viii ACKNOWLEDGMENTS
······················ .....................................................................................
toward the final product, a book. I'm a lucky guy.
Charlie che Cat also provided valued companionship,
lying on the desk next to the computer as I labored
through the final chapters of a long book, keeping me
calm (a technique that is alleged to have also worked
for Mark Twain).
The editing and production team assembled by
Princeton University Press was not only awesome in
its ef/iciency but also a pleasure to work with. Just
look at the book. That's what they did! I want to
thank my editor, Alison Kalett, for her patience and
sound stewardship. In addition, the book could not
1"Í
have happened without the diligent and tenacious
work of Karen Fortgang; Stefani Wexler; my most able
copy editor, Jennifer Harris; and proofreader Barbara
Liguori. Finally, I want to acknowledge Robert Kirk,
. -··• lntroduction
Executive Editor and Group Publisher, Science and
Reference at Princeton University Press. lt was Rob-
ert who proposed the idea of a tropical text to me. lt
was Robert who convinced me to do it. When I asked
Robert what he thought the book's title ought to be,
he looked at me and said something like, "lt's about
tropical ecology so I think we should call ir Tropical
Ecology." Well done, Robert, and thanks.
What Is Tropical Ecology?
Asking the question, What is tropical ecology? may seem akin to asking questions
such as, Who is buried in Grant's tomb? Tropical ecology is the study of the ecology
of tropical regions. But so what? Consider these questions: First, what is ecology?
What are its paradigms, what are its principies, and what is its place in the bio-
logical sciences? What unique insights has ecology added to biology? Second and
perhaps more important, is tropical ecology unique within the study of ecology? In
other words, if you studied the ecology of the eastern deciduous forest, the arctic
tundra, the Mojave Desert, or the seas that surround Antarctica, what paradigms
and principies might you miss because none of these ecosystems are tropical?Take
a few minutes to think about these questions, and then read on.
Readers of this book should be familiar with ecology and thus need little
coaching as to what ecology is. Ir is, in a nutshell, the study of how organisms exist,
adapt, and interact within their environments within the context of both abiotic
and biotic factors. It is studied at widely different spatial and temporal scales. Ecol-
ogists revea! and attempt to explain patterns of species richness and explore how
top-clown and bottom-up forces influence the structure and stability of food webs,
struccuring whole ecological communities. Ecologists use experimental techniques
and mathematical models to study competition, predation, mutualism, and the as-
sembly, persistence, and stability of ecological communities. Ecology has matured
in that disciplines such as landscape ecology, restoration ecology, and conservation
ecology ali use theory from basic ecology. Each of rhese areas of applied ecology has
made an impact on issues in ti:opical ecology.
The various topics cited in the previous paragraph are the grist for ecological
study and make up che "usual suspects" in ecology texts. Once one understands
these topics, one is considered proficient in ecology. Do these topics form the para-
digms of ecology? Not really. They are best considered principies, areas of study
around which the discipline is organized. Ecology arguably has no paradigms.
(1argue this point throughout my book The Balance of Nature: Ecology's Enduring
Myth. ) On close inspection, ecology is a prominent branch of evolutionary biology.
Not much of what ecologists study makes sense without placing the data firmly
within evolutionary context. This reality will become apparent in the study of trop-
ical rain forest ecology, especially regarding species' origins, interactions, and inter-
dependencies. The paradigm within which ecology is included is organic evolution,
and it, along with cell theory, are the only real paradigms of biology. Cell theory
and evolution are what biology is. Ali the rest is detail.
Where <loes this leave tropical ecology? What will you learn by going to Pan-
ama, or Ecuador, or Indonesia, or Queensland, Australia, that could not be learned
7
INTROOUCTION
······················· ···· ······························•·························•················· ·······•················· ·································
by going to central Spain, or Norway, or Japan, or
New Jersey? Yes, the species are different, but what
of it? The pattern of seasonality is different; precipita-
tion patterns are different; lots of things are different.
What of it?
This "what of it?" is the subject of this book. By
the conclusion of your study, you should be competent
to voice an educated opinion regarding the question
of just how unique tropical ecology is. Learn and en-
joy the journey.
Overview
When Charles Darwin was fresh out of college, he vis-
ited a tropical rain forest in eastern Brazil as pare of
his famous voyage aboard the HMS Beagle. Like oth-
ers who carne befare and since, his first impressions
were vivid. He wrote of this experience in che tropical
rain forest:
When quietly walking along the shady path-
ways, and admiring each successive view, 1
wished to find language to express my ideas.
Epithet after epithet was found too weak to
convey to those who have not visited the in-
tertropical regions the sensation of delight
which the mind experiences. (Darwin 1845J
In other words, Darwin was pretty overwhelmed by
what he saw, felt, heard, even what he smelled. He
commented on the hazy air within the foresr, evidence
of the high humidity. A tropical rain forest <loes leave
you with a vivid "first impression." It gets your atten-
tion both rationally and emotionally. But years 1ater,
a much older Charles Darwin had a different perspec-
tive when he was composing his autobiography:
In my journal l wrote that whilst standing in
the midst of the grandeur of a Brazilian for-
est, "it is not possible to give an adequate
idea of the higher feelings of wonder, admira-
tion, and devotion which fil! and elevare the
mind." I well remember my conviction that
there is more in man than the mere breath of
his body. But now the grandest scenes would
not cause any such convictions and feelings to
rise in my mind. (Darwin 1887)
What changed Darwin was, of course, evolution.
When Darwin returned from the Beagle voyage, he
quickly became convinced of the truth of organic evo-
lution. Ali present species have arisen from previous
species and were not specially created. This was soon
followed by his discovery of the mechanism for evo-
lutionary change, natural selection, and so he began
to view the world differently. Of course, not everyone
agreed with Darwin. Sorne rejected his theory our-
right; orhers agreed that evolution might be true but
did not agree that a blind mechanism called natural
selection caused species to evolve into different forms.
From 1859, when On the Origin of Species was pub-
lished, until Darwin's death at the age of 73 in 1882,
he was strongly attacked in speeches and in print over
his assertions concerning evolution.
What does ali this Darwinian history have to do
with tropical ecology? The great twentieth-century
evolutionary biologist Theodosius Dobzhansky, who
spent a fair amount of time collecting fruit flies in
the tropics, said "Nothing in biology makes sense ex-
cept in the light of evolution" (Dobzhansky 1973, p.
449). Darwin's big idea has become the most impor-
tant paradigm in biology, and without it, the immense
complexities that characterize a tropical rain forest or
any other ecosystem would seem hopelessly beyond
rational explanation.
Therefore, much of this book will be about evolu-
tion as it is evident in the tropics, and in rain forests
in particular. There is nothing unique about this ap-
proach, as I fail to see how it could be otherwise for
any kind of ecological study. If this book were about
the ecology of Antarctica, evolution would still be the
organizing paradigm. The noted ecologist G. Evelyn
Hutchinson once titled a book about general ecology
The Ecological Theater and the Evolutionary Play. 1
doubt that many who visit a rain forest or any other
tropical ecosystem today need fear Josing their sense
of awe (as Darwin apparently did) simply because
of acquiring an understanding of how the process of
evolution is manifest in the multitudes of morpholo-
gies and interactions that characterize rhe ecological
community. Darwin, when elderly, apparently lost his
awe of nature, perhaps embittered by ali the criticism
directed toward him. You need not worry, however.
Learning the science behind the landscape really adds
to awe-it <loes not detraer from it. This book aims to
teach the science behind the landscape.
As author of the book you hold, I am in a unique
position. lt's my book, and l wrote it my way. Orher
books about the tropics take different approaches from
what you will read within this volume. This book reflects
my views from study and field experience (particularly
in ornithology) over my four decades asan ecologist. Let
me explain something of how the book carne to be.
In 1989, Princeton Umvers1ty Press publtshed my
book tided A Neotropical Companion. The book wás
modest in size and had a bright green cover, which led
ro irs nickname, the "Little Green Book." lt amazed
me while writing that book how much wonderful in-
formation ecologists had garnered in their collective
research efforts and how little of chis really cool st.uff
was finding its way to more popular-based audiences,
including into the hands and minds of students. My
goal was to be selective and make sorne of rhat amaz-
ing information available.
The success of the Little Green Book led to a ma-
jor revision and expansion of A Neotropical Compan-
ion that appeared in 1997, again published by Prince-
ton University Press. This volume was bigger, had new
iliustrations and sorne color photographs, and, most
iinportantly, covered more topics in greater derail,
with much additional treatment of Sourh America.
Eager to learn from my readers, 1 included my
e-mail address in the revised volume, and many re-
sponded. I have hada few errors of both omission and
cammission pointed out to me, as well as many com-
ments as to how the book could be improved. Sorne
v::rnt more maps; sorne want more illustrations; sorne
want greater emphasis on particular tapies. But the
most-repeated comment (the title of the book notwith-
standing) is that the book is too narrow, as it covers
only the Neotropics and omits the Old World tropics
of Africa, Southeast Asia, and Australia/New Guinea.
How does the Neotropics compare with other tropical
regions of Earth? How are these regions biogeograph-
ically distinct, one from another? Are the ecological
processes that characterize Neotropical ecosystems
the same as those in other regions?
Thus, it seemed both to me and to my editors at
Princeton University Press that it was time for yet an-
other revision and expansion, this time going "the full
Monty" toward a more comprehensive book cover-
ing the ecology of the global tropics. And that is how
Tropical Ecology carne to be.
Ah, but as alwa ys, the devil is in the details. My
previous book was specifically about the Neotropics,
and the vast majority of my field experience has been
in that region. True, I have visited Africa (Tanzania),
Southeast Asia (Indonesia), as well as the Australian
tropics in Queensland and Darwin, but most of my
reading and field study is from Central and South
America and the Caribbean. And, l dare say, much if
not most of the published literature continues to be
INTROD UCTION
from studies in the Neotropics. So I cannor pretend to
take an "even-handed" approac;h to the global tropics.
This book will expose my bias, as it reflects my con-
fidence leve] in my expertise-and that bias remains
clearly toward the Neotropics and, in particular, birds.
But such a bias has a pragmatic component in that
many students will like]y use this text in conjunction
with an actual field experience, a college course that
includes a trip to experience a tropical region. And
most North American college courses in tropical ecol-
ogy visir sorne area within the Neotropics. My contin-
ued emphasis on that region will benefit such courses.
The organization of the work is mine, my way
to tell the story of tropical ecology. But the informa-
tion contained herein is largely rhe work of others, the
thousands of researchers in rhe field and lab whose
work, often repetitive and always exacting, deals with
sorne aspect of tropical ecology. Far the most pare, 1
rely on primary source material, papers from jour-
nals- some of which, like Biotropica, publish only
papers about tropical ecology, and sorne of which,
like Ecology and American Naturalist, publish many
papers that deal with the tropics. I also include many
citations from Science and Nature, both of which
publish numerous major papers on various aspects of
tropical biology. Each topic covered is illustrated by
selected examples, so obviously many worthy papers
are not cited. My apologies to these authors, but this
book is not meant as a comprehensive review of ali
topics tropical. lt is nice to have an embarrassment of
riches when ir comes t0 information about the tropics,
and I have had to carefully pick and choose. Examples
were by and large chosen for their overall interest, po-
tential significance, clarity, and robustness. And, since
researching the second edition of A Neotropical Com-
panion, I am amazed by the remarkable increase in
'the number of tropical studies represented in profes-
sional publications of ali kinds. Another point worth
emphasizing is that my area of expertise is in ornithol-
ogy, the study of birds, so whoever reads my text will
encounter quite a few examples that involve birds. But
consider that birds are fairly easy to observe in the
field, and thus much data about species interactions in
the tropics does, indeed, focus on birds.
The first question any author faces is, Who is the
audience for this book? This work is aimed at students
and others who take a serious interest in learning
about tropical ecology. lt assumes a basic knowledge
of introductory biology, including ecology and the te-
nets of evolurionary biology. I present an overview of
each tapie and then note where information gaps exist

.. .....
1
-·• What and Where Are the Tropics?
Beginnings
The world was once a larger place, or so it seemed. Vast areas lay unexplored, and
following the Renaissance, European explorers began to fan out across the Earth
in a narionalistic search for conquest, treasure, and lands to claim for their respec-
tive countries. With the onset of global colonialism and imperialism carne an age
of exploration. By the eighteenth and nineteenth centuries, man y new regions were
becoming known, and many of these were in the tropics.
Science, in the philosophical sense, traces its roots back to the emergence of sci-
ence in ancient Greece, the Middle East, and Asia. But in a more contemporary sense,
science really began during the Renaissance, when curiosity led to inductive and de-
ductive reasoning, observacional and experimental analysis, and most importantly,
empírica! investigation and materialistic explanations for natural phenomena. The
study of the tropics was formaLized only after che ages of global exploracion and dis-
covery and when che scientific mechod was used to investigare phenomena.
The exploracion of che world's tropics by European nations brought with it
much specimen collection and cacaloging. Thus, che roots of tropical ecology lie
most deeply in museum cabinets housing vast numbers of carefully labeled plants
and animals broughc back by man y devoted and indefatigable nacuralists, including
such figures as Charles Darwin and Alfred Russel Wallace (Place 1-1). When Cap-
tain James Cook chose to have two naturalists, Joseph Banks and Douglas Solander,
PLATE 1-1
(a) CHARLES DARWIN
(b) ALFRED RUSSEL WALLACE (a) (b)
WHAT ANO WHERE ARE THE TROPICS?
.......................
·· ··• •" ' ' ' "" ' '
• l1ided on che voyage of che Endeavor
me .
(1768),
.
he set elevation Andean slopes as he traveled from Venezu-
-~ immensely valuable precedent. lt 1s thac acnon chat ela through the northern Andes. He visited central
•l;imacely led to Darwin's being invited on the Beagle Mexico, Cuba, and the United States. Humboldt
~oyage (1831-1836) and thus indirectly was respo·n- climbed sorne of the .,higher Andean peaks, reach-
sible for che Origin of Species (1859).. ing elevations of 5,800 meters (19,000 feet). lt was
One of the most noceworthy contnbutors to early Humboldt who first discovered and described the
knowledge of the cropics was Alexander von Hum- unique nocturnal oilbird (Steatornis caripensis) (see
boldt (Piare 1-2), sometimes called che founder of the Chapter 7) in the caves around Caripe, Venezuela.
science of biogeography (Jackson 2009). Biogeography Humboldt and his botanist colleague Aimé Bon-
is che study of how organisms vary among regions, pland described elevational changes evident on Andean
even when climate is similar. For example, there are no slopes. They realized that as climate changed by eleva-
apes in che New World tro~ics, but there are no mon- tion, so did plant and animal communities, and thus
keys with prehens1le ta1ls m the Old World trop1cs. they were the first to elucidare what is termed the life
This sort of example reflects differing evolutionary zone concept. Much later, c~s concept was formalized
histories among regions. Biogeography will be a focus by Leslie R. Holdridge (1947). Alife zone is an ecologi-
úf the next chapter. Humboldt led successful and bold cal area defined by clima tic variables such as mean an-
expeditions to difficult places ranging from Russia to nual temperature, total annual precipitation, and the
northern South America. He published a massive five- · ratio of mean annual evapotranspiration (a measure
\·o!ume tome titled Kosmos in which he attempced to of heat) to mean total annual precipitation. (Transpira-
summarize the entire scope of scientific knowledge of /ion is a term referring to the evaporative water loss
¡¡·;e world. from plants. Plants are adapted to pul! water from
Humboldt's travels in Souch America took him soil via roots, and eventually that moisture is lost by
irom rain forests along che Orinoco River to high- evaporation from the surface parts of the plan t.) In the
Andes (as with al! mountains), these variables change
with elevation and result in differing ecosystems such
as páramo (an ecosystem of wet grassland and shrubs)
at high elevations and lush rain forest ac low elevation
(Piares 1-3 and 1-4). Holdridge's life zone concept will
be further discussed later in this chapter.
Naturalists were often moved to comment on the
rich and diverse nature of flora and fauna within tropi-
cal rain forests. Their observations, by today's stan-
dards, were largely (and understandably) anecdotal,
though not lacking in important details. The modern
student of tropical ecology will profit from reading
sorne of che narratives of figures such as Thomas Belt,
Henry Walter Bates, the aforementioned Darwin and
Wallace, as well as others. For example, Henry Walter
Bates discove red the form of animal mimicry (described
in Chapter 8) that bears his name. These early explorers
displayed much tenacity in capturing decail, transcribing
vast amounts of information, and collecting and trans-
porting specimens often under difficult circumstances.
HISTORICAl REFERF.NCES
These books, all still available (at least in libraries), were authored by
sorne ol thepioneersof tropical ecologyand are stronglyrecommended
as insighttul, entertaining,andinspiring resources.
f-lATE 1-2 Bales, H. W. 1892. 7118 Na/JJralist oo lheRilfr Amazons. Looooo:.))hn Murray. Classic
Al.EXANDER VONHUMBOLDT acrounl of Amazon~nna!Ural histCI')'. and quite1\'0ndert~.
 INTRODUCTION
and where controversies have arisen. The book is
meant to be provocative and poses many questions,
the answers to which are still to be revealed or are
presently being hotly debated. Sorne of che studies 1
describe come to different conclusions about che same
copie. Science is always a work in progress, and this
book strongly takes that approach.
AII science books should be subject to peer review,
and this book certainly has enjoyed that advantage
(see the Acknowledgments). Others, expert in vari-
ous areas of tropical biology, ha ve read chapter drafts
and made numerous critica! comments. I have taken
man y of these suggestions, bue [ have also chosen not
to take sorne. No two tropical biologists would ever
write the "same" book. Our views are influenced by
our biases, varied interests, and experiences, ro say
nothing of our intellectual outlooks. We make judg-
ments, and we often have to agree to disagree on
various poincs. Any problems identilied with rhis rext
should be directed to me (jkricher@wheatonma.edu),
and I take sote responsibiliry for any errors that may
be detected.
Sorne material that has remained robust has been
taken and often updated from A Neotropical Com-
panion, but most of che material is new to this book.
That said, any book such as this is old before it is
published. Right now, any number of good studies are
occurring or have occurred that alter sorne of what is
written on these pages. That is the narure of science,
the strength of science, the very essence of science. Sci-
ence is cumulative, and anyone entering into rhe study
of a lield of science prolits from acquiring a lirm foun-
dation from which to begin critica! study. I hope rhis
book will serve that purpose for its readers.
Organization
of This Book
This book consists of 15 chapters, a challenge for a
one-semester course, bue hopefully the writing is suf-
/iciemly engaging that the task of moving through the
chapters will prove to be relatively pleasant. There
are numerous ligures, maps, tables, and color photo-
graphs throughout to add to the clarity of che text.
1begin (Chapter 1) with a broad overview of what
and where the tropics are. This provides sorne infor-
mation about how tropical ecology emerged from
global exploration into the realm of scienri/ic srudy. lt
also describes something of the climatic variables that
......,
are the overarching determinants of tropical ecosys-
tem characteristics, and it sets the stage for what is to
come throughout the book.
Next (Chapter 2) 1discuss biogeography and evo-
lurion in the tropics. For those already well versed in
these ropics, this chapter may serve as a helpful re-
view. But for many, and I expect most studenrs, it will
be essential in explaining how and why tropical re-
gions differ, as well as in clarifying basic evolutionary
principies, particularly speciation.
Then it is time to entera rain forest (Chapter 3) and
examine the physical srructure of this unique ecosys-
tem. This chapter discusses the various characteristics
of rain forest vegetation, from tree sbapes to buttressed
roots. lt should be the primer that will allow any sru-
denr to see the many details and general characteristics
common to all rain forests.
Chapters 4 and 5 both explore ideas as to why
tropical regions and rain forests in panicular are so
rich in their numbers of species. Biodiversity is one
of the most signi/icant arcas of tropical research, and
thus two chapters are devoted to research focused on
this complex topic. Chapter 4 discusses how diversiry
is partitioned within and between habitats and con-
siders whether rhe high diversiry of tropical regions
is an example of high rates of speciation (or low rates
of extinction). Whar factors allow tbe continued exis-
tence of such complex communities? Chapter 5 looks
at the range of hypotheses attempting to explain che
uniquely high tree species richness in tropical forests,
one of the most diflicult theoretical problems in tropi-
cal ecology.
The effect of disturbance is explored next (Chap-
ter 6). Periodic natural disturbance may result in a
shifting mosaic of different aged parches of forest such
rhat bigh species richness is maintained over large
arcas. This chapter deals in part with gap phase dy-
namics. Gaps in a forestare created by forces ranging
from single fallen trees to major blowdowns caused
by storms or fue. Gaps are exposed to high levels of
sunlight and thus have a differenr ecology from closed
rain forest, where canopy crees strongly attenuate the
light striking the forest íloor. Last, this chapter consid-
ers che topic of secondary s11ccessio11, which is of ma-
jor importance in assessing how rain:forest regrows
following disturbance.
The richness of species in tropical forests results
in numerous complex forms of interactions among
species-some rather casual, sorne sufliciently com-
plex that the species are evolutionarily interdepen-
dent. Chapter 7 presents examples of sorne of the
INTRODUCTION
-, r striking biotic imeracrions involving such topics vey other tropical ecosyscems and also examine topics
n.os d d.
~ fruir consumption as it relates to see 1spersa an
1 d having to do with conservation science in the tropics.
5
' propensity of many tropical species to evolve mu- Arcas of grassland with scattered trees are called
ne1 · 1udes a focus on savanna. Chapter 11 discusses tropical savannas and
tistic interacrions. This cha pter mc
rua . b
coevolution, when two or more spec1es ecorne mutu- dry forest ecosystems. This discussion examines rwo
ally interdependent. views of how savannas and dry forests coexist in the
Food webs and the adaptations associated with tropics. One is that grassland, savanna, and dry forest
rhem form the core of Chapter 8. This chapter looks represent a moist11re gradient e((ect (where grassland
ar various top-clown and bottom-up inreractions that is found in the most arid arcas, and dry forest where
contribute to scructuring food webs and trophic dy- there is more annual precipitation), and the other
namics in rain forests. Topics include the evólution of views them as alternative stable states dependent on
characteristics such as cryptic and warning coloration. local condirions. The importance of lire and grazing
The high rate of photosynthesis evident in tropi- are also discussed.
cal rain forests results in high rates of primary produc- Chapter 12 surveys other major tropical ecosys-
tivity, and rhis is the subject of Chaprer 9. How much tems, beginning with montane ecosystems and con-
carbon do lush tropical forests take in and how much tinuing to riverine forests and coastal mangrove forest.
do rhey emit (as carbon dioxide)? The carbon µux of This chapter also considers the connectivity between
tropical forests is an important area of research. As . montane and lowland ecosystems from the standpoint
currenr climate change continues dueto ongoing addi- of biodiversiry maintenance.
tion of atmospheric car bon dioxide, what is che poten- Humans evolved in tropical Africa, and Chaprer
tisi of tropical forests to act as carbon sinks, absorb- 13 discusses human ecology in the tropics. Topics
ing and storing excess carbon? This chapter explores include hunter-gatherer communities, simple agricul-
sorne of the sometimes contradictory research about ture in the tropics, and some of the traditional ways
h,w tropical forests might act to mitiga te the effects rhar humans have affected the overall ecology of the
of increasing atmospheric carbon dioxide. rropics.
Rich tropical forests often occur on poor soils. The last rwo chapcers <leal with conservation is-
Chapter 10 examines nutrienr cycling and tropical sues in the rropics. Chapter 14 focuses on forest frag-
soils. lt describes how vital nutrients such as phospho- mentation and biodiversity. Chapter 15 discusses
rus, nitrogen, and calcium cycle in tropical ecosystems deforestation and forest degradation, as well as the ef-
and the essential function of decomposer organisms. fects of fire on tropical forests. lt also examines emer-
This chapter includes discussion of various kinds of gent pathogens and invasive species and concludes
tropical soils and contrasts forests on µoodplains, with an overview of a currenr debate among ecolo-
where annual flooding renews soil fertiliry, with those gists abour the likely future of tropical forests. While
on terre firme, off floodplains. not meant to be comprehensive, these two chapters
Chapters 3 chrough 10 ali focus on che ecology of taken together cover the most fundamental issues and
rl;e tropical rain forest. The remaining live chapters sur- questions facing tropical conservation science.
 8 CHAPTER 1 WHAT ANO WHERE ARE THE TROPICS? 9
..................................................,. ····•······················· ...................................................... , ····•········......................... ,., . ····· ................................. ··•·······•·····"·· ...... ..................................................................... .
.····· ..
..·• Henry Walter Bates versus Curl-Crested Aracaris

The following text from Bates's account in The Nat11- tree in a dark glen in rhe foresr, and enrered 1hc thicker
ralist on the River Amazons illustrates che curl-crested where the bird had fallen ro secme my booty. lt was
only wounded, and on my attcmpting to seize it, ser up
aracari (Pterog/ossus beauhamaesii), a species of toucan, a loud scream. In an instanr, as if by magic, 1he shady
and the memorable encounter that Bares had with a f\ock nook seemed alive wi1h 1hese birds, although thcre was
of 1hese birds (Figure 1-1). certainly none visible when I enrered the jungle. They
The Curl-crested Toucan (Pttrog/ossus Beauharnaisii).- descended rowards me, hopping from bough 10 bough,
Of the four smaller roucans, or arassaris, found near Ega, sorne of rhem swinging on thc loops and cables of woody
lianas, and all croaking and 0urtering their wings. (Bates
1he P1eroglossus flavirosrris is perhaps 1he mosr beauriful
in i1s colours, its breasr being adorned wi1h broad belrs 1892, pp. 335-336)
of rich crimson and black; bui the most curious species
by far is 1he Curl<resred, or Beauharnais Toucan. The
feathers on rhe head of 1his singular bird are transformed
into horny pla1es, of a lus1rous black colour, curled up a1
rhe ends, and resembling shavings of sreel or ebony wood,
the curly crest being arranged on the crown in the form
of a wig. Me. Wallace and I fust met with this species on
ascending the Amazons, a1 the mou1h of the Solimoens;
from that poinr it continues as a rarher common bird
on the terra firma, at leas, on the south side of the rive~
as far as the Fonte Boa, but I did not hear of irs being
found furrher 10 rhe wesr. lt appears in large flocks in
the fotests near Ega in May and June, when it has com-
pleted its moult. I did nor find these bands congregared
ar fruit rrees, but always wandering through the forest,
hopping from branch 10 branch amongst the lower trees,
and parrly concealed amongs1 1he foliage. None of 1he
arassaris ro my knowledge make a yelping noise like thar
PLATE 1-3 uttered by the larger 1oucans (Ramphasros); the notes
SNOW ON PARAMO of rhe curl-cresred species are very singula~ resembling
Snow is common ar high clevarions in equarorial regions. Here, bunch grasses of rhe high páramo life zone are partially snow-covered. From rhe croaking of frogs. I had an amusing adventure one
Ecuador. day with thcse birds. I had shot one from a rather high

Beebe, W. 1918. .Alngle Peace. Loodoo: w~. Beebe's ¡wrey from tfle West lndies
to the rai1 lorest riJng~') of GiJyana makes lor engrossing reading, 0000119 a
IIOOderfu iKXXllrtof the IXld Hoatlilbird(Olapter 12).
_ . 1921. &Jge o/ t/le J.rlgle. New\'oo<: Helvy tt>I. L~ the prelious >dt.me,
coo1a¡¡¡ short, del1ghtftJ essays Otl tr~ ecology. aassi:
Belt, T. !18741 1985 reissue. 1/Je Na/uralist ;¡ Nicaragua. Chi::ago: Uni'lersity ol Chi-
cago Press. Ooe of the besl of the classi: exploratory accoonts, IOCtlSed entiety
Otl Ceroal Ameri::a
~ . f. M. 1938. U/e i1 an,'ifCas//e. New York: ,tw~t0t1.century. H~lvy readable
1~th much inlormation, pallicu~y or, troptal birds.
Oanm, C. R 110061 1959. 1/Je ~¡age o/ t/le /Jeag/e. Reprnt Loooon: J. M. Dent and
Sons. Ore ol the besl c1assi: - I S ol travel llroughout Sootll Mier1ca Ma11f
re¡xinted eótxlns are a'l3iable. Must reaoo;¡.
Wallace, A R. 1869. 1/Je MatayAlchipetago. LClldoo: Macmilan. This is Walace's roost
lruoous book, an engrossing accooot ol his e~h1 years of travel throughout what
is now lrrlorlesia. The book abooods ~i:ll informaoon atru wiklde as wel as the (a) (b)
varixJs ~ Ylalace lifXXIUOtered. FIGURE 1-1
- -. 1895. Na/uf¡¡/ Se!ecoon and TrOl)ical Nature. Lorxfoo: Macmillan.This deUght- (a) Curl-cresrcd toucan. (b) Mobbed by curl-cres1ed roucans.
PLATE 1-4
lul book oontains ,oo descriptions, in gklrious Vttorian prose, ol waiace·sexpen-
FOREST CANOPY
mis in Proalooia.
The dense canopy of a low-elevarion, humid tropical moist foresr,
······································........................................................................................ , .............·
l\'ateslon, C. 11825) 1983. V ~ i'1 Sootl'IAmerica Reprilt Loooon: Century Ptb-
lishing. Aveij entertailing narratr,oe by aralfler eccenlric bu1 pe,ceptive expi¡xer. such as this one in the Arima Valley in Trinidad, is typically irregu•
lar, wtth sorne emergenr rree species rising above orhers. Foresr gaps
creared by fallen trees also add ro canopy hererogeneity.
 10 CHAPTER 1
WHAT ANOWHERE ARE THETROPICS? 11
..... ' .. ' ' .....~.'.' .............. ' .....' ......' ...................................... .................................................' ........... ' ... ·•· ...... .

Most eminent naturalists of the nineteenth cen-
tury honed their skills through participation in voy-
ages of discovery. For example, Thomas Henry Hux-
ley and Joseph Hooker, both clase friends of Darwin,
each made an extensive voyage for the express pur-
pose of expanding his horizons. These voyages were
insightful, but they normally did not permit careful,
systematic study within a given region.
It was not until the twentieth century that system-
atic srudy of regions within the tropics was really pos-
sible. What was needed were permanent field stations
where sciemists could go and conduct their work.
In 1923, following the completion of the mam-
morh engineering project known as the Panama Canal,
a hilltop area of about 1,500 hectares (3,706 acres)
on the Panamanian isthmus remained above water, a
newly formed island surrounded by the newly created
Gatun Lake. This was named Barro Colorado Island.
Since 1946, it has been a Jield station administered by
the Smithsonian Institution, and it has served numer-
ous researchers, a function that continues to the pres-
ent day. As you continue through this book, you will
notice numerous citations regarding Barro Colorado
Island, usually referred to simply as BCI.
Another unique field station devoted to tropical
biology is Simia, in the Arima Valley on the island of
Trinidad. While Simia is not one of the most promi-
nent si tes toda y, ir is noteworthy for its founder, the

.. • Williarn Beebe . .... .. ...... ........... ..... ..... ... ........ ......... .... . ....... .. ....... ... . . .... ... .. .. . . ...... .

William Beebe (1877-1962) (Plate 1-5) did far more than
change the directioo of research in tropical ecology. He
had a long and highly distinguished career as an explorer,
scienrist, and writer. He spent most of his career asso-
ciated with the New York Zoological Society. His pro-
ductivity was amazing, having authored 24 books and
PLATE 1-5
WILLIAM BEEBE
... .. ... ... ...... .................................
about 800 scientific arrides over the course of his career.
Beebe began asan ornithologist, and one of his first ma-
jor books bore the title The Bird, Its Form and Function
(1906). Arguably bis srrongest contribution to ornithol-
ogy (he had many) was his classic book A Monograph
o( the Pheasants, published in four volumes from 1918
to 1922.
Beebe's tropical research was prolific and began in
British Guiana (now Guyana) in 1916. He traveled widely,
including to rhe Galapagos Islands, a trip that resulted in
another memorable book, Galapagos: World's End (1924).
He also authored two books about tropical ecology, one of
which, Jungle Peace (1918), greatly impressed Theodore
Roosevelr, himself an explorer of the Neotropics.
In 1928, Beebe established a field station in Bermuda,
an event that led· to another major chapter in his caree¡;
that of oceanic explorer. Beebe was frustrated that animals
hauled up from the depths of the seas were not only dead
on arrival but also muti!ated by the pressure change expe-
rienced while being hauled to the surface. After consulta-
tion with Theodore Roosevelt, Beebe eventually focused on
a submersible that he called a batlrysphere, a thick metal
chamber that could withstand the great pressures of the
ocean depths and that could be lowered and raised from
a ship. He subsequently made about 30 descents, accom-
panied in the tiny chamber by the bathysphere's invento¡;
Otis Barton. In 1934, Barton and Beebe reached a depth :
of 922.93 meters (3,028 feet), an amazing accomplishment
that was not equaled for many years.
Beebe was a popularizer of science but, at the same
time, a very productive scientific researcher. lt has been
said of him that he really founded the science of tropical
ecology.
··············"······· ........... ..... ······•······.. .. , ....... ................·
• www.teamnetwork.org/en/about). The Organization
naturalist•explorer William Beebe, and for its initial
pc1rpose. Though Trinidad appears to be part of the
West lndies, it is biogeographically distinct, as it is
actually on the continental shelf, part of Venezuela,
just as the island of Martha's Vineyard is really part
of Massachusetts. Thus, the flora and fauna are less
Wcst Indian than they are South American. Recogniz-
ing this, Beebe established Simia in 1950. Beebe pur-
chased the land atan advanced age (he was 73) and
began the research station. His strong belief was that
researchers muse work on live animals and not merely
collect specimens. At that time, specimen collection
for museums essentially dominated tropical research;
thus, Beebe's approach helped change the direction of
ecological work from specimen collection to detailed
ecological studies of organisms in the field.
Many tropical field stations exist today and more
are being established. In the Western Hemisphere
tropics, in addition to Barro Colorado lsland, sorne
of the most prolific research sites are La Selva (Costa
Rica), Cocha Cashu (Peru), Manu Biological Preserve
(Peru), Mínima! Critica! Size of Ecosystems (Brazil),
Los Tuxtlas (Mexico), and Luquillo (Puerto Rico).
The Tropical Ecology Assessment and Monitoring
Network (TEAM) includes 122 sites throughout the
tropical world, from Colombia to China (see hnp://
PLATE 1-6
The OTS field srarion ar La Selva,
for Tropical Studies (usually referred to as OTS) main-
tains not only the La Selva Biological Station in Costa
Rica (Plate 1-6) but also biological stations at Las
Cruces and Palo Verde. Many other research stations
also are found throughout the rropics, sorne of which
will be noted for studies described within chis book.
In many areas within the tropics, huge challenges
ranging from marginal living conditions, lack of facili•
ties, and political instability still face the field worker.
Nonetheless, it is fair to say chat many if not most field
sites now have sufficient infrastructure and facilities
ro permit cutting edge research to be conducted. What
was, once che age of exploration and description has
become the age of experimental design, data collec-
tion, and hypothesis testing.
Modern Tropical Ecology
Many college-level tropical ecology courses now in-
elude field opportunities at one or more research sta-
tions somewhere within the world's tropical regions.
The very fact that undergraduate and graduare stu-
dents now routinely visit and work at tropical re-
search stations has vastly increased tropical ecological
research.
7 12 CHAPTER 1 WHAT ANO WHERE ARE THE TROPICS? 13
1 ,•••• •••••• •• ••• ••••••••• ••••••••••• ••• ••• ••"••••••oO• •• •• oO • • ••••••••• ., • • • •• ••••••• •• • •••• • •• • ••••• • • • ••••••• • •• • •••• • •••••• ,OO•• • •• • ••••••••oO••••••,O • • ••,O•••• ••,o • • ............... ······· ·············•··············•························ ···•·······••·························· ····························
1
....... .. ,, ..
scudies about ali tropical regions, but the majority of The View from the Canopy ·········"····· ················••····· ······ ···················•························•·········
che work reponed is about the Neotropics. Another
important journal is the journal of Tropical Ecology,
firsc published in 1985. This journal, like Biotropica,
publishes research papers about ali tropical regions,
but most papers tend to be about Africa and Asia.
Taken together, a recent survey showed that the high-
est percencage of papers in those cwo journals were
from scudies conducted in Brazil, Costa Rica, Mexico,
Panama, Malaysia, Puerto Rico, Australia, French
Guiana, Venezuela, Ecuador, Uniced $taces, Peru, In-
donesia, Colombia, and India, in that order (Stocks et
al. 2008) (Figure 1-2). The dearth of scudies from Af.
PLATE 1·7 rican nations is obvious, as is the dominance of studies
The harpy eagle (Harpía Harpyia) released on Barro Colorado Island, from the Neotropics.
Panama, is monirored by radio tracking.

Professional researchers work many months at a The Tropical Rain Forest-

time at tropical research stations with well-equipped
laboratories, computers with Internet access, global
A Pioneering Book
positioning systems, and so on (Piare 1-7). Expedi- The study of tropical ecology became formalized in
tions into remete areas where services are few to none 1952 when P. W. Richards, a botany professor from
still occur, however, as much of the tropics still de- Great Britain, published The Tropical Rain Forest: An
mands field work away from major research stations. Ecological Study (Cambridge, UK: Cambridge Uni-
Far a wonderful and insightful look into how tropi- versity Press). As the title suggests, this book focused
cal research is conducted and the dedicarion and drive on rain forests bue also included discussions of coastal
r1r,uRE 1·3
of the researchers, read The Tapir's Morning Bath by mangrove forescs and savannas. Reílecring the field 1 nis sketch illustratcs rhe use of a large industrial crane ro examine tropical forest canopies: a is rhe tower, b is the counretjib, <is the
Elizabeth Royce (Boston: Houghton Mifflin, 2001). of ecology as it was at that time, che book is largely , o,nrerweight, d is rhe operator's cab, ande is the gondola, where a rcsearcher would be located.
lt chronicles che life and dedication of researchers at
Barro Colorado Island.
404
In the Western Hemisphere, most tropical re- 30 Tropical research has been facilirated in recent years by construction, allow mobile access to the canopy (Parker
searchers focus on the Neotropics, their work taking "t> ¡vwers, walkways, and cranes that allow direct access et al. 1992) (Figure 1-3; Piares 1-8 to 1-11).
¡1 them anywhere from Mexico to southern Amazonia f" 25 ro the foresr canopy. Ir has been frusrrating to tropical
as well as to the Caribbean islands. In the Eastern :¡; .cologists to realize that much of the biological diver-
Hemisphere, much research is carried on in Africa and 5. 20 siry found in tropical forests is confined to the forest
Australasia. The intensiry of tropical research varíes ~ 217 canopy, the dense and irregular !ayer of foliage repre1;:
considerably from one country to another, and thus 'g 15 senting numerous tree species, usually located 30 to 40
ÜÍ 153 merers (about 98 ro 130 feet) from the ground. Seores
there are sorne areas that are far less studied than oth- o 10 ,,f arthropod and other invertebrate species as well as
ers (Stocks et al. 2008). e~
96 96
birds, mammals, reptiles, and amphibians spend most
Professional papers detailing tropical research are ~ 68 62 51 4 · 4'
o b403938333230 t ali of their lives well above ground leve!. The same
routinely published in leading science journals such "
ll.
is true for epiphytes, the myriad species of bromeliads,
as Science and Nature. In addition, journals such as tchids, cacti, and other plant species that live on the
American Naturalist, Ecology, Oecologia, Conserva- bark of tree trunks and branches. (Epiphytes live on
tion Biology, and many others routinely include pa- bark and branches of other plants but are not directly
pers about research from che tropics. ,,arasitic.) But ecologists now do havc more access to
The first professional society devoted to tropical the forest canopy. The construction of high towers,
research was the Bombay N¡tural Hiscory Sociery, orne associated with tall, emergen! ttees, allows one to PLATE 1·8
founded in 1883. The lnternational Sociery for Tropi- FIGURE 1-2 ht at canopy lcvcl and make observations. Even more CANOPY WALKWAY ANO FOREST
cal Biology was founded in 1956 (Chazdon 2002). This graplúc depictS che number oí srudies published in two proíes- access is provided by canopy walkways that connect This rall and lengthy canopy walkway ar Sacha Lodge in Ecuador
sional journals, Biotropica and Joumal o/Tropical Biology, from ,arious trees or that are supported by metal towers. provides saíe and easy access thar permits srudy of species largely
The Association for Tropical Biology and Conser- 1995 ro 2004. [t shows how many srudies were conducred in each In sorne areas, tall cranes, such as are used in building conlined ro rhe forest canopy.
vation, founded in 1963, publishes the journal Bio- oí various counrries. Noce thar mosr srudies are íocuscd in rhe Neo•
tropica. The papers published in Biotropica include rropies. (Nurnbe~ on bars are actual numbers oí published srudies.)
 14 CHAPTER 1 WHAT AN D WHERE ARE THE TROPICS? 15

.......... .................. .............................................. ......................

Canopy research has helped not only to document cal variables, such as elevated carbon dioxide levels and
patterns of biodiversity but also to demonstrate how the production of biogenic aerosol compounds (Ozanne
tropical tree species vary in their response to physiologi- et al. 2003).

1 !1
FIGURE 1-4
This is a sketch of a forest
profile in what was then
British Guiana and is now
the independent oation of
The expanse and structural complexity of Ecuador's lowland for- Guyana. This is adapted from
est is readily evident, as observed from a canopy walkway. one of severa! classic forest
Viewed from the ground, the canopy walkway is seen crossing a profiles induded in Richards's
major forest gap. famous book The Tropical
Rain Forest.
....................................................... ................................................................................
, ······················• ··················

PLATE 1-11
WOODEN CANOPYTOWER
Canopy towers allow more restricted access than walkways but,
if strategicaUy placed, afford excellent opportunities for canopy
study. Note the bread branching pattern of the tree. The tower
is essentially built around the tree. From Sacha Lodge, Ecuador.

············· ········································ .............., ........

descriptive, but it does an admirable job of compar- sizing concepts. It paved the way for much research,
ing global equatorial forests and identifying common and it is still valuable and insightful. This classic book
patterns-particularly of physiognomy, climate, and was revised and a second edition was published in
soils- that they share to varying degrees (Figures 1-4 1996 by Cambridge University Press.
and 1-5). What is obviously missing from the discus-
sion is treatment of the complex roles of animals in FIGURE 1-5
rain forest ecology. Pollination and seed dispersal- Geographi c Definiti on of the Tropics Also adapted from Richards's
both of which, in the tropics, are usually dependent In today's world, the tropics are the region located classic volume, rhis is a profile
on animals-are not listed in the index, for example, of a rnixed Dipterocarp forest
between the Tropic of Cancer (23º27'N) and the
in Borneo. Compare this with
and are barely touched upon. Richards's book was Tropic of Capricorn (23°27'S), a latitudinal band of Figure 1-4. Both forests are
nonetheless the first serious attempt at sumrnarizing approximately 47°. This belt of laritude encircles the strucrurally similar bur contain
knowledge about rain forest ecology as well as synthe- Earth at its widest circumference. At either extreme, emirely diiferent tree species.
16 CHAPTER 1
........................... ········· ·········" ······.. ······.. ...... ··············.... ·... ·.................................. .
• Tropical Worlds Long Ago: Rain Forest in Denver
and aVery Large Snake from South America
Denver, Colorado, is located along che front range of
che Rocky Mountains, an area well within che temper-
ate zone. Forescs of spruce, fir, pine, and aspen charac-
terize the region. Bur had Denver existed 64.1 million
years ago (a rime called rhe Paleogene), not long after the
mass extinction rhat ended che Cretaceous period, a very
different forest would be evident-a tropical rain forest
(see Appendix). Ata site called Castle Rock, near Denver,
fossil planes (particularly leaves) have been unearthed,
plants that are clearly tropical Uohnson and Ellis 2002)
(Figure 1-6). The diversity of tropical planes at this fos-
sil site shows that tropical conditions then prevailed at
a region far from today's tropics. Plant diversity had re-
FIGURE 1·6
This drawing illusttates the morphotypes of tropical leaves found near Casrle Rock, Denver, dating back 64.l million years. Only the
leaves in the upper right are nondicot leaves. By this time, modern plants, dominated by dicots, dominated tetresttial plant communities.
,
....... .
covered quickly (within 1.4 million years) after the mass
extinction that ended rhe Mesozoic era.
The largest snakes found in the tropics are the py-
thons (Pythonidae) of Africa and Australasia and che
boas and anaconda (Boidae) of the Neotropics. The yel-
low anaconda (Eunectes murinus) of the Neotropics is
arguably the world's largest extant snake, reaching a
length of 9 meters (about 29.5 feet) (Piare 1-12). The
great lengths reached by pythons and boas are facilitated
by the warm tropical climate in which they live (Head
et al. 2009). Snakes are poikilothermic, ectothermic
vertebrates, meaning that their body temperature and
rhus their metabolism is dependent on ambient air
w;.perature. (Poikilothermic and ectothermic are dif-
ferrnt in meaning. Poikilothermic means specifically
thar body temperature changes as ambient temperature
ch,, ,,ges, whereas ectothermic means that the animal
<loes not have physiological ability to regulare its body
tem;,erature to stay warm.) Mathematical models have
ecos¡;tems become subtropical. Tropical regions re-
pres~nt about half of Earth's surface and thus has
an irnmense effect on Earrh's climare (Huber 2009).
Fo, much of Earth's history, even before the inicial
evolution of flowering plants (angiosperms) in the
mid- ro late Cretaceous period (144 ro 65.5 million
years ago), Earth has been even more tropical rhan it
is today (see Appendix). Fossil palm trees and croco-
diles have been unearthed in what is now Alaska and
Siberia (Huber 2009), This is because the climate of
Eanh, for a variety of reasons discussed in che next
chapter, is not constanr. In times when climate was
WHAT AND WHERE ARE THE TRDPICS? 17
been developed that correlate the difference in maxi-
mum body size of snake species occurring in different
regions with mean annual temperature of each region.
The warmer the mean annual temperarure is, the larger
the snakes are able to grow. Thus, it was of great interese
when a fossil boid snake dated at 58 to 60 million years
old was unearthed from a site in northeastern Colom-
bia and the fossils (vertebrae) indicated that the snake
reached an estimated length of 13 meters (about 42,6
feet). This serpent, named Titanoboa, was estimated to
have weighed about 1.27 tons. That's a lot of snake,
When included within the model that correlates today's
boid snakes and air temperarure, the results suggest that
Titanoboa required a mínimum annual mean tempera-
ture of berween 30ºC to 34ºC (86ºF to 93.2ºF) to sur-
vive. This would make che climate at that time about
6ºC to 8ºC warmer than it is currently and suggests that
.a much higher leve! of ~arbon dioxide (a gas that con-
tributes strongly to retention of atmospheric heat) was
present in the atmosphere when Titanoboa lived,
These two examples show that the distribution and
intensity of tropical climate have changed throughout
Earth's history, as it appears to be doing today,
warmer, the distribution of che tropics was globally
broader than is rhe case today.
Well before the evolution of flowering planes,
the world was largely tropical. In the Jurassic pe-
riod (206 to 144 million years ago), che giant long-
necked sauropod dinosaurs lived in a largely tropi-
cal world, where there was no polar ice and where
the average global temperature as well as oxygen
and carbon dioxide levels exceeded those found to-
day (Piare 1-13).
During the Cenozoic era (65 million years ago
.
u~til the present), a time of proliferation of insects,
PLATE 1-13
CAMPTOSAURUS, APATOSAURUS,
STEGOSAURUS, DRYOSAURUS,
CAMARASAURUS ANO ALLOSAURUS
(LEFT TORIGHT)
Each of rhese dinosaurs cohabited tropi-
cal regions in western North America
during the late Jurassic period. The
tropics then would bear scant resem-
blance to the tropics of toda y.
 18 CHAPTER 1
Saceilite image of Earth, showing che band of cropics around che equacor.
flowering plants, and mammalian diversity, global cli-
mate cooled, begiiming strongly during che Miocene
epoch (23.8 to 5.3 miUion years ago) and accelerat·
ing during the Pleistocene (1.8 million to 8,000 years
ago), commonly referred to as ice ages (see Appendix).
During times of cooling, tropical forests presumably
shrank in area, replaccd by other more seasonal tropi•
cal ecosystems such as grassland, savanna (an arca
$0Uftf
P,\CIFIC
OC(.UI
~OVtN ( RII
~IGURE 1-8
1 ;he Kornodo dragon, which sometimes reaches a length of nearly
l 0G tical
· map of Earth, showmg
· nacions that occur wichin the cropics.
........ ····•·" ''
where grasses predominare but with varying amounts
of scattered trees), dry forest (forests that experience
significant dry season), or deciduous forest (forests
where most trees synchronously shed leaves for part
of the year). Earth is currently considered to be within
an interglacial period, and there is strong evidence for
climate change due to human activities, a ropic to be
considered later in this book (Figures 1-7 and 1-8).
i considered Jacer in the texr.
0C f ,J,N
WHAT ANO WHERE ARE THE TROPICS' 19
B,•cause of rhe geological proccss of piare tectonics half of lndia is within the tropics, and it is unique be-
(des.;ribed in the following chapter), the Earth's con- cause India separated from Gondwanaland (a massive
. •.. unired 248 mdlton years ago at the boundary southern continent once consisring of South America,
r1ne11.3 , .
of che Paleozoic and Mesozo1c eras, have gradually Africa, Madagascar, India, Antarctica, Australia, and
fragrnented and separated, with large areas o_f ocean New Zealand) in the early Cretaceous period (about
· olatÍnºO them ro varymg dcgrees. Th1s separat1on mto 140 rnillion years ago) and drifted asan island unril
IS .
ass,ve continents and varymg-s1ze. d 1s. 1ands has re- it united with Asia about 40 million years ago, in the
: lted in evolutionary consequences, particularly rhe early Cenozoic. Likewise, Madagascar has been iso-
svolution of endemic species. An endemic species is lated from the rest of Africa to the point where it has
: species whose range is confined with_in a limited numerous endemic species such as the primate group
geographic area. For example, the _1mposmg Komodo known as lemurs (Lemuridae).
dragon lizard (Varam,s konodoensts) 1s endem1c to ln- Each of these realms contains unique assemblages
dones·a, specifically to che Lesser Sunda lslands, most of plant and animal species (Primack and Corlett 2005).
particularly to Komodo (Piare 1-14 ). Biogeographers
recogmze various b10geograph1c realms throughouc
che planet. Piare tectonics and biogeography will be Tropical Climate
discussed in greater detail in che ncxt chapter.
Middle America and South America, along with · In two words, the tropics (especially where there is
rhe various islands of the Caribbean and tropical At- lush forest) are warm and wet (Figure 1-9). Tropical
lantic Ocean, constitute the Neotropic realm, the so- regions experience high average temperature, and most
called \!ew World tropics. The Paleotropic realm, or (dry grassland and deserrs being exceptions) experi-
Old World tropics, comprises Africa and Madagascar ence substantial precipitation in rhe form of rainfall.
(Malagasy Republic), southern India, and Australasia. Tropical ecosystems experience seasonality in the form
The ]:;¡rer is further divided inro the Sourheast Asian of varying and often predictable amounts of rainfall
tropics (extreme southern China, Myanmar [Burma], throughout the year. In conuast to the seasonal pat·
Viern2in, Laos, Thailand, Malaysia, Sumatra, Borneo, cern chat typifies the temperare zone, the cropics vary
Philip:,ines) and the Australian tropics (including Su- primarily in precipitation amounts, not temperature
lawesi and other islands of eastern Indonesia as well flucruation. There are no periods of protracted cold,
as New Guinea). The poim of separacion berween the no frost, no snow in tropical ecosystems except at high
biogeographic realms of tropical Asia and Australia elevations. Bue there are distinct and pronounced dry
is cal1ed Wallace's Line, after Alfred Russel Wallace, and wet seasons. lf the dry season is moderare, forests
who first described it (see Cha peer 2). Biogeographers will remain evergreen throughout the year. If che dry
differ on exactl y where to place Wallace's Line be- season is protracted and severe, forests will have less
cause plant species are not as sharply separated by biomass and more deciduous species, or the ecosystem
it as animal species (Whitmore 2002). The sourhern will not be forest bue will be savanna, shrubland, or
' grassland. The ecosystem type of a given region is also
s¡rongly influenced by soil characteristics, a point co be
Les lie Holdridg e and the Lite
Zone Con cept
One of the most significant papers in the science of
climatology and ecology was published in Scíence
by Leslie R. Holdridge in 1947. The riele of the pa-
per, "Determination of world plant formations from
simple climatic data," addressed a profound truth:
lt is possible by knowing three broad climatic vari-
ables to predice what major kind of terrestrial eco-
PLATE 1-1 4 system will prevail in any given geographical rcgion.
Those variables are (1) mean annual biotemperature
(defined as the average air temperature after values
metel'.$ J9.8 feet), is endemic to the Lesser Sunda lslands of Indonesia.
 20 CHAPTER 1 WHAT ANO WHERE ARE THE TROPICS? 21

Te:nperatJre ('F1 R;·n (in.) ¡~;,.,, r¡~,~

Temperature ('F) Rain (in.)

1
' 1
'I ' AVERAGE TOTAL
700 Daily maximum tcmperature 20 100 ~ : 1 ' ' ., l t.• ¡ -: .-- ANNUAL
., 1 .. 1
.~~EC[P~ATION (CM)
Daily mi11imum temperature Daily maximum
75 ---- ----...... _--------------- 13 75 lemperature 15 <1.5 0.125-1.5 6.25- 75
Polar
, ......,-- .. , Subpolar (alpine) Tundra
50 50 ,, ''
10
, / Oaily minfmum',, 10 Dry 1.5-3 1-2 Subhumid 6.25-12.5
Rain/al: / temperature ', Moist J.5-3 0.5-1 Humid 12.5-25
, ''
25 5 25 ,, ' Wet 1.5-3 0.25-0.50 Perhumid 25-50
---- - ~,/ RainíaH ', 5
Rain 1.5- 3 0.125- 0.25 Superhumid 50-100
o'---'--'--'---'--'--'---'-_,__-'-_,__,___, o'---'--'--'---'----'-'---'---'--'---'--'---' 3-6 2-4 Semiarid 6.25- 12.5
J F MA ~ J J A S O N D J F MA M J J A S O N D Bore~l (subalpine) Desert
Belén,, Brazil Toronto, Canada Boreal Dry scrub 3-6 1-2 Subhurnid 12.5-25
FIGURE 1·9 Forest
Boreal
These graphs contrast the climate in tropical Brazíl with that of temperate Canada. Note the strong seasonality of the tropics with regard to
precipitation bue che s1eady temperature that prevaíls in the tropics. The opposite pa11ern charac1erizes the temperare wne. Moist Puna 3-6 0.50-1 Humid 25-50
Wet páramo 3- 6 0.25-0.50 Perhumid 50- 100
1I below 0ºC or above 30ºC are removed); (2) total an- each of which is positioned according to where it fits Rain páramo 3-6 0.125-0.25 Superhumid 100-200
nual precipitation; and (3) ratio of mean potential among the varia bles, each of which represents one Desert 6-12 4-8 Arid 6.25-12.5
1,1 Cool temperate
evapotranspiration (a function of moisture and tem- side of the triangle (Figure 1-10).
perature) to mean annual precipitation. Holdridge Cool temperate Desert scrub 6-12 2-4 Semiarid 12.5-50
1.
Holdridge's 38 life zones were broadly divided
illustrated the concept with a triangle of hexagons, into polar, subpolar, boreal, cool temperate, warm Cooi temperate Tundra dry 6-12 1-2 Subhumid 25-50
Co,·i temperate Forest
Moist 6-12 0.5-1 Humid 50-100
Wet 6- 12 0.25-0.5 Perhumid 100-200
Rain 6-12 0.125- 0.25 Superhumid 200-400
Subopical Desert 12- 24 8- 16 Perarid 6.25- 12
Subtropical Desert scrub 12-24 4-8 Arid 12-25
Subtropical Thorn woodland 12-24 2-4 Semiarid 25-50
Latitudinal Altitudinal
regions belts Subtropical Forest
Polar Alvar
Dry 12-24 1-2 Subhumid 50- 100
1.5ºC Moist 12-24 0.5-1 Hurnid 100-200
Subpolar Alpine
Wet 12-24 0.25- 0.5 Perhumid 200-400
Rain 12-24 0.125- 0.25 Superhumid 400-800
3'C
Boreal Subalpine
Tropical Desert >24 16- 32 Superarid 6.25-12
Desert scrub >24 8-16 Perarid 12- 25
6'C
1
>

Cool temperale Monlane

Tropical Thorn woodland >24 4-8 Arid 25- 50
Tropical Forest
~ 12'C
Warm lemperate

t
<§5 Lower montane 2-4 Semiarid 50-100
I Very dry >24
Sublropical / ~ Premontane Dry >24 1-2 Subhumid 100-200
Tropical
.Rain f 24ºC
~~'-"-~~"---'-=~~~-'-'='--"--=
lor~est'--'.: Critica! Moist >24 0.5-1 Humid 200-400
·. Superarkf ·. Perarid Arid ·. Semiarid Humid · Perhumid .,Supeffiumid temperalure
line
Wet >24 0.25- 0.5 Perhumid 400-800
Humidity provinces
FI GURE 1-10 Rain >24 0.125-0.25 Superhumid >800
This is the famous Hofdridge diagram that iliustrates the relationship between ecosys1em types as determined by la1itude, elevation, and com-
bination of precipitation and temperature. FIGURE 1-1O
(continued)
 22 CHAPTER 1 r ·····••'
WHAT ANO WHERE ARE THETRO PlCS? 23

- 15
21

-10 19

-5
Co'd

e
2
E 1
"o.
E i
2 \(J) 1
¡¡ 1::r 1
~ 10 !~ \
e \óí1111
~ 1s 1R \ r
g 1s

l\
!
:a 12
20W 1 (-~a 1
• ~n-- -+-+-+'- - -
20 i? g- \--i (~avanna) \ PLATE 1-16

FIGURE 1· 11 ;:,.~ ¡g;:,. ¡ií : !

Diagram showing the relationship between
ecosystem type, mean annual tcmpera-
rure (ºC), and mean annual precipitation
(ccntimeters). Note that tropical rain !ores,
occurs where both d imatic variables are
highest.
temperare, subtropical, and tropical. Eight life zones
are represented in subtropical and eight life zones in
tropical latitudinal regions. The tropical life zones are
.. tropical desert, tropical desert scrub, tropical thorn
woodland, tropical very dry forest, tropical dry forest,
tropical moist forest, tropical wet forest, and tropical
rain forest (Figure 1-11).
PLATE 1-15
Where rain/ali is highly seasonal, tropical
ecosystems are ofren dry !oresr, with many
species of deciduous teces. Forest srarure
is generally small, and trees are usually
widely spaced. From Venezuela.
¡3 1(12) \
Tropical
Northern Australia is an area of vast
17 -~ , 1 seasonai dry foresr with tall, scattered termite
2 0
~ \2 1 9: l forest mounds. This is the ecOS)'Stem typi-
\a 9 : 8 ~ i 2 cally called 011tba,k.
\..._ l C. j
30L-l..:::.:::.:::;.;;;~~====:;:'.::=::::;:==:==-~~
t:O0 450
24ºC (75.2ºF), has a potential evaporation ratio of Savannas and dry forests (Piare 1-20) will be discussed
1:2 (defined as the ratio of annual potential evapo- in detail in Chapter 11.
transpiration [PET] to mean total annual precipita- Ir is importan! to realize that Holdridge's life zones
Consider the difference between tropical rain for- tion-this meaos that dry ecosystems in tropical re- are not merely latitudinal but also altitudinal. A walk
est and tropical dry forest, both of which may occur gions ha ve higher ratios and humid ecosystems lower from low Andean rain forest up to an elevation of
at the same latitude. Dry forest occurs where there is ratios), is subhumid, and has an average total annual about 3,500 meters (about 11,483 feet) will take you
a significant dry season for part of the year. Tropical precipitation of between 100 and 200 centimeters through severa] life zones, from hot, humid, dense for-
dry forest (Piares 1-15 and 1-16), under Holdridge's (39.4 to 78.7 inches). In comrast, rain forest has a est to cold, windswept páramo, a vastly different
classification, is defined as an ecosystem that experi- mean annual biotemperature also greater than 24ºC
ences a mean annual biotemperature of greater than but has a potential evaporation ratio of only 0.125
to 0.25, is superhumid, and experiences greater than
800 centimeters (315 inches) of average total annual
pre.:ipitation. Notice that Holdridge's classification
makes true rain forest an extremely wet ecosystem.
In fact, most rich lowland tropical forests routinely
called rain forests tend to fall imo what, in Hold-
ridge's classification, would be termed tropical moist
or tropical wet forests (Figure 1-11; Piare 1-17).
Sorne tropical ecosystems may be much drier
than dry forest, with less annual precipitation, most
of it highly seasonal. Tropical thorn woodland (un-
der Holdridge's classification) includes ecosystems
that are dominated by grasses with varying densities
of trees scattered throughout. These kinds of ecosys-
PLATE 1-17
tems are called sauannas and are found in most tropi-
Rapidly growing rree species such as Cecropia rrees, shown
cal regions. The best known savannas occur in Africa,
here in the loreground, typify the forest edge, where vegecation
where the last of the world's megafauna (diversity of scructure is dense. This would be considered tropical moisr forest
large animals) is still found (Piares 1-18 and 1-19). in the Holdridge classificacion. From Trinidad.
24 CHAPTER 1 WHAT ANO WHERE ARE THE TROPICS? 25

...........
............ ........................................................................ .......................... ........ .... . ..
,

ecosystem. Likewise, life zones would markedly change F }~·:..r"-1¡:~1:;r n1J

if you were in Tanzania and scaled Mount Kilimanjaro Mete Ar,nua! P-2J p1!,. < ~ro ,
iemperature (= sDil t3mfJ:~-a;1r3)
(whose summit is 5,895 meters, or 19,341 feet). '-04:
1.5'
Differences in elevation result in differences in
average temperature and precipitation, and that, in t'
4()()()
turn, results in different ecosystems. Within the South
5'
American country of Venezue)a, for example, severa! 7'
major elevational ecosystem types are apparent. At 3000 . 9'
highest elevations (what would be described as al-
13'
pine), a wet and windswept shrubland called páramo
(Figure 1-12) is found. Lower in elevation is forest 2000
1r
of low-stature gnarled trees, and lower still, is cloud
forest, a forest shrouded in cool mist much of the 21'
PLATE 1-18 1000 FIGURE 1-12
time. Below cloud forest ma y be semievergreen for-
Herds o/ wildebeest move with rain/ali patterns in a vast seasonal This diagram shows the discribution
est, deciduous forest, or thorn forest, depending on 25'
of ecosystem types in Venezuela
migration on the Easr African savanna. From Tanzania.
soil and water availability and local seasonal patterns as thcy relate 10 elevation and
(Figure 1-12). Tropical rain forest occurs in lowland climatic variables.
areas where there is ample precipitation throughout ........•........ ...
'

the year. Elevation, average annual temperature, and

average annual precipitation coupled with seasonality 2. Poly/epis is a unique endemic tree genus that forms to southern Argentina. However, for a variety of
ali combine to determine what sort of ecosystem will forests within relatively sheltered areas of the An- reasons, Polylepis forests are much reduced from
be present. des Mountains, normally at elevations of around their former abundance and are considered en-
What follows are sorne brief examples of eco- 4,500 meters (14,764 feet). The trees typically are dangered ecosystems.
systems along an elevational gradient of the Andes short in stature, gnarled, and twisted, and they of- 3. Puna is a grassland ecosystem of the high Andes
Mountains, beginning at high elevation and descend- ten grow on slopes, as shown in Place 1-22. The Mountains (mean elevation about 4,000 meters,
ing to wet forest. These examples are meant only to trees are dense with epiphytes, and the forest is or 13,123 feet). In Peru and Chile, it is called al-
introduce these ecosystems. They will be treated in usually bathed in a cool mist. Many kinds of ani- tiplano. lt is much drier than páramo, with less
greater detail in subsequent chapters. mals tend to specialize in living among Polylepis, precipitation and constant winds that stimulate
and Incan peoples make much use of the trees for continuous evaporative water loss. About 75% of
1. Páramo is a high-elevation (between 3,800 and various medicinal purposes. There are 15 species ali precipitation occurs from December to March.
PLATE 1·19
5,000 meters, or 12,467 to 16,404 feet) ecosys- within the genus Polylepis, ranging from Venezuela ln sorne areas such as Chile, the mean annual
Savannas are open areas dominated by grasses and scattered trees,
particularly in che genus Acacia. Here gira/fes are /eeding on acacia tem dominated by grasses and shrubs. Climate
leaves on the East African savanna. From Tanzania. is generally cool and wet, with frost often oc-
curring at night. In many regions, the dominant
grasses are collectively called tussock grasses,
their blades typically sharp (Plate 1-3). Various
shrubs and ferns also are found in páramo. Es-
peletias are unique shrub-sized members of the
Composite family (a huge family that includes
daisies, goldenrod, and asters) (Piare 1-21). The
/lowers of Espeletias, like those of other /lower-
ing plants at high elevation, attract a diversity
of hummingbird and insect species. Páramo re-
gions are typically wet with scattered acidic bogs
of heath-like vegetation. Because of the relative
isolation among various mountain ranges, much
endemism is evident in páramo plant species.
PLATE 1·20
Sorne high-elevation ecosystems in Africa, Indo- PLATF 1-21 PLATE 1·22
Austcalia has many endemic plant species, including hundreds o/ nesia, and New Guinea are structurally similar Members of rhe Composite /amily, Espeletias are unique shrubs of At high elevations in the Andes Mountains where moisrure is
Eucalyptus species and grasses such as Spinifex. This image shows to páramo, but the species composition is quite the high páramo o/ the Andes Moumains. Severa! hummingbird normally abundam, gnarled foresrs o/ Polylepis trees are supported.
dry forest with Spinifex in south central Australia. different. species specialize on feeding on Espeletia nectar. From Venezuela. From Ecuador.
 26 CHAPTER 1
cemperature is only about 5ºC (41ºF). January is
the warmest month, and July is the coldest (re-
mernber that south of the equator, July is a time
of deep winter). Various grasses and sedges, many
endemic, characterize the ecosystem. Among the
most noteworthy animals of puna, the vicuña
(Vicugna vicugna), a member of the carne! family,
can be found in small herds throughout much of
the region.
4. Cloud forest is found at midelevations through-
out the global tropics and varíes considerably in
species composition from one biogeographic re-
gion to another. As the name implies, cloud for-
est is typified by frequent immersion in clouds, a
characteristic that keeps the ecosystem continu-
ally moist (Place 1-23). There are many cree spe- PLATE 1-24
Many tree species are found in cloud forests, along with numerous
cies, and they range in size from short stacure at
epiphytes such as orchids, cacti, bromeliads, and various vines. The
higher elevations to nearly the size of lowland ecosystem is normally cloud-covered for at least part of each day.
rain forest trees at lower elevations. Masses of From Ecuador.
epiphytes of many species live on che boughs and
trunks of cloud forest trees (Plate 1-24). In South rain forest. These forests differ in the amount of
America, along parts of the Andes Mountains, average annual precipitation they receive:
cloud forest is habitat for such species as che en- Tropical moist forest 200-400 centimeters
demic spectacled bear (Tremarctos ornatus). In (79- 157 inches)
East A/rica, in Uganda, cloud forest is habita! for
mountain gorilla (Gorilla beringei), also an en- Tropical wet forest 400-800 centimecers
(157-312 inches)
demic species.
5. Wet forest- when most people think of what they Tropical rain forest > 800 centimeters
call tropical rain forest, they are usually lumping (312 inches)
together three categories within the Holdridge life In spite of these differences, each of these forests is lush
zone classification: moist forest, wet forest, and and they share the characteristic of having uniquely
high numbers of species, not only of trees, buc of many
other plant and animal groups (Plates 1-25 and 1-26).
I will generally treat them together in this book and
not attempt to strictly separare them.
Thesignificanceof Holdridge's classificationschema
is that it demonstrates the extreme importance of di-
mate as an overriding evolutionary selection pressure
on plant growth forms. Plants are sessile and must
adapt to climate. The reason why plants look as they
do, are adapted as they are, and experience differing
patterns of growth changes with latitude and eleva-
tion, is ultimately due to selection pressures generated
by regional climatic variables.
Seasonality in the Trop ics:
A Closer Look
PLATE 1-23
Clouds shroud much of the mountainsides along the east slope of In the Amazon Basin, the very heart of the Neotropics,
the Andes Mountains in parts of South America, suppon ing dense, climate is permanently hot and humid, with the tem-
species-rich ecosystems called cloud forests. From Ecuador. perature averaging 27.9ºC (82.2ºF) during dry season
PLATE f-25
The complexity of srructure and dimsity of species characterizes
tropical moist, wet, arid rain foresrs throughout the world. This
foresc is ii; central Brazil.
and 25.8ºC (78.4ºF) in rainy season. Ic is cooler in
rainy season mostly due to clouds occluding rhe sun.
In che tropics, daily temperature fluctuation exceeds
average annual seasonal fluctuation, and air humidity
is high, being about 88% in rainy season and 77% in
dry season (Junk and Furch 1985).
The range of seasonality within the tropics is
evident when comparing places such as Singapore, in
Malaysia, a region that supports tropical moist forest,
and Darwin, in northern Australia, where the preva-
lent ecosystem is tropical savanna (Figure 1-13). In
Singapore, the average temperature remains at about
27ºC [80.6ºF) throughout the year, every day, every
PLAY( 1·26
The rerm jungle refcrs to a disrurbed area where an abundance of
sunlight results in a dense array of many plant species, often so
thick rhat it is difficult to penetrare withour the use of a machete.
Jungles typify areas of moist forest. From Ecuador.
WHAT AND WHERE ARE THE TROPICS? 27
month. Daily temperature fluctuations are wider rang-
ing than annual fluctuations. Rainfall, as well, is abun-
dant and relatively constant, averaging 2,415 millime-
ters (about 95 inches) annually. In contrast, Darwin is
strongly seasonal, its average rainfall of 1,538 milli-
meters (about 60.6 inches) being largely concentrated
in the wet season between the months of April and
October. Rainfall drops to almost nothing in Novem-
ber through early March. But note that in Darwin, as
in Singapore, temperature remains nearly constant, its
annual average being about 28ºC (82.4ºF).
At La Selva Biological Station in Costa Rica, Au-
gust is the month with the highest mean temperature,
27.lºC (81ºF), while January has the lowest mean
temperature, 24.7ºC (76.SºF). Relative humidity, as
noted earlier, is generally high in the tropics, especially
in lowland rain forest, where humidities ranging from
90% to 95% at ground leve! are common. Humidity
is lower in the rain forest canopy, usually no higher
than 70%.
Tropical regions are warm and generally wet be-
cause the sun's radiation falls most directly and most
constantly upon the equator, thus warming the Earth
more in the tropics than at other latitudes. As one
travels either north or south from the equator, the
Earth's axial tilt of 23º27' results in part of the year
being such that the sun's rays fall obliquely and for
much shorter periods of time, thus the well-known
cycles of day length associated with the chang-
ing seasons of temperare and polar regions. Day
length varies much less in tropical regions. Equato-
rial regions experience close to 12 hours of daylight
throughout the year. At the equator, every day lasts
exactly 12 hours. North of the equator, days become
,a little longer in the northern summer and shorrer in
winter, but this means only that summer sunset is at
6:15 or 6:20 rather that 6:00 P.M.
At the equator, heat builds up, and cherefore rhe
air rises, carrying warmth (Figure 1-14). Water is
evaporated, and therefore water vapor rises as well.
The warm moist air is cooled as it rises, condensing
the water, which then fa lis as precipitation, accounting
for the rainy aspect of tropical climates. The normal
flow of warm, moisture-laden air is from the equa-
tor to more northern and southern latitudes. As the
air cools, it not only loses its moisture to precipita-
tion, but also becomes more dense and falls, creating
a backward flow toward the equator. At the equator,
two major air masses, one from the north and one
from the south, along with major ocean currents,
form the Intertropical Convergence Zone (ITCZ), the
28 CHAPTER 1 WHAT ANO WHERE ARE TH E TROPICS? 29
··••······························································ .................. ............................................ .

.
-----
Dar.vin
----- ------------------------¡,------;}--

Humid tropical domaln

D Savaona d'•,isicn [ ] Ra'n foresl division FIGURE 1-14
111 Savanna re,:;;me m0untain, lf Earth were not rotating, this would be the
(a)
distribution of primary and secondary air
movement of the planer.
lfomi~ tropical
............................
Tropical s;.vanna Tropical rain foros/
Darwln. t\u~tralia (3¿, rr)
mT íl'm
Singapore (3 m)
281 'G 1538 mm 300 27.2 ·C 2415mm 300
20(' 20,;
100 100

80 80
ºC ··e
so 60 3G 60
FIGURE 1-13
These clima te diagrams from {a) Darwin, 20 AJ 2(1 40
Australia, and (b) Singapore illustrate the
extreme contrast between savanna and 10 · 20 10 20
rain forest. But note that temperature is
about the same, and relatively constant in o o o+ - ~ ~ ~ ~ ~ ~ ~~....-+ o
both places. What varíes is the seasonality J F !J A M J J A S O N O F M A iv' J A S ON O
of precipitation. (b)

major climatic heat engine on the planet. The large air
masses that drive che process are termed Hadley cel/s
(Figure 1-15).
Tropical areas fall within the trade wind belts
(so named because winds were favorable for sailing
ships trading their goods) except near the equator, at
che ITCZ. This region is called che doldrums, where
winds are usually light (often becalming sailing ships).
From the equator to 30ºN, the eastern trade winds
blow steadily from the northeast, a direction deter-
mined because of the constant rotation of the Earth
from west to east. South of che equator to 30°S, the
eastern trades blow from the southeast, again due
to the rotacional motion of che planet. As the Earth,
tilted at about 23.5° on its axis, moves in its orbit
around che sun, its direct angle to the sim's radia-
tion varíes with latitude, causing seasonal change,
manifested in the tropics by changing heat patterns
of air masses around the ITCZ that result in seasonal
FIGURE 1-1 5
rainfall (Figure 1-16). In che Western Hemisphere, Earth's daily rotation deílects the Hadley cell
from July throughout October, severe wind and rain ;Gle currents and creates the trade wind belts. Note
storms called hurricanes can occur in parts of the ~ ct0:sLire cnd ·1hi belts ·J! th~ v·or'f'J the lack of wind at the equator.
 30 CHAPTER 1
' ,• ... ............... ,
\
e E,~;--------
, ·;·r::·~ic of
C~p1irorn
1:
FIGURE 1-16
The angle of the Earth relative to the
snn varíes during the course of Earth's
annual orbit, crcating rhe seasonality
that characterizes Earth.
Neotropics. Similar kinds of storms are referred to as
monsoons in the OId World tropics.
At the point where tropics meet subtropics, at
about 30° norch and south latitude, deserts are com-
mon. This is because the moisture contained in the air
masses (Hadley cells) moving away from the equator
has been depleted by the time the air reaches 30º lati-
tude. Thus, skies tend to be clear with low humidity
and deserts form. Longitude also concributes to des-
ert formation. If a landmass is far from the sea, or if
mountain ranges block moisture circulation, moisture
evaporated from the sea will not reach inland areas,
and deserts will result. The combination of 30º lati-
tude and distance from rhe sea produces many of the
world's vast deserts, including the Sabara in Africa.
In the Amazon Basin, precipitation ranges be-
tween 1500 and 3000 millimeters (59 to 118.1
inches) annually, averaging around 2000 millimeters
(78.7 inches) in central Amazonia (Salati and Vose
1984) (Figure 1-17). About half of the total precipi-
tation is brought to the basin by the eastern trade
winds blowing in from the Atlantic Ocean, while the
WHAT ANO WHERE ARE THE TROPICS? 31
··············.. ·················· ·····,,, ..................................................................
······•······ ·•··· ·
........ ·· •• ' .
······
~ \
3
Cit. 2..:;
,'),.; 1 ,, .,.. z..~, ~-,,):'"
e
\
1 '.! , ; ,', .¡
5~2,:J-1
1 1
1
1
1
1
1
1 N
1
1
1
Eq
1
1
1
1
1
1
1
1
\ 1
s
1::: ~:""'
MJ;?·
Sur¡ ove¡ eqws1oí
\
15' FIGURE 1-17
This diagram shows how climate varies in
the tropics. Note the absence of seasonality
:.1.:9 S'.:-:- O'.--• r;o J ~-: at the equaror.
other half is the result of evapotranspiration from
the vast forest that covers the basin (Salati and Vose
1984; Junk and Furch 1985). Up to 75% of the rain
falling within a central Amazonian rain forest may
come directly from evapotranspiration (Junk and This phenomenon has become increasingly evident rupted. When that happens, warm waters flow along
Furch 1985), an obviously tight recycling of water. since the mid-twentieth century, but its exact cause the norrnally cold South American coast. Weather sys-
Much of the reason for the large amounc of recycled remains elusive. El Niño occurs periodically, approxi- te,[11S change, resulting in heavy downpours and flood-
moisture lies in the nature of the forest itself. This mately every two to seven years, when a high-pressure ing in sorne regions and droughts where there should
vast assemblage of trees transpires far more moisture weather system that is normally stable over the east- be rainfall, effects that range ecologically from mildly
than other kinds of ecosystems such as savannas or ern Pacific Ocean breaks clown, destroying the pattern stressful to highly significant.
agricultura! land. The forest functions such that of thc westward-blowing trade winds. Trade winds The El Niño of 1982-1983 was considered up
precipitation and the water-recycling system are es- weak~n, sornetimes reversing from their normal west- to then to be the most powerful of this century, esti-
sentially in equilibrium, though large-scale defores- ward direction. Warm water from the western Pacific mated to have caused S8.65 billion worth of darnage
tation (a topic that will be discussed in more detail flows eastward, causing an influx of abnormally warm worldwide. There have been severa! other El Niños
later in Chapters 13 to 15) could significantly upset Water to the western coast of South America. The wa- since 1982, eight major El Niño events since 1945,
the system (Solati and Vose 1984). ter column reaches a depth of up to about 150 me- and at least 20 during this century. A severe El Niño
ters (about 500 feet), and because it is warm, it is less occurred in 1986-1987. Another El Niño occurred in
dense and thus flows over and blocks the colder, more- 1994-1995, comparable to those of 1982-1983 and
Short -Term Climate Change: ENSO nutrient-rich waters below. This prevents the upwell- 1986-1987. Satellite data indicated that the northern
Short-term but sometimes majar climatic changes ing of nutrients into the upper water column, where Pacific Ocean was nearly 20 centimeters (8 inches)
also occur in the tropics, the most noteworthy at- they would normally be available to phytoplankton. higher than normal, dueto the influx of warm surface
tributed to the El Niño/Southern Oscillation (ENSO). The result is that oceanit food chains are severely dis- waters. Yet another El Niño occurred in 1997-1998,
l 32 CHAPTER 1
• El Niño Comes to Panama
A major El Niño affecred Barro Colorado lsland in Pan-
ama during 1982-1983 (Leigh 1999). One of rhe mosr
drama tic effecrs of the El Niño was rhe failure of trees co
produce fruit during the second of two fruiting seasons.
The failure of the fruir crop resulted in a cascade of se-
vere effecrs on various anirnals. Normally wary species
such as collared peccaries {Tayassu tajacu), coatimundis
{Nasua ilarica), tapirs {Tapirus baridii), and kinkajous
{Polos fiavus) all made regular visits to the laboratory
area where food had been pur out for them. Sorne of
these animals appeared emaciated, obviously under stress
from lack of food (Plate 1-27).
One researcher wrote
The spider monkeys, which normally visit rhe laborarory
clearing ar least once every day, now launched an ali-out
1
1
1 the west coast of South America. The cessation of an at multiple time scales: short-term, as is the case with
1
1
1 El Niño/Southern Oscillation, or ENSO). Though with major global climate change.
(a)
PLATE 1·27
(a) KINKAIOUANO(b) SPIOER MONKEY
,.,.......
WHAT ANO WHERE ARE THE TROPICS? 33
assault on food resources inside rhc buildings, leaming
for the first time ro open doors and make quick forays
to the dining room table, where rhey sought bread and
bananas, ignoring the meat, poraroes, and canned fruit
cocktail, and brushing aside rhe startled biologisrs at their
dinner (Foster 1982).
Dead animals were also encounrered far more frequently
than usual:
The mosr abundant carcasses were rhose of coaris,
agoutis, peccaries, howler monkeys, opossums, arma•
dillos, and porcupines; there were only occasional dead
two-toed sloths, three-roed slorhs, white-faced mon-
keys and pacas. Ar rimes ir was difficulr ro avoid the
stench: neirher the rurkey vulrures nor rhe black vul-
tures seemed able to keep up with the abundance of
carcasses (Fosrer 1982).
FIGURE ·¡ -18
This map shows the general locarion of the lTCZ.
and its combined global effects are estimated to have normal, carrying warm surface water toward Asia.
resulted in 2,100 human casualties and property dam- This creares enhanced upwelling of colder, deeper,
age tetaling a staggering $33 billion (Suplee 1999). more nutrient-rich water along the South American
In 2009-2010, still another El Niño altered weather coast. Because the colder waters evaporare more
patterns around the world. slowly, rain is reduced and drought may result. La
The causal factors responsible for the periodic- Niña followed the 1982-1983, 1986-1987, 1995,
ity of El Niños are thus far unknown (Canby 1984; and 1998 El Niños.
Graham and White 1988), but it is clear that the In- Ecologists realize that ENSO events cause short·
tertropical Convergence Zone (Figure 1-18 ), a com- term but nonetheless significant perturbations in eco-
plex system of oceanic and air currents, migrares, systems. ENSO events represent one of severa! major
to a lower latitude, raising sea surface temperatures global climatic patterns that may vary periodically, af-
and destroying the normal upwelling pattern along • fecting tropical ecosystems. Climate variations occur
El Niño occurs when the ITCZ returns northward an abnormally cold or snowy winter; moderate, as is
to its normal position (hence the alternative terrn che case with ENSO events; or long-term, as is thecase
El Niño has global effects, tropical ecosystems in
panicular can be anywhere from moderately to se-
verely affected.
El Niños tend to alternare with another climatic
The Scope of Tropical
{b)
phenomenon rhat produces largely the opposite ef-
fecrs and is called La Niña. Like El Niño, La Niña
Ecology
systems begin when normal trade winds are altered. The word tropics is often synonymous with tropical
In this case, however, the westward trade winds gain rain forest, bue such a linkage, as I hope chis chapter
abnormal strength and move farther westward than makes clear, is too limited. Tropical rain forest is one
7
34 CHAPTER 1 WHAT AND WHERE ARE THE TROPICS? 35
·········•"'
················ ............................................. , ···· ··· · · •"''" ''' " ' ' " " ' " ' "' ................................ .
. .' ... ... '

kind of ecosystem that is found within the tropics. Be- nisms of speciation (and generation of high levels of Eu:SJr:01n l;r:i.:n
cause of its potentially unique nature, and because of biodiversity), food web dynamics, and mainrenance {Pop..ila1lon: 3iild n,:1 ·" ,

its amazing richness of species, tropical rain forest (or of biodiversity. Tropical rain forest tree species rich- 100+
more generally, tropical hurnid forest) will be the eco- ness is decidedly higher than that of any orher kind 95-99
F
system most discussed in this text. But the tropics are of forest and the question of how so many species
90-94 )1, \Í
'
85-89
more than tropical rain forest. Ecosystems ranging from coexist is a majar area of investigation, one that en- 80-84
75-79
deserts to rain forests occur in equatorial latitudes. compasses a lively debate. Species richness patterns 70-74
Because parteros of rainfall and temperature vary in animals ranging from beetles to birds are also a 65-69
60-64
with elevation, because soil characteristics differ due majar copie of investigarion. Indeed, the causal pat- 55.59 >

in part to regional geological differences, and because terns and function of che uniquely high biodiversity g 50-54
:': 45-49
disturbance faccors such as storms (monsoons, hur- seen in many tropical ecosystems is a tapie of very 40-44
ricanes) and fire vary regionally in frequency, tropical active research. 35.39
:
30.34
ecosystems vary from one place to another. Natural The role of disturbance as it relates to biodiver- 2S 29

l
25-29 '-
disturbance may be sufficiently frequent to prevent sity continues to occupy researchers. Traditionally and 20-24 -
15-19 9 '--
sorne, perhaps most, ecosystems from attaining a point naively, the rropics were once considered a region of 10-i4 ¡
where species composition ceases to change. Add to
that differences in historical biogeography from one
great ecological stability. This nocion followed from
the assumption that climate (considered warm and
5.9
o-~L - ,-1."_ o-
__JLL__.1_..__,-::'----,,':- -=,;e
5-J
r. t, t= - 60
place to another, and the reality is that ecological vari- wet and relatively invariable) was not a strong selec-
ation is extensive within che tropics. tive force in tropical regions, and thus species evolved 100+
Considera comparison between savanna and rain to the point where they became distinct from one an- 95-JS
forest in Africa. These two very different ecosystems 90.94 M
other. But disturbances occur constantly and at vary- 85-89
may be found at different longitudes but at the same ing scales in space and time. Such disturbances force 8C-f4

1~
latitude. They have little in common except that they 75.;r
changes in ecosystems and are thus selective forces 7C·74
1

are both in the same biogeographic region. Now con- relative to both ecological trends and ultimately evo- 65-69
sider rain forest in Papua New Guinea as compared ێ-64
lutionary direction. o 55.59 :
with that in Alta Floresta, Brazil. The structure and lt is obvious that che potencial far significant and 8 50.54 (

overall look of the two forests will be similar, but that "' 45.49
original research of tropical regions, where the abun- 40-44 4¡.,4 1 1
similarity is superficial. Because these regions have dance and diversity of both plants and animals ex- 35-39 3,,.,,9 1
30-34 3•)-'j 1
been isolated for nearly 100 million years, evolution- ceeds that found anywhere else, is very high. There is 23-29
25-29 1 1
ary patterns are different and species composition is still more to know than is known. 20-24 2C-2 1 1 1
15-1 g 1
10-1,
quite distinct between them. Because che species com- 15-12 1
10-i4 1 1
position is different, there will be significant differ- .¡.
ences in biological interactions and evolutionary pat-
5.9
0-4 L__ 0-'' 11

-
1

terns (see Chapter 2). Conservation ~

.l..__
W
_L_L...J.__L..i, - - - - ,',-----;!
20
(Populrtion: to1 m;u:on)
~ 60 ¿o 4ü
(Pcpuiat:cn. 1 298 m'lilon)
20 40 60

Ecological analysis of tropical ecosystems requires

full consideration of biogeographical history, climate,
in the Tropics 100+
95.99
edaphic factors such as soil characteristics, and distur- The world's tropical regions include sorne of the 9C-94 M M F
bance history, frequency, and magnitude. densest human populations, and sorne of che least
wealthy. The growth rate of human populations is
85-fJ
8~·84
75.79
,t~
70-74
high in most tropical regions. The human population 65-6$ 1 1
is predicted to grow from 6.3 billion {in 2003) to 8.9 6<)-34 1 1
Current Research billion by 2050 (Cohen 2006) (Figure 1-19). Much
o 55.59
~ 50·5!'. 1
1 1
1

Directions in Tropical of that increase will be in equatorial countries. Ac-

cording to the Population Reference Bureau's web-
N 45.;9

40-44 1
35.39
Ecology site (www.prb.org, 2007): ,0--1; > 1
1 1

25·2}
"C-21
,.'
20-'.i
1
1
This text summarizes current research in tropical Between 2000 and 2030, nearly 100 percent í5 ;g 15·1ºe 1
ecology, but at this juncture it is helpful to make a few 1C-1 4 1
of chis annual growth will occur in the less 5-S ¡ 1
generalizations. Tropical researchers study intricare developed countries inAfrica,Asia, and Latin J. 1
interactions among incerdependent species, carbon America, whose population growth rates are
0·' L__
,0
_L__
'O
_L_J__J_...J.___j__ _L._--:
2C 2', tcO 50 '
fO
1
40 n
1
,
1
,O
,,1 sr
sequestration in forests and other tropical ecosys- much higher than those in more developed FIGURE 1-1 9
tems, effects of climate change on ecosystems, mecha- countries. Growth rates of 1.9 percent and Hunian population size and age disttibution contrasted between developed and underdeveloped nations, with projections to 2050.
 WHAT AND WHERE ARE THE TRDPICS? 37
36 CHAPTER 1
......................... ............................................... ...............................
..........
,, ....
. · ¡ forest cover for 1990 and 1997 and mean annual change estimates during thai time period. Ali figures are in
higher mean that populations would double an additional 2.3 +/- 0.7 hecrares were visibly de- Hum1dtrop1ca
TABLE1
in about 36 years or less, if these rates con- graded (Achard et al. 2002). millions oft,ectares.
tinue. Demographers do not believe they Predicting the future of tropical forests based on
will. Projections of growth rates are lower current rates of deforestation is both perplexing and
than 1.9 percent because birth rates are de- complex. For example, it has been argued that as the : :, t ' t:.
clining and are expecred to continue to do century proceeds, birth rates in tropical regions will
so. The populations in the less developed re- decline and urbanization will increase, and thus ru- 1155 337 446 1937
Total srudy area
gions will most likely continue to command ral population growth will decline, slowing the rate
a larger proportion of the world total. While of deforestation (Wright and Muller-Landau 2006a 669 ± 57 198 ± 13 283 ± 31 1150 ± 54
Forest .:over in 1990
Asia's share of world population may con- and 20066). If this were the case, tropical forests
tinue to hover around 55 percent through would begin to regrow on once-cleared land, much 653 ± 56 193 ± 13 270 ± 30 1116 ± 53
Foresr cover in 1997
the next century, Europe's portion has de- as forests in eastern North America have regrown
clined sharply and could drop even more after forest clearance in the seventeenth and eigh- 2.5 ± 1.4 0.85 ± 0.30 2.5 ± 0.8 5.8 ± 1.4
Anni ¡ deforested area
during the 21st century. Africa and Latín teenth cenruries. This pattern would, in essence, be
America each would gain part of Europe's a reversa! of deforestation, but these forests would 0.38% 0.43% 0.91% 0.52%
Rare
porrion. By 2100, Africa is expected to cap- be secondary forests, not the same as the old-growth
ture the greatest share. primary forests that preceded them. This view is 0.28 ± 0.22 0.14 ± 0.11 0.53 ± 0.25 1.0 ± 0.32
Annual regrowth area
11 highly controversia!, and a lively debate has been
Humans will inevitably have an increasing impact on generated. Sorne argue that as indusrrialization in- 0.04% 0.07% 0.19% 0.08%
I ¡ Rn
tropical ecosystems, and as such the study of tropical creases in tropical areas, even with demographic
1 ecology must consider anthropogenic factors. shifts from rural to urbanized, large-scale forest Anm,al net cover change -2,2 ± 1.2 - 0.71 ± 0.31 -2.0 ± 0.8 -4.9 ± 1.3
Humans evolved in the rropics, emerging from clearance by industrial interests will continue the
tropical Africa and colonízing the remainder of the loss of tropical forests (Laurance 2006; Brook et al. 0.33% 0.36% 0.71% 0.43%
Rate
terrestrial planet. Human populations have adapted 2006). Because of the high biodiversity contained
to survive and prosper in tropical regions, and much within tropical forests, such widespread loss would Annual degraded area 0.83 ± 0.67 0.39 ± 0.19 1.1 ± 0.44 2.3 ± 0.71
insight inro ecological processes can be gained from inevitably result in high rates of exrinction (Brook
the study of tropical anrhropology. et al. 2006). lt is imperative to understand what is 0.13% 0.21 % 0.42% 0.20%
R~te
For example, the use of the land for agriculrure, occurring and what can be done to mitigare species
ranging from slash and bum techniques to susrained losses, but at this point, the future of tropical for-
terraced rice fields, demonstrates rhat humans can ex- ests is, ar best, uncertain (Gardner et al. 2006). This
rract continuous resources from tropical ecosystems. topic will be reviewed in Cbapter 15. diseases such as Ebola have put the African great apes,
Sorne communities demonstrare a highly sophisticated The rate of forest loss in the tropics, whatever ir the common chimpanzee (Pan troglodytes), and the
knowledge of the species patterns in rain forest. The may be, has strong implications (Table 1-1). A signifi- gorilla (Gorilla gorilla) (Piare 1-28), in serious peri!
use of various plants as pharmaceuticals and roxins cant and conrinuing loss diminishes populations of (Walsh et al. 2003 ). In the African country of Gabon,
has given rise ro the science of ethnobotany (rhe study numerous species, reduces biodiversity, and has the by the year 2000 ape popularions dropped to less rhan
of how local cultures master and develop various uses potential to seriously disrupt trophic dynamics and half what they had been in 1983. The aurhors of the
for plant species) and rhe search for potent drugs from rhe overall funcrioning of remaining forest fragments. study documenting the decline argue thar these species
tropical plants (Chapter 13). In addition, it reduces the capacity of forests to seques- should be elevared to a status of critically endangered.
Tropical ecology represents an area of basic re- ter excess carbon, creating a diminishing carbon sink. Such a starement could be construed as somewhat po-
search within ecology that can be applied ro conser- · But the issue of climate change, carbon sequestration, lirical, but it is not. The conclusion is based on data
vation issues. The discipline of conservation biology and the function of tropical forests in relation to di- and follows directly from an analysis of such data,
is essenrially derivative of basic ecology, and much mate change is an area that is still poorly understood given, of course, the premise that extinction of chim-
research in tropical ecology deals in sorne way with (Clark 2004). panzees and gorillas is neirher in rheir best interests
issues relared to conservation. One exa mple is land Yet another conservation issue is the effect of nor in the best interest of human society.
use and its effects, particularly deforestation and for- hunting and logging on forest ecology (Chapter 15). Conservation issues in tropical ecology will be
est fragmenration (Chapter 14). Deforestation rates in In many tropical regions, large vertebrare animals are PLATE 1·28 treared in much greater derail in Chapter 15.
tropical regions are generally high. In a broad srudy routinely hunted to the point where their numbers are GORIUJ..
based on global imaging from satellires, it was learned seriously diminished. Throughout the tropical world,
that from 1990 to 1997, 5.8 +/- 1.4 million hectares bushmeat, as it is termed, is widely used for food.
of humid tropical forest were lost annually and that Hunting pressure, logging, and rhe emergence of viral
 BIOGEOGRAPHYANO EVOLUTION IN THE TROPICS 39

.. .... ·········· ............................................................................................ ···············································

.......··········

2 Biogeography and Evolution in the Tropics

Chapter Overview
For much of the past 250 million years of Earth's history, tropical ecosystems have
extended north and south well beyond the latitudes they encompass coday. Bu1
climate has changed over the millions of years of the planet's history, and tha1
facc influences rhe distribution of tropical ecosystems. In addition, because of pla1e
tectonics, continents have separated and moved apart as Earth has become rr,ore
temperare, such that the tropical regions are currently confined within about 50° of
equatorial latitudes. Climatic change, the separation of conrinents, and continuous
tectonic activity such as mountain uplift have contributed to ongoing patterns of
evolution and are responsible for complex biogeographic distributions of flora and
fauna that typify today's tropical regions. This chaprer will discuss plate tecto1•ics
(che movement and rearrangement of Earth's crusr), vicariance (che separation and
subsequent evolurion of populations), and endemism (species restricred to a cenain
area) as each relates to biological evolution. lt will also discuss how these events
contribute ro the evolution of spccies.

What Is Biogeography?
Biogeography is the study of the distributions of organisms as they vary from one
region on Earth to another.
For example, ali of the world's 71 extant lemur species occur exclusively on rhe
island of Madagascar, off the eastern coast of Africa (Piares 2-la and 2-16). ln ,1d·
dicion, five familes of birds are found only on that island. Certain planr species are
also confined only to Madagascar. PLATE 2-1
(a) Coquerel's Sifaka (PropithecU5
Such unique assemblages of species have long stimulated thought among nam· coquere/1), a lemur species common to
ral historians. On a larger scale, many elements of both flora and fauna differ dca· low-elevation, dry deciduous forcsts
matically in diversity and distribution among tropical zones of different regicns in Madagascar. (b) Ring•1ailed lemurs
around che equator. For example, approximately 445 species of stately dipterocarp (Lemur catta) are highly social.
trees (Dipterocarpaceae) are distributed throughout che lowland forests of tropical .... . . .. . ·············· .. ········· ... ............................... ......... ....... ............................... ·· ····················
Asia and New Guinea, while a mere 30 dipterocarp species are found in Africa, aod
bue a single species is found in norrheastern South America. Suéh a pactern begs
fo r explanation. Charles Darwin devoted two chapters in On the Origin ofSpecies
(1859) to a consideration of geographical distributions of various groups of organ· Beginning with Alexander von Humboldt, bio- realms, che Neotropical, African, Oriental, and Aus-
geographers have long recognized different biogeo- tralian, include tropical regions (Primack and Corbett
isms, as evidence in support of his theory of evolution. Darwin and Alfred Russel
graf,l:!ic realms (Lomolino, Dov, and Brown 2004) 2005). Wallace is best known for denoting Wallace's
Wallace (who each independently conceived of che cheory of natural selection) both Line, che demarcacion of separation between the Ori-
(Figure 2-1). Wallace proposed six realms encompass-
linked biogeography with che process of evolution.
ing the world's terrestrial environments. Four of chese ental and Australian fauna! realms.
 40 CHAPTER 2 BIOGEOGRAPHY ANO EVOLUTION IN THE TROPICS 41
. ········•······················•······················ ·······································•·······································
·····••"•'

Wallace's Une

(a) Proboscis monkey

Fauna! Realms 6(f Scale by latitude

LJHolarcüc L] Afro--Tethyan :io)í, /

- Major region boundaiy o 1,000 2,000 míles
LJLalinAmerican LJlsland O 1,610 3,220 kilometers

FIGURE 2-1
This map shows the major biogeographic regions of thc world. Each is distinct from the others because each has various endemic groups of
plants and animals.

(b) Flying lizard (e) Tree kangaroo

.-···• Wallace's Une ·········· ...................................................... ······.. ·······················.. ···················••.

Alfred Russel Wallace (1823-1913) devoted eight years of Indonesia (Figure 2-2). He noted that animals found
to the study of what was then called the Malay Archi- west of the large island of Sulawesi were typical of those
pelago, an area that today is mostly within the country found in southeast Asia (Figures 2-3a to 2-3c). Such

h11 ines LJº miles

O 300 kilomelers

(e) Bornean bristlehead (f) Spotted cuscus

FIGURE 2-2 Asia Australia

Wallace's Line was originally developed
by Alfred Russel Wallace based on the
distribution of animal groups. Those FIGURE 2·3 .
typical of tropical Asia occur on the Lamples of animals found on either side of WaUace's Une. West of the line, nearer tropical Asia, one finds species such as (a) proboscts
west side of the line; those typical of mcnkey (Nasa/is /arvatus), (b) flying lizard (Draco sp.), (e) Bornean bristlehead (Pityriasis gymnocephala). East of the hne one finds sucb
Australia and New Guinea occur on the spccies as (d) yellow-crested cockatoo (Cacatua sulphurea), (e) various tree kangaroos (Dendro/agus sp.), and (0 spotted cuscus
east side of the line. (S,•ilocusws rnaa,lates). Sorne of these species are either threatened or endangered.

42 CHAPTER 2
islands as Surnatra, Borneo, and Bali ali ace exarnples.
These islands harbor various primate species, including
one lacge ape (orangutan) as well as severa! gibbons and
rnacaques. But animals found on and to the east of Su-
lawesi are rnoce cepcesentarive of those found in New
Guinea and Australia (Figures 2-3d to 2-31). No primates
occur other than hurnans, but species of arboreal marsu-
pials such as tcee kangaroos and possurns ace found on
sorne islands. Thus Wallace, by noting distributions of
animals on various islands, surmised that the Malay Ar-
chipelago is actually two separate biogeographic realrns,
\ ........... , ................................................................................................................................................
Biogeographers were initially puzzled, if not per-
plexed, by the complex intercontinental distributions
of plants and animals. Why, for example, are marsupial
mammals essentially conlined to the continents of Aus-
tralia and South Arnecica, which are quite widely sepa-
rated? (One marsupial species, the opossum [Dide/phis]
occurs in North Arnerica, but its ancestors clearly dis-
persed there from South Arnerica.) The tacit assump-
tion was made that ali laxa (groups of organisms that
are related-for example, palms, parrots) evolved ini-
tially in what were caUed centers o( origin. Such areas
were designated based on criteria such as where the
greatest number of species were to be found (within
a taxon), the greatest concentration of individuals, the
presence of pcimitive forms, and the existence of mi-
gratory routes. From centers of origin, the belief was
that taxa radiated away (to various degrees) to inhabit
different regions. The dipterocacp example cited earlier
would strongly suggest a tropical Asian origin for dip-
terocarps, with subsequent dispersa! to and coloniza-
tion of Africa and later South Arnerica.
Early biogeographers attemped to explain how
different organisms dispersed from their centers of or-
igin to other conrinents. Of coucse, these early bioge-
ographers were making the assumption (quite reason-
able at that time) that Earth's continents have always
occupied the same locations. Plate tectonics, explained
in the next section, demonsuates that such an assump·
tion was unwarramed. Nonetheless, three dispersion
models were suggested to explain the pattcrns of bio-
geography: the broad corrido,, tbe filter bridge, and
the sweepstakes route. Each of these models helps to
explain how dispersion occurs.
A broad corridor would be typified by the Pleis-
tocene (ice age) connection between northeastern Asia
and extreme northwestern Norch America, a landmass
now largely submerged and commonly called Beringia.
BIOGEOGRAPHYANO EVOLUTION IN THE TROPICS 43
·•·························· ························· ..... ......................... ........... ·················· ................. .
one represenring wbac he terrned tbc Oriental province
and thc othec tbe Au.stralian province. Wbere the two
provinces come togcther, Wallace noted sorne rnixing of
animals from both cealrns, a result of natural dispersa!
arnong populations. Today biogeographecs infocrnally
honor WaUace by ccfcrcing to thc islands from Java to
New Guinea as Wallacea.
Because of differences in distribution patterns of
plants and animals, the exact boundaries of Wallace's
Linc havc bcen altcrcd a bit since WaUacc 6m dcvcloped
the concept, but bis thcory remains gcnerally accurate.
(b)
Most organisms, including humans, who colonized
North Arnerica from Asia at that time, would pre-
sumably be able to travel such a corridor, pecmitting
extensive mixing between biota fromdifferent biogeo-
graphic realms-in this case, Asia and North America.
Broad corridors obviously require that the oceans be-
tween the continents must be sufficiently shaUow so
that a global drop in sea leve! will expose the cocridor.
This is the case with the Beringia corridor.
Afilter bridge would be exemplified by tbe Greater
and Lesser Sunda lslands of Indonesia, the very area of
Wallace's Line. As these islands become more disranr
eastward from Asia (one source of colonization) and (di
(e)
westward from Australia and New Guinea (anotber
source of colonization), they become increasingly spe-
cies poor. This is partly because species typically must
island hop from one landmass to another as they colo-
nize. Thus the species richness (number of species) of
various taxonomic groups on an island is related ro
tbe island's disrance from its source of colonizers, a
topic developed in greater detail in Chapter 4.
A s1ueeptakes route would be typical of how or-
ganisms colonize isolated islands, such as tbe Galápa-
gos Archipelago (Piares 2-2a ro 2-20. Presumably, an-
cestral iguanid lizards from South America that would
eventually evolve to be Galápagos marine and land
iguanas were accidently transponed over water on
detached vegeration from areas in Central and South (e) (O
Arnerica- the ocean currents steering them, quite by
chance, to landfall somewbere among the Galápagos. PLAH 2·2 .
Currents are directionally favorable for such unique Thesc vmcbratc animals are cach cndcmic to thc Galápagos lslands, but each uaccs its anccsuy to animals living in South Amcnca.
(a) and (b) Galápagos torto,sc (Geochelone mgra). Thesc two ,magcs show (a) a saddle-shellcd tortoisc and (b) a dome-shelle_d tort_o,sc.
events to occur. Another sweepstakes route would be
Each is a distinct subspecies. Oomed tortoises inhabit moist highland arcas, and saddle-shelled torroises_ frequent 101:land, and reg,ons
when seeds of plants contained in mud become stuck of th, islands. (e) Galápagos marine iguana (Amblyrhy11clms crista/lls). The flattened face adapts the am~als to grazing on marme algae
to the feet of migrating waterfowl- in the coursc of the attached to rocks. (d) Galápagos land iguana (Conolophus sub,ristatus). Land iguanas occupy_ low-e~evanon, and areas of the 1Slands,
birds' migration, the seeds are transponed bundreds if wherc they fetd extensi,·cly on cactus. (el Galápagos flightlcss cormorant (Phalacrocorox harrml- ThlS spcc,cs is found on!y ,n the
not thousands of miles before being derached and ger- coldtr watcrs 0 ¡ the western islands. (ij Hood mockingbird (Nesomimus ma,donald,). This spwcs IS found only on Espanola
minating. Charles Darwin performed experiments to (fo,merly Hood Island) in ,he Galápagos. lt spends much of its time on thc ground.
 44 CHAPTER 2 BIOGEOGRAPHY ANO EVOLUTION IN THE TROPICS 45
·········· " ''' ' '' ········ •""''' ' .... ...................
••• • •••••••••••HO<•••>> •• ••••••••• "• ••••••·•••••• • • ••• .. ••••••• • ••••• •• • ••••• .. •"••••••••• .. • • ••••• .. • • •• ••• ••••• ••• •••••• •• • O••••• • o• O• •• • • •• l ••• • • • .. •• • •••oo

test the viability of seeds under varying conditions to The problem that faced early biogeographers, as
test this concept. o·s
noted earlier, was that they assumed the continents of
The obvious difficulty with the preceding sum- Earth to have always been in their present positions. Seyclielles
mary is that in numerous cases, it falls far short of Sorne eady scientists were bold enough to suggest oth- lf
explaining causal reasons for organismal distribution erwise (see the section on "Plate Tectonics"), though no
among continents. Where, for example, is the corridor one had any idea how a huge continent could move. lt 10'S 10'S
between Australia and South America that the marsu- was not until 1957 that physical evidence was put forth
Africa Scuth
pials presumably used? Is it now, like the fabled island validating what had come to be called continental dri~, Equatolial
of Atlantis, lost beneath the seas? The answer is no. Cweot
brought about by a dynamic process termed plate tec-
The sea is sufficiently deep in most places that such a tonics. The understanding of piare tectonic theory for-
corridor could never have existed. No drop in sea leve! ever altered biogeography and made it into a far more 20'S
Mozambique
2o·s
would have exposed it. powerful and predictive science.

,.·· • ASweepstakes Ticket to Madagascar ... ... . . ................... .. .............

Toe island of Madagascar, off the east coast of Africa,

is inhabited by numerous endemic planrs and animals.
Only four mammal groups are found on rhe island (not
including sorne recently imported by humans), each hav-
ing evolved into unique endemic species. These groups are
rhe tenrecs, rodenrs, carnivores, and primates. While rhe
present representatives of each of these groups (lemurs, SO'E
MadagaSC8!
&sin

60ºE
30ºS
-'
Africa

30'E
,,O
//Agulhas
/ Current

4-0"E SO'E
30'S

60'E
for example, are primates) are ali endemic ro Madagascar, Longilude
O 200-2,000 meters D 2,000-4,000 meters [J >4,000 meters ~ Current flow

g ~ ,000 I\ ,-\
Rodents i: r .., \
'5. 2 000 , &f~~tava ,>•-~,-¡.-:,,,.,,....,.-....,..,,,_-+~=-:-1.,
24-20 Ma
Ó 3.000•~--·-_
--_•'-""
seaa 'o:.,·st]."s"'--- - - - - -- - - - - - 1
ª"as
Lemurs
1970'$ 13.00'S
60-50 Ma 41.40.E
~ LSO'E - 750 kdomalers

(b)
FIGb.it 2•4
(cori1i,1ued)

the ,•.roups rhemselves are evolutionarily related to African
mammals.The logical assumption is that Madagascar was
col•, 1ized by each of rhese four groups at sorne time duri.Q_g
Mozambique its history. Bur how did mammals colonize Madagascar?
Aclose look at the mammals of Madagascar has long
Channe/ suggested that they evolved from Cenozoic mammals, not
fro!t• the more ancient mammals of rhe Cretaceous pe-
riod. But Madagascar had been separated from Africa by
the time of rhe Cenozoic. So how did the mammals cross
rhe Mozambique Channel? Perhaps they walked. Bioge-
ographers have long suggesred rhe possibiJiry of a land
bri¿ge connecting Madagascar and Africa during parr of
the Cenozoic. Recent work on the likely topography of
(a)
FIGURE 2·4
the seafloor during the Cenozoic argues against such a
possibiliry. Perhaps rhese animals floated over on rafts of
Figure 24(a) illustrates the suggested rafting route by which sorne mammalian groups may have colonized Madagascar. The numbers
detached vegeration. This sweepstakes mode would ex-
indicare estimates of when (in millions of years ago) the colonizations may have occurred. Other groups such as elephants, zebras, an-
telopes, and apesare considered unlikely colonists, and indeed, none are found on Madagascar. Figure 2-4(6) illustrates how today's pla,n why larger animals, such as elephants, zebras and
ocean currents would minimize dispersa! potencial from mainland A/rica to Madagascar, but such was not always rhe case. orher hoofed mammals, and apes never colonized rhe
island. Only smaller crearures would have been able ro lit
on small rafts of vegetacion. The present pattern of ocean
currents would not allow an easy crossing, however.
Research modeling of Cenozoic climate and currents
strongly supports rhe sweepstakes route hypothesis (Ali
and Huber 2010). Surface water currents during the Ce-
nozoic appear to have been difieren! from roday, moving
in a direction thar would facilitate sweepstakes transport.
The data suggest that lemurs colonized via a sweepstakes
voyage across the Mozambique Channel approximately
60 to 50 million years ago. Tenrecs arrived around 42 to
25 million years ago, carnivores 26 to 19 million years
ago, and rodents 24 to 20 million years ago.
Tbe sweepstakes model for mammalian colonization
of Madagascar was first proposed by rhe eminent twenrieth-
century paleontologist George Gaylord Simpson (Krause
2010), and while still not proveo, evidence now seems to
suppott Simpson's conjecture (Figures 2-4a and 2-46).
 46 CHAPTER 2
.................................................................................................................................... ..................
Plate Tectonics
On a globe, look at South Arnerica, and notice that
the east coast of South America looks as though it
fits rather well against the west coast of Africa. If the
globe shows relief features, notice that the Himalayan
Mountains form a rugged boundary all along the bor-
der of India and Asia. Australia sits by itself, alone-
an immense island continent in the southern Pacific
Ocean. But notice that if sea leve! was lower, north-
eastern Australia (the Cape York Península) would
11 connect to New Guinea. Ali along the west coast of
South America extending north through Mexico to
Alaska are volcanically active mountain chains where
earthquakes are common. Frequent volcanic activity
is also typical where Wallace's Line occurs in Indone-
sia. These seemingly disparate observations are each
explained by the process of piare tectonics.
1 1 The occurrence of cataclysmic events such as
earthquakes and volcanic eruptions demonstrates
Earth's ceaseless active geology. Unlike Earth's moon
(Plate 2-3) or planees such as Mars, Earth is geologi-
cally dynanúc, continuously rearranging its surface
because of processes occurring in its interior. The term
1 into a system of rigid piares, as is the case with Earth.
1 Amenca. The southern beech tree, Nothofagus, occurs in
1 continents have always been in their present locations.
i' .... • The Discovery of Continental Drift
The intriguing distribution of continents had been noticed
from the time of Francis Bacon (1561-1626), who may
have been che first to muse about the possibility that South
America and Africa were once joined. He is reputed to
have noticed that che two continents had complementary
shapes. In 1912, a German meteorologist named Alfred
Wegener published Die Entstehung der Kontinente und
Ozeane (The Movements of che Contents and che Oceans),
a book in which Wegener made a bold and, to geologists at
the time (and for sorne time thereafter), preposterous sug-
gestion. Wegener proposed that che continents move rela-
tive to one another; drifting on the surface of the planet. If
true, this would mean chat the continents were not always
in their present positions. By implication, they might still
be drifting, each in its own particular direction (Figure 2-6).
But how could something as immense as a continent
actually move? Geologists could identify no mechanism
to account for such motion, and many authorities at che
time ridiculed Wegener's theory, which soon carne to be
called continental drift, a name that is still applied.
PLATE 2-3
Earth's moon, unlike Earth, lacks piare tectonics.
for this process of continua! change is plate tectonics.
Earth is one of the few known tectonic planets. Both
the planet Venus and Jupiter's large moon Europa
have dynamic tectonics, but they are not organized
Wcgener's principal motive in arguing for continen-
tal drift was the compelling fit that the continents seem
to have when you literally cut them out like pieces of a
jigsaw puzzle and rearrange them. Prior to Wegener, in
the late nineteenth century, Eduard Suess (1831-1914),
an Austrian geologist, suggested chat the southern con-
tinents were once fused into one gigantic continent.
This huge landmass was named Gondwana, a romantic-
sounding name chat refers to an area within India with
characteristic rocks that were thought to date back to
when such a hypothetical supercontinent might have
existcd. Suess based his hypothesis on fossil ferns of che
genus Glossopteris, found in rocks from South America,
Africa, and India, now each widely separated. Glossop-
teris fossils also occur in Antarctica, which likely wciuld
have pleased Suess.
The fit between Africa and South America is fairly
obvious, but Wegener was able to show how ali of the
continents could, if arranged carefully, roughly fit to-
gether into a single supercontinent, Pangaea. North of
·····················
' a and originally fused with it, was rhc north•
Gond,n , · .. h'd
. • continent of Lauras1a (contammg w at 1s to ay
ern ,u,. r h . W,
North ~merica, Europe, and nort ern Asia). egener
loss for a mechanism to account for how, from
wu ~' , . . ratites, a term referring to cheir lack of a broad bony keel
, ¡wo massive continents, today s vaned contments
ane. o J and drifted apart, bh ut e ms1ste t at such a
' ' dh
separa.•
ssibility should be considered.
po In addition to Wegener's view that the continents
could ! e assembled rogecher like a kind of planet-sized
.. ,w
11gs, puzzle, there was much ecolog,cal ev1dence to. sug-
est that at sorne time in che past, sorne of the contments
~er< oace literally joined. As early explorers stud1ed the
uniqur• p!ants and animals discovered during the age of
Jobal exploration (see Chapter 1), the1r study of b1oge-
~grap,.; revealed many enigmatic examples of puzzling
distrib,;tions among both plants and ammals. ·
humerous examples showed that dosely related spe-
cies ,,:.::ur on different continents. How could that be? For
example, there are magnolias (Magnolia spp.) in che south-
eastei, United $tates and in China. Three tapir species
(Tapirus spp.) occur in South America, bue another dosely
relattú tapir species occurs in Soucheast Asia. Ostriches
(Strutio camelus), found in Africa, are closely related to
emus ¡Dromaius novaehollandiae) and rheas (Rhea spp.),
but e ""s are found only in Australia and rheas in South
sud- .,J1ces as southern South America, southern Austra-
lia and Tasmania, and New Zealand. Such a widespread
distr:~ution is difficult to explain if che assumption is chat
Lét us examine one case more closely. As noted ear-
lier, there are severa! extant large flightless birds that
seem anatomically and behaviorally very similar: the
PLn E 2-4
(a) Ostrich (Struthio camelus) and (b) emu (Dromaius novaehollandiae) are two examples of ratite birds. They are very similar in
siie and anatomy, but ostriches are endemic to Africa and emus to Australia.
. .. . . ' ................ ~. '... . . . . . . . ... . .. . .. ... ... . . .. . . . . . . . . . . . .... . . . . . ... .. ... .... .. .... . ... . . . . .. .. ......... .. . .
BIOGEOGRAPHY ANO EVOLUTION IN THE TROPICS 47
ostrich of Africa, che two rhea species of South America,
che emu of Australia, the cassowa;ies (Casuarius spp.) of
Australia and New Guinea, and the kiwis (Apteryx spp.)
of New Zealand. Collectively, this group is named the
on the sternum, a trait typical of large flightless birds
(Plates 2-4a and 2-46). (The keel, or carina, is a prom-
inent blade-like protrusion from che ventral surface of
che breastbone. lt is very well developed in ali birds that
fly because it is the site of attachment of the large flight
muscles. In ratite birds, the carina is absent.) Each ratite
species is generally large and skeletally similar; has a body
covered in down-like feathers, has undersized wings, and
lacks the aforementioned large bony keel on che breast-
bone for the support of llight muscles. These birds nei-
ther fly nor swim. How can large llightless birds disperse
worldwide?
As described earlier, sorne biogeographers suggested
che existence of now (presumably) submerged land
bridges that once connected continents now separated by
ocean. There was no evidence that such bridges existed,
only speculation. Other biogeographers argued that ex-
amples such as the flightless birds noted earlier were ex·
amples of convergent evolution, where distantly related
species evolve similar appearances due to being exposed
to similar environmental selection pressures. But modern
genetic studies of both mitochondrial and nuclear DNA
sequences have demonstrated unequivocally that the cat-
ites forma monophyletic group, meaning chat they are ali
descended from the same comrnon ancestor. The logical
hypothesis is that ratite birds originated in and dispersed
from Gondwana and rode the separating continents over
time to their present positions.
(b)
.. .

48 CHAPTER 2
Piare tecronics, now supporred by multiple lines
of physical evidence, recognizes that Earth's crust is
divided inro approximarely eighr large and sorne ad-
dirional small basaltic plates. Some of these places are
entirely on the seafloor, some have continents resting
arop rhem, and sorne have parts of continents. For ex-
ample, che Pacific piare is huge and is covered mostly
by the Pacific Ocean, except thar the extreme western
part of North America rests on the most eastern part
of rhe Pacific plate. The remainder of North America
sits atop che North American plate. What this means
is that places such as Los Angeles and San Francisco
are actually situated on the Pacific piare. The North
American piare is moving ever so slightly west-south-
west, and rhe Pacific plate is moving more directly
north. The two piares meet in an area extending from
the central coast of Mexico to almosr rhe Oregon bor-
der. Part of rhis area is called the Sa11 A11dreas fault
• (Figure 2-5). Because the piares are moving in some-
whar different directions, rhere is fricrion between
them, the result of which is known to most Califor-
t ¡ 1 nians: earthquakes occur.
BIOGEOGRAPHY ANO EVOLUTION IN THE TROPICS 49
......... .................
······· .......... .................................. .......... .
····•·"
Permian Triassic
The heat generared from Earth's interior; a region 200 million years ago
called the asthe11osphere within Earth's mande, pro- 225 million years ago
duces immense subterranean convection currents rhat
keep rhe basaltic piares moving in slow motion (Figure
2-6). For example, Europe and North America are mov-
------- e Lauras i a
ing apart by a mere 17 millimeters peryear. But in a mil- 1
lion years, they will have separated by 17 million mil- o,
- Telhys Equalor
limerers, or about 17 kilometers, roughly 10.55 miles. Sea
Why then are there such large differences in the flora
and fauna of Africa compared with South America? Q.
Geological evidence suggests that these continents began
their slow but continuous separation as long ago as the
early Cretaceous period, sorne 144 million years ago.
Tectonic motion occurs because the heat from
Jurassic Cretaceous
Earth's interior forces new basalt to the surface ar 65 million years ago
135 million years ago
active ridges rhroughout the oceans. The lengthy
mid-Atlantic ridge brings up new material from the
interior, a process caUed seafioor spreadi11g (Figure
2-7). The North American and European continents
actually ride on the piares, being passively carried
along as the piares move. Old seafloor plunges back Equalor
_ Equalor
into the bowels of the Earth at deep oceanic trenches,

Present day

N~
Point Reyes
/ America
San Francisco

America
..l..- Australia
7- FIGURE 2-6
\
These maps illusrrare the changcs in
continental positions from the Permian
A[lla¡cti_ca,,,. to the present.
.................... ············••·······•················································ ...............................................................................

sorne in excess of seven miles deep. The Andes Moun-
tains have risen in large part due ro the impact of
Pacific Ocean tectonic piares, subducting beneath western South
America and forcing rhe rise of the Andes and ali of
the subsequent volcanism and earthquakes in places
such as Chile.
FIGURE 2-5
You can actually mimic this process in a crude
The San Andreas faulr is well known for irs relatively frequent recronic acrivity, rcsulting in eatthquakcs of varying magnitudes. Note where
the boundary between the Notth American piare and Pacific piare occurs. Way if you heat sorne tomara sauce and cover its sur-
face with crackers. The convection currents from the
heat rising through the sauce will cause the surface
crackers to move about, somewhat like Earth's basal-
tic piares move in response to interna! forces generat-
ing convectional heat.
Given the insights of piare tectonics, the notion
rhar continenrs do move now has credence. Geolo-
gists have traced rhe morions of the continents back
 50 CHAPTER 2
..
·• :• .............................., ..... ····•··"···· ..... .... ................... .... ..... ... .... .. .... .. .... . .....
through time. About 245 rrúllion years ago, at the on-
set of the Mesozoic era, Pangaea began to break up,
splitting into Laurasia and Gondwana. The former gi-
ant continent consisted of what toda y is North Amer-
ica, Europe, and northern Asia. Gondwana consisted
of what is now South America, Africa, Madagascar,
1 j lce!and
BIOGEOGRAPHY ANO EVOLUTION IN THE TROPICS 51
" . ...... .. ............ ..... ······· .. ..•. ...... ' .. · · · · · · • " ' ' ' º ' ' ' ' ' " º ' ' ' º' ' ' ' ' ' ' ' " ' ' ' ' ' ' " ' ' ' ' ' ' ' ' '
Antarctica, India, Australia, and New Zealand. Lau- TheUttle Animal That Proved Continental Drift .... .......................... .
rasia and Gondwana split throughout the Mesozoic.
Antarctica, once at a temperare latitude, eventually
drifted ro the South Pole, where we see it toda y. Asma!I synapsid (mammal-.like) reptile na".'e~ Lystrosau- Asia. When Lystrosaurus was discovered in Antarctic rocks
lndia drifted northeast and, abour 40 rrúllion - 11d mammal-like for 11s anacom1cal smularmes w11h in 1969-1970, che scory made the front page of che New
rJIS (<ª e .mto true mamma1s) was abun- York Times, so compelling was it as support for continen-
years ago, literally crashed into Asia, causing the uplift
che grouPthar would evolve
. . . . tal drifr. lt was obvious that Lystrosaurus could not have
of the Himalayan Mountains. Mount Everest, Earth's danr in Anrarctica dunng che Tnass,c penod of che Meso-
tallest moumain (8,848 meters [29,028 feet]), zoic era, when the continent was far from 1ts pres:nt pos1- existed on Antarccica if che continent had always been at its
was created in the course of this cotlision. . (Figure 2-8). Its fossils, found m Antarcac sedimentary present south polar location-the linle creatures, about che
nonk size of a spaniel, would have been frozen solid.
The large island of Madagascar, today a ha- roe , arealso found in southern Mrica, India, and southern
ven of endemic lemurs, birds, and plants, sepa-
rated from Africa about 160 million years ago
and from what would become lndia 90 million
years ago. Thus Madagascar has been alone, un-
attached to any other landmass, for 90 million
years, ample time for unique evolutionary trajec-
tories to develop.
Continental movement caused by piare tec-
tonics has profound consequences for Earth's
ecology. Without such constant movement,
Earth's climatic history would have been very
different and less variable. As the continents
move about the surface of the planet, they af-
fect changes in ocean currents, air currents, and
climate in general. As more coast!ine is exposed,
coastal shallow-water species such as corals
tend to proliferare. But when coastline is rrúni-
mized, as when continents fuse, extinctions of
such organisms seem common. This point can-
not be overstated. Piare tectonics is a driving
force in the generation of evolutionary selecrion
pressures.
Separation of the continents acts ro geo-
graphically separare organisms (known as vi- FIGC 2•8
cariance; see the following section}, stimulating Fossil remains of che synapsid (somewhat maounal-!ike) reptile Lystrosaurus have been found in sucb widely separated arcas as
evolutionary change and allowing evolution to Anm :uca and sourhern Africa. When Lystosaurus lived, in the l'ermian and Triassic periods, these arcas were part of the massive
proceed along varied and different pathways, conrinent Gondwana.
evolving endemic groups varying from conánent
to continent and island to island. The isolation
of Australia, for example, led to the furrher di- grear biodiversity of mammals, and in particular, pla- such as Madagascar, New Guinea, and Borneo, and
versification of marsupial mammals (recall that centa! mammals, evident in rhe fossil record through- sorne small, such as Komodo, New Caledonia, and
marsupials also occur in South America}, mak- our the Cenozoic is likely amibutable ro continental Cocos.
ing Australia unique as the land o( diverse mar- separacion stimulating high levels of speciation among A visir to a tropical moist forest in Queensland,
supia/s. Likewise, eucalyptus trees of over 600 groups physically isolated from one anorher. Australia, will differ from a visir ro a similar ecosys-
species occur in Australia and nowhere else (ex- tem sourh of rhe Orinoco River, in Venezuela, rhough
cept when transplanted by humans). Part of the both forests may, at first glance, appear structurally
Vicariance and Endemism much the same. There will be impressively tall trees,
buttressed roots, many vines, and epiphyres attached
FIGURE 2-7
Tropical ecosystems occur on four widely separated ro the trees. There will be palms and strangler figs.
The mid-Aclantic ridge is part of the extensive ridge system
that characterizes areas where new seafloor comes to the conrinents (South America, Africa, Asia, and Australia) Colorful birds may be seen, as well as equally striking
surface. and many islands. Among the islands, sorne are large, butterflies. Ants will be common. But a comprchensive
 52 CHAPTER 2 BIOGEOGRAPHY ANO EVOLUTION IN THE TROPICS 53
···· ······························ ·········•······•················ ·········· ···························•·························································· ................. ..................... .................................... ·················· "·" ··· .. ····· ·· ..............................

species lisr from rhese two widely separated forests will mammals, insects, and orhers will be highly disrinct
contain few (if any) species in common. lnstead, not between Australian rain forest and Venezuelan rain
only will species differ, bur also whole families of birds, forest (Piares 2-5a to 2-5f). The forests appear struc-
turally similar because natural selection favors certain
phenotypic characters in regions of high temperature,
high humidity, and abundant rainfall. Broad-leaved
evergreen trees share various characteristics that con-
fer high fitness in such areas, just as various forms
of highly colorful butterflies have high fitness within
broad-leaved forests. But why are the species, the gen-
era, and indeed the families so different?
¡I The answer is vicariance, a word referring to factors
,1 1
resulting in geographical separation (also called allopa-
try) and the evolurionary consequences of such separa-
tion. When allopatry occurs, populations are unable to
interbreed and thus may diverge evolutionarily. This re-
alicy forms pan of che basis for the biological species

(e) (0
PLATE M
(a) (b) (contin:ied)

concept (BSC), to be described later in this chapter. Sepa-
ration of continents results in allopatry on a large scale.
Similarly, island flora and fauna are unique in large part
due to random colonization followed by evolution in,,
isolation. Vicariance means that taxonomic origins dif-
fer as a result of separation caused by sorne son of bar-
rier. On a continent, such a barrier could be a mountain
tange ora broad river, or perhaps a desert. On a broader
scale, oceans form barriers between continents and is-
lands, resulcing in che evolution of vicariants, groups of
closely related species isolated from other groups. Al-
lopatry imposed by vicariance creares a barrier to gene
(e) (d)
exchange, and thus groups of related species evolve iso-
PLATE 2-5
IMAGES OF VENEZUELAN ANIMALS ANO AUSTRALIAN ANIMALS
lated from other groups of related species.
From Venezuela: (a) male Guianan cock-of-the-rock (Rupicola mpicola); (b) royal flycatcher (Onychorhynchus mexicanus); and (e) Brazilian Evolucion of species occurs wherever life exists.
porcupine (Coendou prehensilis), Froni Australia: (d) male Victoria's riflebird; (e) noisy pitta (Pitta versicolor); and (f) common ring-tailed Thus in landforms that are isolated, one from another,
possum. speciation results in che evolution of unique assem-
blages of plants and animals. When understood as
a produce of historical vicariance biogeography, che
large differences among biogeographic realms become
much more clear.
When a species' evolutionary history, its very gene-
alogy, is unique to a given region, and the species re-
mains only in thar region, it is terrned endemic. Ende-
mism is cypically high among island species because of
tbe relative isolation of islands. Thus it is understand-
able that the Galápagos Archipelago, with its large as-
semblage of endemic plants, birds, and reptiles, inspired
Charles Darwin in 1835. Endemism refers to raxa that
are native to and restricted to a single area. The Galá-
pagos penguin (Spheniscus mendiculus) (Piare 2-6d) is
endemic to che Galápagos Islands, found nowhere else.
The okapi (Okapia johnstoni), related to giraffes, is en-
demic to central African forests. Whole groups such as
 54 CHAPTER 2
••,,:• ......... .
(a)
' 1
, 1
1 i
(e)
PLATE 2·6
These three bird species- (a) Bahama swallow, (b) Bahama woodstar, and (e) Bahama yellowthroat-each endemic only to certain of the
Bahama lslands, are each closely relared to more widely ranging species. (d) Toe Galápagos penguin (Spheniscus mendiculus) is the most
northern of the world's 17 penguin species and the only one to reach the equator.
.... ................................................................................................................................... .........................
the Hawaiian finches (Drepanididae) in Hawaii, ground
antbirds (Formicariidae) in South America, and gibbons
(Hylobatidae) in tropical Asia, including India, are en-
demic where they occur. Continents and islands al1 ex-
hibir various degrees of endemism at both the species
and subspecies leve!. For example, of the approximately
100 resident bird species in the Bahama lslands, three
are endemic (Plates 2-6a to 2-6c). There are 26 endemic
subspecies of birds found in the Bahamas.
Natural Selection
Modern evolutionary theory defines natural selection
as dífferential reproduction among genotypes. Genes,
the long, coiled molecules of DNA that concain hered-
BIOGEOGRAPHY AND EVOLUTION IN THE TROPICS 55
f . ni saying that a coin weighted ·1
to heads
. f
d1·fferent ioffpped consistently come up ta, s more o -
will, when , ,
han heads. . ..
ten t . variabihty ongmates through the ran-
GenetIC .
, 5 of mutation, a sud den and unpredJCt-
d h ge in a gene. Mutatton 1s non 1reet1ona1:
om proce. . . d. .
abl: e anents do not cause or produce useful mutants
env1
. ronmnse to need. The vana . b·¡ · 1. f
1 1ty resu tmg rom
m respon is Pnhanced by recombination of alleles in
mutatJO • • · · 1
reproducing spec1es. Selection can act on y on
sexua lly . f .
whatever genetic vanants are present. I sorne vanants
b tter adapted than others, they w1ll be passed
are inehighe.r proport1on.. Se1ewon,. unl"k I e mutat1on,
.
00
· ot a random process (any more than a we1ghted
isn·n flip is considered random m . tts
. outcome), be-
~::se only certain members of a population are best
ited for a given environment and thus have a non-
su
random chance of surv1vors . h"1p and reproduet1on. . In
recent years, botb geneticists and molecular biologists
have established beyond doubt that large amounts of
genetic variability exist in most populations. Thus in
most cases, there is ample raw material for natural
selection tu act on.
Adaptation
An adaptation is any anatomical, physiological, or be-
(d)
havioral characteristic that can be shown to enhance
either tbe survival or reproduction of an organism.
Such traits, as they translate into successful repro- (a)
duction, comprise what is called evolutionary fitness.
Fitness is reflected in adaptations, ali of which result
from the action of natural selection.
Any organism may be viewed as a cluster of vari-
itary information, were quite unknown to Darwin, al- ous adaptations. For instance, opossums, Neotropical
though he knew that phenotypic traits were inherited. porcupines, kinkajous, as well as many monkeys of
As long as a population contains genetic variability the American tropics possess a prehensile tail. Such a
among irs members, natural selection can act, since structure iunctions effectively as a lifth lirnb, lending
there are genetic variants that will respond differendy security and mobiliry to the animal as it moves through
depending on abiotic and biotic selection pressures the treetops. It is easy to see intuitively that the pre-
imposed. Natural selection acts only on the present, hensile tail is an adaptive structure. Tailless rnonkeys
never "planning" for the future. Those phenotypes or opossums would face a smaller lifetime reproductive
that survive relatively better than others do so only success because of the added risk of falling. But note
because conditions, whatever they· may be, suir them that Old World monkeys also are lit within their ar-
better than others in the population. Natural selection boreal environments, but none possess prehensile tails.
is a general statistical truth: individuals with genes Thus a prehensile tail, while adaptive to those that have
PLATE 2-7
(b)
that confer reproductive advantage will tend to leave evolved it, is not absolutely required to survive a lile in
(a) The wooUy monkey (Lagothrix (/avica11da) found in South
the most progeny, and thus those genes will propor- the treetops (Plates 2-?a and 2-?b).
America has a prehensile tail and is demonstrating how to use ir.
tional1y increase generation after generation relative Not ali adaptations are obvious. In tropical parts of (b) The Guereza colobus monkey (Colobus guereza) of Africa lacks
to others in the population's gene pool. This is litde Africa, many humans carry a gene that when present in a prehensile tail; no African primates have prehensile tails.
 56 CHAPTER 2
double dose (one inherited frorn the mother, one from
the father), produces defective hemoglobin molecules,
resulting in m.isshapen red blood celis called sickle ce/Is.
These victims of sickle cell anemia usually die before
reaching reproductive age. However, other individu-
als in the population who possess only one dose of the
sickle cell gene and one dose of the normal gene for
beta hemoglobin are not seriously disabled and are
more resistant to malaria rhan others in the population,
including those individuals with two normal genes for
hemoglobin. Thus the sickle cell gene is adaptive (since
the protozoan that causes malaria is part of this envi-
rorunent), but only when in a single dose. In a double
dose, it is lerhal.
' 1
Testing Adaptations
Because rhey can seem so obvious, adaptations are of-
ten inferred, and such inference may be true, but rig-
orous testing is rhe only way in which adaptation can
actually be demonstrated. For example, orb-weaver
spiders (family Araneidae) throughout che Neotropics
and temperare zone make large webs in which rhere
are sorne areas of obvious, thickened, zigzag strands
called stabilimenta (Plate 2-8). Spiders must use con-
siderable energy to synthesize the silk for the web,
especially the dense stabilimenta. Why should spiders
invest in making stabilimenta? Are these conspicuous
zigzag strands adaptive? The spiders that make srabili-
menta are those whose webs remain intact through-
out the da ylight homs. (Many spiders make new webs
each evening and take them down ar dawn.) Biolo-
gists have hypothesized that srabilimenta, which make
webs easily visible to humans, have the same effect
on birds (which are also visuaUy oriented), thus al-
lowing a ílying bird to avoid an orb-weaver spider's
web, saving the spider from having to remake the web,
which would be significantly damaged should a Aying
bird strike it (and would yield no food for the spider-
birds are too big to captme and eat). But spiders that
invest energy in making srabilimenta rather than risk
having to start from scratch and make a whole new
web would be less prone to "bird accidents." There-
fore, stabilimenta could represent an adaprarion to
energy saving in an environment where birds pose a
risk to the secmity of the web. As such, rhe hyporhesis
sounds plausible, but without testing, it is just a story,
an educated guess. How could it be tested?
First, an observer could simply watch spider webs
and note bird behavior. This was done, both in Pan-
PLATE 2-8
Orb-weaver spider female in a web with stabilamenta visible.
ama and in Florida, and birds were observed to rake
short-range evasive action when approaching webs
with srabilimenra (Eisner and Nowicki 1983; Craig
1989). Second, using webs thar do not have srabili-
menra, researchers altered sorne webs, adding artifi-
cial stabilimenra, keeping orher webs without stabili-
menra as concrols. The webs wirhout srabilimenta did
nor remain intact during the day, while rhose with
stabilimenta generally <lid, Further direct observations
implicated birds as the major threat to webs, though
other large animals ranging from burterílies to <leer
could also be forewarned by the presence of stabili-
menta. This work demonstrated the adaptivcness of
srabilimenta.
Not ali traits are adaprive. Organisms represent
che combined effects of thousands of genes working in
concert. The anatomy, physiology, and behavior of an
organism represent various compromises imposed hy
BIOGEOGRAPHY ANO EVOLUTION IN THE TROPICS 57
·········· ··· ··· ···• ·······• "' ' ''' ' "' ' '' "' '' ················
........
adapred by size and srrength to fuacrion terrestrially.)
·nreracn·ve effecrs of rhe genes rbat formcd ir,.Is
t he 1 . chat hurnans generally lack body hatr? Tree slorhs descend from trees only about once a week,
.11 dapuve
ª· e poss1·bty so (to allow loss of body heat .m a and only rhen to defecare ar the base of the rree. Once
Qui1 savanna climate, for example), but no that is done, rhey climb back into che tree. Today's rree
bot' sunny,
ll acceprable adapnve · exp1anat1·on for ha1·r- slorhs are totally arboreal, skillfully moving upside
genera yhas been rested. Om hairlessness could be down from branch to branch. In his well-known ac-
lessness uct of our development, w1r. b no adapnve . count, Wanderings i11 South America (1825), Waterton
a by-prod . wrote of rhe slorh, "This singular animal is desrined by
function now or_ ever m the past. •
A given rra1t may, however, have been adap- nature ro be produced, to live and to die in the trees;
. · rhe past, but not any longer. There are large and to do justice to him, naruralists must examine him
nve 1n . f d . . in rhis his upper element." Waterton then went on to
· from about 30 spec1es o rrees an vmes.w
f~m describe in derail how well adapted the sloth is for its
Costa Rica rhat produce large and Aeshy fru1ts.
These fruits drop from rhe rree, and most essco- ar boreal life. Warerton put adaptation in context.
rially just rot. That is very o_dd considering that
fruits function to attract seed d1spersers (Chapter 7).
So why are rhese fruits nor consumed, and the1r
eeds dispersed? Some years ago, Daniel Janzen
:nd Paul Martín hypothesized that che large fruits
represent what they called ecological anachro-
nisms, fruits adapted to be d1spersed by ammals
that are now extinct Uanzen and Martín 1982).
During rhe Pleistocene, sorne 10,000 years ago,
about 15 species of large (megafaimal) mammals
(including horses and ground slorhs) became ex-
tinct in what is now Costa Rica. These included a
unique elephant-like group called gomphotheres.
Gomphorheres may have been essenrial dispersers
of seeds from large-fruired planrs. As a result of
gomphothere extinction, planrs such as guanacasre
(Enterolobi11m cyclocarpmn) were left without any
agent of dispersa!. Their fruits, once extremely well
adapted, are no longer well adapted.
Adaptati ons in Envi ronmental Context
An adaptation must be viewed in the context of
the environment. Obviously, the sickle cell gene is
nonadaptive in human populacions living in envi-
ronmems free of malaria.And a beautifully adapted
creature may appear awkward, a "misrake of na-
ture," if seen out of its proper environment. Charles
Waterton (1782-1865), a rarher eccentric Brirish
explorer who traversed the Amazon during the early
nineteenth cenrury, noted that che three-toed sloth
(Bradyp11s variegatus) appears very poorly suited to
survive when seen struggling over the ground (Piare
2-9). Tree sloths, however, do not normally peram- PLATE 2- 9
bulate on rhe ground, ar least not in modero rimes. The three-toed sloth (Bradyp11s variegat11s) is wcUadapred 10 an
(Large ground sloths once did, and rhey were well arboreal life but poorly adapted tO being on the ground.

58 CHAPTER 2
Speciation
Humaniry has long recognized variabiliry in narure
and that organisms fall into natural groupings. Lin•
naeus generated a system organizing the many life
forms based on rhe degree of shared similarities: king·
dom, phylum, class, order, family, genus, and species.
The Linnaean system, though much modified in mod-
ern classification studies, still prevails roday, and the
branch of biology that deals with it is called systemat·
ics, the classification of organisms.
1
1 1 The species is the basic unit of the Linnaean sys·
tem. When we survey an ecosystem, we rypically mea-
sure the species richness (number of species) of trees,
birds, moths, ants, or whatever groups we find to be
(a)
PLATE 2-1 0
MALE ANO FEMAL E ECLECTUS PARROTS.
(a) The male is green with an orange bill. (b) The female is rnostly red and blue, with a dark bill.
of most interest. Thus it is important for ecologiscs
to understand what, exactly, a species happens ro be.
The quesrion is complex, because ir is often difficult to
determine when organisms are members of the sam~
spccies or are of different species.
Suppose that you were wandering the rain forest of
New Guinea and you look up and observe a big red par-
rot with blue on its wings and belly, a strik.ingly hand-
some bird, arop a palm tree. Soon you see a second bird
that looks about the same size and shape as the first but
is brilliantgreen, with red below its wings.The two birds,
both wonderfully bright, have totally different plumages.
Be that as it may, you are observing a pair of eclectus
parrots (Eclectus roratus). Males are mostly green, an<l
females are red and blue (Plates 2-lOa and 2-lOb).
(b)
......................................... .......... .......................................................... .................
............, ..
·/J
(' l -.._•
·) .,."J.ttt\
' ;..,
y Why are eclectus parrots and African swallowtail
~
')
ti
\ /
. ·-~
PLATE 2-1 1
This image shows, in the bottom thrce rows, unpalatable butterfly
model species in the family Danaidae (left) and palatable rnimccic
forms of female Papilio darda111,s (right), an African swallowtail
species. At top-left is the Papilio dardanus male; at top-right is
a nonmimetic, male-like female of the same species. The poly·
morphic, female-limited Batesian mimicry was first described by
Roland Trimen in 1869.
When males and females do not look alike, they
are considered examples of sexual dimorphism, but
they are still obviously members of the same species.
Many animals, including insects, spiders, fish, reptiles,
birds, and marnmals (including, of course, humans),
are to varying degrees externally sexually dimorphic.
In sorne cases, sexual dirnorphism is the result of sex-
ual selection, which will be discussed in Cbapter 7.
Now suppose that you are wandering in Africa,
looking at butterflies. You sec rhe colorful swallow·
tail Papí/io darda1111s, a large, conspicuous butterlly
(Piare 2-11).
Nearby is a second butterlly that looks quite dif-
ferent. Would it surprise you ro learn that it is tbe
same species? lt mighr very well be. In this casr rh•
BIOGEOGRAPHY ANO EVOLUTION INTHE TROPICS 59
fema le butterllies of P. dardanus rypically mimic other
butterlly species, those that are noxious to predarors
such as birds. In sorne cases, the female looks like the
male, with a "swallow-rail," bur more rypically, fe-
males look nothing like the males. Mimicry will be
discussed in the Chapter 7.
butterllies considered single species even though in•
dividuals within these populations are quite distincr?
The answer is that they successfully interbreed. Male
eclectus parrots may be green and females red and
blue, but they know each other when it is time to form
a pair bond. The same is true of African swallowrail
butterllies. Ali animals are genetically adapted to rec·
ognize othcrs within their species. The same is true, in
a more passive way, for such organisms as llowering
plants. The size and shapes of the llowers as well as
rhe exacr riming of when they bloom acr ro aid in dis-
persing pollinators among the various species.
This realiry forms the basis of the most widely
used definition of a species: populations of actually
or potentially interbreeding organisms. In this defini-
tion, callcd the biological species concept (BSC), spe·
cies designation is based on an assumption of sorrs-
namely, that rhe organisms themselves will revea! ro
which species they belong rhrough their reprod uctive
habirs. Under the BSC, rhe species category becomes
a natural unit in the Linnaean system because the
organisms to which it pertains actually recognize ir
themselves, at leasr in a manner of speaking. lt is un·
likcly that an eclectus parrot knows it is a "parrot"
compared with, say, a hawk or a stork, or if it ap·
preciares that it is an animal and nota fungus. But its
genetics make it competem to recognize and interact
meaningfully witb another eclectus parrot. The BSC is
a theoretically strong definition, heurisrically satisfy-
ing, but somecimes difficult to apply in the field, es-
pecially in areas where popularions are separared by
allopatry. IJ rwo populations look alike and seem ro
be reasonably likely to be part of one species, but if
these populations are separared geographically, how
do we know whether they could mate and produce
ferrile offspring?
In spite of these obvious difficulties, the BSC has
endured as the mosr useful definition of a species. In
practica! terms, it is true that characteristics other
than reproductive isolation, such as anatomical dif-
ferences, voice, and behavior, must be frequenrly em·
ployed to assess whether rwo populations are one or
two specics. Today it is common to "'" ~···•-=- ·'
 60 CHAPTER 2
BIOGEOGRAPHY ANOEVOLUTION IN THE TROPICS 61
.................... ........................... ......................................................................
, ..... .,........ .
Marmosets
The b1·oiogical species concept •
prevails. through-
..
. ¡ y but another spec1es concept 1s gammg
out bto og ' . . Tamarins
. • <upport. Many systematlsts are now usmg
1ncreasmg • .
h ·que known as phylogenettc systematrcs, or
.
Capuchins
atecn1 h .
. ¡·es Cladistics does not depend on t e estabhsh- Squirrel Monkeys
cladIS 1 • ' . .
.reproductive 1solauon between two popula-
ment
.
of 'd d
· order for them to be cons1 ere separate spe- ~ - -- Night Monkeys
0
uons
. When I
applying clad.1st1.cs, the b'101og1.st measures Uakaris
c,es. aracteristics of orgamsms · (and the character- Sakis
many Ch · . . Howlers
.1st1C
• s are eaually
• we1ghted, . not biased .as more or less
• ortant) and then appltes an algonthm, a mathe- Spider Monkeys
1 1 ,m!rical model, which allows a computer to produce Woolly Monkeys
1 1 mbranching diagram, called a cladogram, showing ~ - - - Macaques
(a)
:he degree to which individuals share various suites
1 1 - ? Mangabeys
PLATE 2-12 of characters (Figure 2-1 O). Characters may be ana-
(a) Savanna elephant compared with (b) forest elephant. tomical or molecular: the rechnique is often applied to
1 </)
Mandrills
>,
1

1
DNA, to designate species, a clear modifi~arion of the
1 1 BSC. African elephants provide an example.
Elephants are widely distributed throughout East
and Central Africa. As large and unmistakable as these
ponderous beasts are, it may surprise you to learn that
until quite recently, taxonomists were apparently mis-
taken about exacdy how many species of elephants
currently inhabit Africa. lt was assumed that there is
but a single species, the African elephant, Loxodonta
africana, when there are (or at least, rhere may be) two.
Studies by Alfred L. Roca and colleagues (2001) suggest
data such as was generated in the preceding elephant w
example. ln designating species, an attempt is made to
.:,;
e Baboons
Compared with L. africana, now renamed the African o
separate populations on the basis of the most recently :;;
savanna elephant, the forest elephant is smaller in body 1:J
size, has rounded, not pointed ears, and has straighter derived genetic characteristic that two populations no ~
~ Colobines
tusks (Piares 2-12a and 2-126). The scientists used a Ionger shcire in common. Such a definition thus puts </) 1:J
rhe focus indirectly on the most recent point, termed a w 5 Langurs
technique called dart-biopsy, aJlowing them to harm- e
Probaseis and
lessly sample the DNA of free-ranging wild elephants. nade, when the rwo populations diverged genetically. f
~ Leaf Monkeys
They sampled 21 populations and collected data from This approach to species designation is called the phy- ;;
ü
195 individual animals, examining DNA from mito- /ogenetic species concept (PSC). - - -- - - - Gibbons
chondria as well as sequences from four genes from the The kry distinction between the BSC and the PSC
nucleus (representing 1,732 base pairs). The results of is that reproductive isolation is not required when ap- w
plying che PSC. What the PSC is based on is demon- Chimpanzees
the molecular analysis suggest that forest and savanna ·e"'
1:J

4 strongly that there is a second species of African ele- elephants are as distinct from one another as lions are stration of a genetically distinct, diagnosable suite of .E
o Bonobos
phant, the African forest elephant, Loxodonta cyclotis. from tigers. The genetic distinction between forest and characteristics (which can be anatomical or based on I
savanna elephants is 58% for the same molecuh1, sequence differences). Molecular distance, Humans
gene sequences. This separation is ap- the degree of difference between molecular sequences
proximately equal to that seen when in two pcpulations, is measured for known species and
Afncan · comparing African with Asian elephants, then used as a kind of yardstick to determine whether
LCY5 forest long considered separate species, the lar- two populations should be designated as separate spe-
n=1 ter in the genus Elephas. cies. Usi:1g the PSC, there would be many more species Lemurs
Given such a large difference in recognized for taxa such as birds and mammals, where

lft\ DNA, there is thus little doubt that

there are really two, not one, species of
African elephant (Figure 2-9). The com-
many species exhibir what are termed subspecies, or
races. (A subspecies [or race] is a population within
a species that is genetically distinct but not reproduc-
Rufled Lemurs
Aye-Ayes

bination of genetic, morphological, and tively is0lated from other similar populations.) Most lndris and Sifakas
As1an [ EMA1 ecological differences led the research- subspecies, even though not necessarily reproductively ._____ Mouse and
tsolated, would likely attain foil species status under Dwart Lemurs
ers to conclude that there are really two
n=13 the PSC. Bushbabies
species of elephants currently in Afriq. 1

Lorises
...... , ................... , ,,. Speciation as a Process
African FIGURE 2-9 FIGURE 2-10
savanna This cladogram, based on genetic similarity,
Speciation is the essence of evolution, the production
This is an example of a cladogram- in this case, for primates.
illustrates that the savanna and forest elephants of new forms of life, different from ancestral popula-
The shorter the distance from branch to branch, the more closely
are about as different from one another as either tions. !t can happen in two ways. A population can relared are the species. For example, note the top of rhe cladogram.
is from an Asian elephant. change wirh time, so much so that it deserves to be Tama rins and marmosets are each other's closesrrelatives.
 62 CHAPTER 2 BIOGEOGRAPHY AND EVOLUTION IN THE TROPICS 63

called a new species. In this model, the population fol-
lows a linear trajectory and gradually changes, under
the iníluence of natural selection, into something else.
It was once believed that human evolution was best
described in this way. But another, and likely more
common, method of speciation is mulriplication of
species. From a single species, new species formas var-
ious branches off a bush. This model seems to describe
Speciation by vicariance
1 1
1 1
' 1
¡ lations become geographically
·················
most of speciation and is now the model that seems
to accurately account for human evolution. About 1.8
million years ago, the fossil record suggests that there
were up to six species of hominids coexisting in A/- Great Rifl Valley
rica, ali evolved from an ancestral species that dates Volcano
back to just over four million years ago (Tattersal
1995; Gibbons 2009). Though Hamo sapiens is the
only descendant species, we need to understand that
the history of humanity involves
many genetic branches, only one
of which persists to this day.
The common model for spe-
Strong selection for ciation, largely based on the bio-
reproductive logical species concept and or.
isolation in zone vicariance, involves severa! steps
of sympahy
(Figure 2-11 ). First, because of an
event causing vicariance, popu- lndian
Plate
isolated. Second, geographic iso-
lation prevents gene ílow, allow-
Populations ing genetic differences between
recontact each
other- sympatry
the isolates to accumulate with
reestablished time. Natural selection may act to
promote such genetic divergence.
Last, genetic differences between
an isolate popularion and its pa-
Somalian
rental population increase ro the Plate
point that, should they establish
Populalions evolve secondary contact, natural selec-
separately- no tion acts to promete reproduc-
gene flow between
tive isolation between them, thus
them
establishing them as separare spe-
cies. It should be emphasized rhat
this process does not require a
FIG URE 2-12
great <leal of genetic difference to This map shows the location of the
lndian
occur between the isolated popu- Ocean three majar lakes of the Great Rift

Vicariance event
lations, only enough to establish
reproductive isolarion. Speciation .................... .
......................... .. ····· ··· ····· ······•" "' ''' ' ' .............................
Valley in Africa.

makes populalion can be surprisingly quick.
allopatric
FIGURE 2·1 1
Beginning with TI, note how a vicari-
am event such as a mountain rising may
separate populations geographically, thus
allowing for gcnetic differences to develop
over time between the now-isolated popu·
Single population-
lations. At time T4, the populations again
become partially sympatric and selcction
all individuals
pressures acr to enhance their separation.
sympatric
At ti.me T5, rhere are reproductive-iso-
lating mechanisms that select to prevent
hybridization and ro mainrain the rwo
populations as full species.
In the model described here, geographic isolation
promotes genetic divergence and subsequent speciation.
1t is believed that most speciation, especially among an-
imals, occurs in this manner. But recent work suggests
that it is possible for natural selection to drive the pro-
cess of speciation and promete reproductive isolation,
even in the absence of complete geographic isolation.
Cichli i Fish: An Example
of Mu:tiple Speciation
Cichlids are diverse bass-like fish found in tropi-
cal waters around rhe· world (Barlow 2000; Stiassny
and Meyer 1999). There are about 300 species in the
Americas, including one that occurs ás far north as
Texas. But most cichlid species are found in East Af-
rica, clustered in three lakes that formed in the Great
Rift Valley: Lake Victoria (more than 400 species),
Lake Tanganyika (200 species), and Lake Malawi
(300 to 500 species) (Figure 2-12).
The oldesr of the three African lakes is Lake Tan-
ganyika, which formed 9 to 12 mi Ilion years ago, and
the youngest is Lake Victoria, whose origin is between
250,000 and 750,000 years ago. Considering only
Lake Tanganyika, if one assumes ir to be 12 million
years old, the highest estimare of its age, and that it
 64 CHAPTER 2
I • ~ • • .'•• • • • • • • • • • •• . • •• : . • • • • • • • • ' • • • •••• • • • ••• •• . • • "
PLATE 2-13
Cichlid fish o/ 1hc Great Rift Valley
11 1 lakes demonstratc an amazingly high
species richness.
• ooO•• • ••• • • •• •• • • • Ol ••• • •• O••••••• • •••••H• ••••••• •• •••• • •••••••• •• •••• •• • ••••••• •• •• •••••• • ••••••• ••••••• 0,ooo, ,,,, •• •••• ••••••••• • • • • • •• • • • • •••• •••· ••• • •• .. • ••••• • ••••• •• •••• ..
1 1
has 200 cichlid species, that is a speciation rate of
about 1 species every 60,000 years. However, Lake
Victoria is at most 750,000 years old, and it has 400
cichlid species. If ali of them evolved in that lake, the
speciation rate is about one species every 1,875 years.
But even that number is too low. Researchers are in
agreement that che amazing diversity of cichlids in the
African rift lakes arose recently, within the past few
million years, demonstrating that speciation can occur
with impressive rapidity (Verheyen et al. 2003).
Studies of the mitochondrial DNA of the cichlids
from Lake Victoria have demonstrated that this clus-
ter of species is genetically very closely related, show-
ing that they have ali recently evolved from a common
ancestor (Piare 2-13). What is stunning is that the to-
tal genetic variation among these 400 species is less
than that found throughout Horno sapiens-humans!
In other words, the genetic distance between you and
a stranger you might meet is likely to be greater than
that between two separare species of cichlids from
Lake Victoria. Given the esrimated rates at which
mutations occur in mitochondrial DNA, researchers
believe that the entire assemblage of cichlid species in
Lake Victoria has arisen within 200,000 years. That is
equivalent to about one species every 500 years.
One reason for such rapid speciarion is that the
African lakes have repeatedly experienced periods of
• • •• • . • • ••• • • • •• • • • • •• •• • . • •• • • . • " •• • • •• • • • • • • • • • . • •• •• • • • • • • ••• • • • •• • • • ••• • • •, • • • •••• • • • • • •' • •
... ·························
······
sure is sufficient to allow for speciation to occur
pres function of depth, without requiring any form of
as. ariance (geograp h"1c 1.so1at10n
. ).
vica Natural selecnon. has greatly .nfl dh
I uence t e spec1es
. The most likely precursor to speciation in either
clusters of cichlids _in r_he African lakes in ways mher
han speciation. C1chhds have adapted and d1vers1-
~ed ro become specialists and occupy many ecological
iches. As an example, there are cichlids that feed by
~ipping off the scales of other cichlids. These predatory
fish are actually specialized, by species, to nip scales ei-·
ther frorn rhe right or from the left side of their prey!
The evolutionary process whereby one kind of or-
ganism-in this case, an ancestral cichlid-evolves into
numerous species, each uniquely specialized, is called
adaptive radiation. There are numerous examples of
adaptive radiation both in the fossil record and among
extant animals and plants. The famous Darwin's
finches of the Galápagos Islands forrn the classic ex-
ample of adaptive radiation. Less famous, but no less
interesting, are the Scalesias of the Galápagos lslands,
plants in the daisy family that have evolved from a
single colonizing ancestral species into sorne 20 species
found throughout the islands. But 13 species of Dar-
win's finch and 20 species of Scalesia pale in compari-
extreme dryness, followed by replenishment. Evidence son with the amazingly rapid and diverse evolution of
suggests that Lake Victoria, fo r example, was nearly the cichlids of the Great Rift lakes in Africa.
dry as recently as 14,000 years ago. During dry pe- The African cichlids have not only undergone a
riods, cichlid populations would become smaller and stunning adaptive radiation in each of the three large
be isolated from one another, ideal conditions for ge- lakes, but they have also evolved an amazing conver-
netic divcrgence and subsequent speciation. (Compare gence. Species not closely related from different lakes
this with the refugia model for speciation, described ha ve come to bear a remarkable resemblance to one
later in this chapter.) With repeated waves of dryness another, occupying essentially equivalent ecological
would come repeated bouts of speciation, increasing roles (niches).
the number oí species. But that's not ali. The combined effects of rapid speciation and nat-
Cichlids in the African lakes show strong evidence llfal selection on a rime scale of about 200,000 years
of natural selection acting on rheir sensory systems as resulted in the extraordinary diversity of cichlid fishes
they perceive one another, something termed sensory in the African lakes. Evolution is not sluggish.
drive speciation. As males have diverged in color, fe-
male cichlids show strong preferences for male color
Spe~iation in the Tropics
types. This is likely due to the fact that color penetra-
tion in the various lakes varíes with depth, and thus Speciation under the BSC involves establishment of
reproductive isolation between two populations. For
the perception of color must vary. Red becomes more
predominant with depth, blue less predominant. Thus this to happen, the f!ow of genes between populations
red males are easier for females to see in deeper wacers. must be prevented or severely reduced for a sufficient
lt has been learned that certain genes, termed opsin number of generations to permit genetic divergence
genes, are responsible for adjusting the color detection adequate to establish reproductive isolation. Once
of the fish, and red-biased opsin genes typify fish of two populations have genetically diverged such that
deeper depths. Thus females at shallow depths prefer they are unable to produce fertile offspring, they be-
bluer males, while females at greater depths tend to come separare species. Speciation under the PSC re-
mate mostly with reddcr males. The study (Seehausen quires that populations diverge genetically to the
et al. 2008) concludes by suggesting that the selection point where they are clearly distinct, though such a
BIOGEOGRAPHY AND EVOLUTION IN THE TROPICS 65
disrinction may not necessarily be obvious. Cryptic
species are known in which members of each species
appear highly similar but are geneiically divergent.
case is vicariance, the occurrence of physical barri-
ers that fragment populations and prevent gene ílow.
Mountains, rivers, deserts, and savannas ali represent
possible barriers between rain forest sites. Other fac-
tors, most notably climatc changes, can fragment spe-
cies' ranges, creating vicariance (see the section "Age,
Endemism, and Refugia: Many Questions," later in
this chapter). lf a mountain is formed by uplifting of
Earth's crust, as has happened extensively in the Andes
Mountains in South America, what was once a con-
tiguous area of forest will be fragmented by the moun·
tain range. Individuals on one side of a mountain, or in
an isolated valley, or even at various elevations on the
-mountainside are prevented from mating with other
populations because they cannot easily reach them.
Once populations are separated by geographic factors,
there is the strong possibility of genetic divergence
with time between the fragmented populations.
For instance, a population on the west side of a
mountain range may be subjected to different selec-
cion pressures than a population on the easr side of the
range. Each popularion will be selected for somewhat
different characteristics, and thus for different genes.
Specifically, this will result in genetic differences from
one population to another among alleles at specific
chromosomal locations. (An al/ele is one or more al-
ternative forms of a gene that occur at a specilic site,
or loc11s, on a chromosome.) Since vicariance prevents
exchanging of genes, fragmented populations diverge.
They may also diverge genetically due to random loss
and fixation of alleles, a process called genetic drift.
But genetic drift is not as powerful in causing diver·
gence as natural selection, and it does not result in
adaptational change.
The Andes Mountains, Amazonia,
Vicariance, and Speciation
Much of South America remains geologically active.
The Andes Mountain chain, which has been uplifting
since at least the Mesozoic era and became particu-
larly active during the Cenozoic era (approximately
65 mili ion years), is responsible for the inicial creation
of diversified habitats as well as providing numerous
climatic and physical barriers that greatly enhance
geographic isolation among populations (Piare 2-14 ).
More recently, approximately 20 million years ago,
 66 CHAPTER 2
···················· ·······················•··············.. ·····-····"···················· ······················· ···················•·······························
1 1
PLATE 2· 14
1 1 The complex topography of the Andes
Moumain chain has resulted in numer-
ous vicariant speciation events through~
out the full range of the Andes.
···················· ··· ······· ····· ·· ····················· ·············· ·········
additional uplift events affected dispersa! patterns,
isolating ecological communities and stimulating evo-
lution (AntoneUi et al. 2009).
Geologists havc determined that elevarional
changcs throughout the history of the Andes have oc-
curred in short time spans, from J to 4 million years,
························ ........ .......... ·······••·······. ············••·•····························· ················ ········································............
7'l'W
16'S
FIGURE 2-13
The altiplano region of the central Andes.
..... ····· ... ............................................ ·····•· ........................................................................ .
OL-.-L-U-'--~- - ~ - - ~ - ~ -- ~ -- ~ - - ~ - ~ - - ~ ~
o 00 3(,)
with the mountain chain remaining stable for long in-
tervals berween such times of rapid change (Garzione
et al. 2008). The topography of the central Andes is FIGURE 2-13
(continutdl
complex, consisting of eastern and western cordilleras
separaced by an expansive altiplano (Figure 2-13). The
copographic complexity of the Andes chain makes for a veritable engine driving speciation. Consider the fol-
lowing exJ1nples.
7(f'W 68'W
A likely cxample of vicariant speciation can be
6f!/'W 64'W
demonstrated for tapirs. Baird's tapir (Tapirus bairdii)
is found ,mly in lowland forest on the west sidc of the
Andes, extending inro Central America as far north
as tropical Mexico (Figure 2-14a). The similar Brazil-
16'S ian tapir (T. terrestris) occurs only on the east side of
the Andes, occupying the entire Amazon Basin (Figure
2-14b). The Andes Mounrains geographically isolate
these two species, and likely provided the main factor
in the initial split of the ancestral species inro isolate
18"S populations. Last, there is a third eocropical tapir
species, che mountain tapir (T. pi11chaq11e), which as
the common name suggests inhabits montane (mean-
ing moumainous) fotests at mid- ro high elevations of
the central and eastern cordilleras of the Andes, in Co-
20'S
lombia and Ecuador (Eisenbcrg 1989). The mounrain
tapir is isolated both by range and elevation from the
other rwo species.
Chat-tyrants are common insectivorous birds in
22'S the huge family (almost 400 spccies) of tyrant fly-
catchers (Tyrannidae). Figure 2-15 illusrrates the dis-
tribution of a cluster of chat-tyrant species found in
the Andes Mountains from Colombia to Bolivia. The
(a) group is composed of four closely related, phenotypi-
cally similar species (with subspecies) divided inro two
superspecies. A superspecies is a very closely related,
BIOGEOGRAPHY ANO EVOLUTION IN THE TROPICS 67
..
. ,. i'J¡.) 8'JC
(b)
recenrly evolved cluster of species, cypically quite sim-
ilar in morphology. Such a pattern is possible from
isolation and subsequent genetic divergence of local
populations ar various rangcs and elcvations. The
(ro11talis superspecies occurs on the west side of the
Andes and is composed of the four birds on the lefc in
the figure, and the diadema superspecies is found on
the east side and consists of the three birds on the right
in the figure. 1ore the similarity among them-such a
complex pattern suggests that speciarion is strongly
facilitated by Andean topography.
As another example, metaltail hummingbirds (ge-
nus Metall11ra) show a rapid and complex evolucionary
patrern. By comparing 345 base pairs of the mitochon-
drial cytochrome b gene and using cladistic analysis, rc-
searcbers have determined that there are seven species
at various ranges in che Andes, one of which exhibits
seven subspecies (Figure 2-16). Mitochondrial DNA
in birds shows an average mutation cate of 2% every
million years. Using that "clock," the metaltail spe-
cies and various subspecies ali evolved wichin the past
rwo million years. This demonstrates rapid evolution
and is characteristic of other srudics of Andean bird
groups (Price 2008). The stimulus for speciation seems
to have bcen mountain uplift and/or climate change,
isolating sorne popularions of hummingbirds on differ-
enr mountain peaks, while at least one specics occupies
an extensive mid-elevational belt. The midelevational
 68 CHAPTER 2 BIOGEOGRAPHYANO EVOLUTION IN THE TROPICS 69
..... ..................... ..................... ........................ ........... , ......................... .
,.\
•• •• ••• ••• l •••• • • o l l •••• • • • •I •• •• • . . • ••••• ••• • •
·······•··········

1 1

' 1

FIGURE 2·15
The char-tyranr genus Ochthoeca shows
a complex panero of differenriation in the
(a) Andes from southern Colombia ro central
Bolivia. The birds on the left are patt of the
O. fro11talis group (called a superspecies
because there are different disrinet
populations). The group on the right is part
of 1he O. diadema superspecies group, found
on the eastern slopes of the Andes. Note
how sinúlar 1he species are 10 one another, a
result of recem evolutionary divergcnce. This
A is an actively evolving group of genetically
closely related populations, separated to
various degrees by their ranges within the
Andes chain.
............. , ..................................................... ·························· ····················· ····································· ············••············

FIGURE 2-16
Viridian D The cladogram on the left depicrs relation-
Viridian ships among a group of hummingbird
1
Violet-lhroated species of the genus Metal/11ra (rhe meraltail
Neblina 1 hummingbirds). The cladograrn is bastd on
1 comparisons of sections of mitochondrial
Coppery D DNA, a commonly ustd index to ascertain
Fire-throaled 1 degrees of genetic relationship. The colored
squares correspond to the ranges of the
Scaled 1 various species shown in color on the map.
1
Species differ in elevation as well as geo-
2
graphic location. The dashed line indicates
3 two million years, so 1he cladogram depiru
4 Tyrian D a group"of recently evolved species. Note the
(bJ 5 high rate of genetic differenriation occurring
FIGURE 2·14 6 in the Tyrian metahail, shown in the lower
(a) The range of Baird's tapir. Note 1ha1 it occurs throughour Central America and on 1he most western side of the Andes in Colombia and 7
cladogram, and note 1he cxrensive range of
Ecuador. (b) The range of 1he lowland tapir. that species shown in the map.
 70 CHAPTER 2 BIOGEOGRAPHY ANO EVOLUTIDN IN THE TROPICS 71

·- :. :-.. . _) ..r_ _._ _,. .:l

J •• • :,:· · · · · · · •••••• , ••••• •••••••• ••• ••••••••••••• ••••••••• ••••••• • ••••••••• , ••••••••••••••••••••••••••••••••••••• •• •• • •••• ••••••• ••••

·······
arcas of probable endemism throughour rhe zon and irs tributaries is sufficicnt ro isolare popula-
"-..,,... \ ~ ; _'/
. ·.
rnerouS
nu . r Each of the postulared areas of endem1sm
onunen . f . f d.
. tions of birds whose individual members are relucrant
.. \. '\• .....-, ... e . unique assemblages o spec1es nor oun m to cross such a wide expanse of water (Haffer 1985;
0 1a1ns .
2 'ºrher ar .
eas This helps ro accounr for rhe umquely
.
Haffer and Fitzpatrick 1985). The Amazon River iso·
• •• J.,. •,.•·,,,•_J -___.__,...., , .,-..,~ O>'-.:": • ' ... .. ,
º. h diversity of birds throughout rhe connnent, and lates rwo similar specics of antbirds. The dusky anrbird
~,g srs rha1evolurionary parteros have been com- (Cercomacra tyrannina) occurs norrh of rhe Amazon,
,r sugge d 86 while che similar blackish anrbird (C. 11igresce11s) oc-
lex (Figures 2-18a an 2-1 ). .
P With rhe rise of rhc Andes Mountams, the Ama- curs south of rhe river. Borh species occur rogerher only
River began its flow from wesr ro east. The mas- along a secrion of the northern bank of rhe river, bur
. • ~azon River and 1.ts numerous w1'de tri.buranes
zoo . here the blackish anrbird inhabirs wer varzea foresrs,
8 s1vee"'"serve<l ro isolate rracts of foresr and savanna, and and rhe dusky antbird favors second-growrh vegera-
hav
60 •
g1ve 0
rhe scdenrary narure of many anuna · 1popu1at1·ons
.
tion of the terre firme (Haffer 1985). (Terre firme is a
10 • Amazonia, rivers have probably served as unportant term used for areas of forest and savanna that occur
::rces of geographic isolarion. Toe widrh of the Ama- off che riverine floodplain.) Asimilar partero is evidenr
1 in che distribution of rhe razor-billed curassow super-
species (Figures 2-19a and 2- 19b).
14
A comparison of bird species evolurion in Sourh
America and Africa demonstrates che grearer diversi·
fication rhar has occurred in South America. Over rhe
16

76
. \::''..CS; pasr 12.5 million years, rhe rare of speciarion minus
extinction has been far greatcr in Sourh America rhan
in sub-Saharan Africa (which acrually exceeds Sourh
America in arca). South America is esrimared ro have
3,150 land and freshwater bird species compared with
o 1,623 in sub-Saharan Africa (Price 2008). Using mo·
lecular analysis, ir has also been learned rhat bird spe-
cies from Andean areas of endemism are in general
more recently evolved than those of lowland Amazo-
nia. Yo1111g species (in rhis specific contexr) are defined
(a)
as belonging ro a clade (a cluster of closely relared
species) of ar least 10 species whose common ances-
tor dates back ro abour six million years or younger,
and old species are rhose whose ancesrry is further re-
moved in rime.

10
FIGURE 2-17
Thcsc two figures dcpict rhe
ranges of 10 spccies oí birds 12
Age, Endemism, and
that, r.kcn rogcrh,~ define thc
lnambari ccntcr of cndcm1sm. 7
Refugia: Many Questions
The bird spccics are from 14
widcly distinct groups. Arcas
How does rhe high diversiry wirhin tropical regions
oí cndcmism b<comc apparcnt rela re ro geological- long periods whcn speciarion
whcn rangcs oí many spccics 1 ¡·_ .• ,_, • ~. ••• • • /-\~-- - \.\~- ~
-~ mighr exceed extinction? Conrrasring Sourh America
converge, as shown in the 76 °z_ '-.• \ ..•._.. wirh orher major tropical regions, one finds rhat as
figures. , , . with bird species richness, plant species richness is
.......................................................................... also unequivocally highest in Sourh America. Bur how
····································································· ·············· .....
did rhar happen? Whar factors are responsible? South
specics could provide dispersing individuals that colo- Figure 2-17 shows such an analysis for part of South (b) American species richness, as extreme as ir is, has long
nized various high montane peaks. America. 1ore rhe high correlarion among rhe ranges posed vexing quesrions for cvolutionary biologisrs and
FIGURE 2·18
Byexamining rhe ranges of many species, ir is pos- of rhe 1Obird species illustrared on rhe maps. Postui,,ed areas oí cndcmism for birds in South Amcrica. (a) Arcas biogeographers. Since rhe 1970s, a debate has been
sible ro asccrtain geographic areas where groups of Analysis of bird disrribution patrerns throughour oí posrulated cndcmism in nonhcrn South Amcrica. (b) Areas of ongoing berween rhose who believe mosr speciarion
species trace rheir origins, called ce11ters of e11demism. South America by Joel Cracrafr has demonsrrated POStulated endcmism for Amazonia and thc Andes Mountains. in Sourh Amcrica (and by implication, elsewhere in
 "I!

72 CHAPTER 2 BIOGEOGRAPHY ANO EVOLUTION IN THE TROPICS 73

\ ............. ~-· ....... ' .... ' .... . ·········· ·•·· ··· •···
......................................
• :1·

the global tropics) is recent, dating primarily to events
in the Pleistocene and immediately before, and those
who argue that che data do not support such asser-
tions and that much of the speciation occurred many
millions of years before the Pleistocene.
lt is unlikely that the equatorial tropics were climat-
ically stable and constant throughout the Pleistocene
(from 1.64 million to 10,000 years ago), undisturbed
by the giant glaciers bearing clown on northern tem-
perare areas. Thomas Belt, in 1874, discussed the pos-
sible effects of northern glaciation on the tropics. More
recently, studies from geomorphology (che historical
development of present landforms}, paleobotany (the
study of past patterns of plant distribution), and bioge-
ography suggest dramatic changes in Amazonia during
the Pleistocene (Colinvaux 1989a and 19896; Haffer
1969, 1974, 1985, and 1993; Prance 1985; Simpson
and Haffer 1978). From these studies has emerged the
refugia hypothesis for high speciation rates in the trop-
ics, and particularly in the Neotropics.
lt is hypothesized that during glacial advances
in northern latitudes, the tropics may have become
cooler and drier. Por example, during part of the Pleis-
tocene, temperature in the Ecuadoran foothills, east
of che Andes, was between 4ºC and 6ºC cooler than
ar present (Colinvaux 1989a and 19896). A possible
ecological result of the climatic cooling was to alter
and shift the distribution of ecosystems. Ecosystems
such as savannas presumably enlarged, and moist for-
ests presumably contracted. Large continuous traces

1 ¡ ................. .
tyrannina

1
Mitu tuberosa
11 1
1
'1 1

O 500 1000 kilometers

(b)

/
.,; lcontinued)
......................................................................................... ·······························•···················

of lowland rain forest were therefore fragmented into
varying-sized "forest islands" surrounded by "seas" of
savanna or dry woodland. Savanna expansion, in the
refugia model, created vicariance among forest tracts.
Because of the repeated shrinking and fragmenting
of forests, forest organisms periodically became geo-
graphically isolated from populations in other for-
est arcas, promoting speciation (Nores 1992). The
Amazon Basin became a climatically dynamic "archi-
pelago" of varying-sized rain forest islands. In other
o 500 1ooo kilometers
words, rain forest species were confined to fragmented
refuges during the driest periods. These small, island-
(a)
like "rcfuges" promoted speciation among plants and
FI GURE 2-19 animals.
The distributions of two superspecies groups of Amazonian birds dosely follow the geography of the river systems in Amazonia. (a) This Studies based on the current distribution of cer-
map shows the distribution of rwo dosely related antbird species in Amazonia. Note that they are separated based on occurrence either
north or south of the Ama20n River. (b) This figure depicrs the distribution of the razor-billed curassow species group. The species are
tain kinds of birds (Haffer 1969, 1974, 1985, and
morphologically similar but are regionally separated. Note that the Amazon River separares the southern Mitu tuberosa from the more 1993; Haffer and Fitzpatrick 1985; Simpson and
northem populations, and note the isolated Mitu mitu, found in the Brazilian Atlantic forest region. Haffer 1978) postulate that at least nine major and
numerous smaller forest island refuges were present
in Amazonia during che Pleistocene (Figure 2-20).
They assert thar many taxonomic groups subse-
,quently went through periods of rapid speciation
because there were repeated episodes of rain forest
shrinkage and expansion. During interglacial periods
forests expanded, and secondary contact was estab-
lished becween newly speciated populations, explain-
ing why so many extremely similar species can be
found today in Amazonia.
The refugia model has been and continues to be
the subject of much debate (Haffer 1993). Evidence
supporting it is based mostly on present distribution
and diversity patterns. Por example, Prance (1982a
and 19826) has closely examined woody plant diver-
sity and concluded that 26 probable foresr refuges
existed for these plants. Kinzey (1982) concluded
that present primate distribution fits predictions of
 - 74 CHAPTER 2 BIOGEOGRAPHY ANO EVOLUTION IN THE TRO PICS 75

............. ................. .............. .

,

1
1 1

1
1 1
1 1

1 1 i
1 1
FIGURE 2-20
This series ol maps depicts the presumed forest refugia during the dry phases ol the Pleistocene. The left-most map dcpicts refugia based
on bird distribution, the ccnter map is based onAmazonian lizards, and rhe right-mosr map is based on butterflies o! rhe genus Heliconius.
Note that the presumed refugia differ substantially among the groups.
· · · · · · , , , , , , , , , , , , , , , , , , , ,,,,,,,, , , . . . . . . . .. .. . . . . . . . . . . , ., .,,., .. . . . . . . . . . . ,, . , . .• , ..... , ... . , . . . . . ; ; , 1 .. . ... . . . . . . . . . . . . . . . . . . . . .. .... . . . . . . . . . . . . . . . ... , ..•.•..•.

the refugia model, and Haffer (1974), Haffer and the refugia model (Bush and de Oliveira 2006). Pollen
Fitzpatrick (1985), and Pearson (1977 and 1982) is resistant to decomposition, particularly in anaerobic PLATE 2-15
These photos oí fossil plants demon-
have presented evidence from present bird distribu- sediments of lake basins. As pollen accumulates year strare morphological diversity in rhe
tion in support of the model. In a series of papers after year, it forms a vertical poi/en profile, a historical Laguna del Hunco flora from about
by Keith Brown, cited by Prance (19826 and 1985), indicator of which plant species were present in a re- 52 million years ago.
heliconid butterflies have been shown to have 44 gion. (Lake sediments are cored, and pollen at tlie bot-
centers of endemism throughout Amazonia and sur- tom of the core is oldest, pollen at the top youngest.)
rounding areas. But it is important to note that cen- Pollen profiles from various Arnazonian lakes showed
ters of endemism, the presumed Pleistocene refuges, that although. the region likely cooled by about 5ºC, Another objecrion to the refugia model focuses (Willis and Whittaker 2000; Mayle 2004). Rather, evi-
do not precisely overlap among taxa, which lessens it remained fully forested, not broken up by savanna. on the ages of species. The refugia model suggesrs dence now seems to favor the temperate zone as the
support for the refugia model. Supporters of the re- Other strong objections to the refugia theory have thar sr eciation has been recent. As mentioned ear- region where refugia may have been important drivers
fugia model respond that different taxa have differ- been raised (Willis and Whittaker 2000). One analy- lier, m~lecular analysis (unavailable when the refugia. of speciation.
ent dispersa! powers and different generation times sis of Amazonian sedimentary patterns, for example, model was proposed) indicates that many bird species
and thus would be expected to differ sornewhat with suggests that there was no substantive climatic change in Amazonia trace their origins to more than six mil-
regard to the degree of regional endernism (Prance or reduction in forest cover in Amazonia during the
1982a and 1985). Pleistocene (lrion 1989). However, what is suggested
lion years ago, well before the Pleistocene. A complex
of frog species of the genus Leptodactylus exhibics
The Great American
The refugia model has been subject to strong criti-
cism. Colinvaux (1989a) summarizes the arguments
by sediment patterns is not a stable Amazonia, but
rather strong oscillations throughout the Pleistocene
high species richness dating back beyond the Pleis- Interchange
toceno. Speciation appears to have occurred in the
against the Pleistocene-based refugia model, pointing in the distribution of land, water, floodplain forests, mid-Tertiary period (Maxson and Heyer 1982; Heyer The Panamanian land bridge, now called che Isthmus
out that at least one study showed that for plants, ref- and terre firme. Sea-leve! changes are thoughno have and Maxson 1982). Studies of plant fossils and sub- of Panama, formed approximately three million years
uge locations coincide with areas in which sampling of led to the alternation of huge Amazonian lakes with sequcnt analysis indicate that the high plant species ago beca use of a combination of uplift of the northern
plants for herbaria specimens has been historically most strong valley cutting, geological events that would richness traces back fully to 52 million years ago, in Andes and a global drop in sea leve], perhaps as much
intense. This, of course, would suggest that the refuges presumably have a strong impact on flora and fauna. the Eocene epoch of the Tertiary period (Wilf et al. as 50 meters, a result óf the increasing size of the po-
are artifactual derivatives of uneven sampling effort. This view sees the Amazonian rain forest as subject 2003¡ (Plate 2-15). . lar ice caps (Marshall et al. 1982). Thus it was just
Analysis of pollen from sediments taken from the to vicariance, though not necessarily reduced to scat- At present, the evidence does not strongly support prior to the onset of the ice age that the continents of
Amazon lakes in Ecuador and Brazil do not support tered refuges. the refugia model as a species pump in tropical regions Norrh and South Arnerica were no longer separated
 76 CHAPTER 2
BIOGEOGRAPHY ANOEVOLUTION IN THETROPICS 77
·······································••············ ···••·····•····························•································ ....................................
. ..... ..................... ······.................... ············•········ ···············································································
by water. The Panamanian land bridge profoundly pendemly of one anocher for at least 40 million years che !and bridge formed, sorne South Ameri- The ground sloths were probably killed by humans as
alrered the ecology of South America, much more so but their mingling, once che land bridge developed' 0 nce Is moved norrhward as early as 2.5 million the human population spread southward. The orhers
chan ic did chat of Nonh America. Consider that the
fauna of North and South America had evolved inde-
was completed witbin a mere two miUion years (Webb can anrmaO
. to Nortb Amenca.
Jicerally walkmg . These have simply dropped out of che fossil record.
1978) (Figure 2-21). Yearsag, . . d Many North American animals walked across che
. d d rwo armad1llo spec1es, a g1ypto ont, two spe-
1nc 1 · a ¡arge capy-
. u felar"e uround sloths, a porcupme, land bridge to South America. Their impacc on the
.............................................. , ....................... ...... ··························•············· ············••········••····················
O
cies nl o~e phorusrhacoid bird. Ground sloths, native fauna was substantial. The list of invaders in-
bara, a extinct were very 1arge terrestrta . 1s1oths w1.th eludes skunks, peccaries, horses, dogs, saber-toorhed
nowa 11 ' .
ws rhat likely fed on fo!tage. They were closely cats, other cats, tapirs, camels, <leer, rabbits, cree squir-
Iongcla
ced to the much smaller tree s!oths of the Neo- rels, bears, and an odd group of elephant-like mam-
1
re ª •es Glvptodonts looked like gigantic armadillos, · mals, the gomphotheres. Add to chis list the field mice,
tropt · , . .
ered with bony armor, whose ta1ls (dependmg on or cricerid rodents, whose travel rouce to Souch Amer-
covecies) sometimes tenrnnate · dm . a mace-l'k1 e cIub. ica is still debated, but who have since radiated into
I' ¿apybaras are cavimorph rodents, such as chinchillas 54 living genera, and you begin to see why the effect
and guinea pigs. The extant capybara (Hydroch_oerus of North American mamrnals on South American eco-
1 1
Jrydrochaeris), abundant U1 parts of Sourh Amenca, 1s systems was so great (Marshall 1988).
1 1 the world's largest rodent. Phorusrhacoid birds, now Various amphibians, reptiles, and bird groups
ali extinct, have earned the nickname terror birds. also migrated in eirher direction, sorne from the norrh,
Sorne srood almost as tall as a human being. They ·sorne from rhe south. The fauna! interchange alrered
were flightless but could move quickly, running on ecological communities on both continents.
long !egs. Their huge, raptor-like beaks were easily The exchange scrongly altered the look of rhe
capable Jf killing small to moderate-sized mammals. South American fauna such that it is today much
At the time they lived, they were among the top car- more like rhat of North America. As Marshall (1988)
nivores of Souch America (Figures 2-22a and 2-22b). so aptly summarizes (about mammals), "the Great
Other invaders included Didelphis, the familiar American lnterchange resulred in a majar resrructur-
1 opossum and North America's only marsupial mam- ing. Nearly half of the families and genera now on
mal. Opossums invaded approximately 1.9 million the Souch American continem belong to groups that
years ago (Marshall 1988). Lasr, from South America emigrated from North America during the last 3 mil-
carne at least one species of toxodon, an odd maroma! lion years."
that beh nged to a group named notoungulates. Tox- lt is quite unclear how che inllux of North Ameri-
odons were bulky mammals whose appearance was can mammals, as diverse as it was, affected the abun-
suggestive of a cross berween a cow and a hippopota- dances of South American mammal species. For ex-
mus. Thc collective impact of che South American in- ample, many hoofed mamrnals invaded from North
vaders was modest at best. Of their host, only armadil- America. However, their ecological counterparts in
los and opossums remain, both of which are thriving. South America, the litopterns and notoungulates,
FIGURE 2-21
This figure illustrates the ....................... .... ···············••··················· ·····•·····•····················· .................................................................................
pcincipal groups of mammals
involved wich che Great
American lnterchange
that resulted in mixing the
mammalian (and other animal)
faunas of North and South
America once the lsthmus
of Panama became exposed
approximately 3 million years
ago. The upper maps illustrate
how che posítíons of che
Amerícas changed from the
late Cretaceous rhrough the
Míocenc, bríngíng the conrínents FIGURE 2-22
sufficíencly close thar when Phorusrhacoíd bírd (leh) and toxodon (right).
glacíatíon occurred, the land These animals represent unique components
bridge was esrablíshed. of the faunas of South Ameríca before the
Great American lnterchange took place.
 78 CHAPTER 2
•',,,' ~··· ······· ·· ········· ··················
FIGURE 2-23
Smilodon (left) and Thylacosmilus (right) were
both morphologically similar, particularly
regarding the enlargcd sabcr-like canine teeth,
But Smilodon was a placenta] mammal, whereas
Tlrylacosmilus was a marsupial mammal. Their
1 1 similarity is a clear case of convergent evolution.
! complex physical structure and high biodiversity. This chapter describes the vari-
1 1
1 i
were declining in numbers and diversity long be-
fore the camels and horses arrived. (Litopterns com-
prised a camel-like group that included that unique
Machrauchenia, an animal that looked like a carne!
with an elephant-like proboscis.) lt is not considered
likely that the North American ungulares were the
¡1 .' primary cause of the extinction of the various South
American groups (Marshall 1988). Among predatory
manunals, the large, wolveri¡¡e-like South American
borhyaenoids were essentially outcompeted by the
terror birds, not by mammalian invaders from the
norrh. However, the phorusrhacoid terror birds may
have evemually lost out to invading placenta! mam-
mals. There is really only one relatively clear-cut ex-·
ample where it appears that there was direct competi-
tion between a North American species and a South
American species. This is the case of the marsupial
saber-toothed cat, Thylacosmilus. This animal bore a
striking anatomical similarity to the placenta! saber-
toothed cat, Smilodon, an example of convergem evo-
lution (Figures 2-23).
Smilodon crossed the land bridge into South
America, and the extincrion of Thylacosmilus coin-
cides closely with the arrival of Smilodon. Much of
the extinction of large mammals on both cominents is
probably explainable by the proliferation of humans
during the Pleistocene (Marshall 1988) and not due
to competition among the animal groups themselves.
The Great American lnterchange demonstrates
that ecological communities are highly subject to ma-
jor change over time. The mixing of the two· fauna
created new communities. One of the most challeng-
ing questions in ecology, one related both to long-term
biogeographic history as well as to present ecologi-
cal imeractions, is to understand how various factors
structure communities.
lnside Tropical Rain Forests: Structure
Chapter Overview
Rain forests stand out among other tropical terrestrial ecosystems because of their
ous structural attributes of rain forests and provides an introduction to rain forest
structural diversity. In Chaptm 4 and 5, we will take a detailed look at biodiversity.
Why Rain Forest Is Unique
Rain forest is a major terrestrial biome. A biome is an ecosystem type that has
distinctive and characteristic plant and animal species and that is largely deter-
mined by climatic characteristics. (Compare chis definition with the Holdridge life
zone system described in Chapter 1.) Rain forest represents a type of biome, with
characteristics distinguishing ir from other biomes such as desert, savanna, tundra,
or various kinds of temperate forests (e.g., boreal needle-leaved forests and broad-
leaved deciduous forests). Biomes are fundamentally the result of global differences
in climate (see Chapter 1). Tropical rain forest biomes occur equatorially, where the
temperature remains warm and conditions are generally wet throughout the year.
There are two broad characteristics that are evident in rain forest. The first is
complex physiognomy. Physiognomy, when used in ecology, refers to the physical
structure of an ecosystem. Rain forests are voluminous, high in biomass, and struc-
turally elaborate. The stature of the trees is usually tall (between 30 and 35 meters,
and occasionally taller), and from the canopy to the ground, the physical structure
is considerably more varied than in other forest types (Place 3-1). This reality mea ns
that numerous life forms, both plant and animal, have the opportunity to evolve
and specialize in different vertically structured microhabitats in rain forest. That
contributes to the other major characteristic-high biodiversity of numerous taxo-
nomic groups.
Biodiversity has various meanings, depending on precisely how the term is
used. The most basic meaning is that biodiversity refers to the number of species
within a habitar, termed species richness. But the term may also refer to how species
richness is distributed relative to species' population sizes, termed species evenness.
Concepts such as species diversity; evenness; and alpha, beta, and gamma diversity
will be discussed in the next chapter. Biodiversity may sometimes refer to genetic
diversity among populations within a species (or individuals within a population),
an important parameter in much conservation research.
No matter where you are on Earth, once you entera rain forest, biodiversity be-
comes higher, ofren much higher, than is found in adjacent or nearby non-rain-forest