandibulata: Hexapoda: Heteroptera: Aradidae;
Aradus albicornis)
4.4 Indian python (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Reptiliomorpha: Amniota: Sauropsida: Diapsida: Lepidosauromorpha: Squamata:
Scleroglossa: Autarchoglossa: Anguimorpha: Serpentes: Boidae; Python molurus)
4.5 Timber rattlesnake (Lepidosauromorpha: Squamata: Scleroglossa: Autarchoglossa:
Anguimorpha: Serpentes: Viperidae: Crotalinae; Crotalus horridus)
5 Convergent structure and function: ELECTRIC-FIELD “VISION” (electroreceptive
organs allowing the detection of moving prey by the electrical-field activity associated with
muscle contractions)
5.1 Cephalaspid osteostracans (Vertebrata: Osteostraci: Cephalaspidae; Hemicyclaspis
murchisoni †Devonian)
5.2 Marbled electric ray (Vertebrata: Chondrichthyes: Elasmobranchii: Batoidea:
Torpediniformes: Torpedinidae; Torpedo marmorata)
34 Chapter 2
Table 2.5
(continued)
5.3 Winter skate (Chondrichthyes: Elasmobranchii: Batoidea: Rajiformes: Rajidae;

Raja ocellata)
5.4 Pacific stargazer (Vertebrata: Osteichthyes: Actinopterygii: Teleostei: Perciformes:
Uranoscopidae; Astroscopus zephyreus)
5.5 Banded knifefish (Actinopterygii: Teleostei: Gymnotiformes: Gymnotidae;
Gymnotus carapo)
5.6 Peters’ elephant-nose fish (Actinopterygii: Teleostei: Osteoglossiformes:
Mormyridae; Gnathonemus petersii)
5.7 African electric catfish (Actinopterygii: Teleostei: Siluriformes: Malapteruridae;
Malapterurus electricus)
5.8 Duckbill platypus (Osteichtyes: Sarcopterygii: Reptiliomorpha: Amniota: Synapsida:
Therapsida: Mammalia: Monotremata: Ornithorhynchidae; Ornithorhynchus anatinus)
5.9 Australian echidna (Mammalia: Monotremata: Tachyglossidae; Tachyglossus
aculeatus)
toads. These toads burrow into the ground during the day, and emerge
to hunt at night.
In the lepidosaurian reptiles, the great majority of both lizards and
snakes have eyes with round pupils. But among the lizards, we find
vertical-slit pupils in the helmeted geckos (whose bizarre eyes are unique
in many ways) and in all the nocturnal species of the gecko family. On
opposite sides of the planet, the xantusid night lizards of North America
and many species of the pygopodid legless lizards of Australia, like Bur-
ton’s legless lizard, have eyes with vertical-slit pupils. Among the snakes,
the two deadliest snake groups have convergently evolved eyes with
vertical-slit pupils: the boid constrictors and the venomous pit vipers.
Both of these predator groups have also evolved infrared-sensitive pit
organs, and are able to hunt warm-blooded animals in total darkness (a
trait that we shall consider in more detail in a moment).
The advanced archosaurian reptiles, such as the crocodiles and
alligators, are aquatic carnivores that have independently evolved eyes
with vertical-slit pupils. Since the great majority of the archosaurian
dinosaurs are extinct, we may never know what type of eyes they pos-
sessed. This fact does not prevent the makers of science fiction movies
from usually portraying extinct theropod predators, such as the veloci-
raptors, with eyes with vertical-slit pupils. In this context, it is interesting
to note that only one species of the modern avian dinosaurs has eyes
with vertical-slit pupils: the black skimmer Rhynchops niger (Zusi and
Bridge 1981). The many species of modern owls, which are avian noctur-
nal predators, all have eyes with round pupils. Zusi and Bridge (1981)
suggest that the wide range of light intensities encountered by the black
skimmer, which seeks prey during both the dark of night and brilliant
daylight, led it to evolve vertical-slit pupils to protect its eyes during
daylight fishing.
In the placental mammals, eyes with vertical-slit pupils are most char-
acteristic of the small cats, such as the European wildcat Felis sylvestris
or the North American bay lynx Lynx rufus. All the large cats that hunt
during daylight, such as the African lion Panthera leo, have eyes with
round pupils. However, eyes with vertical-slit pupils have been conver-
gently evolved within the canid predators as well, even though these
animals are so unlike the cats in many ways. One example is the modern
red fox, Vulpes vulpes. The small foxes are in essence honorary cats,
converging on a felid way of life. Many of the marine carnivores, such as
the harp seal, have also evolved eyes with pupils that are round when
the animal is hunting in the water, but that contract to vertical slits
when in the air. Last, eyes with vertical-slit pupils have independently
evolved within the primates themselves, and are present in the lorisiform
nocturnal carnivore the slow loris, Nycticebus coucang (Malmström and
Kröger 2006).
In summary, the vertical-slit pupil has repeatedly evolved in carnivores
that are nocturnal, hunting at night, or crepuscular, hunting in the dim
light of twilight or just before sunrise, as well as in those that hunt in
environments that normally have low light intensities, as in many aquatic
habitats. Malmström and Kröger (2006) have argued that in animals with
multifocal lenses, the slit-pupil system allows the full usage of the diam-
eter of the lens of the eye in bright light as well as in very low light
intensities, with well-focused images at all light intensities. In contrast,
the round-pupil system shades the peripheral zones of the lens of the
eye as it is constricted, leading to the loss of well-focused images at
wavelengths of light that are normally focused in those regions of the
lens. Zusi and Bridge (1981) report that, in the eyes of small cats, the
slit-pupil system allows a greater reduction in pupil size than is possible
with a round-pupil sphincter system; this is particularly important to
animals that have very large pupils when the pupillary opening is at its
maximum dilation. Thus, for animals with large pupils, the slit-pupil
system offers much better protection to the retina of the eye in bright
light conditions than a round-pupil system.
The catch in the reasoning of these two studies is this: it does not
matter how the slit in the slit-pupil system is oriented in order to obtain
both of these superior-focus and retinal-protection advantages. That is, a
36 Chapter 2
horizontal-slit pupil would work just as well as a vertical-slit pupil. And

indeed we shall see that many animals—both carnivores and herbi-
vores—have convergently evolved eyes with horizontal-slit pupils. Why,
then, do we find the vertical-slit pupil only in carnivore eyes, and never
in herbivores? The answer appears to involve the type of motion detec-
tion that is important to a predator. The vertical-slit-pupil system in the
eye is particularly good at detecting motion in a horizontal plane. And
since most prey animals are moving in a near-horizontal plane, parallel
to the surface of the ground or to the sea floor, predators with vertical-
slit pupils in their eyes have an advantage in spotting prey motion, even
in very low light intensities.
Although the frogs and toads are carnivores, the great majority of frogs
and toads have eyes with horizontal-slit pupils, not vertical. These include
the true toads (bufonids), most of the true frogs (ranids), and most of
the true tree frogs (hylids); see table 2.5. In these animals, the horizontal-
slit-pupil system is particularly good at detecting motion in a vertical
plane. Because many of their prey species are flying insects, which move
in three dimensions off the surface of the Earth, the horizontal-slit
pupils of toads and frogs give these predators an advantage in spotting
flying prey. Many also are diurnal predators, hunting during the daylight
hours, as opposed to nocturnal or crepuscular predators.
Otherwise, the horizontal-slit pupil is rare in predators. The catlike
Asian palm civet has convergently evolved horizontal-slit-pupil eyes, but
its adaptive significance is open to question, as these animals hunt in a
fashion similar to true cats, which have vertical-slit-pupil eyes. Curiously,
eyes with horizontal-slit pupils have also convergently arisen in a few
marine predators like the bottlenose dolphin and the octopus, animals
with radically different evolutionary backgrounds but with similar eyes.
As with the palm civets, it is unclear what selective advantage these eye
systems may possess, if any, in the habitats frequented by the bottlenose
dolphin and the octopus.
In addition to pupil type, the size of the eye itself is subject to conver-
gent evolution depending on the hunting behavior of the animal. Noc-
turnal predators have much larger eyes, relative to the size of their skulls,
than diurnal predators. Within the avian clade, the diurnal-hunting hawk
has small eyes, while the nocturnal-hunting owl has enormous eyes.
Within the felid clade, the diurnal-hunting lion has small eyes, while the
nocturnal-hunting small cats have enormous eyes. I have listed only a
few notable examples of this convergent trait in table 2.5, as it is rife
throughout the animal kingdom. What is not common among animals is
an extreme, but highly advantageous, form of nocturnal “vision”: the

ability to “see” infrared light.
The evolution of infrared vision would obviously give a predator an
extremely deadly advantage over warm-blooded, endothermic prey
animals in that the predator would be able to detect the prey animal
even in total darkness simply by detecting its body heat. Yet no predator
has evolved eyes that are capable of seeing the infrared spectrum of light.
It may well be that there are limits on the maximum wavelength of light
that can be detected by the vertebrate eye (the same appears to be true
for the compound eyes of the insects, as we shall see in a moment). In
the classic work On Growth and Form, Thompson (1942) suggested that
there is a minimum size to the vertebrate eye, below which the eye simply
cannot function given the wavelengths of visible light (around 400 nm
to 700 nm for humans), and Purnell (1995) has marshaled evidence from
divergent vertebrates that indicates this limit is in the range of 1.0 to 1.5
millimeters in diameter. Thus, at the maximum end of the visible light
spectrum, it may well be impossible to evolve an eye that can detect
electromagnetic radiation with wavelengths greater than 800 nanome-
ters, the beginning of the infrared spectrum.
However, two convergent groups of snakes have evolved pit organs
capable of detecting infrared radiation (table 2.5). In the crotaline pit
vipers, a heat-sensitive facial pit is located between the eye and the
nostril on each side of the head. In boid constrictors, heat-sensitive pits
are located in the lip scales. The sensory information from the pit organs
is processed in the same region of the snake’s brain as the sensory infor-
mation from the eyes; thus, this may be the only way in which vertebrates
may “see” infrared light. This type of convergent “vision” may also be
attainable only by ectothermic predators, as an endothemic predator’s
own body heat may overwhelm, and hence render useless, the heat-at-a-
distance detection ability of such an infrared-perception organ. Although
at first glance the convergent evolution of infrared “vision” via heat-
sensitive facial pits in two separate groups of snakes is remarkable, the
selective advantage of being able to detect endothermic prey in total
darkness leads one to wonder why more ectothermic predators have not
convergently evolved this capability.
There exists another group of animals that have evolved the capability
of “seeing” infrared light—the bizarre pyrophilous beetles and flat
bugs—but they also cannot detect infrared light with their compound
eyes. Twice within the buprestid beetle family (Evans 1966; Schmitz
et al. 2008) and once within the aradid flat bug family (Schmitz et al.
38 Chapter 2
2008), three groups of insects have convergently evolved specialized

infrared-receptor organs (table 2.5). These insects actually hunt for forest
fires, seeking out burnt wood in order to feed on fast-growing postfire
fungi. As such, they are not strictly predators because their intended prey
are fungi, not other animals. But, as the fungi are much more closely
related to metazoans than either of these two groups are to the plants
(see appendix of this book), I include them here.
The fact that both the camera-eyed vertebrates and the compound-
eyed arthropods independently evolved the capability of detecting infra-
red radiation—but not with their eyes—argues for functional constraint
rather than developmental constraint in the absence of infrared vision
in animals. Surely, if it were functionally possible, some group of animals
with these very different types of eyes would have evolved a modified
eye structure that could detect electromagnetic radiation with wave-
lengths greater than 800 nm in the past 600 million years of evolution.
But could the absence of infrared vision also be considered a develop-
mental constraint peculiar to Earth-type life? In an interesting case of
deep homology, it is now known that all animal eyes—both camera and
compound—contain highly conserved transcription factor Pax-6 gene
homologs (a fact that we shall consider in more detail later in the
chapter). Might some alien forms of life elsewhere in the universe, life
forms that evolved eyes with a different genetic coding and totally dif-
ferent eye structure for detecting electromagnetic radiation, easily see
long-wavelength light as well as short-wavelength light? We shall return
to this question in chapter 7, when we consider the possibility that there
might exist organic forms that function perfectly well, but that neverthe-
less cannot be developed by Earth life.
An even more exotic prey-detection system is the electric fishes’ ability
to “see” an electric field itself, and not just electromagnetic radiation, in
that the same regions of the brain that process visual information also
process electrosensory information in these fish (Bastian 1982; Moller
1995). Seven different lineages of fish have convergently evolved the
capability of detecting an electric field (table 2.5). These include two
different groups of rays and skates in the cartilaginous fishes, four dif-
ferent groups of modern bony fishes, and one group of ancient agnathan
fishes. This latter group, the cephalaspid osteostracans, is particularly
interesting. These primitive jawless fishes possessed three sensory fields
on their bony head shields, two lateral and one on the top of the head.
Thick bundles of nerves led from these sensory fields back to the brain,
but appear to have led to the auditory region of the brain (Stensiö 1963)
rather than to the visual region. These ancient fishes therefore may have
evolved the capability to “hear” the static of an electric field, rather than
“seeing” the electric field.
Modern electric fishes are predators, and use their electroreceptive
organs to locate moving prey by detecting the electric-field activity asso-
ciated with muscle contractions in the prey organisms. Many are able to
“see” the electric field of their prey even in total darkness, similar to the
ability of the pit viper to “see” the infrared radiation emitted by endo-
thermic prey species in total darkness. For example, the African electric
catfish is a voracious nocturnal predator, but in contrast to most noctur-
nal predators, it has very small eyes (even smaller than those of diurnal
fishes). Instead of using visual detection to hunt its prey, it uses electro-
sensory detection.
Water is an excellent conductor of electricity, and seven groups of
fishes have independently evolved the capability to detect an electric
field in water. But these fishes are not alone: two separate groups of
primitive monotreme mammals—the duckbill platypus and the Austra-
lian echidna—have also evolved the capability of detecting electric fields
(table 2.5). The duckbill platypus is a semiaquatic predator capable of
hunting in total darkness under water (Gregory et al. 1987; Proske and
Gregory 2003); it has thus converged on the same hunting strategy used
by the electric fishes. The Australian echidna is a land-dwelling animal,
however, and electric field intensities in air are very weak compared to
those that can be developed in water. The echidna uses the electrorecep-
tors on its beak to detect an electric field in the moist soil of its rainforest
habitat, produced by moving earthworms (Manger et al. 1997).
Charles Darwin found the convergent evolution of electroreception in
the electric fishes to be so unusual and so improbable that he included
it in On the Origin of Species in his list of difficulties for the theory of
natural selection to explain: “The electric organs offer another and more
serious difficulty; for they occur in only about a dozen fishes, of which
several are widely remote in their affinities. Generally when the same
organ appears in several members of the same class, especially if in
members having very different habits of life, we may attribute its pres-
ence to inheritance from a common ancestor; and its absence in some of
the members to its loss through disuse or natural selection. But if the
electric organs had been inherited from one ancient progenitor thus
provided, we might have expected that all electric fishes would have been
specially related to each other” (Darwin 1859, 193). Today we know that
the convergent distribution of electrosensory organs in fish is much more
40 Chapter 2
widespread than Darwin realized, and that they have convergently

appeared even in the mammals. Rather than posing a difficulty for the
theory of natural selection, the convergent evolution of electrosensory
organs in distantly related animals is seen today as a prime example of
functional constraint and the process of natural selection in action
(Zakon and Unguez 1999, Hopkins 2008).
Many animals on Earth have developed the capability of detecting a
magnetic field as well, but I know of no animal that uses this capability
to locate prey (presumably because prey animals do not produce detect-
able magnetic fields). Rather than hunting, animals use magnetosensory
capabilities to orient themselves relative to the magnetic field of the
Earth, particularly in migratory species.
In addition to electromagnetic waves, many carnivores can detect pres-
sure waves in water or air when locating prey. The possession of the
tympanic-membrane system of hearing, or sound detection, in land
animals is a trait that can be traced back to the early tetrapods in the
Late Devonian and Carboniferous (more on the convergent evolution
of this trait later in the chapter). However, tympanal hearing systems
have been convergently modified in surprising ways by more derived
tetrapod predators, particularly nocturnal predators. For example, as
flying nocturnal predators, owls need to be able to locate prey animals
in all three spatial dimensions; otherwise, they might overshoot or under-
shoot the prey animal when they swoop down out of the sky. Norberg
(1977) has argued that asymmetric ear systems for three-dimensional
stereophonic hearing have been convergently evolved by at least five
separate groups of owls, independently of one another (table 2.6). In the
tytonid owls, asymmetric ears have been independently evolved by
species of the genus Tyto and the genus Pholidus. In the strigid owls,
asymmetric ears of similar structure are found in owl species in the
genera Bubo, Strix (such as the Eurasian tawny owl, S. aluco), and
Ciccaba (such as the mottled owl, C. virgata). Asymmetric ears of differ-
ent structure are found in strigid owl species of the genera Asio, Rhinop-
tynx (such as the striped owl, R. clamator), and Pseudoscops (such as the
Jamaican owl, P. grammicus). Last, yet another group of strigid owls,
consisting of species of the genus Aegolius, have asymmetric ears that
are different from the two other strigid owl groups.
An even more radical alteration of the hearing system occurs in
animals that have essentially evolved an organic form of sonar (table
2.6). These animals can produce very high-pitched ultrasonic sound
waves that are bounced off the surrounding environment—including
Table 2.6
Convergent evolution of predator ear structures and auditory systems
1 Convergent structure and function: ASYMMETRIC EARS (allows three-dimensional

stereophonic hearing—distance perception in both the vertical and horizontal plane—for
precise prey location)
1.1 Barn owl (Archosauromorpha: Archosauria: Ornithodira: Dinosauria: Saurischia:
Theropoda: Maniraptora: Aves: Strigiformes: Tytonidae; Tyto alba)
1.2 Asian bay owl (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves: Strigiformes:
Tytonidae; Phodilus badius)
1.3 Great horned owl (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves:
Strigiformes: Strigidae; Bubo virginianus)
1.4 Long-eared owl (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves:
Strigiformes: Strigidae; Asio otus)
1.5 Boreal owl (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves: Strigiformes:
Strigidae; Aegolius funereus)
2 Convergent structure and function: ULTRASONIC HEARING (“sonar” adaptation:
three-dimensional depth perception by echolocation for precise prey location)
2.1 Edible-nest swiftlet (Amniota: Sauropsida: Diapsida: Archosauromorpha:
Archosauria: Ornithodira: Dinosauria: Saurischia: Theropoda: Maniraptora: Aves:
Apodiformes: Apodidae; Aerodramus fuciphagus)
2.2 Oilbird (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves: Caprimulgiformes:
Steatornithidae; Steatornis caripensis)
2.3 Lesser hedgehog tenrec (Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
Afrotheria: Afrosoricida: Tenrecidae; Echinops telfairi)
2.4 Eurasian shrew (Mammalia: Eutheria: Laurasiatheria: Eulipotyphles: Soricidae;
Sorex araneus)
2.5 Mouse-eared bat (Eutheria: Laurasiatheria: Chiroptera: Microchiroptera:
Vespertilionidae; Myotis myotis)
2.6 Bottlenose dolphin (Eutheria: Laurasiatheria: Cetartiodactyla: Cetacea: Odontoceti:
Delphinidae; Tursiops truncatus)
Note: For data sources see text.
potential prey animals—and have ears that can hear the ultrasonic
echoes, or return waves. Echolocation thus allows these animals not only
to hunt prey animals by ultrasound but to safely locomote without sight.
The most spectacular example of echolocation is that seen in bats flying
after insects in the dim light of an early summer evening (while we can
still see them). Two separate groups of birds have also convergently
evolved ultrasonic hearing and echolocation capabilities: the apodid
swiftlets of the tropical Indo-Pacific and the steatornithid oilbirds of
South America. The edible-nest swiftlets are crepuscular-to-nocturnal
flying insectivores that inhabit dark caves, and thus are ecological equiva-
lents to bats. Yet the swiftlet is an avian dinosaur, whereas the bat is a
placental mammal. The closest common ancestor for the two is found
back in the Carboniferous, over 340 million years ago, yet they have
42 Chapter 2
convergently evolved the ability to hear ultrasound. The South American

oilbirds are included here as an anomaly—they have independently
evolved echolocation, as they are birds that are nocturnal and inhabit
caves, but they are frugivores, not carnivores. Thus, they do not use echo-
location to hunt.
Ground-dwelling organisms have also convergently evolved echoloca-
tion. Many of the fascinating little tenrecs of Madigascar, such as the
lesser hedgehog tenrec, have evolved the capability to locate their insect
prey using echolocation (Gould 1965). They produce very high-pitched
squeaks and then analyze the pattern of the ultrasound echoes, a capabil-
ity that has been convergently evolved by species of two genera of
insectivorous shrews: Sorex and Blarina. Nevertheless, these two groups
of mammals are very distantly related: the shrews are laurasiatherians,
while the tenrecs are afrotherians. Last, ocean-dwelling animals have
convergently evolved echolocation. The predaceous toothed cetaceans,
such as the bottlenose dolphin, produce high-frequency clicks and
can use the subsequent submarine echoes to echolocate as well as to
detect prey.
Carnivores: Prey Capture
The possession of teeth in animals is a symplesiomorphic trait that can

be traced back to the early gnathostomes in the Devonian, and the pos-
session of claws is a symplesiomorphic trait that can be traced back to
the early tetrapods in the Carboniferous. However, both teeth and claws
have been repeatedly, convergently, modified to produce highly efficient
killing structures in animal predators. The most widely cited example of
remarkable convergence in predator dentary evolution is the convergent
modification of normally conical fang teeth into saber teeth in three
separate groups of mammals in the Cenozoic (table 2.7). Elongation of
the fang teeth in creodont oxyaenid carnivores, such as Machaeroides
simpsoni, began during the Paleocene and is seen as a precursor to the
evolution of saber-tooth predators in the placental mammals (Turner
1997). The Creodonta are a sister lineage to the true Carnivora, and by
the late Eocene the first saber-tooth predators evolve in the nimravids,
or false cats, such as Hoplophoneus mentalis. Saber-tooth nimravid pred-
ators persisted throughout the Oligocene, producing the lion-sized saber-
tooth Barbourofelis fricki in the Miocene (Turner 1997). The first true
cats evolve in the Oligocene, and by the late Miocene are convergently
producing saber-tooth species, culminating in the Pleistocene saber-tooth
great cat, Smilodon fatalis. As such, it would appear that the evolution
of saber-tooth predators in the nimravids and felids is an interesting, but
hardly remarkable, case of parallel evolution in the Carnivora. Not so.
During the same interval of time in which felid saber-tooth cats roamed
North America, Europe, and Asia, the marsupial mammals of South
America evolved a saber-tooth predator, Thylacosmilus atrox, that is
almost identical in morphology to Smilodon fatalis (Turner 1997, 136).
Although we have to go back in time 100 million years or so, to the mid-
Cretaceous, to find a common ancestor between the placental and mar-
supial mammals, the two groups converged on the same predator
morphology in the Pliocene and Pleistocene.
Elongated, saber-shaped teeth are an adaptation for producing shear-
ing bites, bites that can tear off whole chunks of prey flesh and that
produce gaping wounds, leading to massive blood loss and shock in the
prey animal. Shearing bites to the neck can rip out the entire neck of the
prey animal below the cervical vertebrae, or a saber-tooth bite can crush
and collapse the windpipe of large prey animals, quickly causing uncon-
sciousness due to oxygen deprivation (Turner 1997, 125). The extinction
of the saber-tooth predators in the Pleistocene appears to be ecologically
linked to the extinction of the large-prey species that formed their chief
source of food.
The modern cats are well known for another highly derived trait:
retractable claws. Most predators, like bears and wolves, have claws that
can cause serious damage to a prey animal, but that are very dull com-
pared to the razor-sharp claws of the cat. The cat’s claws are usually
retracted, and thus do not come into contact with the ground when they
are walking; hence they are not dulled by constant wear and abrasion,
like the claws of the dog. When the cat deploys its claws, its entire paw
expands as the toe digits are extended, and the sickle-shaped claws may
be used for either slashing the prey animal or clinging to it while the cat’s
fang teeth are in use.
Interestingly, retractable claws were first evolved by the dinosaurs, not
by the cats (table 2.7). Both the dromaeosaurids and the troodontids
evolved a retractable claw on the second digits of their hind feet. Like
the cat’s claws, these claws were sickle-shaped and were retracted
when not in use. The dromaeosaurids and troodontids were related mani-
raptoran theropods, but their retractable claws were independently
evolved in parallel and differ in several aspects of their anatomy (Var-
ricchio 1997). A third example of convergent evolution of this trait is
seen in the very plesiomorphic Late Cretaceous bird Rahonavis ostromi,
44 Chapter 2
Table 2.7
Convergent evolution of predator killing structures
1 Convergent structure and function: SABER TEETH (fang teeth elongated and laterally
flattened for shearing bites to produce gaping wounds)
1.1 Marsupial saber-tooth “cat” (Mammalia: Marsupialia: Sparassodonta:
Thylacosmilidae; Thylacosmilus atrox †Pliocene)
1.2 False saber-tooth “cat” (Mammalia: Eutheria: Laurasiatheria: Carnivora: Feliformia:
Nimravidae; Barbourofelis fricki †Miocene)
1.3 True saber-tooth cat (Eutheria: Laurasiatheria: Carnivora: Feliformia: Felidae;
Smilodon fatalis †Pleistocene)
2 Convergent structure and function: RETRACTABLE CLAWS (protection of claws
when not in use, enhances sharpness of sickle claws for ripping prey)
2.1 Dromaeosaurs (Amniota: Sauropsida: Diapsida: Archosauromorpha: Archosauria:
Ornithodira: Dinosauria: Saurischia: Theropoda: Maniraptora: Dromaeosauridae;
Velociraptor mongoliensis †Cretaceous)
2.2 Troodontids (Dinosauria: Saurischia: Theropoda: Maniraptora: Troodontidae;
Saurornithoides mongoliensis †Cretaceous)
2.3 Rahonaves (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves: incertae sedis;
Rahonavis ostromi †Cretaceous)
2.4 Red-legged seriema (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves:
Gruiformes: Cariamae: Cariamidae; Cariama cristata)
2.5 Nimravid “cats” (Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
Laurasiatheria: Carnivora: Feliformia: Nimravidae; Hoplophoneus mentalis †Eocene)
2.6 True cats (Mammalia: Eutheria: Laurasiatheria: Carnivora: Feliformia: Felidae;
Proailurus lemanensis †Oligocene)
2.7 Banded palm civet (Eutheria: Laurasiatheria: Carnivora: Feliformia: Viverridae;
Hemigalus derbyanus)
3 Convergent structure and function: RAPTORIAL BEAK (laterally compressed, hooked
beak for piercing and tearing flesh)
3.2 Snapping turtle (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Reptiliomorpha: Amniota: Sauropsida: Anapsida: Testudines: Chelydridae; Macroclemys
temmincki)
3.3 Peregrine falcon (Sauropsida: Diapsida: Archosauromorpha: Archosauria:
Ornithodira: Dinosauria: Saurischia: Theropoda: Maniraptora: Aves: Falconiformes:
Falconidae; Falco peregrinus)
3.4 Red-tailed hawk (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves:
Accipitriformes: Accipitridae; Buteo jamaicensis)
3.5 Secretary bird (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves:
Accipitriformes: Sagittaridae; Sagittarius serpentarius)
3.6 Osprey (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves: Accipitriformes:
Pandionidae; Pandion haliaetus)
3.7 Turkey vulture (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves:
Ciconiiformes: Cathartidae; Cathartes aura)
3.8 Great horned owl (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves:
Strigiformes: Strigidae; Bubo virginianus)
3.9 Red-legged seriema (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves:
Ralliformes: Cariamae: Cariamidae; Cariama cristata)
Table 2.7
(continued)
4 Convergent structure and function: POISON-INJECTING FANGS (hollow fangs

connected with poison glands for injection of poison into prey)
4.1 Black widow spider (Bilateria: Protostomia: Ecdysozoa:Arthropoda: Cheliceriformes:
Arachnida: Araneae: Theridiidae; Latrodectus mactans)
4.2 Amazonian giant centipede (Arthropoda: Mandibulata: Myriapoda: Chilopoda:
Scolopendridae; Scolopendra gigantea)
4.3 European ant lion larvae (Arthropoda: Mandibulata: Hexapoda: Neuroptera:
Myrmeleontidae; Euroleon nostras)
4.4 Four-spotted owlfly larvae (Arthropoda: Mandibulata: Hexapoda: Neuroptera:
Ascalaphidae; Ululodes quadripunctatus)
4.5 Timber rattlesnake (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Reptiliomorpha: Amniota: Sauropsida: Diapsida: Lepidosauromorpha: Squamata:
Scleroglossa: Autarchoglossa: Anguimorpha: Serpentes: Viperidae: Crotalinae; Crotalus
horridus)
5 Convergent structure and function: POISONOUS STINGERS (stingers coated with
poison or connected with poison glands for injection of poison into prey)
5.1 Portuguese man-of-war (Metazoa: Cnidaria: Siphonophora: Physaliidae; Physalia
physalis)
5.2 Striated cone-shell snails (Metazoa: Bilateria: Protostomia: Lophotrochozoa:
Mollusca: Gastropoda: Prosobranchiata: Neogastropoda: Conidae; Conus striatus)
5.3 Babylonian auger-shell snails (Mollusca: Gastropoda: Prosobranchiata:
Neogastropoda: Terebridae; Terebra babylonia)
5.4 Giant desert hairy scorpion (Protostomia: Ecdysozoa: Arthropoda: Cheliceriformes:
Arachnida: Scorpiones: Iuridae; Hadrurus arizonensis)
5.5 Giant cicada-killer wasp (Arthropoda: Mandibulata: Hexapoda: Hymenoptera:
Sphecidae; Sphecius speciosus)
5.6 Giant short-tailed stingray (Bilateria: Deuterostomia: Chordata: Chondrichthyes:
Elasmobranchii: Batoidea: Dasyatidae; Dasyatis brevicaudata)
6 Convergent structure and function: ELECTRIC-FIELD GENERATION (generation
of an electric field to stun or kill prey)
6.1 Marbled electric ray (Vertebrata: Chondrichthyes: Elasmobranchii: Batoidea:
Torpediniformes: Torpedinidae; Torpedo marmorata)
6.2 Pacific stargazer (Vertebrata: Osteichthyes: Actinopterygii: Teleostei: Perciformes:
Uranoscopidae; Astroscopus zephyreus)
6.3 South American electric eel (Actinopterygii: Teleostei: Gymnotiformes:
Electrophoridae; Electrophorus electricus)
6.4 African electric catfish (Actinopterygii: Teleostei: Siluriformes: Malapteruridae;
Malapterurus electricus)
46 Chapter 2
which looked very similar to the Late Jurassic bird Archaeopteryx

lithographica (it still possessed three fingers with claws on its wing,
and had caudal vertebrae along with tail feathers), but differed
markedly in possessing retracted, enlarged sickle-shaped claws on the
second digits of its feet. This last convergence has been questioned by
some, who argue that Rahonavis was not a bird but a dromaeosaurid
with wings, and thus conclude that its sickle claw was not a convergent
character. That proposal, however, would require the convergent evolu-
tion of the wing within the maniraptors. And indeed, Mayr et al. (2005)
have proposed that birds—winged maniraptors—independently evolved
twice, although that conclusion remains controversial (Mayr and
Peters 2006).
The dromaeosaurids, troodontids, and rahonaves did not survive the
mass extinction at the end of the Cretaceous, but some of their living
maniraptoran relatives—the seriema birds—have convergently evolved
a retracted, sickle-shaped claw in the second digits of their feet that is
eerily similar to that seen in the second hind digits of the troodontids.
This convergent retracted sickle claw is well developed in the feet of the
modern red-legged seriema, Cariama cristata, of South America. The
seriemas are long-legged, mostly terrestrial predators that are ecologi-
cally convergent with the secretary birds of Africa. Only two species of
seriemas survive today, but these cariamaen birds are part of a long
lineage of terrestrial avian predators that includes the Eocene “terror
bird” Phorusrhacos longissimus, a flightless raptor that stood almost 3
meters tall.
Is the reappearance of the retracted sickle claw in the Cenozoic car-
iamaen birds an example of convergence due to functional or to devel-
opmental constraint? The number of ways a claw might be modified is
limited, and a sickle claw serves a clear function. Yet why not develop
the retracted sickle claw on the third digit, or the fourth digit, of the foot?
Is there a developmental predisposition within the cariamaen birds to
modify the second digit of the foot, like their ancient troodontid relatives,
and not the third or fourth? Or does the sickle claw simply function best
if it is placed in the second digit of the foot?
Within the placental mammals, fully retractable claws evolved in paral-
lel in three separate lineages during the Cenozoic: the nimravid false cats,
the true cats, and the civets (table 2.7). Not all of the civets have retract-
able claws and, of those that do, the number of digits on the foot that
have retractable claws is variable from species to species. Thus, the
pattern of convergent evolution of retractable claws in the civets is
strikingly different from the pattern seen in the troodontids and the
cariamaen birds. On the other hand, the pattern of convergent evolution
in the true cats and the nimravid false cats is identical—both groups
developed fully retractable claws on all of their digits.
Not all predators have teeth like the sharks in the sea and the cats on
the land. The cephalopods, such as the modern octopus, are ancient mol-
luscan predators that grasp their prey with multiple tentacles, and bite
with a beak that is surprisingly similar in form to the raptorial beak seen
in a hawk (table 2.7). But because the cephalopods evolved in the Late
Cambrian, some 340 million years before the evolution of the first avian
dinosaurs in the Late Jurassic, it would be more correct to state that the
hawks have beaks like the octopus rather than vice versa. Raptorial
beaks appear to be a synapomorphy for the entire Cephalopoda, where
the beaks seen in modern coleoid cephalopods are more similar to the
extinct ammonites than to the modern living ectocochleate cephalopod
Nautilus pompilius.
Raptorial beaks have been convergently evolved by gnathostome ver-
tebrates that have secondarily lost an ancient characteristic trait of most
jawed animals—their teeth. The raptorial beak is well developed in many
types of turtles, themselves very ancient sauropsids, as seen in the modern
snapping turtle. But the raptorial beak is most characteristically, and
convergently, developed in a more derived group of sauropsids, the the-
ropod dinosaurs, as seen in the modern hawks (table 2.7).
The accipitriform raptors include a very diverse group of bird species:
hawks, eagles, kites, and Old World vultures (Accipitridae), the secretary
birds (Sagittaridae), and the ospreys (Pandionidae). The hawks, eagles,
and kites hunt on the wing, whereas the secretary birds are largely ter-
restrial hunters. The ospreys are specialized piscivores that catch fish with
the talons of their feet (the single species Pandion haliaetus is considered
to be so different from the other accipitriform birds that it constitutes
the entire family Pandionidae). And the Old World vultures are mostly
scavengers instead of hunters. Although these birds differ greatly in their
morphologies and ecologies, the raptorial beak may be a synapomorphy
for the entire Accipitriformes, rather than three separate convergences
in the three separate families of accipitriforms (table 2.7).
The falconiform raptors, the falcons and caracaras, are more distantly
related to the accipitriforms, and a more substantial argument may be
made for the convergent development of the raptorial beak in this group.
The New World vultures, such as the familiar turkey vulture in the skies
of North America, are ciconiiform birds even more distantly related to
48 Chapter 2
the accipitriforms (and the ecologically convergent Old World vultures),

and have convergently evolved the raptorial beak. The strigiform owls,
specialized nocturnal predators, are more closely related to humming-
birds than they are to hawks (Lecointre and Le Guyader 2006), yet they
too have convergently developed the raptorial beak. And last, the modern
seriemas and their ancient “terror bird” cousins, Phorusrhacos longissi-
mus, are basal neoavians very distantly related to all the other modern
bird groups (Benton 2005), yet they too have convergently evolved the
raptorial beak. In all, some five to seven different groups of raptorial
birds have developed beaks convergent on those seen in the ancient
turtles and cephalopods.
Some carnivores have evolved even more specialized structures for
killing prey animals than the convergent teeth, claws, and beaks that we
have considered so far. These animals have evolved fangs like hypoder-
mic needles: fangs that are capable of injecting poison directly into the
body of the prey animal. Poison-injecting fangs have been independently
evolved in widely separated groups of arthropods (table 2.7). The most
familiar to us are the spiders and the centipedes. These two predator
groups are very distantly related; the spiders are cheliceriforms and the
centipedes are mandibulates. Yet they have evolved very similar poison-
injecting fangs, as have the larvae of two additional groups of neurop-
teran hexapods: the ant lions and the owlflies. Although one might
suppose that the development of such a specialized structure as a hypo-
dermic fang must surely be confined to the arthropods and their exoskel-
etal mode of growth, this is not the case. The vertebrate snakes, such as
the timber rattlesnakes and their pit viper kin, have convergently evolved
hollow, poison-injecting fangs.
Other predators have taken a backward approach to poisoning their
prey: rather than attacking with poisonous fangs on the front of their
bodies, they use poisonous stingers located on the back of their bodies
(table 2.7). Of these, the most familiar to us are the scorpions and the
wasps. Once again, although these two groups are both arthropods, they
are very distantly related: the scorpions are cheliceriforms and the wasps
are mandibulates. The wasps are a very highly diverse group of flying
predators, yet some have even devolved their wings and become ground
dwellers. One particularly notable example is the velvet ant, also known
as the cow killer ant, which is not an ant at all (as anyone who tries to
pick one up will painfully discover). Just as with poison-injecting fangs,
poisonous stingers are not confined to the arthropods. The vertebrate
stingrays, a group of cartilaginous fishes, have also convergently evolved

a poisonous stinger located on the rear of the animals.
Stingers are not confined to the rear end of animals (table 2.7). The
cone-shell snails, voracious predators of other marine invertebrates, have
a long tonguelike proboscis with tiny radulae, or teeth. Most snails use
these barbed tongues to drill holes into the bodies of their prey, but the
cone-shell snails have evolved radulae that can inject an extremely
deadly neurotoxin. This poison is so deadly it can kill animals as large as
humans, animals much larger than the snail can actually use as a food
resource. A related group of snails, the auger-shell snails, have also
evolved stinger tongues, but the venom they inject is not nearly as poi-
sonous as that of the cone-shell snails.
Stingers are also not confined to highly derived, protostomous and
deuterostomous bilaterian animals. Near the base of the metazoan phy-
logenetic tree, the cnidaria have evolved barbed, poisonous stingers that
are contained in specialized explosive cells, the cnidocytes. The tentacles
of the cnidarians are lined with cnidocytes, which project the stingers into
the body of the prey animal, injecting a paralyzing poison. The cnidarian
then contracts its tentacles to pull the prey to its mouth. Some cnidarians,
like the Portuguese man-of-war jellyfish, are mobile predators of fish and
zooplankton, while others, such as the corals, are sessile predators that
simply wait for prey animals to come within striking distance.
Perhaps the most exotic prey-capture capability has been indepen-
dently evolved by four of the groups of electric fishes: the ability to
generate an electric field strong enough to stun, and some cases even kill,
the prey animal (table 2.7). These fishes have gone beyond the already
unusual ability to detect an electric field in prey hunting that we consid-
ered previously (table 2.5). For example, the freshwater electric eels and
electric catfish can use their electric organs to discharge amplitudes
between 300 to 600 volts, and the marine electric rays and stargazers
discharge amplitudes between 15 to 50 volts (Moller 1995). The electric
eels and electric catfish are thus in many cases able to directly kill their
target frogs and fish with an electric shock, while the electric rays and
stargazers stun their prey into immobility or disorientation, and then
proceed to attack. Interestingly, the electric catfish no longer possess the
poisonous stingers on their fins (a trait we shall consider in more detail
later in the chapter) that are found in so many other groups of catfish—
since they can directly use their electric-field generative ability to defend
themselves, the stingers have been lost.
50 Chapter 2
Herbivores: Plant Processing
Land plants cover extensive areas of the surface of the Earth; thus,
life would at first glance seem to be much easier for a plant-eating her-
bivore than for a carnivore. That is, plant food is generally abundant,
and plants cannot run away, as hunted prey can. The major problem
for herbivores is that plant food is difficult to digest, and essential nutri-
ents are difficult to obtain in sufficient quantities from plants. For
example, consider many plant seeds and grains. They are rich in nutrients,
but they are also very hard, particularly when dried, and can crack your
teeth if you try to chew them. Yet think of the common domestic chicken,
Gallus gallus, an animal that survives on a diet of seeds and grains but
has no teeth. The chicken has evolved an alternative method of grinding
up hard plant material—it has a gastric mill, or gizzard. The chicken
actually deliberately swallows sharp stone fragments, which are then
held in the muscular walls of its gizzard and are used to grind up seeds
and grains.
Gastric mills have been convergently evolved by an astonishing variety
of animals, from earthworms to mammals (table 2.8). The gizzard of the
domestic chicken is most familiar to us, and some consider it to be a
delicacy. But gizzards are not just found in the galliform birds or just in
herbivorous or granivorous birds. The gaviiform red-throated loon,
Gavia stellata, is more closely related to a penguin than a chicken, yet it
also possesses a gastric mill that it uses to grind up the bones of frogs
and the exoskeletons of crustaceans that form part of its diet. The plesio-
morphic paleognathaean ostrich, Struthio camelus, a modern large flight-
less bird, also possesses a gastric mill that it uses to grind the grasses,
roots, and leaves that it normally eats, plus the occasional insect, lizard,
or small mammal.
Curiously, we even have a fossil record for gastric mills. The sharp
stones located within the gizzard are termed “gastroliths,” or stomach
stones, and they become worn and rounded in a characteristic fashion
with usage. Gastroliths are occasionally found within the skeletal remains
of fossil birds and other animals, indicating that these animals possessed
gastric mills in life. Thus, we know that the Cretaceous maniraptoran
dinosaur Caudipteryx zoui had a gizzard—and it was not a bird, although
closely related (note, however, that Maryańska et al. [2002] consider the
oviraptorosaurians to be secondarily flightless birds, and so the gastric
mill of Caudipteryx zoui may be an avian synapomorphy rather than
convergent within the maniraptorans).
Table 2.8
Convergent evolution of gastric mills in herbivores and other animals that require
mechanical assistance to digestion
Convergent structure and function: GASTRIC MILLS (muscular gizzards with embedded
stones or teeth for grinding/processing plant material)
1 Common European earthworm (Bilateria: Protostomia: Lophotrochozoa: Annelida:
Oligochaeta: Lumbricidae; Lumbricus terrestris)
2 Great pond snail (Lophotrochozoa: Mollusca: Gastropoda: Pulmonata: Lymnaeidae;
Lymnaea stagnalis)
3 Priapulid worm (Protostomia: Ecdysozoa: Introverta: Priapulida: Priapulidae;
Priapulus caudatus)
4 Hermit crab (Arthropoda: Mandibulata: Malacostraca: Decapoda: Paguridae;
Eupagurus bernhardus)
5 Madagascar hissing cockroach (Arthropoda: Mandibulata: Hexapoda: Blattodea:
Blattellidae; Gromphadorhina portentosa)
6 Striped mullet (Bilateria: Deuterostomia: Chordata: Osteichthyes: Actinopterygii:
Perciformes: Mugilidae; Mugil cephalus)
7 American gizzard shad (Osteichthyes: Actinopterygii: Clupeiformes: Clupeidae;
Dorosoma cepedianum)
8 Nile crocodile (Osteichtyes: Sarcopterygii: Reptiliomorpha: Amniota: Sauropsida:
Diapsida: Archosauromorpha: Crurotarsi: Crocodylia: Crocodylidae; Crocodylus
niloticus)
9 Sauropodomorph dinosaur (Archosauromorpha: Ornithodira: Dinosauria: Saurischia:
Sauropodomorpha: Plateosauridae; Plateosaurus engelhardi †Triassic)
10 Caudipteryx oviraptor (Dinosauria: Saurischia: Theropoda: Maniraptora:
Oviraptorosauria: Caudipterygidae; Caudipteryx zoui †Cretaceous)
11 Ostrich (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves: Paleognathae:
Struthionidae; Struthio camelus)
12 Psittacosaurid dinosaur (Dinosauria: Ornithischia: Cerapoda: Marginocephalia:
Psittacosauridae; Psittacosaurus mongoliensis †Cretaceous)
13 Giant scaly anteater (Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
Laurasiatheria: Pholidota: Manidae; Manis gigantea)
From the fossil record, we know that gastric mills were convergently
evolved by the second major branch of the saurischian dinosaurs, the
herbivorous sauropodomorphs. Characteristic gastroliths have been
found with the skeletal remains of plesiomorphic prosauropods, like
Plateosaurus engelhardti, of the Triassic and at the very base of the sau-
ropodomorph clade. Gastroliths are later found in the rib cages of giant
herbivorous sauropod dinosaurs, like Seismosaurus, that apparently
swallowed all of their plant food whole and ground it up internally in
their gastric mill (McIntosh 1997). As many as 50 polished gastroliths
have been found in several psittacosaur skeletons, indicating that gastric
mills were convergently evolved in the herbivorous ornithischian dino-
saurs as well (Sereno 1997).
52 Chapter 2
Distant cousins to the dinosaurs, the modern crocodiles and alligators,

have also evolved gastric mills, but not for grinding tough plant material
or hard seeds. These animals are carnivores, but they swallow their food
whole, including the prey’s very hard bones. The crocodiles thus have
convergently evolved a gastric mill for the same reason, but not for the
same food type, as the extinct sauropodomorphs. The bones of their prey
are cracked and ground up by their gastric mills, and then sent onward
to their very high-acid-content stomachs for further processing.
Gastric mills are not a convergent phenomenon to be found only in
the archosaurian clade of tetrapods. Among the mammals, the giant scaly
anteater has evolved a gastric mill to grind the carapaces of the ants that
form the bulk of its diet. It actually has no teeth in its mouth and swal-
lows stones just as granivorous birds do, depending upon its gizzard to
process the hard exoskeletons of its food. The gizzard of the giant scaly
anteater contains an additional element to aid in its grinding function—
stomach teeth. This gizzard innovation, too, has been convergently
evolved at least twice in the arthropods. The stomatogastric systems of
the decapod malacostracans contain gizzards with three hard, chitinous
stomach teeth that assist in the grinding function of the gastric mill.
Diverse decapods such as the red swamp crayfish, Procambarus clarkii,
a common freshwater dweller in streams of southeastern North America,
and the familiar hermit crab, Eupagurus bernhardus, all possess gizzards
with teeth. Many of the hexapod insects, such as the spectacular Mada-
gascar hissing cockroach, have convergently evolved gizzards with chitin-
ous teeth similar to those found in the decapods.
Gastric mills have been convergently evolved in molluscs as well (table
2.8). Many snails, such as the great pond snail, deliberately ingest sand
for their gizzards, and use these gizzards to process tough plant material.
Even the earthworms have evolved gizzards to grind the detritus and
plant fragments that form their diet. The peculiar priapulid worms in the
oceans, more carnivores than detritivores, have also convergently evolved
gizzards to grind up the carapaces of their prey, as well as ingested detri-
tus from sediment. The priapulid worms are ecdysozoans, more closely
related to arthropods than to the lophotrochozoan annelid worms, yet
the two distantly related groups of worms independently evolved similar
gastric mills.
Last, gastric mills have been convergently evolved at least twice in the
fishes. The striped mullet and the American gizzard shad both ingest sand
for their gizzards, and use these gizzards to grind the small invertebrates
and detritus that form their diets. The striped mullet is a perciform fish
and the gizzard shad is a clupeiform fish, and their gizzards are indepen-
dently evolved.
The alternative to grinding up plant material in an internal gizzard is
to grind it up in the mouth itself. In the modern world, mammals are the
undisputed masters of processing food in the mouth. The characteristic
heterodont dentition of the mammals is a veritable toolbox, in that
mammals have specialized teeth for slicing, piercing, shearing, crushing,
and grinding. Moreover, these specialized teeth meet in a precise arrange-
ment when the jaws of the mammal are closed, an arrangement known
as dental occlusion. Mammals represent an extreme in the development
of dental occlusion, where the complex crowns of the molar teeth in the
upper and lower jaw fit together in a precise mortar-and-pestle fashion,
even to the extent of facilitating tooth wear on the crowns to produce
wear facets resulting from tooth-to-tooth contact (DeMar and Bolt 1981;
Benton 2005, 292). In particular, the mammals possess tribosphenic
molars in which a cusp in the upper-jaw molar, the protocone, fits into
an opposing basin in the lower-jaw molar, the talonid (Benton 2005,
307). These complex molars are capable of both shearing and grinding
occlusal functions, and it was long thought that they were a unique
autapomorphy of the therian clade of extinct and extant marsupial and
placental mammals. However, we now know that an extinct group of
australosphenid mammals, the Ausktribosphenida, convergently evolved
the tribosphenic molar (table 2.9). Luo et al. (2001) argue that the
tribosphenic molar evolved vicariantly in the southern continent of
Gondwana (as seen in Ausktribosphenos nyktos), and in the northern
continent of Laurasia (as seen in marsupials and placentals) during the
Cretaceous. In addition, yet another group of even more distantly related
Jurassic mammals, the Shuotheriidae, evolved a reversed, or “pseudo-
tribosphenic,” molar in which the position of a pseudo-talonid basin is
shifted from the posterior part of the molar to the anterior in order to
receive the pseudo-protocone cusp (Luo et al. 2007). The functional end
result is the same as that for the tribosphenic molar, and Luo et al. (2007)
argue that early mammalian dental evolution may be much more itera-
tive and parallel than currently recognized.
Hunter and Jernvall (1995) point out that many of the more derived
tribosphenic mammals have added an additional cusp, a hypocone, to the
original three cusps found in the upper molar. They argue that the acqui-
sition of the hypocone is a key evolutionary adaptation for herbivory,
and that it has been convergently evolved over twenty separate times
within the tribosphenic mammals (see table 1 in Hunter and Jernvall
54 Chapter 2
Table 2.9
Convergent evolution of teeth and dental systems in herbivores and other animals that
require mechanical assistance to digestion
1 Convergent structure and function: TRIBOSPHENIC MOLARS (complex molars

capable of both shearing and grinding occlusal functions)
1.1 Shuotheriid mammal (Mammalia: Yinotheria: Shuotheriidae; Pseudotribos robustus
†Jurassic)
1.2 Ausktribosphenid mammal (Mammalia: Australosphenida: Ausktribosphenida:
Ausktribosphenidae; Ausktribosphenos nyktos †Cretaceous)
1.3 Virginia opossum (Mammalia: Theria: Marsupialia: Didelphimorphia: Didelphidae;
Didelphis virginiana)
2 Convergent structure and function: OCCLUSAL DENTITION (dentition capable of
interlocking occlusion for crushing and chewing food)
2.1 Idiognathodontid conodont (Chordata: Craniata: Conodonta: Ozarkodinida:
Idiognathodontidae; Idiognathodus claviformis †Carboniferous)
2.2 Queensland lungfish (Chordata: Vertebrata: Osteichthyes: Sarcopterygii: Dipnoi:
Ceratodontidae; Neoceratodus forsteri)
2.3 Procolophonoid reptile (Sarcopterygii: Tetrapoda: Reptiliomorpha: Amniota:
Sauropsida: Anapsida: Procolophonoidea; Procolophon trigoniceps †Triassic)
2.4 Trilophosaurid reptile (Sauropsida: Diapsida: Archosauromorpha: Trilophosauridae;
Trilophosaurus buettneri †Triassic)
2.5 Chimaerasuchid notosuchian (Archosauromorpha: Crurotarsi: Crocodylomorpha:
Notosuchia: Chimaerasuchidae; Chimaerasuchus paradoxus †Cretaceous)
2.6 Pakasuchid notosuchian (Archosauromorpha: Crurotarsi: Corcodylomorpha:
Notosuchia: Pakasuchidae; Pakasuchus kapilimai †Cretaceous)
2.7 Diademondontid cynodont (Amniota: Synapsida: Therapsida: Cynodontia:
Diademodontidae; Scalenodon angustifrons †Triassic)
2.8 Traversodontid cynodont (Therapsida: Cynodontia: Traversodontidae;
Massetognathus pascuali †Triassic)
2.9 Tritylodontid cynodont (Therapsida: Cynodontia: Tritylodontidae; Kayentatherium
wellesi †Triassic)
2.10 Triassic mammal (Therapsida: Cynodontia: Mammalia: Adelobasileus cromptoni
†Late Triassic)
3 Convergent structure and function: DENTAL BATTERIES (multiple rows of teeth in
jaw to form a grinding/shearing surface for processing plant material)
3.1 Hadrosaurid dinosaur (Archosauria: Ornithodira: Dinosauria: Ornithischia:
Cerapoda: Ornithopoda: Hadrosauridae; Parasaurolophus walkeri †Cretaceous)
3.2 Ceratopsid dinosaur (Dinosauria: Ornithischia: Cerapoda: Marginocephalia:
Ceratopsia: Ceratopsidae; Chasmosaurus belli †Cretaceous)
1995). Jernvall (2000) argues that only small developmental changes are
needed to produce major changes in cusp numbers and sizes in mam-
malian teeth, and that this explains the frequent convergent evolution of
new cusps.
Just as the tribosphenic molar in mammals is convergent, dental occlu-
sion itself is not a unique trait of the mammals. In fact, mammals are
only one of four separate groups of cynodont therapsids that conver-
gently evolved complex dental occlusion in the Triassic (table 2.9). The
tritylodontids, such as Kayentatherium wellesi, are the closest relatives to
the most ancient mammal known to us, Adelobasileus cromptoni (Benton
2005, 290), but they independently evolved dental occlusion. The traver-
sodontids are more distantly related, and the diademondontids even
more so, yet they too independently evolved dental occlusion (Rowe
1993; Martinez et al. 1996). Interestingly, all three of these cynodont
groups were secondarily herbivorous, in that most of the cynodonts were
carnivores or omnivores. Thus, in the Triassic, evolution produced four
separate and different cynodont experiments in complex dental occlu-
sion and extensive food processing in the mouth, only one of which
survives to the present day (the mammals).
Dental occlusion is also not a unique trait of the synapsid amniotes.
The other branch of the amniote animals, the reptilian sauropsids, also
evolved independent lineages of animals with at least partial dental
occlusion. This is particularly interesting in that the sauropsids usually
replace their teeth continually throughout life, whereas the mammals
have only two sets of teeth—the early deciduous teeth, and the later set
of permanent teeth. Occlusion is particularly difficult to maintain in a
jaw that contains teeth of differing ages, and hence different shapes and
positions, as is commonly the case in sauropsids.
Even so, the convergent evolution of dental occlusion has been
reported in some extinct archosaurs and primitive sauropsids. Wu et al.
(1995, 678) describe a Cretaceous notosuchian crocodylomorph from
China, Chimaerasuchus paradoxus, that has teeth “very similar to that of
the postcanine teeth of tritylodontid synapsids and represents a particu-
larly striking example of convergent evolution.” This animal was second-
arily herbivorous, as most of the crocodile-like animals were carnivores,
and thus also represents an ecological convergence to the tritylodontid
cynodonts, which were also secondary herbivores. A second, indepen-
dent, evolution of molariform teeth and occlusal precision in Cretaceous
notosuchians is reported from Tanzania for Pakasuchus kapilimai (table
2.9). This species belongs to a separate notosuchian clade than the
56 Chapter 2
Chinese Chimaerasuchus paradoxus, yet it also evolved molariform teeth

“paralleling the level of occlusal complexity seen in mammals” (O’Connor
et al. 2010, 748). Noting that ancient mammals were rare on the southern
supercontinent Gondwana in the Cretaceous, in contrast to the northern
supercontinent Laurasia, O’Connor et al. (2010, 748) suggest that these
“notosuchians probably filled niches and inhabited ecomorphospaces
that were otherwise occupied by mammals on northern continents.” (The
phenomenon of ecological-niche convergence by other disparate phylo-
genetic lineages will be considered in detail in chapter 4.) Finally, the
convergent evolution of “complex, multi-cusped, mammal-like teeth” has
also been reported in another Cretaceous crocodylomorph from Africa
(Clark et al. 1989, 1064), and the evolution of dental occlusion has been
reported in even more distantly related archosauromorph diapsids, Trilo-
phosaurus buettneri and Tricuspisaurus thomasi (Robinson 1956; DeMar
and Bolt 1981), and in a procolophonoid anapsid (Carroll and Lindsay
1985), all from the Triassic (table 2.9).
Leaving terrestrial ecosystems, we find that occlusal dentition has only
rarely been reported for aquatic animals. The Queensland lungfish devel-
ops a series of tooth plates within its mouth, which are cusped and which
occlude to form a crushing surface that it uses to eat shelled invertebrates
such as snails and prawns (Kemp 1977). Far back in time, the Carbonifer-
ous conodont Idiognathodus claviformis apparently crushed food with
the complex, interpenetrative oral surfaces of its molariform Pa ele-
ments. These were toothlike structures reminiscent of gastric mill teeth
in that, because the conodonts had no jaws, these element structures
were located in the throat posterior to other teethlike elements near the
opening of the mouth. Still, microwear patterns on their oral surfaces
indicate that these conodonts “developed dental occlusion of mammal-
like complexity” (Donoghue and Purnell 1999, 58).
In the end, mammals still remain the evolutionary masters of chewing
efficiency. Or perhaps they are not? There exist two groups of herbivo-
rous dinosaurs that clearly rivaled—if not surpassed—the mammals in
specialized masticatory adaptation. These were the hadrosaurids and the
ceratopsids. Of all the sauropsids, only the ornithopod ornithischians
were able to chew their food with the efficiency of a mammal, although
they did so in a different fashion (Benton 2005, 207). The highly derived
hadrosaurid ornithopods further increased their chewing efficiency by
evolving multiple rows of teeth in each jaw, not just the normal one row.
These multiple rows of teeth, known as dental batteries, contained 40 to
60 teeth per battery, with the result that the typical hadrosaur contained
hundreds of teeth in its mouth. Moreover, the hadrosaurs continued to

grow and replace teeth in the dental batteries throughout their lifespan
(a plesiomorphic sauropsid feature), so a typical animal produced thou-
sands of teeth, replacing them constantly as the older teeth were worn
down. The hadrosaurs have been argued to have been “Mesozoic ungu-
lates” (Carrano et al. 1999), ecologically and morphologically convergent
on modern ungulate mammals (we shall consider ecological convergence
in detail in chapter 4). Unlike the hadrosaurs, modern grazing ungulates
that eat tough grass, like the horse or cow, have only one set of perma-
nent teeth. Because these teeth have very high crowns, they can last the
lifetime of the animal, as the tooth is worn away progressively from the
top down (Janis 1990).
Dental batteries are not unique to the hadrosaurs. The ceratopsid
ornithischians also evolved multiple rows of teeth, but their dental bat-
teries are different in structure from those of the hadrosaurs. The sur-
faces of the ceratopsid dental batteries are more bladelike, rather than
the rasplike batteries of the hadrosaurs, and were used more for shearing
than for grinding. The ceratopsids are descendants of the psittacosaurs,
animals that simply swallowed their food whole and processed it inter-
nally in a gastric mill (Sereno 1997). Thus, the peculiar evolution of
multiple rows of teeth in the jaw occurred twice, independently, in the
ornithischian dinosaurs (table 2.9).
A major problem for herbivores is the digestion of cellulose. Some-
times mechanical crushing and shredding of cellulose is simply not suf-
ficient, and so chemical fermentation is used as an alternative digestive
route. Multiple lineages of mammals have convergently evolved special-
ized digestive structures to house and protect anaerobic bacteria, and to
allow these bacteria to ferment cellulose within the structures. Specifi-
cally, two types of cellulose fermentation systems have evolved: stom-
achal-fermentation systems, in which specialized stomach compartments
house the anaerobic bacteria, and cecal-fermentation systems, in which
an intestinal cecum houses the anaerobic bacteria (Lecointre and Le
Guyader 2006). Stomachal-fermentation systems have convergently
arisen in three groups of cetartiodactyls—the ruminants, the camels, and
the hippopotamuses—one group of primates—the colobus monkeys—
and even in one type of herbivorous bird (table 2.10). Lecointre and Le
Guyader (2006) argue that the stomachal-fermentation system evolved
independently and in parallel in the three cetartiodactyl groups, and that
it is not a synapomorphy for these groups. The four-chambered stomachs
of the many ruminant groups are quite different from those of the
58 Chapter 2
Table 2.10
Convergent evolution of chemical digestive structures and systems in herbivores
1 Convergent structure and function: STOMACHAL-FERMENTATION SYSTEM

(specialized stomach compartments house anaerobic bacteria for fermentation of cellulose
in digestion)
1.1 Hoatzin (Amniota: Sauropsida: Diapsida: Archosauromorpha: Ornithodira:
Dinosauria: Saurischia: Theropoda: Aves: Opisthocomiformes: Opisthocomidae;
Opisthocomus hoatzin)
1.2 Dromedary camel (Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
Laurasiatheria: Cetartiodactyla: Tylopoda: Camelidae; Camelus dromedarius)
1.3 Red deer (Mammalia: Eutheria: Laurasiatheria: Cetartiodactyla: Ruminantia:
Cervidae; Cervus elaphus)
1.4 Hippopotamus (Eutheria: Laurasiatheria: Cetartiodactyla: Hippopotamidae;
Hippopotamus amphibius)
1.5 Western red colobus monkey (Eutheria: Euarchontoglires: Primates: Cercopithecidae;
Colobus badius)
2 Convergent structure and function: CECAL-FERMENTATION SYSTEM (intestinal
cecum houses anaerobic bacteria for fermentation of cellulose in digestion)
2.1 Yellow-spotted hyrax (Mammalia: Eutheria: Afrotheria: Hyracoidea: Procaviidae;
Heterohyrax brucei)
2.2 Mountain zebra (Eutheria: Laurasiatheria: Perissodactyla: Equidae; Equus zebra)
2.3 European rabbit (Eutheria: Euarchontoglires: Lagomorpha: Leporidae; Oryctolagus
cuniculus)
3 Convergent structure and function: RUMINANT-STOMACH SYSTEM (multiple-
stomach food processing to allow delayed mastication, thus decreasing actual feeding time
and predator exposure)
3.1 Red kangaroo (Mammalia: Marsupialia: Diprotodontia: Macropodidae; Macropus
rufus)
3.2 Red deer (Mammalia: Eutheria: Laurasiatheria: Cetartiodactyla: Ruminantia:
Cervidae; Cervus elaphus)
3.3 Dromedary camel (Eutheria: Laurasiatheria: Cetartiodactyla: Tylopoda: Camelidae;
Camelus dromedarius)
Note: For data sources, see text.
tylopod camels, guanocos, and their kin, and yet more different from the
stomachs of the hippopotamuses. The colobus monkeys are primates,
very different animals from the cetartiodactyls, yet they too have con-
vergently evolved a stomachal-fermentation system to deal with their
exclusively leaf-eating herbivorous diet. And last, the peculiar hoatzin or
“stink bird” is an avian dinosaur, very far removed from the clade of the
mammals, yet it also has independently evolved a stomachal-fermenting
system (table 2.10).
The cecal-fermentation system also has been convergently evolved by
a phylogenetically diverse group of mammals (table 2.10): euarchonto-
glirean lagomorphs (rabbits, hares, and pikas), afrotherian hyracoids
(hyraxes), and laurasiatherian perissodactyls (horses, rhinos, and tapirs).

The lagomorphs are particularly interesting in that they recycle the cel-
lulose-rich part of their diet: the food passes first through the cecum,
where the bacteria degrade the cellulose and the cecum absorbs the
released nutrients, leaving behind a soft fecal pellet enriched in vitamins
and bacteria. These soft pellets are defecated, reswallowed, and redi-
gested by the rabbit, so that the pellet’s vitamins are absorbed and the
symbiotic bacteria are digested as food. This peculiar digestive process,
termed “caecotrophy” (Lecointre and Le Guyader 2006, 438), is conver-
gent—at least in its digestive effects—on the regurgitation and rediges-
tion process used by the ruminant cetartiodactyls.
Rumination has been convergently evolved by three separate groups
of mammals (table 2.10): twice independently in two groups of placental
cetartiodactyls (ruminants and camels), and a third time in the very
distantly related marsupial mammals, the kangaroos (Lecointre and Le
Guyader 2006, 469). Ruminants minimize feeding time and predator
exposure by eating their food rapidly, and thus usually chew it poorly.
Later, in more protected settings, the partially digested food is regurgi-
tated, remasticated, and reswallowed for further digestive processing.
Before moving on to consider convergent antipredator adaptations, let
us consider the pattern of evolution of one last feeding structure: animal
beaks. We are all familiar with the distinctive beaks of modern ducks and
parrots. Parrots use their pincerlike, hooked beaks to manipulate, crack
open, and extract food from seeds and nuts. Ducks use their flattened
beaks to filter food material from water. Both of these modern animals
are highly derived theropod dinosaurs. In the Cretaceous another group
of dinosaurs, the psittacosaurs and their ceratopsian descendants, inde-
pendently evolved beaks that are astonishingly similar to parrot beaks,
yet these dinosaurs were highly derived marginocephalian ornithischi-
ans, not theropods (table 2.11). Analyses of tooth wear in a recently
discovered, exceptionally well-preserved psittacosaur from Mongolia
has led Sereno et al. (2009) to the conclusion that the diet of this
psittacosaur was identical to that of a parrot, that is, it was a “nucivore”
(nut eater).
The closest common ancestor for these two groups of animals (psit-
tacosaurs and parrots) lived over 240 million years ago, in the Middle
Triassic. In the Triassic there existed another group of animals that con-
vergently evolved parrotlike beaks, the rhynchosaurid reptiles. Parrotlike
beaks are not confined to the sauropsid amniotes, however, as in the
Permian a group of dicynodontid therapsids in the synapsid amniote
60 Chapter 2
Table 2.11
Convergent evolution of parrot and duck beaks
1 Convergent structure and function: PARROT BEAK (laterally flattened, dorsoventrally

flared, pincerlike hooked beak for holding, cracking open, and extracting food from seeds
and nuts)
1.1 Rhynchosaurid reptile (Amniota: Sauropsida: Diapsida: Archosauromorpha:
Rhynchosauria: Rhynchosauridae; Hyperodapedon gordoni †Triassic)
1.2 Scarlet macaw (Archosauromorpha: Archosauria: Ornithodira: Dinosauria:
Saurischia: Therapoda: Maniraptora: Aves: Psittaciformes: Psittacidae; Ara macao)
1.3 Psittacosaurid dinosaur (Dinosauria: Ornithischia: Cerapoda: Marginocephalia:
Psittacosauridae; Psittacosaurus mongoliensis †Cretaceous)
1.4 Dicynodontid therapsid (Amniota: Synapsida: Therapsida: Dicynodontia:
Dicynodontidae; Pristerodon mackayi †Permian)
2 Convergent structure and function: DUCK BEAK (dorsoventrally flattened, laterally
flared, horn beak for filtering food from water)
2.1 Wood duck (Amniota: Sauropsida: Diapsida: Archosauromorpha: Ornithodira:
Dinosauria: Saurischia: Theropoda: Maniraptora: Aves: Anseriformes: Anatidae; Aix
sponsa)
2.2 Hadrosaurid dinosaur (Dinosauria: Ornithischia: Cerapoda: Ornithopoda:
Hadrosauridae; Anatotitan copei †Cretaceous)
2.3 Duckbill platypus (Amniota: Synapsida: Therapsida: Mammalia: Monotremata:
Ornithorhynchidae; Ornithorhynchus anatinus)
lineage also independently evolved parrotlike beaks. All of these animals

were herbivorous.
There are no living synapsids that still possess parrotlike beaks, but
there are some that possess beaks that are identical to those of ducks—
the duckbill platypuses (table 2.11). As occurred with parrot beaks, a
group of distantly related ornithischian dinosaurs, the hadrosaurs, inde-
pendently evolved beaks like those of theropod ducks. Mechanically, the
mammalian platypus uses its beak exactly as a duck does, filtering small
aquatic animals out of the water with the transverse ridges of its beak,
ridges that are similar to the teethlike serrations of the duck’s beak. The
hadrosaurs, however, were not aquatic animals. Their beak usage was
probably more like that of modern geese, which use their ducklike beaks
to crop grass and other plant material.
Defense: Antipredator Adaptations
Many of the same structures that predators use to kill prey can also be
used to defend themselves from other predators. In this section of the
chapter, I will concentrate on the convergent evolution of structures and
systems that are primarily for defense. Although a few predators have
evolved eyes with horizontal-slit pupils (as we have previously discussed),
eyes of this type are very widespread in herbivores. Most familiar,
perhaps, are the odd-looking eyes of domestic goats, although the same
type of eye is present in sheep, cows, deer, elks, hippopotamuses, and so
on. It is a near universal trait of the cetartiodactyl mammals. However,
horses and their kin also have eyes with horizontal-slit pupils; thus, this
characteristic has clearly evolved independently in the perissodactyl
mammals. A third group of mammals, the afrotherian manatees, have
convergently evolved the horizontal-slit pupil in the eye as well (table
2.12). In all cases, this type of eye is used by herbivores for predator
detection at a distance, in all directions, and thus for predator
avoidance.
A typical herbivore strategy of predator avoidance is simply to run
away—as fast as possible. The evolution of fast running has anatomical
consequences, a functional constraint that leads directly to convergent
evolution. Bakker (1983) has demonstrated that six morphological
Table 2.12
Convergent evolution of herbivore predator-avoidance systems
1 Convergent structure and function: HORIZONTAL-SLIT PUPILS IN EYES (allows

[1] full usage of the diameter of the lens of the eye in bright light as well as in very low
light intensities, with well-focused images in all light intensities, and [2] in combination with
near 360-degree vision, continuous viewing of almost total horizon for predator detection
in all light intensities)
1.1 African manatee (Mammalia: Eutheria: Afrotheria: Sirenia: Trichechidae; Trichechus
senegalensis)
1.2 Mountain zebra (Mammalia: Eutheria: Laurasiatheria: Perissodactyla: Equidae;
Equus zebra)
1.3 Domestic goat (Eutheria: Laurasiatheria: Cetartiodactyla: Ruminantia: Bovidae;
Capra aegagrus)
2 Convergent structure and function: DIGIT-NUMBER REDUCTION (reduction of
muscle mass in the distal parts of limbs, elongation of the bones in the feet, loss of digits
in the foot, and the aquisition of hooves to protect the digits from shock in running, to
produce fast-running spindly-limbed animals)
2.1 Modern horse (Mammalia: Eutheria: Laurasiatheria: Perissodactyla: Equidae;
Equus caballus)
2.2 Litoptern “horse” (Eutheria: Laurasiatheria: Meridiungulata: Litopterna:
Proterotheriidae; Thoatherium minusculum †Miocene)
2.3 Red deer (Eutheria: Laurasiatheria: Cetartiodactyla: Ruminantia: Cervidae; Cervus
elaphus)
2.4 Guanaco (Eutheria: Laurasiatheria: Cetartiodactyla: Tylopoda: Camelidae; Lama
guanicoe)
62 Chapter 2
transitions occur in the evolution of fast runners from slow, and that
these morphological changes have repeatedly occurred in distantly
related animal groups that convergently evolved fast runners. These
evolutionary changes involve lengthening and attenuation of the distal
bone shafts in the limbs while the proximal bones of the limbs become
shorter, lengthening and attenuation of the metapodial bones in the feet
as the animals locomote first on their toes and then on the tips of their
toes, shortening of the phalangeal bones in the foot, and transformation
of claws into hooves to protect the ends of the toes from shock and
damage while running. Last, the number of hoofed toes in the foot are
reduced, leading in some cases to the animal standing upon a single
hoofed toe. The spindly legs of red deer are a familiar example of this
morphological transition: the upper bone of the hind limb (femur) is
short and surrounded by a compact mass of muscle (which controls the
limb) held next to the body of the deer, the lower limb bone shafts (tibia
and fibula) are elongated away from the deer’s body and contain much
less muscle mass, the foot is enormously elongated into a thin shaft of
very low mass, and the animal stands on the tips of only two hoofed toes.
This form of limb, so common in many ruminant cetartiodactyls, has been
independently evolved in the tylopod cetartiodactyls as well, even to the
extent of reduction of the number of digits in the foot to two (table 2.12).
The more distantly related perissodactyls have not only independently
evolved the same type of running limb but have further reduced the
number of hoofed digits in the foot to one, as is seen in the modern horse.
In the Miocene, a group of distantly related South American meridiun-
gulates convergently evolved limbs like those seen in modern horses. The
ancient one-toed litopterns that ran across the plains of South America
looked like horses, but were not (table 2.12).
Another antipredator adaptation is to acquire armor. An armored
body can be of the passive, tanklike form or the actively offensive, spine
armor form. Very distantly related animals have repeatedly evolved
tanklike body armor: the saurposid turtles, the dinosaurian ankylosaurs,
and the mammalian glyptodonts, armadillos, and pangolins (table 2.13).
When confronted by a predator, these animals retreat within their armor
and passively wait until the predator gives up trying to attack them and
moves on to more vulnerable prey. The morphological convergence
between the Cretaceous ankylosaurid dinosaurs and the Pleistocene
glyptodontid mammals is astonishing—not just in the similarity of the
body armor, but even down to the offensive club located at the ends of
their tails, which these animals used to strike out at a predator while
Table 2.13
Convergent evolution of body armor in animals
1 Convergent structure and function: TANKLIKE BODY ARMOR (shell of rigid to

semirigid plates that enclose soft body tissues for defense against predators)
1.1 Eastern box turtle (Amniota: Sauropsida: Anapsida: Testudines: Emydidae;
Terrapene carolina)
1.2 Ankylosaur (Diapsida: Archosauromorpha: Ornithodira: Dinosauria: Ornithischia:
Thyreophora: Ankylosauridae; Ankylosaurus magniventris †Cretaceous)
1.3 Glyptodont (Amniota: Synapsida: Therapsida: Mammalia: Eutheria: Xenarthra:
Cingulata: Glyptodontidae; Doedicurus clavicaudatus †Pleistocene)
1.4 Nine-banded armadillo (Mammalia: Eutheria: Xenarthra: Dasypodidae; Dasypus
novemcinctus)
1.5 Giant scaly anteater (Mammalia: Eutheria: Laurasiatheria: Pholidota: Manidae;
Manis gigantea)
2 Convergent structure and function: SPINOSE BODY ARMOR (external covering of
sharp spines that enclose soft body tissues for defense against predators)
2.1 Australian echidna (Mammalia: Monotremata: Tachyglossidae; Tachyglossus
aculeatus)
2.2 Greater hedgehog tenrec (Mammalia: Eutheria: Afrotheria: Afrosoricida:
Tenrecidae; Setifer setosus)
2.3 European hedgehog (Eutheria: Laurasiatheria: Eulipotyphles: Erinaceidae;
Erinaceus europaeus)
2.4 North American porcupine (Eutheria: Euarchontoglires: Rodentia: Caviomorpha:
Erethizontidae; Erethizon dorsatum)
2.5 Crested porcupine (Eutheria: Euarchontoglires: Rodentia: Hystricomorpha:
Hystricidae; Hystrix cristata)
3 Convergent structure and function: CHEMICAL BODY ARMOR (poison concentrated
in skin or body tissues for defense against predators)
3.1 Mamor nudibranch snail (Bilateria: Protostomia: Lophotrochozoa: Mollusca:
Gastropoda: Opisthobranchia: Nudibranchia: Dendrodorididae; Dendrodoris
grandiflora)
3.2 Monarch butterfly (Protostomia: Ecdysozoa: Arthropoda: Mandibulata: Hexapoda:
Lepidoptera: Danaidae; Dananus plexippus)
3.3 Margined blister beetle (Arthropoda: Mandibulata: Hexapoda: Coleoptera:
Meloidae; Epicauta pestifera)
3.4 Pyralis firefly (Arthropoda: Mandibulata: Hexapoda: Coleoptera: Lampyridae;
Photinus pyralis)
3.5 Puffer fish (Bilateria: Deuterostomia: Chordata: Osteichthyes: Actinoptergyii:
Tetraodontiformes: Tetraodontidae; Takifugu vermicularis)
3.6 Bumblebee poison-arrow frog (Osteichtyes: Sarcopterygii: Tetrapoda:
Batrachomorpha: Lissamphibia: Batrachia: Anura: Dendrobatidae; Dendrobates
leucomelas)
3.7 Fire salamander (Lissamphibia: Batrachia: Urodela: Salamandridae; Salamandra
salamandra)
3.8 Crested auklet (Tetrapoda: Reptiliomorpha: Amniota: Sauropsida: Diapsida:
Archosauromorpha: Ornithodira: Dinosauria: Saurischia: Theropoda: Maniraptora:
Aves: Charadriiformes: Alcidae; Aethia cristatella)
3.9 Hooded pitohui (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves:
Passeriformes: Pachycephalidae; Pitohui dichrous)
64 Chapter 2
crouched down under their armor. Both creatures had to live with deadly
predators—the ankylosaurids were preyed upon by the tyrannosaurs,
and the glyptodontids had to deal with saber-tooth cats—and they
evolved the same morphological solution, even though they are vastly
different animals.
Besides passively protecting the bearer, spinose body armor also
presents the danger of injury to the predator. Animals bristling with
sharp spines have repeated evolved in widely separated lineages—
monotreme echidnas, laurasiatherian hedgehogs, and afrotherian
tenrec “hedgehogs”—and twice independently have evolved among the
rodents—the North American porcupines and the African porcupines
(table 2.13).
A third type of body armor is not mechanical but chemical. Animals
with chemical defenses are also some of the prettiest, most colorful, and
most deadly in nature (table 2.13). Their brightly colored bodies are a
visible warning to predators that their body tissues are loaded with
poison. In the oceans, the wildly colorful nudibranch snails not only
manufacture their own poison, such as sesquiterpene toxin (Cimino et
al. 1985), but can store and concentrate poisons found in their environ-
ment. The bizarre puffer fish have deadly tetrodotoxin in their skin and
organs; if eaten, the predator swiftly dies.
On land, diverse arthropods such as pretty orange-and-black monarch
butterflies store and accumulate cardenolide poison from the milkweeds
on which they feed (we shall consider poisonous plants in detail in the
next chapter); blister beetles can burn with cantharidin toxin if eaten;
and brilliant flashing fireflies contain lucibufagin poison—and insecti-
vores avoid them all (Agosta 1996). In the amphibians, the beautifully
colored poison-arrow frogs are deadly, and toxin from their skin has been
used by humans to poison the tips of arrows (hence their name). Fire
salamanders have independently evolved poison glands along the tops
of their backs. Even a few birds have convergently evolved chemical
body armor; the skin and feathers of the hooded pitohui are poisonous
(Agosta 1996), and the crested auklet secretes antiectoparasite toxins
(Douglas et al. 2005), creating its own mosquito repellant.
A last line of antipredator defense is to fight back. In the seas, five
different groups of bony fishes have evolved poisonous stinger spines for
defense (table 2.14). The scorpion fishes are diverse and deadly: some
are very pretty, like the red lionfish that is popular in aquariums even
though it is dangerous; others are nondescript and cryptic, like the stone-
fish Synanceia horrida, which can kill a human if inadvertently stepped
Table 2.14
Convergent evolution of specialized defensive structures in animals
1 Convergent structure and function: STINGER SPINES (poison-bearing spines for

defensive use against predators)
1.1 Red lionfish (Osteichthyes: Actinopterygii: Teleostei: Scorpaeniformes:
Scorpaenidae; Pterois volitans)
1.2 Weever fish (Actinopterygii: Teleostei: Perciformes: Trachinidae; Trachinus vipera)
1.3 Coral rabbitfish (Actinopterygii: Teleostei: Perciformes: Siganidae; Siganus
corallinus)
1.4 Poison toadfish (Actinopterygii: Teleostei: Batrachoidiformes: Batrachoididae;
Thalassophryne megalops)
1.5 Striped eel catfish (Actinopterygii: Teleostei: Siluriformes: Plotosidae; Plotosus
lineatus)
1.6 Duckbill platypus (Osteichthyes: Sarcopterygii: Tetrapoda: Reptiliomorpha:
Amniota: Synapsida: Therapsida: Mammalia: Monotremata: Ornithorhynchidae;
Ornithorhynchus anatinus)
2 Convergent structure and function: STINGERS (poison stingers for defensive use
against predators)
2.1 Honeybee (Arthropoda: Mandibulata: Hexapoda: Hymenoptera: Apidae; Apis
mellifera)
2.2 Fire ant (Arthropoda: Mandibulata: Hexapoda: Hymenoptera: Formicidae;
Solenopsis geminata)
2.3 Long-horned beetle (Arthropoda: Mandibulata: Hexapoda: Coleoptera:
Cerambycidae; Onychocerus albitarsis)
3 Convergent structure and function: CHEMICAL SPRAY (poisonous chemical sprays
for defensive use against predators)
3.1 Common millipede (Bilateria: Protostomia: Ecdysozoa: Arthropoda: Mandibulata:
Myriapoda: Diplopoda: Spirobolidae; Narceus annularis)
3.2 Flat-backed millipede (Arthropoda: Mandibulata: Myriapoda: Diplopoda:
Polydesmidae; Polydesmus angustus)
3.3 Neotropical arboreal termite (Arthropoda: Mandibulata: Hexapoda: Isoptera:
Termitidae; Nasutitermes corniger)
3.4 Bombardier beetle (Arthropoda: Mandibulata: Hexapoda: Coleoptera: Carabidae;
Brachinus explodens)
3.5 Green stinkbug (Arthropoda: Mandibulata: Hexapoda: Hemiptera: Pentatomidae;
Acrosternum hilare)
3.6 Striped skunk (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Reptiliomorpha: Amniota: Synapsida: Therapsida: Mammalia: Eutheria: Laurasiatheria:
Carnivora: Mephitidae; Mephitis mephitis)
3.7 European polecat (Eutheria: Laurasiatheria: Carnivora: Mustelidae; Mustela
putorius)

66 Chapter 2
on. The weever fishes, rabbitfishes, toadfishes, and catfishes have all con-
vergently evolved this line of defense—as has the duckbill platypus on
land, with stinger spurs on its hind feet.
We have previously considered predators that sting; three groups of
arthropods have convergently evolved stingers for defense. These are the
bees, some ants, and some long-horned beetles (table 2.14). The social
bees in particular will defend their nests by stinging, which is an act of
self-sacrifice as the bee dies afterward. Most ants bite to defend their
nests, but a few have independently modified their ovipositors to produce
stingers, in parallel evolution with their relatives the bees. The long-
horned beetle Onychocerus albitarsis has evolved hypodermic antennae;
that is, it uses its antennae as stingers with “a delivery system almost
identical to that found in the stinger of a deadly buthid scorpion” (Berkov
et al. 2008, 257).
Rather than use a stinger, which requires close contact with the preda-
tor, other animals have convergently evolved systems to spray poisonous
chemicals at the predator from a distance (table 2.14). In the mammals,
the New World skunks and Old World polecats are legendary as well as
interesting, in that most mammals do not use chemical defenses. In the
arthropods, the bombardier beetle is most famous because it not only
sprays noxious liquids with quinone (Agosta 1996) but heats the liquid
to boiling before spraying. The common millipede also will spray quinone
if disturbed, whereas the polydesmid millipedes spray deadly hydrogen
cyanide.
Why is chemical defense so rare in terrestrial mammals and birds,
but so common in terrestrial panarthropods and amphibians? From the
perspective of theoretical morphology, one can easily visualize a non-
existent poison-skinned mouse or sparrow, convergent on a poison-
arrow frog, that has evolved a chemical defense to deter a cat or hawk
predator. Yet the poison-arrow frog exists, whereas the poison-skinned
mouse or sparrow does not. Does a developmental constraint exist
for the generation and maintenance of skin poisons in endothermic
animals? Or is the constraint functional, in that other forms of defense,
perhaps less metabolically costly and complicated, are adequate in endo-
thermic animals?
Organ Systems
We have previously considered the convergent evolution of pupil shapes

in the eyes of carnivores (table 2.5) and herbivores (table 2.12). But what
about the eye itself? In the classic study of the convergent evolution of
the eye, Salvini-Plawen and Mayr (1977) presented evidence for the
definite independent evolution of light-detection organs in 40 separate
animal lineages, and the probable independent origination of light-
detection organs in 20 additional lineages, bringing the total to 60 sepa-
rate convergences. They also document the convergent evolution of the
process of eye evolution itself: from single photoreceptor cells scattered
across the body of an animal, to distinct aggregations of photoreceptor
cells into eyespots, to ocellar pits, to ocellar cups, to simple eyes, and to
complex compound and camera eyes. This same evolutionary sequence
can be observed in multiple independent lineages of animals.
In table 2.15 I have conservatively taken the 40 definite cases of con-
vergent eye evolution given by Salvini-Plawen and Mayr (1977), reana-
lyzed each case in terms of a phylogenetic classification of life (see
appendix), and added additional convergent cases that modern phyloge-
netic analyses have revealed. These additional factors have altered the
reasoning of Salvini-Plawen and Mayr (1977) somewhat; for example,
because Salvini-Plawen and Mayr considered the arthropods to be
descendants of the annelids, they considered the possible homology of
light-detection structures found in both groups of animals. We now know
that the arthropods are ecdysozoan protostomes more closely related to
introvert worms like the priapulids and nematodes, whereas the annelids
are lophotrochozoan protostomes more closely related to molluscs
(Lecointre and Le Guyader 2006), and thus the two are not closely
related. These considerations have resulted in bringing the number of
definite convergences to 49 (table 2.15).
It is clear that eyes and other light-detection organs are ancient inven-
tions of the animals. Simple eyespots, more complex ocellar cups, and
even complex camera eyes can be found in some of the most plesiomor-
phic animals still alive, the most simple of the eumetazoa, the cnidarians.
Complex camera eyes in our own lineage, the Chordata, were developed
over 500 million years ago in the craniate conodonts of the Cambrian
(Purnell 1995).
Light-detection organs have been repeatedly evolved, lost, and
reevolved to varying degrees of complexity throughout the past 600
million years. Thus, we find simple, isolated photoreceptor cells scattered
across the tissues of bilaterian animals as widely divergent as lophotro-
chozoan protostome rock chitons, ecdysozoan protostome earthworms,
and even in the tails of the ammocoete larvae of deuterostome sea
lampreys (table 2.15). These are all highly derived animals compared
68 Chapter 2
Table 2.15
Convergent evolution of eyes and light-detection systems in animals
1 Convergent structure and function: PHOTORECEPTOR CELLS (cells with enlarged

surface areas containing photopigment for light detection)
1.1 Common European earthworm (Bilateria: Protostomia: Lophotrochozoa: Annelida:
Oligochaeta: Lumbricidae; Lumbricus terrestris)
1.2 Common rock chiton (Lophotrochozoa: Mollusca: Polyplacophora: Chitonidae;
Acanthopleura spiniger)
1.3 Soft-shelled clam (Mollusca: Bivalvia: Myoida: Myidae; Mya arenaria)
1.4 Marine nematode (Protostomia: Ecdysozoa: Introverta: Nematozoa: Nematoda:
Oncholaimidae; Oncholaimus vesicarius)
1.5 Sea lamprey (Bilateria: Deuterostomia: Chordata: Vertebrata: Petromyzontiformes:
Petromyzontidae; Petromyzon marinus)
2 Convergent structure and function: EYESPOTS (photoreceptor cells aggregated into
distinct spot regions for increased light-detection abilities)
2.1 Pandeid jellyfish (Metazoa: Cnidaria: Hydrozoa: Anthomedusae: Pandeidae;
Leuckartiara octona)
2.2 Asplanchnid rotifer (Metazoa: Bilateria: Protostomia: Lophotrochozoa: Syndermata:
Rotifera: Asplanchnidae; Asplanchna brightwelli)
2.3 Green spoon worm larvae (Lophotrochozoa: Annelida: Echiuria: Bonelliidae;
Bonellia viridis)
2.4 Palola worm (Annelida: Polychaeta: Eunicemorpha: Eunicidae; Palola viridis)
2.5 Sabellid bristle worm (Annelida: Polychaeta: Sabellidae; Notaulax rectangulata)
2.6 Serpulid bristle worm (Lophotrochozoa: Sipuncula: Sipunculidea: Sipunculidae;
Sipunculus nudus)
2.7 Planarian flatworm (Lophotrochozoa: Parenchymia: Platyhelminthes: Turbellaria:
Microstomidae; Microstomum lineare)
2.8 Lampert’s sea cucumber (Bilateria: Deuterostomia: Echinodermata: Holothuroidea:
Apoda: Synaptidae; Synaptula lamperti)
3 Convergent structure and function: OCELLAR PIT (photoreceptor cells located in an
open pit, connected to an optic nerve, for directional light detection)
3.1 Spiral tufted bryozoa larvae (Bilateria: Protostomia: Lophotrochozoa: Bryozoa:
Cheilostomata: Bugulidae; Bugula neritina)
3.2 Megathyrid lamp shell larvae (Lophotrochozoa: Brachiopoda: Terebratulida:
Megathyrididae; Argyrotheca cistellula)
3.3 Loxosomatid goblet worm larvae (Lophotrochozoa: Spiralia: Entoprocta:
Loxosomatidae; Loxosomella cochlear)
3.4 Common starfish (Bilateria: Deuterostomia: Echinodermata: Asteroidea:
Forcipulatida: Asteriidae; Asterias rubens)
3.5 Tornarian larvae of Balanoglossus proterogonius (Deuterostomia: Phryngotremata:
Hemichordata: Enteropneusta: Tornaria: Tornaria ancoratae)
4 Convergent structure and function: OCELLAR CUP (photoreceptor cells located in
partially enclosed cup-shaped structure, connected to an optic nerve, for increased precision
in directional light detection)
4.1 Bougainvillid jellyfish (Metazoa: Cnidaria: Hydrozoa: Anthomedusae:
Bougainvillidae; Bougainvillea principis)
Table 2.15
(continued)
4.2 Tiaropsid jellyfish (Cnidaria: Hydrozoa: Leptomedusae: Tiaropsidae; Tiaropsis

multicirrata)
4.3 Moon jellyfish (Cnidaria: Scyphozoa: Semaeostomae: Ulmaridae; Aurelia aurita)
4.4 Pectinariid trumpet worm (Metazoa: Bilateria: Protostomia: Lophotrochozoa:
Annelida: Polychaeta: Terebellomorpha: Pectinariidae; Pectinaria gouldii)
4.5 Opheliid bristle worm (Annelida: Polychaeta: Opheliidae; Armandia brevis)
4.6 Nephthyid bristle worm (Annelida: Polychaeta: Phyllodocemorpha: Nephthyidae;
Nephthys ciliata)
4.7 Lineid ribbon worm (Lophotrochozoa: Parenchymia: Nemertea: Lineidae; Lineus
ruber)
4.8 Echinoderid kinorhynch (Protostomia: Ecdysozoa: Introverta: Cephalorhyncha:
Kinorhyncha: Echinoderidae; Echinoderes aquilonius)
4.9 Salpid sea squirt (Bilateria: Deuterostomia: Chordata: Urochordata: Thaliacea:
Salpidae; Salpa cylindrica)
4.10 Florida lancet (Chordata: Myomerozoa: Cephalochordata: Branchiostomidae;
Branchiostoma floridae)
5 Convergent structure and function: SIMPLE EYE (globular eyes with pinhole openings,
closed openings, or closed openings with a simple lens for increased precision in directional
light detection)
5.1 Onitochiton (Protostomia: Lophotrochozoa: Mollusca: Polyplacophora: Chitonidae;
Onitochiton neglectus)
5.2 Queen scallop (Mollusca: Bivalvia: Pectenacea: Pectenidae; Pecten maximus)
5.3 Pterotracheid snail (Mollusca: Gastropoda: Caenogastropoda: Littorinimorpha:
Pterotracheidae; Pterotrachea mutica)
5.4 Common garden snail (Mollusca: Gastropoda: Pulmonata: Helicidae; Helix aspersa)
5.5 Paragordian hair worm (Protostomia: Ecdysozoa: Introverta: Nematozoa:
Nematomorpha: Gordiidae; Paragordius varius)
5.6 Peripatid velvet worm (Panarthropoda: Onychophora: Peripatidae; Peripatus
antiguensis)
5.7 Sagittid arrow worm (Protostomia: Chaetognatha: Aphragmophora: Sagittidae;
Sagitta enflata)
6 Convergent structure and function: COMPOUND EYE (complex eye with multiple
lenses for image resolving)
6.1 Sabellid bristle worm (Protostomia: Lophotrochozoa: Annelida: Polychaeta:
Sabellidae; Bispira volutacornis)
6.2 Sabellid bristle worm (Annelida: Polychaeta: Sabellidae; Branchiomma bombyx)
6.3 Sabellid bristle worm (Annelida: Polychaeta: Sabellidae; Megalomma vesiculosum)
6.4 Sabellid bristle worm (Annelida: Polychaeta: Sabellidae; Pseudopotamilla occelata)
6.5 Noah’s ark shell (Mollusca: Bivalvia: Arcoida: Arcidae; Arca noae)
6.6 Mydocopid ostracode (Protostomia: Ecdysozoa: Arthropoda: Mandibulata:
Crustacea: Ostracoda: Mydocopida: Azygocypridina lowryi)
6.7 Praying mantis (Arthropoda: Mandibulata: Hexapoda: Mantodea: Mantidae;
Mantis religiosa)
6.8 Eukrohnid arrow worm (Protostomia: Chaetognatha: Phragmophora: Eukrohniidae;
Eukrohnia hamata)
70 Chapter 2
Table 2.15
(continued)
7 Convergent structure and function: CAMERA EYE (complex eye with a single lens for
image resolving)
7.1 Carybdeid box jellyfish (Metazoa: Cnidaria: Cubozoa: Carybdeidae; Carybdea
marsupialis)
7.2 Alciopod annelid worm (Metazoa: Bilateria: Protostomia: Lophotrochozoa:
Annelida: Polychaeta: Alciopidae; Torrea candida)
7.3 Common octopus (Lophotrochozoa: Mollusca: Cephalopoda: Coleoidea:
Octopodidae: Octopus vulgaris)
7.4 Jumping spider (Protostomia: Ecdysozoa: Arthropoda: Cheliceriformes: Arachnida:
Salticidae; Portia fimbriata)
7.5 Ogre-faced spider (Arthropoda: Cheliceriformes: Arachnida: Deinopidae; Dinopsis
subrufus)
7.6 Dawn conodont (Bilateria: Deuterostomia: Chordata: Myomerozoa: Craniata:
Conodonta: Cordylodontidae; Eoconodontus notchpeakensis †Cambrian)
to the nonbilaterian cnidarians, where one might expect to find only

simple photoreceptor cells and eyespots—but that is not so: the simple
carybdeid box jellyfish have convergently evolved camera eyes (Martin
2002).
The independent evolution of complex camera eyes in the chordates
(humans and our kin) and in the molluscs (the octopus and their kin) is
a classic case of amazing convergent evolution, oft cited along with the
convergent evolution of wings in tetrapods. Yet we now know that camera
eyes have also been evolved by alciopid annelid worms (Wald and
Raypart 1977) and by two separate groups of spiders (Williams and
McIntyre 1980; Laughlin 1980), in addition to some box jellyfish. The
ogre-faced spiders are nocturnal predators, and have the largest single-
lens eyes described among all the arthropods (Laughlin 1980), a large-
eyed predatory convergence seen repeatedly in vertebrate nocturnal
carnivores (table 2.5). The telescopic eyes of the jumping spider Portia
fimbriata are convergent on those of raptorial birds, who use a similar
morphological trick to increase the focal length of their eyes: “a parallel
to the pit [of the eye] of P. fimbriata is the deep convexiclivate fovea
found in some vertebrates, notably falconiform birds. . . . [B]oth a group
of vertebrates and an invertebrate have therefore adopted the same
strategy to improve visual acuity despite a restricted cephalic space”
(Williams and McIntyre 1980, 580).
Phylogenetic analyses by Fitzhugh (1989) have revealed that com-
pound eyes have independently evolved four separate times in sabellid
bristle worms, not just once. Molecular analyses by Oakley and Cun-
ningham (2002) have revealed that the compound eyes of mydocopid
ostracodes have an origin independent from those in other arthropods.
Thus, Salvini-Plawen and Mayr’s (1977) upper estimate of 60 indepen-
dent origins of light-detection organs may ultimately prove to be closer
to the actual mark. Modern analyses do not always increase the number
of convergences discovered, however. Fitzhugh’s (1991) phylogenetic
analysis revealed that the pygidial eyes of the sabellid bristle worms
Fabriciola and Pseudofabriciola were a synapomophy, and not indepen-
dently derived.
It is interesting to note that the complex camera eye has been inde-
pendently evolved in deuterostome chordates, protostome molluscs and
arthropods, and nonbilaterian cnidarians (table 2.15), but that complex
compound eyes have appeared only in the protostomes. Compound eyes
have independently evolved eight different times within the three major
types of protostomes (lophotrochozoans, ecdysozoans, and chaeto-
gnaths), but not outside of the protostome clade, to my knowledge. Is
there a developmental constraint limiting compound eyes to the proto-
stome lineage? This question leads to the consideration of the develop-
ment of the eye itself.
Modern developmental studies have revealed that all animal eyes
contain highly conserved transcription factor Pax-6 gene homologs,
and the ubiquitous presence of the photopigment opsin proteins (Salvini-
Plawen 2008). This discovery has led to the proposal of two different
models for eye evolution: (1) prototype photoreceptor structures and
their developmental genes evolved only once, and further evolution
has elaborated them along independent lines, or (2) photoreceptor
structures evolved independently multiple times, each time co-opting
homologous genes for use in developing those structures (Oakley and
Cunningham 2002, 1429). The first model is, in essence, a model of paral-
lel evolution of the eye, and the second model is a model of convergent
evolution of the eye. Both models thus can be used to argue for func-
tional and developmental constraints in the evolution of the eye—eyes
function only if they are constructed in a limited number of ways, and
there are a limited number of genes available that can be used for their
development.
Both Oakley and Cunningham (2002) and Salvini-Plawen (2008) argue
for the second model. Oakley and Cunningham (2002) point out that the
first model requires a photoreceptor structure to be present in all
common ancestors, whereas their analyses indicate that compound eyes
72 Chapter 2
were not present in ostracode ancestors. Conway Morris (2003) has also
argued for the second model, and points out that in mammals alone the
Pax-6 gene is redeployed multiple times for other functions, as it is
associated with the development of the brain, nose, pituitary gland, pan-
creas, and the gut. The Pax-6 gene is, in essence, equivalent to the homol-
ogous tetrapod limb that has convergently evolved into wings for flying
in three separate groups of vertebrates (table 2.2), and into fins and
paddle-form swimming appendages in seven other groups of vertebrates
(table 2.1).
We have previously considered the convergent evolution of asymmet-
ric positioning and ultrasonic hearing in the ears of carnivores (table 2.6).
But what about the ear itself? As an organ for detecting pressure waves
in air, or for hearing, the ear is an adaptation of land animals. The ances-
tors of land animals, the fish, possess lateral line systems for detecting
pressure waves in water, which is much denser than air, and these lateral
line systems are still present in many aquatic amphibians, the tetrapod
descendants of fish. Experimentation has shown that lateral line systems
are extremely sensitive to minute differences in water pressure (Dijkgraaf
1967), and that fish and aquatic amphibians have well-developed direc-
tional “hearing” under water (Kuroki 1967).
Lateral line systems are not a unique synapomorphy of the verte-
brates, however. Both molluscs and arthropods have evolved species that
have convergently evolved lateral line systems (table 2.16). In the mol-
luscs, many cephalopods have evolved lines of ciliated tissue on their
heads and arms, and these epidermal lines have been shown to function
like the lateral line systems of fish, allowing the cephalopods to have
directional hearing underwater. In addition to the convergent evolution
of the complex camera eye (table 2.15), these epidermal lateral line
systems are “another example of convergent evolution between a sophis-
ticated cephalopod and vertebrate sensory system” (Budelmann and
Bleckmann 1988, 1). In the marine arthropods, some pelagic decapod
crustaceans have evolved very long antennae that are held parallel to
the long axis of the body of the animal on either side, and these structures
also function as lateral line systems for pressure-wave detection (Denton
and Gray 1985).
The thin gaseous mixture of the Earth’s atmosphere has very low pres-
sure compared to underwater habitats; consequently, land tetrapods have
evolved tympanal ear systems in order to hear in air, and have lost the
lateral line system. However, some tetrapods that have secondarily
returned to aquatic habitats have convergently reevolved a lateral line
Table 2.16
Convergent evolution of ears and sound-detection systems in animals
1 Convergent structure and function: LATERAL LINE SYSTEMS (lateral line row of
pores with nerve endings for detecting minute changes in water pressure, used for
directional hearing in aqueous habitats)
1.1 Shallow-water brief squid (Bilateria: Protostomia: Lophotrochozoa: Mollusca:
Cephalopoda: Coleoidea: Loliginidae; Lolliguncula brevis)
1.2 Penaeid shrimp (Protostomia: Ecdysozoa: Arthropoda: Mandibulata: Malacostraca:
Decapoda: Penaeidae; Acetes sibogae)
1.3 Astrapid jawless fish (Bilateria: Deuterostomia: Chordata: Vertebrata:
Pteraspidomorphi: Astrapida: Astrapidae; Astrapis desiderata †Ordovician)
1.4 Florida manatee (Vertebrata: Osteichthyes: Sarcopterygii: Reptiliomorpha:
Amniota: Synapsida: Therapsida: Mammalia: Eutheria: Afrotheria: Sirenia: Trichechidae;
Trichechus manatus)
2 Convergent structure and function: TYMPANAL EAR SYSTEMS (tympanum-based
hearing system for detecting minute changes in air pressure, used for directional hearing
in terrestrial habitats)
2.1 Field cricket (Bilateria: Protostomia: Ecdysozoa: Arthropoda: Mandibulata:
Hexapoda: Orthoptera: Gryllidae; Gryllus pennsylvanicus)
2.2 Cricket-parasite tachinid fly (Arthropoda: Mandibulata: Hexapoda: Diptera:
Tachinidae; Ormia ochracea)
2.3 Green lacewing (Arthropoda: Mandibulata: Hexapoda: Neuroptera: Chrysopidae;
Chrysopa carnea)
2.4 Grand western cicada (Arthropoda: Mandibulata: Hexapoda: Hemiptera: Cicadidae;
Tibicen dorsata)
2.5 Praying mantis (Arthropoda: Mandibulata: Hexapoda: Mantodea: Mantidae;
Mantis religiosa)
2.6 Tiger beetle (Arthropoda: Mandibulata: Hexapoda: Coleptera: Cincindela marutha)
2.7 Ornate tiger moth (Arthropoda: Mandibulata: Hexapoda: Lepidoptera: Noctuoidea:
Arctiidae; Apantesis ornata)
2.8 Mallow moth (Arthropoda: Mandibulata: Hexapoda: Lepidoptera: Geometroidea:
Geometridae; Larentia clavaria)
2.9 Red postman butterfly (Arthropoda: Mandibulata: Hexapoda: Lepidoptera:
Papilionoidea: Nymphalidae; Heliconius erato)
2.10 Butterfly moth (Arthropoda: Mandibulata: Hexapoda: Lepidoptera: Hedyloidea;
Hedylidae; Macrosoma nigrimacula)
2.11 Balanerpeton amphibian (Bilateria: Deuterostomia: Chordata: Osteichthyes:
Sarcopterygii: Tetrapoda: Batrachomorpha: Dendrerpetonidae; Balanerpeton woodi
†Carboniferous)
2.12 Eastern box turtle (Tetrapoda: Reptiliomorpha: Amniota: Sauropsida: Anapsida:
Testudines: Emydidae; Terrapene carolina)
2.13 Granite night lizard (Diapsida: Lepidosauromorpha: Squamata: Scleroglossa:
Autarchoglossa: Scincomorpha: Xantusidae; Xantusia henshawi)
2.14 Nile crocodile (Diapsida:Archosauromorpha: Crurotarsi: Crocodylia: Crocodylidae;
Crocodylus niloticus)
2.15 Duckbill platypus (Amniota: Synapsida: Therapsida: Mammalia: Monotremata:
2.16 Mouse-eared bat (Mammalia: Eutheria: Laurasiatheria: Chiroptera:
Microchiroptera: Vespertilionidae; Myotis myotis)
74 Chapter 2
system. In addition to losing their tetrapods and reevolving fins, the

manatees have acquired a body covering of sinus tactile hairs (which are
usually confined to the head region) that function as a lateral line system
underwater (Reep 2002).
Two major groups of animals have invaded the land from the water:
the panarthropods and the tetrapods. Among the tetrapods, many of the
early amphibians still possessed lateral line systems—evidence that they
continued to spend a great deal of time in the water. With the evolution
of the amniotes, tetrapod invasion of the land began in earnest. It was
long thought that tetrapod tympanal ear systems were inherited from
the first fully terrestrial amphibian ancestors, and thus that ears evolved
very early and only once in the tetrapod lineage. But this is not so. The
phylogenetic analyses of Clack (2002) have revealed that tympanal ear
systems evolved independently in at least five different tetrapod lineages:
once in the ancient batrachomorph amphibians and still present in their
lissamphibian descendants (frogs, salamanders, and kin), three additional
separate times in the sauropsid amniotes (ancestors of living turtles,
lizards, and archosaurs), and independently in the synapsid amniotes
(ancestors of mammals). It is now known that the mammalian tympanal
ear, with its characteristic inner-ear stapes, incus, and malleus bones,
evolved twice independently in the synapsid lineage: once in the mono-
tremes, and independently in the therians (table 2.16). This fact was
revealed by the discovery of the basal monotreme species Teinolophus
trusleri from the Early Cretaceous of Australia, in which the three inner-
ear bones had not yet separated from the mandible of the animal
(Rich et al. 2005). Therefore, the divergence of the australosphenids
and monotremes from the therian lineage occurred before the evolution
of the characteristic mammalian typanal ear, and the separation of the
three inner-ear bones from the mandible and the evolution the mam-
malian inner ear occurred independently in the monotreme and therian
lineages.
We can hear birds sing, and birds can hear humans sing, but our ears
do not have a common origin. As summarized by Clack (2002, 300): “The
evolution of hearing in air was a slow and complicated process; far from
being achieved and perfected in the earliest tetrapods, it was separately
invented many times by different groups—and a long time after tetra-
pods first gained the land.” Clack (2002, 300) surmises that a natural-
selective impetus for the evolution of tympanal ears may have been the
evolution of noisy, buzzing insects in the Carboniferous: “the timing of
the evolution of the characteristics of an ear capable of receiving air-
borne sound well certainly corresponds to the huge radiation of insects

that can be seen in the fossil record of the Late Permian.” Tetrapods that
could hear insects, as well as see them, would certainly be more successful
in hunting insects for food than a tetrapod that could not hear.
Insectivory-triggered ear evolution is not unique to the tetrapods.
Within the hexapods themselves, today there exist the cricket-parasite
tachinid flies, which have evolved tympanal ears that allow them specifi-
cally to hear the crickets that they hunt for food (Robert et al. 1992;
Robert and Willi 2000). Tympanal ears have independently evolved in
seven different orders of hexapods, and four times independently within
the lepidopteran order itself (Hoy and Robert 1996), for a total of at
least ten convergent originations (table 2.16). The hexapods show much
greater inventiveness in the evolution of their ears than tetrapods, as they
variously have placed the tympanum on their abdomens, their thoraxes,
their wing bases, and so on. The early evolution of ears in hexapods may
have had more to do with their social interactions—calling, courtship,
and territorial signals—than with predation, with subsequent ear evolu-
tion as antipredator adaptations, such as ultrasonic hearing in moths as
a means of evading bats (Hoy and Robert 1996).
In addition to sight and hearing, another major animal sensory capabil-
ity is that of smell. To my knowledge, very little attention has been given
to the phylogenetic distribution of different olfactory organs as a means
to determining how often these organs have been independently evolved.
Most mentions of convergence in olfactory organs are of the notably
unusual convergences, as in star-nosed moles (Conway Morris 2003).
Although possible convergences in the olfactory organs of hexapods and
vertebrates have been discussed, at present “a definitive answer is not
yet possible, but given the differences in the molecular genetics of olfac-
tory receptor coding, the evidence points strongly to convergence”
(Conway Morris 2003, 180).
Related to the chemosensory process of detecting molecules in air, or
smelling, is the process of breathing itself in the thin atmosphere of the
terrestrial realm. The two major groups of animals that have invaded the
land from the water have evolved two different organ systems to deal
with gas exchange. The panarthropods use a tracheal system of tubes to
carry oxygen from the outside air deep into their tissues, and to remove
carbon dioxide waste from those tissues and expel it to the outside envi-
ronment. The tetrapods use lungs instead, with oxygen absorbed across
the surface area of the lung tissue and diffused into the blood of the
circulatory system, where it is then transported to interior tissues. Carbon
76 Chapter 2
dioxide waste is transported in the reverse direction; it is expelled across

the surface area of the lung tissue, and exhaled to the outside environ-
ment. Thus, the circulatory system of tetrapods serves the dual function
of transporting both oxygen and nutrients to interior tissues and remov-
ing both carbon dioxide and chemical metabolic wastes from those
tissues. The tracheal system of the panarthropods is used only for gas
exchange; chemical metabolic wastes are removed by other excretory
systems.
As such, these two breathing systems would be an example of diver-
gent evolution in two separate groups of animals, and of no particular
interest in the consideration of convergent evolution. However, it appears
that the panarthropods have evolved the tracheal system of breathing
not just once but five separate times: in the Onychophora (land-dwelling
velvet worms), in Cheliceriformes (land-dwelling scorpions and arach-
nids), and three times in the Mandibulata—the land-dwelling myriapods,
malacostracan isopods, and the hexapods (Klok et al. 2002). Klok et al.
(2002) give further evidence that the ability to open and close the spir-
acles of the trachael system (to produce discontinuous gas exchange) has
independently evolved four separate times in separated species of tra-
cheated panarthropods. In all cases, the evolution of the tracheal system
of breathing is linked to the developmental invagination of the respira-
tory surfaces of the aquatic panarthropod forms (Pritchard et al. 1993).
Convergent evolution has also occurred in the lung-breathing tetra-
pods. The plesiomorphic condition is to have lungs like bellows and to
breathe in a biflow system, where oxygen is inhaled and carbon dioxide
is exhaled. In a biflow breathing system, the same pathway for air trans-
port is used both in inhaling and in exhaling. The birds, however, have a
much more efficient uniflow system of breathing, where air is first inhaled
into a system of interior air sacs, then moved from the air sacs to the
lungs, then exhaled from the lungs to the outside environment. The
pathway of air through the bird’s breathing system thus moves only in
one direction, in a looped pathway rather than in and out along the same
pathway, as with other tetrapods. This efficient uniflow system of breath-
ing and oxygen extraction allows birds to easily breathe in regions of
low-oxygen partial pressure, as is encountered high in the sky, whereas
humans at the same altitude may need the assistance of breathing masks
and oxygen cylinders.
It was long thought that the uniflow system of breathing was an auta-
pomorphy of the birds, and related to the evolution of flight in dinosaurs.
However, exceptionally well-preserved fossils of the plesiomorphic the-
ropod Majungatholus atopus reveals that the animal had the avian pul-
monary air-sac system, and that the avian style of breathing evolved very
early in the evolution of the theropod clade of dinosaurs (O’Connor and
Claessens 2005). Now, a more careful examination of other skeletons of
the ornithodires in general reveals that the pterosaurs also had a pulmo-
nary air-sac system and uniflow respiration (Claessens et al. 2009). In
essence, the pterosaurs evolved the uniflow system of breathing 70
million years before the birds did. As the pulmonary air-sac system is
demonstrably absent in dinosauromorphs and in the herrerasaurids (the
basal dinosaurs), the breathing system clearly evolved convergently
twice within the archosaurs—once in the pterosaurs, and once in the
theropod dinosaurs.
It is interesting that the panarthropods have never evolved lungs.
This may well be another example of developmental constraint on the
number of evolutionary possibilities open to a group of animals, and a
limit to convergent evolution. The tracheal system of breathing limits the
size of an animal, and only in the Carboniferous—when oxygen partial
pressure in the atmosphere was higher than at present—did very large
insects evolve, such as meganeurid dragonflies with wingspans of 75
centimeters.
The small size of insects has allowed these animals to converge on a
thermoregulatory system that is usually thought to be possessed only by
the birds and mammals—endothermy. Like all ectotherms, insects use a
variety of behavioral mechanisms to warm up, such as basking in the sun
or shivering. Due to the insects’ small size, the heat produced by their
muscular actions alone is large relative to the volume of their bodies,
and the insects have evolved a series of organ systems to retain their
heat once it is acquired: insulating air sacs around their thoracic muscles,
an insulating air sac between their thorax and their abdomen, a circula-
tory system with countercurrent heat exchange, and external body
insulation, like the scales of moths and the setae fuzz of bees (Heinrich
1993, 1996).
Large ectotherms, like amphibians and reptiles, have a much larger
volume of body tissue to warm up, and a much smaller body surface to
absorb heat by basking. The larger the animal, the more serious the
problem, because volume increases as a cubic function of dimension,
whereas surface area increases only as a square function. One way to
solve this problem would be to somehow increase the surface area of the
body, in order to absorb more heat while basking. At least six different
groups of tetrapods have increased their body surface areas by vastly
78 Chapter 2
elongating the neural spines of their vertebrae, thus creating a large “sail
fin” structure on their backs (table 2.17). Interestingly, both the derived
synapsid pelycosaurs and the ancient batrachomorph amphibians con-
vergently evolved this surface-area-increase structure in the Permian.
Sail fins were convergently evolved twice, at the same time, by the pely-
cosaurs, as the sail fins of the carnivorous sphenacodontids are structur-
ally very different from those of the herbivorous edaphosaurids (Bennett
1996). The fact that the large pelycosaurs evolved sail fins is taken as
evidence that these synapsids were still ectotherms, and that endothermy
had not yet evolved in the synapsid lineage.
Even more interesting, three separate groups of dinosaurs indepen-
dently evolved sail fins (table 2.17). As I shall discuss in a moment, the
dinosaurs have been argued to have been endotherms, so why would they
evolve sail fins like an ectotherm? If the dinosaurs were endotherms, it
may well be that the sail fin structure was primarily used to shed excess
heat and to cool the body down, rather than to absorb heat.
Two modern groups of highly active, energetic animals are known for
their endothermic metabolisms—the birds and the mammals. But when
did endothermy evolve in these two separate lineages? The first bird,
Archaeopteryx lithographica, dates from the Late Jurassic. However, the
discovery of fossil non-avian coelurosaurian theropods with feathers but
no wings indicates that feathers evolved first as body insulation to retain
heat, and that the coelurosaurs were endothermic. Is endothermy thus a
theropod trait? The fact that many small ornithopod ornithischians, such
as the hypsilophodonts, have fibrolamellar bone (an endothermic char-
acteristic) has been used to argue that they also were endothermic. Did
endothermy convergently arise twice in the dinosaurs, once in the the-
ropod saurischians and once in the ornithopod ornithischians, or is endo-
thermy a synapomorphy of the entire clade of dinosaurs? Bakker (1986)
has been the most vocal advocate of the latter position, though others
still dispute the uniform distribution of endothermy within all of the
dinosaurs (see the discussion in Benton 2005, 219–223).
The ornithodiran pterosaurs have also been argued to have been endo-
thermic (see Benton 2005, 226–228). As discussed above, the uniflow
system of breathing using lungs and pulmonary air sacs was indepen-
dently evolved by the pterosaurs and the birds (Claessens et al. 2009),
and the evidence is strong that endothermy also independently evolved
in the pterosaurs and the dinosaurs (table 2.17). Endothermy clearly
arose independently in the synapsid amniote lineage, far removed from
the sauropsid dinosaur and pterosaur lineages. Most date the evolution
Table 2.17
Convergent evolution of animal thermoregulation systems
1 Convergent structure and function: HEAT TRANSFER SURFACE (organic surface-

area structures for absorbing heat from the external environment or shedding excess heat
to the external environment)
1.1 Dissorophid amphibians (Tetrapoda: Batrachomorpha: Dissorophidae; Platyhystrix
rugosus †Permian)
1.2 Spinosaur (Tetrapoda: Reptiliomorpha: Amniota: Sauropsida: Diapsida:
Archosauromorpha: Ornithodira: Dinosauria: Saurischia: Theropoda: Carnosauria:
Spinosauridae; Spinosaurus maroccanus †Cretaceous)
1.3 Amargasaur (Dinosauria: Saurischia: Sauropodomorpha: Dicraeosauridae;
Amargasaurus cazaui †Cretaceous)
1.4 Ouranosaur (Dinosauria: Ornithischia: Cerapoda: Ornithischia: Iguanodontidae;
Ouranosaurus nigeriensis †Cretaceous)
1.5 Edaphosaurid synapsid (Amniota: Synapsida: Edaphosauridae; Edaphosaurus
cruciger †Permian)
1.6 Sphenocodontid synapsid (Synapsida: Sphenacodontidae; Dimetrodon grandis
†Permian)
2 Convergent system and function: ENDOTHERMY (active internal generation and
maintenance of body heat via metabolism)
2.1 Salmon shark (Vertebrata: Chondrichthyes: Elasmobranchii: Lamnidae; Lamna
ditropis)
2.2 Common thresher shark (Chondrichthyes: Elasmobranchii: Alopidae; Alopias
vulpinus)
2.3 Tuna (Vertebrata: Osteichthyes: Actinopterygii: Teleostei: Scombridae; Thunnus
alalunga)
2.4 Butterfly mackerel (Actinopterygii: Teleostei: Scombridae; Gasterochisma
melampus)
2.5 Swordfish (Actinopterygii: Teleostei: Xiphiidae; Xiphias gladius)
2.6 Pterosaur “hairy devil” (Osteichthyes: Sarcopterygii: Reptiliomorpha: Amniota:
Sauropsida: Diapsida: Archosauromorpha: Ornithodira: Pterosauria:
Rhamphorhynchidae; Sordes pilosus †Jurassic)
2.7 King penguin (Archosauromorpha: Ornithodira: Dinosauria: Saurischia: Theropoda:
Aves: Sphenisciformes: Spheniscidae; Aptenodytes patagonica)
2.8 Polar bear (Amniota: Synapsida: Therapsida: Mammalia: Eutheria: Laurasiatheria:
Carnivora: Ursidae; Thalarctos maritimus)
80 Chapter 2
of endothermic metabolisms in the synapsids to the evolution of the

mammalian clade itself; thus it is intriguing that endothermic metabo-
lisms arose three times, independently, in amniote tetrapods (pterosaurs,
dinosaurs, and mammals) at the same time—in the Late Triassic.
Yet endothermy is not, as we might conclude, an advanced, derived
condition of active terrestrial animals, because it has also convergently
arisen in the fishes. Block et al. (1993) have demonstrated that endother-
mic metabolisms have convergently evolved five separate times in the
fishes, twice in the cartilaginous fishes and three times in the bony fishes
(table 2.17). In the fishes, it has been argued that the evolution of endo-
thermy is not primarily a result of the adaptive benefits of energetic
lifestyles with sustained high activity levels, but rather the niche expan-
sion of these fish groups into cold-water regions of the oceans (Watson
1993; Block et al. 1993).
Last in our consideration of organ systems, it has been recently pro-
posed (Schierwater et al. 2009) that the nervous system itself is conver-
gent, and has independently evolved in two separate metazoan lineages
(table 2.18). In the current classification used in this book (see appendix),
the nervous system evolved in the cnidarians, and is thus a synapomor-
phy for subsequent metazoan evolution (table 2.18). In the cnidarians,
the nerve cells form a network without a central nervous system. In the
Table 2.18
The alternative classification of metazoans proposed by Schierwater et al. (2009)
Current classification Alternative classification
Metazoa Metazoa
– Placozoa – Diploblasta
– Demospongiae – – Placozoa
– Hexactinellida – – – Porifera
– Calcarea – – – – Calcarea
– Eumetazoa – – – – Demospongiae
– – Ctenophora – – – – Hexactinellida
– – Cnidaria* – – – Ctenophora
– – Bilateria – – – Cnidaria*
– – – Protostomia – Bilateria*
– – – Deuterostomia – – Protostomia
– – Deuterostomia
Note: The alternative classification of Schierwater et al. (2009) would require the indepen-
dent evolution of nervous systems in the cnidarians and the bilaterians (marked by
asterisks in the table).
bilaterians, a central nerve system is present, organized around a cephalic

ganglion from which the central nerve cord emerges; it is considered to
have evolved from the simpler cnidarian nervous system (Lecointre and
Le Guyader 2006).
Schierwater et al. (2009) have argued that two basal monophyletic
groups of metazoans exist: the Diploblasta and the Bilateria (table 2.18).
In the current classification, an evolutionary grade exists at the base of
the metazoan clade, running from the simple placozoans through to the
more complex cnidarians (table 2.18). All of these animals have a roughly
radial symmetry, and are sometimes referred to as the “Radiata,” a taxon
that has no formal status, as it is paraphyletic. In the alternative classifica-
tion, the placozoans are proposed to be the basal group of a monophy-
letic clade consisting of three phyla: the Porifera (which are paraphyletic
in the current classification, table 2.18), the Ctenophora, and the Cni-
daria. In this alternative classification, the cnidarians are not ancestral to
the bilaterians; therefore, the evolution of the network nervous system
of the cnidarians would be independent of the evolution of the central
nervous system of the bilaterians, as these two groups of animals occur
in two separate metazoan lineages (table 2.18).
Needless to the say, the argument of Schierwater et al. (2009) remains
controversial at present. The peculiar placozoans play a pivotal part in
the alternative classification (table 2.18). These organisms are known
from a single species, Trichoplax adhaerens, which was first discovered
in a saltwater aquarium in Austria and subsequently has been found in
marine waters worldwide. Lecointre and Le Guyader (2006) suggest that
placozoan morphological simplicity may be secondary, and that they are
derived from more complex eumetazoans, based upon preliminary
molecular studies. Schierwater et al. (2009, 37) argue that placozoan
morphological simplicity is primary, and that these organisms possess “a
living fossil genome” that shows that “complex genetic tool kits arise
before morphological complexity” (in that they were present prior to the
Diploblasta-Bilateria bifurcation in evolution) and that “these kits may
form similar morphological structures in parallel,” such as nervous
systems.
Reproduction
“Classic textbook examples of convergence, such as the evolution of

flight, and fusiform body shapes in aquatic vertebrates, pale beside those
relating to viviparity and matrotrophy” (Blackburn 1992, 320). This
82 Chapter 2
observation, surprising as it might appear at first glance, is true, as we

shall see in this final section of the chapter. Viviparous animals give live
birth, and matrotrophic females provide nutrients for the internally
developing embryo. Both of these conditions, complex though they may
be, have repeatedly and independently evolved in an astonishing number
of animals.
In the plesiomorphic condition for animals, the female produces the
egg, the male produces sperm, and the fertilization of the egg by sperm
takes place external to the female’s body. In marine invertebrates, the
egg may contain only enough nutrient for development to the larval
stage, after which the larva is on its own and must find its own food to
survive, a condition known as planktotrophy. Alternatively, the female
can develop a larger egg, one that contains enough yolk nutrient for
substantial larval development, a condition known as lecithotrophy. In
terms of bioeconomics, it costs the female more energy to produce leci-
thotrophic larvae than planktotrophic, but in terms of reproductive
success lecithotrophic larvae have a better chance of survival than
planktotrophic.
In external fertilization, the female fish lay a gelatinous mass of eggs
over which the male fish sprays a milky cloud of sperm. This is the plesio-
morphic reproductive condition for many oviparous, egg-laying fish and
some aquatic amphibians. A more certain form of fertilization, and one
that is less costly in terms of sperm numbers, is internal, in which the
sperm is introduced into the body of the female, fertilization of the eggs
occurs internally, and the female then lays the fertilized eggs. Simple
internal fertilization requires the alignment of the female and male
cloacal vents so that sperm may be transferred between them. Many
male fish, such as sharks, have evolved clasping appendages from the
pelvic fins which can be used to hold onto the female during mating,
maintaining contact between the male and female cloacae. Terrestrial
tetrapods are able to hold one another by their appendages, and males
have evolved the ability to somewhat protrude their cloaca so that it fits
a slight distance within the cloaca of the female. Some primitive saurop-
sids have added barbs to their cloaca, which can be used to hold the
male’s everted cloaca within the female’s during fertilization.
In the most efficient mode of internal fertilization, the male evolves
an inflatable copulatory organ that can deposit the sperm as close to the
eggs within the female’s body as possible, and that can be deflated when
it is not in use so that it is not a liability in locomotion: in other words,
a penis (Kelly 2002). The hydrostatic penis has been convergently evolved
Table 2.19
Convergent evolution of male internal fertilization organs
Convergent structure and function: HYDROSTATIC PENIS (inflatable copulatory organ

for passing sperm directly into the body of the female)
1 Golden apple snail (Bilateria: Protostomia: Lophotrochozoa: Mollusca: Gastropoda:
Caenogastropoda: Ampullariidae; Pomacea bridgesii)
2 Cayenne caecilian (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Tetrapoda: Batrachomorpha: Lissamphibia: Gymnophiona: Typhlonectidae;
Typhlonectes compressicauda)
3 North American tailed frog (Lissamphibia: Batrachia: Anura: Ascaphidae; Ascaphus
truei)
4 Snapping turtle (Tetrapoda: Reptiliomorpha: Amniota: Sauropsida: Anapsida:
Testudines: Chelydridae; Macroclemys temmincki)
5 Green anole (Sauropsida: Diapsida: Lepidosauromorpha: Squamata: Iguania:
Polychrotidae; Anolis carolinensis)
6 Nile crocodile (Archosauromorpha: Crurotarsi: Crocodylia: Crocodylidae; Crocodylus
niloticus)
7 Ostrich (Archosauromorpha: Ornithodira: Dinosauria: Saurischia: Theropoda:
Maniraptora: Aves: Paleognathae: Struthionidae; Struthio camelus)
8 Argentine lake duck (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves:
Neognathae: Anseriformes: Anatidae; Oxyura vittata)
9 Nine-banded armadillo (Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
Xenarthra: Dasypodidae; Dasypus novemcinctus)
at least six separate times by amniote animals: four separate times within
the sauropsids, and once within the synapsids (table 2.19). A single penis
is present in anapsid turtles and archosaurian crocodiles but, curiously,
two penises, or “hemipenes,” were evolved by the lepidosaurian lizards
and passed down to their snake descendants. Although the lizards and
snakes have two penises, they only use one at a time in intercourse.
Equally curious, the penis was lost in highly derived archosaurs like the
avian dinosaurs, which reverted back to the use of the cloaca-versus-
cloaca method of fertilization. Only a very few birds, such as the ostrich
and the duck, have secondarily reevolved a penis organ (King 1981; Kelly
2002).
In the nonamniote tetrapods, males of the caecilian amphibians have
an evertable rear part of the cloaca, or phallodeum, that they use to
transfer sperm into the cloaca of the female (Wake 1993). Among the
frogs, species members of only one genus, the tailed frog Ascaphus, have
evolved a similar penislike extension of the cloaca. Of the invertebrate
animals, the snails are unique in that both dioecious and hermaphroditic
species of gastropods have evolved a penis. Some other marine inverte-
brates, such as the male octopus, secondarily use one of their tentacles
84 Chapter 2
as a pseudopenis, or gonopod, to transfer sperm packets to the female.

Most curious, the female dwarf seahorse Hippocampus zosterae, a pecu-
liar bony fish, uses a gonopod-like organ to transfer eggs from her body
into the body of the male seahorse, where they are fertilized (Mason-
jones and Lewis 1996), in what could be argued to be the convergent
evolution of a pseudopenis by a female animal.
Internal fertilization is a precursor to the evolution of viviparity, in
which the female does not lay the eggs but rather gives live birth to the
offspring. As in fertilization modes, there is a spectrum of modes or
degrees of viviparity. The simplest is lecithotrophic viviparity, in which
the female produces eggs with large yolks but does not lay them; instead,
the embryo develops within the body of the female (this reproductive
mode is also sometimes called ovoviviparity). A very simple placenta is
present for gas exchange between the female and the eggs, as the devel-
oping embryos must respire. The opposite extreme is obligate placentot-
rophy, in which the female produces small eggs with little or no yolk and
the necessary nutrient is directly provided by the female across the pla-
centa to the developing embryos (Thompson and Speake 2006). The
spectrum between these two extrema is usually divided into four modes
or types of matrotrophy (Blackburn 1993): Type I (lecithotrophic vivipar-
ity), Type II, Type III, and Type IV (obligate placentotrophy).
As mammals, many of us have the impression that giving live birth is
a unique mammalian trait. This is not true—the convergent evolution of
viviparity is rampant in all kinds of animals, from very simple to very
complex (tables 2.20, 2.21, 2.22). It is often also thought that viviparity
is associated only with highly derived land animals, the amniotes who
have adapted to life in harsh terrestrial habitats, but this also is not true,
as many marine animals have evolved viviparity (table 2.20). Viviparity
is not even a vertebrate trait, for it has independently evolved in both
main branches of the invertebrate protostomes, from lophotrochozoan
marine bristle worms (Baskin and Golding 1970) to ecdysozoan velvet
worms, and in invertebrate deuterstomes as well, such as marine starfish
and sea cucumbers, where it has independently evolved at least four
separate times (Byrne 2005).
The evolution of viviparity is also ancient, as fossils of the Devonian
placoderm fish Materpiscis (mother fish) have revealed (Long et al.
2008). In the living sharks, descendants of the ancient placoderms, phy-
logenetic analyses have revealed that viviparity has independently
evolved in nine separate lineages (Dulvy and Reynolds 1997). In table
2.20, I have listed examples of basal or near-basal species for each inde-
pendent evolution of viviparity in sharks from the cladogram of Dulvy

and Reynolds (1997); subsequent evolution within some of these clades
has resulted in the presence of viviparity in 40 percent of all living shark
species. Fossils of extinct Carboniferous chondrichthyan fish also reveal
that viviparity has been sporadically evolved within the cartilaginous
fishes for over 300 million years (Dulvy and Reynolds 1997).
In living actinopterygian bony fishes, viviparity has independently
evolved in 14 families, or about 3 percent of living actinopterygian
species, still not nearly as prevalent as in the cartilaginous fishes (Wourms
and Lombardi 1992). It also has been convergently evolved in the sar-
copterygian bony fish, and is present in the living-fossil coelocanths
(Wourms and Lombardi 1992).
In the transition to terrestrial habitats, viviparity has convergently
evolved within both the protostome panarthropods and the deutero-
stome amphibians. Living insects exhibit sporadic incidences of the inde-
pendent evolution of viviparity, as it is present in ten separate groups
(Hagan 1951). All three groups of living nonamniote tetrapods, the
amphibians, have convergently made the evolutionary transition from
oviparity to viviparity (table 2.20): twice each in frog lineages and cae-
cilian lineages, but only once within the salamanders (Wake 1992, 1993).
In both groups, however, viviparous species are quite rare compared to
the large number of species that still lay eggs.
Of the two lineages of amniote tetrapods, the sauropsids and the syn-
apsids, the sauropsids exhibit an astonishing number of separate acquisi-
tions of viviparity, a trait normally thought of as characteristic of synapsid
mammals. Estimates of the independent evolution of viviparity in squa-
mate sauropsids have ranged from 98 times (Blackburn 1992) to 100
times (Blackburn 1993) to 105 times (Blackburn 2000). Rigorous phylo-
genetic analyses for all sauropsids are currently not available, but Lee
and Shine (1998) have conducted the most complete phylogenetic analy-
sis of the incidence of viviparity within the reptiles. Phylogenetic rela-
tionships within the speciose clades of scincid lizards, colubrid snakes,
and elapids snakes were too inadequate to be included in their study, but
phylogenetic analyses of the remaining reptiles (Lee and Shine 1998)
reveals the independent evolution of viviparity 39 separate times (table
2.21). In table 2.21, I have listed examples of basal or near-basal species
for each independent evolution of viviparity in sauropsids from the
cladograms of Lee and Shine (1998); future inclusion of the colubrid and
elapid snake and scincid lizard species will certainly increase the length
of this list. Evolution of viviparity in the sauropsids is also ancient, as
86 Chapter 2
Table 2.20
Convergent evolution of animal reproductive systems: viviparity in nonamniote animals
Convergent structure and function: VIVIPARITY (development of embryos within the

body of the female in order to produce more highly developed offspring at birth to enhance
their potential for survival)
1 Oligohaline bristle worm (Bilateria: Protostomia: Lophotrochozoa: Annelida:
Polychaeta: Nereidae; Nereis limnicola)
2 Peripatid velvet worm (Protostomia: Ecdysozoa: Panarthropoda: Onychophora:
Peripatidae; Peripatus antiguensis)
3 Pacific beetle-mimic cockroach (Panarthropoda: Arthropoda: Mandibulata:
Hexapoda: Blattodea: Blaberidae; Diploptera punctata)
4 Megathrips (Arthropoda: Mandibulata: Hexapoda: Thysanoptera: Phlaeothripidae;
Megathrips lativentris)
5 Body louse (Arthropoda: Mandibulata: Hexapoda: Anoplura: Pediculidae; Pediculus
corporis)
6 Viviparous caddisfly (Arthropoda: Mandibulata: Hexapoda: Ephemeroptera:
Ephemeridae; Notanatolica vivipara)
7 Brown soft-scale insect (Arthropoda: Mandibulata: Hexapoda: Homoptera: Coccidae;
Lecanium hesperidum)
8 Viviparous moth (Arthropoda: Mandibulata: Hexapoda: Lepidoptera: Tineidae;
Tinea vivipara)
9 Flesh fly (Arthropoda: Mandibulata: Hexapoda: Diptera: Sarcophagidae; Sarcophaga
carnaria)
10 Leaf beetle (Arthropoda: Mandibulata: Hexapoda: Coleoptera: Chrysomelidae;
Chrysomela varians)
11 Ichneumon fly (Arthropoda: Mandibulata: Hexapoda: Hymenoptera: Ichneumonidae;
Paniscus testaceus)
12 Viviparous earwig (Arthropoda: Mandibulata: Hexapoda: Dermaptera:
Hemimeridae; Hemimerus vosseleri)
13 Sea star (Bilateria: Deuterostomia: Echinodermata: Asteroidea: Asterinidae;
Cryptasterina hystera)
14 Deep-water feather star (Echinodermata: Crinoidea: Comasteridae; Comatilia
iridometriformis)
15 Sea cucumber (Echinodermata: Holothuroidea: Apodida: Synaptidae; Leptosynapta
clarki)
16 Small brittle star (Echinodermata: Ophiuroidea: Amphiuridae; Amphipholis
squamata)
17 Placoderm “mother fish” (Deuterostomia: Chordata: Vertebrata: Gnathostomata:
Placodermi: Ptyctodontida: Materpiscis attenboroughi †Devonian)
18 Carboniferous ratfish (Vertebrata: Chondrichthyes: Holocephali; Delphyodontus
dacriformes †Carboniferous)
19 Blunt-nose cow shark (Chondrichthyes: Elasmobranchii: Squalea: Hexanchiformes:
Hexanchidae; Hexanchus griseus)
20 Short-tailed nurse shark (Chondrichthyes: Elasmobranchii: Galeomorphii:
Orectolobiformes: Ginglymostomatidae; Pseudoginglymostoma brevicaudatum)
21 Cobbler wobbegong shark (Chondrichthyes: Elasmobranchii: Galeomorphii:
Orectolobiformes: Orectolobidae; Sutorectus tentaculatus)
22 Goblin shark (Chondrichthyes: Elasmobranchii: Galeomorphii: Lamniformes:
Mitsukurinidae; Mitsukurina owstoni)
Table 2.20
(continued)
23 African sawtail catshark (Chondrichthyes: Elasmobranchii: Galeomorphii:

Carcharhiniformes: Scyliorhinidae; Galeus polli)
24 Speckled catshark (Chondrichthyes: Elasmobranchii: Galeomorphii:
Carcharhiniformes: Scyliorhinidae; Halaelurus boesemani)
25 Cuban ribbontail catshark (Chondrichthyes: Elasmobranchii: Galeomorphii:
Carcharhiniformes: Proscylliidae; Eridacnis barbouri)
26 Slender smooth-hound shark (Chondrichthyes: Elasmobranchii: Galeomorphii:
Carcharhiniformes: Proscylliidae; Gollum attenuatus)
27 Barbeled hound shark (Chondrichthyes: Elasmobranchii: Galeomorphii:
Carcharhiniformes: Leptochariidae; Leptocharius smithii)
28 Reef ocean perch (Vertebrata: Osteichthyes: Actinoptergyii: Teleostei:
Scorpaeniformes: Sebastidae; Helicolenus percoides)
29 Large Baikal golomyanka (Actinoptergyii: Teleostei: Scorpaeniformes:
Comephoridae; Comophorus baicalensis)
30 Four-eyed fish (Actinoptergyii: Teleostei: Cyprinodontiformes: Anablepidae;
Anableps anableps)
31 Black sailfin goodeid (Actinoptergyii: Teleostei: Cyprinodontiformes: Goodeidae;
Giardinichthys viviparous)
32 One-sided livebearer (Actinoptergyii: Teleostei: Cyprinodontiformes: Jenynsidae;
Jenynsia lineata)
33 Live-bearing tooth-carp fish (Actinopterygii: Teleostei: Cyprinodontiformes:
Poeciliidae; see Table 2.21)
34 Mousey klipfish (Actinopterygii: Teleostei: Perciformes: Clinidae; Fucomimus mus)
35 Pile perch (Actinopterygii: Teleostei: Perciformes: Embiotocidae; Rhacochilus
vacca)
36 Redrump blenny (Actinopterygii: Teleostei: Perciformes: Labrisomidae; Xenomedea
rhodopyga)
37 Viviparous eelpout (Actinopterygii: Teleostei: Perciformes: Zoacidae; Zoarces
viviparous)
38 Viviparous brotula (Actinopterygii: Teleostei: Ophidiiformes: Brotulidae;
Dinematichthys ilucoeteoides)
39 Viviparous bythithid (Actinopterygii: Teleostei: Ophidiiformes: Bythitidae; Bythites
islandicus)
40 False cusk (Actinopterygii: Teleostei: Ophidiiformes: Parabrotulidae; Parabrotula
plagiophthalmus)
41 Viviparous halfbeak (Actinopterygii: Teleostei: Beloniformes: Hemiramphidae;
Zenarchopterus gilli)
42 Coelacanth (Osteichthyes: Sarcopterygii: Actinistia: Coelacanthidae; Latimeria
chalumnae)
43 Cayenne caecilian (Sarcopterygii: Tetrapoda: Batrachomorpha: Lissamphibia:
Gymnophiona: Typhlonectidae; Typhlonectes compressicauda)
44 Banded caecilian (Lissamphibia: Gymnophiona: Scolecomorphidae; Scolecomorphus
vittatus)
45 African live-bearing toad (Lissamphibia: Batrachia: Anura: Bufonidae;
Nectophynoides occidentalis)
46 Golden coqui (Lissamphibia: Batrachia: Anura: Leptodactylidae; Eleutherodactylus
jasperi)
47 Fire salamander (Lissamphibia: Batrachia: Urodela: Salamandridae; Salamandra
salamandra)
88 Chapter 2
Table 2.21
Convergent evolution of animal reproductive systems: viviparity in amniote animals
Convergent structure and function: VIVIPARITY (development of embryos within the

body of the female in order to produce more highly developed offspring at birth to enhance
their potential for survival)
1 Ichthyosaur (Amniota: Sauropsida: Diapsida: Ichthyosauria: Ichthyosauridae;
Icthyosaurus platyodon †Jurassic)
2 Plesiosaur (Sauropsida: Diapsida: Lepidosauromorpha: Sauropterygia: Pachypleuro-
sauridae; Keichousaurus hui †Triassic)
3 Black-headed dwarf chameleon (Lepidosauromorpha: Lepidosauria: Squamata:
Iguania: Chameleonidae; Brachypodion melanocephalum)
4 Side-striped chameleon (Lepidosauria: Squamata: Iguania: Chameleonidae;
Chamaeleo bitaeniatus)
5 Pygmy lizard (Squamata: Iguania: Agamidae; Cophotis ceylanica)
6 Arabian toad-headed lizard (Squamata: Iguania: Agamidae; Phrynocephalus
arabicus)
7 Helmeted basilisk (Squamata: Iguania: Agamidae; Corytophanes cristatus)
8 Patagonian lizard (Squamata: Iguania: Iguanidae; Phymaturus patagonicus)
9 Rock horned lizard (Squamata: Iguania: Iguanidae; Phrynosoma ditmarsi)
10 Chihuahua desert horned lizard (Squamata: Iguania: Iguanidae; Phrynosoma
orbiculare)
11 Santa Cruz Island spiny lizard (Squamata: Iguania: Iguanidae; Sceloporus angustus)
12 Trans-volcanic bunchgrass lizard (Squamata: Iguania: Iguanidae; Sceloporus
bicanthalis)
13 Crevice swift lizard (Squamata: Iguania: Iguanidae; Sceloporus torquatus)
14 Mexican emerald spiny lizard (Squamata: Iguania: Iguanidae; Sceloporus formosus)
15 Florida worm lizard (Squamata: Scleroglossa: Amphisbaenia: Amphisbaenidae;
Rhineura floridana)
16 Jewelled gecko (Squamata: Scleroglossa: Gekkonta: Gekkontidae; Naultinus
gemmeus)
17 Skinks (Squamata: Scleroglossa: Autarchoglossa: Scincomorpha: Scincidae; see
Table 2.22)
18 Cape grass lizard (Squamata: Scleroglossa: Autarchoglossa: Scincomorpha:
Cordylidae; Chamaesaura anguina)
19 Arizona night lizard (Squamata: Scleroglossa: Autarchoglossa: Scincomorpha:
Xantusiidae; Xantusia arizonae)
20 Common lizard (Squamata: Scleroglossa: Autarchoglossa: Scincomorpha: Lacertidae;
Lacerta vivipara)
21 Gobi racerunner lizard (Squamata: Scleroglossa: Autarchoglossa: Scincomorpha:
Lacertidae; Eremias przewalskii)
22 California apod lizard (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Anguidae: Anniellinae; Anniella pulchra)
23 Peloponnesa slow-worm lizard (Squamata: Scleroglossa: Autarchoglossa:
Anguimorpha: Anguidae: Anguinae; Anguis cephallonica)
24 Striped grass lizard (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Anguidae: Diploglossinae; Ophiodes striatus)
25 Imbricate alligator lizard (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Anguidae; Barisia imbricata)
26 Flathead knob-scaled lizard (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Xenosauridae; Xenosaurus platyceps)
27 Mosasaur (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha: Mosasauridae;
Platecarpus ictericus †Cretaceous)
Table 2.21
(continued)
28 Blind snake (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha: Serpentes:

Typhlopidae; Typhlops diardi)
29 False coral snake (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha: Serpentes:
Aniliidae; Anilius scytale)
30 Western sand boa (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Serpentes: Boidae: Erycinae; Eryx jaculus)
31 Banded dwarf boa (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Serpentes: Tropidophiidae; Tropidophis feicki)
32 Rasp-skinned water snake (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Serpentes: Acrochordidae; Acrochordus granulatus)
33 Brown snake (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha: Serpentes:
Colubridae; Storeria dekayi)
34 Northern death adder (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Serpentes: Elapidae; Acanthophis praelongus)
35 Massasauga rattlesnake (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Serpentes: Viperidae: Crotalinae; Sistrurus catenatus)
36 Green pit viper (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha: Serpentes:
Viperidae: Crotalinae; Trimerosurus albolabris)
37 Gabon viper (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha: Serpentes:
Viperidae: Viperinae; Bitis gabonica)
38 Common European adder (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Serpentes: Viperidae: Viperinae; Vipera berus)
39 Russell’s viper (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha: Serpentes:
Viperidae: Viperinae: Daboia russelli)
40 Mongolian multituberculate (Amniota: Synapsida: Therapsida: Mammalia:
Multituberculata: Djadochtatheriidae; Kryptobaatar daszevegi †Cretaceous)
41 Virginia opossum (Mammalia: Theria: Marsupialia: Didelphimorphia: Didelphidae;
Didelphis virginiana)
seen in the fossil record (Carter 2008) and in particular in the marine
reptiles that convergently returned to the marine realm from the ter-
restrial in the Mesozoic—the ichthyosaurs, plesiosaurs, and mosasaurs
(Lee and Shine 1998; Cheng et al. 2004; Carter 2008).
The numerous incidences of convergent evolution of viviparity within
the squamates are evidence of the adaptive value of this reproductive
mode, or the role of functional constraint in convergent evolution, but
evidence of developmental constraint may be present as well. Thompson
and Speake (2006) argue that the unique development of an intravitel-
line mesoderm and the formation of an isolated yolk sac makes it par-
ticularly easy for squamates to make the developmental shift to viviparity,
which they have done repeatedly (table 2.21). On the other hand, the
highly speciose clades of both the plesiomorphic turtles and highly
derived archosaurs have remained oviparous. Lee and Shine (1998)
suggest that this also may be due to developmental constraint, in that
90 Chapter 2
embryonic diapause is widespread in both turtle and archosaur species,

but not in squamates. In embryonic diapause, the egg will suspend devel-
opment if retained in the uterus of the female, and only resumes develop-
ment after it is laid. Thus, prolonged retention of the egg within the
female will not result in significant development of the embryo in utero,
a process that is a precursor to hatching the egg in utero and live birth.
In synapsid amniotes, it was long thought that lecithotrophic viviparity
was evolved once by the marsupials (table 2.21) and obligate placentot-
rophy once by the eutherians (table 2.22). Evidence now suggests that
the extinct multituberculate mammals independently evolved lecithotro-
phic viviparity as well (Kielan-Jaworowska 1979), bringing the total to
twice for the mammals as opposed to 39 times for the reptiles.
The evolution of extensive matrotrophy (Types III and IV) with
complex placentae and obligate placentotrophy is much rarer than the
evolution of viviparity (Thompson and Speake 2006). Obligate placen-
Table 2.22
Convergent evolution of animal reproductive systems: complex placentae
1 Convergent structure and function: COMPLEX PLACENTAE (extensive matrotrophy

for the production of highly developed offspring at birth to enhance their potential for
survival)
1.1 Dwarf livebearer fish (Osteichthyes: Actinopterygii: Teleostei: Cyprinodontiformes:
Poeciliidae; Heterandria formosa)
1.2 Live-bearing tooth-carp fish (Actinopterygii: Teleostei: Cyprinodontiformes:
Poeciliidae; Poeciliopsis paucimaculata)
Poeciliidae; Poeciliopsis prolifica)
Poeciliidae; Poeciliopsis turneri)
2 Convergent structure and function: OBLIGATE PLACENTOTROPHY (Type IV full
matrotrophy for the production of highly developed offspring at birth to enhance their
potential for survival)
2.1 Brazilian scincid lizard (Amniota: Sauropsida: Diapsida: Lepidosauromorpha:
Lepidosauria: Squamata: Scleroglossa: Autarchoglossa: Scincomorpha: Scincidae;
Mabuya heathi)
2.2 Italian three-toed skink (Squamata: Scleroglossa: Autarchoglossa: Scincomorpha:
Scincidae; Chalcides chalcides)
2.3 Australian grass skink (Squamata: Scleroglossa: Autarchoglossa: Scincomorpha:
Scincidae; Pseudemoia entrecasteauxii)
2.4 Western serpentiform skink (Squamata: Scleroglossa:Autarchoglossa: Scincomorpha:
Scincidae; Eumecia anchietae)
2.5 Placental mammal “dawn mother” (Amniota: Synapsida: Therapsida: Mammalia:
Theria: Eutheria; Eomaia scansoria †Cretaceous)
totrophy was long considered to be an autapomorphy and defining char-

acteristic of the eutherians, the placental mammals, the first of which
evolved 125 million years ago in the Early Cretaceous (Ji et al. 2002).
This view was overturned with the discovery of Type IV matrotrophy in
the Brazilian scincid lizard Mabuya heathi (Blackburn et al. 1984), and
subsequent discoveries of obligate placentotrophy in three other skink
species (Blackburn 1993, 2000; Flemming and Blackburn 2003). Thus,
both the synapsid and sauropsid lineages of amniotes have convergently
evolved obligate placentotrophy (table 2.22). The mammalian evolution-
ary pathway from oviparity (monotremes) to lecithotrophic viviparity
(marsupials) to obligate placentotrophy (eutherians) has been indepen-
dently traveled by the squamate reptiles.
Is complex placentation thus a unique trait of land animals? Perhaps
not. Reznick et al. (2002) have developed a “matrotrophy index” (MI)
for species comparisons, which they have used to argue that complex
placentation has convergently evolved four separate times in fish (table
2.22). They point out that species with lecithotrophic viviparity typically
have an MI of 0.6 to 0.7, whereas the fish species Poeciliopsis retropinna
has an MI of 117. Although the argument for the evolution of complex
placentation remains controversial (see Morell 2002), many of the vivip-
arous fish species remain to be analyzed using the methodology of
Reznick et al. (2002).
In concluding this section of the chapter on the convergent evolution
of reproductive traits in animals, I would like to mention the land plants.
Plants are very different organisms from animals; for one thing, they can
synthesize their own food. Nevertheless, as we shall see in the next
chapter, plants faced the same functional problems that animals faced in
their transition from the sea to the land, and thus their evolution is as
rife with convergence as we have seen in this chapter among the animals.
That convergence extends even to the plants’ mode of reproduction, for
the early free-spore-reproducing land plants were as dependent on a
source of water for their reproduction as were the amphibian animals,
and the evolution of the seed by land plants was the equivalent of the
evolution of the amniote egg by tetrapods, a key trait in their respective
successful invasion of terrestrial habitats (Niklas 1997).
3 Convergent Plants
Phenotypic correspondence among unrelated species provides strong circum-

stantial evidence for adaptive evolution because it shows that organisms differing
in genetic and developmental capabilities can converge on comparable solutions
to life’s exigencies when confronted with the same or very similar selection pres-
sures. Here we shall consider a much touted example of convergent evolution—
arborescence, the tree growth habit.
—Niklas (1997, 316)
Arborescence
The reverse of the old adage “not seeing the forest for the trees”—the
condition in which people overlook a larger overall pattern because they
are too narrowly focused on individual particulars—is “not seeing the
trees for the forest.” The trees in a forest do vary—an oak tree is differ-
ent from a maple tree—but the apparent similarity of tree form that can
be seen throughout thousands of trees in a forest may give one the
impression that all trees are variants of a single common ancestor. This
is not so. Nine separate groups of plants have independently and con-
vergently evolved the tree form (Niklas 1997; Donoghue 2005; Stein
et al. 2007).
Trees are an adaptation to life on land. Plant growth in a two-
dimensional plane soon leads to crowding and overgrowth, with one
plant shading out another plant in the competition for light from the sun,
the energy source for the survival of the plant. Just as in human cities,
when crowding occurs in the two-dimensional plane of the land surface,
the solution is to move into the third dimension above the land surface—
to construct towering skyscraper buildings or to evolve trees.
In order to grow into the third dimension of height, a major force must
be overcome—gravity. Tree forms all have a central support structure,
the trunk, that rises vertically from the land surface. At some distance
94 Chapter 3
above the ground, branches extend out from the tree trunk in order to
capture as much sunlight as possible for the tree’s survival and, in order
to ensure the survival of the species, to facilitate fertilization and disper-
sal of the tree’s offspring.
The first trees appeared in the Middle Devonian, over 390 million
years ago. Plants began their full-scale invasion of the terrestrial realm
in the Late Devonian and Carboniferous, and huge forests spread across
the landscapes of the Earth. The trees in those forests were very different
from those that constitute the forests of the Earth today. The famous
Gilboa fossil forest in New York State, from the Middle Devonian, gives
us a glimpse of the earliest forests on Earth. Numerous upright trunks
of trees are preserved as sandstone casts, each one broken off about a
meter above the ancient ground surface. The trunks range in size from
about a half meter to a meter in diameter with a wider, bulbous base
with numerous small anchoring roots, revealing that these earliest trees
had no taproot system for anchoring. The morphology of the upper por-
tions of the trees was completely unknown until the spectacular discov-
ery in 2007 of fossils of the complete tree, logs that had fallen over and
were preserved horizontally, in rock outcrops only 13 kilometers east of
the Gilboa forest (Stein et al. 2007). Some of these trees were over 8
meters tall and, most surprisingly, had no laminar leaves or horizontal
branches. Instead, they were topped with a crown resembling a spiky
pipe brush or shaving brush, consisting of numerous digitately branched
offshoots held at an acute angle to the trunk stem. The complete tree
Wattieza (Eospermatopteris) erianus belongs to an extinct group, the
Cladoxylopsida (table 3.1), that is related to modern surviving horsetail
rushes (see the appendix). Stein et al. (2007, 906) note that the cladoxy-
lopsid tree’s “body plan now stands unequivocally as the oldest known
arborescent terrestrial plant form” and that it is “therefore interesting to
see how instantly recognizable and, in a significant sense, ‘modern’ the
tree-like architecture of Wattieza seems to be.”
Both the more plesiomorphic lycophyte club mosses and the
cladoxylopsid-related equisetophyte horsetail rushes independently
evolved towering trees (table 3.1), although the tallest members of these
groups today is around 10 centimeters. In the lycophytes, the ancient
lepidodendrid scale trees constructed their trunks with two outer layers
of cortex beneath an outermost layer of leaf cushions, whereas in the
equisetophytes, the ancient calamite horsetail trees had trunks supported
by internal wedges of wood (Niklas 1997).
Convergent Plants 95
Table 3.1
Convergent evolution of arborescent plant morphologies
Convergent structure and function: TREE TRUNK (vertical support structure to resist
gravity for the elevation and spatial dispersal of photosynthetic and reproductive
substructures)
1 Lepidodendrid scale tree (Tracheophyta: Lycophyta: Lepidodendracea;
Lepidodendron rhodumnense †Carboniferous)
2 Gilboa-forest cladoxylopsid tree (Tracheophyta: Euphyllophyta: Moniliformopses:
Cladoxylopsida: Pseudosporochnaceae; Wattieza (Eospermatopteris) erianus †Middle
Devonian)
3 Calamite horsetail tree (Euphyllophyta: Moniliformopses: Equisetophyta:
Calamitaceae; Calamites cistiiformes †Carboniferous)
4 Psaronis-type tree ferns (Euphyllophyta: Moniliformopses: Filicophyta: Marattiaceae;
Psaronius schopfii †Carboniferous)
5 Tempskya-type tree ferns (Euphyllophyta: Moniliformopses: Filicophyta:
Tempskyaceae; Tempskya dernbachii †Cretaceous)
6 Australian tree fern (Euphyllophyta: Moniliformopses: Filicophyta: Cyatheaceae;
Cyathea cooperi)
7 Archaeopterid-type woody trees (Euphyllophyta: Lignophyta: Archaeopteridales:
Archaeopteridaceae; Archaeopteris hibernica †Late Devonian)
8 Cycad palm (Euphyllophyta: Lignophyta: Spermatophyta: Cycadophyta: Cycadaceae;
Cycas media)
9 Date palm (Euphyllophyta: Lignophyta: Spermatophyta: Angiospermae:
Euangiosperms: Monocotyledons: Aracaceae; Phoenix dactylifera)
Three separate groups of the ferns have convergently evolved the tree
form through time: the psaronis-type tree ferns in the Carboniferous, the
tempskya-type tree ferns in the Cretaceous, and the tree ferns that
survive today (table 3.1). Each group invented a different type of trunk
in their evolution: the Carboniferous tree ferns had trunks supported by
an outer mantle of adventitious roots, the Cretaceous tree ferns had
trunks consisting of interweaved stems bound together by adventitious
roots, and living tree ferns have a lower columnar base of adventitious
roots and an upper trunk supported by an outer layer of cortex and
external layer of leaf bases (Niklas 1997).
The first of the lignophyte trees that are so abundant today appeared
in the Devonian, particularly the archaeopterid-type woody trees of the
Late Devonian, which still reproduced with spores but had trunks sup-
ported by innermost concentric layers of heartwood (Niklas 1997). These
trees gave rise to the first seed plants in the Late Devonian, a reproduc-
tive event in plant evolution equivalent to the evolution of the amniote
egg in the tetrapods that we considered in the previous chapter, thus a
96 Chapter 3
vital event in the plant conquest of the terrestrial realm. The spermato-
phyte seed plants are incredibly diverse. Although they descend from the
original lignophyte trees, at least two groups of spermatophytes have lost
and secondarily reevolved the tree form: the cycad palms and the date
palms (table 3.1). The cycad palms have trunks that are supported by an
outermost layer of persistent leaf bases, whereas the date palms have
trunks supported by a fibrous complex of primary vascular bundles,
increasing in density of occurrence from the center to the outer margin
of the trunk (Niklas 1997), very different from the heartwood trunks of
most lignophyte trees today.
Niklas (1997, 316) argues strongly for the role of functional constraint
in the evolution of the tree form: “convergence among so many different
plant lineages for many biomechanical traits rather than only one or a
few important ones strongly suggests that these traits are the result of
adaptive evolution and not fortuitous expressions of diffusive evolution.”
Would not these same functional constraints produce the same evolu-
tionary result in alien worlds, as they have nine independent times here
on Earth? If sessile multicellular photoautotrophic organisms evolve on
distant Earth-like worlds, and if those organisms emerge from the seas
to invade the land, then they should necessarily—and convergently—
evolve the geometry of the tree form. Or, in the spirit of theoretical
morphology, can we conceive of an alternative geometry that these
organisms might evolve to solve the problems of overcrowding, gravity,
and maximizing surface area for photosynthesis? We shall return to this
question in the next chapter, when we consider the possible evolution of
balloon-like plants, plants that could float in the air, and why such hypo-
thetical plants do not exist on the Earth.
Niklas (1997) points out that both the ancient lepidodendrid scale
trees and the calamite horsetail trees not only convergently evolved tree-
trunk support structures, but also independently evolved traits such as
wood tissue, leaves, cones for sporangia, and heterosporous reproduc-
tion. These ancient trees depended upon a readily available and depend-
able supply of water for their free-spore mode of reproduction, and thus
were confined to swampy regions of the terrestrial realm—a develop-
mentally imposed constraint similar to that seen in early tetrapod invad-
ers of the terrestrial realm, the amphibians (as discussed in the previous
chapter). And just as the tetrapods escaped that constraint in the evolu-
tion of the amniotes, the plants escaped it in the evolution of the seed
plants, an event in reproductive evolution that we shall consider in detail
later in this chapter.
Because land plants are surrounded by air, not water, they require a
system to transport water from the soil below to the plant tissues above,
where it is needed. The simplest of the embryophytes, land plants such
as liverworts, have no roots and no vascular system. The derived and
highly successful vascular land plants are classified as tracheophytes, as
they possess tracheids, specialized cells for the transport of water up the
stem of the plant against the force of gravity (Lecointre and Le Guyader
2006). Advanced embryophytes, such as Cooksonia hemisphaerica from
the Late Silurian and Aglaophyton major from the Devonian, are now
considered to be transitional to the earliest true tracheophytes (Niklas
1997). These horneophytes had no roots or leaves, but did have small
stems that stood up against the force of gravity and bifurcated to produce
multiple tips that contained the sporangia. The tracheids of the earliest
tracheophytes were arranged in a series of circular rings, stacked one
atop another up the stem of the plant, or in a continuous climbing
helix. The walls of the tracheids were lignified, and thus also provided
some support to hold the stem of the plant erect against the force of
gravity. The question is still open as to whether tracheids are a synapo-
morphy, arising only once in plant evolution, and whether the Tra-
cheophyta are monophyletic. A possible alternative is that tracheids are
convergent, independently evolved in the lycophytes and the euphyllo-
phytes (Niklas 1997).
The next evolutionary step beyond tracheids for water transport within
the plant is the evolution of vessels for rapid and higher-volume water
transport. The evolution of vessels is clearly convergent: at least nine
separate phylogenetic lineages of tracheophytes have independently
evolved vessels (table 3.2). The angiosperms alone have convergently
evolved vessels several times (Niklas 1997); for example, the basal angio-
sperm nymphaealeans have independently evolved vessels at least twice
(Schneider and Carlquist 1995). Rates of water and food transport within
a plant are size limiting; with the convergent evolution of vessels, plants
began to achieve large sizes, and with the convergent evolution of trunks
(table 3.1) the invasion of land was in full force.
Photoautotrophs: Light Capture
Plants are photoautotrophs: they produce their own food via photosyn-
thesis. Early land plants, such as Cooksonia hemisphaerica some 425
million years ago, had small green stems and no leaves. A stem is a small
cylinder, with a fairly large surface area where photosynthesis can take
98 Chapter 3
Table 3.2
Convergent evolution of water-conducting structures in plants
Convergent structure and function: VESSELS (water-conducting cellular structures for

rapid water transport within the plant)
1 Resurrection plant (Tracheophyta: Lycophyta: Selaginellaceae; Selaginella
lepidophylla)
2 Woodland horsetail (Tracheophyta: Euphyllophyta: Moniliformopses: Equisetopyta:
Equisetaceae; Equisetum sylvaticum)
3 Bracken fern (Euphyllophyta: Moniliformopses: Filicophyta: Dennstaedtiaceae;
Pteridium aquilinum)
4 Gigantopterid seed fern (Euphyllophyta: Lignophyta: Spermatophyta:
Gigantopteridales: Gigantopteridaceae; Gigantopteris nicotianaefolia †Permian)
5 Arizona ephedra (Euphyllophyta: Lignophyta: Spermatophyta: Gnetophyta:
Ephedraceae; Ephedra fasciculata)
6 Malaysian “water lily” (Euphyllophyta: Lignophyta: Spermatophyta: Angiospermae:
Nymphaeales: Nymphaeaceae; Barclaya kunstleri)
7 Giant “water lily” (Euphyllophyta: Lignophyta: Spermatophyta: Angiospermae:
Nymphaeales: Nymphaeaceae; Victoria amazonica)
8 Pumpkin (Euphyllophyta: Lignophyta: Spermatophyta: Angiospermae:
Euangiosperms: Eudicots: Rosidae: Cucurbitaceae; Cucurbita maxima)
place, and a small volume of internal tissue in need of food. The key word
in the last sentence is “small”: the larger the cylinder, the smaller its
surface area relative to its volume, as surface area increases as a square
function of linear dimension whereas volume increases as a cubic func-
tion. Thus, the larger the stem of the plant, the smaller its surface area,
where photosynthesis occurs, relative to its volume of internal tissue,
where food is needed.
The solution to this classic area-volume problem is to somehow
increase the surface area of the plant that is capable of photosynthesizing
food—that is, to evolve leaves. The evolution of the leaf has occurred
independently in at least nine different phylogenetic lineages of land
plants (table 3.3). Interestingly, eight of these lineages convergently
developed leaves in the Devonian, from the simple liverworts to the first
of the derived lignophyte trees. Thus, the many leaves we see around us
each day, from ferns to maple tree leaves, can be traced back in time to
that explosion of convergent evolution in the Devonian.
Botanists divide leaves into two types: microphylls, which are small
with only one vein and no leaf gaps, and macrophylls, which are large
with complex venation and leaf gaps (Tomescu 2008). Not surprisingly,
microphylls were convergently evolved by Devonian liverworts, mosses,
Table 3.3
Convergent evolution of leaves in plants
Convergent structure and function: LEAVES (expanded-surface-area phototrophic

structures for light capture and photosynthesis)
1 Devonian liverwort leaf (Embryophyta: Marchantiophyta: Pallavincinaceae;
Pallavicinites devonicus †Late Devonian)
2 Devonian moss leaf (Embryophyta: Stomatophyta: Hemitracheophyta: Bryophyta
sensu stricto: Sporogitales; Sporogonites exuberans †Early Devonian)
3 Devonian lycopod leaf (Hemitracheophyta: Polysporangiophyta: Tracheophyta:
Lycophyta: Protolepidodendraceae; Leclercqia andrewsii †Middle Devonian)
4 Devonian horsetail leaf (Tracheophyta: Euphyllophyta: Moniliformopses:
Equisetophyta: Sphenopsidaceae; Sphenophyllum involutum †Late Devonian)
5 Whisk fern leaf (Euphyllophyta: Moniliformopses: Filicophyta: Psilotaceae; Psilotum
nudum)
6 Devonian fern leaf (Euphyllophyta: Moniliformopses: Filicophyta: Zygopteridaceae;
Ellesmeris sphenopteroides †Late Devonian)
7 Devonian aneurophyte leaf (Euphyllophyta: Lignophyta: Aneurophytales:
Pseudosporochnaceae; Rellimia thomsonii †Middle Devonian)
8 Devonian archaeopterid leaf (Euphyllophyta: Lignophyta: Archaeopteridales:
Archaeopteridaceae; Archaeopteris hibernica †Late Devonian)
9 Carboniferous seed-fern leaf (Euphyllophyta: Lignophyta: Spermatophyta:
Lyginopteridales: Lyginopteridaceae; Telangium bifidum †Early Carboniferous)
and lycopods (Gensel and Kasper 2005); that is, in land plants that are
small and thus less in need of increased surface area for photosynthesis.
The analyses of Boyce and Knoll (2002) demonstrate that macrophylls
evolved convergently at least four separate times in the equisetophytes,
filicophytes, “progymnosperms” (nonspermatophyte lignophytes), and
spermatophytes. In table 3.3 I have listed a leaf species that is one of the
most ancient for each of these groups, taken from the analyses of Boyce
and Knoll (2002). Tomescu (2008) agrees that macrophylls originated
independently a minimum of four times, but presents phylogenetic analy-
ses that suggest the actual number of convergences may be nine. He
demonstrates that leaves were independently developed at least twice in
the ferns (separately in the Psilotaceae and the Zygopteridaceae), and
perhaps as many as five times. Likewise, he argues that leaves arose
independently in both the aneurophyte and archaeopterid nonspermato-
phyte lignophytes (table 3.3).
Before leaving this section on photosynthetic light-capture structures
in plants, I would like to touch briefly on the convergent evolution of a
different type of light-capture structure: eyes. Eyes are usually con-
sidered to be unique to animals, and we examined their convergent
100 Chapter 3
evolution in detail in the previous chapter (see table 2.15). However, not
only have eyes evolved in the unikontan branch of the eukaryote phy-
logenetic tree—the branch leading to animals—but they have conver-
gently evolved in the bikontan branch as well, the branch that leads to
the plants (see the appendix). And not just once—simple eyespots have
independently evolved in a wide variety of phototactic algae and algal
phytoflagellates (table 3.4). Eyespots have independently evolved no less
than 11 separate times in groups as diverse as green algae, brown algae,
yellow-green algae, and many algal flagellates (Foster and Smyth 1980;
Kivic and Walne 1983). Kivic and Walne (1983) document the convergent
evolution of different types of eyespots in the algae (table 3.4), and
Foster and Smyth (1980, 619) also demonstrate that the phototaxis
Table 3.4
Convergent evolution of eyes in algae and algal phytoflagellates
Convergent structure and function: EYESPOTS (photoreceptor cells aggregated into

distinct spot regions for light-detection abilities)
1 EYESPOT TYPE A:
1.1 Chroomonas cryptophyte (Eukarya: Bikonta: Chromoalveolata: Cryptophyta;
Chroomonas salina)
1.2 Perdinium dinoflagellate (Eukarya: Bikonta: Chromoalveolata: Dinophyta;
Peridinium westii)
1.3 Chlamydomonas green alga (Eukarya: Bikonta: Green eukaryotes: Chlorobionta:
Ulvophyta: Chlorophyceae; Chlamydomonas reinhardtii)
1.4 Platymonas flagellated alga (Eukarya: Bikonta: Green eukaryotes: Chlorobionta:
Prasinophyta; Platymonas subcordiformis)
2 EYESPOT TYPE B:
2.1 Pavlova coccolithophorid (Eukarya: Bikonta: Chromoalveolata: Haptophyta;
Pavlova helicata)
2.2 Ochromonas golden alga (Eukarya: Bikonta: Chromoalveolata: Stramenopiles:
Chrysophyceae; Ochromonas villosa)
2.3 Fucus brown alga (Eukarya: Bikonta: Chromoalveolata: Stramenopiles:
Phaeophyceae; Fucus serratus)
2.4 Heterococcus yellow-green alga (Eukarya: Bikonta: Chromoalveolata:
Stramenopiles; Xanthophyceae; Heterococcus pleurococcoides)
3 EYESPOT TYPE C:
3.1 Euglena euglenophyte (Eukarya: Bikonta: Excavobionta: Euglenobionta; Euglena
gracilis)
3.2 Wolozynskia dinoflagellate (Eukarya: Bikonta: Chromoalveolata: Dinophyta;
Wolozynskia coronata)
3.3 Polydriella eustigmatophyte alga (Eukarya: Bikonta: Chromoalveolata:
Stramenopiles: Eustigmatophyceae; Polydriella helvetica)

control system for the motile species often has “dramatically different
machinery” and has been independently derived in each group.
While present in some of the most plesiomorphic Chlorobionta, like
the green algae, eyespots are lost in higher plants, particularly sessile land
plants. Land plants have evolved a series of phototropic mechanisms
instead, as seen in plant species that slowly turn in the light, tracking the
path of the sun across the sky, or opening and closing their flowers in
response to the rising and setting of the sun. Charles Darwin was fasci-
nated with heliotropic movement in plants, and discovered in his experi-
ments that his plants stopped tracking the sun during the day if a filter
removing blue wavelength light was placed between them and the sun.
We now know that the genes encoding for heliotropic movement in
plants have been convergently evolved in animals, where they are used
in maintaining circadian rhythms (this example of molecular conver-
gence will be considered further in chapter 5).
Photoautotrophs: Carbon Processing
Life has evolved a variety of autotrophic pathways: chemoautotrophic

eubacteria obtain energy for organic synthesis from oxidizing methane,
ammonia, nitrites, sulfites, or other inorganic mineral compounds present
on the Earth. Photoautotrophic eubacteria obtain the energy they need
from a source not on the Earth, namely by capturing photons from the
sun. The oxygenic type of photosynthesis was evolved by the cyanobac-
teria over 3.5 billion years ago, as witnessed by the presence of stromato-
lite bacterial mounds in the Earth’s fossil record; the subsequent
production of free oxygen by these bacteria has transformed the very
atmosphere of the planet. In the Cenozoic, some 3.47 billion years later,
oxygen levels in the atmosphere have risen to such high levels, and
carbon dioxide content fallen to such low levels, that they actually inter-
fere with the photosynthetic process the cyanobacteria evolved such a
long time ago.
The chloroplasts in the cells of all photosynthetic eukaryotes, and
hence all plants, are symbiotic cyanobacteria. Carbon fixation for the
production of hydrocarbon molecules, or food, by plants is catalyzed
by Rubisco (ribulose-1,5-bisphosphate carboxylase/oxygenase) in an
ancient process known as C3 photosynthesis. In the normal C3 photosyn-
thetic pathway, carbon is fixed directly by Rubisco carboxylation in
mesophyll-tissue chloroplasts. However, if carbon dioxide partial pres-
sures in the atmosphere are low and oxygen partial pressures are high,
102 Chapter 3
the presence of oxygen can interfere with Rubisco carboxylation and

produces oxygenation instead; fixed carbon is then wastefully lost in
photorespiration.
About 30 to 50 million years ago this threshold was reached in the
evolution of the Earth’s atmosphere. This event was partially due to the
activity of land plants themselves, as they have been extracting carbon
dioxide from the atmosphere and releasing free oxygen back into the
atmosphere for the past 425 million years, ever since the rapid spread of
land plants in the Late Silurian. It was also partially due to the formation
of the gigantic Himalayan mountain chain in the Cenozoic, produced by
the plate-tectonic-driven collision of the Indian subcontinent with Asia,
and subsequent weathering processes that have removed a substantial
amount of carbon dioxide from the Earth’s atmosphere. In any event,
plants began to experience difficulty in fixing carbon using the C3 pho-
tosynthetic pathway.
To deal with the problem of Rubisco oxygenation, plants have evolved
a more efficient C4 photosynthetic pathway, one that builds upon the
original C3 pathway but adds an additional carbon dioxide–concentrating
pathway. In the C4 photosynthetic pathway, carbon is primarily fixed by
PEPC (phosphoenol pyruvate carboxylase) carboxylation in the meso-
phyll tissue, and then four carbon acids are shuttled into bundle-sheath
tissue regions of the leaf where Rubisco and the chloroplasts have been
concentrated (Sage 1999, 2001a). This more efficient C4 photosynthetic
pathway of carbon dioxide concentration can overcome photorespira-
tion that would normally be triggered by high oxygen levels in the
atmosphere.
Evolution of the C4 photosynthetic pathway is not trivial. Land plants
were developmentally constrained by the original ancient Rubisco-based
biochemistry and cellular systems evolved over 425 million years ago
(Sage 1999). Thus, it might appear that only a few types of plants (or
perhaps only one) could have made the biochemical and cellular changes
necessary to evolve a new photosynthetic pathway. But this is not the
case: the C4 photosynthetic pathway has been independently evolved,
over and over, by different plant groups in the past 30 million years (table
3.5). The phylogenetic analyses of Kellogg (1999) have revealed that the
C4 photosynthetic pathway was independently evolved by 18 separate
angiosperm families, and that it has convergently arisen at least four
independent times in the grass family Poaceae alone (Sinha and Kellogg
1996). In table 3.5 I have listed basal or near basal species for each clade
of plants that have independently evolved C4 photosynthesis from the
cladograms of Kellogg (1999) and Sinha and Kellogg (1996). At present,

the known number of convergences of C4 photosynthesis stands at 33,
but future analyses of previously unstudied angiosperm groups are
almost certainly going to lengthen the list of convergences in the future
(Sage 2001a, 2001b).
Is C4 photosynthesis developmentally constrained to the flowering
plants? Sage (2001b) notes that it is surprising that C4 photosynthesis has
not been evolved by the nonangiosperm tracheophytes, particularly the
ferns, and suggests that only the flowing plants have the developmental
flexibility to evolve the biochemical and cellular changes necessary for
C4 photosynthesis. In addition, some geochemical models suggest that
atmospheric oxygen and carbon dioxide levels may have approached
today’s levels back in the Carboniferous, largely due to the effects of
the explosive spread of the Earth’s first forests. The fact that tracheal-
breathing size-limited hexapods, such as the dragonflies, could evolve
species with 75-centimeter wingspans supports the model predictions of
higher oxygen levels in the Carboniferous. Yet no evidence exists that
any of the Carboniferous plants groups (all nonangiosperm, for the
angiosperms evolved much later in the Cretaceous) evolved the C4 pho-
tosynthetic pathway (Sage 1999).
Although at present it is known that only about 3 percent of angio-
sperm species are C4 photosynthesizers, that 3 percent accounts for
about 30 percent of total terrestrial primary productivity on a global
scale, and up to as much as 80 percent of the primary productivity in
certain ecosystems, such as tropical grasslands (Sage 2001b). Following
the expectations of the theory of natural selection, one would predict the
further convergent evolution of C4 photosynthesis in new angiosperm
groups, and continued ecological expansion of existing C4 photosynthe-
sizers (table 3.5) into the future. However, Sage (2001a) has noted that
human burning of fossil fuels in the past several hundred years has led
to progressively increasing levels of carbon dioxide in the Earth’s atmo-
sphere, a process that favors the original C3 photosynthesizers, while
other human activities, such as slash-and-burn agriculture and deforesta-
tion in the tropics, has favored the spread of C4 grasslands and weeds.
Thus, predicting the future direction of plant evolution appears to be
more a function of predicting the future activity of humans than that
of plants.
In the past two sections of the chapter we have considered some of
the functional problems associated with a photoautrophic mode of life,
and the convergencies that necessarily result from those functional
104 Chapter 3
Table 3.5
Convergent evolution of the C4 photosynthetic pathway in plants in the past 30 million
years
Convergent pathway and function: C4 PHOTOSYNTHESIS (more efficient photosynthetic

pathway to counteract Rubisco oxidation and photorespiration)
1 Water thyme (Euangiosperms: Monocotyledons: Alismatales: Hydrocharitaceae;
Hydrilla verticillata)
2 Switchgrass (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Poaceae; Panicum virgatum)
3 Maize (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Poaceae; Zea mays)
4 Australian eriachne grass (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Poaceae; Eriachne triodioides)
5 Sand dropseed (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Poaceae; Sporobolus cryptandrus)
6 Flatspike sedge (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Cyperaceae; Abildgaardia ovata)
7 Egyptian papyrus (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Cyperaceae; Cyperus papyrus)
8 Amphibious sedge (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Cyperaceae; Eleocharis vivipara)
9 Rhynchospora (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Cyperaceae; Rhynchospora rubra)
10 Mongolian calligonum (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Polygonaceae; Calligonum mongolicum)
11 Pindan pink (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Caryophyllaceae; Polycarpaea longiflora)
12 Tassel flower (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Amaranthaceae; Amaranthus edulis)
13 Black saxual (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Chenopodiaceae; Haloxylon aphyllum)
14 Suaeda halophyte (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Chenopodiaceae; Suaeda aralocaspica)
15 Four-wing saltbush (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Chenopodiaceae; Atriplex canescens)
16 Slender snakecotton (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Chenopodiaceae; Froelichia gracilis)
17 Bienertia halophyte (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Chenopodiaceae; Bienertia sinuspersici)
18 Little hogweed (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Portulacaceae; Portulaca oleracea)
19 Australian anacampseros (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Portulacaceae; Anacampseros australiana)
20 Shoreline purslane (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Aizoaceae; Sesuvium portulacastrum)
21 Scarlet spiderling (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Nyctaginaceae; Boerhavia coccinea)
22 Naked-stem carpetweed (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Molluginaceae; Mollugo nudicaulis)
23 Hawaiian spurge (Euangiosperms: Eudicots: Rosidae: Eurosids I: Euphorbiaceae;
Chamaecyce forbesii)
Table 3.5
(continued)
24 Desert forb (Euangiosperms: Eudicots: Rosidae: Eurosids I: Zygophyllales:

Zygophyllaceae; Zygophyllum simplex)
25 Arizona poppy (Euangiosperms: Eudicots: Rosidae: Eurosids I: Zygophyllales:
Zygophyllaceae; Kallstroemia grandiflora)
26 Purple mistress (Euangiosperms: Eudicots: Rosidae: Eurosids II: Brassicaceae;
Moricandia arvensis)
27 Spider wisp (Euangiosperms: Eudicots: Rosidae: Eurosids II: Brassicaceae; Cleome
gynandra)
28 Pineland heliotrope (Euangiosperms: Eudicots: Asteridae: Euasterids I:
Boraginaceae; Heliotropium polyphyllum)
29 Blepharis (Euangiosperms: Eudicots: Asteridae: Euasterids I: Acanthaceae;
Blepharis ciliaris)
30 Anticharis bush (Euangiosperms: Eudicots:Asteridae: Euasterids I: Scrophulariaceae;
Anticharis linearis)
31 Australian yellow weed (Euangiosperms: Eudicots: Asteridae: Euasterids II:
Asteraceae; Flaveria australasica)
32 Narrow-leaf lemon weed (Euangiosperms: Eudicots: Asteridae: Euasterids II:
Asteraceae; Pectis linifolia)
33 Rock anethum (Euangiosperms: Eudicots: Asteridae: Euasterids II: Asteraceae;
Glossocardia bosvallea)
constraints on possible evolutionary pathways. One way around these

constraints would be to switch from an autotrophic mode of life to a
heterotrophic one; that is, to evolve a mode of life similar to that seen in
the animals and the fungi. In the next chapter we shall consider plants
that have done just that: the convergent evolution of carnivorous plants
(plants that eat animals) and herbivorous plants (plants that eat other
plants).
Defense: Antiherbivore Adaptations
Herbivores are to plants what carnivores are to animals. A typical animal

anticarnivore defense is to run away, as discussed in the previous chapter
(see table 2.12), but this option is not open to plants. Being unable to
flee, plants have evolved both physical and chemical defensive systems,
in many ways convergent on the defensive spines, stingers, and poison
defenses evolved by the animals (see tables 2.13, 2.14).
The leaves of a plant are its food factories, and one could predict that
there is a selective advantage for plants that defend those structures. It
is not surprising that a wide range of distantly related plants all have
convergently evolved leaf spines, from the nonflowering cycads to highly
derived asterid angiosperms (table 3.6). The similarity of some of these
106 Chapter 3
Table 3.6
Convergent evolution of physical defenses in plants
1 Convergent structure and function: LEAF SPINES (spines on apices of leaf lobes or
around leaf margins to discourage herbivore biting)
1.1 Cycad (Euphyllophyta: Lignophyta: Spermatophyta: Cycadophyta: Cycadales:
Zamiaceae; Encephalartos ferox)
1.2 Aloe yucca (Spermatophyta: Angiospermae: Euangiosperms: Monocotyledons:
Eumonocotyledons: Liliales: Liliaceae; Yucca aloifolia)
1.3 Cogon grass (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Poaceae; Imperata cylindrical)
1.4 Black oak (Euangiosperms: Eudicots: Rosidae: Eurosids I: Fagaceae; Quercus
velutina)
1.5 Scorpion weed (Euangiosperms: Eudicots: Asteridae: Euasterids I: Boraginaceae;
Phacelia purshii)
1.6 European holly (Euangiosperms: Eudicots: Asteridae: Euasterids II: Aquifoliaceae;
Ilex aquifolium)
1.7 Bull thistle (Euangiosperms: Eudicots: Asteridae: Euasterids II: Asteraceae; Cirsium
vulgare)
2 Convergent structure and function: TWIGS/BARK WITH THORNS/SPINES (thorns
or spines on twigs and/or trunk bark to discourage herbivores)
2.1 Candle cactus (Euangiosperms: Eudicots: Core eudicots: Caryophyllales: Cactaceae;
Pilocereus lanuginosus)
2.2 Knotweed (Euangiosperms: Eudicots: Core eudicots: Caryophyllales: Polygonaceae;
Polygonum pennsylvanicum)
2.3 Madagascar spiny plant (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Didiereaceae; Alluaudia montagnacii)
2.4 Canary Island spurge (Euangiosperms: Eudicots: Rosidae: Eurosids I: Euphorbiaceae;
Euphorbia canariensis)
2.5 French rose (Euangiosperms: Eudicots: Rosidae: Eurosids I: Rosaceae; Rosa
gallica)
2.6 European buckthorn (Euangiosperms: Eudicots: Rosidae: Eurosids I: Rhamnaceae;
Rhamnus cathartica)
2.7 Honey locust (Euangiosperms: Eudicots: Rosidae: Eurosids I: Fabaceae; Gleditsia
triacanthos)
2.8 Common prickly-ash (Euangiosperms: Eudicots: Rosidae: Eurosids II: Rutaceae;
Zanthoxylum americanum)
2.9 Bottle tree (Euangiosperms: Eudicots: Asteridae: Euasterids I: Apocynaceae;
Pachypodium lealii)
3 Convergent structure and function: STINGERS (stinging poison hairs and needles to
discourage herbivores)
3.1 Stinging nettle (Euangiosperms: Eudicots: Rosidae: Eurosids I: Urticaceae; Urtica
dioica)
3.2 Spurge stinging nettle (Euangiosperms: Eudicots: Rosidae: Eurosids I:
Euphorbiaceae; Cnidoscollus stimulosus)
3.3 Cowhage vine (Euangiosperms: Eudicots: Rosidae: Eurosids I: Fabaceae; Mucuna
pruriens)
Table 3.6
(continued)
4 Convergent structure and function: HOSTED INSECT DEFENDERS (mutualistic

relationship with animals who also serve to defend the host plant)
4.1 Ant rattan (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Aracaceae; Daemonorops formicaria)
4.2 Whistling-thorn acacia (Euangiosperms: Eudicots: Rosidae: Eurosids I: Fabaceae;
Acacia drepanolobium)
4.3 Valley oak (Euangiosperms: Eudicots: Rosidae: Eurosids I: Fagaceae; Quercus
lobata)
4.4 Common figs (Euangiosperms: Eudicots: Rosidae: Eurosids I: Moraceae; Ficus
carica)
4.5 Ant plant (Euangiosperms: Eudicots: Asteridae: Euasterids I: Rubiaceae;
Hydnophytum formicarum)
biological structures to human deadly devices is often apparent; for

example, the aloe yucca is also known as the “Spanish bayonet” because
of the shape of its sharp-edged leaves. Cogon-grass blades contain silica
crystals that render the margins of the leaf as sharp as the blade of a
serrated saw, and every child quickly learns not to run barefoot around
holly trees. Scotland is said to have chosen the thistle as its national
emblem because ancient Scots were warned of approaching invaders in
the night when one of the invaders inadvertently stepped on a thistle
and let out a loud yelp of pain.
Other plants have convergently evolved defensive structures for the
twigs and bark of the plant as well as its leaves (table 3.6). Cactus and
spurge spines are familiar convergent examples of armor covering the
entire plant. Rose thorns are a hazard every gardener learns to be wary
of, and this same family of plants produces the cockspur hawthorn, Cra-
taegus crus-galli, a name illustrating the convergence of a plant structure
to one seen in animals. Another animal-form convergent example in the
family of the thorny honey locusts is the catclaw acacia, Acacia greggii,
with sharp sickle-shaped spines that can shred the clothing and skin,
ruining many a hiking experience in southwestern North America.
Still other plants add poison to their defensive structures (table 3.6),
producing stingers convergent to those found in animals (see tables 2.7,
2.14). The poisonous tiny hairs of the stinging nettle produce painful
burns if brushed against skin tissue; this same family produces the Aus-
tralian stinging tree, Dendrocnide moroides, with tiny spines containing
neurotoxin. The “cnide” in the plant’s generic name is apt, as the sting
108 Chapter 3
of this tree is like that of a cnidarian jelly fish and can actually be fatal
(Stewart 2009). The Euphorbiaceae family is famous for its eerie conver-
gence on numerous cactus forms, which we shall consider later in this
chapter, but it has also produced a convergent stinging nettle that is
virtually identical to a true nettle (table 3.6).
Last, some plants have recruited animals to defend them (table 3.6).
The defender species are sometimes ants, like those hosted by the
whistling-thorn acacia, or wasps, hosted by the valley oak. This conver-
gent mutualism is not common; much more abundant is the convergent
assistance of animals in plant reproduction, a topic that will be taken up
later in the chapter.
Convergent with animals (see table 2.13), plants have evolved chemi-
cal defenses as well. Unlike animals, however, plants have independently
acquired a staggering number of chemical defense systems (table 3.7).
This is not surprising, as the number of defensive strategies available to
organisms that cannot flee or actively fight is more limited. In table 3.7
I have listed numerous examples of the convergent evolution of poison
in different plant structures, from leaves to seeds to roots, taken largely
from the works of Stewart (2009) and Agosta (1996) and phylogeneti-
cally cross-checked using the APG II (2003). The most poisonous plant
families are the nightshades (Solanaceae), cashews (Anacardiaceae),
spurges (Euphorbiaceae), and parsleys (Apiaceae), groups that Stewart
(2009, 71) collectively terms “botanical crime families.” But this designa-
tion is only from our perspective; from the plant’s perspective, it is simply
defending itself. Although these families contain the most numerous, and
most infamous, species of poisonous plants, poison tissues are found scat-
tered across the plant phylogenetic spectrum, from nonflowering pines
and cycads, to more basal core angiosperms like magnoliids and mono-
cotyledons, to highly derived eudicots like rosids and asterids. For each
separate clade of plants, I have listed a single species example in table
3.7, regardless of whether the clade contains many poisonous species or
only a few, in order to illustrate the extent of convergent evolution of
chemical defenses across all clades.
The poison produced in convergent chemical defense is not always the
same, but even here biochemical convergence is widespread (Stewart
2009). For example, cyanide has been independently developed in mono-
cotyledons (Johnsongrass leaf), eurosids (rubber tree seeds), and asterids
(black elderberry), to name only a few examples. Poisonous alkaloid
neurotoxins like nicotine and cocaine have been independently evolved
by monocotyledons (betel nut), basal eudicots (monkshood plant), rosids
Table 3.7
Convergent evolution of chemical defenses in plants
1 Convergent structure and function: LEAF/STEM POISON (poison concentrated in

leaves and/or stems to protect phototrophic structures from herbivores)
1.1 Jack-in-the-pulpit leaf (Euangiosperms: Monocotyledons: Alismatales: Araceae;
Arisaema triphyllum)
1.2 False hellebore leaf (Euangiosperms: Monocotyledons: Eumonocotyledons: Liliales:
Liliaceae; Veratrum viride)
1.3 Yellow lady’s slipper leaf (Euangiosperms: Monocotyledons: Eumonocotyledons:
Asparagales: Orchidaceae; Cypripedium calceolus)
1.4 Blue agave stem (Euangiosperms: Monocotyledons: Eumonocotyledons:
Asparagales: Agavaceae; Agave tequilana)
1.5 Johnsongrass leaf (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Poaceae; Sorghum halepense)
1.6 Swallowwort leaf (Euangiosperms: Eudicots: Ranunculales: Papaveraceae;
Chelidonium majus)
1.7 Mayapple leaf (Euangiosperms: Eudicots: Ranunculales: Berberidaceae;
Podophyllum peltatum)
1.8 Curare vine leaf (Euangiosperms: Eudicots: Ranunculales: Menispermaceae;
Chondrodendron tomentosum)
1.9 Kratom leaf (Euangiosperms: Eudicots: Core eudicots: Caryophyllales: Cactaceae;
Trichocereus pachanoi)
1.10 Soapwort leaf (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Caryophyllaceae; Saponaria officinalis)
1.11 Marijuana leaf (Euangiosperms: Eudicots: Rosidae: Eurosids I: Cannabinaceae;
Cannabis sativa)
1.12 Khat shrub leaf (Euangiosperms: Eudicots: Rosidae: Eurosids I: Celastraceae;
Catha edulis)
1.13 Coca shrub leaf (Euangiosperms: Eudicots: Rosidae: Eurosids I: Erythroxylaceae;
Erythroxylum coca)
1.14 Ratbane tree leaf (Euangiosperms: Eudicots: Rosidae: Eurosids I: Dichapetalaceae;
Dichapetalum toxicarium)
1.15 Pencil cactus stem (Euangiosperms: Eudicots: Rosidae: Eurosids I: Euphorbiaceae;
Euphorbia tirucalli)
1.16 Upas tree leaf (Euangiosperms: Eudicots: Rosidae: Eurosids I: Moraceae; Antiaris
toxicaria)
1.17 Black locust bark (Euangiosperms: Eudicots: Rosidae: Eurosids I: Fabaceae;
Robinia pseudoacacia)
1.18 Tansy mustard leaf (Euangiosperms: Eudicots: Rosidae: Eurosids II: Brassicaceae;
Descurainia pinnata)
1.19 Poison sumac leaf (Euangiosperms: Eudicots: Rosidae: Eurosids II: Anacardiaceae;
Toxicodendron vernix)
1.20 Tea leaf (Euangiosperms: Eudicots: Asteridae: Ericales: Thaeaceae; Thea sinensis)
1.21 Common rhododendron leaf (Euangiosperms: Asteridae: Ericales: Ericaceae;
Rhododendron ponticum)
1.22 Strychnine vine leaf (Euangiosperms: Eudicots: Asteridae: Euasterids I:
Loganiaceae; Strychnos toxifera)
1.23 Tobacco leaf (Euangiosperms: Eudicots: Asteridae: Euasterids I: Solanaceae;
Nicotiana tabacum)
1.24 Diviner’s sage leaf (Euangiosperms: Eudicots: Asteridae: Euasterids I: Lamiaceae;
Salvia divinorum)
110 Chapter 3
Table 3.7
(continued)
1.25 Kombe vine leaf (Euangiosperms: Eudicots: Asteridae: Euasterids I: Apocynaceae;

Strophanthus kombe)
1.26 Sweet woodruff leaf (Euangiosperms: Eudicots: Asteridae: Euasterids I: Rubiaceae;
Galium odoratum)
1.27 Indian tobacco leaf (Euangiosperms: Eudicots: Asteridae: Euasterids II:
Campanulaceae; Lobelia inflata)
1.28 White snakeroot leaf (Euangiosperms: Eudicots: Asteridae: Euasterids II:
Asteraceae; Eupatorium rugosum)
1.29 English ivy leaf (Euangiosperms: Eudicots: Asteridae: Eurasterids II: Araliaceae;
Hedera helix)
1.30 Blister bush leaf (Euangiosperms: Eudicots: Asteridae: Euasterids II: Apiaceae;
Peucedanum galbanum)
2 Convergent structure and function: SEED/FRUIT POISON (poison concentrated in
seeds and/or fruit to protect reproductive structures from herbivores)
2.1 Sweet shrub seed (Euangiosperms: Magnoliidae: Laurales: Calycanthaceae;
Calycanthus floridus)
2.2 Betel nut (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Arecaceae; Areca catechu)
2.3 Opium poppy seedpod (Euangiosperms: Eudicots: Ranunculales: Papaveraceae;
Papaver somniferum)
2.4 Blue cohosh seed (Euangiosperms: Eudicots: Ranunculales: Berberidaceae;
Caulophyllum thalictroides)
2.5 Castor bean (Euangiosperms: Eudicots: Rosidae: Eurosids I: Euphorbiacea; Ricinus
communis)
2.6 Rosary pea seed (Euangiosperms: Eudicots: Rosidae: Eurosids I: Fabaceae; Abrus
precatorius)
2.7 Habanero chile pod (Euangiosperms: Eudicots: Rosidae: Eurosids I: Solanaceae;
Capsicum chinense)
2.8 Squirting cucumber (Euangiosperms: Eudicots: Rosidae: Eurosids I: Cucurbitaceae;
Ecballium elaterium)
2.9 Coyotillo berries (Euangiosperms: Eudicots: Rosidae: Eurosids I: Rhamnaceae;
Karwinskia humboldtiana)
2.10 Cacao bean (Euangiosperms: Eudicots: Rosidae: Eurosids II: Malvaceae;
Theobroma cacao)
2.11 Fetid buckeye (Euangiosperms: Eudicots: Rosidae: Eurosids II: Sapindaceae;
Aesculus glabra)
2.12 Lime fruit (Euangiosperms: Eudicots: Rosidae: Eurosids II: Rutaceae; Citrus
aurantifolia)
2.13 Coffee bean (Euangiosperms: Eudicots: Asteridae: Euasterids I: Rubiaceae;
Coffea arabica)
2.14 Suicide-tree seed (Euangiosperms: Eudicots: Asteridae: Euasterids I: Apocynaceae;
Cerbera odollam)
2.15 Lantana berries (Euangiosperms: Eudicots: Asteridae: Euasterids I: Verbenaceae;
Lantana camara)
2.16 Morning glory seed (Euangiosperms: Eudicots: Asteridae: Euasterids I:
Convolvulaceae; Ipomoea tricolor)
2.17 Black elderberry (Euangiosperms: Eudicots: Asteridae: Euasterids II: Adoxaceae;
Sambucus nigra)
Table 3.7
(continued)
3 Convergent structure and function: ROOT/BULB POISON (poison concentrated in

roots and/or bulbs to protect nutrient uptake and reproductive structures from herbivores)
3.1 Fly-poison bulb (Euangiosperms: Monocotyledons: Eumonocotyledons: Liliaceae;
Amianthium muscaetoxicum)
3.2 Bloodroot root (Euangiosperms: Eudicots: Ranunculales: Papaveraceae;
Sanguinaria canadensis)
3.3 Pokeberry root (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Phytolaccaceae; Phytolacca americana)
3.4 Cassava root (Euangiosperms: Eudicots: Rosidae: Eurosids I: Euphorbiaceae;
Manihot esculenta)
3.5 Potato tuber (Euangiosperms: Eudicots: Asteridae: Euasterids I: Solanaceae;
Solanum tuberosum)
3.6 Iboga root (Euangiosperms: Eudicots: Asteridae: Euasterids I: Apocynaceae;
Tabernanthe iboga)
3.7 Water hemlock root (Euangiosperms: Eudicots: Asteridae: Eurasterids II: Apiaceae;
Cicuta maculata)
4 Convergent structure and function: ENTIRE PLANT POISON (poison distributed
throughout entire plant to repel herbivores)
4.1 Sago palm (Euphyllophyta: Lignophyta: Spermatophyta: Cycadophyta: Cycadaceae;
Cycas revoluta)
4.2 European yew tree (Euphyllophyta: Lignophyta: Spermatophyta: Pinophyta:
Taxaceae; Taxus baccata)
4.3 Split-leaf philodendron (Spermatophyta: Angiospermae: Euangiosperms:
Monocotyledons: Alismatales: Araceae; Philodendron selloum)
4.4 Autumn crocus (Euangiosperms: Monocotyledons: Eumonocotyledons: Liliaceae;
Colchicum autumnale)
4.5 Monkshood (Euangiosperms: Eudicots: Ranunculales: Ranunculaceae; Aconitum
napellus)
4.6 Peyote cactus (Euangiosperms: Eudicots: Core eudicots: Caryophyllales: Cactaceae;
Lophophora williamsii)
4.7 Black snakeroot (Euangiosperms: Eudicots: Rosidae: Geraniales: Melanthiaceae;
Zigadenus venenosus)
4.8 Daphne shrub (Euangiosperms: Eudicots: Rosidae: Eurosids II: Thymelaeaceae;
Daphne mezereum)
4.9 Mountain laurel (Euangiosperms: Eudicots: Asteridae: Ericales: Ericaceae; Kalmia
latifolia)
4.10 Oakleaf hydrangea shrub (Euangiosperms: Eudicots: Asteridae: Cornales:
Hydrangeaceae; Hydrangea quercifolia)
4.11 Deadly nightshade (Euangiosperms: Eudicots: Asteridae: Euasterids I: Solanaceae;
Atropa belladona)
4.12 Buttonbush (Euangiosperms: Eudicots: Asteridae: Euasterids I: Rubiaceae;
Cephalanthus occidentalis)
4.13 Common foxglove (Euangiosperms: Eudicots: Asteridae: Euasterids I:
Plantaginaceae; Digitalis purpurea)
4.14 Butterfly weed (Euangiosperms: Eudicots: Asteridae: Euasterids I: Asclepiadaceae;
Asclepias tuberosa)
4.15 Yellow jessamine vine (Euangiosperms: Eudicots: Asteridae: Euasterids I:
Gelsemiaceae; Gelsemium sempervirens)

112 Chapter 3
(coca leaf), and asterids (tobacco leaf). Deadly paralysis-inducing muscle

toxins have been independently developed in plants ranging from basal
eudicots (curare vine leaf) to highly derived asterids (strychnine vine
leaf), and heart-muscle poisons have independently evolved in rosids
(upas tree leaf) and asterids (kombe vine leaf). Psychoactive toxins,
targeting the brain, have been independently developed by basal core
eudicots (kratom leaf), rosids (marijuana leaf), and asterids (diviner’s
sage leaf). Powerful skin-destructive irritants have been independently
developed by monocotyledons (blue agave stem), rosids (poison sumac
leaf), and asterids (blister bush leaf).
Some of these plants are not really dangerous to humans; morning
glory seeds, for example, contain only tiny amounts of the psychoactive
toxin LSD, and hydrangea bushes only small amounts of cyanide. But
then, morning glories and hydrangea bushes are not defending them-
selves against large animals like humans; they have smaller animals in
target. Others are surprising: cassava roots and potato tubers are poison-
ous (the latter less so than the former), but are rendered harmless if they
are cooked properly, and they are a widely used food source for humans.
Humans even deliberately enjoy the effects of the neurotoxin caffeine
in tea leaves and coffee beans, and some enjoy the neurotoxin nicotine
in tobacco leaves (although the delivery system, burning cigarettes, even-
tually kills them).
Defense: Antidehydration Adaptations
Dehydration is a very serious problem for land-dwelling organisms.

Many of the morphological and physiological changes that took place in
the evolution of tetrapods from fish were adaptations for preventing
water loss in the dry, hostile terrestrial environment. (Other changes
were linked to two other serious problems on land: the crushing force of
gravity and rapid temperature fluctuations, as compared to the sea,
where neutrally buoyant fish feel no gravity and temperatures vary very
little, even on seasonal scales.) Even so, if an animal feels thirsty, it is able
to go hunting for water and to replenish its water loss by drinking as
much water as it can find.
In contrast, plants cannot go hunting for water. Plants are entirely
dependent on the vagaries of terrestrial weather patterns in bringing rain
and groundwater, weather patterns that are usually unreliable. A wet
spring this year might be followed by a drought next year. One way to
deal with an unreliable source of water is to store up as much water as
possible when it is available, so the plant can survive periods of time

when water is scarce. One such water-storage adaptation among plants
is the evolution of succulent stems and leaves, structures that store large
amounts of water in fleshy pith and cortex cells (Niklas 1997), resulting
in a plant that possesses stems and leaves that appear swollen or fat.
Plants of widely differing phylogenetic lineages have repeatedly and
convergently evolved succulent stems and leaves (table 3.8); interest-
ingly, even some animals have converged on a succulent water-storage
adaptation, such as the desert-adapted dromedary camel, Camelus dro-
medarius, which can store enormous amounts of water in its tissues. In
table 3.8 I have listed examples from families containing the largest
numbers of succulent species known to me, such as the Cactaceae (cac-
tuses), Euphorbiaceae (spurges), Aizoaceae (stone plants), and Crassu-
laceae (jade plants). Many, many other families of plants have a few
species, or sometimes only one, that have also convergently evolved
succulent stems or leaves while living in semiarid environments; thus
the list given in table 3.8 cannot be taken in any sense to be exhaustive.
Table 3.8
Convergent evolution of water-retention structures in plants
Convergent structure and function: SUCCULENT STEMS/LEAVES (water-storage

structures to ensure sufficient water for plant survival in semiarid environments or
environments in which the water source is unreliable)
1 Quiver tree (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Asphodelaceae; Aloe dichotoma)
2 Agave (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Agavaceae; Agave flexispina)
3 Rosette plant (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Bromeliaceae; Abromeitiella lorentziana)
4 Candle cactus (Euangiosperms: Eudicots: Core eudicots: Caryophyllales: Cactaceae;
Pilocereus lanuginosus)
5 Stone plant (Euangiosperms: Eudicots: Core eudicots: Caryophyllales: Aizoaceae;
Carpobrotus edulis)
6 Madagascar spiny plant (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Didiereaceae; Alluaudia montagnacii)
7 Jade plant (Euangiosperms: Eudicots: Core eudicots: Saxifragales; Crassulaceae;
Crassula ovata)
8 Canary Island spurge (Euangiosperms: Eudicots: Rosidae: Eurosids I: Euphorbiaceae;
Euphorbia canariensis)
9 Bottle tree (Euangiosperms: Eudicots: Asteridae: Euasterids I: Apocynaceae;
Pachypodium lealii)
10 Pickle plant (Euangiosperms: Eudicots: Asteridae: Euasterids II: Asteracea; Senecio
stapeliiformis)

114 Chapter 3
It can, however, illustrate how phylogenetically widespread the phenom-

enon is—independently arising in numerous monocotyledons, core eudi-
cots, and highly derived rosids and asterids—and how geographically
widespread—from Africa (quiver tree) to North America (candle
cactus) to South America (rosette plant) to Madagascar (bottle tree;
Niklas 1997).
The most spectacular examples of the convergent evolution of identi-
cal morphological structures to prevent dehydration are the desert-
adapted cactuses of the Western Hemisphere and spurges of the Eastern
Hemisphere (table 3.9). Niklas (1997) documents that both groups
started out from morphologically similar beginnings, from tropical broad-
leafed trees, such as the leaf cactus and the honey spurge (table 3.9). In
subsequent evolution in more arid environments, both groups of plants
reduced the amount of their woody stem tissues, the number of their
branches, and the sizes of their leaves—progressing from tall, treelike
cactus forms to shorter, sparsely branched cactus forms, to squat, barrel-
shaped stems, and finally to rounded sea-urchin-shaped domes (table
3.9). Both groups of plants lost their leaves and evolved thorns (cactuses
evolved thorns from modified leaves and spurges from modified
branches) that not only deter herbivores (table 3.6) but also dissipate
heat and break up wind currents from the stem surface, both of which
can increase water loss (Niklas 1997).
Table 3.9
Convergent evolution of cactus forms in arid-adapted desert plants of the Western
Hemisphere (Cactaceae) and the Eastern Hemisphere (Euphorbiaceae)
Ancestral morphological condition: tropical broad-leafed tree species

Cactaceae: Leaf cactus (Pereskia sacharosa)
Euphorbiaceae: Honey spurge (Euphorbia mellifera)
1 Tall tree-forms with multiple succulent branches from a central trunk
Cactaceae: Candle cactus (Pilocereus lanuginosus)
Euphorbiaceae: Canary Island spurge (Euphorbia canariensus)
2 Shorter, sparsely branched succulent stems with zigzag vertical ridges and furrows
Cactaceae: Snake cactus (Peniocereus serpintinus)
Euphorbiaceae: Devil’s backbone spurge (Euphorbia cryptospinosa)
3 Squat, barrel-shaped succulent stem with flat ridges separating recessed furrows
Cactaceae: Compass-barrel cactus (Ferocactus cylindraceus)
Euphorbiaceae: Velvet spurge (Euphorbia valida)
4 Rounded, sea-urchin-shaped succulent dome
Cactaceae: Sea-urchin cactus (Astrophytum asterias)
Euphorbiaceae: Baseball spurge (Euphorbia obesa)

Reproduction
In the last chapter we saw that animals have repeatedly and indepen-
dently evolved the same reproductive systems, particularly viviparity. The
same is true of plants, except the convergent reproductive system for
these organisms is heterospory. The repeated evolution of heterospory
in independent phylogenetic lineages has been called “the most iterative
key innovation in the evolutionary history of the plant kingdom”
(Bateman and DiMichele 1994, 345).
The original Devonian tracheophytes had sporophytes that shed their
spores freely into the air; the spores, which were all the same size, then
produced free-living gametophytes that produced the next generation of
sporophytes. The free-living gametophytes were bisexual, and required
moist conditions not only for their vegetative growth but also for sperm
transfer and fertilization of their eggs, as well as the survival of the devel-
oping new sporophyte embryo (Niklas 1997). The production of bisexual
gametophytes is an ancient reproductive condition known as homospory,
which still exists in many living species of club mosses (lycophytes),
horsetails (equisetophytes), and ferns (filicophytes). In many ways,
homosporous plants “may be thought of as the amphibians of the plant
kingdom in the sense that the completion of their sexual life cycle
requires a ‘return to water’” (Niklas 1997, 190).
A major evolutionary shift in plants’ mode of reproduction occurred
independently in many separate plant lineages during the Late Devonian
(table 3.10). This was the convergent evolution of heterospory, where
plants now produced two distinctly different types of spores: numerous,
very small microspores and relatively few but much larger megaspores.
Specialized plants producing the new microspores developed into sperm-
producing gametophytes, and plants producing the larger megaspores
developed into egg-bearing gametophytes; thus, heterosporous plants
had now evolved gametophytes that were unisexual, not bisexual like the
homosporous plants (Niklas 1997). The evolution of heterospory and
unisexual gametophytes occurred independently at least eleven separate
times in disparate groups of plants (Bateman and DiMichele 1994). Two
separate groups of lycopods, one major group of equisetophytes, and
three separate groups of filicophytes convergently became heterospo-
rous (table 3.10), although homosporous species of these types of plants
exist to the present day. The filicophytes are particularly interesting in
that they have convergently evolved heterospory repeatedly throughout
geologic time: at least once in the Late Devonian, at least once in the
116 Chapter 3
Table 3.10
Convergent evolution of heterosporous reproduction, and ultimately of the seed
Convergent structure and function: HETEROSPORES (differentiated microspores and

megaspores for the production of unisexual gametophytes)
1 Devonian barinophyte spores (Tracheophyta: Lycophyta: Barinophytales:
Barinophytaceae; Protobarinophyton pennsylvanicum †Late Devonian)
2 Devonian selaginellalean spores (Tracheophyta: Lycophyta: Selaginellales:
Protolepidodendraceae; Barsostrobus famennensis †Late Devonian)
3 Carboniferous calamite spores (Tracheophyta: Euphyllophyta: Moniliformopses:
Equisetophyta: Equisetales: Archaeocalamitaceae; Protocalamostachys farringtonii
†Early Carboniferous)
4 Devonian stauropterid spores (Euphyllophyta: Moniliformopses: Filicophyta:
Stauropteridales: Gillespieaceae; Gillespiea randolphensis †Late Devonian)
5 Giant water-fern spores (Euphyllophyta: Moniliformopses: Filicophyta: Salviniales:
Salviniaceae; Salvinia molesta)
6 European water-clover fern spores (Euphyllophyta: Moniliformopses: Filicophyta:
Marsileales: Marsileaceae; Marsilea quadrifolia)
7 Australian platyzoma-fern spores (Euphyllophyta: Moniliformopses: Filicophyta:
Filicales: Platyzomataceae; Platyzoma microphylla)
8 Devonian aneurophyte spores (Euphyllophyta: Lignophyta: Aneurophytales:
Chaleuriaceae; Chaleuria cirrosa †Middle Devonian)
9 Devonian archaeopterid spores (Euphyllophyta: Lignophyta: Archaeopteridales:
Archaeopteridaceae; Archaeopteris latifolia †Late Devonian)
10 Carboniferous noeggerathialean spores (Euphyllophyta: Lignophyta:
Noeggerathiales: Noeggerathiaceae; Noeggerathiostrobus vicinalis †Late Carboniferous)
11 Carboniferous cecropsidalean spores (Euphyllophyta: Lignophyta: Cecropsidales:
Cecropiaceae; Cecropsis luculentum †Late Carboniferous)
12 Devonian elkinsia seeds (Euphyllophyta: Lignophyta: Spermatophyta:
Lagenstomales: Elkinsiaceae; Elkinsia polymorpha †Late Devonian)
Late Cretaceous (the ancestors of living giant water ferns), and sepa-
rately in the modern European water-clover ferns and Australian
platyzoma ferns (table 3.10). Four separate groups of lignophytes inde-
pendently evolved heterospory in the span of time from the Middle
Devonian to the Late Carboniferous: the major plant lineages of the
aneurophytales and archaeopteridales, and the more enigmatic noeg-
gerathiales and cecropsidales, all of which are now extinct. In the midst
of this rampant convergent evolution of heterospory in plant lineages,
one plant group took the next major evolutionary step in plant reproduc-
tion: the evolution of the seed (table 3.10).
The evolution of heterospory, the partitioning of resources between
numerous small microspores and a few large megaspores, set the stage
for the evolution of the seed (Niklas 1997). Production of microspores
required minimal resources, and because they were so very small they
were easily and widely distributed by wind, ensuring an enlarged geo-

graphic dispersal of genetic information and the reduced probability of
inbreeding in plants. The concentration of more resources in a few mega-
spores, or in only one megaspore, increased the probability of successful
production of new sporophytes by these megagametophytes. The next
step in plant reproductive evolution would be for the sporophyte to
retain the megaspores rather than releasing them to produce free-living
megagametophytes. Hence the evolution of the seed: “The seed is an
adaptive solution to life on land functionally analogous to the way amni-
otic animals reproduce. . . . The evolution of the seed unquestionably
released vascular plant reproduction from the ecological requirement for
an external source of water for the growth and survival of the megaga-
metophyte, the fertilization of its eggs, and the early development of the
sporophyte embryo. Thus the seed opened the door for the adaptive
radiation of vascular plants into habitats drier than those favoring the
survival and reproductive success of free-sporing plants” (Niklas 1997,
343–344).
The first known spermatophytes, the seed plant Elkinsia polymorpha,
are found as fossils in Late Devonian strata of the Appalachian Moun-
tains of eastern North America. Just as multiple plant lineages converged
on the evolution of heterospory, the fossil record suggests that several
plant lineages were converging on the evolution of the seed. Spermato-
phytes are heterosporous plants that retain the megagametophyte, and
its developing embryos, within sporophyte tissues. Niklas (1997) argues
that heterosporous species of the ancient lepidodendrid lycophyte Lepi-
docarpon and the ancient calamite equisetophyte Calamocarpon were
well on the way to developing the seed habit. However, as far as we yet
know from the fossil record, only one lineage of plants successfully
developed the seed, and that group, the spermatophytes, dominates the
terrestrial plant biosphere just as the amniotes dominate the terrestrial
animal biosphere.
The spermatophytes evolved in the Late Devonian and the amniotes
soon after in the Early Carboniferous. The amniotes quickly diverged
into two distinct phylogenetic lineages—the sauropsids and the synap-
sids (see appendix)—and the subsequent evolution of members of these
lineages has been discussed in chapter 2. The spermatophytes were more
divergent in their evolution: they produced three distinct lineages that
are now extinct (the lyginopteridales, medullosanales, and gigantopteri-
dales) and the five living ginkgophyte, cycadophyte, pinophyte, gneto-
phyte, and angiosperm lineages (see appendix). Just as the spermatophytes,
118 Chapter 3
the seed plants, are the most successful of all the land plants, the angio-
sperms, the flowering plants, are the most successful of all the
spermatophytes.
Why flowers? The plant seed, like the animal amniote egg, is a repro-
ductive structure clearly linked to antidehydration adaptations in the
hostile dry air of terrestrial habitats. Much of the success of flowering
plants is linked to solving not dehydration problems but a different
reproductive problem: the problem of fertilization in organisms that
cannot move about in search for a mate.
The problem of fertilization in sessile organisms is an ancient one that
extends back to life in the sea. Many sessile marine animals simply
release their gametes directly into the sea water and depend upon water
currents to transport them to other animals for fertilization. These
animals are often astonishingly convergent to plants in their morpholo-
gies: crinoids are deuterostomous bilaterians, closely related to chordate
animals such as ourselves, yet many ancient crinoids had root systems
that looked like those of land plants, stems like plant trunks, and branch-
ing arms like plant branches. Plant similarities are often seen in both the
colloquial and scientific names of sessile marine animals: stalked echino-
derms are known as “sea lilies” because their long stems and calyxes look
like the stems and flowers of lilies, the bryozoa are named “moss animals”
because their colonies look like moss, the gorgonian cnidarian “sea fans”
whose central colonial stem and lateral branches look like palm leaves,
the colorful polyps of an anthozoan cnidarian colony appear similar to
a garden of “sea flowers,” and so on.
The apparent convergence in form between many sessile marine
animals and land plants is in fact reversed: it is the land plants that have
converged on the sessile marine animals, not the other way around, as
the marine animals evolved these forms first and the plants converged
on them secondarily. Rather than currents in the dense water of marine
habitats, land plants must depend upon wind in the thin atmosphere of
terrestrial habitats to disperse their gametes. Just as many sessile marine
animals have elevated themselves above the sea floor in order to catch
stronger marine currents, plants have evolved stems and trunks to
elevate them up in the air in order to catch the wind with their spores
and pollen.
Many land plants are thus not only dependent upon the vagaries of
terrestrial weather patterns in bringing rain; they are also dependent
upon these same weather patterns in bringing wind for successful fertil-
ization. The spermatophytes largely depended upon wind for fertilization
for over 240 million years, from the Late Devonian until the evolution
of the first angiosperms some 115 to 125 million years ago (Thien et al.
2009). At that time, the angiosperm spermatophytes evolved an interest-
ing new fertilization strategy: they began to involve a sexual partner who
was mobile. Although a few nonangiosperm spermatophytes use pollen-
eating insects as pollination assistants, the really elaborate plant-animal
reproductive pas de deux began with the evolution of the first flowering
plants in the Early Cretaceous (and with the evolution of the dioecious
angiosperms, which have separate male and female plants, the plant-
animal interaction becomes a ménage à trois).
The basal angiosperms consist of three sequential cladogram branches,
in what has been termed the ANITA grade of angiosperm evolution
(Thien et al. 2009). These branches are: first, the Amborellales, second,
the Nymphaeales, and third, the Austrobaileyales (APG II 2003; Lecoin-
tre and Le Guyader 2006), based upon molecular phylogenies. Although
molecular evidence places the Amborellales at the very base of the
angiosperm clade, the oldest fossil flowers found thus far belong to the
Nymphaeales, one branch up from the base of the clade.
The oldest flower yet discovered is quite small, only about 1 centimeter
in diameter. Somewhat younger nymphaealean flowers, around 90 million
years old, have fossil morphologies very similar to modern-day flowers
that are pollinated by beetles, reinforcing long-held views that beetle
pollination was a major form of plant-animal reproductive interaction
in the earliest of the flowering plants (Bernhardt 2000; Thien et al.
2009). Examination of living members of the basal angiosperms reveal
that flies are the primary pollinators of six families, beetles the primary
pollinators of five families, and bees are major pollinators in only one
family (Thien et al. 2009). Plant-animal interactions surely must have
evolved in the 115 to 125 million years that have passed since the evolu-
tion of the first flowering plant, and so the reproductive pattern seen in
living basal angiosperms cannot be taken as proof of identical relation-
ships in the Early Cretaceous. Still, it is interesting that living basal
angiosperms are pollinated preferentially by flies and beetles, and only
rarely by bees. More derived angiosperms not only added bees to the list
of sexual partners but added butterflies, moths, wasps, birds, and mammals
as well.
Four different types of flower morphologies have been evolved by
angiosperms that are pollinated exclusively by beetles: bilobate, brush,
chamber-blossom, and painted-bowl flowers (Bernhardt 2000). These
cantharophilous flower types have been convergently evolved over and
120 Chapter 3
over again in even highly derived angiosperms (table 3.11); that is, beetle
pollination is not a reproductive mode confined to the basal angiosperms
or plesiomorphic magnoliids. Nine independent groups of monocotyle-
dons, eight of them derived eumonocotyledons, have independently
evolved cantharophilous flowers (table 3.11). In the derived eudicot
angiosperms, cantharophilous flowers have convergently appeared 12
separate times, ranging from basal eudicots like the Ranunculalaes to the
highly derived Rosidae and Asteridae (Bernhardt 2000).
Giant flowers are rarer and more specialized than cantharophilous
flowers, yet they too have convergently evolved multiple times in the
angiosperms (table 3.12). Floral gigantism is defined as the possession of
blossoms—either flowers or flowerlike inflorescences—that exceed 30
centimeters in diameter (Davis et al. 2008). Very large flowers can be up
to a meter in diameter, such as those found in rafflesia, or up to three
meters in height, like as those found in the corpse lilies (table 3.12). These
giant flowers attract a wide range of pollinators: beetles, flies, moths,
birds, fruit bats, and even lemur primates (Davis et al. 2008). However,
the great majority of pollinators of giant flowers are flies and small-
bodied beetles. It appears that the large flower size facilitates the tem-
porary trapping of these smaller animals by the flower, and the animals
subsequently pollinate the flower as they seek an escape route. The phy-
logenetic distribution of the convergent evolution of floral gigantism is
across the entire spectrum of the angiosperm clade: from the most basal
angiosperms, like the Nymphaeales, to the highly derived Eurasterids
(table 3.12).
Even more specialized than giant flowers are flowers that smell like
rotting meat. One would think that such a specialized type of flower,
specifically adapted to the restricted role of attracting carrion-eating
animals, would be unique in the evolution of flowering plants, but this is
not the case: sapromyiophilous, or carrion-mimic, flowers have conver-
gently evolved in the angiosperms no less than seven separate times
(table 3.13). Some carrion-mimic flowers are also giant, like the corpse
lilies and rafflesia (table 3.12), but others are normal sized or even small
(Davis et al. 2008; Stewart 2009). Sapromyiophilous flowers have inde-
pendently evolved in plesiomorphic angiosperms like the magnoliid
pawpaws; in monocotyledons like the corpse lilies and carrion flowers;
in basal eudicots like the stinking hellebore; and in highly derived euaste-
rids like the Zulu giant carrion plant (table 3.13).
Derived flowers have evolved shapes, colors, and odors to attract the
attention of potential pollinators. Flowers also offer food rewards to
Table 3.11
Convergent evolution of cantharophilous flowers for plant pollination by beetles in
angiosperms
Convergent structure and function: CANTHAROPHILOUS FLOWERS (evolution of

specialized flower morphologies that facilitate pollination by beetles)
1 Corpse lily (Euangiosperms: Monocotyledons: Alismatales: Araceae; Amorphophallus
titanum)
2 Split-leaf cyclanthus (Euangiosperms: Monocotyledons: Eumonocotyledons:
Pandanales: Cyclanthaceae; Cyclanthus bipartitus)
3 Sun’s eye tulip (Euangiosperms: Monocotyledons: Eumonocotyledons: Liliales:
Liliaceae; Tulipa agenensis)
4 Peacock iris (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Iridaceae; Moraea glaucopis)
5 Giant orchid (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Orchidaceae; Pteroglossaspis ecristata)
6 Star of Bethlehem (Euangiosperms: Monocotyledons: Eumonocotyledons:
Asparagales: Hyacinthaceae; Ornithogalum umbellatum)
7 Cape star (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Hypoxidaceae; Spiloxene capensis)
8 Orchidantha (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Lowiaceae; Orchidantha maxillarioides)
9 Root-spine palm (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Arecaceae; Crysophila warscewiczii)
10 Persian buttercup (Euangiosperms: Eudicots: Ranunculales: Ranunculaceae;
Ranunculus asiaticus)
11 Corn poppy (Euangiosperms: Eudicots: Ranunculales: Papaveraceae; Papaver
rhoeas)
12 Sundew (Euangiosperms: Eudicots: Core eudicots: Caryophyllales: Droseraceae;
Drosera pauciflora)
13 Autograph tree (Euangiosperms: Eudicots: Rosidae: Eurosids I: Clusiaceae; Clusia
rosea)
14 Tropical chestnut (Euangiosperms: Eudicots: Rosidae: Eurosids II: Malvaceae;
Sterculia stipulata)
15 Malasian vatica (Euangiosperms: Eudicots: Rosidae: Eurosids II: Dipterocarpaceae;
Vatica parvifolia)
16 Indian mastixia (Euangiosperms: Eudicots: Asteridae: Cornales: Cornaceae; Mastixia
arborea)
17 Indian persimmon (Euangiosperms: Eudicots: Asteridae: Ericales: Ebenaceae;
Diospyros malabarica)
18 Ballhead ipomopsis (Euangiosperms: Eudicots: Asteridae: Ericales: Polemoniaceae;
Ipomopsis congesta)
19 Unamkodi (Euangiosperms: Eudicots: Asteridae: Euasterids I: Convolvulaceae;
Erycibe paniculata)
20 African iodes (Euangiosperms: Eudicots: Asteridae: Euasterids I: Icacinaceae; Iodes
africana)
21 Naked bluebell (Euangiosperms: Eudicots: Asteridae: Euasterids II: Campanulaceae;
Wahlenbergia gymnoclada)

122 Chapter 3
Table 3.12
Convergent evolution of floral gigantism in plants
Convergent structure and function: FLORAL GIGANTISM (blossoms with a diameter of

30 centimeters or greater that facilitate pollination by flies and small-bodied beetles)
1 Giant “water lily” (Euphyllophyta: Lignophyta: Spermatophyta: Angiospermae:
Nymphaeales: Nymphaeaceae; Victoria amazonica)
2 Big-leaf magnolia (Angiospermae: Euangiosperms: Magnoliidae: Magnoliales:
Magnoliaceae; Magnolia macrophylla)
3 Pelican flower (Euangiosperms: Magnoliidae: Piperales:Aristolochiaceae; Aristolochia
grandiflora)
titanum)
5 Lady slipper orchid (Euangiosperms: Monocotyledons: Eumonocotyledons:
Asparagales: Orchidaceae; Phragmipedium grande)
6 Sacred lotus (Euangiosperms: Eudicots: Proteales: Nelumbonaceae; Nelumbo
nucifera)
7 Rafflesia (Euangiosperms: Eudicots: Rosidae: Eurosids I: Rafflesiaceae; Rafflesia
arnoldii)
8 Beobab (Euangiosperms: Eudicots: Rosidae: Eurosids II: Malvaceae; Adansonia
digitata)
9 Zulu giant carrion plant (Euangiosperms: Eudicots: Asteridae: Euasterids I:
Apocynaceae; Stapelia gigantea)
Table 3.13
Convergent evolution of sapromyiophilous flowers that smell like rotting meat
Convergent structure and function: CARRION-MIMIC FLOWERS (production of

rotting-meat smell to attract carrion-eating animals for pollination purposes)
1 Pawpaw (Euangiosperms: Magnoliidae: Magnoliales: Annonaceae; Asimina triloba)
titanium)
3 Carrion flower (Euangiosperms: Monocotyledons: Eumonocotyledons: Liliales:
Liliaceae; Smilax herbacea)
4 Stinking iris (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Iridaceae; Iris foetidissima)
5 Stinking hellebore (Euangiosperms: Eudicots: Ranunculalaes: Ranunculaceae;
Helleborus foetidus)
arnoldii)
7 Zulu giant carrion plant (Euangiosperms: Eudicots: Asteridae: Euasterids I:
Apocynaceae; Stapelia gigantea)

animal pollinators: the ancient flowers offered only excess pollen for
pollen-eating animals, a trait that is actually older than the angiosperms,
as it is found in some nonangiosperm spermatophytes like cycads and
conifers as well (Labandeira et al. 2007). Derived flowers offer not only
excess pollen but also sweet nectar, starchy food bodies, stigmatic secre-
tions, perianth segments, and edible stamenodia (Bernhardt 2000; Thien
et al. 2009).
One particularly specialized type of food reward system, known as
ornithophily, is adapted specifically to attract nectar-feeding birds. Orni-
thophilous, or hummingbird, flowers have no smell; they are of a brick-
red monocolor, are trumpet-shaped, and produce copious amounts of
nectar. The evolution of hummingbird flowers in western North America
has been extensively studied by Grant and Grant (1968), who document
the independent evolution of ornithophilous flowers in 18 separate fami-
lies of angiosperms in this region of the world alone (table 3.14). For
each family, I have listed a single species representative, but some fami-
lies have more hummingbird-flowered species than others. For example,
the Scrophulariaceae contain eight genera and 74 species with humming-
bird flowers, and a more detailed phylogenetic analysis of the family
would be needed to determine how many of these hummingbird-
flowered species are independently derived and how many are synapo-
morphic. Other families, like the Cactaceae, Nyctaginaceae, Saxifragaceae,
Fouquieriaceae, Convolvulaceae, and Rubiaceae contain a single species
with hummingbird flowers, and it is clear that each of these is an inde-
pendent convergence on ornithophily (Grant and Grant 1968). Last, of
the 18 convergently evolved hummingbird flowers listed in table 3.14,
seven belong to species that are the sole red-flowered representative of
all of the other variously hued flowers belonging to other species within
each particular genus: Silene laciniata, Ribes speciosum, Astragalus coc-
cineus, Gilia aggregata, Ipomoea coccinea, Monardella macrantha,
Mimulus cardinalis; and only two red-flowered species are found in the
generally blue- and white-flowered species of the genus Delphinium, D.
cardinale and D. nudicaule (Went 1971).
Worldwide, some 65 families of angiosperms have evolved ornithophi-
lous flowers (Cronk and Ojeda 2008). It will take a detailed analysis of
the phylogenetic distribution of these flowers to determine how many
are independent originations of ornithophily. For example, species in all
eight families of the monocotyledon order Zingiberales are pollinated
by birds (Cronk and Ojeda 2008). Do these eight families represent eight
independent convergencies on ornithophily within the Zingiberales, or
124 Chapter 3
Table 3.14
Convergent evolution of ornithophilous hummingbird flowers of western North America
Convergent structure and function: HUMMINGBIRD FLOWERS (odorless, brick-red,

trumpet-shaped flowers rich in nectar that attract nectar-feeding birds for pollination
purposes)
1 Firecracker plant (Euangiosperms: Monocotyledons: Eumonocotyledons: Liliales:
Liliaceae; Brodiaea ida-maia)
2 Schott’s century plant (Euangiosperms: Monocotyledons: Eumonocotyledons:
Asparagales: Agavaceae; Agave schottii)
3 Scarlet larkspur (Euangiosperms: Eudicots: Ranunculalaes: Ranunculaceae;
Delphinium cardinale)
4 Indian pink (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Caryophyllaceae; Silene laciniata)
5 Hedgehog cactus (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Cactaceae; Echinocereus triglochidiatus)
6 Scarlet four o’clock (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Nyctaginaceae; Allionia coccinea)
7 Fuchsia-flowered gooseberry (Euangiosperms: Eudicots: Core eudicots: Saxifragales:
Saxifragaceae; Ribes speciosum)
8 California fuchsia (Euangiosperms: Eudicots: Rosidae: Myrtales: Onagraceae;
Zauschneria californica)
9 Locoweed (Euangiosperms: Eudicots: Rosidae: Eurosids I: Fabaceae; Astragalus
coccineus)
10 Ocotillo (Euangiosperms: Eudicots: Asteridae: Ericales: Fouquieriaceae; Fouquieria
splendens)
11 Scarlet skyrocket (Euangiosperms: Eudicots: Asteridae: Ericales: Polemoniaceae;
Gilia aggregata)
12 Star-glory (Euangiosperms: Eudicots: Asteridae: Euasterids I: Convolvulaceae;
Ipomoea coccinea)
13 Scarlet monardella (Euangiosperms: Eudicots: Asteridae: Euasterids I: Lamiaceae;
Monardella macrantha)
14 Crimson monkey-flower (Euangiosperms: Eudicots: Asteridae: Euasterids I:
Scrophlariaceae; Mimulus cardinalis)
15 Desert honeysuckle (Euangiosperms: Eudicots:Asteridae: Euasterids I:Acanthaceae;
Anisacanthus thurberi)
16 Scarlet bouvardia (Euangiosperms: Eudicots: Asteridae: Euasterids I: Rubiaceae;
Bouvardia glaberrima)
17 Cardinal flower (Euangiosperms: Eudicots: Asteridae: Euasterids II: Campanulaceae;
Lobelia cardinalis)
18 Arizona honeysuckle (Euangiosperms: Eudicots: Asteridae: Euasterids II:
Caprifoliaceae; Lonicera arizonica)

is ornithophily a synapomorphy for the entire order, with nonornithophi-

lous species within the order having secondarily lost the bird-pollination
mode of reproduction? In other cases convergent evolution has clearly
occurred, as with the species-rich eudicot family Asteraceae that pos-
sesses a single South American genus, Mutisia, that is bird-pollinated
(Cronk and Ojeda 2008).
In the Western Hemisphere, ornithophilous flowers are pollinated by
species of hummingbirds (family Trochilidae), in Africa and Asia by
species of sunbirds (family Nectariniidae), and in Australia and New
Zealand by species of honeyeaters (family Meliphagidae)—a striking
ecological convergence that we shall examine in more detail in the next
chapter. Regardless of the particular pollinating bird species, ornithophi-
lous flowers around the world are remarkably convergent in form and
color. The nectar-producing capabilities of the flowers are easily under-
standable, as this is the food reward system that the plants use to attract
nectar-feeding birds. But why are the flowers all red? Curiously, the red
color is thought to be as much for bee deterrence as much as it is for
bird attraction (Cronk and Ojeda 2008). Bees see color up to light wave-
lengths of 550 nanometers, whereas birds can see colors that extend to
wavelengths of 660 nanometers, much further into the red end of the
electromagnetic spectrum. The red color of many ornithophilous flowers
has a wavelength of 585 nanometers, outside the visual range of bees but
clearly visible to birds (Cronk and Ojeda 2008).
Other aspects of ornithophilous flower morphology also act to deter
bees. Hummingbirds with long, curved bills can easily drink nectar from
trumpet- or tubular-shaped flowers, whereas bees cannot. Since hum-
mingbirds feed while hovering, hummingbird flowers typically have no
landing platforms that bees might use. In addition, most birds have a very
poor sense of smell (excepting carrion hunters, such as vultures), and so
ornithophilous flowers have no scent that might attract unwanted insects.
In summary, convergent fertilization systems are rife in the seed plants.
Even such highly specialized systems as cantharophily, sapromyiophily,
and ornithophily have been evolved over and over again in independent
plant lineages. There appear to be a limited number of successful fertil-
ization systems, and plant evolution has been constrained to reusing
these systems repeatedly in independent plant lineages.
While there are a limited number of ways to achieve fertilization in
spermatophytes, there also appear to be a limited number of ways to
disperse the product of fertilization: the seed. Species whose members
are geographically widespread are more resistant to extinction than
126 Chapter 3
geographically endemic, locally restricted species. Whereas animal

species can expand their geographic ranges by simply walking or flying
to new territories, plants cannot. Thus, the sessile nature of plants pre-
sents two reproductive problems, not just one: the problem of fertiliza-
tion in organisms that cannot move about in search for a mate, which we
have just considered, and the problem of attaining geographic dispersal
for the plant’s offspring. Spermatophytes have evolved two pathways to
try to obtain geographic dispersal for their fertile seeds: physical seed-
dispersal systems and the evolution of fruit, in which the plant once again
uses an attract-and-reward system to elicit the help of a mobile partner
in carrying the plant’s offspring away to other territories.
Just as many plants use wind for pollination, they also depend upon
wind for dispersal of their seeds. Some of these plants have evolved seeds
with fine, filamentous tufts with much larger surface areas than those of
the seed; these tufts can catch breezes which lift the seed up into the air
to be carried away by the wind (Loewer 1995; Stewart 2009). Tufted seeds
have independently appeared in both the rosid and asterid eudicot
clades, and are perhaps most familiar in the cottonwoods and common
dandelion (table 3.15). Other plants have seeds that develop wings—
enlarged surface-area structures that extend away from the seed cen-
ter—that can not only catch breezes to be lifted up into the air, but can
also maintain themselves in the air for considerable distances due to the
helicopter-lift effect of their spinning wings even if the initial breeze has
died. Winged seeds have been independently developed throughout the
clade of seed plants, from nonangiosperm spermatophytes like pino-
phytes to plesiomorphic angiosperm spermatophytes like magnoliids and
to highly derived asterid eudicots (table 3.15). The spinning seeds of
maple trees are perhaps the most familiar example of this type of seed-
dispersal system in eastern North America. However, the East Indian
cucurbit vine, which aptly bears the species name macrocarpa, has seeds
with wings that are 15 centimeters in total length, and the falling seeds
descend in a spiral path with a 7-meter diameter (Loewer 1995), a feat
our most massive maples cannot duplicate.
An alternative physical-dispersal system that does not use wind is the
development of seeds that are capable of hitching a ride on mobile
animals. In this case the animal is not rewarded for carrying the seed
away, as often the animal is not even aware that a seed has attached itself
to its body. I have termed this type of seed-dispersal system the “Velcro
system,” as it is a historical fact that the idea for the creation of Velcro
came from the observation of the adhesive capabilities of the thistle-
shaped burrs of the greater burdock, Arctium lappa (Stewart 2009).

Many different plant lineages, from the monocotyledon foxtails and
needle grasses to highly derived asterid devil’s claws, beggar’s lice, and
cockleburs have independently evolved seeds with tiny barbs or hooks
that attached themselves to moving animals (table 3.15).
A last physical-dispersal system is to produce seed containers that
actually explode, sending the seeds of the plant rocketing off into the
distance. The eastern dwarf mistletoe is capable of blasting its seeds away
at the velocity of 90 kilometers an hour, and gorse seedpods explode with
a noise like that of a gunshot (Stewart 2009). The wild geranium can
throw its seeds some 3 meters away; squirting cucumbers and Indian
sword beans can propel their seeds up to 6 meters away; exploding witch
hazel seedpods send their seeds up to 10 meters away; and the exploding
fruit of the sandbox tree is able to propel its seeds up to 100 meters into
the distance (Loewer 1995; Stewart 2009). The exploding-seed-container
dispersal system has been developed multiple times independently
throughout the eudicot clade (table 3.15). The list of convergent species
given in table 3.15 is not exhaustive, but was chosen to illustrate the
widespread phylogenetic distribution of independent origins of the same
dispersal system across the spermatophyte clade. A rigorous analysis of
the phylogenetic distribution of all physical-dispersal systems in sper-
matophytes has yet to be conducted. However, such analyses have been
conducted, at least in part, for those plants that have evolved an alterna-
tive dispersal system: fleshy fruit.
By far the most familiar angiosperm seed-dispersal system is the evo-
lution of fruit, in which the mobile animal partner is enticed to eat the
sweet fruit and swallow the enclosed seeds, which are defecated by the
animal much later, hopefully (from the plant’s point of view) to then
sprout and grow at some distance removed from the original fruit-
bearing plant. Fleshy fruit and berry structures have convergently and
independently evolved in so many plant lineages that it will take many
detailed systematic analyses to sort out the many separate phylogenetic
originations of the fruit habit. The fruit habit is also not an exclusively
angiosperm trait: the odiferous fruit of the nonangiosperm spermato-
phyte Ginkgo biloba is familiar to many, as are the aromatic “berries” of
the common juniper, a nonangiosperm pinophyte spermatophyte.
Commercially produced fruit that are popular with human consumers
have convergently evolved in many branches of the eudicot clade. The
highly derived Eurosid-I family Rosaceae in particular contains many
fruit-bearing species, giving us apples, plums, peaches, pears, and cherries.
128 Chapter 3
Table 3.15
Convergent evolution of physical seed-dispersal systems in plants
1 Convergent structure and function: TUFTED SEEDS (fine, filamentous, parachute-like

tufts for geographic dispersal of seeds by wind)
1.1 Fireweed (Euangiosperms: Eudicots: Rosidae: Myrtales: Onagraceae; Epilobium
angustifolium)
1.2 Eastern cottonwood (Euangiosperms: Eudicots: Rosidae: Eurosids I: Salicaceae;
Populus deltoides)
1.3 Common milkweed (Euangiosperms: Eudicots: Asteridae: Euasterids I:
Asclepiadaceae; Asclepias syriaca)
1.4 Dogbane (Euangiosperms: Eudicots: Asteridae: Euasterids I: Apocynaceae;
Apocynum cannabinum)
1.5 Common dandelion (Euangiosperms: Eudicots: Asteridae: Euasterids II: Asteraceae;
Taraxacum officinale)
2 Convergent structure and function: WINGED SEEDS (helicopter-like wings for
geographic dispersal of seeds by wind)
2.1 Eastern white pine (Euphyllophyta: Lignophyta: Spermatophyta: Pinophyta:
Pinaceae; Pinus strobus)
2.2 Yellow poplar (Spermatophyta: Angiospermae: Euangiosperms: Magnoliidae:
Magnoliales: Magnoliaceae; Liriodendron tulipifera)
2.3 Curly dock weed (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Polygonaceae; Rumex crispus)
2.4 Winged elm (Euangiosperms: Eudicots: Rosidae: Eurosids I: Ulmaceae; Ulmus
alata)
2.5 East Indian cucurbit vine (Euangiosperms: Eudicots: Rosidae: Eurosids I:
Cucurbitaceae; Macrozanonia macrocarpa)
2.6 Silver maple (Euangiosperms: Eudicots: Rosidae: Eurosids II: Sapindaceae; Acer
saccharinum)
2.7 Alianthus (Euangiosperms: Eudicots: Rosidae: Eurosids II: Simaroubaceae;
Alianthus altissima)
2.8 Common hoptree (Euangiosperms: Eudicots: Rosidae: Eurosids II: Rutaceae;
Ptelea trifoliata)
2.9 Alpine rockcress (Euangiosperms: Eudicots: Rosidae: Eurosids II: Brassicaceae;
Arabis alpina)
2.10 Horseradish tree (Euangiosperms: Eudicots: Rosidae: Eurosids II: Moringaceae;
Moringa pterygosperma)
2.11 White ash (Euangiosperms: Eudicots: Asteridae: Euasterids I: Oleaceae; Fraxinus
americana)
2.12 Princess tree (Euangiosperms: Eudicots: Asteridae: Euasterids I: Scrophulariaceae;
Paulownia tomentosa)
3 Convergent structure and function: VELCRO SEEDS (geographic dispersal of seeds by
adhesion to mobile animals)
3.1 Foxtail (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Poaceae; Hordeum murinum)
3.2 Sand-tick trefoil (Euangiosperms: Eudicots: Rosidae: Eurosids I: Fabaceae;
Desmodium lineatum)
3.3 Devil’s claw (Euangiosperms: Eudicots: Asteridae: Euasterids I: Pedaliaceae;
Harpagophytum procumbens)
3.4 Beggar’s lice (Euangiosperms: Eudicots: Asteridae: Euasterids I: Boraginaceae;
Cynoglossum officinale)
Table 3.15
(continued)
3.5 Unicorn plant (Euangiosperms: Eudicots: Asteridae: Euasterids I: Martyniaceae;

Proboscidea louisianaca)
3.6 Queen Anne’s lace (Euangiosperms: Eudicots: Asteridae: Euasterids II: Apiaceae;
Daucus carota)
3.7 Cocklebur (Euangiosperms: Eudicots: Asteridae: Euasterids II: Asteraceae;
Xanthium strumarium)
4 Convergent structure and function: EXPLODING SEED CONTAINERS (exploding
seed pods or fruit for geographic dispersal of seeds)
4.1 Eastern dwarf mistletoe (Euangiosperms: Eudicots: Core eudicots: Santalales:
Santalaceae; Arceuthobium pusillum)
4.2 Witch hazel (Euangiosperms: Eudicots: Core eudicots: Saxifragales: Hamamelidaceae;
Hamamelis virginiana)
4.3 Wild geranium (Euangiosperms: Eudicots: Rosidae: Geraniales: Geraniaceae;
Geranium maculatum)
4.4 Sandbox tree (Euangiosperms: Eudicots: Rosidae: Eurosids I: Euphorbiacea; Hura
crepitans)
4.5 Gorse (Euangiosperms: Eudicots: Rosidae: Eurosids I: Fabaceae; Ulex europaeus)
4.6 Squirting cucumber (Euangiosperms: Eudicots: Rosidae: Eurosids I: Cucurbitaceae;
Ecballium elaterium)
4.7 Creeping woodsorrel (Euangiosperms: Eudicots: Rosidae: Eurosids I: Oxalidaceae;
Oxalis corniculata)
4.8 Pale touch-me-not (Euangiosperms: Eudicots: Asteridae: Ericales: Balsaminaceae;
Impatiens pallida)
But melons are produced by a different Eurosid-I family, Cucurbitaceae,

and grapes are even more distantly placed, in the basal rosid family
Vitaceae. Tart persimmons come from a much more distantly related
family in the asterid clade (Ebenaceae), as do our popular tomatoes and
eggplants (Solanaceae). A modern molecular phylogeny for the family
Brassicaceae in the eudicot Eurosid-II clade has revealed that fruit has
evolved three separate times independently in this single family alone
(Mummenhoff et al. 2008).
While sorting out the many convergent occurrences of the fruit habit
in the huge clade of the eudicots is going to take some time, rigorous
phylogenetic analyses of convergent fruit evolution have been com-
pleted for the clade of the monocotyledons (Givnish et al. 2005). Fleshy
fruits have independently evolved in 22 separate lineages within the
monocolyledons (table 3.16). In table 3.16 I have listed basal or near-
basal species for each independent evolution of fruit in the cladogram
of Givnish et al. (2005); subsequent evolution in some of these clades
have produced numerous fruit-bearing species (such as the clade of
130 Chapter 3
Table 3.16
Convergent evolution of fleshy fruit in monocotyledon angiosperms
Convergent structure and function: FLESHY FRUIT (geographic dispersal of seeds by

fruit-eating animals)
1 Settler’s flax (Euangiosperms: Monocotyledons: Alismatales: Araceae; Gymnostachys
anceps)
2 White bat flower (Euangiosperms: Monocotyledons: Eumonocotyledons: Dioscoreales:
Dioscoreaceae; Tacca integrifolia)
3 Screwpine (Euangiosperms: Monocotyledons: Eumonocotyledons: Pandanales:
Pandanaceae; Pandanus tectorius)
4 Giant lily (Euangiosperms: Monocotyledons: Eumonocotyledons: Liliales: Liliaceae;
Cardiocrinum giganteum)
5 Supplejack vine fruit (Euangiosperms: Monocotyledons: Eumonocotyledons: Liliales:
Ripogonaceae; Ripogonum scandens)
6 Night heron (Euangiosperms: Monocotyledons: Eumonocotyledons: Liliales:
Colchicaceae; Disporum cantoniense)
7 Large-flowered trillium (Euangiosperms: Monocotyledons: Eumonocotyledons:
Liliales: Melanthiaceae; Trillium grandiflorum)
8 Forest smilax (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Agavaceae; Behnia reticulata)
9 Smooth Solomon’s seal (Euangiosperms: Monocotyledons: Eumonocotyledons:
Asparagales: Ruscaceae; Polygonatum biflorum)
10 Hyacinthina (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Amaryllidaceae; Griffinia hyacinthina)
11 Cardwell lily (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Amaryllidaceae; Proiphys amboinensis)
12 Scrambling lily (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Hemerocalidaceae; Geitonoplesium cymosum)
13 Palm grass (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Hyperoxidaceae; Curculigo capitulata)
14 Tasmanian Christmas bells (Euangiosperms: Monocotyledons: Eumonocotyledons:
Asparagales: Blandfordiaceae; Blandfordia punicea)
15 Neuwiedia orchid (Euangiosperms: Monocotyledons: Eumonocotyledons:
Asparagales: Orchidaceae; Neuwiedia veratrifolia)
16 ‘Ohe (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Poales: Joinvilleaceae; Joinvillea ascendens)
17 Whip vine (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Poales: Flagelliariaceae; Flagellaria indica)
18 Brocchinia bromeliad (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Poales: Bromeliaceae; Brocchinia reducta)
19 Rattan palm (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Arecales: Arecaceae; Calamus laoensis)
20 Malaysian hanguana (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Commelinales: Hanguanaceae; Hanguana malayana)
21 Amischolotype commelinid (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Commelinales: Commelinaceae; Amischolotype hispida)
22 Wild ginger (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Zingiberales: Zingiberaceae; Siphonochilus aethiopicus)

the Arecaceae), while others have not (such as the clade of the
Ruscaceae).
Interestingly, Givnish et al. (2005) also have demonstrated that former
fruit-bearing species have secondarily lost the fruit habit 11 separate
times within the clade of the monocotyledons. Their careful ecological
analyses of the phylogenetic distribution of fruit occurrence within the
monocotyledons reveals that species that have invaded shaded habitats
have often subsequently evolved fleshy fruit, whereas species possessing
fruit have often lost the fruit habit if those species returned to open
habitats. Apparently animal dispersal of seeds is needed for plants
growing under closed canopies, whereas in open habitats less costly
mechanisms of wind dispersal seem to suffice. Up to 95 percent of woody
understory species in neotropical rain forests have evolved fleshy fruits,
a phenomenon that Givnish et al. (2005, 1481) have termed “concerted
convergence”; they point out that this pattern is among “the strongest
ever demonstrated for evolutionary convergence in individual traits and
the predictability of evolution.”
Even more specialized than the evolution of fruit and subsequent seed
dispersal by vertebrate animals is the evolution of seed dispersal by ants,
or myrmecochory (Dunn et al. 2007). Myrmecochorous plants have
evolved elaiosome, a specialized lipid-rich structure that ants use as a
handle in carrying the seed from the plant back to their nest. At the nest,
the ants remove the elaiosome (which is the plant’s reward to the ants),
and the seed is discarded in a midden or outside of the nest itself. The
elaiosome has clearly been convergently evolved by numerous different
plants, as the structure has multiple developmental origins (Dunn et al.
2007).
As phylogenetic relationships within the clade of the monocotyledons
have now been resolved (Givnish et al. 2005), Dunn et al. (2007) have
used this phylogeny to study the distribution of elaiosome origins. Myr-
mecochory has independently evolved 19 separate times within the clade
of the monocotyledons (table 3.17). In table 3.17 I have listed basal or
near-basal species for each independent evolution of elaiosomes in the
cladogram of Dunn et al. (2007). Unlike monocotyledons that have con-
vergently evolved fruit (table 3.16), Dunn et al. (2007) have demon-
strated that the evolution of myrmecochory in monocotyledons is
independent of whether habitats are shaded or open, and hence inde-
pendent of the evolution of fleshy fruits themselves. Rather, the evolu-
tion of myrmecochory in plants appears to be directly linked to the
evolutionary diversification and increase in population sizes of ants
132 Chapter 3
Table 3.17
Convergent evolution of myrmecochory in monocotyledon flowering plants
Convergent structure and function: MYRMECOCHORY (geographic dispersal of seeds

by ants)
1 Large-flower bellwort (Euangiosperms: Monocotyledons: Eumonocotyledons:
Liliales: Colchicaceae; Uvularia grandiflora)
2 Men-in-a-boat (Euangiosperms: Monocotyledons: Eumonocotyledons: Liliales:
Colchicaceae; Androcymbium striatum)
3 Dogtooth violet (Euangiosperms: Monocotyledons: Eumonocotyledons: Liliales:
Liliaceae; Erythronium americanum)
4 Yellow Star-of-Bethlehem (Euangiosperms: Monocotyledons: Eumonocotyledons:
Liliales: Liliaceae; Gagea lutea)
5 Bigelow’s adder’s tongue (Euangiosperms: Monocotyledons: Eumonocotyledons:
Liliales: Liliaceae; Scoliopus bigelovii)
6 Large-flowered trillium (Euangiosperms: Monocotyledons: Eumonocotyledons:
Liliales: Melanthiaceae; Trillium grandiflorum)
7 Palm grass (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Hypoxidaceae; Curculigo capitulata)
8 Cyanastrum (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Tecophilaeaceae; Cyanastrum cordifolium)
9 Turbinate hensmannia (Euangiosperms: Monocotyledons: Eumonocotyledons:
Asparagales: Hemerocallidaceae; Hensmannia turbinata)
10 Summer snowflake (Euangiosperms: Monocotyledons: Eumonocotyledons:
Asparagales: Amaryllidaceae; Leucojum aestivum)
11 Yellow iris (Euangiosperms: Monocotyledons: Eumonocotyledons: Asparagales:
Iridaceae; Iris pseudacorus)
12 Hooker’s dasypogon (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Dasypogonaceae; Dasypogon hookeri)
13 Pennsylvania sedge (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Poales: Cyperaceae; Carex pensylvanica)
14 Switchgrass (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Poales: Poaceae; Panicum virgatum)
15 Silver vase bromeliad (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Poales: Bromeliaceae; Aechmea fasciata)
16 Blushing bromeliad (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Poales: Bromeliaceae; Nidularium fulgens)
17 Spiral ginger (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Zingiberales: Costaceae; Costus barbatus)
18 Zebra plant (Euangiosperms: Monocotyledons: Eumonocotyledons: Commelinidae:
Zingiberales: Marantaceae; Calathea zebrina)
19 Dancing girl ginger (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Zingiberales: Zingiberaceae; Globba winitii )

beginning in the Late Eocene, some 37 million years ago. The peak in
the evolution of fleshy fruit in the monocotyledons began earlier, some
85 million years ago, and was triggered by the spread of Late Cretaceous
forests and closed-canopy habitats (Givnish et al. 2005).
Although myrmecochorous plants are equally likely to be found in
either open habitats or closed-canopy habitats, Dunn et al. (2007) do
demonstrate that monocotyledons in open habitats that are fire-prone,
such as in Mediterranean ecosystems, have preferentially evolved myr-
mecochory. They suggest that transport of seeds to ant nests confers the
selective advantage of reducing the risk of seed mortality due to fire, in
addition to the selective advantage of geographic dispersal of the seeds.
This conclusion is supported by the documented convergent evolution
of myrmecochory in the fire-prone shrub lands of both South African
and Australian ecosystems (Milewski and Bond 1982), which provides
an example of ecosystem convergence as well, the subject of the next
chapter of this book.
In summary, the incidences of reproductive convergence in plants, just
as in the animals we considered in the previous chapter, vastly over-
shadow the classic examples of convergent evolution that are usually
offered for both groups. The spectacular convergent evolution of wings
in animals pales in comparison with the astonishing number of indepen-
dent origins of viviparity, and the convergent evolution of towering trees
in plants pales in comparison with the astonishing number of indepen-
dent origins of the fruit habit. I predict the 22 separate convergences on
the fruit habit discovered in the monocotyledons alone (table 3.16) are
going to be a small number indeed when rigorous phylogenetic analyses
are eventually applied to mapping the number of independent origins of
fleshy fruit in the huge clade of the eudicots.
In many ways, plant evolution is more complex and more interesting
than animal evolution because of its overtly multispecies aspect. Plants
coevolve with the numerous animal partners that they have involved in
their reproduction and defense, and these beetles, flies, bees, wasps, but-
terflies, moths, ants, birds, and mammals have themselves evolved in
response to plant evolution. In the next chapter we shall focus on the
phenomenon of convergent evolution at the multispecies level: that is,
the convergent evolution of entire ecosystems.
4 Convergent Ecosystems
In chemistry, the urge of scientists to order and classify natural phenomena

resulted in the well-known periodic table of the elements, which allowed
chemists to predict new elements and their chemical properties. . . . Some ecolo-
gists wonder whether something like a “periodic table of niches” might be
possible.
—Pianka (1978, 267)
One Ecological Role, Many Convergent Players
Imagine a universe in which there are an unlimited number of ways to

make a living. In such a universe, each species would have its own unique
way of making a living, different from all other species. When we examine
the ecological structure of living organisms on Earth, we can clearly see
that we do not inhabit such a universe.
In our universe the number of ways of making a living, of ecological
roles or niches, is demonstrably limited. Multiple species are constrained
in their evolution to playing the same ecological role, filling the same
ecological niche, as best they can given their own phylogenetic and
developmental backgrounds. The ecological role is the same, but the
species players are from many different evolutionary pathways that have
independently converged on filling that niche.
Ecological niche convergence is best recognized by first considering
truly bizarre ways of making a living—ecological roles that are so strange
that, at first glance, it would seem probable that only one species would
have evolved such a restricted pathway. And then one realizes that,
astonishingly, the ecological evolution of life on Earth is so constrained
that multiple species have converged on that pathway.
One such bizarre pathway is that of plants that eat animals—an eco-
logical role so strange that it fascinated Darwin (1875). Carnivorous
plants have evolved adaptations to actively trap, kill, and digest animals.
136 Chapter 4
Although such adaptations seem so improbable that one might think

they could have evolved only once in the plants, Darwin’s (1875) own
researches convinced him that the carnivorous plants were polyphyletic,
and led him to believe that plants had independently evolved carnivory
in three separate lineages. Over a century and a quarter later, modern
phylogenetic analyses reveal that carnivory has arisen independently in
at least six separate plant clades: within the commelinid monocotyledons,
the caryophyllalean core eudicots, the eurosids-I eudicots, the ericalean-
asterid eudicots, and at least twice in the eurasterids-I eudicots (Albert
et al. 1992; Cameron et al. 2002; Ellison and Gotelli 2009).
Not only is the trait of carnivory convergent in plants, but the means
of carnivory itself is also convergent, in that independent plant lineages
have evolved the same mechanism for it. Although six separate plant
clades have evolved carnivory, carnivorous plants use only three differ-
ent types of animal traps: flypaper traps, pitcher traps, and mechanical
traps (table 4.1). Carnivorous plants with flypaper traps have leaves
or hairs that ooze sticky droplets to capture animals that touch them.
Flypaper-type traps have convergently evolved at least five separate
times (table 4.1), and it is possible that they have independently arisen
more than once in the clade of the Caryophyllales. Within the carnivo-
rous Caryophyllales, the sundews (Droseraceae) are basal; thus, at first
one might conclude that flypaper traps are a synapomorphy for all plants
in this clade with that trait, and so I have listed its evolution only once
in table 4.1. However, higher in the clade, the dewy pines (Drosophyl-
laceae) and the African lianas (Triphyophyllum peltatum, Dioncophyl-
laceae) may have reevolved flypaper traps a second time, in that the
Nepenthaceae are below these two families in the clade of the Caryo-
phyllales, and the Nepenthaceae do not have flypaper-type traps (Ellison
and Gotelli 2009).
Carnivorous plants with pitcher traps have tubular leaves or flowers
with a pool of digestive fluids at the base of the tube. Animals that enter
the plant’s tubes fall into the cup at its base to drown and be consumed.
In some cases the plant’s tubes have downward-pointing hairs or spines
that prevent the prey from trying to climb back out of the tube structure
to escape. Pitcher traps have convergently evolved at least four separate
times (table 4.1). In some plant lineages, pitcher traps are developmental
modifications of preexisting flypaper traps (Asian pitcher plants in the
Caryophyllales), whereas in other lineages pitcher traps have evolved
independently of any lineage possessing flypaper traps (Australian
pitcher plants, Brocchinia bromeliads). While insects are the prey of
Convergent Ecosystems 137
Table 4.1
Convergent evolution of carnivory in plants
Convergent ecological role: CARNIVOROUS PLANTS (“Venus flytrap niche”)

1 Convergent lineages with “flypaper” traps:
1.1 Sundew (Euangiosperms: Eudicots: Core eudicots: Caryophyllales: Droseraceae;
Drosera intermedia)
1.2 Roridula (Euangiosperms: Eudicots: Asteridae: Ericales: Roridulaceae; Roridula
gorgonias)
1.3 Devil’s claw (Euangiosperms: Eudicots: Asteridae: Euasterids I: Martyniaceae;
Ibicella lutea)
1.4 Rainbow plant (Euangiosperms: Eudicots: Asteridae: Euasterids I: Byblidaceae;
Byblis gigantea)
1.5 Butterwort (Euangiosperms: Eudicots: Asteridae: Euasterids I: Lentibulariaceae;
Pinguicula vulgaris)
2 Convergent lineages with “pitcher” traps:
2.1 Brocchinia bromeliad (Euangiosperms: Monocotyledons: Eumonocotyledons:
Commelinidae: Bromeliaceae; Brocchinia reducta)
2.2 Asian pitcher plant (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Nepenthaceae; Nepenthes truncata)
2.3 Australian pitcher plant (Euangiosperms: Eudicots: Rosidae: Eurosids I:
Cephalotaceae; Cephalotus follicularis)
2.4 American pitcher plant (Euangiosperms: Eudicots: Asteridae: Ericales:
Sarraceniaceae; Sarracenia purpurea)
3 Convergent lineages with “mechanical” traps:
3.1 Venus flytrap (Euangiosperms: Eudicots: Core eudicots: Caryophyllales:
Droseraceae; Dionaea muscipula)
3.2 Bladderwort (Euangiosperms: Eudicots: Asteridae: Euasterids I: Lentibulariaceae;
Utricularia inflata)
3.3 Corkscrew plant (Euangiosperms: Eudicots:Asteridae: Euasterids I: Lentibulariaceae;
Genlisea margaretae)
choice for most carnivorous plants, species of the pitcher-type carnivores

are known to trap and eat animals as large as mice (Stewart 2009).
The most complicated of carnivorous-plant capture mechanisms are
the mechanical traps, the most famous of which is the snap trap of the
Venus flytrap (table 4.1). These snap traps are highly modified leaves
with projecting antennae-like spikes around their perimeters. Animals
alighting on the leaf, or touching the spikes in approaching the leaf,
trigger the snap and the leaf folds shut, trapping the animal inside, with
a rapidity that is the fastest movement known to exist in plants (Cameron
et al. 2002). Darwin (1875) argued that the snap traps of the Venus
flytrap, Dionaea muscipula, and the waterwheel, Aldrovanda vesiculosa,
were independently evolved. Ecologically the two species are very dif-
ferent: the Venus flytrap is a terrestrial plant whereas the waterwheel is
fully aquatic and does not possess roots; in addition, the Venus flytrap
138 Chapter 4
captures flying animals whereas the waterwheel captures underwater

swimming animals. However, Darwin’s (1875) hypothesis of the conver-
gent evolution of snap traps has been disproved by modern phylogenetic
molecular analyses, which reveal that the Venus flytrap and the water-
wheel are sister taxa, and that the snap trap is a synapomorphy, thus not
convergent (Cameron et al. 2002; Ellison and Gotelli 2009).
The bladderworts are also aquatic, but rather than having snap-trap
leaves like the waterwheel, they have modified their leaves into bladder-
like structures with a snap-trigger lid. When triggered, these bladders
develop an internal negative pressure that is capable of sucking swim-
ming animals into the bladder and closing the lid, trapping the animal
inside to be digested. The Venus flytrap usually eats only insects, but the
bladderworts can trap and eat animals as large as tadpoles, and they can
reset their traps every thirty minutes, making them one of the most vora-
cious of the carnivorous plants (Stewart 2009).
The corkscrew plants have modified some of their leaves into geomet-
ric traps, twin corkscrews that meet at a feeding pit containing digestive
fluid. These corkscrewed leaves are actually developed underground, in
the place of roots, and primarily capture bacteria. Upon entering the
corkscrew, the prey organism is unable to leave because inward-pointing
hairs and the geometry of the screw prevent backward movement, and
the prey can only move forward to its death in the feeding pit.
The bladderworts and the corkscrew plants both belong to the
family of the Lentibulariaceae (table 4.1), yet the nature of their me-
chanical leaf traps is so different that it is believed that the two types
of trap were independently developed; that is, neither plant is in an
ancestor-descendant relationship to the other, and their different traps
represent separate modifications of the flypaper-type trap of the basal-
lentibulariacean butterworts (Ellison and Gotelli 2009).
What is convergent in all of these plants is the ecological role of car-
nivory, rather than the overall morphology of the plants (as in chapter
3). There is an underlying ecological trigger, a selective advantage, that
is found in most carnivorous plants: they are able to invade nitrogen-
poor habitats preferentially over other plants. Most carnivorous plants
are found in boggy or swampy environments where nutrients are in short
supply, and therefore these plants have evolved a new way to acquire
additional nutrients, namely by eating animals. Nitrogen, in particular, is
preferentially allocated to reproductive structures in many carnivorous
plants (Ellison and Gotelli 2009), triggering the differential reproductive
process of natural selection in their favor.
Plants by definition are multicellular photoautotrophs—that is, they

produce their own food by the process of photosynthesis. Even the car-
nivorous plants still have green leaves and photosynthesis, so they have
not fully abandoned photoautotrophy. Still other plants have taken
another bizarre evolutionary pathway: they have become heterotrophic,
like animals. Such plants have drastically reduced their reliance on pho-
tosynthesis, and some have eventually lost it entirely, becoming parasitic
like many animal species or saprophytic like many fungal species, who
are close relatives to the animals. These are plants that, in essence, eat
other plants.
Perhaps the most notorious of the parasitic plants is the dodder vine
(table 4.2), a parasite that one botanist describes as resembling “an alien
life form that has come to suck the life out of [E]arth’s vegetation”
(Stewart 2009, 147). Dodder vines lack the green color of chlorophyll,
and instead are eerie shades of orange, red, and yellow. Dodder leaves
have been reduced to near-microscopic scalelike structures, such that
the vines appear to be naked, brightly colored strings smothering the
plant that is being parasitized. The dodder vine is a highly derived asterid
eudicot, yet the very distantly related, near-basal-Euangiosperms clade
of the magnoliids has convergently evolved its ecological equivalent in
the macabre-named “love vine,” although its embrace is just as unloving
as that of the dodder, at least from the point of view of the parasitized
plant (table 4.2). All in all, at least seven independent phylogenetic
Table 4.2
Convergent evolution of heterotrophy in plants
Convergent ecological role: PARASITIC AND SAPROPHYTIC PLANTS (“dodder

niche”)
1 Love vine (Euangiosperms: Magnoliidae: Laurales: Lauraceae; Cassytha filiformis)
2 Coral root orchid (Euangiosperms: Monocotyledons: Eumonocotyledons:
Asparagales: Orchidaceae; Corallorhiza maculata)
3 Eastern dwarf mistletoe (Euangiosperms: Eudicots: Core eudicots: Santalales:
Santalaceae; Arceuthobium pusillum)
arnoldii)
5 Indian pipe (Euangiosperms: Eudicots: Asteridae: Ericales: Ericaceae; Monotropa
uniflora)
6 Witch weed (Euangiosperms: Eudicots: Asteridae: Euasterids I: Scrophulariaceae;
Striga asiatica)
7 Dodder (Euangiosperms: Eudicots: Asteridae: Euasterids I: Convolvulaceae; Cuscuta
epithymum)

140 Chapter 4
lineages of plants have converged upon the ecological role of heterotro-

phy, the most aptly named of which is perhaps the “witch weed”
(table 4.2).
Striking and seemingly improbable examples of ecological niche con-
vergence are not limited to the plants. Consider the ecological role of
insectivory. Insects are an abundant source of food on land, and so it is
not surprising that many animals have evolved insectivorous modes of
life. Ecological convergence exists at all levels and types of insectivory,
such as the fairly restricted ecological role of eating-insects-on-the-wing,
in which the insectivore captures and eats insects in the air without
pausing from insect to insect. Dragonflies, which are themselves insects
(Odonata), avian swifts (Apodidae), and mammalian bats (Microchirop-
tera), to name only a few separate phylogenetic lineages, are all species
that have converged on filling this same ecological niche.
Let us consider an extremely specialized form of insectivory: eating
insects found only in tree bark. There exists a highly specialized group
of species that have evolved adaptations to fill this ecological niche: the
avian woodpeckers (Picidae). These birds have evolved long, sharp,
pointed beaks for boring and digging into tree bark, and very long
tongues for extracting exposed insects. The jaws and cranium of the
woodpecker are structured in a fashion that cushions the brain of the
bird as it hammers away, preventing concussion. In the foot of the wood-
pecker, the fourth digit has been rotated backward, joining the first digit,
such that the bird has feet with two toes pointing forward and two back-
ward. These clawed feet, combined with the stiffened tail feathers of the
woodpecker, allow the bird to firmly anchor itself against a tree and to
forcefully hammer the bark of the tree without becoming dislodged from
its triangular base of two grasping feet and supporting prop of tail
feathers.
Thus, the woodpeckers are unique, highly specialized animals that
have evolved to fill an extremely specialized and restricted ecological
role, yes? Not so. They are not unique, in that no less than six additional
phylogenetic lineages of animals have converged on filling the ecological
niche of eating wood-boring insects (table 4.3). Three of these lineages
are avian, but they are very different types of birds from woodpeckers,
with very different adaptations. Both the akiapola’au honeycreepers of
Hawaii and the sickle-billed vangas of Madagascar extract insects from
tree bark with their bills, but their bills are not like those of woodpeckers.
The akiapola’au honeycreeper has a thick lower beak that is only half as
long as its curved upper beak. It drills and digs into tree bark with its
Table 4.3
Convergent evolution of woodpecker ecological equivalents in animals
Convergent ecological role: EATERS OF WOOD-BORING INSECTS (“woodpecker

niche”)
1 Red-headed woodpecker (Amniota: Sauropsida: Archosauromorpha: Dinosauria:
Saurischia: Theropoda: Maniraptora: Aves: Neognathae: Neoaves: Piciformes; Picidae;
Melanerpes erythrocephalus)
2 Akiapola’au honeycreeper (Aves: Neognathae: Neoaves: Passeriformes: Drepanididae;
Hemignathus munroi)
3 Sickle-billed vanga (Aves: Neognathae: Neoaves: Passeriformes: Vangidae; Falculea
palliata)
4 Woodpecker finch (Aves: Neognathae: Neoaves: Passeriformes: Emberizidae;
Cactospiza pallida)
5 Striped possum (Amniota: Synapsida: Therapsida: Mammalia: Marsupialia:
Diprotodontia: Petauridae; Dactylopsila trivirgata)
6 Apatemyid (Mammalia: Eutheria: Boreoeutheria: Apatemyidae; Apatemys chardini
†Paleogene)
7 Aye-aye (Eutheria: Euarchontoglires: Primates: Lemuriformes: Daubentoniidae;
Daubentonia madagascariensis)
lower beak, and then extracts insects from the bark with the hooked
upper beak. As their name implies, the sickle-billed vangas have sickle-
shaped beaks, which they insert into bark cracks and openings; they then
use a lever action to pull out the insect.
The woodpecker finches of the Galapagos have evolved a totally
different approach to extracting wood-boring insects—they use tools.
These birds use their beaks, not to drill with, but to hold a long, sharp
tool, usually a cactus spine. They insert the cactus spine into bark
cracks and openings, and also use a lever action to pull out insects, but
using the tool rather than a specialized beak morphology. We shall con-
sider the convergent evolution of tool usage by animals in detail in
chapter 6.
Not only did tool usage and three distinctly different types of beak
morphologies evolve in independent bird lineages to fill the same eco-
logical niche, but no less than three separate lineages of mammals have
also converged on the same restricted ecological role of eating wood-
boring insects (table 4.3). All three of these mammal groups use very
elongated fingers, tipped with claws, to insert into tree bark and to pull
out enclosed insects. The particular fingers used, however, are different
in each group. The striped possum of New Guinea, a marsupial mammal,
has evolved a very elongated fourth digit (the ring finger in humans) for
this function, whereas the aye-aye of Madagascar, a placental mammal,
142 Chapter 4
has evolved an elongated third digit (the middle finger in humans) for
the same function. These two groups of mammals are alive today, but
there existed a third group, the extinct apatemyids, that also converged
on this same ecological role (table 4.3). The apatemyids evolved two
elongated fingers rather than one, the second digit (the pointing finger
in humans) and the third digit, for use in extracting insects from tree
bark (Koenigswald and Schierning 1987).
Again, what is convergent in all of these animals is the ecological role
of eating wood-boring insects, not the morphologies of the animals as
previously considered in detail in chapter 2. Birds, bats, pterosaurs, and
butterflies have all convergently evolved a morphologically similar struc-
ture, namely wings, to accomplish the same function: powered flight (see
chapter 2). Here, greatly different morphological structures and behav-
iors have evolved in independent phylogenetic lineages in order to con-
vergently fill the same ecological niche: eating wood-boring insects. A
bat superficially looks like a bird, but a red-headed woodpecker looks
very different from a striped possum. Morphologically, these animals are
very different, but ecologically they are equivalents.
Consider yet another highly specialized form of insectivory: attacking
and eating ants in their nests. Huge numbers of ants can inhabit an ant
colony, and they are highly aggressive in defending that colony. Thus,
attacking an ant nest is not a trivial task, and one might conclude that
only one type of highly specialized animal would fill the niche of such a
restrictive mode of food acquisition. But within the mammals alone, five
separate groups have converged on myrmecophagy, three times indepen-
dently in the placental mammals, once in the marsupial mammals, and
once in the monotreme mammals (table 4.4). Interestingly, in the placen-
tal mammals both morphological convergence and ecological conver-
gence have occurred, in that the South American giant anteater, the
Asian pangolin, and the African aardvark have all evolved long claws
and strong forelimbs for tearing into ant nests. All three groups of animals
have evolved tubular-shaped mouths that contain a very long, sticky
tongue for capturing ants. The anteaters and pangolins have totally lost
their teeth, and the aardvark retains only a few teeth. All three groups
have highly developed salivary glands, which assist in swallowing copious
amounts of ants, and all three have convergently evolved strongly mus-
cular pyloric regions in their stomachs to assist in digesting ants (Lecoin-
tre and Le Guyader 2006). Yet these animal groups are only very distantly
related: the anteaters are xenarthrans, basal eutherians; the aardvarks
are afrotherians; and the pangolins are laurasiatherians (table 4.4).
Table 4.4
Convergent evolution of myrmecophagous ecological equivalents in animals
Convergent ecological role: ANT-EATING ANIMALS (“anteater niche”)

1 Thorny devil lizard (Amniota: Sauropsida: Diapsida: Lepidosauromorpha: Squamata:
Iguania: Agamidae; Moloch horridus)
2 Desert horned lizard (Lepidosauromorpha: Squamata: Iguania: Phrynosomatidae;
Phrynosoma platyrhinos)
3 “Desert bird” alvarezsaurid (Diapsida: Archosauromorpha: Dinosauria: Saurischia:
Theropoda: Coelurosauria: Maniraptora:Alvarezsauridae; Shuvuuia deserti †Cretaceous)
4 Long-beaked echidna (Amniota: Synapsida: Therapsida: Mammalia: Monotremata:
Tachyglossidae; Zaglossus bruijni)
5 Numbat (Mammalia: Marsupialia: Myrmecobiidae; Myrmecobius fasciatus)
6 Giant anteater (Mammalia: Eutheria: Xenarthra: Myrmecophagidae; Myrmecophaga
tridactyla)
7 Aardvark (Eutheria: Afrotheria: Tubulidentata: Orycteropodidae; Orycteropus afer)
8 Giant pangolin (Eutheria: Laurasiatheria: Pholidota: Manidae; Manis gigantea)
Although today these animal groups are geographically separated

ecological equivalents, the enigmatic fossil species Eurotamandua
joresi, from the Eocene Messel strata in Germany, indicates that the
xenarthrans once extended their range into that of the laurasiatherian
pangolins (Gaudin and Branham 1998).
The marsupial mammals and the plesiomorphic, egg-laying mono-
treme mammals are even more distantly related, but they too have
converged on the ecological role of myrmecophagy with the numbat of
Australia and the long-beaked echidna of New Guinea (table 4.4). Like
the placental mammals, the myrmecophagous echidnas have evolved
toothless, tubular-shaped mouths with long, sticky tongues for capturing
ants (Lecointre and Le Guyader 2006).
Last, the ecological niche of myrmecophagy has also been conver-
gently filled by three independent groups of animals in the clade of
sauropsid amniotes, the reptiles. The desert horned lizards of North
American, phrynosomatid lepidosaurs, and the thorny devil lizards of
Australia, agamid lepidosaurs, have independently evolved myrmecoph-
agy (table 4.4). Both of these lizard groups have reduced their dentition,
a morphological trait convergent on the myrmecophagous mammals,
have expanded their stomachs to assist in digesting large numbers of ants,
and have strikingly similar dorsoventrally flattened bodies covered with
sharp spines (Pianka and Parker 1975). These two lizard groups have also
convergently evolved much quicker prey capture and processing behav-
iors than any other iguanian lizards; with their particularly rapid tongue
144 Chapter 4
protrusion and retraction, the thorny desert lizards can consume up to

2,000 ants per day (Meyers and Herrel 2005).
Myrmecophagy also appears to have evolved within the clade of the
dinosaurs, as evidenced by the peculiar alvarezsaurids of the Cretaceous
(table 4.4). One of the best preserved of these is the “desert bird,”
Shuvuuia deserti, which is not a true bird but a related maniraptoran
theropod. It has powerful forearms and hands with a single finger, tipped
with a very large, hooked claw that is hypothesized to have been used to
rip open ant or termite nests, as the living mammalian anteaters do. Its
snout is also elongated and roughly tubular, but the morphology of the
tongue contained inside is unknown. These Cretaceous animals are enig-
matic in that ants did not become a significant part of the terrestrial
ecosystem until the Eocene, as discussed in chapter 3 in the discussion
of the convergent evolution of myrmecochory in plants (see table 3.17).
Thus, they may not have been exclusively specialized for eating only ants,
but may have also preyed on wood-nesting termites (Longrich and
Currie 2008).
In general for the major heterotrophic organisms, animals are different
types of carnivores or herbivores, ultimately using living organisms as a
source of food, whereas the closely related fungi are saprophytes, feeding
on dead and decaying organisms. However, there exist some animals, the
carrion-eaters, who have converged on the saprophytic, necrophagous
mode of life. The corpse-seeking carrion beetles and hyenas are very
different-looking types of animals, one an arthropod and the other a
mammal, yet they are ecological equivalents (table 4.5). However, the
most spectacular convergence on the ecological role of necrophagy is
exhibited by the birds, where morphological convergence has paralleled
ecological convergence. The Eurasian black vulture looks very similar to
the North American turkey vulture, but the accipitrid Eurasian black
vulture is a member of the clade of the Falconiformes, closely related to
hawks and eagles, whereas the cathartid North American turkey vulture
is a member of the clade of the Ciconiiformes, closely related to the
storks and flamingos (Lecointre and Le Guyader 2006). Nevertheless,
both bird groups have independently evolved featherless, bald heads and
necks that allow them to push their heads into rotting carcasses without
fouling their body feathers. Both bird groups have also evolved extremely
acidic stomachs to protect them from bacterial poisoning in eating
carrion. Even more amazing, both bird groups have independently
evolved similar behaviors: both Eastern Hemisphere and Western Hemi-
sphere vultures aggregate in social flocks in trees, and both have evolved
Table 4.5
Convergent evolution of necrophagous ecological equivalents in animals
Convergent ecological role: CARRION-EATING ANIMALS (“vulture niche”)

1 American carrion beetle (Bilateria: Protostomia: Ecdysozoa: Arthropoda: Hexapoda:
Coleoptera: Staphylinoidea: Silphidae; Necrophila americana)
2 Eurasian black vulture (Bilateria: Deuterostomia: Chordata: Osteichthyes:
Sarcopterygii: Reptiliomorpha: Amniota: Sauropsida: Archosauromorpha: Dinosauria:
Saurischia: Theropoda: Maniraptora: Aves: Neognathae: Neoaves: Falconiformes:
Accipitridae; Aegypius monachus)
3 Turkey vulture (Aves: Neognathae: Neoaves: Ciconiiformes: Cathartidae; Cathartes
aura)
4 Teratorn (Aves: Neognathae: Neoaves: Ciconiiformes: Teratornithidae; Argentavis
magnificens †Miocene)
5 Striped hyena (Amniota: Synapsida: Therapsida: Mammalia: Eutheria: Laurasiatheria:
Carnivora: Feliformia: Hyaenidae; Hyaena hyaena)
the soaring-in-circular-formation flight pattern in searching for rotting

corpses (we shall consider the convergent evolution of behavior in
animals in detail in chapter 6).
The North American ciconiiform cathartid vultures also contain the
largest flying bird species alive on the Earth today, the California condor,
Gymnogyps californianus, with a wingspan of three meters. Yet the cico-
niiform birds apparently converged on the ecological role of necrophagy
at least twice in their evolutionary history (table 4.5): in the Miocene,
some 10 million years ago, they produced another family of gigantic
vultures, the Teratornithidae, with wingspans of 6 meters, the largest
flying birds known in Earth’s history (Benton 2005).
Compared to necrophages, convergent animals who survive by eating
rotting corpses, the ecological role of eating nectar would seem to be a
sweet alternative. Nectarivory is a very highly specialized ecological
niche, one that (at first glance) would not seem to be easy to fill. Nectar
is very sugar-rich, but it is very difficult to survive on a diet of sugar. The
proteins, amino acids, vitamins, and trace minerals necessary for animal
growth and reproduction are present in nectar in extremely low levels,
if at all (Gartrell 2000). The best-known nectarivores, the hummingbirds,
have evolved a suite of adaptations directly related to the difficulty of
surviving on nectar: small body sizes, lowered metabolic rates, lowered
protein requirements, and substantial changes in both their digestive and
renal physiologies (Gartrell 2000). All this is in addition to their evolu-
tion of extremely long beaks and tubular tongues in order to feed from
flowers, and glossy feathers packed tightly to their bodies, an adaptation
146 Chapter 4
believed to prevent their feathers from being soiled by sticky nectar

(Gartrell 2000).
Thus, the hummingbird would seem to be an exquisite but unique
animal highly adapted to a very restrictive ecological role. But it is not
unique—no less than eleven other phylogenetic lineages of animals have
independently converged on the ecological role of nectarivory (table
4.6). The hummingbirds, avian family Trochilidae, are found only in the
Western Hemisphere, in both North and South America (Cronk and
Ojeda 2008). In the tropics of the Eastern Hemisphere, from Africa to
Asia and Australia, a separate group of birds have independently evolved
to fill the nectarivore ecological role: the sunbirds, avian family Nectari-
niidae (table 4.6). In Australia, New Zealand, and Pacific islands as far
away as Hawaii, a third group of birds have convergently filled the nec-
tarivore niche: the honeyeaters, avian family Meliphagidae.
The trochilids, nectariniids, and meliphagids constitute the three
main phylogenetic lineages of birds that have converged on filling the
Table 4.6
Convergent evolution of flying nectarivorous ecological equivalents in animals
Convergent ecological role: FLYING NECTAR FEEDERS (“hummingbird niche”)

1 Hummingbird clearwing moth (Bilateria: Protostomia: Ecdysozoa: Arthropoda:
Hexapoda: Lepidoptera: Sphingidae; Hemaris thysbe)
2 Ruby-throated hummingbird (Bilateria: Deuterostomia: Chordata: Osteichthyes:
Sarcopterygii: Reptiliomorpha: Amniota: Sauropsida: Archosauromorpha: Dinosauria:
Saurischia: Theropoda: Maniraptora: Aves: Neognathae: Neoaves: Apodiformes:
Trochilidae; Archilochus colubris)
3 Ruby-cheeked sunbird (Aves: Neognathae: Neoaves: Passeriformes: Nectariniidae;
Chalcoparia singalensis)
4 Red-throated honeyeater (Aves: Neognathae: Neoaves: Passeriformes: Meliphagidae;
Myzomela eques)
5 Red-legged American honeycreeper (Aves: Neognathae: Neoaves: Passeriformes:
Thraupidae; Cyanerpes cyaneus)
6 ‘I’iwi Hawaiian honeycreeper (Aves: Neognathae: Neoaves: Passeriformes:
Drepanididae; Vestiaria coccinea)
7 Japanese white-eyes (Aves: Neognathae: Neoaves: Passeriformes: Zosteropidae;
Zosterops japonicus)
8 South African sugarbird (Aves: Neognathae: Neoaves: Passeriformes: Promeropidae;
Promerops cafer)
9 Tennessee warbler (Aves: Neognathae: Neoaves: Passeriformes: Parulidae: Vermiuora
peregrina)
10 Swift parrot (Aves: Neognathae: Neoaves: Psittaciformes: Psittacidae; Lathamus
discolor)
11 Mexican long-tongued bat (Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
Laurasiatheria: Chiroptera: Microchiroptera: Phyllostomidae; Choeronycteris mexicana)

nectarivore ecological niche (Cronk and Ojeda 2008). But they are not
alone, for no less than five other families of passeriform birds and one
family of psittaciform birds have convergently evolved nectarivory (table
4.6). In the Western Hemisphere the red-legged American honeycreep-
ers (Thraupidae) and Tennessee warblers (Parulidae) have convergently
evolved nectarivory, and coexist with hummingbirds. In the Eastern
Hemisphere, the South African sugarbirds (Promeropidae) and Japanese
white-eyes (Zosteropidae) have convergently evolved nectarivory, and
coexist with the sunbirds. In Australia the swift parrots (Psittacidae) and
in Hawaii the ‘I’iwi honeycreepers (Drepanididae) have convergently
evolved nectarivory, and coexist with honeyeaters. All of these birds have
independently evolved many of the nectarivorous adaptations seen in
the hummingbirds: lowered metabolic rates and protein requirements in
honeyeaters; glossy feathers and slender, thin beaks in swift parrots;
extensible brush tongues independently in both honeyeaters and swift
parrots, and so on (Gartrell 2000).
Convergence on this extremely specialized way of surviving is not
limited to the birds. In North America, the hummingbird clearwing moth
is an insect that has not merely converged on the ecological role of nec-
tarivory, it is also a hummingbird morphological mimic. This moth has
converged so closely on the shape and color of a hummingbird that it is
often mistaken for one—only when one notices a pair of antennae on
the head of the animal does one suddenly realizes that it is not a bird at
all. And last, even the mammals have evolved nectarivory (table 4.6). The
Mexican long-tongued bat, aptly named as its tongue can be as long as
one-quarter of its body length, has convergently filled the nectarivore
ecological niche even though it looks nothing like a hummingbird.
I will use one last—and major—example of ecological niche conver-
gence to conclude this section of the chapter, and to set the stage for the
next section, in which we will consider the ecological convergence of
entire ecosystems. This is the convergent evolution of ground-dwelling,
flightless birds. Many different phylogenetic lineages of birds have inde-
pendently devolved the ability to fly, losing the characteristic set of traits
that sets them apart from the other theropod dinosaurs (Roff 1994;
Roots 2006). Their wings become reduced and vestigial, with the shorten-
ing and weakening of the humerus, ulna, and radius within the wing.
Their sterna become flat, losing the prominent keel for anchoring power-
ful flight muscles. Flight feathers in the wing and tail are shortened until
they become useless for flight. Finally, the feathers lose their aerody-
namic asymmetry and their smooth aerodynamic contours, which were
148 Chapter 4
maintained by barbules and microscopic hooklets, and become loosely

constructed and hairlike, giving a shaggy appearance to many flightless
birds (Feduccia 1996).
Not all changes in the evolution of flightlessness in birds are trait
losses. Because they no longer fly and no longer need to be lightweight,
constraints on body size are lifted and flightless birds can become quite
large. To adapt to their new life as bipedal ground dwellers, they develop
larger legs and powerful thigh muscles, not only to outrun predators or
to run down prey but also to defend themselves with powerful kicks of
their hind limbs (Roots 2006). As summarized by Feduccia (1996, 289),
“whatever the group of birds, when flightlessness occurs, the same general
morphological features appear time and time again, convergently, pro-
viding a veritable lesson in the rules of evolutionary design.”
From an ecological perspective, what is particularly striking is the
convergent evolution of multiple mammalian ecological equivalents in
flightless birds—these birds do not converge on a single mammalian
ecological role, but rather have evolved to fill a suite of niches usually
occupied by the ground-dwelling mammals (table 4.7). Ground-dwelling
birds have evolved carnivores, insectivores, omnivores, herbivores, and
frugivores that convergently play the mammalian ecological roles of cats,
moles, badgers, ungulates, and tapirs. Other flightless birds have become
swimmers, occupying both marine and freshwater habitats, converging
on the ecological role of pinniped mammals (table 4.7). Ecologically
considered, the penguin is a bird that has convergently evolved into
a seal.
Not long after the destruction of the dinosaurian ecosystem in the
end-Cretaceous mass extinction, a major group of ground-dwelling pred-
atory birds, the phorusrhacids or “terror birds,” evolved in the Paleocene
(Benton 2005). Anatomically, some of these birds are eerily similar to
the Mesozoic dromaeosaurs and troodontids, as discussed in chapter 2
(see table 2.7). Therefore, rather than saying that the birds converged on
the mammalian role of a ground-dwelling felid carnivore in the Ceno-
zoic, it would be more accurate to argue that the felids convergently filled
the ecological niche of the extinct coelurosaurs. Some of the extinct
phorusrhacids were quite large; the typical species of Phorusrhacos were
1.5 meters tall, but the Eocene Phorusrhacos longissimus reached almost
3 meters and the titan phorusrhacid, Titanus walleri, was slightly taller
than 3 meters (Feduccia 1996; Benton 2005). The red-legged seriema of
South America is a surviving cousin of the extinct phorusrhacids, but
this ground-dwelling predator is only about 40 centimeters in height

(Feduccia 1996).
The phorusrhacids survived from the Paleocene to the Pliocene and,
although early members of the group are found in Europe, the family
diversified and flourished in isolation in South America. Another group
of ground-dwelling avian carnivores, the diatrymids, were more success-
ful in the Northern Hemisphere, but persisted only from the Paleocene
through the Eocene. Some of these birds were also quite large: the
Eocene Diatryma gigantea stood at a height of 2 meters (table 4.7). Their
diet may have been more diversified than the felid-like phorusrhacid
carnivores in that some may have been primarily scavengers and some
more omnivorous, eating some plant material as well as animal prey
(Feduccia 1996; Benton 2005).
No other group of birds has produced as many convergent flightless
species as the rails (family Rallidae; table 4.7). At least 24 living species
of rails have independently evolved flightlessness (Roff 1994; Roots
2006), and the number of extinct species of flightless rails that existed on
tropical Pacific islands before the arrival of humans (and the mammalian
predators they brought with them) is estimated to have been between
500 to 1,600 (Steadman 2006). However, the New Zealand weka rails are
formidable predators that have added human-introduced rats and mice
to their own diet, rather than the other way around (Feduccia 1996). I
have listed some 13 species of rails whose diet primarily consists of
animal prey (although they also consume some plant material) in table
4.7. Rather than felids, these 13 rail species are probably more conver-
gent on omnivorous predators like the canids, and may be considered as
ecological equivalents of foxes. Other species of rails clearly are more in
the category of foraging omnivores, and a few are herbivores (table 4.7).
The South Island takahe of New Zealand is the largest living rail, yet its
diet is exclusively vegetarian (Roots 2006).
The living ostriches, rheas, and the extinct palaeotids are plesiomor-
phic paleognath flightless birds that have converged on filling the eco-
logical niche of foraging omnivores. Their diet consists primarily of plant
material (for which they convergently evolved gizzards to process, as
discussed for the ostrich in chapter 2), but they consume a considerable
number of animals as well, chiefly insects, small mammals, and small
reptiles. Captive rheas in zoos are even known to have developed a taste
for sparrows (Roots 2006). As such, these ground-dwelling birds have
converged on mammalian ecological roles ranging from badgers’ to wild
150 Chapter 4
Table 4.7
Convergent evolution of mammalian ecological equivalents in birds
1 Convergent ecological role: FLIGHTLESS, GROUND-DWELLING AVIAN

CARNIVORES (“cat niche, fox niche”)
1.1 Giant diatrymid (Aves: Neognathae: Neoaves: Gastornithiformes: Diatrymidae;
Diatryma gigantea †Eocene)
1.2 Titan phorusrhacid (Aves: Neognathae: Neoaves: Gruiformes: Cariamae:
Phorusrhacidae; Titanus walleri †Pliocene)
1.3 Red-legged seriema (Aves: Neognathae: Neoaves: Gruiformes: Cariamae:
Cariamidae; Cariama cristata)
1.4 North Island weka rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae;
Gallirallus australis grayi)
1.5 Black weka rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Gallirallus
australis australis)
1.6 Stewart Island weka rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae;
Gallirallus australis scotti)
1.7 New Britain rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Gallirallus
insignis)
1.8 New Caledonia wood rail (Aves: Neognathae: Neoaves: Guiformes: Rallidae;
Gallirallus lafresnayus)
1.9 Calayan rail (Aves: Neognathae: Neoaves: Guiformes: Rallidae; Gallirallus
calayanensis)
1.10 Okinawa rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Rallus okinawa)
1.11 Henderson Island rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Porzana
atra)
1.12 Aldabra rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Canirallus cuvieri
aldabranus)
1.13 Snoring rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Aramidopsis
plateni)
1.14 Guam rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Rallus owstoni)
1.15 Lord Howe Island rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae;
Tricholimnas sylvestris)
1.16 Auckland Island rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Rallus
pectoralis muelleri)
INSECTIVORES (“mole niche”)
2.1 Kiwi (Aves: Paleognathae: Struthioniformes 3: Apterygidae; Apteryx haasti)
2.2 Kagu (Aves: Neognathae: Neoaves: Gruiformes: Rhynochetidae; Rhynchetos
jubatus)
FORAGING OMNIVORES (“badger niche, wild pig niche”)
3.1 Palaeotis (Aves: Paleognathae: Lithornithiformes: Palaeotidae; Palaeotis weigelti
†Eocene)
3.2 Ostrich (Aves: Paleognathae: Struthioniformes 1: Struthionidae; Struthio camelus)
3.3 Rhea (Aves: Paleognathae: Struthioniformes 2: Rheidae; Rhea americana)
3.4 New Guinea flightless rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae;
Megacrex inepta)
3.5 Inaccessible Island rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae;
Atlantisia rogersi)
Table 4.7
(continued)
3.6 Woodford’s rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Nesoclopeus

woodfordi)
3.7 Samoan moorhen (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Gallinula
pacificus)
3.8 San Cristobal moorhen (Aves: Neognathae: Neoaves: Gruiformes: Rallidae;
Gallinula silvestris)
3.9 Gough Island moorhen (Aves: Neognathae: Neoaves: Gruiformes: Rallidae;
Gallinula nesiotis comeri)
3.10 Giant coot (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Fulica gigantea)
3.11 Zapata rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Cyanolimnus
cerverai)
HERBIVORES (“ungulate niche”)
4.1 Emu (Aves: Paleognathae: Struthioniformes 3: Dromaiidae; Dromaius
novaehollandiae)
4.2 Moa-nalo (Aves: Neognathae: Galloanserae: Anseriformes: Anatidae;
Chelychelynechen quassus †Holocene)
4.3 Kakapo parrot (Aves: Neognathae: Neoaves: Psittaciformes: Psittacidae; Strigops
habroptilus)
4.4 Broad-billed parrot (Aves: Neognathae: Neoaves: Psittaciformes: Psittacidae;
Lophopssittacus mauritianus †Holocene)
4.5 South Island takahe (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Notornis
mantelli hochsteteri)
4.6 Tasmanian native hen (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Tribonyx
mortieri)
4.7 Invisible rail (Aves: Neognathae: Neoaves: Gruiformes: Rallidae; Habroptila
wallaci)
FRUGIVORES (“tapir niche”)
5.1 Eremopezus (Aves: Paleognathae: Aepyornithiformes: Eremopezidae; Eremopezus
eocaenus †Oligocene)
5.2 Elephant bird (Aves: Paleognathae: Aepyornithiformes: Aepyornithidae;
Mullerornis agilis †Holocene)
5.3 Cassowary (Aves: Paleognathae: Struthioniformes 3: Casuariidae; Casuarius
casuarius)
5.4 Dodo (Aves: Neognathae: Neoaves: Columbiformes: Raphidae; Raphus cucullatus
†Holocene)
5.5 Solitaire (Aves: Neognathae: Neoaves: Columbiformes: Raphidae; Pezophaps
solitaria †Holocene)
6 Convergent ecological role: FLIGHTLESS, SWIMMING AVIAN PISCIVORES
(“pinniped niche”)
6.1 Patagopteryx (Aves: Ornithothoraces: Patagopterygiformes: Patagopterygidae;
Patagopteryx deferrariisi †Cretaceous)
6.2 Hesperornis (Aves: Ornithurae: Hesperornithiformes: Hesperornithidae;
Hesperornis regalis †Cretaceous)
6.3 Steamer duck (Aves: Neognathae: Galloanserae: Anseriformes: Anatidae; Tachyeres
brachypterus)
152 Chapter 4
Table 4.7
(continued)
6.4 Auckland Islands teal (Aves: Neognathae: Galloanserae: Anseriformes: Anatidae;

Anas aucklandica)
6.5 Law’s diving goose (Aves: Neognathae: Galloanserae: Anseriformes: Anatidae;
Chendytes lawi †Holocene)
6.6 King penguin (Aves: Neognathae: Neoaves: Sphenisciformes: Spheniscidae;
Aptenodytes patagonica)
6.7 Flightless cormorant (Aves: Neognathae: Neoaves: Pelecaniformes:
Phalacrocoracidae; Nannopterum harrisi)
6.8 Murres (Aves: Neognathae: Neoaves: Charadriiformes: Alcidae; Uria aalge)
6.9 Titicaca grebe (Aves: Neognathae: Neoaves: Podicipediformes: Podicipedidae;
Rollandia microptera)
6.10 Junín grebe (Aves: Neognathae: Neoaves: Podicipediformes: Podicipedidae;
Podiceps taczanowskii)
6.11 Atitlán grebe (Aves: Neognathae: Neoaves: Podicipediformes: Podicipedidae;
Podilymbus gigas †Holocene)
pigs’, depending upon the proportions of plant and animal content in

their diet.
It was long thought that the living “ratite” ground-dwelling birds—
ostriches, rheas, emus, cassowaries, and kiwis—all descended from a
single common ancestor. Today they exist geographically scattered across
the Earth: ostriches in Africa, rheas in South America, emus and casso-
waries in Australia, and kiwis in New Zealand. The classic explanation
for this puzzling geographic dispersal was that the flightless ancestors of
these birds evolved on the Mesozoic supercontinent of Gondwana, which
consisted of the present-day landmasses of South America, Africa, Mad-
agascar, India, Australia, the New Zealand islands, and Antarctica all
combined. It is known that Gondwana began to break up in the Creta-
ceous, and thus it was hypothesized that the present-day ratites evolved
separately, each on its own fragment of Gondwana, a process of vicariant
evolution driven by plate-tectonic continental drift.
However, the monophyletic origin of the ratites came into question
when molecular phylogenies for these birds began to be constructed.
Recent molecular analyses reveal that the ratites are polyphyletic, and
that the close morphological similarities between these paleognath birds
are not synapomorphies, but rather are due to convergence (Harshman
et al. 2008). In retrospect, this is not surprising, given the incredible
number of independent morphological convergences seen in other flight-
less birds (table 4.7). Molecular evidence now indicates that the ostriches
(Struthionidae) and the rheas (Rheidae) independently evolved flight-
lessness. It also indicates that the living emus (Dromaiidae), cassowaries

(Casuariidae), and kiwis (Apterygidae), as well as the extinct moas
(Dinornithidae), are a monophyletic group that evolved flightlessness
independently of both the ostriches and the rheas (Harshman et al.
2008). Thus, in table 4.7, I have divided the Struthioniformes—the
“ostrich-like” birds—into the Struthioniformes 1, 2, and 3 to indicate that
they are polyphyletic and that they represent three separate convergent
lineages.
The end result is not only two independent morphological convergen-
ces on flightlessness in the ostrichs and the rheas, but also two indepen-
dent ecological convergences on foraging omnivory (table 4.7). In
contrast, the Australian emus and their New Zealand cousins the moas
(Dinornis giganteus, now extinct due to human activities) are considered
to be ecologically equivalent to browsing ungulate mammals (Daugherty
et al., 1993). In addition to some rail species, the ecological role of brows-
ing herbivory is convergently filled by two species of flightless parrots
and one species of a flightless browsing “goose,” the moa-nalo. These
ground-dwelling birds convergently filled the ecological niche of ungu-
late mammals in widely separated regions of the Earth: the kakapo
parrot still survives in New Zealand, whereas the recently extinct broad-
billed parrot was a native of Mauritius, and the moa-nalo lived in Hawaii
(Roots 2006).
Rather than these ground-dwelling birds converging on the ecological
role of browsing herbivorous mammals, it may well be that the Cenozoic
browsing herbivorous mammals were themselves convergent on the eco-
logical role of the Mesozoic browsing herbivorous dinosaurs. Barrett
(2005) argues that the ornithomimosaurs were herbivores, and points out
that many traits now associated with herbivory in living birds, such as
gastric mills and large intestines, were also a feature of these extinct
theropods. Thus, rather than being convergent on mammals, the living
herbivorous ground-dwelling birds may have reevolved a more ancient
ecological role, that of the ornithomimosaur niche.
Even though the emu-cassowary-kiwi clade is now proved to be
monophyletic (Harshman et al. 2008), the members of the clade are
ecologically divergent (table 4.7). The peculiar New Zealand kiwis are
not herbivores, but rather are ecologically convergent on the mammalian
mole ecological role, specialized in hunting down and eating animals
that dwell in the soil. While they do not tunnel through the soil like
moles, they do live in burrows and hunt for insects at night in the dark,
using their highly developed sense of smell and long beaks to locate
154 Chapter 4
earthworms and other burrowing prey underground, as moles do subter-

raneally (Daugherty et al. 1993). Far to the north of New Zealand, in
New Caledonia, another group of ground-dwelling birds have conver-
gently filled the mole ecological niche. These are the kagu, which special-
ize in locating underground prey in a manner similar to the kiwis (Roots
2006).
The Australian cassowaries are also not herbivores, but frugivores
(table 4.7). They are the last surviving large ground-dwelling birds that
have convergently filled the mammalian frugivore niche. The large mam-
malian frugivores themselves are in danger of extinction—the largest
remaining South American frugivore is the 300-kilogram Brazilian tapir,
compared to the extinct 7,580-kilogram gomphothere and the extinct
6,300-kilogram giant ground sloth (Hansen and Galetti 2009). The
recently extinct elephant bird of Madagascar was a 450-kilogram frugi-
vore (Hansen and Galetti 2009) that may have been related to the more
ancient giant flightless eremopezus, a bird of Egypt. Eremopezus is
argued not to have been related to the African ostrich (Rasmussen et al.
2001), and molecular evidence indicates that the elephant bird also was
“clearly not the result of a recent divergence from the ostrich, or any
other ratite lineage” (Cooper et al. 2001). Yet another group of birds, the
columbiform pigeons, independently filled the ground-dwelling frugivore
ecological niche on isolated islands further to the east from Madagascar
and Africa. The recently extinct dodo was a large, ground-dwelling fru-
givore on Mauritius Island, and the recently extinct solitaire indepen-
dently filled the ground-dwelling frugivore niche on Rodriguez Island
(Feduccia 1996). In all cases, the extinction of these large frugivores—
whether mammal or flightless bird—has led to serious disruption of
seed-dispersal interactions in the affected ecosystems (Hansen and
Galetti 2009).
Ground-dwelling birds are not the only birds to have converged on
mammalian ecological roles. Multiple lineages of birds have become
flightless swimming and diving piscivores, converging on the pinniped
mammal niche (table 4.7). Just as the seals and sea lions have devolved
their walking legs into flippers for swimming (as discussed in chapter 2;
see table 2.1), so have the penguins devolved their wings. Unlike the
ground-dwelling flightless birds, the penguins have not lost the strong
keel on their sternum, nor have they lost their powerful pectoral muscles.
Instead of using these traits for flight, however, they are used for swim-
ming in water. As water is much denser than air, swimming through this
dense medium is even more energetic than flying (Roots 2006). One has
only to observe the graceful, seemingly effortless, swimming of seals in

any zoo large enough to have underwater viewing stations in the seal
exhibit area, and to repeat those observations in the penguin exhibit
area, to appreciate the power of convergent evolution (I recommend the
Schönbrunn zoo in Vienna, my favorite, where the seal and penguin
exhibition areas are side by side).
Other swimming birds did not devolve their wings into flippers, and
thus have convergently lost their keeled sternum and powerful pectoral
muscles, as the ground-dwelling birds have (Roff 1994; Roots 2006).
Instead of using their wings as flippers, they have evolved powerful leg
muscles and large, webbed feet. They hold their vestigial wings tight
against their bodies and smoothly propel themselves through the water
with their feet. The large flightless cormorant of the Galapagos, the large
steamer ducks of South America, and the murres of the North Atlantic
swim in this fashion (table 4.7). Two other groups of the duck family have
independently converged on the swimming piscivore niche: the Auckland
Island teal and the recently extinct Law’s diving goose of western North
America.
The convergent evolution of flightless, swimming piscivorous birds is
not confined to the marine realm. Three species of grebes are known to
have independently filled this ecological niche in freshwater lakes in
South America, two of which survive today (table 4.7). How many other
instances there may be of the convergent evolution of swimming birds
in freshwater habitats is unknown, as the fossil record of these environ-
ments is very poor.
However, the fossil record does reveal that the convergent evolution
of flightless, swimming birds is an ancient phenomenon that predates the
destruction of dinosaurian ecosystems in the Cretaceous and the subse-
quent evolution of the mammalian ecosystems of the Cenozoic (Benton
2005). Two independent lineages of birds are known to have devolved
their wings and converged on the swimming piscivore niche in the Cre-
taceous (table 4.7), long before the evolution of pinniped mammals. Thus,
the pinniped mammals may themselves be viewed as having converged
on a more ancient ecological role, the hesperornithid niche.
One Ecosystem Play, Many Convergent Casts of Actors
In the previous section of the chapter, we have seen that the number of
ecological roles or niches available for Earth organisms is demonstrably
limited, in that species from many different phylogenetic lineages have
156 Chapter 4
been constrained in their evolution to filling the same ecological niche,

even if that ecological role is extremely specialized. Ecological niche
convergence is the rule, rather than the exception, in evolution.
In this section we shall consider the ecological level of communities
and ecosystems, which are assemblages of multiple ecological roles. Com-
parisons of ecosystems from different regions of the Earth, both in space
and in time, reveal a remarkable similarity in the assemblages of ecologi-
cal roles within those ecosystems. One might think of an ecosystem as
being analogous to a theatrical performance: in any given region of the
Earth, an ecosystem play is being performed with a script specifying the
interaction of the various ecological roles in that play. What is unex-
pected is one’s observation that the ecosystem play—the script of eco-
logical roles—seems to be very similar in many parts of the Earth, even
though the casts of animal and plant actors in those ecological roles may
be very different.
The flightless, ground-dwelling birds of New Zealand reveal a conver-
gent phenomenon that is larger than that of the “one ecological role,
many convergent players” thesis of the previous section of this chapter.
The kiwis have convergently filled the mammalian mole niche, and three
other bird lineages have convergently filled mammalian ungulate-herbi-
vore niches: the recently extinct moas filled the antelope niche of a
browsing ungulate, and the kakapo parrots and takahe rails have filled
the sheep niche of a grazing ungulate (Daugherty et al. 1993). The suite
of ecological roles filled by mammals elsewhere in the world are thus
filled by birds in New Zealand.
New Zealand split off early in the fragmentation of the Gondwanan
supercontinent in the Mesozoic, and was geographically isolated from
subsequent mammalian evolution in the Cenozoic following the destruc-
tion of the dinosaur-dominated ecosystems of the Earth in the end-
Cretaceous mass extinction. There are no native ground-dwelling
mammals on New Zealand, and there are only two species of bats, whose
ancestors flew there across the waters (Roots 2006). In the absence of
mammalian moles, antelopes, and sheep, the birds convergently evolved
the ecological roles of these animals. It is as if these ecological roles exist
in the absence of either mammals or birds, and that they would be con-
vergently found by evolution within some other phylogenetic lineage if
they had not been discovered by birds in New Zealand and mammals in
other regions of the world.
The peculiar fauna of New Zealand therefore suggest that ecological
convergence exists at the level of ecosystems and communities as well
as at the level of individual ecological roles. In the absence of mammalian
predators, other phylogenetic lineages have filled predatory niches in

New Zealand—not only the Weka rail, but also ancient reptiles, giant
molluscs, and giant myriapod arthropods. These nonmammal predators
have convergently evolved many mammalian traits—such as long life
spans, low reproductive rates, and large body sizes—that are not charac-
teristic of normal molluscs and arthropods (Daugherty et al. 1993). The
predatory tuatara, Sphenodon punctatus, is a reptilian “living fossil”
(Benton 2005, 237–239) that is not a lizard (although it resembles a
lizard) but rather is the sole remaining survivor of a lepidosaurian sister
group to the Squamata, the Sphenodontia (see appendix). The tuatara
has a life span of over 70 years, and reproduces only every four years
after taking 13 years to reach sexual maturity (Daugherty et al. 1993), a
life history more similar to that of a mammal than of a reptile. Its ances-
tors survived the end-Cretaceous mass extinction, while the New Zealand
non-avian dinosaurs did not. The giant predatory snail Paryphanta busbyi,
which has a shell 115 millimeters in diameter, can live up to 40 years and
takes 15 years to reach sexual maturity (Daugherty et al. 1993). The giant
predatory centipede Cormocephalus rubriceps can grow to the length of
250 millimeters, and its females brood only about 20 eggs (Daugherty
et al. 1993). In the omnivore ecological role, the cricket-like giant
weta insect, Deinacrida heteracantha, has converged on the rat niche,
reaching the comparable body length of 150 millimeters and a weight of
70 grams. Like rats, they forage at night and congregate in diurnal shel-
ters (Daugherty et al. 1993).
The New Zealand islands are very unusual, but not unique. The island
of Madagascar has not been separated from the African mainland as long
as New Zealand has been isolated by water, but ecosystem convergences
have occurred there as well, on a smaller scale. There are no laurasiathe-
rian hedgehogs and shrews on Madagascar. Instead, afrosoricid mammals
have convergently evolved these same ecological roles in the greater
hedgehog tenrec, Setifer setosus, and in the lesser long-tailed shrew
tenrec, Microgale longicaudata. On the African mainland itself, the lau-
rasiatherian otter niche and mole niche have been convergently filled by
the afrosoricid giant otter “shrew,” Potamogale velox, and the cape
golden “mole,” Chrysochloris asiatica (Lecointre and Le Guyader 2006).
Thus, the same suite of ecological niches—hedgehog, shrew, otter, mole—
exist in African ecosystems, but they are not filled by hedgehogs, shrews,
otters, or moles.
The classic, most frequently cited example of ecosystem convergence
is provided by the marsupial fauna of Australia and Tasmania. The iso-
lated marsupial mammals of these regions have independently evolved
158 Chapter 4
Table 4.8
Convergent evolution of ecological-analog compositions of marsupial-dominated ecosys-
tems in isolated Australian and Tasmanian regions and placental-dominated ecosystems in
the rest of the world
1 Convergent ecological analog: LARGE AMBUSH PREDATOR (“great cat niche”)

Placental-mammal ecosystems: Lion (Eutheria: Felidae; Panthera leo)
Australian ecosystem: Marsupial lion (Marsupialia: Thylaconidae; Thylacoleo carnifex
†Pleistocene)
2 Convergent ecological analog: SMALL AMBUSH PREDATOR (“small cat niche”)
Placental-mammal ecosystems: Wild cat (Eutheria: Felidae; Felis sylvestris)
Australian ecosystem: Marsupial quoll “cat” (Marsupialia: Dasyuridae; Dasyurus
viverrinus)
3 Convergent ecological analog: PURSUIT PREDATOR (“wolf niche”)
Placental-mammal ecosystems: Wolf (Eutheria: Canidae; Canis lupus)
Australian ecosystem: Tasmanian wolf (Marsupialia: Thylacinidae; Thylacinus
cynocephalus †Holocene)
4 Convergent ecological analog: NOCTURNAL FORAGING PREDATOR (“wolverine
niche”)
Placental-mammal ecosystems: Wolverine (Eutheria: Mustelidae; Gulo gulo)
Australian ecosystem: Tasmanian devil (Marsupialia: Dasyuridae; Sarcophilus harrisi)
5 Convergent ecological analog: ANT EATER (“anteater niche”)
Placental-mammal ecosystems: Anteater (Eutheria: Myrmecophagidae; Myrmecophaga
tridactyla)
Australian ecosystem: Numbat (Marsupialia: Myrmecobiidae; Myrmecobius fasciatus)
6 Convergent ecological analog: FOSSORIAL INSECTIVORE (“mole niche”)
Placental-mammal ecosystems: Mole (Eutheria: Talpidae; Talpa europea)
Australian ecosystem: Marsupial mole (Marsupialia: Notoryctidae; Notoryctes typhlops)
7 Convergent ecological analog: SMALL COMMUNAL-NESTING OMNIVORE (“rat
niche”)
Placental-mammal ecosystems: Black rat (Eutheria: Muridae; Rattus rattus)
Australian ecosystem: Brown marsupial rat (Marsupialia: Dasyuridae; Antechinus
stuartii)
8 Convergent ecological analog: TINY COMMUNAL-NESTING OMNIVORE (“mouse
niche”)
Placental-mammal ecosystems: Mouse (Eutheria: Muridae; Mus musculus)
Australian ecosystem: Fat-tailed marsupial mouse (Marsupialia: Dasyuridae;
Sminthopsis crassicandata)
9 Convergent ecological analog: GLIDING ARBOREAL OMNIVORE (“flying squirrel
niche”)
Placental-mammal ecosystems: Flying squirrel (Eutheria: Sciuridae; Glaucomys volans)
Australian ecosystem: Marsupial flying opossum (Marsupialia: Petauridae; Petaurus
australis)
10 Convergent ecological analog: SEMI-FOSSORIAL HERBIVORE (“woodchuck
niche”)
Placental-mammal ecosystems: Woodchuck (Eutheria: Sciuridae; Marmota monax)
Australian ecosystem: Wombat (Marsupialia: Phascolomidae; Phascolomys ursinus)
11 Convergent ecological analog: MID-SIZED FAST-RUNNING BROWSING
HERBIVORE (“deer niche”)
Placental-mammal ecosystems: Red deer (Eutheria: Cervidae; Cervus elaphus)
Australian ecosystem: Red kangaroo (Marsupialia: Macropodidae; Macropus rufus)
Table 4.8
(continued)
12 Convergent ecological analog: LARGE HERDING MIXED-FEEDING HERBIVORE

(“bison niche”)
Placental-mammal ecosystems: Bison (Eutheria: Bovidae; Bison bison)
Australian ecosystem: Diprotodon (Marsupialia: Diprotodontidae; Diprotodon optatum
†Pleistocene)
13 Convergent ecological analog: SMALL MIXED-FEEDING HERBIVORE (“rabbit
niche”)
Placental-mammal ecosystems: Rabbit (Euarchontoglires: Lagomorpha: Leporidae;
Oryctolagus cuniculus)
Australian ecosystem: Wallaby (Marsupialia: Macropodidiae; Macropus agilis)
an ecological analog structure, a series of 13 ecological roles, that mirror

those found in placental mammalian ecosystems in other regions of
the world (table 4.8). The isolated marsupials convergently evolved both
of the two principal placental-mammal predatory strategies: the stalk-
and-ambush strategy of the placental felids, in which the stalked prey
animal is not aware that it is in danger until the split second in which
the felid pounces upon it, and the pursuit-and-charge strategy of the
placental canids, in which the prey animal is well aware of its pack pursu-
ers, who run it to exhaustion and then charge upon it en masse. Both
the large marsupial lion and the Tasmanian wolf are extinct (the latter
only recently), but the small marsupial quoll “cat” still survives (Turner
1997; Lecointre and Le Guyader 2006). In addition, the marsupials
have convergently evolved the nocturnal foraging predator niche,
with the voracious Tasmanian devil filling the role of the placental
wolverine.
Two insectivore niches have been convergently evolved by the Austra-
lian marsupials, with the numbat filling the anteater niche and the mar-
supial mole filling the mole niche. These two highly specialized ecological
roles have independently appeared a surprising number of times in
ecosystems in widely separated regions of the Earth (see tables 2.4, 4.4,
and 4.7).
Three omnivore ecological analogs have been convergently evolved
by the isolated Australian fauna, with the brown marsupial rat and the
fat-tailed marsupial mouse respectively filling the placental rat niche and
mouse niche (Lecointre and Le Guyader 2006), and the marsupial flying
opossum has independently evolved the ecological role of the placental
flying squirrel (table 4.8).
160 Chapter 4
Last, Australian marsupials have independently evolved four herbi-

vore ecological roles that are analogs of those found in placental-mam-
mal ecosystems: the wombat not only fills the ecological niche of the
woodchuck, it even resembles a woodchuck. The red kangaroo does not
look like a deer, yet “it lives roughly the same way” (Benton 2005, 313).
In addition, both the deer and the kangaroo have independently evolved
the ruminant-stomach system of digestion (see table 2.10). The wallaby
is the native Australian player of the rabbit ecological role, a role that is
now increasingly being filled by the invasive European rabbit (and to the
detriment of the wallaby). The Australian diprotodons are now gone,
dying out in the extinction of many large land animals in the Pleistocene,
but these bison-sized marsupials once formed great herds (Benton
2005) of mixed-feeding herbivores, eating the twigs and leaves of shrubs
as well as grass, like the once-great herds of bison in placental ecosystems
(table 4.8).
As in the previous discussion of the peculiar fauna of New Zealand,
it is as if these 13 ecological roles exist in the absence of either placental
or marsupial mammals and, since the placentals were not present in
Australia, the marsupials independently discovered them in their own
separate evolution. This impression can be tested by considering the
pattern of animal evolution on yet another isolated, continental-sized
island: South America. South America is not an island today, but for most
of the 65 million years of the Cenozoic it was surrounded by water, cut
off from both Africa to the east (from which it split apart in the frag-
mentation of the supercontinent Gondwana in the Mesozoic) and from
North America to the north (Benton 2005).
During their period of isolation, the mammals of South America inde-
pendently evolved a series of at least 12 ecological roles that are analogs
of those found in terrestrial ecosystems in other regions of the world
(table 4.9). Some of these convergences are spectacular: not only did the
ambush-predator niche independently evolve in North and South
America, as it did in Australia, but this ecological role was played by
highly specialized saber-toothed placental cats in North America and
almost identical-appearing saber-toothed marsupial “cats” in South
America (as discussed in chapter 2, table 2.7). And not only did the fast-
running, browsing herbivore niche independently evolve in North and
South America, but in both regions the animals also convergently reduced
the number of digits on their feet down to one large hoofed toe—Plio-
hippus pernix in the north and Thoatherium minusculum in the south—
and at the same time, in the Miocene. The two animals never met one
another, yet they existed at the same time and were remarkably similar
in appearance.
Many of the South American ecological convergences were made by
meridiungulate mammals, a group of mammals that are now extinct (see
appendix). Completely in isolation, the meridiungulates convergently
filled the ecological niches of wild pigs, woodchucks, mastodons, horses,
rabbits, camels, hippopotamuses, beavers, and tapirs with their own
analogs (Prothero and Schoch 2002).
It is interesting that 13 ecological niches in Australia and 12 ecological
niches in South America, each evolving in isolation, converged with
niches in other regions of the world. However, by comparing the lists of
ecological analogs given in tables 4.8 and 4.9, we can see that the two
sets of niches listed are not all the same. In fact, only 5 of the 12 (or 13)
in the two sets are the same: the great cat, anteater, woodchuck, deer,
and rabbit niches. The South American mammals evolved a series of
ecological roles that are analogs of those independently evolved in North
America, but that appear not to have been discovered by the Australian
marsupials. These include the elephant, camel, hippopotamus, tapir,
beaver, and wild pig niches. Amphibious herbivores and large frugivores,
in particular, appear to be missing in Australian ecosystem evolution.
These absent ecological roles may well be due to the aridity of Australian
ecosystems and its effect on the Australian flora. Yet if that is the reason,
why did the Australian marsupials not evolve the camel niche, an ecologi-
cal role that independently appeared in both South and North American
ecosystems?
In the reverse comparison, several predator niches appear to be
missing in South American ecosystem evolution. Why did the South
American marsupials not evolve the wild cat, wolf, and wolverine
niches? Why were there no South American marsupial mole, rat, or
flying squirrel analogs? Some of these absences may be artificial, and
simply due to the vagaries of fossil preservation. All of the Australian
animals listed in table 4.8 either are alive today or have only relatively
recently become extinct (that is, going back in time at most about two
million years, to the beginning of the Pleistocene). In contrast, the pat-
terns of South American ecosystem evolution examined in table 4.9
extend back in time to the Eocene, some 55 million years into the past
(Prothero and Schoch 2002). Thus, some of these missing niches may
in fact have been filled in the past, but their fossil species either have
not been found yet or were never preserved in the fossil record in the
first place.
162 Chapter 4
Table 4.9
Convergent evolution of ecological-analog compositions of terrestrial ecosystems in iso-
lated South American regions and terrestrial ecosystems in the rest of the world in the
Cenozoic
1 Convergent ecological analog: LARGE SABER-TOOTHED AMBUSH PREDATOR

(“sabertooth tiger niche”)
North American ecosystem: Sabertooth tiger (Eutheria: Felidae; Smilodon fatalis
†Pleistocene)
South American ecosystem: Marsupial sabertooth tiger (Marsupialia: Thylacosmilidae;
Thylacosmilus atrox †Pliocene)
Asian ecosystem: Giant pangolin (Laurasiatheria: Manidae; Manis gigantea)
South American ecosystem: Giant anteater (Xenarthra: Myrmecophagidae;
Myrmecophaga tridactyla)
3 Convergent ecological analog: FORAGING OMNIVORE (“wild pig niche”)
North American ecosystem: Musk-hog peccary (Cetartiodactyla: Suina: Tayassuidae;
Pecari tajacu)
South American ecosystem: Notoungulate “warthog” (Meridiungulata: Notoungulata:
Toxodonta: Isotemnidae; Thomashuxleya rostrata †Eocene)
4 Convergent ecological analog: SUBTERRANEAN HERBIVORE (“gopher niche”)
North American ecosystem: Plains pocket gopher (Rodentia: Sciuromorpha: Geomyidae;
Geomys bursarius)
South American ecosystem: Chilean coruro (Rodentia: Hystricomorpha: Octodontidae;
Spalacopus cyanus)
5 Convergent ecological analog: SEMIFOSSORIAL HERBIVORE (“woodchuck niche”)
North American ecosystem: Woodchuck (Rodentia: Sciuridae; Marmota monax)
South American ecosystem: Archaeohyrax (Meridiungulata: Notoungulata: Typotheria:
Archaeohyracidae; Archaeohyrax concentricus †Eocene)
6 Convergent ecological analog: LARGE HERDING AND BROWSING HERBIVORE
(“elephant niche”)
North American ecosystem: Mastodon (Afrotheria: Proboscidea: Mammutidae;
Mammut americanum †Pleistocene)
South American ecosystem: Pyrotherian “mastodon” (Meridiungulata: Pyrotheria:
Pyrotheriidae; Pyrotherium romeri †Oligocene)
7 Convergent ecological analog: ONE-TOED, FAST-RUNNING BROWSING
HERBIVORE (“horse niche”)
North American ecosystem: Western horse (Perissodactyla: Equidae; Pliohippus pernix
†Miocene)
South American ecosystem: Litoptern “horse” (Meridiungulata: Litopterna:
Proterotheriidae; Thoatherium minusculum †Miocene)
8 Convergent ecological analog: SMALL MIXED-FEEDING HERBIVORE (“rabbit
niche”)
North American ecosystem: Eastern cottontail rabbit (Euarchontoglires: Lagomorpha:
Leporidae; Sylvilagus floridanus)
South American ecosystem: Notoungulate “rabbit” (Meridiungulata: Notoungulata:
Typotheria: Hegetotheriidae; Propachyrucos ameghinorum †Oligocene)
9 Convergent ecological analog: SEMI-ARID-ADAPTED FORAGING HERBIVORE
(“camel niche”)
North American ecosystem: Western camel (Cetartiodactyla: Camelidae; Camelops
hesternus †Pleistocene)
South American ecosystem: Litoptern “camel” (Meridiungulata: Litopterna:
Macraucheniidae; Macrauchenia patagonica †Pleistocene)
Table 4.9
(continued)
10 Convergent ecological analog: LARGE HERDING AMPHIBIOUS HERBIVORE

(“hippopotamus niche”)
North American ecosystem: Anthracotherium “hippopotamus” (Cetartiodactyla:
Anthracotheriidae; Anthracotherium magnus †Oligocene)
South American ecosystem: Notoungulate “hippopotamus” (Meridiungulata:
Notoungulata: Toxodonta: Toxodontidae; Toxodon platensis †Pleistocene)
11 Convergent ecological analog: SMALL AMPHIBIOUS HERBIVORE (“beaver
niche”)
North American ecosystem: Canadian beaver (Euarchontoglires: Rodentia: Castoridae;
Castor canadensis)
South American ecosystem: Notoungulate “beaver” (Meridiungulata: Notoungulata:
Typotheria: Mesotheriidae; Mesotherium cristatum †Pleistocene)
12 Convergent ecological analog: LARGE FRUGIVORE (“tapir niche”)
North American ecosystem: Tapir (Perisodactyla: Tapiridae; Tapirus californicus
†Pleistocene)
South American ecosystem: Astropotherian “tapir” (Meridiungulata: Astrapotheria:
Astrapotheriidae; Astrapotherium magnum †Miocene)
Isochronous and Heterochronous Ecosystem Convergence
The pattern of ecosystem convergence seen in the Australian example,

table 4.8, would support the hypothetical analogy of a convergent ter-
restrial ecological play being in existence on the Earth at the present
time, with its ecological roles being filled by a cast of marsupial mammals
in Australia and by a cast of placental mammals in other regions of the
Earth. Such a pattern of ecosystem convergence can be technically
termed isochronous, that is, at the same point in time. In contrast, the
pattern of ecosystem convergence seen in the South American example,
table 4.9, suggests that the scripted roles in the ecological play have
remained the same in time, that is, these roles have not significantly
changed in the past 55 million years. The wild pig niche existed in the
Eocene, and was filled by a notoungulate “warthog” in South America,
and the one-toed litoptern “horse” ran in the grasslands of the Miocene,
much as modern horses do today (Prothero and Schoch 2002). Such a
pattern of ecosystem convergence is heterochronous, that is, occurring in
different periods of time.
Has the script, the cast of ecological roles, in the terrestrial ecosystem
play remained unchanged for the past 55 million years? Or is this appar-
ent heterochronous ecosystem convergence simply due to the fact that
the roles are all being played by the same types of animals during this
164 Chapter 4
span of the Cenozoic? The extinct South American meridiungulates were

laurasiatherian mammals, after all, and thus related to the modern peri-
sodactyl and cetartiodactyl mammals that are still playing these ecologi-
cal roles today (see appendix).
One way to examine the question of the reality of heterochronous
ecosystem convergence is to go even further back in time, to a time
before terrestrial ecosystems were dominated by mammals of any type
at all, to a time when the dinosaurs reigned instead. The Mesozoic and
Cenozoic eras of time are separated by the second-greatest ecological
disruption of terrestrial ecosystems in Earth’s history (McGhee et al.
2004). Surely the dinosaur-dominated ecosystems of the Mesozoic must
have been fundamentally different from the mammal-dominated ecosys-
tems of the Cenozoic—the script of ecological roles in the Mesozoic must
have described a completely different ecological play.
Not so. The dinosaurs of the Mesozoic independently evolved a series
of at least 12 ecological roles that are analogs of those found in Cenozoic
terrestrial ecosystems (table 4.10). Since the Mesozoic world was so dif-
ferent from the Cenozoic, I have expanded the ecological comparisons
between these two eras of time to include fluvial, marine, oceanic, and
aerial ecological roles, and not just the roles of ground-dwelling terres-
trial animals. All in all, a series of at least 19 ecological roles existed in
the Mesozoic world that are analogs of those played by Cenozoic animals
(table 4.10).
Dinosaurs independently evolved the ecological roles of lions, wild
cats, wolves, wolverines, anteaters, elephants, deer, bison, rhinoceroses,
glyptodonts, goats, and ground sloths. However, since the dinosaurs filled
these ecological niches long before the mammals, it is more correct to
say that the Cenozoic mammals have convergently refilled the ecological
niches of allosaurs, coelophysises, velociraptors, troodonts, alvarezsau-
rids, sauropods, hypsilophodonts, hadrosaurs, triceratopses, ankylosaurs,
pachycephalosaurs, and therizinosaurs. Although some of these two
groups of animals resemble each other morphologically (triceratopses
and rhinoceroses, ankylosaurs and glyptodonts, therizinosaurs and
ground sloths), the others appear radically different—the mammalian
predators are all quadrupeds, whereas the dinosaurian predators walked
on their hind legs only. Yet they all were ecological analogs of each other,
making their living in roughly the same way. The goats, for example, have
a ruminant stomach system to process plant material, while the pachy-
cephalosaurs had a gizzard-like gastric mill to accomplish the same
purpose. The similarity to ungulate mammals in the herbivorous ecologi-
cal roles of many of the ornithopod dinosaurs, such as the hypsiloph-

odonts and hadrosaurs, has prompted Carrano et al. (1999, 256) to
conclude: “Although late Mesozoic and late Cenozoic terrestrial ecosys-
tems were profoundly different in terms of both animal and plant taxa,
there may be universal constraints on the ecological roles played by large
herbivores, resulting in convergence in morphology (and, by implication,
behavioral ecology) between groups as taxonomically distinct as dino-
saurs and mammals.” Thus, Carrano et al. (1999) suggest that Mesozoic
hadrosaurs and Cenozoic ungulates perhaps provide an example of both
ecological and behavioral convergence (we shall consider behavioral
convergence in animals in detail in chapter 6).
Not all dinosaurian ecological roles are filled by Cenozoic mammals,
and not all mammalian roles were filled by dinosaurs. Although the otter
is a fluvial piscivore, the gavialid crocodile is a much closer modern eco-
logical analog to the toothy spinosaur (table 4.10). The dinosaurs were
exclusively terrestrial animals, but if we examine marine and oceanic
habitats, we find that other Mesozoic animals independently evolved
ecological roles that are now played by Cenozoic mammals: the plesio-
saur, ichthyosaur, and pliosaur ecological niches have been convergently
refilled by Cenozoic sea lions, porpoises, and killer whales. In the air, the
insectivorous pterosaur and frugivorous pterosaur niches have been
independently reevolved by the microbats and the frugivorous bats. And
the modern pelican, itself an avian dinosaur, convergently refilled the
niche of the Mesozoic pteranodon and even resembles a small pteran-
odon in appearance. Mammals also have convergently evolved flying
piscivores, such as the greater bulldog bat, Noctilio leporinus, although
it looks nothing like a pteranodon (excepting the wings).
In these 19 ecological roles (table 4.10), the Cenozoic ecosystem play
is scripted in the same way as the Mesozoic ecosystem play. The roles of
the ecological play have not changed, although the casts of actors have
changed dramatically following the end-Cretaceous mass extinction. Yet,
just as in the comparison of the Australian and South American ecosys-
tems, there are missing roles in the Mesozoic. In the terrestrial realm,
where are the dinosaurian moles, rats, woodchucks, rabbits, camels, and
beavers? Were these missing niches, these ecological roles, in fact filled
by dinosaurs, whose fossil species have yet to be discovered? Or are these
ecological roles newly evolved by the Cenozoic mammals? If the latter,
then the Cenozoic ecosystem is not entirely convergent upon the Meso-
zoic—it would, instead, be a “revised” play. Given the ancientness of the
dinosaur-dominated ecosystem, some 65 to 220 million years ago, we may
166 Chapter 4
Table 4.10
Convergent evolution of ecological-analog compositions of Mesozoic dinosaur-dominated
ecosystems and Cenozoic mammal-dominated ecosystems
1 Convergent ecological analog: LARGE AMBUSH PREDATOR (“great cat niche”)

Cenozoic ecosystem: Lion (Mammalia: Felidae; Panthera leo)
Mesozoic ecosystem: Allosaur (Dinosauria: Allosauridae; Allosaurus fragilis †Jurassic)
2 Convergent ecological analog: SMALL AMBUSH PREDATOR (“small cat niche”)
Cenozoic ecosystem: Wild cat (Mammalia: Felidae; Felis sylvestris)
Mesozoic ecosystem: Coelophysis (Dinosauria: Coelophysidae; Coelophysis rhodesiensis
†Triassic)
3 Convergent ecological analog: PURSUIT PREDATOR (“wolf niche”)
Cenozoic ecosystem: Wolf (Mammalia: Canidae; Canis lupus)
Mesozoic ecosystem: Velociraptor (Dinosauria: Dromaeosauridae; Velociraptor
mongoliensis †Cretaceous)
4 Convergent ecological analog: NOCTURNAL FORAGING PREDATOR (“wolverine
niche”)
Cenozoic ecosystem: Wolverine (Mammalia: Mustelidae; Gulo gulo)
Mesozoic ecosystem: Troodont (Dinosauria: Troodontidae; Saurornithoides mongoliensis
†Cretaceous)
Cenozoic ecosystem: Anteater (Mammalia: Myrmecophagidae; Myrmecophaga
tridactyla)
Mesozoic ecosystem: “Desert bird” alvarezsaurid (Dinosauria: Alvarezsauridae;
Shuvuuia deserti †Cretaceous)
6 Convergent ecological analog: LARGE HERDING BROWSING HERBIVORE
(“elephant niche”)
Cenozoic ecosystem: Elephant (Mammalia: Elephantidae; Loxodonta africana)
Mesozoic ecosystem: Sauropod (Dinosauria: Titanosauridae; Titanosaurus
madagascariensis †Cretaceous)
7 Convergent ecological analog: MID-SIZED FAST-RUNNING BROWSING
HERBIVORE (“deer niche”)
Cenozoic ecosystem: Deer (Mammalia: Cervidae; Cervus elaphus)
Mesozoic ecosystem: Hypsilophodont (Dinosauria: Hypsilophodontidae; Hypsilophodon
foxii †Cretaceous)
8 Convergent ecological analog: LARGE HERDING MIXED-FEEDING HERBIVORE
(“bison niche”)
Cenozoic ecosystem: Bison (Eutheria: Bovidae; Bison bison)
Mesozoic ecosystem: Hadrosaur (Dinosauria: Hadrosauridae; Parasaurolophus walkeri
†Cretaceous)
9 Convergent ecological analog: LARGE HORNED GRAZING HERBIVORE
(“rhinoceros niche”)
Cenozoic ecosystem: Rhinoceros (Mammalia: Rhinocerotidae; Rhinoceros unicornus)
Mesozoic ecosystem: Triceratops (Dinosauria: Ceratopsidae; Triceratops albertensis
†Cretaceous)
10 Convergent ecological analog: LARGE ARMORED GRAZING HERBIVORE
(“glyptodont niche”)
Cenozoic ecosystem: Glyptodont (Mammalia: Glyptodontidae; Doedicurus clavi-
caudatus †Pleistocene)
Mesozoic ecosystem: Ankylosaur (Dinosauria: Ankylosauridae; Ankylosaurus
magniventris †Cretaceous)
Table 4.10
(continued)
11 Convergent ecological analog: MID-SIZED RUMINANT GRAZER (“goat niche”)

Cenozoic ecosystem: Goat (Mammalia: Bovidae; Capra hircus)
Mesozoic ecosystem: Pachycephalosaur (Dinosauria: Pachycephalosauridae;
Pachycephalosaurus wyomingensis †Cretaceous)
12 Convergent ecological analog: GIANT FRUGIVORE (“ground sloth niche”)
Cenozoic ecosystem: Giant ground sloth (Mammalia: Megatheriidae; Nothrotheriops
shastensis †Pleistocene)
Mesozoic ecosystem: “Sloth-claw” therizinosaur (Dinosauria: Therizinosauridae;
Nothronychus mckinleyi †Cretaceous)
13 Convergent ecological analog: FLUVIAL PISCIVORE (“gavialid niche”)
Cenozoic ecosystem: Gavialid crocodile (Archosauria: Crocodilidae; Gavialis
gangeticus)
Mesozoic ecosystem: Spinosaur (Dinosauria: Spinosauridae; Spinosaurus maroccanus
†Cretaceous)
14 Convergent ecological analog: SHALLOW-MARINE PISCIVORE (“sea-lion niche”)
Cenozoic ecosystem: Sea lion (Mammalia: Otariidae; Zalophus californianus)
Mesozoic ecosystem: Plesiosaur (Diapsida: Sauropterygia: Cryptocleididae;
Cryptocleidus oxoniensis †Jurassic)
15 Convergent ecological analog: SMALL OPEN-OCEAN CARNIVORE (“porpoise
niche”)
Cenozoic ecosystem: Porpoise (Mammalia: Phocaenidae; Phocaena phocaena)
Mesozoic ecosystem: Ichthyosaur (Diapsida: Ichthyosauria: Ichthyosauridae;
Ichthyosaurus platyodon †Jurassic)
16 Convergent ecological analog: LARGE OPEN-OCEAN CARNIVORE (“killer whale
niche”)
Cenozoic ecosystem: Killer whale (Mammalia: Delphinidae; Orcinus orca)
Mesozoic ecosystem: Pliosaur (Diapsida: Sauropterygia: Rhomaleosauridae;
Rhomaleosaurus megacephalus †Jurassic)
17 Convergent ecological analog: FLYING INSECTIVORE (“bat niche”)
Cenozoic ecosystem: Bat (Mammalia: Vespertilionidae; Myotis myotis)
Mesozoic ecosystem: “Frog-jaw” pterosaur (Ornithodira: Pterosauria: Aneurognathidae;
Batrachognathus volans †Jurassic)
18 Convergent ecological analog: FLYING FRUGIVORE (“fruit bat niche”)
Cenozoic ecosystem: Fruit bat (Mammalia: Pteropodidae; Rousettus aegyptiacus)
Mesozoic ecosystem: “Bakony dragon” pterosaur (Ornithodira: Pterosauria:
Azhdarchidae; Bakonydraco galaczi †Cretaceous)
19 Convergent ecological analog: FLYING PISCIVORE (“pelican niche”)
Cenozoic ecosystem: Pelican (Aves: Pelecanidae; Pelecanus occidentalis)
Mesozoic ecosystem: Pteranodon (Ornithodira: Pterosauria: Pteranodontidae;
Pteranodon longiceps †Cretaceous)
168 Chapter 4
never be able to reconstruct all the complexities of its structure, particu-

larly for small animals whose fragile bones have a very low probability
of preservation in the fossil record.
Thus far we have examined and compared lists of ecological roles that
have independently evolved in ecosystems that are separated from one
another in space or time. An alternative approach to examining the
phenomenon of convergent ecosystem evolution is to examine the eco-
logical-role composition of ecosystems that have evolved in equivalent
physical environmental regions, or biomes, such as tundra regions, desert
regions, and so on. The hypothesis to be tested here is whether ecosys-
tems that develop in equivalent biomes will also develop equivalent
compositions of ecological roles, or niches. Several isochronous studies
have examined Mediterranean-style climatic regions of the Mediterra-
nean, of course, and equivalent biomes in Chile and California (Went
1971; Mooney 1977; Peet 1978). The studies included in Mooney (1977)
documented extensive ecosystem convergence in the types of ecological
roles played by both plants and vertebrate animals, but much less con-
vergence in the roles played by invertebrate animals, particularly insects.
On the other hand, Milewski and Bond (1982) document the convergent
evolution of the same seed-dispersal ecological roles in insects in
Mediterranean-style climatic regions in Australia and South Africa
(particularly myrmecochory, which we considered in the last chapter).
Extensive ecosystem convergence is not a phenomenon confined to
terrestrial biomes. McKinney (2007) and McKinney et al. (2007) docu-
ment the heterochronous convergent evolution of Paleozoic-structured
marine ecosystems in the unusual oligotrophic (low-nutrient level)
biomes in the Northern Adriatic Sea of Europe. Only 2 percent of the
species present are survivors of the ancient Paleozoic faunas; the other
species are all modern, yet they have converged on lifestyles that are 250
million years out of date! These include 30-centimeter-high polychaete
worms, Sabella spallanzanii, that have convergently refilled the ecologi-
cal niche of Paleozoic crinoids, and numerous bivalve molluscs, such as
the scallop Aequipecten opercularis, that have convergently refilled the
ecological niche of Paleozoic brachiopod shellfish. While the ecological
roles are the same as those in the Paleozoic, the players of those roles
are modern marine species.
A Periodic Table of Niches?
In the previous section of the chapter we examined data suggesting that

the script of ecological roles in nature seems to exist in the absence of
the identity of the players, whether they are dinosaurs or mammals (table
4.10), placentals or marsupials (table 4.8), and that these players have
convergently filled those roles independently in different positions in
both space and time. While stressing the importance of the phenomenon
of ecological convergence, Pianka (1978: 300) also cautions that “evolu-
tionary convergence can easily be read into a situation by placing undue
emphasis upon superficial similarities but failing to appreciate fully
the inevitable dissimilarities between pairs of supposed ecological
equivalents.”
Thus, an alternative approach to the comparison of existent ecosys-
tems would be to attempt to compare existent ecosystems with nonexis-
tent ecosystems; then we could seek to understand what it is about
existent ecosystems that has led organisms to construct them repeatedly
rather than constructing theoretically possible but nonexistent ecosys-
tems. That is, we can apply the analytical techniques of theoretical mor-
phology (McGhee 2007) to ecology. As Maclaurin and Sterelny (2008,
128–129) argue, “one way of thinking about ecological diversity is in
terms of a phenomenological ecospace. The dimensions of that space
. . . define a space of possibilities,” analogous to a theoretical morpho-
space; but they also caution that “[a]s with morphospace, though, a total
ecospace is of high and somewhat arbitrary dimensionality” (Maclaurin
and Sterelny 2008, 111).
Pianka (1978) in fact used this approach in his attempt not only to
create an empty niche—to specify an ecological role a priori without
reference to any living organism—but also to arrange those empty niches
into a “periodic table of niches,” a space of potential ecological roles,
again analogous to a theoretical morphospace. He used two dimensions
to construct that space: a trophic-niche dimension and a life-history-
niche dimension (table 4.11). The trophic-niche dimension spans the
possible trophic spectrum of autotrophy (primary producers) to heter-
otrophy (herbivores and carnivores). The life-history-niche dimension
spans the possible spectrum from r-selected organisms (those with high
reproductive rates, low parental care for offspring, and short life spans)
to K-selected organisms (those with low reproductive rates, high parental
care for offspring, and long life spans).
Within this space, Pianka (1978) thus created nine permutation posi-
tions, empty niches, or potential ecological roles, in the absence of any
actual living organisms. The next step in the analysis is to ask: are these
potential niches actually empty in nature? The answer is no—Pianka
(1978) discovered that actual living organisms have evolved each of the
nine potential ecological roles (table 4.11). Pianka (1978) then added an
170 Chapter 4
Table 4.11
The ecological periodic table of niches
Trophic niche
Ecological life-history
niche Primary producers Herbivores Carnivores
r-selected organisms Annual plants Aphids, Mantids

Caterpillars
Midpoint of r-K Shrubs Lemmings Weasels
continuum
K-selected organisms Perennial plants Deer, Bison Cougars, Wolves
(especially trees)
Note: Modified from Pianka (1978).
additional dimension to the analysis: the dimension of space utilization.

This dimension spans the spectrum of organisms that exploit space in
only two dimensions to those that exploit space in three (table 4.12). As
in the previous analysis (table 4.11), he discovered that the additional six
potential empty niches of three-dimensional space utilization by hetero-
trophs were in fact filled by actual animals in nature (table 4.12).
However, the three ecological niches of three-dimensional space
exploitation by primary producers are not filled—Pianka (1978) had
created possible but nevertheless nonexistent ecological roles (table
4.12), although he did not comment further upon these possibilities. In
the spirit of theoretical morphology, it is possible to predict what such
an organism might look like: a plant that has evolved a gas-filled bladder
(or multiple bladders) that allows it to float in the air. Such an organism
could easily photosynthesize in midair (in fact, it could avoid competition
for light from other plants by doing just that, and thus would have a
selective advantage over ground-dwelling plants), but it would also
require water and nutrients. Water is obtainable from water vapor in
clouds or from the atmosphere itself in humid regions of the Earth.
Nutrients are a problem, though. Such a hypothetical plant would prob-
ably have to be carnivorous, obtaining its necessary nutrients by preying
on flying insects or other animals, in a manner convergent on that of
ground-dwelling carnivorous plants (table 4.1).
A balloon-like, floating carnivorous plant has never evolved on Earth;
this potential ecological role remains empty (table 4.12). Why? The
reason is probably the extreme instability of such a habitat in the Earth’s
existent weather patterns; sessile ground-dwelling plants have had to
evolve complex mechanisms to solve their own serious weather-related
Table 4.12
Pianka’s inclusion of organisms that exploit space in three dimensions (3D), as opposed
to just two dimensions (2D), in his periodic table of niches
Trophic niche
Ecological
life-history niche Primary producers Herbivores Carnivores
r-selected 2D: Annual plants 2D: Aphids, 2D: Mantids

organisms Caterpillars
3D: — 3D: Bees, 3D: Dragonflies
Butterflies
Midpoint of r-K 2D: Shrubs 2D: Lemmings 2D: Weasels
continuum 3D: — 3D: Squirrels, 3D: Flycatchers,
Fruit bats Insectivorous
bats
K-selected 2D: Perennial plants 2D: Deer, Bison 2D: Cougars,
organisms (especially trees) Wolves
3D: — 3D: Parrots 3D: Eagles,
Falcons
Note: Modified from Pianka (1978). Note that the three possible ecological niches of
three-dimensional space exploitation by primary producers are not filled.
problems, as discussed in chapter 3. These problems might be exacer-

bated, perhaps fatally, for a hypothetical floating plant with no control
over where it might be carried by the winds; if blown into desert regions,
it could die of dehydration; if blown out over the ocean, it could perish
by being unable to feed on flying insects. But might such a plant exist on
another world, a world with weather patterns radically different from
those found on Earth? The concept of entirely possible ecological roles,
roles that are nevertheless nonexistent on Earth, has implications for the
concept of convergent evolution itself, implications that will be examined
in detail in chapter 7.
Pianka (1978) noted in passing that a potential periodic table of niches
for marine organisms would differ from that for terrestrial organisms in
that there are relatively few K-selected marine organisms. Bambach
(1983) used a similar approach to Pianka (1978) to create an ecospace
for marine organisms by using a similar trophic-niche dimension, but he
abandoned the r-K life-history-niche dimension of Pianka (1978) in favor
of a habitat-mobility-niche dimension. In table 4.13, I have simplified
Bambach’s original trophic-niche dimension by creating a detritivore
niche, one that combines the organic-detritus collection mechanisms of
filter feeding and deposit feeding. I have further simplified the habitat-
mobility-niche dimension to cover the spectrum from pelagic (habitats
172 Chapter 4
Table 4.13
Occupation of marine-niche ecospace in the Cambrian
Trophic niche
Habitat-mobility
niche Detritivores Herbivores Carnivores
Pelagic Trilobites — —
Epibenthic
1. Sessile Brachiopods, — —
Eocrinoids
2. Mobile Trilobites, Ostracodes Monoplacophorans, —
Ostracodes
Endobenthic
1. Mobile, shallow Trilobites, Bristle — Bristle
worms, Brachiopods worms
2. Mobile, deep — — —
Note: Modified from Bambach (1983).
in the water column) to epibenthic (habitats on the sea bottom) to endo-

benthic (habitats within the sediments below the sea bottom). On the
other hand, I have made his original analysis a bit more complicated by
exhausting all possible permutations of these ecological states along
these two ecological dimensions and considering all of these permuta-
tions as possible ecological roles (table 4.13). In his original analysis,
Bambach (1983) omitted some of these permutations, a point I will
return to later.
Further, Bambach (1983) added an important new dimension not
present in Pianka’s analyses: geological time. He then proceeded to
analyze the pattern of marine ecosystem evolution for the past 545
million years since the evolution of multicellular heterotrophs, that is,
the animals (autotrophs were not included in the analysis). Of the 15
possible ecological roles in Bambach’s ecospace, only six were evolved
by animals in the Cambrian—the other nine niches were empty (table
4.13). By the middle of the Paleozoic, animals had evolved 11 of the
possible ecological roles, and only four niches remained empty (table
4.14). And by the Mesozoic, only three possible ecological roles remained
unfilled (table 4.15).
Note that although the permutations of ecological roles are exactly
the same in tables 4.13, 4.14, and 4.15, the animals playing those roles
change through time. Trilobites played the pelagic-detritivore role in the
Cambrian but were replaced by conodonts and hemichordates in the
mid-Paleozoic, who themselves were replaced by crustaceans and
Table 4.14
Occupation of marine-niche ecospace in the Middle Paleozoic
Trophic niche
Habitat-mobility
niche Detritivores Herbivores Carnivores
Pelagic Conodonts, — Cephalopods,

Hemichordates Placoderm fishes,
Sea scorpions
Epibenthic
1. Sessile Brachiopods, Crinoids, — Cnidarians
Bryozoans, Sponges
2. Mobile Agnathan fishes, Echinoids, Cephalopods,
Monoplacophorans, Gastropods, Crustaceans,
Gastropods Ostracodes Starfish
Endobenthic
1. Mobile, shallow Trilobites, Bivalves, — Merostomes,
Bristle worms Bristle worms
2. Mobile, deep Bivalves — Bristle worms
Table 4.15
Occupation of marine-niche ecospace in the Mesozoic
Trophic niche
Habitat-
mobility niche Detritivores Herbivores Carnivores
Pelagic Crustaceans, Osteichthyans Osteichthyans,

Gastropods Chondrichthyans,
Marine reptiles,
Cephalopods
Epibenthic
1. Sessile Bivalves, — Cnidarians
Bryozoans,
Barnacles,
Sponges
2. Mobile Bivalves, Gastropods, Gastropods,
Gastropods, Polyplacophorans, Crustaceans,
Crustaceans Crustaceans Starfish
Endobenthic
1. Mobile, Bivalves, — Gastropods,
shallow Echinoids, Sea Crustaceans,
cucumbers Bristle worms
2. Mobile, Bivalves, — Bristle worms
deep Bristle worms

174 Chapter 4
gastropods in the Mesozoic. Brachiopods were important sessile epiben-

thic detritivores throughout the Paleozoic (tables 4.13 and 4.14), and
indeed are known as the shellfish of the Paleozoic. Their ecological role
was taken over by the bivalve molluscs (table 4.15) following the end-
Permian mass extinction, the greatest ecological disruption in Earth
history, both in the sea and on the land (McGhee et al. 2004).
Marine reptiles, especially ichthyosaurs and pliosaurs, were important
players of the pelagic-carnivore role in the Mesozoic (table 4.15). Today
that same ecological role still exists, but it is now being played by marine
mammals, such as porpoises and killer whales. That last conclusion we
had already reached by comparing the actual existent Mesozoic and
Cenozoic ecosystems (table 4.10). Bambach’s analysis is importantly dif-
ferent, however, in that he created the pelagic-carnivore niche not by
examining lists of actual animals and their ecologies, but by considering
the possible permutations of ecological states along a spectrum of two
different ecological dimensions, trophic-niche and habitat-mobility-
niche. This is the same analytic procedure that is used in theoretical
morphology (McGhee 2007), and Bambach and his colleagues are now
engaged in the explicit construction of a theoretical ecospace that is
analogous to a theoretical morphospace (Bush et al. 2007).
Originally, however, Bambach (1983, 726) considered some of the
permutations of his two ecological dimensions to be “not biologically
practical adaptive strategies.” These include the three possible but cur-
rently empty niches that are listed in table 4.15. Two of these roles belong
to mobile endobenthic herbivores, an ecological role unfilled in marine
ecosystems. Yet this same ecological role has been evolved in terrestrial
ecosystems (the subterranean herbivore, or gopher, niche; see table 4.9).
The absence of animals filling this niche in marine ecosystems probably
reflects the different morphologies and ecologies of marine plants, which
generally do not have extensive root systems, and terrestrial plants, which
do. Thus, while this same ecological role may be nonfunctional in marine
ecosystems, it is clearly functional in terrestrial ecosystems.
The last empty niche listed in table 4.15 is the ecological role of a
sessile epibenthic herbivore. This ecological-dimension permutation may
indeed represent an impossible ecological role. Sessile epibenthic carni-
vores exist—the cnidarians—who simply wait for prey animals to come
within reach of their tentacles. A sessile herbivore has never evolved on
Earth, mainly because on Earth the plants themselves are sessile and
thus would never move to be within reach of the sessile herbivore. But
might such an animal exist on another world, a world that has evolved
mobile plants? Might this odd ecological role be possible after all, if only
theoretically? In chapter 7 we will explore in more detail the implications
of theoretically possible, but nevertheless nonexistent, ecological roles
on Earth for the concept of convergent evolution.
The phenomenon of ecosystem convergence supports the idea that
there may be “universal constraints on the ecological roles” played in
those ecosystems (Carrano et al. 1999, 256), just as the phenomenon of
morphological convergence argues for evolutionary constraint on the
spectrum of possible forms available to animals and plants, as discussed
in chapters 2 and 3. In the case of ecosystems, the nature of these poten-
tial universal constraints is far from clear, and is part of the larger debate
concerning the degree of integration and structure of ecosystems them-
selves (for an overview of this larger question, see Maclaurin and Sterelny
2008, 106–131). The creation of theoretical ecospaces could add much
more rigor to the debate concerning ecosystem convergence by revealing
the total spectrum of what is ecologically possible, and what is not, in
potential ecosystems. The works of Pianka (1978), Bambach (1983), and
Bush et al. (2007) are first steps in this direction, an analytic approach
to theoretical morphospace that needs much more attention from
ecologists.
Thus far in this book, we have considered the phenomena of conver-
gence in individual animal and plant forms and, on a larger scale, in the
ecological roles of entire assemblages of animal and plant species. In the
next chapter we shall consider convergence on much smaller scales; that
is, in the very molecules that make up living animals and plants.
5 Convergent Molecules
Recent advances in protein chemistry suggest that at least one set of biological
forms—the basic protein folds—is determined by physical laws similar to those
giving rise to crystals and atoms. . . . If it does turn out that a substantial amount
of higher biological form is natural, then the implications will be radical and
far-reaching. It will mean that physical laws must have had a far greater role in
the evolution of biological form than is generally assumed . . . underlying all the
diversity of life is a finite set of natural forms that will recur over and over again
anywhere in the cosmos where there is carbon-based life.
—Denton and Marshall (2001, 417)
Convergent Molecules?
The source code for all life on Earth is contained in the deoxyribonu-
cleic-acid molecule, DNA. DNA codes for RNA molecules, ribonucleic
acids, and RNA codes for the assembly of amino acids into proteins,
the building blocks of life. Each of the essential molecules of life has a
finite number of possible states. Since DNA and RNA molecules contain
only four different nucleotides, the code for Earth life is a base-four
system. As such, the probability, p, of convergent molecular evolution of
the same nucleotide at the same site in two different DNA or RNA
molecules, via random mutation, is one in four ( p = 0.25 per site). Protein
molecules contain only 20 different amino acids; thus, the probability of
the convergent molecular evolution of the same amino acid at the same
site in two different protein molecules is one in twenty ( p = 0.05 per
site). Proteins are complex molecules, and the function of the protein is
often determined not only by its amino acid composition but also by the
complex manner in which the molecule is folded. Basic protein folds also
exist in a finite number of geometries, and the number of fold geometries
that are potentially functional are only a tiny subset of all possible folds
(Axe 2004). Denton and Marshall (2001) estimate that the number of
178 Chapter 5
different protein-fold geometries that exist in life on Earth are between

500 and 1,000; therefore, the probability of independently evolving
protein molecules with the same fold geometry is one in five hundred
( p = 0.002) to one in a thousand ( p = 0.001).
Even though some of these probabilities are quite high, many biolo-
gists consider the occurrence of convergent molecular evolution in
nature to be quite unlikely. Biological molecules are typically very large,
and the possibility of many independent site convergences is improbable.
For example, even though the probability of the convergent evolution of
the same nucleotide at the same site in two different DNA molecules is
one in four ( p = 0.25), the probability of five such independent site con-
vergences occurring in two molecules is only two in 10,000 ( p = p1 × p2
× p3 × p4 × p5 = 0.0002). Numerous independent site convergences are
thus unlikely.
But what if the probabilities are not independent? Castoe et al. (2009,
8990) point out that mitochondrial DNA codes for metabolic genes that
are functionally related, and that “directional selection may thus tend to
affect many mitochondrial genes at once, leading to large-scale conver-
gence if similar selective events occur in different lineages.” Weinreich
et al. (2006) point out that pervasive biophysical pleiotropic effects, or
intermolecular interactions, in proteins constrain much of protein evolu-
tion. In two separate protein studies, they point out that only 15 to 29
percent of all possible mutational trajectories in these proteins are acces-
sible to natural selection, and thus the probability that evolution will
convergently occur along “largely identical mutational trajectories” is
quite high (Weinreich et al. 2006, 113). Further theoretical analyses of
mutational pathways in molecular adaptive landscapes reveal that few
pathways are evolvable, leading to the implication that “evolution might
be more reproducible than is commonly perceived, or [might] even be
predictable” (Poelwijk et al. 2007, 386). Rokas and Carroll (2008, 1943)
stress that the twin roles of positive selection for particular amino acid
substitutions in proteins and purifying selection against others “con-
strains substitutions to a small number of functionally equivalent amino
acids,” and that the action of natural selection has therefore produced
“frequent and widespread parallel evolution of protein sequences.”
DNA
The analysis of convergent evolution at the level of the DNA molecule

is difficult at present due to the lack of data. Although we have entered
Convergent Molecules 179
the “age of complete genomes,” Castoe et al. (2010) point out that only
in the past five to seven years have substantial numbers of mitochondrial
genome data sets become available for the vertebrates, and that com-
plete nuclear genome data sets number only in the tens for these highly
derived animals (bacterial genomes are densely sampled, but these
organisms have ancient phylogenies and thus the number of changes
separating their genomes are large, making the detection of convergent
evolution difficult; Castoe et al. 2010).
The few data that do exist are startling. Vision in vertebrate animals
is due to the activation of the visual pigments in the eye by impacting
photons. The visual pigments are composed of opsin proteins and a
chromophore. Most vertebrates have a single type of chromophore, but
five groups of opsin proteins exist: rhodopsin (RH1) and RH1-like
(RH2), which are maximally activated by light in the blue-green part of
the spectrum (about 480 nm to 530 nm); short-wavelength-sensitive
opsin-1 (SWS1), which is maximally activated by violet-blue wavelengths
(about 410 nm to 490 nm); short-wavelength-sensitive opsin-2 (SWS2),
which is maximally activated by ultraviolet-violet wavelengths (about
355 nm to 440 nm); and a long-wavelength- to medium-wavelength-
sensitive opsin (LWS/MWS), which is maximally activated by green-red
wavelengths (about 490 nm to 570 nm; Zhang 2003; Bowmaker and Hunt
2006).
The DNA coding for the opsin proteins has been intensively investi-
gated, and early studies suggested that the evolution of red-sensitive
opsins in humans and some fish independently evolved from green-
sensitive opsins by the same nucleotide substitutions in the encoding
DNA (Yokoyama and Yokoyama 1990). Subsequent studies of the evolu-
tion of trichromatic color vision in primates revealed that the precise
same three substitutions in DNA nucleotides that coded for LWS/MWS
opsins independently evolved in Old World primates (Catarrhini) and
New World primates (Platyrrhini; table 5.1). Parallel evolution in the
LWS/MWS opsin amino acid sequences in both primate groups (Hunt
et al. 1998) resulted in the replacement of serine with alanine at site 180
(substituting G for T in the first position in the DNA codon for serine),
the replacement of tyrosine with phenylalanine at site 277 (substituting
T for A in the second position in the DNA codon for tyrosine), and the
replacement of threonine with alanine at site 285 (substituting G for A
in the first position in the DNA codon for threonine).
These DNA coding changes resulted in a 30-nanometer spectral shift
in the maximum activation of the LWS/MWS opsin, from 530 to 560
Table 5.1
Convergent evolution of identical nuclear DNA molecules coding for photosensitive opsin
proteins
1 Convergent molecule and function: LWS/MWS OPSIN–ENCODING DNA (30 nm

spectral shift in vision produced by the identical three nucleotide substitutions in the DNA
coding for the long-wavelength/medium-wavelength-sensitive cone opsin protein)
1.1 Capuchin monkey (Primates: Simiiformes: Platyrrhini: Cebidae; Cebus apella)
1.2 Human (Primates: Simiiformes: Catarrhini: Hominidae; Homo sapiens)
2 Convergent molecule and function: SWS1 OPSIN–ENCODING DNA (66 nm spectral
shift in vision produced by the identical nucleotide substitution in the DNA coding for the
short-wavelength-sensitive cone opsin protein)
2.1 Tammar wallaby (Mammalia: Marsupialia: Diprotodontia: Macropodidae; Macropus
eugeni)
2.2 Cow (Mammalia: Eutheria: Laurasiatheria: Cetartiodactyla: Ruminantia: Bovidae;
Bos taurus)
3 Convergent molecule and function: SWS1-ultraviolet OPSIN–ENCODING DNA (35
nm spectral shift in vision produced by the identical nucleotide substitution in the DNA
coding for the ultraviolet-wavelength-sensitive cone opsin protein)
3.1 Rhea (Aves: Paleognathae: Struthioniformes 2: Rheidae; Rhea americana)
3.2 Herring gull (Aves: Neognathae: Neoaves: Ciconiiformes: Laridae; Larus argentatus)
3.3 Zebra finch (Aves: Neognathae: Neoaves: Passeriformes: Passeridae; Taeniopygia
guttata)
3.4 Gray parrot (Aves: Neognathae: Neoaves: Psittaciformes: Psittacidae; Psittacus
erithacus)
4 Convergent molecule and function: SWS1-violet OPSIN–ENCODING DNA (spectral
tuning shifts in vision produced by the identical nucleotide substitutions in the DNA coding
for the violet-wavelength-sensitive cone opsin protein)
Convergent lineages (three identical site changes):
4.1.1 African clawed frog (Tetrapoda: Batrachomorpha: Lissamphibia: Batrachia:
Anura: Pipidae; Xenopus laevis)
4.1.2 Human (Tetrapoda: Reptiliomorpha: Amniota: Synapsida: Mammalia: Eutheria:
Euarchontoglires: Primates: Hominidae; Homo sapiens)
Convergent lineages (one identical site change):
4.2.1 Zebra finch (Amniota: Sauropsida: Archosauromorpha: Ornithodira: Dinosauria:
Saurischia: Theropoda: Aves: Passeriformes: Passeridae; Taeniopygia guttata)
4.2.2 Budgerigar (Aves: Psittaciformes: Psittacidae; Melopsittacus undulatus)
4.2.3 Human (Amniota: Synapsida: Mammalia: Eutheria: Euarchontoglires: Primates:
Hominidae; Homo sapiens)
Convergent lineages (one identical site change):
4.3.1 Malawi fish (Osteichthyes: Actinopterygii: Teleostei: Perciformes: Cichlidae;
Metriaclima zebra)
4.3.2 African clawed frog (Osteichthyes: Sarcopterygii: Tetrapoda: Batrachomorpha:
Lissamphibia: Batrachia: Anura: Pipidae; Xenopus laevis)
4.3.3 Chicken (Tetrapoda: Reptiliomorpha: Amniota: Sauropsida: Archosauromorpha:
Ornithodira: Dinosauria: Saurischia: Theropoda: Maniraptora: Aves: Galliformes:
Phasianidae; Gallus gallus)
5 Convergent molecule and function: RH1 RHODOPSIN–ENCODING DNA (10 nm
spectral shift in vision produced by the identical nucleotide substitution in DNA coding
for the RH1 rhodopsin protein)
5.1 Longfin Baikal sculpin (Osteichthyes: Actinopterygii: Teleostei: Scorpaeniformes:
Cottocomephoridae; Cottocomephorus inermis)
5.2 Baikalian deepwater sculpin (Osteichthyes: Actinopterygii: Teleostei:
Scorpaeniformes: Abyssocottidae; Abyssocottus korotneffi)

nanometers. Hunt et al. (1998, 3299) suggest that these changes were
adaptive, the functional result of visual foraging for yellow and orange
fruits against a green foliage background, and that “the separate origin
of trichromacy in New and Old World primates would indicate that the
selection of these three [amino acid] sites is the result of convergent
evolution.”
The catarrhine and platyrrhine primates are closely related—both are
members of the simiiform clade (table 5.1)—and the two lineages
diverged only some 35 million years ago in the Eocene (Benton 2005).
One could thus argue that these two lineages demonstrate parallel
molecular evolution; that is, independent parallel nucleotide substitu-
tions in similar DNA molecules. In contrast, the marsupial mammals and
placental mammals diverged some 125 million years ago, in the Early
Cretaceous (Benton 2005). Yet even in these two ancient, independent
lineages of mammals, convergent evolution in opsin-encoding DNA has
been demonstrated (table 5.1).
Deeb et al. (2003) report that the identical nucleotide substitution in
DNA coding for the SWS1 opsin amino acid sequence occurred inde-
pendently in the marsupial tammar wallaby and in the placental cow:
replacement of phenylalanine by tyrosine at site 86 (substituting A for
T in the second position in the DNA codon for phenylalanine). This
DNA coding change resulted in a 66-nanometer spectral shift in the
maximum activation of the SWS1 opsin, from 358 to 424 nanometers.
The tammar wallaby grazes on grass and low tree branches in the late
afternoon, and Deeb et al. (2003) suggest that the convergent evolution
in SWS1-encoding DNA observed in it and in the placental grazing
ungulates enables the animals to visually distinguish features in grasses
with differing shades of green.
Primates are unique among mammals in having trichromatic color
vision, in that most mammals have only dichromatic vision. In contrast,
some fish, reptiles, and birds have tetrachromatic vision, and may be
able to see twice the number of colors that even trichromatic primates
can see (Ödeen and Håstad 2003). Many of these same animals can
see ultraviolet light, which has much shorter wavelengths than the
400-nanometer violet light that we are able to see. Ödeen and Håstad
(2003) have demonstrated that ultraviolet vision evolved four times inde-
pendently in the birds (table 5.1) by the identical nucleotide substitution
in DNA coding for the SWS1 opsin amino acid sequence: replacement
of serine by cysteine at site 90 (substituting G for C in the second posi-
tion in the DNA codon for serine). This DNA coding change results in
182 Chapter 5
a 35-nanometer spectral shift, from 406 down to 371 nanometers, in the

maximum activation of the SWS1 opsin. Ödeen and Håstad (2003, 859)
suggest that the observed convergent evolution is adaptive, although its
precise function is not clear: the difference in ultraviolet and violet vision
“is quite dramatic and changes not only the perception of objects that
reflect light solely in the UV or violet ranges but also the perception of
objects that reflect both UV/violet and longer wavelengths. This should
have important consequences for foraging, habitat use, social signaling,
and mate choice.”
An exhaustive analysis of the evolution of ultraviolet vision in the
vertebrates (Shi and Yokoyama 2003; Zhang 2003) has revealed even
more convergent molecular evolution in the SWS1 opsin–encoding DNA
(table 5.1). Such phylogenetically divergent lineages as frogs and pri-
mates have convergently evolved the same three nucleotide substitutions
in DNA coding for the SWS1 opsin: the replacement of phenylalanine
by leucine at site 49 (substituting A or G for T or C in the third position
in the DNA codon for phenylalanine), replacement of threonine by
proline at site 93 (substituting C for A in the first position in the DNA
codon for threonine), and replacement of serine by threonine at site 118
(substituting A for T in the first position in the DNA codon for serine).
Primates and two different groups of avian dinosaurs have indepen-
dently evolved an SWS1 opsin that replaced alanine with glycine at site
114 (substituting G for C in the second position in the DNA codon for
alanine). And the very distantly related fish, frog, and chicken lineages
have independently evolved an SWS1 opsin that replaced leucine with
valine at site 116 (substituting G for C in the first position in the DNA
codon for leucine).
The SWS1 opsin protein contains over 300 amino acids, yet Shi and
Yokoyama (2003) argue that only nine amino acid positions are involved
in spectral shifting. For all the opsin proteins, Bowmaker and Hunt (2006,
R484) argue that “there are only a limited number of sites within opsin
that can be altered without producing a non-functional pigment.” They
demonstrate that natural selection targets these sites in deepwater fish,
producing “classic examples of spectral tuning of visual pigments within
specific opsin classes which correlate with photic environments” (Bow-
maker and Hunt 2006, R486). Adaptation of vision to different light
intensities in different water depths extends to the rods of the eye as well
as the color-sensitive cones; for example, Bowmaker and Hunt (2006)
show that two separate species of deepwater Lake Baikal fish indepen-
dently evolved the same RH1 rhodopsin protein by replacing aspartic
acid with asparagine at site 83 (substituting A for G in the first position

in the DNA codon for aspartic acid) in the RH1-encoding DNA, produc-
ing a spectral shift from 505 to 495 nanometers (table 5.1).
Convergent evolution has also been demonstrated to occur in the
genes encoding for photoreceptor proteins that are not primarily used
in vision, but in the maintenance of circadian rhythms in animals and
heliotropic movement in plants (Cashmore et al. 1999). Charles Darwin
himself discovered that his plants ceased their daily movement if a filter
was placed between them and the sun, a filter that removed blue wave-
length light. It is now known that cryptochrome genes, cry, encode for
cryptochrome proteins, photoreceptors that respond to light in the
400-to-500-nanometer wavelength. These genes are present throughout
the green eukaryotes, from algae to land plants. Cashmore et al. (1999,
763) demonstrate that they have independently evolved in the metazoa:
“cryptochromes represent an example of repeated evolution, a special
case of convergent evolution in which a new genetic function arises
independently in two different lineages from orthologous (or paralo-
gous) genes. . . . This phenomenon contrasts with classic convergent
evolution, where the ancestral genes are unrelated.” As discussed in
chapter 1, the phenomenon in a strict sense would be termed “parallel
evolution.”
Other examples of convergent molecular evolution come from the
analysis of digestive proteins. Five independent lineages of animals have
convergently evolved stomachal-fermentation systems, as discussed in
chapter 2 (see table 2.10). Three of these lineages have also experienced
convergent molecular evolution of the enzymes found in those special-
ized stomachs (table 5.2), in which a conventional lysozyme was inde-
pendently recruited for digestive purposes in the stomach (Stewart et al.
1987; Kornegay et al. 1994; Zhang and Kumar 1997). The production of
the stomach lysozyme was independently produced in these three lin-
eages by the identical two nucleotide substitutions in the DNA coding
for amino acid sequences in the stomach lysozomes: replacement of
asparagine with aspartic acid at site 75 (substituting G for A in the first
position of the DNA codon for asparagine), and replacement of aspartic
acid with asparagine at site 87 (substituting A for G in the first position
of the DNA codon for aspartic acid). Interestingly, a third parallel nucle-
otide substitution occurred in the stomach lysozymes of the hoatzin, a
bird, and the ungulate mammals—the replacement of alanine with
glycine at site 76 (substituting G for C in the second position of the DNA
codon for alanine)—which does not occur in the langur monkey. This
184 Chapter 5
Table 5.2
Convergent evolution of identical nuclear DNA and mitochondrial DNA molecules coding
for digestive proteins
1 Convergent molecule and function: STOMACH LYSOZYME–ENCODING DNA

(conversion of conventional lysozymes to digestive stomach lysozymes produced by the
identical two nucleotide substitutions in the DNA coding for the stomach lysozymes)
1.1 Hoatzin (Amniota: Sauropsida: Archosauromorpha: Ornithodira: Dinosauria:
Saurischia: Theropoda: Aves: Opisthocomiformes: Opisthocomidae; Opisthocomus
hoatzin)
1.2 Cow (Amniota: Synapsida: Therapsida: Mammalia: Eutheria: Laurasiatheria:
Cetartiodactyla: Ruminantia: Bovidae; Bos taurus)
1.3 Langur monkey (Mammalia: Eutheria: Euarchontoglires: Primates: Cercopithecidae;
Presbytis entellus)
2 Convergent molecule and function: PANCREATIC RIBONUCLEASE–ENCODING
DNA (decrease in pH for the maximum ribonucleolytic activity of the enzyme produced
by the identical three nucleotide substitutions in the DNA coding for the pancreatic
RNase)
2.1 Asian douc langur monkey (Primates: Cercopithecidae: Colobinae; Pygathrix
nemaeus)
2.2 African guereza monkey (Primates: Cercopithecidae: Colobinae; Colobus guereza)
3 Convergent molecules and function: METABOLIC PROTEIN–ENCODING
MITOCHONDRIAL DNA (44 nonrandom parallel amino acid substitutions across all 13
mtDNA genes coding for oxidative phosphorylation metabolic proteins)
3.1 Australian bearded dragon (Lepidosauria: Squamata: Iguania: Agamidae; Pogona
vitticepes)
3.2 Blind snake (Lepidosauria: Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:
Serpentes: Typhlopidae; Typhlops diardi)
suggests a closer convergence in stomachal-fermentation function in the

hoatzins and ungulates, as the statistical analyses of Zhang and Kumar
(1997) found that the patterns of amino acid substitutions at these sites
in the enzyme were significantly unusual and could be taken as evidence
for the action of nonneutral natural selection. Similar to the convergent
molecular evolution in opsin proteins, Kornegay et al. (1994, 927) suggest
that there may be a limited number of ways to convert a conventional
vertebrate lysozyme-c gene into a stomach lysozyme gene, “a set of spe-
cific structural changes that are necessary and sufficient to convert a
lysozyme for stomach function even when different genes have been
separated by more than 300 million years of independent evolution,” as
in the case of the avian dinosaurs and the placental mammals (table 5.2).
Further study of the stomachal-fermenting colobine monkeys reveals
parallel evolutionary changes in pancreatic ribonuclease–encoding DNA
that occurred independently in Asian colobine monkeys and African

colobine monkeys (table 5.2), two lineages that diverged some 13 million
years ago (Prud’homme and Carroll 2006). Zhang (2006) reports that
three identical nucleotide substitutions in the DNA coding for the pan-
creatic RNase amino acid sequence independently occurred in these two
lineages: replacement of arginine with glutamine at site four (substituting
A for G in the second position in the DNA codon for arginine), replace-
ment of lysine by glutamic acid at site six (substituting G for A in the
first position in the DNA codon for lysine), and replacement of arginine
by tryptophan at site 39 (substituting T for C or A in the first position in
the DNA codon for arginine). This DNA coding change resulted in the
lowering of the pH of the maximum ribonucleolytic activity of the
enzyme from 7.4 to 6.3 (Zhang 2006). The end result is a pancreatic
RNase with maximum activities at more acid conditions than usual,
conditions that match those of the acidic environment of the colobine
monkey’s small intestine. Zhang (2006, 819) concludes that “the new
genes acquired enhanced digestive efficiencies through parallel amino
acid replacements driven by darwinian selection” and that “molecular
evolution has a certain degree of repeatability and predictability under
the pressures of natural selection.”
As in the nuclear genome data sets, initial analyses of some of the
mitochondrial genome data sets that do exist have produced startling
results. Castoe et al. (2009, 2010) have documented that 44 amino acid
substitutions have occurred in parallel in all 13 mitochondrially encoded
metabolic proteins in agamid lizards and snakes, an unexpected result
that they summarize by stating: “convergent molecular evolution in mito-
chondria can occur at a scale and intensity far beyond what has been
documented previously” (Castoe et al. 2009, 8986). Although they are
both squamate reptiles, the agamid lizards and the snakes are only dis-
tantly related: the agamid lizards are iguanians, a basal group of squa-
mates, whereas the snakes are highly derived anguimorphs, located many
nodes up in the squamate clade (table 5.2; see also the appendix).
Further, Castoe et al. (2009, 2010) argue that the observed convergent
evolution of mitochondrial DNA, mtDNA, is driven by natural selection.
The convergent sites occur primarily at the first and second positions in
the mtDNA codons, as is the case in the previously discussed cases of
nuclear DNA convergence. The affected mtDNA codes for core oxida-
tive phosphorylation proteins, which are some of the most highly con-
served proteins in the genome. As in the previously discussed cases of
convergent molecular evolution, Castoe et al. (2009, 2010) argue that
186 Chapter 5
there appear to be a limited number of ways in which amino acid

sequences can be modified and yet remain functional in these proteins,
and that sustained selective pressure may preferentially target the
mtDNA encoding these proteins if there are only a few possible ways to
achieve the same functional innovation. In addition, Naylor et al. (1995)
and Yee (1999) have argued that the mtDNA molecule itself may be
limited in its number of nucleotide permutations. They point out that the
functional requirement for hydrophobicity in many mitochondrial pro-
teins biases the second codon position in the encoding mtDNA to either
T or C, and that molecular evolution may be constrained to cycle between
these two nucleotide states at the second codon position. In summary,
Castoe et al. (2010, 67) conclude that “molecular convergence can happen
en masse in nature, affecting multiple genes . . . [and] the result implies
that the protein adaptive landscape is sometimes highly constrained.”
The extent of the newly discovered convergence in mtDNA has pro-
found implications for molecular phylogenetic analyses, implications that
Castoe et al. (2009, 8990) find worrisome: “contrary to widespread belief,
. . . non-neutral convergence can be a major force in molecular evolution,
and . . . it should be considered more seriously as a cause of phylogenetic
incongruence among datasets.” They point out that mtDNA molecular
phylogenies would place agamid lizards and snakes in the same clade,
whereas both morphological data and nuclear genome data place the
agamid lizards with the basal Iguania and the snakes many nodes away
in the highly derived Anguimorpha (table 5.2). But what about the
numerous cases in nature where the morphological data and the nuclear
genome data clash? In those cases, it is generally assumed that similari-
ties in morphological traits have arisen by convergent evolution and are
not synapomorphies, because the nuclear genome data can be trusted to
be almost free from molecular convergences. What if this is not true?
Other studies have identified parallel amino acid replacements across
families of closely linked genes, such as the histocompatibility complex
in humans (Yeager and Hughes 1999) and toxin-resistant sodium chan-
nels in puffer fish (Jost et al. 2008), that are also argued to be due to
nonrandom natural selection.
In an intriguing study, Rodríguez-Trelles et al. (2003) document the
convergent evolution of the aldehyde oxidase gene in two independent
gene-duplication events combined with identical amino acid replace-
ments in the duplicated DNA (table 5.3). The aldehyde oxidase gene, Ao,
which codes for the aldehyde oxidoreductase enzyme, AOX, is known to
have evolved by gene duplication of the xanthine dehydrogenase gene,
Table 5.3
Convergent evolution of identical nuclear DNA molecules coding for aldehyde oxidore-
ductase enzymes
Convergent molecule and function: ALDEHYDE OXIDOREDUCTASE–ENCODING

DNA (identical 10 nucleotide substitutions in the DNA coding for the aldehyde
oxidoreductase enzyme that catalyzes the oxidation of aldehydes into acids)
1 Asian rice (Eukarya: Bikonta: Chlorobiota: Embryophyta: Euangiosperms:
Monocotyledons: Poaceae; Oryza sativa)
2 Puffer fish (Eukarya: Unikonta: Metazoa: Vertebrata: Osteichthyes: Actinoptergyii:
Tetraodontidae; Fugu rubripes)
Xdh, in eukaryotes before the evolution of multicellular life. The AOX

enzyme is present in metazoans up to the evolution of the primitive
chordates, as it is still present in the urochordate sea peach Ciona intes-
tinalis, after which point it was evidently lost. The AOX enzyme was then
independently reevolved in the osteichthyan fishes by a second duplica-
tion event in the Xdh gene, producing a new Ao gene, and is present in
subsequent vertebrates, including humans.
What is particularly interesting in these gene duplication events is
the fact that “after each duplication, the Ao duplicate underwent a
period of rapid evolution during which identical sites across the two
molecules . . . were subjected to intense positive Darwinian selection”
(Rodríguez-Trelles et al. 2003, 13413), shifting the substrate acted upon
by the enzyme from hypoxanthine in the ancestral gene coding to alde-
hydes in the new genes. In total, 10 sites in the DNA of the two indepen-
dent Ao genes have identical nucleotide substitutions, as seen in the
match of the two independent AOX enzyme amino acid sequences. The
two convergent Ao genes “originated from Xdh by duplications sepa-
rated by approximately one billion years (i.e., the time span from the
origin of multicellular eukaryotes to the last common ancestor of verte-
brates),” yet the very same nucleotide substitutions still occurred in
the DNA of the independent gene duplicates due to natural selection
for the same enzymatic function (Rodríguez-Trelles et al. 2003, 13417).
At present, relatively few examples of adaptive convergent molecular
evolution at the level of DNA nucleotide sequences are known. However,
Castoe et al. (2010) point out that (1) we are only now beginning
to acquire large, complete genome data sets that can be analyzed for
convergence, (2) systematic surveys for molecular convergence are
usually not conducted, and (3) it is difficult to distinguish random, neutral
188 Chapter 5
molecular convergence from nonneutral, adaptive molecular conver-

gence. They suggest that the phenomenon may be much more widespread
than believed, and that “greater detection of adaptive convergence events
in nature may also provide important insight into the diversity of protein
adaptive landscapes in nature, and help to explain why in some cases
these landscapes appear highly constrained” (Castoe et al. 2010, 69).
Last, Stern and Orgogozo (2009, 747) also argue that the phenomenon
of parallel genetic evolution is much more common than believed, as in
their study they examined 350 different mutations that produce pheno-
typic variation and found that “more than half of these represent cases
of parallel genetic evolution.” Noting that some of these clearly involve
DNA coding for gene products (such as photosensitive optical proteins
and digestive proteins that we have considered previously; see tables 5.1
and 5.2, respectively), they point out that “gene function explains part
but not all of the observed pattern of parallel genetic evolution. In
several cases, parallelism has been observed even though mutations in a
large number of genes can produce similar phenotypic changes.”
Stern and Orgogozo (2008, 2009) and Stern (2010) point out that there
are two fundamentally different types of DNA coding for genes: one
DNA region that codes for the gene product (protein or RNA), and
another adjacent cis-regulatory DNA region that encodes the instruc-
tions determining when and where the gene product will be produced.
They point out that, given the molecular data presently at hand, the
overwhelming number of mutations that produce morphological varia-
tion between species occur in these cis-regulatory regions, and not in the
gene product regions. Stern and Orgogozo (2009, 750) further argue that
“[a]lthough mutations are thought to occur randomly in the genome, the
distribution of mutations that cause biological diversity appears to be
highly nonrandom” (that is, preferentially in cis-regulatory DNA), and
that the “genetic basis of phenotypic evolution thus appears to be some-
what predictable.” Convergent molecular evolution is here seen to be a
product of preferential mutational changes in a limited subset of cis-
regulatory DNA regions, which they term “hot spots”: “evolutionary
relevant mutations are expected to accumulate in a few hot-spot genes
and even in particular regions within a single gene” that control develop-
ment (Stern and Orgogozo 2009, 750).
To summarize this section of the chapter, at the DNA level of molecu-
lar evolution we have examined examples of (1) the convergent evolu-
tion of identical nuclear DNA molecules and (2) the convergent evolution
of identical mitochondrial DNA molecules, both produced by identical
nucleotide substitutions occurring in the molecules of very distantly

related organisms. The DNA molecule possesses only four nucleotides,
and has a relatively simple geometry: the double helix. In the next section
we shall examine the convergent evolution of the complex geometries
and chemical compositions of the huge protein molecules.
Proteins and Protein Functions
The independent evolution of two convergent protein molecules with

the same amino acid sequences does not necessarily mean that the DNA
coding for those two convergent proteins is itself exactly the same. This
is due to the fact that most amino acids are coded for by more than one
DNA codon; for example, the six codons TCT, TCC, TCA, TCG, AGT,
and AGC all code for the same amino acid, serine. Identical amino acid
sequences may be encoded by two quite different DNA nucleotide
sequences. Thus, the phenomenon of convergent evolution in protein
molecules is at least semi-independent of the phenomenon of convergent
evolution at the level of the DNA molecules.
Indeed, Chen et al. (1997) have demonstrated the convergent evolu-
tion of an antifreeze protein molecule in two independent groups of
cold-water fishes that are encoded by totally different DNA (Logsdon
and Doolittle 1997). The convergent molecule is an antifreeze glycopro-
tein, AFGP, found in paracanthopterygian gadiform fish in the Arctic
waters of the Northern Hemisphere of the Earth; it is also found in
acanthopterygian perciform fish in Antarctic waters on the opposite
side of the planet (table 5.4). The two groups of fishes are adapted for
existence in subfreezing temperatures, and have evolved independently
in isolation at the two poles of the planet, geographically separated
by the wide expanse of temperate- and tropical-temperature waters
in between.
The AFGP of the Antarctic notothenioid fish is encoded by a gene
that is largely constructed of previously noncoding DNA, combined with
bits of DNA coding for a pancreatic trypsinogen gene that is used in
digestion. This gene may have been modified for preventing freezing in
the intestinal fluid, a function that was later expanded throughout the
circulatory system in the evolution of the new AFGP gene, whereas the
AFGP gene of the Arctic cod fish shows no DNA nucleotide sequence
identity with the trypsinogen gene, and arose from a completely unre-
lated genomic locus (Chen et al. 1997; Logsdon and Doolittle 1997).
Nonetheless, the Antarctic notothenioid and Arctic cod AFGP molecules
190 Chapter 5
Table 5.4
Convergent evolution of identical antifreeze protein molecule in independent lineages, and
of different molecules that serve the same antifreeze molecular function
A Convergent molecule (same molecule, different convergent lineages)

1 Convergent protein and function: AFGP (blood serum glycoprotein used to lower the
freezing temperature of tissue to prevent cell rupture)
1.1 Artic cod (Osteichthyes: Actinopterygii: Teleostei: Euteleosti: Paracanthopterygii:
Gadiformes: Gadidae; Boreogadus saida)
1.2 Antarctic notothenioid (Osteichthyes: Actinopterygii: Teleostei: Euteleosti:
Acanthopterygii: Perciformes: Notothenidae; Dissostichus mawsoni)
2 Convergent protein and function: AFP Type II (blood serum protein used to lower the
freezing temperature of tissue to prevent cell rupture)
2.1 Atlantic herring (Osteichthyes: Actinopterygii: Teleostei: Clupeomorpha:
Clupeiformes: Clupeidae; Clupea harengus)
2.2 American smelt (Osteichthyes: Actinopterygii: Teleostei: Euteleosti:
Protocanthopterygii: Osmeriformes: Osmeridae; Osmerus mordax)
2.3 Sea raven (Osteichthyes: Actinopterygii: Teleostei: Euteleosti: Acanthopterygii:
Scorpaeniformes: Cottidae; Hemitripterus americanus)
B Convergent molecular function (same function, different convergent molecules):
Molecular function: Antifreeze protection
Convergent molecules:
1 AFP Type I
Lineage: Winter flounder (Osteichthyes: Actinopterygii: Teleostei: Euteleosti:
Acanthopterygii: Pleuronectiformes: Pleuronectidae; Pleuronectes americanus)
2 AFP Type III
Lineage: Ocean pout (Osteichthyes:Actinopterygii:Teleostei: Euteleosti:Acanthopterygii:
Perciformes: Zoarcidae; Macrozoarces americanus)
3 AFP Type IV
Lineage: Longhorn sculpin (Osteichthyes: Actinopterygii: Teleostei: Euteleosti:
Acanthopterygii: Scorpaeniformes: Cottidae; Myoxocephalus octodecimspinosus)
4 AFP Cf
Lineage: Spruce budworm (Arthropoda: Mandibulata: Hexapoda: Lepidoptera:
Tortricidae; Choristoneura fumiferana)
5 AFP Tm
Lineage: Yellow mealworm beetle (Arthropoda: Mandibulata: Hexapoda: Coleoptera:
Polyphaga: Tenebrionidae; Tenebrio molitor)

are nearly identical in amino acid composition and are comprised mainly
of threonine-alanine-alanine repeats, usually encoded by ACA-GCT/G-
GCA codons in the Antarctic notothenioids, whereas the Arctic cod use
DNA codons rarely observed in the notothenioid fish (Logsdon and
Doolittle 1997). The convergent amino acid triplet-repeating structure of
AFGP appears to be required for antifreeze-ice interaction to occur, and
probably evolved in the notothenioids about 10 to 14 million years ago,
as the Antarctic Ocean began to freeze, and about two to three million
years ago in the cod, during the Arctic glaciation (Chen et al. 1997;
Logsden and Doolittle 1997).
Subsequent work has revealed the convergent evolution of yet another
antifreeze molecule: AFP Type II. This molecule evolved independently
in cold-water clupeomorph fish (Atlantic herring) and in two clades of
euteleostian fishes, the protocanthopterygian American smelts and the
acanthopterygian sea ravens (table 5.4). Unlike in the case of AFGP, the
convergent evolution of AFP Type II in these three fish groups appears
to have been a case of parallel independent changes in the same C-type
lectin genes ancestrally found in all three groups (Fletcher et al. 2001).
Convergence in antifreeze protein function does not always produce
convergent evolution, or parallel evolution, of the same antifreeze mol-
ecules, as in the case AFGP or AFP Type II. Consider the convergent
evolution of wings in insects, pterosaurs, birds, and bats (see table 2.2),
all of which serve the common function of generating lift for powered
flight. The structure of the wing is different in each animal group, but the
function is the same. Fletcher et al. (2001) point out that three additional
antifreeze proteins have evolved in the cold-water fishes—AFP Types I,
III, and IIV—and Davies et al. (2002) discuss the evolution of two more
antifreeze proteins that have evolved in the arthropods, AFP Cf and Tm
(table 5.4).
The amino acid sequences in these additional antifreeze proteins are
different, but aspects of their geometries are very similar (Davies et al.
2002), and their functions are the same. In particular, Davies et al. (2002,
932) point out that the helical geometries of AFP Cf and AFP Tm are
such that “their ice-binding sites are virtually superimposable” even
though AFP Cf is a left-coiled β-helix and AFP Tm is a right-coiled β-
helix. Davies et al. (2002, 932) describe this geometric result as a “remark-
able example of convergent evolution” in that the “two ranks of threonine
residues line up perfectly because the threonines are in the same rota-
meric configuration” in the helices.
192 Chapter 5
Preventing tissue freezing is a specialized problem in a restricted eco-

logical setting, namely the polar regions of the Earth. Transporting
oxygen from one region of tissues to another is a functional challenge
that is ubiquitous to life itself. Three different types of proteins have
convergently evolved to fulfill the same oxygen-binding function: hemo-
globin, hemocyanin, and hemerythrin (table 5.5). Of these three, hemo-
globin is by far the most widespread in life on Earth (Hardison 1996).
All three molecules are ancient: Bailly et al. (2008) argue that α-helical
bundles were the earliest protein folds to emerge in the evolution of life,
as they are well adapted to binding metal ions. Thus, the globins and the
hemerythrins represent two ancient families of molecules that evolved
the functions of sequestering reduced iron and sensing and controlling
oxygen concentrations, as corrosive oxygen was lethal to the earliest
Table 5.5
Convergent evolution of similar oxygen-binding protein molecules in independent lin-
eages, and of different molecules that serve the same oxygen-binding molecular function
A Convergent molecule (same molecule, different convergent lineages):

1 Convergent protein and function: HEMOCYANIN (protein that uses copper to bind
oxygen for the function of oxygen transport)
1.1 Common octopus (Protostomia: Lophotrochozoa: Eutrochozoa: Mollusca:
Cephalopoda: Octopodidae; Octopus vulgaris)
1.2 Horseshoe crab (Protostomia: Ecdysozoa: Arthropoda: Cheliceriformes: Limulidae;
Limulus polyphemus)
2 Convergent protein and function: HEMERYTHRIN (co-option of the ancient nonheme
oxygen-sensor molecule hemerythrin for the function of oxygen transport)
2.1 Terebratulid brachiopod (Protostomia: Lophotrochozoa: Lophophorata:
Brachiopoda: Cancellothyrididae; Terebratulina retusa)
2.2 Serpulid bristle worm (Protostomia: Lophotrochozoa: Eutrochozoa: Spiralia:
Sipuncula: Sipunculidae; Sipunculus nudus)
2.3 Priapulid worm (Protostomia: Ecdysozoa: Introverta: Priapulida: Priapulidae;
Priapulus caudatus)
3 Convergent protein and function: two-CCP-domain HAPTOGLOBIN (modified
hemoglobin-binding plasma protein with the ability to form high-molecular-weight
oligomers)
3.1 Cow (Eutheria: Laurasiatheria: Cetartiodactyla: Ruminantia: Bovidae; Bos taurus)
3.2 Human (Eutheria: Euarchontoglires: Primates: Catarrhini: Hominidae; Homo
sapiens)
B Convergent molecular function (same function, different convergent molecule):
Molecular function: oxygen transport; convergent molecule: HEMOGLOBIN
Lineage: Earthworm (Protostomia: Lophotrochozoa: Eutrochozoa: Spiralia: Annelida:
Lumbricidae; Lumbricus terrestris)

forms of anaerobic life (Bailly et al. 2008). With the evolution of aerobic
life, transport of oxygen within the organism for metabolic purposes is a
critical function, and van Holde et al. (2001, 15566) hypothesize that
oxygen-transport proteins developed “in several independent ways,
hemoglobins from myoglobins, hemerythrins from myohemerythrins,
and the two kinds of hemocyanins from two different classes of phenol
oxidases.”
Hemocyanins evolved twice, independently, in the clade of the mol-
luscs (lophotrochozoans) and the clade of the arthropods (ecdysozoans;
table 5.5). In the molluscs, Class 3a hemocyanins are used for oxygen
transport in the blood of cephalopods, such as the octopus, and of gas-
tropods, including both marine and terrestrial snails. In the arthropods,
Class 3b hemocyanins are present in the blood of cheliceriformes, such
as the horseshoe crab, and mandibulate crustaceans, such as the true
crabs (van Holde et al. 2001).
The hemocyanins bind oxygen with copper (thus, the blood of an
octopus is bluish-green when it is oxygenated, not red like ours), but
copper used as an oxygen-binding metal is not as efficient as iron. Both
hemoglobin and hemerythrin use iron, but the latter is a nonheme protein
(despite its deceptive name) unlike hemocyanin and hemoglobin, and is
less efficient in oxygen transport than hemoglobin. The hemerythrins are
patchily present throughout life, from bacteria to archaea to the eukarya,
and are used in a variety of functions, the most ancient of which is oxygen
sensing (Bailly et al. 2008). Only late in the evolution of life have three
groups of protostome invertebrates—the brachiopods, serpulids, and pri-
apulids (table 5.5)—independently co-opted the hemerythrin protein for
the primary function of oxygen transport. Blood sufficiently enriched in
hemerythrin is pale violet when oxygenated, not blue-green as for hemo-
cyanin or red as for hemoglobin.
Last, mammals possess a specialized plasma protein, haptoglobin, that
can be used to inhibit hemoglobin’s oxidative activity, if needed. A
variant of this protein, two-CCP-domain haptoglobin, independently
evolved in the ruminant mammals and, curiously, in humans (table 5.5).
Because the two-CCP-domain haptoglobin is found throughout the
clade of the ruminants, the mutation producing this protein occurred
early in the Cenozoic for these mammals (Wicher and Fries 2007). The
same protein evolved much more recently in humans, for it is not present
in our close primate cousins. In both cows and humans, the convergent
duplication of the gene segment coding for the CCP domain of the α-
chain in the haptoglobin molecule resulted in a much longer α-chain in
194 Chapter 5
the new protein, which contains two CCP domains in tandem. Wicher
and Fries (2007) point out that the new protein can form much higher-
molecular-weight oligomers than the conventional protein, and suggest
that the different binding properties of this molecule is useful in prevent-
ing infection in the ruminant mammals. Its function in humans is less
clear, but data do exist that indicate humans with the mutated haptoglo-
bin molecule are more resistant to life-threatening streptococcus infec-
tions, and possibly to malarial infections (Wicher and Fries 2007).
Preventing tissue death by freezing or oxygen starvation is a matter
of life or death, yet convergent molecular evolution also occurs in situ-
ations that are not so dire. Mundy (2005) reviews the recent discovery
that melanic polymorphism in three groups of distantly related wild birds
has evolved in parallel, in that the plumage color variations seen in these
birds are produced by point mutations in amino acid substitution in the
same gene coding for the MC1R protein (table 5.6). Many birds are well
known for the dazzling colors of their plumage, and so it was unexpected
to find parallel changes in the same gene to be producing the color varia-
tions in such distantly related species. Mundy (2005, 1638) concludes by
noting that the “examples discussed here clearly only represent the tip
of the iceberg as far as avian plumage colouration genetics in the wild is
concerned. Over 300 avian species exhibit plumage polymorphisms”;
thus, he implies that more examples of parallel evolution of the MC1R
protein may soon be discovered.
Likewise, Protas et al. (2005) have shown that mutations in the
protein-encoding sequences of the same gene, the one coding for
Table 5.6
Convergent evolution of similar melanin-controlling protein molecules
1 Convergent protein and function: MC1R (protein producing plumage color variations
in birds)
1.1 Snow goose (Aves: Neognathae: Galloanserae: Anseriformes: Anatidae; Anser
caerulescens)
1.2 Bananaquit (Aves: Neognathae: Neoaves: Passeriformes: Coerebidae; Coereba
flaveola)
1.3 Arctic skua (Aves: Neognathae: Neoaves: Charadriiformes: Stercorariidae;
Stercorarius parasiticus)
2 Convergent protein and function: OCA2 (protein producing albinism in fish)
2.1 Molino cave populations of the Mexican tetra, Astyanax mexicanus
2.2 Pachón cave populations of the Mexican tetra, Astyanax mexicanus
2.3 Japonés cave populations of the Mexican tetra, Astyanax mexicanus

OCA2, produces albinism in three independent populations of cave fish

(table 5.6). Animals living in the perpetually dark environment of caves
routinely lose the function of their eyes and experience pigment loss. This
same evolutionary trend repeatedly occurs across phylogenetically
diverse lineages, in very distantly related organisms such as spiders,
isopods, fish, and salamanders. As a result, it was a surprise when, in
studying pigment loss in the Mexican tetra cavefish in two widely sepa-
rated caves in Mexico, Molino and Pachón, Protas et al. (2005, 109)
discovered that “albinism evolved independently in these two caves, by
convergent evolution in the same gene.” Preliminary analyses indicate
that yet another independent cave population of these fish, Japonés,
also has experienced melanin loss by convergent evolution of the gene
producing OCA2. The independent evolution of albinism was an expected
given, but the fact that three independent populations of cavefish
experienced mutations in the very same gene to produce that albinism
was not.
An even more unexpected demonstration of convergent evolution in
the same gene has been discovered by Colosimo et al. (2005) in 14 inde-
pendent freshwater populations of the three-spine stickleback fish
species Gasterosteus aculeatus (table 5.7). The Gasterosteus aculeatus
species complex contains two strikingly different end-member pheno-
types: a heavily armored ecomorph, typically possessing a row of over
30 bony armor plates extending along the length of the body of the fish,
and a lightly armored ecomorph, possessing at most nine plates only at
the anterior end of the fish. The heavily armored ecomorph is found
exclusively in marine waters, and the lightly armored ecomorph in fresh-
water streams and lakes. These end-member phenotypes are so different
that they were classified as separate species by the great French mor-
phologist Georges Cuvier in 1829, but subsequent observation has dem-
onstrated that the two ecomorphs freely interbreed and hybridize
(Colosimo et al. 2005).
Of particular interest is the fact that the lightly armored ecomorph
repeatedly and independently evolved from members of the heavily
armored marine ecomorph that invaded the new freshwater stream
and lake habitats which had been created around the world following
the retreat and melting of the glaciers from 20,000 to 10,000 years
ago. The greater maneuverability and flexibility of the lightly armored
ecomorph is believed to be of selective advantage in freshwater pred-
atory regimes, as opposed to open-ocean marine ones (Colosimo
et al. 2005).
196 Chapter 5
Table 5.7
Convergent evolution of the EDA protein in independent freshwater populations of the
three-spine stickleback fish species Gasterosteus aculeatus
Convergent protein and function: EDA (protein used in constructing rows of bony armor
plates on the bodies of stickleback fish)
Convergent populations:
A European populations
1 Schwale (Germany) freshwater G. aculeatus
2 Loch Fada (Scotland) freshwater G. aculeatus
3 Norway stream (Norway) freshwater G. aculeatus
4 Blautaver (Iceland) freshwater G. aculeatus
B Eastern North American population
5 Olmstead Park (Massachusetts) freshwater G. aculeatus
C Western North American populations
6 Santa Clara River (California) freshwater G. aculeatus
7 Friant (California) freshwater G. aculeatus
8 San Francisco Creek (California) freshwater G. aculeatus
9 Conner Creek (Washington) freshwater G. aculeatus
10 Paxton Lake (Canada) freshwater G. aculeatus
11 Salmon River (Canada) freshwater G. aculeatus
12 Wallace Lake (Alaska) freshwater G. aculeatus
13 Alaska stream (Alaska) freshwater G. aculeatus
D Japanese population
14 Nakagawa Creek (Japan) freshwater G. aculeatus
The protein EDA is used in constructing the body armor of Gaster-

osteus aculeatus, and Colosimo et al. (2005) have demonstrated that
four mutations in the Eda protein-encoding gene repeatedly and inde-
pendently occur in 13 freshwater populations of lightly armored eco-
morphs around the world. One exception to the convergent evolution
of the freshwater stickleback Eda gene is the Nakagawa Creek popula-
tion in Japan (table 5.7), a population with a lightly armored phenotype
that “is due to an independently derived allele of Eda” different from
those of the other 13 freshwater populations examined (Colosimo et al.
2005, 1929).
Other examples of the convergent evolution of structural proteins are
due not to convergent changes in the same gene in independent lineages,
but to identical structural repeats in the amino acid sequences compris-
ing those proteins. Four different types of fibroelastic structural proteins
have independently evolved in the arthropods and vertebrates, all of
which have multiple glycine-glycine-leusine-glycine-tyrosine amino acid
repeats in the protein (table 5.8). The amino acid sequence convergences
Table 5.8
Convergent evolution of similar structural protein molecules
1 Convergent proteins and functions: OOTHECIN, CHORION CLASS B, LAMPRIN,

and ELASTIN (fibroelastic structural proteins used in the construction of chordate
cartilages, arthropod silks, and insect egg cases)
1.1 American cockroach (Bilateria: Protostomia: Ecdysozoa: Arthropoda: Hexapoda:
Blattodea: Blatellidae; Periplaneta americana)
1.2 Domestic silk moth (Arthropoda: Hexapoda: Lepidoptera: Bombycidae; Bombyx
mori)
1.3 Sea lamprey (Bilateria: Deuterostomia: Chordata: Vertebrata: Petromyzontiformes:
1.4 Norwegian rat (Chordata: Vertebrata: Osteichthyes: Sarcopterygii: Reptiliomorpha:
Amniota: Synapsida: Therapsida: Mammalia: Eutheria: Muridae; Rattus norvegicus)
2 Convergent proteins and functions: QID74 and BR3 (proteins used to construct cell
walls by filamentous fungus and to construct external tube walls by midges)
2.1 Mycoparasitic filamentous fungus (Unikonta: Fungi: Eumycetes: Ascomycota:
Hypocreaceae; Trichoderma harzianum)
2.2 Midge (Unikonta: Metazoa: Protostomia: Arthropoda: Hexapoda: Chironomidae;
Chironomus tentans)
3 Convergent protein and function: APOLIPOPROTEIN(a) (adaptive function uncertain;
possibly protects against certain infectious diseases by inhibiting plasminogen activation.
Otherwise similar to lipoproteins that bind fibrin and deliver cholesterol in reconstructing
damaged tissue)
3.1 European hedgehog (Eutheria: Laurasiatheria: Eulipotyphles: Erinaceidae;
Erinaceus europeus)
3.2 Human (Eutheria: Euarchontoglires: Primates: Catarrhini: Hominidae; Homo
sapiens)
seen in these proteins are “driven by structural or functional properties

imparted to the protein by the shared sequence,” producing the ability
of these “predominantly hydrophobic proteins to self-organize into a
polymeric, fibrillar matrix” (Robson et al. 2000, 1751). Twenty-eight of
the amino acids in the cockroach oothecin protein are in precisely the
same sequence in the sea lampry lamprin protein, an amino acid sequence
convergence that “may represent one of the best examples of primary
sequence convergence so far identified” (Robson et al. 2000, 1751). These
two animals are very distantly related: the cockroach is a protostomous
bilaterian, and the sea lampry is a deuterostome.
Rey et al. (1998, 6212) report an “unexpected homology” found in
comparing a fungal and an insect protein: fully 25.3 percent of the amino
acid sequence is identical in the fungal Qid74 protein and the insect BR3
protein (table 5.8). The fungal protein is used in cell wall construction,
198 Chapter 5
whereas the insect protein is secreted in the midge’s saliva, and is used
in constructing external tube walls within which the insect lives. Rey et
al. (1998, 6215) note that the two genes coding for the two proteins, one
fungal and one insect, had to have originated independently of each
other, and that the “ancestor of each new gene created de novo has to
be a noncoding DNA sequence that abounds in every eukaryotic
genome”; furthermore, the observed protein convergence “was not
between the genes but between the similar noncoding repeats whence
the two [new] genes sprung quite independently of each other.”
More enigmatic is the convergent evolution of the protein
apolipoprotein(a), or apo(a), in the hedgehog and in humans (table 5.8).
Apo(a) is part of the low-density lipoprotein complex (LDL) that is a
known risk factor in the development of atherosclerosis in humans,
hence its research interest in medicine. It is known that apo(a) is found
in Old World primates (Catarrhini), but it was a surprise to find the same
protein in the very distantly related European hedgehogs: Lawn et al.
(1997, 11992) conclude that it is “a remarkable example of ‘parallel gene
evolution.’ . . . By apparent remodeling of a plasminogen-like gene,
hedgehog and human ancestors independently evolved an apo(a)
protein” that functions identically in both organisms. Unfortunately,
some of those functions increase the risk of cerebral stroke in humans,
which is one of the reasons the protein was under study in the first place.
Last, we have the convergent evolution of protein function in which
multiple independent proteins with their own independent functions are
co-opted to serve the convergent protein function, a process known as
gene sharing. Convergent evolution of protein function by gene sharing
was first discovered in the crystallin proteins of the eye lens in animals:
“The term ‘gene sharing’ was used initially to generalize the finding that
crystallins with a structural, optical function in the lens may also be
expressed in other tissues [in the body], where they have a metabolic,
nonrefractive function. Thus, the protein encoded in the identical gene
may perform entirely different functions depending upon its expression
pattern . . . gene regulation is a critical factor for gene sharing, leading
to the use of a metabolic [enzyme] or stress protein as a structural
crystallin. It is ironic that the abundantly expressed crystallins, long
considered inert bricks as it were, squirreled away in a tiny transparent
tissue . . . have become landmarks for multifunctional uses of widely
expressed proteins that illustrate the dynamism of evolutionary change!”
(Piatigorsky 2007, 55–56; author’s emphasis).
Subsequent analysis has revealed the convergent evolution of J-

crystallins, Ω-crystallins, S-crystallins, and drosocrystallins in invertebrate
animals (table 5.9). The α-, β-, and γ-crystallins are found in the eye lenses
of all known vertebrates, but the vertebrate βγ-crystallins are believed
to have evolved from the Ci-βγ-crystallin of the urochordates (table 5.9),
and thus are not independent originations (Piatigorsky 2007). Within the
vertebrates, δ-crystallins independently evolved in the sauropsid clade,
and are present in modern reptiles and avian dinosaurs. Other crystallins
are more restricted in their phylogenetic distribution, and are called
“taxon-specific.” These include the ρ-, π-, ι-, and ρB-crystallins in some
frogs and geckos, τ-crystallins in some fish and turtles, and ε-crystallins
in the archosaurs (table 5.9). Within the mammals, υ-crystallins indepen-
dently evolved in the monotremes, μ-crystallins in the marsupials, and
η-, ζ-, and λ-crystallins in separate branches of the placental mammal
clade (table 5.9). In conclusion, Piatigorsky (2007, 65) states: “the diver-
sity and taxon-specificity of lens crystallins throughout the animal
kingdom are consistent with overlapping functions by different proteins,
indicative of convergent evolution for crystallin roles.”
To summarize this section of the chapter, at the protein level of molec-
ular evolution we have examined examples of (1) the convergent evolu-
tion of similar amino acid sequences in unrelated proteins, (2) the
convergent evolution of similar structural geometries in proteins with
different amino acid sequences, and (3) the convergent evolution of
similar protein function by gene sharing, in which the same protein-
encoding gene is used for two entirely different functions depending
upon its regulatory expression. Proteins converge in every possible way.
Enzymes and Enzyme Functions
Enzymes are specialized protein molecules that catalyze the chemical

reactions of other molecules. Theoretically, it is the catalytic function
itself that is subject to natural selection, not the specific molecule produc-
ing that function, hence opening the possibility that different types of
molecular structures might be selected if they could be modified to
produce the same catalytic function. Thus, the phenomenon of conver-
gent evolution in enzyme function is at least semi-independent of the
phenomenon of convergent evolution of identical protein molecules.
We have previously examined the convergent evolution of some
enzymes, such as stomach lysozyme (table 5.1) and aldehyde oxidore-
200 Chapter 5
Table 5.9
Convergent evolution of crystallin protein molecules by gene sharing
Convergent molecular function: Construction of a transparent eye lens that optimizes the
refractive index for focusing and image formation
1 J1-, J2-, J3-CRYSTALLINS
Lineage: Box jellyfish (Metazoa: Cnidaria: Cubozoa: Carybdeidae; Tripedalia
cystophora)
2 Ω/L-CRYSTALLIN
Lineage: Sea scallop (Metazoa: Bilateria: Protostomia: Lophotrochozoa: Mollusca:
Bivalvia: Pectinidae; Placopecten magellanicus)
3 SL11/Lops4-, S-CRYSTALLINS
Lineage: Octopus (Protostomia: Lophotrochozoa: Mollusca: Cephalopoda: Octopodidae;
Octopus vulgaris)
4 DROSOCRYSTALLIN
Lineage: Fruit fly (Protostomia: Ecdysozoa: Arthropoda: Hexapoda: Drosophilidae;
Drosophila melanogaster)
5 Ci-βγ-CRYSTALLIN
Lineage: Sea squirt (Bilateria: Deuterostomia: Urochordata: Ascidiacea: Cionidae;
Ciona intestinalis)
6 τ-CRYSTALLIN
Lineage: Sea lamprey (Deuterostomia: Chordata: Vertebrata: Petromyzontiformes:
7 ρ-CRYSTALLIN
Lineage: Leopard frog (Chordata: Osteichthyes: Sarcopterygii: Tetrapoda:
Batrachomorpha: Lissamphibia: Anura: Ranidae; Rana pipiens)
8 π-, ι-, ρB-CRYSTALLINS
Lineage: Striped day gecko (Tetrapoda: Reptiliomorpha: Amniota: Sauropsida:
Lepidosauromorpha: Scleroglossa: Gekkonidae; Phelsuma lineata)
9 ε-CRYSTALLIN
Lineage: Crocodile (Sauropsida: Archosauromorpha: Crurotarsi: Crocodylidae;
Crocodylus niloticus)
10 δ1-, δ2-CRYSTALLINS
Lineage: Chicken (Sauropsida: Archosauromorpha: Ornithodira: Dinosauria: Saurischia:
Theropoda: Maniraptora: Aves: Galliformes: Phasianidae; Gallus gallus)
11 υ-CRYSTALLIN
Lineage: Platypus (Amniota: Synapsida: Therapsida: Mammalia: Monotremata:
12 μ-CRYSTALLIN
Lineage: Kangaroo (Mammalia: Marsupialia: Diprotodontia: Macropodidae; Macropus
rufus)
13 η-CRYSTALLIN
Lineage: Elephant shrew (Mammalia: Eutheria: Afrotheria: Macroscelidea:
Macroscelididae; Elephantulus rozeti)
14 ζ-CRYSTALLIN
Lineage: Camel (Mammalia: Eutheria: Laurasiatheria: Cetartiodactyla: Camelidae;
Camelus dromedarius)
15 λ-CRYSTALLIN
Lineage: Rabbit (Mammalia: Eutheria: Euarchontoglires: Lagomorpha: Leporidae;
Oryctolagus cuniculus)

ductase (table 5.3), that is caused by the convergent evolution of the

DNA directly coding for those proteins. Therefore, in these instances
the convergent enzyme function is a result of convergent evolution of
the same protein producing that function. In a case of “convergent evolu-
tion of similar enzymatic function on different protein folds,” Bork et al.
(1993, 31) argue that the “enzymatic function of sugar phosphorylation
appears to have evolved independently . . . by convergent evolution” on
three different protein geometries: the hexokinase, ribokinase, and galac-
tokinase families of protein folds (table 5.10). Although proteins in each
of these three families have different three-dimensional folds and strik-
ingly different sequence patterns, “each catalyzes chemically equivalent
reactions on similar or identical substrates” (Bork et al. 1993, 31). More-
over, enzymatic functions for specific sugars have independently evolved
within the three kinase families: fructoskinase specificity has indepen-
dently evolved in both the hexokinase and ribokinase families, and
glucose specificity has evolved several times independently within the
hexokinase family. Bork et al. (1993, 38) argue that two different types
of molecular convergence are exhibited by the kinase enzymes: (1) from
different protein structures to the same enzymatic function, and (2) from
Table 5.10
Convergent evolution of different enzyme molecules that serve the same metabolic
catalytic function
1 Convergent enzyme function: SUGAR PHOSPHORYLATION (enzymes that catalyze

the first step in sugar metabolism)
1.1 HEXOKINASE family
Lineage: Human (Eukarya: Unikonta: Metazoa: Deuterostomia: Vertebrata: Hominidae;
Homo sapiens)
1.2 RIBOKINASE family
Lineage: E. coli bacterium (Eubacteria: Proteobactia: γ-Proteobacteria; Escherichia
coli)
1.3 GALACTOKINASE family
Lineage: Lactobacillus (Eubacteria: Firmicutes: Bacillus group; Lactobacillus helveticus)
2 Convergent enzyme function: β-ELIMINATION MECHANISM (enzymes that catalyze
the β-elimination of sugar uronic acids)
2.1 POLYGALACTURONIC ACID LYASE Pel10A
Lineage: Cellvibrio soil bacterium (Eubacteria: Proteobactia: γ-Proteobacteria;
Cellvibrio japonicus)
2.2 POLYGALACTURONIC ACID LYASE Pel1C
Lineage: Erwinia plant-pathogen bacterium (Eubacteria: Proteobactia: γ-Proteobacteria;
Erwinia chrysanthemi)

202 Chapter 5
different functional specificities to similar functional specificities within

the same protein structure.
In examining enzymes that catalyze the β-elimination of sugar
uronic acids, Charnock et al. (2002, 12067) found that the “active center
provides a stunning example of convergent evolution” between two bac-
terial enzymes (table 5.10). The catalytic center of the Cellvibrio soil
bacterium’s Pel10A enzyme is “isostructural with the catalytic center of
the totally unrelated family PL-1 enzyme Pel1C” of the Erwinia plant-
pathogen bacterium (Charnock et al. 2002, 12067), an “essentially identi-
cal disposition of six active-center groups despite no topological similarity
between these enzymes” in a case of “convergent evolution of catalytic
mechanism” (Charnock et al. 2002, 12070).
Serine protease enzymes serve a variety of important functions, from
digestive to blood clotting. Reardon and Farber (1995) demonstrate that
three very different serine proteases have converged on the same enzy-
matic function, peptide bond hydrolysis, in bacteria, plants, and animals
(table 5.11). In their analysis of the α/β barrel proteins, they cite these
three serine proteases as “the best example of convergent evolution to
a similar active site,” noting that the “three proteins have entirely differ-
ent 3-dimensional structures” and yet the “catalytic triad is essentially
identical in the three proteins” (Reardon and Farber 1995, 501). They
further suggest that this convergence is due more to physics than selec-
tion: “convergent molecular evolution occurs when there is only one
energetically reasonable pathway for a chemical reaction,” such as
peptide bond hydrolysis (Reardon and Farber 1995, 497).
Tripp et al. (2001) document the convergent evolution of carbonic
anhydrases in the archaea, plants, and animals (table 5.11). The α-, β-,
and γ-classes of carbonic anhydrase found in these independent lineages
have no significant amino acid sequence identity, and their very different
fold geometries “underscore their independent origins. Despite gross
structural differences, the active sites of all three classes function with
[a] single zinc atom . . . [and] all three classes employ a two-step iso-
mechanism” (Tripp et al. 2001, 48615).
We have previously examined the convergent evolution of the stomach
lysozymes (table 5.1), used in the digestion of cellulose. Another difficult
molecule to digest is chitin. Shen and Jacobs-Lorena (1999) document
the convergent evolution of chitinases in animals and plants (table
5.11). Arthropods use chitinase to digest the chitin of their exoskeletons
during molting, whereas plants use chitinase to defend themselves against
fungal and insect infections by breaking down the structural chitin
Table 5.11
Convergent evolution of different enzyme molecules that serve the same catalytic
function
1 Convergent enzyme function: PEPTIDE BOND HYDROLYSIS (enzymes that catalyze

the hydrolysis of a peptide bond using a Ser nucleophile)
1.1 SUBTILISIN
Lineage: Hay bacillus (Eubacteria: Firmicutes: Bacillus group; Bacillus subtilis)
1.2 WHEAT SERINE CARBOXYPEPTIDASE
Lineage: Bread wheat (Eukarya: Bikonta: Chlorobionta: Tracheophyta: Angiospermae:
Poaceae; Triticum aestivum)
1.3 CHYMOTRYPSIN
Lineage: Cow (Eukarya: Unikonta: Metazoa: Deuterostomia: Vertebrata: Bovidae; Bos
taurus)
2 Convergent enzyme function: CO2 HYDRATION (enzymes that catalyze the reversible
hydration of carbon dioxide)
2.1 γ-Class CARBONIC ANHYDRASE
Lineage: Methanosarcinalean archaeon (Archaea: Euryarchaeota: Methanosarcinales;
Methanosarcina thermophila)
2.2 β-Class CARBONIC ANHYDRASE
Lineage: Pea (Eukarya: Bikonta: Chlorobiota: Tracheophyta: Angiospermae: Fabaceae;
Pisum sativum)
2.3 α-Class CARBONIC ANHYDRASE
Lineage: Human (Eukarya: Unikonta: Metazoa: Deuterostomia: Vertebrata: Hominidae;
Homo sapiens)
3 Convergent enzyme function: CHITIN DIGESTION (digestive enzymes that break
down glycosidic bonds in chitin)
3.1 Plant CHITINASE CBD
Lineage: Tobacco (Bikonta: Chlorobiota: Tracheophyta: Angiospermae: Solanaceae;
Nicotiana tabacum)
3.2 Animal CHITINASE CBD
Lineage: Domestic silk moth (Unikonta: Metazoa: Protostomia: Arthropoda: Hexapoda:
Bombycidae; Bombyx mori)
present in these organisms. The plant and animal chitinase proteins are
very different in structure, and the chitin binding domains (CBDs) of the
two enzymes do not share significant amino acid sequence similarity;
nevertheless, the fact that “plant and invertebrate CBDs share a common
core structure and chitin-binding mechanism while being unrelated in
evolution, argue[s] in favor of convergent evolution” of these molecules
(Shen and Jacobs-Lorena 1999, 343).
Last, convergent evolution also exists in enzymatic functions involving
RNA molecules, as well as the proteins. Within the cell, the aminoacyl-
tRNA synthetases are responsible for the synthesis of the substrates for
204 Chapter 5
ribosomal translation of messenger RNA, with 20 aminoacyl-tRNA syn-

thetases corresponding to the 20 amino acids used in protein synthesis.
Terada et al. (2002, 257) document the convergent evolution of two lysyl-
tRNA synthetases in independent archaean and bacterial lineages (table
5.12), and point out that the “identical, sophisticated reaction occurs
within two unrelated architectures” of the lysyl-tRNA synthetases-I
(LysRS-I) and lysyl-tRNA synthetases-II (LysRS-II) molecules; further-
more, the “functional convergence of the two enzymes . . . is illustrated
by the mirror images of LysRS-I and LysRS-II structures. . . . LysRS-I
and LysRS-II can be docked simultaneously on the tRNALys molecule
without any steric clashes” (Terada et al. 2002, 260).
The “hammerhead” macromolecular ribozyme structure occurs spo-
radically in both viruses and animals (table 5.13), and Salehi-Ashtiani
and Szostak (2001) argue that the hammerhead ribozyme has evolved
independently many times. Their experiments with this molecule “firmly
suggest that the hammerhead ribozyme is the simplest RNA motif
capable of catalysing RNA strand cleavage at rates between 1.0 and
Table 5.12
Convergent evolution of different enzyme molecules that serve the same RNA catalytic
function
Convergent enzyme function: LYSYL-tRNA SYNTHESIS (enzymes that catalyze the

synthesis of the substrate for tRNALys)
1 LYSYL-tRNA SYNTHETASE-I
Lineage: Thermococcalean archaeon (Archaea: Euryarchaeota: Thermococcales;
Pyrococcus horikoshii)
2 LYSYL-tRNA SYNTHETASE-II
Lineage: E. coli bacterium (Eubacteria: Proteobactia: γ-Proteobacteria; Escherichia
coli)
Table 5.13
Convergent evolution of the same ribozyme molecular structure
Convergent enzyme and function: HAMMERHEAD RIBOZYME (enzyme that catalyzes

RNA strand cleavage at rates between 0.1 and 1.0 per minute)
1 Cave cricket (Bilateria: Protostomia: Ecdysozoa: Arthropoda: Hexapoda: Orthoptera:
Rhaphidophoridae; Dolichopoda linderi)
2 Newt (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Batrachomorpha: Urodela: Salamandridae; Notophthalamus viridescens)

1.0” per minute (Salehi-Ashtiani and Szostak 2001, 82). They argue that
their “results suggest that the evolutionary process may have been chan-
nelled . . . towards repeated selection of the simplest solution to a bio-
chemical problem” (Salehi-Ashtiani and Szostak 2001, 82), and that
“purely chemical constraints (that is, the ability of only certain sequences
to carry out particular functions) can lead to the repeated evolution
of the same macromolecular structures” (Salehi-Ashtiani and Szostak
2001, 84).
In summary, we have examined examples of (1) the convergent evolu-
tion of the same enzyme function produced by the convergent evolution
of the same protein producing that function, (2) the convergent evolu-
tion of the same catalytic core structure in enzyme molecules that have
no similarity in amino acid sequence or in the three-dimensional struc-
ture of the molecules, and (3) the convergent evolution of the same
macromolecular structure in unrelated enzymes. Enzymes converge in
every possible way.
Convergent Evolution of Cellular and Tissue Structures?
The self-assembly of molecules into crystals is a natural process driven

by the laws of physics and, as we believe the laws of physics to be invari-
ant in space, the same molecules will self-assemble into the same crystal
structure throughout the universe. That is, a salt crystal (NaCl) that forms
on Mars will have the identical crystal structure as a salt crystal that
forms on the Earth. Denton and Marshall (2001) have argued that the
same might be true for protein folds: the same amino acids will self-
assemble into the same protein-fold geometries throughout the universe,
as they are driven by the laws of physics to do so. They imply that con-
vergent molecular evolution may be driven more by the laws of physics
than by natural selection; in essence, natural selection has no choice.
Selection for a given function will necessarily result in convergent molec-
ular evolution.
Denton et al. (2002) further argue that, above the molecular level of
protein-fold geometries, two other microbiological levels exhibit evi-
dence of self-assembly that is determined by the laws of physics, similar
to the nonbiological self-assembly of molecules into crystals. First are the
cellular microtubules, whose forms “are not specified directly in the
DNA, nor does any genetic program direct their assembly;” rather, they
are generated by the dynamic interactions of a few basic components of
206 Chapter 5
the cell (Denton et al. 2002, 339). Second is cell form itself which, when
experimentally altered by microsurgery, can “recover its proper form by
searching a conformational space for its correct conformation, just like
a folding protein” (Denton et al. 2002, 340). Therefore, like the molecules
composing them, the macromolecular structures of the cell itself
converge.
West et al. (1999) assert that biological convergence in form due to
the laws of physics extends upward from the cell to entire tissue systems,
such as capillary, intestinal, and lung geometries and surface areas in
animals and root, branch, and leaf geometries and surface areas in
plants. They note that biological form follows an allometric quarter-
power scaling, rather than the expected third-power from Euclidean
geometric scaling, and argue that this follows from the fractal-like archi-
tecture of all vascular branching systems in living organisms: “The vast
majority of organisms exhibit scaling exponents very close to 3/4 for
metabolic rate and 1/4 for internal times and distances. These are the
maximal and minimal values, respectively, for the effective surface
area and linear dimensions for a volume-filling fractal-like network . . .
it is testimony to the severe geometric and physical constraints on meta-
bolic processes, which have dictated that all of these organisms obey a
common set of quarter-power scaling laws” (West et al. 1999, 1679). They
continue: “[u]nlike the genetic code, which has evolved only once in the
history of life, fractal-like distribution networks that confer an additional
effective fourth dimension have originated many times” (West et al. 1999,
1679).
In summary, convergent evolution occurs across the entire spectrum
of molecules that make up life. We have studied examples of the conver-
gent evolution of identical nucleotide substitutions in nuclear and
mtDNA molecules of distantly related organisms, similar amino acid
sequences in unrelated protein molecules, similar structural geometries
in proteins with different amino acid sequences, and similar protein func-
tions by gene sharing; the convergent evolution of the same enzyme
function produced by the convergent evolution of the same protein
producing that function; the same catalytic core structure in enzyme
molecules that have no similarity in amino acid sequence or in the three-
dimensional structure of the molecules; and the convergent evolution of
the same macromolecular structure in unrelated enzymes. The effects of
functional constraint can be seen throughout molecular evolution, in that
selection for a given function repeatedly produces convergent molecular
compositions and/or molecular geometries. The number of functional
molecular evolutionary pathways available to life is not endless but is

quite restricted, and convergent evolution is the direct result.
In the next chapter we shall examine what, at first glance, might seem
to be even more improbable than the convergent evolution of molecules.
Animals have minds; we all think (some animals more than others). Can
it be that convergence extends to minds, and even to the ways in which
we think?
6 Convergent Minds
Discussions of the evolution of intelligence have focused on monkeys and apes

because of their close evolutionary relationships to humans. Other large-brained
social animals, such as corvids, also understand their physical and social worlds.
. . . Because corvids and apes share these cognitive tools, we argue that complex
cognitive abilities evolved multiple times in distantly related species with vastly
different brain structures in order to solve similar socioecological problems. . . .
[T]his convergent evolution of cognition has not been built on a convergent
evolution of brains.
—Emery and Clayton (2004, 1903)
Convergent Minds?
How can minds be said to converge? We consider our mental states to

be a function of our brain structures and sensory inputs. At first glance,
then, organisms with radically different brain structures would be
expected to have radically different minds. Take, for example, the brains
of a magpie and a human. One is a bird brain, possessed by an avian
dinosaur in the sauropsid clade of amniotes, and the other is a primate
brain, possessed by a placental mammal in the synapsid clade of amni-
otes. The sauropsid and synapsid lineages diverged back in the Carbon-
iferous, and have evolved along independent pathways ever since. Magpie
brains and human brains are structurally very different, separated by a
vast chasm of 340 million years of independent evolution. Yet it can be
demonstrated that, in many ways, magpies think like humans (Emery
and Clayton 2004; Prior et al. 2008). Our minds have converged.
Problem-Solving Behavior
It was long thought that the tool was the unique invention of the human
mind. Only the human mind had the ability to understand the limits to
210 Chapter 6
accomplishing a task using the body alone, and thus to conceive of

extending those limits by using an external object to accomplish that
task. We now know that this long-held belief is not true; other minds
have made that conceptual leap as well. Quite a few other minds—no
less than 12 independent lineages of animals use tools (table 6.1).
Within our own lineage, we can trace the manufacture of stone tools
back some 2.6 million years ago to our ancestor Homo habilis, aptly
named the “handyman” (evidence of tool usage in our lineage goes as
far back as Australopithecus afarensis, some 3.39 million years ago, but
it is not clear if A. afarensis manufactured stone tools or simply used
naturally occurring sharp stones as tools; see McPherron et al. 2010).
Table 6.1
Convergent evolution of tool-using and tool-making behavior
1 Convergent behavior and function: TOOL USAGE (using an existing external object as
a tool to attain an immediate goal)
1.1 Black-and-yellow mud wasp (Bilateria: Protostomia: Ecdysozoa: Arthropoda:
Mandibulata: Hexapoda: Hymenoptera:Vespoidea: Sphecidae; Sceliphron caementarium)
1.2 Forest ant (Arthropoda: Mandibulata: Hexapoda: Hymenoptera: Formicidae:
Myrmicinae; Aphaenogaster rudis)
1.3 Woodpecker finch (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Reptiliomorpha: Amniota: Sauropsida: Archosauromorpha: Dinosauria: Saurischia:
Theropoda: Maniraptora: Aves: Neognathae: Neoaves: Passeriformes: Emberizidae;
Cactospiza pallida)
1.4 Egyptian vulture (Aves: Neognathae: Neoaves: Falconiformes: Accipitridae;
Neophron percnopterus)
1.5 Asian elephant (Amniota: Synapsida: Therapsida: Mammalia: Eutheria: Afrotheria:
Proboscidea: Elephantidae; Elephas maximus)
1.6 Sea otter (Mammalia: Eutheria: Laurasiatheria: Carnivora: Mustellidae; Enhydra
lutris)
1.7 Polar bear (Mammalia: Eutheria: Laurasiatheria: Carnivora: Ursidae; Ursus
maritimus)
1.9 Black-striped capuchin monkey (Mammalia: Eutheria: Euarchontoglires: Primates:
Platyrrhini: Cebidae; Cebus libidinosus)
1.10 Yellow baboon (Mammalia: Eutheria: Euarchontoglires: Primates: Catarrhini:
Cercopithecoidea: Cercopithecidae; Papio cynocephalus)
2 Convergent behavior and function: TOOL CONSTRUCTION (modifying an existing
external object or objects to create a tool to attain an immediate goal)
2.1 New Caledonian crow (Amniota: Sauropsida: Archosauromorpha: Dinosauria:
Saurischia: Theropoda: Maniraptora: Aves: Neognathae: Neoaves: Passeriformes:
Corvidae; Corvus moneduloides)
2.2 Orangutan (Amniota: Synapsida:Therapsida: Mammalia: Eutheria: Euarchontoglires:
Primates: Catarrhini: Hominoidea: Hominoidae: Pongidae; Pongo pygmaeus)

Convergent Minds 211
Before the invention of these Paleolithic tools, species in the Hominini

certainly must have created tools out of wood and other organic matter,
tools that decomposed and thus are not preserved in the fossil record.
The classic studies of field anthropologists in Africa, beginning with Jane
Goodall (Wade 2006), have demonstrated to us that our close Panini
cousins, the chimpanzees (Pan troglodytes) and the bonobos (Pan panis-
cus), also create and use tools made of wood and other organic matter;
thus, tool construction is found throughout the Homininae. Since then,
gorillas (Gorilla gorilla) and now orangutans (Pongo pygmaeus) have
been observed making tools (van Schaik et al. 1996), and so tool-
construction behavior is a synapomorphy for the entire clade of the
Hominoidae, all of the great apes, not just ourselves.
Rather than being a unique product of the human mind, is tool con-
struction a unique product of the great ape mind? Not so. Both labora-
tory experimentation and field observation have demonstrated that
corvid birds, such as the New Caledonian crow, make tools (table 6.1).
When in the wild, these birds are particularly adept at making hook tools
by carefully choosing suitable twigs and then modifying them to form
hooks, and will do the same thing with metal wires in laboratory condi-
tions (Emery and Clayton 2004). We are not unique; the vastly different
brains of these avian dinosaurs can also conceive of taking an external
object and transforming it into another object for the purpose of solving
a problem or achieving a goal. Our minds have converged.
Ten other lineages of animals are known to use tools, but at present
are not known to make them (table 6.1). Tool usage is a somewhat
simpler mental concept, in which the animal looks for an existing object
to use as a tool but does not conceive of modifying that object into
another object to use as a tool. Old World monkeys (Catarrhini) and
New World monkeys (Platyrrhini) have independently evolved tool-
using behavior; capuchin monkeys pick up and carry stones with the
express purpose of using them later to hammer nuts, and baboons do the
same in hammering hard fruit, as well as using sticks for various purposes
(Baber 2003). Nonprimate minds have also converged on hammer-using
behavior: sea otters use stones to hammer mollusc shells, Egyptian vul-
tures use stones to hammer ostrich eggs, and polar bears have been
observed to drop stones on seals to stun them (Baber 2003). Even the
vastly different, and tiny, mind of the wasp has convergently evolved
hammer-using behavior: the black-and-yellow mud wasp will collect and
carry “stones” (to them; small grains of sand to us) in their mandibles,
and use these to hammer mud into burrows in which it has laid its eggs
212 Chapter 6
to protect the eggs from predators (Baber 2003). Other insects, such as
the forest ant, collect and use leaves as containers in which to carry food.
In chapter 4 we considered the case of the finch that has converged on
the ecological role of the woodpecker (see table 4.3). The woodpecker
finch searches for and carries away a tool, a cactus spine, which it uses
to extract wood-boring insects. And last, the mighty elephant has been
routinely observed to carry and use branches to scratch itself (Baber
2003), and the ocean-dwelling bottlenose dolphin collects and uses
sponges to protect its skin from abrasion while foraging on the sea
bottom (Hansell and Ruxton 2008). Animals with extremely different
brains, scattered across very distantly related phylogenetic lineages, have
independently evolved tool-using behavior (table 6.1).
Taking a different perspective, Hansell and Ruxton (2008, 77) argue
that entirely too much emphasis has been placed on tool behavior in
animals because it is always interpreted in terms of human evolution:
“Those who feel that tools are ‘special’ might be correct—but they might
be more special to researchers than to the animals that use them.” As
they point out, a great deal of attention was given to the discovery that
gorillas make tools, while at the same time researchers ignored the fact
that these animals build complicated nest structures out of modified
branches and twigs on a daily basis. Likewise, relatively few birds make
or use tools (table 6.1), but the majority of birds build nests, some of
which are structurally very complex.
Before considering the convergent evolution of architectural behavior
in animals, let us briefly review the evolution of architectural behavior
in the most diversified animal architect on the planet at present, namely
humans. Our ancestors, like our living cousins the chimpanzees and
bonobos, surely built nests in trees, habitats constructed to provide a safe
and comfortable place to sleep, high above the ground and thus out of
the reach of ground-dwelling predators. In contrast, many of our ground-
dwelling ancestors looked for preexisting natural structures, such as
caves, that could be used as habitats. Caves and other naturally occurring
habitats are rare, and with further growth in human population numbers,
humans took the next step: the construction of an artificial habitat on
the ground. Two potential construction materials are readily available:
vegetative matter, such as wood, and dirt. Humans began to build habi-
tats made of modified wood, an architectural behavior still widely in use
today, or with adobe, which many people also still use today. Only later
did we begin to build habitats with cut stone, fired bricks (which are
much more durable than adobe), and concrete (artificial stone). At
present, we have added metal, glass, and plastic to the construction mate-
rials used for our habitats.
Other animals have evolved the exact same sequence of architectural
behavior, and they did so long before we existed (table 6.2). Hansell
(2005, 33) argues that “[a]rchitecture results from the application of
behaviour to materials”; for this reason, I have omitted the behavioral
step where animal species simply hunt for preexisting structures that can
be used for habitats, like hermit crabs that use snail shells, and will move
directly to animals that actively gather and modify material to construct
a habitat. Like humans, other animal species construct habitats with
material that is either vegetative or earthen. Unlike humans, however,
animals have not evolved the capability of working with materials like
metal, glass, or plastic in their construction, although some do steal and
inventively use our worked materials in their habitats!
Hansell (2005) identifies five architectural behaviors found in nest-
building birds, listed here in terms of increasing behavioral complexity:
stacking, entangling, Velcro-fastening, stitching, and weaving. At present
there exists no comprehensive phylogenetic analysis of how many inde-
pendent times each of these behaviors have evolved in separate groups
of birds. Thus, I will discuss as many convergent examples that I am aware
of, knowing that future phylogenetic analyses will almost certainly
expand that list.
The eagle (table 6.2) is an example of a bird that constructs its nest by
first searching for and collecting branches of the desired length and
diameter, and then transporting those branches back to the tree it has
chosen for the nest. The eagle simply stacks the branches atop one
another until a bowl-shaped nest has been created. This same nest-
building architectural behavior is used by some ground-nesting birds as
well, such as the swan (table 6.2). In general, many of the large tree-
dwelling or ground-dwelling birds use stacking to build their nests: storks
(Cicioniidae), pelicans (Pelicanidae), and herons (Ardeidae), in addition
to eagles and swans (Hansell 2005).
Nests built by entangling architectural behavior are sturdier than those
built by simple stacking, as the ends of the branches or twigs in these
nests are actively bent and pulled back into the core of the nest by the
bird. The familiar wood pigeon (table 6.2) is an example of a bird that
constructs its nest using entangling (Hansell 2005). The same method is
used by the gorilla (table 6.2) and chimpanzee (and presumably many
of our ancestors), who search out a spot in the trees with an upright fork
and then bend branches back into this foundation while holding them in
214 Chapter 6
Table 6.2
Convergent evolution of architectural behaviors
1 Convergent behavior and function: CONSTRUCTION USING VEGETATIVE

MATERIAL (collecting and modifying vegetative material to build a wicker-like bowl,
globe, baglike, or tentlike construction for use as a protective shelter)
1.1 Veined octopus (Bilateria: Protostomia: Lophotrochozoa: Mollusca: Cephalopoda:
Coleoidea: Octopodidae; Amphioctopus marginatus)
1.2 Arctic yellow jacket (Ecdysozoa: Arthropoda: Mandibulata: Hexapoda:
Hymenoptera: Vespoidea: Vespidae: Polistinae; Dolichovespula norwegica)
1.3 Neotropical paper wasp (Arthropoda: Mandibulata: Hexapoda: Hymenoptera:
Vespoidea: Sphecidae; Microstigmus comes)
1.4 Mute swan (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Theropoda: Maniraptora: Aves: Neognathae: Galloanserae: Anseriformes: Anatidae;
Cygnus olor)
1.5 Bald eagle (Aves: Neognathae: Neoaves: Falconiformes: Accipitridae; Haliaeetus
leucocephalus)
1.6 Wood pigeon (Aves: Neognathae: Neoaves: Columbiformes: Columbidae; Columba
palumbus)
1.7 European penduline tit (Neognathae: Neoaves: Passeriformes: Remizidae; Remiz
pendulinus)
1.8 Long-tailed tit (Aves: Neognathae: Neoaves: Passeriformes: Aegithalidae; Aegithalos
caudatus)
1.9 Gray-backed camaroptera (Aves: Neognathae: Neoaves: Passeriformes: Cisticolidae;
Camaroptera brevicondata)
1.10 Village weaverbird (Aves: Neognathae: Neoaves: Passeriformes: Ploceidae:
Ploceinae; Ploceus cucullatus)
1.11 Montezuma oropendola (Aves: Neognathae: Neoaves: Passeriformes: Fringillidae:
Icterinae; Gymnostinops montezuma)
1.12 Great bowerbird (Aves: Neognathae: Neoaves: Passeriformes: Ptilonorhynchidae;
Chlamydera nuchalis)
1.13 Dusky-footed wood rat (Amniota: Synapsida: Therapsida:Mammalia: Eutheria:
Euarchontoglires: Rodentia: Myomorpha: Cricetidae; Neotoma fuscipes)
1.14 European harvest mouse (Mammalia: Eutheria: Euarchontoglires: Rodentia:
Myomorpha: Muridae; Micromys minutus)
1.15 Gray squirrel (Mammalia: Eutheria: Euarchontoglires: Rodentia: Sciuromorpha:
Sciuridae; Sciurus carolinensis)
1.16 Peter’s tent-making bat (Mammalia: Eutheria: Laurasiatheria: Chiroptera:
Microchiroptera: Phyllostomidae; Uroderma bilobatum)
1.17 Short-nosed fruit bat (Mammalia: Eutheria: Laurasiatheria: Chiroptera:
Megachiroptera: Pteropodidae; Cynopterus sphinx)
1.18 Gorilla (Mammalia: Eutheria: Euarchontoglires: Primates: Catarrhini: Hominoidea:
Hominidae: Gorillinae; Gorilla gorilla)
2 Convergent behavior and function: CONSTRUCTION USING ANIMAL-
GENERATED MATERIAL (collection and modification of animal-generated material
to build a woven bowl, globe, baglike, or tentlike construction for use as a protective
shelter)
2.1 Common tailorbird (Amniota: Sauropsida: Archosauromorpha: Dinosauria:
Saurischia: Theropoda: Maniraptora: Aves: Neognathae: Neoaves: Passeriformes:
Cisticolidae; Orthotomus sutorius)
Table 6.2
(continued)
2.2 Chaffinch (Aves: Neognathae: Neoaves: Passeriformes: Fringillidae; Fringilla

coelebs)
2.3 Modern human (Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
Euarchontoglires: Primates: Catarrhini: Hominoidea: Hominidae: Homininae; Homo
sapiens)
3 Convergent behavior and function: CONSTRUCTION USING EARTHEN
MATERIAL (collecting and modifying earthen material to build an adobe-like dome,
cylindrical, pyramidal, or globelike construction for use as a protective shelter)
3.1 Neotropical mud wasp (Bilateria: Protostomia: Ecdysozoa: Arthropoda:
Mandibulata: Hexapoda: Hymenoptera: Vespoidea: Sphecidae; Trigonopsis cameronii)
3.2 Malasian hover wasp (Arthropoda: Mandibulata: Hexapoda: Hymenoptera:
Vespoidea: Vespidae: Stenogastrinae; Liostenogaster flavolineata)
3.3 Ceylonese potter wasp (Arthropoda: Mandibulata: Hexapoda: Hymenoptera:
Vespoidea: Vespidae: Eumeninae; Paraleptomenes mephitis)
3.4 Warlike termite (Arthropoda: Mandibulata: Hexapoda: Isoptera: Termitidae:
Macrotermitinae; Macrotermes bellicosus)
3.5 Cliff swallow (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Theropoda: Maniraptora: Aves: Neognathae: Neoaves: Passeriformes: Hirundinidae;
Hirundo pyrrhonota)
3.6 White-necked rockfowl (Aves: Neognathae: Neoaves: Passeriformes: Picathartidae;
Picathartes gymnocephalus)
3.7 Rufous hornero (Aves: Neognathae: Neoaves: Passeriformes: Furnariidae; Furnarius
rufus)
3.8 North American beaver (Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
Euarchontoglires: Rodentia: Castorimorpha: Castoridae; Castor canadensis)
3.9 Modern human (Mammalia: Eutheria: Euarchontoglires: Primates: Catarrhini:
Hominoidea: Hominidae: Homininae; Homo sapiens)
place with their feet; finally, they tuck leafy twigs growing around the rim
of the nest back into the core of the nest (von Frisch 1974). Unlike birds,
however, apes construct their sleeping nests on a daily basis.
Many plants have convergently evolved seeds with the tiny barbs and
hooks that provided the inspiration for the modern fastening product
Velcro (see table 3.15). Some birds use this same fastening principle in
building their nests: they actively seek out materials with tiny projecting
processes and position these in the nest so that they ensnare each other,
thereby creating even sturdier nests than those built using entangling.
The long-tailed tit (table 6.2) uses small-leaved mosses for Velcro, and
the related bushtit (Psaltiparus minimus) uses lichen (Hansell 2005).
These nests are so sturdy that they can be totally enclosed into a baglike
or socklike nest, with an opening at one end, as seen in the nests built
by European penduline tits (table 6.2; von Frisch 1974). Rather than
216 Chapter 6
using Velcro-fastening, other birds actually stitch their nests together,

pushing thin plant fiber through leaf margins, grasping it on the other
side, and pushing it through again until separate leaves are stitched
together. The gray-backed camaroptera (table 6.2) is an example of a
bird that uses stitching behavior to build its nests (Hansell 2005).
Last, the most complex architectural behavior seen in birds is actual
weaving, a suite of multiple behaviors including creating “loops, half-
hitches, hitches, bindings, slip knots, and overhand knots, as well a more
regular over-and-under pattern of weft through warp that defines human
weaving” (Hansell 2005, 73–74), as can be seen in the nests built by the
village weaverbird (table 6.2). Unlike most bird architectural behavior,
which appears to be totally instinctual, weaving involves a learning
component: young weaverbirds must practice and observe older weav-
erbirds in order to perfect the weaving technique (Hansell 2005). Yet this
most complex architectural behavior in birds is not unique: “Weaving
has arisen independently in two taxa of birds, the Old World weaver
birds (Ploceinae) and the New World oropendolas, orioles, and caciques
(Icterini, Fringillidae)” (table 6.2; Hansell 2005, 74).
In addition to using vegetative material, many birds also use animal-
generated materials such as silk and mammal hair to construct nests.
Hansell (2005) notes that 56 percent of passeriform families of birds
actively search for arthropod silk, particularly spider silk, to use as strong
threads in their nest building; this is seen in the stitching-constructed
nests of the common tailorbird (table 6.2). Other birds, such as the chaf-
finch (table 6.2), use mammal hair. Gould and Gould (2007, 194) docu-
ment the case of a Carolina wren (Thryothorus ludovicianus) that
constructed its nest using human-made fishing line, rather than spider
silk, to build its nest—incorporating the fishing hooks and lead sinkers
into the nest! Humans have converged on these avian habitat architec-
tures in the building of tents constructed of woven wool and other animal
hairs.
The birds are avian dinosaurs that evolved in the Jurassic, some 100
million years before the destruction of the dinosaurian ecosystem and
its replacement by the mammalian, as discussed in chapter 4. In the
Cenozoic Era, mammalian architectural behavior has converged on that
of the more ancient avian, as seen in the entangling-constructed nests
built by modern apes. The European harvest mouse (table 6.2) builds a
globular-shaped enclosed nest, with an entrance hole on one side, sus-
pended in the branches of shrubs or the stems of oats (von Frisch 1974).
Gray squirrels (table 6.2) build similar, but larger enclosed globular nests
in trees; these myomorph and sciuromoph rodents have converged on

the architecture of the weaverbird.
The tent-making bats have evolved an architectural behavior not seen
in birds. These bats seek out suitably placed large leaves in trees and then
bite into the veins of the leaves, creating a line of perforation along which
the leaf folds down on either side under the pull of gravity (Hansell
2005). The bats use different bite techniques for different leaf types in
order to create the three-dimensional tent structure. Odd as this archi-
tectural behavior is, it is not unique. Tent-making behavior has indepen-
dently evolved in both Neotropical microbats, like Peter’s tent-making
bat in Costa Rica, and in Paleotropical megabats, like the short-nosed
fruit bat in India (table 6.2; Kunz et al. 1994).
The wasps have also evolved an architectural behavior not seen in
birds. They use vegetative material in their nest constructions, but it is a
material that they manufacture themselves—paper. These wasps chew
plant material to produce pulped and blended fragments that adhere
together, and that are further held together by the wasp’s dried spit, to
form sheets of paper that are both light in weight but also strong in
tension (Hansell 2005). Complex paper-constructed nests are built by
both social wasps, such as the vespid Arctic yellow jacket, and solitary
wasps, such as the sphecid Neotropical paper wasp (table 6.2). Humans
have converged on these wasp architectures in the constructing of rice-
paper shoji that are used for lightweight walls in many Japanese
buildings.
Like humans, animals also use vegetative material to build habitats on
the ground level, but these architectures are not as common as those that
are constructed with earthen material. The dusky-footed wood rat (table
6.2) builds lodges of branches and twigs that can be up to one meter high,
and that contain as many as five separate chambers or rooms that are
used for different purposes, such as a nursery, a living room, and a lava-
tory (von Frisch 1974). The bowerbirds of Australia and New Guinea
(table 6.2) are famous for their elaborate ground constructions that,
unlike all the other architectural examples discussed here, are not used
as functional protective habitats but as artistic constructions designed to
attract mates. These display stages—and stage it is—may be furnished
with “curtains” made of leaves or vines, decorated with “floodlight”
borders of flowers, colored stones, berries, or snail shells. They have
enormous (for the size of the bird) arched stage roofs of branches that
are sometimes painted with berry juice, and are fronted by “theatre aisle”
walkways leading to the stage. If a female appears, the male gives an
218 Chapter 6
elaborate performance of dancing and singing on the stage. If the female

is unimpressed, she walks away; if she is attracted, she joins the male in
the dance.
Have the bowerbirds convergently invented art? Gould and Gould
(2007, 221–222) argue that bowerbird behavior “may be very close to
crossing some important line between the mental experiences of humans
and other animals. Bowerbirds can indulge in this frivolity, if that is what
it is, because there are few predators and little competition in their
habitat much of the year, and many species have a very long or continu-
ous breeding season—factors that have allowed artistic creativity to
flourish among humans as well. . . . At the cognitive level, these creations
are the most complex seen in birds. Only beavers and humans undertake
work with more steps and greater flexibility in design, materials, and
execution.”
Perhaps even more striking, however, is the recent discovery that the
veined octopus (table 6.2), a marine mollusc species, has innovatively
seized human-made vegetative objects to build habitats: coconut half-
shells. In the last several decades, an abundance of this new building
material appeared in the octopus’s shallow marine realm, as humans eat
the coconuts and then discard the clean and lightweight shell halves into
the waters surrounding their coastal communities in Indonesia. Over 20
individual veined octopi have been observed collecting, transporting, and
assembling the coconut shell halves into protective habitats to hide in.
When flushed from the shells by human scuba divers, the octopi quickly
return and reassemble the shell-constructed habitats, leaving Finn et al.
(2009, R1070) to conclude that “even marine invertebrates engage in
behaviours that we once thought the preserve of humans.”
Returning to the terrestrial realm, and on the ground level, many
humans lived in caves first (in essence, holes in the ground), and then
began to build adobe habitats (in essence, artificial caves). Many birds
have followed this same behavioral pathway in their evolution, with the
exception that their adobe habitats are built on vertical cliff faces rather
than on horizontal ground surfaces. Molecular phylogenies of swallows
and martins (Hirundinidae; table 6.2) reveal the multiple independent
evolution of the behavioral pathway of first dwelling in preexisting cliff
cavities, then constructing open mud cups attached to cliff faces as nests,
and finally constructing enclosed adobe globes attached to cliff faces as
nests (Gould and Gould 2007). In contrast to the small adobe nests of
the cliff swallow, the picathartid white-necked rockfowl (table 6.2) builds
a two-kilogram adobe nest, even though the bird itself weighs only 200
grams (Hansell 2005). Gould and Gould (2007, 184) comment: “The
invention of adobe cups that then develop into enclosed cavities, so
evident in the progression from barn to cliff swallows, is also seen in
some birds that build in trees.” The most striking example of a tree-
dwelling adobe builder is the rufous hornero of South America (table
6.2), nicknamed the “ovenbird” because its vertical adobe nest looks like
a baker’s oven (von Frisch 1974). The globe-shaped nest contains a large
interior room that is used as a brooding chamber and a narrow, hall-like
entrance passage that opens into a ten-centimeter-diameter hole to the
outside. Both the male and female birds cooperate in its construction,
which takes them over two weeks to complete (von Frisch 1974).
Wasps have evolved a similar behavioral pathway in their habitat
constructions, as noted by Hansell (2005, 232): “I have identified two
alterations in nest building behaviour that are associated with the transi-
tion from solitary to social behaviour of wasps. . . . The first is the transi-
tion from an excavated to a constructed nest. . . . This occurred in both
the Sphecidae and Vespidae (Eumeninae). The second change was in the
choice of the building material from mud to paper.” Both paper-building
behavior and adobe-building behavior is convergent in wasps (table 6.2).
An interesting transition between excavating tunnels for below-ground
habitats to constructing above-ground adobe habitats is seen in the
eumenin funnel wasp (Paralastor emarginatus) of Australia. This wasp
first digs a tunnel, lines it with mud, and then extends these mud walls
above ground to build an external cylinder, which it increases in diameter
to build a downward-pointing funnel opening (Gould and Gould 2007).
The funnel serves the purpose of concealing the entrance to the tunnel
from parasites while the female is hunting for caterpillars to paralyze
and place in the tunnel as food for her later-developing larvae. Solitary
sphecid wasps, such as the Neotropical mud wasp (table 6.2), do no
excavating but build adobe brood chambers entirely above ground.
While adobe building is the norm for many sphecid wasps, the eusocial
Malasian hover wasp (Stenogastrinae) also builds with adobe, as does
the Ceylonese potter wasp (Eumeninae) within a subfamily of wasps that
normally builds with paper (table 6.2; Hansell 2005).
Termites also have converged on the extension of their below-ground
excavations to above-ground adobe-building behavior (table 6.2). Com-
pared to the size of their inhabitants, termite adobe dwellings are huge:
the Australian compass termite, Amitermes meridionalis, builds above-
ground nests that are up to 5 meters high and 3 meters long, but only
around 30 centimeters thick. The species is given the name “compass”
220 Chapter 6
because its slablike buildings are always oriented with the long direction
north-south, such that the habitat always has a broad side warmed by
the sun (either to the east or the west, depending upon the time of day),
and an opposite side that is cool and shaded (von Frisch 1974). The
temperature difference between the two broad sides of the habitat pro-
duces air flow from the cool side of the habitat through to the warm
side—the termites have invented buildings with air conditioning! And
not just one type of air-conditioning system—the much-studied African
warlike termite (table 6.2) builds five-meter-high pyramidal-shaped
dwellings with hard, cement-like adobe outer walls and an interior adobe
nest that is suspended within the outer walls by a series of pillars, arches,
and buttresses. The inner dwelling is surrounded by an air space, through
which warm air rises from the bottom of the building, creating a lower-
pressure region that then draws cooler, denser air into the interior of the
building from the outside. The source of heat at the bottom of the dwell-
ing is the fermentation in fungus gardens arranged in special adobe
chambers—the termites have thus also invented agriculture (a point that
we shall consider in more detail below). A large habitat of the warlike
termite can contain as many as two million termites. The various air-
conditioning systems that termites build also serve the purpose of flush-
ing the carbon-dioxide-rich air created by their breathing out the top of
the building; oxygen-enriched air is drawn in through the ventilation
holes in the side of the building (von Frisch 1974).
Australian compass termites and African warlike termites inhabit gen-
erally arid regions; in wet tropical regions, termites change the architec-
ture of their dwellings to deal with the problem of frequent rainfall. The
termite species Cubitermes fungifaber builds tall, cylindrical-shaped
adobe dwellings with multiple roofs that are slanted downward and
extend out beyond the diameter of the main habitat (Gould and Gould
2007). These multiroofed towers are reminiscent of the pagodas build by
humans in Asia.
Last, beavers have evolved diversified architectural behaviors that
converge on those of humans: they build with logs, adobe, and stone. The
beavers are the largest rodents alive today, and may weigh up to 30
kilograms (von Frisch 1974). The North American beaver habitat, or
lodge, is constructed of both vegetative and earthen material; it is archi-
tecturally very similar to the log cabins constructed by early European
settlers in North America (table 6.2). The beavers first build a domelike
habitat with cut tree branches and logs, and then carefully plaster and
caulk the openings between the branches in the walls of the lodge with
mud and clay. Unlike a log cabin, however, the door to the lodge is under
water, and the lodge itself is located on an artificially created island in a
pond or lake. Both of these features serve to protect the beavers’ dwell-
ing from invasion by predators. These round dome-shaped dwellings,
isolated from the shore out in a lake, are reminiscent of the crannogs
built by prehistoric humans in Scotland.
The pond or lake that surrounds the beavers’ lodge is often the cre-
ation of the beavers themselves: “beavers are experts not only in the
building of dwellings but also in hydro-engineering, and have performed
tremendous feats in this line long before man attempted anything of the
kind” (von Frisch 1974, 268). Blocking a flowing stream or river with a
dam is not an easy task, yet beavers do so by first ramming cut branches
and logs into the river bed, supporting these against the flow of the river
with forked branches pointing upstream, and adding further anchoring
by bringing heavy stones to hold the dam superstructure in place. They
further excavate the river bed on the upstream side of the dam, which
not only reduces the speed of water flow but also provides the beavers
with the earthen material they need to finish sealing the dam. The fin-
ished dam is higher in the center than at the termini, thus forcing the
dammed river to flow over the dam near the two river banks. The beavers
then actively control the level of the lake by either lowering or raising
these two overflow sluices—they frequently lower the level of the lake
during winter to created a breathing space under the ice of the frozen
surface of the lake, an air space that extends the entire area of the lake,
providing them with a swimming region protected from the winter cold.
Von Frisch (1974) reports that the largest beaver dam on record was 700
meters long, and was strong enough to support the weight of human
riders on horseback.
Do convergent architectures reveal convergent minds? Gould and
Gould (2007, 271–272) argue that “[m]ental activity is, by its nature,
private; what goes on in the brain has to be inferred. In tracing the evolu-
tion of cognitive strategies, the most tangible evidence is found among
animals that build. . . . These abilities seem to have evolved independently
in several different groups, but always apparently in about the same
order, and to serve analogous ends.” That is, given the same problem—
such as building a protective habitat—animals with radically different
brains have evolved the same architectural behaviors. Those analogous
behaviors thus reflect analogous mental activities and cognitive strate-
gies taking place in independent lineages, which is convergent mental
evolution.
222 Chapter 6
On the other hand, do humans really think like termites when they
build an adobe habitat? Gould and Gould (2007, 5) also note that
“[w]e know that animal building, like most tasks animals accomplish,
depends on many different neural mechanisms. For most creatures,
though, instinct rather than learning seems to guide behavior.” This con-
clusion was revealed in the pioneering work with animal instinctive
behavior by Konrad Lorenz, Niko Tinbergen, and Karl von Frisch, work
for which they were awarded the 1973 Nobel prize in physiology or
medicine. Behavioral evolution by the trial-and-error process of natural
selection requires very long periods of time and generations of reproduc-
tive cycles. Behavioral evolution by learning, in contrast, is very rapid:
one generation simply teaches the next generation the new behavioral
innovations discovered by the parent generation. Cognitive strategies are
transmitted by teachers, not by genes. Yet stereotypic instinctive behav-
iors and flexible learned behaviors have both converged on the same
result in architecture (table 6.2).
The discipline of theoretical morphology (McGhee 2007) takes a dif-
ferent perspective on the question of convergent architectural behavior
by examining the geometries of the architecture itself: Are there only a
limited number of ways to build structures? If so, convergent behavior
must necessarily result because independent lineages of organisms have
a limited number of building options to discover. The convergent usage
of the cantilever and the arch by such disparate organisms as mud wasps,
termites, cliff swallows, and humans results from the fact that cantilevers
and arches are necessary in the construction of adobe architectures.
Convergent minds are here the product of the functional constraints of
geometry. The implications of this theoretical morphologic perspective
on convergent evolution will be explored in more detail in chapters 7
and 8, but I would point out here that we should not be surprised to find
cantilevers and arches on Earth-like worlds elsewhere in the universe,
produced by totally alien minds.
To conclude this section of the chapter, let us consider one last
extremely complex problem-solving behavior: agriculture. Modern
humans, Homo sapiens, are the only vertebrate species that has evolved
agriculture. Agricultural behavior is convergent, however, in that several
populations of humans made the transition from hunter-gatherers to
agriculturalists independently of each other around the world about
10,000 years ago: in the Tigris and Euphrates valleys in the Near East, in
South Asia, in Central Asia, and in Central America (Gupta 2004).
Was agriculture clearly an idea whose time had come, beginning about
10,000 years ago, as a unique behavioral trait of the human species? Not
so. In contrast to the single human species, the hexapod arthropods have
independently evolved agriculture nine separate times in different lin-
eages (table 6.3). The transition from hunter-gatherer to agriculturalist
was made once by the ants, once by the termites, and no less than seven
separate times by the ambrosia beetles (Farrell et al. 2001; Mueller et al.
2005). And they did so long before humans existed: the platypodine
ambrosia beetles evolved agricultural behavior 60 million years ago
Table 6.3
Convergent evolution of agricultural behaviors
1 Convergent behavior and function: CROP AGRICULTURE (cultivation of crops for

nourishment)
1.1 Attine ant (Bilateria: Protostomia: Ecdysozoa: Arthropoda: Mandibulata: Hexapoda:
Hymenoptera: Formicidae: Myrmicinae: Attini: Paleoattini; Apterostigma auriculatum)
1.2 Xyleborine ambrosia beetle (Arthropoda: Mandibulata: Hexapoda: Coleoptera:
Curculionidae: Xyleborini; Dryocoetoides cristatus)
1.3 Hyorrhynchine ambrosia beetle (Arthropoda: Mandibulata: Hexapoda: Coleoptera:
Curculionidae: Hyorrhynchini; Sueus niisimai)
1.4 Platypodine ambrosia beetle (Arthropoda: Mandibulata: Hexapoda: Coleoptera:
Curculionidae: Platypodini; Australoplatypus incompertus)
1.5 Bothrosternine ambrosia beetle (Arthropoda: Mandibulata: Hexapoda: Coleoptera:
Curculionidae: Bothrosternini; Cseninus lecontei)
1.6 Xyloterine ambrosia beetle (Arthropoda: Mandibulata: Hexapoda: Coleoptera:
Curculionidae: Xyloterini; Indocryphalus pubipennis)
1.7 Scolytoplatypodine ambrosia beetle (Arthropoda: Mandibulata: Hexapoda:
Coleoptera: Curculionidae: Scolytoplatypodini; Scolytoplatypus macgregori)
1.8 Corthyline ambrosia beetle (Arthropoda: Mandibulata: Hexapoda: Coleoptera:
Curculionidae: Corthylini; Gnathotrupes quadrituberculatus)
1.9 Black-winged subterranean termite (Arthropoda: Mandibulata: Hexapoda:
Isoptera: Termitidae: Macrotermitinae; Odontotermes formosanus)
1.10 Modern human (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Reptiliomorpha: Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
sapiens)
2 Convergent behavior and function: ANIMAL HUSBANDRY (cultivation of animals for
nourishment)
2.1 Japanese ant (Bilateria: Protostomia: Ecdysozoa: Arthropoda: Mandibulata:
Hexapoda: Hymenoptera: Formicidae: Formicinae; Formica yessensis)
2.2 Modern human (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
sapiens)

224 Chapter 6
(Farrell et al. 2001), the ants did so 50 million years ago (Schultz and
Brady 2008), and the termites evolved agricultural behavior between 34
to 24 million years ago (Mueller et al. 2005).
Agricultural behavior is incredibly complex. It includes (1) preparing
the substrate to be used for growing the crops, (2) planting the crops, (3)
monitoring the growth and potential disease status of the crops, (4) pro-
tecting the crops from disease, (5) protecting the crops from crop-eating
species other than the farmers, (6) weeding invasive species out of the
crops, (7) using chemical herbicides for weed control, (8) using microbes
for biological pest control, (9) using microbial symbionts to procure
nutrients for the crops, and (10) the sustainable harvesting of the crops
for food (Mueller et al. 2005). This list is not a list of agricultural behav-
iors found in humans—it is a list of behaviors found in ants! In fact,
Mueller et al. (2005, 564) argue that “[b]ecause of the universality of crop
diseases in both human and insect agriculture, it may be fruitful to
examine the . . . solutions that have evolved convergently in insect agri-
culture for possible application to human agriculture.” Here we should
note that ant agriculturalists use microbes for biological pest control,
whereas most human agriculturalists (thus far) use chemical pesticides.
There is one major difference in human and hexapod agriculture:
hexapods cultivate fungus crops, whereas humans usually cultivate plant
crops (although we do also farm mushrooms). The most advanced
hexapod agriculturalists have in fact evolved a new way to digest cellu-
lose: rather than digesting it internally with the usage of anaerobic bac-
teria and specialized stomachs like many vertebrate animals (itself a
convergent trait; see table 2.10), advanced hexapods use fungus gardens.
Indeed, Schultz and Brady (2008, 5435) argue that a colony of leaf-cutter
ants is “the ecological equivalent of a large mammalian herbivore in
terms of collective biomass, lifespan, and quantity of plant material
consumed.”
Fungi are the only major group of organisms that are able to digest
cellulose in the presence of free oxygen (Gould and Gould 2007). They
also do not need light, as they are not photoautotrophs like plants, but
they do need heat and humidity for efficient digestion. Both ants and
termites have independently evolved the behaviors needed to create
warm, moist agricultural chambers for fungus gardens, the ants under-
ground and the termites in their above-ground adobe buildings. The ants
are particularly important in analyzing the evolution of agricultural
behavior, in that living ants still use five distinctly different agricultural
systems that can be demonstrated to have evolved in a sequence of
increasing agricultural sophistication (Schultz and Brady 2008). “Lower

agriculture,” in which a wide range of fungal species are cultivated, first
appeared in paleoattine ants some 50 million years ago (table 6.3). About
30 million years later, there was a bifurcation in agricultural behavior.
“Yeast agriculture,” in which neoattine ants specialize in cultivating leu-
cocoprineaceous fungi in the single-celled yeast phase, evolved 20 million
years ago, and “coral fungus agriculture,” in which paleoattine ants spe-
cialize in cultivating nonlecocoprineaceous fungi in the clade of Pteru-
laceae, evolved 15 million years ago. Also about 20 million years ago a
separate clade of neoattine ants evolved “domesticated agriculture,” in
which the fungal species have coevolved so much with their ant cultiva-
tors that they are no longer capable of existing on their own outside of
the ant gardens. These domesticated fungi also produce nutritious swollen
hyphal tips, or gongylidia, that the ants harvest for food (Schultz and
Brady 2008). Last, some 12 million years ago the leaf-cutting neoattine
ant farmers evolved. The leaf-cutter ants specialize in cutting and pro-
cessing fresh vegetation for substrates on which to grow their fungus
crops, unlike all the other ants, who use dead organic detritus. These ants,
who in effect “eat” living plants, are the dominant “herbivores” in the
New World tropics in terms of the quantity of the plant material that
they remove from the ecosystem (Schultz and Brady 2008).
Unlike the versatile ant farmers, termite farmers (table 6.3) cultivate
species of a single genus of fungus, Termitomyces, which they have
domesticated (Mueller et al. 2005). Termite domestic fungi produce veg-
etative nodules that are harvested by the termites for food, similar to the
domestic fungi of the ants. Like many ant farmers, termites grow their
fungus crops in specialized agricultural chambers on a prepared substra-
tum of dead plant material, and have not (yet) evolved the ability to
process living plants as the leaf-cutter ants have.
Unlike ants and termites, the ambrosia beetles do not build specialized
agricultural chambers for growing fungus gardens. They bore into tree
trunks and excavate extensive galleries of tunnels within the living
phloem layer of the tree, and then grow their fungal crops along the walls
of their galleries. Besides providing food for the beetles, the fungi also
block the chemical pesticides produced by the tree to defend itself from
infection—thus, invasion by these beetles can be devastating to forests
(Farrell et al. 2001).
Also unlike the ants and termites, agricultural behaviors have inde-
pendently evolved in seven separate lineages of beetles (table 6.3). Dif-
ferent beetle lineages farm different species of the ophiostomatoid fungi,
226 Chapter 6
some species of which they have domesticated (Farrell et al. 2001). Dif-
ferent beetle lineages also use different tree hosts, in that some infest
species of conifers, and others attack species of angiosperms.
Animal husbandry in agriculture, where animals are domesticated
instead of crops, is also convergent, but rare (table 6.3). Humans
domesticated sheep, goats, cattle, and pigs for food some 7,000 years ago
(Gupta 2004). In the hexapods, the formicine and dolichoderine ants in
particular are known for their domestication of aphids (Aphididae), soft
scale insects (Coccidae), tree hoppers (Membracidae), and mealy bugs
(Pseudococcidae; Delabie 2001; Stadler and Dixon 2005).
Aphids are herbivores that feed on the phloem sap of plants, which is
rich in sugar but poor in nitrogen. To obtain enough nitrogen, aphids
have to consume large amounts of sap, and they excrete the excess sugar
as honeydew. This honeydew waste product is a food source for ants, who
collect it from the aphids. Thus begins the evolution of the trophobiotic
relationship between the aphid herbivore and the ant farmer. When
tended by the ant farmers, aphids produce much more honeydew, and
higher-quality honeydew, than they do when solitary. To offset this meta-
bolic cost, the aphids benefit by being protected from predators by the
ant farmers. The ants are so good at protecting their aphid herbivores
that farmed aphids can become very numerous and gregarious. The cost
of protecting the aphids is therefore offset by the production of aphid
“herds,” which provide a food source for the ants that is concentrated in
a small farmed area—rather than having a diffuse source of food scat-
tered over a wide geographic region. Some aphids have become so
domesticated that they depend upon the ant farmer for existence; for
example, the aphid Stomaphis quercus is cultivated by the ant Lasius
fuliginosus, and is only found on oak trees farmed by this ant (Stadler
and Dixon 2005).
Ants are known to construct protective shelters out of plant debris for
some of their farmed insects, both to shelter the insects from the weather
and to conceal them from predators. The ant Formica obscuripes goes
further and shelters its herd of aphids in galleries within its own under-
ground dwelling, leading the herd of aphids out to “pasture” during the
day, and herding them back underground for the night (Delabie 2001).
Other ants will pick up and transport their farmed insects, particularly
soft scale insects, to more favorable plant “pastures” for them to feed.
When an ant farmer feeds, it will either drum the aphid’s abdomen with
its antennae or stroke the aphid’s sides to encourage it to excrete hon-
eydew droplets, which the ant collects (Delabie 2001).
At first glance, it would appear that ants have converged on the animal
husbandry behavior of the human dairy farmer (table 6.3). The dairy
farmer benefits from a herd of cows, which are a concentrated source of
milk for food. The cows benefit from being protected from wolves and
other predators by the farmer, and farmed cows produce much more
milk than wild ones. The farmer leads the herd of cows out of a barn
shelter to the open pasture on a daily basis, and later strokes the mammary
glands of the cows to encourage them to excrete milk, which the farmer
collects. The problem with this analogy is the fact that ants evolved aphid
farming in the Early Oligocene, some 30 million years before the exis-
tence of humans (Stadler and Dixon 2005). It is the human dairy farmer
that has converged on the animal husbandry behavior of ants, not the
other way around.
In summary, it took ant agriculturalists 30,000,000 years to go from
farming fungus to domesticating fungus: from the Early Eocene paleoat-
tine ants to the Early Miocene neoattine ants (Schultz and Brady 2008).
In contrast, human farmers have accomplished the same feat with plants
in less than 10,000 years. Yet, as we also saw with the evolution of archi-
tectural behaviors, the evolution of stereotypic instinctive behaviors in
ants and flexible learned behaviors in humans have both converged on
the same result: farming of domesticated crops (table 6.3). Ants and
humans also convergently evolved animal husbandry, but in this case the
fossil record does not reveal how long it took the ants to accomplish the
feat of domesticating aphids.
The complex of agricultural behaviors independently evolved by
hexapod arthropods and vertebrate humans is astonishingly convergent.
However, as with architecture, using the analytical technique of theoreti-
cal morphology to examine the question of convergent agricultural
behavior provides a different perspective by focusing on the cultivation
process instead: Is there only a limited number of ways to successfully
cultivate crops? If so, convergent behavior must necessarily result, for
independent lineages of organisms have a limited number of farming
options to discover. The convergent usage of chemical herbicides by such
disparate organisms as ants and humans results from the fact that such
pesticides are necessary for the cultivation of crops. As with architecture,
convergent agricultural minds are here the product of functional con-
straint on evolution.
It is difficult for a single individual to run a really efficient farm.
Farming is labor intensive, and is usually the sum product of many indi-
viduals who have taken on different tasks. Task partitioning and the
228 Chapter 6
division of labor in farming have evolved convergently in both humans

and hexapods, but some hexapods have taken it to an extreme not seen
in humans—they have become eusocial. Eusociality is an ultimate form
of group behavior and, as we shall see in the next section of the chapter,
it is also convergent.
Group Behavior
In the first edition of On the Origin of Species, Darwin considered

convergent minds to present special difficulties for his theory of
natural selection. He was troubled by observed cases of identical behav-
ior, or “instincts,” in animals within independent evolutionary lineages,
and the very worst of these cases was that of the eusocial hexapods: “No
doubt many instincts of very difficult explanation could be opposed to
the theory of natural selection . . . instincts almost identically the same
in animals so remote in the scale of nature, that we cannot account for
their similarity by inheritance from a common parent, and must there-
fore believe that they have been acquired by independent acts of
natural selection. . . . I allude to the neuters or sterile females in insect-
communities” (Darwin 1859, 235–236). The very heart of the theory of
natural selection is the effect of differential reproductive success of indi-
viduals with different phenotypes; thus, how can natural selection produce
social behaviors in which the majority of individuals do not reproduce
at all? Yet eusocial cooperative societies, in which the majority of indi-
viduals in those societies have sacrificed their own reproduction, have
convergently evolved at least 17 independent times (table 6.4).
In honeybee societies, for example, there is only one fertile female, the
queen. The rest of the society consists of numerous female workers, who
are the sterile daughters of the queen, and the male drones, who are the
fertile sons. Darwin knew nothing of genetics, yet it is in the peculiar
haplodiploid genetics of honeybee reproduction that a clue can be seen
in the evolution of eusocial behavior. The queen produces both fertilized
eggs, which develop into diploid females, and unfertilized eggs, which
develop into haploid males. In the classic model of eusocial evolution,
Hamilton (1964) pointed out that the haplodiploid genetic system pro-
duces sisters that are more closely related to one another than they
would be to their own offspring. This is due to the fact that all sisters
share an identical set of chromosomes from their father (as the male is
haploid), and thus are 75 percent related to one another, whereas they
would share only half their chromosomes with a daughter (50 percent
Table 6.4
Convergent evolution of eusocial behavior
Convergent behavior and function: EUSOCIALITY (cooperative social system in which

only one female is reproductive, and labor behaviors are complexly divided among the
numerous other individuals within the colony, resulting in a highly stable and competitive
colonial organization able to sustain high population densities)
1 Regalis sponge-dwelling shrimp (Bilateria: Protostomia: Ecdysozoa: Arthropoda:
Mandibulata: Malacostraca: Decapoda: Alpheidae; Synalpheus regalis)
2 Chacei sponge-dwelling shrimp (Arthropoda: Mandibulata: Malacostraca: Decapoda:
Alpheidae; Synalpheus chacei)
3 Small paraneptunus sponge-dwelling shrimp (Arthropoda: Mandibulata:
Malacostraca: Decapoda: Alpheidae; Synalpheus “paraneptunus small”)
4 Honeybee (Arthropoda: Mandibulata: Hexapoda: Hymenoptera: Apoidea: Apidae:
Apini; Apis mellifera)
5 Stingless bee (Arthropoda: Mandibulata: Hexapoda: Hymenoptera: Apoidea: Apidae:
Meliponini; Mellipona compressipes)
6 Halictus sweat bee (Arthropoda: Mandibulata: Hexapoda: Hymenoptera: Apoidea:
Halictidae; Halictus (Halictus) quadricinctus)
7 Lasioglossum sweat bee (Arthropoda: Mandibulata: Hexapoda: Hymenoptera:
Apoidea: Halictidae; Lasioglossum (Dialictus) figueresi)
8 Augochlorella sweat bee (Arthropoda: Mandibulata: Hexapoda: Hymenoptera:
Apoidea: Halictidae; Augochlorella pomoniella)
9 Yellow-jacket wasp (Arthropoda: Mandibulata: Hexapoda: Hymenoptera: Vespoidea:
Vespidae: Vespinae; Vespula maculifrons)
10 Hover wasp (Arthropoda: Mandibulata: Hexapoda: Hymenoptera: Vespoidea:
Vespidae: Stenogastrinae; Parischnogaster mellyi)
11 Dracula ant (Arthropoda: Mandibulata: Hexapoda: Hymenoptera: Vespoidea:
Formicidae: Ponerinae; Amblyopone pallipes)
12 Galling thrip (Arthropoda: Mandibulata: Hexapoda: Thysanoptera: Tubuliferidae:
Phlaeothripinae; Oncothrips tepperi)
13 Horned-soldier aphid (Arthropoda: Mandibulata: Hexapoda: Hemiptera:
Hormaphididae: Cerataphidini; Pseudoregma sundanica)
14 Australoplatypus ambrosia beetle (Arthropoda: Mandibulata: Hexapoda:
Coleoptera: Curculionidae: Platypodini; Australoplatypus incompertus)
15 Damp-wood termite (Arthropoda: Mandibulata: Hexapoda: Isoptera: Termopsidae;
Zootermopsis nevadensis)
16 Naked mole rat (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Euarchontoglires: Rodentia: Hystricognatha: Bathyergidae; Heterocephalus glaber)
17 Damaraland mole rat (Mammalia: Eutheria: Euarchontoglires: Rodentia:
Hystricognatha: Bathyergidae; Cryptomys damarensis)

230 Chapter 6
related; Holmes et al. 2009). Under the predictions of the theory of

natural selection, it is more advantageous for sisters to help raise more
sisters (produced by their mother, the queen) than it is for them to
produce their own offspring. Of the 17 evolutionary lineages that have
convergently evolved eusocial behavior (table 6.4), over half are haplo-
diploid: the bees, in five independent lineages (Cameron and Mardulyn
2001; Brady et al. 2006); the wasps, two independent lineages (Hines
et al. 2007); the ants, one lineage (Thorne and Traniello 2003); and the
thrips, one lineage (Crespi et al. 1997).
At first glance, the convergent evolution of eusocial behavior
might be seen as a function of developmental constraint, in that it is
associated with the haplodiploid genetic system. However, eusocial
behavior has also convergently evolved in lineages that are diploid
(table 6.4), lineages that have no asymmetric degrees of relationships
between siblings and offspring of either sex: shrimp, in three lineages
(Duffy et al. 2000); aphids, one lineage (Shingleton and Foster 2001);
beetles, one lineage (Mueller et al. 2005); termites, one lineage (Johns
et al. 2009); and rodents, two lineages (Scantlebury et al. 2006; Holmes
et al. 2009).
In contrast to haplodiploid lineages, the convergent evolution of euso-
cial behavior in diploid lineages has been argued to be the product of
functional constraint: “the inability of the haplodiploid hypothesis to
explain all eusocial evolution has returned some attention to the ecologi-
cal benefits of eusocial behavior” (Johns et al. 2009, 17452). In addition,
many have argued that eusocial species have a competitive advantage in
highly crowded environments (Duffy et al. 2000; Wilson and Hölldobler
2005). Indeed, all of the 17 eusocial lineages are found in organisms that
inhabit sheltered, enclosed nesting sites: sponge-dwelling shrimp, plant-
gall-dwelling aphids and thrips, wood-boring beetles, hive-dwelling bees
and wasps, and subterranean-dwelling ants, termites, and mole rats (table
6.4). Because similarly protected nesting sites may be difficult to find or
construct, Holmes et al. (2009) assert, it is less risky to remain at the
original site than to seek another, leading to overlapping generations of
crowded adults living in a single group, cooperative breeding within the
group, and division-of-labor behaviors within the group.
There still exists no universally accepted causal hypothesis for the
evolution of all eusocial behavior. Some suggest that the phenomenon
is analogous to the convergent self-assembly of molecules (Camazine
et al. 2001), while others suggest that it is analogous to the evolution of
multicellularity (Boomsma 2009). Wilson and Hölldobler (2005) propose
a mix of selective causal factors—an interplay of group selection, kin

selection, and individual selection. Others stress ecological factors: the
“convergent similarities in behavior . . . suggest that highly eusocial orga-
nization has limited permutations. Each tribe has been channeled along
a similar behavioral track, responding in similar fashion to similar con-
tingencies” (Cameron and Mardulyn 2001, 209).
In summary, the independent evolution of eusocial behavior may be
produced by developmental constraint (intrinsic genetic systems), func-
tional constraint (extrinsic habitat ecologies), or the action of the two
constraints in concert. The convergent evolution of hive minds can be
spectacularly successful: the number of species of haplodiploid ants and
diploid termites constitutes only 2 percent of all known insect species,
yet they together constitute more than half of the biomass of all insects
on Earth (Wilson and Hölldobler 2005).
Another highly convergent form of group behavior is the collective
mind of the herd (table 6.5): “Herding is a form of convergent social
behaviour that can be broadly defined as the alignment of the thoughts
or behaviours of individuals in a group (herd) through local interaction
and without centralized coordination” (Raafat et al. 2009, 420). Everyone
in Europe and North America is familiar with the huge flocks of migrat-
ing European starlings that flow through our skies like a veritable river
of birds. Unlike a river of water acting under the influence of gravity, a
bird flock is a river of minds that are acting in concert: “Consider a flock
of starlings under attack by a peregrine falcon: The flock contracts,
expands, and even splits, continuously changing its density and structure.
Yet, no bird remains isolated, and soon the flock reforms as a whole”
(Ballerini et al. 2008, 1232). This group behavior is not unique to the birds
nor to the terrestrial realm—far out in the oceans, “striking similarities
between the internal organization of bird flocks and fish schools” can be
seen in the silvery schools of fish responding to the attack of a shark
predator (Couzin and Krause 2003, 37).
Couzin (2007, 715) considers the collective minds of a moving herd to
function like “an integrated self-organizing array of sensors,” in that
critical information, like the detection of a source of food or an oncoming
predator, “may often be detected by only a relatively small proportion
of group members. . . . [B]ehavioural coupling among near neighbours,
however, allows a localized change in direction to be amplified, creating
a rapidly growing and propagating wave of turning across the group. This
positive feedback results from the ability of individuals to influence and
be influenced by others, and allows them to experience an ‘effective
232 Chapter 6
Table 6.5
Convergent evolution of collective animal behavior
1 Convergent behavior and function: HERDING IN TWO DIMENSIONS (behavioral

coupling in a large group of animals to produce an integrated self-organizing array of
sensors distributed over a large two-dimensional geographic area to detect food or
predators)
1.1 Desert locust (Bilateria: Protostomia: Ecdysozoa: Arthropoda: Mandibulata:
Hexapoda: Orthoptera: Acrididae; Schistocerca gregaria)
1.2 Leptothorax ant (Arthropoda: Mandibulata: Hexapoda: Hymenoptera: Vespoidea:
Formicidae: Myrmicinae; Leptothorax albipennis)
1.3 Alamosaur (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Sauropoda: Titanosauridae; Alamosaurus sanjuanensis †Cretaceous)
1.4 Maiasaur (Dinosauria: Ornithischia: Cerapoda: Ornithopoda: Hadrosauridae;
Maiasaura pebblesorum †Cretaceous)
1.5 African elephant (Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
Afrotheria: Proboscidea: Elephantidae; Loxodonta africana)
1.6 Blue wildebeest (Mammalia: Eutheria: Laurasiatheria: Cetartiodactyla: Ruminantia:
Bovidae; Connochaetes taurinus)
2 Convergent behavior and function: HERDING IN THREE DIMENSIONS (behavioral
coupling in a large group of animals to produce an integrated self-organizing array of
sensors distributed over a large three-dimensional spatial volume to detect food or
predators)
2.1 Antarctic krill (Bilateria: Protostomia: Ecdysozoa: Arthropoda: Mandibulata:
Malacostraca: Euphausiacea: Euphausiidae; Euphausia superba)
2.2 Honeybee (Arthropoda: Mandibulata: Hexapoda: Hymenoptera: Apoidea: Apidae:
Apini; Apis mellifera)
2.3 Monarch butterfly (Arthropoda: Mandibulata: Hexapoda: Lepidoptera: Danaidae;
Danaus plexippus)
2.4 Atlantic herring (Bilateria: Deuterostomia: Chordata: Osteichthyes: Actinopterygii:
Clupeiformes: Clupeidae; Clupea harengus)
2.5 European starling (Osteichthyes: Sarcopterygii: Reptiliomorpha: Amniota:
Sauropsida: Archosauromorpha: Dinosauria: Saurischia: Theropoda: Maniraptora:
Aves: Neognathae: Neoaves: Passeriformes: Sturnidae; Sturnus vulgaris)
2.6 Bottlenose dolphin (Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
Laurasiatheria: Cetartiodactyla: Cetacea: Odontoceti: Delphinidae; Tursiops truncatus)
range’ of perception much larger than their actual sensory range.” In the
European starling (table 6.5), this behavioral coupling is a function of
topological distance: each bird interacts with a fixed number of neighbor-
ing birds, typically six or seven, instead of with all birds contained in a
fixed volume of metric space (Ballerini et al. 2008). This topological
interaction allows the size of the flock of birds to expand and contract
without altering the behavioral coupling between the birds.
Field studies with the desert locust (table 6.5) have revealed a critical
threshold effect that triggers herd behavior. Buhl et al. (2006, 1402)
observed that each locust behaves individually until the number of
locusts present per square meter of land area reaches the number eight,
at which point “a rapid transition occurs from [the] disordered move-
ment of individuals within the group to highly aligned collective motion.”
Buhl et al. (2006) further argue that this critical density effect is a key
triggering factor in all herding behavior, although the threshold number
is different for different animal groups.
In table 6.5 I have listed as many disparate convergent lineages that
have evolved collective minds, or herd behavior, that I am aware of. I
have made no effort to try to determine how many times herd behavior
may have convergently arisen within these major lineages; future phylo-
genetic studies of within-lineage convergence in herd behavior will cer-
tainly expand the list given in table 6.5. Within the mammalian clade, I
have listed examples of convergent lineages in the three major branches
of existent mammals: the afrotherians, laurasiatherians, and euarchonto-
glires. Some ruminant ungulates form enormous self-organized groups,
as seen in the behavior coupling that produces the wavelike front of
100,000 moving individuals in a wildebeest herd in Africa (Couzin and
Krause 2003). Elephants form smaller herds, and even we humans exhibit
markedly different behavior when in groups than when alone, and
unconsciously adopt coordinated behavior when we are moving as
pedestrians along the sidewalks of our cities (Raafat et al. 2009). At the
opposite extreme, swarming ants and marching locust juveniles can form
even larger herds that cover many kilometers in area even though the
individual animals themselves are tiny (Buhl et al. 2006; Conradt and
Roper 2005).
The independent evolution of herding behavior by the Mesozoic dino-
saurs is yet another example of the ecosystem convergence that exists
between the dinosaurian and mammalian ecosystems (see table 4.10). In
table 6.5 I have listed examples of convergent lineages in the two major
branches of the dinosaur clade, the saurischians and the ornithischians.
234 Chapter 6
Preserved fossil trackways demonstrate that sauropods, such as the ala-

mosaur, moved in herds of over 30 individuals. The actual size of the herd
was undoubtedly much larger. In the ornithischians, trackways of more
than 80 ornithopods in a herd have been found (Martin 2006). In one
spectacular example, a herd of maiasaurs was poisoned by gases from a
volcanic eruption, producing a mass-kill horizon containing over 10,000
individuals in a bone bed preserved under a layer of volcanic ash in
present-day Montana. A range of ages is evident in the herd, from juve-
nile maiasaurs only 3 meters in length to older adults over 7.5 meters in
length (Weishampel and Horner 1990).
Flocking starlings and schooling fish, like the Atlantic herring (table
6.5), have already been mentioned in the introduction to this section of
the chapter. Here I simply wish to emphasize that although birds and
fish are radically different kinds of animals and live in radically different
environments (air versus water), their minds have still converged on the
same group behavior. These animals are vertebrates, but the same con-
vergence occurs in the arthropods, as seen in the swarming Antarctic krill
in the oceans and the flocking monarch butterflies in our skies (table
6.5). Consensus decision in groups is equally found in large-brained
dolphin vertebrates and the tiny-brained honeybee arthropods (Conradt
and Roper 2005).
In summary, Sumpter (2006, 5) argues that all herding behavior is
driven by the same behavioral principles: “These principles, such as posi-
tive feedback, response thresholds and individual integrity, are repeat-
edly observed in very different animal societies. The future of collective
behaviour research lies in . . . asking why they have evolved in so many
different and distinct natural systems.” Buhl et al. (2006, 1402) note that
“models from theoretical physics have predicted that mass-migrating
animal groups may share group-level properties, irrespective of the type
of animals in the group.” Taking these points into consideration, Couzin
(2007, 715) concludes that “group behaviour holds clues about the evolu-
tion of sociality, and also for the development of novel technological
solutions, from autonomous swarms of exploratory robots to flocks of
communicating software agents that help each other navigate through
complex and unpredictable data environments.”
Swarms of robots? Robots are machines, not living organisms. Yet
we could construct and program mechanical robots to sense and
respond to the presence of, say, six to seven neighboring robots in a
topological distance algorithm just like that of European starlings. Would
large numbers of such programmed robots spontaneously self-assemble
to produce swarm behavior? In addition, from the perspective of

theoretical morphology, could it be that there exist only a limited
number of ways in which a group of objects can move in a two-dimen-
sional plane, or a three-dimensional volume of space? How much of
convergent group behavior in animals is driven by the laws of physics
and geometry?
Herding behavior (table 6.5) can serve multiple, disparate functions:
predator avoidance, food detection, selective protection of the juvenile
members of the herd from predators, successful mass migration from one
distant region of the Earth to another without significant loss of group
members, and so on. However, there exists one type of group behavior
that has a very specific function: pack hunting. Pack hunting is a form of
social, coordinated hunting that enables a group of small predators to
kill prey animals that are much larger than they are. In table 6.6 I have
listed all of the disparate convergent lineages that have evolved pack-
hunting behavior that I am aware of. In one instance we know that a
Table 6.6
Convergent evolution of social hunting behavior
1 Convergent behavior and function: PACK HUNTING (social, coordinated hunting by

a group of small predators, enabling them to kill prey animals that are much larger than
they are)
1.1 New World army ant (Bilateria: Protostomia: Ecdysozoa: Arthropoda: Mandibulata:
Hexapoda: Hymenoptera: Vespoidea: Formicidae: Ecitoninae; Eciton burchelli)
1.2 Deinonychosaur (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:
Theropoda: Maniraptora: Dromaeosauridae; Deinonychus antirrhopus †Cretaceous)
1.3 White pelican (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves: Neognathae:
Neoaves: Pelicaniformes: Pelicanidae; Pelecanus erythrorhynchus)
1.4 African wild dog (Amniota: Synapsida: Therapsida: Mammalia: Eutheria:
Laurasiatheria: Carnivora: Caniformia: Canidae; Lycaon pictus)
1.5 Wolf (Mammalia: Eutheria: Laurasiatheria: Carnivora: Caniformia: Canidae; Canis
lupus)
1.6 African lion (Mammalia: Eutheria: Laurasiatheria: Carnivora: Feliformia: Felidae;
Panthera leo)
2 Convergent behavior and function: CARRION SCAVENGING (carrion scavenging in
social groups for reasons that are not well understood)
2.1 Eurasian black vulture (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves:
Neognathae: Neoaves: Falconiformes: Accipitridae; Aegypius monachus)
2.2 Turkey vulture (Dinosauria: Saurischia: Theropoda: Maniraptora: Aves: Neognathae:
Neoaves: Ciconiiformes: Cathartidae; Cathartes aura)
236 Chapter 6
previously accepted example of the convergent evolution of pack hunting

turns out to be not convergent at all: the army ants. It was long thought
that pack-hunting behavior in army ants was a convergent trait, and that
the New World army ants (Ectoninae) and two lineages of Old World
army ants (Aenictinae and Dorylinae) all independently evolved collec-
tive minds. This hypothesis has been disproved by the molecular phylo-
genetic analyses of Brady (2003), which revealed that all the army ants
come from a single common ancestor that evolved pack-hunting behav-
ior 105 million years ago, before the Cretaceous breakup of the southern
supercontinent Gondwana. Thus, the current existence of army ants in
Africa and South American is due to continental drift, and not due to
convergent evolution. Army ants hunt in the tens of thousands, and liter-
ally tear apart the prey animals that they attack. Apocryphal tales of
unstoppable army ant hordes that are capable of entirely defleshing farm
animals or even humans abound in Africa.
Pack-hunting behavior has also convergently evolved in the sauropsid
dinosaurian and synapsid mammalian lineages of amniote vertebrates,
but these animals hunt in much smaller packs. In table 6.6 I have listed
one Mesozoic dinosaurian lineage, the dromaeosaurs, for which we have
solid fossil evidence of pack-hunting behavior. It is possible that social
hunting may have independently evolved in other Mesozoic theropods
as well, such as the coelurids (i.e., Coelophysis baurri; Martin 2006), and
may be revealed in future field work. The dromaeosaur Deinonychus
antirrhopus hunted in packs of six or more individuals, enabling this rela-
tively small predator (50 to 100 kilograms) to kill large ornithopod prey
such as Tenontosaurus tilleti, which weighed 1,000 kilograms (Ostrom
1990; Martin 2006). Most modern avian dinosaur predators are solitary
hunters, like the hawks and the owls. A few birds have convergently
evolved coordinated hunting behavior, but not as an adaptation for
hunting large prey animals. The white pelican (table 6.6) is a herding
animal, migrating in large numbers, but in addition “[t]hey often capture
fish cooperatively, forming a long line, beating their wings and driving
the prey into shallow water, where they seize the fish in their large,
pouched bills” (Bull and Farrand 1988, 425).
In the mammalian lineage, pack-hunting behavior has independently
evolved at least twice in the canids (table 6.6). Ecologically, the modern
wolf is the mammalian equivalent of a Cretaceous dromaeosaur (see
table 4.10; Ostrom 1990). The smaller African wild dogs are even more
vicious than wolves. Nicknamed the “super beast of prey,” packs of these
canids are seemingly unstoppable when they have chosen a prey animal,
and will run down, attack en masse, rip to shreds, and entirely eat a
gazelle in ten minutes (Wilson 1980). As such, they approach the deflesh-
ing abilities of the army ants.
The top carnivores of the mammalian ecosystem, the felids, are mostly
solitary predators. But one lineage, the African lion, has convergently
evolved pack-hunting behavior (table 6.6). These cats form social prides,
or herds, of numerous females and one to two males. Pack-hunting
behavior in the lions is also a gendered behavior, in that the females hunt
as a cohesive social unit but the males do not. Last, modern humans are
pack hunters but the reverse of the lions, in that the males were tradition-
ally the hunters and the females the gatherers. Pack-hunting behavior
enabled our small ancestors (1.5 to 1.8 meters tall, 50 to 100 kilograms
in weight) to bring down large prey like the woolly mammoths (Mam-
muthus primigenius) in Europe, which were 3 to 4 meters tall and weighed
8,000 kilograms.
Another odd form of social hunting has convergently evolved in the
carrion-eating birds (table 6.6). As discussed in chapter 4, Old World and
New World carrion-eating birds have independently evolved many mor-
phological features associated with their feasting on carcasses. Curiously,
they also have converged in aspects of their behavior: Eurasian black
vultures and North American turkey vultures are gregarious, roosting
together in large numbers at night, and often assembling in large groups
during the day as well, perched in trees or on top of buildings. Yet Eur-
asian vultures are related to hawks and eagles, which are usually not
gregarious birds, unlike the storks and flamingos that are the close rela-
tives of the turkey vultures. In both Eurasia and North America, these
birds use a soaring-in-circular-formation flight pattern in searching for
rotting corpses and, when a carcass is discovered, they will often continue
to soar in circles until other vultures arrive. Once a group has been
assembled, the flock descends to feed.
The function of this convergent behavior is not well understood. Some
have proposed that the vultures continue to circle in order to make sure
the target animal is indeed dead, or that there are no ground-dwelling
predators in the area that might pose a danger to the bird. However, the
frequency of this behavior seems to lend more weight to the suggestion
that the vultures are actually waiting for other birds to arrive, or even
summoning them. They certainly do not need a larger number of vultures
to complete the kill, because the prey animal is already dead. Carrion
is an unpredictable food resource, and vultures must often wait long
periods between feedings. When a carcass is discovered, a vulture will
238 Chapter 6
often gorge itself to the point that it has difficulty flying away. It might
seem advantageous to the bird to selfishly keep the discovery of a carcass
to itself, but at the same time, a large animal carcass is usually too much
food for a single bird to consume by itself. Thus, it might be advantageous
to share the information with other hungry vultures, in anticipation that
they will return the favor in the future when they discover a carrion
source, in a case of reciprocal altruism. Whatever the real function
of vulture group behavior, it is clear the minds of these birds have
converged.
Convergent Mentalities
When confronted with a mirror, or a window acting as a mirror, a territo-

rial bird like a cardinal (Cardinalis cardinalis) clearly thinks that it is in
the presence of another cardinal, and will attack the image in the mirror
as if it were another bird encroaching on its territory. Many animals,
when viewing a mirror, realize that another animal of their own kind is
present (kittens will often try to play with the kittens they see in the
mirror), but they never make the mental leap to the conclusion that
the animal they are viewing in the mirror is themselves. They are not
self-aware.
Humans are self-aware, and self-awareness was once thought to be a
unique human trait. As is the case with so many other mental traits that
were once thought to be unique to humans, we now know that self-
awareness has independently evolved in other lineages of animals. Within
the primates, self-awareness—just like tool construction and usage (table
6.1)—is a synapomorphy for the clade of the Hominoidae, all of the great
apes, and not confined to the human species. We did not evolve self-
awareness with the evolution of our large brains; we simply inherited it
from our ancestors.
Experiments with mirror self-recognition have thus far revealed that
elephants, cetaceans, and corvid birds are also consciously aware of their
own individual existence (table 6.7): Asian elephants (Prior et al. 2008),
bottlenose dolphins (Marino 2002), and magpies (Prior et al. 2008) all
have been demonstrated to be aware that the animal they are seeing in
the mirror is themselves, and will use the mirror to investigate their own
bodies. All of these animals have relatively large brains, but curiously it
is difficult to directly compare their brain sizes. The brain of the magpie
is tiny compared to the brain of a human, but then the body size of the
magpie is also tiny compared to that of the human. If we devise a metric
Table 6.7
Convergent evolution of self-awareness and metacognition
1 Convergent mentality: SELF-AWARENESS (the ability of an animal to be consciously

aware of its own existence)
1.1 Magpie (Amniota: Sauropsida: Archosauromorpha: Dinosauria: Saurischia:
Theropoda: Maniraptora: Aves: Neognathae: Neoaves: Passeriformes: Corvidae; Pica
pica)
1.2 Asian elephant (Amniota: Synapsida: Therapsida: Mammalia: Eutheria: Afrotheria:
Proboscidea: Elephantidae; Elephas maximus)
1.4 Orangutan (Mammalia: Eutheria: Euarchontoglires: Primates: Catarrhini:
Hominoidea: Hominoidae: Pongidae; Pongo pygmaeus)
2 Convergent mentality: METACOGNITION (the ability of an animal to monitor or
regulate its own cognitive state)
2.2 Barbary macaque monkey (Mammalia: Eutheria: Euarchontoglires: Primates:
Catarrhini: Cercopithecoidea: Cercopithecidae; Macaca sylvanus)
2.3 Human (Mammalia: Eutheria: Euarchontoglires: Primates: Catarrhini: Hominoidea:
Hominidae; Homo sapiens)
measuring the weight of the brain relative to the weight of the body
(Prior et al. 2008), the magpie has a brain-body weight index of 31, which
is greater than the index of 21 for a human. The bottlenose dolphin has
an index of only 9.0, and the Asian elephant an index of 1.6. Clearly the
brain-body weight index favors animals of small size over those of large
(Prior et al. 2008).
In chapter 5 we considered the effect of size and allometric scaling in
producing convergent tissue structures (West et al. 1999). If we plot the
logarithm of brain mass versus the logarithm of body mass, most mammals
and birds fall along a single linear function of allometric scaling (Gould
and Gould 2007). Primates, cetaceans, corvids, and elephants plot above
this line, however, indicating that they have larger brains than are strictly
necessary for neural control of their body masses. Humans and porpoises
have brains that plot the farthest away from the general allometric
scaling, but even here problems of brain-size comparison arise. The por-
poise lives in a state of weightlessness, of neutral buoyancy in water, and
does not need the large muscle masses (which require neural control)
that land animals must have in order to exist in the gravitational field of
the Earth. Much of their body mass is actually blubber fat, which they
240 Chapter 6
use for thermal insulation in the cold waters of the ocean (Gould and
Gould 2007). Thus, their large brains may be even more unusual than
that of humans.
Although self-aware animals all have unusually large brains (in the
allometric sense), their brain structures are radically different: “primate-
cetacean cognitive and behavioral convergence is a dramatic example of
functional convergence in the face of profound structural or mechanistic
divergence” (Marino 2002, 30). Even though the elaboration of the corti-
cal circuitry in primates and cetaceans is very different, they are both
members of the clade of the eutherian mammals and share much of their
subcortical neuroanatomy (Marino 2002). The corvid birds, however, do
not even have a prefrontal cortex! Although these animals are avian
dinosaurs, with vastly different brain structures from those of mammals,
“intelligence in both corvids and primates has evolved through a process
of divergent brain evolution yet convergent mental evolution . . . [showing
that] intelligence can evolve in the absence of a prefrontal cortex”
(Clayton and Emery 2008, 138–139).
Another mental state associated with high degrees of intelligence is
metacognition, the ability of an animal not only to be aware of its own
existence, but to think about its own thinking. Humans know when they
are uncertain, are aware of that cognitive condition, and are able to think
about how to deal with their uncertainty—to seek further information,
or to defer a response and to wait (Smith 2009). Thus far in this new area
of research, experimentation has revealed that the bottlenose dolphin
and the Barbary macaque monkey are aware of their own cognitive
states (table 6.7), that capuchin monkeys (Cebus apella) show equivocal
evidence of metacognition, and that pigeons (Columba livia) show no
evidence of being aware of their cognitive state (Smith 2009). Smith
(2009, 395) notes that the corvid birds, “having proven themselves cog-
nitively sophisticated, would be the usual suspects for testing in this
area,” as are elephants and great apes other than humans. It is known
that both New World and Old World monkeys have convergently evolved
tool use (see table 6.1), and the evidence that macaque monkeys have
metacognition suggests that these monkeys should be carefully exam-
ined for evidence of possible mirror self-recognition as well.
Finally, another mental state associated with convergent intelligence
is the awareness of death, as well as the experience of grief and sorrow
at the loss of another individual (table 6.8). Both gorillas and dolphins
have been observed to mourn the death of their parent or their offspring.
Table 6.8
Convergent evolution of mourning
1 Convergent mentality: FAMILIAL MOURNING (parent mourning the death of an

offspring, or offspring mourning the death of a parent)
1.2 Gorilla (Mammalia: Eutheria: Euarchontoglires: Primates: Catarrhini: Hominoidea:
Hominidae: Gorillinae; Gorilla gorilla)
2 Convergent mentality: GROUP MOURNING (mourning the death of one’s own kind,
even if only distantly related)
2.1 African elephant (Mammalia: Eutheria: Afrotheria: Proboscidea: Elephantidae;
Loxodonta africana)
2.2 Chimpanzee (Mammalia: Eutheria: Euarchontoglires: Primates: Catarrhini:
Hominoidea: Hominidae: Homininae: Panini; Pan troglodytes)
2.3 Neanderthal human (Mammalia: Eutheria: Euarchontoglires: Primates: Catarrhini:
Hominoidea: Hominidae: Homininae: Hominini; Homo neanderthalensis †Holocene)
2.4 Modern human (Hominidae: Homininae: Hominini; Homo sapiens)
Both in the wild and in zoos, gorilla and chimpanzee mothers carefully
carry around their dead infants for days, caressing them, or placing them
on the ground and pacing in circles around them, checking over and over
again to see if they have resumed breathing. Dolphin mothers repeatedly
lift their infants up in the water for days in apparent attempts to help
the calf breathe. The mothers do not eat during this period, and make
repeated distress cries.
Other animals are known to extend their mourning behavior to other
members of their own kind that are not members of their immediate
family (table 6.8). Elephants repeatedly touch the skulls and tusks of
long-dead elephants, gently lifting the bones up and down or moving
them around. They also repeatedly pace away and then return to the
remains of the dead elephant. In a stunning photograph, Monica Szczu-
pider (2009, 12–13) captured the facial expressions and body language
of 16 chimpanzees who were mourning the death of an elder female at
the Sanaga-Yong Chimpanzee Rescue Center.
Two human species evolved a mourning behavior not seen in our
great ape cousins: the burial of our dead. The oldest fossils of anatomi-
cally modern humans found to date are 200,000 years old, and are from
northeast Africa (modern Ethiopia). The Neanderthal humans have long
roots in Europe, going back some 400,000 years. It is thought that the
Neanderthal humans evolved from Homo heidelbergensis in Europe,
242 Chapter 6
sometimes described as “archaic humans,” and that H. heidelbergensis is

derived from H. erectus, the first species of human to become geographi-
cally widespread outside of Africa (Benton 2005). Modern humans, in
contrast, are thought to have evolved from H. ergaster back in Africa
(Wade 2006).
Thus, modern humans and Neanderthal humans do not share a
common species ancestor, but are cousins in the human family tree. Yet
both Neanderthal humans and modern humans buried their dead—a
mourning behavior that independently evolved in these two separately
derived species. We have no evidence that either H. ergaster, our ances-
tor, or H. heidelbergensis, the Neanderthal’s ancestor, buried their dead.
Our two geographically separated species came into contact with one
another around 100,000 years ago, in the first attempt of our species to
migrate out of Africa through the region that is now Israel. We were
stopped by the Neanderthals, who appear to have destroyed all the emi-
grant modern human populations (Wade 2006). Only 50,000 years ago
did modern humans successfully migrate out of Africa, and the last
of the Neanderthal human populations went extinct 30,000 years ago
(Wade 2006).
In summary, Emery and Clayton (2004, 1903) propose that animals
have evolved intelligence not to solve physical problems, but rather to
solve social ones: “we argue that complex cognitive abilities evolved
multiple times in distantly related species with vastly different brain
structures in order to solve similar socioecological problems.” That is,
rather than evolving as a function of the limited number of ways that
exist to solve physical problems with objects, as seen in the convergent
evolution of tool-using behavior or architectural behavior, convergent
intelligence evolves as a function of the limited number of ways that exist
to solve interactive social problems with other individuals. Greater
awareness of the behavior and potential motivation of other individuals
is hypothesized to lead to the greater awareness of one’s own existence
as an individual, and of one’s own behavior and motivation.
Convergent social systems have evolved in radically different environ-
mental systems on Earth: on land and in the oceans. Whitehead (2008,
114) notes that “terrestrial and oceanic environments provide a tough
challenge for convergence. When traits do converge, something remark-
able has occurred. Despite the radically different physical environments
. . . the social structures and cultures of sperm and killer whales have
much in common with those of elephants and humans,” and “there are
also non-mammalian species, especially birds, that have social, cognitive,
cultural, and life-history characteristics in common with the apes, odon-

tocetes [toothed cetaceans], and elephants” (Whitehead 2008, 157). The
extension of these observations to alien ecologies and environments is
obvious: If the same convergent minds arise in such vastly different eco-
logical settings as the oceans and terrestrial regions of the Earth, what
is the probability that they will also arise on extrasolar planets circling
alien stars? Will alien minds converge on ours, even though their brain
structures may be radically different?
7 Functional and Developmental Constraint in Convergent
Evolution
Our results strongly support the hypothesis that the essential elements of organic
structure are highly constrained by geometric rules, growth processes, and the
properties of materials. This suggests that, given enough time and an extremely
large number of evolutionary experiments, the discovery by organisms of “good”
designs—those that are viable and that can be constructed with available mate-
rials—was inevitable and in principle predictable.
—Thomas and Reif (1993, 342)
Was mich eigentlich interessiert, ist, ob Gott die Welt hätte anders machen
können.
—Einstein (quoted in Seelig 1956, 72)
Albert Einstein once mused, “What really interests me is whether God

could have made the world in a different way.” To Einstein, God repre-
sented the laws of nature. He was asking whether the evolution of the
universe was so constrained by the initial conditions of the Big Bang, by
the observed constants of nature, that only the present universe could
have evolved? Or were alternative universes possible, universes that
would have evolved along physical pathways not followed by our present
universe?
This question can be framed with respect to biological evolution as
well. Is the evolution of life so constrained by the geometry of the uni-
verse, by the physical constants of nature, that its outcome is predictable?
Or are so many alternative evolutionary pathways possible that it would
never be possible to predict the trajectory of the evolution of life? We
can easily visualize a universe in which every species is morphologically
different from every other species, and in which each species has its own
unique ecological role, or niche, in nature. That universe does not exist.
Instead, we live in a universe where convergence in evolution is rampant
at every level, from the external forms of living organisms down to the
246 Chapter 7
very molecules from which they are constructed, from their ecological
roles in nature to the way in which their minds function.
Since convergent evolution is so ubiquitous in nature, as we have seen
in the previous five chapters, the total extent of convergent evolution
might best be revealed by studying its opposite: unique evolution. That
is, rather than compiling lists of convergences, we might compile lists of
solitary evolutionary innovations in species that have not been indepen-
dently discovered by other species in their evolutionary pathways.
Vermeij (2006) set out to do just that, and compiled a list of evolutionary
innovations said to be unique. He discovered that “purportedly unique
innovations either arose from the union and integration of previously
independent components or belong to classes of functionally similar
innovations” and that “important ecological, functional, and directional
aspects of the history of life are replicable and predictable” (Vermeij
2006, 1804).
What are the possibilities for evolution in our universe? Can we
even think about considering the total spectrum of what is possible and
not possible in biological evolution? The answer is yes, by using the
analytical techniques of theoretical morphology, in particular by the
construction of theoretical morphospaces (McGhee 2001, 2007). The
concept of the theoretical morphospace originated in evolutionary
biology (McGhee 1999), but it has subsequently caught the attention of
philosophers (Maclaurin 2003), linguists, cultural anthropologists, and
neuroscientists (Hauser 2009) who are seeking to explore the spectrum
of both possible and impossible languages and cultures. Here we will
use the concept to analyze the phenomenon of convergent evolution
with respect to the spectrum of existent, nonexistent, and impossible
biological form.
Convergent Evolution in Theoretical Morphospace
The analytical techniques of theoretical morphology allow us to take a

spatial approach to the concept of convergent evolution. Any given bio-
logical form, or f in abbreviation, may be described by a set of measure-
ments taken from that form—how tall is it, how wide, how long? Each
type of measurement (height, width, length, etc.) can be considered as a
dimension of form. The total set of the possible dimensions of form can
be used to construct a hyperdimensional morphospace of possible form
coordinates (figure 7.1). Each point within this theoretical morphospace
represents a specific combination of form measurements that will produce
Functional and Developmental Constraint in Convergent Evolution 247
Figure 7.1
A theoretical hyperdimensional space of possible form (modified from McGhee 2007).
Each dimension of the space represents a morphological trait that may be measured on a
given biological form, f. All possible coordinate combinations (points) within the theoreti-
cal morphospace represent the set of all possible biological forms. Although only eight
dimensions are shown in this schematic diagram, the dimensionality of an actual hyper-
space of form will be much larger.
the form coordinate for a hypothetical form f. Convergence occurs when

forms originally present in different regions of the morphospace evolve
in such a way that they move to the same spatial region in the morpho-
space (figure 7.2).
Returning to figure 7.1, we can begin to consider the effects of
evolutionary constraint in producing convergent evolution. First, we
know that some types of forms function in nature: these are the myriad
forms of life that surround us here on Earth. The opposite of this
observation is the concept that there exist forms that do not function
in nature, and that if a living organism were to produce one of these
forms, it would be lethal. Mutations that produce nonfunctional, lethal
forms are well known in biology. We can thus conceptually divide
the spectrum of form within the morphospace into spatial regions
that contain nonfunctional form and functional form (figure 7.3). The
boundary between the spatial regions of form separates nonfunctional
from functional forms, and is itself a spatial representation of the
concept of functional constraint. That is, if the evolving form is to
remain functional, it must remain within the functional region of forms
in the morphospace (an aside: these boundaries are not mere idle
248 Chapter 7
Figure 7.2
Convergent evolution of form within a theoretical morphospace. Organisms in different
regions of the morphospace, and thus possessing different initial morphologies, have evo-
lutionary trajectories that take them to the same region within the morphospace, and their
morphologies converge.
Figure 7.3
Functional constraint in theoretical morphospace. Functional forms are found within the
rectangular region in the morphospace, and nonfunctional forms are found outside the
rectangular region. The boundaries of the rectangle thus delimit the functional constraint
boundary on possible form within the morphospace.
speculation, but can actually be mapped in theoretical morphospaces;

see McGhee 2007).
Second, we know that organisms develop biological form from an
original cell. We also know that the possible types of form that can be
developed from a given cell are limited, that they depend upon the DNA
coding within the cell and the interaction of the different molecules and
tissue geometries produced as the cell grows. These observations lead to
the concept of developmental constraint: the different types of forms that
different organisms can develop are limited.
Now let us consider the spatial representation of these two types of
evolutionary constraint within the morphospace (figure 7.4). The func-
tional constraint boundary in figure 7.4 is the same as in figure 7.3, but
now we have added a developmental constraint boundary as well (the
dotted line in figure 7.4). That is, the evolving form must remain within
the developmentally possible regions of forms in the morphospace. As
can be seen in figure 7.4, the functional constraint boundary within the
morphospace does not have to coincide spatially with the developmental
constraint boundary.
Figure 7.4
Combining functional and developmental constraint in theoretical morphospace. The
boundaries of the dotted-line rectangle delimit the developmental constraint boundary on
possible form within the morphospace: forms within the dotted-line rectangle are devel-
opmentally possible, while forms outside the dotted-line rectangle are developmentally
impossible. Forms f:0 are thus both nonfunctional and developmentally impossible, forms
f:1 (shaded region) are both functional and developmentally possible, forms f:2 are non-
functional but developmentally possible, and forms f:3 are functional but developmentally
impossible. See text for discussion.
250 Chapter 7
We now are in the position to consider the spectrum of all possible

existent, nonexistent, and impossible biological forms. Let us consider
the developmental constraint boundary present in figure 7.4 to spatially
portray the spectrum of all biological forms that can be developed by
life on Earth. That is, forms within this region can be developed by organ-
isms present on Earth, but forms outside this region cannot. We now see
that four different regions of potential form exist within the morpho-
space, in the form of a Venn diagram:
(1) Biological forms that do not function and that cannot be developed
by Earth life, abbreviated f:0. These are all of the hypothetical forms
shown in figure 7.1 that fall outside the two intersecting rectangles shown
in figure 7.4.
(2) Biological forms that do function and that can be developed by
Earth life, abbreviated f:1. These are all of the hypothetical forms shown
in figure 7.1 that fall in the region where the two rectangles shown in
figure 7.4 intersect (the shaded region).
(3) Biological forms that could be developed by Earth life, but that do
not function and thus are lethal, abbreviated f:2. These are all of the
hypothetical forms shown in figure 7.1 that are located in figure 7.4 in
the region bounded by the dotted-line rectangle but not located in the
region where it intersects with the solid-line rectangle (the shaded
region).
(4) Biological forms that do function but that cannot be developed by
Earth life, abbreviated f:3. These are all of the forms shown in figure 7.1
that are located in figure 7.4 in the region bounded by the solid-line
rectangle but are not in the region where it intersects with the dotted-line
rectangle (the shaded region).
Now let us consider these four regions of possible form from the point
of view of actual existent life on Earth. The myriad forms of life that
surround us here on Earth clearly are both functional and developable,
and we can collect all of these forms into a set of forms { f:1}. Every living
thing on Earth, from butterflies to bacteria, belong to the form set { f:1}.
Needless to say, all of the convergent forms of life that we have consid-
ered throughout this book are also members of the form set { f:1}. The
phenomenon of convergent evolution immediately reveals to us that the
size of form set { f:1} is not infinite, in that life on Earth has been con-
strained to repeatedly reevolve the same forms within this set over and
over. The hypothetical universe in which every species has its own unique
functional morphology, is morphologically different from every other

species, does not exist.
Biologists over the ages have studied the many different mutant forms
of life—two-headed snakes, three-legged frogs, and so on—that are lethal
mutations (which are important in that they give us valuable clues about
the process of development; see the discussions in Alberch 1989 and
Blumberg 2009). Developmental abnormalities like two-headed snakes
and frogs with three hind legs instead of two are real and are also non-
functional, in that they do not survive in the wild (as opposed to in the
laboratory). We can collect all such developmental “freaks of nature”
(Blumberg 2009) into the set of forms { f:2}.
Visualizing the types of forms that are both nonfunctional and that
cannot be developed by life on Earth is a bit more difficult, but it can be
done mathematically (see McGhee 2007), and we can place all of these
hypothetical form coordinates within the morphospace into the form set
{ f:0}. They do not really matter for our discussion here, but they do need
to be listed so that we will have considered the complete spectrum of all
possible existent, nonexistent, and impossible biological forms. So, thus
far we have three sets of form with respect to Earth life: one functional
set of forms of Earth life, { f:1}; one nonfunctional set of forms of Earth
life, { f:2}; and one impossible set of forms, { f:0}.
The last possible set of forms is crucial to our understanding of the
implications of the phenomenon of convergent evolution. These are the
possible forms of life that are functional, that work just fine in nature,
but that nevertheless cannot be developed by life on Earth. We will place
these possible but nonexistent forms on Earth in the final set of forms,
{ f:3}. We have now arrived at the critical question in the analysis of con-
vergent evolution: Does the form set { f:3} exist in the universe? We know
that it does not exist on Earth, but do living forms exist elsewhere on
alien worlds that belong to the form set { f:3}?
At first this question may seem to be so abstract as to be of no real
importance, but it is of critical importance in our consideration of the
question of predictability in evolution (which we will consider in detail
in the next chapter). We know that much of the convergent evolution of
life on Earth is driven by developmental constraint, yet we do not know
the answer to two of the most fundamental questions concerning devel-
opmental constraint itself: “How did development originate [on Earth]?”
and “How did the developmental repertoire evolve?” (Müller 2007, 944).
Only when we have the answers to those two questions can we make
predictions about the possible evolution of alien developmental
252 Chapter 7
repertoires, and whether or not those developmental repertoires might

be similar to those found in Earth life. The existence of the form set { f:3}
would imply that there exists somewhere in the universe an alien set of
functional biological forms that nevertheless cannot be developed by
Earth life, and thus these alien life forms would not be convergent on any
of the forms of life seen on Earth.
Life as We Know It
All life on Earth belongs to the form set { f:1}, biological forms that both
function and can be developed by Earth life. As we have seen in chapters
2 through 6 in this book, life on Earth is highly convergent. The spectrum
of forms in the set { f:1} have been reevolved over and over again in the
history of life on Earth.
The set of existent forms of life on Earth, { f:1}, is a function of the
functional constraint boundary and the developmental constraint bound-
ary in the theoretical morphospace (figure 7.4). We believe the functional
constraints of physics and geometry to be the same throughout the uni-
verse and to be extrinsic with respect to biology (McGhee 2007); there-
fore, this constraint boundary should also apply to life forms throughout
the universe. That is, the spatial position of the functional constraint
boundary shown in figure 7.4, relative to the spectrum of hypothetical
forms shown in figure 7.1, should remain the same throughout the uni-
verse. Alien life forms should belong to the same set of forms { f:1} that
Earth life forms belong to, given the same physical conditions on that
alien planet.
Thus, we could predict that those alien organisms with powered flight
will have evolved wings, just as animals on Earth have done (table 2.2),
and fast-swimming organisms in alien seas will evolve streamlined, fusi-
form bodies (table 2.1). Land-dwelling, sessile, photoautotrophic organ-
isms will evolve tree forms, just as plants on Earth have done (table 3.1),
and those tree forms will have leaflike structures (table 3.3). The laws of
physics and geometry also extend to activities, in that there are a limited
number of ways to build structures (table 6.2) and to successfully culti-
vate crops (table 6.3). The geometries of alien buildings, and the algo-
rithms of farming procedures, should be similar to those found on
Earth. If they live in highly social groups, natural selection should drive
evolution to produce aliens that are self-aware, just as it has on Earth
(table 6.7).
Life as We Do Not Know It
It is the second boundary in theoretical morphospace—the developmen-

tal constraint boundary—that undermines our ability to confidently
predict that life throughout the universe will converge on the same forms
seen here on Earth. Developmental constraints are intrinsic (McGhee
2007), imposed by the laws of biology of specific organisms, in this case
the biology of Earth organisms. Will these constraints be the same for
alien organisms?
Even in the case of Earth organisms themselves, developmental con-
straint introduces an element of uncertainty into our ability to precisely
predict convergence. We can be confident in predicting that animals that
fly will have wings—but how many? Birds, bats, and pterosaurs are devel-
opmentally constrained to have only two wings, as their tetrapod phylo-
genetic legacy provides them with only two forelimbs that can be modified
in convergent evolution. This developmental constraint does not apply
to insects, however, and so the dragonfly has four wings, not just two. The
mythical centaurs have never been evolved by mammals; their tetrapod
phylogenetic legacy provides them with only four appendages to modify,
not six. The insects, however, are developmentally provided with six
appendages, and they have indeed independently evolved centaur forms
twice: the mantids and the mantispids (as discussed in chapter 2). Hence,
alien organisms, with a totally different developmental repertoire, might
evolve forms never even dreamed of in human mythology.
We have already considered one possible example of a potential func-
tional structure that apparently cannot be developed by Earth life: eyes
that can see infrared light (chapter 2). Multicellular life with eyes has
existed on Earth for over 600 million years, yet never has been able to
develop eyes that can see light with long wavelengths, wavelengths in the
infrared, even though possession of those eyes would give a significant
selective advantage to predators. Both camera-eyed animals and com-
pound-eyed animals have had to evolve alternative organ systems in
order to “see” infrared (table 2.5). And we humans have constructed
machines that enable us to “see” into the infrared region of the spectrum.
Could an alien life form exist that has been able to do so organically? If
so, that alien would have a form that belongs to form set { f:3}. Or does
alien life elsewhere experience the same developmental constraints that
forms of life on Earth do? If so, the form set { f:3} does not exist—there
are no functional forms of life that cannot be developed by life on Earth
254 Chapter 7
simply because life on other planets cannot develop these forms either.
Several lines of reasoning, however, suggest that this may not true, and
that the set of forms { f:3} potentially exists in the universe. NASA, the
space exploration agency of the United States, is also very interested in
the possible existence of non-Earth-type life, as discussed in Peter Ward’s
book Life as We Do Not Know It (2005). In trying to conceive of possible
forms of life that are nonexistent on Earth, the scientists at NASA are
engaging in classic theoretical morphology.
Let us also engage in a theoretical-morphology thought experiment.
Development of life on Earth starts from a single cell and, as far as we
currently understand it, depends upon the genetic coding within the cell
and the interaction of the different molecules and tissue geometries
produced as the cell grows. Fundamentally, we are dealing with mole-
cules and geometries: all Earth life is carbon-based, uses DNA to code
for amino acid assembly, and is constructed of amino acids that have
left-coiled geometries; furthermore, multicellular life develops complex
tissue geometries from simple, spherical, topologically solid cell clusters
(Newman 2010). So, let us consider each of these characteristics in
turn with regard to the question: Could hypothetical life elsewhere be
different? Can we conceive of possible life forms that are nonexistent
on Earth?
All life on Earth is carbon-based: non-carbon-based life would surely
experience different developmental constraints than those present in
Earth life, and therefore could possibly develop functional biological
forms that Earth life cannot. The most widely discussed chemical alterna-
tive to carbon is silicon-based life. From the periodic table of elements,
we see that silicon can form four bonds like carbon (for example, in the
formation of monosilane, SiH4, the analog of methane, CH4), which is so
important for building the carbon-backboned molecules of Earth life.
Unfortunately, silicon bonding simply cannot produce the huge macro-
molecules that carbon can—macromolecules that are very strong and
stable at normal Earth temperatures and pressures. Curiously, however,
it is thought that silicon-based life might be possible at either very high
temperatures or very low temperatures (Ward 2005). At very low tem-
peratures, silicon can form polymers, large macromolecules that are
analogs to those formed by carbon. Rather than using water as a solvent,
as with carbon-based life, silicon-based life could use liquid methane or
nitrogen. As a result, it has been hypothesized that silicon-based life
might exist in the methane lakes that exist on Titan, Saturn’s planet-like
moon that possesses its own nitrogen atmosphere, or in possible liquid
nitrogen lakes on Triton, the larger moon of Neptune, and, obviously, in

similar worlds elsewhere in the universe (Ward 2005). In particular,
McKay and Smith (2005) have shown that is energetically possible for a
hypothetical Titan life form to metabolize acetylene (C2H2), present on
the surface of Titan, using atmospheric hydrogen (H2) to produce
methane (CH4), analogous to Earth life that metabolizes sugar (C6H12O6)
with atmospheric oxygen (O2) to produce carbon dioxide (CO2). They
further predict that the presence of such a hypothetical methanogenic
life form on Titan should result in chemically anomalous depletions of
acetylene on the surface of Titan, and the anomalous depletion of hydro-
gen in Titan’s lower atmosphere. Interestingly, as of this writing, both of
these predictions have been observed to occur on Titan by the orbiting
Cassini spacecraft (Strobel 2010; Clark et al. 2010), precipitating renewed
debate concerning possible life on Titan.
Such hypothetical life forms on cold planets would have developmen-
tal pathways radically different from those present in Earth life, and thus
could evolve functional morphologies in their cold habitats totally unlike
any seen on Earth. If these biological forms exist, they would belong to
form set { f:3}. Since Earth life cannot exist on Titan or Triton, it is obvious
that Earth life cannot develop a biological form that could function in a
lake of liquid methane.
Ward (2005) points out that the zone within our solar system contain-
ing cold-type environments in which methane and nitrogen are liquids
is much wider than the narrow zone in which water is liquid. Only the
Earth and Mars exist in this zone today, and Mars is at its very outermost
fringe. Consequently, silicon-based life, if it exists, would have a much
larger suite of worlds and moons within our own solar system upon which
it could exist compared to carbon-based life, and this will be true of other
solar systems as well. Could silicon-based life be more abundant in the
universe than carbon-based life?
Conway Morris (2003) argues that we should expect to find that life
is carbon-based wherever life exists in the universe. According to this
argument, functional constraints in molecular evolution will always drive
natural selection for the uniquely versatile carbon atom in the formation
of polymers, or complex macromolecules, throughout the universe. No
other element in the periodic table of elements has carbon’s chemical
ability to form four bonds that are strong, to form double bonds, to
link together to form stable rings and chains, and to form a gas by com-
bining with oxygen (CO2; silicon readily combines with oxygen, but pro-
duces rigid crystalline SiO2, which presents serious problems if the
256 Chapter 7
silicon-based life is to use oxygen in its metabolism). If molecular evolu-

tion always converges on carbon in the formation of life, then all life
elsewhere will be confined to Earth-like worlds with liquid water. This
reasoning also leads Conway Morris (2003) to conclude that life could
be quite rare in the universe, as the temperature zone around a star that
could contain Earth-like worlds is narrow (Ward 2005).
In summary, if silicon-based life exists in the universe, it could poten-
tially develop forms that belong to form set { f:3}, life forms that could
potentially not be convergent with any life found on Earth. On the other
hand, if all life in the universe is carbon-based and found only on Earth-
like worlds, then life elsewhere could be highly convergent on the same
life forms found on Earth.
All life on Earth is coded for by base-four DNA, coding with the
nucleotides ATCG, which are read in a triplet. Could life exist elsewhere
that is not thus coded? Life that uses a different coding system would
surely experience different developmental constraints than base-four
DNA-coded life, and so could possibly develop functional biological
forms that Earth life cannot. The easiest way to envision such life forms
would be to keep the same sugar-phosphate chains that are present in
DNA, but then to add or subtract the number of nucleotides that code
for protein. Laboratory experiments have actually produced synthetic
base-six DNA, molecules having six nucleotides rather than four (Ward
2005). Could life have arisen elsewhere that uses base-six DNA coding
as part of its developmental system?
Conway Morris (2003) argues that we should expect to find that life
uses base-four DNA with triplet codons wherever life exists in the uni-
verse. Just as with the element carbon, Conway Morris (2003) points out
that computer simulations of alternative coding systems have shown that
base-four DNA coding is vastly more efficient than any other hypotheti-
cal coding system: less than base-four coding is too limited and restricted,
and greater than base-four coding is too cumbersome and error-prone.
Freeland and Hurst (1998) have further experimented with base-four
DNA with triplet codons by randomizing the codon positions to produce
alternative genetic codes. They discovered that, in a sample of over a
million alternative codes, only one code was more efficient than the
natural code found in life on Earth. While this result can be used to
support the view that carbon-based life elsewhere will convergently
evolve the same genetic code that was evolved by life on Earth (Conway
Morris 2003), from the perspective of theoretical morphology the dis-
covery of a more efficient but nonexistent code is of more interest. Even
here, however, the nonexistent code “shows behavior very similar to that
of the natural code” (Freeland and Hurst 1998, 247); thus, it is not clear
whether life evolved using this alternative code would be that much dif-
ferent from life on Earth.
The amino acid molecules that are used as the building blocks of pro-
teins in Earth life are all left-coiled. Could life exist elsewhere that uses
right-coiled amino acids? Experiment shows that chemical selection in
protein synthesis produces homochirality, the use of only one type of
amino acid geometry. But proteins can be constructed in the laboratory
with either all left-coiled or all right-coiled amino acids, and it is not clear
why life on Earth has chosen the left-coiled chemical pathway (Castelvec-
chi 2007). Both right-coiled and left-coiled amino acids are produced by
inorganic chemical processes in nature, and so perhaps chance has pro-
duced life forms elsewhere that are constructed of proteins made with
right-coiled amino acids. Would such life produce biological forms that
are simply the chemical mirror-image of Earth life forms? Or would
those forms be different from anything seen on Earth, and might they
belong to form set { f:3}?
Just as the element carbon has unique properties and the DNA coding
system found in Earth life is highly efficient, some have argued that
molecular evolution preferentially produces left-coiled amino acids.
Experiments have suggested (1) that left-coiled amino acids are slightly
more stable than right-coiled, and thus could have accumulated more
readily in prebiotic environments than right-coiled, and (2) that left-
coiled amino acids crystallize slightly faster than right-coiled, and thus
could have become more enriched in prebiotic environments than right-
coiled (Castelvecchi 2007). Either process would result in prebiotic envi-
ronments containing more left-coiled amino acids than right-coiled
amino acids as potential building blocks for proteins. Still, a chirality-
biasing process would not rule out the possibility that life constructed
with right-coiled amino acids could exist in the universe; it would simply
mean that such life forms would be less abundant in the universe than
life forms with left-coiled amino acids.
Modern molecular studies have revealed that a great deal of conver-
gent evolution of life here on Earth is in fact due to developmental
constraint: the phenomenon of “deep homology” that we considered in
chapter 1. “Studies of deep homology are showing that new structures
need not arise from scratch, genetically speaking, but can evolve by
deploying regulatory circuits that were first established in early animals.
. . . The more that researchers look, the more they will find that the same
258 Chapter 7
tools have been used to build a great variety of structures long thought
to have independent histories” (Shubin et al. 2009, 822). Would life else-
where in the universe, even if it is also carbon-based, coded for by base-
four DNA, and composed of proteins made of left-coiled amino acids,
have evolved the same developmental regulatory systems seen in Earth
life? If so, the form set { f:3} would not exist, but this at first seems
unlikely. The deep-homology regulatory circuits discussed by Shubin
et al. (2009) can be traced back to the Cambrian diversification of animal
life some 540 to 600 million years ago. Yet life existed on Earth 3.5 billion
years ago, and animals are a late addition to the history of life. It took
roughly 3 billion years of evolution for life to evolve the deep-homology
regulatory circuits present in animal life on Earth—what is the probabil-
ity that the very same regulatory system would convergently evolve with
the convergent evolution of multicellular, heterotrophic life elsewhere?
Multicellular, heterotrophic life forms—animals—that have evolved
with a different set of deep-homology regulatory circuits could poten-
tially be able to develop forms that cannot be developed by life on Earth,
forms that could belong to the set { f:3} and that would not be convergent
on Earth life.
Because we do not know the answer to the question of how develop-
ment originated in multicellular life forms on Earth, we also do not know
how development might originate in life forms elsewhere in the universe.
Theoretical modeling, however, suggests that a significant part of devel-
opmental constraint might indeed be a function of physical and geomet-
ric constraints (Newman et al. 2006; Müller 2007; Newman 2010), and
therefore that developmental systems in alien life might be similar to
those found in Earth life. In modeling the evolution of the development
of multicellular organisms, Newman et al. (2006) start with four different
kinds of physical and chemical patterning mechanisms: diffusion gradi-
ents, sedimentation gradients, reaction-diffusion mechanisms, and chemi-
cal oscillation mechanisms. Most importantly, these four patterning
mechanisms are found in nonliving as well as living chemical systems.
Then Newman et al. (2006) add two basic cell properties: differential
adhesion and cell polarity. Differential adhesion in pairs of tissues that
differ in cohesivity produces mixing behavior like that seen in immiscible
liquids, a behavior that is found in nonliving chemical systems and thus
not unique to life. Likewise, the acquisition of polarization in the adhe-
sion of cells is a mechanism not unique to life. Newman et al. (2006) next
explored the possible interactions of the four patterning mechanisms
with the two cell properties in a manner similar to that used in theoretical
morphology, producing a spectrum of hypothetical morphological out-

comes. In one spectrum, hollow spheres of tissue are produced when
diffusion gradients are combined with the cellular properties of differ-
ential adhesion and polarity in development, whereas invaginated
spheres are produced when sedimentation gradients are applied instead.
In the other spectrum, radially periodic tissue structures are produced
when reaction-diffusion mechanisms are combined with the cell proper-
ties, whereas serially periodic structures are produced when chemical
oscillation mechanisms are applied instead. Newman et al. (2006, 289)
point out that the spectrum of these hypothetical forms, produced by
physical and chemical patterning mechanisms that are not unique to life,
are very similar to “the hollow, multilayered and segmented morpho-
types seen in the gastrulation stage embryos of modern-day metazoa as
well as in Ediacaran fossil deposits of approximately ≈600 Ma.” In an
extended study, Newman (2010, 285) argues that nine “dynamical pat-
terning modules” (or DPMs) in particular exist within the spectrum of
hypothetical developmental forms, and that “the DPMs, in conjunction
with cell-type-defining and switching networks, transformed simple,
spherical, topologically solid cell clusters into hollow, multilayered, elon-
gated, segmented, folded, and appendage-bearing structures. They thus
founded the pathways that evolved into the developmental programs of
modern animals.”
In conclusion, Newman et al. (2006) and Newman (2010) suggest that
the evolution of development on Earth may have been a two-stage
process: metazoans originated from multicellular forms and structures
first assembled by predominantly physical mechanisms, and then subse-
quently evolved genetic mechanisms to perpetuate the functionally suc-
cessful morphologies formed in the first stage. The first step in this
process is subject to the laws of physics and geometry, which are assumed
to be same throughout the universe. Assembling tissues on alien worlds
should also form hollow spheres, invaginated spheres, radially periodic
structures, and serially periodic structures in the early evolution of mul-
ticellular alien life forms, just as on Earth.
If the early stages of developmental constraint are indeed a function
of physical and geometric constraints, then the set { f:3} might indeed be
empty—multicellular alien life forms may develop (at least initially)
in a very similar fashion to those found on Earth. However, the second
step in this hypothetical process—the evolution of genetic mechanisms
to perpetuate the functionally successful morphologies formed in the
first step—returns us to the question of the evolution of the coding
260 Chapter 7
mechanism of life that we have considered previously: How likely is it

that life elsewhere will evolve the DNA coding system present in Earth
life?
On Earth, almost all of the developmental-genetic “tool kit” genes
used by the multicellular metazoans (Unikonta: Opisthokonta: Choano-
zoa: Metazoa; see appendix) are found in the genome of the unicellular
choanoflagellate Monosiga brevicollis (Unikonta: Opisthokonta: Cho-
anozoa: Choanoflagellata); therefore, the metazoan tool-kit genes
predate the evolution of animals themselves (King et al. 2003). Only a
few additional tool-kit genes appear in the origin of the Metazoa (the
placozoans and sponges), and a few more in the evolution of the simplest
Eumetazoa (ctenophores and cnidarians), and then “all the triploblastic
metazoan body plans emerged within the space of no more than 20
million years” in the Cambrian explosion of animal evolution (Newman
2010, 284). Since they are determined by the laws of physics and geom-
etry, the multicellular DPMs should be the same for alien life, but could
alien life duplicate the feat of the Cambrian explosion in encoding the
DPMs into heritable body plans without the preexisting DNA coding
system of tool-kit genes? Or does it matter—regardless of the coding
system evolved by alien life, will that system not also have to accomplish
the exact same result in achieving multicellularity if that alien life is to
survive and evolve?
A Periodic Table of Life?
An alternative approach to analyzing the potential existence of the bio-

logical form set { f:3} elsewhere in the universe would be to try to visual-
ize what these functional forms that are nonexistent on Earth might look
like. One of the key features of theoretical morphology is the ability to
create both existent and nonexistent form by considering all the possible
permutations of form dimension parameters. We have previously seen in
chapter 4 that nonexistent ecological roles on Earth could be examined
by such a procedure (see table 4.12), resulting in the consideration of the
functional possibility of a nonexistent plant that floats in the air by means
of a gas-filled bladder or bladders, and in the conclusion that such a plant
would have to be carnivorous in addition to being photoautotrophic.
Analogous to the “periodic table of niches” (Pianka 1978) that we
considered in chapter 4, it is possible to create a “periodic table of life”
in a simple theoretical-morphology thought experiment (McGhee 2008).
The periodic table of elements allows chemists to predict not only the
behavior of existent elements but the behavior of nonexistent yet

possible elements, such as the heavy elements that chemists have
subsequently created in nuclear laboratories, elements that do not exist
in the natural state. In essence, the rows of the periodic table of elements
are based on the complexity of the atomic structure of the elements:
elements in the first row have only the electron shell K, elements in
the second row have the electron shells K and L, third-row elements
have electron shells K, L, and M, and so on. These rows also reflect the
evolutionary sequence of the appearance of the elements, with elements
in the first row (hydrogen and helium) appearing first in the evolution
of the universe, elements in the second row evolving next, and so on. We
can use the chemical concepts of elemental complexity and evolutionary
sequence in an analogous fashion by arranging the major groups of
multicellular life in a similar series of rows of morphological complexity
and biological evolutionary sequence (table 7.1).
The columns of the periodic table of elements can be considered to
characterize the mobility of the elements in those rows, with highly
mobile elements in some columns (elements that chemically combine
readily, such as the column containing hydrogen, lithium, sodium, etc.)
and low-mobility elements in other columns (elements that are chemi-
cally inert, such as the column containing helium, neon, argon, etc.). In
an analogous fashion, we can consider the mobility of multicellular forms
Table 7.1
A periodic table of life, based upon locomotory type and evolutionary sequence of
origination
Spectrum of
locomotion
None 2D locomotion 3D locomotion
Swimming
Sequence of Crawling Walking (fusiform Flying
evolution Sessile (legless) (legs) body) (wings)
Plants Plants — — — —
Invertebrates Barnacles Worms Arthropods Squid Insects
Amphibians — Caecilians Amphibians Tadpoles —
Reptiles — Snakes [reptiles] Ichthyosaurs Pterosaurs
Dinosaurs — — [dinosaurs] Penguins Birds
Mammals — — [mammals] Porpoises Bats
Note: Modified from McGhee (2008). See text for discussion.

262 Chapter 7
of life on the basis of locomotory type arranged in a series of columns

(table 7.1; I thank George Ellis of Capetown, South Africa, for suggesting
to me that I add the column of “no locomotion” to the original thought
experiment in McGhee 2008).
Even such a simple attempt to create a periodic table of life immedi-
ately reveals major incidences of convergent evolution (table 7.1). Not
only is the convergent evolution of fast-swimming fusiform morpholo-
gies in tetrapod vertebrates (reptilian ichthyosaurs, mammalian por-
poises, and dinosaurian penguins) apparent, but we also see that certain
invertebrates animals have also convergently evolved this same fast-
swimming morphology (most notably in modern-day squid and cuttlefish
cephalopods and their extinct orthoconic and belemnitellid relatives).
The major convergences in the evolution of wing structures for powered
flight in arthropods (insects), reptiles (pterosaurs), dinosaurs (birds), and
mammals (bats) are also apparent in the table.
Two major groups of animals have convergently evolved leg structures
for walking: the arthropods and the ancestral amphibians. (The tetrapod
reptiles, dinosaurs, and mammals are listed in brackets in the walking
column in table 7.1 because their legs are not independent convergences
but rather symplesiomorphic structures simply inherited from their
amphibian ancestors.) Note, however, that both the amphibians and the
reptiles have separately and independently reevolved legless morpholo-
gies (amphibian caecilians and reptilian snakes) and morphologically
have converged on annelid worms. All of the life forms listed in table 7.1
exist on the Earth and belong to the form set { f:1}.
Of major interest are the empty permutations in table 7.1—these are
potential candidates for functional life forms that are nonexistent on
Earth but could belong to the form set { f:3}. Currently, there are no flying
frogs on Earth, but we do have gliding frogs that are heading in that
evolutionary direction (see table 2.2). No legless, feathered avian-snake
forms nor furry mammalian-snake forms exist on Earth, though then
again weasels and ferrets—with their elongated bodies and small legs—
are headed in this evolutionary direction. These are also life forms that
could conceivably be developed by existent frogs, birds, and mammals,
and so is their absence on Earth due entirely to functional constraint,
and not to developmental constraint?
The major predicted life forms that are nonexistent on Earth revealed
in table 7.1 are mobile plants and sessile terrestrial animals. We have
previously considered the possibility of flying plants in chapter 4, and
concluded that the absence of such forms on Earth was probably due to
functional constraint (i.e., inability to deal with the habitat instability

produced by weather patterns existent on the Earth). That conclusion
might be wrong—could it be instead that Earth plants are unable to
develop the gas bladders that would be needed by a flying plant? The
closest thing we have in terrestrial ecosystems to the sessile animals that
exist in marine ecosystems are the web-spinning spiders. These spiders
are mobile, but when feeding, they simply sit sessile in the centers of their
webs and wait for prey animals to impact and stick to the web. Ecologi-
cally, they are like marine sessile carnivores such as corals. Is the absence
of truly sessile terrestrial predators due entirely to functional constraint
(i.e., inability to obtain sufficient prey by an immobile carnivore)?
Writers of science fiction have been engaging in theoretical morphol-
ogy for years, although they are unaware of it. Hypothetical walking
plants, the triffids, were created by the science fiction writer John
Wyndham in his novel The Day of the Triffids (1951). These plants grew
normally, using roots for nutrient uptake and leaves for photosynthesis,
until they reached maturity, at which time they uprooted and used three
specialized root prongs for clumsy walking. The springlike root-prong
mechanism used by the triffids for walking is not unlike the springlike
trap mechanisms present in Venus fly traps. Interestingly, Wyndham con-
cluded that his mature triffids would have to be carnivorous, just as we
concluded in chapter 4 that flying plants would have to use carnivory for
nutrient acquisition. Larry Niven (1968a) created hypothetical flying
plants, called stage trees, that used rocket propulsion for seed dispersal
in his short story “A Relic of the Empire.” These plants grew as low
shrubs or bushes until their reproductive phase, when they grew tall tree
trunks enriched with organic explosives. During the dry season, forest
fires would ignite the stage trees, which would lift off like rockets and
explode in the skies, scattering their seeds over huge geographic dis-
tances. Niven also created sessile mammal-like terrestrial predators, the
grogs, in his short story “The Handicapped” (1968). These hypothetical
animals evolved a unique adaptation to lure prey animals within striking
distance of the sessile predator’s long, prehensile tongue (I will not be a
spoiler and reveal what the adaptation is). If they existed, triffids, stage
trees, and grogs would fill the empty permutations of walking plants,
flying plants, and sessile mammals in table 7.1, and they certainly are
developmentally impossible for life forms on Earth. Are they candidates
for the form set { f:3}?
In summary, using the techniques of theoretical morphology, we can
visualize possible, but nonexistent, biological forms. But we do not know
264 Chapter 7
if the nonexistence of those forms is simply due to functional constraint

(that is, that those forms belong to form set { f:2}), or if the nonexistence
of those forms is due to developmental constraint (that they belong to
form set { f:3}). The possibility that life elsewhere, evolving with different
developmental constraints, might be able to develop functional forms
that cannot be developed by life on Earth limits our ability to predict
the outcome of evolution with certainty. Only when we fully understand
the evolution of the developmental regulatory system present in Earth
life will we able to make predictions about possible alien developmental
systems.
A strong case has been made in this chapter, however, for the conver-
gent evolution of similar developmental constraints in alien life forms
that are carbon-based and that inhabit Earth-type worlds. That is, I
suggest that developmental constraints in this restricted context are a
function of physical and geometric constraints, just as functional con-
straints are. If this is true, we should expect radically different develop-
mental constraints to have been evolved only in life forms that are not
carbon-based, or that inhabit non-Earth-type planets that are either very
hot or very cold, or are gas giants. In our own solar system, Titan is
perhaps the best possible candidate for exploration for life forms unlike
any seen on Earth.
8 Philosophical Implications of Convergent Evolution
Complexity theory also suggests a new take on an old question, long a staple of
science fiction and speculative science: When we do find aliens, or they find us,
what will they look like? By revealing many forms of Earth life to be governed
by deep geometrical rules of self-organization in nature, complexity suggests a
universal geometry of life that should transcend worlds. . . . There is no way to
predict precisely what aliens will look like, but the fractal geometry of life gives
us reason to believe that when they do finally land on the White House lawn,
whatever walks or slithers down the gangplank may look strangely familiar.
—Grinspoon (2003, 272–273)
The laws of physics, it is believed, are the same everywhere in the universe. This
is unlikely to be true of biology.
—Crick (1988, 138)
Most scientists tend to ignore philosophers. In evolutionary biology, the

scientist Ernst Mayr (1964, xi–xii) traced this tendency back to Charles
Darwin himself: “No one resented Darwin’s independence of thought
more than the philosophers. How could anyone dare to change our
concept of the universe and man’s position in it without arguing for or
against Plato, for or against Descartes, for or against Kant? Darwin had
violated all the rules of the game by placing his argument entirely outside
the traditional framework of classical philosophical concepts and termi-
nologies. . . . No other work advertised to the world the emancipation of
science from philosophy as blatantly as did Darwin’s Origin.” Among
other things, philosophers concern themselves with questions about
purpose and meaning in life, questions that are unanswerable using the
materialistic methodology of science. Because metaphysical questions
are unanswerable by science, many scientists consider such questions to
be meaningless.
I am a scientist, yet I also find philosophy to be interesting. A scientific
commitment to a materialistic methodology does not mean one also has
266 Chapter 8
to commit oneself to a materialistic philosophy, although many scientists

do. In this chapter I will touch on philosophical questions concerning
freedom, purpose, design, destiny, teleology, spirituality, and even God—
not usual topics for a scientist. I do so because I am not alone—these
questions have been raised by other scientists as well, with reference to
a very specific natural phenomenon: convergent evolution.
The difference of opinion between the astrobiologist David Grinspoon
and Francis Crick, co-discoverer of the molecular structure of DNA,
given in the two epigraphs at the beginning of this chapter, sets the stage
for the spectrum of topics to be considered here. In the chapter I will
follow the model of discussion used by John Casti in his philosophical-
scientific book Paradigms Lost (1989), in which arguments for and
against a point of view are presented as if in a trial in a courtroom, fol-
lowed by a ruling from the judge on the bench. I, of course, shall act as
judge and, of course, the reader is free to disagree with the bench!
The Argument for Unpredictability: Creative Freedom or Chaotic Randomness?
The best-known evolutionary essayist of the twentieth century, the

Harvard paleontologist Stephen Jay Gould, was fond of a thought experi-
ment of his own that he called “replaying life’s tape” (Gould 1989). That
is, consider the history of the evolution of life on Earth to be similar to
a videotape of a popular movie. Imagine what would happen if you could
take a copy of the videotape and rewind it to a point early in the movie,
erasing everything on the tape that happened after that point, and then
could rerun the tape to see what would happen this time on the now
blank tape. Would the historical sequence of events in the evolution of
life in the rerun of the tape resemble the original? Or would evolution
take radically different pathways in the new narrative, producing animal
and plant forms totally unlike those of the original? Gould (1989, 51)
argues strongly for the second scenario: “Any replay of the tape would
lead evolution down a pathway radically different from the road actually
taken. . . . The diversity of possible itineraries does demonstrate that
eventual results cannot be predicted at the outset. Each step proceeds
for cause, but no finale can be specified at the start, and none would ever
occur a second time in the same way, because any pathway proceeds
through thousands of improbable stages.”
In short, in Gould’s view, evolution is totally unpredictable: evolution
has no predictable direction and no predictable destination. An evolu-
tionary trend in time is solely a chain of contingent historical events, and
Philosophical Implications of Convergent Evolution 267
could not have been deduced from any laws of nature: “Contingency is
the affirmation of control by immediate events over destiny. . . . Our own
evolution is a joy and a wonder because such a curious chain of events
would probably never happen again, but having occurred, makes eminent
sense” (Gould 1989, 284–285).
Some evolutionary biologists find Gould’s view of evolution as an
unpredictable, nonrepeatable, and historically contingent process to be
liberating and uplifting. For example, Stuart Kauffman (2008, 5) com-
ments on Gould’s view that the path of evolution cannot be deduced
from any laws of nature: “a ‘natural law’ is a compact description before-
hand of the regularities of a process. But if we cannot even prestate the
possibilities, then no compact descriptions of these processes beforehand
can exist. These phenomena, then, appear to be partially beyond natural
law itself. This means something astonishingly and powerfully liberating.
We live in a universe, biosphere, and human culture that are not only
emergent but radically creative. We live in a world whose unfoldings we
often cannot prevision, prestate, or predict—a world of explosive creativ-
ity on all sides.”
Similarly, in writing of Gould’s view that evolution could not be
repeated, Kauffman (2008, 130) expands on this concept with obvious
joy: “The vast nonrepeatability, or nonergodicity, of the universe at all
levels of complexity above atoms—molecules, species, human history—
leaves room for a creativity in the way the universe unfolds at these
levels, a creativity that we cannot predict. . . . Now I want to make my
outrageous claim: the evolution of the biosphere is radically nonpredict-
able and ceaselessly creative.”
And last, as I promised at the beginning of this chapter, the concepts
of spirituality and of God also arise from this consideration of evolution
as an unpredictably creative process: “I believe we can reinvent
the sacred. We can invent a global ethic, in a shared space, safe to all
of us, with one view of God as the natural creativity in the universe”
(Kauffman 2008, xiii).
Other biologists welcome Gould’s view of evolution as the ultimate
dismissal of the older idea of teleology, that evolution is design-like and
progressive, and thus has a direction and ultimately a destination. But
Reiss (2009, 174) goes beyond Gould to attack the very idea of adaptive
improvement in evolution as being teleological, as in his discussion of
Sewall Wright’s concept of the adaptive landscape and Darwin’s concept
of natural selection: “Wright seems never to have realized that while in
the context of domestic breeds it may be reasonable to speak of ‘improve-
268 Chapter 8
ment’ by selection of ‘superior’ herds, where the standard of value is

determined by the goals of the breeder, no such a priori standard of
overall value exists in nature. In fact, by focusing on the hypothetical
ways in which organisms could evolve toward a future state of improved
adaptation, Wright introduced a teleological element into his evolution-
ary theory, just as Darwin had into his.” He further argues, “Darwin
introduced a teleological determinism into the heart of his theory. This
teleology is expressed in two related conceptions: (1) that evolution is a
process going from a less-adapted to a better-adapted state and (2) that
natural selection is a deterministic force, or agent, that directs the evo-
lutionary process toward this better-adapted state” (Reiss 2009, 140).
Rather than leading to thoughts of spirituality and God, as in the case
of Kauffman (2008), the view that evolution is unpredictable and direc-
tionless produces a totally different reaction in Reiss (2009, 356): “In
short, when we recognize that organisms exist only by virtue of the fact
that they are satisfying their conditions of existence, and that features of
organisms likewise exist because they are satisfying their conditions for
existence, we are freed to consider evolution in a dispassionate manner.
Life is not designed, or at least it shows no evidence of design for any-
thing other than continued existence, which needs no designer.”
Gould (1989, 290) himself was well aware that his stance on evolution
was partially a reaction to “traditionalist” and teleological ideas of evolu-
tion: “ultimately, the question of questions boils down to the placement
of the boundary between predictability under invariant law and the
multifarious possibilities of historical contingency. Traditionalists like
[Charles Doolittle] Walcott would place the boundary so low that all
major patterns of life’s history fall above the line into the realm of pre-
dictability (and, for him, direct manifestation of divine intentions). But
I envision a boundary sitting so high that almost every interesting event
in life’s history falls into the realm of contingency.”
As one might expect, reaction to Gould’s argument from those scien-
tists whose view of evolution he characterized as “traditional” has not
been positive; Gould’s idea that evolution is unpredictable, contingent,
nonrepeating, and directionless has been viewed with horror. Rather
than agreeing with Kauffman (2008, 5), who felt Gould’s view of evolu-
tion is “liberating” and “radically creative,” Gould’s critics see an abyss
of chaotic randomness and meaninglessness: “To say that any of this is
in the broadest sense ‘adaptive’ simply begs the question of how the
world came to be ordered in the first place. It is, of course, no accident
that the incidental, the chance occurrence, the contingent happenstance,
is so influential in our deracinated and nihilistic culture, especially as

reflected in the biological sciences that have spent the last century trying
to square the circle of a meaningless process, that is, evolution, leading
to the appearance of a sentient species that sees meaning all around
itself” (Conway Morris 2008, 61).
The Argument for Predictability: Comforting Certainty or Depressing

Inevitability?
Surprisingly, some of the strongest statements for evolution as a predict-

able, law-driven process come from the molecular biologists: “Whether
or not there are other sets of lawful organic forms, there is no doubt that
the universe of protein folds represents a Platonic universe . . . a universe
where abstract rules, like the rules of grammar, define a set of unique
immaterial templates which are materialized into a thousand or so
natural forms—a world of rational morphology and pre-ordained evolu-
tionary paths. . . . [A]s far as the 1000 protein folds are concerned, we
may be sure that they will be present everywhere in the cosmos where
there is carbon-based life, utilizing the same 20 protogenic amino acids”
(Denton et al. 2002, 340). These authors further claim: “For the lawful
nature of the folds provides for the first time evidence that the laws of
nature may not only be fine-tuned to generate an environment fit for life
(the stage) but may also be fine-tuned to generate the organic forms (the
actors) as well, in other words that the cosmos may be even more bio-
centric than is currently envisaged . . . raising the possibility that all
organic forms and indeed the whole pattern of life may finally prove to
be the determined end of physics and life a necessary feature of the
fundamental order of nature” (338).
Denton et al.’s view of an evolutionary process with “pre-ordained
evolutionary paths” and a lawlike behavior such that “the whole pattern
of life may finally prove to be the determined end of physics” is the very
antithesis of the view of Gould (1999) and Kauffman (2008). Yet it is
mirrored by the statements of other molecular biologists: “Darwinian
evolution can follow only very few mutational paths to fitter proteins.
. . . [W]e conclude that much protein evolution will be similarly con-
strained. This implies that the protein tape of life may be largely repro-
ducible and even predictable” (Weinreich et al. 2006, 111); and: “Our
results show that, despite the dominance of contingency (historical acci-
dent) in some recent discussions of evolutionary mechanisms (Gould
1989), purely chemical constraints (that is, the ability of only certain
270 Chapter 8
sequences to carry out particular functions) can lead to the repeated

evolution of the same macromolecular structures” (Salehi-Ashtiani and
Szostak 2001, 84).
On the macroscopic scale of biological form, the paleontologist Simon
Conway Morris (2003, 309) argues that most of the potential hyperspace
of form is empty, and that the pathways through that hyperspace have
been repeatedly traveled in evolution due to evolutionary constraint:
“The phenomenon of evolutionary convergence indicates that . . . the
number of alternatives is strictly limited, with the interesting implication
that the vast bulk of any given ‘hyperspace’ not only never will be visited
during evolutionary exploration but it never can be. These are the
howling wildernesses of the maladaptive, the 99.9% recurring of biologi-
cal space where things don’t work, the Empty Quarters of biological
non-existence.” That is, life is constrained to evolve in only a tiny fraction,
0.1 percent, of the potential form hyperspace, and it is this constraint that
produces convergent evolution.
Conway Morris (2003, 309–310) further argues that the available evo-
lutionary pathways in the hyperspace, and the nodes within the hyper-
space where functional forms can exist, have been preset from the very
beginning of the universe: “It is my suspicion that . . . the nodes of occu-
pation are effectively predetermined from the Big Bang.” The idea that
evolution must follow predetermined pathways, set by the constants and
laws of nature from the very beginning of the universe, leads to the idea
that evolution in these predetermined pathways is not only predictable
but also inevitable: “The principal aim of this book has been to show that
the constraints of evolution and the ubiquity of convergence make the
emergence of something like ourselves a near-inevitability” (Conway
Morris 2003, 328).
Is evolution predictable, preordained, and inevitable? If so, it should
have not only a direction but also an ultimate destination. This was cer-
tainly the view of another paleontologist, Pierre Teilhard de Chardin, as
described by Aczel (2007, 75–76): “the ideas of evolution became so
powerful that they convinced him that everything in the universe, inani-
mate objects and living systems alike, was in constant flux, ever evolving
as decreed by God. The goal was a point where everything would con-
verge to form the body of Christ. This was Teilhard’s Omega Point.”
Teilhard was religious, a Jesuit priest, and it is notable that both Conway
Morris’s Life’s Solution (2003) and Barlow’s Let There Be Sight! A Cel-
ebration of Convergent Evolution (2005) have a religious component to
their argumentation.
Other scientists have also come to the conclusion that evolution is

predictable and lawlike, but from a materialist rather than religious
viewpoint, such as the Russian zoologist Leo Berg in his book Nomo-
genesis: Or Evolution Determined by Law (1922). From an admittedly
materialist viewpoint, the astrobiologist David Grinspoon (2003, 412)
also writes of predictable, goal-directed evolution: “what I mean by spiri-
tuality is the religious impulse stripped of religion. . . . I believe the
phenomenon of humanity on Earth is a local example of a trend toward
higher consciousness and spiritual enlightenment that transpires all over
this universe. . . . Cosmic spiritual advancement by Darwinian natural
selection!” Kauffman (2008) was able to find spirituality in the view that
evolution was unpredictable, contingent, nonrepeating, and directionless,
as discussed in the previous section of this chapter. Thus, it is very inter-
esting that spirituality has also been found in the exact opposite view
that evolution is predictable, preordained, and inevitable.
It is clear that many who believe that evolution is predictable, preor-
dained, and inevitable find comfort in the lawlike certainty of this point
of view. Others react with horror to this point of view, and find it depress-
ingly deterministic, stultifying rather than creative, enslaving rather than
liberating. Where is individual freedom, where is creativity, in an evolu-
tionary process that already has a predetermined destination?
Judging the Arguments: Evolutionary Views on Trial
We have now reviewed the arguments for the view that evolution is
unpredictable, contingent, nonrepeating, and directionless and those for
the view that evolution is predictable, predetermined, and inevitable.
What are the implications of the phenomenon of convergent evolution,
the subject of this book, with regards to these radically different views
of evolution?
First, the view that the evolutionary process is nonrepeating (noner-
gotic) is demonstrably false. Chapters 2 through 6 of this book have
demonstrated that evolution is a highly repeatable process. In fact,
repeatability in evolution is rampant at every level of life on Earth, from
tiny organic molecules to entire ecosystems of species, and even to the
ways in which we think.
Second, the view that evolution is entirely historically contingent, and
thus unpredictable (and nonrepeating), is demonstrably false. Conver-
gent evolution—the repeated, independent, evolution of the same trait
in multiple phylogenetic lineages at different points in geological time—
272 Chapter 8
reveals to us the limits imposed upon evolution by functional constraints.

Once we know what those limits are, we can use them to predict the
trajectory of future evolution by predicting that those constraints will
continue to operate in the future, just as they have in the past. Given an
Earth-like world elsewhere in the universe, we can also predict that the
same functional constraints that exist on Earth will be in operation on
that world as well. Each step in the evolution of a Jurassic ichthyosaur
and a Cenozoic porpoise has its unique contingent aspects, yet the end
result was the same: a fusiform body, the same body form that is found
in a shark or a swordfish (see table 2.1). If any large, fast-swimming
organisms exist in the oceans of Jupiter’s moon Europa, swimming under
the perpetual ice that covers their world, I predict with confidence that
they will have streamlined, fusiform bodies; that is, they will look very
similar to a porpoise, an ichthyosaur, a swordfish, or a shark.
Third, the view that evolution is directionless needs more careful
examination. Reiss’s view that there is no adaptive improvement in evo-
lution is demonstrably false. Reiss (2009, 140) objects to the idea that
“evolution is a process going from a less-adapted to a better-adapted
state” under the influence of the “deterministic force” of natural selec-
tion because he considers that idea to be teleological. Is it teleological
that water flows downhill? That it goes from a state of higher potential
energy to a state of lower potential energy under the influence of the
deterministic force of gravity? Philosophers may argue about whether
that phenomenon is teleological or not, but in science it is an empirical
observation, an established fact.
It is an empirical observation that adaptive improvement occurs in
much of biological evolution. In fact, the adaptive landscape concept of
Sewall Wright (which Reiss objects to as teleological) can be used to
analyze actual cases of adaptive improvement in biological evolution
(McGhee 2007). However, there does exist one model of evolution that
corresponds to Reiss’s view: the Red Queen hypothesis (Van Valen
1973), named after the Red Queen in Lewis Carroll’s Alice’s Adventures
in Wonderland (1865), who told Alice that she had to constantly keep
running just in order to stay in the same place. In such a scenario, a
species can only evolve fast enough to exactly maintain its same position
on a moving adaptive peak; that is, it evolves fast enough to avoid sliding
downslope with time (and thus to eventual extinction), yet it cannot
evolve fast enough to climb upslope on the peak to higher levels of
adaptation (McGhee 2007, 27–28). The species is constantly evolving, but
perpetually stuck at the same degree of adaptation. The Red Queen
hypothesis is a rather bleak evolutionary possibility, in that no adaptive

improvement ever takes place, and the probability of the species going
extinct is always a constant regardless of whether the species is young,
in existence only a few hundred years, or has been present on the Earth
for millions of years.
Reiss (2009, 140) further objects to the idea that natural selection is a
deterministic force that directs the evolutionary process toward a better
adapted state because that idea appears to be goal directed, or teleologi-
cal. Is it teleological that water flows in rivers toward the future “goal”
of reaching a sea or lake? The flow of water is mindless, and water has
no goal in sight, but nevertheless it will reach its lowest possible potential
energy state under the deterministic force of gravity, which means that
it will wind up in a sea or lake.
Part of Reiss’s goal-directed objection stems from the fact that both
Darwin and Wright used the analogy of human artificial selection in
trying to explain the process of nonhuman natural selection. As quoted
earlier, Reiss complains: “Wright seems never to have realized that while
in the context of domestic breeds it may be reasonable to speak of
‘improvement’ by selection of ‘superior’ herds, where the standard of
value is determined by the goals of the breeder, no such a priori standard
of overall value exists in nature” (Reiss 2009, 174). That is, Reiss argues
that a human breeder may have a standard value or goal that he or she
is trying to achieve in artificial selection, but that no such standard value
or goal exists in nature. This argument is demonstrably false, because the
phenomenon of convergent evolution abundantly demonstrates that the
environment does have a priori standards of overall value. The laws of
physics impose the functional constraints, the a priori standards, that
fast-swimming organisms must be fusiform in shape (see table 2.1) and
that flying organisms must have wings (see table 2.2). And so on, for all
the myriad examples of convergent evolution that have been considered
in chapters 2 through 6 of this book.
However, there is a real difference between the a priori standards of
human artificial selection and nonhuman natural selection: the human
mind. The a priori standards of nature are mindless functional constraints
imposed by the laws of physics and geometry. It is part of Darwin’s genius
that he realized that the mindless a priori standards of nature would sort
out organic variation in a process of mindless natural selection (for a
book-length discussion of modern misconceptions regarding teleology
and adaptation, see Ruse 2003). Natural selection has a direction only in
the sense that it will, in general, operate to move evolving organisms up
274 Chapter 8
the slopes of the adaptive landscape to higher states of adaptation

(McGhee 2007). It is the empirical observation that the number of these
higher states of adaptation, or adaptive peaks, is limited that gives evolu-
tion its direction. If the number of adaptive peaks were infinite, then each
species on Earth would be morphologically different from every other
species, and each species would have its own unique ecological role, or
niche. Such an Earth does not exist. Instead, repeated evolutionary con-
vergence on similar morphologies, niches, molecules, and even mental
states is the norm on Earth.
Fourth: while evolution is not directionless, it has not been proved that
the constraint on evolutionary pathways available for life to evolve along
was predetermined from the beginning of the universe. It is certainly a
possibility, as Einstein himself wondered if there was any freedom left
for alternative ways for the physical universe to evolve once the con-
stants and laws of nature were established at the Big Bang, and the same
could be true for biological evolution, as discussed in chapter 7. Thus,
one could indeed argue that functional constraints, which are themselves
a function of the laws of physics and geometry, have been invariant since
the beginning of the universe if those same laws of physics and geometry
have been.
It is the role of developmental constraint in convergent evolution that
leads to uncertainty in the argument for predetermined constraints.
Although Stern (2010) has argued for a “predictable genome” and
Brakefield (2006, 364) notes that “patterns of parallel morphological
evolution given similar ecological opportunities are to be expected and
that, given sufficient knowledge of developmental processes, they could
be predictable,” this possible predictability refers solely to the develop-
mental regulatory processes evolved by Earth life. Would life elsewhere
in the universe evolve the same developmental constraints? How much
of developmental constraint is lawlike and predictable, a function of
the laws of physics and geometry at the level of molecules, cells, and
tissues (Newman et al. 2006; Müller 2007; see also chapter 5), and how
much is it a function of chance in the evolution of life specific to the
Earth, and unpredictable? As we saw in chapter 7, the possibility that
life elsewhere, evolving with different developmental constraints,
might be able to develop functional forms that cannot be developed by
life on Earth limits our ability to predict the outcome of evolution with
certainty.
Fifth: it has not been proved that biological evolution is inevitable, that
it has a destination. And even if it did, it is not clear how we might deter-
mine what that destination might be. Even our models for the physical
evolution of the universe have changed radically over the past century:
once it was thought the universe was steady-state, and would infinitely
persist. Then it was thought that the universe would eventually cease to
expand, and then collapse upon itself in a sort of reverse-Big-Bang fire-
ball. Now it is generally thought the universe will continue to expand
infinitely and that all the stars will eventually burn out, one by one, in a
cold, dark universal death. But, as we currently still do not truly under-
stand dark matter or dark energy, that model of the ultimate destination
of the physical evolution of the universe may change in the future as
well. As yet, we have no equivalent models for the ultimate destination
of biological evolution. If we cannot, at present, confidently predict the
ultimate destination of the physical evolution of the universe, how can
we hope to predict the ultimate destination of the vastly more complex
process of biological evolution?
There are hints that such a destination might exist: the increase in
complexity produced by evolution in the past 3.5 billion years of Earth
history, and the evolution of self-awareness in the past 5 to 10 million
years in multiple independent lineages (see table 6.7). The universe has
become aware of itself (Grinspoon 2003). Is such a trend intrinsic to the
process of biological evolution? The self-assembling properties of bio-
logical evolution, and its constant building upon previous evolutionary
steps via the process of natural selection, would lead us to predict that
such a process should produce increasing complexity in other life forms
as well, on other worlds.
Have life forms on other worlds become self-aware? Has matter awak-
ened, has the universe become aware of itself elsewhere? We have no
evidence that this has occurred, although we keep searching the skies for
possible electromagnetic or other signals that would point to the exis-
tence of extraterrestrial civilizations of self-aware organisms in search of
other self-aware organisms. On Earth, the convergent evolution of self-
awareness has been linked to the convergent evolution of complex social
systems, as discussed in chapter 6. The fact that, on Earth, life in the sea
and life on land—life in radically different environments—both have
evolved convergent social systems and self-awareness would lead us to
predict that self-awareness would evolve on extraterrestrial worlds as
well.
Or is self-awareness an intrinsic property of the developmental regula-
tory systems present in animal life on Earth? Would life elsewhere,
evolving under different developmental constraints, have evolved social
276 Chapter 8
systems similar to those convergently evolved by Earth life, and hence

self-awareness?
The Question of Extraterrestrial Life
The phenomenon of convergent evolution demonstrates that life repeat-

edly evolves in a finite number of preferred directions and, as such, is in
principle predictable. Analysis of that phenomenon reveals that it is
driven by functional and developmental constraints on evolution.
However, we have only one example of that phenomenon in operation:
Earth life. Thus, given the existence of an extraterrestrial planet with
Earth-like physical conditions, we can predict that the same functional
constraints that have shaped the evolution of life on Earth will operate
in a similar fashion on that extraterrestrial planet. Life on that planet
will look “strangely familiar,” as David Grinspoon (2003) argues in the
epigraph at the beginning of this chapter.
Francis Crick (1988) may also be right, as well. Convergent evolution
of life forms on Earth has also been driven by developmental constraints.
Totally different life forms, such as silicon-based life, would undoubtedly
evolve with developmental constraints different from those that operate
in life on Earth. The “laws of biology” would be different for those life
forms, and they might evolve forms that are totally unlike those seen on
Earth.
The question of the existence of extraterrestrial life is of crucial impor-
tance to the question of the predictability of evolution. Short of intelli-
gent extraterrestrials contacting us and discussing their biochemistry
with us, we must search for such life ourselves. Within our own solar
system (since, at present, we can only dream of spacecraft that could take
us to other star systems), if life is present on Mars or even in the water
oceans of Jupiter’s moon Europa, it is likely to be convergent on the
carbon-based model found on Earth. Only the far-distant planet-like
moons Titan and Triton offer physical conditions and chemistries radi-
cally different from those found on Earth, and the possibility of discover-
ing life “as we do not know it.”
In Conclusion: A Rewrite of Darwin’s View of Life
Darwin (1859, 490), in a philosophical mood, closed the first edition of

On the Origin of Species with the sentence: “There is grandeur in this
view of life, with its several powers, having been originally breathed into
a few forms or into one; and that, whilst this planet has gone cycling on
according to the fixed law of gravity, from so simple a beginning endless
forms most beautiful and most wonderful have been, and are being,
evolved.”
We now have the benefit of an additional 150 years of analysis of
biological evolution. The phenomenon of convergent evolution leads us
to reconsider Darwin’s closing sentence, at least for carbon-based life,
and to suggest that it be rewritten (with the modern alterations in the
original sentence indicated by italics): “There is grandeur in this view of
life in the universe, with its several powers of functional and developmen-
tal constraint, having been originally breathed into a few forms of life or
into just the one that is carbon based; and that, whilst this planet and
others have gone cycling on according to the fixed law of gravity, from
so simple a beginning limited forms most beautiful and most wonderful
have been, and are being, re-evolved throughout the universe.”
Appendix: A Phylogenetic Classification of Life
Table A.1
A phylogenetic classification of life forms discussed in this book
Eubacteria
Archaea
Eukarya
– Bikonta
– – Excavobionta
– – Rhizaria
– – Chromoalveolata
– – Green eukaryotes
– – – Glaucophyta
– – – Metabionta
– – – – Rhodobionta
– – – – Chlorobiota > see Table A.3
– Unikonta
– – Amoebozoa
– – Opisthokonta
– – – Fungi
– – – Choanozoa
– – – – Choanoflagellata
– – – – Metazoa
– – – – – Placozoa
– – – – – Demospongiae
– – – – – Hexactinellida
– – – – – Calcarea
– – – – – Eumetazoa
– – – – – – Ctenophora
– – – – – – Cnidaria
– – – – – – Bilateria
– – – – – – – Protostomia
280 Appendix
Table A.1
(continued)
– – – – – – – – Lophotrochozoa
– – – – – – – – – Lophophorata
– – – – – – – – – – Bryozoa
– – – – – – – – – – Phoronozoa
– – – – – – – – – – – Phoronida
– – – – – – – – – – – Brachiopoda
– – – – – – – – – Eutrochozoa
– – – – – – – – – – Syndermata
– – – – – – – – – – – Rotifera
– – – – – – – – – – – Acanthocephala
– – – – – – – – – – Spiralia
– – – – – – – – – – – Entoprocta
– – – – – – – – – – – Annelida
– – – – – – – – – – – Sipuncula
– – – – – – – – – – – Mollusca
– – – – – – – – – – – Parenchymia
– – – – – – – – – – – – Platyhelminthes
– – – – – – – – – – – – Nemertea
– – – – – – – – Cuticulata
– – – – – – – – – Gastrotricha
– – – – – – – – – Ecdysozoa
– – – – – – – – – – Introverta
– – – – – – – – – – – Nematozoa
– – – – – – – – – – – – Nematoda
– – – – – – – – – – – – Nematomorpha
– – – – – – – – – – – Cephalorhyncha
– – – – – – – – – – – – Kinorhyncha
– – – – – – – – – – – – Loricifera
– – – – – – – – – – – – Priapulida
– – – – – – – – – – Panarthropoda
– – – – – – – – – – – Onychophora
– – – – – – – – – – – Tardigrada
– – – – – – – – – – – Arthropoda
– – – – – – – – – – – – Cheliceriformes
– – – – – – – – – – – – Mandibulata
– – – – – – – – Chaetognatha
– – – – – – – Deuterostomia
– – – – – – – – Echinodermata
– – – – – – – – Pharyngotremata
– – – – – – – – – Hemichordata
A Phylogenetic Classification of Life 281
Table A.1
(continued)
– – – – – – – – – Chordata
– – – – – – – – – – Urochordata
– – – – – – – – – – Myomerozoa
– – – – – – – – – – – Cephalochordata
– – – – – – – – – – – Craniata
– – – – – – – – – – – – †Conodonta
– – – – – – – – – – – – Myxinoidea
– – – – – – – – – – – – Vertebrata > see Table A.2
Note: Major clades used frequently in the text for comparative purposes are in boldface
type. Extinct taxa are marked with a †. Classification modified from Lecointre and Le
Guyader (2006) and Benton (2005).
Table A.2
A phylogenetic classification of vertebrate taxa discussed in this book
Vertebrata
– Petromyzontiformes
– †Pteraspidomorphi
– †Osteostraci
– Gnathostomata
– – †Placodermi
– – Chondrichthyes
– – †Acanthodii
– – Osteichthyes
– – – Actinopterygii
– – – Sarcopterygii
– – – – Dipnoi
– – – – Crossopterygii
– – – – – Actinistia
– – – – Tetrapodomorpha
– – – – – Tetrapoda
– – – – – – Batrachomorpha
– – – – – – – †Dendrerpetontidae
– – – – – – – †Dissorophidae
– – – – – – – Lissamphibia
– – – – – – – – Gymnophiona
– – – – – – – – Batrachia
– – – – – – – – – Anura
– – – – – – – – – Urodela
– – – – – – †Lepospondyli
282 Appendix
Table A.2
(continued)
– – – – – – Reptiliomorpha
– – – – – – – †Seymouriamorpha
– – – – – – – †Diadectomorpha
– – – – – – – Amniota
– – – – – – – – Sauropsida
– – – – – – – – – Anapsida
– – – – – – – – – – Testudines
– – – – – – – – – Diapsida
– – – – – – – – – – †Ichthyosauria
– – – – – – – – – – Lepidosauromorpha
– – – – – – – – – – – †Sauropterygia
– – – – – – – – – – – Lepidosauria
– – – – – – – – – – – – Sphenodontia
– – – – – – – – – – – – Squamata
– – – – – – – – – – – – – Iguania
– – – – – – – – – – – – – Scleroglossa
– – – – – – – – – – – – – – Amphisbaenia
– – – – – – – – – – – – – – Gekkonta
– – – – – – – – – – – – – – Autarchoglossa
– – – – – – – – – – – – – – – Scincomorpha
– – – – – – – – – – – – – – – Anguimorpha
– – – – – – – – – – – – – – – – †Mosasauridae
– – – – – – – – – – – – – – – – Serpentes
– – – – – – – – – – Archosauromorpha
– – – – – – – – – – – Archosauria
– – – – – – – – – – – – Crurotarsi
– – – – – – – – – – – – Ornithodira
– – – – – – – – – – – – – †Pterosauria
– – – – – – – – – – – – – Dinosauria
– – – – – – – – – – – – – – Saurischia
– – – – – – – – – – – – – – – †Sauropodomorpha
– – – – – – – – – – – – – – – Theropoda
– – – – – – – – – – – – – – – – †Ceratosauria
– – – – – – – – – – – – – – – – Tetanurae
– – – – – – – – – – – – – – – – – †Carnosauria
– – – – – – – – – – – – – – – – – Coelurosauria
– – – – – – – – – – – – – – – – – – Maniraptora
– – – – – – – – – – – – – – – – – – – Aves
– – – – – – – – – – – – – – – – – – – – Paleognathae
– – – – – – – – – – – – – – – – – – – – Neognathae
– – – – – – – – – – – – – – – – – – – – – Galloanserae
– – – – – – – – – – – – – – – – – – – – – Neoaves
Table A.2
(continued)
– – – – – – – – – – – – – – †Ornithischia
– – – – – – – – – – – – – – – †Fabrosauridae
– – – – – – – – – – – – – – – †Genasauria
– – – – – – – – – – – – – – – – †Thyreophora
– – – – – – – – – – – – – – – – †Cerapoda
– – – – – – – – – – – – – – – – – †Ornithopoda
– – – – – – – – – – – – – – – – – †Marginocephalia
– – – – – – – – Synapsida
– – – – – – – – – †Ophiacodontidae
– – – – – – – – – †Edaphosauridae
– – – – – – – – – †Sphenacodontidae
– – – – – – – – – Therapsida
– – – – – – – – – – †Dinocephalia
– – – – – – – – – – †Dicynodontia
– – – – – – – – – – †Gorgonopsia
– – – – – – – – – – Cynodontia
– – – – – – – – – – – Mammalia
– – – – – – – – – – – – †Volaticotheria
– – – – – – – – – – – – †Yinotheria
– – – – – – – – – – – – Australosphenida
– – – – – – – – – – – – – †Ausktribosphenida
– – – – – – – – – – – – – Monotremata
– – – – – – – – – – – – †Multituberculata
– – – – – – – – – – – – Theria
– – – – – – – – – – – – – Marsupialia
– – – – – – – – – – – – – Eutheria
– – – – – – – – – – – – – – Xenarthra
– – – – – – – – – – – – – – Epitheria
– – – – – – – – – – – – – – – Afrotheria
– – – – – – – – – – – – – – – Boreoeutheria
– – – – – – – – – – – – – – – – †Apatemyidae
– – – – – – – – – – – – – – – – Laurasiatheria
– – – – – – – – – – – – – – – – – Eulipotyphles
– – – – – – – – – – – – – – – – – Scrotifera
– – – – – – – – – – – – – – – – – – Chiroptera
– – – – – – – – – – – – – – – – – – Fereungulata
– – – – – – – – – – – – – – – – – – – Ferae
– – – – – – – – – – – – – – – – – – – – †Creodonta
– – – – – – – – – – – – – – – – – – – – Carnivora
– – – – – – – – – – – – – – – – – – – – Pholidota
284 Appendix
Table A.2
(continued)
– – – – – – – – – – – – – – – – – – – Perissodactyla
– – – – – – – – – – – – – – – – – – – †Meridiungulata
– – – – – – – – – – – – – – – – – – – Cetartiodactyla
– – – – – – – – – – – – – – – – Euarchontoglires
– – – – – – – – – – – – – – – – – Glires
– – – – – – – – – – – – – – – – – – Lagomorpha
– – – – – – – – – – – – – – – – – – Rodentia
– – – – – – – – – – – – – – – – – Euarchonta
– – – – – – – – – – – – – – – – – – Dermoptera
– – – – – – – – – – – – – – – – – – Primates
– – – – – – – – – – – – – – – – – – – Strepsirrhini
– – – – – – – – – – – – – – – – – – – Haplorrhini
– – – – – – – – – – – – – – – – – – – – Tarsiiformes
– – – – – – – – – – – – – – – – – – – – Simiiformes
– – – – – – – – – – – – – – – – – – – – – Platyrrhini
– – – – – – – – – – – – – – – – – – – – – Catarrhini
– – – – – – – – – – – – – – – – – – – – – – Cercopithecoidea
– – – – – – – – – – – – – – – – – – – – – – Hominoidea
– – – – – – – – – – – – – – – – – – – – – – – Hylobatoidae
– – – – – – – – – – – – – – – – – – – – – – – Hominoidae
– – – – – – – – – – – – – – – – – – – – – – – – Pongidae
– – – – – – – – – – – – – – – – – – – – – – – – Hominidae
Guyader (2006) and Benton (2005).
Table A.3
A phylogenetic classification of plant taxa discussed in this book
Chlorobionta
– Ulvophyta
– Prasinophyta
– Streptophyta
– – Chlorokybophyta
– – Klebsormidiophyta
– – Phragmoplastophyta
– – – Zygnematophyta
– – – Plasmodesmophyta
– – – – Chaetosphaeridiophyta
– – – – Charophyta
Table A.3
(continued)
– – – – Parenchymophyta
– – – – – Coleochaetophyta
– – – – – Embryophyta
– – – – – – Marchantiophyta
– – – – – – Stomatophyta
– – – – – – – Anthocerophyta
– – – – – – – Hemitracheophyta
– – – – – – – – Bryophyta sensu stricto
– – – – – – – – Polysporangiophyta
– – – – – – – – – †Horneophyta
– – – – – – – – – Tracheophyta
– – – – – – – – – – Lycophyta
– – – – – – – – – – Euphyllophyta
– – – – – – – – – – – Moniliformopses
– – – – – – – – – – – – †Cladoxylopsida
– – – – – – – – – – – – Equisetophyta
– – – – – – – – – – – – Filicophyta
– – – – – – – – – – – Lignophyta
– – – – – – – – – – – – †Aneurophytales
– – – – – – – – – – – – †Archaeopteridales
– – – – – – – – – – – – Spermatophyta
– – – – – – – – – – – – – †Lyginopteridales
– – – – – – – – – – – – – †Medullosanales
– – – – – – – – – – – – – †Gigantopteridales
– – – – – – – – – – – – – Ginkgophyta
– – – – – – – – – – – – – Cycadophyta
– – – – – – – – – – – – – Pinophyta
– – – – – – – – – – – – – Gnetophyta
– – – – – – – – – – – – – Angiospermae
– – – – – – – – – – – – – – Amborellales
– – – – – – – – – – – – – – Nymphaeales
– – – – – – – – – – – – – – Austrobaileyales
– – – – – – – – – – – – – – Euangiosperms
– – – – – – – – – – – – – – – Chloranthales
– – – – – – – – – – – – – – – Ceratophyllales
– – – – – – – – – – – – – – – Magnoliidae
– – – – – – – – – – – – – – – – Magnoliales
– – – – – – – – – – – – – – – – Laurales
– – – – – – – – – – – – – – – – Canellales
– – – – – – – – – – – – – – – – Piperales
286 Appendix
Table A.3
(continued)
– – – – – – – – – – – – – – – Monocotyledons
– – – – – – – – – – – – – – – – Acorales
– – – – – – – – – – – – – – – – Alismatales
– – – – – – – – – – – – – – – – Eumonocotyledons
– – – – – – – – – – – – – – – – – Miyoshiales
– – – – – – – – – – – – – – – – – Dioscoreales
– – – – – – – – – – – – – – – – – Pandanales
– – – – – – – – – – – – – – – – – Liliales
– – – – – – – – – – – – – – – – – Asparagales
– – – – – – – – – – – – – – – – – Commelinidae
– – – – – – – – – – – – – – – Eudicots
– – – – – – – – – – – – – – – – Ranunculalaes
– – – – – – – – – – – – – – – – Sabiales
– – – – – – – – – – – – – – – – Proteales
– – – – – – – – – – – – – – – – Buxales
– – – – – – – – – – – – – – – – Trochodendrales
– – – – – – – – – – – – – – – – Core eudicots
– – – – – – – – – – – – – – – – – Berberidopsidales
– – – – – – – – – – – – – – – – – Santalales
– – – – – – – – – – – – – – – – – Dilleniales
– – – – – – – – – – – – – – – – – Caryophyllales
– – – – – – – – – – – – – – – – – Saxifragales
– – – – – – – – – – – – – – – – – Vitales
– – – – – – – – – – – – – – – – – Rosidae
– – – – – – – – – – – – – – – – – – Crossosomatales
– – – – – – – – – – – – – – – – – – Geraniales
– – – – – – – – – – – – – – – – – – Myrtales
– – – – – – – – – – – – – – – – – – Eurosids I
– – – – – – – – – – – – – – – – – – Eurosids II
– – – – – – – – – – – – – – – – – Asteridae
– – – – – – – – – – – – – – – – – – Cornales
– – – – – – – – – – – – – – – – – – Ericales
– – – – – – – – – – – – – – – – – – Euasteridae
– – – – – – – – – – – – – – – – – – – Euasterids I
– – – – – – – – – – – – – – – – – – – Euasterids II
Guyader (2006), Donoghue (2005), APG II (2003), and Niklas (1997).
References
Aczel, A. D. 2007. The Jesuit and the skull. New York: Riverhead Books.
Agosta, W. 1996. Bombardier beetles and fever trees: A close-up look at chemical warfare
and signals in animals and plants. Reading, Mass.: Helix Books.
Alberch, P. 1989. The logic of monsters: Evidence for internal constraint in development
and evolution. Geobios (Lyon, France): Memoire Special 12:21–57.
Albert, V. A., S. E. Williams, and M. W. Chase. 1992. Carnivorous plants: Phylogeny and
structural evolution. Science 257:1491–1495.
APG II. 2003. An update of the Angiosperm Phylogeny Group classification for the orders
and families of flowering plants: APG II. Botanical Journal of the Linnean Society
141:399–436.
Axe, D. D. 2004. Estimating the prevalence of protein sequences adopting functional
enzyme folds. Journal of Molecular Biology 341:1295–1315.
Baber, C. 2003. Cognition and tool use: Forms of engagement in human and animal use of
tools. London: Taylor and Francis.
Bailly, X., S. Vanin, C. Chabasse, K. Mizuguchi, and S. N. Vinogradov. 2008. A phylogenetic
profile of hemerythrins, the nonheme diiron binding respiratory proteins. BMC Evolution-
ary Biology 8:244, doi:10.1186/1471-2148-8-244.
Bakker, R. T. 1983. The deer flees, the wolf pursues: Incongruencies in predator-prey coevo-
lution. In Coevolution, ed. D. J. Futuyma and M. Slatkin, 350–382. Sunderland, Mass.:
Sinauer Associates, Inc.
Bakker, R. T. 1986. The dinosaur heresies. New York: William Morrow.
Ballerini, M., N. Cabibbo, R. Candelier, A. Cavagna, E. Cisbani, I. Giardina, V. Lecomte, A.
Orlandi, G. Parisi, A. Procaccini, M. Viale, and V. Zdravkovic. 2008. Interaction ruling
animal collective behavior depends on topological rather than metric distance: Evidence
from a field study. Proceedings of the National Academy of Sciences of the United States of
America 105 (4): 1232–1237.
Bambach, R. K. 1983. Ecospace utilization and guilds in marine communities through the
Phanerozoic. In Biotic interactions in recent and fossil benthic communities, ed. M. J. S.
Tevesz and P. M. McCall, 719–746. New York: Plenum.
Barlow, C. 2005. Let there be sight! A celebration of convergent evolution. http://www
.thegreatstory.org.
Barrett, P. M. 2005. The diet of ostrich dinosaurs (Theropoda: Ornithomimosauris). Palae-
ontology 48 (2): 347–358.
Baskin, D. G., and D. W. Goldring. 1970. Experimental studies on the endocrinology
and reproductive biology of the viviparous polychaete annelid, Nereis limnicola Johnson.
Biological Bulletin 139:461–475.
288 References
Bastian, J. 1982. Vision and electroreception: Integration of sensory information in the optic
tectum of the weakly electric fish Apteronotus albifrons. Journal of Comparative Physiol-
ogy 147A:287–297.
Bateman, R. M., and W. A. DiMichele. 1994. Heterospory: The most iterative key innovation
in the evolutionary history of the plant kingdom. Biological Reviews of the Cambridge
Philosophical Society 69:345–417.
Bennett, S. C. 1996. Aerodynamics and thermoregulatory function of the dorsal sail of
Edaphosaurus. Paleobiology 22 (4):496–506.
Benton, M. J. 2005. Vertebrate palaeontology. 3rd ed. Oxford: Blackwell Publishing.
Berg, L. S. 1922. Nomogenesis; or, evolution determined by law. Cambridge, Mass.: Massa-
chusetts Institute of Technology Press (1969 reprint).
Berkov, A., N. Rodríguez, and P. Centeno. 2008. Convergent evolution in the antennae of
a cerambycid beetle, Onychocerus albitarsis, and the sting of a scorpion. Naturwissen-
schaften 95 (3):257–261.
Berman, D. S., R. R. Reisz, D. Scott, A. C. Henrici, S. S. Sumida, and T. Martens. 2000. Early
Permian bipedal reptile. Science 290:968–972.
Bernhardt, P. 2000. Convergent evolution and adaptive radiation of beetle-pollinated
angiosperms. Plant Systematics and Evolution 222:293–320.
Blackburn, D. G. 1992. Convergent evolution of viviparity, matrotrophy, and specializations
for fetal nutrition in reptiles and other vertebrates. American Zoologist 32:313–321.
Blackburn, D. G. 1993. Chorioallantoic placentation in squamate reptiles: Structure, func-
tion, development, and evolution. Journal of Experimental Zoology 266:414–430.
Blackburn, D. G. 2000. Reptilian viviparity: Past research, future directions, and appropriate
models. Comparative Biochemistry and Physiology Part A 127A:391–409.
Blackburn, D. G., L. J. Vitt, and C. A. Beuchat. 1984. Eutherian-like reproductive specializa-
tions in a viviparous reptile. Proceedings of the National Academy of Sciences of the United
States of America 81:4860–4863.
Block, B. A., J. R. Finnerty, A. F. R. Stewart, and J. Kidd. 1993. Evolution of endothermy in
fish: Mapping physiological traits on a molecular phylogeny. Science 260:210–214.
Blumberg, M. S. 2009. Freaks of nature: What anomalies tell us about development and
evolution. Oxford: Oxford University Press.
Boomsma, J. J. 2009. Lifetime monogamy and the evolution of eusociality. Philosophical
Transactions of the Royal Society of London: Series B, Biological Sciences B364:3191–
3207.
Bork, P., C. Sander, and A. Valencia. 1993. Convergent evolution of similar enzymatic func-
tion on different protein folds: The hexokinase, ribokinase, and galactokinase families of
sugar kinases. Protein Science 2:31–40.
Bowmaker, J. K., and D. M. Hunt. 2006. Evolution of vertebrate visual pigments. Current
Biology 16 (13):R484–R489.
Boyce, C. K., and A. H. Knoll. 2002. Evolution of developmental potential and the
multiple independent origins of leaves in Paleozoic vascular plants. Paleobiology 28
(1):70–100.
Brady, S. G. 2003. Evolution of the army ant syndrome: The origin and long-term
evolutionary stasis of a complex of behavioral and reproductive adaptations. Proceedings
of the National Academy of Sciences of the United States of America 100 (11):6575–
6579.
Brady, S. G., S. Sipes, A. Pearson, and B. N. Danforth. 2006. Recent and simultaneous origins
of eusociality in halictid bees. Philosophical Transactions of the Royal Society of London:
Series B, Biological Sciences B273:1643–1649.
Brakefield, P. M. 2006. Evo-devo and constraints on selection. Trends in Ecology and Evolu-
tion 21 (7):362–368.
References 289
Budelmann, B., and H. Bleckmann. 1988. A lateral line analogue in cephalopods: Water
waves generate microphonic potentials in the epidermal head lines of Sepia and Lolligun-
cula. Journal of Comparative Physiology 164A:1–5.
Buhl, J., D. J. T. Sumpter, I. D. Couzin, J. J. Hale, E. Despland, E. R. Miller, and S. J. Simpson.
2006. From disorder to order in marching locusts. Science 312:1402–1406.
Bull, J., and J. Farrand. 1988. The Audubon Society field guide to North American birds. New
York: Alfred A. Knopf.
Bush, A. M., R. K. Bambach, and G. M. Daley. 2007. Changes in theoretical ecospace utiliza-
tion in marine fossil assemblages between the mid-Paleozoic and late Cenozoic. Paleobiol-
ogy 33 (1):76–97.
Byrne, M. 2005. Viviparity in the sea star Cryptasterina hysteria (Asterinidae): Conserved
and modified features in reproduction and development. Biological Bulletin 208:81–91.
Camazine, S., J.-L. Deneubourg, N. R. Franks, J. Sneyd, G. Theraulaz, and E. Bonabeau.
2001. Self-organization in biological systems. Princeton: Princeton University Press.
Cameron, K. M., K. J. Wurdack, and R. W. Jobson. 2002. Molecular evidence for the common
origin of snap-traps among carnivorous plants. American Journal of Botany 89 (9):
1503–1509.
Cameron, S. A., and P. Mardulyn. 2001. Multiple molecular data sets suggest independent
origins of highly eusocial behavior in bees (Hymenoptera: Apinae). Systematic Biology 50
(2):194–214.
Carrano, M. T., C. M. Janis, and J. J. Sepkoski. 1999. Hadrosaurs as ungulate parallels: Lost
lifestyles and deficient data. Acta Palaeontologica Polonica 44 (3):237–261.
Carroll, L. [C. W. Dodgson]. 1865. Alice’s adventures in Wonderland. London: Macmillan.
Carroll, R. L., and W. Lindsay. 1985. Cranial anatomy of the primitive reptile Procolophon.
Canadian Journal of Earth Sciences 22:1571–1587.
Carroll, S. B. 2006. Endless forms most beautiful: The new science of EvoDevo. New York:
W. W. Norton and Co.
Carter, A. M. 2008. What fossils can tell us about the evolution of viviparity and placenta-
tion. Placenta 29:930–931.
Cashmore, A. R., J. A. Jarillo, Y.-J. Wu, and D. Liu. 1999. Cryptochromes: Blue light recep-
tors for plants and animals. Science 284:760–765.
Castelvecchi, D. 2007. Alien pizza, anyone? Biochemistry may have taken a different turn
on other worlds. Science News 172:107–109.
Casti, J. L. 1989. Paradigms lost: Tackling the unanswered mysteries of modern science.
New York: Avon Bard.
Castoe, T. A., A. P. J. de Koning, H.-M. Kim, B. P. Noonan, G. Naylor, Z. J. Jiang, C. L. Par-
kinson, and D. D. Pollock. 2009. Evidence for an ancient adaptive episode of convergent
molecular evolution. Proceedings of the National Academy of Sciences of the United States
of America 106 (22):8986–8991.
Castoe, T. A., A. P. J. de Koning, and D. D. Pollock. 2010. Adaptive molecular convergence:
Molecular evolution versus molecular phylogenetics. Communicative and Integrative
Biology 3 (1):67–69.
Charnock, S. J., I. E. Brown, J. P. Turkenburg, G. W. Black, and G. J. Davies. 2002. Convergent
evolution sheds light on the anti-β-elimination mechanism common to family 1 and 10
polysaccharide lyases. Proceedings of the National Academy of Sciences of the United States
of America 99 (10):12067–12072.
Chen, L., A. L. DeVries, and C.-H. C. Cheng. 1997. Convergent evolution of antifreeze
glycoproteins in Antarctic notothenioid fish and Arctic cod. Proceedings of the National
Academy of Sciences of the United States of America 94:3817–3822.
Cheng, Y.-N., X.-C. Wu, and Q. Ji. 2004. Triassic marine reptiles gave birth to live young.
Nature 432:383–386.
290 References
Cimino, G., S. De Rosa, S. De Stefano, R. Morrone, and G. Sodano. 1985. The chemical
defense of nudibranch molluscs: Structure, biosynthetic origin and defensive properties of
terpenoids from the dorid nudibranch Dendrodoris grandiflora. Tetrahedron 41
(6):1093–1100.
Clack, J. A. 2002. Gaining ground: The origin and evolution of tetrapods. Bloomington:
Indiana University Press.
Claessens, L. P. A. M., P. M. O’Connor, and D. M. Unwin. 2009. Respiratory evolution
facilitated the origin of pterosaur flight and aerial gigantism. PLoS ONE 4 (2):e4497. www
.plosone.org.
Clark, J. M., L. L. Jacobs, and W. R. Downs. 1989. Mammal-like dentition in a Mesozoic
crocodylian. Science 244:1064–1066.
Clark, R. N., J. M. Curchin, J. W. Barnes, R. Jauman, L. Sonderblom, D. P. Cruikshank,
R. H. Brown, et al. 2010. Detection and mapping hydrocarbon deposits on Titan. Journal
of Geophysical Research-Planets, doi:10.1029/2009JE003369.
Clayton, N. S., and Emery, N. J. 2008. Canny corvids and political primates: A case for
convergent evolution in intelligence. In The deep structure of biology, ed. S. Conway Morris,
128–142. West Conshohocken, Pa.: Templeton Foundation Press.
Colosimo, P. F., K. E. Hosemann, S. Balabhadra, G. Villarreal, M. Dickson, J. Grimwood,
J. Schmutz, R. M. Myers, D. Schluter, and D. M. Kingsley. 2005. Widespread parallel
evolution in sticklebacks by repeated fixation of Ectodysplasin alleles. Science 307:1928–
1933.
Conradt, L., and T. J. Roper. 2005. Consensus decision making in animals. Trends in Ecology
and Evolution 20 (8):449–456.
Conway Morris, S. 2003. Life’s solution: Inevitable humans in a lonely universe. Cambridge:
Cambridge University Press.
Conway Morris, S. 2008. Evolution and convergence: Some wider considerations. In The
deep structure of biology, ed. S. Conway Morris, 46–67. West Conshohocken, Pa.: Templeton
Foundation Press.
Cooper, A., C. Lalueza-Fox, S. Anderson, A. Rambaut, J. Austin, and R. Ward. 2001. Com-
plete mitochondrial genome sequences of two extinct moas clarify ratite evolution. Nature
409:704–707.
Couzin, I. D. 2007. Collective minds. Nature 445:715.
Couzin, I. D., and J. Krause. 2003. Self-organization and collective behavior in vertebrates.
Advances in the Study of Behavior 32:1–75.
Crespi, B. J., D. A. Carmean, and T. W. Chapman. 1997. Ecology and evolution of galling
thrips and their allies. Annual Review of Entomology 42:51–71.
Crick, F. H. C. 1988. What mad pursuit: A personal view of scientific discovery. New York:
Basic Books.
Cronk, Q., and I. Ojeda. 2008. Bird-pollinated flowers in an evolutionary and molecular
context. Journal of Experimental Botany 59 (4):715–727.
Darwin, C. 1859. On the origin of species by means of natural selection, or the preservation
of favoured races in the struggle for life. London: John Murray.
Darwin, C. 1875. Insectivorous plants. London: John Murray.
Daugherty, C. H., G. W. Gibbs, and R. A. Hitchmough. 1993. Mega-island or micro-conti-
nent? New Zealand and its fauna. Trends in Ecology and Evolution 8 (2):437–442.
Davies, P. L., J. Baardsnes, M. J. Kuiper, and V. K. Walker. 2002. Structure and function of
antifreeze proteins. Philosophical Transactions of the Royal Society of London: Series B,
Biological Sciences B357: 927–935.
Davis, C. C., P. K. Endress, and D. A. Baum. 2008. The evolution of floral gigantism. Current
Opinion in Plant Biology 11:49–57.
References 291
Deeb, S. S., M. J. Wakefield, T. Tada, L. Marotte, S. Yokoyama, and J. A. M. Graves. 2003.

The cone visual pigments of an Australian marsupial, the Tammar wallaby (Macropus
eugenii): Sequence, spectral tuning, and evolution. Molecular Biology and Evolution 20
(10):1642–1649.
Delabie, J. H. C. 2001. Trophobiosis between Formicidae and Hemiptera (Sternorrhyncha
and Auchenorrhyncha): An overview. Neotropical Entomology 30 (4):501–516.
DeMar, R., and J. R. Bolt. 1981. Dentitional organization and function in a Triassic reptile.
Journal of Paleontology 55:967–984.
Denton, E. J., and J. Gray. 1985. Lateral-line-like antennae of certain of the Penaeidea
(Crustacea, Decapoda, Natantia). Philosophical Transactions of the Royal Society of
London: Series B, Biological Sciences B226:249–261.
Denton, M., and C. Marshall. 2001. Laws of forms revisited. Nature 410:417.
Denton, M. J., C. J. Marshall, and M. Legge. 2002. The protein folds as Platonic forms: New
support for the pre-Darwinian conception of evolution by natural law. Journal of Theoreti-
cal Biology 219:325–342.
Dijkgraaf, S. 1967. Biological significance of the lateral line organs. In Lateral line detectors,
ed. P. H. Cahn, 83–95. Bloomington: Indiana University Press.
Donoghue, M. J. 2005. Key innovations, convergence, and success: Macroevolutionary
lessons from plant phylogeny. Paleobiology 31 (2 Supplement):77–93.
Donoghue, P. C. J., and M. A. Purnell. 1999. Mammal-like occlusion in conodonts. Paleobiol-
ogy 25 (1):58–74.
Douglas, H. D., III, J. E. Co, T. H. Jones, W. E. Conner, and J. F. Day. 2005. Chemical odorant
of colonial seabird repels mosquitos. Journal of Medical Entomology 42 (4):647–651.
Duffy, J. E., C. L. Morrison, and R. Ríos. 2000. Multiple origins of eusociality among sponge-
dwelling shrimps (Synalpheus). Evolution: International Journal of Organic Evolution 54
(2):503–516.
Dulvy, N. K., and J. D. Reynolds. 1997. Evolutionary transitions among egg- laying, live-
bearing and maternal inputs in sharks and rays. Philosophical Transactions of the Royal
Society of London: Series B, Biological Sciences B264:1309–1315.
Dunn, R. R., A. D. Gove, T. G. Barraclough, T. J. Givnish, and J. D. Majer. 2007. Convergent
evolution of an ant-plant mutualism across plant families, continents, and time. Evolution-
ary Ecology Research 9:1349–1362.
Eisner, T., M. Eisner, and M. Siegler. 2005. Secret weapons: Defenses of insects, spiders,
scorpions, and other many-legged creatures. Cambridge, Mass.: Belknap Press.
Ellison, A. M., and N. J. Gotelli. 2009. Energetics and the evolution of carnivorous plants—
Darwin’s “most wonderful plants in the world.” Journal of Experimental Botany 60
(1):19–42.
Emery, N. J., and N. S. Clayton. 2004. The mentality of crows: Convergent evolution of
intelligence in corvids and apes. Science 306:1903–1907.
Evans, W. G. 1966. Perception of infrared radiation from forest fires by Melanophilia acu-
minata De Geer (Buprestidae, Coleoptera). Ecology 47:1061–1065.
Farrell, B. D., A. S. Sequeira, B. C. O’Meara, B. B. Normark, J. H. Chung, and B. H. Jordal.
2001. The evolution of agriculture in beetles (Curculionidae: Scolytinae and Platypodinae).
Evolution: International Journal of Organic Evolution 55 (10):2011–2027.
Feduccia, A. 1996. The origin and evolution of birds. New Haven, Conn.: Yale University
Press.
Finn, J. K., T. Tregenza, and M. D. Norman. 2009. Defensive tool use in a coconut-carrying
octopus. Current Biology 19 (23):R1070.
Fitzhugh, K. 1989. A systematic revision of the Sabellidae-Caobangiidae-Sabellongidae
complex (Annelida: Polychaeta). Bulletin of the American Museum of Natural History
192:1–104.
292 References
Fitzhugh, K. 1991. Further revisions of the Sabellidae subfamilies and cladistic relationships
among the Fabriciinae (Annelida: Polychaeta). Zoological Journal of the Linnean Society
102:305–332.
Flemming, A. F., and D. G. Blackburn. 2003. Evolution of placental specializations in vivipa-
rous African and South American lizards. Journal of Experimental Zoology 299A:33–47.
Fletcher, G. L., C. L. Hew, and P. L. Davies. 2001. Antifreeze proteins of teleost fishes.
Annual Review of Physiology 63:359–390.
Foster, K. W., and R. D. Smyth. 1980. Light antennas in phototactic algae. Microbiological
Reviews 44 (4):572–630.
Freeland, S. J., and L. D. Hurst. 1998. The genetic code is one in a million. Journal of
Molecular Evolution 47:238–248.
Gartrell, B. D. 2000. The nutritional, morphologic, and physiologic bases of nectarivory in
Australian birds. Journal of Avian Medicine and Surgery 14 (2):85–94.
Gaudin, T. J., and D. G. Branham. 1998. The phylogeny of the Myrmecophagidae (Mam-
malia, Xenarthra, Vermilingua) and the relationship of Eurotamandua to the Vermilingua.
Journal of Mammalian Evolution 5 (3):237–265.
Gensel, P. G., and A. E. Kasper. 2005. A new species of the Devonian lycopod genus,
Leclercqia, from the Emsian of New Brunswick, Canada. Review of Palaeobotany and
Palynology 137:105–123.
Givnish, T. J., J. C. Pires, S. W. Graham, M. A. McPherson, L. M. Prince, T. B. Patterson, H.
S. Rai, et al. 2005. Repeated evolution of net venation and fleshy fruits among monocots
in shaded habitats confirms a priori predictions: Evidence from an ndhF phylogeny. Philo-
sophical Transactions of the Royal Society of London: Series B, Biological Sciences
B272:1481–1490.
Gould, E. 1965. Evidence for echolocation in the Tenrecidae of Madagascar. Proceedings
of the American Philosophical Society 109 (6):352–360.
Gould, J. L., and C. G. Gould. 2007. Animal architects: Building and the evolution of intel-
ligence. New York: Basic Books.
Gould, S. J. 1989. Wonderful life: The Burgess Shale and the nature of history. New York:
W. W. Norton.
Grant, K. A., and V. Grant. 1968. Hummingbirds and their flowers. New York: Columbia
University Press.
Gregory, J. E., A. Iggo, A. K. McIntyre, and U. Proske. 1987. Electroreceptors in the platypus.
Nature 326:386–387.
Grinspoon, D. 2003. Lonely planets: The natural philosophy of alien life. New York:
HarperCollins Publishers Inc.
Gupta, A. K. 2004. Origin of agriculture and domestication of plants and animals linked to
early Holocene climate amelioration. Current Science 87 (1):54–59.
Hagan, H. R. 1951. Embryology of the viviparous insects. New York: Ronald Press.
Hamilton, W. D. 1964. The genetical evolution of social behavior I, II. Journal of Theoretical
Biology 7:1–52.
Hansell, M. 2005. Animal architecture. Oxford: Oxford University Press.
Hansell, M., and G. D. Ruxton. 2008. Setting tool use within the context of animal construc-
tion behaviour. Trends in Ecology and Evolution 23 (2):73–78.
Hansen, D. M., and M. Galetti. 2009. The forgotten megafauna. Science 324:42–43.
Hardison, R. C. 1996. A brief history of hemoglobins: Plant, animal, protist, and bacteria.
Proceedings of the National Academy of Sciences of the United States of America
93:5675–5679.
Harshman, J., E. L. Braun, M. J. Braun, C. J. Huddleston, R. C. K. Bowie, J. L. Chojnowski,
S. J. Hackett, et al. 2008. Phylogenetic evidence for multiple losses of flight in ratite birds.
References 293

105:13462–13467.
Hauser, M. D. 2009. The possibility of impossible cultures. Nature 460:190–196.
Heinrich, B. 1993. The hot-blooded insects: Strategies and mechanisms of thermoregulation.
Cambridge, Mass.: Harvard University Press.
Heinrich, B. 1996. The thermal warriors: Strategies of insect survival. Cambridge, Mass.:
Harvard University Press.
Hines, H. M., J. H. Hunt, T. K. O’Connor, J. J. Gillespie, and S. A. Cameron. 2007. Multigene
phylogeny reveals eusociality evolved twice in vespid wasps. Proceedings of the National
Academy of Sciences of the United States of America 104 (9):3295–3299.
Holmes, M. M., B. D. Goldman, S. L. Goldman, M. L. Seney, and N. G. Forger. 2009. Neu-
roendocrinology and sexual differentiation in eusocial mammals. Frontiers in Neuroendo-
crinology 30:519–533.
Hopkins, C. D. 2008. Commentary: Evolution of electric organs. Journal of Physiology, Paris
102:162–163.
Hosler, J. 2003. The Sandwalk adventures: An adventure in evolution told in five chapters.
Columbus, Ohio: Active Synapse.
Hoy, R. R., and D. Robert. 1996. Tympanal hearing in insects. Annual Review of Entomology
41:433–450.
Hunt, D. M., K. S. Dulai, J. A. Cowing, C. Julliot, J. D. Mollon, J. K. Bowmaker, W.-H. Li,
and D. Hewett-Emmett. 1998. Molecular evolution of trichromacy in primates. Vision
Research 38:3299–3306.
Hunter, J. P., and J. Jernvall. 1995. The hypocone as a key innovation in mammalian evolu-
tion. Proceedings of the National Academy of Sciences of the United States of America
92:10718–10722.
Janis, C. M. 1990. The correlation between diet and dental wear in herbivorous mammals,
and its relationship to the determination of diets in extinct species. In Evolutionary paleo-
biology of behavior and coevolution, ed. A. J. Boucot, 241–259. Amsterdam: Elsevier Press.
Jernvall, J. 2000. Linking development with generation of novelty in mammalian teeth.
97:2641–2645.
Ji, Q., Z.-X. Luo, C.-X. Yuan, J. R. Wible, J.-P. Zhang, and J. A. Georgi. 2002. The earliest
known eutherian mammal. Nature 416:816–822.
Johns, P. M., K. J. Howard, N. L. Breisch, A. Rivera, and B. L. Thorne. 2009. Nonrelatives
inherit colony resources in a primitive termite. Proceedings of the National Academy of
Sciences of the United States of America 106 (41):17452–17456.
Jost, M. C., D. M. Hillis, Y. Lu, J. W. Kyle, H. A. Fozzard, and H. H. Zakon. 2008. Toxin-
resistant sodium channels: Parallel adaptive evolution across a complete gene family.
Molecular Biology and Evolution 25 (6):1016–1024.
Kauffman, S. A. 2008. Reinventing the sacred: A new view of science, reason and religion.
New York: Basic Books.
Kellogg, E. A. 1999. Phylogenetic aspects of the evolution of C4 photosynthesis. In C4 plant
biology, ed. R. F. Sage and R. K. Monson, 411–444. San Diego: Academic Press.
Kelly, D. A. 2002. The functional morphology of penile erection: Tissue designs for increas-
ing and maintaining stiffness. Integrative and Comparative Biology 42:216–221.
Kemp, A. 1977. The pattern of tooth plate formation in the Australian lungfish, Neocera-
todus forsteri Krefft. Biological Journal of the Linnean Society 60:223–258.
Kielan-Jaworowska, Z. 1979. Pelvic structure and nature of reproduction in Multitubercu-
lata. Nature 277:402–403.
King, A. S. 1981. Phallus. In Form and function in birds. Vol. 2, ed. A. S. King and J. McLel-
land, 107–147. London: Academic Press.
294 References
King, N., C. T. Hittinger, and S. B. Carroll. 2003. Evolution of key cell signaling and adhesion
protein families predates animal origins. Science 301:361–363.
Kivic, P. A., and P. L. Walne. 1983. Algal photosensory apparatus probably represent mul-
tiple parallel evolutions. Bio Systems 16:31–38.
Klok, C. J., R. D. Mercer, and S. L. Chown. 2002. Discontinuous gas-exchange in centipedes
and its convergent evolution in tracheated arthropods. Journal of Experimental Biology
205:1019–1029.
Koenigswald, W. v., K. D. Rose, L. Grande, and R. D. Martin. 2005. First apatemyid skeleton
from the Lower Eocene Fossil Butte Member, Wyoming (USA), compared to the Euro-
pean apatemyid from Messel, Germany. Palaeontographica Abteilung A272:149–169.
Koenigswald, W. v., and H.-P. Schierning. 1987. The ecological niche of an extinct group of
mammals, the early Tertiary apatemyids. Nature 326:595–597.
Kornegay, J. R., J. W. Schilling, and A. C. Wilson. 1994. Molecular adaptation of a leaf-eating
bird: Stomach lysozyme of the Hoatzin. Molecular Biology and Evolution 11 (6):921–928.
Kuhn-Schnyder, E., and H. Rieber. 1986. Handbook of paleozoology. Baltimore: Johns
Hopkins University Press.
Kunz, T. H., M. S. Fujita, A. P. Brooke, and G. R. McCracken. 1994. Convergence in tent
architecture and tent-making behavior among Neotropical and Paleotropical bats. Journal
of Mammalian Evolution 2:57–58.
Kuroki, T. 1967. Theoretical analysis of the role of the lateral line in directional hearing. In
Lateral line detectors, ed. P. H. Cahn, 217–237. Bloomington: Indiana University Press.
Labandeira, C. C., J. Kvač ek, and M. B. Mostovski. 2007. Pollination drops, pollen, and insect
pollination of Mesozoic gymnosperms. Taxon 56:663–695.
Laughlin, S., A. D. Blest, and S. Stowe. 1980. The sensitivity of receptors in the posterior
median eye of the nocturnal spider, Dinopis. Journal of Comparative Physiology
141:53–65.
Lawn, R. M., K. Schwartz, and L. Patthy. 1997. Convergent evolution of apolipoprotein(a)
in primates and hedgehog. Proceedings of the National Academy of Sciences of the United
Lecointre, G., and H. Le Guyader. 2006. The tree of life: A phylogenetic classification. Cam-
bridge, Mass.: Belknap Press of Harvard University Press.
Lee, M. S. Y. 1998. Convergent evolution and character correlation in burrowing reptiles:
Towards a resolution of squamate relationships. Biological Journal of the Linnean Society
65:369–453.
Lee, M. S. Y., and R. Shine. 1998. Reptilian viviparity and Dollo’s law. Evolution: Interna-
tional Journal of Organic Evolution 52 (5):1441–1450.
Loewer, P. 1995. Seeds: The definitive guide to growing, history, and lore. Portland, Ore.:
Timber Press.
Logsdon, J. M., and W. R. Doolittle. 1997. Origin of antifreeze protein genes: A cool tale in
molecular evolution. Proceedings of the National Academy of Sciences of the United States
of America 94:3485–3487.
Long, J. A., K. Trinajstic, G. C. Young, and T. Senden. 2008. Live birth in the Devonian
period. Nature 453:650–653.
Longrich, N. R., and P. J. Currie. 2008. Albertonykus borealis, a new alvarezsaur (Dinosauria:
Theropoda) from the Early Maastrichtian of Alberta, Canada: Implications for the system-
atics and ecology of the Alvarezsauridae. Cretaceous Research 30 (1):239–252.
Luo, Z.-X., R. L. Cifelli, and Z. Kielan-Jaworowska. 2001. Dual origin of tribosphenic
mammals. Nature 409:53–57.
Luo, Z.-X., Q. Ji, and C.-X. Yuan. 2007. Convergent dental adaptations in pseudo-tribos-
phenic and tribosphenic mammals. Nature 450:93–97.
References 295
Maclaurin, J. 2003. The good, the bad and the impossible. Biology and Philosophy
18:463–476.
Maclaurin, J., and K. Sterelny. 2008. What is biodiversity? Chicago: University of Chicago
Press.
Malmström, T., and R. H. H. Kröger. 2006. Pupil shapes and lens optics in the eyes of ter-
restrial vertebrates. Journal of Experimental Biology 209:18–25.
Manger, P. R., R. Collins, and J. D. Pettigrew. 1997. Histological observations on presumed
electroreceptors and mechanoreceptors in the beak skin of the long-beaked echidna,
Zaglossus bruijnii. Philosophical Transactions of the Royal Society of London: Series B,
Biological Sciences B264:165–172.
Marino, L. 2002. Convergence of complex cognitive abilities in cetaceans and primates.
Brain, Behavior and Evolution 59:21–32.
Martin, A. J. 2006. Introduction to the study of dinosaurs. Oxford: Blackwell.
Martin, V. J. 2002. Photoreceptors of cnidarians. Canadian Journal of Zoology 80:
1703–1722.
Martinez, R. N., C. L. May, and C. A. Forster. 1996. A new carnivorous cynodont from the
Ischigualasto Formation (Late Triassic, Argentina), with comments on eucynodont phylog-
eny. Journal of Vertebrate Paleontology 16:271–284.
Maryańska, T., H. Osmólska, and M. Wolsan. 2002. Avialan status for Oviraptorosauria.
Acta Palaeontologica Polonica 47:97–116.
Masonjones, H. D., and S. M. Lewis. 1996. Courtship behavior in the dwarf seahorse, Hip-
pocampus zosterae. Copeia (3):634–640.
Mayr, E. 1964. Introduction. In A facsimile of the first edition of On the origin of species,
ed. E. Mayr, vii–xxvii. Cambridge, Mass.: Harvard University Press.
Mayr, G., and D. S. Peters. 2006. Response to comment on “A well-preserved Archaeopteryx
specimen with theropod features.” Science 313: 1238c, doi: 10.1126/science.1130964.
Mayr, G., B. Pohl, and D. S. Peters. 2005. A well-preserved Archaeopteryx specimen with
theropod features. Science 310: 1483–1486.
McGhee, G. R. 1999. Theoretical morphology: The concept and its applications. New York:
Columbia University Press.
McGhee, G. R. 2001. Exploring the spectrum of existent, nonexistent and impossible bio-
logical form. Trends in Ecology and Evolution 16:172–173.
McGhee, G. R. 2007. The geometry of evolution: Adaptive landscapes and theoretical
morphospaces. Cambridge: Cambridge University Press.
McGhee, G. R. 2008. Convergent evolution: A periodic table of life? In The deep structure
of biology, ed. S. Conway Morris, 17–31. West Conshohocken, Pa.: Templeton Foundation
Press.
McGhee, G. R., P. M. Sheehan, D. J. Bottjer, and M. L. Droser. 2004. Ecological ranking of
Phanerozoic biodiversity crises: Ecological and taxonomic severities are decoupled. Pal-
aeogeography, Palaeoclimatology, Palaeoecology 211:289–297.
McIntosh, J. S. 1997. Sauropoda. In Encyclopedia of dinosaurs, ed. P. J. Currie and K. Padian,
654–658. San Diego: Academic Press.
McKay, C. P., and H. D. Smith. 2005. Possibilities for methanogenic life in liquid methane
on the surface of Titan. Icarus 178:274–276.
McKinney, F. K. 2007. The Northern Adriatic ecosystem: Deep time in a shallow sea.
New York: Columbia University Press.
McKinney, F. K., S. J. Hageman, and A. Jaklin. 2007. Crossing the ecological divide:
Paleozoic to modern marine ecosystem in the Adriatic Sea. Sedimentary Record 5
(2):4–8.
296 References
McPherron, S. P., Z. Alemseged, C. W. Marean, J. G. Wynn, D. Reed, D. Geraads, R. Bobe,

and H. A. Béarat. 2010. Evidence for stone-tool-assisted consumption of animal tissues
before 3.39 million years ago at Dikika, Ethiopa. Nature 466:857–860.
Meyers, J. J., and A. Herrel. 2005. Prey capture kinematics of ant-eating lizards. Journal of
Experimental Biology 208:113–127.
Milewski, A. V., and W. J. Bond. 1982. Convergence of myrmecochory in Mediterranean
Australia and South Africa. In Ant-plant interactions in Australia, ed. R. C. Buckley, 89–98.
The Hague: Dr. W. Junk Publishers.
Moller, P. 1995. Electric fishes: History and behavior. London: Chapman and Hall.
Mooney, H. A. 1977. Convergent evolution in Chile and California: Mediterranean climate
ecosystems. Stroudsburg, Pa.: Dowden, Hutchinson and Ross, Inc.
Moore, J., and P. Willmer. 1997. Convergent evolution in invertebrates. Biological Reviews
of the Cambridge Philosophical Society 72:1–60.
Morell, V. 2002. Placentas may nourish complexity studies. Science 298:945.
Mueller, U. G., N. M. Gerardo, D. K. Aanen, D. L. Six, and T. R. Schultz. 2005. The evolution
of agriculture in insects. Annual Review of Ecology Evolution and Systematics
36:563–595.
Müller, G. B. 2007. Evo-devo: Extending the evolutionary synthesis. Nature Reviews: Genet-
ics 8:943–949.
Mummenhoff, K., A. Polster, A. Mühlhausen, and G. Theißen. 2008. Lepidium as a model
system for studying the evolution of fruit development in Brassicaceae. Journal of Experi-
mental Biology online, doi:10.1093/jxb/ern304.
Mundy, N. I. 2005. A window on the genetics of evolution: MC1R and plumage colouration
in birds. Philosophical Transactions of the Royal Society of London: Series B, Biological
Sciences B272:1633–1640.
Naylor, G. J. P., T. M. Collins, and W. M. Brown. 1995. Hydrophobicity and phylogeny. Nature
373:565–566.
Nevo, E. 1999. Mosaic evolution of subterranean mammals: Regression, progression, and
global convergence. Oxford: Oxford University Press.
Newman, S. A. 2010. Dynamical patterning modules. In Evolution: The extended synthesis,
ed. M. Pigliucci and G. B. Müller, 281–306. Cambridge, Mass.: Massachusetts Institute of
Technology Press.
Newman, S. A., G. Forgacs, and G. B. Müller. 2006. Before programs: The physical origina-
tion of multicellular forms. International Journal of Developmental Biology 50:289–299.
Niklas, K. 1997. The evolutionary biology of plants. Chicago: University of Chicago Press.
Niven, L. 1968. The handicapped. In Neutron star, 209–235. New York: Ballantine Books.
Niven, L. 1968a. A relic of the empire. In Neutron star, 29–49. New York: Ballantine
Books.
Norberg, R. Å. 1977. Occurrence and independent evolution of bilateral ear asymmetry in
owls and implications on owl taxonomy. Philosophical Transactions of the Royal Society of
London: Series B, Biological Sciences B280:375–408.
Oakley, T. H., and C. W. Cunningham. 2002. Molecular phylogenetic evidence for the inde-
pendent evolutionary origin of an arthropod compound eye. Proceedings of the National
O’Connor, P. M., and L. P. A. M. Claessens. 2005. Basic avian pulmonary design and flow-
through ventilation in non-avian theropod dinosaurs. Nature 436:253–256.
O’Connor, P. M., J. J. W. Sertich, N. J. Stevens, E. M. Roberts, M. D. Gottfried, T. L. Hiero-
nymus, Z. A. Jinnah, R. Ridgely, S. E. Ngasala, and J. Temba. 2010. The evolution of
mammal-like crocodyliforms in the Cretaceous Period of Gondwana. Nature 466:
748–751.
References 297
Ödeen, A., and O. Håstad. 2003. Complex distribution of avian color vision systems revealed
by sequencing the SWS1 opsin from total DNA. Molecular Biology and Evolution 20
(6):855–861.
O’Reilly, J. C., D. A. Ritter, and D. R. Carrier. 1997. Hydrostatic locomotion in a limbless
tetrapod. Nature 386:269–272.
Ostrom, J. H. 1990. Dromaeosauridae. In The Dinosauria, ed. D. B. Weishampel, P. Dodson,
and H. Osmólska, 269–279. Berkeley: University of California Press.
Peet, R. K. 1978. Ecosystem convergence. American Naturalist 112:441–459.
Pianka, E. R. 1978. Evolutionary ecology. 2nd ed. New York: Harper and Row.
Pianka, E. R., and W. S. Parker. 1975. Ecology of horned lizards: A review with special refer-
ence to Phrynosoma platyrhinos. Copeia 1975:141–162.
Piatigorsky, J. 2007. Gene sharing and evolution. Cambridge, Mass.: Harvard University
Press.
Poelwijk, F. J., D. J. Kiviet, D. M. Weinreich, and S. J. Tans. 2007. Empirical fitness landscapes
reveal accessible evolutionary paths. Nature 445:383–386.
Popov, I. Y. 2002. “Periodical systems” in biology (a historical issue). Verhandlungen zur
Geschichte und Theorie der Biologie 9:55–68.
Prior, H., A. Schwarz, and O. Güntürkün. 2008. Mirror-induced behavior in the magpie
(Pica pica): Evidence of self-recognition. PLoS Biology 6 (8):1642–1650.
Pritchard, G., M. H. McKee, E. M. Pike, G. J. Scrimgeour, and J. Zloty. 1993. Did the first
insects live in water or in air? Biological Journal of the Linnean Society 49:31–44.
Proske, U., and E. Gregory. 2003. Electrolocation in the platypus: Some speculations. Com-
parative Biochemistry and Physiology: Part A, Molecular and Integrative Physiology 136
(4):821–825.
Protas, M. E., C. Hersey, D. Kochanek, Y. Zhou, H. Wilkens, W. R. Jeffery, L. I. Zon, R.
Borowsky, and C. J. Tabin. 2005. Genetic analysis of cavefish reveals molecular convergence
in the evolution of albinism. Nature Genetics 38(1): 107–111.
Prothero, D. R., and R. M. Schoch. 2002. Horns, tusks, and flippers: The evolution of hoofed
mammals. Baltimore: Johns Hopkins University Press.
Prud’homme, B., and S. B. Carroll. 2006. Monkey see, monkey do. Nature Genetics 38
(7):740–741.
Purnell, M. A. 1995. Large eyes and vision in conodonts. Lethaia 28:187–188.
Raafat, R. M., N. Chater, and C. Frith. 2009. Herding in humans. Trends in Cognitive
Sciences 13 (10):420–428.
Rasmussen, D. T., E. L. Simons, F. Hertel, and A. Judd. 2001. Hindlimb of a giant terrestrial
bird from the Upper Eocene, Fayum, Egypt. Palaeontology 44 (2):325–337.
Reardon, D., and G. K. Farber. 1995. The structure and evolution of α/β barrel proteins.
Journal of the Federation of American Societies for Experimental Biology 9:497–503.
Reep, R. L., C. D. Marshall, and M. L. Stoll. 2002. Tactile hairs on the postcranial body in
Florida manatees: A mammalian lateral line? Brain, Behavior and Evolution 59:141–154.
Reiss, J. O. 2009. Not by design: Retiring Darwin’s watchmaker. Berkeley: University of
California Press.
Rey, M., S. Ohno, J. A. Pintor-Toro, A. Llobell, and T. Benitez. 1998. Unexpected homology
between inducible cell wall protein QID74 of filamentous fungi and BR3 salivary protein
of the insect Chironomus. Proceedings of the National Academy of Sciences of the United
Reznick, D. N., M. Mateos, and M. S. Springer. 2002. Independent origins and rapid evolu-
tion of the placenta in the fish genus Poeciliopsis. Science 298:1018–1020.
Rich, T. H., J. A. Hopson, A. M. Musser, T. F. Flannery, and P. Vickers-Rich. 2005. Indepen-
dent origins of middle ear bones in monotremes and therians. Science 307:910–914.
298 References
Robert, D., J. Amoroso, and R. R. Hoy. 1992. The evolutionary convergence of hearing in
a parasitoid fly and its cricket host. Science 258:1135–1137.
Robert, D., and U. Willi. 2000. The histological architecture of the auditory organs in the
parasitoid fly Ormia ochracea. Cell and Tissue Research 301:447–457.
Robinson, P. L. 1956. An unusual saurposid dentition. Zoological Journal of the Linnean
Society 43:283–293.
Robson, P., G. M. Wright, J. H. Youson, and F. W. Keeley. 2000. The structure and organiza-
tion of lamprin genes: Multiple-copy genes with alternative splicing and convergent evolu-
tion with insect structural proteins. Molecular Biology and Evolution 17 (11):1739–1752.
Rodríguez-Trelles, F., R. Tarrio, and F. J. Ayala. 2003. Convergent neofunctionalization by
positive Darwinian selection after ancient recurrent duplications of the xanthine dehydrog-
enase gene. Proceedings of the National Academy of Sciences of the United States of America
100 (23):13413–13417.
Roff, D. A. 1994. The evolution of flightlessness: Is history important? Evolutionary Ecology
8:639–657.
Rokas, A., and S. B. Carroll. 2008. Frequent and widespread parallel evolution of protein
sequences. Molecular Biology and Evolution 25 (9):1943–1953.
Roots, C. 2006. Flightless birds. Westport, Conn.: Greenwood Press.
Rose, R. D., and R. J. Emry. 2005. Extraordinary fossorial adaptations in the Oligocene
palaeanodonts Epoicotherium and Xenocranium (Mammalia). Journal of Morphology
175:33–56.
Rowe, T. 1993. Phylogenetic systematics and the early history of mammals. In Mammal
phylogeny: Mesozoic differentiaton, multituberculates, monotremes, early therians, and mar-
supials, ed. F. S. Szalay, M. J. Novacek, and M. C. McKenna, 129–145. Berlin: Springer-
Verlag.
Ruse, M. 2003. Darwin and design: Does evolution have a purpose? Cambridge, Mass.:
Harvard University Press.
Sage, R. F. 1999. Why C4 photosynthesis? In C4 plant biology, ed. R. F. Sage and R. K.
Monson, 3–16. San Diego: Academic Press.
Sage, R. F. 2001a. C4 plants. In Encyclopedia of biodiversity. Vol. 1, ed. S. A. Levin, 575–598.
San Diego: Academic Press.
Sage, R. F. 2001b. Environmental and evolutionary preconditions for the origin and diver-
sification of the C4 photosynthetic syndrome. Plant Biology 3:202–213.
Salehi-Ashtiani, K., and J. W. Szostak. 2001. In vitro evolution suggests multiple origins of
the hammerhead ribozyme. Nature 414:82–84.
Salvini-Plawen, L. von. 2008. Photoreception and the polyphyletic evolution of photore-
ceptors (with special reference to Mollusca). American Malacological Bulletin 26
(1/2):83–100.
Salvini-Plawen, L. von, and E. Mayr. 1977. On the evolution of photoreceptors and eyes.
Evolutionary Biology 10:207–263.
Scantlebury, M., J. R. Speakman, M. K. Oosthuizen, T. J. Roper, and N. C. Bennett. 2006.
Energetics reveal physiologically distinct castes in a eusocial mammal. Nature
440:795–797.
Schierwater, B., M. Eitel, W. Jakob, H.-J. Osigus, H. Hadrys, S. L. Dellaporta, S.-O. Koloko-
tronis, and R. DeSalle. 2009. Concatenated analysis sheds light on early metazoan evolution
and fuels a modern “urmetazoan” hypothesis. PLoS Biology 7 (1):36–44.
Schmitz, A., M. Gebhardt, and H. Schmitz. 2008. Microfluidic photomechanic infrared
receptors in a pyrophilous flat bug. Naturwissenschaften 95:455–460.
Schneider, E. L., and S. Carlquist. 1995. Vessel origins in Nymphaeaceae: Euryale and
Victoria. Botanical Journal of the Linnean Society 119:185–193.
References 299
Schultz, T. R., and S. G. Brady. 2008. Major evolutionary transitions in ant agriculture.
Proceedings of the National Academy of Sciences of the United States of America 105
(14):5435–5440.
Seelig, C. 1956. Helle Zeit—Dunkle Zeit. In Memoriam Albert Einstein. Zurich: Europa
Verlag.
Sereno, P. 1997. Psittacosauridae. In Encyclopedia of dinosaurs, ed. P. J. Currie and
K. Padian, 611–613. San Diego: Academic Press.
Sereno, P., Z. Xijin, and T. Lin. 2009. A new psittacosaur from Inner Mongolia and the
parrot-like structure and function of the psittacosaur skull. Philosophical Transactions of
the Royal Society of London: Series B, Biological Sciences, doi:10.1098/rspb.2009.0691.
Shen, Z., and M. Jacobs-Lorena. 1999. Evolution of chitin-binding proteins in invertebrates.
Journal of Molecular Evolution 48: 341–347.
Shi, Y., and S. Yokoyama. 2003. Molecular analysis of the evolutionary significance of
ultraviolet vision in vertebrates. Proceedings of the National Academy of Sciences of the
United States of America 100 (14):8308–8313.
Shingleton, A. W., and W. A. Foster. 2001. Behaviour, morphology and the division of labour
in two soldier-producing aphids. Animal Behaviour 62:671–679.
Shubin, N., C. Tabin, and S. Carroll. 2009. Deep homology and the origins of evolutionary
novelty. Nature 457:818–823.
Sinha, N. R., and E. A. Kellogg. 1996. Parallelism and diversity in multiple origins of C4
photosynthesis in the grass family. American Journal of Botany 83 (11):1458–1470.
Smith, J. D. 2009. The study of animal metacognition. Trends in Cognitive Sciences 13
(9):389–396.
Spinney, L. 2007. Back to their roots. New Scientist 194 (2608):48–51.
Stadler, B., and A. F. G. Dixon. 2005. Ecology and evolution of aphid-ant interactions.
Annual Review of Ecology Evolution and Systematics 36:345–372.
Steadman, D. W. 2006. Extinction and biogeography of tropical Pacific birds. Chicago:
University of Chicago Press.
Stein, W. E., F. Mannolini, L. V. Hernick, E. Landing, and C. M. Berry. 2007. Giant cladoxy-
lopsid trees resolve the enigma of the Earth’s earliest forest stumps at Gilboa. Nature
446:904–907.
Stensiö, E. 1963. The brain and the cranial nerves in fossil lower craniate vertebrates.
Skrifter utgitt av det Norske Videnskaps-Akademi i Oslo. I. Matematisk-Naturvidenskapelig
Klasse 13:1–120.
Stern, D. L. 2010. Evolution, development, and the predictable genome. Greenwood Village,
Colo.: Roberts and Company.
Stern, D. L., and V. Orgogozo. 2008. The loci of evolution: How predictable is genetic evolu-
tion? Evolution: International Journal of Organic Evolution 62 (9):2155–2177.
Stern, D. L., and V. Orgogozo. 2009. Is genetic evolution predictable? Nature 323:
746–751.
Stewart, A. 2009. Wicked plants. Chapel Hill: Algonquin Books.
Stewart, C. B., J. W. Schilling, and A. C. Wilson. 1987. Adaptive evolution in the stomach
lysozymes of foregut fermenters. Nature 330:401–404.
Strobel, D. F. 2010. Molecular hydrogen in Titan’s atmosphere: Implications of the mea-
sured tropospheric and thermospheric mole fractions. Icarus, doi:10.1016/j.icarus.2010.03.003.
Sumpter, D. J. T. 2006. The principles of collective animal behaviour. Philosophical Transac-
tions of the Royal Society of London: Series B, Biological Sciences B361:5–22.
Szczupider, M. 2009. Visions of Earth: Cameroon Sanaga-Yong chimpanzee rescue center.
National Geographic 216 (5):12–13.
300 References
Terada, T., O. Nureki, R. Ishitani, A. Ambrogelly, M. Ibba, D. Söll, and S. Yokoyama. 2002.
Functional convergence of two lysyl-tRNA synthetases with unrelated topologies. Nature
Structural Biology 9 (4):257–262.
Thien, L. B., P. Bernhardt, M. S. Devall, Z.-D. Chen, Y.-B. Luo, J.-J. Fan, L.-C. Yuan, and
J. H. Williams. 2009. Pollination biology of basal angiosperms (ANITA grade). American
Journal of Botany 96 (1):166–182.
Thomas, R. D. K., and W.-E. Reif. 1993. The skeleton space: A finite set of organic designs.
Evolution: International Journal of Organic Evolution 47:341–360.
Thompson, D’A. W. 1942. On growth and form. Cambridge: Cambridge University
Press.
Thompson, M. B., and B. K. Speake. 2006. A review of the evolution of viviparity in lizards:
Structure, function and physiology of the placenta. Journal of Comparative Physiology
176B:179–189.
Thorne, B. L., and J. F. A. Traniello. 2003. Comparative social biology of basal taxa of ants
and termites. Annual Review of Entomology 48:283–306.
Tomescu, A. M. F. 2008. Megaphylls, microphylls and the evolution of leaf development.
Trends in Plant Science 14 (1):5–12.
Tripp, B. C., K. Smith, and J. G. Ferry. 2001. Carbonic anhydrase: New insights for an ancient
enzyme. Journal of Biological Chemistry 276 (52):48615–48618.
Turner, A. 1997. The big cats and their fossil relatives. New York: Columbia University
Press.
van Holde, K. E., K. I. Miller, and H. Decker. 2001. Hemocyanins and invertebrate evolu-
tion. Journal of Biological Chemistry 276 (19):15563–15566.
van Schaik, C. P., E. A. Fox, and A. F. Sitompui. 1996. Manufacture and use of tools in wild
Sumatran orangutans. Naturwissenschaften 83 (4):186–188.
Van Valen, L. 1973. A new evolutionary theory. Evolutionary Theory 1:1–30.
Varricchio, D. J. 1997. Troodontidae. In Encyclopedia of dinosaurs, ed. P. J. Currie and
K. Padian, 749–754. San Diego: Academic Press.
Vermeij, G. J. 2006. Historical contingency and the purported uniqueness of evolutionary
innovations. Proceedings of the National Academy of Sciences of the United States of
America 103 (6):1804–1809.
von Frisch, K. 1974. Animal architecture. New York: Harcourt Brace Jovanovich.
Wade, N. 2006. Before the dawn: Recovering the lost history of our ancestors. New York:
Penguin Press.
Wake, M. H. 1992. Evolutionary scenarios, homology and convergence of structural spe-
cializations for vertebrate viviparity. American Zoologist 32:256–263.
Wake, M. H. 1993. Evolution of oviductal gestation in amphibians. Journal of Experimental
Zoology 266:394–413.
Wald, G., and S. Raypart. 1977. Vision in annelid worms. Science 196:1434–1439.
Ward, P. D. 2005. Life as we do not know it. New York: Viking Penguin.
Watson, T. 1993. Why some fishes are hotheads. Science 260:160–161.
Weinreich, D. M., N. F. Delaney, M. A. DePristo, and D. L. Hartl. 2006. Darwinian evolution
can follow only very few mutational paths to fitter proteins. Science 312:111–114.
Weishampel, D. B., and J. R. Horner. 1990. Hadrosauridae. In The Dinosauria, ed. D. B.
Weishampel, P. Dodson, and H. Osmólska, 534–561. Berkeley: University of California
Press.
Went, F. W. 1971. Parallel evolution. Taxon 20:197–226.
West, G. B., J. H. Brown, and B. J. Enquist. 1999. The fourth dimension of life: Fractal
geometry and allometric scaling of organisms. Science 284:1677–1679.
References 301
Whitehead, H. 2008. Social and cultural evolution in the ocean: Convergences and contrasts
with terrestrial systems. In The deep structure of biology, ed. S. Conway Morris, 143–160.
West Conshohocken, Pa.: Templeton Foundation Press.
Wicher, K. B., and E. Fries. 2007. Convergent evolution of human and bovine haptoglobin:
Partial duplication of the genes. Journal of Molecular Evolution 65:373–379.
Williams, D. S., and P. McIntyre. 1980. The principal eyes of a jumping spider have a tele-
photo component. Nature 288:578–580.
Wilson, E. O. 1980. Sociobiology: The abridged edition. Cambridge, Mass.: Harvard Univer-
sity Press.
Wilson, E. O., and B. Hölldobler. 2005. Eusociality: Origin and consequences. Proceedings
of the National Academy of Sciences of the United States of America 102 (38):
13367–13371.
Wourms, J. P., and J. Lombardi. 1992. Reflections on the evolution of piscine viviparity.
American Zoologist 32:276–293.
Wu, X.-C., H.-D. Sues, and A. Sun. 1995. A plant-eating crocodyliform reptile from the
Cretaceous of China. Nature 376:678–680.
Wyndham, J. 1951. The day of the triffids. New York: Doubleday and Company.
Yanoviak, S. P., R. Dudley, and M. Kaspart. 2005. Directed aerial descent in canopy ants.
Nature 433:624–626.
Yeager, M., and A. L. Hughes. 1999. Evolution of the mammalian MHC: Natural selection,
recombination, and convergent evolution. Immunological Reviews 167:45–58.
Yee, M. S. Y. 1999. Molecular phylogenies become functional. Trends in Ecology and Evolu-
tion 14 (5):177–178.
Yokoyama, R., and S. Yokoyama. 1990. Convergent evolution of the red- and green-like
visual pigment genes in fish, Astyanax fasciatus, and human. Proceedings of the National
Zakon, H. H., and G. A. Unguez. 1999. Development and regeneration of the electric organ.
Journal of Experimental Biology 202:1427–1434.
Zhang, J. 2003. Paleomolecular biology unravels the evolutionary mystery of vertebrate
UV vision. Proceedings of the National Academy of Sciences of the United States of America
100 (14):8045–8047.
Zhang, J. 2006. Parallel adaptive origins of digestive RNases in Asian and African leaf
monkeys. Nature Genetics 38 (7):819–823.
Zhang, J., and S. Kumar. 1997. Detection of convergent and parallel evolution at the amino
acid sequence level. Molecular Biology and Evolution 14 (5):527–536.
Zusi, L., and D. Bridge. 1981. On the slit pupil of the Black Skimmer (Rhynchops niger).
Journal of Field Ornithology 52 (4):338–340.
Index of Common Names
Aardvark, African, 142 Butterfly, monarch, 64, 234

Acacia Butterwort, 138
catclaw, 107
whistling-thorn, 108 Cactus, 113–114
Ant Camel
animal-domesticating, 226–227 dromedary, 57–58, 113
army, 236 western, 161–162
farming, 223–225, 227 Cassowary, 152–154
fire, 66 Cat
forest, 212 European wild, 35, 43, 158–159, 166
gliding, 22 false saber-tooth, 42–44, 160, 162
leaf-cutter, 224–225 marsupial, 158–159
velvet, 48 true saber-tooth, 42–44, 160, 162
Anteater Catfish, African electric, 39, 49
great scaly, 52 Centaur, 25–26, 253
South American giant, 142, 158–159, 162, Chicken, common domestic, 50
166 Chimpanzee, 211–212, 241
Aphid, 226–227, 230 Chiton, rock, 67
Armadillo, 29 Civet, palm, 36, 46
Auklet, crested, 64 Coca, 112
Aye-aye, 141 Cockroach, 52, 197
Coffee, 112
Bat Condor, California, 145
greater bulldog, 165 Corkscrew, 138
Mexican long-tongued, 147 Cormorant, flightless, 155
short-nosed fruit, 217 Coruro, Chilean, 29, 162
tent-making, 217 Cow, common domestic, 57, 181, 193, 227
Bayonet, Spanish, 107 Crab
Bear, polar, 8, 211 hermit, 52, 213
Beaver, Canadian, 161, 163, 218, 220–221 horseshoe, 193
Beetle Crayfish, red swamp, 52
ambrosia, 223, 225–226, 230 Cricket, mole, 28
blister, 64 Crocodile
bombardier, 66 gavialid, 167
fire, 37 Nile, 34, 52
long-horned, 66 Crow, New Caledonian, 211
Bison, 159–160, 166
Bladderwort, 138 Deer, red, 58, 62, 158, 160, 166
Bonobo, 211–212 Dinosaur (excluding avian dinosaurs)
Bowerbird, 217–218 allosaur, 164, 166
Bug, flat, 37 alvarezsaur, 144, 166
Burdock, greater, 127 ankylosaur, 62–64, 164, 166
304 Index of Common Names
Dinosaur (cont.) toad, 66

ceratopsid, 56–57, 59, 164, 166 weever, 66
coelophysid, 164, 166, 236 Flying lemur, 20
coelurosaur, 78, 148 Flying squirrel
dromaeosaur, 43, 148, 236 marsupial, 158–159
fabrosaur, 24 scaly-tailed, 20
hadrosaur, 56–57, 164–166 true, 20, 158–159
herrerasaur, 24, 77 Fox, red, 35
hypsilophodont, 164–166 Frog
maiasaur, 234 Costa Rican flying tree, 21
ornithomimosaur, 153 Guinea shovel-snout, 29
oviraptor, 50 poison arrow, 64
pachycephalosaur, 164, 167 tailed, 31, 83
psittacosaur, 51, 59 tree, 31
sauropod, 51, 166, 234 Wallace’s flying tree, 21
spinosaur, 164, 167
therizinosaur, 164, 167 Gecko
troodontid, 43, 148, 164, 166 flying, 21
velociraptor, 164, 166 helmeted, 34
Dodder, 139 Geosaur, 16
Dodo, 154 Goat, domestic, 61, 167
Dolphin, bottlenose, 6, 13, 36, 42, 212, 234, Gopher, pocket, 29, 162
238–241 Gorilla, 211, 213, 240–241
Dragon Grass
Komodo, 28 cogon, 107
mythical, 19–20, 25 Johnson, 108
Dragonfly, 77, 103, 140, 253 Grog, 263
Duck
Aukland Island teal, 155 Hagfish, 15
steamer, 155 Hawthorn, cockspur, 107
wood, 59 Hedgehog, European, 64, 198
Duckbill platypus, 39, 59, 66 Hellebore, stinking, 120
Heron, 213
Eagle, 213 Herring, Atlantic, 191
Earthworm, common, 52, 67 Hippopotamus, 57–58, 61
Echidna, 39, 64, 143 Honeybee, 66, 228–230, 234
Eel Honeycreeper
electric, 49 akiapola’au, 140
European, 15 ‘i’iwi, 147
Elderberry, black, 108 red-legged, 147
Elephant Honeyeater, 146
African, 166, 233, 241 Horse, modern, 57, 62, 160, 162
Asian, 212, 238–239 Human
Emu, 152–154 modern, 23, 179, 193–194, 198, 209–212,
218, 220–225, 227, 237–239, 241–242
Fern Neanderthal, 241–242
European water-clover, 116 Hummingbird, 145–146
giant water, 116 Hydrangea, 112
Finch, woodpecker, 141, 212 Hyrax, yellow-spotted, 59
Firefly, 64
Fish Ichthyosaur, 13, 167, 262, 272
flying Gunnard, 21
puffer, 64 Jellyfish
rabbit, 66 box, 70
red lion, 65 Portuguese man-of-war, 49
scorpion, 65
stickleback, 195–196 Kagu, 154
stone, 65 Kangaroo, red, 24, 59, 158, 160
Kiwi, 152–154, 156 Nettle, stinging, 107–108

Krill, Antarctic, 234 Numbat, 143, 158–159
Lamprey, sea, 15, 67, 197 Octopus

Leopard, snow, 8 common, 36, 47, 83, 193
Lily, corpse, 120 veined, 218
Lion Oilbird, 41
African, 35, 158–159, 166, 237 Orangutan, 211
marsupial, 158–159 Ostrich, 50, 149, 152–153
Lizard Ovenbird, 219
brown basilisk, 24 Owl
desert horned, 143 European tawny, 40
Florida worm, 28 Jamaican, 40
flying, 20 mottled, 40
legless, 28, 34 striped, 40
night, 34
skink, 91 Pangolin, Asian, 142, 162
thorny devil, 143–144 Parrot
Locust, desert, 233 broad-billed, 153
Loon, red-throated, 50 kakapo, 153, 156
Loris, slow, 35 swift, 147
Lungfish Pawpaw, 120
Queensland, 56 Peccary, musk-hog, 161–162
South American, 16 Pelican, 213, 236
Lynx, North American bay, 35 Penguin, king, 17, 148, 154–155, 262
Pigeon, wood, 213, 240
Magpie, 209, 238–239 Pitcher plant
Mammoth, 237 Asian, 136
Manatee Australian, 136
African, 17, 61 Pithui, hooded, 64
Florida, 74 Plesiosaur, 16, 167
Mantidfly, green, 26, 253 Pliosaur, 167
Marijuana, 112 Polecat, European, 66
Martin, 218 Porcupine
Mastodon, 162 African, 64
Millipede North American, 64
common, 66 Porpoise, harbor, 1, 3, 5, 13, 167, 262, 272
flat-backed, 66 Possum, striped, 141–142
Mistletoe, dwarf, 127 Praying mantis, 26, 253
Mole Pteranodon, 165, 167
Cape golden, 29, 157 Pterosaur, 17, 77–78, 167, 253, 262
European, 29, 153–154, 158
southern marsupial, 29, 158 Rabbit, European, 58–59, 159–160,
Monkey 161–162
Barbary macaque, 240 Rail
capuchin, 211, 240 South Island takahe, 149, 156
colobus, 57–58, 184–185 wekka, 149, 157
langur, 183 Rat
Monkshood, 108 bamboo, 29
Morning glory, 112 black, 158–159
Mosasaur, 16 kangaroo, 24
Moth, hummingbird clearwing, marsupial, 158–159
147 mole, 29, 230
Mouse Ray
fawn hopping, 24 electric, 49
harvest, 216 spotted eagle, 31
marsupial, 158–159 sting, 49
Mullet, striped, 52 Rhea, 149, 152–153
306 Index of Common Names
Rhinoceros, 6, 59, 166 Tenrec

Rockfowl, white-necked, 218 hedgehog, 42, 64, 157
shrew, 157
Scorpion Termite
land, 66, 76 compass, 219–220, 224–225
sea, 16 subterranean, 223, 224–225, 230
Seahorse, dwarf, 84 warlike, 220
Seal, harp, 35, 148, 154–155 Tit, 215
Sea lion, California, 17, 167 Toad, burrowing, 28, 31
Sea otter, 16, 211 Tobacco, 112
Sea raven, 191 Tree
Sea serpent, 16 bottle, 114
Seriema, red-legged, 46 fern, 95
Shad, gizzard, 52 horsetail, 96, 117
Shark maple, 98
angel, 31 palm, 96
great white, 13 quiver, 114
whitetip, 31 rubber, 108
Shrimp, sponge-dwelling, 230 sandbox, 127
Skimmer, black, 34 scale, 96, 117
Skunk, striped, 66 stage, 263
Sloth, giant ground, 167 stinging, 107
Smelt, American, 191 Triffid, 263
Snail Trout, rainbow, 31
auger-shell, 49 Tuatara, 157
cone-shell, 49 Tuco-tuco, Rio Negro, 29
great pond, 52 Turtle
Snake sea, 16
constrictor, 34, 37 snapping, 47
coral, 8
flying paradise, 21 Vanga, sickle-billed, 140–141
milk, 8 Venus flytrap, 137–138, 263
pit viper, 34, 37, 48 Vulture
Spider Egyptian, 211
jumping, 70 Eurasian black, 144–145, 237–238
ogre-faced, 70 North American turkey, 47, 144–145,
web-spinning, 263 237–238
Springhaas, 24
Spurge, 113–114 Walkingstick, giant, 8
Squirrel, gray, 216 Wallaby, 159–160, 181
Starfish, 84 Warbler, Tennessee, 147
Stargazer, 49 Wasp
Starling, European, 231–234 funnel, 219
Stink bird (hoatzin), 58, 183 hover, 219
Stork, 213 mud, 211, 219
Sugarbird, South African, 147 paper, 217, 219
Sumac, poison, 112 potter, 219
Sunbird, 146 Waterwheel, 137–138
Sundew, 136 Weaverbird, 216–217
Swallow, 218–219 Whale
Swan, 213 killer, 167, 242
Swiftlet, edible-nest, 41 sperm, 242
Swordfish, 1, 3, 13, 272 White-eyes, Japanese, 147
Wild dog, African, 236–237
Tailorbird, 216 Wildebeest, 233
Tapir, Brazilian, 154, 161, 163 Witch hazel, 127
Tasmanian devil, 158–159 Witch weed, 140
Wolf
European, 158, 166, 236
marsupial, 158–159
Wolverine, 158–159, 166
Wombat, 158, 160
Woodchuck, 158, 160, 161–162
Woodpecker, red-headed, 140, 142
Worms
bristle, 71, 84
velvet, 76, 84
Wren, Carolina, 216
Yellow jacket, 217

Yucca, aloe, 107
Zokor, Siberian, 29
Index of Species
Abildgaardia ovata, 104 Ambystoma talpoideum, 27

Abromeitiella lorentziana, 113 Amianthium muscaetoxicum, 111
Abrus precatorius, 110 Amischolotype hispida, 130
Abyssocottus korotneffi, 180 Amitermes meridionalis, 216
Acacia drepanolobium, 107 Amorphophallus titanium, 121–122
A. greggii, 107 Amphioctopus marginatus, 214
Acanthophis praelongus, 89 Amphipholis squamata, 86
Acanthopleura spiniger, 68 Anableps anableps, 87
Acer saccharinum, 128 Anacampseros australiana, 104
Acetes sibogae, 73 Anas aucklandica, 152
Aconitum napellus, 111 Anatotitan copei, 60
Acrobatus pygmaeus, 18 Anax junius, 18
Acrochordus granulatus, 89 Androcymbium striatum, 132
Acrosternum hilare, 65 Anguilla anguilla, 14–15
Adansonia digitata, 122 Anguis cephallonica, 88
Adelobasileus cromptoni, 54–55 Anilius scytale, 89
Aechmea fasciata, 132 Anisacanthus thurberi, 124
Aegithalos caudatus, 214 Ankylosaurus magniventris, 63, 166
Aegolius funereus, 41 Anniella pulchra, 27, 88
Aegypius monachus, 145, 235 Anolis carolinensis, 83
Aequipecten opercularis, 168 Anomalurus derbianus, 18
Aerodramus fuciphagus, 41 Anser caerulescens, 194
Aesculus glabra, 110 Antechinus stuartii, 158
Aethia cristatella, 63 Anthracotherium magnus, 163
Aetobatus narinari, 32 Antiaris toxicaria, 109
Agalychnis spurrelli, 18 Anticharis linearis, 105
Agave flexispina, 113 Apantesis ornata, 73
A. schottii, 124 Apatemys chardini, 141
A. tequilana, 109 Aphaenogaster rudis, 210
Aglaophyton major, 97 Apis mellifera, 65, 229, 232
Aix sponsa, 60 Apocynum cannabinum, 128
Alamosaurus sanjuanensis, 232 Aptenodytes patagonica, 14, 79, 152
Aldrovanda vesiculosa, 137 Apterostigma auriculatum, 223
Alianthus altissima, 128 Apteryx haasti, 150
Allionia coccinea, 124 Ara macao, 60
Allosaurus fragilis, 166 Arabis alpina, 128
Alluaudia montagnacii, 106, 113 Aradus albicornis, 33
Aloe dichotoma, 113 Aramidopsis plateni, 150
Alopias vulpinus, 79 Arca noae, 69
Amaranthus edulis, 104 Arceuthobium pusillum, 129, 139
Amargasaurus cazaui, 79 Archaeohyrax concentricus, 162
Amblyopone pallipes, 229 Archaeopteris hibernica, 95, 99
310 Index of Species
A. latifolia, 116 Bonellia viridis, 68

Archaeopteryx lithographica, 46, 78 Boreogadus saida, 190
Archilochus colubris, 146 Bos taurus, 180, 184, 192, 203
Arctium lappa, 127 Bougainvillea principis, 68
Ardea herodias, 18 Bouvardia glaberrima, 124
Ardipithecus ramidus, 23 Brachinus explodens, 65
Areca catechu, 110 Brachypodion melanocephalum, 88
Argentavis magnificens, 145 Branchiomma bombyx, 69
Argyrotheca cistellula, 68 Branchiostoma floridae, 69
Arisaema triphyllum, 109 Brocchinia reducta, 130, 137
Aristolochia grandiflora, 122 Brodiaea ida-maia, 124
Armandia brevis, 69 Bubo virginianus, 33, 41, 44
Ascaphus truei, 32, 83 Bufo bufo, 33
Asclepias syriaca, 128 Bugula neritina, 68
A. tuberosa, 111 Buteo jamaicensis, 44
Asimina triloba, 122 Byblis gigantea, 137
Asio otus, 41 Bythites islandicus, 87
Aspergillus clavatus, 6
A. oryzae, 6 Cactospiza pallida, 141, 210
Asplanchna brightwelli, 11, 68 Calamites cistiiformes, 95
Asterias rubens, 68 Calamocarpon, 117
Astragalus coccineus, 123–124 Calamus laoensis, 130
Astrapis desiderata, 73 Calathea zebrina, 132
Astrapotherium magnum, 163 Calligonum mongolicum, 104
Astrophytum asterias, 114 Calycanthus floridus, 110
Astroscopus zephyreus, 34, 45 Camaroptera brevicondata, 214
Astyanax mexicanus, 194 Camelops hesternus, 162
Atlantisia rogersi, 150 Camelus dromedarius, 58, 113, 200
Atriplex canescens, 104 Camponotus heathi, 19
Atropa belladonna, 111 Canirallus cuvieri aldabranus, 150
Augochlorella pomoniella, 229 Canis lupus, 158, 166, 235
Aurelia aurita, 69 Cannabis sativa, 109
Ausktribosphenos nyktos, 53–54 Capra aegagrus, 61
Australopithecus afarensis, 210 C. hircus, 167
Australoplatypus incompertus, 223, 229 Capsicum chinense, 110
Azygocypridina lowryi, 69 Carcharodon carcharias, 13–14
Cardinalis cardinalis, 238
Bachia oxyrhinus, 26, 28 Cardiocrinum giganteum, 130
Bacillus subtilis, 203 Carex pensylvanica, 132
Bakonydraco galaczi, 167 Cariama cristata, 44, 46, 150
Balanerpeton woodi, 73 Carpobrotus edulis, 113
Balanoglossus proterogonius, 68 Carybdea marsupialis, 70
Barbourofelis fricki, 42, 44 Cassytha filiformis, 139
Barclaya kunstleri, 98 Castor canadensis, 163, 215
Barisia imbricata, 88 Casuarius casuarius, 151
Barsostrobus famennensis, 116 Catha edulis, 109
Basiliscus basiluscus, 24 Cathartes aura, 44, 145, 235
Batrachognathus volans, 167 Caudipteryx zoui, 50–51
Behnia reticulata, 130 Caulophyllum thalictroides, 110
Bienertia sinuspersici, 104 Cebus apella, 180, 240
Bison bison, 159, 166 C. libidinosus, 210
Bispira volutacornis, 69 Cecropsis luculentum, 116
Bitis gabonica, 89 Cellvibrio japonicus, 201–202
Blandfordia punicea, 130 Cemophora coccinea, 27
Blepharis ciliaris, 105 Cephalanthus occidentalis, 111
Boerhavia coccinea, 104 Cephalotes atratus, 19
Bombyx mori, 197, 203 Cephalotus follicularis, 137
Cerbera odollam, 110 Costus barbatus, 132

Cervus elaphus, 58, 61, 158, 166 Cottocomephorus inermis, 180
Chalcides chalcides, 90 Crassula ovata, 113
Chalcoparia singalensis, 146 Crataegus crus-galli, 107
Chaleuria cirrosa, 116 Crocodylus niloticus, 32, 51, 73, 83, 200
Chamaecyce forbesii, 104 Crotalus horridus, 32–33, 45
Chamaeleo bitaeniatus, 88 Cryptasterina hystera, 86
Chamaesaura anguina, 88 Cryptocleidus oxoniensis, 167
Charcharhinus longimanus, 32 Cryptomys damarensis, 229
Chasmosaurus belli, 54 C. hottentotus, 27
Chelidonium majus, 109 Crysophila warscewiczii, 121
Chelonia mydas, 14 Cseninus lecontei, 223
Chelychelynechen quassus, 151 Ctenomys rionegrensis, 27
Chendytes lawi, 152 Cubitermes fungifaber, 220
Chimaerasuchus paradoxus, 54–56 Cucurbita maxima, 98
Chironomus tentans, 197 Curculigo capitulata, 130, 132
Chlamydera nuchalis, 214 Cuscuta epithymum, 139
Chlamydomonas reinhardtii, 100 Cyanastrum cordifolium, 132
Chlamydophorus truncatus, 27 Cyanerpes cyaneus, 146
Choeronycteris mexicana, 146 Cyanolimnus cerverai, 151
Chondrodendron tomentosum, 109 Cyathea cooperi, 95
Choristoneura fumiferana, 190 Cycas media, 95
Chroomonas salina, 100 C. revoluta, 111
Chrysochloris asiatica, 27, 157 Cyclanthus bipartitus, 121
Chrysomela varians, 86 Cygnus olor, 214
Chrysopa carnea, 73 Cynocephalus volans, 18
Chrysopelea paradisi, 18 Cynoglossum officinale, 128
Ciccaba virgata, 40 Cynopterus sphinx, 214
Cicuta maculata, 111 Cyperus papyrus, 104
Cincindela marutha, 73 Cypripedium calceolus, 109
Ciona intestinalis, 187, 200
Cirsium vulgare, 106 Daboia russelli, 89
Citrus aurantifolia, 110 Dactylopsila trivirgata, 141
Cleome gynandra, 105 Dactylopterus volitans, 19
Clupea harengus, 190, 232 Daemonorops formicaria, 107
Clusia rosea, 121 Dananus plexippus, 63, 232
Cnidoscollus stimulosus, 106 Daphne mezereum, 111
Coelophysis baurri, 236 Dasyatis brevicaudata, 45
C. rhodesiensis, 166 Dasypogon hookeri, 132
Coelurosauravus jaekeli, 18 Dasypus novemcinctus, 63, 83
Coereba flaveola, 194 Dasyurus viverrinus, 158
Coffea arabica, 110 Daubentonia madagascariensis, 141
Colchicum autumnale, 111 Daucus carota, 129
Colobus badius, 58 Deinacrida heteracantha, 157
C. guereza, 184 Deinonychus antirrhopus, 235–236
Columba livia, 240 Delphinium cardinale, 123–124
C. palumbus, 214 D. nudicaule, 123
Comatilia iridometriformis, 86 Delphyodontus dacriformes, 86
Comophorus baicalensis, 87 Dendrobates leucomelas, 63
Connochaetes taurinus, 232 Dendrocnide moroides, 107
Conus striatus, 45 Dendrodoris grandiflora, 63
Cooksonia hemisphaerica, 97 Descurainia pinnata, 109
Cophotis ceylanica, 88 Desmodium lineatum, 128
Corallorhiza maculata, 139 Diatryma gigantea, 149–150
Cormocephalus rubriceps, 157 Dichapetalum toxicarium, 109
Corvus moneduloides, 210 Didelphis virginiana, 54, 89
Corytophanes cristatus, 88 Digitalis purpurea, 111
Dimetrodon grandis, 79 Erythroxylum coca, 109

Dinematichthys ilucoeteoides, 87 Eryx jaculus, 89
Dinopsis subrufus, 70 Escherichia coli, 201, 204
Dinornis giganteus, 153 Eudibamus cursoris, 23–24
Dionaea muscipula, 137 Euglena gracilis, 100
Diospyros malabarica, 121 Eukrohnia hamata, 69
Diploptera punctata, 86 Eumecia anchietae, 90
Dipodomys californicus, 23 Eupagurus bernhardus, 51–52
Diprotodon optatum, 159 Eupatorium rugosum, 110
Disporum cantoniense, 130 Euphausia superba, 232
Dissostichus mawsoni, 190 Euphorbia canariensis, 106, 113–114
Doedicurus clavicaudatus, 63, 166 E. cryptospinosa, 114
Dolichopoda linderi, 204 E. mellifera, 114
Dolichovespula norwegica, 214 E. obesa, 114
Dorosoma cepedianum, 51 E. tirucalli, 109
Draco melanopogon, 18, 20 E. valida, 114
Dromaius novaehollandiae, 151 Euroleon nostras, 45
Drosera intermedia, 137 Eurotamandua joresi, 143
D. pauciflora, 121 Eurypterus remipes, 14
Drosophila melanogaster, 200 Exocetus volitans, 19
Dryocoetoides cristatus, 223
Falco peregrinus, 44
Ecballium elaterium, 110, 129 Falculea palliata, 141
Echinocereus triglochidiatus, 124 Felis sylvestris, 35, 158, 166
Echinoderes aquilonius, 69 F. sylvestris catus, 32–33
Echinops telfairi, 41 Ferocactus cylindraceus, 114
Eciton burchelli, 235 Ficus carica, 107
Edaphosaurus cruciger, 79 Flagellaria indica, 130
Electrophorus electricus, 45 Flaveria australasica, 105
Eleocharis vivipara, 104 Formica obscuripes, 226
Elephantulus rozeti, 200 F. yessensis, 223
Elephas maximus, 210, 239 Fouquieria splendens, 124
Eleutherodactylus jasperi, 87 Fraxinus americana, 128
Elkinsia polymorpha, 116–117 Fringilla coelebs, 215
Ellesmeris sphenopteroides, 99 Froelichia gracilis, 104
Encephalartos ferox, 106 Fucomimus mus, 87
Enhydra lutris, 16, 210 Fucus serratus, 100
Eoconodontus notchpeakensis, 15, 70 Fugu rubripes, 187
Eomaia scansoria, 90 Fulica gigantea, 151
Eoraptor, 24 Furnarius rufus, 215
Ephedra fasciculata, 98
Epicauta pestifera, 63 Gagea lutea, 132
Epilobium angustifolium, 128 Galago senegalensis, 33
Epoicotherium unicum, 27 Galeus polli, 87
Equisetum sylvaticum, 98 Galium odoratum, 110
Equus caballus, 61 Gallinula nesiotis comeri, 151
E. zebra, 58, 61 G. pacificus, 151
Eremias przewalskii, 88 G. silvestris, 151
Eremopezus eocaenus, 151 Gallirallus australis australis, 150
Erethizon dorsatum, 63 G. australis grayi, 150
Eriachne triodioides, 104 G. australis scotti, 150
Eridacnis barbouri, 87 G. calayanensis, 150
Erinaceus europaeus, 63, 197 G. insignis, 150
Erpetoichthyes calabaricus, 14 G. lafresnayus, 150
Erwinia chrysanthemi, 201–202 Gallus gallus, 50, 180, 200
Erycibe paniculata, 121 Gasterochisma melampus, 79
Erythronium americanum, 132 Gasterosteus aculeatus, 195–196
Gavia stellata, 50 Heterandria formosa, 90

Gavialis gangeticus, 167 Heterocephalus glaber, 229
Geitonoplesium cymosum, 130 Heterococcus pleurococcoides, 100
Gelsemium sempervirens, 111 Heterohyrax brucei, 58
Genlisea margaretae, 137 Hexanchus griseus, 86
Geomys bursarius, 27, 162 Hippocampus zosterae, 84
Geosaurus giganteus, 14 Hippopotamus amphibius, 58
Geranium maculatum, 129 Hirundo pyrrhonota, 215
Giardinichthys viviparous, 87 Holocentrus adscensionis, 33
Gigantopteris nicotianaefolia, 98 Homo erectus, 242
Gilia aggregata, 123–124 H. ergaster, 242
Gillespiea randolphensis, 116 H. habilis, 210
Ginkgo biloba, 127 H. heidelbergensis, 242
Giraffa camelopardalis, 23 H. neanderthalensis, 241–242
Glaucomys volans, 18, 20, 158 H. sapiens, 180, 192, 197, 201, 203, 215,
Gleditsia triacanthos, 106 222–223, 232, 235, 239, 241–242
Globba winitii, 132 Hoplophoneus mentalis, 42, 44
Glossocardia bosvallea, 105 Hordeum murinum, 128
Gnathonemus petersii, 34 Hura crepitans, 129
Gnathotrupes quadrituberculatus, 223 Hyaena hyaena, 145
Gollum attenuatus, 87 Hydnophytum formicarum, 107
Gorilla gorilla, 211, 214, 241 Hydrangea quercifolia, 111
Griffinia hyacinthina, 130 Hydrilla verticillata, 104
Gromphadorhina portentosa, 51 Hyla cinerea, 33
Gryllotalpa gryllotalpa, 27 Hyperodapedon gordoni, 60
Gryllus pennsylvanicus, 73 Hypsilophodon foxii, 166
Gulo gulo, 158, 166 Hystrix cristata, 63
Gymnogyps californicus, 145
Gymnostachys anceps, 130 Ibicella lutea, 137
Gymnostinops montezuma, 214 Icarosaurus siefkeri, 18, 20
Gymnotus carapo, 34 Ichthyophis glutinosus, 26
Ichthyosaurus platyodon, 13–14, 88, 167
Habroptila wallaci, 151 Idiognathodus claviformis, 54, 56
Hadrurus arizonensis, 45 Ilex aquifolium, 106
Halaelurus boesemani, 87 Impatiens pallida, 129
Haliaeetus leucocephalus, 214 Imperata cylindrical, 106
Halictus (Halictus) quadricinctus, 229 Indocryphalus pubipennis, 223
Haloxylon aphyllum, 104 Iodes africana, 121
Hamamelis virginiana, 129 Ipomoea coccinea, 123–124
Hanguana malayana, 130 I. tricolor, 110
Harpagophytum procumbens, 128 Ipomopsis congesta, 121
Hedera helix, 110 Iris foetidissima, 122
Helicolenus percoides, 87 I. pseudacorus, 132
Heliconius erato, 73
Heliotropium polyphyllum, 105 Jenynsia lineata, 87
Helix aspersa, 69 Joinvillea ascendens, 130
Helleborus foetidus, 122
Hemaris thysbe, 146 Kallstroemia grandiflora, 105
Hemicyclaspis murchisoni, 33 Kalmia latifolia, 111
Hemigalus derbyanus, 44 Karwinskia humboldtiana, 110
Hemignathus munroi, 141 Kayentatherium wellesi, 54–55
Hemimerus vosseleri, 86 Keichousaurus hui, 88
Hemisus guineensis, 27 Kryptobaatar daszevegi, 89
Hemitripterus americanus, 190 Kuehneosaurus latus, 18, 20
Hensmannia turbinata, 132
Herrerasaurus ischigualastensis, 23–24 Lacerta vivipara, 88
Hesperornis regalis, 151 Lactobacillus helveticus, 201
Lama guanicoe, 61 Mammuthus primigenius, 237

Lamna ditropis, 79 Manihot esculenta, 111
Lampropeltis triangulum celaenops, 8 Manis gigantea, 51, 63, 143, 162
Lantana camara, 110 Mantis religiosa, 26, 69, 73
Larentia clavaria, 73 Marasuchus illoensis, 23–24
Larus argentatus, 180 Marmota monax, 158, 162
Lasioglossum (Dialictus) figueresi, 229 Marsilea quadrifolia, 116
Lasius fuliginosus, 226 Massetognathus pascuali, 54
Lathamus discolor, 146 Mastixia arborea, 121
Latimeria chalumnae, 87 Materpiscis attenboroughi, 84, 86
Latrodectus mactans, 45 Megacrex inepta, 150
Lecanium hesperidum, 86 Megalomma vesiculosum, 69
Leclercqia andrewsii, 99 Megaphasma dentricus, 8, 23
Lepidocarpon, 117 Megathrips lativentris, 86
Lepidodendron rhodumnense, 95 Melanerpes erythrocephalus, 141
Lepidosiren paradoxa, 14, 16 Melanophila acuminata, 33
Leptocharius smithii, 87 Mellipona compressipes, 229
Leptosynapta clarki, 86 Melopsittacus undulates, 180
Leptothorax albipennis, 232 Mephitis mephitis, 65
Lesothosaurus, 24 Merimna atrata, 33
Leuckartiara octona, 11, 68 Mesotherium cristatum, 163
Leucojum aestivum, 132 Methanosarcina thermophila, 203
Lialis burtonis, 26, 32 Metriaclima zebra, 180
Limulus polyphemus, 192 Microgale longicaudata, 157
Lineus ruber, 69 Micromys minutus, 214
Liostenogaster flavolineata, 215 Microstigmus comes, 214
Liriodendron tulipifera, 128 Microstomum lineare, 68
Lobelia cardinalis, 124 Microtus arvalis, 27
L. inflata, 110 Micruroides euryxanthus, 8
Lolliguncula brevis, 73 Mimulus cardinalis, 123–124
Lonicera arizonica, 124 Mitsukurina owstoni, 86
Lophophora williamsii, 111 Mollugo nudicaulis, 104
Lophopssittacus mauritianus, 151 Moloch horridus, 143
Loxodonta africana, 166, 232, 241 Monardella macrantha, 123–124
Loxosomella cochlear, 68 Monosiga brevicollis, 260
Lumbricus terrestris, 51, 68, 192 Monotropa uniflora, 139
Lycaon pictus, 235 Moraea glaucopis, 121
Lymnaea stagnalis, 51 Moricandia arvensis, 105
Lynx rufus, 35 Moringa pterygosperma, 128
Mucuna pruriens, 106
Mabuya heathi, 90–91 Mugil cephalus, 51
Macaca sylvanus, 239 Mullerornis agilis, 151
Machaeroides simpsoni, 42 Muraenosaurus leedsii, 14
Macrauchenia patagonica, 162 Mus musculus, 158
Macroclemys temmincki, 44, 83 Mustela putorius, 65
Macropus agilis, 159 Mutisia, 125
M. eugeni, 180 Mya arenaria, 68
M. rufus, 23, 58, 158, 200 Myospalax myospalax, 27
Macrosoma nigrimacula, 73 Myotis myotis, 18, 41, 73, 167
Macrotermes bellicosus, 215 Myoxocephalus octodecimspinosus, 190
Macrozanonia macrocarpa, 126, 128 Myrmecobius fasciatus, 143, 158
Macrozoarces americanus, 190 Myrmecophaga tridactyla, 143, 158, 162, 166
Magnolia macrophylla, 122 Myxine glutinosa, 15
Maiasaura pebblesorum, 232 Myzomela eques, 146
Majungatholus atopus, 77
Malapterurus electricus, 34, 45 Nannopterum harrisi, 152
Mammut americanum, 162 Narceus annularis, 65
Nasutitermes corniger, 65 Pagophilus groenlandicus, 32

Naultinus gemmeus, 88 Pakasuchus kapilimai, 54–55
Nautilus pompilius, 47 Palaeotis weigelti, 150
Necrophila americana, 145 Pallavicinites devonicus, 99
Nectophynoides occidentalis, 87 Palola viridis, 68
Nelumbo nucifera, 122 Pan paniscus, 211
Neoceratodus forsteri, 54 P. troglodytes, 211, 241
Neophron percnopterus, 210 Pandanus tectorius, 130
Neotoma fuscipes, 214 Pandion haliaetus, 44, 47
Nepenthes truncata, 137 Panicum virgatum, 104, 132
Nephthys ciliata, 69 Paniscus testaceus, 86
Nereis limnicola, 86 Panthera leo, 35, 158, 166, 235
Nesoclopeus woodfordi, 151 Papaver rhoeas, 121
Neuwiedia veratrifolia, 130 P. somniferum, 110
Nicotiana tabacum, 109, 203 Papio cynocephalus, 210
Nidularium fulgens, 132 Parabrotula plagiophthalmus, 87
Noctilio leporinus, 165 Paradoxurus hermaphroditus, 33
Noeggerathiostrobus vicinalis, 116 Paragordius varius, 69
Notanatolica vivipara, 86 Paralastor emarginatus, 219
Notaulax rectangulata, 68 Paraleptomenes mephitis, 215
Nothronychus mckinleyi, 167 Parasaurolophus walkeri, 54, 166
Nothrotheriops shastensis, 167 Parischnogaster mellyi, 229
Notomys cervinus, 23 Paryphanta busbyi, 157
Notophthalamus viridescens, 204 Patagopteryx deferrariisi, 151
Notornis mantelli hochsteteri, 151 Paulownia tomentosa, 128
Notoryctes typhlops, 27, 158 Pavlova helicata, 100
Nycticebus coucang, 32, 35 Pecari tajacu, 162
Pecten maximus, 69
Ochromonas villosa, 100 Pectinaria gouldii, 69
Octopus vulgaris, 32–33, 44, 70, 192, 200 Pectis linifolia, 105
Odontotermes formosanus, 223 Pedetes capensis, 23–24
Oncholaimus vesicarius, 68 Pediculus corporis, 86
Oncorhychus mykiss, 31 Pelecanus erythrorhynchus, 235
Oncothrips tepperi, 229 P. occidentalis, 167
Onitochiton neglectus, 69 Peniocereus serpintinus, 114
Onychocerus albitarsis, 65–66 Pereskia sacharosa, 114
Ophiodes striatus, 88 Peridinium westii, 100
Ophisaurus ventralis, 27 Peripatus antiguensis, 69, 86
Opisthocomus hoatzin, 58, 184 Periplaneta americana, 197
Orchidantha maxillarioides, 121 Petauroides volans, 18
Orcinus orca, 167 Petaurus australis, 158
Ormia ochracea, 73 P. breviceps, 18
Ornithogalum umbellatum, 121 Petromyzon marinus, 15, 68, 197, 200
Ornithorhynchus anatinus, 34, 60, 65, 73, Peucedanum galbanum, 110
200 Pezophaps solitaria, 151
Ornithosuchus longidens, 23 Phacelia purshii, 106
Orthotomus sutorius, 214 Phascolomys ursinus, 158
Orycteropus afer, 143 Phelsuma lineata, 200
Oryctolagus cuniculus, 58, 159, 200 Philodendron selloum, 111
Oryza sativa, 187 Phlegethonia linearis, 26, 28
Osmerus mordax, 190 Phocaena phocaena, 13–14, 167
Ouranosaurus nigeriensis, 79 Phodilus badius, 41
Oxalis corniculata, 129 Phoenix dactylifera, 95
Oxyura vittata, 83 Phorusrhacos longissimus, 46, 48, 148
Photinus pyralis, 63
Pachycephalosaurus wyomingensis, 167 Phragmipedium grande, 122
Pachypodium lealii, 106, 113 Phrynocephalus arabicus, 88
Phrynosoma ditmarsi, 88 Psaronius schopfii, 95

P. orbiculare, 88 Pseudemoia entrecasteauxii, 90
P. platyrhinos, 143 Pseudoginglymostoma brevicaudatum, 86
Phyllomedusa tarsius, 32 Pseudomymex gracilis, 19
Phymaturus patagonicus, 88 Pseudopotamilla occelata, 69
Physalia physalis, 45 Pseudoregma sundanica, 229
Phytolacca americana, 111 Pseudoscops grammicus, 40
Pica pica, 239 Pseudotribos robustus, 54
Picathartes gymnocephalus, 215 Psilotum nudum, 99
Pilocereus lanuginosus, 106, 113–114 Psittacosaurus mongoliensis, 51, 60
Pinguicula vulgaris, 137 Psittacus erithacus, 180
Pinus strobus, 128 Ptelea trifoliata, 128
Pisum sativum, 203 Pteranodon longiceps, 167
Pitohui dichrous, 63 Pteridium aquilinum, 98
Placopecten magellanicus, 200 Pteroglossaspis ecristata, 121
Platecarpus ictericus, 14, 88 Pterois volitans, 65
Plateosaurus engelhardi, 51 Pterotrachea mutica, 69
Platyhystrix rugosus, 79 Ptychozoon kuhli, 19
Platymonas subcordiformis, 100 Pygathrix nemaeus, 184
Platyzoma microphylla, 116 Pyrococcus horikoshii, 204
Pleuronectes americanus, 190 Pyrotherium romeri, 162
Pliohippus pernix, 160, 162 Python molurus, 32–33
Ploceus cucullatus, 214
Plotosus lineatus, 65 Quercus lobata, 107
Podiceps taczanowskii, 152 Q. velutina, 106
Podilymbus gigas, 152
Podophyllum peltatum, 109 Rafflesia arnoldii, 122, 139
Poeciliopsis paucimaculata, 90 Rahonavis ostromi, 43–44, 46
P. prolifica, 90 Raja ocellata, 34
P. retropinna, 91 Rallus okinawa, 150
P. turneri, 90 R. owstoni, 150
Pogona vitticepes, 184 R. pectoralis muelleri, 150
Polycarpaea longiflora, 104 Rana catesbeiana, 33
Polydesmus angustus, 65 R. pipiens, 200
Polydriella helvetica, 100 Ranunculus asiaticus, 121
Polygonatum biflorum, 130 Raphus cucullatus, 151
Polygonum pennsylvanicum, 106 Rattus norvegicus, 197
Pomacea bridgesii, 83 R. rattus, 158
Pongo pygmaeus, 210–211, 239 Rellimia thomsonii, 99
Populus deltoides, 128 Remiz pendulinus, 214
Portia fimbriata, 70 Rhacochilus vacca, 87
Portulaca oleracea, 104 Rhacophorus nigropalmatus, 18
Porzana atra, 150 Rhamnus cathartica, 106
Potamogale velox, 157 Rhea americana, 150, 180
Presbytis entellus, 184 Rhineura floridana, 26, 88
Priapulus caudatus, 51, 192 Rhinoceras unicornus, 166
Pristerodon mackayi, 60 Rhinophrynus dorsalis, 27, 32
Proailurus lemanensis, 44 Rhinoptynx clamator, 40
Proboscidea louisianaca, 129 Rhizomys sumatraensis, 27
Procambarus clarkii, 52 Rhododendron ponticum, 109
Procolophon trigoniceps, 54 Rhomaleosaurus megacephalus, 167
Proiphys amboinensis, 130 Rhynchetos jubatus, 150
Promerops cafer, 146 Rhynchops niger, 32, 34
Propachyrucos ameghinorum, 162 Rhynchospora rubra, 104
Protobarinophyton pennsylvanicum, 116 Ribes speciosum, 123–124
Protocalamostachys farringtonii, 116 Ricinus communis, 110
Psaltiparus minimus, 215 Ripogonum scandens, 130
Robinia pseudoacacia, 109 Spiloxene capensis, 121

Rollandia microptera, 152 Spinosaurus maroccanus, 79, 167
Roridula gorgonias, 137 Sporobolus cryptandrus, 104
Rosa gallica, 106 Sporogonites exuberans, 99
Rousettus aegyptiacus, 167 Squatina californica, 31–32
Rumex crispus, 128 Stapelia gigantea, 122
Steatornis caripensis, 41
Sabella spallanzanii, 168 Stercorarius parasiticus, 194
Sagitta enflata, 69 Sterculia stipulata, 121
Sagittarius serpentarius, 44 Stomaphis quercus, 226
Salamandra salamandra, 63, 87 Storeria dekayi, 89
Salpa cylindrica, 69 Striga asiatica, 139
Salvia divinorum, 109 Strigops habroptilus, 151
Salvinia molesta, 116 Strix aluco, 40
Sambucus nigra, 110 Strophanthus kombe, 110
Sanguinaria canadensis, 111 Struthio camelus, 50–51, 83, 150
Saponaria officinalis, 109 Strychnos toxifera, 109
Sarcophaga carnaria, 86 Sturnus vulgaris, 232
Sarcophilus harrisii, 33, 158 Suaeda aralocaspica, 104
Sarracenia purpurea, 137 Sueus niisimai, 223
Saurornithoides mongoliensis, 44, 166 Sutorectus tentaculatus, 86
Scalenodon angustifrons, 54 Sylvilagus floridanus, 162
Scaphiopus couchii, 27, 32 Synalpheus chacei, 229
Sceliphron caementarium, 210 S. “paraneptunus small,” 229
Sceloporus angustus, 88 S. regalis, 229
S. bicanthalis, 88 Synanceia horrida, 64
S. formosus, 88 Synaptula lamperti, 11, 68
S. torquatus, 88
Schistocerca gregaria, 232 Tabernanthe iboga, 111
Sciurus carolinensis, 214 Tacca integrifolia, 130
Scolecomorphus vittatus, 87 Tachyeres brachypterus, 151
Scoliopus bigelovii, 132 Tachyglossus aculeatus, 34, 63
Scolopendra gigantea, 45 Taeniopygia guttata, 180
Scolytoplatypus macgregori, 223 Takifugu vermicularis, 63
Selaginella lepidophylla, 98 Talpa europaea, 27, 158
Senecio stapeliiformis, 113 Tapirus californicus, 163
Sesuvium portulacastrum, 104 Taraxacum officinale, 128
Setifer setosus, 63, 157 Tarentola chazaliae, 32
Shuvuuia deserti, 143–144, 166 Tarsius syrichta, 33
Siganus corallinus, 65 Taxus baccata, 111
Silene laciniata, 123–124 Teinolophus trusleri, 74
Siphonochilus aethiopicus, 130 Telangium bifidum, 99
Sipunculus nudus, 68, 192 Tempskya dernbachii, 95
Sistrurus catenatus, 89 Tenebrio molitor, 190
Smilax herbacea, 122 Tenontosaurus tilleti, 236
Smilodon fatalis, 43–44, 162 Terebra babylonia, 45
Sminthopsis crassicandata, 158 Terebratulina retusa, 192
Solanum tuberosum, 111 Termitomyces, 225
Solenopsis geminata, 65 Terrapene carolina, 63, 73
Sordes pilosus, 18, 79 Thalarctos maritimus, 8, 79
Sorex araneus, 41 Thalassophryne megalops, 65
Sorghum halepense, 109 Thea sinensis, 109
Spalacopus cyanus, 27, 162 Theobroma cacao, 110
Spalax leucodon, 27 Thermonectes marmoratus, 14
Sphecius speciosus, 45 Thoatherium minusculum, 61, 160, 162
Sphenodon punctatus, 157 Thomashuxleya rostrata, 162
Sphenophyllum involutum, 99 Thryothorus ludovicianus, 216
Thunnus alalunga, 79 Victoria amazonica, 98, 122

Thylacinus cynocephalus, 158 Vipera berus, 89
Thylacoleo carnifex, 158 Volaticotherium antiquum, 18
Thylacosmilus atrox, 43–44, 162 Vulpes vulpes, 32, 35
Tiaropsis multicirrata, 69
Tibicen dorsata, 73 Wahlenbergia gymnoclada, 121
Tinea vivipara, 86 Wattieza (Eospermatopteris) erianus, 94–95
Titanosaurus madagascariensis, 166 Wolozynskia coronata, 100
Titanus walleri, 148, 150
Tornaria ancoratae, 68 Xanthium strumarium, 129
Torpedo marmorata, 33, 45 Xantusia arizonae, 88
Torrea candida, 70 X. henshawi, 32, 73
Toxicodendron vernix, 109 Xenocranium pileorivale, 27
Toxodon platensis, 163 Xenomedea rhodopyga, 87
Trachinus vipera, 65 Xenopus laevis, 180
Tribonyx mortieri, 151 Xenosaurus platyceps, 88
Triceratops albertensis, 166 Xiphias gladius, 13–14, 79
Trichechus manatus, 73
T. senegalensis, 14, 61 Yucca aloifolia, 106
Trichocereus pachanoi, 109
Trichoderma harzianum, 197 Zaglossus bruijni, 143
Tricholimnas sylvestris, 150 Zalophus californianus, 14, 167
Trichoplax adhaerens, 81 Zanthoxylum americanum, 106
Tricuspisaurus thomasi, 56 Zauschneria californica, 124
Trigonopsis cameronii, 215 Zea mays, 104
Trillium grandiflorum, 130, 132 Zenarchopterus gilli, 87
Trilophosaurus buettneri, 54, 56 Zeugomantispa minuta, 26
Trimerosurus albolabris, 89 Zigadenus venenosus, 111
Tripedalia cystophora, 200 Zoarces viviparous, 87
Triphyophyllum peltatum, 136 Zootermopsis nevadensis, 229
Triticum aestivum, 203 Zosterops japonicus, 146
Tropidophis feicki, 89 Zygophyllum simplex, 105
Tulipa agenensis, 121
Tursiops truncatus, 13, 33, 41, 210, 232, 239,
241
Typhlonectes compressicauda, 83, 87
Typhlops diardi, 89, 184
Tyto alba, 41
Ulex europaeus, 129

Ulmus alata, 128
Ululodes quadripunctatus, 45
Uncia uncia, 8
Uria aalge, 152
Uroderma bilobatum, 214
Ursus maritimus, 210
Urtica dioica, 106
Utricularia inflata, 137
Uvularia grandiflora, 132
Varanus komodensis, 28
Vatica parvifolia, 121
Velociraptor mongoliensis, 44, 166
Veratrum viride, 109
Vermiuora peregrina, 146
Vespula maculifrons, 229
Vestiaria coccinea, 146
Index of Topics
Acetylene digestion, 255 β-elimination, 201–202

Adaptive improvement, 267–268, 272–273 Biological form
Adriatic ecosystems, 163 convergent (see Convergent evolution)
Agricultural limitations, 227 impossible, 250–252
Agriculture, 222–228 nonexistent, 12, 16, 19, 21, 25, 29, 38, 66,
Aldehyde oxidoreductase, 186–187 96, 103, 170–171, 174–175, 250–254,
Animals 260–264
armored, 63–64 Bipedalism, 23–25
carnivores, 30–49 Bird-pollinated flowers, 123–125
convergent, 13–91 Birds, ground-dwelling, 147–155
herbivores, 50–60 Breathing systems, 75–77
locomotion and, 13–30 Burrowing forms, 26–29
poisonous, 45, 48–49, 64–66
reproduction, 81–91 C4 photosynthesis, 101–105
Antidehydration Cactus forms, 112–114
cactus forms and, 112–114 Camera eyes, 38, 70–72
succulent structures and, 112–114 Camouflage, 8–9
Antifreeze proteins, 189–192 Cantharophilous flowers, 119–121
Arborescence, 93–97 Carbon-based life, 254–256
Architectural limitations, 222 Carnivores
Architecture, 212–222 animals, 30–49
Armor plants, 135–138
chemical, 63–64 Carrion-mimic flowers, 120–122
spinose, 63–64 Cecal fermentation, 57–59
tanklike, 63–64 Cellular structures, 205–206
Art, 218 Cellulose digestion, 57–59
Astrobiology, 254, 265–266, 276 Chemical armor, 63–64
Asymmetric ears, 40–41 Chitin digestion, 202–203
Australian ecosystems, 157–160 Claws, retractable, 43–47
CO2 hydration, 202–203
Beaks Compound eyes, 38, 69–72
ducklike, 59–60 Constraints. See Evolutionary constraints
parrotlike, 59–60 Contingency, 265–277
raptorial, 44, 47–48 Convergent evolution
Beetle-pollinated flowers, 119–121 animals and, 13–91
Behavior ecosystems and, 135–175
agricultural, 222–228 minds and, 209–243
architectural, 212–222 molecules and, 177–207
artistic, 218 plants and, 93–133
eusocial, 228–231 recognition of, 1–5
herding, 231–235 theoretical morphospaces and, 246–252
hunting, 235–238 Crystallin proteins, 198–200
320 Index of Topics
“Deep homology,” 6, 38, 71, 101, 257–260 Endothermy, 77–80

Defense Enlarged eyes, 33, 36
antidehydration, 112–114 Enzyme functions
antiherbivore, 105–112 β-elimination, 201–202
antipredator, 60–66 chitin digestion, 202–203
Determinism, 265–277 CO2 hydration, 202–203
Developmental constraint, 7–8, 19, 21, 25, hammerhead ribozyme, 204–205
29, 38, 46, 66, 71, 77, 103, 245–262 lysyl-tRNA synthesis, 203–204
Developmental inertia, 25 peptide-bond hydrolysis, 202–203
Digestion sugar phosphorylation, 201–202
acetylene and, 255 Eusociality, 228–231
cecal-fermentation, 57–59 Evolution
cellulose and, 57–59 convergent (see Convergent
chitin and, 202–203 evolution)
digestive proteins, 183–185 parallel, 3–7, 71, 81, 178, 183, 188
gastric mills and, 50–53 predictability of, 265–277
ruminate, 57–59 reverse, 4–5
stomachal-fermentation, 57–59 unique, 246
Directionality of evolution, 265–277 Evolutionary constraints
DNA developmental, 7–8, 19, 21, 25, 29, 38, 46,
aldehyde oxidoreductase–encoding, 66, 71, 77, 103, 245–262
186–187 extrinsic, 252
cis-regulatory, 188 functional, 7–8, 13, 17, 22, 25, 29, 38, 46,
mitochondrial, 185–186 66, 96, 125, 245–264
opsin-encoding, 179–183 intrinsic, 253
pancreatic ribonuclease–encoding, Evolutionary development (evo-devo), 7
184–185 Exploding seeds, 127, 129
rhodopsin-encoding, 179–183 Extraterrestrial life
stomach lysozyme–encoding, 183–184 developmental constraint and, 258–260,
Ducklike beaks, 59–60 276
non-carbon-based, 254–256
Ears non-DNA-coded, 256–257
asymmetric, 40–41 right-coiled amino acid–constructed,
tympanal, 40–41, 72–75 257–258
Ecological equivalents Extrinsic constraints, 252
birds and mammals, 147–153 Eyes
insectivorous, 140–142 camera, 38, 70–72
marsupials and placentals, 160–163 compound, 38, 69–72
Mesozoic and Cenozoic, 57, 153, 155, crystallin proteins, 198–200
163–168, 216, 233–234 enlarged, 33, 36
myrmecophagous, 142–144 horizontal-slit pupils, 32, 36, 61
necrophagous, 144–145, 235, 237–238 photoreceptor proteins, 179–183
nectarivorous, 145–147 types of, 67–72
South American, 62 vertical-slit pupils, 31–35
Ecospace, 168–175
Ecosystems Flowers
Adriatic, 163 cantharophilous (beetle-pollinated),
Australian, 157–160 119–121
convergent, 155–175 gigantism and, 120–122
Madagascar, 157 ornithophilous (bird-pollinated),
Mesozoic and Cenozoic, 160–163 123–125
New Zealand, 156–157 sapromyiophilous (carrion-mimic),
South American, 160–163 120–122
EDA protein, 195–196 Flying forms, 17–20, 260–264
Electric fields Form
detection, 33, 38–40 convergent (see Convergent evolution)
generation, 45, 49 impossible, 250–252
Index of Topics 321
nonexistent, 12, 16, 19, 21, 25, 29, 38, 66, Methane, 254–255
96, 103, 170–171, 174–175, 250–254, Mimicry, 8–9
260–264 Minds, convergent, 209–243
Freedom, 265–277 Mitochondrial DNA, 185–186
Fruit, 127, 129–133 Molecules
Functional constraint, 7–8, 13, 17, 22, 25, convergent, 177–209
29, 38, 46, 66, 96, 125, 245–264 DNA, 178–189
enzymes, 199–208
Gastric mills, 50–53 proteins, 189–199
Gliding forms, 20–22 Morphology
Ground-dwelling birds, 147–155 convergent (see Convergent evolution)
impossible, 250–252
Hammerhead ribozyme, 204–205 nonexistent, 12, 16, 19, 21, 25, 29, 38, 66,
Haptoglobin, 192–193 96, 103, 170–171, 174–175, 250–254,
Hemerythrin, 192–193 260–264
Hemocyanin, 192–193 Morphospaces. See Theoretical
Hemoglobin, 192–193 morphospaces
Herbivores, 50–60 Mourning, 240–242
Herding, 231–235 Myrmecochory, 131–133
Heterosporous reproduction, 115–118
Homology, 2 Nervous systems, 80–81
“deep homology,” 6, 38, 71, 101, 257–260 New Zealand ecosystems, 156–157
Homoplasy, 2 Niches
Horizontal-slit pupils, 32, 36, 61 carnivorous plant, 135–138
Hunting, 235–238 insectivore, 140–142
Hydrogen, 255 myrmecophage, 142–144
necrophage, 144–145, 235, 237–238
Impossible forms, 250–252 nectarivore, 145–147
Infrared detection, 37–38 parasitic plant, 139–140
Insectivore niche, 140–142 periodic table of, 168–175
Intrinsic constraints, 253 Nonexistent forms, 12, 16, 19, 21, 25, 29, 66,
103, 170–171, 174–175, 250–252, 254,
Lateral-line system, 72–73 260–264
Leaves, 97–99
Life Occlusive teeth, 53–57
carbon-based, 254–256 Opsin, 179–183
periodic table of, 260–264 Ornithophilous flowers, 123–125
silicon-based, 254–256, 276 Oxygen, 255
Titan and, 254–255, 264, 276
Locomotion Pancreatic ribonuclease, 184–185
burrowing, 26–29 Parallel evolution, 3–7, 71, 81, 178, 183,
flying, 17–20, 260–264 188
gliding, 20–22 recognition of, 3–7
swimming, 13–17, 260–264 Parasitic plants, 139–140
walking, 22–26, 260–264 Parrotlike beaks, 59–60
Lysyl-tRNA synthesis, 203–204 Penis, 82–84
Peptide-bond hydrolysis, 202–203
Madagascar ecosystems, 157 Periodic table
Marine-niche ecospace, 171–175 of life, 260–264
Marsupial ecological equivalents, 160–163 of niches, 168–175
Melanin-controlling proteins, 194–195 Photosynthesis, 101–105
Mentalities Placentae, 90–91
metacognitive, 239–240 Placental ecological equivalents, 160–163
mourning, 240–242 Plants
self-aware, 238–240 arborescent, 93–97
Mesozoic-Cenozoic ecosystems, 160–163 carnivorous, 135–138
Metacognition, 239–240 convergent, 93–133
322 Index of Topics
Plants (cont.) velcro-like, 128

dehydration and, 112–114 winged seeds and, 127–128
eyespots and, 99–101 Self-recognition, 238–240
herbivores and, 105–112 Silicon-based life, 254–256, 276
heterotrophic, 139–140 South American ecosystems, 160–163
leaves, 97–99 Spines, 105–108
photosynthesis, 101–105 Spinose armor, 63–64
poisonous, 108–112 Spirituality, 265–277
reproduction, 115–133 Spores, 115–118
spines/thorns, 105–108 Stomachal fermentation, 57–59
vessels, 97–98 Stomach lysozyme, 183–184
Poison Structural proteins, 196–198
fangs, 45, 48 Succulent forms, 112–114
plants, 108–112 Sugar phosphorylation, 201–202
sprayers, 65–66 Swimming forms, 13–17, 260–264
stingers, 45, 48–49, 64–66
Predictability of evolution, 265–277 Tanklike armor, 62–64
Proteins Teeth
antifreeze, 189–192 batteries, 54, 56–57
crystallin, 198–200 occlusive, 53–57
digestive, 183–185 poison fangs, 45, 48
EDA, 195–196 saber, 42–44, 162
haptoglobin, 192–193 tribosphenic, 53–55
hemerythrin, 192–193 Teleology, 265–277
hemocyanin, 192–193 Theoretical morphology, 12, 16, 19, 25, 29,
hemoglobin, 192–193 66, 96, 169–170, 175, 222, 227, 235, 246,
melanin-controlling, 194–195 254, 260, 263
photoreceptor, 179–183 Theoretical morphospaces, 168–175,
structural, 196–198 246–252, 260–264
Pupils convergent evolution and constraint in,
horizontal-slit, 32, 36, 61 246–252
vertical-slit, 31–35 Thermoregulation, 77–80
Thorns, 105–108
Red Queen hypothesis, 272–273 Tissue structures, 205–206
Reproduction Titan, possible life and, 254–255, 264, 276
animals and, 81–91, 119–133 Tools, 209–212
heterosporous, 115–118 Trees, 93–97
penis and, 82–84 Tribosphenic teeth, 53–55
placentae and, 90–91 Tufted seeds, 126, 128
plants and, 115–133 Tympanal ears, 40–41, 72–75
viviparous, 84–91
Respiration Ultrasound detection, 41–42
hydrogen and, 355
oxygen and, 355 Velcro seeds, 128
Retractable claws, 43–47 Vertical-slit pupils, 31–35
Reverse evolution, 4–5 Viviparous reproduction, 84–91
Rhodopsin, 179–183
Ribozyme, hammerhead, 204–205 Walking forms, 22–26, 260–264
Rumination, 57–59 Winged seeds, 127–128
Wings, 17–20, 260–264
Saber teeth, 42–44, 162
Sapromyiophilous flowers, 120–122
Seed dispersal
exploding seeds and, 127, 129
fruit and, 127, 129–133
myrmecochory, 131–133
tufted seeds and, 126, 128

Convergent Evolution Limited Forms Most Beautiful Ebook3000 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Convergent Evolution Limited Forms Most Beautiful Ebook3000 PDF

Uploaded by

Copyright:

Available Formats

www.ebook3000.

The Evolution of Cognition, edited by Cecilia Heyes and Ludwig Huber,

Limited Forms Most Beautiful

George R. McGhee Jr.

The MIT Press

For information about special quantity discounts, please email special_sales@mitpress

Library of Congress Cataloging-in-Publication Data

1 What Is Convergent Evolution? 1

4 Convergent Ecosystems 135

5 Convergent Molecules 177

6 Convergent Minds 209

7 Functional and Developmental Constraint in Convergent Evolution 245

8 Philosophical Implications of Convergent Evolution 265

Appendix: A Phylogenetic Classification of Life 279

Biology is becoming the leading science in this century. As in all other

the University of Vienna. The KLI fosters research projects, workshops,

Gerd B. Müller, University of Vienna and KLI

In 1859, Charles Darwin concluded his momentous book On the Origin

The question “How common is convergence?” remains unanswered and may be

Recognizing Convergent Evolution

Z also arises, but by evolutionary modification of the preexisting trait B,

been modified in two separate lineages to produce the parallel conver-

possess trait B as species 4 and 5 do (figure 1.4). At present, the impor-

Convergent Evolution versus Parallel Evolution?

In modern phylogenetic systematics, we see that parallel evolution is

but the birds modified dinosaurian forelimbs to wings whereas bats

Evolutionary Constraint and Convergent Evolution

Convergent evolution was once thought to be almost exclusively pro-

developmental constraints in the evolutionary process: “Patterns of par-

Mimicry, Camouflage, and Convergent Evolution

Mimicry is a form of convergent evolution in which one species indepen-

and camouflage in nature is sometimes spectacular, the fundamental

On the Organization of This Book

This book is not intended to be an “Encyclopedia of Convergent Evolu-

Some of the most spectacular examples of convergent evolution are

Note: The geological age of extinct species is marked with a †.

The evolution of an ichthyosaur or a porpoise morphology is not

as can be seen in the South American lungfish Lepidosiren paradoxa,

4.3 Kuhl’s flying gecko (Tetrapoda: Reptiliomorpha: Amniota: Sauropsida: Diapsida:

groups modified their forelimbs in the evolution of wings because only

pair of forelimbs. For the insects, such a morphological transition would

modification of the same glider morphology has been evolved by the

Last, convergent gliding morphologies have evolved in at least three

1 Convergent structure and function: LEGS (articulated, muscle-controlled lever

evolved prehensile hands in their forelimbs. Another possible bipedal

rabbit” although it is neither a rabbit nor a kangaroo. Last, bipedalism

impossible, but highly improbable for vertebrates) and, second, develop

1 Convergent structure and function: SNAKE-SHAPED BODIES (penetrator adaptations:

1.6 California legless lizard (Squamata: Scleroglossa: Autarchoglossa: Anguimorpha:

millions of years in the Late Devonian and Early Carboniferous. So it is

Carnivores: Prey Detection

As heterotrophic organisms, animals are incapable of synthesizing their

bodies, many fish still possess a limited degree of binocular vision

2.2 Common toad (Bilateria: Deuterostomia: Chordata: Osteichthyes: Sarcopterygii:

5.3 Winter skate (Chondrichthyes: Elasmobranchii: Batoidea: Rajiformes: Rajidae;

horizontal-slit pupil would work just as well as a vertical-slit pupil. And

an extreme, but highly advantageous, form of nocturnal “vision”: the

2008), three groups of insects have convergently evolved specialized

widespread than Darwin realized, and that they have convergently

1 Convergent structure and function: ASYMMETRIC EARS (allows three-dimensional

Note: For data sources see text.

convergently evolved the ability to hear ultrasound. The South American

Carnivores: Prey Capture