H O T S P O T S

R E V I S I T E D

SERIES PRODUCER
CEMEX BOOKS ON NATURE
PATRICIO ROBLES GIL

RUSSELL A. MITTERMEIER
PATRICIO ROBLES GIL
MICHAEL HOFFMANN
JOHN PILGRIM
THOMAS BROOKS
CRISTINA GOETTSCH MITTERMEIER
JOHN LAMOREUX
GUSTAVO A.B. DA FONSECA

PREFACE
PETER A. SELIGMANN

FOREWORD
HARRISON FORD

CEMEX
2004
In recent years, thanks to the efforts made by CEMEX and other orga-
nizations that share with us the commitment to promote biodiversity
protection, it has been possible to define new, innovative environmen-
tal strategies and policies for curbing the deterioration of hotspots,
Earth’s most endangered terrestrial ecoregions.
The hotspots concept was developed in 1988 by British ecologist Nor-
man Myers; since then, it has been supported by organizations such as
Conservation International. The dissemination and impact of this con-
cept were widened in 1999, with the presentation of our book Hotspots.
Earth’s Biologically Richest and Most Endangered Terrestrial Ecoregions,
the second volume in a series of works published jointly with Conser-
vation International and Agrupación Sierra Madre.
The scientific information gathered to date has made it possible to
identify new areas of great ecological importance to be included within
the hotspots classification, as well as to obtain relevant data on zones
that had originally formed part of this category. Thus, we have been
able to make progress in what definitely should be considered a global
priority: the conservation of biotic resources. In the diversity of life-
forms we find the very possibility of maintaining the wide range of al-
ternatives for achieving sustainable development.
Convinced that the goal of ensuring a living planet requires a fore-
sighted strategy along with collective, coordinated efforts, at CEMEX we
are very proud to work once again with Conservation International and
Agrupación Sierra Madre, presenting this new book, Hotspots Revisited.
On the pages of this volume, you will find clearly stated the ways and
spaces in which we should focus our attention and resources in order to
achieve the survival of the greatest possible number of species and
ecosystems. Included among the new terrestrial ecoregions considered
as hotspots is the Sierra Madre Oriental and Occidental of Mexico.
Located there is El Carmen, a project of wide scope that CEMEX has
promoted for five years now for the purpose of having a direct bearing
on its conservation.
We hope that the enhancement of the hotspots concept and the strate-
gies proposed in this new book will contribute towards thinking on a
world scale and acting on a local one so as to successfully meet the chal-
lenge posed by biodiversity conservation.
As a worldwide enterprise fully committed to promoting sustainable
development, at CEMEX we are confident that the resources devoted to-
day to this goal will entail the best possible investment for ensuring the
permanence of biodiversity on our planet.

CEMEX

7
The Garden of Eden was green, lush, and rich in
wildlife, and a sanctuary where humankind and nature
lived in harmony. Sadly, this image is no longer a real-
ity for those unique places which harbor the majority
of life on Earth. We share our world with at least five
million species of animals and plants, and perhaps as
many as 30 million. This biodiversity is our most pre-
cious resource —our living heritage. All cultures revere
this diversity in one way or another, through music, art,
literature or tourism. All religions charge humanity
with caring for life on Earth. Biodiversity is what dis-
tinguishes our planet from the rest of the universe; in-
deed, it defines us. Biodiversity is unique and irre-
placeable.
Around the world, however, greed and poverty con-
spire to extinguish this variety of life. Overconsumption
is the most grievous of these insults. Our world’s forests
and other natural habitats are being decimated to feed
the lust of the so-called “developed” world for timber,
minerals, exotic pets, luxury foods. Meanwhile, unjust
distribution of wealth forces billions of people in the
“developing” world to depend on exploiting the re-
maining scraps of nature. They hunt the last wildlife
and clear the last habitats for marginal agriculture. As
the world becomes more tightly interconnected, the
threat of invasive species becomes ever more potent.
Such exotics are the death knell for those species that
have evolved without such predators or competitors.
Cataclysmic global warming looms on the horizon.
Neither the places where species live nor the threats
that we impose on them are uniform around the plan-
et. Those places holding the greatest concentrations of
biodiversity also face some of the most intense pres-
sures. More than half of the world’s plant species and
more than a third of Earth’s mammals, birds, reptiles,
and amphibians are found in just 2.3% of its land sur-
face. Most of these areas are tropical, like the Atlantic
Forest, the Caribbean, Madagascar, Sundaland, the Philip-
pines, and the Tropical Andes. They are called the bio-
diversity hotspots. No matter how successful conserva-
tion activities are elsewhere on the planet, unless
threats are reduced in these soon, we will lose at least
half of Earth’s diversity of life.
These biodiversity hotspots hold some of the high-
est human population densities on the planet and
some of its poorest people. Their poverty is a direct re-
sult of the destruction of forests, erosion of soils, pol-
lution of rivers, and overharvesting of wildlife. Some
environmental destruction is fatal. For example, de-
forestation often causes disastrous flooding, but also
the emergence of infectious diseases. Most of the
world’s bloodiest wars are unleashed in the hotspots,
too, often driven by resource conflicts. Somalia, Af-
ghanistan, Palestine, northern Iraq, Timor, Haiti: all
lie in hotspots harboring exceptional numbers of plant
species found nowhere else.
Thoughtless exploitation of the Earth’s plenty has
destroyed the Garden of Eden. Human actions have
exterminated the harmony of nature and depleted the
gifts of clean water, productive soils, clean air, and
the abundance of life-forms required for people to lead
healthy lives.
The challenge which Conservation International has
accepted is to work in these hotspots to prevent irre-
versible loss of biodiversity and to eradicate absolute
poverty. Fortunately, many of the solutions to these
two problems coincide. The highest priority is mainte-
nance of the remaining natural vegetation cover in the
hotspots. This provides habitat for biodiversity and
ecosystem services like clean water for people. In the
longer term, it will be necessary to restore the natural
habitats across much of the hotspots. Meanwhile, re-
duction to sustainable levels of the harvesting of natu-
ral food, fiber, and fuel will be essential if these re-
sources are to persist.
Of course, these solutions will not come for free. Pro-
tection, restoration, and sustainable use of the natural
ecosystems of the hotspots will impose short-term costs
on those least able to afford them. They will only be
possible, therefore, with a dramatic increase of re-
sources flowing from the “developed” countries to the
hotspots of the “developing” world. But these resources
are far from unaffordable for the former. Over the next
10 years, an investment of $100 billion is needed in
these priority hotspots. By comparison, the war in Iraq
2003-2004 has so far cost the United States $168 billion
in military spending alone.
Among the most important ways to meet this chal-
lenge is through partnerships at many different levels,
including the private sector. In this regard, we are par-
ticularly pleased with our long-term relationship with
CEMEX. Their efforts to reduce their industry’s envi-
ronmental footprint and their support for biodiversity
conservation in Mexico have demonstrated their com-
mitment and leadership, and their support for this out-
standing series of books, now in its twelfth year, has
had a major impact around the world.
The hotspots concept was first developed in 1988 by
the ecologist Norman Myers. Now, 16 years later, and
after several revisions and updated analyses, it has
emerged as the dominant paradigm for global conser-
vation strategy. As we enter this new millennium, the
time has come for the world to recognize that the geog-
raphy and solutions of global poverty align closely with
those of the biodiversity hotspots. Only by tackling
these two agendas together will we truly be able to end
poverty and conserve life on Earth.
PETER A. SELIGMANN
Chairman of the Board and CEO
Conservation International
11
“I’m not a biologist, nor an expert on global strategies to
safeguard threatened species and vanishing habitats.
But I am deeply concerned about the future of our
planet,” Harrison Ford tells us in the Foreword of this
book.
This concern is shared by many of us who fight for
the permanence of life on Earth and are united in a
search for solutions. Yet usually we comprise a concert
of voices that does not manage to be heard due to the
tremendous complexity of the world in which we live.
Thus, we need a single voice, such as that of Harrison
Ford, so that we may be heard.
Among the different priorities set on an internation-
al level, the conservation movement is like the “new
kid in the neighborhood,” whose popularity is growing,
but who is far from gaining the attention he needs. This
movement has to compete for resources allocated to
needs deemed to be the most pressing for humankind:
food, health, education, religion.
Nevertheless, nature has always been closely linked
to traditional human preoccupations. Nature, or redis-
covering it, now defines many religions in the world,
and has constituted a major theme for our artistic and
cultural expressions. Lastly, no one can question the
fact that meeting our basic needs in terms of health and
food depends on whether or not our natural environ-
ment remains sound.
This book provides us with a clear strategy for con-
serving biodiversity, and indicates a line of action to be
followed, but society as a whole is who utilizes natural
resources and who can promote and achieve a change
in the way these are consumed. And society is also the
one to determine the need for maintaining the diversi-
ty of life on this planet and the extraordinary variety of
opportunities it affords. If we want biodiversity conser-
vation to be adopted as one of the most important social
priorities, we must organize and carry out an intense
campaign to “sell it” as an attractive, extremely essen-
tial product.
The hotspots concept was presented in another title
of the collection we have been producing for CEMEX.
In the proposal we are making today, that concept has
been strengthened thanks to an updating of informa-
tion on the former hotspots and data on new areas that
have been defined as such, which altogether more than
justify the publication of this volume. However, there is
another reason which, in my opinion, is also very im-
portant: the rules of the market.
As a conservation movement, we have forgotten that
we live in a consumers’ society and that any product we
wish to introduce onto the market should be handled
according to certain rules. Moreover, we must continu-
ally redesign the product we are selling —in this case,
the new priority ecoregions of the world— to keep our
public interested and thus gain greater penetration and
have a better market position.
For this reason, one of the problems we confronted
in producing this book has turned into an obstacle
keeping us from really being effective in communica-
tional terms. The scientific bases defining this conser-
vation strategy are novel, compelling, and foresighted,
but that is not enough to make them reach decision-
makers. Every decision calls for the support of a major-
ity of society, sometimes running against the econom-
ic interests of minority groups that could be affected.
The need to transform the natural world into a prod-
uct that society demands leads us to another matter.
Biodiversity includes all plant and animal species, not nec-
essarily only those that are known or even charismatic.
Nature as people are familiar with it, that which seems
most appealing to us, is but a minimal fraction of bio-
logical diversity. Countless books and magazines have
filled our eyes with images of beautiful flagship species
—some of which have even become icons for conserva-
tion campaigns—, while the great majority of species,
among them endemic ones, are virtually unknown.
When producing this book, we found ourselves faced
with the huge problem of a lack of images for many
species and regions that have not been photographed
by professionals because they are not appealing
enough for most publications. In that sense, we are in-
spired by the work of Cristina Mittermeier, one of the
compilers of this book, in getting together a group of
photographers that will open up new perspectives for
communicating the most pressing priorities for nature
conservation.
For their part, scientific institutions should incorpo-
rate the concept of marketing in their structure; in fact,
communication departments of conservation organiza-
tions will have to become true “advertising agencies.”
That is the only way our product, the natural world,
will benefit from an attractive, novel form of promotion
that may successfully compete in the aggressive world
of sales.
They say that money makes the world go round. To-
day we have the opportunity to define our future and,
by taking strategic advantage of marketing and adver-
tising, we may decide what course we wish our planet
to take in the future.
PATRICIO ROBLES GIL
President
Agrupación Sierra Madre and
Unidos para la Conservación
13
 FOREWORD
Five years ago, I was asked by my colleagues at Con-
servation International to write the Foreword for
Hotspots, an ambitious assessment of our planet’s at-
risk biodiversity. Although I have had the honor and
privilege to serve on CI’s board of directors for the past
12 years, I was reluctant to accept the assignment. I’m
not a biologist, nor an expert on global strategies to
safeguard threatened species and vanishing habitats.
But I am deeply concerned about the future of our
planet.
It is clear that any effort to address these complicat-
ed issues would demand the most complete and
scrupulous scientific understanding. As part of the ini-
tial analysis, all of the world’s major terrestrial envi-
ronments had been inventoried, with 25 emerging as
having exceptional endemism, being severely threat-
ened, and in need of immediate conservation attention.
Some of these regions were regarded as obvious priori-
ties right from the start. Others came as surprises as the
data was analyzed.
The original Hotspots had a remarkable effect. BBC
Wildlife Magazine, in celebration of its 40th anniver-
sary, declared Hotspots one of the Top 40 Wildlife Clas-
sics published during the last four decades. Shortly af-
ter publication of Hotspots, the World Bank, the Global
Environmental Facility (GEF), the MacArthur Founda-
tion, and the Government of Japan joined with Con-
servation International to create the Critical Ecosystem
Partnership Fund (CEPF), and to date have committed
$125 million specifically for conserving the world’s
hotspots. Meanwhile, a growing number of institutions
continue to incorporate the results into their programs.
Perhaps just as importantly, the hotspots analysis has
spawned serious debates about where today’s precious
conservation dollars should be spent to achieve the
greatest future impact. The study focused squarely on
one issue —preserving biodiversity— and then system-
atically measured, assessed, and assigned priorities.
One could argue that it would be better to focus on
global issues of birth control, poverty, disease, unsus-
tainable energy use or atmospheric pollution. Un-
doubtedly, these are clearly critical in the long term
to the quality of life on Earth. But in the race to prevent
the extinction of the greatest portion of Earth’s living
heritage, nothing advances our understanding and
guides our strategies as well as the hotspots model.
That’s why this updated analysis is so important and
why it has quickly become another essential element
in the conservationist’s tool chest. The experts have
added several new regions to an already substantial list
of global priorities. These new hotspots range from the
Pine-Oak Forests of Mexico and the southwestern U.S.
to the Horn of Africa, from the Himalaya to Japan.
Some of these regions, like those of the Solomons and
Vanuatu that form the new East Melanesian Islands
Hotspot, didn’t make the original hotspot list because
their habitats were still largely intact. They weren’t
considered threatened at the time. Yet in the blink of an
eye, their condition has essentially slipped from more
or less stable to critical.
The most significant challenge of our time is to pre-
serve and protect our biotic legacy. CI’s mission to de-
fend biodiversity places its staff in some 40 countries
around the world, in partnership with an amazing di-
versity of institutions and individuals from various
backgrounds and cultures. The strategies they have ap-
plied are informed by sound, independent science.
They have developed broad experience in determining
how best to respond to crisis situations, while at the
same time planning for the future in incremental steps.
I wish you good reading and hope that this book
helps broaden your perspective and understanding of
the challenge of conserving the natural world.
HARRISON FORD
15
EARTH’S BIOLOGICALLY RICHEST AND MOST ENDANGERED TERRESTRIAL ECOREGIONS

CAUCASUS MOUNTAINS OF
148 CENTRAL ASIA
MEDITERRANEAN 297
BASIN
CALIFORNIA 144 MOUNTAINS OF
FLORISTIC PROVINCE MADREAN PINE-OAK SOUTHWEST CHINA
118 WOODLANDS HIMALAYA 159
IRANO-ANATOLIAN 309
205
287 JAPAN
CARIBBEAN 333
ISLANDS
112 TAIWAN
361
INDO-
WESTERN
BURMA PHILIPPINES
MESOAMERICA GHATS AND
323 179
103 SRI LANKA
EASTERN 152 POLYNESIA-MICRONESIA
AFROMONTANE 197
HORN
241 OF AFRICA
GUINEAN
TUMBES- FORESTS OF 277
EAST
CHOCÓ- WEST AFRICA MELANESIAN
POLYNESIA-MICRONESIA
MAGDALENA 123 COASTAL FORESTS
197 ISLANDS
80 OF EASTERN AFRICA
CERRADO SUNDALAND 347
231
93 164
MADAGASCAR WALLACEA
AND THE INDIAN 172 QUEENSLAND
TROPICAL ANDES OCEAN ISLANDS WET TROPICS
73 138 369

ATLANTIC NEW CALEDONIA

FOREST 193
84 MAPUTALAND-
CHILEAN WINTER SUCCULENT
PONDOLAND-ALBANY SOUTHWEST
RAINFALL - KAROO
219 AUSTRALIA
VALDIVIAN FORESTS 134 CAPE NEW ZEALAND
99 FLORISTIC 183 187
REGION
130
 INTRODUCTION
Global Priority Setting for Biodiversity Conservation
Life on Earth faces a crisis of historical and planetary
proportions. Unsustainable consumption in many north-
ern countries and crushing poverty in the tropics are
destroying wild nature. Expanding agriculture, indus-
try, and urbanization are fragmenting and eliminating
natural environments; accidental and deliberate intro-
duction of exotic species is wreaking havoc on native
communities; pollution is altering complex biogeo-
chemical and climate cycles through the land, air, and
water; and hunting, trade, and overfishing are decimat-
ing the last populations of large vertebrate species (Vi-
tousek et al. 1997). Biodiversity is besieged.
Extinction is the gravest aspect of the biodiversity
crisis: it is irreversible. While extinction is a natural
process that is part of the history of this planet, the fos-
sil record indicates that, in the absence of humans, the
life span of a species averages one million years (May
et al. 1995). Now, however, human impacts have ele-
vated the rate of species extinction by at least a thou-
sand, possibly several thousand times the natural back-
ground rate (Pimm et al. 1995). Mass extinctions of this
magnitude have only occurred five times before in the
history of our planet; the last, probably caused by a cat-
aclysmic asteroid impact 65 million years ago, brought
the end of the dinosaur age (Álvarez et al. 1980).
It is easy to imagine the disasters that humanity
would face if the rate of other natural processes, such
as the frequency of floods or disease transmission rates,
increased a thousand-fold. The world as we know it
would be devastated. But what exactly do we lose with
the catastrophic extinction of other species? Foreclos-
ing future resource-use options is perhaps the most ob-
vious consequence. Scientists have recognized only a
fraction (maybe less than 10%; perhaps even as little as
1%) of the species with which we share our planet, and
know the biology of even fewer (Novotny et al. 2002).
Thus, with species extinction we destroy a vast genetic
storehouse (Myers 1983) that could one day be found to
hold, for instance, a cure for AIDS. Current develop-
ments in the treatment of malaria, one of the world’s
biggest killers, use artemisinin-based compounds de-
rived from the plant Artemisia annua (Sachs 2002). The
biodiversity crisis could be compared with burning
down the world’s libraries without knowing the titles of
90% of the books or the content of most of the pages
of the known books. Less tangibly, but no less impor-
tantly, species extinction inflicts a deep cultural, spiri-
tual, and moral wound on humanity. All of the world’s
societies value species for their own sake, over and
above any utilitarian purpose, and wildlife —especially
the larger vertebrates and many plants— are an integral
part of the fabric of all human cultures (Wilson 1984;
Wilson and Kellert 1993).
The road to extinction is also perilous to people. For
example, the destruction of montane forests causes fre-
quent, massive landslides with dreadful human cost
—witness the tragic mudslides that killed 10 000 people
following Hurricane Mitch in Honduras in 1998 (Hellin
et al. 1999)— while the Severe Acute Respiratory Syn-
drome (SARS) outbreak in East Asia has been directly
linked to the trade in wildlife for human consumption
(Guan et al. 2003). Other consequences of biodiversity
loss are more subtle and cumulative, but equally signif-
icant, such as the progressive deterioration of the nat-
ural basis for sustainable economic growth. A number
of high-profile studies (e.g., Costanza et al. 1997) have
placed the annual economic value of ecosystem ser-
vices such as climate and water regulation, pollination,
and recreation in the tens of trillions of dollars —costs
that society would have to bear if these generally free
but unaccounted services were no longer accessible. A
significant fraction of such services can be attributed di-
rectly to biodiversity. Balmford et al. (2002) concluded
that conversion of natural ecosystems to anthropogenic
landscapes roughly halves their economic value.
In order to stem the extinction crisis most effective-
ly, we must prioritize where we should act first. To a
large extent conservation is, and always will be, local.
People care most about what is happening in their own
backyards. We believe that all biodiversity is important
and that all nations and communities, large or small,
should do everything possible to conserve the biologi-
cal riches on which they depend. However, some local
efforts have planetary consequences and thus justify
priority allocation of scarce financial resources.
The establishment of priorities for biodiversity con-
servation is a complex issue (Margules and Pressey
2000). The problem can best be framed by a question:
In which areas would a given conservation dollar con-
tribute the most towards slowing the current rate of ex-
tinction of global biodiversity? Species are distributed
unevenly around the world (Gaston 2000), which
means that mapping this variation is essential if we are
to address the question. However, we can not simply
measure the numbers of species living in particular ar-
eas. This is because several species-rich areas might
hold a large fraction of the same species, meaning that
the overall number that could be conserved within
such areas might be rather small (Pressey and Nicholls
1989). Instead, we must measure not species richness
but endemism: the degree to which species are only
found in a given place. This can be thought of as a mea-
sure of irreplaceability —in essence, the number of geo-
graphic options one has for the conservation of the
On p. 1, in the State of Coahuila,
Mexico, there is a conservation
model that has been carried out by
ranchers in the Serranías del Burro
for forty years. It has facilitated the
recovery of biodiversity in these lands.
© Patricio Robles Gil/Sierra Madre
On p. 2, Eungella National Park is one
of Queensland’s most ecologically
diverse parks with 860 plant species,
including the Alexandra palm
(Archontophoenix alexandrae).
This mountain refuge lies close to
the boundary between subtropical
and tropical rainforests.
© Günter Ziesler
On pp. 4-5, the Sumatran orang-
utan (Pongo abelii), here seen in
Gunung Leuser National Park, is
now considered a full species, distinct
from the Bornean orang-utan. It is one
of the most threatened great apes.
© Anup Shah/naturepl.com
On p. 6, laurisilva forest in the
Frontera Rural Park (El Hierro Island,
Canary Islands), comprised mainly
of flowering laurel (Laurus azorica),
Indian bay (Persea indica),
Canarian laurel (Ilex canariensis),
tilo tree (Ocotea foetens), faya or
firetree (Myrica faya), and delfino
(Myrsine canariensis).
© Francisco Márquez
On pp. 8-9, the Maderas del
Carmen Protected Area for Flora
and Fauna in the El Carmen
Mountains in the State of Coahuila
(Mexico), part of CEMEX’s
comprehensive program in this region.
© Patricio Robles Gil/Sierra Madre
On the opposite page,
the spectacular and Endangered
golden lion tamarin
(Leontopithecus rosalia) is found
only in forest fragments in the
lowlands of the State of Rio de
Janeiro, and is perhaps Brazil’s
best-known monkey species. Along
with Brachyteles, it represents one
of the two primate genera endemic
to the hotspot.
© Frans Lanting/Minden Pictures
19

On p. 10, the Mentawai Archipelago
is made up of four islands off the
coast of west Sumatra. The largest
and most northerly of the islands is
Siberut, which is inhabited by some
23 000 Mentawai people.
© Alain Compost
On p. 12, the magnificent Tamul
Waterfall (110 m) in the Huasteca
region of the State of San Luis Potosí
(Mexico) is in the Sierra Madre.
© Patricio Robles Gil/Sierra Madre
On p. 14, the leopard (Panthera
pardus) poses an interesting
question to biogeographers. This big
cat occurs on Java and also the
Malay Peninsula, but is
surprisingly absent in between, not
being known from either fossil or
recent records from Sumatra
or Borneo.
© Alain Compost
On pp. 20-21, the leaf-tailed gecko
(Uroplatus fimbriatus) is the
largest member of its genus, one of
about 50 reptile genera endemic to
Madagascar.
© Piotr Naskrecki
22
species found in a given area (Pressey et al. 1994).
Since we can not conserve a species that is endemic to
a given area anywhere except in that area, the area is
wholly irreplaceable at a global scale.
A further problem concerns which species we should
evaluate. We know that we can not map all species be-
cause we have not even named most of them. Quite
fortuitously, vascular plants and vertebrate animals
—the species we know best— tend to be large, and play
prominent roles in structuring ecosystems (Terborgh
1988), although species that we know less about are
also vital for ecosystem processes (Wilson 1987).
Whether or not the distributions of plants and verte-
brates are mirrored by the myriad of unknown terres-
trial invertebrate species remains an open question, al-
though some evidence suggests that they may be
(Howard et al. 1998). Some taxa, such as tiger beetles,
seem to exhibit excellent congruence with many other
groups (Pearson and Cassola 1992; Carroll and Pearson
1998), while others show less clear patterns (Van
Jaarsveld et al. 1998). These comparisons are also heav-
ily dependent on scale (Reid 1998). At regional scales
we often see much greater congruence than at fine
scales (Pearson and Carroll 1999). Thanks to recent ad-
vances in bioinformatics, we will soon be able to use
massive datasets on invertebrate species distributions
to delimit the boundaries of biologically unique areas
more precisely than ever (Meier and Dikow 2004). One
of the largest gaps in our current knowledge remains in
the aquatic realm, which is in critical need of effective
conservation action. Distributions of marine and fresh-
water species remain largely unknown, although ongo-
ing projects are addressing this issue.
Our ultimate goal is to keep nature intact, which
means that we must stop anthropogenic species ex-
tinctions. To approach this goal, we must slow the rate
of species extinction as much as possible (over and
above simply conserving as many species as we can)
with whatever conservation resources we have at our
disposal, which requires incorporating threats (or vul-
nerability) and costs into priority setting. Like species,
threats are hard to measure. The extent of habitat
destruction is one useful metric, given the well-
documented relationship between the size of an area
of habitat and the number of species it retains (Brooks
et al. 1997). Other measures, such as human popula-
tion density (Balmford et al. 2001), are also used.
Threats and costs are generally related to each other;
the more threatened an area is, the more it will cost to
conserve (Ando et al. 1997). However, the relationship
is not always linear, as it depends in large part on the
economic conditions of the country and immediate lo-
cale in which the priority area is located (Balmford et al.
2003). Some extremely threatened areas can still be
conserved at low cost (e.g., much of Madagascar and
some areas in Southeast Asia), often by addressing un-
derlying poverty simultaneously with biodiversity con-
servation, while others tend to be quite expensive (e.g.,
the California Floristic Province and New Caledonia).
However, because economic opportunity costs vary
dramatically across the landscapes of hotspots and
wilderness areas, there still exist areas of relatively low
cost in all hotspots, offering great conservation oppor-
tunities, as well as areas of high cost in wilderness areas
requiring immediate attention to threats (Chomitz et al.
2004).
We still face a paradox in determining how to incor-
porate threats, costs, and opportunities into conser-
vation priorities. Intuitively, we want to conserve
the most threatened areas first, to avoid losing them the
soonest. But we also want to get the greatest return for
our conservation dollar, which in theory would mean
targeting the areas of lowest cost, greatest opportunity,
and least threat first. This paradox can be resolved by
the measurement of irreplaceability —or the degree of
endemism (Mittermeier et al. 2003a). Thus, we identi-
fy those areas that hold species found nowhere else
and that are guaranteed to lose species if the areas are
not conserved. Among these, we rank our actions based
on threats, with the most threatened biodiversity re-
ceiving the most urgent action. Wherever we have
choices, or equal levels of endemism, we should select
opportunities for attending to areas that are the least
expensive to conserve (and often the least threatened).
In effect, we need a dual conservation strategy that al-
ways prioritizes endemic-rich areas and ensures that
we protect the most threatened places with species
that we will otherwise lose, while preemptively pro-
tecting equally unique places that are not yet under ex-
treme threat.
History of the Hotspots Concept
A seminal paper by Norman Myers (1988) first articu-
lated the principles of irreplaceability and threat to in-
form terrestrial conservation priorities on a global
scale. Myers identified ten tropical forest “hotspots”
characterized both by exceptional levels of plant en-
demism and by uncommon rates of habitat loss, al-
though without quantitative criteria as to what exactly
constituted a hotspot. Subsequently, Myers (1990) added
a further eight hotspots, including four Mediterranean-
type ecosystems. Conservation International adopted
Myers’ hotspots as its institutional blueprint in 1989,
making minor modifications and additions over the
next seven years. In 1996, Conservation International
made the decision to undertake a reassessment of the
hotspots concept, including an examination of whether
key areas had been overlooked. This was done in col-
laboration with Myers and took three years. A prelimi-
nary report (Mittermeier et al. 1998) was followed by
an extensive global review (Mittermeier et al. 1999), a
scientific analysis (Myers et al. 2000), and a detailed on-
line publication (www.biodiversityhotspots.org). These
efforts introduced quantitative thresholds for the desig-
nation of hotspots. To qualify as a hotspot, a region had
to meet two strict criteria: it had to contain at least
1 500 species of vascular plants (>0.5% of the world’s
total) as endemics, and it had to have 30% or less of its
original vegetation (extent of historical habitat cover)
remaining. This analysis identified 25 hotspots, collec-
tively holding as endemics no less than 44% of the
world’s plants and 35% of terrestrial vertebrates (mam-
mals, birds, reptiles and amphibians) in an area that
formerly covered only 11.8% of the Earth’s land sur-
face. However, the fulcrum around which these star-
tling results were presented was that this land area had
been reduced by 87.8% of its original extent, such that
this amazing wealth of biodiversity was restricted to
only 1.4% of Earth’s land surface.
Concurrent with the development of the hotspots
strategy was the recognition of the advantages to in-
vesting in the least threatened —and cheapest— highly
biodiverse areas. In fact, Myers (1988) was the first to
notice that three endemic-rich regions of tropical forest
remained largely intact —he called these “good news
areas.” Similarly, Mittermeier (1988) called attention to
several high biodiversity tropical rainforest regions that
were still in relatively intact condition. He later broad-
ened the concept to address three regions —Amazonia,
the Congo Forests of Central Africa, and the island of
New Guinea— and referred to them as “major tropical
wilderness areas” (McNeely et al. 1990; Mittermeier et al.
1998). While threatened to a much lesser extent than
the hotspots, these areas are nevertheless under grow-
ing pressure from human activities.
Recently, the emphasis on biodiversity-rich wilder-
ness has been reassessed against the background of all
of Earth’s wilderness areas, quantitatively defined as re-
taining at least 70% of their original habitat and holding
human population densities of less than five people per
square kilometer (Mittermeier et al. 2002, 2003b). This
analysis found that while 44% of Earth’s land area can
still be considered wilderness, only five of these regions
(covering just 6.1% of that area) are “high biodiversity
wilderness areas” with more than 1 500 plant species as
endemics. These are Amazonia, the Congo Forests of
Central Africa, the island of New Guinea, the North
American Deserts of the Southwestern United States
and Northern Mexico, and the Miombo-Mopane Wood-
lands and Savannas of Southern Africa. Together, they
hold 17% of the planet’s plants and 8% of terrestrial
vertebrates as endemics.
Based on these analyses, Conservation International
uses a two-pronged strategy for global conservation pri-
oritization, simultaneously focusing on the threatened
and irreplaceable hotspots and on the high biodiversity
wilderness areas, which are irreplaceable but still large-
ly intact, and as such represent important conservation
opportunities. The decision, at any given point in time,
as to whether we should allocate particular resources to
a hotspot or to a high biodiversity wilderness area de-
pends on numerous factors, including donor interest,
immediate political, economic or social opportunity or
need, and other conservation benefits (e.g., protection
of major ecosystem services). However, all of these ar-
eas —the hotspots and the high biodiversity wilderness
areas combined— are on Conservation International’s
priority list, and the organization’s programs employ a
strategic mix of both.
A consistent concern for conservation practitioners
is that data for aquatic species have yet to be synthe-
sized at a global scale across many taxa and aquatic
habitats. It was not until 2002 that the first compre-
hensive global assessment of conservation priorities
for an aquatic system —coral reefs— was published
(Roberts et al. 2002). This analysis identified 18 centers
of endemism (across four assemblages —1 700 coral
reef fish, all 804 scleractinian coral species, three mol-
lusk families, and 69 lobster species) and highlighted
ten of these regions as hotspots facing high threats. Re-
markably, the study found that eight of the ten reef
hotspots (and 14 of the 18 centers of endemism) lie ad-
jacent to terrestrial hotspots, raising an intriguing pos-
sibility that terrestrial hotspots may actually reflect
aquatic ones rather well. The publication of Roberts et al.
(2002) has attracted much-needed attention to marine
hotspots, although data on these areas remains sparse
compared with information on terrestrial systems
(Lambshead 1993). Our lack of knowledge about fresh-
water systems is even more pronounced —where even a
first look at global conservation prioritization has yet to
be carried out. These areas constitute one of the world’s
most endemic-rich and threatened biomes (McAllister et
al. 1997), making such an analysis most urgent.
Impact of Hotspots
The impact of the hotspots concept has been astound-
ing. One measure of this is scientific. Searching the
Web of Science for all citations including the word
“hotspot” in the title yields numerous scientific papers.
While many of these concern geology, astronomy, or
genetics (and a few, behavioral ecology or remote sens-
ing), nearly 100 citations use the word “hotspot” to re-
fer to biodiversity conservation priorities. Analyzing
these citations over time reveals a clear pattern of grad-
ual increase following Myers’ (1988, 1990) original
work, a rapid acceleration with the publication of My-
ers et al. (2000), and an increasing number of publica-
tions on marine hotspots following Roberts et al.
(2002). In addition, the number of times that Myers et
al. (2000) has been cited in the peer-reviewed scientif-
ic literature has shown a steady increase since its pub-
lication, and by January 2004 totaled 438 instances.
More importantly, the impact of the hotspots con-
cept in terms of investment in conservation has been
dramatic (Myers and Mittermeier 2003). As indicated
above, Conservation International adopted hotspots as
its central strategy in 1989. In the same year, the Chicago-
based John D. and Catherine T. MacArthur Foundation
adopted hotspots as its primary global investment strat-
egy (Mittermeier et al. 1998). In 2000, the World Bank and
the Global Environment Facility joined Conservation
23

On pp. 24-25, the alternative name
of the Hamadryas baboon
(Papio hamadryas) is the sacred
baboon, which derives from the fact
that this primate was held sacred in
ancient Egypt and was even
mummified. Although this species
now lives on hillsides and
escarpments bordering the southern
part of the Red Sea and the Gulf of
Aden, its distribution during the
earliest days of Egyptian
civilization may have extended to
the Egyptian border area.
© Patricio Robles Gil/Sierra Madre
On the opposite page, the cabbage
tree (Cordyline australis) is one of
the most distinctive trees in the New
Zealand landscape. In 1987,
a mystery disease started to kill off
cabbage trees on North Island. The
disease is called “sudden decline,”
and the cause is parasitic bacteria
called phytoplasmas. But ti kouka
(one of the Maori names for the
cabbage tree) in natural forest
patches continues to do well.
© Konrad Wothe/Minden Pictures
26
International to establish the Critical Ecosystem Part-
nership Fund, a conservation finance mechanism focused
explicitly on the hotspots (Dalton 2000; www.cepf.net);
the MacArthur Foundation and the Japanese Govern-
ment have since joined the partnership, bringing the
total amount available to $125 million. Conservation In-
ternational’s $100-million Global Conservation Fund,
supported by the Gordon and Betty Moore Foundation,
also uses hotspots (along with high biodiversity wilder-
ness areas) to guide its investments. More than $750
million has been devoted to saving hotspots over the
last 15 years, perhaps the largest financial investment
in any single conservation strategy (Myers 2003). The
hotspots concept has also entered the mainstream as a
tool for forward-thinking private sector businesses
that have adopted biodiversity conservation policies
for their operations and supply chain systems. For ex-
ample, Office Depot explicitly gives preference to
pulp and paper vendors that protect natural forests in
the biodiversity hotspots and high biodiversity wilder-
ness areas.
Biodiversity conservation efforts in hotspots often re-
quire the ability to withstand and adapt to a rapidly
changing socio-political climate. While it can be tempt-
ing to write off high-risk areas, experience demon-
strates both the importance and the potential for oper-
ating and maintaining a conservation presence in
hotspots that are undergoing political difficulties. Mada-
gascar, one of the most important hotspots, was almost
abandoned by conservationists in the early to mid-
1980s, and again during a brief period of political strife
in 2001 and 2002. Fortunately, several organizations
persevered, notably the World Wildlife Fund, USAID,
and the World Bank and, beginning in the early 1990s,
Conservation International and the Wildlife Conserva-
tion Society. This resolve paved the road for a positive
environment for the new President, Marc Ravalo-
manana, to give conservation a high priority in his gov-
ernment’s development plans. At the Fifth World Parks
Congress in September 2003, President Ravalomanana
committed to tripling the country’s protected area net-
work over the next five years, and just five months af-
ter this pledge he announced the establishment of 14
new areas increasing coverage by 65%. At the time of
his announcement, President Ravalomanana also re-
quested the involvement of the international commu-
nity in creating a $50-million trust fund for conserva-
tion over the next five years; seven months later, a total
of $24 million in commitments has already been made.
Liberia, one of the most important countries in the
heavily impacted Guinean Forests of West Africa Hot-
spot, has, until very recently, been written off by most
of the international conservation community. Nonethe-
less, Fauna and Flora International and Conservation
International operated there through some of the worst
periods of instability and violence. Largely because of
their efforts, the Liberian Senate in 2003 enacted legis-
lation expanding the country’s protected areas net-
work, and the stage is now set to make forest policy re-
form and conservation a priority for Liberia’s recon-
struction. Such cases provide excellent illustrations
of the conservation return on investment produced by
the hotspots strategy.
An Updated Hotspots Analysis
The hotspots analysis is in constant evolution. There
are two major ways in which hotspots can change over
time. The first is a real effect. Threats and their impacts
change, meaning that some places may become more
threatened, while others, if conservation efforts are
successful, may eventually recover. The second is that
our knowledge of biodiversity, threats, and costs is con-
tinually improving; new species are discovered, new
populations are found, and higher-resolution land cov-
er data is collected. Over the last few years, in concert
with the information revolution and the emergence of
the Internet, this data has become better compiled
(Sugden and Pennisi 2000). Now, several years after the
publication of the reassessment of the hotspots strategy
(Mittermeier et al. 1999; Myers et al. 2000), it is time to
revisit the hotspots themselves in light of new data re-
garding species distributions and changing conditions
of the planet’s ecosystems.
We should emphasize from the outset that the cur-
rent effort is not a reworking of the entire hotspots
concept. Rather, the aims of this analysis are to revis-
it the status of the existing hotspots, refine their bound-
aries, update the information associated with them
and, most importantly, consider a number of potential
hotspots that may qualify as additions to the existing
list of 25. Consequently, the criteria for what qualifies
as a hotspot remain unchanged. There continues to be
much debate in the literature concerning total vascu-
lar plant diversity, with lower estimates ranging from
270 000 to 320 000 (May 1992; Prance et al. 2000), and
higher estimates ranging up to 422 000 (Govaerts 2001;
Bramwell 2002). For now, we have retained the lower
figure of 300 000 used by Myers et al. (2000), given
that the higher figure remains controversial (Thorne
2002).
When Myers et al. (2000) published the results of
their analysis, they noted that a number of areas har-
bored exceptional plant endemism and were also under
unusual threat, but were insufficiently documented to
meet the hotspots criteria: the Ethiopian Highlands,
the Angola Escarpment, southeastern China, Taiwan,
and the forests of the Albertine Rift. The Queensland
Wet Tropics in northeastern Australia were also men-
tioned as having a remarkably high species-to-area ra-
tio, but insufficient endemic plant species to qualify as
a hotspot. Additional data on Taiwan and the Queens-
land Wet Tropics now confirm that neither reaches the
threshold of being a hotspot. However, because both
are globally important and come so close to meeting
the criteria, we include special treatment of them in
this book (pp. 361 and 369, respectively). Furthermore,
investigation of the definition of a hotspot for the rain-
forests of eastern Australia continues, and it is likely
that an expanded interpretation of this region will re-
sult in confirmation of its hotspot status in the future.
New data also demonstrates the hotspot status of the
Ethiopian Highlands and Albertine Rift (discussed be-
low). Finally, the Angola Escarpment and southeastern
China remain little known, and thus it is still not possi-
ble to ascertain whether or not the regions qualify as
hotspots.
One major finding of this updated analysis is that
six previously overlooked areas qualify for hotspot
status. These are the Madrean Pine-Oak Woodlands of
northern Mexico and the southwestern United States
(p. 205), southern Africa’s Maputaland-Pondoland-Al-
bany region (p. 219), the Horn of Africa (p. 277), the
Irano-Anatolian region (p. 287), the Mountains of Cen-
tral Asia (p. 297), and Japan (p. 333). In addition, one
existing hotspot is divided into two. Mittermeier et al.
(1998) first suggested the Himalaya (p. 309) and Indo-
Burma (p. 323) regions as separate hotspots. In Mit-
termeier et al. (1999) and Myers et al. (2000) these
were combined, but data is now sufficient to show that
they contain quite distinctive biotas. That a number of
these changes are in Asia is explained partly by the
fact that biodiversity data for the continent has histor-
ically been less thoroughly synthesized than has data
for the Americas and Africa, and partly because much
of the data that does exist for key countries such as
China and Japan has, at least until recently, been in-
accessible to conservation scientists outside of these
regions.
An important modification to the hotspots strategy
presented here is the reconfiguration of several African
hotspots. One problem we have always grappled with is
that the combination of the East African Coastal Forests
(Burgess and Clarke 2000) with Tanzania’s Eastern Arc
Mountains (Lovett and Wasser 1993) as a single hotspot
is somewhat incongruous biogeographically (Myers et al.
1999). Furthermore, recognition by Myers et al. (2000) of
the potentially high levels of endemism in the Ethiopian
Highlands and Albertine Rift meant these regions need-
ed to be re-evaluated with better data. It is now apparent
that none of the montane areas —the Ethiopian High-
lands, Albertine Rift, or Eastern Arc Mountains— quali-
fy as hotspots on their own, because they do not meet
the cut-off of 1 500 endemic plants. However, the classic
work on African biogeography by Frank White (1983)
provides a simple solution. The biogeographic affinities
of these regions suggest that they are best considered as
a single unit, the Afromontane Region, despite their frag-
mented geography. Thus, we identify this region as the
Eastern Afromontane Hotspot, encompassing the East-
ern Arc Mountains and Southern Rift (p. 245), the Al-
bertine Rift (p. 255), the Ethiopian Highlands (p. 262),
and a few outliers. This leaves the Coastal Forests of
Eastern Africa, running from southern Somalia south
through Kenya, Tanzania, and Mozambique, as a unique
hotspot in its own right (p. 231).
The final change revealed in our reassessment of
the hotspots is truly terrifying. Less than a decade ago,
the islands of eastern Melanesia —the Bismarcks,
Solomons, and Vanuatu— while known to be extreme-
ly endemic-rich, still held largely intact habitat. Since
then, rampant logging and establishment of oil palm
plantations have devastated these islands, leaving only
30% of their forests remaining, a situation mirroring
the fate of Indonesia’s forests a decade ago (Holmes
2000). Thus, the primary cause of the identification of
the East Melanesian Islands Hotspot (p. 347) is a worse-
ning threat over a very short period of time.
In revisiting the boundaries of the hotspots, we have
tried to achieve a balance between what is scientifical-
ly defensible, and what is pragmatically acceptable.
The hotspots are based on plant endemism, and so,
as far as possible, our decision regarding where or
whether to include a particular area or island within a
hotspot is determined by the floristic affinities of the re-
gion in question. As before, the landmark publication
Centers of Plant Diversity (Davis et al. 1994-1997) has
been instrumental in guiding and influencing some of
our decisions in this regard. However, in some cases,
we have seen fit to deviate from this ideal, in order to
accommodate tropical islands —many of which have
very high proportions of threatened species— that
might otherwise slip through the net of conservation
priorities. For this reason, we have grouped certain is-
lands with their closest-lying hotspots including, for ex-
ample, Galápagos and Malpelo with Tumbes-Chocó-
Magdalena, Juan Fernández with the Chilean Winter
Rainfall-Valdivian Forests, the Azores and Cape Verde
Islands (both part of the Macaronesian Islands along
with the Canaries and Madeira) with the Mediter-
ranean Basin; and Lord Howe and Norfolk islands with
New Zealand. This is done solely for purposes of prag-
matic convenience, and with full recognition that the
floristic affiliations of these islands with their associat-
ed landmasses are tenuous at best. On the opposite page,
the king protea (Protea cynaroides)
is the national flower of South
Africa. Widely distributed in the
Synthesis of the Updated Hotspots Data
region of the Western Cape, the king
protea is now rarely found in the
In total, this updated analysis reveals the existence of wild —most of them are cultivated.
34 biodiversity hotspots, each holding at least 1 500 © Haroldo Castro
endemic plant species, and having lost at least 70%
of its original habitat extent. Overall, the 34 hotspots On pp. 30-31, the ornate hawk-eagle
(Spizaetus ornatus), in Calakmul
once covered a land area of 23 490 101 km2, 15.7% of
Biosphere Reserve, South Yucatán,
the Earth’s land surface, an area equivalent in size to is a magnificent raptor that is
Russia and Australia combined. Their individual areas declining, but fortunately still
spanned two orders of magnitude. Three of the regions relatively common and widespread.
historically covered more than two million square kilo- Here, the hawk-eagle is taking
meters each (Indo-Burma, the Mediterranean Basin, recently cut leaves to its nest to cool
its chicks.
and the Cerrado), and a further six, more than a million.
© Patricio Robles Gil/Sierra Madre
The smallest, New Caledonia, covered only 18 972 km2,
and three others were smaller than 100 000 km2. The
average original size was 690 885 km2, and the median
size 385 316 km2. This extent of habitat has now been
reduced to 3 379 246 km2, a mere 2.3% of the planet’s
29
 TABLE 1. Original extent, remaining habitat, and percentage of remaining habitat for each hotspot TABLE 2. Numbers of plant and vertebrate species
(determined using an equal-area projection) and its predominant biome type (Olson et al. 2001) endemic to (E) and occurring in (O) each of the 34 hotspots (percentages in parentheses)
Percentage Plants Mammals Birds Reptiles Amphibians Freshwater fishes
Original extent Remaining remaining Hotspot E O E O E O E O E O E O
Hotspot (km2) habitat (km2) habitat Predominant biome type Tropical Andes 15 000 30 000 75 569 584 1 728 275 610 664 1 155 131 380
Tropical Andes 1 542 644 385 661 25 Tropical and Subtropical Moist Broadleaf Forests; Montane Grasslands (50) (13) (34) (45) (57) (35)
and Shrublands Tumbes-Chocó-Magdalena 2 750 11 000 10 283 112 892 98 325 29 204 115 251
Tumbes-Chocó-Magdalena 274 597 65 903 24 Tropical and Subtropical Moist Broadleaf Forests (25) (4) (13) (30) (14) (46)
Atlantic Forest 1 233 875 99 944 8 Tropical and Subtropical Moist Broadleaf Forests Atlantic Forest 8 000 20 000 71 263 148 936 94 306 286 475 133 350
Cerrado 2 031 990 432 814 22 Tropical and Subtropical Grasslands, Savannas, and Shrublands (40) (27) (16) (31) (60) (38)
Chilean Winter Rainfall-Valdivian Forests 397 142 119 143 30 Mediterranean Forests, Woodlands, and Shrubs; Temperate Broadleaf Cerrado 4 400 10 000 14 195 16 605 33 225 26 251 200 800
and Mixed Forests (44) (7) (3) (15) (10) (25)
Mesoamerica 1 130 019 226 004 20 Tropical and Subtropical Moist Broadleaf Forests Chilean Winter Rainfall-Valdivian Forests 1 957 3 892 14 65 12 226 27 41 29 43 24 43
Madrean Pine-Oak Woodlands 461 265 92 253 20 Tropical and Subtropical Coniferous Forests (50) (22) (5) (66) (67) (56)
Caribbean Islands 229 549 22 955 10 Tropical and Subtropical Dry Broadleaf Forests Mesoamerica 2 941 17 000 66 440 213 1 124 240 686 353 575 340 509
California Floristic Province 293 804 73 451 25 Mediterranean Forests, Woodlands, and Shrubs; Temperate Coniferous Forests (17) (15) (19) (35) (61) (67)
Madrean Pine-Oak Woodlands 3 975 5 300 6 328 23 525 37 384 50 218 18 84
Guinean Forests of West Africa 620 314 93 047 15 Tropical and Subtropical Moist Broadleaf Forests
(75) (2) (4) (10) (23) (21)
Cape Floristic Region 78 555 15 711 20 Mediterranean Forests, Woodlands, and Shrubs
Caribbean Islands 6 550 13 000 41 89 167 607 468 499 164 165 65 161
Succulent Karoo 102 691 29 780 29 Deserts and Xeric Shrublands
(50) (46) (28) (94) (99) (40)
Maputaland-Pondoland-Albany 274 136 67 163 25 Tropical and Subtropical Moist Broadleaf Forests; Montane Grasslands
California Floristic Province 2 124 3 488 18 151 8 341 4 69 25 54 15 73
and Shrublands
(61) (12) (2) (6) (46) (21)
Coastal Forests of Eastern Africa 291 250 29 125 10 Tropical and Subtropical Moist Broadleaf Forests
Guinean Forests of West Africa 1 800 9 000 67 320 75 793 52 206 83 246 143 512
Eastern Afromontane 1 017 806 106 870 11 Tropical and Subtropical Moist Broadleaf Forests; Montane Grasslands (20) (21) (9) (25) (34) (28)
and Shrublands Cape Floristic Region 6 210 9 000 4 90 6 324 22 100 16 51 14 34
Horn of Africa 1 659 363 82 968 5 Tropical and Subtropical Grasslands, Savannas, and Shrublands (69) (4) (2) (22) (31) (41)
Madagascar and the Indian Ocean Islands 600 461 60 046 10 Tropical and Subtropical Moist Broadleaf Forests Succulent Karoo 2 439 6 356 2 74 1 227 15 94 1 29 0 28
Mediterranean Basin 2 085 292 98 009 5 Mediterranean Forests, Woodlands, and Shrubs (38) (3) (0) (16) (3) (0)
Caucasus 532 658 143 818 27 Temperate Broadleaf and Mixed Forests Maputaland-Pondoland-Albany 1 900 8 100 5 193 0 541 36 205 12 80 20 73
Irano-Anatolian 899 773 134 966 15 Temperate Broadleaf and Mixed Forests (23) (3) (0) (18) (15) (27)
Mountains of Central Asia 863 362 172 672 20 Temperate Grasslands, Savannas, and Shrublands; Montane Grasslands Coastal Forests of Eastern Africa 1 750 4 000 11 198 12 636 54 250 8 102 32 219
and Shrublands (44) (5) (2) (22) (8) (15)
Western Ghats and Sri Lanka 189 611 43 611 23 Tropical and Subtropical Moist Broadleaf Forests Eastern Afromontane 2 356 7 598 104 490 110 1 325 93 347 79 285 617 893
Himalaya 741 706 185 427 25 Tropical and Subtropical Coniferous Forests; Montane Grasslands and Shrublands (31) (21) (8) (27) (28) (69)
Mountains of Southwest China 262 446 20 996 8 Temperate Coniferous Forests Horn of Africa 2 750 5 000 20 219 25 704 93 284 7 53 10 100
Indo-Burma 2 373 057 118 653 5 Tropical and Subtropical Moist Broadleaf Forests (55) (9) (4) (33) (13) (10)
Sundaland 1 501 063 100 571 7 Tropical and Subtropical Moist Broadleaf Forests Madagascar and the Indian Ocean Islands 11 600 13 000 144 155 183 313 367 381 226 228 97 164
Wallacea 338 494 50 774 15 Tropical and Subtropical Moist Broadleaf Forests (89) (93) (58) (96) (99) (59)
Philippines 297 179 20 803 7 Tropical and Subtropical Moist Broadleaf Forests Mediterranean Basin 11 700 22 500 25 224 32 497 77 228 27 86 63 216
Japan 373 490 74 698 20 Temperate Broadleaf and Mixed Forests (52) (11) (6) (34) (31) (29)
Southwest Australia 356 717 107 015 30 Mediterranean Forests, Woodlands, and Shrubs Caucasus 1 600 6 400 18 130 2 381 20 87 4 17 12 127
East Melanesian Islands 99 384 29 815 30 Tropical and Subtropical Moist Broadleaf Forests (25) (14) (1) (23) (24) (9)
New Zealand 270 197 59 443 22 Temperate Broadleaf and Mixed Forests Irano-Anatolian 2 500 6 000 10 141 0 364 13 116 4 21 30 90
New Caledonia 18 972 5 122 5 Tropical and Subtropical Moist Broadleaf Forests (42) (7) (0) (11) (19) (33)
Polynesia-Micronesia 47 239 10 015 21 Tropical and Subtropical Moist Broadleaf Forests Mountains of Central Asia 1 500 5 500 6 143 0 493 1 59 4 9 5 27
(27) (4) (0) (2) (44) (19)
Total 23 490 101 3 379 246 Western Ghats and Sri Lanka 3 049 5 916 18 140 35 457 176 265 138 179 139 191
(52) (13) (8) (66) (77) (73)
Himalaya 3 160 10 000 12 300 15 979 49 177 41 124 33 269
land surface. Its size is slightly more than the country the most endemics are the Tropical Andes and Sunda- (32) (4) (2) (28) (33) (12)
of India or a fraction less than the five largest Ameri- land, with no less than 15 000 endemic plant species Mountains of Southwest China 3 500 12 000 5 237 1 611 15 94 40 98 23 92
can states combined (Alaska, Texas, California, Mon- (Table 2). Two other hotspots —the Mediterranean (29) (2) (0) (16) (41) (25)
Indo-Burma 7 000 13 500 73 433 73 1 277 204 518 139 311 553 1 262
tana, and New Mexico = 3 392 950 km2). In all, 86% of Basin, and Madagascar and the Indian Ocean Islands— (52) (17) (6) (39) (45) (44)
the hotspots’ habitat has already been lost, and only also exceed 10 000 endemic plant species; and five Sundaland 15 000 25 000 173 381 146 771 244 449 172 242 350 950
14% remains. Table 1 details these statistics hotspot-by- more exceed 5 000. The four hotspots richest in en- (60) (45) (19) (54) (71) (37)
hotspot. demic plants are also the most speciose, with 20 000 or Wallacea 1 500 10 000 127 222 265 650 99 222 32 58 50 250
(15) (57) (41) (45) (55) (20)
The distribution of the hotspots among biomes is more plant species occurring in each. Plant numbers Philippines 6 091 9 253 102 167 185 535 160 235 74 99 67 281
greatly skewed towards tropical forests (Table 1). Of the per hotspot are derived from specialist estimates rather (66) (61) (35) (68) (75) (24)
34 hotspots, 22 (65%) are predominantly tropical forest than from species-by-species lists, which makes it im- Japan 1 950 5 600 46 91 15 368 28 64 44 58 52 214
biomes, ranging from very wet hotspots (like the East possible to calculate the total number of endemics or (35) (51) (4) (44) (76) (24)
Southwest Australia 2 948 5 571 12 57 10 285 27 177 19 33 10 20
Melanesian Islands) to sparsely wooded savanna and even the number of species occurring in hotspots (53) (21) (4) (15) (58) (50)
grassland (as in the Cerrado). Six hotspots (18%) pri- (Table 3). This is because these estimates do not ac- East Melanesian Islands 3 000 8 000 39 86 154 365 54 114 38 44 3 52
marily hold temperate forest, five (15%) Mediterranean- count for the plants that are shared between hotspots. (38) (45) (42) (47) (86) (6)
type ecosystems (two of which also have temperate for- In other words, if we were to attempt to produce such New Zealand 1 865 2 300 2 4 89 198 37 37 4 4 25 39
(81) (50) (45) (100) (100) (64)
est elements), and one (3%) is desert. totals, we would underestimate overall plant endemism New Caledonia 2 432 3 270 6 9 23 105 62 70 0 0 9 85
Among them, the hotspots hold no less than 150 000 by failing to include species confined to multiple hot- (74) (67) (22) (89) — (11)
plant species as single-hotspot endemics (Table 3). That spots while inflating total hotspot richness by single oc- Polynesia-Micronesia 3 074 5 330 11 15 170 300 31 61 3 3 20 96
is 50% of the world’s total. By far, the two hotspots with currences counted more than once. (58) (73) (57) (51) (100) (21)
32 33
 While data for most invertebrate groups remains
sparse, we can produce much more accurate summary
statistics for terrestrial vertebrates (Table 2). The grow-
ing accuracy of the figures provided for terrestrial ver-
tebrates in the hotspots is largely due to major ad-
vances in the reliability of species distribution data.
We have relied on two main sets of data here: the
IUCN (the World Conservation Union) Red List part-
nership and data synthesized across terrestrial ecore-
gions by the World Wildlife Fund-U.S. (Olson et al.
2001). The new data presented for amphibians per
hotspot derives entirely from the former, namely the
groundbreaking work of the Global Amphibian As-
sessment. Mammal data is based on work initiated by
the Global Mammal Assessment, conducted through
the same partnership, and mapped to ecoregions. Dis-
tribution data for birds has always been the most ad-
vanced of the four terrestrial vertebrate groups, thanks
to the pioneering research of BirdLife International, an-
other member of the IUCN Red List partnership, and
was likewise expanded to synthesize the distributions
of non-threatened species across ecoregions. Reptile
data remains poor (only crocodilians, turtles, and
tortoises having been relatively well assessed), but
a comprehensive online taxonomic reference exists
(www.embl-eidelberg.de/~uetz/LivingReptiles.html),
preliminary data has been compiled across ecoregions,
and a Global Red List Assessment is expected to com-
mence in 2005. Species regularly occurring in a given
hotspot were included in the analysis, and so vagrants
or passage migrants were not incorporated. Since
seabirds spend significant time on land where they
nest, their nesting ranges were included, while marine
mammals and marine reptiles were wholly excluded.
Thus, a number of wide-ranging seabirds are consid-
ered endemic to hotspots in which their entire breed-
ing ranges are confined.
Overall, the total number of terrestrial vertebrates
endemic to a single hotspot is 10 413, representing
36% of all terrestrial vertebrate species. With rapid in-
creases in data quality and synthesis, we can now de-
rive species lists for each hotspot, above and beyond
estimated species numbers. We are, therefore, able to
sum not only single-hotspot endemics but also those
species that are confined to multiple hotspots: thus,
12 066 species in total, 42% of all terrestrial verte-
brates, are endemic to the 34 hotspots combined. Fi-
nally, we list the exact number of species occurring in
these hotspots (i.e., species occurring within the
boundaries of, but not endemic to, the hotspots):
22 319 —77% of the planet’s total. We break these num-
bers down by class —mammals, birds, reptiles, and
amphibians— in Table 3. Clearly, reptiles and, espe-
cially, amphibians, are more prone to hotspot en-
demism than are the more wide-ranging mammals
and birds, but the overall similarity between taxonom-
ic groups is reassuring. All of these are truly astound-
ing numbers, demonstrating the extremely high con-
centration of life-forms in hotspots.
The current analysis includes the first assessment of
inland fishes across all hotspots, thereby completing
the coverage of vertebrates. In order to achieve this ad-
vance, we have relied on William Eschmeyer’s excel-
lent Catalogue of Fishes (1998; www.calacademy.org/
research/ichthyology) and an increasing number of at-
lases and checklists for large fish faunas (e.g., Reis et al.
2003). In some cases, we assembled preliminary lists
for the hotspots from country lists available in the
FishBase online database (Froese and Pauly 2003;
www.fishbase.org). Distributions of species in the coun-
try lists were then checked against the boundaries of
the hotspots. The data given in Table 2 for fishes re-
flects species considered valid in the Catalogue of Fish-
es as of February 2004; all inland fishes native to the
hotspots are included, without regard to their salinity
preferences. This is clearly a significant underestimate
of the actual number of fishes in the hotspots, given
that new species of freshwater fishes are being discov-
ered at a rate of close to two hundred species per year.
Future analyses will include estimates of undiscovered
species and will increasingly be based on species map-
ping, as is already the case for most other vertebrates.
Overall, 28% of the world’s 12 070 recognized freshwa-
ter fishes (3 418) are endemic to individual hotspots;
3 550 (29%) are endemic to all hotspots; and 6 689
species (55%) occur in the hotspots.
There are only two invertebrate groups for which
comprehensive global data allow assessment across
hotspots. These are tiger beetles (Pearson and Carroll
1998), for which 1 326 (58% of all species) are endemic On pp. 34-35, a buttress tree
to individual hotspots, and Nasutitermes termites (Abe (Dracontomelon vitiense) and a
et al. 2000), among which 194 (30% of all species) are large hanging vine in the Vatthe
Conservation Area, Espiritu Santo,
single-hotspot endemics, 229 (36%) are endemics over-
Vanuatu Islands, Melanesia
all, and 439 (69%) occur in the hotspots (Table 4). These Hotspot.
proportions are remarkably similar to those found for © Patricio Robles Gil/Sierra Madre
vertebrates and plants, and endemism to individual
hotspots is also highly congruent. More generally, how- On the opposite page, an orchid of
ever, an acceleration of efforts to provide conservation the genus Sobralia, which is native
to Central and South America, in
practitioners with data is under way for invertebrate
Corcovado National Park on the
taxa, with global taxonomic and distribution databases Osa Peninsula, Costa Rica.
of selected invertebrate groups now available (e.g., for © Kevin Schafer
ants www.antbase.org, katydids, www.tettigonia.com,
and beetles www.coleoptera.org), and numerous others
TABLE 3. Numbers of plant and vertebrate species endemic to single hotspots,
endemic to any hotspot(s), and occurring in any hotspot(s). The first row gives
the total global number of species in each group, following Myers et al. (2000)
for plants, and the World Wildlife Fund-U.S. database of terrestrial vertebrates
in ecoregions and relevant IUCN assessments for terrestrial vertebrates
Plants Mammals Birds Reptiles Amphibians
Total number worldwide 300 000 4 932 10 253 8 163 5 454
Single-hotspot endemics 150 000 1 357 2 910 3 305 2 841
% of all species in taxon 50 27 28 40 52
Endemic to any hotspot(s) — 1 569 3 551 3 723 3 223
% of all species in taxon — 32 35 46 59
Occurring in any hotspot(s) — 3 744 8 385 5 779 4 411
% of all species in taxon — 76 82 71 81
37
 TABLE 4. Numbers of tiger beetle and Nasutitermes termite species can, cautiously, use taxonomy as a surrogate for phy-
endemic to (E) and occurring in (O) each of the 34 hotspots logeny. Genera tend to be more distinct from one an-
Tiger beetles Nasutitermes termites other than are species, families more distinct than gen-
(n = 2 304) (n = 640) era, and so on, although the exact phylogenetic differ-
Hotspot E O E O ence between taxonomic levels varies considerably
Tropical Andes 48 63 7 29 within and between classes (Avise and Johns 1999).
Tumbes-Chocó-Magdalena 28 42 3 7 Table 5 presents the numbers of plant and vertebrate
Atlantic Forest 63 88 11 67 genera and Table 6, the numbers of plant and verte-
Cerrado 23 64 28 103 brate families endemic to and occurring in hotspots.
Chilean Winter Rainfall-Valdivian Forests 1 5 0 0 Table 7 emphasizes those hotspots holding endemic
Mesoamerica 107 149 17 34 genera and families. The results are remarkable: the
Madrean Pine-Oak Woodlands 43 71 0 3 hotspots hold particularly large numbers of endemic
Caribbean Islands 9 23 7 14 genera and families (Table 8), even relative to their
California Floristic Province 5 19 0 0 high levels of species endemism (Table 2). Indeed, sim-
Guinean Forests of West Africa 15 52 2 23
ulation tests (for birds) show levels of generic and fa-
Cape Floristic Region 4 21 0 2
milial endemism in the hotspots to be significantly
Succulent Karoo 2 10 0 3
higher than expected given their numbers of endemic
Maputaland-Pondoland-Albany 15 36 0 7
Coastal Forests of Eastern Africa 2 11 1 8
species (Brooks et al. 2004). At the higher level, the dis-
Eastern Afromontane 19 38 3 20 tantly isolated hotspots like Madagascar and the Indian
Horn of Africa 28 46 0 2 Ocean Islands, New Caledonia, New Zealand, Polyne-
Madagascar and the Indian Ocean Islands 211 213 7 9 sia-Micronesia, and the Caribbean Islands stand out in
Mediterranean Basin 24 46 0 1 particular (Table 7). The exceptional nature of Mada-
Caucasus — — 0 0 gascar and the Indian Ocean Islands, with a total of 24
Irano-Anatolian 0 18 0 0 endemic families (eight plant families, 16 endemic ver-
Mountains of Central Asia — — 0 0 tebrate families) can not be overemphasized; its closest
Western Ghats and Sri Lanka 101 139 15 26 competitors are New Zealand, Southwest Australia, and
Himalaya 34 72 10 15 the Chilean Winter Rainfall-Valdivian Forests, each
Mountains of Southwest China — — 0 0 with seven endemic families. The same is true of en-
Indo-Burma 167 279 15 39
demic genera, with Madagascar and the Indian Ocean
Sundaland 96 149 47 81
Islands topping the world list with 478 (310 plant genera,
Wallacea 79 106 3 8
168 vertebrate genera); the Caribbean Islands are sec-
Philippines 113 132 16 31
Japan 6 24 0 0 ond with 269 and Sundaland third with 199 (although
Southwest Australia 39 43 0 17 here it should be noted that data is incomplete for some
East Melanesian Islands 11 20 1 7 of the most diverse hotspots, e.g., the Tropical Andes
New Zealand 14 14 0 1 and the Atlantic Forest, for which comprehensive in-
New Caledonia 15 16 0 1 formation on endemic plant genera were simply not
Polynesia-Micronesia 4 5 1 1 available).

are at different stages of development. Recently, Con-

On the opposite page, a large fern in
the Hengduan Shan Mountains,
in Yunnan Province, China.
© Patricio Robles Gil/Sierra Madre
38
servation International has launched an Invertebrate
Diversity Initiative to help initiate and coordinate ef-
forts to gather and share data on invertebrate distribu-
tions, as well as to develop innovative tools for inverte-
brate identification.
While the 34 hotspots clearly hold astounding levels
of species endemism, this is not sufficient to describe
the extent to which they represent the history of life.
This is important because it could be argued that an-
other measure of biodiversity, phylogenetic diversity
(the length of time of independent evolution of a given
species), better represents evolutionary potential, eco-
logical diversity, and the range of options for future hu-
man use than does species endemism per se. Sechrest
et al. (2002) provided the first test of this for two mam-
malian lineages —primates and carnivores—, and found
that hotspots hold more unique evolutionary history
than expected given their numbers of endemic species.
Phylogenetic data of this kind is not yet available for
most groups of species, but for terrestrial vertebrates we
Threats and Responses in the Hotspots
Threats to hotspots are similar to, although more in-
tense than, the threats facing biodiversity worldwide.
Habitat destruction, by definition, is a pervasive threat
affecting hotspots, and is already causing extinctions in
many (Brooks et al. 2002). Accelerating anthropogenic
climate change will undoubtedly magnify the effects of
habitat destruction and fragmentation (Thomas et al.
2004), although its specific effects on biodiversity have
yet to be assessed for most hotspots (Midgely et al.
2002). Predatory invasive species have already had a
devastating impact on the island hotspots, where spe-
cies evolved in the absence of animals such as cats and
rats (Steadman 1995). Introduction of exotic plant spe-
cies into hotspots, particularly those of Mediterranean-
type vegetation, is also having massive ecosystem ef-
fects, wholly changing hydrology, biogeochemical
cycles and, ultimately, biodiversity (Groves and Di Cas-
tri 1991). Direct exploitation of species for food, medi-
 TABLE 5. Numbers of plant and vertebrate genera endemic to (E) and occurring in (O) each of the 34 hotspots.
A question mark (?) denotes gaps in knowledge
Plants Mammals Birds Reptiles Amphibians Freshwater fishes
Hotspot E O E O E O E O E O E O
Tropical Andes ? ? 5 205 69 623 3 127 10 79 3 136
Tumbes-Chocó-Magdalena ? ? 1 147 13 455 5 114 0 41 7 141
Atlantic Forest ? ? 12 135 22 470 8 105 23 69 10 68
Cerrado ? ? 3 124 0 364 0 91 0 43 18 240
Chilean Winter Rainfall-Valdivian Forests ? ? 5 41 2 132 0 8 5 12 3 17
Mesoamerica 65 2 523 3 175 20 493 11 167 14 59 25 157
Madrean Pine-Oak Woodlands ? ? 2 133 3 265 1 114 0 28 1 42
Caribbean Islands 205 2 500 15 57 35 205 8 46 1 7 5 67
California Floristic Province 52 ? 0 62 0 180 0 36 0 15 4 37
Guinean Forests of West Africa ? ? 7 132 6 311 1 73 7 48 11 146
Cape Floristic Region 160 942 0 76 0 190 0 46 2 16 0 19
Succulent Karoo 80 1 002 0 60 0 141 0 42 0 11 0 12
Maputaland-Pondoland-Albany 39 1 524 0 123 0 267 1 79 2 26 0 31
Coastal Forests of Eastern Africa 28 1 050 0 115 0 287 1 95 1 28 0 78
Eastern Afromontane 44 1 578 12 181 11 361 1 106 17 52 93 185
Horn of Africa 60 970 5 115 1 285 6 94 1 20 3 48
Madagascar and the Indian Ocean Islands 310 ? 40 61 43 157 49 69 22 25 14 59
Mediterranean Basin ? ? 1 94 0 201 4 68 1 17 6 75
Caucasus 17 ? 1 64 0 161 0 33 1 9 0 63
Irano-Anatolian ? ? 0 77 0 167 0 46 1 8 0 43
Mountains of Central Asia 64 ? 0 78 0 195 0 31 1 4 0 16
Western Ghats and Sri Lanka 81 2 218 4 77 0 236 22 88 10 29 8 45
Himalaya 71 ? 1 155 1 344 1 73 0 28 33 104
Mountains of Southwest China 20 ? 0 128 0 253 0 45 0 22 2 50
Indo-Burma ? ? 7 172 3 405 12 152 3 52 30 373
Sundaland 117 ? 17 136 10 305 24 125 6 41 25 314
Wallacea 12 ? 22 83 26 235 3 83 0 14 7 90
Philippines 26 1 308 23 85 7 227 6 75 0 24 9 118
Japan 20 ? 6 46 1 174 0 35 1 16 4 107
Southwest Australia 87 697 4 35 0 158 2 51 4 12 3 18
East Melanesian Islands ? ? 3 37 7 136 6 41 4 10 0 32
New Zealand 35 390 1 3 15 71 5 6 1 1 1 15
New Caledonia 107 807 0 5 3 65 11 23 0 0 1 49
Polynesia-Micronesia 63 802 0 7 28 111 2 25 0 1 0 25

cine, and the pet trade is a serious threat to all hotspots,
particularly in the Guinean Forests of West Africa
(Bakarr et al. 2001) and several Asian hotspots such as
Indo-Burma (Van Dijk et al. 2000; Martin and Stiles
2002). Another grave concern is the severe decline of
amphibians worldwide (Houlahan et al. 2000), the cause
of which remains unclear but may be linked to fungal
disease, pesticides, UV-B radiation, climate change, and
synergistic interactions.
Overall, researchers have found that the hotspots
hold more people than expected when compared to
global population density (Cincotta et al. 2000). Over-
laying estimates of human population for 2002 on the
updated hotspots reveals a similar result (Table 9). But
the relationship between people and biodiversity is
not simply one where more people lead to greater im-
pacts on biodiversity. Despite the shared characteristic
among hotspots of high habitat loss, population densi-
man density but rather in human activity, with many
behaviors enabling relatively few people to have par-
ticularly adverse impacts on other species. A good ex-
ample of this is the Cerrado, with a population density On pp. 40-41, the Endangered Fiji
of only 13 persons per km2 in 2002 but considerable banded iguana (Brachylophus
habitat loss due to the widespread commercial agricul- fasciatus) is one of two iguana
species found in the Pacific
ture that dominates this broad savanna (Kaimowitz and
Islands. This one occurs on both Fiji
Smith 2001). and Tonga, while the other is found
Of greater concern than general human presence in only on a few tiny islands
the hotspots is human settlement near protected ar- in Fiji. Their nearest relatives
eas, the repositories for much of the remaining biodi- live way across the Pacific in
versity in these regions. For the updated hotspots, pro- South America.
© Patricio Robles Gil/Sierra Madre
tected areas with assigned IUCN categories had an
estimated 313 million people within 10 km of their On the opposite page, buttress tree
borders in 2002. Mean population densities near pro- in the Ivory Coast Forest,
tected areas again varied considerably, from 2 people Guinean Forests of West Africa
per km2 to nearly 300 per km2, with population densi- Hotspot.
ties within 10 km of protected areas greater than for © Patricio Robles Gil/Sierra Madre

ty among hotspots varies widely, from four persons per the hotspots overall. Such a large number of people
km2 to 336 persons per km2. Much of our understand- near protected areas places sources of high demand
ing of human-biodiversity interactions lies not in hu- close to localities of particular importance for biodi-
43
 TABLE 6. Numbers of plant and vertebrate families
endemic to (E) and occurring in (O) each of the 34 hotspots. A question mark (?) denotes gaps in knowledge
Plants Mammals Birds Reptiles Amphibians Freshwater fishes
Hotspot E O E O E O E O E O E O
Tropical Andes ? ? 0 44 0 72 0 27 0 12 0 37
Tumbes-Chocó-Magdalena ? ? 0 36 0 69 0 24 0 10 0 54
Atlantic Forest ? ? 0 32 0 70 0 22 1 11 0 20
Cerrado ? ? 0 33 0 67 0 23 0 10 0 40
Chilean Winter Rainfall-Valdivian Forests 3 ? 1 19 0 40 0 5 1 4 2 12
Mesoamerica 3 279 0 34 0 80 1 31 0 14 1 58
Madrean Pine-Oak Woodlands ? ? 0 30 0 60 0 25 0 11 0 13
Caribbean Islands 1 186 2 19 2 56 0 19 0 4 0 39
California Floristic Province 0 ? 0 22 0 55 0 14 0 10 0 17
Guinean Forests of West Africa ? ? 0 35 0 79 0 18 0 12 0 44
Cape Floristic Region 5 164 0 26 0 72 0 14 0 7 0 11
Succulent Karoo 0 168 0 25 0 58 0 14 0 7 0 7
Maputaland-Pondoland-Albany 1 243 0 37 0 80 0 19 0 10 0 16
Coastal Forests of Eastern Africa 0 ? 0 37 0 82 0 21 0 10 0 34
Eastern Afromontane ? 233 0 39 0 83 0 20 0 12 0 24
Horn of Africa 2 170 0 38 0 79 0 21 0 9 0 30
Madagascar and the Indian Ocean Islands 8 ? 6 17 5 60 1 11 2 7 2 22
Mediterranean Basin 2 ? 0 32 0 67 0 19 0 10 1 25
Caucasus 0 ? 0 20 0 57 0 3 0 8 0 19
Irano-Anatolian ? ? 0 25 0 57 0 14 0 6 0 11
Mountains of Central Asia 0 ? 0 21 0 63 0 12 0 3 0 2
Western Ghats and Sri Lanka ? 217 0 28 0 76 0 17 1 8 0 14
Himalaya ? ? 0 35 0 84 0 15 0 8 0 30
Mountains of Southwest China 2 ? 0 30 0 66 0 12 0 9 0 11
Indo-Burma ? ? 1 37 0 95 0 20 0 11 1 74
Sundaland 1 ? 0 36 0 83 2 20 0 7 0 71
Wallacea ? ? 0 20 0 71 0 17 0 5 1 45
Philippines 0 194 0 25 1 77 0 15 0 7 0 49
Japan 3 250 0 18 0 59 0 10 0 8 0 41
Southwest Australia 4 139 2 15 0 54 0 9 0 2 1 13
East Melanesian Islands ? ? 0 10 0 59 0 10 0 4 0 18
New Zealand 1 139 1 2 3 33 1 3 1 1 0 11
New Caledonia 5 ? 0 2 1 36 0 4 0 0 0 27
Polynesia-Micronesia 1 159 0 4 1 38 0 10 0 1 0 3

On pp. 44-45, pair of southern
cassowaries (Casuarius casuarius)
mating. This Vulnerable species is
being severely impacted by ongoing
habitat fragmentation in Australia
(where fewer than 2 000 individuals
remain), and often suffers from
road fatalities.
© Patricio Robles Gil/Sierra Madre
On the opposite page,
the dwarf chameleons of the genus
Brookesia are entirely endemic to
Madagascar. The species seen here,
B. superciliaris, inhabits the entire
eastern coast of the island.
© Piotr Naskrecki
46
versity conservation. Further, much of the land near
protected areas in hotspots is not particularly suitable
for agriculture. Although this low suitability means
that the cost of biodiversity conservation in terms of
lost crop production is low, it also implies that the
capacity of people living near protected areas to sup-
port themselves is limited —in many cases, requiring
an expansion of cultivated land or harvesting other
resources to meet local demands that can compromise
nearby biodiversity.
Hotspots are also notable centers of violent conflict,
for example, Mesoamerica (Chiapas and Guatemala),
the Caribbean Islands (Haiti), the Tropical Andes and
Tumbes-Chocó-Magdalena (Colombia), the Guinean
Forests of West Africa (Liberia, Sierra Leone, Côte
d’Ivoire), the Eastern Afromontane rifts (Rwanda, east-
ern Democratic Republic of Congo), the Horn of Africa
(Somalia), the Caucasus (Chechnya), the Irano-Anatolian
region (Iraq, Iran), the Mountains of Central Asia (Af-
ghanistan), Indo-Burma (Myanmar), Sundaland (Aceh),
Wallacea (Timor), the southern Philippines, and the
East Melanesian Islands (Solomons). Further, two of
the major areas affected by recent violent conflict not
within hotspots are the Congo and West Papua; both
are situated within high biodiversity wilderness areas,
and in the former case the conflict that has spilled
over into the Congo largely originates in the Albertine
Rift region of the Eastern Afromontane Hotspot. The
degree to which these conflicts cause biodiversity loss
or are caused by biodiversity loss is unclear (Dudley et
al. 2002), but the correlation is unnerving (Hauge and
Ellingsen 1998).
The primary response to the biodiversity crisis, in
the hotspots and elsewhere, must be the establishment
and effective management of protected areas (Bruner
et al. 2001), whether as strict reserves or other safe-
guard mechanisms (such as community conservation
agreements in Melanesia or indigenous reserves in the
Andean countries) to ensure the persistence of an
area’s biodiversity. Using the recently released World
Database on Protected Areas (WDPA 2003) —a consor-
tium of the American Museum of Natural History,
 TABLE 7. Hotspots ranked by numbers of endemic
plant and vertebrate genera and families. A question mark (?) denotes gaps in knowledge
Endemic genera Endemic families
Hotspot Total Plants Vertebrates Total Plants Vertebrates
Madagascar and the Indian Ocean Islands 478 310 168 24 8 16
Caribbean Islands 269 205 64 5 1 4
Sundaland 199 117 82 3 1 2
Eastern Afromontane 178 44 134 0 0 0
Cape Floristic Region 162 160 2 5 5 0
Mesoamerica 138 65 73 5 3 2
Western Ghats and Sri Lanka 125 81 44 1+ ? 1
New Caledonia 122 107 15 6 5 1
Himalaya 107 71 36 0 0 0
Southwest Australia 100 87 13 7 4 3
Polynesia-Micronesia 93 63 30 2 1 1
Tropical Andes 90+ ? 90 0 0 0
Succulent Karoo 80 0 80 0 0 0
Horn of Africa 76 60 16 2 2 0
Atlantic Forest 75+ ? 75 1+ ? 1
Philippines 71 26 45 1 0 1
Wallacea 70 12 58 1 0 1
Mountains of Central Asia 65 64 1 0 0 0
New Zealand 58 35 23 7 1 6
California Floristic Province 56 52 4 4+ ? 4
Indo-Burma 55+ ? 55 2+ ? 2
Maputaland-Pondoland-Albany 42 39 3 1 1 0
Guinean Forests of West Africa 32+ ? 32 0 0 0
Japan 32 20 12 3 3 0
Coastal Forests of Eastern Africa 30 28 2 0 0 0
Tumbes-Chocó-Magdalena 26+ ? 26 0 0 0
Mountains of Southwest China 22 20 2 2 2 0
Cerrado 21+ ? 21 0 0 0
East Melanesian Islands 20+ ? 20 0 0 0
Caucasus 19 17 2 0 0 0
Chilean Winter Rainfall-Valdivian Forests 15+ ? 15 7 3 4
Mediterranean Basin 12+ ? 12 2 1 1
Madrean Pine-Oak Woodlands 7+ ? 7 0 0 0
Irano-Anatolian 1+ ? 1 0 0 0

BirdLife International, Conservation International,
Fauna and Flora International, The Nature Conser-
vancy, the World Commission on Protected Areas, the
United Nations Environment Programme-World Con-
servation Monitoring Center, the Wildlife Conserva-
tion Society, the World Resources Institute, and the
World Wildlife Fund—, we have been able to overlay
the distribution of the approximately 100 000 protect-
ed areas mapped worldwide onto hotspots, and deter-
mine how much of each hotspot is under some form of
protection. In addition, because protected areas may
not retain their species if they are used for extraction
of natural resources (Peres and Lake 2003), we have
repeated this analysis considering only protected areas
classified under IUCN categories I-IV (IUCN 1994) be-
cause the management objectives of these categories
imply constraints on human occupation or resource
use. The results of both analyses are given in Table 10. On the opposite page,
The average protected area coverage of hotspots is wild geraniums in the central
10.1% of their original extent, and considering only Caucasus Mountains of Georgia.
© Pat O’Hara
IUCN categories I-IV, 5.0% of original extent. However,
the percentage coverage is much less important than the

TABLE 8. Numbers and percentages of vertebrate genera (G) and families (F)
endemic to single hotspots, endemic to any hotspot(s), and occurring in any hotspot(s)
Mammals Birds Reptiles Amphibians Freshwater fishes
G F G F G F G F G F
Total number worldwide 1 091 124 2 110 195 1 034 63 445 46 1 886 201
Single-hotspot endemics 201 13 326 13 193 4 138 6 326 9
% of all genera/families in taxon 18 11 16 7 19 6 31 13 17 4
Endemic to any hotspot(s) 241 15 433 13 266 6 172 7 350 13
% of all genera/families in taxon 22 12 21 7 26 10 39 15 19 7
Occurring in any hotspot(s) 973 118 1 795 171 850 55 382 43 1 539 180
% of all genera/families in taxon 89 95 85 88 82 87 86 94 82 90
49
 TABLE 9. Estimates of population and population density in hotspots and nearby protected areas, 2002.
Population figures rounded to the nearest 100 000 (or to the nearest 1 000 for values smaller than 50 000)
Entire hotspot Nearby protected areas (<10 km)
Hotspot Population Persons/km2 Population Persons/km2
Tropical Andes 56 700 000 37 9 700 000 53
Tumbes-Chocó-Magdalena 13 900 000 51 500 000 19
Atlantic Forest 106 800 000 87 4 400 000 137
Cerrado 26 800 000 13 1 800 000 76
Chilean Winter Rainfall-Valdivian Forests 14 700 000 36 400 000 8
Mesoamerica 81 400 000 72 22 500 000 101
Madrean Pine-Oak Woodlands 14 700 000 32 4 100 000 98
Caribbean Islands 35 600 000 155 7 600 000 140
California Floristic Province 35 600 000 121 13 100 000 132
Guinean Forests of West Africa 84 700 000 137 5 000 000 127
Cape Floristic Region 4 000 000 51 3 100 000 90
Succulent Karoo 400 000 4 100 000 16
Maputaland-Pondoland-Albany 19 300 000 71 7 500 000 124
Coastal Forests of Eastern Africa 15 000 000 51 1 600 000 109
Eastern Afromontane 96 900 000 92 10 700 000 99
Horn of Africa 38 100 000 23 1 700 000 35
Madagascar and the Indian Ocean Islands 19 200 000 32 2 400 000 46
Mediterranean Basin 232 200 000 111 34 600 000 211
Caucasus 36 100 000 68 7 500 000 75
Irano-Anatolian 52 000 000 58 5 000 000 71
Mountains of Central Asia 36 000 000 42 2 400 000 29
Western Ghats and Sri Lanka 49 400 000 261 12 700 000 225
Himalaya 91 300 000 123 15 700 000 124
Mountains of Southwest China 8 500 000 32 1 200 000 28
Indo-Burma 316 900 000 134 38 600 000 116
Sundaland 229 400 000 153 38 300 000 226
Wallacea 27 500 000 81 5 300 000 124
Philippines 81 000 000 273 20 700 000 265
Japan 125 400 000 336 33 000 000 297
Southwest Australia 1 700 000 5 9 000 2
East Melanesian Islands 1 300 000 13 10 000 16
New Zealand 3 800 000 14 3 300 000 16
New Caledonia 200 000 10 100 000 24
Polynesia-Micronesia 2 800 000 58 400 000 54
Total 1 959 100 000 83 312 500 000 116

On pp. 50-51, male and female
mountain nyalas (Tragelaphus
buxtoni) in Bale Mountains
National Park, an area that
represents the last stronghold for
this Endangered species.
© Patricio Robles Gil/Sierra Madre
On the opposite page, the
Endangered diademed sifaka
(Propithecus diadema) is found in
the eastern rainforests of
Madagascar and is one of the two
largest living lemurs, reaching
weights of up to 7 kg. The animal
depicted here is an undescribed
form from the Tsinjoarivo region,
not far from the capital of
Antananarivo.
© Russell A. Mittermeier
52
location of the protected areas (Rodrigues et al. 2004).
Conservation action in the coming years must focus
heavily on ensuring the long-term persistence of the ar-
eas already protected, particularly those that contain ir-
replaceable habitat, while at the same time adding new
parks and reserves in the highest priority portions of
unprotected intact habitat. The creation of new protect-
ed areas needs to be carried out in a strategic manner,
focusing on filling gaps and making biodiversity cover-
age for each hotspot as complete as possible.
Establishing protected areas that remain resilient
against the increasing threat of climate variability is a
further major challenge. Climate change, driven by hu-
man activity yearly emitting tens of billions of tons of
heat-trapping gases into the atmosphere, forces species
to shift their ranges according to changes in their pre-
ferred habitat conditions, but this movement may be
difficult or impossible in heavily fragmented land-
scapes such as those of the hotspots. Further, the rate
and magnitude of the current climate change is such
that many species may be unable to disperse quickly
enough. For instance, many of the hotspots are mon-
tane regions, and higher elevation species often have
nowhere to disperse to as temperatures rise and other
climatic conditions change. Thus, protecting species
where they currently exist is only the beginning —pro-
tection is needed where species will be in the future, as
well as connections in the landscape between the two.
Biologists are now able to create computer models of
species’ range shifts that can be used to plan corridors
—comprised of additional parks or multiple-use areas
such as forest reserves— to limit the damage of climate
change. To be successful, landscape conservation of
this sort requires a large percentage of remaining nat-
ural habitat. Proactive investment in these areas will
help prevent accelerating extinctions in the near future
due to the “double whammy” of climate change in a
fragmented landscape.
In hotspots that have been especially devastated
by habitat loss, we must improve our abilities to restore
habitats if many already threatened species are to per- TABLE 10. Extent and proportion of coverage by all protected
sist in the face of climate change. Examples include nu- areas and by protected areas in IUCN categories I-IV, for each hotspot.
merous threatened frog species in Sri Lanka that are Percentages are given for original extent of each hotspot
hanging on within small fragments of the country’s 5% All protected areas IUCN categories I-IV
of remaining rainforest. These species may already be Hotspot Area (km2) % Area (km2) %
affected by decreasing precipitation and warmer tem-
Tropical Andes 246 871 16.0 121 650 7.9
peratures, conditions that increasingly wring small
Tumbes-Chocó-Magdalena 34 338 12.5 18 814 6.9
forest fragments dry. In these cases, restored forests Atlantic Forest 50 370 4.1 22 782 1.8
around existing patches will provide water-retentive Cerrado 111 051 5.5 28 736 1.4
buffers locally, as well as corridors to forest remnants Chilean Winter Rainfall-Valdivian Forests 50 745 12.8 44 388 11.2
that respond differently to a changing climate. Mesoamerica 142 103 12.6 63 902 5.7
Conservation success depends on working effective- Madrean Pine-Oak Woodlands 27 361 5.9 8 900 1.9
ly with people. Many residents of the Earth’s most bio- Caribbean Islands 29 605 12.9 16 306 7.1
diverse places are poor, living on less than one dollar California Floristic Province 108 715 37.0 30 002 10.2
per day and directly depending on the products of Guinean Forests of West Africa 108 104 17.4 18 880 3.0
healthy ecosystems, harvesting wild plants and animals Cape Floristic Region 10 859 13.8 10 154 12.9
for their food, fuel, clothing, medicine, and shelter. A Succulent Karoo 2 567 2.5 1 890 1.8
large portion of the sites with remaining biodiversity is Maputaland-Pondoland-Albany 23 051 8.4 20 322 7.4
Coastal Forests of Eastern Africa 50 889 17.5 11 343 3.9
made up of traditional lands of indigenous peoples.
Eastern Afromontane 154 132 15.1 59 191 5.8
The economies, identities, spiritual and cultural values,
Horn of Africa 145 322 8.8 51 229 3.1
and forms of social organization of indigenous and tra-
Madagascar and the Indian Ocean Islands 18 482 3.1 14 664 2.4
ditional peoples are often tied closely to maintaining Mediterranean Basin 90 242 4.3 28 751 1.4
the biodiversity and ecosystems of the areas they Caucasus 42 721 8.0 35 538 6.7
inhabit. Living resources have a unique place in indig- Irano-Anatolian 56 193 6.2 25 783 2.9
enous culture, but are also singular from a biological Mountains of Central Asia 59 563 6.9 58 605 6.8
conservation perspective. These species are irreplace- Western Ghats and Sri Lanka 26 130 13.8 21 259 11.2
able: they occur nowhere else in the world. Their loss Himalaya 112 578 15.2 77 739 10.5
represents not only a loss of global biodiversity, but of Mountains of Southwest China 14 034 5.3 4 273 1.6
cultural patrimony as well. In short, many people and Indo-Burma 235 758 9.9 132 283 5.6
many species share a common vulnerability and strug- Sundaland 179 723 12.0 77 408 5.2
gle for survival. Protecting the critical ecosystems and Wallacea 24 387 7.2 19 702 5.8
Philippines 32 404 10.9 18 060 6.1
biodiversity on which the rural poor and indigenous
Japan 62 025 16.6 21 918 5.9
peoples depend therefore has significant environmen-
Southwest Australia 38 379 10.8 38 258 10.7
tal, economic, and social benefits.
East Melanesian Islands 5 677 5.7 0 0
How much might it cost to complete the protected New Zealand 74 260 27.5 59 794 22.1
area system in order to conserve biodiversity across the New Caledonia 4 192 22.1 497 2.6
hotspots? Myers et al. (2000) suggested $20 million Polynesia-Micronesia 2 436 5.2 2 088 4.4
per hotspot a year which, although much higher than
current expenditure, is probably an underestimate.
Pimm et al. (2001) estimated higher, suggesting that a (1999) to provide data on these protected area funding
one-time investment of $1 billion per hotspot (roughly shortfalls. We estimate that effective management of
$50 million per year) would be necessary. However, the existing protected area systems in all 34 hotspots
there are dramatic economies of scale driven by the would cost a total of approximately $3.3 billion per On the opposite page, a typical
size of protected areas (Frazee et al. 2003). Further, lo- year (around $100 million per hotspot). Expansion of fragmented forest landscape in the
cal and national economic conditions make for differ- the protected area systems to the highest priority heavily impacted northeastern
portion of Brazil’s Atlantic Forest.
ences in the costs (as well as the benefits) of conserva- unprotected sites would increase management costs
The Atlantic Forest has long been
tion (Balmford et al. 2003), and so variation among the to perhaps $5.5 billion per year ($160 million per the agricultural center of Brazil, as
hotspots is considerable. Effective management of hotspot). Costs of initial land acquisition for this ex- typified in this photograph of a
smaller protected areas in the hotspots found within pansion, through purchase or compensation, might be large sugarcane plantation
relatively prosperous countries may require annual as much as $100 billion, although this is probably a sig- interspersed with forest patches
per-hectare expenditures easily 100 times greater than nificant overestimate. These costs are driven in large taken from Usina Serra Grande in
the State of Alagoas.
those needed for larger protected areas in hotspots oc- part by the factors noted above —many priority ex-
© Luiz Claudio Marigo
curring in low-income countries. pansions must take place in relatively expensive con-
There are crippling shortfalls in funding for pro- texts. On the other hand, it is critical to note that there
tected area management across the hotspots, even are opportunities for extremely low-cost expansions
where management is less expensive. Studies pro- and improvements in management even in high-cost
duced by ourselves, Andrew Balmford, Daan Vreug- regions, so strategic investment of far more limited
denhil, and others at the Fifth World Parks Congress in funds can make a major contribution to hotspot con-
Durban, South Africa build on the work of James et al. servation.
55

On pp. 56-57, the Ethiopian
Highlands attract large amounts of
rainfall, which has important
implications for humans that
depend on the water originating
there. The Djema River is one a
number of rivers and streams that
join to form seven major rivers,
including the Great Abbai (Blue
Nile), the Awash, the Wabe
Shebelle, the Juba, the Ghibie and
Omo, and the Sobat.
© Patricio Robles Gil/Sierra Madre
On the opposite page, young
shepherd girl with goat in the
vicinity of Bale Mountains National
Park. Overgrazing by goats and
other domestic animals poses a
major threat to the remaining
natural Afromontane vegetation.
© Patricio Robles Gil/Sierra Madre
58
A final important consideration in the evaluation of
costs of conservation is the impact of “perverse subsi-
dies.” Myers (1998) estimated that the global cost of
subsidies which are economically harmful totals $2 tril-
lion per year, of which approximately two-thirds are
also environmentally harmful. Mobilization of the po-
litical will to remove these subsidies would clearly go a
long way towards balancing the global shortfall in con-
servation investment.
Hotspots as Priorities within
Bioregional Classification
Obviously, hotspots are not the only system devised for
assessing global conservation priorities (Fonseca et al.
2000). For example, the World Wildlife Fund-U.S. has
derived a system called the Global 200 Ecoregions
(www.worldwildlife.org/ecoregions/), which represent
142 priorities for conservation among 867 terrestrial
ecoregions (Olson et al. 2001), plus 53 freshwater and
43 marine ecoregions. The aim of the Global 200 (Ol-
son and Dinerstein 1998) is to prioritize ecoregions for
conservation within each of 12 terrestrial, three fresh-
water, and four marine major habitat types (MHTs
—these equal the biomes of Table 1). Although not de-
fined using any quantitative criteria, Olson and Din-
erstein (1998:509) “…chose the set of ecoregions based
on the following parameters: species richness, en-
demism, taxonomic uniqueness (e.g., unique genera
or families, relict taxa or communities, primitive lin-
eages), unusual ecological or evolutionary phenome-
na (e.g., intact large vertebrate faunas or migrations,
extraordinary adaptive radiations), and global rarity of
MHT.” Given the breadth of characteristics used in
the Global 200 assessment, the degree to which the
142 Global 200 priority terrestrial ecoregions overlap
with hotspots is surprising. All hotspots contain at
least one Global 200 ecoregion; and 60% of Global 200
terrestrial ecoregions (85 of 142) incorporate hotspots
(www.biodiversityhotspots.org). Further, we should
note that threat and cost are not considered in the
Global 200; if we considered the high biodiversity
wilderness areas (Mittermeier et al. 2003b) as well as
the hotspots, this overlap would be even higher.
Hotspots, high biodiversity wilderness areas, and
the Global 200 require bioregional classification, which
is by no means an exact science. Bioregional classifi-
cation subdivides environmental space, a continuum,
into discrete units. Thus, a line must be drawn to rep-
resent a transition between two habitats that might be
easily discernible (e.g., the treeline on a mountain) or
rather vague (e.g., the difference between woodland
and savanna). Analytical techniques can draw such
lines using cutoffs (e.g., elevation) which aim to rep-
resent a transition in a replicable way. However, the
actual selection of cutoff parameters is arbitrary (e.g.,
3 000 m elevation), and while it may represent the
transition between two habitats well in some places, it
may not do so across the entire border between those
habitats (thus, 3 000 m may represent the treeline at
one point in a mountain range, but not throughout the
entire range). Other bioregional classifications rely
heavily on knowledge of a particular region and on
maps of vegetation cover interpreted from survey
transects or aerial photography. The problem with
these approaches is that they can not be repeated by
following the same methods. While the results of such
bioregional classifications often model the landscape
very accurately, the lack of replicability is unsatisfac-
tory and can lead to competing classifications for the
same area. For example, there are a variety of biore-
gional classifications in use in Mesoamerica (Diner-
stein et al. 1995). A related issue faced by bioregional
classification is that it does not provide a framework
for establishing scientifically justified targets for con-
servation on the ground —it is not possible to ascertain
whether “representation” of an ecoregion need involve
the conservation of as little as 1% of its extent or as
much as 99%.
The ecoregional classification of the World Wildlife
Fund-U.S. is now the most widely-used system of its
kind for global-scale bioregionalization (Wikramanayake
et al. 2002). In order to facilitate analysis, interoperabil-
ity, and collaboration, we have therefore gone to consid-
erable lengths to ensure that both the boundaries of the
hotspots and those of the high biodiversity wilderness
areas (Mittermeier et al. 2003b) correspond directly to
those of the World Wildlife Fund-U.S. ecoregions (Olson
et al. 2001). Thus, our calculations of hotspot extents are
based on ecoregional data. In total, the hotspots incor-
porate 374 ecoregions (www.biodiversityhotspots.org),
43% of the total. The only significant exceptions to the
match between hotspots and ecoregions are now in
three cases where recent maps update ecoregional
boundaries: the Albertine Rift (Plumptre et al. 2003a);
Maputaland-Pondoland-Albany (Van Wyk and Smith
2001); and the Caucasus (Krever et al. 2001).
The Future: Hotspots and Species Distributions
The massive acceleration of the compilation of species
data will soon reach the point where bioregional classi-
fication becomes an increasingly unnecessary surro-
gate for species data. Within just a few years, the iden-
tification of hotspots will no longer rely on the current
criteria of plant endemism and remaining habitat as
surrogates for irreplaceability and threat, respectively,
but will be founded on accurate species distribution
data and better understanding of threats and costs. Pre-
liminary analyses have indicated that the biodiversity
hotspots, as currently defined, show a remarkable de-
gree of congruence with areas of restricted-range (less
than 50 000 km2) species across three vertebrate groups
(mammals, birds, and amphibians). Even for well-
known taxonomic groups, there remains considerable
uncertainty (Wheeler and Meier 2000) over species
concepts (how exactly a species is defined) and hence
whether groups of taxa should be considered one or mul-
tiple species. Nevertheless, the species is the most stable
unit of biodiversity —indeed, many believe species to
be the fundamental unit of biodiversity (Wilson 1992).
The Species Survival Commission (SSC) of IUCN
(www.iucn.org/themes/ssc/) provides the overarching
framework for the ongoing revolution in species data
compilation and availability. This compilation is fo-
cused around the assessment of conservation status,
but also contributes to our understanding of taxonomy,
ecology, and —critically for any kind of priority set-
ting— distribution (e.g., Boitani et al. 1999). All of the
approximately 10 000 bird species, always the best-
known taxonomic class, have been comprehensively
assessed in terms of conservation by BirdLife Interna-
tional (2000). Some families of mammals, especially
the larger species, have already been comprehensively
assessed through the SSC network of Specialist Groups
(www.iucn.org/themes/ssc/sgs/sgs.htm), and initial
assessments for all species were made for the 1996
IUCN Red List of Threatened Animals (Baillie and
Groombridge 1996). As noted earlier, the Global Mam-
mal Assessment is now under way to cover all of the
approximately 5 000 species in the class, while the re-
cently completed Global Amphibian Assessment covers
all of the world’s approximately 5 500 amphibian spe-
cies. Most ambitiously, the Parties to the Convention on
Biological Diversity have set a target date of 2010 to
complete comprehensive assessment of all of the
world’s 300 000 vascular plant species, to be led by
IUCN. Similar assessments for reptiles and fish are now
in the planning stages.
Simultaneous to the development of these assess-
ments of relatively well-studied groups, major initia-
tives within the museum and herbarium communities
have stimulated the availability of taxonomic and geo-
graphic data for all taxa. These include the All Species
Initiative (www.all-species.org), the SALVIAS project
(www.salvias.net) and the Global Biodiversity Informa-
tion Facility (www.gbif.org), which aim to provide up-
to-date, electronic catalogues of known species (Bisby
et al. 2002). This data also allows the incorporation of
environmental data (e.g., from remote sensing) to pro-
duce inductive range models (Raxworthy et al. 2003),
which greatly increase the resolution and accuracy of
distribution data (Peterson et al. 2002).
BirdLife International produced the first direct ap-
plication of comprehensive species data to conser-
vation prioritization at a global scale (ICBP 1992;
Stattersfield et al. 1998). This study mapped the distri-
butions of all bird species with global range sizes of
less than 50 000 km2, and then defined “Endemic Bird
Areas” (EBAs) as regions in which the distributions of
two or more such restricted-range species overlapped.
In total, Stattersfield et al. (1998) identified 2 623 re-
stricted-range bird species (27% of all birds known at
the time) confined to 218 EBAs and 138 secondary ar-
eas (holding just one restricted-range species). Stat-
tersfield et al. (1998) also incorporated measurements
of threat in order to prioritize among EBAs. Given that
EBAs focus on endemism, the large degree of overlap
with hotspots (www.biodiversityhotspots.org) —all but
three hotspots incorporate at least one EBA, and over-
all 170 EBAs (78%) are encompassed by hotspots— is
perhaps unsurprising, despite the fact that EBAs are
defined using bird data alone.
A more recent assessment of global priorities using
species data was prepared for the Fifth World Parks
Congress (Rodrigues et al. 2004). This analysis used
four major datasets: BirdLife International’s (2000)
threatened birds, draft data for all mammals and am-
phibians compiled by the Global Mammal and Amphib-
ian Assessments, and the new World Database on Pro-
tected Areas (WDPA 2003). Overlaying these geographic
datasets, Rodrigues et al. (2004) were able to assess the
representation (coverage) of these terrestrial vertebrate
species in the existing protected areas system, and pri-
orities for its expansion and consolidation based on ir-
replaceability and vulnerability. Again, the congruence
between hotspots and the priority (protected and un-
protected) areas identified by this global gap analysis is
very high (Rodrigues et al. 2004), with no less than 80%
of the high-priority area for the establishment of pro-
tected areas occurring in hotspots. One would expect
this result, given the focus of both systems on irre-
placeability and vulnerability (although the latter analy-
sis explicitly incorporates the extent of existing protec-
tion, which was compiled for, but not a criterion for
hotspots). Importantly, such a framework for assessing
global priorities also offers the potential for more ex-
plicit incorporation of spatial data on costs of conser-
vation, to prioritize among areas where conservation
choices exist.
The core of the data produced by the ongoing IUCN-
SSC species assessments concerns conservation status
(IUCN 2003; www.redlist.org). Building from a long his-
tory of Red List assessment (Fitter and Fitter 1987), the
SSC has developed quantitative criteria under which
the probability of extinction is estimated for each spe-
cies (Mace and Lande 1991). Today, the IUCN Red List
(www.redlist.org) represents the global standard for
species conservation status (Lamoreaux et al. 2003),
and the SSC network of scientists now subjects each
species under consideration to these established and On the opposite page, purple-bellied
rigorous measures of conservation status (IUCN 2001). lories (Lorius hypoinochrous) for
The categories for threatened status are Critically En- sale as pets in a market in Rabaul,
New Britain, in the East
dangered (CR), Endangered (EN), and Vulnerable (VU);
Melanesian Islands Hotspot.
these three categories cover all species that have a high © Patricio Robles Gil/Sierra Madre
probability of extinction in the wild in the medium-
term future. Further categories are Extinct (EX) and Ex-
tinct in the Wild (EW), Near Threatened (NT), and
Least Concern (LC) for non-threatened species. Data
Deficient (DD) and Not Evaluated (NE) are categories
applied to species which can not or have not been eval-
uated, respectively (IUCN 2001).
As might be expected, very large proportions of threat-
ened species occur within —and are often endemic
61
 TABLE 11. Numbers of threatened mammal, bird, and amphibian species endemic to and occurring in (in parentheses) individual TABLE 12. Numbers and percentages of threatened mammal, bird,
hotspots, following IUCN (2003) and provisional 2004 listings for amphibians. VU-Vulnerable; EN-Endangered; CR-Critically Endangered and amphibian species endemic to single hotspots, endemic to any hotspot(s),
and occurring in any hotspot(s), following IUCN (2003) and provisional 2004 listings
Mammals Birds Amphibians
for amphibians. VU-Vulnerable; EN-Endangered; CR-Critically Endangered; EX-Extinct
VU VU VU
+ + + Mammals Birds Amphibians
EN EN EN EN EN EN
+ + + + + + VU VU VU
+ + +
Hotspot CR CR CR CR CR CR CR CR CR
EN EN EN EN EN EN
Tropical Andes 15 (68) 8 (22) 4 (6) 101 (149) 58 (68) 12 (13) 255 (315) 177 (200) 75 (84) + + + + + +
Tumbes-Chocó-Magdalena 6 (30) 1 (5) 1 (2) 21 (48) 10 (16) 5 (6) 7 (31) 4 (15) 0 (5) CR CR CR EX CR CR CR EX CR CR CR EX
Atlantic Forest 20 (37) 13 (17) 5 (7) 63 (86) 39 (45) 10 (14) 54 (58) 36 (37) 15 (15) Total number worldwide 1 128 514 180 74 1 186 513 182 129 1 543 891 341 25
Cerrado 4 (21) 1 (4) 0 (0) 7 (24) 3 (8) 3 (5) 3 (4) 3 (3) 2 (2) Single-hotspot endemics 529 278 103 53 798 386 139 107 1 143 718 278 18
Chilean Winter Rainfall-Valdivian Forests 3 (10) 0 (3) 0 (0) 6 (8) 1 (2) 1 (1) 14 (17) 7 (9) 4 (5) % of all species in taxon 47 54 57 72 67 75 76 83 74 81 82 72
Mesoamerica 29 (48) 20 (27) 4 (5) 32 (42) 15 (17) 5 (5) 178 (231) 123 (155) 59 (67) Endemic to any hotspot(s) 564 288 107 53 867 401 142 107 1 221 756 288 18
Madrean Pine-Oak Woodlands 2 (25) 2 (12) 0 (3) 2 (10) 2 (3) 1 (1) 39 (99) 32 (69) 13 (24) % of all species in taxon 50 56 59 72 73 78 78 83 79 85 84 72
Occurring in any hotspot(s) 809 366 119 53 1 039 439 152 107 1 367 823 307 18
Caribbean Islands 18 (18) 8 (8) 2 (2) 48 (52) 28 (28) 10 (10) 135 (135) 101 (101) 39 (39)
% of all species in taxon 72 71 66 72 88 86 84 83 89 92 90 72
California Floristic Province 4 (7) 3 (3) 1 (1) 4 (9) 2 (3) 2 (3) 7 (12) 0 (3) 0 (0)
Guinean Forests of West Africa 31 (48) 19 (26) 3 (3) 31 (36) 12 (13) 4 (4) 41 (60) 32 (47) 11 (14)
IUCN (2003), although using provisional 2004 listings threatened species identified above. The sites meeting
Cape Floristic Region 1 (10) 1 (5) 1 (2) 0 (9) 0 (0) 0 (0) 7 (11) 4 (6) 2 (3)
Succulent Karoo 1 (11) 0 (4) 0 (2) 0 (9) 0 (0) 0 (0) 1 (2) 0 (0) 0 (0)
for amphibians. these criteria are called “key biodiversity areas,” and
Maputaland-Pondoland-Albany 2 (21) 1 (6) 0 (1) 0 (10) 0 (2) 0 (0) 6 (10) 5 (7) 1 (2) Overall, 56% of the world’s EN and CR mammals become targets for conservation action on the ground.
Coastal Forests of Eastern Africa 6 (20) 6 (11) 2 (4) 2 (17) 2 (5) 0 (0) 3 (16) 2 (4) 0 (0) are endemic to the hotspots (59% of just the CR An important characteristic of the process of defining
Eastern Afromontane 48 (78) 20 (28) 14 (14) 35 (45) 15 (22) 2 (3) 39 (59) 20 (27) 2 (4) mammals). The equivalent percentages for birds are key biodiversity areas is that it is a locally led process
Horn of Africa 10 (27) 4 (10) 3 (5) 9 (21) 5 (5) 3 (3) 1 (1) 0 (0) 0 (0) 78%, and for amphibians 85%. This represents three- —although strictly following global standards— and is,
Madagascar and the Indian Ocean Islands 50 (51) 27 (27) 11 (11) 54 (55) 30 (30) 15 (15) 58 (58) 28 (28) 8 (8) quarters of all Critically Endangered and Endangered therefore, a “bottom-up” process organizationally as
Mediterranean Basin 15 (40) 5 (11) 4 (5) 7 (17) 5 (8) 2 (4) 12 (15) 5 (5) 1 (1) terrestrial vertebrates; in absolute numbers, this adds well as ecologically. A further advantage is that this
Caucasus 2 (14) 2 (3) 1 (1) 0 (8) 0 (2) 0 (1) 2 (1) 1 (1) 0 (0) up to 1 445 CR and EN terrestrial vertebrate species process links to BirdLife International’s Important Bird
Irano-Anatolian 3 (16) 3 (5) 0 (0) 0 (9) 0 (1) 0 (0) 4 (5) 1 (1) 0 (0) found only in the hotspots. These numbers empha- Areas process, which has been under way for more
Mountains of Central Asia 3 (19) 1 (7) 0 (2) 0 (9) 0 (1) 0 (0) 1 (2) 1 (1) 0 (0) size the need for immediate conservation action to than a decade (e.g., Fishpool and Evans 2001), and
Western Ghats and Sri Lanka 14 (35) 11 (16) 3 (4) 10 (25) 2 (6) 0 (2) 95 (100) 79 (81) 33 (33) take place in the hotspots if the rate of extinction is covers the identification of key biodiversity areas for
Himalaya 4 (46) 3 (17) 2 (3) 8 (43) 2 (6) 1 (2) 21 (30) 5 (7) 1 (1) to be slowed. birds. To date, processes for identifying key biodiversi-
Mountains of Southwest China 3 (39) 3 (16) 1 (1) 1 (23) 0 (1) 0 (0) 25 (35) 8 (11) 2 (2) ty areas have been initiated in more than half the
Indo-Burma 25 (70) 17 (31) 9 (12) 16 (70) 9 (18) 2 (4) 41 (66) 10 (18) 3 (4) hotspots.
Sundaland 60 (81) 34 (40) 12 (14) 40 (59) 14 (18) 10 (12) 54 (57) 13 (13) 3 (3) Beyond Hotspots: Identifying Conservation A small, but extremely important, subset of key bio-
Wallacea 44 (50) 14 (16) 1 (1) 49 (51) 27 (27) 7 (7) 5 (7) 0 (0) 0 (0) Targets on the Ground diversity areas are those that hold threatened species as
Philippines 47 (49) 20 (21) 7 (7) 55 (61) 21 (22) 11 (11) 13 (14) 4 (4) 0 (0) endemics to a single site. To tackle these extraordinari-
Japan 21 (24) 15 (17) 3 (3) 10 (27) 3 (9) 2 (2) 14 (14) 13 (13) 2 (2) While hotspots and other global prioritization systems ly high-site conservation priorities, an Alliance for Zero
Southwest Australia 6 (10) 3 (4) 1 (1) 3 (5) 1 (1) 0 (0) 3 (3) 1 (1) 1 (1) are extremely important in informing the flow of con- Extinction (AZE; www.zeroextinction.org) of conserva-
East Melanesian Islands 20 (23) 5 (5) 4 (4) 32 (35) 6 (6) 2 (2) 3 (3) 0 (0) 0 (0) servation resources, they do not provide any guidance tion organizations has formed over the last year. AZE On pp. 62-63, the Critically
New Zealand 2 (2) 0 (0) 0 (0) 57 (62) 22 (22) 5 (5) 4 (4) 2 (2) 1 (1) as to how and where within these large regions conser- aims to identify and conserve all sites holding the en- Endangered golden-crowned sifaka
New Caledonia 3 (3) 2 (2) 0 (0) 7 (10) 7 (7) 3 (3) — — — (Propithecus tattersalli) at the edge
vation should be focused on the ground. Thus, a further tire global population of one or more Critically Endan-
Polynesia-Micronesia 7 (8) 6 (6) 4 (4) 88 (93) 45 (46) 21 (21) 1 (1) 1 (1) 0 (0) of an area devastated by gold
crucial advantage bestowed by the IUCN-SSC species gered or Endangered species. An initial draft of site
mining activities in the Daraina
assessments is the ability to move from the global to the identification has now been completed for terrestrial region of northeastern Madagascar.
to— hotspots. For the three groups for which assess- phibians are tropical mountains, especially in the local scale of conservation planning. In Conservation vertebrate species, revealing 365 AZE sites worldwide. This species, which was not
ments of distribution and conservation status have Neotropics —the Tropical Andes and Mesoamerica— International, this planning process is known as estab- A very high proportion of these sites —nearly 80%— described by science until 1988,
been conducted, we can measure these proportions although the Caribbean Islands and the Western Ghats lishing targets for conservation outcomes (Conserva- fall within the hotspots. The AZE sites therefore repre- now has a very small remaining
with a high level of accuracy (Table 11). The groups and Sri Lanka also hold major concentrations of tion International 2004). We define conservation out- sent refugia for species that could potentially repopu- range, but has been the focus of a
comprehensive protected area
show rather different patterns of threat across the threatened amphibians. Table 12 summarizes the comes at three scales of ecological organization: species late areas of restored habitat. So ensuring the continued
program by the Malagasy
hotspots. Threatened birds are concentrated in the is- overall numbers and percentages of Extinct, Critically (where we strive for “Extinctions Avoided” outcomes); existence of these unique, highly threatened nodes is a conservation organization
land hotspots (although the Tropical Andes and At- Endangered, Endangered, and Vulnerable mammals, sites (where the targets are “Areas Protected” out- vital component of conserving the overall biodiversity Association Fanamby.
lantic Forest are very important as well), with the birds, and amphibians endemic to single hotspots, en- comes); and landscapes (at which our aims are “Corri- of each hotspot. © Pete Oxford/naturepl.com
Caribbean Islands, Madagascar and the Indian Ocean demic to any hotspot(s), or occurring in any hotspot(s). dors Consolidated” outcomes). While the achievement of species- and site-scale
Islands, Sundaland, Wallacea, and the Philippines all In all cases, there are many more threatened species Targets for “Extinctions Avoided” outcomes apply to conservation outcomes is essential if conservation is to
having large numbers of threatened species present occurring within or endemic to the hotspots than we those species facing the highest risk of extinction, and succeed, it is not sufficient. A large body of ecological
and endemic. Not surprisingly, the same holds true for expect based on the equivalent statistics for all spe- are listed as threatened on the IUCN Red List. The es- literature demonstrates that conservation action is
threatened mammals, but forested mainland regions cies. Consistently, more than two-thirds of threatened tablishment of targets for “Areas Protected” outcomes also necessary at the landscape scale, for the purposes
also harbor large proportions of threatened mammals, species are single-hotspot endemics, and approaching is more complex, and requires the careful evaluation of maintaining ecological and evolutionary processes
especially the Atlantic Forest, Mesoamerica, the Guin- 90% of threatened species occur in at least one hot- of sites that are actually or potentially managed for on which species and sites depend, mitigating region-
ean Forests of West Africa, and the Eastern Afromon- spot. Throughout the book, we include annotation of conservation against a set of standard criteria (based al-scale threats (Sanderson et al. 2003), and more ef-
tane Hotspot. The hottest hotspots for threatened am- threat status (VU, EN or CR) in parentheses following on vulnerability and irreplaceability) that focus on the ficiently exploring conservation and development
64 65
 options. The Wildlife Conservation Society has been
making some progress with the measurement of land-
scape-scale conservation targets for wide-ranging
species through their “landscape species” concept
(Sanderson et al. 2002), while other work is beginning
to address abiotic ecological processes (e.g., Cowling et
al. 1999).
not optional. We utterly reject a triage approach of
abandoning the hotspots —it would signal the end of half
of our biodiversity. Instead, we see the successes of the
last fifteen years as a rallying cry for a tenfold increase
in conservation attention, resources, and funding re-
ceived by the hotspots. Nothing less than the diversity
of life on Earth hangs in the balance.

Coda: Hotspots as a Line in the Sand RUSSELL A. MITTERMEIER1

On pp. 66-67, Tijuca National Park,
the largest urban forest in the
world, is entirely surrounded by the
city of Rio de Janeiro, one of the
largest cities in the world. More
than 300 million people live within
10 km of protected areas in the
biodiversity hotspots.
© Frans Lanting/Minden Pictures
On the opposite page, a young
Tibetan woman in the northern
part of China’s Yunnan Province.
© Patricio Robles Gil/Sierra Madre
PATRICIO ROBLES GIL4
Overall, three major conclusions emerge from this up- MICHAEL HOFFMANN 2
dated hotspots analysis. First, while the importance of JOHN PILGRIM 2
the additions to the hotspots list and of the advances THOMAS BROOKS 2
provided by matching hotspots to ecoregion boundaries CRISTINA G. MITTERMEIER 1
and by updated species data are not to be underesti- JOHN LAMOREUX 3
mated, it is clear that the hotspots concept is solidify- GUSTAVO A.B. DA FONSECA1, 2, 13
ing. Compared to the initial conceptual advances of
Myers (1988) and the data synthesis of Myers et al. KEITH ALGER 1
(2000), this update results in few major modifications FREDERICK BOLTZ 1, 14
to the broad global picture of hotspots. While the 25 KATRINA BRANDON 2
hotspots identified by Myers et al. (2000) held 44% of AARON BRUNER 2
all plants and 35% of terrestrial vertebrates as single- JOSÉ MARIA CARDOSO DA SILVA5
hotspot endemics, the sum of single-hotspot endemics ASSHETON CARTER 1
to the 34 hotspots considered here gives increases of ROBERTO CAVALCANTI 1, 15
only 6% and nearly 2% respectively, to yield totals DON CHURCH 1
of 50% of all plants and 37% of all terrestrial vertebrates. MATTHEW FOSTER 2
Second, the amount of biodiversity contained in the CLAUDE GASCON 1
hotspots is extremely high. More than half of the plan- LARRY GORENFLO 2
et’s species are endemic to only 16% of its land area. BRIAN GRATWICKE 6
Based on the evidence from terrestrial vertebrates, it MARIANNE GUERIN-MCMANUS 1
seems that the overall number of species occurring in LEE HANNAH 2
the hotspots is much greater —approaching four-fifths. DAVID KNOX 2
If we consider only the extent of remaining habitat WILLIAM R. KONSTANT 1, 7
—2.3% of the planet’s land surface— these numbers are THOMAS LACHER 2
even more remarkable, although it must be kept in PENNY LANGHAMMER 2
mind that in the long term we aim to attempt to restore OLIVIER LANGRAND 1
much of this degraded land, in addition to the immedi- NICHOLAS LAPHAM 1
ate priority of conserving that habitat which still per- DAN MARTIN 1
sists. By any measure, the hotspots stand out from the NORMAN MYERS 8
rest of the world: even the high biodiversity wilderness PIOTR NASKRECKI 2, 9
areas, the next greatest concentrations of biodiversity MICHAEL PARR 10
(and, admittedly, less well known than the hotspots), DAVID PEARSON 11
hold only 17% of plants and 8% of terrestrial verte- GLENN PRICKETT 1
brates as endemics, in 6.1% of Earth’s land area. Hot- DICK RICE 2
spots hold a concentration of endemics about three ANTHONY RYLANDS 2
times as great for plants and nearly five times greater WES SECHREST 2, 12
for vertebrates. Clearly, if the conservation community MICHAEL LEONARD SMITH 2
is to succeed in preventing extinctions and maintaining SIMON STUART 12
the full range of global biodiversity, a very strong focus JORGEN THOMSEN 1
on hotspots is essential. MICHAEL TOTTEN 1
Finally, and most importantly, hotspots provide us JUSTIN WARD 1
with the real measure of the conservation challenge.
Unless we succeed in conserving this small fraction of
the planet’s land area, we will lose more than half of
our natural heritage —regardless of how successful con-
servation is outside of the hotspots. While conservation
in the hotspots is complex, expensive, and difficult, it is Note: The authors’ addresses are listed on p. 390.

68
 AN UPDATE
OF EXISTING HOTSPOTS
In our last review of the hotspots (Mittermeier
et al. 1999; Myers et al. 2000), we presented in-
formation on 25 hotspots. However, we also
noted that there were a number of regions that
were possible hotspot candidates which could
not be included due to a lack of adequate in-
formation at that time. In this reanalysis of
the hotspots, which began in 2000, we have as-
sessed a number of these areas and have found
that some of them qualify for hotspot status.
In addition, we have reassessed and updated
the previously defined 25 hotspots, providing
new data and making minor changes in the
borders of several. For the most part, however,
23 of them have remained the same.
In the case of the remaining two hotspots in-
cluded in Mittermeier et al. (1999) and Myers
et al. (2000), we have made changes signifi-
cant enough to warrant their inclusion as new
chapters. The original Eastern Arc Mountains
and Coastal Forests of Tanzania and Kenya
Hotspot has undergone major changes. The
Coastal Forests have been separated out as a
new hotspot and extended considerably to the
north into northern Kenya and Somalia, and
south into southern Mozambique (p. 231). The
Eastern Arc Mountains, which on their own
do not qualify as a hotspot, have now been in-
cluded in a more extensive Eastern Afromon-
tane Hotspot based on botanical affinities first
recognized by White (1983) in his classic work
on the botany of Africa. This new hotspot also
incorporates the Southern Rift, the Albertine
Rift, and the Ethiopian Highlands, and is pre-
sented on p. 241.
The original Indo-Burma Hotspot also war-
ranted redefinition to recognize the two dis-
tinct, although overlapping, regions that it
contained. Consequently, it, too, was modified
to such an extent that the two areas had to be
separated as new hotspots. The bulk of this
hotspot, including Myanmar (Burma), Viet-
nam, Cambodia, Laos, Thailand, extreme north-
ern Malaysia north of the Kangar-Pattani Line,
the karst region of extreme southeastern Chi-
na, and the island of Hainan, remains in what
we continue to call the Indo-Burma Hotspot.
Areas to the northwest in northeast India,
Bhutan, Nepal, and southwestern China have
now been included in a new Himalaya Hot-
spot, which also extends further to the west
into Pakistan and northeast Afghanistan than
did the Himalayan section of the original
Indo-Burma Hotspot. These two new hotspots
are covered on pp. 309 and 323.
For the other 23 original hotspots, we pre-
sent here a review of their geographic extent,
with details of minor changes in borders, a
small map, their principal distinguishing char-
acteristics, their biodiversity, threats to their
survival, and a discussion of conservation
measures already in place —with particular
emphasis on protected area coverage. In par-
On the opposite page, the
ticular, we provide updated information on orang-utans are endemic to the
their species richness and endemism in vascu- Sundaland Hotspot, and are
lar plants, mammals, birds, reptiles, and am- declining everywhere. The
Sumatran orang-utan is now
phibians; add new information on freshwater
considered a distinct species from
fishes; and, for the first time, also include data the Bornean orang-utan and is
on endemism at the genus and family levels. Critically Endangered.
© Anup Shah/naturepl.com
This data is combined with data on the new
hotspots and presented in the text and tables of
the Introduction.
71
 TROPICAL ANDES
JOSÉ VICENTE RODRÍGUEZ-MAHECHA 25 • PAUL SALAMAN 25
PETER JØRGENSEN 26 • TRISHA CONSIGLIO 26
EDUARDO FORNO 107 • ANTONIO TELESCA 105 • LUIS SUÁREZ 27
FABIO ARJONA 25 • FRANKLIN ROJAS 106
ROBERT BENSTED-SMITH 27 • VICTOR HUGO INCHAUSTY 107
0 1000 km
ATLANTIC
VENEZUELA OCEAN
COLOMBIA
ECUADOR
PERU
BRAZIL
PACIFIC
OCEAN
BOLIVIA
PARAGUAY
CHILE
ARGENTINA
The Tropical Andes Hotspot is the richest and most di-
verse biodiversity hotspot on Earth. The Andes Moun-
tain Range, its different cordilleras, its vast array of
slopes and peaks, and its isolated valleys provide for a
multiplicity of microhabitats and climatic conditions
that have led to the evolution of an incredible number
of plant and animal species. The hotspot covers a total of
1 542 644 km2 in the countries of Venezuela, Colombia,
Ecuador, Peru, and Bolivia, and extends a very short dis-
tance into extreme northwestern Argentina. The cen-
terpiece of the hotspot is the tropical portion of the
Andes mountain chain, which runs north to south in
Bolivia, Peru, and Ecuador, then splits into three major
cordilleras in Colombia, and extends still further to the
northeast into the northwestern corner of Venezuela.
The hotspot is bounded roughly by the Tropic of Capri-
corn to the south and by the natural termini of the An-
des to the north in Colombia and Venezuela (including
the isolated Sierra Nevada de Santa Marta in Colombia
and the Cordillera de la Costa montane forests in Vene-
zuela). The western border of the hotspot is marked by
the eastern edge of the Tumbes-Chocó-Magdalena
Hotspot, while on the eastern slope of the Andes, in
Ecuador, Peru, and Bolivia, the border extends down to
500 m, a realistic cutoff point between the Andean
slopes and the Amazonian lowlands. The hotspot is
also taken to include the inter-Andean valleys of the
northern cordilleras in Colombia.
The Tropical Andes is sometimes divided into north-
ern and southern zones, with the border between them
being an arid, east-west valley that coincides roughly
with the Ecuador-Peru border in the far northern por-
tion of Peru (Piura) and extending north into neighbor-
ing Ecuador. At this nexus, called the Marañón Gap or
Huancabamba Depression, altitudes drop to around
500 m, creating one of the most important barriers to
faunal and floral migration in the Andes. This gap also
serves as an east-west corridor between the Amazon
and the Pacific (Gentry 1977, 1990).
The vegetation of the Tropical Andes Hotspot fol-
lows a gradient from lowlands to highlands, with tropi-
cal wet and moist forests occurring at 500-1 500 m;
cloud forest formations of various kinds, variously re-
ferred to as yunga, ceja de selva, or ceja de montaña,
which can range in altitude from 800 to 3 500 m (and
covering an area of approximately 250 000 km2 in Peru
alone); and grassland and scrubland systems, which
are mainly paramos in the northern Andes and the dri-
er puna in the southern Andes. Both of the latter begin
at 3 000 to 3 800 m and extend up to between 4 200 and
4 800 m, usually ending at the snowline. Beginning in
the lowlands of the eastern slopes at around 500 m alti-
tude, the sub-Andean forests are similar to those of the
hot, Amazonian lowlands, but have fewer palm species,
fewer lianas, and fewer buttresses, although the canopy
can reach as high as 45 m. Within the sub-Andean for-
est belt, vegetation begins to transition at around
1 500 m, at which point the plant family Lauraceae be-
comes the dominant element (Cuatrecasas 1958; Lan-
gendoen and Gentry 1991; Dodson and Gentry 1991).
Andean forests then begin at approximately 2 000 m,
and are characterized by shorter trees and more abun-
dant epiphytes such as mosses, lichens, ferns, and al-
gae. At 3 000 to 3 800 m, Andean forests then give way
to paramos in the north and puna in the south.
In addition to the main Andean vegetation types,
other systems such as dry forests and arid, warm to cool
non-forest environments —including woodlands, cactus
stands, thornscrub, and matorral— occur adjacent to
the wet montane, paramo and puna formations, in dry
intermontane basins or along the dry Andean slopes of
Peru in particular, usually at altitudes of 2 000 to
3 000 m. Polylepis forests represent another vegetation
type unique to the Andes, this tree genus being re-
stricted to the montane areas of western South Ameri-
ca, and a conspicuous element of some high-elevation
tropical habitats.
In terms of plants and vertebrates, the Tropical An-
des Hotspot leads virtually all others in both species di- On the opposite page, cloud forest in
versity and endemism. Perhaps the most impressive La Planada Nature Reserve, an
figures are those for vascular plants, with an estimated important privately managed
protected area in the southern
30 000-35 000 species, or approximately 10% of the
Colombian Andes.
world’s species, occurring in this hotspot. In fact, © Patricio Robles Gil/Sierra Madre
the Tropical Andes contains 20%-80% of the total plant
species occurring in Bolivia, Ecuador, Peru, and Vene-
zuela. Endemism is equally impressive, with an esti-
mated 50% (and perhaps 60% or higher) of species en-
demic to the hotspot (around 15 000 species), and
peaks in the number of endemic species occurring in
73

The adult male Andean
cock-of-the-rock (Rupicola peruviana),
a spectacular species found in the
Andean Mountains from Venezuela
south to Bolivia, spends much of its
time at communal courtship sites,
called leks, where it puts on
displays for females.
© Patricio Robles Gil/Sierra Madre
On the opposite page, violet-tailed
sylph (Aglaiocercus coelestis), one
of the many hummingbird species
found in the Andes, the world’s
center of hummingbird diversity.
© Patricio Robles Gil/Sierra Madre
74
the Andean regions of each country (Jørgensen and
León-Yánez 1999; Kessler 2000, 2002; Van der Werff
and Consiglio 2004). For example, almost 3 000 of
Ecuador’s 4 000 endemic plant species and around
3 650 of Peru’s 5 400 endemic species are Andean; over
25% of total endemic species for Peru and Ecuador oc-
cur in the 2 500-3 000-m elevation zone alone. These
figures are likely to be an underestimate, especially as
new taxa are being described; for example, there have
been about 440 plant species described between 1999
and 2003 from the Ecuadorian portion of the Tropical
Andes Hotspot alone (out of a total of 532 for the coun-
try as a whole) (D. Neill, pers. comm.). In addition, for
the Orchidaceae, the largest family in Peru and one that
has its peak of endemism in the Tropical Andes, it is es-
timated that an increase of almost 50% of known
species has occurred in the last 10 years (C. Dodson,
pers. comm.). It is likely that we need five times the
number of plant collections that have been carried out
to date to be reasonably certain of the region’s plant di-
versity. Several flagship plant species also occur in the
Tropical Andes. Among the list of endemics is a high
Andean bromeliad species (Puya raimondii) that takes
as long as a century to reach maturity and has the
tallest inflorescence of any plant on Earth, reaching as
much as six meters in height.
The Tropical Andes also has the highest bird diversi-
ty and endemism of any hotspot, perhaps not surpris-
ing given that Colombia, Peru, and Ecuador hold the
first, second, and fourth positions on the global list of
countries with the most avian species. Furthermore,
these high numbers of birds derive largely from the An-
dean portions of these countries (Stotz et al. 1996; Mit-
termeier et al. 1997). The total number of regularly oc-
curring bird species for the Tropical Andes Hotspot is
1 728, of which an astounding 584 species are endemic.
Furthermore, 69 genera are endemic. It is not surpris-
ing, then, that BirdLife International recognizes around
21 different Endemic Bird Areas (EBAs) lying partly or
entirely within this hotspot, among which the Colom-
bian East Andes, with 34 species endemic, is the most
important, having one Extinct (Colombian grebe, Podi-
ceps andinus) and four Critically Endangered species:
the gorgeted wood-quail (Odontophorus strophium),
chestnut-bellied hummingbird (Amazilia castaneiven-
tris), Niceforo’s wren (Thryothorus nicefori), and Colom-
bian mountain grackle (Macrogelaius subalaris). Flag-
ship bird species occurring in this hotspot include the
yellow-eared parrot (Ognorhynchus icterotis, CR), an
enigmatic macaw-sized species that depends on the
Quindío wax palm (Ceroxylon quindiuense, VU), which
is the national tree of Colombia; the Andean condor
(Vultur gryphus), one of the largest flying birds on
Earth; and the diminutive hummingbirds (Trochilidae),
a family of tiny, jewel-like birds that reaches its greatest
diversity in the Tropical Andes.
Endemism and diversity among amphibians and rep-
tiles in the Tropical Andes exceed even the amazing fig-
ures for birds and plants. The Tropical Andes have by
far the highest amphibian diversity of any hotspot on
Earth, with a total of 1 155 species (1 088 frogs and
toads; 28 salamanders, newts and sirens; and 39 caecil-
ians), of which 664 species are endemic. Some of the
amphibian genera reach amazing levels of diversity in
the Tropical Andes, the best example being the wide-
spread genus Eleutherodactylus of the family Lepto-
dactylidae, with 343 species present and 244 endemic.
There are also 10 endemic genera (of the 79 represent-
ed). Unfortunately, this is also a hotspot for amphibian
extinctions, with several taxa already having disap-
peared in recent years, particularly some beautiful har-
lequin toads of the genus Atelopus that tend to be
stream-dwelling species and appear highly sensitive to
local climate change and habitat loss, and are suscepti-
ble to disease (Ron et al. 2003). In terms of reptiles,
there are 610 species native to the Tropical Andes
Hotspot (304 lizards, 294 snakes, eight turtles and tor-
toises, and four crocodilians), of which 275 species and
three genera are endemic.
Mammal diversity and endemism are also notewor-
thy. Of a total of 569 species, some 75 are endemic. As
elsewhere, rodents are the most diverse mammal group
with 220 species, followed by bats with 181 species.
There are curious pockets of higher endemism within
certain habitats of this hotspot, with both the puna and
paramo formations having high mammal endemism.
There are also six endemic genera, each represented by
single species: Garlepp’s mouse (Galenomys garleppi),
the Andean rat (Lenoxus apicalis), little or mountain
coati (Nasuella olivacea), puna mouse (Punomys lemmi-
nus), and fish-eating rat (Anotomys leander, EN), a spe-
cies known only from the Andes of northern Ecuador
and highly specialized for an aquatic existence. The
sixth endemic genus is one of the most important mam-
mal flagship species for the Tropical Andes, the yellow-
tailed woolly monkey (Oreonax flavicauda, CR). It is the
largest mammal endemic to Peru, and is only one of
three primate genera in the Neotropics to be endemic
to a single country. It is restricted to a small area of
cloud forest in the northern Peruvian departments
of Amazonas and San Martín. Other important large
mammal flagships include the spectacled bear (Tre-
marctos ornatus, VU), woolly or mountain tapir (Tapirus
pinchaque, EN), and vicuña (Vicugna vicugna), a cam-
elid that lives at altitudes of 3 000 to 4 800 m, mainly in
the puna ecosystem.
Freshwater fishes are represented by 380 document-
ed species, although many more are expected as explo-
rations extend onto the Amazonian flanks of the moun-
tains. A total of 131 fish species are endemic to the
hotspot, a surprising number for an area centered on
the crest of a mountain range. One major component of
endemism consists of members of the cyprinodont
genus Orestias, which has undergone a significant radi-
ation in Lake Titicaca and nearby drainages, resulting
in a cluster of 43 species endemic to the southern end
of the hotspot. Most remarkable perhaps are the naked
sucker-mouth catfishes (family Astroblepidae) that
inhabit torrential streams from one end of the hotspot
to the other. With the exception of a species in Panama
and a few species that extend to lower elevations, the
family’s 90 species are endemic to the region.
Humans have lived in the Andes for many millennia,
and the region was home to one of the world’s great past
civilizations, the Empire of the Incas. However, the de-
gree of human impact varies considerably within this
region, from areas that have lost almost all of their orig-
inal vegetation (e.g., some of the inter-Andean valleys
in Colombia and Ecuador) to some that are still largely
pristine (e.g., parts of the eastern slopes of the Andes in
Bolivia and Peru). Broadly speaking, the most disturbed
portions of the hotspot are the dry inter-Andean valleys,
where the original forests have largely disappeared and,
on average, less than 10% remains. The inter-Andean
valleys provide the most hospitable environment for
humans in the region, and these areas have been dense-
ly populated since pre-Columbian times.
Other heavily impacted ecosystems within the Andes
are the paramos and the puna. Both have been greatly
modified by seasonal burning and grazing, agriculture,
and mining. The puna also suffers from over-exploita-
tion of certain slow-growing woody plant species for
firewood, especially around urban centers such as Are-
quipa, while puna ecosystems of both Peru and Bolivia
are also affected by the mining industry, with toxic
runoff and water contamination being a major concern.
An additional threat that has emerged in recent
years, especially in the high Andean forests of Colom-
bia, is the cultivation of the opium poppy, in clearings
cut within montane forests to grow this highly prof-
itable illegal crop. Unfortunately, the programs de-
signed to control illegal crops use chemical defoliants
that cause even more damage to biodiversity, as well as
allowing harmful chemical herbicides to enter into
highland ecosystems and trickle into lower altitudes
new and better-managed parks, but also to interconnect
existing parks through what are commonly referred to
as “corridors.” An example of a “corridor” project, and a
model for using corridors as a conservation strategy in
the Andean region, can be found in the Amazon-Andes
interface in southern Peru and adjacent portions of Bo-
livia. One of the first and most important components
of this corridor is Manú, the largest rainforest biosphere
reserve on Earth at 18 812 km2, and which protects ma-
jor areas of puna, cloud forest, and lowland forest. An-
other major piece of this string-of-pearls of Andean
slope ecosystems is the Tambopata-Madidi protected
area complex straddling the Peru-Bolivia border, repre-
senting fully 22 250 km2 of new parks created in the
richest portion of the most diverse biodiversity hotspot
on Earth, an area larger than El Salvador and a very sig-
nificant accomplishment for biodiversity conservation.
The Tropical Andes has benefited from a series of
major conservation investments in the last several
years. For example, the Critical Ecosystem Partnership
Fund has made a commitment to invest some $6 mil-
lion in the Vilcabamba-Amboró Corridor in Peru and
Ecuador over the past three years, with the specific ob-
jective of building civil society capacity to carry out
biodiversity conservation activities in this important
region. At the same time, the Global Conservation
Fund (GCF), based at Conservation International, has
invested $1.273 million in projects in Bolivia, Colom-
bia, Peru, and Ecuador. These projects have helped to On pp. 76-77, Machu Picchu
stimulate the creation of some nearly 3 million hect- Historic Sanctuary, located within
ares of new protected areas in some of the highest- the Vilcabamba-Amboró
Conservation Corridor, is
priority regions of these countries. Included among
surrounded by striking mountains.
the projects supported by the GCF were two debt-for- Machu Picchu receives more than
nature swaps in Peru in 2003 and in Colombia in 2004. 400 000 visitors per year, who come
Both of these were carried out in partnership with the to enjoy one of the most magical
World Wildlife Fund, The Nature Conservancy, and landscapes in the Peruvian Andes.
the United States Government; they will provide $10.6 © Haroldo Castro

through the rivers and streams, a factor that may have
contributed to amphibian and freshwater fish die-offs in
these regions.
As a result of all these pressures, a large portion of the
natural vegetation of the Tropical Andes Hotspot has al-
ready been lost, and it is estimated that the area re-
maining in fully intact condition is likely no more than
25%, or 385 661 km2, and probably much less. The east-
ern slopes of the Andes in Peru and Bolivia have the
most extensive areas of largely intact natural vegetation.
Despite the bleak picture painted above, protected
areas today are conserving some of the most important
remnants of the Tropical Andes Hotspot. In total, these
protected areas cover some 16% of its original extent.
However, many of these protected areas are small and
ineffective, and only 7.9% of the hotspot is protected in
reserves or parks in IUCN categories I to IV. This has
led to the recognition for the need not only to create
million to 10 sites in Peru and $10 million to five sites
in Colombia over a 12-year period. These are just a few
examples of the direct support given to conservation in
recent years.
Other conservation activities in the region are focus-
ing directly on amelioration of some of the most dan-
gerous threats to the environment, e.g., infrastructure
development, while several on-the-ground efforts to
conserve key threatened species (such as the recent re-
covery of the yellow-eared parrot) are also under way.
In conclusion, looking at this region as a whole, there is
considerable room for optimism. Although portions of
the Tropical Andes have been heavily impacted, ex-
tinctions have been relatively few, and there is still
enough time to design and implement conservation ar-
eas at a scale which is likely to ensure the survival of
the vast majority of life-forms that exist in this, the
richest of the planet’s terrestrial hotspots.
On the opposite page, the spectacled
bear (Tremarctos ornatus) is the
only representative of the bear
family in South America, and is
distributed in forested habitats at
higher elevations from Venezuela
and Colombia south through the
Andean mountain range to Bolivia.
© Patricio Robles Gil/Sierra Madre
Above, the booted racket-tail
(Ocreatus underwoodii) is a fairly
common and widespread
hummingbird species occurring
from Colombia to Peru.
© Patricio Robles Gil/Sierra Madre

79

On the opposite page,
the rainforests of the Chocó are
some of the wettest and most diverse
on the planet.
© Patricio Robles Gil/Sierra Madre
80
TUMBES-
CHOCÓ-MAGDALENA
JOSÉ VICENTE RODRÍGUEZ-MAHECHA 25 • PAUL SALAMAN 25
PETER JØRGENSEN 26 • TRISHA CONSIGLIO 26 • LUIS SUÁREZ 27
FABIO ARJONA25 • ROBERT BENSTED-SMITH 27
0 500 km
PANAMA
PACIFIC
OCEAN
Malpelo
Island COLOMBIA
ECUADOR
PERU
The Tumbes-Chocó-Magdalena Hotspot was previously
referred to as the Chocó-Darién-Western Ecuador Hot-
spot in Mittermeier et al. (1999). It has now changed its
name and expanded its boundaries to include several
new areas, most notably the Magdalena Valley in
Colombia. As we now define this hotspot, it originally
covered 274 597 km2 in the northwestern corner of
South America. It begins east of the Panama Canal in
the wet and moist forests of Panama’s Darién Province,
extends south through the Chocó region of western
Colombia, and then on into the moist forests of north-
western Ecuador, where it is bounded by the Pacific
Ocean to the west and the western slope of the Andes
to the east. It then also extends still further south to in-
clude the dry forests of western Ecuador and those in
Tumbes, Piura, and La Libertad departments in the ex-
treme northwestern corner of Peru, south as far as
Huacho. In northern Colombia, the hotspot also follows
the forests of the Chocó as they go east around the
northern Andean termini and into the Magdalena Val-
ley. In addition to the mainland portion of the hotspot,
we also include here the island of Malpelo (8 km2), lo-
cated around 500 km off the coast of Buenaventura,
Colombia, and the Galápagos Islands (7 882 km2), lying
some 960 km west of Ecuador in the Pacific Ocean and
including 13 large islands and six smaller islands lying
right on the equator. Although these islands are volcanic
in origin, they do have some floristic affinities with the
mainland, and therefore are included here with the geo-
graphically nearest hotspot.
This hotspot is characterized by a great variety of
habitats ranging from extensive mangrove areas, beach-
es, rocky shorelines, and coastal wilderness and dry
forests, to the world’s wettest rainforests. The dry forest
region of western Ecuador and Tumbes and Piura in
Peru is especially diverse, with habitats ranging from
arid scrub and desert through deciduous tropical thorn-
scrub forest and deciduous Ceiba trichistandra forest,
to semievergreen C. pentandra forest, semievergreen
lowland and premontane tall forest, to deciduous to
semievergreen intermontane scrub. Punctuating the
otherwise flat coastal plain, this hotspot also contains
numerous smaller mountain systems, including the
Serranía del Sapo, Serranía de los Saltos, and Serranía
del Baudo, which run parallel to the coast in extreme
western Colombia; the Cordillera de San Blas and Ser-
ranía del Darién in southeastern Panama; the Serranía
de Abibe and Serranía de San Lucas in northern Colom-
bia; the Cordillera de la Costa in Ecuador; and the Cer-
ros de Amotape in Peru, all of which represent “islands”
of endemism that add to the wide spectrum of biodi-
versity in this top-priority ecosystem.
Broadly speaking, the Tumbes-Chocó-Magdalena Hot-
spot can be divided into two major phytogeographic re-
gions: the wet and moist forest Chocó and Darién bio-
geographic zones in the north and the Ecuadorian and
Peruvian dry forest zone in the south, with a number of
subtle geographic and biological barriers within these.
The variety of ecosystem types in such a limited geo-
graphic area has given rise to high levels of diversity
and endemism, and overall plant diversity in the
hotspot is estimated at 11 000 species; plant endemism
is estimated at 25%, which gives a figure of 2 750 en-
demic species of vascular plants for this hotspot. Plant
diversity in the Colombian portion of the hotspot alone
reaches an estimated 5 000 total species (G. Galeano,
pers. comm.), and it is thought that the Colombian
Chocó is likely to be the most floristically diverse site in
the Neotropics. Based on an assessment of the Missouri
Botanical Garden’s TROPICOS Database, it has been es-
timated that we need five times the number of plant
collections that have been made to date to be reason-
ably certain of the region’s plant diversity. The flora of
the Galápagos Islands is represented by 699 species
of vascular plants, of which at least 177 species are en-
demic (25.3%), and there are six endemic genera of
flowering plants (Jørgensen and León-Yánez 1999; Va-
lencia et al. 2000).
In terms of vertebrate diversity and endemism, the
Tumbes-Chocó-Magdalena Hotspot is impressive. Bird
diversity in the mainland portion of the hotspot is 892
regularly occurring species, with 112 endemics. There
are also 13 endemic bird genera, 10 of which are monotyp-
ic and, with the exception of the spiny-faced antshrike
(Xenornis setifrons, VU), all are represented by species
considered not threatened by BirdLife International,
which is surprising. BirdLife International also consid-
ers this region to be a very high priority, and recognizes
six Endemic Bird Areas (EBAs) in the hotspot as here
defined, including the Nechi Lowlands, the Darién
Lowlands, the Darién Highlands, the Tumbesian Re-
gion, and the Chocó. The Tumbesian Region EBA, with
17 threatened bird species confined entirely to this EBA
(such as the white-winged guan, Penelope albipennis,
and the Peruvian plantcutter, Phytotoma raimondii), is
considered one of the three EBAs most critically in
need of conservation action. The Chocó EBA has a total
of 51 species confined to it, a total second only to the
Atlantic Forest Lowlands EBA (Stattersfield et al. 1998).
The Galápagos Islands form an EBA in their own right,
with 22 endemic terrestrial species including the 12
species of Darwin’s finches so important to Darwin’s
Theory of Evolution. Flagship species include the
bizarre long-wattled umbrella bird (Cephalopterus pen-
duliger, VU) and the blue-black grass quit (Volatinia jaca-
rina), the latter being the common ancestor of the Galá-
pagos finches.
Mammal diversity and endemism are also high in
this hotspot, with 283 species, of which 10 are endem-
ic, five of them on the Galápagos. The rice rats of the
genus Nesoryzomys are confined entirely to the Galápa-
gos Islands. The location of this hotspot at the transi-
tion zone between Central and South America results
in the occurrence of some largely Central American
mammal species (for instance, pocket gophers of the
family Geomyidae) that can not be found elsewhere in
the South American continent. Also among the mam-
mals of this hotspot are a number of important primate
flagships, including three species of spider monkey of
the genus Ateles (A. fusciceps, A. geoffroyi, and A. hy-
bridus, CR) and three species of bare-faced tamarins of
the genus Saguinus: the cotton-top tamarin or mono tití
blanco (S. oedipus, EN), the rufous-naped or Panama-
nian tamarin or bichichi (S. geoffroyi), and the white-footed
tamarin or tití del Chocó (S. leucopus, VU). On the Galá-
pagos Islands, the most recognizable flagship is the
Galápagos Islands fur seal (Arctocephalus galapagoensis,
VU), the smallest of the pinnipeds, with an adult length
of about 1.5 m.
Reptile diversity is quite high in this hotspot, with
an estimated 325 species, of which 98 are endemic
(including 21 species on the Galápagos). The lizard
genus Anolis is particularly well represented, with 42
species present, 30 of them endemic). There are also
five endemic genera including Emmochliophis, for
which the two species, E. fugleri and E. miops, are
known only from a single male and a single female
specimen from western Ecuador; Teuchocercus, with a
single species, T. keyi, from Ecuador; and Trachyboa,
represented by two species of snakes. The remaining
two endemic genera are also among the hotspot’s
most remarkable flagship species, and both occur in
the Galápagos: the marine iguana (Amblyrhynchus
cristatus, VU) and the two threatened species of land
iguana (Conolophus spp.).
Amphibian diversity is even more impressive, with
204 mainland species, of which 29 are endemic, in-
cluding two species of caecilians (Caecilia antioquiaensis
and C. tenuissima). There are no native amphibians on
the Galápagos, although Fowler’s snouted tree frog (Sci-
nax quinquefasciata) has become established on Santa
Cruz. Among the amphibians, the poison arrow or poi-
son dart frogs (Dendrobates spp.) of the family Dendro-
batidae are important flagships. These beautiful, bright-
ly colored little animals secrete toxic alkaloids through
their skin, their bright aposematic coloration serving
to warn predators that they are off limits. One species,
the golden poison frog (Phyllobates terribilis, EN), a
bright yellow species found only in the Río Saija Basin
in the southern portion of the Colombian Chocó, is
among the three most poisonous vertebrates in the
world, and its toxicity is such that the local Emberá In-
dians poison their blowgun darts simply by rubbing
them along the backs of these little frogs. Unfortunate-
ly, many of the known amphibian endemics have very
limited ranges such as isolated ridge tops only a few
square kilometers in extent, making them particularly
vulnerable to extinction.
The coastal watersheds of northwestern South Amer-
ica have rather sparse fish faunas compared to the great
watersheds of the Atlantic versant, and the hotspot con-
tains only 251 species in 54 families. Miocene fossils
from the Magdalena Basin show that the fauna was
richer in the past and that many characteristic ele-
ments of the Amazon/Orinoco fauna were present pri-
or to uplift of the Andes. Isolation following uplift has
contributed to a moderate level of endemism, with 115
endemic species and seven endemic genera, centered
primarily in the Magdalena and Atrato basins. There is
also a single endemic species (Ogilbia galapagosensis)
on the Galápagos.
As a whole, it is estimated that natural vegetation re-
maining in more or less pristine condition in this hot-
spot is approximately 24% of the original extent, with
much of what remains occurring in the Colombian
Chocó and parts of the Darién. Current threats to the re-
gion are the same as in most other hotspots, and range
from direct conversion of land for both large- and small-
scale agriculture, such as banana and African oil palm,
to climate change and elevated ultraviolet radiation im-
pacting amphibians and perhaps other species as well.
Many different kinds of infrastructure development
(such as roads, dams, and canals) are also planned for
this region, and colonization is on the rise in many ar-
eas. Finally, hunting continues to be an issue in some
parts of the hotspot, especially for several of the larger On the opposite page, broad-billed
bird and mammal species. motmot (Electron platyrhynchum),
The degree of threat varies considerably from region a widespread species. This photo
was taken in Río Ñambí Nature
to region. The Ecuadorian portion of this hotspot is un-
Reserve, Colombia.
der the gravest threat at present, with only about 2% © Patricio Robles Gil/Sierra Madre
of the original forest cover remaining. The situation in
the Panamanian Darién is substantially better, with Above, Galápagos land iguana
some 65% of original forest cover still remaining, (Conolophus subcristatus) on Isla
while the Chocó region of Colombia, and especially the Plaza, one of two species of land
iguanas endemic to the Galápagos.
Department of Chocó, remains largely intact. Man-
© Cristina G. Mittermeier
grove ecosystems are under threat throughout this
hotspot due to overexploitation for timber and fuel-
wood, and because they are cleared to give way to
shrimp aquaculture, an activity that has already de-
stroyed many Ecuadorian mangroves. The Galápagos
Islands are severely impacted by invasive alien
83

Agrias claudina photographed in
the State of Santa Catarina, Brazil.
Butterfly species reach a high level
of diversity and endemism in the
Atlantic Forest.
© Luiz Claudio Marigo
On the opposite page, the Critically
Endangered northern muriqui
monkey (Brachyteles
hypoxanthus), here seen drinking
from a stream on the forest floor, is
one of Brazil’s most threatened
primate species, and one of two
primate genera endemic to the
Atlantic Forest Hotspot.
© Luciano Candisani
84
species, and only three of the larger islands are con-
sidered unaltered by humans.
Portions of this hotspot, and especially the Chocó
and western Ecuador, have been considered among
the planet’s highest priorities for conservation for
more than twenty years (e.g., Gentry 1977, 1979). This
has resulted in a wide range of conservation projects
by national governments, multilateral and bilateral
funding agencies, and international and national con-
servation organizations, which have led to the creation
of a range of protected areas and many other conser-
vation efforts of varying success. Presently, approxi-
mately 12.5% of the hotspot is considered protected;
however, only 6.9% of the hotspot is conserved in
IUCN categories I to IV. A large network of indigenous
reserves and comunas (communal Black ancestral lands)
exists throughout the hotspot, an example being that of
several Awá Indigenous Reserves straddling the Colombia-
Ecuador border. Although indigenous reserves do not
necessarily protect the full range of biodiversity as well
as a fully protected national park or biological reserve,
they do have great significance for conservation and
ensure more sustainable use of natural resources than
Western forms of development. The Global Conserva-
tion Fund at Conservation International is currently
supporting several initiatives in northwestern Ecuador,
focused on protection of remaining intact lowland
forests in and around the Cotacachi Cayapas Ecological
Reserve and Awá Ethnic Reserve. Key interventions in
this area include community-supported land acquisi-
tions, purchase of logging concessions, community
land titling, and development of community-based in-
centive agreements for conservation.
The Critical Ecosystem Partnership Fund (CEPF) has
also made considerable investments in the Chocó-Mana-
bí portion of this hotspot. Since this program began in
January, 2002, some $3.3 million have been awarded to
24 different projects focused on field activities and
strengthening local NGO capacity.
Elsewhere, other significant areas have been pro-
posed for protection as well, including such biological-
ly important zones as the Naya Corridor in the south-
central Colombian Chocó (Departments of Valle and
Cauca), the protection of which would help conserve
an altitudinal transect from the peaks of the western
Andean cordillera to the sea, and represent perhaps the
last, best chance to conserve a representative sample of
the Chocoan forest ecosystems.
Last, but not least, the Galápagos have long been a fo-
cus of conservation activity and investment, and the vast
majority of this unique archipelago is now recognized as
both a World Heritage Site and a Biosphere Reserve.
In summary, the Tumbes-Chocó-Magdalena Hotspot
presents several challenges and opportunities. Given
the relatively intact nature of the Chocó and the Da-
rién, there is a better possibility of setting aside large
blocks of forest habitat than in most other hotspots. At
the same time, in the Ecuadorian portion of this hot-
spot, in both the wet and moist forests and the dry
forests, there exists the need for the same kind of im-
mediate salvage-type operations required in some
of the other highly degraded, endemic-rich hotspots
like the lowland forests of the Philippines and the
northeastern portion of Brazil’s Atlantic Forest Region,
which are also at or under 3% of their original extent.
There is no doubt that the future of these Ecuadorian
forests hangs in the balance, and that they must be
placed at or near the top of any list of global biodiversi-
ty conservation priorities.
ATLANTIC FOREST
GUSTAVO A.B. DA FONSECA1, 2, 13 • ANTHONY RYLANDS2
ADRIANO PAGLIA5 • RUSSELL A. MITTERMEIER1
Fernando
de Noronha
Island
BRAZIL
PARAGUAY
ATLANTIC
ARGENTINA OCEAN
URUGUAY
0 600 km
The Atlantic Forest, or Mata Atlântica, is a unique series
of South American rainforest ecosystems quite distinct
from the more extensive Amazonian Forests. It once ex-
tended almost continuously from the states of Rio
Grande do Norte and Ceará in northeastern Brazil south
to Rio Grande do Sul, and included the typical Atlantic
coastal forest on the narrow, low-lying coastal plain, the
forests on the foothills of the Serra do Mar and related
mountain ranges that run roughly parallel to the coast,
and the forests on the slopes of the Serra do Mar itself,
reaching up to 1 800-2 000 m. In northeastern Brazil,
these forests covered only a narrow coastal strip, rarely
exceeding 50-100 km in width, and were replaced in-
land by drier Caatinga and Cerrado formations typical
of the Brazilian interior. The Atlantic Forest also ex-
tends into eastern Paraguay, Misiones in extreme north-
ern Argentina, and narrowly along the coast into
Uruguay. Due to its widespread destruction, rampant
even in the early sixteenth century, it is now difficult to
know the region’s original extent; however, as we define
it here, it originally covered 1 233 875 km2, including
mangroves and restinga, or sandy coastal scrub, and
those portions in Paraguay, Argentina, and Uruguay.
We also include the offshore archipelago of Fernando de
Noronha and its neighbor Atol das Rocas, which belong
to the Fernando de Noronha Chain of submerged vol-
canic mountains in the equatorial Atlantic Ocean
350 km off the coast of northeastern Brazil.
Altitude determines at least three vegetation types
in the Atlantic Forest: the woodlands of the coastal
plain, the slope forests, and the high-altitude woodlands
(Joly et al. 1991). This gradient influences, in turn, the
pattern of richness and composition of the plant com-
munities. Oliveira-Filho and Fontes (2000) found that
increasing altitude affects the composition of woody
plants, augmenting the relative abundance of species of
the families Asteraceae, Melastomataceae, and Solana-
ceae, and reducing the legumes. In the most inland areas
of this biome, the dominant vegetation is semidecidu-
ous, while mixed formations dominated by Araucaria
angustifolia can be found in the coldest southern areas.
Both in the coastal and high-plain areas, the family with
the highest species diversity is the Myrtaceae (more
than 300 species), while the most speciose genus is Eu-
genia (with more than 100 of the 3 500 woody species
found in this biome). Overall, more than 8 000 of an es-
timated 20 000 species of plants (or 40%) are thought to
be endemic. Endemism in trees is particularly high,
with approximately 54% being restricted to the region,
compared to about 30% for higher plants in general
(Mori et al. 1981).
Vertebrate diversity and endemism in this hotspot are
markedly high. There are 263 mammal species known
to occur in the Atlantic Forest, of which 71 are endemic,
including the thin-spined porcupine (Chaetomys sub-
spinosus, VU) and painted tree rat or cocoa rat (Callis-
tomys pictus), which are both representatives of mono-
typic endemic genera, and the maned sloth (Bradypus
torquatus, EN), an unusual and larger relative of the
widespread three-toed sloths (B. tridactylus and B. varie-
gatus). One particularly notable endemic is the Brazilian
arboreal mouse (Rhagomys rufescens, CR), the sole rep-
resentative of an endemic genus, known only from two
specimens collected from Rio de Janeiro, and very like-
ly Extinct. In total, the Atlantic Forest has a remarkable
12 endemic genera, including also two endemic and
threatened primate genera which represent the region’s
primary flagship species, the lion tamarins (Leontopithe-
cus spp.), of which there are four species, and the
muriquis (Brachyteles spp.), with two. Conservation ef-
forts for the lion tamarins have demonstrated the im-
portance of international cooperation between universi-
ties, zoos, national and international NGOs, and the
Brazilian Government; this has resulted not only in their
protection, but also the preservation of their remaining
forests (Kleiman and Rylands 2002). The two Brachyteles
species are the largest of the New World primates. As
with the lion tamarins, conservation efforts and research
focused on them have resulted in numerous initiatives
protecting primate habitat and have also involved train-
ing of many Brazilian conservationists (Strier 1999).
About 936 birds have been recorded from this hot-
spot, 148 of them endemic. There are also 22 endemic
genera. No fewer than four Endemic Bird Areas (EBAs)
have been identified in the region (Stattersfield et al.
1998), the Atlantic Forest Lowlands being one of the
most important in South America, with a total of 52
species and 10 genera confined to this single EBA. More
recent work has also identified 161 Important Bird Ar-
eas in the hotspot (Bencke and Maurício 2002). There
are several outstanding avian flagships, including the
Alagoas curassow (Mitu mitu, EW), a large cracid of
the lowland Atlantic Forest of northeastern Brazil that
was last seen in the wild in 1987 and is now held only
in captivity (Stattersfield et al. 1998), and the related
red-billed curassow (Crax blumenbachii, EN), found fur-
ther south in the Atlantic Forest Lowlands, especially
in the State of Espírito Santo in the large Sooretama Bi-
ological Reserve. The Brazilian merganser (Mergus oc-
tosetaceus, CR) is an important flagship for the southern
Atlantic Forest in Brazil and Misiones. Finally, one of
the most spectacularly colored of all birds is the seven-
colored tanager (Tangara fastuosa, EN), also from north-
eastern Brazil; it has declined mainly due to habitat de-
struction and live capture for the pet trade.
Reptile species in the hotspot number 306, of which
94 species and eight genera (four of them monotypic)
are endemic. The wormlike subterranean reptiles of
the genus Amphisbaena are particularly well represent-
ed, with 20 species, although endemism is low with
only two species endemic, one of them, A. ridleyi, en-
demic to Fernando de Noronha. Endemism is perhaps
most marked in the genus Bothrops, with nine en-
demics of the 17 species present. Amphibian diversity
is very high, with 483 species having been recorded
from the Atlantic Forest, of which a remarkable 286 are
largely endemic. There are also 23 endemic genera; the
most speciose are Cycloramphus with 25 species and
Hylodes with 20, while seven genera are represented by
single species. There is also a single endemic family,
the Brachycephalidae, represented by six species of the
genus Brachycephalus.
Among the remaining vertebrates, there are at least
350 fish species known to occur in the hotspot, of On the opposite page, the channel-
which 133 are endemic; in addition, 10 of the 68 genera billed toucan (Ramphastos
represented are endemic. Invertebrates are not well vitellinus), a widespread species
that occurs in the Atlantic Forest.
known, although 63 of 88 tiger beetle species recorded
© Luiz Claudio Marigo
from the hotspot are endemic (D. Pearson, pers.
comm.), as are 948 of 2 120 butterfly species (Brown Above, the Vulnerable blue-chested
and Freitas 2000). parakeet (Pyrrhura cruentata)
Prince Maximilian zu Wied-Neuwied of Germany was is endemic to the Atlantic Forest
one of the first scientists to travel through the Atlantic of Brazil.
© Luiz Claudio Marigo
Forest. When he carried out his expedition in the early
nineteenth century, he found some of the richest,
tallest, and most impressive forests on Earth. But even
then some regions had already been colonized for more
than two centuries, and were showing the effects of
87

The Endangered black-hooded
antwren (Formicivora
erythronotos) is endemic to the
coastal restingas of the State of
Rio de Janeiro.
© Luiz Claudio Marigo
On the opposite page, bromeliads
provide important microhabitat for
amphibians and foraging habitats
for lion tamarins. This photograph
was taken in Serra da Bocaina
National Park on the border
between the states of Rio de Janeiro
and São Paulo.
© Tui De Roy/Minden Pictures
On pp. 90-91, a typical view of the
Atlantic Forest Hotspot, with forest
fragments interspersed in a largely
denuded landscape.
© Luciano Candisani
88
widespread habitat destruction. Indeed, the Atlantic
Forest was the first region of Brazil to be colonized. It
has always been the most heavily populated part of
Brazil (108 million or 60% of the population of Brazil),
and is the agricultural (particularly sugar cane, coffee,
cacao, and livestock) and industrial center of the coun-
try. Two of the three largest cities in all of South Ameri-
ca, São Paulo and Rio de Janeiro, are only about 400 km
apart in the central portion of the Atlantic Forest.
Dean (1995) and Coimbra-Filho and Câmara (1996)
reviewed the history of the destruction of these re-
markable forests, which has been especially severe in
the low-lying coastal regions and the long-exploited
northeast (Sergipe, Alagoas, and Paraíba), where less
than 1% remains. In the Brazilian portion of the At-
lantic Forest, only about 78 348 km2 of forest remain,
mostly represented by second-growth forests in ad-
vanced stages of regeneration (INPE and SOS Mata
Atlântica 1997). The best-preserved areas are along the
steep slopes of the coastal mountains of the states of Rio
de Janeiro, São Paulo, and Paraná. In Paraguay, Barboza
et al. (1997) estimated that only 11 618 km2 (13.2%) of
the original 88 050 km2 of forest was still intact (see
Cartes 2003). Considerable deforestation has been tak-
ing place since then. Only about 10 000 km2 of the so-
called Paraná Forest (once covering approximately
22 350 km2) remains in Misiones, Argentina (Chebez
and Hilgert 2003). Combining these figures, we arrive at
a total of 99 966 km2, or about 8.1%, of the Atlantic For-
est left in intact condition. This habitat loss has left
large numbers of the region’s endemic species severely
threatened with extinction (Brooks et al. 1999).
Much of what remains of the Brazilian Atlantic For-
est is under various forms of protection. According
to the World Database on Protected Areas, only 4.1% of
the Atlantic Forest Hotspot is under some form of legal
protection, although this drops to 1.8% when one con-
siders only those in IUCN categories I to IV. A recent
study looking at the state of protected areas in Brazil
revealed that, as of December 2003, there were 100 na-
tional and state parks (19 717 km2), 28 federal and state
biological reserves (1 848 km2), and 77 federal and
state ecological stations and reserves (1 598 km2) in
the Brazilian Atlantic Forest, totaling 205 strictly pro-
tected areas covering 23 163 km2 (Rylands et al. 2004).
If one includes the various areas officially reserved for
sustainable use (mainly Environmental Protection Ar-
eas), this protected area coverage is increased some-
what, although levels of protection vary. In neighbor-
ing countries, eight protected areas totaling 1 392 km2
exist in the Atlantic Forest portion of eastern Paraguay,
and cover 1.6% of its original extent (Cartes 2003),
while in Misiones, there are 60 protected areas of vari-
ous categories, totaling about 4 598 km2, or 21% of the
original extent of Atlantic Forest in that Argentinian
province (Chebez and Hilgert 2003). In addition to
lands protected by the government, there is also an im-
portant legal mechanism for private protection in
Brazil known as the Private Natural Heritage Reserve
Program (Reserva Particular do Patrimônio Natural-
RPPN). Since 1990, more than 500 areas of this kind
(covering about 4 500 km2) have been created, mostly
in the Atlantic Forest and the Cerrado.
In 1997, a groundbreaking study, commissioned by
the World Bank and carried out by the Wildlife Conser-
vation Society and Conservation International, proposed
a major project to establish extensive forest corridors in
the Atlantic Forest and Brazilian Amazonia (Ayres et al.
1997). Some will have over 8 million ha of continuous
habitat, with the existing protected area system as the
backbone. Seven such areas have been identified using
biological indicators such as species richness and levels
of endemism. These corridors, covering only 15%-20% of
the total area of these two major regions, are estimated to
contain over 75% of the rainforest diversity in the coun-
try. With some $44 million coming from the Pilot Pro-
gram to Conserve Brazilian Rain Forests (PP-G7; ad-
ministered by the World Bank), two corridors —one in
the Atlantic Forest and one in the Amazon— will be
established during the next four years to test this
innovative approach. The first Atlantic Forest corridor
focuses on southern Bahia and Espírito Santo. This is
only one of a host of important conservation initiatives
being carried out in the Atlantic Forest.
Although a retrospective of what has happened in the
Atlantic Forest may look rather grim, new conservation
instruments and funding mechanisms —most of them
using up-to-date scientific information to orient invest-
ments—, combined with the largest body of well-trained
conservation professionals in the whole of South Amer-
ica, greatly increase the likelihood that a significant
portion of what remains of the Atlantic Forest’s biodi-
versity will be maintained. In one major development,
the Alliance for the Conservation of the Atlantic Forest
launched a four-year small grants program in 2003 to
help local organizations work directly with private
landowners in managing existing RPPNs and to create
new ones —a move that will substantially increase the
number of private reserves in the two conservation cor-
ridors within this hotspot. The alliance of Conservation
International-Brazil and SOS Mata Atlântica is imple-
menting the program as a core part of the $8-million
Critical Ecosystem Partnership Fund investment strate-
gy for the Brazilian part of this hotspot.
The Atlantic Forest region was also the cradle of the
Brazilian environmental movement, with the growth
of NGO capacity there over the past 30 years being
among the most impressive in the tropical world. The
most recent survey indicated that there were approxi-
mately 700 environmental NGOs active in Brazil; about
30 of these have annual budgets of over $300 000 and
about 20 are national in scope (Fonseca and Pinto
1996). Indeed, although it clearly rates as one of the
highest-priority hotspots on Earth and has already lost
over 90% of what once existed, it is fair to say that of
those hotspots found within developing countries, the
outlook for the Atlantic Forest region is undoubtedly
one of the brightest.

Agrias claudina photographed in
the State of Santa Catarina, Brazil.
Butterfly species reach a high level
of diversity and endemism in the
Atlantic Forest.
© Luiz Claudio Marigo
On the opposite page, the Critically
Endangered northern muriqui
monkey (Brachyteles
hypoxanthus), here seen drinking
from a stream on the forest floor, is
one of Brazil’s most threatened
primate species, and one of two
primate genera endemic to the
Atlantic Forest Hotspot.
© Luciano Candisani
84
species, and only three of the larger islands are con-
sidered unaltered by humans.
Portions of this hotspot, and especially the Chocó
and western Ecuador, have been considered among
the planet’s highest priorities for conservation for
more than twenty years (e.g., Gentry 1977, 1979). This
has resulted in a wide range of conservation projects
by national governments, multilateral and bilateral
funding agencies, and international and national con-
servation organizations, which have led to the creation
of a range of protected areas and many other conser-
vation efforts of varying success. Presently, approxi-
mately 12.5% of the hotspot is considered protected;
however, only 6.9% of the hotspot is conserved in
IUCN categories I to IV. A large network of indigenous
reserves and comunas (communal Black ancestral lands)
exists throughout the hotspot, an example being that of
several Awá Indigenous Reserves straddling the Colombia-
Ecuador border. Although indigenous reserves do not
necessarily protect the full range of biodiversity as well
as a fully protected national park or biological reserve,
they do have great significance for conservation and
ensure more sustainable use of natural resources than
Western forms of development. The Global Conserva-
tion Fund at Conservation International is currently
supporting several initiatives in northwestern Ecuador,
focused on protection of remaining intact lowland
forests in and around the Cotacachi Cayapas Ecological
Reserve and Awá Ethnic Reserve. Key interventions in
this area include community-supported land acquisi-
tions, purchase of logging concessions, community
land titling, and development of community-based in-
centive agreements for conservation.
The Critical Ecosystem Partnership Fund (CEPF) has
also made considerable investments in the Chocó-Mana-
bí portion of this hotspot. Since this program began in
January, 2002, some $3.3 million have been awarded to
24 different projects focused on field activities and
strengthening local NGO capacity.
Elsewhere, other significant areas have been pro-
posed for protection as well, including such biological-
ly important zones as the Naya Corridor in the south-
central Colombian Chocó (Departments of Valle and
Cauca), the protection of which would help conserve
an altitudinal transect from the peaks of the western
Andean cordillera to the sea, and represent perhaps the
last, best chance to conserve a representative sample of
the Chocoan forest ecosystems.
Last, but not least, the Galápagos have long been a fo-
cus of conservation activity and investment, and the vast
majority of this unique archipelago is now recognized as
both a World Heritage Site and a Biosphere Reserve.
In summary, the Tumbes-Chocó-Magdalena Hotspot
presents several challenges and opportunities. Given
the relatively intact nature of the Chocó and the Da-
rién, there is a better possibility of setting aside large
blocks of forest habitat than in most other hotspots. At
the same time, in the Ecuadorian portion of this hot-
spot, in both the wet and moist forests and the dry
forests, there exists the need for the same kind of im-
mediate salvage-type operations required in some
of the other highly degraded, endemic-rich hotspots
like the lowland forests of the Philippines and the
northeastern portion of Brazil’s Atlantic Forest Region,
which are also at or under 3% of their original extent.
There is no doubt that the future of these Ecuadorian
forests hangs in the balance, and that they must be
placed at or near the top of any list of global biodiversi-
ty conservation priorities.
ATLANTIC FOREST
GUSTAVO A.B. DA FONSECA1, 2, 13 • ANTHONY RYLANDS2
ADRIANO PAGLIA5 • RUSSELL A. MITTERMEIER1
Fernando
de Noronha
Island
BRAZIL
PARAGUAY
ATLANTIC
OCEAN
ARGENTINA
URUGUAY
0 600 km
The Atlantic Forest, or Mata Atlântica, is a unique series
of South American rainforest ecosystems quite distinct
from the more extensive Amazonian Forests. It once ex-
tended almost continuously from the states of Rio
Grande do Norte and Ceará in northeastern Brazil south
to Rio Grande do Sul, and included the typical Atlantic
coastal forest on the narrow, low-lying coastal plain, the
forests on the foothills of the Serra do Mar and related
mountain ranges that run roughly parallel to the coast,
and the forests on the slopes of the Serra do Mar itself,
reaching up to 1 800-2 000 m. In northeastern Brazil,
these forests covered only a narrow coastal strip, rarely
exceeding 50-100 km in width, and were replaced in-
land by drier Caatinga and Cerrado formations typical
of the Brazilian interior. The Atlantic Forest also ex-
tends into eastern Paraguay, Misiones in extreme north-
ern Argentina, and narrowly along the coast into
Uruguay. Due to its widespread destruction, rampant
even in the early sixteenth century, it is now difficult to
know the region’s original extent; however, as we define
it here, it originally covered 1 233 875 km2, including
mangroves and restinga, or sandy coastal scrub, and
those portions in Paraguay, Argentina, and Uruguay.
We also include the offshore archipelago of Fernando de
Noronha and its neighbor Atol das Rocas, which belong
to the Fernando de Noronha Chain of submerged vol-
canic mountains in the equatorial Atlantic Ocean
350 km off the coast of northeastern Brazil.
Altitude determines at least three vegetation types
in the Atlantic Forest: the woodlands of the coastal
plain, the slope forests, and the high-altitude woodlands
(Joly et al. 1991). This gradient influences, in turn, the
pattern of richness and composition of the plant com-
munities. Oliveira-Filho and Fontes (2000) found that
increasing altitude affects the composition of woody
plants, augmenting the relative abundance of species of
the families Asteraceae, Melastomataceae, and Solana-
ceae, and reducing the legumes. In the most inland areas
of this biome, the dominant vegetation is semidecidu-
ous, while mixed formations dominated by Araucaria
angustifolia can be found in the coldest southern areas.
Both in the coastal and high-plain areas, the family with
the highest species diversity is the Myrtaceae (more
than 300 species), while the most speciose genus is Eu-
genia (with more than 100 of the 3 500 woody species
found in this biome). Overall, more than 8 000 of an es-
timated 20 000 species of plants (or 40%) are thought to
be endemic. Endemism in trees is particularly high,
with approximately 54% being restricted to the region,
compared to about 30% for higher plants in general
(Mori et al. 1981).
Vertebrate diversity and endemism in this hotspot are
markedly high. There are 263 mammal species known
to occur in the Atlantic Forest, of which 71 are endemic,
including the thin-spined porcupine (Chaetomys sub-
spinosus, VU) and painted tree rat or cocoa rat (Callis-
tomys pictus), which are both representatives of mono-
typic endemic genera, and the maned sloth (Bradypus
torquatus, EN), an unusual and larger relative of the
widespread three-toed sloths (B. tridactylus and B. varie-
gatus). One particularly notable endemic is the Brazilian
arboreal mouse (Rhagomys rufescens, CR), the sole rep-
resentative of an endemic genus, known only from two
specimens collected from Rio de Janeiro, and very like-
ly Extinct. In total, the Atlantic Forest has a remarkable
12 endemic genera, including also two endemic and
threatened primate genera which represent the region’s
primary flagship species, the lion tamarins (Leontopithe-
cus spp.), of which there are four species, and the
muriquis (Brachyteles spp.), with two. Conservation ef-
forts for the lion tamarins have demonstrated the im-
portance of international cooperation between universi-
ties, zoos, national and international NGOs, and the
Brazilian Government; this has resulted not only in their
protection, but also the preservation of their remaining
forests (Kleiman and Rylands 2002). The two Brachyteles
species are the largest of the New World primates. As
with the lion tamarins, conservation efforts and research
focused on them have resulted in numerous initiatives
protecting primate habitat and have also involved train-
ing of many Brazilian conservationists (Strier 1999).
About 936 birds have been recorded from this hot-
spot, 148 of them endemic. There are also 22 endemic
genera. No fewer than four Endemic Bird Areas (EBAs)
have been identified in the region (Stattersfield et al.
1998), the Atlantic Forest Lowlands being one of the
most important in South America, with a total of 52
species and 10 genera confined to this single EBA. More
recent work has also identified 161 Important Bird Ar-
eas in the hotspot (Bencke and Maurício 2002). There
are several outstanding avian flagships, including the
Alagoas curassow (Mitu mitu, EW), a large cracid of
the lowland Atlantic Forest of northeastern Brazil that
was last seen in the wild in 1987 and is now held only
in captivity (Stattersfield et al. 1998), and the related
red-billed curassow (Crax blumenbachii, EN), found fur-
ther south in the Atlantic Forest Lowlands, especially
in the State of Espírito Santo in the large Sooretama Bi-
ological Reserve. The Brazilian merganser (Mergus oc-
tosetaceus, CR) is an important flagship for the southern
Atlantic Forest in Brazil and Misiones. Finally, one of
the most spectacularly colored of all birds is the seven-
colored tanager (Tangara fastuosa, EN), also from north-
eastern Brazil; it has declined mainly due to habitat de-
struction and live capture for the pet trade.
Reptile species in the hotspot number 306, of which
94 species and eight genera (four of them monotypic)
are endemic. The wormlike subterranean reptiles of
the genus Amphisbaena are particularly well represent-
ed, with 20 species, although endemism is low with
only two species endemic, one of them, A. ridleyi, en-
demic to Fernando de Noronha. Endemism is perhaps
most marked in the genus Bothrops, with nine en-
demics of the 17 species present. Amphibian diversity
is very high, with 483 species having been recorded
from the Atlantic Forest, of which a remarkable 286 are
largely endemic. There are also 23 endemic genera; the
most speciose are Cycloramphus with 25 species and
Hylodes with 20, while seven genera are represented by
single species. There is also a single endemic family,
the Brachycephalidae, represented by six species of the
genus Brachycephalus.
Among the remaining vertebrates, there are at least
350 fish species known to occur in the hotspot, of On the opposite page, the channel-
which 133 are endemic; in addition, 10 of the 68 genera billed toucan (Ramphastos
represented are endemic. Invertebrates are not well vitellinus), a widespread species
that occurs in the Atlantic Forest.
known, although 63 of 88 tiger beetle species recorded
© Luiz Claudio Marigo
from the hotspot are endemic (D. Pearson, pers.
comm.), as are 948 of 2 120 butterfly species (Brown Above, the Vulnerable blue-chested
and Freitas 2000). parakeet (Pyrrhura cruentata)
Prince Maximilian zu Wied-Neuwied of Germany was is endemic to the Atlantic Forest
one of the first scientists to travel through the Atlantic of Brazil.
© Luiz Claudio Marigo
Forest. When he carried out his expedition in the early
nineteenth century, he found some of the richest,
tallest, and most impressive forests on Earth. But even
then some regions had already been colonized for more
than two centuries, and were showing the effects of
87

The Endangered black-hooded
antwren (Formicivora
erythronotos) is endemic to the
coastal restingas of the State of
Rio de Janeiro.
© Luiz Claudio Marigo
On the opposite page, bromeliads
provide important microhabitat for
amphibians and foraging habitats
for lion tamarins. This photograph
was taken in Serra da Bocaina
National Park on the border
between the states of Rio de Janeiro
and São Paulo.
© Tui De Roy/Minden Pictures
On pp. 90-91, a typical view of the
Atlantic Forest Hotspot, with forest
fragments interspersed in a largely
denuded landscape.
© Luciano Candisani
88
widespread habitat destruction. Indeed, the Atlantic
Forest was the first region of Brazil to be colonized. It
has always been the most heavily populated part of
Brazil (108 million or 60% of the population of Brazil),
and is the agricultural (particularly sugar cane, coffee,
cacao, and livestock) and industrial center of the coun-
try. Two of the three largest cities in all of South Ameri-
ca, São Paulo and Rio de Janeiro, are only about 400 km
apart in the central portion of the Atlantic Forest.
Dean (1995) and Coimbra-Filho and Câmara (1996)
reviewed the history of the destruction of these re-
markable forests, which has been especially severe in
the low-lying coastal regions and the long-exploited
northeast (Sergipe, Alagoas, and Paraíba), where less
than 1% remains. In the Brazilian portion of the At-
lantic Forest, only about 78 348 km2 of forest remain,
mostly represented by second-growth forests in ad-
vanced stages of regeneration (INPE and SOS Mata
Atlântica 1997). The best-preserved areas are along the
steep slopes of the coastal mountains of the states of Rio
de Janeiro, São Paulo, and Paraná. In Paraguay, Barboza
et al. (1997) estimated that only 11 618 km2 (13.2%) of
the original 88 050 km2 of forest was still intact (see
Cartes 2003). Considerable deforestation has been tak-
ing place since then. Only about 10 000 km2 of the so-
called Paraná Forest (once covering approximately
22 350 km2) remains in Misiones, Argentina (Chebez
and Hilgert 2003). Combining these figures, we arrive at
a total of 99 966 km2, or about 8.1%, of the Atlantic For-
est left in intact condition. This habitat loss has left
large numbers of the region’s endemic species severely
threatened with extinction (Brooks et al. 1999).
Much of what remains of the Brazilian Atlantic For-
est is under various forms of protection. According
to the World Database on Protected Areas, only 4.1% of
the Atlantic Forest Hotspot is under some form of legal
protection, although this drops to 1.8% when one con-
siders only those in IUCN categories I to IV. A recent
study looking at the state of protected areas in Brazil
revealed that, as of December 2003, there were 100 na-
tional and state parks (19 717 km2), 28 federal and state
biological reserves (1 848 km2), and 77 federal and
state ecological stations and reserves (1 598 km2) in
the Brazilian Atlantic Forest, totaling 205 strictly pro-
tected areas covering 23 163 km2 (Rylands et al. 2004).
If one includes the various areas officially reserved for
sustainable use (mainly Environmental Protection Ar-
eas), this protected area coverage is increased some-
what, although levels of protection vary. In neighbor-
ing countries, eight protected areas totaling 1 392 km2
exist in the Atlantic Forest portion of eastern Paraguay,
and cover 1.6% of its original extent (Cartes 2003),
while in Misiones, there are 60 protected areas of vari-
ous categories, totaling about 4 598 km2, or 21% of the
original extent of Atlantic Forest in that Argentinian
province (Chebez and Hilgert 2003). In addition to
lands protected by the government, there is also an im-
portant legal mechanism for private protection in
Brazil known as the Private Natural Heritage Reserve
Program (Reserva Particular do Patrimônio Natural-
RPPN). Since 1990, more than 500 areas of this kind
(covering about 4 500 km2) have been created, mostly
in the Atlantic Forest and the Cerrado.
In 1997, a groundbreaking study, commissioned by
the World Bank and carried out by the Wildlife Conser-
vation Society and Conservation International, proposed
a major project to establish extensive forest corridors in
the Atlantic Forest and Brazilian Amazonia (Ayres et al.
1997). Some will have over 8 million ha of continuous
habitat, with the existing protected area system as the
backbone. Seven such areas have been identified using
biological indicators such as species richness and levels
of endemism. These corridors, covering only 15%-20% of
the total area of these two major regions, are estimated to
contain over 75% of the rainforest diversity in the coun-
try. With some $44 million coming from the Pilot Pro-
gram to Conserve Brazilian Rain Forests (PP-G7; ad-
ministered by the World Bank), two corridors —one in
the Atlantic Forest and one in the Amazon— will be
established during the next four years to test this
innovative approach. The first Atlantic Forest corridor
focuses on southern Bahia and Espírito Santo. This is
only one of a host of important conservation initiatives
being carried out in the Atlantic Forest.
Although a retrospective of what has happened in the
Atlantic Forest may look rather grim, new conservation
instruments and funding mechanisms —most of them
using up-to-date scientific information to orient invest-
ments—, combined with the largest body of well-trained
conservation professionals in the whole of South Amer-
ica, greatly increase the likelihood that a significant
portion of what remains of the Atlantic Forest’s biodi-
versity will be maintained. In one major development,
the Alliance for the Conservation of the Atlantic Forest
launched a four-year small grants program in 2003 to
help local organizations work directly with private
landowners in managing existing RPPNs and to create
new ones —a move that will substantially increase the
number of private reserves in the two conservation cor-
ridors within this hotspot. The alliance of Conservation
International-Brazil and SOS Mata Atlântica is imple-
menting the program as a core part of the $8-million
Critical Ecosystem Partnership Fund investment strate-
gy for the Brazilian part of this hotspot.
The Atlantic Forest region was also the cradle of the
Brazilian environmental movement, with the growth
of NGO capacity there over the past 30 years being
among the most impressive in the tropical world. The
most recent survey indicated that there were approxi-
mately 700 environmental NGOs active in Brazil; about
30 of these have annual budgets of over $300 000 and
about 20 are national in scope (Fonseca and Pinto
1996). Indeed, although it clearly rates as one of the
highest-priority hotspots on Earth and has already lost
over 90% of what once existed, it is fair to say that of
those hotspots found within developing countries, the
outlook for the Atlantic Forest region is undoubtedly
one of the brightest.
 CERRADO
GUSTAVO A.B. DA FONSECA1, 2, 13 • ROBERTO CAVALCANTI 1, 15
ANTHONY RYLANDS 2 • ADRIANO PAGLIA5
BRAZIL
BOLIVIA
ATLANTIC
OCEAN
PARAGUAY
0 500 km
The Cerrado is the second largest of Brazil’s major bi-
omes (after Amazonia), originally covering some
2 031 990 km2, and representing the most extensive
woodland-savanna region in South America. It is en-
tirely tropical, its core occupying the central Brazilian
Plateau, although it extends also marginally into east-
ern Paraguay. Rainfall is typically high (between 1 100
and 1 600 mm annually), but is concentrated in a peri-
od of six to seven months (April to October), the dry
season being pronounced and, in some places, result-
ing in water stress. Many of the typical Cerrado plant
species are drought-adapted.
The Cerrado is a gradient-mosaic of different vegeta-
tion types, determined primarily by soil conditions. The
only true forest formation is cerradão, a closed-canopy
dry forest reaching 7 m in height on the more fertile
soils. Cerrado sensu stricto is a woodland savanna with a
well-developed herbaceous layer, small trees, shrubs,
and trunkless palms, and the dominant vegetation oc-
curs on widespread red and yellow latosols. Other for-
mations include campo sujo (“dirty savanna”) and campo
limpo (“clean savanna”), the former scrub grassland
(sparsely distributed shrubs and short trees) and the lat-
ter, open grassland. Gallery forests are not classified as
typical Cerrado formations, but are prevalent in the land-
scape, comprising an interconnected network of mesic
habitats along the streams and rivers of three of the ma-
jor Brazilian drainages. They are important for the main-
tenance of the diversity of the Cerrado fauna (Redford
and Fonseca 1986), being a refuge and dispersal route for
a significant number of plants and animals of Amazon-
ian and Atlantic Forest origin (Sarmiento 1983).
Fire, both natural and human-induced, is an impor-
tant feature of Cerrado ecology. The flora displays a
number of adaptations to fire, including thick bark,
leathery leaves, and a rapid regeneration capacity. Fire
maintains the balance between grasses and woody veg-
etation; in its absence, trees predominate, resulting in
dry forest. Seed germination is fire-dependent in a
number of species.
The Cerrado is one of the richest of all tropical sa-
vanna regions and has high levels of endemism. Men-
donça et al. (1997) listed 6 387 angiosperms, and total
plant diversity is estimated at around 10 000 species, of
which 4 400 are endemic. The herbaceous species are
almost totally endemic. A conservative estimate for di-
versity of fungi (Dianese et al. 1997) gives 24 000
species, but, being based on six species per host plant,
and considering only 4 000 vascular plant species, it is
likely an underestimate. Several non-endemic plant
species stand out as flagships for this hotspot: for ex-
ample, the conspicuous palms Mauritia vinifera (locally
known as buritís), which grow along the swampy head-
waters of streams and rivers (known as veredas) that
flow through the Cerrado’s open grasslands. Trees of
the genus Tabebuia (Bignoniaceae), known as ipê, are
also spectacular and widespread inhabitants of the Cer-
rado; their brilliant yellow, pink, and purple flowers are
a common sight.
Vertebrate diversity is also high in the Cerrado, but
endemism low when compared to plants. Mammals are
represented by 195 species, although only 14 are en-
demic. Four are representatives of the three endemic
genera in the hotspot: a rodent (Microakodontomys tran-
sitorius), known only from a single specimen collected
in 1986 in the Brasília National Park; the Candango
mouse (Juscelinomys candango), a semi-fossorial rodent
first discovered in 1960 on the site of the capital, Bra-
sília, then under construction, and which has never
again been collected; and the cerrado mouse (Thal-
pomys cerradensis) and hairy-eared cerrado mouse (T. la-
siotis). The Cerrado is a major stronghold for a number
of large, widespread South American mammals that are
seen as flagships for the region: the maned wolf (Chryso-
cyon brachyurus), a large-eared, long-legged canid, On the opposite page, the giant
80 cm in height, which has long, golden-red fur and is anteater (Myrmecophaga
one of the most attractive of all wild canids; giant tridactyla) is an ant and termite
specialist. It uses its powerful claws
anteater (Myrmecophaga tridactyla, VU); giant armadillo
to break open mounds,
(Priodontes maximus, EN); Brazilian tapir (Tapirus ter- and extraordinarily long tongue
restris, VU); pampas deer (Ozotoceros bezoarticus); and —which can be up to 60 cm long—
several cat species, including the jaguar (Panthera onca), to lick up the workers.
puma (Puma concolor), ocelot (Leopardus pardalis), and © Luiz Claudio Marigo
jaguarundi (Puma yagouaroundi).
Above, the maned wolf
Bird diversity is relatively high in the Cerrado, with
(Chrysocyon brachyurus),
605 species regularly recorded, 16 of these being strict distinctive among canids for having
endemics. Two of BirdLife International’s Endemic Bird such remarkably long legs, inhabits
Areas (EBAs) coincide with the Cerrado Hotspot (Stat- the grasslands and scrub forest of
tersfield et al. 1998). One well-known endemic flagship central South America, and is a
is the Brasília tapaculo (Scytalopus novacapitalis), a flagship species of the Cerrado
in particular.
passerine with a very limited range, occurring in gallery
© Luiz Claudio Marigo
forest remnants around Brasília, besides a few locations
in Minas Gerais, including the Serra da Canastra Na-
tional Park. Some of the larger birds are also good flag-
ship species, particularly the rhea (Rhea americana).
There are 33 endemic reptiles (of a total of 225
species), six in the snake genus Apostolepis. As with the
93

On pp. 94-95, buriti palm trees or
aguaje palms (Mauritia flexuosa)
lining the banks of the Rio Formoso
in Emas National Park. The species
is typical of river margins in the
Cerrado Hotspot.
© Luiz Claudio Marigo
Above, a scene not uncommon in
the Cerrado Hotspot: isolated forest
fragments interspersed with
industrial farmland. This
photograph was taken in the region
of Emas National Park.
© Frans Lanting/Minden Pictures
On the opposite page, a group
of children in a small coal factory
where cut wood is brought to burn
in the State of Minas Gerais,
Cerrado, Brazil.
© Luiz Claudio Marigo
96
Atlantic Forest, the genus Amphisbaena is well repre-
sented, with at least 26 species present. Very little is
known about the ecology of amphisbaenians, as they
are largely subterranean. One species that has been
well studied is the widespread giant worm lizard (A.
alba), which can grow to 70 cm in length; feeds on bee-
tles, ants, and spiders; and has a clutch size ranging
from 8-16 eggs, one of the largest known for the family
(Colli and Zamboni 1999). Among amphibians, some
289 species have been recorded, of which 26 are en-
demic, including two species endemic to Serra da
Canastra National Park: the Canastra snouted tree frog
(Scinax canastrensis) and Zagaia tree frog (Hyla sazimai).
Freshwater fish diversity is high, with approximately
800 species present, and an estimated 200 of these en-
demic. Generic diversity is also high, with some 240
genera represented in the hotspot, of which 18 are en-
demic. There is a paucity of information on inverte-
brates, but preliminary data on insect diversity indi-
cates that 10 000 of the 40 000 species of Neotropical
Lepidoptera, 129 of 443 Neotropical termites, and 139
of 547 Neotropical social wasps are found in the Cerra-
do (Raw 1998). Bees number at least 820 species (out of
7 000 for the entire Neotropics). Twenty-three of 64
tiger beetle species present are considered endemic (D.
Pearson, pers. comm.).
The Cerrado has been occupied by humans for at
least 11 000 years (Barbosa and Nascimento 1993), and
perhaps as long as 30 000 (Schmitz 1993). Hunter-gath-
erer societies subsisted on game and native plants, as
clearly indicated by the impressive Holocene rock paint-
ings of animals found especially in the dry areas around
the Rio São Francisco in Bahia and Minas Gerais and in
Goiás and Piauí (Schmitz 1993). During the colonization
of Brazil in the early 1500s, the Portuguese avoided the
interior of the country because of treaties between Spain
and Portugal, and also a cultural predisposition to settle
near the sea (Holanda 1936). The first incursions, in
search of gold and precious stones, were in the eight-
eenth century (Alho and Martins 1995), and opened the
way for cattle-ranching, the primary economic activity
until the mid-1950s, when the government began its
plans for a new capital in Goiás, in the heart of the Cer-
rado, to stimulate the colonization and development of
the region. Improved transport and infrastructure, a ma-
jor commitment to the automobile industry, and agri-
cultural and soil research over the ensuing years, result-
ed in the Cerrado becoming the major agribusiness
(soybean, corn, and irrigated rice) frontier of the coun-
try. Following major failures in large-scale development
programs in Amazonia, as of the 1970s the World Bank
considered the Cerrado to be the alternative for Brazil’s
agricultural expansion (Goodland and Irwin 1975). To-
day, nearly one-quarter of all grain produced in Brazil
comes from the Cerrado (Alho and Martins 1995).
Estimates of remaining natural vegetation indicate
that the Cerrado has been heavily impacted. Mantovani
and Pereira (1998) suggested that by 1993, approxi-
mately 67.1% was either “disturbed” and/or “highly
modified,” with the remaining natural habitat largely
fragmented. They detected only three areas of exten-
sive undisturbed Cerrado (two in the far northeast and
one in Mato Grosso do Sul); 37.3% of the Cerrado had
already been totally converted to human use, and
41.4% was being used for pasture and charcoal produc-
tion. Gallery forests have been particularly heavily im-
pacted. These numbers indicate 78.7% under some
form of human use, meaning that only 21.3%, or
432 814 km2, is still in intact condition.
In spite of its large size and great importance for bio-
diversity conservation, the Cerrado is poorly represent-
ed in Brazil’s protected area system. The World Database
on Protected Areas indicates that only 1.4% is conserved
in protected areas in IUCN categories I to IV, increasing
to 5.5% when including other forms of reserves and
parks. In Brazil, 62 strictly protected areas (state and fed-
eral parks, biological reserves, and ecological stations),
mostly limited in size, cover 42 676 km2 —only 2.1% of
the hotspot’s original extent. The most important parks
include Chapada dos Guimarães (330 km2), Serra da
Canastra (2 000 km2, although only about 720 km2 are
currently protected), Emas (1 330 km2), and Grande
Sertão Veredas (840 km2), and in 2001 a large and valu-
able area of the Cerrado in the State of Tocantins was se-
cured as the Jalapão State Park (1 588 km2). Including
protected areas of sustainable use (mainly Environmen-
tal Protection Areas), this number increases to 102 areas
covering 83 520 km2 or 4.1% of the Cerrado (Rylands et
al. 2004).
Conservation International and its partners, the Fun-
daçao Emas, neighboring municipalities, and farming
cooperatives have a major program under way for the
protection and management of Emas National Park,
probably the most important protected area in the en-
tire hotspot, and threatened by surrounding agriculture,
pesticides, and invasion of exotic grasses. The Brasília-
based Fundação Pró-Natureza-FUNATURA was respon-
sible for the creation in 1989 of the Grande Sertão
Veredas National Park, protecting a significant sample
of the Gerais ecosystems: native grasslands, veredas
(Mauritia palm-dominated bogs), and scrub forests, in
addition to populations of key Cerrado vertebrates.
The Brazilian Government held an expert-based Re-
gional Priority Setting Workshop for the Cerrado and
Pantanal in 1998 (Fundação Pró-Natureza-FUNATURA
et al. 2000). Seventy priority areas for conservation
were identified, ranging from the central highlands of
the country, at the headwaters of Brazil’s major rivers,
through the savannas of Amazonia, to the Pantanal of
southwestern Brazil, down to fragmented Cerrado rem-
nants in the states of São Paulo and Paraná. The meet-
ing concluded with recommendations for immediate
action, with suggestions for changes in legislation in or-
der to create incentives to preserve biodiversity togeth-
er with water sources, consolidate existing parks and
reserves, establish new protected areas, and more in-
tensively survey this vast region that is still so little
known to science.
 CHILEAN
WINTER RAINFALL-
VALDIVIAN FORESTS
MARY T. KALIN ARROYO 126 • PABLO MARQUET 127
CLODOMIRO MARTICORENA128 • JAVIER SIMONETTI 126
LOHENGRIN CAVIERES 129 • FRANCISCO SQUEO 130 • RICARDO ROZZI 131
CHILE
ARGENTINA
ATLANTIC
OCEAN
PACIFIC
OCEAN
0 1000 km
This hotspot includes the area of central-northern Chile
characterized by a winter-rainfall regime, together with
the northern part of southern Chile, characterized by
rainfall all year round. The winter-rainfall area is divided
almost equally between a typical Mediterranean-type cli-
mate area (155 000 km2) and a more arid area of winter-
rainfall deserts (145 000 km2). This hotspot has, there-
fore, been modified from its definition five years ago by
extension to the east, to cover a sliver of similar habitats
in neighboring Argentina, and by addition to the south
of a substantial extra portion of the temperate forests, to
cover the full extent of the Valdivian temperate forests
ecoregion. Although it is somewhat unusual to lump
such climatic variation into a single hotspot, we feel
that since the subunits are part of a continuum with no
clear floristic breaks, it is justified to join them together
in one hotspot. As we define it, then, the hotspot ex-
tends from the Pacific Coast to the crest of the Andean
mountains between 25ºS and 47ºS, to include a narrow
coastal strip from 25ºS to 19ºS (Arroyo et al. 1999), and
also the offshore Juan Fernández Islands (these having
been treated separately in Mittermeier et al. 1999), such
that this hotspot now covers some 397 142 km2.
The uniqueness of this hotspot derives from several
factors, notably its position at the crossroads of two ma-
jor floristic and faunistic regions: the Neotropical and
ancient Gondwanan provinces. The island-like features
of this hotspot, determined by the Pacific Ocean to the
west, the Andean crest to the east, and the absolute
desert to the north, are another determinant of diversity:
with only strong vegetation continuity to the south, in-
terchange with adjacent continental areas has been re-
stricted, unequal among the different groups of organ-
isms, and predominantly in one direction (Arroyo et al.
1996; Villagrán and Hinojosa 1997; Armesto et al. 1998).
Among the more interesting and unusual elements in
the northern desert communities (from the northern ex-
treme to around 30ºS) are the extended band of coastal
fog (camanchaca) desert and the more southerly inland
desierto florido (Squeo et al. 2001). South of 30ºS, the
dominant vegetation types are coastal and inland ma-
torral (jaral) and savannas dominated by Prosopsis chilen-
sis and Acacia caven (Fuentes et al. 1990). On the lower
western slopes of the Andes and the eastern slopes of
the Coast Range, the typical matorral is open and con-
tains a rich assemblage of endemic herbaceous and geo-
phyte species (Arroyo et al. 1995). In the wetter climate
of the western side of the Coast Range, the forests are
closed. At higher elevations, the typical Mediterranean
sclerophyllous vegetation grades up to Nothofagus forest.
In addition, there is a small tongue of coastal rainforest
along the southern coast from 39ºS southwards.
Overall, this hotspot holds 3 892 vascular plant spe-
cies, of which 1 957 (50.3%) are endemic to the hotspot
per se (this is around three-quarters of all vascular plant
species/ endemics known to occur in continental Chile).
The entire southern Valdivian forest region, considering
all habitats, supports 1 284 species; however, based on
present information, a surprisingly small number (less
than 100) appear to be entirely restricted to the rainfor-
est zone on account of continuity with contiguous vege-
tation types. Part of the problem resides in defining the
limits of the rainforest. With more precise limits in hand,
this last number is likely to increase. Incorporation of
new data (since Arroyo and Cavieres 1997) brings plant
diversity in the winter-rainfall area to 3 539 species, of
which 1 769 (50%) are endemic to that subunit. The
Juan Fernández Islands have a flora comprised of 200
native species (Marticorena et al. 1998), as well as over
200 introduced species, many of which, interestingly,
are natives of continental Chile. The isolation of these
offshore islands causes high endemism in their native
floras, with 62% endemism.
A salient feature of this hotspot is the number of en-
demic higher taxa. For example, many woody plant gen- On the opposite page, Cerro Las
era on the continent, including trees, as well as a num- Peinetas, Villarrica National Park.
ber of genera of geophytes and Cactaceae are endemic to This park includes three volcanic
cones and extends from the hills
the hotspot. There are also a number of endemic genera
opposite Pucón, Chile to the border
on the Juan Fernández Islands (Stuessy et al. 1992). In
with Argentina.
addition, three plant families (Aextoxicaceae, Gomorte- © Pablo Valenzuela
gaceae, and Lactoridaceae) are restricted to the hotspot.
Although not rich in species, the Valdivian rainforest Above, flower of the quisquito
habitat is characterized by an outstanding number of (Neoporteria aspillagae), a cactus
endemic and/or monotypic plant genera, with one- that grows along the coast of Norte
third of all woody plant genera endemic to the general Chico (the region between the
area of the hotspot (Arroyo et al. 1996). This rainforest Pacific and the Andes) and in the
central part of Chile.
contains several genera important to our understanding
© Pablo Valenzuela
of angiosperm evolution, as seen in the presence of
primitive members of the Asteraceae (Dasyphyllum) On pp. 100-101, araucaria forest
and representatives of the Andean family Calyceraceae (Araucaria araucana) in the
(Moschopsis), considered to be remnants of the original Nahuelbuta National Park.
sunflower alliance that evolved on East Gondwanaland © Pablo Valenzuela
(Bremer and Gustafsson 1997).
Vertebrate richness is relatively low, but endemism
can be high, particularly among reptiles and amphibians.
There are a total of 43 amphibian species in the hotspot
99
(29 endemic). In addition, five of the 12 genera present
are endemic: Telmatobufo, with three species; Rhinoderma,
with two species; and Insuetophrynus, Caudiverbera, and
Hylorina, each with a single species. Furthermore, this is
one of the few hotspots with an endemic amphibian fam-
ily (Rhinodermatidae), with Darwin’s frog (Rhinoderma
darwini, VU) and the Chile Darwin’s frog (R. rufum, CR)
among the most important flagship species for this re-
gion. An impressive 27 of the 41 reptile species known
from this hotspot are endemic. The lizard genus Liolae-
mus accounts for 30 of the reptile species, with 19 of
them endemic to the hotspot —a striking radiation.
Mammal diversity is low, with 64 species and 13 en-
demics. However, there are five endemic genera. The ro-
dent genus Octodon comprises three species of degus,
while the remaining four genera are all represented by
single species: the Chilean climbing mouse (Irenomys
tarsalis), the Chilean shrew opossum (Rhyncholestes ra-
phanurus), the coruro (Spalacopus cyanus), and the moni-
to del monte (Dromiciops gliroides, VU). The last-named
species is the only genus in an endemic family, the Mi-
crobiotheriidae, and is known only in the proximity of
Concepción south to Chiloé Island, south-central Chile,
and in the mountains east of the Argentine border. One
of the best-known endemic mammals of this hotspot is
the chinchilla (Chinchilla lanigera, VU), an unusual ro-
dent species. Another is Darwin’s fox (Pseudalopex ful-
vipes), a canid species with only two known populations,
one found in the forests of Chiloé Island, and another in
the coastal mountains in Nahuelbuta National Park.
Bird diversity is also low: 226 species, but only 12 en-
demics, including two monotypic endemic genera, Se-
phanoides and Sylviorthorhynchus. The former is repre-
sented by the Juan Fernández firecrown (S. fernandensis,
CR), which is confined to the offshore Juan Fernández Is-
lands, and the latter by Des Murs’ wiretail (S. desmursii),
restricted to the temperate forest in south-central Chile.
The hotspot’s fish fauna is small with only 43 native
species, but it is remarkable in having two endemic
families, the mountain catfishes (Nematogenyidae) and
the perch-like fishes of the genus Percilia (Perciliidae).
Although the majority of species belong to South Amer-
ica’s large endemic fish groups, nearly 20% are relicts
of Gondwanan groups and are shared with southern
Africa, Australia, and New Zealand.
The hotspot faces many major threats to its biodiver-
sity, the most serious being habitat degradation, planta-
tion forestry, forest fires, overgrazing, spread of alien
species, and trade in native species (Armesto et al. 1998).
Although it is difficult to come up with an estimate of
natural vegetation remaining in pristine condition, we
believe that no more than 30% of the original extent of
this hotspot would fall into this category. Current land-
use survey data indicates that highly intensive kinds of
land use occupy 72 000 km2 (16.5%) of the total land
area of the hotspot (derived from figures given in Ar-
royo et al. 2000), with this increasing to as much as 58%
in the southern part of the Mediterranean area. The
bulk of the more pristine land is found in the winter-
rainfall deserts subunit. In the more heavily impacted
Mediterranean-type subunit, remaining pristine vege-
tation amounts to no more than 10 000-13 000 km2. In
the Valdivian forest, the total remaining intact, fairly
pristine area is perhaps about 50 000-67 000 km2.
The National Protected Area System (SNASPE) of Chile
falls into three main categories (national parks, national
reserves, and national monuments). Although 19% of
continental Chile as a whole is currently protected, this
hotspot unfortunately has only 12.8% of its original extent
under some form of official protection. However, particu-
larly in the northern part of the rainforest zone, the cur-
rent protected areas tend to be located at mid-elevations
and concentrated in the Andean Cordillera (Armesto et al.
1998), leaving the richest rainforest areas along the coast
practically unprotected. Pumalin Park, a well-known pri-
vate conservation initiative, is situated in the rainforest
zone, as are several other small parks of this nature.
In November 2003, an important step was taken to-
wards conserving the Valdivian Coastal Range in south-
ern Chile. The Nature Conservancy, in partnership
with CI’s Global Conservation Fund, the World Wildlife
Fund, and local conservation organizations acquired
60 000 ha of biologically rich temperate rainforest in an
open auction. This parcel of land lies between the Chai-
huin and Bueno rivers, in the Coast Range of the Lakes
Region in Chile. A portion of this land will be coopera-
tively managed with three adjacent national reserves
that together will ideally be elevated to national park
status, while the remainder will be owned and man-
aged by a new Chilean conservation organization.
An integrated and successful conservation strategy
will require many different and novel approaches. Col-
laboration among government, the private sector, and
civil society will be essential to maintain the biodiver-
sity of this scenically beautiful hotspot for the future.
MESOAMERICA
RUSSELL A. MITTERMEIER 1 • JAN SCHIPPER 97 • GERRIT DAVIDSE 26
PATRICIA KOLEFF 24 • JORGE SOBERÓN 24 • MANUEL RAMÍREZ 99
BARBARA GOETTSCH 98 • CRISTINA G. MITTERMEIER1
GULF
MEXICO OF MEXICO
On the opposite page, the jaguar
Cozumel
Island (Panthera onca) is a wide-ranging
Revillagigedo
Islands species that is still common over
BELIZE
HONDURAS much of the wilderness portion of
GUATEMALA its range in areas like Amazonia
Clipperton EL SALVADOR and the Pantanal, but is becoming
Island
NICARAGUA PANAMA increasingly rare in hotspots like the
COSTA RICA Atlantic Forest and Mesoamerica.
Coiba
PACIFIC Island © Patricio Robles Gil/Sierra Madre
0 800 km
OCEAN Cocos
Island
The Mesoamerica Hotspot includes all tropical and sub-
tropical natural plant formations from eastern Panama
west and north through Costa Rica, Nicaragua, Hon-
duras, El Salvador, Guatemala, and Belize, as well as in
portions of southern and coastal Mexico. Within Mexico,
103

On pp. 104-105, the great curassow
(Crax rubra), here seen in the
El Cielo Biosphere Reserve in the
State of Tamaulipas, is a large,
commonly hunted species.
© Patricio Robles Gil/Sierra Madre
Above, a male plumed basilisk
lizard (Basiliscus plumifrons) in
Tortuguero National Park, Costa
Rica. Basilisks are almost always
found along small streams in the
forest, and are renowned for their
ability to run across the surface of
the water to escape intruders.
© Kevin Schafer
On the opposite page, the
magnificent red-eyed tree frog
(Agalychnis callidryas) is one of
the world’s best-known and most
photographed frogs, and has
become a “poster child” for rainforest
conservation around the world.
© Piotr Naskrecki
106
the hotspot extends as far north as northern Sinaloa (Río
Fuerte) on the Pacific Coast and as far as the middle of
the Sierra Madre Oriental (west of Tampico) on the
coast of the Gulf of Mexico. It includes all of the tropical
and subtropical moist, seasonal, and dry forest forma-
tions, as well as small portions of paramo montane
grasslands, mangroves, semidesert and thornscrub, and
is usually considered to be part of the northern Neotrop-
ical region. However, the hotspot does not encompass
the true desert and desert scrub formations of the Baja
California Peninsula, Sonora, Chihuahua, and Coahuila
in northwestern and central Mexico, or the subtropical
pine-oak forests in the higher elevations of the Sierra
Madre Occidental and Oriental mountains (here includ-
ed in the Madrean Pine-Oak Woodlands Hotspot).
The Mesoamerica Hotspot covers all of political Cen-
tral America, with the exception of easternmost Panama
(which is included in the Tumbes-Chocó-Magdalena
Hotspot). Within Mexico, the region includes the entire
area of the Mexican states of Chiapas, Yucatán, Quintana
Roo, Tabasco, Campeche, Veracruz, and portions of the
states of Oaxaca, Guerrero, Puebla, México, Michoacán,
Morelos, Querétaro, Jalisco, Nayarit, Colima, Guanajua-
to, San Luis Potosí, Zacatecas, Sinaloa, Durango, Sonora,
Chihuahua, and Tamaulipas. Also included in this hot-
spot are numerous nearshore and offshore islands in both
the Caribbean Sea and Pacific Ocean, which are impor-
tant biologically due to the presence of endemic species
and as nesting areas for seabirds. These islands include
the Revillagigedos (2 965 km2; Mexico), Tres Marías Is-
lands (400 km2; Mexico), Cozumel (539 km2; Mexico),
Islas de la Bahía (373 km2; Honduras), Cocos (24 km2;
Costa Rica), Providencia (17 km2; Colombia), San An-
drés (44 km2; Colombia), and Coiba (50 km2; Panama),
as well as Clipperton Island (7 km2; France). In all, this
hotspot covers some 1 130 019 km2.
Although the Mesoamerica Hotspot does not have
quite the breadth or the great heights of the Tropical
Andes Hotspot, it does have a series of impressive high-
lands and mountain chains stretching along its main axis.
These include ancient, worn-down crystalline uplands;
younger, upthrust, metamorphic mountain ranges; and
strings of frequently active volcanoes. Peaks in several
of the latter two types of highlands reach and some-
times exceed 4 000 m in total height —the highest being
the Pico de Orizaba or Citlaltépetl located in the State of
Veracruz, a part of the neovolcanic chain and Mexico’s
highest peak at 5 610 m.
The main axes of these mountains, locally called
“sierras” or “cordilleras,” have always acted as barriers to
species dispersal, resulting in considerable biological
differences between the Pacific and Caribbean lowlands
to either side, and form the continental divide through-
out much of Central America. Additionally, many of the
mountain chains and volcanoes are isolated from each
other, the result of which is high beta-diversity between
landscapes. The valleys and lowlands running parallel
to the mountains, on the other hand, have long served
as natural corridors for animal and human migrations
and dispersal. The species which today inhabit the
Mesoamerica Hotspot are relicts of the convergence be-
tween the flora and fauna of North and South America
in the late Pliocene (approximately 3 million years ago),
when the land bridge between these two continents was
finally complete at what is today the Isthmus of Panama
—giving rise to one of the greatest species interchanges
in the history of the Americas. The result is a complex
mosaic of plant and animal communities that have con-
verged, adapted and, in many cases, evolved into new
species. Thus, the unique geological and evolutionary
history of Mesoamerica has resulted in some of the
highest levels of diversity and endemism on Earth.
In terms of plants, the latest estimate for total diver-
sity of this hotspot is some 17 000 species with 2 941 en-
demics. This estimate is based upon what is known
from the Flora Mesoamericana Project, which covers
the seven Central American countries as well as the five
southernmost Mexican states of Campeche, Chiapas,
Tabasco, Quintana Roo, and Yucatán. In addition, of
2 523 genera, 65 are endemic (of which 50 are repre-
sented by single species), a number that is perhaps low-
er than expected, but which probably reflects a relative-
ly recent (on the geological time scale) mass movement
of plants from both the north and the south into the isth-
mus. Both the number of species and genera present
and endemic would be considerably higher for the
Mesoamerica Hotspot because a significant number of
genera (such as Plocosperma) just spill over into Oaxaca,
Veracruz, and Tamaulipas. Finally, there are three en-
demic families, each with a single species: Haptan-
thaceae (Haptanthus hazlettii), Lacandoniaceae (Lacan-
donia schismatica), and Ticodendraceae (Ticodendron
incognitum). Lacandonia, which has a close affinity with
the family Triuridaceae, is particularly unique among
flowering plants for having “inside-out” flowers, i.e., the
central androecium (stamens) surrounded by the gy-
noecium (separate pistils). Within the Mesoamerica
Hotspot there are approximately 312 cacti species, of
which 60% are endemic, or nearly so, to the hotspot
(and, of these, 85% are found exclusively within the
Mexican portion of the hotspot) (B. Goettsch, unpubl.).
Vertebrate diversity is even more impressive. There
are 440 species of mammals recorded within the hot-
spot, of which 66 are endemic. Three endemic species,
Van Gelder’s bat (Bauerus dubiaquercus), the Yucatán ves-
per rat (Otonyctomys hatti), and Bangs’ mountain squir-
rel (Syntheosciurus brochus), are all representatives of en-
demic genera. A number of endemics are confined to
offshore islands in the hotspot, including three species
on Cozumel (Reithrodontomys spectabilis, EN; Procyon
pygmaeus, EN; and Nasua nelsoni, EN). The primates of
the Mesoamerican forests, particularly the spider mon-
keys (Ateles spp.), the howler monkeys (Alouatta spp.),
and the squirrel monkeys (Saimiri spp.), are often
among the most conspicuous elements of the local fau-
na, play important ecological roles as seed dispersers,
and have great value as ecotourism attractions. The two
largest Neotropical land mammals, Baird’s tapir (Tapirus
bairdii, EN) and the jaguar (Panthera onca), are rarely
seen in the wild, but are among the most well-known
flagship species for Mesoamerica’s tropical forests.
In terms of bird diversity, there are 1 124 species
known to occur regularly in the hotspot, and of these,
213 are endemic. This region is also a top priority for
BirdLife International, which recognizes no less than 17
Endemic Bird Areas (EBAs) within this hotspot (Statters-
field et al. 1998) covering almost its entire extent.
Among the bird species endemic to this hotspot, perhaps
none is more impressive than the resplendent quetzal
(Pharomachrus mocinno), the most important flagship for
conservation of Mesoamerica’s precious cloud forest, a
habitat upon which it depends for survival and which
also retains the headwaters of critical watersheds for
much of the region. Less well known, but at least as
spectacular, is the horned guan (Oreophasis derbianus,
EN), a monotypic genus endemic to this hotspot and
found only in high-altitude cloud forests in southern
Mexico and Guatemala. Besides the guan, a further 19
genera are endemic. In addition to the endemics, this re-
gion is also a critical trans-regional “flyway” for at least
225 migratory species, as three of the hemisphere’s four
migratory bird routes converge in Mesoamerica.
Reptile diversity is again extremely high, with a total
of 686 species recorded from the hotspot, of which ful-
ly 239 are endemic. Included among these are an en-
demic family, the Dermatemydiidae, with a single spe-
cies, the Central American river turtle (Dermatemys
mawii), and 10 endemic genera, seven of which are rep-
resented by single species (for example, Coloptychon
rhombifer, Scolecophis atrocinctus, and Exiliboa placata).
Along with the Dermatemys, another important reptile
flagship is Morelet’s crocodile (Crocodylus moreletii),
also endemic to the hotspot and occupying much the
same range in rivers of southern Mexico and adjacent
parts of Guatemala and Belize. The hotspot also hosts
many important nesting beaches for marine turtles, in-
cluding Tortuguero, Costa Rica —one of the most im-
portant nesting beaches for green turtles (Chelonia my-
das) in the Western Hemisphere. Other beaches on
both the Caribbean and Pacific coasts also provide im-
portant nesting areas for hawksbill (Eretmochelys imbri-
cata, CR), olive ridley (Lepidochelys olivacea, EN), and
leatherback (Dermochelys coriacea, CR) sea turtles.
Amphibian diversity is comparably impressive, with
a total of 580 species, 353 of which are endemic. Fur-
thermore, 14 of the 60 amphibian genera found in the
Mesoamerica Hotspot are also endemic. Guatemala
and the adjacent Mexican State of Chiapas are particu-
larly rich in salamanders, and are considered a center
of origin and dispersal for tropical salamander species.
In total, the hotspot has 164 species of the order Cau-
data (the newts and salamanders), and 121 are endem-
ic; all endemics belong to the family Plethodontidae
(lungless salamanders), for which a total of 156 species
occur. Among the amphibians endemic to this region
are several important flagship species, including the
golden toad (Bufo periglenes) —an Extinct species that
serves as a flagship for the worldwide decline in am-
phibian populations. Additionally, the hotspot is home
to the golden harlequin frog (Atelopus zeteki, CR) of
Panama, and a number of magnificently colored mem-
bers of the poison dart frog family, Dendrobatidae.
Freshwater fishes form a striking component of the
vertebrate diversity of this hotspot, with a total of
509 species, of which 340 are endemic. In addition, 25
genera are endemic, including several genera of live-
bearing fishes in the family Poeciliidae and 12 endem-
ic genera of cichlid fishes. With more than 200 species
between them, these two families are numerically the
most significant elements of the fish fauna. Both fami-
lies have undergone striking radiations inside the hot-
spot associated with a complex history of hydrographic
changes. Nearly a quarter of the region’s fishes have
distributions restricted to single bodies of water or to
small tributaries in larger watersheds.
Unfortunately, Mesoamerica exhibits some of the
highest deforestation rates in the world, and a large part
of Mesoamerica’s Western Pacific lowland forests was
cleared prior to the twentieth century for export crops
and subsistence agriculture. Deforestation during the
past 100 years has been most severe in the remaining
Caribbean lowland forests, which are converted for ba-
nana production (the Banana Republic), and at higher
elevations on both slopes for coffee production —pre-
dominantly export crops. In many cases, abandoned
plantations make way for cattle production, a boom
and bust industry throughout Central America. In areas
of steep terrain, erosion and eventually landslides add
to the impact of large-scale agriculture —with severe ef-
fects to both the terrestrial and freshwater environ-
ment, as well as human health.
As a result of all these impacts on the forests and
other ecosystems of Mesoamerica, a major portion of
the region’s natural vegetation has already been lost, On the opposite page,
such that estimates for remaining natural vegetation in the resplendent quetzal
the Mesoamerica Hotspot range from 31.5% to 47%. For (Pharomachrus mocinno) is a
magnificent bird that has great
example, a detailed analysis for Costa Rica estimates
cultural importance and is one of
the remaining forest cover of the country at 40.29% the outstanding flagship species for
(CCT/CIEDES 1998). However, these figures include the Mesoamerica Hotspot.
secondary forest formations and plantations of many © Neil Lucas/naturepl.com
different kinds, and greatly overestimate the remaining
natural vegetation types. The percentages remaining in Above, an ornate hawk-eagle
(Spizaetus ornatus) in Calakmul
primary forest in Nicaragua, Honduras, Guatemala,
Biosphere Reserve, Campeche,
Panama, and Belize are likely similar to those for Costa Mexico.
Rica. Those for Neotropical Mexico are probably some- © Patricio Robles Gil/Sierra Madre
what lower, and for El Salvador likely much greater.
Consequently, we estimate that roughly 20% of the On pp. 110-111, a northern
original vegetation remains intact in this hotspot. tamandua (Tamandua mexicana)
killed by a car on a Costa Rican
Currently, a total of 12.6% of the Mesoamerica Hot-
road. Roads cut through tropical
spot is included in protected areas (nearly 11% in pro-
forest can have a heavy impact
tected areas in IUCN categories I to VI), although this on terrestrial fauna.
figure is less than half when one considers only those © Piotr Naskrecki
in categories I to IV (5.7%). In terms of the percentage
of national land area protected, Belize and Costa Rica top
the list with 36.6% and 31% of their land area covered
by protected areas in categories I to VI, respectively.
109

A red-billed streamertail (Trochilus
polytmus), a hummingbird species
endemic to the island of Jamaica.
© Mike Read/naturepl.com
On the opposite page, Ricord’s
iguana (Cyclura ricordi), a
Critically Endangered iguana
species that is endemic to a very
small area in the southern part of
the Dominican Republic on the
border with Haiti. It is sympatric
with the more widespread
rhinoceros iguana (C. cornuta),
from which it can be readily
distinguished by its red eyes.
© Cristina G. Mittermeier
112
Although the Mexican portion of the hotspot accounts
for the largest relative area of land under protection
(41 835 km2), this represents only 6.7% of the total
625 700 km2 of the area of the hotspot in Mexico.
Conservation action to counter the ongoing threats
to the Mesoamerica Hotspot concentrates on the cre-
ation and maintenance of the Mesoamerican Biological
Corridor, as well as numerous other initiatives to inte-
grate sustainable development with biodiversity con-
servation. Adopted by a host of donors, the region’s
governments, and four major international conserva-
tion groups —Conservation International, the World
Wildlife Fund, The Nature Conservancy, and the Wild-
life Conservation Society— during the 1980s, the Meso-
american Biological Corridor has three objectives: to
maintain the integrity of the protected areas that al-
ready exist in the Mesoamerica Hotspot; to determine
where new protected areas need to be created in order
to extend the corridor concept; and to connect these
new and existing protected areas through the establish-
ment of high conservation value forestry plantations
and agroforestry systems that run between the corri-
dor’s existing parks, refuges, and biosphere reserves.
This large-scale corridor strategy has gained wide sup-
port because it presents a simple, yet elegant structure
on which to promote a wide range of conservation ac-
tivities, such as biodiversity protection, buffer zone
management, regional planning, and sustainable pri-
vate enterprises based on regional development goals.
This region is particularly important strategically for
food security, since it is home to wild relatives of im-
portant crops such as maize, cacao, beans, and others.
A regional perspective such as the Mesoamerican Bio-
logical Corridor, among others, offers a much broader
perspective concerning the natural range of variability
seldom captured by national policy and within political
borders. Planning should seek to combine local policy
with the specific objectives of habitat connectivity for
area-sensitive species, such as the jaguar; habitat-special-
ists, such as the resplendent quetzal; and narrow-ranging
species, such as some Atelopus toad species. Equally im-
portant is the representation of the full range of the wide
variety of ecosystems within protected areas. There is
great potential to develop a regional network of trans-
boundary protected areas, such as exists between Costa
Rica and Panama (La Amistad International Park), a con-
cept which is emerging elsewhere as well. The Critical
Ecosystem Partnership Fund (CEPF) is facilitating a re-
gional approach to promote conservation by supporting
bi- and tri-national initiatives that rely on cooperation of
governments and NGOs. The fact that there are eight
countries in this hotspot, each with disparate political
agendas and motivations, makes this a great challenge.
The Mesoamerica Hotspot also has within it a global-
ly recognized model for the value of ecotourism as a for-
eign exchange earner and as an alternative to harmful
extractive industries. For a variety of reasons, Costa Rica
recognized early on the great potential for ecotourism
and structured itself to take advantage of this source of
income. As a result, Costa Rica earned $1.25 billion in
tourism-related industries in 2000, and it is estimated
that 70.7% of tourists visit natural protected areas —an
enormous figure for a small country with a population
of only 3.9 million people. On the other hand, there re-
mains a very delicate balance between tourism poten-
tial and agriculture-based revenue generation.
In conclusion, although the problems in this region
are many and the commitment to conservation varies
from country to country, there exists in Mesoamerica a
variety of mechanisms for regional-level government dia-
logue on the environment which are far better than in
many other hotspots that spread over several sovereign
nations. Other positive indicators include the shared vi-
sion of the Mesoamerican Biological Corridor; the high
percentages of national territory already in parks and
reserves in four of the eight nations (Costa Rica, Guate-
mala, Panama, Belize); the strong and growing local ca-
pacity that exists within the region; the proven value of
alternatives like ecotourism and non-timber forest prod-
ucts; the development of several national-level institu-
tions for biodiversity research and conservation; the
promotion of sustainable use initiatives; and the contin-
uing interest and involvement of a variety of donor
agencies and international conservation organizations.
CARIBBEAN ISLANDS
MICHAEL LEONARD SMITH 2 • S. BLAIR HEDGES 117
WILLIAM BUCK 118 • ARLO HEMPHILL 1
SIXTO INCHAUSTEGUI 119 • MICHAEL A. IVIE 120
DON MARTINA 121 • MICHAEL MAUNDER 122
JAVIER FRANCISCO ORTEGA 122, 123
U.S.A.
BAHAMAS
CUBA DOMINICAN
REPUBLIC ATLANTIC
HAITI OCEAN
JAMAICA PUERTO
RICO Lesser
CARIBBEAN SEA Antilles
0 600 km
The Caribbean Islands Hotspot consists mainly of three
large groups of islands between North and South Ameri-
ca: the Bahamas, the Greater Antilles, and the Lesser An-
tilles. In contrast to the previous definition under the
name Caribbean Hotspot, we now exclude southern Flor-
ida and its islands on the North American Continental
Shelf because that area has greater floristic affinity to the
rest of the continent, has few endemics, and has had
only a marginal role in the phytogeographic history of
the Antilles (Graham 2003). The Caribbean Islands them-
selves vary in their floristic affinities, but have in com-
mon a history of limited contact with the diverse biotas

On pp. 114-115, Los Haitises
National Park in the Dominican
Republic, one of an extensive
network of protected areas in this
small island nation.
© Patricio Robles Gil/Sierra Madre
Above, besides a population
recently discovered on Isla de la
Juventud, the Endangered Cuban
crocodile (Crocodylus rhombifer)
is known only from the Zapata
Swamp, also in Cuba. Three other
highly threatened species are found
only in this extremely important
national park.
© Neil Lucas/naturepl.com
116
of the continents. As a result, their biotas share an “oce-
anic” character marked by a relatively low representation
of higher taxa, but also having extraordinary diversity
within those higher phyletic groups that are present.
The Caribbean Islands Hotspot is commonly referred
to as the West Indies in the English-language literature.
Our definition differs only in a few minor details from a
biotic region formally defined by James Bond in his Birds
of the West Indies (1936). The islands of the West Indies
have a terrestrial surface area of only 229 549 km2. The
four largest islands, Cuba (105 806 km2), Hispaniola
(73 929 km2), Jamaica (11 190 km2), and Puerto Rico
(9 100 km2) make up over 90% of the land area of the
hotspot, but the myriad smaller islands harbor large
numbers of endemic species far out of proportion to their
size. Geologically, the West Indies are a mix of old regions
of the Caribbean Tectonic Plate (Cuba, Jamaica, Hispan-
iola, Puerto Rico, the Virgin Islands, and their satellite is-
lands), the tops of volcanoes (the active volcanic arc from
Saba to Grenada), and slightly raised limestone platforms
(the Caribees from Sombrero to Barbados, plus the Ba-
hamas, the Cayman Islands, southern Dutch islands, and
Aves Island). Elevations range from over 3 000 m (the
formerly glaciated summit of Pico Duarte) to a desert de-
pression 40 m below sea level, both on Hispaniola.
Politically, this is a very diverse hotspot, with thou-
sands of islands governed by 18 nations. Twelve of these
are independent island nations (Commonwealth of the
Bahamas, Republic of Cuba, Jamaica, Republic of Haiti,
Dominican Republic, Federation of St. Kitts and Nevis,
Antigua and Barbuda, Commonwealth of Dominica, Bar-
bados, St. Lucia, St. Vincent and the Grenadines, and
Grenada). Fifteen polities are dependencies of, in free
association with, or integral parts of: the United States
of America (Puerto Rico, U.S. Virgin Islands, Navassa);
the United Kingdom (Cayman Islands, Turks and Caicos
Islands, British Virgin Islands, Anguilla, Montserrat);
the Kingdom of the Netherlands (Aruba and the Nether-
lands Antilles including Curaçao, Bonaire, Saba, St. Eu-
statius, and St. Maartin); the Republic of France (De-
partments of Martinique and Guadeloupe, including its
dependencies St. Martin and St. Barthelemy); the Re-
public of Honduras (Swan Islands), or the Bolivarian
Republic of Venezuela (Aves Island west of Dominica,
and oceanic islands from Blanquilla to Islas las Aves).
The vegetation of the Caribbean Islands is variable
due to the influences of climate and Earth history. Low-
lying islands tend to be semiarid, and most were origi-
nally dominated by dry evergreen bushland and dry
evergreen thicket, with savanna occurring on parts of
Barbuda, Jamaica, and Puerto Rico (where the average
rainfall at low elevations is only 300-600 mm per year).
On the other hand, wetter environments occur where
trade winds encounter the higher Caribbean mountains,
giving rise to a variety of moist tropical forest types in-
cluding marsh forest, various types of seasonal forest,
montane forest, and elfin woodland (Beard 1955). In
moister areas, around lagoons and river mouths, per-
manent brackish and freshwater swamps give way to ex-
tensive mangrove forests. The previously extensive low-
land rainforests have mostly been destroyed.
Plant diversity and endemism in the Caribbean Islands
Hotspot are both very high, with a total of 13 000 species
estimated to occur in the Caribbean region (Davis et al.
1997), including perhaps 6 550 single-island endemics.
Cuba’s flora is particularly rich, with an estimated 6 505
vascular plant species, of which 3 224 are endemic (Davis
et al. 1997); this represents 54% of the endemic plants for
the hotspot as a whole, and indeed would qualify Cuba as
a hotspot in its own right. Plant endemism at the generic
level is also high in these islands. Of an estimated
2 500 genera of seed plants in the Caribbean, 204 an-
giosperm genera and one gymnosperm genus (Microcy-
cas) are endemic to the Greater Antilles. Of these, fully
118 are restricted to single islands. Important families
with endemic genera include Asteraceae (32), Rubiaceae
(30), Euphorbiaceae (14), and Leguminosae (13). There
is one endemic plant family, the Goetziaceae (Davis et al.
1997). Even mosses, with notoriously broad distributions,
have around 500 species in the Caribbean Islands, with
about 10% endemism (Delgadillo et al. 1995).
Vertebrate diversity and endemism in this hotspot
are noteworthy. Mammals are represented by 89 extant
species, of which 41 are endemic, including two en-
demic families: the solenodons (Solenodon spp.), with
two species of rare giant shrews, and a large radiation of
rodents called hutias (family Capromyidae), which are
related to guinea pigs. The region hosts 15 endemic
genera, including the fruit-eating bat genus Brachy-
phylla, with two species.
Birds are represented by 607 species, of which 167 are
endemic, with many of those restricted in total range to
small areas within islands. A remarkable 35 genera are
endemic, and there are also two endemic families: the
palmchat (Dulus dominicus) of the family Dulidae, and
the todies (Todidae). The bird list includes some impor-
tant flagship species, such as the St. Vincent parrot
(Amazona guildingii, VU) from the island of St. Vincent,
the St. Lucia parrot (A. versicolor, VU) from the island
of the same name, and the imperial parrot (A. imperia-
lis, EN) from Dominica; the bee hummingbird (Mel-
lisuga helenae) from Cuba, the world’s smallest bird; and
the ivory-billed woodpecker (Campephilus principalis,
CR), last recorded with any certainty in Cuba in 1987.
The Caribbean Islands Hotspot is particularly rich in
reptiles with 499 native species, of which 468 are endem-
ic. There are several large evolutionary radiations of liz-
ards, such as the anoles (Anolis; 154 species, 150 endem-
ic) with their colorful dewlaps used in displays; dwarf
geckos (Sphaerodactylus; 86 species, 82 endemic) that in-
clude the world’s smallest lizards (e.g., S. ariasae, with a
body length of 18 mm); and curly tails (Leiocephalus; 23
species, all endemic) that hold their tails in a coil as they
run. There are 11 species of rock iguanas (Cyclura), in-
cluding some measuring more than one meter in length,
and one that is blue (C. lewisi); one of these species, the
Jamaican iguana (C. collei, CR), is confined to the Hell-
shire Hills in Jamaica. A species considered to be an
ancient relict in the lizard family Xantusiidae, Cricosaura
typica, occurs only in a remote part of eastern Cuba. Ma-
jor radiations of snakes include the large boas (Epicrates,
nine species); a genus of boldly patterned snakes that
change colors (Tropidophis; 26 species, all endemic); fast-
moving racers (Alsophis; 13 species, all endemic); and
some pencil-thin and smaller burrowing snakes (Typhlops
and Leptotyphlops) that include the smallest snake in the
world (L. bilineata; 108 mm). The Aruba Island rattle-
snake (Crotalus unicolor, CR) is found only on Aruba.
All 165 native species of amphibians present in the
hotspot are frogs, of which 164 (99%) are endemic
to the region. All but a few are endemic to single is-
lands. The frog fauna is dominated by those of the genus
Eleutherodactylus (139 spp.), which are forest frogs that
lay eggs on land and hatch into miniature adults with no
tadpole stage. One Cuban species (E. iberia, CR) is the
smallest tetrapod in the Northern Hemisphere, with a
length of only 10 mm, while a golden-colored species in
Puerto Rico, possibly Extinct, is one of only a few species
of frogs in the world known to be live-bearing. One of the
largest tree frogs (Hylidae) in the world, the Jamaican
snoring frog (Osteopilus crucialis, EN), has a length of
about 120 mm and occurs in Jamaica, where males of this
declining species make a loud snoring call from within
giant, hollow trees. The toads (Bufo; 11 species) have also
radiated in the West Indies, and captive-breeding pro-
grams have been implemented for the Puerto Rican
crested toad (B. lemur, CR). An edible species of frog en-
demic to Dominica and Montserrat, the “mountain chick-
en” (Leptodactylus fallax, CR), is one of the largest frogs in
the Western Hemisphere, but has been rapidly declining
in numbers due to human consumption, habitat loss,
and an outbreak of chytridiomycosis on Dominica.
The hotspot’s inland fishes include 161 species, of
which 65 are endemic to one or a few islands. Nearly
half of the endemics are restricted to very small
ranges, often consisting of a single lake or springhead,
and these sites constitute the sole opportunities for
conservation of these species in nature. As in other
hotspots that are composed of islands, the freshwater
fish fauna consists of two distinctive elements. The
smaller and younger islands have faunas dominated
by species that are widespread in marine waters, but
that enter fresh water to some degree, especially
when obligate freshwater species are absent or few in
number. This component accounts for the relatively
high diversity at the family level. The larger and older
islands of the Greater Antilles differ in having faunas
dominated by several groups that are old enough to
have occupied inland waters of the proto-islands and
continental coasts prior to extensive plate tectonic
movements. These groups (e.g., gars, killifishes, sil-
versides, and cichlids) include significant local radia-
tions, which account for the hotspot’s moderate num-
ber of endemic species. The hotspot’s five endemic
fish genera all have distributions that overlap in west-
ern Cuba, apparently the oldest part of the hotspot
that has remained continuously above sea level.
Humans first populated the islands about 4 000 years
ago, and there is evidence that even the early inhabi-
tants were involved in modifying the biota through di-
rect use or, perhaps more far-reaching, through the in-
troduction of species from the continents. Early species
introductions include the agouti on Dominica, tortoises
on many islands, and possibly even large animals such
as white-tailed deer (Odocoileus virginianus) on Curaçao.
Introductions of species were enormously accelerated
after the arrival of Europeans, whose transportation
technologies led to the import of species —both inten-
tionally and accidentally— from all parts of the globe.
The introduction of the Indian mongoose (Herpestes au-
ropunctatus) from Asia in 1872 resulted in a series of ex-
tinctions as it was moved from island to island. Even
small, uninhabitable islands such as Navassa and Som-
brero now have floras dominated by continental weed
species, and throughout the hotspot native amphibian,
reptile, and bird faunas have been devastated by rats and
domestic cats and dogs. In total, some 36 vertebrate
species are considered to have gone extinct in the region
since 1500, including species like the Cuban macaw (Ara
tricolor), Jamaican giant galliwasp (Celestus occiduus), and
four species of Nesophontes (relatives of the solenodons).
The islands have been subject to Western-style devel-
opment including extensive monoculture for five hun-
dred years, longer than any other part of the New World.
For most of this period, the exploration and develop-
ment of the islands’ natural resources were carried out
by colonial centers outside the Caribbean Basin. In many
cases, Caribbean natural resources were simply liquidat-
ed (e.g., semiprecious hardwoods of Haiti and many oth-
er states). The widespread cultivation of sugarcane (Sac-
charum officinarum) was also conducted for the benefit of
foreign economies and, similarly, resulted in broad trans-
formation of island landscapes. Consequently, as a result
of changes that have taken place since European arrival,
only a small portion of the vegetation that once existed
still remains in more or less pristine condition, and it is A handful of hatchling Cuban
estimated that intact vegetation covers no more than crocodiles from a breeding facility in
10% of the original extent of the land area. Interestingly, Cuba. This Endangered species is
endemic to Cuba and is now
although less than 15% of Cuba’s original forests remain
restricted to the Zapata Swamp and
intact, they still represent the largest forested areas re- Isla de la Juventud.
maining in the Antilles. © Cristina G. Mittermeier
According to the World Database on Protected Areas,
some 12.9% of the hotspot is officially protected in a va-
riety of different conservation units, although only
7.1% is conserved in protected areas classified in IUCN
categories I to IV. In the largest Caribbean country,
Cuba, about 15% of the total land area falls within pro-
tected areas, including the 300-km2 Zapata Swamp,
home to the Cuban crocodile (Crocodylus rhombifer,
EN), the Zapata rail (Cyanolimnas cerverai, EN), the
Zapata wren (Ferminia cerverai, EN), and the dwarf
hutia (Mesocapromys nanus, CR), all threatened species
found nowhere else. The country of Dominica leads in
percentage coverage, with 21.4% of its territory desig-
nated for protection, while other countries also report
relatively high protection (for example, the Dominican
117
 Republic, with 15%). However, in many of these coun-
tries, the existing protected area network is ineffective
and poorly managed, while in other nations in the
Caribbean protected areas are almost nonexistent. Haiti
and Grenada, for example, both have only 1.7% of their
respective land areas under protection, while Barbados
and Aruba Island each have less than 1%. Indeed, in a
recent global gap analysis, the Caribbean Islands emerge
as a region of high urgency for expansion of the pro-
tected areas network (Rodrigues et al. 2003).
Prospects for the protection of biodiversity have
been greatly enhanced by the development of alliances
between major industries, such as tourism, and the
governmental and private organizations that carry out
conservation on the ground. Protected area systems are
now being designed for the dual goals of safeguarding
biodiversity and contributing to island livelihoods. A
significant advance occurred in 2000, when the Proto-
col for Specially Protected Areas and Wildlife (SPAW)
came into force. Created at the initiative of the Car-
ibbean countries themselves, this protocol provides
region-wide standards and mechanisms for harmoniz-
ing conservation efforts across the hotspot’s diverse cul-
tures and political systems.
CALIFORNIA
FLORISTIC PROVINCE
WILLIAM R. KONSTANT 1, 7 • DEAN TAYLOR 100
DAVID A. WAKE 101 • SCOTT ROBBINS LOARIE 102
ROXANNE BITTMAN 103 • BARBARA ERTTER 104
U.S.A.
On the opposite page,
California
Channel
rhododendrons in bloom, Redwood
Islands
National Park, California, U.S.A.
© Carr Clifton/Minden Pictures PACIFIC
OCEAN
Guadalupe
Island
Cedros
0 400 km
Island
The California Floristic Province is one of the five
Mediterranean-type hotspots and the only hotspot that
occurs largely within the borders of the United States of
America. Stretching nearly 1 800 km along the western
coast of North America, most of its 293 804 km2 are
118
found within the State of California, with extensions
into southwestern Oregon and northwestern Baja Cali-
fornia, Mexico. All areas west of the peaks of the Cas-
cade and Sierra Nevada ranges are included within
its boundaries, as are a handful of islands off Califor-
nia’s southwestern coast, including the Channel Islands
(913 km2), and Isla Guadalupe (264 km2), located some
300 km west of Baja California.
As its name implies, the California Floristic Prov-
ince is an ecological construct based on plant species
composition, a unique mixture of northern temperate
and southern xeric elements fostered by a Mediter-
ranean climate of hot, dry summers and cool, wet win-
ters. Four other hotspots share this climate: Central
Chile, the Cape Floristic Region, Southwestern Aus-
tralia, and the Mediterranean Basin (Barbour et al.
1993; Dallmann 1998).
Four subregions within the Province stand out as
centers of exceptionally high plant diversity: the Sierra
Nevada, Transverse Ranges, Klamath-Siskiyou region,
and Coast Ranges (Stebbins 1978; Davis et al. 1997).
Rare plant communities of the southern Sierra Nevada
include the giant sequoia (Sequoiadendron giganteum) for-
est, piute cypress woodland, and lone manzanita shrub-
land (Grossman et al. 1994). The Transverse Ranges are
a narrow strip that runs east to west in southern Cali-
fornia, separating the Coast Ranges to the north from
the Peninsular Ranges to the south. At least 10 rare
plant communities have been recorded from this re-
gion. The Klamath-Siskiyou region bridges the coastal
mountain ranges of California and Oregon, and is home
to approximately 20 rare plant communities, including
the most diverse temperate coniferous tree community
in the world (Vance-Borland et al. 1995-1996). This re-
gion also represents the contact zone between the Pa-
cific Northwest Floristic Province and the California
Floristic Province. The Coast Ranges comprise a wide
variety of habitats, including coastal dune, coastal salt
marsh, maritime chaparral, coastal cypress forest, red-
wood forest, mixed evergreen forest, mixed hardwood-
redwood forest, northern yellow pine forest, southern
oak forest, Calocedrus forest, mixed hardwoods, valley oak
savanna, coastal prairie scrub, vernal pools, and fresh-
water marshes, within which at least a dozen rare plant
communities can be found (Steinhart 1994; Davis et al.
1997). The unusually high plant diversity in the Coast
and Transverse ranges, coupled with the dispropor-
tionate human habitat preference for lowlands and
foothills near the coast, puts these regions at special
MEXICO
risk (Seabloom et al. 2002).
Some of the highest levels of plant diversity within
the California Floristic Province are found in the south-
ern part of the Sierra Nevada Range and in the Kla-
math-Siskiyou region (Davis et al. 1997). In addition,
serpentine soil habitats occur along fault zones in
the Central and North Coast and Cascade ranges, from
sea level to an elevation of 2 900 m. Due to specific
chemical and physical characteristics of the soils,
these habitats are nutrient-poor, and this has led to the

Chaparral yucca (Yucca whipplei),
or Our Lord’s Candle as it is also
known, in an area of burn recovery
in dry coastal sage scrub following
fires in San Diego County,
California, October 2003. There is
an interesting symbiotic
relationship between this species
and the California yucca moth
(Tegeticula maculata).
© Richard Herrmann
On the opposite page, coreopses
(Coreopsis spp.) and California
poppies (Eschscholzia californica),
the state flower, in Antelope Valley
California Poppy Reserve.
© Carr Clifton/Minden Pictures
120
establishment of a highly specialized and diverse flora
(Davis et al. 1997). It has been estimated that serpen-
tine endemic plant species represent 10% of the Cali-
fornia Floristic Province’s endemics (Kruckeberg 1984).
High levels of species diversity and endemism have
developed within this region due to its varied topogra-
phy, climate zones, geology, and soils. Plant diversity is
exceptional, with 3 488 native vascular species, includ-
ing 2 124 endemics. Fifty-two of the region’s plant gen-
era are also unique. The total number of plant species
is greater than that for the central and northeastern
United States and adjacent portions of Canada, an area
almost 10 times as large (Raven and Axelrod 1978;
Raven 1988; Davis et al. 1997).
The region’s impressive endemism is a mixture of
outstanding relicts and newly derived species (Stebbins
and Major 1965). Many of the relicts once had much
larger ranges in the Tertiary, but during the cooling and
drying trends that accompanied the ice age in North
America, these species became restricted to the Cali-
fornia region’s relatively mild climate. Because many of
these relicts have no extant close relatives, they are par-
ticularly important botanically. The most famous rel-
icts that best symbolize the region are the giant sequoia
and coastal redwood (Sequoia sempervirens). The giant
sequoia, known to reach more than 75 m in height and
30 m in circumference, is believed to be the largest
species that ever lived. Although coastal redwoods are
among the world’s tallest trees, with record specimens
surpassing 105 m, they are far less massive than their
relatives. Other noteworthy relict trees include the
ironwood trees (Lyonothamnus floribundus) endemic to
Santa Cruz and Santa Catalina islands, and the Califor-
nia bay laurel (Umbellularia californica). The region’s
newly produced species result from recent outbursts of
speciation within certain genera —primarily annual
herbaceous dicots— in response to climate fluctuations
since the Middle Pliocene in areas increasingly diverse
geographically (Raven and Axelrod 1978). Genera such
as Clarkia, Lasthenia, and Phacelia have undergone re-
markable radiations. The above-mentioned serpentine
endemic species represent examples of such newly pro-
duced endemics.
While plant endemism is impressive, vertebrate di-
versity and endemism are less so. Of approximately 341
resident, breeding, and migrant bird species, only eight
are endemic, as are 18 of the 151 native terrestrial
mammals. Four of the 69 native reptiles are endemic,
including two species found only on Cedros Island,
12.9 km off the Baja California Peninsula: the Cedros Is-
land diamond rattlesnake (Crotalus exsul) and Cedros
Island horned lizard (Phrynosoma cerroense).
The highest levels of endemism are found among
the amphibians, with 25 endemic species out of a to-
tal of 54. The salamander fauna of California is espe-
cially noteworthy for the high degree of endemism
and its uniqueness: 38 described species, 24 of which
are endemic. The region contains one of the two best-
understood complexes, the ensatina (Ensatina eschschol-
tzii), which demonstrate stages in the pattern and
process of species formation. It also contains the arbo-
real salamanders (genus Aneides, with three endemic
species), which ascend to the tops of the tallest red-
woods. The slender salamanders of the genus Batra-
choseps are especially diverse, and only one of the 18
currently recognized species does not occur in the
hotspot. Included in this genus are the only truly desert-
adapted fully terrestrial salamanders (two species), as
well as the distinctive San Gabriel Mountains slender
salamander (B. gabrieli), discovered recently in moun-
tains in the Los Angeles Metropolitan Area.
Isolated from the large eastern North American fish
fauna by the western mountains and deserts, the Cal-
ifornia Floristic Province Hotspot has a relatively small
fauna of inland fishes comprising 73 native species.
One of the most distinctive elements is a concentra-
tion of lamprey species that includes a cluster of lo-
calized landlocked species in the northern mountains.
Lampreys represent the deepest offshoot of living ver-
tebrates, and eight of the world’s 43 surviving species
occur in this hotspot. Together with two sturgeons that
represent another of the deepest vertebrate lineages,
these species cause the hotspot to rank with Japan as
global leaders in phyletic rarity for fishes.
Among vertebrates, there are some important flag-
ship species. The grizzly bear (Ursus arctos) has been
the California state symbol for more than 150 years,
and is immortalized in the names of creeks, mountains,
peaks, and valleys throughout California. Some might
consider it a classic flagship species for this hotspot, de-
spite the fact that it has been extirpated; the last Cali-
fornia grizzly was shot sometime in the 1920s (Wirka
1994). Also, of the world’s 13 subspecies of elk or wapi-
ti, two —Roosevelt’s elk (Cervus elaphus roosevelti) and
the tule elk (C. e. nannodes), the largest and the small-
est of the North American subspecies, respectively—
are native to the California Floristic Province and
served as flagships for conservation throughout the last
century. One species that has received a great deal of
attention is the endemic California tiger salamander
(Ambystoma californiense, VU), currently a major point
of contention in rapidly growing Sonoma and Santa
Barbara counties. Animals perhaps less well-known,
but very effective symbols for local action, include the
San Francisco garter snake (Thamnophis sirtalis tetratae-
nia), California gnatcatcher (Polioptila californica), and
island fox (Urocyon littoralis), the latter with six subspe-
cies all confined to the six largest of the eight California
Channel Islands. In salt marshes around San Francisco
Bay, the salt-marsh harvest mouse (Reithrodontomys ra-
viventris, VU) faces dual pressures from encroaching
development and feral house cats.
The most important flagship, however, remains the
California condor (Gymnogyps californianus, CR), North
America’s largest bird, and a strong symbol for wildlife
conservation in the western part of the continent.
Native Americans considered the condor a sacred ani-
mal with supernatural powers, but westward-moving
pioneers shot these birds for target practice, collected
their eggs, and poisoned the animals upon which they
fed. The species was a hair’s breadth away from extinc-
tion. Today, however, thanks to a collaborative captive-
breeding and reintroduction effort involving the U.S.
Fish and Wildlife Service, California Department of Fish
and Game, National Audubon Society, Zoological Soci-
ety of San Diego, Los Angeles Zoo, and The Peregrine
Fund, captive-bred California condors are being re-
turned to the wild.
California is the most populated (estimated at 35 mil-
lion people in 2002; U.S. Census Bureau) and fastest-
growing state in the United States, and this severely im-
pacts biodiversity within the hotspot (Jensen et al.
1993). Urbanization, air pollution, agricultural expan-
sion, logging, strip mining, oil extraction, road con-
struction, livestock grazing, the spread of non-native
plants, an increasing use of off-road vehicles, and the
suppression of natural fires all pose significant threats.
Human population pressures have rendered California
one of the four most ecologically degraded states in the
country, with all or part of the nation’s eight most
threatened ecosystems represented: beach and coastal
strand, southern California coastal sage scrub, large
streams and rivers, California riparian forests and wet-
lands, California native grasslands, old-growth pon-
derosa pine forests, cave and karst systems, and the an-
cient forests of the Pacific Northwest, which include the
coastal redwoods (Noss 1994; Noss and Peters 1995).
Today, native grasslands and vernal pool habitats
have been reduced to perhaps 1% of their original ex-
tent (Holland and Jain 1990), and wetlands, riparian
woodlands, and southern maritime sage scrub to 10%
or less. Saltwater and freshwater wetlands once cov-
ered almost two million hectares in California, but
have since declined by about 90% due largely to land-
filling and the diversion of river systems. As a result,
so have significant shellfish, fish, and waterfowl pop-
ulations (Barbour et al. 1993). Along California’s low-
land rivers, riparian forests provide essential wildlife
habitat. However, the 4 000 km2 or more of riparian
forests found two centuries ago in California’s Central
Valley have been reduced by 90% due to logging, graz-
ing, and industrial development. Along the coast,
redwoods evolved to thrive in a mild, foggy maritime cli-
mate until logging began in earnest in the mid-nine-
teenth century, largely to supply construction needs
in San Francisco. Mechanized timber extraction has
effectively reduced original stands by 85% since cut-
ting began, although, as Sequoia is one of the only
stump-sprouting conifers, many of these stands have
regenerated. In total, we estimate that no more than
25% of the California Floristic Province remains in
pristine condition.
California has a long history of conservation. In
1864, U.S. President Abraham Lincoln established
California’s Yosemite Valley and the Mariposa Grove
of Giant Sequoias as the first national area to receive
protection explicitly for public use (Turner 1991).
Together with Sequoia National Park, it offers protec-
tion for outstanding biodiversity within the southern
Sierra Nevada. Redwood National Park, officially es-
tablished in 1968 (and expanded in 1978), has evolved
over more than 35 years of intensive conservation ef-
fort. By contrast, the 1 010-km2 Channel Islands Na-
tional Park provides protection for nesting colonies of
seabirds and breeding populations of seals and sea li-
ons. In total, approximately 37% of this hotspot has
some form of official protection, although protected
areas offering a stricter level of protection (IUCN cat-
egories I to IV) cover only 10.2% of the hotspot’s sur-
face area. The creation of many protected areas was
brought about largely through the actions of groups
such as the Sierra Club, Wilderness Society, and The
Nature Conservancy. Unfortunately, while a consider-
able degree of protection has resulted for the high-
elevation flora, much less, if any, protection has been
accomplished for the serpentine or vernal pool sub-
sets of the endemic flora.
In conclusion, while the California Floristic Prov-
ince lies largely within one of the world’s richest na-
tions and contains some of Earth’s most famous and
most popular national parks, it suffers from threats
similar to those operating in hotspots found within
countries that are much more disadvantaged eco-
nomically. Furthermore, a great deal remains to be
done in order to ensure that the unique and threat-
ened biodiversity of this hotspot is adequately safe-
guarded in suitable protected areas. Biodiversity loss
clearly is not a problem unique to developing tropical
nations.
GUINEAN FORESTS
OF WEST AFRICA
MOHAMED BAKARR 55 • JOHN F. OATES 56 • JAKOB FAHR 57
MARC PARREN 58 • MARK-OLIVER RÖDEL 59 • RON DEMEY 60
GUINEA BENIN On the opposite page, West African
TOGO
COTE NIGERIA forests still harbor many species
SIERRA D IVOIRE new to science, such as this
GHANA
LEONE attractive unnamed toad of the
CAMEROON
LIBERIA
Bioko Bufo togoensis complex.
(EQUATORIAL GUINEA)
ATLANTIC OCEAN © Piotr Naskrecki
SAO TOMÉ
0 800 km AND PRÍNCIPE
The Guinean Forests of West Africa Hotspot extends
from Guinea and Sierra Leone eastwards to the Sanaga
River in Cameroon. These forests originally covered an
estimated 620 314 km2 and form the westerly part of
the Guineo-Congolian Regional Center of Endemism
defined by White (1983). Although biogeographically
distinct, the hotspot comprises two forest blocks: the
123

On pp. 124-125, a small stream
flowing along the floor of the
rainforest in the Guinean Forests
of West Africa Hotspot.
© Piotr Naskrecki
Above, a hingeback tortoise
(Kinixys erosa) in Tai National
Park, Cote d’Ivoire. Hingebacks are
widely distributed in sub-Saharan
Africa, especially in forested areas,
but are rarely seen.
© Patricio Robles Gil/Sierra Madre
On the opposite page, Lowe’s
guenon (Cercopithecus lowei)
in the Tai Forest region of Cote
d’Ivoire. This species is endemic
to the Guinean Forests of West
Africa Hotspot.
© Patricio Robles Gil/Sierra Madre
126
Upper Guinea block, which extends from Guinea to
Togo; and the Nigeria-Cameroon block, extending to the
Sanaga River. The two units are separated by the Da-
homey Gap in Benin, an area that is now a mixture of
farmland, derived savanna woodland, and relict patch-
es of dry forest.
Although the West African forest region traditionally
is separated from the Central African forests by the
Cross River in eastern Nigeria, the hotspot as here de-
fined extends into Cameroon as far as the Sanaga River
(Oates 1986; Martin 1991; IUCN 1996), and includes
such high-priority areas as Korup National Park and the
Cameroon Highlands. The Nigeria-Cameroon border
region is under a level of threat comparable to that of
Upper Guinea. The hotspot also includes several im-
portant montane regions, such as the aforementioned
Cameroon Highlands (the high point being Mount Cam-
eroon, at 4 095 m) and the Upper Guinea, or Greater Nim-
ba, Highlands, between Guinea, Sierra Leone, Liberia,
and Côte d’Ivoire.
The two major forest blocks in this hotspot also cor-
respond to two important centers of endemism: Upper
Guinea, which corresponds to that portion of the hot-
spot west of the Gap, and the Nigeria-Cameroon border
region plus Bioko Island. The region from Ghana to the
Cross River is transitional and supports overlapping
generalist species shared by the two forest blocks, such
as the mona monkey (Cercopithecus mona) and chim-
panzee (Pan troglodytes, EN).
Also included in this hotspot are four islands located
in the Gulf of Guinea to the west of the southern limits
of the region (and sometimes referred to as the “Galá-
pagos of Africa”): Bioko (Fernando Po; 2 017 km2) and
Annobon (Pagalu; 17 km2), both now part of Equatorial
Guinea; and São Tomé (857 km2) and Príncipe (139 km2),
which together constitute an independent nation.
Bioko is a continental-shelf island, supporting a much
more diverse flora and fauna, but with a relatively low
rate of endemism at the species level, whereas the re-
maining three are oceanic and relatively depauperate
because of their isolation, but rich in endemic taxa
(Jones 1994).
The Guinean forests include a range of distinct veg-
etation zones varying from moist forests along the
coast, freshwater swamp forests (for example, around
the Niger Delta), to dry, semi-deciduous forests in-
land with prolonged dry seasons. Of all West African
countries, only Liberia lies entirely within the moist
forest zone, although a substantial portion of Sierra
Leone also falls within the boundaries. In Ghana, de-
rived savanna reaches the coast around Accra, where
the average annual rainfall falls below 1 200 mm.
This savanna merges into the highly threatened
dry forest of the Togo-Volta Highlands and the Da-
homey Gap, which ends in eastern Benin. Moist for-
est continues from there —originally uninterrupted—
through the Cross River region in southern Nigeria to
Cameroon.
Vascular plant diversity for the entire Guinea-Congo-
lian phytochorion has been estimated to be close to
12 000 species, with 6 400 endemics (Davis et al. 1994).
Upper Guinea alone holds about 2 800 species, of which
nearly 25% are lianas, while some 530 vascular plants
are endemics (Poorter et al. 2004). Site-specific studies
such as those of Cable and Cheek (1998), who recorded
2 435 plant species from Mount Cameroon (49 of them
endemic), suggest that overall plant diversity in the
Guinean forest ecosystem may be as high as 9 000
species. Based on these figures, knowledge of the 185
endemics recorded from the Gulf of Guinea, and using
endemism in the Upper Guinea block as an indication
of endemism overall, the figure for vascular plant en-
demism for the region as a whole is estimated at 20%,
or 1 800 species.
The Guinean Forests Hotspot also contains a rich
and unique faunal assemblage. An important feature of
the Guinean forest fauna is that many of the endemic
species tend to have highly restricted ranges within the
hotspot, making many of these species extremely vul-
nerable to forest destruction. For example, the Upper
Guinea Highlands are home to an exceptionally high
number of endemic bats and amphibians with restrict-
ed ranges, all of which are Critically Endangered or En-
dangered.
About 63 of the estimated 320 mammal species are
endemic to the region; in addition, there are seven en-
demic genera, including monotypic genera such as the
Liberian mongoose (Liberiictis kuhni, EN). Another mono-
typic genus, the pygmy hippo (Hexaprotodon liberiensis,
VU), an appealing miniature of its huge cousin, is en-
demic to Guinean forests, with one subspecies in Upper
Guinea and one in the Niger Delta. This hippo is an im-
portant flagship species, as are a handful of other Upper
Guinea endemics such as Jentink’s duiker (Cephalophus
jentinki, VU), the zebra duiker (C. zebra, VU), the Diana
monkey (Cercopithecus diana, EN), and Preuss’ guenon
(C. preussi, EN).
Avian diversity is also particularly high, with an esti-
mated 793 species that are considered to occur regular-
ly in this hotspot, of which 75 species and six genera
are endemic. BirdLife International recognizes most of
the Guinean Forests as a conservation priority for birds,
with five Endemic Bird Areas (EBAs) lying partly or en-
tirely within the hotspot. These include the Upper
Guinea Forests, with 15 species confined entirely to
this EBA, including the Liberian greenbul (Phyllastre-
phus leucolepis, CR); and the Cameroon Mountains,
with a remarkable 27 endemic species, among them
the enigmatic Mount Kupe bush-shrike (Telophorus ku-
peensis, EN). São Tomé, Príncipe, and Annobon are all
considered distinct EBAs.
Reptile and amphibian diversity is less well docu-
mented. Preliminary minimum estimates indicate 206
species of reptiles and 246 species of amphibians, of
which roughly 52 reptiles and 83 amphibians are en-
demic to the region as defined. There are also seven
endemic amphibian genera, and a single endemic rep-
tile genus (Cophoscincopus) comprising three lizard
species. Many more species are likely to be discovered;
for example, during the last ten years, 11 new frog
species have been described from the region and many
more still await description. Spectacular species in this
region are the goliath frog (Conraua goliath, EN), the
world’s largest anuran; and the Mount Nimba toad
(Nimbaphrynoides occidentalis, CR), a species outstand-
ing for its reproductive biology in that, following a ges-
tation period of nine months, they give birth to fully de-
veloped toadlets.
There are an estimated 512 species of freshwater
fishes in this hotspot, about 35% of which are endem-
ic, making it a truly remarkable area for fish biodiver-
sity. About 25% of the world’s 350 species of killifish
occur in this hotspot, and about half of these are en-
demic. Cichlid fishes also feature prominently in this
hotspot, with 37 (out of 62) endemic species and five
endemic genera. Four of these endemic genera (Ko-
nia, Myaka, Pungu, and Stomatepia) are known only
from Lake Barombi Mbo, a crater lake in northwest
Cameroon.
The Guinean Forests Hotspot is one of the most se-
verely threatened forest systems in the world, with ex-
treme habitat fragmentation and degradation through-
out most of the region. According to Forest Resources
Assessment 2000 figures, remaining forest cover in each
of the countries within the hotspot totals 716 200 km2,
although this includes all of Cameroon, which falls only
partly in the hotspot. Other earlier estimates put re-
maining forest cover at 15% for the entire Guinean for-
est ecosystem, including the islands (Sayer et al. 1992;
WRI 1992; Parren and De Graaf 1995). Taking the more
or less pristine forests into consideration, we believe
that the total is considerably less, but in the absence of
more solid figures we use the figure of 15% in our glob-
al analysis.
Presently, the major threat is deforestation, due to
commercial logging and slash-and-burn agriculture,
both of which are prevalent in all Guinean forest coun-
tries. Plantation agriculture (e.g., oil palm, rubber, ba-
nanas, and cacao) has been very significant in replac-
ing forest in Côte d’Ivoire, parts of Ghana, southern
Nigeria, western Cameroon, and Bioko. There is al-
ready a negative feedback of deforestation on the re-
gional and local climate, resulting in decreased annu-
al precipitation and prolonged dry seasons. Most of
the moist and dry forests within the Guinean region
are already at the climatic margin where the respec-
tive forest types can persist. The coupled effects of
global and regional climate change seriously threaten
vast stretches of the remaining areas that today are
still forested.
Both small-scale and industrial-scale mining are also
serious threats to the remaining Guinean forests. In
many of the countries, particularly in the Upper Guinea
Highlands, forests are located on substrates that are of-
ten rich in high-grade iron ore, diamonds, gold or baux-
ite. In addition, the Gulf of Guinea harbors one of the
richest oil fields on a global scale.
Bushmeat hunting is one of the major threats to larg-
er animal species in the Guinean forests. Although
bushmeat has long been an important component of
the diet of rural West Africa, growing urban popula-
tions, improved road networks, and increased access to
forests have created a huge commercialized trading sys-
tem for it both nationally and internationally. Numer-
ous studies have indicated that the bushmeat trade in
the region is enormous; estimates of its value in Ghana
run as high as US$400 million per year and for Côte
d’Ivoire, $500 million (see Bakarr et al. 2001). Although
it is unlikely that the bushmeat trade can ever be com-
pletely controlled in an area where it is such an integral
part of the culture, efforts should be made to curb or
eliminate hunting of the most extinction-prone species,
such as red colobus monkeys (Oates et al. 2000). If
such efforts are not promoted, then we will likely see
full realization of an “Empty Forest Syndrome” where-
by, structurally, forest looks rather undisturbed, but in
which the larger mammals are either reduced to very
small numbers or completely extirpated (see, for exam-
ple, Oates 1999).
The region’s political instability has also contributed
to forest destruction. In the 1990s, the United Nations
High Commissioner for Refugees documented over a
million refugees fleeing civil wars and persecution in
Liberia and Sierra Leone and seeking safety in Guinea,
Côte d’Ivoire, and Ghana. This number has increased
significantly in recent years with the outbreak of con-
flict in Côte d’Ivoire. Refugee communities impact
forests by settling in uninhabited areas, placing in-
creasing pressure on resources through demands for
fuelwood and food resources such as bushmeat. Envi-
ronmental impacts of refugees do not necessarily end
when peace accords are signed, and indeed can be ex-
acerbated. Many times, there is a need for more land
during repatriation and resettlement. Very often the On the opposite page, an African
only uninhabited land available is in forest parks and palm civet (Nandinia binotata) sold
reserves. for food by a roadside vendor near
Monrovia, Liberia. A wide range of
Weak and inefficient governance constitutes a prob-
species are sold for the bushmeat
lem in most of the countries in this region, again trade in this region.
meaning that laws and protected areas on paper may © Haroldo Castro
not be well enforced. One of the outcomes is that ille-
gal logging is on the rise since timber is becoming Above, a family of lowland gorillas
scarce. For instance, Ghana is able to produce one mil- (Gorilla gorilla gorilla) killed for
meat in Cameroon. The bushmeat
lion cubic meters sustainably from its forest reserves
trade is widespread in West and
and the wider agricultural landscape. However, in Central Africa and is a major threat
2002, Ghana’s Minister for Lands and Forestry admit- to wildlife throughout the region.
ted that “the total quantity of logs removed in 1999 © Karl Ammann/naturepl.com
amounted to 3.7 million cubic meters, four times the
annual allowable cut.”
Since the late 1960s, efforts have been under way in
all the countries to establish more strictly controlled,
129

The South African bowsprit or
angulate tortoise (Chersina
angulata) is the most common of
the endemic tortoises of South
Africa, and is particularly
abundant in parts of the
Cape Floristic Region.
© Tony Heald/naturepl.com
On the opposite page, a spring
bloom at Ramskop Nature Reserve
in the Clanwilliam area of the
Western Cape. Namaqualand’s
spectacular flowers attract visitors
from around the world
every September.
© Gerald Cubitt
130
and effectively managed, protected areas. However,
the area of forest incorporated into strict protected
areas is still small compared to the area of closed forest
cover remaining in each of the countries. An analysis
of the protected areas coverage in this hotspot yields
some interesting results. Approximately 17% of the
original extent of the hotspot is considered to have
some form of protection; however, when one considers
only those classified in IUCN categories I to IV, then
the percentage drops to just 3%. The management of
protected areas such as national parks has been almost
universally lax, with rampant hunting still occurring,
except in limited areas where there are long-term
wildlife research projects.
To ensure long-term survival of many forest spe-
cies, it is imperative that remaining forest areas be giv-
en more adequate protection and that as many exist-
ing forest reserves as possible be elevated to National
Park status. While this is being done, it is also impor-
tant to consider broader landscape approaches, identi-
fying mechanisms like corridors, in order to connect
these fragmented protected areas with one another
through whatever means possible (Parren et al. 2002).
In 1999, Conservation International held a priority-
setting workshop in Elmina, Ghana, which defined
priority actions and areas for the Upper Guinea forest
block (Bakarr et al. 2001). The following year, the U.S.-
based World Wildlife Fund organized and led a similar
process for the Congo Basin forests, and extended it to
the Nigeria-Cameroon forest block. The two work-
shops have together established a regional vision for
biodiversity conservation in the entire Guinean for-
ests, and should stimulate transboundary conserva-
tion initiatives and foster integration among West
African countries.
Although civil strife in Côte d’Ivoire, Liberia, and
Sierra Leone continues to pose major constraints for
conservation investment in the hotspot, moderate
progress has been made with the regional conserva-
tion vision. With the five-year, $5-million investment
in the Upper Guinea forest block from the Critical
Ecosystem Partnership Fund (CEPF), most priority ar-
eas and actions identified in the region are already be-
ing addressed. The CEPF investment has been partic-
ularly crucial in mobilizing locally based conservation
organizations as well as civil society groups, mainly
through partnerships with international organizations.
In addition, CEPF investment has also catalyzed new
investments from bilateral donors and the private sec-
tor. For example, CI and local partners are imple-
menting conservation activities in the Upper Guinea
Highlands with support from the U.S. Agency for In-
ternational Development (USAID) and Rio Tinto. The
Global Conservation Fund at CI is also contributing
significantly through innovative mechanisms to in-
crease the area of forests under protection across the
hotspot.
CAPE
FLORISTIC REGION
RICHARD M. COWLING 28 • SHIRLEY M. PIERCE 29
SOUTH
AFRICA
INDIAN
OCEAN 0 300 km
The Cape Floristic Region is located at the southwestern
tip of the African Continent and lies entirely within the
borders of South Africa. It is one of the five Mediter-
ranean-type systems on the hotspots list, and is one of
only two hotspots that encompass an entire floral king-
dom (the other being New Caledonia). Indeed, despite
having extremely infertile soils, this region has the great-
est extratropical concentration of plant species in the
world, with 9 000 plant species, 6 210 of them endemics,
crammed into just 78 555 km2. Furthermore, diversity
and endemism are impressive not just at the species lev-
el, but at the generic and familial levels as well, with this
region accounting for five of South Africa’s 12 endemic
plant families. Among the hotspots, only Madagascar
and the Indian Ocean Islands (8) and New Caledonia (5)
have comparable family-level endemism.
The characteristic and most widespread vegetation
of the Cape, covering some 46 000 km2, is fynbos, an
Afrikaans word that translates as “fine bush” (Kruger
1979). Fynbos is a shrubland comprising hard-leafed,
evergreen, fire-prone shrubs, which are, in geological
terms, a relatively recent feature (Cowling and Richard-
son 1995). Before the predominance of these shrub-
lands, the region was covered with a lush rainforest of
mixed Gondwanan-Paleotropical affinities (Linder et al.
1992). With the development of the cold Benguela Cur-
rent along the Cape west coast about 15 million years
ago, there was a subsequent deterioration in climate,
resulting in a retreat of the forests. They were replaced
by flammable sclerophyllous plants, relatives of the an-
cient lineages that had persisted on locally dry sites
since the birth of flowering plants.
Many different fynbos vegetation types have been
described by plant ecologists (Cowling et al. 1997), and
fynbos is characterized by four major plant types:
restioids, ericoids, proteoids, and bulbs. Restioids, main-
ly members of the Gondwanan family Restionaceae,
are evergreen rush or reed-like plants that are the
uniquely diagnostic plant type of fynbos. The ericoids
include more than 3 000 species of small-leafed shrubs
(0.5-2 m tall), which give fynbos a heath-like appear-
ance. The proteoids are the tallest fynbos shrubs (2-
4 m), and comprise showy members of the Proteaceae,
another Gondwanan family, among them the king pro-
tea (Protea cynaroides), South Africa’s national flower.
Finally, fynbos includes more than 1 500 species of
bulbs or geophytes, many of which have been devel-
oped worldwide as valuable horticultural plants, e.g.,
freesias, agapanthus, gladioli, and ixias.
The Cape also includes several non-fynbos vegeta-
tion types. Of these, Renosterveld (Afrikaans for “rhi-
noceros veld,” referring to the presence of the black
rhinoceros [Diceros bicornis], that used to browse there
but is now extinct in this region) is the most extensive,
covering some 20 000 km2. This community comprises
a low shrub layer (1-2 m tall) of mainly ericoids, usual-
ly dominated by the renosterbos (Elytropappus rhinoce-
rotis, Asteraceae), with a ground layer of grasses and
seasonally active bulbs.
Trees are very rare in pristine Cape landscapes and
true forests occupy a mere 3 850 km2, mostly in moist,
fire-protected sites on the southern coastal forelands and
lower mountain slopes. The Cape forests, 10-30 m tall,
are essentially outliers of the Afromontane forests of the
high mountains of tropical Africa, but also include relics
from the mesic Tertiary times. A form of dwarf forest,
3-5 m tall and locally known as thicket, occupies about
4 500 km2 of fire-protected habitat (mainly coastal dunes
and river valleys) in drier areas; thicket is composition-
ally related to the subtropical forests of the Indian Ocean
coastline (Vlok et al. 2003). In contrast to the plants of
the fynbos and renosterveld, forest and thicket plants
are not chemically and structurally adapted to burn, and
new individuals that germinate from bird-dispersed
fruits establish only over extended, fire-free periods.
The Cape owes its status as a distinct floral kingdom
to the presence of five endemic plant families (of a to-
tal of 164), and 160 endemic genera (17% of 942 gen-
era). Certain genera have undergone massive diversifi-
cation —the 10 largest genera account for 21.5% of the
flora—, with the two most speciose being Erica (Eri-
caceae: 658 species) and Aspalathus (Fabaceae: 257
species). Species richness and local endemism is great-
est in the southwest; the Cape Peninsula (471 km2)
alone supports 2 256 species (including 90 endemics).
Diversity and endemism among the fauna of the Cape
Floristic Region appears to be much lower than in
plants, although very little is known about some inver-
tebrate groups (Johnson 1992). There are 90 mammal
species recorded from the Cape Floristic Region, of
which four are endemic. The list of endemic species in-
cludes two species of golden moles: the Fynbos golden
mole (Amblysomus corriae) and Van Zyl’s golden mole
(Cryptochloris zyli, CR). The best mammal flagship in the
region is the bontebok (Damaliscus pygargus pygargus,
VU), a beautiful animal whose name in Afrikaans means
“brightly colored antelope” and which was saved from
extinction only by the foresight of conservationists.
Bird diversity, too, is not particularly high, owing to
the structural uniformity of the vegetation and the
shortage of food (McMahon and Fraser 1988). Only 324
regularly occurring species have been recorded from
the region, and just six of these are endemics. Nonethe-
less, the area is considered an Endemic Bird Area by
BirdLife International (Stattersfield et al. 1998), and is
home to a number of true fynbos species such as the
Cape sugarbird (Promerops cafer), one of only two
species in the family Promeropidae; the orange-breast-
ed sunbird (Nectarinia violacea); and the Protea canary
(Serinus leucopterus).
Reptile diversity, on the other hand, is moderately
high at 100 species, of which 22 are endemic. Among
reptiles, the tortoises are the best flagship species, with
five species occurring almost entirely within the Cape
Floristic Region, including the South African bowsprit
or angulate tortoise (Chersina angulata), the leopard tor-
toise (Geochelone pardalis), and the geometric tortoise
(Psammobates geometricus, EN). The latter is among the
most threatened tortoises in the world, having already
lost some 97% of its original habitat to extensive agri-
cultural development; its remaining habitat amounts to
no more than 5 000 ha, and its population is estimated
at only 2 000-3 000 individuals in 31 different localities
(Baard 1997).
Amphibians and freshwater fish, though low in over-
all diversity, exhibit high endemism. In all, there are 51
species of amphibians (in 16 genera), 16 of them en-
demic (although no genera or families are endemic), and
they include species like the enchanting arum lily frog
(Hyperolius horstocki, VU) which, as its name suggests,
lives within the flower of the arum lily (Zantedeschia
aethiopica). In addition, two amphibian genera are en-
demic, both represented by single species: the micro
frog (Microbatrachella capensis, CR), which is found in
sandy, coastal fynbos heathland, and the montane
marsh frog (Poyntonia paludicola), a species of moun-
tain fynbos heathland. Of the 34 native fish species, 14
are endemic. Some distinctive fishes in the clear moun-
tain streams characteristic of this region are the Cape
galaxias (Galaxias zebratus), a peculiar, elongated and On the opposite page, George or
scaleless fish, as well as several endemic species of Scarborough lily (Cyrtanthus
redfin minnows (Pseudobarbus spp.) (Skelton 1993). elatus), Outeniqua Mountains,
South Africa.
Much less is known about the invertebrate fauna of
© Haroldo Castro
the Cape. However, the few groups that have been stud-
ied suggest very high levels of endemism. For example, Above, a succulent, Crassula
of the 234 species of butterflies in the region, 72 are en- columnaris, from the Western
demic (Rebelo 1992). One regional study, carried out Cape.
on the Cape Peninsula, recorded 111 invertebrate en- © Gerald Cubitt
demics in 471 km2, a higher number than for plant
endemics (Picker and Samways 1996).
The Cape Floristic Region is seriously threatened by
a battery of human activities that have seen the Coast
Renosterveld and Sandplain Fynbos, both lowland habi-
tats, reduced by 48% and 83% of their original extent,
respectively; much of what remains exists in small and
133

Succulents reach amazing levels of
diversity in the Succulent Karoo,
and include a wide range of tiny
species such as this Crassula
congesta, photographed near
Oudtshoorn in the Little
Karoo region.
© Olivier Langrand
On the opposite page, aloes are very
diverse in the Succulent Karoo,
and are often very restricted in
range by small differences in
altitude and soil condition.
© Haroldo Castro
134
isolated fragments in a matrix of chemically treated
agriculture (Rouget et al. 2003a). Even in the moun-
tains, where impoverished soils previously limited agri-
culture, farming based on indigenous crops such as
rooibos tea (Aspalathus linearis), honeybush tea (Cyclo-
pia spp.), and cut flowers (mainly Proteaceae) is rapid-
ly encroaching on natural habitat. Nonetheless, the
greatest threat to the Cape Flora overall is undoubtedly
the invasion of alien plants (Richardson et al. 1992),
mainly originating from other Mediterranean-type
climate regions such as southern Australia (wattles
[Acacia spp.], myrtle [Leptospermum laevigatum]), the
Mediterranean Basin (Pinus halepensis, P. pinaster), and
California (P. radiata). Originally introduced to supple-
ment the meager Cape tree flora, these fire-adapted
species have invaded about 70% of both mountain and
lowland fynbos. Overall, some 33% or 23 600 km2 of the
Cape Floristic Region has been transformed by agricul-
ture, urbanization, and dense stands of alien plants
(Rouget et al. 2003a). However, of the remaining area,
only about 20% can be regarded as pristine in the sense
that it is entirely free of alien plants and subjected to
appropriate fire and grazing regimes.
An analysis of the World Database on Protected Areas
reveals that approximately 14% of the hotspot has some
level of protection, of which at least 13% of the hotspot
area is classified in IUCN categories I to IV. However, ac-
cording to Rouget et al. (2003b), approximately 25%, or
19 350 km2, of the Cape Floristic Region (including the
Little Karoo enclave) is conserved in 189 protected ar-
eas. Statutory conservation areas (protected areas with
secure organizational and institutional support) com-
prise 49% of this area, whereas non-statutory reserves
(mainly private conservation initiatives) comprise the
remainder. Despite a relatively healthy proportion un-
der some form of conservation management, these re-
serves were established in an ad hoc manner and are
not entirely representative of the Cape’s biota. Thus,
some 50% of mountain landscapes are conserved, ex-
ceeding targets for many individual habitat types,
whereas only 9% of the lowlands are conserved. Fur-
thermore, the conservation area does a relatively poor
job in protecting the ecological and evolutionary pro-
cesses that maintain the Cape’s unique biodiversity. Ob-
vious priorities include remnant habitat in the highly
vulnerable and poorly conserved lowlands, as well as
the expansion of upland reserves to accommodate the
full spectrum of species and processes.
However, there have been some positive develop-
ments on the conservation front, including: the success-
ful initiation of the “Working for Water Programme,”
which aims to deliver a major fynbos ecosystem service
—clean and plentiful supplies of water— cost-effectively
by removing all alien plants from mountain catchments;
the expansion of protected areas through proclamation
of the Cape Peninsula National Park and the Cape Agul-
has National Park; and investment by the Critical
Ecosystem Partnership Fund (CEPF) and the Global En-
vironmental Facility (GEF) to develop and implement a
strategic plan for the sustainable conservation of the
Cape Flora and associated biota (Cowling and Pressey
2003). Of particular interest within CEPF’s $6-million in-
vestment portfolio is support for the region’s three
mega-reserves: Baviaanskloof, Cederberg, and Gouritz,
where funds have been used to prepare strategic man-
agement and business plans to ensure that these areas
will be able to meet future conservation challenges. Em-
phasis has also been placed on building the capacity of
previously disadvantaged local conservationists through
a $1-million CEPF small-grants fund administered by the
Table Mountain Fund. If the programs mentioned above
continue to be successful, then there is every hope that
future generations will be able to benefit from and mar-
vel at this most extraordinary biological phenomenon.
SUCCULENT KAROO
PHILIP DESMET 37 • RICHARD M. COWLING 133
NAMIBIA
SOUTH
AFRICA
ATLANTIC
OCEAN
0 200 km
Arid lands tend to have highly endemic floras and fau-
nas, but are usually relatively species poor. The Succu-
lent Karoo is an extraordinary exception, and is one of
only two arid regions in the world that can be considered
a true biodiversity hotspot (the other being the Horn of
Africa, which is recognized here as a new hotspot). It
harbors a staggering 6 356 plant species (in 168 families
and 1 002 genera) (Driver et al. 2003) and the richest suc-
culent flora in the world (Van Jaarsveld 1987).
This predominantly winter-rainfall desert, which oc-
cupies 102 691 km2 on the arid fringes of the Cape
Floristic Region, has been divided by biogeographers
into two zones, the strongly winter-rainfall Namaqua-
land-Namib Domain of the Cape west coast and south-
ern Namibia (generally known as Namaqualand) and
the Southern Karoo Domain, where rainfall often peaks
in the spring and autumn months (Jürgens 1991). In
broad phytogeographical terms, the Succulent Karoo
forms part of the Karoo-Namib Regional Center of En-
demism (White 1983). However, the area is largely tran-
sitional between the Cape Floristic Region and the
Nama-Karoo, and shows a particularly strong floristic
affinity with the Cape Floristic Region, to the point that
some have argued convincingly for the region’s inclu-
sion as part of a greater Cape Flora (Jürgens 1991; J.
Born, P. Linder, and P. Desmet, unpubl.)
The typical vegetation of the Succulent Karoo is a
dwarf shrubland dominated almost entirely by leaf suc-
culents in the Mesembryanthemaceae, Crassulaceae,
Asteraceae and Liliaceae families (Milton et al. 1997).
Some 1 700 out of a total of 1 843 estimated succulents
are leaf succulents in the Succulent Karoo, and this dom-
inance is unique among the deserts of the world (Jür-
gens 1986). The recent and explosive diversification of
the Mesembryanthemaceae, the largest group, has been
described as an event unrivaled among flowering plants
(Ihlenfeldt 1994; Desmet et al. 1998; Klak et al. 2004). In-
deed, the exceptionally high vascular plant diversity and
endemism (2 539 species, or 40%, and 80, or 8% of gen-
era) of the Succulent Karoo is associated with massive
speciation within the predominantly succulent and bul-
bous genera. Stem succulents, comprising some 130
species, include species of Euphorbia, Tylecodon, Othon-
na, Pelargonium, and numerous stapeliads. Seasonal
bulbs (1 143 species) and annuals (390 species), which
appear in the open spaces between the shrubs, provide
magnificent spring displays, especially in Namaqualand.
The quartz gravels of the Knersvlakte, as well as simi-
lar habitats elsewhere in Namaqualand and in the Little
Karoo, support a unique vegetation type characterized by
an extremely sparse cover of locally endemic Mesem-
bryanthemaceae and other succulents. Nearly all of
these are stemless, embedded forms (stone plants). The
Succulent Karoo is home to about 700 such species be-
longing to the genera Argyroderma, Bulbine, Conophytum,
Crassula, Haworthia, and Lithops. Another unusual fea-
ture of much of the Succulent Karoo is the presence of
heuweltjies or mima-like mounds (Dean and Milton
1999). The mounds are of zoogenic origin, originally cre-
ated by termites (Microhodotermes viator) and often colo-
nized by a wide variety of burrowing animals. As a result,
nutrient-enriched subsoil is brought to the surface, cre-
ating a substratum physically and chemically very dif-
ferent from the intervening matrix. Thus, heuweltjies
support a flora that is markedly distinct in appearance
and composition from that of the surrounding vegetation.
Among the most famous of all the attractive succulent
species found in the Succulent Karoo is the halfmens
(“half human”) (Pachypodium namaquanum), a stem suc-
culent endemic to the Richtersveld and that grows to a
height of up to 4 m. The stems of these plants are in-
clined to the north, thus creating an uncanny resem-
blance to humans peering northwards to the distant
plains of southern Namibia. Khoi-khoi folklore has it that
the plants are the remnants of an ancient tribe, banished
by warfare from their original home in Namibia, which
gaze nostalgically across the Orange River to their home-
land. The scientific explanation for this unusual orienta-
tion is equally engaging (Rundel et al. 1995). Growing on
shaded slopes, the plants lean northwards in order to en-
sure that their leaves and developing flowerheads, pro-
duced during the cool, foggy winter months, are maxi-
mally exposed to the sun’s warming rays.
The fauna of the Succulent Karoo is rich in endemic
species, especially among arachnids, hopliniid beetles,
aculeate Hymenoptera, and reptiles (Vernon 1999). Esti-
mates from available data suggest that endemism in in-
sects is especially high (>50%) for groups examined
(Hymenoptera [Apoidea] and Isoptera) (Driver et al.
2003). The predictable rainfall appears to have resulted
in many resident forms of invertebrates and small
vertebrates, which has led to isolation and speciation.
Several invertebrate groups are particularly well repre-
sented, especially those that reach high levels of diversi-
ty in arid regions, such as scorpions (with 18 of the 70
scorpions in the Succulent Karoo endemic; Driver et al.
2003). Monkey beetles (Rutelinae: Hoplini), a group
largely endemic to southern Africa, are concentrated in
the Succulent Karoo, where some genera are important
pollinators of daisies and mesembs.
Among the vertebrates, there are, at present, 74 mam-
mals recorded from the Succulent Karoo, of which two
are endemic: De Winton’s golden mole (Cryptochloris
wintoni, VU) and the Namaqua dune mole-rat (Bathyer-
gus janetta). Birds are represented by 227 regularly oc-
curring species, including one endemic, the recently de-
scribed Barlow’s lark (Certhilauda barlowi). In addition,
the black harrier (Circus maurus, VU), with its distinctive
black-and-white plumage, has the most restricted range
of the world’s 13 harrier species (Harrison et al. 1997)
and is frequently observed hunting over the Namaqua-
land plains. Amphibians are poorly represented, with
only 29 species of frogs, one of which, the desert rain frog
(Breviceps macrops, VU) is endemic, occurring on the
Namaqualand coast of South Africa north to Luderitz in
coastal southwestern Namibia. Reptiles are very diverse,
with 94 species, of which 15 are endemic. These include
58 lizard and gecko species with 15 endemics (the genera
Cordylus and Pachydactylus are especially rich, with eight
and 12 species and six and two endemics, respectively)
and 29 snakes, of which none is endemic. The freshwater
fish fauna includes 26 indigenous species, though none
are endemic (modified from D. Impson et al., unpubl.).
Considering this wealth of unique plant species, and On the opposite page, another aloe
high endemism, it is encouraging that the Succulent Ka- plant from Little Karoo, Oudtshoorn
roo is one of the more intact hotspot systems. Over 90% of region, South Africa.
© Gerald Cubitt
the region is used for natural grazing, a form of land use
that is, theoretically, compatible with maintenance of bio- Above, a Cape sugarbird
diversity and ecosystem processes. It is difficult to make (Promerops cafer) perched on a
accurate estimates of the amount of intact vegetation. It is Protea flower. One of only two
known that at least 5% of the biome has been irreversibly members of the subfamily
lost to mining (3%) and cropping agriculture (2%), and Promeropinae, this species is
endemic to the Cape Region of
about 0.1% to urbanization (Driver et al. 2003). The
South Africa.
amount transformed or degraded by overgrazing is un-
© M. Watson/Ardea
known, although one estimate states that approximately
70 000 km2, especially the communally owned parts of
Namaqualand, have been seriously overgrazed (Davis et
al. 1994). Thus, optimistically, only some 30 000 km2 of
the Succulent Karoo, or about 29%, exist in a relatively
pristine state, although this figure is probably much lower.
137

The striking angonoka
(Geochelone yniphora) is
Madagascar’s largest tortoise and is
endemic to the Bay of Baly region in
a remote northwestern corner of the
country. Considered one of the most
endangered tortoises in the world,
it has been the subject of a
successful captive-breeding and
reintroduction project by the Durrell
Wildlife Conservation Trust.
© Pete Oxford/naturepl.com
On the opposite page, the
Endangered black-and-white ruffed
lemur (Varecia variegata) is found
in the rainforests of eastern
Madagascar, and is threatened by
habitat destruction and
hunting. Recent information
indicates that it may be divided into
three distinct subspecies.
© Russell A. Mittermeier
138
A number of threats are likely to take on greater im-
portance in this region in the future. The rise of the os-
trich farming industry has resulted in the degradation
of thousands of hectares of veld in the Little Karoo (Dri-
ver et al. 2003). Also, the expected increase in the ex-
tent of communal land in Namaqualand and elsewhere
in the region will undoubtedly accelerate desertifica-
tion, too (Hoffman et al. 1999), while dryland farming
in Namaqualand is practiced where the annual rainfall
is as low as 150 mm per year. And, of course, diamond
mining has had a major impact on the Namaqualand
coastline and alluvial terraces of the lower Orange Riv-
er Valley (Davis et al. 1994). Approximately 65% of the
South African, and almost all Namibian Atlantic coast-
line of the Succulent Karoo, has been mined for dia-
monds. The extraction of heavy minerals, an equally
destructive process, continues to develop as a major in-
dustry on the coast of southern Namaqualand.
Remarkably, for a country like South Africa renowned
for its well-established game parks and reserves, pro-
tected area coverage in this hotspot is poor. According to
the World Database on Protected Areas, only 2.5% of the
area has some form of protection (with only 1.8% in
IUCN categories I to IV). Regional information proves
the situation is not much better, with only 3.5% of the
Succulent Karoo formally protected in category I re-
serves; these are statutory reserves managed primarily
for biodiversity conservation, and include National
Parks and Provincial Nature Reserves (Driver et al.
2003). An additional 2.3% of the biome is protected in
category II reserves. These are statutory and non-statu-
tory reserves managed for biodiversity conservation
and/or other land uses. A statutory reserve managed for
biodiversity conservation and other land uses would be
classified as a category II reserve rather than a catego-
ry I reserve. This is the case with the Richtersveld Na-
tional Park, a contractual national park in which both
conservation and other land uses occur. Other category
II reserves include municipal reserves and conservan-
cies (voluntary agreements among private landowners).
Considering that the region has more than 936 Red Data
Listed plant species (Driver et al. 2003), the reserve sys-
tem for the Succulent Karoo is grossly inadequate for
conserving the region’s biodiversity. It is estimated that
59% of the region’s quarter-degree grid cells would be re-
quired in a reserve system to represent each plant
species at least once (Lombard et al. 1999).
Recently, there have been two positive developments
regarding the conservation status of the Succulent Karoo.
First, the reserve system is being expanded, although at
the current rate of expansion it will take 130 years to
achieve the reservation targets set by the Succulent Ka-
roo Ecosystem Program (SKEP) project (P. Desmet, un-
publ.). The creation of the Namaqua National Park
(ca. 600 km2) in the central uplands of Namaqualand is
a positive development. This park is set to expand west-
wards to encompass Sandveld habitats on the coastal
plain as well as the marine zone. Secondly, the SKEP
project has created much awareness in the region, as
well as interest in conserving biodiversity, and fortu-
nately funds have become available recently to fuel this
awareness and interest. The Critical Ecosystem Part-
nership Fund (CEPF) has allocated $8 million for this
hotspot over five years. Already action is taking place in
numerous geographic priority areas, with technical assis-
tance in project design and proposal preparation provid-
ed to numerous civil society groups, and grant funds
being disbursed in support of conservation initiatives
throughout the region. Fortunately, the situation in the
Succulent Karoo is such that timely action can still en-
sure that a strong, representative cross section of its rich
and diverse desert landscapes will survive into the future.
MADAGASCAR AND THE
INDIAN OCEAN ISLANDS
RUSSELL A. MITTERMEIER 1 • OLIVIER LANGRAND 1
PORTER P. LOWRY II 26 • GEORGE SCHATZ 27 • JUSTIN GERLACH 47, 48
STEVEN GOODMAN 49, 50 • MARC STEININGER 2 • FRANK HAWKINS 52
NORO RAMINOSOA 51 • OLGA RAMILIJAONA 51
LUCIANO ANDRIAMARO 52 • HARISON RANDRIANASOLO 52
HARISON RABARISON 52 • ZO LALAINA RAKOTOBE 52
SEYCHELLES
COMOROS
MADAGASCAR INDIAN
OCEAN
MAURITIUS
Rodrigues
Réunion
0 800 km
The island of Madagascar, at approximately 590 000 km2
the fourth largest on Earth, has long been recognized as
one of the world’s highest priority hotspots and one of
the top megadiversity countries. Although located only
about 400 km from the east coast of Africa, the island,
another chip off the supercontinent Gondwana, has
been isolated from other landmasses for more than 160
million years. Consequently, most of the plant and ani-
mal species occurring there have evolved in long isola-
tion, and are unique and found nowhere else. Levels of
endemism in most groups of organisms are exception-
ally high, not just at the species level, but often at the
generic and even the family levels as well.
In addition to Madagascar itself, we include in this
hotspot the neighboring western Indian Ocean Islands
which, although much smaller, are also very important
biologically. These include the independent nations of
Seychelles (454 km2; including Aldabra), the Comoros
(1 862 km2), Mauritius (2 040 km2; including Rodri-
gues), and the French overseas departments of Réunion
(2 535 km2) and Mayotte (371 km2; one of the Com-
oros), which is a departmental collectivity of France.
Also included are the Iles Esparses, all of which belong
to France: Europa (30 km2), Les Glorieuses (7 km2), Juan
de Nova (5 km2), Bassas da India, and Tromelin (1 km2).
These Indian Ocean islands add considerably to the over-
all importance of the hotspot, and bring its total land
area to 600 461 km2.
The natural vegetation of this hotspot is quite diverse.
Madagascar is characterized by tropical rainforest on the
eastern side, dry deciduous forests on the western side,
and spiny desert in the far south of the country. In the far
north, there is a mosaic of dry and moist forest forma-
tions, and a series of mountains are found in the north
and east (Lowry et al. 1997). The Indian Ocean islands
are composed of a range of relatively recent volcanic is-
lands (the Mascarenes and the Comoros), fragments of
continental material (the main group of the Seychelles),
and the coral cays of the Amirantes and the atolls of the
Farquhar, Cosmoledo, and Aldabra groups, as well as
the five Iles Eparses. The continental and volcanic islands
have high peaks that in the recent past were covered by
dense forest; indeed, the Comoros (up to 5 600 mm per
year on Grande Comore) and the Mascarenes (up to
6 000 mm per year on Réunion) are sometimes subjected
to very high levels of rainfall. The highest peak in the In-
dian Ocean is the Piton des Neiges on Réunion (3 069 m),
which received the heaviest downpour on record (4.9 m
of rain in one week in 1980). By contrast, the Seychelles
are relatively dry (up to 2 400 mm per year on Mahé).
In terms of its biodiversity, Madagascar’s most strik-
ing feature is its high levels of endemism, particularly at
the generic and family levels. Madagascar also has very
high species diversity in certain groups of organisms, es-
pecially given its relatively small size. Both of these
characteristics are best represented in Madagascar’s flo-
ra: current plant diversity is estimated to be at least
12 000 species, and possibly as many as 14 000, of which
around 90% are endemic (G. Schatz and P. Lowry, un-
publ.). In addition, there are seven endemic plant fami-
lies, which is unmatched by any other country; indeed,
only Australia, New Caledonia, and South Africa are
comparable in terms of plant endemism at the family
level. Madagascar recently made headlines in the botan-
ical world with the rediscovery of Takhtajania perrieri,
the only Afro-Malagasy member of the primitive family
Winteraceae, in the northeast of the country. It is fitting
that Madagascar’s signature endemic plant, the travel-
er’s tree (Ravenala madagascariensis), is pollinated by
the island’s flagship vertebrate species, the lemurs.
Among vertebrates, some 283 avian species have
been recorded, of which 209 breed on the island, and
109 of these are endemic (Hawkins and Goodman
2003); and, not surprisingly, no fewer than five En-
demic Bird Areas have been recognized in the country
(Stattersfield et al. 1998). At the higher level, 34 out of
148 resident genera are endemic, and there is also high
family-level endemism, with five bird families restrict-
ed to the island. There are some extraordinary ancient
relict bird species on Madagascar, such as the ground-
rollers, cuckoo-rollers, and mesites.
Mammals are represented by 131 species, of which 30
are bats; all but 12 species are endemic (Goodman et al.
2003; Eger and Mitchell 2003). No other country or hot-
spot comes close in terms of primate family-level en-
demism, with some five families and 15 genera endemic,
and a total of 72 taxa in all (with more still to be de-
scribed). This incredible primate radiation constitutes the
best known of the region’s charismatic species, including
the aye-aye (Daubentonia madagascariensis, EN), the indri
(Indri indri, EN), and Madame Berthe’s mouse lemur (Mi-
crocebus berthae), at 30 g the world’s smallest primate.
Although there is only one endemic amphibian family
(Mantellidae) and one endemic reptile family (Opluri-
dae), the reptiles and amphibians exhibit very high levels
of endemism. Indeed, of the 340 reptile species, 314 are
endemic (Raxworthy 2003), as are an amazing 215 of the
217 described amphibian species. Apart from being ex-
tremely rich in these groups, Madagascar may even be
the place of origin for some; for example, it has been pro-
posed recently that all the world’s chameleons originated
here. Among the flagship amphibians and reptiles are the
beautiful frogs of the genera Mantella and Scaphiophryne,
the tomato frog (Dyscophus antongilii), a bright red, bull-
frog-sized animal found only in a tiny area in northeast-
ern Madagascar and, of course, the chameleons.
It is noteworthy that many new species have been dis-
covered on Madagascar in the last decade or so, and, in-
deed, since the publication of the last Hotspots book in
1999. For example, no fewer than 22 new mammal species
and subspecies have been described from Madagascar in
the last 15 years (including seven full lemur species be-
tween 1997 and 2003). Furthermore, many new species re-
main to be discovered and described; for example, C. Rax- On the opposite page, the leaf-tailed
worthy (pers. comm.) notes that at least 100 reptile species geckos of the genus Uroplatus are
and about 100 amphibian species await formal description. endemic to Madagascar and rank
among the most cryptically shaped
Most of the other fauna on Madagascar is poorly
and colored lizards in the world.
known. However, some of the non-marine invertebrate This species, the aptly-named
groups that are reasonably well known are: terrestrial U. phantasticus, is from the region
snails (651 species, all endemic); scorpions (40 species, of Ranomafana National Park.
all endemic); spiders (459 species, 390 endemics); drag- © Piotr Naskrecki
onflies and damselflies (181 species, 132 endemics);
Above, the Vulnerable helmet vanga
lacewings (163 species, 119 endemics); tiger beetles (211
(Euryceros prevostii) is endemic
species, 209 endemics [D. Pearson, pers. comm.]); scarab to the northern portion of the
beetles (148 species, all endemic); true butterflies (300 eastern rainforests of Madagascar.
species, 211 endemics); freshwater crayfish (six species, It is a member of the endemic
all endemic); and freshwater shrimp of the family Atyi- monophyletic family Vangidae
dae (26 species, 20 endemics). Overall, total species rich- —famous for its amazing variation
in bill shape and diversity of
ness for macroinvertebrate groups covered in a recent re-
ecological niches.
view of the natural history of Madagascar is slightly more
© Pete Oxford/naturepl.com
than 5 800 species, of which 86% are endemic to the is-
land, although several speciose groups of invertebrates
are not covered in the volume (e.g., the vast majority of
beetle families) (Goodman and Benstead, pers. comm.).
The smaller neighboring Indian Ocean islands are bi-
ologically closely linked to Madagascar, are under heavy
141

Aerial view of deforested landscape
in the central plateau region of
Madagascar.
© Haroldo Castro
On the opposite page, forest
destruction for slash-and-burn
agriculture, or tavy as it is known
in Madagascar, in spiny desert
near Ifaty.
© Nick Garbutt/naturepl.com
142
pressure, and add important endemic biodiversity with-
out adding significantly to the land area covered by this
hotspot. For example, in terms of plant diversity and en-
demism, the flowering plants are represented by about
2 200 to 2 400 species (ca. 1 300 in the Mascarenes, 1 000
in the Comoros, and 310 in the Seychelles, several hun-
dred of which are shared among two or more of the is-
land groups); around 810 (34%-37%) of these are en-
demic (about 585 in the Mascarenes, 150 in the
Comoros and 75 in the Seychelles), along with one fam-
ily endemic to the Seychelles (Medusagynaceae).
Space does not permit detailed discussion of the verte-
brate diversity of all the islands, although much research
has been conducted on Seychelles recently, and these is-
lands are mentioned briefly here. The Seychelles add a
further 104 native breeding vertebrates, with endemics
comprising one fish, 11 amphibian, 27 reptile, 14 bird, and
four mammal species. In addition, the Seychelles have
one endemic amphibian family —the Sooglossidae, an
ancient group endemic to this area with its closest relative
in the Western Ghats of India— and six endemic amphib-
ian genera. The presence of seven species of caecilian
amphibians is bizarre, as these are entirely absent from
Madagascar and the other western Indian Ocean islands.
The invertebrate fauna of the Seychelles comprises
3 555 recorded species, with an estimated total of per-
haps 5 100 species; of these, approximately 80% are en-
demic. One truly unique and amazing invertebrate flag-
ship is the endemic giant tenebrionid beetle (Polposipus
herculeanus, CR), restricted to one small island in the
Seychelles, and one of the largest terrestrial inverte-
brates in the world. The region also supports the largest
millipede (Sechelleptus seychellarum) and the only se-
cure population of the world’s largest terrestrial inver-
tebrate, the coconut or robber crab (Birgus latro).
The hotspot includes two distinctive components of
freshwater fishes. The smaller islands have depauperate
faunas dominated by species that have wide marine dis-
tributions and that enter brackish and freshwater habitats
to some degree. The few endemics on these islands in-
clude a cyprinodont, a chandid, and gobies. Madagascar
differs in having a freshwater fauna that includes fishes of
continental origin. These groups have undergone radia-
tion on the island, resulting in 93 endemic species (of the
hotspot’s 97 species), and these also account for all of
the hotspot’s 14 endemic genera and two endemic families.
Overall, the Madagascar and the Indian Ocean Islands
Hotspot (i.e., including not only Madagascar and the
Seychelles, but also the remaining Indian Ocean Islands)
has a total of at least 155 mammals, 144 of which are en-
demic; about 313 regularly occurring birds, 183 of them
endemic; 381 reptiles, 367 endemic; and at least 228 am-
phibians, of which 226 are endemic. Plants total at least
13 000 species, and probably as many as 15 000, with a
staggering 11 600 (and perhaps 13 500) endemics, and at
least 310 endemic genera. The number of endemic fam-
ilies in this hotspot is truly exceptional, totaling 24 in all
(16 for vertebrates, including fishes, and eight for
plants), and far surpasses that of any other hotspot. This
once again attests to the great global importance of
Madagascar and its neighboring islands, and highlights
its role as one of the highest priority hotspots on Earth.
The threats to Madagascar and the Indian Ocean is-
lands are well documented, with forest destruction
through slash-and-burn agriculture, mining, and logging
being among the main causes of habitat loss. In Mada-
gascar, it is estimated that around 90 000 km2 of closed-
canopy primary forest and woodland remained as of
2000, with an average annual rate of loss during the 1990s
of 0.9% per year (Steininger et al., unpubl.). Assuming
that 90% of Madagascar was once forested (Perrier de la
Bathie 1936), this equates to roughly 17% of original pri-
mary vegetation remaining. The most heavily impacted
habitats are lowland rainforest, dry deciduous forests,
and spiny forest. Wetlands, including lakes, rivers, and
marshes, are under threat from transformation to rice
fields, siltation resulting from soil erosion, and intro-
duced species. The latter have accounted for several ex-
tinctions recently. Lemurs, some birds, and smaller
mammals are very susceptible to hunting. The pet trade
has also had a serious impact on endemic plants and an-
imals of Madagascar, especially amphibians, reptiles, and
succulent plants. The proliferation of exotic plant species
is also recognized as a major threat affecting the biodi-
versity of Madagascar and the western Indian Ocean is-
lands, and freshwater ecosystems, in particular, have
been seriously impacted by alien plants such as Eichhor-
nia crassipes (Langrand and Goodman 1995).
In the Comoros, at least 80% of the native vegetation
has been destroyed since human colonization first be-
gan over 1 000 years ago. During the period 1990-1995,
deforestation reached 5.8% per annum, the fourth high-
est rate of any country (Jolly and Fukuda-Parr 2000).
Today, plantations dominate the land. Mayotte Island
also was once forested, with dry, humid, and transi-
tional types recognized. Most of this was destroyed be-
fore 1900, leaving fragments of native forest that cov-
ered only 3% of the island in 1999 (Pascal 2002), most
of which (6.7 km2) is humid forest. A further 14.8 km2 of
the island’s vegetation has been classified as secondary
forest up to 300 m. Hunting is a major problem for the
avifauna, and species such as the Comoro olive pigeon
(Columba polleni) are becoming rare as a result.
Réunion was permanently colonized in 1646. Since
then, humans have brought about the loss of 75% of the
native vegetation area (around 650 km2 remain) and 50%
of the native vertebrate fauna (including 55% of the birds).
As is often the case on recent volcanic islands, the intro-
duction of exotic species of fauna and flora is having a
huge impact on the survival of endemic species. On the
Seychelles, much of the original lowland vegetation was
cleared for timber production or agriculture, particularly
for coconut plantations and cinnamon exploitation in the
granitic islands, although significant reservoirs of biodiver-
sity have survived thanks to the steep terrain of the high is-
lands. The Amirantes, and Cosmoledo and Astove atolls,
have also been greatly impacted by coconut plantations.
Given the importance of the hotspot, it is not surpris-

The El Hierro giant lizard (Gallotia
simonyi machadoi) is considered
the most threatened reptile in
Europe, as well as one of the five
most threatened reptiles in the
world. It is currently the flagship
species for El Hierro Island (Canary
Islands, Spain), and an example of
the recovery and conservation of a
species that is unique. Today its
population is estimated at between
1 000 and 1 600 individuals, in the
Tibataje Special Natural Reserve.
© Francisco Márquez
On the opposite page,
the Cabrera Archipelago,
south of Mallorca Island, Spain,
is one of the few sites where
Mediterranean-type forests have
found refuge, thanks to which this
area was declared a national park
and, above all, because for many
years a military base operated
there, hindering the development of
this zone, which now has different
endemic flora species.
© Patricio Robles Gil/Sierra Madre
144
ing that there have been major efforts in biodiversity con-
servation, and the hotspot, in particular Madagascar, may
be entering a new era in terms of biodiversity protection.
In Madagascar, the government is about to embark on the
third phase of the National Environmental Action Plan,
with a five-year program of conservation and sustainable
management actions. There are many conservation
organizations involved in this effort, both international
and local, and the World Bank, Global Environmental Fa-
cility (GEF), United Nations Development Program, and
French, U.S., German, and Swiss bilateral aid is all being
deployed specifically for biodiversity conservation.
In the Comoros, political instability has meant that con-
servation action has been rather piecemeal until recently,
with only limited GEF and IUCN support for the estab-
lishment of a marine national park and some work on
Livingstone’s fruit bat (Pteropus livingstonii, CR) by local
and international NGOs. The climate for collaboration is
set to change, however, and further investment in estab-
lishing threatened species conservation programs,
through CI and other partners, is under way; and there is
a plan to establish a terrestrial national park in Mount
Karthala. The Seychelles, by contrast, have several very
active conservation NGOs that have had a dramatic
impact on the conservation status of many of their threat-
ened species recently. A notable example is the Sey-
chelles magpie robin (Copsychus sechallarum, CR), res-
cued from the edge of extinction over the last ten years by
Nature Seychelles and partners. A similar situation ap-
plies in the Mascarenes; while there are many species on
the verge of extinction, considerable effort has been de-
voted to captive-breeding and release programs. This is
particularly evident in Mauritius, where the Mauritian
Wildlife Foundation, Durrell Wildlife Conservation Trust,
and other partners have been prominent in rescuing spe-
cies such as the pink pigeon (Columba mayeri, EN) and
the echo parakeet (Psittacula eques, CR) from extinction.
As of mid-2002, the protected area network of Mada-
gascar included 46 legally protected areas (national
parks, strict nature reserves, and special reserves) cov-
ering 16 131 km2 or 2.7% of the country (Randrianandia-
nina et al. 2003). However, on September 17, 2003, at the
World Parks Congress in Durban, South Africa, Mada-
gascar’s President, Marc Ravalomanana, made history
when he declared his intention to triple protected area
coverage over the next five years and to seek assistance
from the international community for a $50-million
trust fund to make this a reality. Demonstrating once
again the interest of the international community, fully
$24 million in commitments were made to this trust
fund in the first six months following this announce-
ment. The Global Conservation Fund (GCF) at Conser-
vation International is assisting in the design and capi-
talization of the trust fund. The GCF and the Critical
Ecosystem Partnership Fund (CEPF) are also actively
supporting the efforts of local partners to identify prior-
ity areas for conservation, as well as to plan and create
new protected areas in irreplaceable forests in the
Menabe region of Western Madagascar, the extensive
rainforests of Makira in northeastern Madagascar, and
the Daraina region in the extreme northeast. CEPF’s
contribution of $4.25 million for conservation in Mada-
gascar has played a major role in helping local partners,
such as Association Fanamby, engage in biodiversity
conservation at multiple levels. The partnership’s sup-
port to local groups is set to further expand under a new,
three-year small grants program to help Malagasy orga-
nizations undertake conservation efforts at a local scale.
A number of the Indian Ocean Islands also have at
least some protected area coverage. For example, Ré-
union has 21 protected areas, all in IUCN category IV
and totaling 231 km2, while in the Seychelles, 208 km2
(46% of the land area) is designated as national parks,
with a further 228 km2 of marine national parks. These
include two World Heritage Sites: the Vallée de Mai and
Aldabra. Overall, an analysis of the World Database on
Protected Areas shows that there are around 100 pro-
tected areas in the hotspot, the majority in IUCN cate-
gories I to IV, covering 3% of the surface area.
Although this extremely important hotspot still faces
many threats and challenges, recent developments,
particularly in Madagascar, give cause for more opti-
mism than ever before. Indeed, the major new com-
mitments being made for Madagascar could mean that
this wonderful island, once considered almost a lost
cause for conservation, could quickly be transformed
into a global model. This renewed interest in the region
as a whole will hopefully have a significant impact in
the neighboring islands as well.
MEDITERRANEAN BASIN
FRÉDÉRIC MÉDAIL 61 • NORMAN MYERS 8
0 1500 km
FRANCE
ITALY BLACK SEA
Azores
Islands
SPAIN TURKEY
PORTUGAL
SYRIA
ATLANTIC
Madeira Sicily
CYPRUS
OCEAN MEDITERRANEAN
Canaries MOROCCO SEA
ALGERIA EGYPT
CAPE LIBYA
VERDE
The Mediterranean Basin Hotspot covers some
2 085 292 km2 and stretches from Portugal to Jordan
and from Morocco to northern Italy. It encompasses
over 90% of Greece, Lebanon, and Portugal, but less
than 10% of France, Algeria, and Libya. In Spain, 6 000
of the country’s 7 500 plant species occur within the
Mediterranean climate zone; in Israel, 1 500 out of
2 200; and in Morocco, 3 800 out of 4 200 (Quézel 1985;
Greuter 1991). The hotspot also includes the Canary Is-
lands, Madeira, and the Selvages (Selvagens) Islands
and, in contrast to the former definition of this hotspot

The silhouette of a large conifer
on the Sierra de Cazorla
in southern Spain.
© Patricio Robles Gil/Sierra Madre
146
(Myers and Cowling 1999), the region is here consid-
ered to also include the Azores and Cape Verde Islands,
even though the floristic affinities of these two Mac-
aronesian island groups lie more closely with Europe
and Africa, respectively.
Of overwhelming importance in understanding the
origins and diversity of the Mediterranean Basin biota
is the region’s location at the intersection of two major
landmasses, Eurasia and Africa. Indeed, the collision
between these two continental plates during the mid-
Tertiary is responsible for the Basin’s spectacular
scenery. The Basin’s violent geological history has pro-
duced an unusual geographical and topographical di-
versity, with high mountain ranges (more than 4 500 m
in elevation), peninsulas, and one of the largest archi-
pelagos in the world (the Mediterranean Sea includes
several hundred islands and islets). The physiographic
diversity of the region has resulted in a wide range of
local climates, with mean annual rainfall ranging from
100 mm up to 3 000 mm (Blondel and Aronson 1999).
The typical and most widespread vegetation type is
maquis or matorral, a hard-leaved shrubland dominat-
ed by Cistus, Erica, Genista, Juniperus, Myrtus, Phillyrea,
Pistacia, and other evergreens, and similar in appear-
ance to the chaparral of California and the matorral of
Chile (Di Castri and Mooney 1973). Although maquis
now covers more than half of the region, much of it has
been derived from forest formations created by human-
induced disturbances. Frequent burning of maquis re-
sults in depauperate vegetation dominated by Kermes
oak (Quercus coccifera), Cistus spp. or Sarcopoterium spi-
nosum, all of which regenerate rapidly after fire by
sprouting or mass germination.
Shrublands, including maquis and the aromatic, soft-
leaved and drought phrygana of Rosmarinus, Salvia, and
Thymus, persist in the semiarid, lowland, and coastal re-
gions of the Basin. However, prior to the onset of signifi-
cant human impact, which started some 8 000 years ago,
most of the Mediterranean Basin was covered by some
form of forests (Quézel and Médail 2003), including: ever-
green oak forests (Quercus ilex, Q. suber, and Q. coccifera
ssp. calliprinos); deciduous forests (Quercus canariensis, Q.
faginea, Q. frainetto, Q. ithaburensis, Q. petraea, Q. pubes-
cens, Q. pyrenaica, and Fagus sylvatica); and conifer forests
(Abies spp., Cedrus spp., Juniperus spp., and Pinus spp.).
The flora of the Mediterranean Basin is comprised of
around 25 000 species of vascular plants, 13 000 of which
are endemic (Quézel 1985; Greuter 1991). These figures
include taxonomically doubtful taxa (6% to 9%) and nat-
uralized exotics (less than 3%). As we are considering
here only confirmed native species, we subtract 10% for
a figure of 22 500 (and 11 700 endemics). The plant
species endemic to the Mediterranean Basin are not a
random assemblage in terms of their taxonomic affini-
ties, biology, habitat requirements, and geographical
distribution. Rather, the flora comprises a complex ad-
mixture of Mediterranean woody plants belonging to
pre-Mediterranean lineages (start of the Tertiary) (Verdú
et al. 2003) and localized neoendemics composed pre-
dominantly of herbs and subshrubs in the families
Asteraceae, Brassicaceae, Caryophyllaceae, Cistaceae,
Fabaceae, Lamiaceae, Poaceae, Ranunculaceae, and so
on. Endemics are mainly concentrated on some Tertiary
and Pleistocene refugia on islands, peninsulas, rocky
cliffs, and mountain peaks (Médail and Verlaque 1997).
Nevertheless, endemism at the higher level is very re-
duced, with only two endemic families (Aphyllan-
thaceae and Drosophyllaceae), both represented by sin-
gle species, Aphyllanthes monspeliensis and Drosophyllum
lusitanicum. The Mediterranean Region also harbors a
high degree of tree richness and endemism (290 indige-
nous tree species with 201 endemics) (Quézel and Mé-
dail 2003). A number of trees are important flagships, in-
cluding the cedars (such as the famous cedar of Lebanon,
Cedrus libani); the argan tree (Argania spinosa), a species
in the Souss region of southwest Morocco; oriental sweet
gum (Liquidambar orientalis); and Cretan date palm
(Phoenix theophrasti) in Greece and western Turkey.
The principal foci in the Mediterranean are 10 region-
al mini-hotspots within the larger hotspot, characterized
by areas of high plant richness and narrow endemism of
more than 10% (Médail and Quézel 1997, 1999): the At-
las Mountains in North Africa; the Rif-Betique range in
southern Spain and two coastal strips of Morocco and Al-
geria; Maritime and Ligurian Alps of the French-Italian
border; Tyrrhenian Islands; southern and central Greece;
Crete; southern Turkey/Cyprus; Israel and Lebanon;
Cyrenaica in Libya; and the Canary/Madeira Islands.
These 10 areas cover about 22% of the Basin’s total area,
yet account for almost 5 500 endemic plants, i.e., about
47% of total Mediterranean endemics (Médail and
Quézel 1999). Considering the redefinition of the hotspot
boundary, the inclusion of the Cape Verde Islands
(4 071 km2) and the Azores (2 407 km2) with the Ca-
nary/Madeira Islands to form an expanded Macarone-
sian mini-hotspot appears warranted. Clearly, these are
priority sites for conservation of these plant components
of Mediterranean-Macaronesian biodiversity.
As with the other Mediterranean-climate hotspots, di-
versity and endemism among vertebrates is much low-
er than for plants (Blondel and Aronson 1999). The
mammal and bird faunas are largely derived from extra-
Mediterranean biogeographical zones, with Eurasian
and African elements dominating the mammal fauna,
whereas Eurasian and semiarid southern elements
dominate the avifauna. The North African mammal fau-
na has closer affinities with tropical Africa than with
the Mediterranean Basin. On the other hand, the reptile
and amphibian faunas comprise mainly Mediterranean
species, and have higher levels of endemism. Many en-
demic species and genera are archaic lineages, which
have probably remained unchanged since their differ-
entiation before the Late Tertiary onset of Medi-
terranean climate conditions.
The present number of land mammals in the region
is about 224, of which 25 are endemic, including sever-
al standouts like the Barbary deer (Cervus elaphus
barbarus); Barbary macaque (Macaca sylvanus, VU);
Mediterranean monk seal (Monachus monachus, CR),
now thought to number fewer than 400 individuals;
and Iberian lynx (Lynx pardinus, CR), the world’s most
threatened felid with no more than 250 individuals sur-
viving. The region’s avifauna includes about 497 regu-
larly occurring species of which only 32 are endemic,
among them two Critically Endangered species: the
Raza Island lark (Alauda razae), found only on that is-
land in the Cape Verde Islands; and the Madeira or
Zino’s petrel (Pterodroma madeira), which has an esti-
mated breeding population of 20-30 pairs in the central
mountain massif of Madeira. A few small portions of
the hotspot also appear as priorities in BirdLife Inter-
national’s recent global analysis of Endemic Bird Areas
(EBAs) (Stattersfield et al. 1998), namely Cyprus,
Madeira, and the Canary Islands (with eight species
confined to this EBA), and Cape Verde.
Endemism is much better developed in reptiles, with
228 species, 77 of which are endemic; there are also four
endemic genera, namely Algyroides, Trogonophis, Macro-
scincus, and Gallotia (the latter being a genus of lizard
unique to the Canary Islands). Of the 86 species of am-
phibians in the hotspot, 27 are endemic. Reptile diversi-
ty is highest in the drier, eastern part of the European
portion of the basin and in North Africa, whereas the op-
posite is true of amphibians. For both groups, the Med-
iterranean Basin is an important center of diversity and
endemism for some families. These include, for amphib-
ians, the Discoglossidae (11 of the world’s 12 currently
recognized species, with nine endemic) and, for reptiles,
the Lacertidae (63 species, or 23% of the world total; 25
endemic), Testudinidae (five species, 16% of the world
total; one endemic, Testudo weissingeri), and Viperidae
(19 species, 8% of the world total; five endemic).
The inland fishes of this hotspot represent small sub-
sets of the rich Eurasian and African fish faunas from
which they are isolated. Consequently, the fauna is
species-poor (216 native species), although it includes
63 endemic species, six endemic genera, and even an
endemic family, Valenciidae, the tooth carps of the
Iberian and Greek peninsulas. These two peninsulas
provide the primary centers of fish endemism in the
hotspot, together containing 86% of its endemic fishes.
The present human population of the Mediterranean
Basin is some 300 million and increasing, and the im-
pact of a long history of human assault on Mediter-
ranean ecosystems has been huge. Perhaps the most se-
vere transformation has been the conversion of forests,
especially primeval deciduous forests, to agricultural
lands, evergreen woodlands, and maquis. Tourism, too,
has had a serious impact on semi-natural areas in west-
ern and southern Turkey, and in Cyprus, Tunisia, and
Morocco, a list that may shortly be joined by Greece
among several other countries, particularly as concerns
Mediterranean Islands such as the Balearics, Corsica,
Sardinia, Sicily, Crete, and the Canary and Madeira Is-
lands. Although reforestation is occurring in some of
the more developed areas of the Basin, e.g., southern
France and parts of northern Spain, Italy, and Croatia,
the pace of land degradation is accelerating in North
Africa, where human populations are still increasing at
a rapid rate (2.1% per year, way higher than the 1.6%
for developing countries as a whole, and the population
could surge from 150 million in 2003 to 206 million in
2025) (Population Reference Bureau 2003).
Much of the Basin’s current vegetation reflects the
influence of humans for several millennia. To this pro-
found extent, many, if not most, present-day land-
scapes are an example of co-evolution between nature
and humans. It would be erroneous in the case of the
Mediterranean to regard only purely pristine vegeta-
tion as “worthwhile” primary vegetation in the sense of
undisturbed original vegetation when, even by the year
1000, much of the Mediterranean Basin was bearing
widespread marks of human activity. For this reason,
today’s environments have been divided into pristine
(extremely rare, 1%-2%), semi-natural (fairly frequent,
albeit in small patches, for the most part uncultivated,
but somewhat managed, 20%-25%), and sub-natural
(the predominant form, with limited natural vegetation
and of contrasted relevance to the conservation of bio-
diversity, 73%-79%) (Naveh and Kutiel 1990).
The Mediterranean Basin Hotspot, therefore, emerges
as one of the hottest, having exceptionally high plant en-
demism, and one of the lowest percentages of natural
vegetation remaining in pristine condition (no more
than 5%) of any hotspot. Furthermore, the hotspot has a
surprisingly low protected area coverage, with only 4%
under some form of protection, and protected areas in
IUCN categories I to IV make up an extremely low 1% of
the surface area of the hotspot.
Protected areas have had a long history in the Medi-
terranean. As much as 2 000 years ago, several societies
set aside areas for protective purposes or, rather, as “re-
source reserves,” some of them surviving to the present
time. They were “developed as an ancient acknowledg-
ment of the scarcity of renewable resources and a need
to conserve and use them widely in support of sustain-
able rural economic development” (Sulayem 1994). To-
day, most countries of the Basin are planning substantial
increases in their protected area systems. This applies
especially to the Levant countries of Turkey, Syria,
and Lebanon. Because of the demands of agriculture and A group of sheep going out to
other activities that absorb large tracts of natural envi- pasture in the highland meadows
ronment, many protected areas are small, but a good deal of the Sierra de Cazorla, Spain.
© Patricio Robles Gil/Sierra Madre
is being accomplished through more flexible arrange-
ments of Biosphere Reserves, especially in those areas
where state authorities recognize their value. Biosphere
Reserves seek to safeguard natural environments in ar-
eas where there is a moderate degree of intrusion by hu-
mans. There is also a host of conservation efforts under
way, such as the European Union’s Habitats Directive,
popularly known as Natura 2000, which requires the
Mediterranean countries of the European Union to iden-
tify the more important natural sites and to formulate
conservation responses. Indeed, there is still much to be
hopeful about in this hotspot, where there is so much
biodiversity at stake.
147

On the opposite page, morning light
on the forests of the Vashlovani
Nature Reserve, in eastern Georgia.
© Pat O’Hara
148
CAUCASUS
NUGZAR ZAZANASHVILI 108 • GIORGI SANADIRADZE 108
ALEXANDER BUKHNIKASHVILI 109 • ANDREI KANDAUROV 108
DAVID TARKHNISHVILI 110
RUSSIA
BLACK
SEA
CASPIAN
GEORGIA SEA
AZERBAIJAN
ARMENIA
TURKEY
IRAN
0 400 km
The Caucasus region covers a total area of some
532 658 km2 in the nations of Georgia, Azerbaijan, and
Armenia; the North Caucasian portion of the Russian
Federation (including the Dagestan, Chechnya, In-
gushetia, Northern Ossetia, Kabardino-Balkaria, Karachai-
Cherkesia, and Adigea Autonomous Republics); the north-
eastern part of Turkey; and a part of northwestern Iran.
The definition of this region is somewhat different from
what was presented in Zazanashvili et al. (1999). The
northern boundary still conforms largely to the original
definition, but has been modified slightly in accordance
with Krever et al. (2001), while the southern boundary of
the hotspot intergrades with the Irano-Anatolian Hotspot
to the south, in accordance with WWF ecoregions (such
that Armenia now falls within both hotspots). In ad-
dition, the Caucasus Hotspot now includes the Hyrcan-
ian mixed forests of the south Caspian Coastal Plain and
northern slopes of the Elburz Mountains.
In terms of its origin, the Caucasus Isthmus is part of
a huge mountain belt, the Alpine Orogene, which em-
braces the whole of Eurasia from the Pyrenees and the
Atlas Mountains in the west to the Malay Peninsula and
Vietnam in the east. It is a region of natural contrasts,
and is composed of several prominent elements. These
include the North Caucasus Plain (the eastern part of
which is below sea level); the Greater Caucasus Range
(with the highest peak being Mt. Elburz at 5 642 m); the
Transcaucasian Depression; the Lesser Caucasus
Mountain Chain (to 4 000 m); and the South Caucasian
Uplands (covering parts of the Asia Minor, Armenian,
and Iranian uplands). Glaciers are concentrated main-
ly in the Greater Caucasus Range, with over 2 000 of
them covering 1 450 km2. Not surprisingly, the climate
is also very variable. Mean annual rainfall in the south-
western part of the region is quite high, exceeding
4 000 mm at some places in the coastal mountains of
the Black Sea; in the eastern part of the region, on the
Caspian Coast, it rarely exceeds 150 mm.
The vegetation of the Caucasus is quite diverse, in-
cluding broad-leaved and coniferous forests, subalpine
and alpine meadows, montane steppes and woodlands,
and semideserts, and depends on both the physical
features discussed above and the evolutionary history
of the flora. There are two refugia of Tertiary flora in
the region: the Colchis in the catchment basin of the
Black Sea, and the Hyrcanian in the extreme south-
eastern end of the Caucasus on the Caspian Sea coast.
Even now, many relict forms still appear as dominants
or codominants in a number of plant communities.
Species diversity and endemism are exceptionally
high for a temperate zone and relate to the geographic
position of the Caucasus, which lies at the junction of
two distinct zoogeographic regions: the Euro-Siberian,
and Irano-Turanian. Endemic species are common in
the western part of both the Greater and Lesser Cauca-
sus Ranges, as well as in the Talysh and Elburz moun-
tains on the southern coast of the Caspian Sea.
In terms of vascular plant diversity, some 6 400
species have been recorded in the Caucasus Hotspot, of
which more than 1 600 are endemic. In addition, there
are 17 endemic genera. About 35% of the endemic spe-
cies are thought to have originated in the Greater Cau-
casus Range, and many of these are high-mountain and
mountain xerophytic forms, as well as those growing on
rocks and scree (Dolukhanov 1966; Gagnidze et al.
2002). Genera with large numbers of endemic species
include: Saxifraga, Draba, Delphinium, Astragalus, Rosa,
Pyrus, Onobrychis, Scutellaria, Campanula, Symphyan-
dra, Pyrethrum, Primula, Heracleum, Jurinea, Psephellus,
and Cirsium. The region also harbors a remarkable con-
centration of economically important plants, particu-
larly wild crop relatives, such as wheat, rye, and barley
(but also fruits like walnuts, apricots, and apples).
In all, the Caucasus Hotspot has an estimated 130
mammal species; of these, 18 are endemic to the hot-
spot. As with other young mountain regions, the Cau-
casus has many newly evolved species, but also quite a
few relict species as well, as evidenced by the variety of
primitive species such as the unusual long-clawed
mole-vole (Prometheomys schaposchnikowi), the only
representative of an endemic genus, and the genera
Mesocricetus, Apodemus, and Sicista. The latter genus is
represented by four endemic species of unicolor birch
mice (Sicista caucasica, S. kluchorica, S. kazbegica, and S.
armenica); the Armenian birch mouse (S. armenica, CR)
is known only from the type locality. The Caucasian
tur (Capra caucasica, EN), a member of the goat family,
is also endemic to this hotspot.
Bird diversity is only moderate compared to the oth-
er hotspots, with 381 species, and endemism is low,
with only two endemics, the Caucasian black grouse
(Tetrao mlokosiewiczi) and the Caucasian snowcock (Te-
traogallus caucasicus). BirdLife International has listed
the Caucasus as an Endemic Bird Area (Stattersfield et
al. 1998). The Caucasus is also important for migratory
species, with two major migration routes passing
through the region —the east coast of the Black Sea and
the west coast of the Caspian Sea. The Batumi and
North East Anatolia bottlenecks are particularly good
places to observe migrants. Every spring and autumn,
millions of birds of several species fly over the Cauca-
sus Isthmus en route to their winter homes. Globally
threatened waterbird species in the region include the
marbled duck (Marmaronetta angustirostris, VU), lesser
white-fronted goose (Anser erythropus, VU), and white-
headed duck (Oxyura leucocephala, EN).
Reptiles are represented by 87 species, of which 20 are
endemic. The genera Lacerta and Darevskia, from the
family Lacertidae, exhibit particular diversity: of the 60
known species in the world, 28 species occur in the Cau-
casus, and 15 are endemic. Several of the Darevskia
species (D. dahli, D. armeniaca) have parthenogenic
populations, meaning there are no males and the fe-
males are able to reproduce entirely on their own. Rep-
tile endemics are concentrated in the xeric portions of
the hotspot.
Amphibian diversity is relatively low, with 17 species,
but four of these are endemic; three of these endemics
are found in the forest belt in the western part of the
Lesser Caucasus Range. The endemic Caucasian sala-
mander (Mertensiella caucasica, EN), sole representative
of its genus (although probably a species complex;
Tarkhnishvili et al. 2000), is a graceful and colorful ani-
mal, and the best example of an amphibian flagship spe-
cies in the region. The most remarkable species, howev-
er, is the Gorgan salamander (Batrachuperus gorganensis,
CR), which is restricted entirely to the Shir-Abad Cave
and the stream flowing from it, in northwestern Iran, and
is estimated to number only 100 breeding adults.
The Caucasus region has 127 fish species, only 12 of
which are endemic. Among the many interesting fea-
tures of this fish assemblage are the three lamprey
species: Caspiomyzon wagneri, Eudontomyzon mariae,
and Lampetra lanceolata. Lampreys are ancient, jawless,
scaleless fishes that date back 280 million years and have
the highest number of chromosomes of all vertebrates
(164-174) (Hardisty 1986). Another ancient group of fish-
es that is well represented in this hotspot are seven
species of sturgeon, including the famous Beluga stur-
geon (Huso huso), which is the largest freshwater fish,
weighing as much as 1 000 kg (Frimodt 1995). Popula-
tions of all sturgeon species have been reduced through
overharvesting, primarily for the high-value caviar, while
other threats include water pollution and damming that
restricts anadromous migrations (Birstein 1993).
The invertebrates, especially insects, are diverse
and, in the uplands, one can observe spectacular exam-
ples of the varied insect life, including an endemic but-
terfly (Parnassius nordmanni) and the Rosalia longicorn
beetle (Rosalia alpina, VU). Some endemic insects are
found in the foothills, including the Caucasian zerinthia
(Allancastra caucasica).
Humans have inhabited the Caucasus for a very long
time and, for that reason, natural ecosystems have
been affected by humans in many ways. Since the so-
cial and economic crises that began in 1992, the situa-
tion in the Caucasus has deteriorated significantly. Due
to a lack of energy resources, the population now con-
sumes 200%-300% more firewood per family every
year than it did in the 1980s, and illegal forest cutting
and timber exports have increased as well. More than
30% of the pastureland of the region suffers from vary-
ing levels of erosion; overgrazing, mainly by sheep, in
the subalpine and alpine summer ranges of the Central
Caucasus and in the winter ranges of the steppes and
semideserts increased by a staggering 250% at the end
of the 1980s. As a result, secondary communities now
occupy two-thirds of the pasturelands in the subalpine
belt and arid zones. Nonetheless, pristine areas remain
in a number of places, particularly in inaccessible
gorges and other remote, high-altitude areas. However,
looking at the region as a whole, we estimate that about
145 000 km2, or 27% of the land area, remains in rea-
sonable condition, while still less, only about 11%-13%
of the original extent of natural vegetation in the Cau-
casus, remains in pristine condition.
Poaching is also a serious problem that increased sig-
nificantly in the 1990s, with animals like the leopard
(Panthera pardus), brown bear (Ursus arctos), wolf (Canis
lupus), Caucasian red deer (Cervus elaphus maral), be-
zoar goat (Capra aegagrus, VU), Caucasian chamois
(Rupicapra rupicapra caucasica, VU) and tur being most
heavily impacted. The leopard is now down to no more
than 40 individuals in the entire region (WWF Technical
Reports 2002-2004). The Caucasian tur, which still num-
bered in the hundreds of thousands in the middle of this
century, has now been drastically reduced because of
the combination of poaching and greatly increased de-
velopment of pastureland. Today, there are only about
4 000 of the eastern subspecies and 6 000-10 000 of the
western subspecies still in existence (Weinberg 1997).
Fortunately, there is a long tradition of nature con-
servation in the Caucasus, with the first State (Strict) Na-
ture Reserve (Zapovednik), the Lagodekhi Gorge in Geor-
gia, having been created in 1912 on the southeastern
slope of the Greater Caucasus Range. Subsequently, es-
tablishment of strict nature reserves became a key part
of conservation activities in the Russian Empire and
then the Soviet Union. According to the World Database
on Protected Areas, in this hotspot protected areas in
IUCN categories I to IV cover approximately 6.7% of its
land area. Taking into account other forms of managed On the opposite page,
reserves (sanctuaries or wildlife refuges), where some the mountains of the Racha region
kind of protective regime is also in place, around 8% of in Georgia.
© Pat O’Hara
the hotspot is afforded some form of protection. Regimes
for protection generally forbid logging, drainage of wet- Above, an azure butterfly
lands, use of chemicals, and any other intensive forms (Celastrina sp.) photographed in
of nature use, but often permit hunting and fishing. Lagodekhi Nature Reserve, Georgia.
And there are positive developments in the region, © Pat O’Hara
too. In a presidential statement on the occasion of the
launch of the “WWF-2000-Living Planet Campaign,” a
pledge was made that by the year 2010, Georgia would
increase forest protected areas (IUCN I to IV) coverage
to 15% of its total forest area as a “Gift to the Earth.”
Azerbaijan’s Ministry of Ecology and Natural Resources
151
 has been actively developing a protected areas system
in the country, and several existing nature reserves
have been expanded and national parks established.
There are concrete plans for creating new protected ar-
eas and connecting corridors in the Caucasus part of
the Russian Federation and in Armenia. Turkish activi-
ties have focused on strengthening the protected areas
system, while sustainable resource use programs in the
Camili region are ongoing.
In 2003, the Critical Ecosystem Partnership Fund
(CEPF) began investing in the Caucasus Hotspot. A joint
initiative of Conservation International, the Global En-
vironment Facility, the Government of Japan, the John
D. and Catherine T. MacArthur Foundation, and the
World Bank, CEPF’s five-year, $8.5-million investment in
the Caucasus will target a slightly larger region than dis-
cussed here. CEPF will focus on conserving the hot-
spot’s globally threatened species, the majority of which
are found in specific sites in five target corridors:
Greater Caucasus, Caspian, West Lesser Caucasus, East
Lesser Caucasus, and Hyrcan. This is one example of
conservation plans that have been developed for the re-
gion based on wide stakeholder involvement, but these
processes are still in crucial need of further financial
and technical support from the international donor
community and conservation organizations.
WESTERN GHATS
AND SRI LANKA
AJITH KUMAR 88 • ROHAN PETHIYAGODA 89 • DIVYA MUDAPPA 90
Hanuman langurs
(Semnopithecus entellus)
are found across India, Pakistan,
Bangladesh, Burma, and Sri Lanka INDIA
(the endemic subspecies S. e. priam
is pictured here).
© Patricio Robles Gil/Sierra Madre INDIAN
OCEAN
BAY OF
On the opposite page, BENGAL
the Endangered lion-tailed macaque
(Macaca silenus), once considered
to be on the verge of extinction, is
now holding its own thanks to more
than 25 years of successful
conservation action. SRI
LANKA
© Elio Della Ferrera/naturepl.com
0 400 km
The Western Ghats of southwestern India and the
highlands of southwestern Sri Lanka, separated by
400 km (including the Palk Strait and Gulf of Mannar),
are somewhat similar in their geology, climate, and
evolutionary history. The montane forest ecosystems
that characterize both regions are subject to monsoon
152
weather patterns that mediate the warm, tropical cli-
mate and contribute to the high levels of biodiversity
in the region.
The Western Ghats, known locally as the Sahyadri
Hills, are formed by the Malabar Plains and the chain of
mountains running almost parallel to India’s western
coast, approximately 30-50 km inland, traversing the
states of Tamil Nadu, Kerala, Karnataka, Goa, Maharash-
tra, and Gujarat. In all, the Western Ghats stretch over
a distance of 1 600 km, interrupted only by the 30-km
Palghat Gap.
The Western Ghats (which have their highest point
at 2 695 m; Anamudi Peak) mediate the rainfall regime
of peninsular India by intercepting monsoon storm sys-
tems. Areas to the west of the higher elevations receive
the greatest annual rainfall, 3 000 mm on average, with
80% of it falling during the period of the southwest
monsoon (June-September) and the balance during the
northeast monsoon (October-November). Annual rain-
fall decreases considerably along the eastern slopes of
the Western Ghats; plains along the foothills and with-
in the rainshadow, although only 10 km from the wet-
test areas, receive as little as 500 mm each year. Rainfall
also decreases from south to north, especially north of
the Palghat Gap.
The wide variation of rainfall patterns in the West-
ern Ghats, coupled with the region’s complex geogra-
phy, produces a great variety of vegetation types.
These include scrub forests in the low-lying rainshad-
ow areas and the plains, deciduous and tropical rain-
forests up to about 1 500 m, and a unique mosaic of
montane forests and rolling grasslands (shola) above
1 500 m. Tropical rainforests represent primary centers
of species richness and endemism within the Western
Ghats, and cover approximately 20 000 km2. Dry, moist
deciduous, and scrub forests cover another 20 000 km2
(Collins et al. 1991).
The most detailed study of the Western Ghats rainfor-
ests is that of Pascal (1988), who identified six major as-
sociations at low altitudes (less than 850 m), four at
medium elevations (850-1 500 m), and one at elevations
above 1 500 m. Of these, the Dipterocarpus bourdilloni-D.
indicus-Anacolosa densiflora type (confined to altitudes
below 700 m, including the coastal zone and areas south
of the Palghat Gap) is believed to be the richest one, ho-
mologous to the lowland dipterocarp forests of South-
east Asia. Unfortunately, today this forest remains only
as sacred groves and very small remnants in various
stages of degradation, and D. bourdilloni itself has be-
come extremely rare.
Sri Lanka is a continental island separated from the
southern tip of peninsular India by the 20-m-deep Palk
Strait. The island, some 67 654 km2 in size, has been re-
peatedly connected with India between successive in-
terglacials, the most recent one, in which sea level was
about 120 m below present (Siddall et al. 2003), result-
ing in a 140-km-wide land bridge 20 000 years ago. De-
spite these repeated connections, the fauna and flora of
the island show remarkably high endemism against a
background of a more generalized biota representative
of southern India (Davis et al. 1995; Wikramanayake et
al. 2002).
Precipitation across Sri Lanka is dependent on mon-
soonal winds, resulting in much of the island’s experi-
encing relatively low rainfall (less than 2 000 mm per
year), except for the southwestern “wet zone” quarter,
where precipitation ranges from 2 000-5 000 mm per
year. While dry evergreen forests occupy almost the
entirety of the “dry zone,” dipterocarp-dominated rain-
forests dominate the lowlands of the wet zone, and
some 220 km2 of tropical montane cloud forest still per-
sist in the central hills, which rise to a maximum alti-
tude of 2 524 m.
In terms of plant diversity, the Western Ghats har-
bors approximately 5 000 species of flowering plants
belonging to nearly 2 200 genera and 217 families;
about 1 700 species (34%) are endemic. There are also
58 endemic plant genera, 49 of which are monotypic
and some highly speciose (e.g., Niligrianthus with 20
species). Some prominent genera and families are rep-
resented by large numbers of endemic species, such as
Impatiens with 76 of 86 species endemic, Dipterocarpus
with 12 of 13 species endemic, and Calamus with 23 of
25 species endemic. Of the 490 tree species (over 10 cm
girth at breast height, or GBH) recorded from low- and
mid-elevation forests, 308 (63%) species representing
58 families are endemic (Ramesh and Pascal 1991). The
only gymnosperm tree, Podocarpus (=Nageia) walli-
chianus, is also endemic. Of the 267 species of orchids
(representing 72 genera), 130 are endemic. Similarly,
plant diversity and endemism in Sri Lanka is quite
high, with 3 210 flowering plant species in 1 052 gen-
era, of which 916 species and 18 genera are endemic
(Dassanayake and Fossberg 1980-2002). In addition, the
island’s ferns (although not recently assessed) are esti-
mated to number about 350 species. Approximately
433 plant species, and at least five genera, are confined
to Sri Lanka and the Western Ghats combined. Overall,
an estimated total of 3 049 species are endemic to the
hotspot (out of a minimum of 5 916 species), as are at
least 81 genera.
In the Western Ghats, the Agasthyamalai Hills in the
extreme south are believed to harbor the highest levels
of plant diversity and endemism at the species level.
Nearly 87% of the region’s flowering plants are found
south of the Palghat Gap (37% being exclusive to this
sub-region); these figures decrease to about 60% and
5%, respectively, in the Nilgiri Hills. In Sri Lanka, di-
versity, richness, and endemism across all taxa are
much higher in the wet (including the montane) zone
than in the dry zone. Indeed, the wet zone, which ac-
counts for only a quarter of Sri Lanka’s territory, con-
tains 88% of the flowering plants occurring in the is-
land, and 95% of its angiosperm endemics. Convincing
evidence is emerging from recent phylogenetic and
phytogeographic studies of Dipterocarpaceae (Guna-
sekera and Dayanandan 2003), and molecular dating
studies of Crypteroniaceae (Conti et al. 2003) and the
frog family Ranidae (Roelants et al., in press), that
the wet zone biota of Sri Lanka is a unique relict of the
Deccan-Gondwanic biota and, therefore, of consider-
able biogeographic significance.
Vertebrate diversity and endemism across the hot-
spot is impressive. Approximately 140 mammal species
have been recorded (49 of them bats), although en-
demism is relatively low, with only 18 endemic spe-
cies. In addition, four genera are endemic, each rep-
resented by single species. Three are confined to Sri
Lanka: Pearson’s long-clawed shrew (Solisorex pear-
soni, EN), Kelaart’s long-clawed shrew (Feroculus fero-
culus, EN), and the Ohiya rat (Srilankamys ohiensis).
The bat Latidens salimalii (CR) is endemic to the High
Wavy Mountains in the Western Ghats. Among flag-
ship mammal species, the most prominent are the
lion-tailed macaque (Macaca silenus, EN) and the en-
demic Nilgiri tahr (Hemitragus hylocrius, EN). One of
the most threatened Indian mammals, the Malabar
civet (Viverra civettina, CR), is known only from the
Malabar Plains, which are densely populated and the fo-
cus of most developmental activities. The Western
Ghats is also home to the largest population of Asian
elephant (Elephas maximus), with about 11 000 ani-
mals, while in Sri Lanka the species has now been al-
most completely extirpated in the wet zone and re-
duced to an estimated population of around 2 500
elsewhere in the island.
A total of 457 bird species occur regularly within the
Western Ghats and Sri Lanka, but only 35 are endemic.
Of the endemics, seven occur in low-elevation forests,
including species such as the grey-headed bulbul (Pyc-
nonotus priocephalus), white-bellied tree pie (Dendrocit-
ta leucogastra), and Malabar parakeet (Psittacula colum-
boides), while seven others are associated with higher
elevations, among them the white-bellied shortwing
(Brachypteryx major, VU), Nilgiri flycatcher (Eumyias
albicaudata), and broad-tailed grassbird (Schoenicola
platyura, VU). In recognition of the significant regional
avian diversity, both the Western Ghats and Sri Lanka On the opposite page,
are considered as Endemic Bird Areas by BirdLife In- the Endangered Nilgiri tahr
ternational (Stattersfield et al. 1998); there are no few- (Hemitragus hylocrius) occurs in
very fragmented habitat in the
er than 14 bird species restricted entirely to the wet
Western Ghats of India, with the
zone of Sri Lanka. majority of its population in the
As is true for a number of other biodiversity hot- Eravikulam National Park. Most of
spots, the highest levels of vertebrate endemism within the species’ habitat has been lost,
this hotspot are among amphibians and reptiles. Of the but remarkably it persists in some
179 species of amphibians reported, 138 are endemic. heavily disturbed areas.
© Patricio Robles Gil/Sierra Madre
In the case of Sri Lanka, amphibian diversity is only
now becoming better known, and the country’s wet Above, in general, hornbills are
zone may contain as many as 140 endemic species frugivorous and arboreal forest
alone (Meegaskumbura et al. 2002). Amphibian en- dwellers, like these two Indian or
demism is also impressive at the generic level, with 10 Malabar pied hornbills
genera (of a total of 29) occurring only here. Recently, (Anthracoceros coronatus)
from Sri Lanka.
a new burrowing anuran family, Nasikabatrachidae,
© Patricio Robles Gil/Sierra Madre
closely related to the Sooglossidae from the Seychelles,
has been described following its discovery in the Iddu-
ki District in Kerala in the Western Ghats (Biju and
Bossuyt 2003); this represents the only endemic family.
155

While the Asian elephant (Elephas
maximus) remains at the center of
species conservation activity in
Sri Lanka, a growing human
population and poorly planned
land-use changes have resulted in
an average of three elephants being
killed each week in protection of
crops. Sri Lanka’s elephants show
remarkable genetic diversity and
differ significantly from the Indian
mainland population.
© Patricio Robles Gil/Sierra Madre
On the opposite page, forest loss in
the Western Ghats has been driven
largely by shifting agriculture, and
by conversion to plantations of tea
(pictured here), coffee, teak,
eucalyptus, and wattle.
© Patricio Robles Gil/Sierra Madre
156
About 265 species of reptiles have been recorded, and
175 of these are endemic; one-quarter (22) of all genera
represented are endemic, and nine of these are mono-
typic. Families such as Uropeltidae (46 of 47 species),
Gekkonidae (19 of 30), and Agamidae (20 of 26) exhibit
very high endemism.
Besides terrestrial vertebrates, the Western Ghats and
Sri Lanka Hotspot harbors 191 species of strictly fresh-
water fishes belonging to 45 genera; 139 species and eight
genera are endemic (including one genus in Sri Lanka,
Malpulutta). In the Western Ghats, the southern region is
known to be more diverse than the northern and central
regions (Daniels 2001; Dahanukar et al. 2004).
Although knowledge of invertebrate diversity is poor,
levels of endemism within certain groups in the West-
ern Ghats and in Sri Lanka are believed to be signifi-
cant; for example, among the tiger beetles, around 101
of 139 recorded species are endemic (D. Pearson, pers.
comm.). This may not hold true across groups: the
number of butterfly species in this region is relatively
low, with only 37 endemics of 330 (Kunte et al. 1999) in
the Western Ghats, and 24 of 234 species endemic in Sri
Lanka (D’Abrera 1998). The wet zone of Sri Lanka is
known to contain several unique endemics such as
Krombeinictus nordenae, a pollen-provisioning sphecid
wasp (Krombein and Norden 1997) and Aneuretus si-
moni, the sole surviving representative of one of the
world’s 14 ant subfamilies (Jayasuriya and Traniello
1985).
The high levels of biodiversity in the Western Ghats
and Sri Lanka face the same pressures as other hot-
spots, particularly being part of one of the hotspots with
the highest population density (Cincotta et al. 2000),
and it is likely that no more than 25% of the original
vegetation of this hotspot remains intact. In the late
1980s, India’s remaining forest cover already stood at
less than 20% of its original extent. Forest loss has been
driven largely by shifting agriculture; conversion to
plantations of tea, coffee, teak, eucalyptus, and wattle,
as well as for the creation of reservoirs; and construc-
tion of roads and railways. There are no reliable esti-
mates for recent rates of total forest loss for the Western
Ghats. A study that estimated changes in forest cover
between 1973 and 1995 in the southern part of the
Western Ghats using satellite data found that the area
(approximately 40 000 km2) had lost 25.6% of its forest
cover (Jha et al. 2000). In all, approximately 20% of the
natural forest vegetation of the Western Ghats is
thought to remain (Collins et al. 1991), but in a highly
fragmented state, and the only tracts greater than
200 km2 are found in the Agasthyamala Hills, Car-
damom Hills, Silent Valley-New Amarambalam Forests,
and southern parts of the South Kannada District in
Karnataka State.
On Sri Lanka, aquatic habitats are increasingly
threatened by invasive alien species (more than 90% of
the freshwater-fish biomass comprises exotics; Pethi-
yagoda 1994), and forests are threatened by encroach-
ment and illegal produce extraction (50% of the island’s
energy demand is met with uncultivated firewood).
While the wet zone harbors greater diversity and en-
demism than the dry zone, the former contains 67% of
the island’s 19.3 million inhabitants (Population Refer-
ence Bureau 2003) —i.e., a density of 740 people per
km2, which is exceptional by standards of other global
hotspots (Cincotta et al. 2000). The main threats to the
island’s remaining rainforests are habitat fragmenta-
tion, together with invasive species, pesticide influx,
and edge effects. Although 33% of Sri Lanka’s territory
is still forested, much of this forest is in the biodiversi-
ty-poor dry zone. Only 4.6% of the wet zone (800 km2)
now contains closed-canopy rainforest, the remainder
having been cleared for the cultivation of cinchona,
coffee, tea, and rubber during the past 150 years. The
remaining expanse of wet-zone natural forest compris-
es some 140 fragments, the largest three of which are
Peak Wilderness (250 km2), the Knuckles Hills (175 km2),
and the Sinharaja World Heritage Site (90 km2), but at
least 120 of these fragments are less than 10 km2 in
extent.
For the purpose of addressing future threats to biodi-
versity conservation in this region, a major priority is
the establishment of protected areas. A national pro-
tected area system has existed in India for more than a
century, and was strengthened considerably with pas-
sage of the Wildlife Protection Act of 1972 and the Na-
tional Wildlife Action Plan of 1983. Of India’s 10 bio-
geographic zones (Rodgers et al. 2002), the Western
Ghats has the highest percentage of protected area cov-
erage on the mainland (15%), which is represented in
20 national parks and 68 sanctuaries (Rodgers et al.
2002). Sri Lanka prides itself in that 13% of its territory
comprises protected areas, although this is an underes-
timate, because a further 16% of the territory is pro-
tected as forest reserves which, since an administrative
ban on logging in 1989, are now in effect protected (not
yet formally so designated). The challenge of biodiver-
sity conservation in this country, therefore, lies largely
in the conservation of the remaining extent of wet-zone
natural forest. Overall, approximately 23% of this hot-
spot has some level of protection, although only 11% is
represented in protected areas classified in IUCN cate-
gories I to IV.
Biodiversity conservation in the region has a long
history, and one that is enshrined to some degree in re-
ligion, culture, and mythology. There is clearly an in-
creased awareness among local people regarding nat-
ural resources and the need to conserve them, which
provides an outlook for biodiversity conservation that,
at least in the case of India, is perhaps much brighter in
this region than elsewhere in the country. Problems do
loom on the horizon, however, including government
plans for new hydroelectric projects, roads, and rail-
ways, as well as possible movements by local groups in
India to reclaim rights to forest lands for traditional and
subsistence use.
 MOUNTAINS OF
SOUTHWEST CHINA
DAVID E. BOUFFORD 94 • PETER PAUL VAN DIJK71 • LU ZHI 72
CHINA
INDIA
MYANMAR
(BURMA)
0 300 km
The Mountains of Southwest China Hotspot, perhaps
more appropriately referred to as the Hengduan Shan,
covers the temperate to alpine montane region lying be-
tween the easternmost edge of the Qinghai-Xizang (Ti-
betan) Plateau and the Central Chinese Plain. It lies to
the north of the Indo-Burma Hotspot, and to the imme-
diate east of the Himalaya Hotspot, and is bounded to
the northwest by the dry Qinghai-Xizang Plateau, to the
north by the Tao River of extreme southern Gansu, and
to the east by the Sichuan Basin and the plateau of east-
ern Yunnan.
The hotspot is characterized by extreme topography
of steep, often precipitous mountains, with altitudes
ranging from less than 2 000 m in some valley floors to
7 558 m at the summit of Gongga Shan. The mountain
ridges are oriented in a generally north-south direction,
perpendicular to the main Himalayan chain. The region
includes the Hengduan Shan, Gaoligong Shan, and Nu
Shan mountains of western Yunnan; the Nyainqentang-
lha Shan, Ningjing Shan, Taniantaweng Shan, and others
at the southeastern edge of the Qinghai-Xizang Plateau;
the Shaluli Shan, Daxue Shan (including Gongga Shan),
Chola Shan, and Qionglai Shan mountain systems of
Sichuan; and the Min Shan on the Sichuan-Gansu bor-
der. The Ailao Shan and Wuliang Shan of central Yunnan
are not part of this hotspot (being instead included in
the Indo-Burma Hotspot). The total surface area of the
Mountains of Southwest China Hotspot (which, accord-
ing to the delineation of WWF ecoregions, results in the
exclusion of some high-altitude regions that we include
here) is approximately 262 466 km2.
The complex topography of the region results in a
wide range of local climatic conditions. Correlated to al-
titude, latitude, and exposure, temperatures range from
frost-free throughout the year in parts of Yunnan and
short, frost-free periods at the northern boundary of the
region, to permanent glaciers on the high mountain
peaks of Sichuan, Yunnan, and Xizang. Annual average
rainfall in the region exceeds 1 000 mm on southwest-
ern slopes at higher altitudes in Yunnan, while areas of
the northwestern part of the region, in the rainshadow
of the Qinghai-Xizang Plateau, rarely receive more than
400 mm annually.
Given the wide range of topographic and climatic
conditions, the mountainous area of western Sichuan
possesses a great variety of vegetation types. These in-
clude broad-leaved and coniferous forests, bamboo
groves, scrub communities, savanna, meadow, prairie,
freshwater wetlands, and alpine scrub and scree com-
munities (Sichuan Vegetation Study Group 1980). Fur-
thermore, as a result of the complex local geomorphol-
ogy, the north-south orientation of the mountains, and
the huge vertical differences in topography, vertical
zonation of vegetation formations is also well devel-
oped. The extremes of topography, vegetation, and cli-
mate, coupled with the isolation of the numerous steep
peaks and ridges —like islands in the sky— have served
as a stimulus to evolution, and are responsible for the
high species diversity and endemism of plants, verte-
brates, and invertebrates found in the region.
Vascular plant diversity is particularly impressive in
this hotspot, with the mountains of western Sichuan,
Yunnan, and southeastern Xizang being home to as
much as 40% of the total for all of China, or about 12 000
species. Of these, about 3 500 species (29%) and at least
20 genera are endemic, including about 100 endemic
ferns. More than a quarter of the world’s Rhododendron,
Primula, Corydalis, Anaphalis, Delphinium, Gentiana,
Saussurea, and Sorbus species occur in the Hengduan
Mountains of western Sichuan, while over half the spe-
cies of Ligularia, Cremanthodium, Cotoneaster, and Pedicu-
laris have been recorded in this hotspot. Of particular
interest is the occurrence of two endemic families, Cir-
caeasteraceae (two genera, Circaeaster and Kingdonia) On the opposite page, the
and the monotypic Acanthochlamydaceae (Wu 1988; Vulnerable golden or snub-nosed
Xiong 1989; Xu 1991; Wang et al. 1993, 1994; Zhou 1994). monkey (Rhinopithecus roxellana)
lives in temperate forests and moves
In terms of vertebrate diversity, some 237 mammal
in large groups that can exceed 300
species occur in this hotspot, of which only five are en- individuals. This photograph was
demic. For a relatively poorly known region, this hot- taken at the Beihai Park, Sichuan.
spot has some of the world’s most interesting and best- © Patricio Robles Gil/Sierra Madre
known flagship species, of which the giant panda
(Ailuropoda melanoleuca, EN), perhaps the world’s num- Above, mother and infant
Yunnan snub-nosed monkey
ber one wildlife symbol, is the most famous. At present,
(Rhinopithecus bieti), an
this species survives in fragmented populations con- Endangered species found only in a
fined to over 40 reserves stretching from western few isolated forests in northern
Sichuan to southern Gansu and southern Shaanxi. Oth- Yunnan and adjacent Tibet. It has
er important mammal flagships include the golden one of the highest altitudinal ranges
monkey (Rhinopithecus roxellana, VU) and the Yunnan of any primate species, occurring
up to 4 200 m above sea level.
or black snub-nosed monkey (R. bieti, EN), which lives
© Patricio Robles Gil/Sierra Madre
at higher altitudes (as much as 4 500 m) than any other
nonhuman primate; the takin (Budorcas taxicolor, VU),
a 300-kg goat-antelope with a coarse, dense brown coat;
and Bailey’s goral (Nemorhaedus baileyi, VU), a goat spe-
cies endemic to the Gaoligong Shan.
The bird fauna of the Mountains of Southwest China
Hotspot is quite rich, with 611 species recorded, includ-
159

Woman collecting mushrooms
in Shiang Gu Ching village,
Yunnan Province.
© Patricio Robles Gil/Sierra Madre
On the opposite page, view
of Hengduan Shan Mountain Range
in Yunnan Province, China.
© Patricio Robles Gil/Sierra Madre
On pp. 162-163, the Mountains
of Southwest China Hotspot has
very high mushroom diversity and
many of these species are used as a
source of food. Here, a man from
Shigu, in Yunnan Province,
carrying mushrooms to market.
© Cristina G. Mittermeier
160
ing residents, migrants, and vagrants. Of these, only a
single species is endemic. No fewer than four Endemic
Bird Areas (EBAs) have their boundaries largely over-
lapping with this hotspot (Stattersfield et al. 1998). A re-
cent study by Lei et al. (2003) has confirmed that the
Hengduan Shan is a center for avian endemism. This
region also has the world’s richest variety of pheasants
and their relatives, with 27 species inhabiting the hot-
spot, among them the Lady Amherst pheasant (Chryso-
lophus amherstiae). For sheer brilliance, however, noth-
ing exceeds the iridescent splendor of two monal
pheasants, Sclater’s monal (Lophophorus sclateri, VU)
and the Chinese monal (L. lhuysii, VU).
The hotspot’s herpetofauna is surprisingly diverse,
given the generally temperate climate and the physio-
logical constraints on ectothermic animals. In all, this
hotspot is inhabited by about 94 reptile species, of which
15 are endemic; amphibians are represented by around
98 species, of which 40 are endemic. Particularly re-
markable among amphibians are the radiations of two
closely related megophryid frog genera, Scutiger and
Oreolalax; of the 12 species of the former that are pres-
ent, nine are endemic, while all 11 species of Oreolalax
present are endemic (70% of the entire genus). Some of
these species occur at relatively high altitudes; for exam-
ple, the Xizang alpine toad (Scutiger boulengeri) is found
to elevations of more than 5 000 m above sea level.
The Mountains of Southwest China are important be-
cause they feed the most species-rich river systems in
Asia, including both temperate and tropical examples.
Major river systems that traverse or originate in the
hotspot include the Jingshajiang, Yalongjiang, Daduhe,
and Minjiang, all branches of the Yangtze River leading
to the East China Sea, while Lancangjiang (Mekong Riv-
er), which passes through Guangxi Province, Laos, Cam-
bodia, and Vietnam, leads to the South China Sea.
Nujiang reaches the Indian Ocean through Guangxi Prov-
ince and Burma. Within the hotspot itself, the fish fauna
comprises 92 species, of which 23 species and two gen-
era are endemic. The fish fauna is represented mainly
by species of the families Cyprinidae (50) and Balitoridae
(18), while the endemic species are predominantly in
two genera: Schizothorax (eight) and Triplophysa (three).
Despite the inaccessibility of much of the region, the
major threat to biodiversity in this mountainous hotspot
before 1998 was commercial logging. Overall forest cov-
er in Sichuan, one of China’s main sources of timber, de-
clined from 19% in the 1950s to 12.6% in 1988 (Smil
1992). Almost all this loss has been in the more remote
and mountainous western part of Sichuan, since forests
elsewhere in the province were cut long ago (MacKinnon
et al. 1996). Removal of trees is not limited to timber
species. 1998 saw the launch of a national logging ban by
the government, since which commercial logging has
ceased. However, forests are still being cut at a significant
rate for fuelwood and house construction by local resi-
dents who live in alpine areas in the hotspot. The high
valleys are bitterly cold in winter and transportation fa-
cilities are limited, so local collection and stockpiling of
firewood is essential. In recent years, as economic con-
ditions have improved, houses built by local villagers
have been getting bigger and bigger, and it is estimated
that each consumes nearly 300 m3 of timber per year.
The volume of firewood collected and timber cut for
house construction may exceed past logging harvests in
some areas and is a primary cause of deforestation today
(MacKinnon et al. 1996). Furthermore, nomadic Tibetans
wander throughout the entire region with their herds of
yak, sheep, and goats to utilize the natural pastures. Tra-
ditionally, they ascend to alpine pastures in summer and
return to valleys before winter. Nearly all areas without
natural forest cover, or unsuitable for crops, are heavily
grazed by livestock, the only exceptions being scree
slopes and vertical cliffs (MacKinnon et al. 1996).
Looking at what remains in terms of natural vegeta-
tion, MacKinnon et al. (1996) calculated that remaining
areas of forest covered 23% of Yunnan Province, 12% of
Sichuan Province, and 5.1% of Xizang Province in 1992.
The forest cover figure for Sichuan is probably still a rea-
sonable estimate for the entire hotspot, although much
of the region has not been covered by forest since the Hi-
malayan uplift. However, the intensity of grazing has
had a conspicuous impact nearly everywhere, to the
point that it is difficult to know what exactly constitutes
natural vegetation. Certainly, less than 10% of the entire
hotspot area is in pristine condition, and quite likely it is
as low as 5%-6%, although, for purposes of this analysis,
we use an estimate of 8% remaining. Since the 1980s,
and particularly from the late 1990s onwards, forest re-
generation has been taking place on logged sites and
“wasteland.” While the forested areas have increased, an-
other threat has arisen in the form of monoculture plan-
tations, sometimes with alien species such as Eucalyptus.
Over 150 protected areas covering more than
70 000 km2 have been declared in the area that is in-
cluded here in the hotspot. The region’s most famous
protected area is Wolong Nature Reserve in western
Sichuan, which covers about 2 000 km2. Included in the
list of protected areas are 28 nature reserves designated
for giant panda conservation, totaling about 16 800 km2;
these reserves account for nearly half of all forest land
under protection in this hotspot. Interestingly, Wolong
has a great diversity of ancestral species from which
many fruit trees and shrubs were domesticated, such as
apples, pears, plums, cherries, gooseberries, black cur-
rants and strawberries, representing a genetic resource
of inestimable value (MacKinnon 1996). In total, 4 000
plant species occur in the reserve, including many rare,
endemic, and otherwise notable species such as Tetra-
centron sinense, Chinese yew (Taxus chinensis), and Chi-
nese katsura (Cercidiphyllum japonicum var. sinense)
(Zhao et al. 1990).
Other important protected areas in the hotspot in-
clude: Emei Shan, an isolated limestone mountain on the
eastern edge of the hotspot that rises to 3 099 m in
height, is considered to be one of the botanically richest
and most diverse mountains in the Northern Hemi-
sphere, and also harbors an important population of the

Griffith’s leaf monkey
(Trachypithecus villosus villosus)
is a wide-ranging species that
occurs in much of the Sundaland
Hotspot, with several closely-related
subspecies extending into the Indo-
Burma Hotspot. It is especially
common in mangroves and other
coastal habitats.
© Bernard Walton/naturepl.com
On the opposite page, Wagler’s pit
viper (Tropidolaemus wagleri)
from Sabah, Borneo. The species
ranges from southern Thailand,
through peninsular Malaysia to
Borneo, Sumatra, Sulawesi, and the
southern Philippines.
© Nick Garbutt/naturepl.com
164
Tibetan macaque (Macaca thibetana), being one of the
few places where these animals can readily be seen; Luo
Ji Shan Nature Reserve, which contains more than 2 000
species of higher plants, including more than 50 species
of Rhododendron; and Gaoligong Shan Nature Reserve.
The expansion of Gaoligong Shan (4 052 km2) north-
westward into Xizang and eastward across the Nu Jiang
(Salween) and Lancang-Jiang (Mekong) rivers to link
with the Bai Ma Xue Shan Nature Reserve (2 816 km2)
and the east bank of Jinsha River (part of the Yangtze)
has created the famous Three Parallel Rivers World Her-
itage Site, which provides protection for the many differ-
ent ridges and valleys and their highly distinctive floras.
Although some important areas are currently pro-
tected in this hotspot, much more needs to be done.
The enormous watershed value of this region in and of
itself provides more than adequate justification for in-
creased protection. Furthermore, the scientific impor-
tance of this hotspot can hardly be overestimated, since
it provides a living laboratory for the study of biogeo-
graphical patterns and evolutionary processes critical
to evaluating hypotheses on the diversification of
plants and animals.
In spite of the importance of this hotspot, time is
short. The many old pressures on fragmented natural
habitats from grazing, clearance, hunting, and collec-
tion of forest produce remain, and new threats, such as
dam building on all main rivers in the hotspot, mining,
and unplanned mass tourism development accompa-
nied by road expansion and wildlife consumption are
emerging. This means that the extinction of many of
the restricted-range species of plants and animals is a
realistic and immediate possibility. Since the late 1990s,
the Chinese government has implemented intensive
land protection policies in the hotspot area, including
the logging ban, the Grain to Green Project (conversion
of sloping farmland to forest and grassland, under way
in the provinces of Sichuan, Shaanxi, and Gansu), and
the Endangered Species and Protected Areas Program,
along with the financial investment of billions of dol-
lars. Many new protected areas have been planned.
NGOs, working with the government, have also made
significant contributions in this great movement, espe-
cially on capacity building and applied research to pro-
vide a sounder scientific basis for both conservation
and development policies, designation of protected ar-
eas, and the use of local plant and animal resources.
One other significant feature of this hotspot is its rich
cultural diversity. There are 16 ethnic groups living in
this area, and the majority of the region is inhabited by
Tibetans. In Tibetan culture, sacred landscape is a tra-
ditional land protection system that still exists today.
Thousands of villages and monasteries each have their
own sacred site —mountain, lake or forest— where old-
growth forests can be found. However, this tradition has
been challenged by outside influences and the demand
for economic development. Preserving and reviving
such a tradition will provide much impetus to protect
the biodiversity in this hotspot.
SUNDALAND
TONY WHITTEN 69 • PETER PAUL VAN DIJK 71
LISA CURRAN 45 • ERIK MEIJAARD 46 • PETER WOOD 70
JATNA SUPRIATNA44 • SUSIE ELLIS 1
SOUTH
CHINA SEA
0 800 km
THAILAND
M A L A Y S I A
BRUNEI PHILIPPINES
Simeulue
SINGAPORE Sulawesi
Nias
Batu Kalimantan
(Borneo)
Mentawai Sumatra Sulawesi
Islands
INDONESIA
Java Bali
INDIAN
OCEAN
Cocos Christmas
Island
Islands
The Sundaland Hotspot covers the western half of the
Indo-Malayan archipelago (1.5 million km2), an arc of
some 17 000 islands lying north and south of the equa-
tor and spanning a distance of 5 000 km between the
Asian mainland and Australia. The hotspot is dominat-
ed by Borneo (725 500 km2) and Sumatra (427 300 km2),
the third and sixth largest islands on Earth, with the re-
mainder mostly occupied by the Malay Peninsula and
the island of Java (126 700 km2). Sundaland sits on the
Sunda Shelf, an extension of the Asian continent where
the sea is shallow, 120 m deep or less, although the sea
level has risen and fallen on several occasions during
the past. As a result, for long periods of time, the islands
of Sundaland were an extension of the Asian continen-
tal landmass, which is why the fauna and flora have
much in common and why the Malay Peninsula is con-
sidered as part of this hotspot.
The hotspot is bordered by three other hotspots. The
boundary between the Sundaland Hotspot and the Indo-
Burma Hotspot to the northwest is here taken as the
Kangar-Pattani Line, which lies near the Thailand-
Malaysia border (Van Steenis 1950; Whitmore 1984).
Wallacea lies immediately to the east of the Sundaland
Hotspot, separated by the famous Wallace’s Line, while
the 7 100 islands of the Philippines Hotspot lie imme-
diately to the northeast.
In political terms, Sundaland covers a small portion
of southern Thailand (provinces of Pattani, Yala, and
Narathiwat) south of the Kangar-Pattani Line; nearly all
of the megadiversity country of Malaysia (most of Penin-
sular Malaysia and the East Malaysian states of Sarawak
and Sabah in northern Borneo); Singapore at the tip of
the Malay Peninsula; all of Brunei; and all of the west-
ern half of the megadiversity country of Indonesia, in-
cluding the Indonesian portion of Borneo, Sumatra,
Java, and Bali, and the many other smaller islands that
are part of western Indonesia (e.g., the Bunguran group
or Natuna Islands). The Nicobar Islands, which are
jurisdictionally controlled by India, form part of this
hotspot (having floristic affinities with this hotspot;
Davis et al. 1995). The Cocos (Keeling) and Christmas
Islands are also included, although both are oceanic is-
lands that are politically part of Australia.
Sundaland shows tremendous variation in vegeta-
tion, with formations differing along bedrock, drainage,
and altitudinal gradients (from sea level to 4 101 m). At
lower altitudes, lowland rainforest occurs, and along
sandy and rocky coasts, stands of beach forest can be
found, while along muddy shores there grow stilt-rooted
and other types of mangrove trees. Behind the man-
groves in coastal Borneo and eastern Sumatra are large
areas of peat swamp forest where the depth of the peat
can exceed 20 m. In natural raised ridges and depres-
sions, alluvial bench forests and freshwater swamps oc-
cur with just a thin layer of peat above the mineral soil,
but these forest formations have been favored for agri-
cultural use and are now rare.
Limestone habitats are widespread, but geographi-
cally isolated; these limestone formations have high
levels of endemism, and exhibit high levels of alpha-
and beta-diversity of their associated flora and fauna.
For example, the slipper orchid (Paphiopedilum sande-
rianum) is a beautiful species with corkscrew petals up
to one meter long, and is known only from two small
populations in the limestone hills of Gunung Mulu Na-
tional Park, Sarawak (Whitten 2002).
On the slopes, one encounters montane forest in
which the limbs of the trees are often swathed with a
thick growth of moss, hanging garlands of lichens, and
many exquisite orchids. Still higher up is the scrubby,
but often colorful subalpine forest dominated by rhodo-
dendrons and their relatives. And, finally, one arrives
at the bare exposed peaks of the higher mountains,
with Mt. Kinabalu (4 101 m) on Borneo being the high-
est of all, and harboring a remarkable concentration of
endemic animals and especially plants.
In terms of vascular plant diversity and endemism,
no truly reliable estimates exist. However, by extrapo-
lating from what is known of the principal islands and
countries that comprise the hotspot, total vascular plant
diversity is estimated at roughly 25 000 species, and the
number of endemics at 15 000. In terms of higher-level
plant endemism, there is just one endemic family, the
Scyphostegiaceae, represented by a single species of
tree, and at least 117 endemic plant genera. Borneo
leads in endemic genera with 59, Sumatra has 17, and
Java none, but the Malay Peninsula has 41. Borneo also
has some 3 000 species of trees, including 267 species of
the economically important Dipterocarpaceae, of which
fully 155 (58%) are endemic to the island; Borneo also
has over 2 000 species of orchids. Sundaland’s flora in-
cludes several spectacular species, most notably Raffle-
sia, the best botanical flagship for the lowland forests,
particularly Rafflesia arnoldii, whose one-meter “petals”
make it the largest flower in the world.
Looking at vertebrate diversity, Sundaland has a total
of 381 mammal species, of which 173 are endemic (27
to Borneo); in addition, 17 of the 136 genera present are
endemic. Many of the mammals occurring in Sunda-
land are globally recognized flagship species, foremost
of which are the two species of orang-utan, the Bornean
(Pongo pygmaeus, EN) and the Sumatran (P. abelii, CR),
both seriously threatened by habitat loss and hunting
(Rijksen and Meijaard 1999). Other famous mammal
flagships include the proboscis monkey (Nasalis larva-
tus, EN), also endemic to the island of Borneo, and the
Sumatran rhinoceros (Dicerorhinus sumatrensis, CR)
and Javan rhinoceros (Rhinoceros sondaicus, CR), the lat-
ter with only some 40-50 individuals left in the world.
Also of special note are the four Mentawai Islands off
the west coast of Sumatra (Siberut, Sipora, North Pagai,
and South Pagai). These small islands, covering only
5 951 km2, are home to fully four endemic species and
seven endemic taxa of primates, including the endem-
ic genus Simias, the pig-tailed langur.
A total of 771 bird species are thought to occur regu-
larly in Sundaland, and no less than 146 are endemic.
There are at least 30 endemic species on Borneo, most
of which are montane species and, as such, the Bornean
Mountains are considered one of five Endemic Bird Ar-
eas (EBAs) recognized by BirdLife International (Stat-
tersfield et al. 1998) for this hotspot, in addition to
Sumatra and Peninsular Malaysia, with 20 species con-
fined to this EBA; Enggano (two species); the Java and
Bali Forests (20 species, including the Bali starling, Leu-
copsar rothschildi; CR); and the Javan Coastal Zone (Ja-
van plover, Charadrius javanicus).
Reptile diversity is estimated at some 449 species in
125 genera, of which 249 species and 24 genera are en-
demic. There are also three endemic reptile families,
Anomochilidae and Xenophidiidae (two families of
snakes, each with a single genus with two species each)
and the monotypic Lanthanotidae, represented by the On the opposite page, the Sumatran
Bornean earless monitor (Lanthanotus borneensis), a orang-utan (Pongo abelii) is now
very rare and little-known lizard. Amphibian diversity recognized as a distinct species,
separate from its cousins on Borneo.
is also impressive, with around 242 species in some 41
This one was photographed in
genera, of which 172 species (including 14 species of Gunung Leuser National Park, one
caecilians) and six genera are endemic. of the last major strongholds of this
New species of fish are still being discovered in the Critically Endangered great ape.
rivers, lakes, and swamps of the area, nearly 200 in © Art Wolfe
the last ten years alone. Currently, around 1 000 spe-
Above, the Critically Endangered
cies are known (and as many as 1 400 projected) from
Sumatran rhinoceros
Sundaland, of which 280 are restricted to one or more (Dicerorhinus sumatrensis),
of the main islands. Yet again, Borneo boasts the most photographed in the Way Kambas
species, about 430, with 164 endemic, while Sumatra National Park. Numbers have
has 270 species, 42 endemic (Kottelat and Whitten declined over 50% due to poaching
1996). The best flagship for the fresh waters of Sunda- over the last 15 years. Fewer than
300 Sumatran rhinoceros survive in
land is the dramatic Asian bony tongue or golden
very small and highly fragmented
arowana (Scleropages formosus, EN), a very highly prized populations in Peninsular
aquarium fish, with prices for large individuals reach- Malaysia, Sabah, and Sumatra.
ing thousands of dollars. © Haroldo Castro
Forest destruction rates as the major threat to the
biodiversity of Sundaland. The vast majority of forest
loss has occurred over the last 30 years or so, mainly as
the result of commercial logging, followed by massive
agricultural projects (e.g., rubber, oil palm, industrial
167
timber for pulp production, etc.) and smallholder agri-
culture. The infamous Indonesian program of transmi-
gration, in which people from Java and neighboring is-
lands were translocated to other less densely inhabited
parts of the country, also had a major impact, as has
civil war in areas such as Nangroe Aceh, Darussalam
Province, which has caused large numbers of refugees
to move into North Sumatra Province.
A forest cover mapping effort conducted in 1999 by
the Indonesian government with support from the
World Bank concluded that the average annual defor-
estation rate between 1985 and 1997 in Indonesia was
about 1.7 million ha, with Sumatra and Kalimantan both
losing more than 20% of their forest cover. Around 35%
of Sumatra and 60% of Kalimantan was considered
forested in 1997, although these regions have experi-
enced continued and accelerating declines. Particularly
hard hit have been the more accessible lowland forests,
with both islands having estimated losses of more than
60% between 1985 and 1997. More recent estimates in-
dicate that Kalimantan’s protected lowland forests de-
clined by 56% between 1985 and 2001 primarily from
logging (Curran et al. 2004), and that less than 33% of
lowland forest and peat swamp remains across all of In-
donesian Borneo (L. Curran, unpubl.), although mon-
tane forest appears intact. Logging has also taken place
within protected areas; for example, forest loss within
Bukit Barisan Selatan National Park between 1985 and
1999 averaged around 2% per year (Kinnaird et al. 2003)
and as much as 9.5% per year in Gunung Palung Na-
tional Park between 1999 and 2002 (Curran et al. 2004).
Some protected areas, such as Kutai National Park, have
been completely deforested through a combination of
the effects of logging, and also drought and fires associ-
ated with the 1982-1983 and later 1997-1998 El Niño
phenomenon (see, for example, Siegert et al. 2001).
At current deforestation rates, lowland forest in
Sumatra will disappear by about 2005 and in Kaliman-
tan soon after 2010 (The World Bank 2001; FWI/GFW
2002; and see Wikramanayake et al. 2002). At the time
of press, Malaysia is considering raising the upper-
altitude ban on logging from 300 to 800 m altitude on
the Peninsula.
Examining the region as a whole, some 700 000 km2
of forest still remains, but highly fragmented, with
many of these patches too small to support the full com-
plement of species over the long term, and most in less
accessible montane regions rather than in the species-
rich lowland forests. For example, a recent study of
mass extinctions in Singapore in relation to habitat loss
(which has exceeded 95% over 183 years) has revealed
that forest reserves comprising only 0.25% of Singa-
pore’s 540 km2 of original vegetation now harbor over
50% of the residual native biodiversity (Brook et al.
2003). Of the remaining forest, only about 100 000 km2
can be considered more or less pristine. This accounts
for only 6.7% of the original extent of this hotspot, with
the bulk of this pristine forest being found in the interi-
or portions of Borneo and in key protected areas in
Sumatra and Peninsular Malaysia, although many of
these comprise montane rather than lowland forests.
Currently, the World Database on Protected Areas re-
veals that protected areas cover about 12% of the land
area of the hotspot. However, throughout Sundaland, the
tendency is for the least threatened, least utilized, and
least valuable habitats (such as upper montane forest) to
be the best protected. Furthermore, protection of parks
and reserves is highly variable, with some such as Taman
Negara in Peninsular Malaysia, Kinabalu in Sabah, and
Ujung Kulon and Gunung Gede Pangrango in Java being
well protected, while others are still subjected to a vari-
ety of pressures including legal and illegal timber ex-
traction. Indeed, less than half of the protected areas
(representing a coverage of 5% of the hotspot) are clas-
sified in IUCN categories I to IV.
Given the pressures, special efforts and new ap-
proaches are needed for long-term and effective conser-
vation in this hotspot. An interesting development on
Borneo is the implementation of transboundary conser-
vation areas or “peace parks.” For example, the Lanjak-
Entimau Wildlife Sanctuary in Sarawak (168 758 ha) is
contiguous with Batang Ai National Park (24 040 ha),
also in Sarawak, and Betung Kerihun National Park
(800 000 ha) in West Kalimantan. This is one of the
largest contiguous areas of forest remaining in Indone-
sian Borneo, but it is under threat of illegal logging,
while hunting for protected species is common. Other
important clusters are Ulu Muda Forest Reserve, Hala-
Bala Wildlife Sanctuary, Belum National Park, and Te-
mengor Forest Reserve between Malaysia and Thailand.
Initiatives are also under way to expand existing in-
country protected area networks; for example, in the
Tesso Nilo/Bukit Tigapuluh landscape in central Suma-
tra, the Critical Ecosystem Partnership Fund is support-
ing acorridor-level initiative to link four existing protect- On pp. 168-169, the world-
ed areas and one new protected area into a 3-million-ha renowned island of Krakatau, off
haven of lowland forest. Such collaborative conservation Sumatra, site of one of the largest
volcanic eruptions in historic times
efforts increase the effective size of protected areas to a
(1883).
great extent, but can only be effective in the long term if © Alain Compost
management strategies deal with the socio-economic
and political drivers of land-use change. On the opposite page, logging of
In Indonesia, a new and encouraging policy initia- intact tropical rainforest continues
tive by the central government aims to reduce the to be a major conservation problem
in Sundaland. This photograph was
number of existing logging forest concessions to half.
taken in Sabah.
Some cancelled or abandoned concessions have been
© Neil Lucas/naturepl.com
converted to protected areas (for example, the Seban-
gau area in Central Kalimantan) while others will be de- Above, the Critically Endangered
signed for sustainable forest harvesting programs. In Sumatran tiger (Panthera tigris
addition, a new policy framework allows for declaration sumatrae) is the last surviving
subspecies of the three historically
of a protected area by local government officials, rather
found in Indonesia, the Javan and
than by a process initiated by the national government. Bali races having become extinct in
This model, being used to create new protected areas the twentieth century.
such as the Batang Gadis National Park in Sumatra, © John Downer/naturepl.com
is particularly important because the impetus comes
from local stakeholders who already support the Park’s
establishment and who feel a sense of ownership in the
process, rather than a protected status imposed by
the national government without consultation.
171
 BirdLife Indonesia and other members of the Bird-
Life International network are pursuing another novel
initiative to acquire the rights to manage a logging con-
cession in central Sumatra for conservation and reha-
bilitation. The approach is an important test of the ef-
fectiveness and cost-efficiency of an alternative to
conventional protected areas, and will produce impor-
tant lessons for management of the large areas of low-
land forest in the hotspot, which are under logging li-
censes that will expire over the next decade.
Although this hotspot, one of the richest on Earth, still
faces many serious problems, recent changes in the re-
gion, and especially in Indonesia, which has most of this
hotspot within its borders, may bode well for the future.
If biodiversity can be identified as a top priority and in-
tegrated into the policies of the government, much could
be accomplished. However, making this possible will re-
quire a far greater effort on the part of everyone, from
civil societies in these countries of Sundaland to the mil-
itary and new political leadership that is taking hold un-
der decentralization, with simultaneous increases in
commitment and involvement from international donor
agencies and conservation organizations alike. An indi-
cation of the changing interests in conservation in the re-
gion comes from Kerinci-Seblat National Park, which
receives more than 200 requests from people interested
in serving as volunteer rangers/interpreters every year
—something that was unheard of only a few decades ago.
WALLACEA
TONY WHITTEN 69 • JATNA SUPRIATNA44
RIA SARYANTHI 70 • PETER WOOD 70
Talaud
Islands
Sangihe
Islands Morotai
Borneo
Halmahera
Sulawesi Sula
Obi
Islands
Buru
Ceram
On the opposite page, the
Endangered Sulawesi crested I N D O N E S I A
macaque (Macaca nigra), formerly
known as the Celebes black ape, is
Sumbawa Flores
one of seven species of macaque
found in the Wallacea Hotspot. EAST TIMOR
© Alain Compost Sumba
Timor
0 500 km
Wallacea is the biogeographical name given to a region
comprising thousands of islands lying between Java,
Bali, and Borneo to the west and Papua to the east.
The hotspot covers the central part of Indonesia and the
whole of Timor Leste. It includes the large island of Su-
172
lawesi (178 700 km2), and the smaller islands of two ex-
tensive archipelagos: the Moluccas (or Maluku in In-
donesian) and the Lesser Sundas (the Indonesian re-
gion of Nusa Tenggara and Timor Leste). In total, the
land area of this region is around 338 494 km2, and it is
one of two hotspots found within Indonesia, the other
being Sundaland to the west.
Wallacea is named after the great nineteenth-century
English explorer-naturalist Alfred Russel Wallace, who
found that the Indonesian Archipelago (then called the
Malay Archipelago) was inhabited by two distinct fau-
nas, one found in the east and one in the west, and pro-
posed a line running between Borneo and Sulawesi and
Bali and Lombok that separated these two faunas (Wal-
lace 1876); this line has long been known as Wallace’s
Line, and the region between it and the island of New
Guinea has come to be called Wallacea (e.g., Dickerson
1928). In certain ways, Wallacea is a transition zone be-
tween the great Indo-Malayan (or Oriental) and Aus-
tralasian biogeographical realms, rather than a distinct
biogeographical entity in itself; nonetheless, it has a fas-
cinating and highly endemic fauna of its own.
In terms of natural vegetation, Sulawesi and the
Moluccas are largely in a tropical rainforest zone, but in
large parts of the Lesser Sundas, rainforest formations
are found only at high elevations and in areas facing
the rain-bearing winds, while significant areas are cov-
ered in savanna woodland, even including some Euca-
lyptus forests. The wetter lowland and hill forests have
the highest number of tree species, but differ from the
commercially valuable forests of Sundaland in having
only a handful of dipterocarp species. The major trees
of commercial value are the tall kauri (Agathis spp.),
the magnificent yellow-flowered legume Pterocarpus in-
dicus in the more seasonal areas, and the gum tree Eu-
calyptus deglupta, usually found in riverine habitats and
PACIFIC used extensively in reforestation projects. In some low-
OCEAN
land areas, such as in eastern Sulawesi, there are unusu-
al and infertile ultrabasic soils with high concentrations
of iron, magnesium, aluminum, and heavy metals. The
New
lowland forests on these nutrient-poor ultrabasic soils
Guinea
have rather short trees, and appear to be dominated by
the myrtle family. Animal life in ultrabasic forests is, in
general, rather limited.
Biological data for Wallacea is patchy. There are few-
er botanical specimens per-unit-area collected there
than on any other major islands in Indonesia, for ex-
ample, and the amphibians are poorly known. For oth-
er groups, species lists for islands may be near com-
plete, but there is very little information on distribution
or the impact of habitat loss except where conservation
organizations (such as the Wildlife Conservation Soci-
ety in Sulawesi) have worked intensively. The Moluc-
cas, in particular, are lacking up-to-date information on
the status of species and habitats.
Total vascular plant diversity in Wallacea is estimated
at 10 000 species, of which perhaps 1 500 species are
endemic; there are at least 500 species endemic to Su-
lawesi alone, 120 to the Lesser Sundas, and 300 to the
Moluccas. There are also at least 12 endemic genera in
the region, seven on Sulawesi, three in the Lesser Sun-
das, and two in the Moluccas.
In terms of vertebrate diversity, Wallacea has a to-
tal of 223 native mammal species, 126 of which are
endemic. If the 124 bat species are excluded, 87 of
the 99 non-flying mammals, or 88%, are endemic. Su-
lawesi has the highest number of mammals (136), of
which 82 species and about one-quarter of the genera
are endemic. The list of Sulawesi endemic mammals
includes flagship species such as the anoa (Bubalus
depressicornis, EN) —a dwarf buffalo— and babirusa
(Babyrousa babyrussa, VU), an enigmatic pig with
long, recurved upper tusks that penetrate through
the skin of the upper lip. In addition, the primates of
Sulawesi are all important flagship species, with at
least seven species of macaques (Macaca spp.) unique
to the island and at least five species of tarsier (Tar-
sius spp.).
There are 650 bird species in Wallacea, of which 265
species are endemic, again very high numbers given
the land area of the region. Of the 235 genera repre-
sented, 26 are endemic, with 16 genera (15 of them
monotypic) restricted to Sulawesi and its satellite is-
lands. Sulawesi has the largest bird fauna, with 356
species, including 96 endemics, among them the ma-
leo (Macrocephalon maleo, EN), a distinctive megapode
currently thought to number between 4 000 and 7 000
breeding pairs. Ten Endemic Bird Areas (identified by
BirdLife International) are found entirely within Wal-
lacea (Stattersfield et al. 1998), and BirdLife Indonesia
has recently identified 112 Important Bird Areas
(IBAs) —priority areas for avian conservation through-
out the region— including 33 on Sulawesi, 36 in the
Moluccas, and 43 in the Lesser Sundas (Rombang et al.
2002, in prep.).
Reptile diversity is also quite high, with 222 species,
99 of which are endemic. These include 118 lizards,
with 60 endemics; 98 snakes, of which 37 are endemic;
five turtles, two of them endemic; and one crocodilian,
the wide-ranging saltwater or Indo-Pacific crocodile
(Crocodylus porosus). There are also three endemic gen-
era (all snakes): Calamorhabdium, with two species;
and Rabdion and Cyclotyphlops, both monotypic. The
best known reptile in Wallacea, and one of Indonesia’s
most famous species, is the Komodo dragon or ora
(Varanus komodoensis, VU), the heaviest lizard in the
world (males can reach about 2.8 m in length and
weigh about 50 kg), known from only the tiny islands of
Komodo, Padar, and Rinca, and the western end of Flo-
res. Amphibians are represented by 58 native species,
all of them frogs; of these, 32 are endemic. The frog
fauna is a fascinating combination of Indo-Malayan
and Australasian elements, with several local radiations
as well.
With freshwater fishes, most of the 310 species
recorded from the rivers and lakes of Wallacea are tol-
erant of both fresh and salt water to some extent.
Around 75 species are endemic. In the Moluccas and
Lesser Sundas, the fish fauna is poorly known, but
there appear to be around six island endemics. On Su-
lawesi, however, there are 69 known species, of which
53 (77%) are endemic. The complex of deep lakes,
rapids, and rivers which makes up the Malili Lakes in
South Sulawesi has at least 15 endemic and quite beau-
tiful telmatherinid fishes, two of them representing en-
demic genera, three endemic Oryzias, two endemic
halfbeaks, and seven endemic gobies, as well as about
50 endemic mollusks, three endemic crabs, and a num-
ber of endemic shrimps.
The invertebrate fauna of Wallacea remains poorly
known, except for groups such as the enormous bird-
wing butterflies (members of the swallowtail butterfly
family). The birdwings are represented by 82 species
in Wallacea, 44 of which are endemic. There are also
109 tiger beetle species recorded from this hotspot, 79
of which are endemic (D. Pearson, pers. comm.). Wal-
lacea also has the world’s largest bee (Chalocodoma
pluto) in the northern Moluccas, a creature in which
the females can grow to four centimeters in length.
This bee is also remarkable because it nests commu-
nally in inhabited termite nests in lowland forest
trees.
As elsewhere, things have changed dramatically in
Wallacea during the course of the past century. The hu-
man population has nearly quadrupled, and the In-
donesian economy has grown tremendously. In the last
decade, one of the world’s newest countries, Timor
Leste, was created in the hotspot, and many parts of
Wallacea have seen political turmoil and dramatic
changes. The first commercial logging operation in Wal-
lacea began in the early part of the century, and forests
have been cleared for agricultural programs, for industri-
al timber plantations, and for land settlement schemes.
Much of the remaining forest is allocated as timber
concessions and other areas are threatened by mining
developments. Furthermore, as has been so obvious On the opposite page, view
with the El Niño-related fires that have raged through of the Gamalama Volcano
much of Indonesia from mid-1997 to the present, fire (Halmahera Island) from the Panau
Lagoon on Ternate Island, North
continues to be a problem —and is now greatly exacer-
Moluccas.
bated by increased drying because of logging and plan- © Gerald Cubitt
tation agriculture, and sometimes by intentional burn-
ing as well (Brown 1998). Invasive alien species are a Above, the Malayan pangolin
threat that is certainly widespread, but too little under- (Manis javanica) is distributed
stood. Hunting and trapping for the pot, and the exotic from Burma and Thailand to Java
and the Philippines. There is a high
pet trade, are widespread. Terrestrial and marine con-
demand for pangolin scales for
servation issues can not be separated, as the livelihoods traditional medicines in many parts
of a huge proportion of the region’s human inhabitants of the world, and particularly so in
come from the sea and are under pressure from over- Southeast Asia, and the meat is also
exploitation and pollution. eaten by indigenous peoples.
As a result of the different human impacts on the © Michael Pitts/naturepl.com
Sulawesi environment, there has been substantial de-
cline in forest cover, although less than that in most of
the other hotspots. What remains is also partly a func-
tion of dryness and altitude. Lowland areas have suf-
fered more than the highlands and, while dry forest
175

A spectral tarsier (Tarsius
spectrum) in a fig tree in the
Tangkoko-Batuangus-Dua Saudara
National Park in northern Sulawesi.
This is one of at least five species
endemic to the Wallacea Hotspot.
© Tui De Roy/Auscape
On the opposite page, knobbed
hornbill (Aceros cassidix), a large
hornbill species endemic to
Sulawesi.
© Tim Laman/National Geographic
Image Collection
176
types in general have only about 10%-20% remaining,
moist and wet forest types have substantially more.
The Lesser Sundas are thought to have only about 7%
forest cover remaining, while Sulawesi is still about
42% covered in original forest (FWI/GFW 2002). Over-
all, about 45% of Wallacea still has some forest cover;
however, if one considers forest that is still in more or
less pristine condition, the percentage drops to only
15%. This loss of forest habitat, particularly in the low-
lands, has caused dramatic and severe declines in the
populations of numerous forest species (many as
much as 90%); as an example, Wallacea holds around
5% of the world’s threatened birds.
At this point in time, forest protection in Wallacea is
inadequate. For the hotspot as a whole, protected area
coverage is around 24 387 km2, or 7% of its original ex-
tent. Around 6% of the protected areas coverage is rep-
resented by reserves in IUCN categories I to IV. As an
example of the poor representation of biodiversity in
protected areas, only 35 of the 112 IBAs that have been
identified are protected. Of course, establishment of
protected areas is only a beginning. Once created, they
need management and the cooperation of local people,
the government, and the private sector in order to be
successful in conserving biodiversity.
Although little known outside the region, Wallacea
does have a number of interesting conservation sto-
ries. One of these is in the 3 000-km2 Bogani Nani
Wartabone (previously Dumoga Bone) National Park
in northern Sulawesi, one of the most important con-
servation areas on the entire island. In the 1980s, the
World Bank helped WWF to encourage establishment
of this park for the purpose of protecting the upper wa-
tershed of the Dumoga River, which was to be used to
irrigate 110 km2 of rice fields. The park had support
from provincial and district officials for many years,
but recently has suffered from large numbers of small-
scale gold miners, who have poisoned the river with
mercury and have cleared forest. Agricultural en-
croachment and illegal logging, hunting, and rattan
collection are also on the increase. Recently, the gov-
ernment of Gorontalo Province on Sulawesi increased
the size of the Paguyaman Forest, the stronghold of the
babirusa.
On Sulawesi, over the last nine years, Conserva-
tion International has been engaged in community-
based conservation in the Togean Islands. The To-
geans occupy the central portion of Tomini Bay,
stretching over a distance of about 90 km. The main
threats are overfishing and destructive fishing, as
well as illegal logging and small plantation develop-
ments. Recently, the local government has declared
its intention to proclaim the Togeans as a marine
park for tourism, covering 4 000 km2 of marine and
terrestrial habitats.
The Wildlife Conservation Society (WCS) carried out
a three-year island-wide biodiversity survey, covering
eighteen forests, to understand the status of key wildlife
species, their habitats, and the factors contributing to
their distribution and abundance. Based on their find-
ings, recommendations for conservation priorities will
be made to governments at the national, provincial,
and regional levels. WCS also assists the government in
managing several protected areas throughout northern
Sulawesi by providing technical assistance through
wildlife monitoring, joint forest patrols, GIS analyses,
helping to establish a formal collaborative management
scheme, monitoring and increasing breeding success of
maleo birds, and through the Wildlife Crimes Units, a
collaborative program established to strengthen conser-
vation law enforcement.
The Nature Conservancy is active in Komodo Nation-
al Park, which covers 1 730 km2, including 1 320 km2 of
coral reefs and sea at the eastern tip of Flores. Conser-
vation activities have included awareness programs, the
formation of a dive club and lodge which incorporates
ecotourism training facilities, reaching agreement on
how to allocate tourism revenues, capacity-building for
communities and the conservation agency, and alterna-
tive livelihood programs to draw people away from
overfishing, fuelwood collection, setting fires to pro-
mote grazing, and dynamiting reefs to catch fish.
On the Sangihe-Talaud Islands, between northern
Sulawesi and the Phillipines, seven endemic bird spe-
cies depend on two forest areas that are shrinking
even though they are protected. BirdLife Indonesia
and the World Bank-GEF have successfully promoted
a process to resolve community-government conflicts
that had prevented progress in forest conservation,
and are now working with all parties concerned.
BirdLife International and BirdLife Indonesia have
worked together on Sumba Island in the Lesser Sun-
das for the past eight years. Two National Parks have
been declared to protect the most important remain-
ing forests on the island, disputes over community
land inside the park have been resolved, and commu-
nities have started to take action to stop illegal logging
and trapping inside the area. The program is working
to formalize the role of local communities in manage-
ment and protection of the park, and to set up a sus-
tainable management system that involves all local
stakeholders. BirdLife Indonesia has also undertaken
surveys and identified priorities for action throughout
the Lesser Sundas and the Moluccas. On Tanimbar Is-
land in the southeast Moluccas, this has been followed
up with a project to help local government and com-
munities plan the management of their still-extensive
forests (which are home to eight endemic bird spe-
cies). On Halmahera, development of conservation
action to protect the critical forests on the island was
suspended with the violence in 1999 and has been re-
vived in 2004.
Although Wallacea is still in relatively good shape
compared to most other hotspots, much needs to be
done to ensure that its large number of endemic spe-
cies and unique ecosystems are maintained.
 PHILIPPINES
LAWRENCE HEANEY 50 • PERRY ONG 53 • ROMEO TRONO 54
LEONARD CO 54 • THOMAS BROOKS 2
0 500 km
Luzon
SOUTH
CHINA SEA
Mindoro
PHILIPPINES
Panay
Palawan
Negros Bohol
Mindanao
BRUNEI Borneo Sulu
Archipelago
MALAYSIA
The Philippines is one of the few nations that is, in its
entirety, both a hotspot and a megadiversity country.
The Philippines is made up of more than 7 100 islands
(of which 700 are inhabited by humans), covers
297 179 km2, and lies entirely in the tropics, between
5ºN and 21ºN. The archipelago stretches over 1 810 km
from north to south, and measures 1 104 km at its
widest point. Northern Luzon is only 241 km from Tai-
wan (with which it shares some floristic affinities),
and the islands off southwestern Palawan are only
40 km from Malaysian Borneo. Palawan, which is sepa-
rated from Borneo by a channel some 145 m deep, has
floristic affinities with both the Philippines and Borneo
in the Sundaland Hotspot, and strong faunal affinities
with the Sunda Shelf (Esselstyn et al., in press), but is
here included within the Philippines Hotspot.
The archipelago is formed from a series of isolated
fragments that have long and complex geological histo-
ries, some dating back 30-50 million years. With at least
17 active volcanoes, these islands are part of the “Ring
of Fire” of the Pacific Basin, extending from Indonesia
to Japan and eastern Russia, and around the western
edge of both American continents.
Until fairly recently, almost the entire area of the
Philippines (94%) was covered by some form of rain-
forest. Trees of the family Dipterocarpaceae were
diverse (at least 45 species), with this dipterocarp-
dominated forest covering most of the islands from
sea level to about 1 000 m. Above these elevations,
lowland forest grades into montane forest, dominated
by oaks, laurels, and ericaceous plants. Above 1 000 m,
the forest undergoes a dramatic change, with a rise in
rainfall (to as high as 12 000 mm per year), cooler tem-
peratures, an increase in steepness of the terrain and,
at the highest elevations, a dense moss cover and a
number of distinct taxa. It is in this habitat that levels
of endemism are the highest, at least proportionately;
all 16 of the bird species endemic to Mindanao are
present in mossy forest on Mt. Kitanglad. The mon-
tane forests (which originally covered only about
10%-12% of the country), have limited economic val-
ue and so have not been as heavily exploited as the
lowland forest; the montane forests now represent
PACIFIC about half of the primary forest remaining in the
OCEAN
Philippines.
The patchwork of isolated islands, the tropical loca-
tion of the country, and the formerly extensive areas of
rainforest have resulted in high species diversity in
Samar
certain groups of organisms and a very high level of
endemism. There are five major and at least five mi-
Leyte
nor centers of endemism, ranging in size from Luzon,
the largest island at 103 000 km2 which, for example,
has at least 31 endemic species of mammals, to Si-
buyan Island (445 km2) with four endemic mammals,
to tiny Camiguin Island, a 265-km2 speck of land north
of Mindanao, which has at least two species of endem-
ic mammals.
There is some uncertainty as to the number of seed
plant species in the Philippines. The “classic” compila-
tion (Merrill 1923-1926) recognized 7 620 indigenous
species in 1 308 genera and 194 families, and 5 832 spe-
cies (76.5%) as endemic. By way of synonymy and the
addition of newly described taxa through the Flora
Malesiana Project (1948-present), a revised estimate is
in the order of around 8 000 species of flowering plants
or angiosperms, 33 species of gymnosperms, and 1 100
species of ferns and allies (see Fernando et al. 2003)
This approximates the estimate of plant diversity made
earlier by Ashton (1997). Endemic species total a min-
imum of 6 091, comprising 5 800 angiosperms, 6 gym-
nosperms, and 285 ferns and allies. No families are
endemic, although certain families (such as the Orchi-
daceae, the largest family of flowering plants in the
hotspot) reach very high levels of species endemism,
and only 26 genera are endemic (22 of them repre-
sented by single species; Fernando et al. 2003), indi-
cating that endemism in this hotspot is mainly at the
species level.
There are at least 167 native terrestrial mammal
species in the Philippines, of which at least 102 are en- On the opposite page, Mala Palao
demic, one of the highest levels of mammal endemism Island in Bacuit Bay, off Palawan
in any hotspot. The largest and most impressive of the in the Philippines.
© Jean-Paul Ferrero/Auscape
mammal species in the Philippines, and indeed a flag-
ship for the hotspot, is the tamaraw (Bubalus mindoren- Above, the elusive Luzon bleeding-
sis, CR), the dwarf water buffalo of Mindoro Island now heart pigeon (Gallicolumba
thought to number only a few hundred. Other flagship luzonica) is endemic to Luzon in
Philippine endemics include the Visayan spotted deer the Philippines, occurring mainly
(Rusa alfredi, EN), now reduced to only a few hundred in lowland forest.
© Konstantin Mikhailov/naturepl.com
individuals on the islands of Masbate, Negros, and Panay
(and thought to be extinct on Cebu and Guimaras); the
golden-capped fruit bat (Acerodon jubatus, EN), proba-
bly the heaviest bat in the world, weighing up to
1.2 kg; and the rodents, of which 15 new species have
been discovered in the last ten years, and including
179

The greater mouse deer (Tragulus
napa) is widespread in Southeast
Asia. This Endangered subspecies
(T. n. nigricans) is found only on
the island of Balabac in the
Philippines.
© Jean-Paul Ferrero/Auscape
On the opposite page,
Idea leuconoe, commonly called
the tree nymph, has an endemic
subspecies (I. l. princesa) on
Palawan, Philippines.
© Haroldo Castro
180
the spectacular adaptive radiation of giant cloud rats.
Importantly, at least one species of mammal thought
to be extinct in the Philippines, the Negros naked-
backed fruit bat (Dobsonia chapmani), has recently
been rediscovered on Cebu and Negros islands. Mam-
mal endemism is also very high at the generic level; of
the 85 genera of mammals in the Philippines, 23 are
endemic.
Bird diversity is moderate at 535 regularly occur-
ring species, but endemism is very high at 185; among
individual islands, Luzon has the largest numbers of
single-island endemics, and is one of seven Endemic
Bird Areas recognized by BirdLife International that
fall within the Philippines Hotspot (Stattersfield et al.
1998). The most famous of all bird species is the Phil-
ippine eagle (Pithecophaga jeffreyi, CR), the second
largest eagle in the world, and which has been se-
verely affected by habitat loss such that it only sur-
vives on Luzon, Mindanao, and Samar, where the
largest tracts of forest remain. Other particularly en-
demic-rich taxa include pigeons, kingfishers, horn-
bills, babblers, sunbirds, and flowerpeckers. The only
endemic family in the Philippines is the Rhabdornith-
idae, represented by the Philippine creepers (Rhab-
dornis spp.).
Reptile diversity is quite high at 235 species, with
some 160 species and six genera endemic; one of these
(Myersophis) has a single species of snake (M. alpestris)
from Luzon. Among these, an important flagship is the
Philippine crocodile (Crocodylus mindorensis, CR), con-
sidered to be the most threatened of all crocodiles, hav-
ing been reduced to only 100 animals in 1993. Amphib-
ians are moderately diverse with 99 species, of which
74 are endemic. These totals are increasing rapidly
with the continuing description of the Philippine her-
petofauna (Brown et al. 2001).
The hotspot has a moderate-sized inland fish fauna,
with 281 native species in 49 families. Although lakes
are usually uncommon on islands, the high level of geo-
logical activity in the Philippines has produced a large
number of lakes, many of which are landlocked. The
development of lacustrine environments is associated
with endemism and, apparently, with the evolution of
several notable species flocks. The hotspot has nine en-
demic genera of inland fishes and at least 67 endemic
species, many of which are restricted to single lakes.
One such species is Sardinella tawilis, a freshwater sar-
dine found only in Taal Lake.
Among invertebrates, insects are the most speciose
group, with a current count of 20 942 species and an
overall endemicity of 69.8%, in 6 185 genera and 499
families (Gapud 2002). There are 132 species of tiger
beetles in the Philippines, of which 113 species occur
nowhere else (D. Pearson, pers. comm.), while butter-
flies are estimated to number some 915 species, of
which 362 are endemic (Treadaway 1995; Danielsen
and Treadaway 2004).
Besides its remarkable endemism, the Philippines,
unfortunately, is also the world leader in terms of
threat, with about 6%-7% of the original old-growth,
closed-canopy forest remaining, and far less than
that, probably on the order of 3%, estimated to re-
main in the lowland regions (Environmental Science
for Social Change 1999). Thus, lowland rainforests
are the most threatened forest type. A further 14% of
the country remains as second-growth forest in vari-
ous stages of degradation, but still capable of regen-
eration if left alone. Unquestionably, the most dam-
aging practice has been the extensive commercial
logging (both legal and illegal) that has taken place in
the past. As late as 1945, as much as 60%-65% of the
Philippines was covered by old-growth forest, but
the rate of logging accelerated quickly after World
War II, with old-growth forest cover dropping to 55%
in 1950, 30% by about 1975, and 20% by 1988. Re-
cently, lumber exports have declined drastically (by
90% in the last 20 years), but this is principally be-
cause there is virtually nothing left to export. However,
the prospect of a major increase in mining is now an
imminent threat. In 1997, regions where mining ap-
plications took place covered over 25% of the land
area of the country, and included over 50% of the re-
maining primary forest.
Currently, around 11% of the total land area of the
Philippines is under some form of protection; however,
when one considers only protected areas in IUCN cate-
gories I to IV, this figure drops to 6%. Indeed, in the late
1980s, the Haribon Foundation, the country’s foremost
conservation NGO, stated that none of the protected ar-
eas currently in existence met international standards
for protection and management. An IUCN report from
1988 estimated that two-thirds of the parks contained
human settlements, and 27% of their cumulative area
was covered by disturbed habitat or agriculture. In
1992, the government established a National Integrated
Protected Areas System (NIPAS) throughout the coun-
try. To assist in the implementation of this law, a U.S.
$20-million grant was provided by the Global Environ-
ment Facility/World Bank for the Conservation of Pri-
ority Protected Areas Project (CPPAP), which focused on
improving management and protection in 10 top prior-
ity protected areas. One of the ten sites selected was
the Palanan Wilderness Area on the island of Luzon. At
3 500 km2, the national park that has been established
there is by far the largest in the country, and covers
perhaps 7% of the remaining primary rainforest of the
Philippines.
A strong focus on creating effective national protect-
ed areas is the best hope for those few remaining ex-
tensive tracts of forest in the Philippines, for example,
in the Northern Sierra Madre mountains and Central
Cordillera of northern Luzon, on Palawan, and in the
Kitanglad Range of Mindanao, and on some smaller,
but endemic-rich islands such as Sibuyan and Camigu-
in. Meanwhile, local and community protected area
mechanisms are desperately needed to conserve the
last fragments of habitat in the endemic-rich, but heav-
ily populated, Visayan Islands (Cebu, Negros, and
Panay), Mindoro, and Tawi-Tawi. In addition, biodiver-
sity research studies, steps to halt all commercial log-
ging in remaining natural forest (and to allow sec-
ondary forest to regenerate into mature native forest),
effective cooperation with local communities, and pro-
grams that prevent subsistence farmers from moving
into forest, whether primary or partially logged, are
crucial.
In 2000, a national conservation priority setting ex-
ercise was undertaken that involved more than 300 nat-
ural and social scientists from more than 100 local and
international institutions representing major stake-
holders spanning academia, government, civil society,
the donor community, and the private sector. This
exercise led to the identification of 206 conservation
priority areas, of which 170 are terrestrial and inland
waters and 36 are marine areas (Ong et al. 2002). Of
these, only 53 priority areas have some form of protec-
tion because of their inclusion under the NIPAS, anoth-
er 57 priority areas are being processed for inclusion
under NIPAS, and 96 priority areas have no legal pro-
tection as they are outside the NIPAS. Of the 53 priori-
ty areas that are under NIPAS, their boundaries need to
be redrawn as most of them include highly degraded ar-
eas and exclude areas of high biodiversity importance
(Ong 2002). The Critical Ecosystem Partnership Fund
(CEPF) is investing $7 million in three key conserva-
tion corridors identified in the priority-setting process,
as well as in key sites for threatened species that fall
outside those corridors. The corridor conservation ef-
fort in the Sierra Madre region on Luzon has led to a sig-
nificant extension of one protected area and the for-
mation of another, both of which will contribute to
connecting the series of protected areas that run from
the island’s northeastern tip to the watershed of Metro
Manila.
In conclusion, since the Philippines has one of the
densest and most rapidly expanding human popula-
tions in Asia, its need for economic and social reform
that will alleviate the causes of poverty is even more
closely tied to biodiversity conservation than is usu-
ally the case. If ever there was a time and place for
immediate and effective conservation action, it is
now in the Philippines. There is still time to stave off
disaster, but rapid action in this unique and troubled
hotspot is needed now. As an offshoot of the 2000
priority-setting exercise and the lessons learned from
conservation initiatives of the 1990s, new trends of
working together emerged, some of which include
the Network for Nature concept that aims to bring to-
gether stakeholders at the local, regional, and nation-
al levels from a wide spectrum of society so as to con-
serve the biodiversity of the hottest of the hotspots.
Indeed, there is still hope to save the Philippines
Hotspot from being the first country in the world to
experience mass species extinction spasms and envi-
ronmental collapse.
SOUTHWEST AUSTRALIA
JOHN S. BEARD 43
AUSTRALIA
INDIAN
OCEAN
0 300 km
The Southwest Australia Hotspot is located in the enor-
mous State of Western Australia, between latitudes 26º
and 36ºS, and longitudes 114º and 126ºE, and original-
ly covered 356 717 km2 in extent. The climate is Medi-
terranean in character, with most of the rain falling in
winter. The total average winter rainfall varies from
over 1 400 mm in the extreme southwest to 300 mm in
the interior, the number of dry months increasing
from 3-4 to 7-8 as rainfall declines. Inland, the bound-
ary of the hotspot closely approximates the 300-mm
isohyet.
As defined here, the Southwest Australia Hotspot
comprises the Southwest Botanical Province of Beard
(1980, 1990; Davis et al. 1995), but excludes the neigh-
boring Southwestern Interzone. The vegetation of the
Southwest Botanical Province has been mapped in de-
tail to show original natural vegetation before the
arrival of Europeans. The vegetation is almost entire-
ly woody, forming forest, woodland, shrubland, and
heath, and there are no grasslands. Among the princi-
pal vegetation types in this region are Eucalyptus
woodlands (formerly covering 25.9% of the Province,
but of which 89% has been cleared), and the Eucalyptus-
dominated “mallee” shrubland (which formerly cov-
ered 22% of the Province, but has now been 50%
cleared). Kwongan is a term adapted from the Aborig-
inal Nyungar language to cover the various Western
Australian types of Mediterranean shrubland, compa- On the opposite page, heathland
rable with the maquis, chaparral, and fynbos of other vegetation at Toolbrunup Peak and
countries with such systems (Pate and Beard 1984). Mt. Hassell in the Sterling Ranges
National Park in southwestern
The principal structural types of kwongan are thicket,
Australia.
scrub-heath, and heath, which together comprise © Marianne F. Porteners/Auscape
about 30% of the original vegetation. These forma-
tions have also been cleared to a large degree: 59% of
the once extensive kwongan heath formations have
been cleared.
Many of the vegetation units in this Province are of
an endemic character. While Eucalyptus-dominated
forests, woodlands, and mallee occur in eastern Aus-
tralia, the dominant species and a majority of the as-
sociated species in the west are endemic. Kwongan is a
183

The Vulnerable numbat
(Myrmecobius fasciatus) is the
only representative of an endemic
family, the Myrmecobiidae, now
restricted to the Southwest Australia
Hotspot. The species is unique in
that it is the only marsupial
adapted to a diet of termites; it lacks
effective teeth and has a long, sticky
tongue used to lick up those insects.
© Jean-Paul Ferrero/Auscape
On the opposite page, giant
eucalyptus trees, including karri
(Eucalyptus diversicolor) and
jarrah (E. marginata),
in Shannon National Park in
southwestern Australia.
© Jaime Plaza van Roon/Auscape
184
formation unique to Western Australia. Although anal-
ogous “heaths” occur in the east, certain structural
forms of kwongan such as the Acacia-Casuarina-
Melaleuca thickets are endemic. The majority of spe-
cies in the kwongan flora are also endemic.
In terms of plant diversity and endemism, the
Southwest Botanical Province has a total of 5 571
species, of which 2 948 (52.9%) species are entirely
confined to this Province; a further 1 462 species
(26.2% of species present) extend slightly beyond its
borders into the Southwestern Interzone or Eremaea
and can still be considered endemic to a more broad-
ly defined concept of southwestern Australia (Beard
et al. 2000). The flora of the Southwest Botanical Prov-
ince comprises 139 families, of which four are en-
demic: the Ecdeiocoleaceae, Cephalotaceae (repre-
sented by the pitcher plant Cephalotus follicularis,
VU), Emblingiaceae, and Eremosynaceae. In addition,
87 (12.5%) of 697 genera are endemic. The ten largest
families (including the Myrtaceae with 785 species, of
which 92% are endemic, and Proteaceae with 684
species, 96% endemic) comprise 61% of the flora,
while the number of species per genus averages eight,
although the ten largest genera (including Acacia with
397 species, 51% endemic, and Eucalyptus with 246
species, 52% endemic) far exceed this figure (Beard et
al. 2000).
Foremost among a wealth of flagship plant species in
this region are the Banksias of the family Proteaceae
(which are 100% endemic), the “blackboys” of the fam-
ily Xanthorrhoeaceae (so named because these plants
with their tall inflorescences reminded early settlers of
aborigines with spears), and the giant Eucalyptus char-
acteristic of the southern portion of this region, includ-
ing the jarrah (E. marginata), the marri (E. calophylla),
and the karri (E. diversicolor), the latter forming a
canopy at about 70 m, with some individuals attaining
80 m or more, ranking this southwestern Australian en-
demic as one of the tallest trees on Earth.
Vertebrate diversity in this hotspot is not nearly as
high as that of plants. Mammals number 57 species, of
which 12 are naturally endemic to this hotspot. Al-
though overall diversity is low, there are some very in-
teresting higher-level endemics. The honey possum
(Tarsipes rostratus), among the few truly nectivorous
mammals, is the only living representative of the fam-
ily Tarsipedidae, and the numbat (Myrmecobius fascia-
tus, VU) is the only member of the marsupial family
Myrmecobiidae. Another appealing flagship marsupi-
al, this time an endemic genus from the kangaroo
family, Macropodidae, is the quokka (Setonix brachyu-
rus, VU), a small wallaby confined to the mainland,
where it has been declining in numbers, and two small
offshore islands (Rottnest Island and Bald Island).
Bird diversity is relatively low, but endemism is a lit-
tle higher than for most of the other Mediterranean-
type systems. Some 285 species are regularly recorded
from this region, and 10 of these are endemic. BirdLife
International also lists Southwest Australia as one of
their Endemic Bird Areas (Stattersfield et al. 1998).
Among the birds, flagship species include the black
swan (Cygnus atratus), the principal state emblem of
Western Australia, and the noisy scrub bird (Atrichornis
clamosus, VU), so called because the male is particular-
ly vociferous.
Reptile diversity is quite high, not surprisingly,
since Australia, as a country, is considered the world
leader in reptile diversity (Mittermeier et al. 1997). A
total of at least 177 species are found in the Southwest
Australia Hotspot, of which 27 are endemic. The most
interesting of these is the western swamp tortoise
(Pseudemydura umbrina, CR), a monotypic genus en-
demic to the region and now found only in one or two
small swamps at Bullsbrook, near Perth. Another en-
demic genus is the short-nosed snake (Elapognathus
minor), which is confined to the humid coastal plains
of the hotspot. Amphibians are somewhat less diverse,
as is usually the case for a dry region, but the 33 spe-
cies include 19 endemics, four of which represent
endemic genera: Myobatrachus gouldii, Metacrinia ni-
chollsi, Arenophryne rotunda, and Spicospina flammo-
caerulea.
The Southwest Australia Hotspot has a very small
amount of freshwater habitat and a correspondingly
small fish fauna with only 20 native species. It is, how-
ever, one of the most distinctive faunas, 10 (50%) of its
species and three of its genera being endemic. Most re-
markable is the salamanderfish (Lepidogalaxias sala-
mandroides), which constitutes the one endemic family
(Lepidogalaxiidae) that is entirely restricted to this
small hotspot, surviving in the harsh conditions of its
ephemeral pools and highly acidic peat habitats. Locat-
ed near the southern limits of the freshwater world, the
hotspot’s fish fauna is dominated by remnants of an-
cient Gondwanan groups including the southern lam-
preys and galaxiids (Allen et al. 2002).
The biggest threat to the unique biota of Southwest
Australia has been land clearing for agriculture, as the
entire Southwest Botanical Province falls within the In-
tensive Land-use Zone. This has left a legacy of sub-
stantial habitat loss and probably species loss as well.
Nearly all the land that can be farmed economically has
been now utilized, and expansion of farms into virgin
areas has stopped. Clearing of remaining vegetation on
existing farms still takes place, though restrictions
have been placed on certain areas for environmental
purposes such as protection of water supplies. It is es-
timated that 30% of the native vegetation in this hot-
spot remains in the Intensive Land-use Zone, which is
more or less synonymous with the Southwest Botani-
cal Province; the Swan Coastal Plain, wheatbelt, and
mallee regions have been largely cleared, and only
patches of original native vegetation remain (Shepherd
et al. 2002).
Currently, the most serious threat is the spread of
root disease (“jarrah dieback”) caused by the root fun-
gus Phytophthora cinnamomi. This disease was first no-
ticed in the jarrah forests around 1940. Unfortunately,
the pathogen responsible was not identified until
1965, by which time thousands of hectares of forest,
both trees and associated flora, were affected. By
1974, it was estimated that 2 820 km2 had been affect-
ed and that the disease was spreading at the rate of
200 km2 per year, which turned out to be an overesti-
mate. The disease can be severe on sites where soil
drainage is obstructed; most of these have now been
impacted so that only isolated trees continue to be at-
tacked by “jarrah dieback.” However, new concern has
arisen as Phytophthora has been found attacking kwon-
gan habitats outside the forests, in particular the Stir-
ling Range National Park, where it has caused mor-
tality among susceptible plants like the grass trees
(Xanthorrhoea spp.) and members of the Proteaceae,
especially Banksias. As indicated above, these are par-
ticularly important flagship species and their loss
would be tragic.
Conservation in Western Australia is vested in the
hands of the State Department of Conservation and
Land Management (CALM), which maintains a series
of national and state protected areas. Some 10.8% of
the hotspot is under some form of official protection, a
figure that remains unchanged when one includes only
reserves classified in IUCN categories I to IV. An im-
portant issue, however, is that of the representative-
ness of existing protected areas. Biodiversity surveys
were never used to determine appropriate sites for pro-
tected areas, meaning that there is an uneven repre-
sentation of different ecosystem types in this region.
Many existing reserves are small, and are isolated
“islands” of natural vegetation within vast areas of
farmlands. Furthermore, many species endemic to the
Southwest Province have very restricted ranges and a
number of rare species are found only on private land,
meaning that the cooperation of landowners has to be
sought to ensure their protection.
All things considered, the Southwest Australia
Hotspot is one of the least threatened of the hotspots,
and has one of the best opportunities to achieve rep-
resentation of all the region’s biodiversity in protected
areas. Maintaining the integrity of the existing pro-
tected area coverage is clearly one step, but much va-
cant land that is not good for either farming or pas-
tureland, and unlikely to be disturbed in the near
future, could also be conserved. And finally, and per-
haps most important, is an increasing desire on the
part of the government to look into the possibility
of creating new reserves in key pieces of privately
owned land in under-represented vegetation types
—in order to increase restricted-range species and
ecosystem coverage in the region’s protected area net-
work. A variety of approaches to creating such private
reserves through tax and other fiscal incentives al-
ready exist, and could very well be put to use here. If
these steps can be taken and succesfully implemented
over the long term, the Southwest Australia Hotspot
could indeed become one the best-protected hotspots
on Earth.
NEW ZEALAND
DAVID R. GIVEN 38 • ALAN SAUNDERS 39 • DAVE TOWNS 40
ALAN TENNYSON 41 • KERI NIELSON 42
Norfolk Kermadec
Islands
Island
Lord
AUSTRALIA Howe
Islands
North
NEW Island
ZEALAND
South
Island
Chatham
Islands
Tasmania
Stewart Bounty
Islands
Island
Antipodes
Auckland Islands
Islands PACIFIC
Campbell OCEAN
Island
Macquarie
Island
0 800 km
New Zealand (Aoteraroa) is the only hotspot that en-
compasses the entire land area of a developed nation. A
piece of Gondwanaland that separated from Australia
about 82 million years ago (Sutherland 1999), it now
forms an isolated archipelago located some 2 000 km to
the southeast of Australia in the southern Pacific Ocean.
It covers 270 197 km2, with the three main islands,
North Island, South Island, and Stewart Island, making
up 90% of the land area. Smaller outliers are the
Chatham Islands (963 km2) 800 km east of South Island,
the Kermadec Islands (34 km2) to the north, and the
Subantarctic Islands to the south (including the Bounty
Islands, Antipodes Islands, Campbell Island, Snares Is-
lands, Auckland Islands, and Macquarie Island). Also
included here are Lord Howe (14.6 km2) and Norfolk
(36.8 km2) islands, both Australian territories.
Climate is an important determinant of biotic patterns
in New Zealand. The country is strongly influenced by
its mid-temperate location, northeast-southwest orienta-
tion, and hilly to mountainous topography, with about
75% of the land area above 200-m altitude and reaching
a maximum altitude of 3 700 m on Aoraki (Mount Cook).
The mountains form a barrier to westerly airflow, result-
ing in 12 000 mm or more annual rainfall on the western On the opposite page,
flank of the Southern Alps, one of the highest on Earth, the New Zealand pigeon
and heavy falls of winter snow. In striking contrast, rain- (Hemiphaga novaeseelandiae)
is a large and impressive forest
shadow areas east of the Southern Alps can experience
pigeon endemic to New Zealand.
as little as 300 mm annually. The Kermadec Islands There are reports that this pigeon,
have a subtropical climate, with warm, moist conditions which may actually represent two
throughout the year, while the Chatham Islands have a distinct species, is declining in parts
cloudy, humid climate, with cool, wet winters and warm, of its range.
often dry summers. The southern Subantarctic Islands © Patricio Robles Gil/Sierra Madre
have a generally windy, cool, temperate-montane cli-
mate and relatively low levels of sunshine.
New Zealand’s forest ecosystems are greatly depleted,
but perhaps the most impressive of those remaining are
the forests of giant New Zealand kauri (Agathis australis),
which are restricted to the far north of New Zealand,
being replaced further south by forests dominated by
187

On pp. 188-189, lush tree-fern forest
on North Island, New Zealand.
© Patricio Robles Gil/Sierra Madre
Above, the Endangered brown kiwi
(Apteryx mantelli) from Westland,
South Island, New Zealand.
This particular form, the Okarito
brown kiwi, may require
recognition as a distinct species.
© Tui De Roy/Minden Pictures
190
angiosperms. In the southern part of North Island and
on South Island, extensive areas of forest are dominated
by Gondwanan gymnosperms of the family Podocar-
paceae, by southern beech (Nothofagus), and by various
combinations of these with broad-leaved angiosperms.
The podocarp and beech-podocarp forests on the west-
ern flanks of the Southern Alps are among the most ex-
tensive temperate rainforests on Earth. Scrub and shrub-
lands are of widespread occurrence in drier parts of
eastern South Island and North Island; they are often
floristically rich, with numerous endemic species. With
its many offshore islands, New Zealand has a diverse
coastal flora with a significant number of coprophilous
plants associated with nesting seabirds, penguin rook-
eries, and sea mammal colonies. Above the timberline,
snow grasses (Chionochloa spp.) often dominate, with
floristically rich alpine herbfields in wetter sites and on
areas of late snow-lie. At higher altitudes, the nival zone
is characterized by cushion plants, many of them en-
demic and including the peculiar and distinctive “veg-
etable sheep” (Raoulia and Hastia spp.), which are high-
ly compacted shrubs of the family Asteraceae.
New Zealand has relatively low plant species diversi-
ty, with 3 400 species (including 2 300 vascular species
and 1 100 liverworts and mosses), but high endemism.
At least 1 865 vascular plants are endemic (81%) and an
additional 220-440 liverworts and mosses (20%-40%),
the exact number for the latter not yet having been de-
termined. In addition, of the estimated 390 plant gen-
era, there are about 35 endemic genera. Some of the
more interesting plant species on New Zealand include
the endemic fern Loxoma cunninghamii, whose closest
living relatives are three species of the genus Loxomop-
sis from Central America, and the pingao sand sedge,
belonging to the endemic monotypic genus Desmos-
choenus. There is also a single endemic family, the Ixer-
baceae, represented by a single species (I. brexiodes).
Vertebrate diversity in New Zealand is low overall, but
again, there is high endemism. The number of regularly
occurring bird species in New Zealand totals 198, of
which 89 are endemic. New Zealand is thought to have
the most diverse seabird community in the world, with
no fewer than 84 species known to breed there; for ex-
ample, it is estimated that at least three-quarters of the
world’s penguins breed in the New Zealand region (D.
Towns, unpubl.). A recent staggering development has
been the rediscovery of the New Zealand storm petrel
(Oceanites maorianus), in waters just off New Zealand’s
North Island. Birds were seen in January and November
of 2003, the first records of this supposedly Extinct
species, previously known only from fossil material and
three nineteenth-century specimens. New Zealand also
has 15 endemic bird genera (of a total of 71) and three en-
demic extant bird families (Acanthisittidae, Callaeidae,
and Apterygidae), a very high number for a country of
this size. Among the surviving avian fauna, the flightless,
nocturnal kiwis, of which there are three species (Apteryx
spp.), are the most famous New Zealand endemics. In ad-
dition, three very large parrots are endemic to New
Zealand, including the kakapo (Strigops habroptilus, CR),
the most unusual of all psittacine birds, and the kea
(Nestor notabilis, VU), a large, inquisitive, long-beaked
mountain parrot restricted to the mountainous areas of
South Island. Indicative of New Zealand’s importance for
bird conservation is the fact that BirdLife International
recognizes five Endemic Bird Areas (EBAs) for this hot-
spot: North Island; South Island; the subantarctic Auck-
land Islands; the Chatham Islands to the east of South
Island; and Norfolk Island (Stattersfield et al. 1998).
Reptiles are represented by 37 species on New Zea-
land, and all native species are endemic and, remark-
ably, five of the six genera represented are endemic. The
largest terrestrial reptile in this hotspot is the tuatara
(Sphenodon spp.), a member of an endemic order (Sphen-
odontida), and the only case in which an entire reptilian
order is endemic to a single country. It was previously
thought that only one species, S. punctatus, existed, but a
second species, S. guntheri, has now been recognized.
These reptiles, superficially resembling iguana lizards,
are the last survivors of a group that lived side by side
with the dinosaurs and whose heyday was the Triassic
Period some 200 million years ago (May 1990).
Amphibians and mammals are the two groups of ter-
restrial vertebrates that are poorly represented. There
are only four native frog species, all highly primitive
and members of an endemic family, Leiopelmatidae, and
genus, Leiopelma, found only on New Zealand. Native
mammals on New Zealand number only two, both of
them endemic bats, but one, the New Zealand lesser
short-tailed bat (Mystacina tuberculata, VU), is the only
living representative of an endemic family, Mystacinidae.
The hotspot harbors one of the smallest but most dis-
tinctive inland fish faunas of any hotspot, with 39 species
in 15 genera, and 25 endemic species. This fauna is dom-
inated by members of the family Galaxiidae, a group of
coolwater trout-like fishes restricted to the southern tips
of South America, Africa, Australia, New Zealand, and a
few small islands such as Lord Howe and the Campbell
Islands. Of the 51 galaxiid species known worldwide, 19
occur in the hotspot and 16 are restricted to it. A related
family, the Retropinnidae or New Zealand smelts, is rep-
resented in the hotspot by three endemic species includ-
ing the only member of the endemic genus Stokellia.
A distinctive element of the New Zealand biota is the
widespread occurrence of gigantism (Daugherty et al.
1993). Some of the giant forms include the now extinct
flightless moas and Haast’s eagle among the birds, and
also giant insects, myriapods, flatworms, land snails,
centipedes, slugs, earthworms, and some plants. The
world’s heaviest insect, the weta or wingless cricket of
Little Barrier Island (Hauturu) weighs up to 70 g and is
one of 12 species of Deinacrida, the ancestors of which
roamed the Jurassic forests.
However, this biodiversity represents only a small per-
centage of what existed prior to human settlement on
these islands. As is the case with many oceanic islands,
humans arrived fairly late on New Zealand, with the
Maoris first arriving perhaps 700-800 years ago (Ministry
for the Environment 1997) and the Europeans in the ear-
ly part of the nineteenth century. Since then, remarkable
species like the giant moas (which reached nearly 3 m in
height) and the immense Haast’s eagle that preyed
on the moas, have gone extinct. Human impact on the
pristine ecosystems of New Zealand can be divided into
three main categories: predation through hunting, fish-
ing, and gathering; habitat destruction through defor-
estation, wetland drainage, and ecosystem degradation of
various kinds; and, particularly, introduction of alien
species, both plant and animal. Since European settle-
ment alone, some 16 land birds, one native bat species,
one fish, at least a dozen invertebrates and 10 plants are
believed to have gone extinct, while other species such
as the tuataras, the stitchbird (Notiomystis cincta, VU),
and the North Island saddleback (Philesturnus caruncula-
tus rufusater) survive only on offshore islands.
Furthermore, prior to the arrival of humans in New
Zealand, indigenous forest covered some 230 000 km2, or
about 85% of the country, with the remainder being na-
tive grasslands, duneland ecosystems, and wetlands. To-
day, the forest has been reduced to 62 000 km2, or about
23% of the country, and only about 35 000 km2 (13%) of
this is still in more or less pristine condition. Grasslands,
on the other hand, have now increased from 10 000-
20 000 km2 to more than 140 000 km2, or 52% of the
country; however, most of this is grazed or overgrown
with introduced grasses, leaving only about 15 000 km2
in more or less pristine condition. Duneland ecosys-
tems, one of the most threatened ecosystem types in the
country, are now down to no more than 250 km2. Wet-
land systems have been especially heavily impacted;
once covering perhaps as much as 10 000 km2, or almost
4% of the country, they have now been substantially re-
duced in extent, with only about 4 000 km2 still remain-
ing in good condition. In terms of natural habitat, then,
it is estimated that remaining indigenous habitat in more
or less primary condition amounts to 35 000 km2 of
forest, 15 000 km2 of native grassland-scrub, 4 000 km2
of wetlands and other aquatic systems, 2 600 km2 of
smaller island ecosystems, 1 800 km2 of alpine systems,
and about 1 000 km2 of coastal systems, for a total of
59 400 km2 (or 22% of the land surface of the country).
The protected area network of the New Zealand Hot-
spot includes 3 345 protected areas in IUCN categories I to
IV, covering around 22% of the hotspot. The additional
protected areas in IUCN categories V and VI bring the to-
tal surface area of the hotspot under a reasonable level of
protection to 27%, a very high percentage by internation-
al standards. Comparing this figure with the 59 400 km2
estimated above to remain in more or less pristine condi-
tion, it is likely that much of what is left intact in New
Zealand is already under some form of protection. In
part, this is because a lot of New Zealand is mountainous,
and areas like the Southern Alps are protected because
the land can’t be used for anything else; lowlands, on the
other hand, are not nearly as well protected. At least 60
protected areas have been set aside as Nature Reserves or
Wildlife Sanctuaries specifically to protect threatened
species. These include many of the offshore and outly-
ing islands ranging from the large subantarctic Auckland
and Campbell Island groups, Little Barrier (Hauturu), and
Kapiti, to the warm temperate Kermadec Islands.
To deal with the incomplete coverage of the country’s
native ecosystems in the protected area network, sever-
al initiatives have been undertaken. For example, one
important initiative has been the subdivision of New
Zealand into a network of Ecological Regions and Eco-
logical Districts, with subsequent development of rapid
survey techniques to assess them. Under the Protected
Natural Areas Program (PNAP), these methods aim to
identify sites of conservation value which are represen-
tative of the study area. The Department of Conservation
(which is the main government agency that administers
protected areas in New Zealand), local government, and
other agencies then begin landowner discussions, with a
view to protection. Many landowners are coming to val-
ue habitat remnants and rare species, and this opens up
an array of possibilities for innovative conservation prac-
tices, where the energy and interest of individuals can be
combined with the resources of the government and oth-
er agencies. Recently, the New Zealand Government has
encouraged development of “mainland islands” that are
intensively managed, predator-free areas where threat-
ened species can be re-established.
Finally, no discussion on New Zealand would be
complete without a word on invasive species, which
have contributed to the decline and extinction of many
native species in this country. While formal habitat pro-
tection is important, active pest management is re-
quired if further extinctions are to be avoided. Conser-
vation practitioners in New Zealand have earned an
international reputation for their achievements in erad-
icating invasive mammals from islands and, more re-
cently, controlling animal and plant pests at “mainland”
sites. Twelve species of pest mammals and one preda-
tory bird have been successfully eradicated from off-
shore and oceanic islands in the New Zealand region
(Veitch and Bell 1990). Significant recent advances have
involved a new capability to eradicate rodents from
much larger islands using aerial bait application tech- The Vulnerable Brother’s Island
niques, and the use of more effective quarantine and tuatara (Sphenodon guntheri) is
contingency procedures to reduce the risks of further one of two species of the order
Rhynchocephalia, the only living
invasions. For example, Norway rats (Rattus norvegicus)
representatives of one of the six
were recently eradicated from Campbell Island (112 km2), major groups of reptiles. Although
opening the way for important species recovery and these unique creatures date back to
ecological restoration objectives. The Department of the time of the dinosaurs, they are
Conservation is applying a strategy to develop capacity now restricted to a handful of small
to eradicate different suites of invasive species from fur- islands in the Cook Strait, which
separates New Zealand’s North and
ther islands and to refine procedures to minimize inva-
South Islands. The New Zealand
sion risks (Cromarty et al. 2002). Important recent Hotspot is the only biodiversity
progress has also been made in controlling invasive hotspot with an endemic order of
species on the New Zealand “mainland” —sites not sur- terrestrial vertebrates.
rounded by water, where terrestrial pest invasion rates © Russell A. Mittermeier
are higher than on remote islands. Better planned and
more consistently supported pest animal and weed con-
trol programs have resulted in significant conservation
outcomes being recorded (Saunders 2000).
191
 NEW CALEDONIA
PORTER P. LOWRY II 26 • JÉRÔME MUNZINGER 33
PHILIPPE BOUCHET 34 • HUBERT GÉRAUX 35 • AARON BAUER 36
OLIVIER LANGRAND 1 • RUSSELL A. MITTERMEIER 1
Chesterfield
Islands Loyalty
Islands
NEW Matthew
CALEDONIA Isle
Island
Hunter
CORAL of Pines
SEA
Island
PACIFIC
OCEAN
0 400 km
New Caledonia is one of the smallest of the hotspots, and
has some of the highest levels of endemism, especially
in plants. It lies at the southern extremity of the Melane-
sian Region, some 1 200 km east of Queensland, Austra-
lia and 1 700 km northeast of New Zealand. Until recent-
ly, New Caledonia was classified as a French Overseas
Territory, but is now in the process of becoming an Over-
seas Country (“Pays d’Outre-Mer”) with substantial polit-
ical autonomy that stops short of full independence.
Unlike the nearby island nations of the East Melane-
sian Islands Hotspot, which are of volcanic origin, the
main island of New Caledonia, Grande Terre (16 595 km2),
was once part of the great ancient continent of Gond-
wana, from which it became separated 65-80 million
years ago. Of more recent origin are the 1 600-km long
reef (second in the world) that encircles Grande Terre
and the raised limestone Loyalty Islands to the east (Ou-
véa, Lifou, Tiga, and Maré, plus a few uninhabited is-
lands). The hotspot includes the Belep Islands to the
north of Grande Terre, and the Isle of Pines, immediate-
ly south. The Chesterfield Islands further to the west, and
the uninhabited volcanic islands of Matthew and Hunter
to the east, are politically dependent on New Caledonia
and also included here, although their value for terrestri-
al biodiversity conservation is limited. The total land area
of New Caledonia, therefore, comes to 18 972 km2.
Despite its small size, New Caledonia is biologically
very diverse and, like the other hotspots that are also
pieces of Gondwana (Madagascar and New Zealand), has
very high levels of endemism, both at species and high-
er levels, especially among plants and invertebrates.
Plant diversity and endemism are truly outstanding in
global terms, with 3 270 vascular plant species, of which
2 432 species (74.4%) are endemic. These are in 808
genera, of which 108 (13.4%) are endemic (updated from
Jaffré et al. 2001). Furthermore, there are five endemic
families —Amborellaceae, Paracryphiaceae, Strasburge-
riaceae, Oncothecaceae, and Phellinaceae (Morat 1993;
Jaffré et al. 2001)—, a truly amazing number for such a
small area and exceeded only by Madagascar and the
Cape Floristic Region of South Africa. Conservative esti-
mates suggest that as much as 5% of the vascular plants
in New Caledonia remain undescribed, which would
bring the total to more than 3 400 species (Lowry 1998).
The flora occurs in four main natural vegetation types,
namely humid evergreen forest, sclerophyllous forest,
low- to mid-altitude maquis and high-altitude maquis for-
mations. Grassland and niaouli (Melaleuca quinquenervia)
savanna today occupy more than 6 000 km2 or 32% of
the area, and are often mistaken by visitors and residents
as the typical landscape of New Caledonia. These are, in
fact, highly disturbed anthropogenic formations that are
maintained by repeated fire and grazing by cattle and in-
troduced deer; the niaouli, an invasive, non-endemic
(and possibly even non-native) eucalypt, has a thick, pa-
pery bark that makes it resistant to fire.
Humid evergreen forest once covered some 70% of
the territory, or more than 13 000 km2, but has now been
reduced to only about 4 000 km2. Around 2 012 plant
species are found in the rainforest, of which 82.2% are
endemic, making it the richest of New Caledonia’s vege-
tation types (Jaffré et al. 1998, 2001). Sclerophyll forest
once covered about 23% of New Caledonia, or more than
4 400 km2; however, it has now been drastically reduced
to just 45 km2 (Bouchet et al. 1995; Jaffré et al. 1998; H.
Géraux, unpubl.), making sclerophyll forest the most
threatened vegetation type in the territory. Even what
little remains is generally very degraded and fragmented
into small patches of 20-30 ha or less (though there are a
few blocks that exceed 100 ha), surrounded by agricul-
tural land. A recent study suggests that New Caledonia’s
sclerophyll forest is the most threatened tropical dry for-
est in the world (Gillespie and Jaffré 2003). Sclerophyll
forest is not as rich as humid evergreen forest or maquis,
but nonetheless contains numerous endemic species;
some 456 plant species have been recorded from this
forest type (57.5% endemic) (Jaffré et al. 1998, 2001).
New Caledonia’s unusual maquis is a specialized
edaphic formation that is now the most extensive vege-
tation type in the territory. High-altitude maquis is very
limited, but still occupies almost all of its original ex-
tent of 100 km2. Some 200 plant species occur in high-
altitude maquis, of which 91% are endemic to New
Caledonia. In contrast, low- to mid-altitude maquis is
now the most extensive natural formation in the coun-
try. It once occupied only about 5% of the country, but
has now expanded, largely as a result of fire distur-
bance, to cover some 4 400 km2, or 23% of New Cale-
donia. Some 1 144 plant species occur in this kind of
maquis (89% endemic) (Jaffré et al. 1998).
Certain plant groups in New Caledonia are particularly On the opposite page, pines
exceptional: of 44 gymnosperm species, 43 are endemic, (Neocallitropsis pancheri) in
including 13 endemic species of Araucaria, an ancient Madeleine Falls Botanic Reserve,
Southern Province, New Caledonia.
group of Gondwanaland gymnosperms, of which there
© Jean-Paul Ferrero/Auscape
are only 19 worldwide (Setoguchi et al. 1998) and the
world’s only parasitic gymnosperm, Parasitaxus ustus.
The territory also has 31 endemic species of palms, rep-
resenting 15 endemic genera out of a total of 16 (Hodel
and Pintaud 1998). Furthermore, New Caledonia is home
to the endemic, monotypic family Amborellaceae, which
comprises a single species, Amborella trichopoda, recently
shown to represent the basal-most branch in the evolu-
tionary tree of the flowering plants (see, for example,
193

The cheerful yellow-bellied robin
(Eopsaltria flaviventris) is a
common primary forest species
endemic to New Caledonia.
© Jean-Paul Ferrero/Ardea
On the opposite page, the kagu
(Rhynochetos jubatus) is the only
representative of a bird family,
the Rhynochetidae, endemic to New
Caledonia. Now considered
Endangered, the species receives
full legal protection.
© Jean-Paul Ferrero/Auscape
194
Mathews and Donoghue 1999; Qui et al. 1999). Given the
uniqueness and age of New Caledonia’s flora, the region
has been classified as a distinct phytogeographic province
(Morat et al. 1984), or placed in its own floristic entity, the
Neocaledonian Subkingdom, by Takhtajan (1986).
In terms of animals, New Caledonia only has nine
native land mammal species, all of them bats, six of
which are endemic to the island, including a new spe-
cies of long-eared bat (Nyctophilus nebulosus) recently
described from Nouméa (Parnaby 2002). Native am-
phibians are entirely absent. However, what the terri-
tory lacks in these groups, it makes up for in birds, liz-
ards, and invertebrates. Twenty-three of the 105 native
species of birds that occur regularly in the New Cale-
donia region are endemic, and New Caledonia is con-
sidered an Endemic Bird Area by BirdLife Internation-
al (Stattersfield et al. 1998). The native species are
grouped in some 65 genera, of which three are endem-
ic. Two are endemic monotypic genera (Rhynochetos
and Drepanoptila), with the kagu (Rhynochetos jubatus,
EN) being the only representative of an endemic fami-
ly, Rhynochetidae; the third endemic genus (Eunym-
phicus) is represented by only two species.
The reptile fauna of New Caledonia is exceptionally
rich at the species level, but only moderately diverse at
the family level: there are a total of 70 terrestrial species
(with at least 42 more as yet undescribed), nearly all of
them lizards in two families of geckos and one family
of skinks; the only native terrestrial snakes are a blind
burrowing species (Ramphotyphlops willeyi) of the fami-
ly Typhlopidae and the Pacific boa (Candoia bibroni),
both restricted to the Loyalty Islands (Bauer and Sadlier
2000; Sadlier et al. 2002). Sixty-two species (or 89% of
the terrestrial reptile fauna) and 11 of the 23 genera are
endemic. Most New Caledonian reptiles exhibit pat-
terns of micro-endemism and are restricted to small geo-
graphic ranges. The lizards include the largest extant
species of gecko in the world, Rhacodactylus leachianus.
In the marine habitats surrounding New Caledonia at
least 14 sea snakes and three sea turtles occur.
The hotspot has a small inland fish fauna consisting of
85 species (only nine endemic) in 49 genera. The fauna is
dominated by euryhaline species that commonly occur
in marine, brackish or freshwater habitats and are gen-
erally widely distributed among islands of the tropical
Western Pacific. The most notable element is perhaps the
endemic galaxiid Galaxias neocaledonicus, the northern-
most representative of a group that is mostly restricted to
the southern tips of New Zealand, Australia, South Amer-
ica, and Africa. A single genus is endemic, Protogobius.
Among the freshwater crabs and shrimps, 10 of the 33 na-
tive species are also endemic (and four more await de-
scription). The lake system on the Plaine des Lacs (Grand
Lac and Lac en Huit) in the south of the main island has
endemic genera of fishes, snails, and crustaceans.
The New Caledonian invertebrate fauna, like that of
most Pacific islands, has many land snails, including an
exceptional number of endemics. There are currently
approximately 200 described native species (although
there may be as many as 400-600 species), all of which
are endemic, with many endemic genera as well. The
genus Placostylus, locally known as bulime, is among
the largest land snails in the South Pacific, and shows
an interesting disjunct distribution with Lord Howe Is-
land, northern New Zealand, and Fiji.
Some 4 000 insect species have been catalogued to
date, and again these show high endemism at the species
and genus levels (Chazeau 1993); with new species con-
stantly being added to the inventory (e.g., Najt and
Grandcolas 2002), the total insect fauna is projected to be
between 8 000 and 20 000 species (J. Chazeau, pers.
comm.). There are more than 70 native species of but-
terflies, of which at least 11 species and two genera are
endemic (Holloway and Peters 1976), along with more
than 300 species of moths (A. Renevier, pers. comm.),
and there are 16 species of tiger beetles, 15 of which are
endemic (D. Pearson, pers. comm.). Among the true spi-
ders, 194 species have so far been recognized (Platnick
1993); New Caledonia has the only family of spiders en-
demic to a single island, the Bradystichidae.
The geological history that resulted in such an unusu-
al array of organisms also created one of the world’s
largest known deposits of nickel. Mining for nickel forms
the foundation of New Caledonia’s economy, the indus-
try representing some 90% of the territory’s foreign ex-
change. The mining is open-cast, and until recently was
largely unregulated, which resulted in an exceptionally
high degree of erosion that has devastated large areas.
Today, open-cast mining, while still destructive, is better
managed and more localized, and more serious threats
to the biodiversity include bush fires, conversion to agri-
culture, and the introduction of alien species.
There are now 772 exotic plant species established in
New Caledonia, many of which compete with the local
flora. In addition, faunal species have been introduced
and are now widespread on the island, constituting a real
threat to the local fauna and flora. These include 24 gas-
tropod species, 400 insect species, six freshwater fish spe-
cies, one amphibian species, two reptile species, 13 bird
species, and 12 mammal species (Gargominy 2003). The
most serious of these introductions, with regard to envi-
ronmental impact, are probably the Indonesian deer (Cer-
vus timorensis) and the fire ant (Wasmannia auropunctata).
Hunting (New Caledonian imperial pigeons, flying foxes)
and selective illegal collecting of animals (reptiles, parrots)
and plants (palms, orchids) represent additional threats.
Logging is less of a threat today than in times past, but
about 3%-5% of the area of humid evergreen forest was
cleared in the 1970s and 1980s (Morat et al. 1995) and sev-
eral new logging concessions have recently been awarded
in the Northern Province. The remaining humid forest is
mainly on public land, whereas for sclerophyllous forest,
44% is in private holdings and 2% in indigenous tribal ar-
eas, with only 54% on public land. Remaining terrestrial
habitat in more or less pristine condition includes about
3 900 km2 of humid evergreen forest, only 45 km2 of scle-
rophyll forest, ca. 100 km2 of high-altitude maquis (nearly
its full original extent), and about 1 000 km2 of pristine to
lightly impacted low- to mid-altitude maquis, for a total
of ca. 5 050 km2, or 27% of New Caledonia’s land area.
At present, the World Database on Protected Areas
includes for this hotspot a protected area network cov-
ering 22% of its land area, although, remarkably, the 24
reserves classed in IUCN categories I to IV cover only
3% of the hotspot. Regional information suggests that
officially, New Caledonia has a protected area network
of 46 reserves covering 1 070 km2 (5.6% of the land
area) in IUCN categories I to IV, and an additional two
in categories V and VI, as well as three parks now being
established in the Southern Province that will include a
further 570 km2. However, a review of the effectiveness
of these reserves in protecting New Caledonia’s amaz-
ing plant diversity (Jaffré et al. 1998), in particular the
25% of endemic plants that are considered to be at risk,
indicated that 83% of the threatened species did not oc-
cur in any protected areas and that only 11% had their
conservation status improved by these reserves (most
of which lack staff and an adequate management plan).
Procuring the necessary resources to carry out biodi-
versity conservation is hampered by the fact that multi-
lateral and bilateral funding agencies will not support
programs in New Caledonia because it is a political de-
pendency of a wealthy country. While the French Gov-
ernment in Paris is solely responsible for international
relations (including conventions and treaties such as
the Convention on Biological Diversity), natural re-
source management lies exclusively with the provincial
governments. Because of this divided responsibility, the
French Government has not regarded New Caledonia as
a high priority for action, although it is currently devel-
oping a National Biodiversity Strategy and Action Plan
in which New Caledonia is considered as a top priority.
The local governments have focused their attention pri-
marily on environmental policy, with some efforts be-
ing made recently in the area of biodiversity protection.
Aside from local government initiatives, a small num-
ber of New Caledonian residents have recognized the
conservation imperative. The Nouméa-based ASNNC (As-
sociation pour la Sauvegarde de la Nature Neo-Calédoni-
enne, or Association for the Protection of New Caledon-
ian Nature), the oldest group in the territory, has spent
three decades promoting conservation in New Caledonia,
with particular emphasis on environmental education
and studies of flagship species such as the kagu and sea
turtles. More recently, international NGOs have also be-
come involved in New Caledonia. WWF has worked since
1997 with local partners to implement a sclerophyllous
forest conservation program. Conservation International
has supported the Maruia Society’s effort to implement
local conservation in the Mt. Panié massif, and BirdLife
International is starting a two-year IBA assessment. Clear-
ly, as is the case for several other hotspots, an immediate
priority is the establishment of additional, carefully se-
lected protected areas to provide adequate coverage for
New Caledonia’s unique biodiversity, and thereby pre-
vent not only the loss of individual species, but also of the
island’s numerous endemic genera and families.
POLYNESIA-MICRONESIA
ALLEN ALLISON 87 • LUCIUS G. ELDREDGE 87
0 2000 km U.S.A.
Northern MEXICO
Mariana
Hawaii
Guam Islands
PALAU MICRONESIA
MARSHALL
ISLANDS
New
KIRIBATI
PACIFIC
Guinea NAURU OCEAN
TUVALU
French
SAMOA Polynesia
FIJI
Cook
TONGA Islands Easter
Island
AUSTRALIA NEW
ZEALAND
The Polynesia-Micronesia Hotspot consists of all of Mi-
cronesia and tropical Polynesia, and also Fiji, which is
generally considered a transitional area between Mela-
nesia and Polynesia. Furthermore, we include the Ha-
waiian Islands within this hotspot, although its flora is so
distinct, with 87% of 1 170 total native vascular plant
species endemic (updated from Eldredge and Evenhuis
2003), that it is usually treated as a separate floristic
kingdom (Takhtadzhëiìan and Cronquist 1986). Although
it stretches over an enormous area of some 21.6 million
km2 of Pacific Ocean (2.6 times the size of the continen-
tal United States), it is in fact one of the smallest hotspots
in terms of land area (roughly 47 000 km2 in size).
Stretching from the Mariana and Palau archipelagos
in the west to Easter Island (Rapa Nui) in the east, and
from the Hawaiian Islands in the north to the Cook
Isands, Tonga, and Niue in the south, the Polynesia-
Micronesia Hotspot includes at least 1 415 islands be-
longing to 11 countries, eight territories, and one Amer-
ican state. In Polynesia, the independent nations are
the Cook Islands (237 km2), Niue (259 km2), Samoa
(2 935 km2), Tonga (649 km2), and Tuvalu (26 km2), and
Fiji (18 333 km2) and, in Micronesia, the Federated
States of Micronesia (701 km2), Kiribati (811 km2), the
Marshall Islands (181 km2), Nauru (21 km2), and Palau
(488 km2). The territories in Polynesia include the
French territories of Wallis and Futuna (274 km2) and
French Polynesia (3 521 km2), the American territory of
American Samoa (200 km2), the British territory of Pit-
cairn Islands (47 km2), the New Zealand territory of On the opposite page, a heavily
Tokelau (10 km2), and the Chilean territory of Easter Is- forested canyon in the foothills of
land (166 km2). In Micronesia, the American territories Samoa, Upolu Island.
© Patricio Robles Gil/Sierra Madre
are Guam (541 km2) and the Northern Mariana Islands
(477 km2). Finally, the one American state in the re-
gion is Hawaii (16 642 km2) (SPC 2003).
The islands in the region differ widely in geological
composition, ranging from ancient highly weathered
rocky islets, low-lying coral atolls, and uplifted limestone
islands (makatea) to much larger, high volcanic islands
such as those found in Hawaii, Fiji, and Micronesia that
support most of the human population in the region
(Stoddart 1992). Ongoing volcanic activity is found in
Hawaii, Tonga, Samoa, and the Northern Mariana Islands.
197

The Vulnerable masked shining-
parrot (Prosopeia personata) now
survives only on the island of Viti
Levu in Fiji. The total population of
this species is thought to be around
5 000 birds, but may be declining
mainly due to habitat loss.
© Patricio Robles Gil/Sierra Madre
On the opposite page, the crimson
shining-parrot (Prosopeia
splendens) is endemic to Fiji and
Tonga in the Polynesia-Micronesia
Hotspot. The species may be found
singly, in pairs or in groups of up
to 40 individuals outside the
breeding season.
© Patricio Robles Gil/Sierra Madre
198
This region is biologically very diverse, with Pacific
Island plants occurring in 12 principal vegetation asso-
ciations or biomes (Mueller-Dombois and Fosberg
1998). The most widespread association is strand vege-
tation, consisting of salt-tolerant plants found along the
shores of most Pacific Islands. Inland areas with high
rainfall support various types of mesic forest or rain-
forest, generally typified by high stature trees at low el-
evations, and lower stature trees and a higher percent-
age of shrubs and epiphytes at higher elevations. Areas
at the cloud line are dominated by cloud forest, while
areas above the cloud line —which in this region are
limited to mountain summits on Maui and the island of
Hawaii— are covered in an open woodland association
of grassland and xerophytic shrubs. In addition, there
are some 515 wetlands scattered throughout the region,
with 386 in the Hawaiian Islands alone and 129 in the
remainder of the region. Included among the wetlands
are some 500 km2 of mangrove forests, composed of 22
species, declining in diversity from west to east across
the Pacific, reaching a limit in American Samoa, where
only a single mangrove species is present.
In general, there is a high degree of endemism
among both vascular plants and certain animal groups.
The actual number of native vascular plants (ferns and
flowering plants) in the hotspot is unknown. Davis et
al. (1997) listed 3 074 species of endemic vascular
plants. The overall rate of vascular plant endemism in
the hotspot is about 58%; based on this, we estimate
that the hotspot flora includes about 5 330 native spe-
cies of vascular plants; these are represented by 159
families and 802 genera, of which 63 genera are en-
demic to the hotspot (Van Balgooy 1993). One of the
families, the Degeneriaceae, is endemic to the hotspot
and is represented by a single species, Degeneria vitien-
sis, which is endemic to Fiji.
In terms of animal diversity, the terrestrial animal
associations of Polynesia-Micronesia include rich ar-
rays of birds, reptiles, land snails, and insects, all groups
adept at crossing gaps of open sea. Birds are the domi-
nant terrestrial vertebrates in the Pacific, and many
groups have radiated extensively throughout the re-
gion. There are around 300 regularly occurring species
in the region, of which 242 regularly breed within the
hotspot and 170 are endemic (an additional 25 species
that were endemic went extinct since the arrival of Eu-
ropeans) (updated from Pratt et al. 1987). In general,
the number of endemic species increases with the geo-
graphic isolation and topographic diversity of the dif-
ferent archipelagos, with most of the endemic species
being forest birds found in the larger and higher is-
lands. Generic endemism is also high, with 28 of 111
genera endemic. Not surprisingly then, no less than 15
Endemic Bird Areas have been identified within the
Polynesia-Micronesia Hotspot (Stattersfield et al. 1998).
One of the best-known bird groups, and one of world’s
most spectacular cases of adaptive radiation and speci-
ation in an isolated island ecosystem, involves the en-
demic honeycreepers (Drepanididae). Believed to have
originated from a single ancestor from North America,
the family has evolved into 34 extant and recently ex-
tinct species (in 18 genera), although numerous fossil
taxa have also been described (Olson and James 1991;
James and Olson 1991).
Among reptiles, there are 61 terrestrial species found
in Polynesia-Micronesia, including seven species of ter-
restrial snakes, the saltwater crocodile (Crocodylus poro-
sus), and 53 species of lizards, mostly skinks and geck-
os, but also including two iguanas, the Fiji banded
iguana (Brachylophus fasciatus, EN) and Fiji crested
iguana (B. vitiensis, CR), which are endemic to the Fiji-
Tonga area. In addition, there are two land-breeding
species of marine snakes (Laticauda spp.) and six spe-
cies of marine turtles found within the hotspot (updat-
ed from Allison 1996). Thirty-one reptile species are
endemic. This herpetofauna is almost entirely Indo-
Pacific in origin, except for the iguanas, whose closest
living ancestors are in the Americas (Gibbons 1981).
Two reptile genera are endemic to the hotspot: the
genus Brachylophus (two species; see above), and the ge-
nus Ogmodon, represented by a single species, the Fiji
snake (O. vitianus, VU) which, as its name implies, oc-
curs only on Fiji. A third endemic genus, Tachygia,
which included a single skink species formerly re-
stricted to Tonga, is extinct. It is likely that there are no
native terrestrial species found east of the Cook Islands
(Crombie and Steadman 1986).
Other vertebrate groups are characterized by low di-
versity but high endemism. The native terrestrial mam-
mal fauna, for example, consists entirely of bats. Fifteen
living native species of bats are known from this hotspot
(including Myotis insularum, which may not represent a
valid record for the area), and 11 are endemic, most re-
stricted to the high islands (Mariana Islands, Palau,
Chuuk (= Truk), Kosrae, Pohnpei, Yap, Samoa, Fiji, and
the Hawaiian Islands) (Tomich 1986; Flannery 1995).
An endemic bat, the Fijian monkey-faced bat (Pteralopex
acrodonta, CR), is one of the most primitive species of
fruit bat known and is the only mammal endemic to
Fiji. Similarly, amphibian diversity in this hotspot is
also low; only three native amphibians are known to oc-
cur, all of them ranid frogs of the genus Platymantis: the
Palau frog (P. pelewensis), endemic to Palau; and the Fiji
ground frog (P. vitiana, EN) and Fiji tree frog (P. vitien-
sis), both endemic to Fiji (Allison 1996).
There are no truly freshwater fishes in Polynesia and
Micronesia other than some 86 introduced species, such
as tilapia, mosquitofish, and others (Eldredge 2000), but
at least 96 native species are found as adults in freshwa-
ter; 20 species are endemic. Gobioid fishes (Gobiidae
and Eleotridae) comprise the largest element of the
freshwater fish fauna, especially members of the genera
Lentipes, Sicyopterus, Sicyopus, Stiphodon, and Stenogo-
bius. Information on their life history is limited, but it
appears that they all have a pelagic marine larval stage.
Most have a very limited geographic range, and there
are many island or island-group endemics. Hawaii, rep-
resenting the northernmost island group in Polynesia,
has five species of gobioid fishes, whereas Guam in
western Micronesia has four times that of Hawaii
(Fitzsimmons et al. 2002). Four of the five species in
Hawaii are considered endemic (McDowall 2003).
Invertebrate diversity in the region is high for certain
groups. Particularly impressive are the land snails, a con-
spicuous feature of Pacific Island ecosystems. Of the 13
major indigenous pulmonate land snail families on the
Pacific Islands, four are endemic to the central Pacific
(Cowie 1996). The Hawaiian Islands have a total of 763
native species, of which a staggering 748 are endemic
(Cowie et al. 1995); the Samoan Islands have 99 native
species, of which 64 are endemic (Cowie and Robinson
2003). The land snails of the subfamily Achatinellinae
are considered by many to be the most remarkable of all
land snail faunas. Another example of a spectacular
adaptive radiation, these brilliantly colored little jewels
of the Hawaiian forests reach perhaps as many as 2 000
individuals per tree in some areas (Hadfield 1986).
Unfortunately, the biota of the Polynesia-Micronesia
Hotspot is under intense pressure from human-inuced
disturbance and habitat loss, alien species introduc-
tions, and other factors, and its plants and animals are
among the most highly threatened in the world. For ex-
ample, in Hawaii, approximately 189 plant taxa exist in
the wild as multiple populations with fewer than 50 in-
dividuals. At least 14 species of Hawaiian endemic
plants are down to only one individual in the wild; an
additional 46 are down to two to 10 individuals in the
wild; and more than half of the native flora of 1 170
species (which is now exceeded in number by alien spe-
cies) has fewer than 5 000 individuals each in wild pop-
ulations (updated from Allison and Miller 2000).
Alien species, an especially insidious problem, have
contributed to many of the extinctions in the region, in-
cluding 50% of the achatinelline land snail species (or
50 out of a total of 99 species). Due to alien species and
other threats, only 8-10 of these snail species are now
considered extant (Holland and Hadfield 2002) and per-
haps as many as 90% of the total Hawaiian land snails
are extinct (Lydeard et al., in press). Alien plants are
also a significant cause of degradation of Pacific ecosys-
tems. Miconia calvescens, for example, is a small, fast-
growing South American tree that has taken over on
Tahiti, where it has crowded out native species and cov-
ers at least 700 km2, or about 65% of the island (Meyer
and Florence 1997), while other aggressive plants, such
as gorse (Ulex europaeus), banana poka (Passiflora mol-
lissima), and other species, are estimated to have dam-
aged at least 57% of the original forest of the Hawaiian
Islands. In the Society Islands, the introduction of a
predatory snail for biological control resulted in the
loss of 57 of that archipelago’s 61 endemic partulid land
snails (Coote and Loève 2003).
As a result of all of these factors, and also the im-
pact of guano mining and other land disturbances,
more than three-quarters of the original vegetation of
Polynesia-Micronesia has been destroyed or substan-
tially modified, with an estimated 10 000 km2 or 21%
of the original vegetation remaining in more or less
pristine condition for the entire hotspot.
The Polynesia-Micronesia Hotspot has a total of 259
terrestrial protected areas and community-based con-
servation areas that cover a little more than 5 000 km2,
or just under 11% of all land within the region (SPREP
2003). However, Hawaii has 105 (41%) of these protect-
ed areas, or a little more than 58% of the overall area
for the region as a whole. The actual level of protection
that the various parks and reserves receive varies wide-
ly within the region, from those that exist only on pa-
per to the intensely managed national parks of Hawaii.
Protected areas coverage is lower according to the
World Database on Protected Areas (5.2%).
In terms of strategies, conservation activities will be
most effective if undertaken on a regional basis through
cooperation among nations, as many of the Pacific Is-
land nations are simply too small and have too few re-
sources to undertake this work individually. Further-
more, many of the problems are regional in nature. The
twenty nations, territories, and states in Polynesia-
Mironesia have a long history of cooperation in trade
and development through the Secretariat of the Pacific
Community (SPC) which, in 1978, established the South
Pacific Regional Environment Programme (SPREP) to
deal specifically with environmental issues. Equipped
with a five-year, $10-million grant from the Global En-
vironment Facility (GEF), SPREP was able to form the
South Pacific Biodiversity Conservation Programme
(SPBCP), and has used this funding to “establish and ini-
tially manage a series of large, diverse Conservation Ar-
eas (CAs) in which human activities will be guided to
protect ecological function” (Reti 1993).
Much of the biota of the Polynesia-Micronesia Hot-
spot remains unknown, and there is an urgent need for
a comprehensive biological survey to guide and inform
the development of conservation priorities. Even in
Hawaii, which is the best-known archipelago in the
hotspot, more than a third of the biota —at least 7 000
species— have not yet been scientifically named (Alli-
son 2003). Bishop Museum and a host of partners, espe-
cially the U.S. Geological Survey (USGS), are addressing On the opposite page,
this issue in Hawaii through the Hawaii Biological Sur- on Upolu Island, Samoa, the
vey, and are working with the Pacific Science Associa- community of Eufato has declared
a reserve within its lands so as to
tion and governments and NGOs throughout the region
conserve them in the future.
to develop a Pacific Biological Survey. The Pacific Sci- © Patricio Robles Gil/Sierra Madre
ence Association is also working with the USGS, the
Global Biodiversity Information Facility (GBIF), SPREP, Above, a family in Eufato village
and others to develop a Pacific Basin Information Facil- in Samoa.
ity (PBIF) to facilitate Web access to a wide array of cur- © Patricio Robles Gil/Sierra Madre
rently inaccessible biodiversity information.
On pp. 202-203, two young boys
During the past several thousand years, the Pacific with a harvest of coconuts on
has arguably lost more species to extinction than any Upolu Island, in Samoa.
other region on Earth (Steadman 1995). Coordinated re- © Patricio Robles Gil/Sierra Madre
gional efforts that are to being developed to share in-
formation and address common threats, such as inva-
sive alien species, are showing great promise and offer
the best hope for preserving what remains of the extra-
ordinary biota of the Polynesia-Micronesia Hotspot.
201
 MADREAN PINE-OAK
WOODLANDS
The Madrean Pine-Oak Woodlands Hotspot as de-
fined here includes the main mountain chains of
Mexico, namely the Sierra Madre Occidental (in-
cluding the Madrean Sky Islands of southern Ari- U.S.A.
zona and New Mexico), the Sierra Madre Oriental,
the Trans-Mexican Volcanic Belt, the Sierra Madre
del Sur, and the Sierra Norte de Oaxaca, as well as iso-
lated mountaintop islands in Baja California (par-
ticularly around the Sierra de la Laguna). Excluded
are the more mesic pine-oak formations of the Altos
and Sierra Madre de Chiapas, lying to the east of the
Isthmus of Tehuantepec, which are included within GULF
OF
the Mesoamerica Hotspot. Thus, most of this hotspot MEXICO MEXICO

lies within Mexico, with isolated patches occurring

further north, in the southern United States. Many of
the pine-oak woodlands have an insular-type distri- PACIFIC
OCEAN
bution by virtue of being surrounded by more exten-
sive floristic provinces, generally tropical or arid.
This feature is particularly noticeable in the north-
ern Mexican Highlands and the Madrean Sky Island
Archipelago. In total, the original extent of this
hotspot covered 461 265 km2. 0 500 km

The Madrean Pine-Oak Woodlands Hotspot con-

sists of rugged mountainous terrain, high relief, and
deep canyons, evidence of its complex geological
history. The Sierra Madre Occidental, which runs
from Jalisco north through Sonora and Chihuahua
to the Madrean Sky Islands, is mostly composed of
volcanic rocks from the Cretaceous and early Ter-
tiary, covered by other Early-Middle Tertiary materi-
als. Its origin seems to be related to the subduction
of the Farallón Plate beneath the North American
and Caribbean Plates. The Madrean Sky Islands, a
series of about 40 mountain tops, lie between the
Mogollon Rim and the Sierra Madre Occidental.
About half of the mountains occur in the United
States, including the tallest sky island, Mt. Graham
(at 3 300 m), while the remainder occur in northern
Mexico. The Sierra Madre Oriental of eastern Mexico
and southwestern Texas is composed of sedimentary
marine rocks from the Jurassic and Cretaceous that
folded up into complex shapes. The tallest peaks are
Potosí (3 625 m) and Peña Nevada (3 480 m).
The Trans-Mexican Volcanic Belt, which runs
from west to east across central Mexico and serves as
a bridge connecting the Sierra Madre Occidental and
Sierra Madre Oriental, is the tallest mountain chain
in the country, including the peaks of Pico de Oriza-
ba (5 747 m) and the Popocatépetl (5 452 m). As its
name suggests, this chain is volcanic in origin, and
is composed primarily of igneous rock. The Sierra
Madre del Sur, which runs parallel to the Pacific
Coast of Guerrero, is the most geologically complex,
with pre-Cambrian units, volcanoes from the Meso-
zoic, and formations from the Middle Tertiary relat-
ed to the Sierra Madre Occidental. Finally, there is
the Sierra Norte de Oaxaca in northern Oaxaca State, On the opposite page, understory
with the tallest peak being Zempoaltépetl (3 400 m). of pine forest with Agave
Most pine-oak forests in Mexico are found in areas montana and Nolina sp.,
Miquihuana, in the Sierra Madre
with a temperate climate, characterized by average
Oriental.
annual temperatures between 5ºC and 18ºC. Annual © Jack Dykinga
precipitation of temperate zones in Mexico is typi-
cally between 500 mm and 2 500 mm, with lows of
300 mm (P. cembroides forests) and highs of 3 000 mm
(González-Medrano 2004). Humidity varies accord-
ing to altitude and slope. In the Sierra Madre Orien-
205

Montezuma quail (Cyrtonyx
montezumae) in the El Carmen
Mountains of the Sierra Madre
Oriental in Mexico.
© Patricio Robles Gil/Sierra Madre
On the opposite page, pinyon pine
forest in the Chisos Mountains, Big
Bend National Park, Texas.
© Patricio Robles Gil/Sierra Madre
206
tal, for example, the climate is more humid on the
northeastern slope, and subhumid on the western slope
and in the highest portions of the mountain range.
However, the wettest portions are home to cloud forests
that represent the northernmost reaches of this vegeta-
tion in Mexico. Likewise, in the Sierra Madre Occiden-
tal, the western sides generally receive more rainfall
and have a milder winter, although as a whole, the en-
tire mountain range is considered to have mild winters
and wet summers (Brown 1994), with mean annual
rainfall around 550 mm (mostly falling in August). In
the Sierra Madre Oriental, average annual rainfall
ranges from 250-300 mm in the north and from 900-
1 500 mm in the southern parts near Nuevo León,
Mexico. In the Sierra Madre del Sur, annual precipita-
tion is between 800 and 1 600 mm (Rzedowski 1978).
Snow is common on the tallest peaks of mountain
ranges, particularly in the Trans-Mexican Volcanic belt.
As a result of the different climates, aspect and orien-
tation of slopes, type and depth of soil layers, and the
complex geological history of the region, the hotspot
features distinct biotic communities. These charac-
teristics, along with its significant north-south latitudinal
reach, make it difficult to treat any one portion of the
hotspot as representative of the whole. There are, how-
ever, clear affinities of flora and fauna among its subre-
gions. For example, half of the characteristic species in
the Madrean flora found in northern Mexico and the
southern United States extend as far south as Durango,
and one-fifth of these typical species range into southern
Mexico (McLaughlin 1994). These observations, along
with the consistent canopy habitat provided by pines
and oaks, attest to the cohesiveness of the hotspot.
As the name of this hotspot denotes, pine and oak
forests are the characteristic vegetation type. At one ex-
treme, these forests may contain monospecific stands
of either pines (Pinus sp.) or firs (Abies sp.); at the oth-
er extreme, they are represented by almost pure stands
of oak (Quercus sp.). Mexico is considered an important
center of diversity for both Pinus and Quercus, and has
the highest diversity of pine species of any country
in the world, with 44 of the 110 recognized species of
pine occurring there (Farjon and Page 1999). The coun-
try also has 135-150 species of oak, representing more
than 30% of the world’s species of the genus Quercus;
of these, 86 are endemic to Mexico (Nixon 1993).
Across the hotspot, different regions have varying
combinations of species, with some more dominant
than others. On the Sierra Madre Occidental, pine-oak
forests grow on elevations between approximately
1 500 and 3 300 m. The dominant species on the east
slopes is Pinus lumholtzii (15-20 m tall, growing at 1 900-
2 400 m), while in the Madrean Sky Islands the domi-
nant species are P. leiophylla (15-25 m), P. cembroides (6-
12 m), P. ponderosa (23-30 m), Quercus hypoleucoides, Q.
arizonica, Q. emoryi, and Q. rugosa (Valero et al. 2001a).
The Sierra Madre Oriental is dominated by pine-oak
forests growing at altitudes between 1 000 and 3 500 m,
particularly Pinus nelsonii (6-7 m), P. cembroides, P. pseu-
dostrobus (15-25 m), P. ponderosa, and P. greggii (10-25 m,
with a restricted distribution at 1 200-2 700 m), as well
as the oaks Q. castanea and Q. affinis (Valero et al.
2001b). In the Trans-Volcanic Belt, pine forests grow at
elevations of 2 275-2 600 m, pine-oak at 2 470-2 600 m,
and pine-cedar at more than 2 700 m. The dominant spe-
cies in the western reaches include Q. resinosa, Q. mag-
nolifolia, Q. conspersa, and Q. peduncularis (and Q. rugo-
sa and Q. laurina grow at over 2 400 m); the eastern
reaches are composed of Q. mexicana, Q. crassifolia, Q.
laeta, and Q. deserticola. Pine forests comprise mainly P.
montezumae (20-30 m), P. pseudostrobus, P. hartwegii (15-
30 m), and P. teocote (10-20 m). Below 2 000 m, P. oocar-
pa (12-20 m), P. michoacana (20-30 m), P. herrerae (30-
35 m), P. pringlei (20-25 m), and P. leiophylla are common,
while above 3 000 m, the forests are a combination of P.
hartwegii and Abies religiosa (35-45 m, and up to 50-60 m)
at 3 000 m (Valero et al. 2001c).
In the Sierra Madre del Sur, oak forests grow at 1 900-
2 500 m, cloud forests at 2 300 m, and pine-oak and pine
forests at elevations between 2 400 and 2 500 m. Domi-
nant species in oak forest formations include Q. magno-
lifolia and Q. castanea, while in pine forests the most
common species are P. herrerae, P. pseudostrobus, P.
pringlei (15-20 m), and P. ayacahuite (more than 20 m, up
to 45 m), and the endemic P. rzedowskii (15-30 m) (Far-
jon and Styles 1997; Valero et al. 2001d). Finally, in the
Sierra Norte de Oaxaca, pine forests with P. ayacahuite, P.
lawsonii (25-30 m), P. chiapensis (30-40 m), P. devoniana
(20-30 m), and P. pseudostrobus grow at 1 600-2 600 m,
and pine-oak forests characterized by P. hartwegii, P. de-
voniana, P. lawsonii, P. montezumae, Q. laurina, and Q. ru-
gosa grow from 2 000-2 800 m. P. hartwegii, P. ayacahuite,
and P. patula (35-40 m) occur at higher elevations, some-
times in association with cloud forests. Oak forests are
typical between 2 000 and 2 500 m, and include species
such as Q. crassifolia, Q. castanea, Q. crassipes, Q. rugosa,
and Q. laurina. As one approaches the Isthmus, some re-
gions become more humid and the presence of epi-
phytes increases (Valero et al. 2001e).
The Madrean Pine-Oak Woodlands Hotspot is also an
important center of cultural and ethnic diversity. In
Mexico, the natural resources of over half of the nation-
al territory are under the control and use of indigenous
communities (Toledo et al. 2002). Indigenous communi-
ties are also commonly established within the country’s
natural protected areas, and thus have to be taken into
consideration in all plans for management and protec-
tion (CONANP 2002). Among the ethnic groups charac-
teristic of the region are the Chinantec, Cora, Cuicatec,
Mazatec, Mixe, Mixtec, Náhuatl, Popoluca, P’urhepecha,
Tarahumara, and Zapotec (Robles-Gil 1995).
Biodiversity
Flowering plant species richness in the Madrean Pine-
Oak Woodlands Hotspot has been estimated at ap-
proximately 5 300 species (24% of the Mexican flora;
Rzedowski 1993; González-Medrano, pers. comm.), al-
though a compilation of total vascular plant diversity
for the Mexican pine-oak community using herbaria
databases held at the Comisión Nacional para el
Conocimiento y Uso de la Biodiversidad (CONABIO)
suggests that it may contain as many as an astonishing
6 700 species. The total native flora of the Madrean
Sky Islands includes nearly 2 000 species (McLaughlin
1994), although surveys of many such islands are in-
complete.
It is more difficult to estimate level of endemism
within the pine-oak woodlands, particularly because of
the incompleteness of inventories, but it is clear that the
percentage is quite high. Rzedowski (1993) estimates
that a remarkable 75% of the pine-oak flora of Mexico
and the pine-oak regions in the Madrean Sky Island
Archipelago is endemic, although it is unclear whether
this includes pine-oak forests lying east of the Isthmus
of Tehuantepec. While this figure is undoubtedly an
overestimate, it is the best available; we, therefore, use
a figure of 3 975 endemic vascular plant species based
on this percentage.
Examples of endemic genera in the hotspot include:
Arnicastrum, Pionocarpus, Pippenalia, Stenocarpha, and
Trichoryne (Sierra Madre Occidental); Greenmaniella,
Loxothysanus, and Mathiasella (Sierra Madre Oriental);
and Achaenipodium, Microspermum, Hintonella, Omil-
temia, Peyritschia, and Silvia (Trans-Volcanic Belt and
Sierra Madre del Sur) (Rzedowski 1978). The ten most
diverse families are: Asteraceae, Poaceae, Leguminosae,
Orchidaceae, Cactaceae (the pine-oak forests harbor be-
tween 189 and 213 cacti species, of which as many as
44% are thought to be endemic; B. Goettsch, pers.
comm.), Cyperaceae, Euphorbiaceae, Acanthaceae, Fa-
gaceae, and Agavaceae.
Around 525 bird species are known to occur regular-
ly in the hotspot, of which 23 are strict endemics. Most
endemics occur in the middle Sierra Madre Occidental,
the Trans-Mexican Volcanic Belt, and the southern Sier-
ra Madre Oriental. Three bird genera are endemic to
Mexican pine-oak woodlands, two of which are mono-
typic: Euptilotis, represented by the eared quetzal (E.
neoxenus); Xenospiza, with a single species, the Sierra
Madre sparrow (X. baileyi, EN); and Rhynchopsitta, with
two species, the thick-billed parrot (R. pachyrhyncha,
EN) and the maroon-fronted parrot (R. terrisi, VU).
Another perspective on avian endemism in this
hotspot is its coincidence with BirdLife International’s
Endemic Bird Areas (EBAs) (Stattersfield et al. 1998).
white-throated jay (Cyanolyca mirabilis, VU)—, as well
as the white-fronted swift (Cypseloides storeri). The North-
ern Sierra Madre Oriental EBA has two restricted-range
species, including the maroon-fronted parrot, while the
Southern Sierra Madre Oriental EBA has four species
confined to it, including two threatened species: the
bearded wood-partridge (Dendrortyx barbatus, VU) and
the dwarf jay (Cyanolyca nana, VU).
A remarkable 218 amphibian species have been
recorded from this hotspot, with an estimated 50 spe-
cies endemic. Frogs of the genera Hyla and Eleuthero-
dactylus are the most speciose, with 35 and 36 species,
and nine and eight endemics, respectively. Although
salamanders are mostly restricted to northern temper-
ate zone habitats, the families Plethodontidae, with 63
species (25 endemic), and Ambystomatidae, with 14 spe-
cies (five endemic), are well represented in the hotspot.
Salamanders of the genera Pseudoeurycea (25 species
present, 13 endemic), Thorius (17 species, nine endem-
ic), and Chiropterotriton (11 species, two endemic) are
particularly well represented.
In terms of reptile diversity, there 384 species
known from the hotspot, and an estimated 37 species are
endemic. One reptile genus is endemic, Rhadinopha-
nes, with a single species, the graceful mountain
snake (R. monticola). Among lizards, the genus Scelo-
porus is represented by 50 species, of which six are en-
demic; in addition, 14 species have been recorded
from the genus Anolis, two of which are endemic.
Among snakes, the Colubridae are particularly well
represented, with 147 species (17 endemic), including
14 species of Geophis (seven endemic), 11 species of
Rhadinea (three endemic), and 16 species of Tantilla
(five endemic).
There are 328 species of mammals thought to be pres-
ent in the hotspot, only six of which are strictly en-
demic. Remarkably, there are also two endemic genera,
both monotypic: Zygogeomys, represented by the Mi-
choacán pocket gopher (Z. trichopus, EN), and Romero-
lagus, with its single species, the zacatuche or volcano
rabbit (R. diazi, EN), endemic to the Trans-Mexican Vol- On the opposite page, the maroon-
canic Belt. fronted parrot (Rhynchopsitta
Given that the pine-oak woodlands have a general terrisi) is a Vulnerable species that
is one of the most important
“sky islands” character that would not be expected to
endemic flagships for the Madrean
be concordant with fish distributions, it is perhaps sur- Pine-Oak Woodlands. This flock
prising that the hotspot includes 84 fish species and was photographed in the El Taray
that 18 of them are endemic. This reflects the fact that Sanctuary in the Sierra Madre
Mexico’s total fish fauna is remarkably diverse, even at Oriental.
high elevations and in semiarid regions. A major com- © Patricio Robles Gil/Sierra Madre

Four EBAS overlap with this hotspot as we have defined
it: the Sierra Madre Occidental and Trans-Mexican
Range EBA supports seven range-restricted species,
including the thick-billed parrot, the Sierra Madre
sparrow, and the imperial woodpecker (Campephilus
imperialis, CR). The Sierra Madre del Sur EBA contains
four threatened, range-restricted species —the short-
crested coquette (Lophornis brachylopha, CR), the
white-tailed hummingbird (Eupherusa poliocerca, VU),
the Oaxaca hummingbird (E. cyanophrys, EN), and the
ponent of diversity consists of species in the minnow
family Cyprinidae that form a biogeographic track
along the axis of the Sierra Madre Occidental (Miller
and Smith 1986). This group includes twenty montane
species in genera that enter Mexico primarily from the
western United States, although the hotspot’s only en-
demic genus, Codoma, is also included. In central Mex-
ico, diversity is associated with lake-adapted fishes in
the high-elevation lakes that were created by tecton-
ism and volcanism. These lacustrine species include
Above, this subspecies of Durango
chipmunk (Tamias durangae
solivagus) is an endemic to the
Sierra de Guadalupe, Coahuila, in
the transition zone at elevations of
2 550 and 2 850 m.
© Patricio Robles Gil/Sierra Madre
209

On pp. 210-211, Cumbres de
Monterrey National Park at dawn.
© Patricio Robles Gil/Sierra Madre
Above, the volcano rabbit or
zacatuche (Romerolagus diazi) is
a species that is endemic to Mexico.
Its range is restricted to the central
part of the Trans-Volcanic Belt,
specifically around the Popocatépetl,
Iztaccíhuatl, Pelado, and Tláloc
volcanoes. This species is currently
classified as Endangered.
© Patricio Robles Gil/Sierra Madre
On the opposite page, cloud forest
and waterfall in the Sierra de
Tamaulipas, Mexico. Some of the
best remaining stands of pine-oak
woodlands occur in this state.
© Patricio Robles Gil/Sierra Madre
212
several live-bearing fishes in the family Goodeidae,
and a number of silversides in the endemic Mexican
genus Chirostoma.
Among invertebrates, the pine-oak woodlands har-
bor approximately 160-200 butterfly species, of which
45 are endemic to the hotspot (J. de la Maza, pers.
comm.). Perhaps the most noteworthy butterfly in the
pine-oak woodlands is the monarch (Danaus plexippus),
whose spectacular annual migration of up to 4 000 km
from the north culminates in masses of butterflies in
the forests of Michoacán (Brower et al. 2002).
Flagship Species
Not surprisingly, a number of the pine and oak species
from which this hotspot derives its name have excep-
tional value for conservation. For example, the Sierra
Madre Occidental is home to at least two endemic
species of oak, Quercus carmenensis and Q. deliquescens
(Nixon 1993), while forests in the Baja California Penin-
sula hold an important diversity of pine trees, among
them Pinus lambertiana, one of the largest pine trees in
Mexico, reaching 70 m in height and with cones of
70 cm in length.
The most charismatic species in this hotspot is the
endemic zacatuche or volcano rabbit, one of the small-
est and most unusual of rabbits. It sports small, round
ears and —unlike other rabbits— utters high-pitched,
penetrating vocalizations. The zacatón bunchgrasses
are essential to the zacatuche —the rabbit constructs the
entrance to its burrow at the base of a clump of bunch-
grass, and prunes the plants so that they form a dense
roof for protection and cover (Cervantes et al. 1990;
Romero and Velázquez 1994; Velázquez et al. 1996).
The zacatuche lives only in the Trans-Mexican Volcanic
Belt surrounding Mexico City. It currently occurs on
the slopes of the Iztaccíhuatl Volcano and the Popocaté-
petl Volcano, and in 16 small fragments of habitat on
the slopes of the Pelado Volcano and the Tláloc Volcano
in the mountain complex south of Mexico City. Fre-
quent burning of the zacatón grasses to increase the
production of palatable forage for livestock, clearing of
zacatón for thatch (used in brooms and other products),
unsustainable forestry, and ever-increasing encroach-
ment of settlements from the expansion of Mexico City
have all negatively impacted the zacatuche (Portales et
al. 1997; Velázquez et al. 1996). Efforts to save the za-
catuche from extinction will benefit the entire habitat
and all the species occupying the Trans-Mexican Vol-
canic Belt of the Madrean Pine-Oak Woodlands Hot-
spot, and for this reason it has been proposed as a sym-
bol for conservation in Mexico (Velázquez 2001).
This hotspot is also home to one of the world’s most
famous wildlife spectacles, the overwintering mass of
monarch butterflies (Danaus plexippus) in the pine
forests of Michoacán. Each fall, about 100-500 million
monarchs migrate south from eastern North America to
form giant clusters on the boughs and trunks of trees in
the oyamel fir (Abies religiosa) dominated ecosystem
(Brower et al. 2002). There are only about 30 of these
overwintering sites covering 10-25 ha each. It has been
estimated that the spring and summer range of the
monarch is a million times more extensive than their
overwintering habitat (Brower et al. 2002). This scien-
tific and aesthetic wonder attracts over 200 000 visitors
annually and represents a seasonal economic boon to
the communal landholders that manage the main but-
terfly sanctuaries in the Monarch Butterfly Biosphere
Reserve.
During January of 2002, however, disaster struck,
and as many as 250 million dead monarchs littered the
ground in a carpet 20 cm deep in places. In some
colonies, over 80% of all butterflies died (Brower et al.,
in press). The immediate cause of this tragic loss was
a severe winter storm that swept across Mexico, but
deforestation has increased the susceptibility of mon-
archs to such storms. The forests provide substantial
protection, but with increasing fragmentation of these
forests, the vulnerability of the butterflies to cold and
wet conditions intensifies. Despite recent legislation to
prevent logging, the practice continues, and the mira-
cle of the monarchs is in jeopardy.
Farther north, in Cave Creek Canyon in the Chiri-
cahua Mountains of southeastern Arizona, birders flock
to see the elegant trogon (Trogon elegans). This species’
distributional range closely maps onto that of the
Madrean Pine-Oak Woodlands Hotspot, and it reaches
its most northern extension in the sky islands. It is not
alone, as the southeastern mountains in Arizona harbor
more species of hummingbird than any part of the
United States, as well as a host of other flora and fauna
that link this sky island to the pine-oak woodlands to
the south. The uniqueness of the biota in these sky is-
lands has given rise to an expanding ecotourism indus-
try and, indeed, ecotourism, primarily based around
bird watching, is the main economic driver in this part
of Arizona.
Threats
Pines (and, to a lesser extent, oaks) are very important
for the Mexican logging industry and their exploita-
tion for timber has increased, in some areas becoming
indiscriminate. Illegal logging is on the rise in many ar-
eas (Challenger 1998). Additionally, many non-timber
forest products are being used unsustainably within the
hotspot, although these fail to register in any formal
economic accounting. Examples are: a common vascu-
lar epiphytic plant (despite the fact that pine forests
usually do not have epiphytics) that seems like a lichen
(Tillandsia usneoides), which is extracted for Christmas
ornamental purposes in Mexico; and a large variety of
mushrooms in pine-oak forests such as Amanita spp.,
Leccinum spp., Russula spp., and Boletus spp., among oth-
ers, that are extracted for culinary use (Challenger 1998).
Habitat alteration, generally through fire, influences the
understory and impacts both flora and fauna. Fires
are partly a natural process in this ecosystem, but in-
tentional burning has increased greatly to foster re-
generation of fresh sprouts for livestock grazing, and
also for other agricultural purposes. In southern Ari-
zona, fire has been excluded for almost a century, pri-
marily due to suppression by federal and state agen-
cies, and modification of fuels by grazing and other
causes. The apparent result of 90 years of suppression
and fuel modification has, at higher elevations, led to
conifer forests changing from open-grown ponderosa
pine (P. ponderosa) to dense stands of mixed conifers
(Allen 1994).
In most areas of the Madrean Pine-Oak Woodland
Hotspot, local experts highlight the massive disruption
of the ecosystem. In the mountain region to the south of
the Cuenca de México, Romero et al. (1999) plea for re-
forestation programs, and patrolling actions to reduce
wildlife poaching, illegal timber harvest, plant collec-
tion, and soil and rock extractions. In the forests of Mi-
choacán that are home to overwintering populations of
monarch butterflies, illegal deforestation proceeds on a
grand scale. Stattersfield et al. (1998) comment on the
threats occurring within each of the region’s EBAs as
follows: In the Northern Sierra Madre Oriental, “the
mixed conifer forests…are being destroyed by fire, log-
ging and clearance for agriculture.” In the Sierra Madre
Occidental and Trans-Mexican Range EBA there has
been “almost complete destruction of old-growth pine
forest,” which seriously threatens the survival of the
thick-billed parrot and may have led to the extinction of
the imperial woodpecker (currently listed as CR). And
in the Southern Sierra Madre Oriental, “the majority of
the forests…have already been lost or degraded, and
this destruction continues through logging, agricultural
expansion, firewood-gathering, road and associated
tourist developments, sheep ranching and overgrazing,
as well as intensive urbanization.” The consistency of
these reports across the landscape of the Madrean Pine-
Oak Woodlands Hotspot paints a grim picture for its re-
markable biodiversity.
Due to the fragmented nature of the Madrean Pine-
Oak Woodlands Hotspot, arriving at a precise estimate
of the amount of original vegetation remaining intact is
not easy. Originally, pine-oak forests covered around
21% of Mexico (Rzedowski 1978), and it is estimated
that about 50%-67% of this has been lost and that the
remaining forest covers no more than 8% of the coun-
try (Rzedowski 1993). However, these figures do not
take into account those remaining stands that have
been impacted by fire, overgrazing, and other factors.
In total, we estimate that no more than 20% of the re-
maining vegetation can be considered pristine. Other
estimates stating that less than 1% of the original vege-
tation of the Sierra Madre Occidental remains intact
(Lammertink et al. 1997) refer to pristine, old-growth
forests with no history of human intervention, but non-
pristine pine-oak forests are capable of retaining a very
significant portion of the original diversity.
Conservation
Despite the biodiversity value of the Madrean Pine-Oak
Woodlands Hotspot, the region’s protected area cover-
age is poor. Less than 6% of the hotspot is under some
form of official protection, and only 1.9% is in protected
areas that fall within IUCN categories I to IV. Most ter-
restrial vertebrate species endemic to Mexico that are
unrepresented in the country’s protected area system
are located on the southern slopes of the Sierra Madre
del Sur and, to a lesser extent, on the western slopes
of the Sierra Madre Occidental and the eastern slopes of
the Sierra Madre Oriental, highlighting the immediate
need for improved protection in this hotspot.
Among the more important protected areas in the
hotspot is the famous Reserva de la Biosfera Mariposa
Monarca (563 km2), the Monarch Butterfly Biosphere
Reserve, in Michoacán, which was decreed in 1986 and
conserves different types of vegetation such as pine,
pine-oak, Abies, and Juniperus forests. One of the larger
protected areas is the Sierra de Manantlán Biosphere
Reserve (1 396 km2) in Jalisco (and also in the Trans-
Mexican Volcanic Belt); pine-oak forests are very diverse
in this reserve, with some 33 species of Quercus. Close to
this reserve is the La Primavera Forest Reserve (305 km2),
while further to the east are the following National
Parks: Nevado de Colima (96 km2), Nevado de Toluca
(468 km2), and Iztaccíhuatl-Popocatépetl (903 km2), as
well as the Cofre de Perote (117 km2) and Pico de Ori-
zaba National Parks (197 km2) in Veracruz. The largest
protected area in the Sierra Madre Oriental, and one of
the largest in Mexico is the Cumbres de Monterrey Na-
tional Park (1 774 km2), while further north, on the
American side of this hotspot, is the Big Bend National
Park in Texas, covering 3 245 km2.
In the Sierra Madre Occidental, parks conserving no-
table stands of pine-oak formations include the Cum-
bres de Majalca National Park (48 km2), the Cascada de
Basaseachic National Park (58 km2), and the La Michi-
lía Biosphere Reserve (93 km2). Further north, in the On the opposite page, the Near
Madrean Sky Islands of the United States, most of Threatened tufted jay (Cyanocorax
the pine-oak woodland is “protected,” but the majority dickeyi) is a beautiful Madrean
Pine-Oak endemic that has such a
is in U.S. Forest Service land. However, a variety of
restricted range that it was not
much smaller, highly protected reserves occur, some in discovered until 1934. This one was
private ownership (such as The Nature Conservancy), photographed in the Sierra Madre
although most are U.S. National Monuments, or desig- Occidental.
nated as U.S. Wilderness Areas. These include Corona- © Patricio Robles Gil/Sierra Madre
do National Memorial (19 km2), Chiricahua National
Above, the aptly named long-tailed
Monument (42 km2), Chiricahua Wilderness (355 km2),
wood-partridge (Dendrortyx
Miller Peak Wilderness (82 km2), and Pajarita Wilder- macroura) is endemic to montane
ness (31 km2). Finally, in Baja California, the Sierra de oak and pine forest and cloud forest
la Laguna Biosphere Reserve (1 124 km2) was estab- within the central and southern
lished as a protected area in 1994 primarily to protect Mexican section of this hotspot.
this island of intact pine-oak vegetation. Within this range, it appears to be
common, but is poorly known and
A number of conservation NGOs are working in the
surveys are needed.
region. The Fondo Mexicano para la Conservación de la
© Patricio Robles Gil/Sierra Madre
Naturaleza is working to enhance social participation so
as to prevent fires in protected areas and a number of
strategic regions, among them pine-oak forests in the
215

In the Maderas del Carmen project,
run by CEMEX, a radio telemetry
study was conducted with
American black bears (Ursus
americanus) to determine their
movements within this biological
corridor.
© Patricio Robles Gil/Sierra Madre
On the opposite page, the black
bear (Ursus americanus) is a
widespread North American species
that reaches exceptionally high
densities in some parts of the
Madrean Pine-Oak Woodlands.
© Patricio Robles Gil/Sierra Madre
216
Sierra de Manantlán and the monarch butterfly sanctu- The village of San Nicolás Totolapan is perched on a
ary. PRONATURA/PRONATURA NORESTE is an NGO mountain just south of Mexico City. The rapid urban-
established in 1997 that aims to conserve and pro- ization of Mexico City has encroached upon the village;
mote the sustainable use of natural resources in indeed, most of its young people worked dead-end
northeast Mexico. The organization has programs un- jobs in the city and had no connection to the land of
der way in the Sierra Madre Occidental (the Thick- their ancestors. For generations, this small community
billed Parrot Sanctuary in Madera, Chihuahua) and had coexisted with nature through livelihoods based on
Sierra Madre Oriental, where it is contributing to the the sustainability of natural resources. The village’s tra-
government program of reforestation and soil restora- ditions and unique environment were in jeopardy. In
tion in areas affected by forest fires in the villages response, they looked toward a solution that would pro-
of Laguna de Sánchez and Mesa de las Tablas at the tect their local land (ejido) while preserving their cul-
Sierra de Arteaga in the states of Nuevo León and tural heritage. The result has been a remarkably suc-
Coahuila, respectively. cessful ecotourism park —the San Nicolás Totolapan
The Sierra Madre Alliance is a support network of Park— providing a unique experience with nature in an
Mexican and international partners pursuing conserva- environmentally sensitive manner. There are 150 km
tion priorities in the Sierra Madre Occidental in Chi- of bike and hiking trails, photographic routes, a nursery
huahua. The Alliance aims to preserve biodiversity and for native plants (for reforestation), as well as permits for
restore the functioning of the forested ecosystems of camping and fishing. Through ecotourism the commu-
the Sierra Madre through local participation, including: nity is ensuring the survival of a biodiverse ecosystem,
protection of remnant old-growth pine-oak forests, and as well as providing environmental awareness pro-
the establishment of a Tarahumara-managed forest re- grams. In addition to the environmental benefits that
serve in Pino Gordo. Finally, in the United States, the the park provides for the region, it has also had a posi-
Sky Island Alliance is dedicated to preserving this glob- tive impact on the daily lives of the local people, for ex-
ally important region. Formed in 1992, it has worked ample, through increased employment. This example
with agencies in Mexico to produce a Sky Islands Wild- shows how protecting the Madrean Pine-Oak Wood-
lands Network Conservation Plan designed to take into lands Hotspot can be a win-win situation.
account the needs and concerns of all stakeholders in
the region (within the U.S., including the U.S. Forest PATRICIA KOLEFF 24
Service, existing Nature Conservancy and Audubon So- JORGE SOBERÓN 24
ciety reserves, private ranges, etc.) to formulate a ANDREW SMITH 11
framework for conservation of the unique biota in the
region.
Another very positive development has been the
work of CEMEX to protect Maderas del Carmen in
the northern part of Coahuila. Since 1999, the compa-
ny has purchased and effectively protected what is
now a total of 70 000 ha of land, including several new
pieces that bring El Carmen right up to the U.S. border
and make it contiguous with the Big Bend National
Park in Texas. Now that the Mexican and American
pieces have been connected, the result is a 2 000 000-ha
conservation mega-corridor that has enormous signifi-
cance for conservation. Indeed, it includes both some
of the most important remaining tracts of the Chi-
huahuan Desert wilderness and significant portions of
this newly-recognized Madrean Pine-Oak Woodland
Hotspot. Furthermore, the vision of CEMEX is to expand
protection on the Mexican side even more, making this
area a real model for private sector conservation.
Nevertheless, it will take a ground swell of approach-
es, from the NGO, government, and private sectors, to
arrest the decline of the Madrean Pine-Oak Woodlands
Hotspot across its full extent. So perhaps it is best to
conclude with a remarkable story of one village that
has looked for solutions to meet the demands for both
economic development and biodiversity conservation
and, in doing so, has preserved its own future.
 MAPUTALAND-
PONDOLAND-ALBANY
The Maputaland-Pondoland-Albany region lies
along the east coast of southern Africa below the
Great Escarpment. It extends from the extreme
southern parts of Mozambique (Limpopo River)
and Mpumalanga (Olifants River) in the north,
through Swaziland and KwaZulu-Natal, to the East-
ern Cape Province in South Africa in the south. The
Maputaland-Pondoland-Albany region is very di-
verse and complex floristically, climatologically,
and geologically. There are at least three clear foci
of high endemism and high diversity in the area,
the names of which have been amalgamated as the
name of this hotspot. These foci are Maputaland
(= Tongaland) in the north, Pondoland further
south, and Albany in the southwest. The Maputa-
land-Pondoland-Albany region is not only a center
of endemism, but also a marked center of diversity
across the taxonomic spectrum.
The Maputaland-Pondoland-Albany Hotspot is
about 274 316 km2 in size, and its boundaries corre-
spond broadly to White’s (1983) delimitation of the
Tongaland-Pondoland Regional Mosaic. However, as
defined here, it is a bit larger than this region as it
extends further inland to include areas of temperate
grassland and forest lying below 1 800 m along the
Great Escarpment in KwaZulu-Natal and the Eastern
Cape, and below 1 200 m further north in Swaziland,
Mpumalanga, and Limpopo (see Davis et al. 1994;
Van Wyk and Smith 2001). Following Van Wyk and
Smith (2001), it also extends further west in the
south to include the Albany center of plant en-
demism. The Maputaland area has strong floristic
and faunistic connections with the Coastal Forests of
Eastern Africa Hotspot to the north, whereas the
Pondoland region shows some floristic links with
the Cape Floristic Region further south.
The Maputaland-Pondoland-Albany Hotspot bor-
ders several important areas of plant diversity and
endemism, including the Cape Floristic Region
Hotspot, and the Barberton, Wolkberg, and Drakens-
berg Alpine Centers of Plant Endemism (Van Wyk
and Smith 2001), which are considered as outliers of
the broader Eastern Afromontane Hotspot.
The topography of the Maputaland-Pondoland-Al-
BOTSWANA MOZAMBIQUE
SWAZILAND
SOUTH AFRICA
LESOTHO
INDIAN
OCEAN
0 300 km
bany region is very diverse, ranging from ancient
and young sand dunes and low-lying plains in the
north to a series of rugged terraces deeply incised by
river valleys in the central and southern parts. Sev-
eral mountain ranges, including the Sneeuberg
(highest peak in the region: Kompasberg in Sneeu-
berg, 2 502 m), Winterberg (2 369 m), Amatola Moun-
tains (1 937 m), Ngeli Range (2 268 m), Lebombo
Mountains (699 m), and Ngoye Range (486 m), also
occur within the Maputaland-Pondoland-Albany re-
gion, while the Great Escarpment borders it to the
west. The break-up of Gondwana, and subsequent
cycles of uplift and erosion shaped the landscape of
the Maputaland-Pondoland-Albany region. These
processes formed the Great Escarpment, which On the opposite page, Strandloper
receded from the coast after the break-up event and Falls in the Mkambati Nature
the establishment of an effective drainage system. Reserve, Pondoland.
© Patricio Robles Gil/Sierra Madre
Today, the Great Escarpment separates the elevated
interior plateau of southern Africa from the coastal
lowlands. Regional geology consists of basement
granites, gneisses and schists, various sedimentary
deposits, lavas (basalt and dolerite intrusions), and
marine sediments of various ages. The climate of the
219

The blood lily or snake lily
(Scadoxus puniceus) is one of
South Africa’s most striking bulbous
plant species. As with other closely
related species, the bulb of this
species is poisonous and deaths
have been reported following its
ingestion (as the species is widely
used in traditional medicine to treat
coughs and gastrointestinal
problems).
© Cristina G. Mittermeier
On the opposite page, historically,
nyalas (Tragelaphus angasii)
were widespread in thicket and
forest habitats in Mozambique,
occurring as far south as the
Hluhluwe River in KwaZulu-Natal,
South Africa. Their current
distribution is more widespread,
due largely to introductions outside
their former range.
© Patricio Robles Gil/Sierra Madre
220
Maputaland-Pondoland-Albany region ranges from sub-
tropical-tropical in the low-lying, northern coastal areas
to more temperate with frost in winter on higher
ground away from the coast.
Six of South Africa’s eight biomes enter into the
hotspot, and 27 of the 68 vegetation types that occur
within South Africa, Lesotho, and Swaziland, according
to Low and Rebelo (1998), as well as one that they do
not recognize, are represented within the region. One
type of forest (namely Licuáti forest, previously called
sand forest), three types of thicket, six types of bushveld,
and five types of grassland, including the coastal
plateau sourveld grasslands of Pondoland, the short
mistbelt grasslands, and the woody grasslands of Ma-
putaland, are endemic to the Maputaland-Pondoland-
Albany Hotspot.
The forests of the Maputaland-Pondoland-Albany re-
gion, despite their naturally fragmented distribution,
are of special interest. Forest vegetation covers less
than 30 000 km2 in South Africa (of which approxi-
mately 80% falls within the hotspot); these warm tem-
perate forests have by far the highest tree richness of
any of the world’s other temperate forests, with some
598 tree species occurring (Silander 2001). This rich-
ness in tree species is exceeded only in the evergreen
forests of East Asia, where 876 species grow in a much
larger area. South African forests are also between
three and seven times richer in tree species than other
forested areas of the Southern Hemisphere, even though
the other Southern Hemispheric forests cover a much
larger area (Silander 2001).
The thicket biome of southern Africa, the largest
part of which occurs within the Maputaland-Pondoland-
Albany region, is thought to be the most species-rich
formation of woody plants within South Africa. It has
been suggested that thickets are extremely ancient and
include many elements basal to the Cape and Succu-
lent Karoo flora (Vlok et al. 2003). In addition to forest
and thicket, grassland is also of particular importance
in this hotspot, especially as it is the most threatened
and least protected of all the biome types in southern
Africa.
This hotspot also has a remarkable succulent flora
that is mainly concentrated in the Albany region. The
succulent riches of southern Africa are well known (es-
pecially that of the Succulent Karoo Hotspot), with over
46% (4 674 taxa in 58 families) of the world’s succulents
growing naturally in southern Africa (Smith et al. 1997).
This is perhaps not surprising, as large parts of the south-
ern African landscape are prone to regular droughts, to
which numerous plant species have adapted by devel-
oping succulent leaves, stems, and roots. However,
whereas leaf succulents predominate in the western,
mainly winter-rainfall parts of southern Africa, the suc-
culents of the Maputaland-Pondoland-Albany region
are predominantly stem succulents.
The Maputaland-Pondoland-Albany region is not ex-
ceptionally diverse as a region of cultural diversity
when compared with other parts of southern and cen-
tral Africa. However, the region is very densely popu-
lated (over 20 million people), with extensive informal
township and urban development, especially along the
coastline. In some parts the human population reaches
densities of up to 1 900 people per km2 (Durban and
surroundings). Other parts are less densely populated;
the Pietermaritzburg and Port Elizabeth areas, for ex-
ample, have population densities of around 500 people
per km2, while Umtata and East London population den-
sities are up to 100 people per km2. Most of the region
is subjected to high population densities, with only a
third of the land surface area having population densi-
ties of below 50 people per km2.
Nevertheless, the hotspot remains a significant pluri-
cultural, multilingual, and multi-ethnic region. Its oldest
known inhabitants were various tribes of Bushmen
(Saan) and Khoekhoe. Sadly, these cultures are now es-
sentially gone from the region, although limited assimi-
lation of the original Khoesaan languages is still reflect-
ed by the presence of so-called click sounds in widely
spoken Nguni languages. Modern-day cultural groups
centered in the Maputaland-Pondoland-Albany region
include the Zulu, Xhosa, Swazi, Ronga, and Shangaan.
Other cultures, mainly centered in and around urban
areas, include the English and Afrikaners (speakers of
Afrikaans —the only Germanic language to have devel-
oped outside Europe), as well as Indians. As a former
Portuguese colony, Portuguese is one of the official lan-
guages in Mozambique and is widely spoken and un-
derstood in the Mozambican part of the hotspot.
Biodiversity
Based on species numbers, the Maputaland-Pondoland-
Albany region is the second richest floristic region in
Africa, after the Cape Floristic Region. An estimated
8 100 species from 243 families occur within the Ma-
putaland-Pondoland-Albany region, and at least 1 900
(23%) species are endemic to the region. Plant families
rich in endemics are (approximate number of species
endemic to the Maputaland-Pondoland-Albany region
in parentheses): Asteraceae (266), Apocynaceae (203;
including Asclepiadaceae and Periplocaceae), Fabaceae
(200), Asphodelaceae (155), Iridaceae (110), Euphor-
biaceae (96), Scrophulariaceae (81), Lamiaceae (77), and
Mesembryanthemaceae (76). One endemic family oc-
curs within the Maputaland-Pondoland-Albany Hotspot,
the monotypic Rhynchocalycaceae. In all, there are
1 524 vascular plant genera in the hotspot, of which 39
are endemic, including: Acharia (Achariaceae), Berge-
ranthus (Mesembryanthemaceae), Dahlgrenodendron
(Lauraceae), Dermatobotrys (Scrophulariaceae), Emico-
carpus (Asclepiadaceae), Helichrysopsis (Asteraceae),
Heywoodia (Euphorbiaceae), Pseudosalacia (Celastra-
ceae), Rhynchocalyx (Rhynchocalycaceae), Stangeria
(Stangeriaceae), and Umtiza (Caesalpiniaceae).
Vertebrate diversity and endemism are low relative
to most other hotspots. Birds are the most diverse group
of vertebrates in the hotspot, with 541 regularly occur-
ring species. The hotspot forms part of the Southeast
African Coast Endemic Bird Area recognized by Bird-
Life International (Stattersfield et al. 1998), with four
restricted-range species: Rudd’s apalis (Apalis ruddi),
pink-throated twinspot (Hypargos margaritatus), Neer-
gaard’s sunbird (Cinnyris neergaardi), and lemon-breasted
canary (Serinus citrinipectus).
The reptiles are the second most diverse vertebrate
group in the hotspot. Of the 205 species occurring, 36
are endemic. At least seven species of dwarf chameleon
(Bradypodion spp.) occur in the hotspot: the Transkei
dwarf chameleon (B. caffrum), the Kentani dwarf cha-
meleon (B. kentanicum), the black-headed dwarf chame-
leon (B. melanocephalum), the Zululand dwarf chameleon
(B. nemorale), Setaro’s dwarf chameleon (B. setaroi), the
Natal Midlands dwarf chameleon (B. thamnobates), and
the southern dwarf chameleon (B. ventrale). All have
very restricted distributions within the region. There is
one endemic genus (Macrelaps), represented by a sin-
gle species, the Natal black snake (M. microlepidotus).
There are a total of 193 species of mammals, and at
least five species are endemic, including two species of
golden mole, Marley’s golden mole (Amblysomus mar-
leyi) and the giant golden mole (Chrysospalax trevelyani,
EN), and the four-toed elephant shrew (Petrodromus
tetradactylus).
The frogs of the Maputaland-Pondoland-Albany
hotspot number 80 species, of which 12 are endemic.
Two genera are endemic, both represented by single
species: Boneberg’s frog (Natalobatrachus bonebergi,
EN) and Rattray’s or hogsback frog (Anhydrophryne
rattrayi, EN).
Of the 73 indigenous species of freshwater fishes oc-
curring within the Maputaland-Pondoland-Albany re-
gion, 20 are endemic, including four species of barb
(Barbus spp.) (Skelton 2001). The Maputaland-Pondoland-
Albany Hotspot also harbors an exceptionally rich and
diverse invertebrate fauna. Charismatic insect groups
such as butterflies and moths are well represented in
the hotspot, with several rare and localized species.
Among the more spectacular butterflies are Charaxes
pondoensis, a species confined to a small area of coastal
forest in the vicinity of Port St. Johns. The Lycaenidae
is the largest family of butterflies in southern Africa,
and several rare species are endemic to the hotspot.
The pale yellow Bashee River buff (Deloneura immacu-
lata) is known from only three specimens collected
from Fort Bowker on the Bashee River in 1863. The
species has not been recorded since, despite an exten-
sive search by numerous collectors over the years, and
is considered Extinct.
The phylum Onychophora comprises a fascinating
group of ancient, caterpillar-like animals whose fossil
record shows that they have not changed substantially
in 400 million years. They are the most primitive group
of animals to walk with the body raised upon legs
(Hamer et al. 1997). There are two genera endemic to
the larger southern African region, Opisthopatus and
Peripatopsis. Of the nine described species of Ony-
chophora in South Africa, the genus Opisthopatus is
largely confined to the Maputaland-Pondoland-Albany
Hotspot; O. roseus (CR), one of two species in the genus,
is extremely rare and only known from Ngeli Forest
near Kokstad (Hamer et al. 1997). Peripatopsis is repre-
sented in this region by the endemic P. moseley.
Finally, the family Microchaetidae (Oligochaeta)
contains four genera and over 100 described species
of truly amazing earthworms. The family is endemic
to southern Africa, with the Maputaland-Pondoland-
Albany Hotspot a major center of diversity for the
group. Moreover, the monotypic genus Michalakus is
endemic to the region, as are perhaps the majority of
species in the genus Tritogenia (Plisko 1998). Many mi-
crochaetids in the Maputaland-Pondoland-Albany re-
gion are gigantic and inhabit moist, undisturbed, pri-
mary grassland or forest. Perhaps the most remarkable
is Microchaetus vernoni, with adults known to reach a
length of 2.6 m and a diameter of about 10 mm. It is
only known from grassland in Vernon Crookes Nature
Reserve, southern KwaZulu-Natal, a relatively small
(2 189 ha) conservation area to which another two spe-
cies, M. zaloumisi and M. ambitus, are endemic.
Flagship Species
The Maputaland-Pondoland-Albany Hotspot is a source
of numerous plants that have been developed success-
fully as horticultural subjects across the globe, includ-
ing Tecomaria capensis, Plumbago auriculata, Crassula
ovata, Carrissa macrocarpa, and many of the flagship
species discussed here. One such flagship, the bitter
aloe (Aloe ferox), is one of the best known and most
conspicuous floristic elements in the southern parts of
the hotspot. This medium to large, single-stemmed
aloe is arguably the most important medicinal plant in
South Africa. The yellowish brown leaf exudate has
been used for several hundred years as the primary in-
gredient of a purgative drug known commercially as
Cape Aloes.
The bird-of-paradise flower (Strelitzia reginae) is a On the opposite page,
hotspot endemic that grows up to 2 m high in its nat- the Cercopithecus mitis group of
ural habitat in the Eastern Cape coastal bush. Today, it guenons is represented in the
Maputaland-Pondoland-Albany
is a popular horticultural subject in many parts of the
Hotspot by the distinctive endemic
world and has even been adopted as the civic emblem Samango monkey subspecies,
of Los Angeles. This species hybridizes readily with C. m. labiatus.
the sword-leaved crane flower (S. juncea), another © Patricio Robles Gil/Sierra Madre
hotspot endemic. A tree member of the genus, Natal
wild banana (S. nicolai), is also endemic and distrib- Above, poor man’s cycad
(Encephalartos villosus) in the
uted along nearly the whole of the eastern seaboard of
St. Lucia World Heritage Site in
the region. KwaZulu-Natal, South Africa.
Commonly known as Christmas bells, Sanderso- © Patricio Robles Gil/Sierra Madre
nia aurantiaca is a monotypic genus endemic to this
hotspot, which has beautiful orange-yellow flowers and
lily-like growth. In past times, this flower was so plen-
tiful in the region that it was common to see Zulu
women walking with huge bunches that they collected
223

Girl and infant in the Maputo
Elephant Reserve, Mozambique.
© Cristina G. Mittermeier
On the opposite page, in addition to
the irreversible loss of habitat in the
Maputaland-Pondoland-Albany
Hotspot to cultivation,
urbanization, and plantations,
localized mining activities
—specifically titanium extraction
from coastal sand dunes— have
also led to loss of natural vegetation.
© Patricio Robles Gil/Sierra Madre
224
in the field, selling them from door-to-door as cut flow-
ers. Recently, however, it seems that this species is get-
ting increasingly rare.
About 40 species of red-hot poker (Kniphofia spp.) oc-
cur within the Maputaland-Pondoland-Albany Hotspot.
Of these 40 species, at least half are endemic. The cho-
sen flagship species, Kniphofia rooperi, is a large and
sturdy plant that flowers about mid-September. Unlike
the typical expectation of red-hot pokers having bright
red flowers, this species’ flowers are orange-yellow in
color. Already many cultivars of Kniphofia have been
produced by horticulturists, including the use in some
cases of K. rooperi as one of the parents.
An important avian flagship is the southern race of
Cape parrot (Poicephalus robustus robustus), which can
be distinguished from the northern race by its brownish
head. It is dependent on the yellowwoods (Podocarpus
spp.) for both nesting sites and food, although illegal
harvesting of yellowwood timber for the furniture mar-
ket severely threatens the existence of this subspecies.
Another striking flagship is Woodward’s barbet (Stacto-
laema olivacea woodwardi) which, in southern Africa, is
restricted to the Ngoye Forest between Eshowe and
Empangeni in KwaZulu-Natal. The species also occurs
on the Rondo Plateau in Tanzania; however, the precise
taxonomic status of these disjunct populations is not
clear. The pink-throated twinspot and Neergaard’s sun-
bird are both endemics of the northern parts of the
hotspot. The pink-throated twinspot is mostly confined
to low-lying coastal areas, but in Zululand is also found
at the top of the Lebombo Mountains, inhabiting the
edge of the forested slopes. Neergaard’s sunbird is con-
fined to the coastal plains north of Lake St. Lucia. Oth-
er bird flagship species include Gurney’s sugarbird
(Promerops gurneyi), the spectacular Knysna turaco or
Knysna lourie (Tauraco corythaix), and the southern
subspecies of Delegorgue’s pigeon (Columba delegorguei
delegorguei).
Reptilian flagship species include the seven species
of the dwarf chameleons (genus Bradypodion) and
three endemics: the Natal hinged tortoise (Kinixys na-
talensis), the Albany adder (Bitis albanica), and Tas-
man’s girdled lizard (Cordylus tasmani). The Natal
hinged tortoise occurs throughout the Lebombo Moun-
tain range; the Albany adder is a very rare snake con-
fined entirely to the Algoa Bay area of the Eastern
Cape; and the Tasman’s girdled lizard is endemic to the
Algoa Bay area, where it lives under the “apron” of dead
leaves on tall aloes, or on dead aloe stems lying on
rocky slopes.
One of the most notable mammal flagships, and ar-
guably among the most important for the hotspot, is the
southern subspecies of the white rhinoceros (Cera-
totherium simum simum). This species was once com-
mon and widely distributed throughout southern and
East Africa. The reduction in its range in recent times
is largely due to hunting for its prized horn. The south-
ern race narrowly survived extinction in the KwaZulu-
Natal’s Hluhluwe-Umfolozi Park, which acted as a
refuge during the times when the southern white rhino
was at its most vulnerable, being reduced to a few
dozen in number. In one of the greatest conservation
success stories in African conservation, the southern
white rhino has since increased in number to more
than 12 000, with many having been relocated to other
areas.
Two dainty antelope species also serve as important
flagships. The blue duiker (Philantomba monticola) has
a disjunct distribution in southern Africa, the southern
population being confined to the hotspot. The south-
ern form of the blue duiker favors forest, thicket, and
dense coastal bush, and appears to be unaffected by
the availability of drinking water. Blue duikers are se-
verely threatened by habitat destruction and fragmen-
tation of populations, and are heavily poached with
dogs and by snaring. The southern race of the suni
(Neotragus moschatus zuluensis) is endemic to the hot-
spot. It is a secretive, little antelope that is very habitat-
specific, relying on forest with high stem density and
low ground cover. The Zululand suni’s distribution is
currently quite restricted due to the destruction of
its habitat. It is both sensitive to, and dependent on,
the modification of its habitat by factors such as the
feeding behavior of cattle, nyalas, and elephants, har-
vesting activities by humans, and the impact of fire on
forest margins and understory. Three small mammals
deserve mention as flagships, namely the red bush
squirrel (Paraxerus palliatus, VU), and the two golden
mole species.
Frog flagships of the Maputaland-Pondoland-Albany
region include the aforementioned Boneberg’s frog and
Rattray’s frog. The former is restricted to forests along
the coasts where recent housing developments and
sugarcane plantations have destroyed much of its habi-
tat. The latter is known from the Amatola and Katberg
Mountains in the Eastern Cape Province, where it oc-
curs along streams in thick vegetation; commercial tim-
ber plantations are the main threat to this frog’s con-
tinued existence. Other notable species include the
golden spiny reed frog (Afrixalus aureus), Pickersgill’s
reed frog (Hyperolius pickersgilli, EN), and the soprano
or whistling rain frog (Breviceps sopranus), a recently
described species so called because it utters a long,
high-pitched whistle. The golden spiny reed frog, with
its gold-colored back, is endemic to the hotspot (where it
occurs in low-altitude grasslands), while Pickersgill’s
reed frog is also endemic and is found in the coastal
lowlands of KwaZulu-Natal.
Among the flagship freshwater fishes are the Sibayi
goby (Silhouettea sibayi), which occurs in a variety of
freshwater and brackish habitats, and the border barb
(Barbus trevelyani, CR), which is restricted to the
Keiskamma and Buffalo river systems in the Eastern
Cape, and inhabits pools and riffles of clear rocky
streams, where it feeds on insects, seeds, and algae.
The Eastern Cape rocky (Sandelia bainsii, EN) is re-
stricted to the Buffalo, Keiskamma, Great Fish, and
Kowie river systems in the Eastern Cape; interestingly,
the only other species in the genus is endemic to
streams in the Cape Floristic Region.
The hotspot also has a very rich and varied scarab or
dung beetle (Scarabaeidae: Scarabaeinae) fauna. One
rare species, the flightless dung beetle (Circellium bac-
chus), has a very restricted present-day distribution and
has captured the imagination of visitors to the Addo
Elephant Park in the Eastern Cape. The only member
of its genus, it is named after the god of wine, probably
because of its somewhat erratic, stumbling behavior!
This must be one of the few insect species in the world
for which special road signs have been erected, alerting
motorists not to drive the wheels of their vehicles over
elephant and buffalo dung pads in roads.
Threats
Having one of the highest human densities in sub-
Saharan Africa (Esterhuysen 1998a, b), the Maputaland-
Pondoland-Albany Hotspot is threatened by a number
of human activities. Land cover information derived
from satellite data indicates that permanent and com-
plete transformation of habitat has affected 19% of the
region. This has been caused mainly by cultivation
(12.7%), plantation forestry (3.4%), and urbanization
(1.7%). A further 30% of the natural vegetation has
been severely damaged and permanently degraded so
that it now exists only in a secondary state, while about
27% is in a poor, in-between, non-pristine state. This
degradation has mainly been caused by harvesting of
indigenous woodlands, soil erosion, overgrazing and
shifting cultivation, and invasive species. A maximum
of 24.5% of the hotspot can be considered close to the
pristine state.
Degradation of the thicket and grassland biomes has
been particularly severe. By 1981, more than 50% of
the thicket biome in South Africa was seriously over-
grazed and 9% has been permanently transformed (Le
Roux 2002). Since then, these figures have probably in-
creased significantly, all the more worrying since only
5% of the thicket biome is formally protected in South
Africa. Between 60% and 80% of the grasslands of
South Africa are irreversibly transformed, while only
2% are formally conserved (Le Roux 2002). Many of
these are (were) primary grasslands that took hundreds
of thousands of years to develop their present diversity,
and which do not recover their original floristic com-
position after destruction. The same holds true for
those parts of the grassland biome that enter into the
Maputaland-Pondoland-Albany region. For example,
the endemic grassland type “Pondoland coastal plateau
sourveld” is the smallest veld type recognized by
Acocks (1953), and is seriously threatened by overgraz-
ing, sugarcane production, and commercial timber
plantations.
Cultivation practices that threaten and degrade habi-
tats in the hotspot include both large-scale commercial
agriculture and subsistence farming. Subsistence farm-
ing occurs mainly in communal areas, and consists
mostly of shifting cultivation which, while not expand-
ing in the region at the moment, does affect entire
landscapes covering hundreds of square kilometers. In
many of these areas it is impossible to find even a
small portion of the landscape that has not been af-
fected by cultivation. When such areas do exist, they
are often under severe grazing pressure from domestic
livestock. The subtropical areas of this region are also
particularly well-suited to sugarcane production, and
South Africa’s large sugar industry is based entirely
within this hotspot. Consequently, commercial sugar-
cane farming has completely transformed large tracts
of land, especially in the coastal regions north and
south of Durban. Cultivation in South Africa increased
by 122% in area between 1987 and 1994 (Le Roux
2002) and, as human population numbers increase and
the pressure to produce food continues, this trend is
likely to continue into the future, resulting in more
habitat loss.
Industrial timber production is the second largest
cause of habitat loss in the region. Plantations form
very large continuous stands of alien trees along es-
carpment slopes, coastal plains, and midland mist belts,
and are a particularly serious threat to grassland habi-
tats. Several hundred thousand hectares of species-rich
primary grassland in the hotspot have already been de-
stroyed by commercial afforestation, and plans are un-
der way to establish more alien tree monocultures, es-
pecially in parts of the Eastern Cape. The effects of
these commercial tree plantations on biodiversity and
water runoff have been devastating. Besides transform-
ing habitats, the alien trees alter the natural hydrologi- On the opposite page, among a
cal regime by using much more groundwater than the number of important protected
indigenous vegetation and affect the chemical and areas in the Maputaland-
Pondoland-Albany Hotspot is the
physical status of soils. Pine trees (Pinus spp.), euca-
Tembe Elephant Reserve in the
lypts (Eucalyptus spp.), and Australian wattles in the north of KwaZulu-Natal Province in
drier northern areas (Acacia spp.) are the three groups South Africa. Plans are underfoot to
of plantation trees that have the largest negative impact try to link this protected area with
on the grasslands of the region. the Maputo Elephant Reserve in
Urbanization is the third largest threat to natural neighboring southern Mozambique.
© Patricio Robles Gil/Sierra Madre
habitats and vegetation in the region. There are three
major urban centers in the region, namely Maputo in Above, the southern white rhino
southern Mozambique, and Durban and Port Elizabeth (Ceratotherium simum simum)
in South Africa. The growth of cities appears to be an was near extinction at the
unstoppable phenomenon of the modern era, which beginning of the twentieth century,
almost always leads to the formation of unplanned but was brought back from the
brink by what was then called the
sprawling slums on the outskirts of the cities. The
Natal Parks Board. It is now by far
Durban-Pietermaritzburg area in KwaZulu-Natal is one the most abundant rhino species in
of the three largest urban centers in South Africa, and the world —to the point that several
has a population of around three million people. Re- dozen are sold every year to restock
cently, an industrial development zone was designat- other parks and private ranches
ed at Coega, adjacent to Port Elizabeth and, with in South Africa and elsewhere
in the region.
strong political backing, this is likely to promote urban
© Russell A. Mittermeier
and industrial spread.
Almost half of the region is communally owned and
supports livestock numbers far in excess of what is
considered ecologically sustainable. These high live-
stock numbers have caused extensive degradation of
227

Grass crinum (Crinum acaule) is
found in northern KwaZulu-Natal,
mainly on the Maputaland coastal
plain. Much of the habitat of this
species has undergone extensive
afforestation with exotics and
also degradation in some densely
settled areas.
© Patricio Robles Gil/Sierra Madre
On the opposite page, satellite data
has shown that around 20% of the
original vegetation of the
Maputaland-Pondoland-Albany
Hotspot has been irreversibly
transformed through cultivation,
plantations, and urbanization.
In addition, more than half of what
remains has been degraded through
the loss of indigenous woodlands,
soil erosion, overgrazing, and
invasive species.
© Patricio Robles Gil/Sierra Madre
228
the natural rangelands. Vegetation that has not (yet) The existing protected area system does not repre-
been permanently degraded is seriously threatened by sent the biodiversity of the hotspot, in terms of both
overgrazing. In the combined areas of communal and species and the processes required to sustain them.
commercially owned rangelands, overgrazing has de- This is especially true of the Pondoland area, where
graded 25% of the total area covered by the hotspot to only a few small conservation areas are present. Grass-
the point that the vegetation is in a very poor condition, lands, woody grasslands, and coastal forests and thick-
with altered species composition and reduced vegeta- ets are just some of the habitats that are inadequately
tion cover. Extensive invasion by alien plant species protected. A major problem is that the conservation ar-
and localized mining activities (specifically titanium eas of the hotspot have mostly been established with
extraction from coastal sands) are two other threats to the protection of big game in mind, and a number of
natural vegetation in the region. floristically interesting and often unique areas, for ex-
In southern Mozambique, specifically, one of the ma- ample the Noorsveld in the southern parts, therefore
jor threats is the large-scale conversion of trees into still go unprotected.
charcoal to supply the growing demand for firewood for One of the more important private initiatives in the
the larger Maputo Metropolis. Important timber species region is the conservancy program. The conservancy
such as chamfuta (Afzelia quanzensis) are also being concept originated in KwaZulu-Natal in 1978, and in-
harvested extensively from natural vegetation, in many volves the establishment of committees of landowners
cases illegally. who pledge to protect the natural environment, or cer-
tain aspects thereof (they may choose, for example, to
focus on a specific species to look out for and protect)
Conservation on the land they own. There are currently about 218
conservancies in KwaZulu-Natal alone, covering about
An analysis of the World Database on Protected Areas 1.5 million ha, including 167 rural, 38 urban and sub-
reveals that 8.4% of the Maputaland-Pondoland-Albany urban, 4 township, 4 industrial, and 5 marine conser-
Hotspot is conserved in various forms of protected vancies. Members of the conservancy program (cur-
areas. This figure drops only slightly when one con- rently numbering 2 761 members of the public and 372
siders only protected areas classified in IUCN cate- game rangers) attend lectures and participate in vari-
gories I to IV (7.4%). The South African National ous conservation programs.
Parks, a statutory body within the Department of En-
vironmental Affairs and Tourism, manages several na- YOLANDE STEENKAMP 16, 134
tional parks within the area, including the Greater BRAAM VAN WYK 134
Addo Park (24 000 ha) and the Mountain Zebra Na- JANINE VICTOR 13
tional Park (6 536 ha); however, management of most DAVID HOARE 17
protected areas in KwaZulu-Natal falls under the ju- GIDEON SMITH 13, 134
risdiction of Ezemvelo KwaZulu-Natal Wildlife. In TONY DOLD 132
Mozambique, the management of conservation areas RICHARD COWLING 133
falls under the Direcção Nacional de Florestas e Fauna
Bravia (DNFFB) of the Ministério da Agricultura e Pes-
cas. The Greater St. Lucia Wetland Park (256 644 ha)
was declared a World Heritage Site under the World
Heritage Convention of UNESCO, and efforts are un-
der way to establish a transfrontier conservation area
that would link nature reserves in Swaziland, south-
ern Mozambique, and northeastern KwaZulu-Natal.
Numerous other conservation areas —managed by
provincial or local governments or private individu-
als— of varying sizes occur within the hotspot.
Current conservation initiatives include the estab-
lishment of the Baviaanskloof Megareserve, the ex-
pansion of the Greater Addo Park and the Mountain
Zebra Reserve, the wild coast initiative, and the Sub-
tropical Thicket Ecosystem Planning (STEP) project.
The Licuáti Forest Reserve (established in 1943 with
the purpose of protecting woody plant species, partic-
ularly chamfuta) and the Maputo Elephant Reserve
are the primary conservation areas in southern Mo-
zambique. However, law enforcement in these areas is
very poor, and they are not very well protected
(Izidine 2003).
 COASTAL FORESTS
OF EASTERN AFRICA
Scattered along the coastal margins of eastern Africa
are a chain of relict forest and thicket patches set
within savanna woodlands, wetlands, and increasing
areas of farmlands and fallow. These forested areas
are typically tiny and fragmented, but contain re-
markable levels of biodiversity, which often varies
dramatically between forests. During the past twenty
years, studies of the Coastal Forests of Eastern Africa
have resulted in this forest mosaic being recognized
as a globally important conservation priority in a
number of major analyses completed for this region
(Stattersfield et al. 1998; Olson and Dinerstein 1998;
Myers et al. 2000; Burgess et al., in press).
The northern boundary of this hotspot, as defined
here, is located in the remaining small patches of
coastal (riverine) forest along the Jubba and Sha-
belle rivers in southern Somalia (Madgwick 1988;
Clarke 2000a). In Kenya, the hotspot is confined to a
relatively narrow (up to 40 km) coastal strip, except
along the Tana River, where it extends some 120 km
inland to include the forests of the Lower Tana River.
In Tanzania, the hotspot runs along the coast with
coastal forest patches such as Rondo up to 80 km
from the sea. There are also some outliers located up
to 300 km inland at the base of the Udzungwa Moun-
tains of the Eastern Arc mountain chain (Burgess et
al. 1998), and minute patches up to 325 km inland in
northern Mozambique (Timberlake et al. 2004).
The southern boundary of the Coastal Forests of
Eastern Africa is more problematic. White (1983)
previously mapped the vegetation of coastal Somalia,
Kenya, Tanzania, and Mozambique as far south as In-
hambane within his Zanzibar-Inhambane Regional
Mosaic. The World Wildlife Fund (WWF) divided the
Zanzibar-Inhambane Regional Mosaic of White (1983)
into the Northern and Southern Zanzibar-Inhambane
Coastal Forest Mosaic ecoregions, with the boundary
between the Northern and Southern ecoregions in
southern Tanzania and the Southern ecoregion ex-
tending to Maputo in southern Mozambique (Olson
et al. 2001; Burgess et al., in press). Clarke (1998b) de-
fined an entirely new phytogeographical region, the
Swahilian Regional Center of Plant Endemism from
Somalia to northern Mozambique, and a Swahilian-
Maputaland Regional Transition Zone from north-
ern Mozambique to just north of Maputo. For the
0 500 km SOMALIA
KENYA
Lake
Victoria
Lake TANZANIA
Tanganyika
Lake
MALAWI Malawi
MADAGASCAR
MOZAMBIQUE
ZAMBIA
ZIMBABWE INDIAN
OCEAN
purposes of this chapter, the Coastal Forests of East-
ern Africa Hotspot includes the Northern and South-
ern Zanzibar-Inhambane Coastal Forest Mosaic eco-
regions, and the Zambezian Coastal Flooded Savanna
as defined by WWF (Burgess et al., in press), such
that the Limpopo River forms the southern boundary
of the hotspot. This corresponds, roughly, to the
Swahilian Regional Center of Plant Endemism and
the Swahilian-Maputaland Regional Transition Zone
of Clarke (1998b). The hotspot also includes all is-
lands lying immediately offshore, including Zan-
zibar, Pemba, Mafia, and the Bazarruto Archipelago
off Mozambique, such that the total land area covered
by this hotspot is around 291 250 km2.
Much of the region covered by this hotspot was On the opposite page, the
formed from the deposition of marine sediments strikingly beautiful and
since the break-up of Pangea (Clarke and Burgess Endangered Zanzibar red colobus
(Procolobus kirkii) is confined
2000), with younger sediments closer to the coast.
entirely to the island of Zanzibar,
Over the past 30 million years, tectonic activity has re- with its stronghold in and around
sulted in the formation of low ridges and swells, which the Jozani Forest.
has caused the shoreline to move in and out from its © Anup Shah/naturepl.com
present position, with the subsequent uplifting and
erosion of sediments. Faulting associated with the de-
velopment of the Great Rift Valley also affected the re-
gion, with one such fault isolating Pemba Island
through the formation of a deep-water marine trench.
231

The Endangered golden-rumped
elephant shrew or sengi
(Rhynchocyon chrysopygus)
is endemic to the coastal forests of
Kenya north of Mombasa. During
the day, these animals forage on the
forest floor for invertebrates and
spend the night in leaf-litter nests
on the ground. The size of these
animals is about 50 cm long.
© Galen Rathbun
On the opposite page, the crested
guineafowl (Guttera pucherani) is
a wide-ranging species. It is not
globally threatened, but its
population may be affected by
habitat loss and hunting.
© Cristina G. Mittermeier
232
Climatically, the hotspot is largely tropical, although it
is almost subtropical in the southern reaches. The climate
is characterized by high temperatures (23ºC or more
north of the Limpopo River) and high humidity, by inci-
dental sunlight with little seasonal or annual variation,
and variable rainfall. There are two rainy seasons (long,
April-June; short, November-December) in the north,
merging into one (November-April) in the south. Rainfall
ranges from 2 000 mm/year (Pemba and Mafia) down to
500 mm/year in northern Kenya and 800 mm/year in
parts of southern Tanzania/northern Mozambique, al-
though average rainfall in most of the coastal forests is be-
tween 900 and 1 400 mm/year. Dry seasons can be severe
and El Niño effects dramatic. Current daily, monthly, and
annual rainfall fluctuations are thought to be more signif-
icant than those of the Quaternary Period, which appar-
ently were less severe along the eastern African coast
than elsewhere on the continent (Clarke 2000b).
The vegetation of this hotspot is characterized by a
complex mix of moist forests (for example, at the base
of the Eastern Arc Mountains) and drier forests with
coastal thicket, fire-climax savanna woodlands, season-
al and permanent swamps, and littoral habitats. The lit-
toral vegetation includes mangrove vegetation along
some parts of the coast, especially in sheltered bays and
along river mouths. This complexity is partly natural,
although the influence of anthropogenic activities over
thousands of years is believed to have been vital in
shaping the “natural” vegetation now found in the area.
Trees dominate the coastal forest flora, with some of the
more abundant species being Afzelia quanzensis, Albizia
spp., Bombax rhodognaphalon, Combretum schumannii,
Croton spp., Cussonia zimmermannii, Cynometra spp.,
Dialium spp., Diospyros spp., Grewia spp., Hymenaea
verrucosa, Manilkara spp., Millettia stuhlmanni, Nesogor-
donia holtzii, Ricinodendron heudelotii, Scorodophloeus fis-
cheri, Sterculia appendiculata, Sorindeia madagascarien-
sis, Xylia africana, and Zanthoxylum spp. Lianas are also
common as, too, are shrubs, herbs, grasses, sedges,
ferns, and various epiphytes. Most coastal forests are
found up to 500 m above sea level, although in Tanzania
they occur up to 1 030 m on Handeni Hill, but this is
highly unusual (Burgess and Clarke 2000).
The distinction between the coastal forests and the
Eastern Arc Mountains has been a matter of some de-
bate (e.g., Lovett et al. 2000), and J. Lovett (pers.
comm.) has argued that the boundary between the two
forest types is continuous and can not be resolved. A
gradation between the two forest formations is found at
the eastern base of the East Usambara, Uluguru,
Udzungwa, and Nguru ranges. The altitudinal separa-
tion is generally placed between 500 and 800 m (e.g.,
White 1983). Other coastal forests are not contiguous
with mountain forest habitats and are often separated
from the mountains further inland by hundreds of kilo-
meters of drier Zambezian woodlands. Furthermore, the
flora has affinities with that of West Africa, suggesting
an ancient connection with the Guineo-Congolian low-
land forests (Lovett and Wasser 1993; Clarke et al. 2000).
Biodiversity
Studies generally indicate that, for all taxonomic
groups, the region of highest endemism within this
hotspot stretches from northern Kenya to southern Tan-
zania (probably extending into northernmost Mozam-
bique), with other parts being somewhat impoverished
biologically, although additional survey work in Mo-
zambique may yet reveal an unknown wealth of biodi-
versity. At a finer scale, two important subcenters of en-
demism can also be recognized. The first straddles the
border between Kenya and Tanzania —the “Kwale-
Usambara” local center of endemism—, while the sec-
ond is found in southern Tanzania —the “Lindi” local
center of endemism (Burgess et al. 1998; Clarke 2001).
Narrow ranges and disjunct distributions typify the en-
demic species. There is also a huge turnover of species
between forest patches, especially in the less mobile
species. For example, forests that are only 100 km apart
can differ in 80% of their plants (Clarke et al. 2000), and
70% of their millipedes (Hoffman 2000).
The entire Coastal Forests of Eastern Africa Hotspot
contains an estimated 4 050 plant species within around
1 050 plant genera, of which 1 750 plant species and 28
genera are endemic; most of the endemic genera are
monotypic. Around 70% of all endemics (1 225 species)
and 90% of the endemic genera have been recorded from
forest habitats; 92% of all Swahilian endemic species are
recorded from closed canopy vegetation types (forest,
woodland, bushland, and thicket). Indeed, the lowland
forest habitat is the most biologically valuable, with at
least 554 endemic plant species and 18 of the 28 de-
scribed endemic genera confined entirely to it (Clarke et
al. 2000; G.P. Clarke, unpubl.). The forested habitats prob-
ably contain no more than 750 plant genera, given that
their endemic species are distributed among 495 genera.
The non-forested vegetation of the coastal strip of eastern
Africa (i.e., swamp, wooded grassland, coastal margins) is
also important, with at least 812 endemic plants and
members of 10 endemic genera recorded from it. Some
47% of the region’s endemic species have been record-
ed from non-forest vegetation, which covers at least
275 000 km2 of land (0.3 endemic plants per 100 km2
of habitat), whereas the coastal forests cover a total of
6 259 km2 (8.8 endemics per 100 km2 of habitat) (Burgess
et al. 2003b). Clearly, it is the forest patches that have the
highest biodiversity importance per unit area. In addi-
tion, approximately 40% of the endemic plants are con-
fined to a single forest (for example, the Rondo Forest
area in southern Tanzania, has 60 endemic species and
two endemic genera, and the Shimba Hills, Kenya, have
12 endemic species) (Clarke et al. 2000; Clarke 2001).
Similarly, these forest patches are important in terms
of vertebrate diversity and endemism. Birds are repre-
sented by 636 species, of which 12 species are endemic.
Pemba Island, which is considered an Endemic Bird
Area (EBA) by BirdLife International (Stattersfield et al.
1998) contains four endemic bird species: the Pemba
white-eye (Zosterops vaughani), Pemba green-pigeon
(Treron pembaensis), Pemba sunbird (Nectarinia pem-
bae), and Pemba scops-owl (Otus pembaensis). There is
one bird endemic to the Lower Tana River, the Tana Riv-
er cisticola (Cisticola restrictus), and other endemics are
mainly found in the mainland coastal forest remnants
of Kenya and Tanzania, namely the yellow flycatcher
(Erythrocercus holochlorus), Sokoke pipit (Anthus soko-
kensis, EN), Clarke’s weaver (Ploceus golandi, EN), and
Mombasa woodpecker (Campethera mombassica). One
endemic is found in coastal grasslands in Kenya, the
Malindi pipit (Anthus melindae). At the extreme south,
small parts of the Southeast African Coast EBA also fall
within this hotspot, but bird endemics are lacking from
the coastal forest habitats of most of Mozambique.
There are 198 mammal species recorded from this
hotspot, of which 11 are endemic (Burgess et al. 2000).
Endemic mammals include Aders’ duiker (Cephalophus
adersi, EN), Pemba flying fox (Pteropus voeltzkowi, CR),
Kenyan wattled bat (Chalinolobus kenyacola), Dar es
Salaam pipistrelle (Pipistrellus permixtus), golden-rumped
elephant shrew (Rhynchocyon chrysopygus, EN), Tana Riv-
er red colobus (Procolobus rufomitratus, CR), Tana River
mangabey (Cercocebus galeritus, CR), Zanzibar red colo-
bus (Procolobus kirkii, EN), two galagos (Rondo galago, Ga-
lagoides rondoensis, and Diani small galago, G. cocos). In
addition, there is a recently described species of horse-
shoe bat (Rhinolophus maendeleo) described from the
Amboni Caves in Tanga District in Tanzania. These en-
demics are confined to the region from northern Kenya
to southern Tanzania, and northernmost Mozambique.
Among the remaining terrestrial vertebrates, some
250 reptile species are recorded, of which 54 are en-
demic. A single genus, Scolecoseps, is endemic, repre-
sented by three species. In addition, there are 116 am-
phibian species, of which seven are found nowhere
else: the Tana River caecilian (Boulengerula denhardti),
Loveridge’s snouted toad (Mertensophryne micranotis),
Mafia Island toad (Stephopaedes howelli, EN), Loveridge’s
toad (S. loveridgei), Usambara toad (S. usambarensis,
EN), Shimba Hills banana frog (Afrixalus sylvaticus, VU),
and Shimba Hills reed frog (Hyperolius rubrovermicula-
tus, EN). Both the Usambara toad and Loveridge’s toad
are recently described species, the latter representing
an endemic genus. A new genus of Kassina-like frog has
recently been found in the Jozani Forest on Zanzibar
and awaits description (C. Msuya, pers. comm.).
This hotspot has a moderately diverse fish fauna with
219 species. Of the 34 families represented, the min-
nows (family Cyprinidae) are dominant (43 species in
five genera), followed by the killifishes (Nothobranchius
spp.) in the family Aplocheilidae, for which about 75%
of the 33 species present are endemic to the hotspot.
While endemism within vertebrates is moderately im-
pressive, rates of endemism are much higher in inverte-
brate groups such as millipedes (80% of all the forest spe-
cies) and mollusks (68%) (Burgess and Clarke 2000).
Interestingly, in the dragonflies there is a Gondwana relict
species (Coryphagrion grandis) that has its nearest rela-
tives in Central and Southern America (Clausnitzer 2001).
Flagship Species
The mosaic of habitats comprising this hotspot still sup-
port considerable populations of large herbivores such
as elephants (Loxodonta africana, EN), especially in the
larger protected areas and wilderness regions of south-
ern Tanzania and northern Mozambique. Other African
flagship species also occur, including the African wild
dog (Lycaon pictus, EN), lion (Panthera leo, VU), leopard
(P. pardus), and black rhinoceros (Diceros bicornis, CR)
in the Selous. But, despite this megafauna, flagship
species in the coastal forests are typically comprised of
the diurnal and nocturnal primates, some ancient
groups of mammals, endemic birds, and plants.
Among the plants are a few species of African violets
(Saintpaulia spp.). Although these plants are globally
cultivated as house plants, hardly anyone in the devel-
oped world knows that they originate in Tanzanian and
Kenyan forests. There are now 40 000 cultivated vari-
eties of the African violet, forming the basis of a retail
trade of some $100 million, and they all come from just
three of these species (Baatvik 1993). The coastal
forests also contain 11 species of wild coffee, of which
eight are endemic (Clarke et al. 2000). None of these
have been exploited as commercial crops.
The Tana River mangabey and Tana River red colobus
are both restricted to the Tana River region in Kenya. A
1994 survey indicated that there were still some 1 100-
1 300 red colobus and 1 000-1 200 mangabeys in the Tana
River, confined to small patches of gallery forests. The
Zanzibar red colobus has an estimated population of
1 000-1 500 individuals, mainly living in and around the
Jozani Forest, but also in a number of village forests in
close association with people. It is not hunted by the
Muslim inhabitants of this island and has become a sig-
nificant tourist attraction.
Detailed field studies using vocalizations and penile
morphology (Bearder 1999; Bearder et al. 2003) have also On the opposite page, young leopard
revealed unexpected diversity among nocturnal prosimi- (Panthera pardus), a globally
an primates (Grubb et al. 2003). The Zanzibar galago widespread and common species
that is increasingly adapting to live
(Galagoides zanzibaricus) is found in northern coastal
side by side with humans in the
Tanzania, the Rondo galago is found in the southern Tan- outskirts of African and
zanian forests, Grant’s galago (G. granti) in the northern Asian cities.
Mozambique forests, and the Diani small galago ranges © Patricio Robles Gil/Sierra Madre
from northernmost Tanzania into the Kenyan coastal
forests (Bearder et al. 2003). These animals, whose diet Above, Roosevelt’s sable
(Hippotragus niger roosevelti) has
includes the gums of trees, are representatives of an
a patchy distribution in the coastal
ancient group of primates which seems to have its center hinterland of Kenya and Tanzania,
of diversity in the forests of eastern Africa. associated with an extension of
Another ancient mammal group with endemic species miombo-type woodland into Kenya,
in the eastern African forests is the elephant shrew or where this sable subspecies is now
sengi (Macroscelidae). Four species occur, including the found only in Shimba Hills
National Reserve. The Usambara,
endemic golden-rumped elephant shrew in Kenya, so-
Rufiji, and Pangani rivers restrict its
named because of its unmistakable yellow rump and distribution further west and south.
found only in a narrow coastal strip in southeastern © Gerald Cubitt
Kenya, and the black and rufous elephant shrew (Rhyn-
chocyon petersi, EN), found in the coastal and Eastern Arc
Mountain forests of Tanzania. Finally, Aders’ duiker is
confined to woodland in Arabuko-Sokoke Forest (where
235

Spiny reed frog (Afrixalus
fornasinii) in Arabuko-Sokoke
Forest, Kenya.
© David Shale/naturepl.com
On the opposite page, the littoral
vegetation along the coast of East
Africa transitions into mangrove
vegetation, especially in sheltered
bays and along river mouths.
© Gerald Cubitt
236
it is extremely rare and perhaps extirpated) and thicket
forest on Zanzibar Island, particularly the Jozani Forest.
The endemic bird species form the focus of consider-
able interest for local and visiting ornithologists, as well
as numbers of ecotourists in Kenya. The most notable of
these are the Sokoke scops-owl, found in Arabuko-Sokoke
Forest in Kenya and the East Usambaras in Tanzania, and
Clarke’s weaver, which is largely confined to Arabuko-
Sokoke. Other ornithological delights include the Sokoke
pipit and the four endemic bird species on Pemba Island.
An amphibian species of great interest is the endem-
ic Loveridge’s snouted toad, the only member of its
genus. This species is remarkable in that it is one of the
very few amphibian species that breeds by internal fer-
tilization, but lays eggs (instead of giving birth to live
young). It lays its eggs in very small water-filled holes
and crevices (including snail shells).
Fishes living in the temporary swamps and flood-
plains of the coastal areas of this hotspot have evolved
remarkable adaptations to withstand desiccation. The
air-breathing lungfishes Protopterus amphibius and P. an-
nectens can aestivate as adults for over a year at a time
in ‘U’-shaped cocoons underneath the dried-up mud.
The annual killifishes have a very short life cycle, and
grow quickly in food-rich temporary water bodies. They
lay their eggs in the mud and they die when the pool
dries up, but the eggs survive and hatch when the rainy
season begins (Hrbek and Larson 1999).
Threats
The main proximal threats to this hotspot include ex-
panding agriculture, charcoal burning and fuelwood,
uncontrolled fires, unsustainable logging, human set-
tlement, and destructive mining practices (Younge et
al. 2002). All three of the countries of the hotspot are
included among the poorest nations in the world
(World Bank 2002). Basic subsistence activities, such as
agriculture, occupy most of the people living within the
hotspot, except for a small proportion in the larger
cities (Mombasa, Tanga, and Dar es Salaam) and those
engaged in coastal tourism activities (mainly in Kenya).
Poverty, and the consequent reliance on “free” forest
resources, drives much of the forest degradation in the
coastal zone.
The most important threat facing the natural habitats
of eastern Africa, and the coastal forests, is the expan-
sion of agriculture. In general, the soils of coastal east-
ern Africa are poor and can only support subsistence
agriculture. In the past, these soils were ranked as “use-
ful for tree crops or game ranching.” Coastal agriculture
thus involves short-term shifting cultivation concen-
trating on food crops such as cassava, maize, and some
banana, pawpaw, and coconut. The average human
population increase is 2.5%-3.5% per annum, and the
demand for additional farmland is increasing every
year. Plantations of coconut, sisal, and cashew nut also
occupy considerable areas of coastal land, and have
been responsible for the removal of lowland coastal for-
est and other natural habitats.
Charcoal is the major cooking fuel for urban Tanzani-
ans, especially in Dar es Salaam. Burning woody plants
growing in natural vegetation produces most of this
charcoal, hence it is a major cause of habitat loss close to
Tanzanian coastal towns and alongside main roads lead-
ing to them. In places other than towns and roads, this
threat is much less important as local people use fire-
wood for cooking, but the collection of firewood is also
often unsustainable and poses a threat to the forests.
Although fires are a natural phenomenon in eastern
Africa, the majority that occur nowadays are started by
people, who use the fire to clear farmland, to drive an-
imals for hunting, to collect honey, and to reduce tsetse
flies in an area. Fire can invade lowland coastal forest
and thicket patches, and frequent fires can convert this
vegetation to fire-adapted “miombo” woodlands. This
results in a loss of the narrowly endemic coastal forest
specialist species and their replacement by wide-rang-
ing and common species typical of the huge Zambezian
region of eastern and southern Africa.
Logging using pit-sawing techniques occurs in almost
all coastal forests where timber trees remain (many for-
ests have already been logged to exhaustion), regardless
of their protection status. Although some of this logging is
undertaken using licenses obtained from the relevant au-
thorities, much is believed to be illegal. Timber harvest-
ing is occurring in northern Mozambique (Burgess et al.
2003a) and southern Tanzania (from the Rufiji River
southwards) (Milledge and Kaale 2003). This involves the
harvesting of large logs for export as roundwood, reput-
edly to the Far East. Much of this timber is from miombo
woodland elements in the coastal mosaic, but another
part of it comes from the coastal forests themselves. In
Kenya, the large logs are mostly already gone, and timber
extraction concentrates on carving-wood species, particu-
larly Brachylaena huillensis (although much of that used in
Kenya comes across the border illegally from Tanzania).
Despite their low per capita incomes, the countries
of eastern Africa are endowed with a wealth of mineral
resources. In coastal regions these include gas, gem-
stones, iron, titanium, limestone, and kaolin. In coastal
Kenya, Tanzania, and Mozambique, titanium occurs in
coastal sands —the mining of this ore would destroy the
natural vegetation and the species that live there.
Kenyan Government approval has recently been given
to a Canadian-based multinational (Tiomin Inc.) to start
mining titanium in the Kwale area. Other deposits are
under Arabuko-Sokoke forests. High-grade silica sands
for glass manufacture are also mined from deposits in
Msambweni, while iron and manganese are mined on a
small scale in the Kwale Kaya forests of coastal Kenya.
There are also extensive areas of limestone along the
coast. In some other coastal forests of Tanzania (e.g.
Ruvu Forest Reserve, lowland Ulugurus), there are ru-
bies and other precious stones. The mining of these re-
sources has destroyed large areas of natural habitat.
Due to the impact of these threats, total habitat loss in

The Jozani Forest on the island of
Zanzibar in the Indian Ocean is one
of the few remaining forests left on
the island.
© Patricio Robles Gil/Sierra Madre
238
the hotspot is estimated at around 65% of the original
area, based on analyses of land cover maps and the de-
gree of conversion to agriculture or urban land classes
(Burgess et al., in press). The loss of forest habitat is prob-
ably greater than this, and only fragments remain of
what were once large forest patches scattered in climati-
cally suitable positions along the coast (Burgess and
Clarke 2000). As such, we estimate that no more than
10% of the original vegetation can be considered to re-
main intact. Embedded within the remaining natural and
human-made habitats are at least 400 separate patches of
lowland forest, covering around 6 259 km2 (Younge et al.
2002). The remaining forest includes around 2 km2 in the
Jubba River of Somalia, 787 km2 of Kenya, 692 km2 of
Tanzania, and at least 4 778 km2 in Mozambique.
The legal setting of the remaining forest patches
varies dramatically along the coastline. Some forest
patches are found within large protected areas and are
little threatened (e.g., within the Selous Game Reserve
or Sadaani National Park in Tanzania). Other forests are
simply remote and thus largely unthreatened (e.g.,
some of the forests of northern Mozambique and north-
ern Kenya). Many forests are found within government
Forest Reserves managed by District and Central Gov-
ernment forestry authorities, and surrounded by vari-
able densities of rural people who use natural resources
to survive. The most extreme pressures on coastal for-
est habitats are found in the narrow coastal strip of
Kenya, particularly around Mombasa and around Dar
es Salaam. A few forests (e.g. Pugu, Kazimzumbwi,
Vikindu, and Pande) are being engulfed within the
rapidly expanding Dar es Salaam urban metropolis.
Site-specific levels of threat have been assessed for 101
coastal forests in Kenya and 103 coastal forests in Tan-
zania (Younge et al. 2002). Almost all the forests are un-
der some threat, with 57% judged to be highly threat-
ened and 32% very highly threatened. The levels of
threat are very similar in Kenya and Tanzania, but are
believed to be far lower in Mozambique.
Conservation
An analysis of the World Database on Protected Areas re-
veals that protected areas cover approximately 17% of
the hotspot. However, when one considers only those
protected areas that have a higher level of protection
(i.e., those in IUCN categories I to VI), this coverage
drops to 9%, and even lower, to 4%, when only protect-
ed areas in IUCN categories I to IV are included. The two
largest protected coastal forests in Kenya are Arabuko-
Sokoke, with a minimum area of 417 km2 and Shimba
with 63 km2, while in Tanzania there are no protected
coastal forests larger than 40 km2 (Younge et al. 2002), al-
though the Matundu Forest Reserve in the lowlands of
the Udzungwa Mountains contains larger areas of coastal
forest/Eastern Arc transition habitats and some of that is
included within Udzungwa Mountains National Park. In
Mozambique, there are large patches of coastal forest
habitats and the new Quirimbas National Park may con-
tain larger areas of coastal forest than Arabuko-Sokoke.
Within the Kenyan area of the hotspot, there is one
National Park, a 6-km2 area to the northwest of Arabuko-
Sokoke Forest. This “National Park” is, however, some-
what of an anomaly, contains no closed forest, and ex-
ists only on paper. There are also four National Reserves
(Shimba, Tana River, Boni, and Dodori) that fall under
the jurisdiction of the Kenya Wildlife Service (KWS).
Many of the other Kenyan forests are within Forest Re-
serves. As mentioned above, the largest is Arabuko-
Sokoke, which for the last 10 years has been under mul-
ti-institutional management (KWS, FD, Kenya Forestry
Research Institute [KEFRI], and the National Museums
of Kenya, [NMK]). The effectiveness of management
has been variable over time, but has generally been bet-
ter than in the other 17 Forest Reserves within the
Kenyan coastal forest belt. National Monument status
has been given to 39 out of nearly 50 of the sacred Kaya
forests in Kenya (Younge et al. 2002), but the level of
protection gained from this status is below that of the
Forest Reserves. An additional National Monument at
Gede Ruins is not a Kaya. There are numerous Local
Government or County Council Forests, but the protec-
tion of these forests is very poor. A large proportion
(nearly 40%) of the Kenyan coastal forests fall into this
category or are totally unprotected (Younge et al. 2002).
In Tanzania, most closed forests are within Forest
Reserves (80 sites), the majority as national Forest Re-
serves under central government control and a smaller
number as local-authority Forest Reserves under dis-
trict control. On the ground, the little management and
protection that does take place is done by District Nat-
ural Resource staff except in the Tanga Region, where
“catchment” staff also manage some coastal forest sites.
Most of the other sites (20) are unprotected and found
on Village or General Land (Younge et al. 2002). Two
more sites, Zaraninge and the former Mkwaja ranch,
are being incorporated into the new Sadaani National
Park. Some forest and thicket patches are also found in
the Selous Game Reserve, and within the terrestrial por-
tions of Mafia Island and Dar es Salaam Marine Parks.
The coastal forests close to Dar es Salaam face the
biggest conservation challenges. Over the past decade,
significant areas of the Pande Game Reserve (Burgess
and Hipkiss 2002) and Pugu, Kazimzumbwi, and Ruvu
South Forest Reserves have been either heavily degraded
or entirely destroyed by charcoal burning and fuelwood
harvesting. Parts of the Kazimzumbwi and Vikindu For-
est Reserves have also been encroached by farmers who,
despite the interventions of senior government officials,
have remained within the reserves. Further to the south,
in Rufiji District, Namakutwa-Nyamuete, Kiwengoma,
Muhoro, Ruhoi River, and Tongomba-Kilwa Forest Re-
serves, together with three proposed Village Forest
Reserves (Tawi, Mbwara, and Nambunju) and the Kichi
Hills Local Authority Forest Reserve, are being heavily
logged for valuable timber species. Logging is also re-
ported further south in Lindi and Mtwara Districts, with
the majority of the logs being exported from Tanzania as
roundwood. The export of roundwood from Tanzania is a
recent development of questionable legality and, al-
though the government issued a ban on such export
trade in early 2004, the ban has been revoked.
In Mozambique, conservation has become a higher
priority over the past few years, following the cessation
of the civil war in 1992 and a period of reconstruction in
the country. The Quirimbas National Park in Cabo Del-
gado Province was declared in 2002 and covers coastal
forest habitat, in addition to marine areas and extensive
miombo woodlands (Burgess et al. 2003a). Further
south, a large Ramsar site was declared in 2003 within
the Zambezi Delta. The Mozambique Government is
also working to manage its Forest Reserves that cover
some areas of coastal forest habitat, and to map the re-
maining areas of high forest for future management.
The management and protection of forests through-
out the hotspot have suffered from inadequate stake-
holder involvement, conflicts of interest, and corrup-
tion. Where forests are gazetted, the boundaries tend to
be respected, but the forests themselves suffer steady
degradation. The levels of protection achieved on the
ground are strongly dependent on local factors such as
proximity to urban areas, pressure for land, ease of ac-
cess, presence of valuable timber, and the capacity and
morale of local forestry officers. In Kenya, BirdLife In-
ternational (through its local partner, Nature Kenya) has
been involved with conservation of the Arabuko-Sokoke
Forest for over 10 years, and WWF has also provided as-
sistance to the conservation of the Kaya Forests for the
same length of time. In Tanzania, WWF, the Wildlife
Conservation Society of Tanzania, and the Tanzania For-
est Conservation Group have also been working with the
government to conserve important coastal forest patches
for more than a decade, joined more recently by CARE-
Tanzania and the IUCN forest program. Significant fund-
ing for coastal forest conservation has been made avail-
able by the Finnish International Development Agency
(Global Finland) in the East Usambaras and Mtwara Re-
gion, by the Danish International Development Agency
(DANIDA) in Kilwa and Lindi Districts of the Lindi Re-
gion, and by the Global Environmental Facility (GEF) in
Jozani on Zanzibar. Further south in Mozambique, WWF
has assisted the Mozambiquean Government in devel-
oping the new Quirimbas National Park.
In both Kenya and Tanzania there is a general move
towards using various forms of Participatory Forest
Management for forest conservation, which is now re-
ceiving significant funding from DANIDA in Tanzania.
This methodology engages the rural population in the
conservation process in the hope that an exchange of
forest user rights for community management respon-
sibilities and ownership (where appropriate) will lead
to better protection by the people who often know best
what is going on in the forests. Although this hope is
widely held, it has not yet been scientifically tested
within the region, but various projects are exploring the
approach, particularly in Tanzania.
Currently, about $4 million per annum of external
funding is spent within the hotspot on conservation or
related development activities (in addition to that spent
by governments in the region). Bilateral donors are the
largest single source of funding, with a considerable
portion of financial assistance being routed through
conservation NGOs in cooperation with government de-
partments (CEPF 2003). Most government funding is al-
located to salary support, and few funds are available for
conservation activities on the ground. Private invest-
ment for conservation in the coastal forests is also small.
Looking to the future, over the past couple of years a
program coordinated from the WWF office in Nairobi has
brought together representatives from the three coun-
tries supporting large areas of these habitats, Kenya, Tan-
zania, Mozambique. This process has developed a com-
mon vision on the way forward for conservation in the
coastal forests region (Younge et al. 2002; WWF-EARPO
2002). National coastal forest task forces have also been
formed and have assisted in developing an ecoregion-
wide approach to the conservation of these forests. The
national task forces have also assisted in developing con-
servation proposals and sharing experiences.
Recent data gathering, funded through the Critical
Ecosystem Partnership Fund (CEPF), has further re-
fined the conservation priorities in the hotspot and has
brought together a broader network of stakeholders (Na-
ture Kenya and WCST 2003; CEPF 2003). Many of the
NGOs active in conservation in the coastal forests of
eastern Africa are now seeking funds to implement the
findings of these two processes, in collaboration with bi-
lateral aid donors, the government protected area man-
agement authorities, and the local populations living
around the forest remnants. In particular, WWF-Tanza-
nia aims to apply landscape approaches to conservation
so as to enhance the connectivity of fragmented coastal
forest patches in the lowland Usambaras (in partnership
with Tanzania Forest Conservation Group), Gendagen-
da-Msubugwe-Saadani National Park, Matumbi-Kichi
Hills, and Rondo-Noto-Chitoa. Other NGOs active in the
region have similar plans to work on landscape-scale
conservation initiatives in the coastal forest area.
In conclusion, the Coastal Forests of Eastern Africa
Hotspot contains many important patches of lowland
forest habitat, which face great threats from an expand- Charcoal is the major cooking fuel
ing population hungry for land and natural resources. for many urban Tanzanians,
The important forests are primarily managed as Forest especially in Dar es Salaam.
Burning woody plants growing in
Reserves by impoverished forestry departments at the
natural vegetation produces most of
District level, where the primary focus is on income this charcoal, which makes such
generation to support District development projects. activity a major cause of habitat
Conserving these forests and the species they contain loss close to Tanzanian coastal
presents a serious conservation challenge and one that towns and alongside main roads.
will require the input of external funding and expertise © Patricio Robles Gil/Sierra Madre
for some time to come.
NEIL BURGESS 64, 65 IAN GORDON 75, 76
JOHN SALEHE 76 PETER SUMBI 77
NIKE DOGGART 78 ALAN RODGERS 79
G. PHILIP CLARK 80
239
 EASTERN
AFROMONTANE
The montane regions of Africa have always present- 0 1000 km
ed a challenge for the evaluation of biogeography SAUDI
ARABIA
and of conservation priorities. The world’s moun- RED
ERITREA SEA
tain ranges, in general, are clumped —consider, for
YEMEN
example, the Andes and the Himalaya. In contrast,
Africa’s complex geological history has resulted in
SUDAN
its mountains being widely scattered, especially
through the eastern edge of the continent, but with SOMALIA
highly distinctive floras surrounded by other vegeta- ETHIOPIA

tion types. So fragmented is this mountain system

that White (1978, 1983) likened it to islands, calling
DEMOCRATIC UGANDA INDIAN
it the “Afromontane archipelago-like regional center REPUBLIC OCEAN
of endemism.” OF THE CONGO KENYA

Many of these “islands,” including some of the RWANDA Lake

Victoria
largest, are volcanic in origin. Most of the Ethiopian BURUNDI
Highlands are formed of basalt, and the Kenyan High- Lake TANZANIA
lands of various volcanic deposits, including phono- Tanganyika

lite, nephelinite, and basalt. The highlands of the Al- Lake

bertine Rift are largely of Pre-Cambrian rocks, with
islands of volcanic deposits, including those formed
by the still active Virunga Volcanoes. Some of the
more isolated mountains in the region are volcanic,
including Mt. Elgon (4 315 m), Mt. Meru (4 566 m)
and Mt. Kilimanjaro (5 895 m) (White 1983). This vol-
canic and seismic activity was caused by the original
separation of the African and Arabian tectonic plates,
around 35 million years ago, and that resulted in the
formation of the Great Rift Valley system that runs
from Syria to Zimbabwe and Mozambique.
The flora of the Afromontane “archipelago” shows
much uniformity and continuity, its composition
changing with increasing altitude. The lower altitudi-
nal limit is usually taken as between 1 500 and
2 000 m, although this is lower away from the equator,
while the Knysna Forests in the Cape, which are con-
sidered to be the southernmost reaches of the Afro-
montane habitat, are at 300 m. The most widespread
plant genus is Podocarpus (in the broad sense, includ-
ing the recently segregated Afrocarpus), although Ju-
niperus is found in drier forests of northeastern and
eastern Africa. A zone of bamboo is often found be-
tween 2 400 and 3 000 m (Hubbard 1970), above
which there is often a forest zone with Hagenia, ex-
tending as high as 3 600 m (Graham 1960). Many
species common in montane forest, such as trees of
Malawi
MOZAMBIQUE
ZAMBIA MALAWI
MADAGASCAR
ZIMBABWE
the genera Podocarpus and Juniperus, have economic
importance, while several crops including coffee (Cof-
fea arabica) and tef (Eragrostis tef) were domesticated
from the Ethiopian Highlands (Davis et al. 1994).
At the highest elevations, such as the Rwenzori
Mountains, Aberdares, Mounts Kilimanjaro and
Kenya, and Bale and Simien Mountains, Afroalpine On the opposite page, Afroalpine
vegetation occurs, typically from above 3 400 m. vegetation occurs at the highest
Afroalpine vegetation is characterized by the pres- elevations of the Eastern
Afromontane Hotspot (above
ence of giant senecios (Dendrosenecio spp.), giant lo-
3 400 m), such as on the
belias (Lobelia spp.), and Helichrysum scrub. White Rwenzori Mountains, and Bale
(1983) referred to the Afroalpine vegetation as the and Simien Mountains, and is
“Afroalpine archipelago-like region of extreme im- characterized by the presence of
poverishment” because the flora is markedly poor in giant senecios (Dendrosenecio
species. However, endemism is high, possibly due to spp.), giant lobelias (Lobelia spp.
—shown here), and
extended periods of isolation.
Helichrysum scrub.
The greatest biological importance of the Afro-
© Andrew Murray/naturepl.com
montane region derives from three ancient massifs.
The largest of these, the Ethiopian Highlands, lies in
the region’s northern portion, and covers much of
241

On the opposite page, Endangered
chimpanzee (Pan troglodytes) from
the Gombe Stream National Park in
Tanzania, site of Jane Goodall’s
long-term research program.
© Gerry Ellis/Minden Pictures
242
Ethiopia with outliers in Eritrea, Djibouti, and Sudan. It
is bisected by the northernmost extension of Africa’s
Great Rift Valley system. To the southwest lies the Al-
bertine Rift, straddled by the tiny mountain states of
Rwanda and Burundi, and flanked by Uganda and Tan-
zania to the east and the Democratic Republic of Congo
to the west. To the southeast and extending further
south still, the Eastern Arc Mountains go from south-
eastern Kenya to south-central Tanzania, and then are
extended south by the southern Rift mountains of
southern Tanzania and Malawi, with distant outliers in
the Chimanimani Highlands of eastern Zimbabwe, and
Gorongosa of western Mozambique. The geological tur-
moil responsible for the uplift of these mountains has
also resulted in the formation of some of the world’s
most remarkable lakes, with Lakes Tanganyika and
Malawi (Nyasa) abutting the Albertine Rift and South-
ern Rift mountains, respectively. To ensure that these
three remarkable, yet highly threatened, mountain
blocs are treated in sufficient detail, we present each
separately here, but combine them for overall analysis
and consideration as the Eastern Afromontane Hotspot.
Beyond these three main massifs, a number of out-
lying mountains are included in this hotspot. Perhaps
most important are the Neogene Volcanics of the
Kenyan and Tanzanian highlands, a geologically young
formation that includes four of Africa’s five highest
mountains, Mt. Kilimanjaro and Mt. Meru in Tanzania,
and Mt. Kenya (5 199 m) and Mt. Elgon in Kenya. Pos-
sibly because of the relative youth of these mountains,
their endemism is much less marked than that of
the rest of the Eastern Afromontane Hotspot. Far to the
north, the Asir Mountains of southwest Saudi Arabia
and the highlands of Yemen have biological affinities
with the Eastern Afromontane Hotspot, being as they
were at one stage part of the large Ethiopian dome that
began to arise some 75 million years ago (Kingdon
1989). In total, the Eastern Afromontane Hotspot covers
an area of 1 017 806 km2. The Drakensberg Range
of South Africa and Lesotho (the high point being
Thabana-Ntlenyana, 3 482 m), coupled with the Bar-
berton, Wolkberg, Soutpansberg and perhaps Sekhu-
khuneland centers of endemism in southern Africa
(Van Wyk and Smith 2001), form a similar high-altitude
Afromontane outlier to the south, but are, for the mo-
ment, not included in the analysis.
White (1983) also included a number of other regions
within his Afromontane archipelago-like regional center
of endemism, namely the Angola Escarpment, the Jebel
Marra in Sudan (an isolated volcanic massif near the
border with Chad), and the Cameroon and Nimba high-
lands of West Africa, but noted that the inclusion at least
of the Angola Scarp and the high-lying areas west of
Mount Cameroon was problematical due to the presence
of many lowland species. Indeed, we consider the An-
gola Scarp to be too poorly known, at present, for incor-
poration into the hotspot, while the Nimba Highlands
are here considered within the Guinean Forests of West
Africa Hotspot. The Cameroonian Highlands are the
strongest contender for inclusion in this hotspot, but we
retain them in the Guinean Forests Hotspot, restricting
our definition of the Eastern Afromontane Hotspot to
those high-lying regions of the Great Rift.
Overall, the Eastern Afromontane Hotspot holds im-
pressive levels of biodiversity. Its plant diversity sums
to a minimum of 7 598 species, of which at least 2 356
are endemic. For vertebrates, this hotspot holds 490
mammal species, 1 325 birds, 347 reptiles, and 285 am-
phibians, of which 104, 110, 93, and 79 species, respec-
tively, are endemic. Including the Great Rift lakes into
the hotspot also means that it is of phenomenal impor-
tance for freshwater fishes, with a minimum of 893
species occurring, of which at least 617 are endemic.
Further details are given for biodiversity in each of the
sections on the region’s major massifs.
Across the expanse of this complex hotspot, the
threats are largely the same, with habitat loss due to
conversion of land for agriculture, plantations and com-
mercial estates, and logging for commercially valuable
timber species, the primary threat. In addition, a host
of no less insidious factors has led to the loss of much of
the original vegetation in this hotspot, including the im-
pact of fires, mining, infrastructure development, and
collection of firewood and/or plants for medicinal use,
while other threats, such as hunting and disease, have
resulted in serious declines in the populations of many
individual species. Overall, it is doubtful that much
more than 10% of the original vegetation of this hotspot
remains intact.
Rather surprisingly, around 15% of the hotspot is
considered to have some level of official protection,
but this figure is skewed by the inclusion of a number
of protected areas that have limited real protection val-
ue, namely a number of areas in Ethiopia that can not
even be termed “paper parks” because they have still
not been officially gazetted. The true level of effective
protection is likely much closer to the figure of 5.8%,
which includes only those protected areas in IUCN cat-
egories I to IV.
Despite the complex nature of this hotspot, it is be-
lieved that the unification of these unique and floristi-
cally similar montane regions under a single banner
will help to focus increased attention on these threat-
ened habitats. Indeed, the fact that the Eastern Afro-
montane Hotspot is the only “archipelago-like” hotspot
reflects its unique character amongst the list of global
hotspots.
THOMAS BROOKS 2
MICHAEL HOFFMANN 2
NEIL BURGESS 64, 65
ANDREW PLUMPTRE 66
STUART WILLIAMS 67, 68
ROY E. GEREAU 26
RUSSELL A. MITTERMEIER 1
SIMON STUART 12
EASTERN ARC MOUNTAINS
AND SOUTHERN RIFT
Like scattered green islands within an ocean of brown,
the Eastern Arc Mountains and those of the Southern
Rift rise from the African savanna-woodland plains.
These ancient mountain ranges run in a discontinuous
chain from southern Kenya through Tanzania into Ma-
lawi and northern Mozambique. Most of the mountain
blocks are composed of metamorphosed Pre-Cambrian
basement rocks. However, two separate southern moun-
tains have a different geology: Mt. Rungwe is formed
from an ancient volcano, and Mt. Mulanje is the rem-
nant of a giant plutonic intrusion. All the mountain
blocks rise dramatically from the drier African Plateau,
which is the product of hundreds of millions of years of
erosion. Individual mountain blocks have been uplifted
along ancient faults dating back to the breakup of Gond-
wana (more than 180 million years ago), with some up-
lift events occurring at least since the Miocene (about
30 million years ago), and especially over the last seven
million years, associated with the development of the
Rift Valley system (Griffiths 1993).
The Eastern Arc Mountains and Southern Rift region
covers a total of 61 123 km2, and is divided into two
parts based on differences in the climatic regime. The
Eastern Arc Mountains component is by far the biolog-
ically richer and, from north to south, includes the Tai-
ta Hills in Kenya, the North and South Pare, the East
and West Usambaras, the Nguu, the North and South
Ngurus, the Ukaguru, the Rubeho, the Uluguru, the
Udzungwas, and Mahenge in Tanzania. There are also
smaller isolated outliers including Kasigau in Kenya,
and Mafi and Malundwe in Tanzania (Lovett 1988, 1990;
Lovett and Pócs 1993; Lovett and Wasser 1993). The
Southern Rift component is not under the stable Indian
Ocean climatic regime and is biologically poorer. It also
consists of several discontinuous mountain blocks and
plateaus from the Ufipa Plateau and “Southern High-
lands” of Tanzania, along the mountainous margins of
Lake Malawi (Nyasa) to the Mulanje Massif in southern
Malawi. The forests in Ufipa, such as Mbizi, exhibit
strong Congolian and Albertine Rift influences in vege-
tation and in mammalian and invertebrate fauna, and
are thus included in the Albertine Rift. The remaining
part of the Ufipa Plateau is primarily montane grass-
land, more indicative of Southern Rift habitat (T. Daven-
port and A. Rogers, pers. comm.).
The Eastern Arc Mountains rise to 2 635 m at Kim-
handu Peak in the Ulugurus and to a similar height at
Mt. Luhomero in the Udzungwas, although more typi-
cally the maximum altitudes of different blocks are be-
tween 2 200 m and 2 500 m. The Arc Mountains are un-
der the direct climatic influence of the Indian Ocean
(Lovett 1990, 1993), and the climatic regime is believed
to have been stable over millions of years, as indicated
by biogeographical affinities to the forests of West
Africa, Madagascar, and Asia (Hamilton 1982; Lovett
and Friis 1996; Fjeldså and Lovett 1997; Lovett et al.
2004). The Eastern Arc Forests survived the driest and
coldest periods of the last ice ages, as the Indian Ocean
did not cool appreciably and rainfall patterns may not
have been greatly disrupted (Lovett and Wasser 1993).
Today, the climate in the mountains remains wetter,
and less seasonal than the surrounding lowlands. The
Ulugurus receive up to 3 000 mm rain a year on the east-
ern slopes, and annual rainfall on most other mountains
exceeds 1 500 mm in wetter parts. There is, however,
some evidence that the climate has become drier and
more seasonal in recent decades, with a lower likeli-
hood of the forests being enveloped in mist (Hamilton
and Bensted-Smith 1989). Although the mountain soils
are not rich, being old and leached, they are often better
for agriculture than those of the surrounding lowlands.
The favorable climate and moderate soils have attracted
people to the mountains. As such, these areas support
some of the highest population densities in Tanzania.
The eastern-facing slopes of the Arc Mountains are
(or were) forest covered, but most blocks have plateau-
like tops (e.g. Pares, West and East Usambaras, Udzung-
was, and Ulugurus), the higher and colder of which
support montane grasslands/heathlands. The forest
formations of the Eastern Arc Mountains have been di-
vided into upper montane (1 800-2 635 m), montane
(1 250-1 800 m), submontane (800-1 250 m), and low-
land (less than 800 m) (e.g., Pócs 1976). At the lowest
altitudes (generally below 500-800 m, depending on
the block), the forest grades into that more typical of the
lowland Coastal Forests of Eastern Africa Hotspot. At
higher altitudes the canopy height decreases, and
at around 2 400 m altitude the forest generally grades
into Afromontane grassland and heathland plant com-
munities with temperate affinities (Lovett 1993). The
upper altitudinal limit of forest vegetation is deter-
mined by the regular occurrence of frost, and varies On the opposite page, the streams of
between the different mountain blocks. The montane the Udzungwa Mountains are home
forest is characterized by large trees such as Ocotea us- to a number of amphibian species
found nowhere else, among them
ambarensis, Allanblackia stuhlmannii, A. ulugurensis,
the Critically Endangered
Ochna holstii, Podocarpus latifolius, P. falcatus, Ilex mitis, Udzungwa Scarp viviparous toad
Cornus volkensii, Newtonia buchananii, and Synsepalum (Nectophrynoides wendyae),
msolo. In the submontane forests, the timber trees restricted to an area of no more
Khaya anthotheca and Milicia excelsa also become im- than nine hectares, and the
portant. Rubiaceae and Acanthaceae dominate the Critically Endangered Kihansi
spray toad (N. asperginis), which
shrub and ground layers.
occurs only in the two-hectare spray
The Southern Rift is under the climatic influence of zone of the Kihansi Falls in the
Lake Malawi (which is here considered part of the Udzungwa Mountains, but has been
broader hotspot), with some southeast trade wind cli- seriously affected by water
matic elements from southern Africa (Chapman and diversion to the Kihansi Dam.
White 1970). These climatic influences are more vari- © Günter Ziesler
able than those from the Indian Ocean. For example,
Above, African violets (Saintpaulia
climatic instability is indicated by dramatic fluctua- spp.) are well-known flagship
tions in the water levels of Lake Malawi, with major species for the Eastern Arc
recessions before 25 000 years ago and 11 000 years Mountains, with as many as 20
ago, and further large falls 850-750 and 300-150 years ago endemic species.
(Owen et al. 1990). © Patricio Robles Gil/Sierra Madre
Grasslands, commonly attributed to the high fre-
quency and extent of fire, now dominate the natural
245

The silvery-cheeked hornbill
(Ceratogymna brevis) is a
widespread and common species
across Africa.
© Patricio Robles Gil/Sierra Madre
On the opposite page, the
Vulnerable Udzungwa or Uhehe red
colobus (Procolobus gordonorum)
is a beautiful monkey endemic to
the Udzungwa Mountains in
Tanzania’s Eastern Arc Mountains,
where it survives only in a few
small forests.
© Patricio Robles Gil/Sierra Madre
246
habitats of the Southern Rift. Grass species include
Loudetia simplex, Exotheca abyssinica, Monocymbium
ceresiiforme, Themeda triandra, Andropogon spp., Pen-
nisetum spp., and Setaria spp. (Kerfoot 1963). In areas
of impeded drainage, permanent and seasonal bogs
—dominated by grasses and sedges— have rich orchid
floras (Kerfoot 1963; Cribb and Leedal 1982). The for-
ests are mainly found in sheltered valleys and moun-
tain ridges (White 1983; Dowsett-Lemaire 1989) and
contain species such as Newtonia buchananii, Podocar-
pus latifolius, Ilex mitis, and Olea capensis, with Mt. Mu-
lanje containing stands of the conifer Widdringtonia
nodiflora and the endemic Mulanje cedar (W. whytei).
The Southern Rift Mountains mainly rise to between
1 400 and 2 400 m in altitude, but Mt. Rungwe and Mt.
Mtorwi in Tanzania attain 2 961 m, and Mt. Mulanje in
Malawi reaches 3 002 m. Many of the mountain blocks
have plateau-like tops, such as the Ufipa and Nyika
plateaus. The climate regime of these mountains is
largely dictated by Lake Malawi, and for the northwest-
ern areas (Ufipa Plateau), by Lake Tanganyika. Average
annual rainfall ranges from 820 mm at Ufipa to over
2 800 mm in the Livingstone and Poroto mountains of
Tanzania. The mean rainfall of the Malawian Moun-
tains is around 1 500 mm. At the highest altitudes, tem-
peratures as low as minus 7ºC have been recorded, and
frosts are common.
The isolated Mt. Namuli and neighboring formations
in Mozambique have a somewhat ambiguous relation-
ship with the Eastern Arc Mountains and the Southern
Rift regions. Although being under the Indian Ocean
climatic regime, they are not known to possess partic-
ularly high rates of endemism. The Southern Rift also
has similarities to the mountains of the border between
Zimbabwe and Mozambique —including Chimanimani
and Gorongosa (here considered as a southern outlier
of the Southern Rift component). The ancient, complex
biological relationships between these different areas is
illustrated by the plant Necepsia castaneifolia, which
has subspecies in the Eastern Arc Mountains at Kim-
boza; a subspecies in Chirinda, southeast Zimbabwe;
and two subspecies in Madagascar.
Biodiversity
The majority of the biodiversity of this region is con-
centrated in the Eastern Arc Mountains, from the Tai-
ta Hills in Kenya to the Udzungwa Mountains of Tan-
zania (Rodgers and Homewood 1982a, b; Lovett and
Wasser 1993; Burgess et al. 1998; Newmark 2002), an
area that is globally exceptional in terms of its levels
of endemism (Brooks et al. 2002; Burgess et al., in
press).
Current estimates suggest there are over 2 000 plant
species in 800 genera in the Eastern Arc Mountains, of
which at least 800 species (and probably well over
1 000) and around 40 genera are believed to be endem-
ic (Lovett 1998b). There are also high rates of en-
demism in the non-vascular bryophytes, with 32 known
endemics (Pócs 1998). Endemic plants are not only
found in the forests, but also in the montane grass-
lands, wetland areas, on rocky outcrops, and in the dri-
er “rainshadow” (west and north) areas. The Southern
Rift supports at least 1 900 plant species, with 530 on
Mt. Rungwe alone (T. Davenport, unpubl.). However,
the levels of endemism are low, and perhaps only 100
plant species are endemic to the Southern Rift. Re-
duced endemism is most obvious in the forest flora,
whereas the grasslands remain quite rich, especially for
orchids and Protea. As examples, the Nyika Plateau sup-
ports 214 orchid species, with four endemic species and
two endemic subspecies (Kurzweil 2000; Willis et al.
2000). The Kitulo Plateau in the Southern Highlands of
Tanzania supports 350 plant species, three of which are
strictly endemic (Lovett and Prins 1994; Davenport and
Bytebier, in press). A few species of trees and large
shrubs are also endemic to Mt. Mulanje, such as Wid-
dringtonia whytei, Rawsonia burtt-davyi, Ficus modesta,
and Encephalartos gratus (White et al. 2001). In total, we
estimate that about 4 000 plant species in around 1 000
genera occur in the Eastern Arc and Southern Rift
Mountains, of which perhaps 30% of the species and
around 40 genera are endemic. Most of the endemics
are in the Eastern Arc portion.
Among vertebrates, the highest endemism occurs in
the amphibians and reptiles. Some 128 amphibian spe-
cies are recorded, with 22 endemics, mainly in the East-
ern Arc region, although four species are found only on
Mt. Mulanje: Afrana johnstoni (EN), France’s squeaker
(Arthroleptis francei, EN), Broadley’s ridged frog (Pty-
chadena broadleyi, EN), and Broadley’s mountain frog
(Nothophryne broadleyi, EN). The Eastern Arc Moun-
tains are home to 50% of the members of the caecilian
family Scolecomorphidae, among which the genus Sco-
lecomorphus, with three species, is endemic. Besides
Scolecomorphus, the Eastern Arc Mountains and South-
ern Rift are home to an additional seven endemic gen-
era, among them Hoplophryne, Nectophrynoides, Probre-
viceps, Parhoplophryne, Nothophryne, and Callulina; the
last three mentioned comprise single species. In 2002,
a startling discovery was published from the Ukaguru
Mountains: a large, brightly colored tree toad, Churami-
ti maridadi (CR), belonging to a new genus (Channing
and Stanley 2002). New species continue to be discov-
ered, and there are currently seven species of Nec-
tophrynoides under description, while Callulina is in the
process of being split into a number of species (S.
Loader, pers. comm.).
Relatively high rates of endemism are also found in
the other major groups of vertebrates. In the reptiles,
of the 112 species known to occur in the Eastern Arc
and Southern Rift Mountains, there are at least 36 en-
demic species, including eight species of chameleons

Women at the edge of a road in the
Usambara Mountains, Tanzania.
© Patricio Robles Gil/Sierra Madre
248
(six Chamaeleo and two Rhampholeon), three species of
worm snakes (Typhlops), and six species of colubrid
snakes in four genera. Almost all of these are found in
the Eastern Arc Mountains, where endemism is partic-
ularly high for an African mountain system (as cool
and moist habitats are not ideal for exothermic rep-
tiles). Mt. Mulanje also has a number of endemics, in-
cluding the Mulanje mountain chameleon (Bradypodion
mulanjense), Malawi stumptail chameleon (Rhampholeon
platyceps), king dwarf gecko (Lygodactylus rex), and
Mitchell’s flat lizard (Platysaurus mitchelli).
The Eastern Arc and Southern Rift Mountains (as
far South as Mt. Mulanje) are considered to be a single
Endemic Bird Area (EBA) by BirdLife International
(Stattersfield et al. 1998). Of the 639 species known
from this region, 31 are endemic. Four endemic gen-
era (Xenoperdix, Sceptomycter, Modulatrix, and Swyn-
nertonia) also occur. Some bird species have extreme-
ly limited distributions; for example, the Taita thrush
(Turdus helleri, CR) and Usambara akalat (Sheppardia
montana, CR) occur only in a few square kilometers of
forest in the Taita Hills and West Usambaras, respec-
tively. In addition, the Uluguru bush-shrike (Mala-
conotus alius, EN) is confined to one forest reserve on
the Uluguru Mountains, of less than 100 km2 total for-
est area (Burgess et al. 2001). Some bird species have
disjunct distribution patterns covering parts of the
Eastern Arc, the Southern Rift, and the Zimbabwe
Highlands; for example, the monotypic genus Swyn-
nertonia and the long-billed tailorbird (Orthotomus
moreaui).
Mammal endemism is also quite high, with 12 of 204
species endemic, among these three species of pri-
mates —the Sanje mangabey (Cercocebus sanjei, EN),
the Udzungwa red colobus (Procolobus gordonorum,
VU), and the mountain dwarf galago (Galagoides ori-
nus)—, the black-and-red bush squirrel (Paraxerus lu-
cifer), and six shrew species; the desperate shrew (Cro-
cidura desperata, CR), for example, is known only from
the Udzungwa and Rungwe Mountains. Further south,
Vincent’s bush squirrel (P. vincenti, VU) is confined to
Mt. Namuli in Mozambique. A number of new mammal
species have been discovered recently, or are in the
process of being described, including a possible new
species of dwarf galago (Galagoides sp.) on the Taita
Hills (Perkin et al. 2003).
The region contains the very distinctive and highly
endemic fish assemblage of the Lake Malawi Basin, as
well as fishes from upland areas of rivers draining into
the east African coast. Loach catfishes (Amphiliidae),
shellears (Kneriidae), minnows (Cyprinidae), and other
current-loving species tend to dominate the steep
mountain streams in Tanzania, but they have few en-
demic species. In fact, only one of the 90-odd fish spe-
cies and 38 genera occurring in Tanzania, Oreochromis
chungruruensis, is endemic and lives in Lake Chungru-
ru, a crater lake in the Rungwe Mountains, north of
Lake Malawi (Trewavas 1983). In contrast, Lake Malawi
has over 380 fish species, and nearly 90% of them are
endemic. In addition to these cichlids, there are at least
12 large endemic catfishes of the genus Bathyclarias
that live in deeper areas of the lake.
The Eastern Arc Mountains also support an inverte-
brate fauna that is exceptionally rich in endemic
species, although it remains poorly known. Informa-
tion on spiders and millipedes indicate that up to 80%
of invertebrate species (and many genera) may be
strictly endemic to a single mountain (Scharff 1992;
Hoffman 1993). These patterns seem to be repeated
across other invertebrate groups, including butterflies
(De Jong and Congdon 1993). There are 78 butterfly
species endemic to the Eastern Arc (Congdon et al.
2001). In the Southern Highlands, at least seven butter-
fly species are restricted to the Nyika Plateau in north-
ern Malawi, whereas in Tanzania, Neocoenyra petersi is
restricted to the Kitulo Plateau. Among the Odonata,
there are two strictly endemic species in the East Us-
ambaras (Clausnitzer 2001): the Mulanje damselfly
(Oreocnemis phoenix) is an endemic species and genus
to Mulanje and the dragonfly Teinobasis malawiensis is
known only from montane streams in northern Malawi
(Stuart et al. 1990).
Finally, although here considered as a distant outlier
of the Southern Rift, the Chimanimani Mountains and
Mt. Gorongosa area along the Zimbabwe-Mozambique
border supports two endemic birds, Chirinda apalis
(Apalis chirindensis) and Roberts’ prinia (Prinia robertsi).
As such, these mountains comprise their own Endem-
ic Bird Area (Stattersfield et al. 1998). In addition, there
at least two endemic mammals, Arend’s golden mole
(Carpitalpa arendsi) and the Selinda rock rat (Aethomys
silindensis), and five endemic amphibians: the highland
rain frog (Probreviceps rhodesianus, EN), cave squeaker
(Schoutedenella troglodytes, CR), Inyanga frog (Afrana
inyangae, EN), Inyanga toad (Bufo inyangae, EN), and
Chirinda toad (Stephopaedes anotis, EN). Endemic rep-
tiles include the ferocious round-headed worm lizard
(Zygaspis ferox), Zimbabwe girdled lizard (Cordylus
rhodesianus), FitzSimons’ dwarf gecko (Lygodactylus
bernardi), and Arnold’s skink (Proscelotes arnoldi) in
montane grasslands, and Marshall’s dwarf chameleon
(Rhampholeon marshalli) in montane forests. The re-
gion harbors 24 inland fish species, none of which are
endemic, but the fish fauna is fairly distinctive, with
relict populations of the loach catfish Amphilius ura-
noscopus and the shellear Kneria auriculata (Bell-Cross
and Minshull 1988).
Flagship Species
As with the Coastal Forests of Eastern Africa Hotspot,
traditional flagship mammals, such as the elephant
(Loxodonta africana, EN), buffalo (Syncerus caffer), and
leopard (Panthera pardus) are found in this region,
mainly within Tanzania. However, as with the Coastal
Forests of Eastern Africa Hotspot, the critical elements
of biodiversity are smaller and less obvious, and main-
ly found in the plants, primates, smaller mammals,
birds, and amphibians.
The Udzungwa red colobus has a restricted distribu-
tion in forests in the Udzungwa Mountains, where it
survives in riverine, foothill, and montane forest
patches. Although this colobus was once considered
rare, a recent survey has estimated its total population
at around 20 000 individuals (Dinesen et al. 2000). The
Sanje mangabey was discovered in the early 1980s
(Rodgers and Homewood 1982a), and is restricted to
forests of the Udzungwas Massif from 300-1 600 m alti-
tude. The most recent population estimate for this
species is 2 500 individuals (Dinesen et al. 2000). Oth-
er notable mammal species include Abbott’s duiker
(Cephalophus spadix, VU), which is confined largely
(although not entirely) to the montane parts of the
Eastern Arc Mountains and the Southern Highlands in
Tanzania, and the eastern tree hyrax (Dendrohyrax
validus, VU), which also occurs in the Coastal Forests
of Eastern Africa Hotspot. The near endemic black
and rufous elephant shrew or sengi (Rhynchocyon pe-
tersi, EN) is also a notable mammal of the Eastern Arc
Mountains.
Many of the endemic bird and chameleon species
of the Eastern Arc Mountains also attract internation-
al attention. In the birds, attention focuses on the
Udzungwa forest-partridge (Xenoperdix udzungwensis,
EN), which is known from a few forests in the
Udzungwa Mountains and one in the Rubeho Moun-
tains; the Uluguru bush-shrike; and the rufous-winged
sunbird (Nectarinia rufipennis, VU). The Nyika Nation-
al Park on the Nyika Plateau in the Southern Rift sup-
ports the world’s largest breeding population of blue
swallow (Hirundo atrocaerulea, VU) (Dowsett-Lemaire
et al. 2001), which breeds also on Mt. Rungwe, the
Mbeya Range, and the Kitulo Plateau (T. Davenport,
unpubl.).
Of the incredible diversity of amphibians, the genus
Nectophrynoides requires particular mention, since it
is endemic to the Tanzanian part of the Eastern Arc
Mountains and Southern Rift and includes the majori-
ty of viviparous (live-bearing) frogs in the world (the
only other confirmed viviparous frogs are two species
endemic to Mt. Nimba in Liberia and Guinea). One
of these species, the Kihansi spray toad (Nectophry-
noides asperginis, CR), occurs only in the two-hectare
spray zone of the Kihansi Falls in the Udzungwa
Mountains (Poynton et al. 1998), where it has been af-
fected by water diversion to the Kihansi Dam and pop-
ulations may now be under 100 individuals (K. How-
ell, pers comm.). In addition, one other species, the
Udzungwa Scarp viviparous toad (Nectophrynoides
wendyae, CR), appears to be restricted to a tiny area of
no more than nine hectares, on the Udzungwa Moun-
tain escarpment.
Among plants, the best-known flagship species are
the African violets (Saintpaulia spp.), with up to 20 en-
demic species in the Eastern Arc Mountains, and
African primroses (Streptocarpus spp.), with 13 en-
demics. There are also at least 50 endemic species of
balsam (Impatiens spp.) or “bizzy-Lizzy,” and a number
of endemic begonias (Begonia spp.). Orchids, too, are
charismatic, with more than 500 species in the South-
ern Rift (La Croix et al. 1991; Davenport and Bytebier,
in press).
No discussion of flagships would be complete with-
out mention of the amazing diversity of cichlid fishes in
Lake Malawi. Many of these cichlids are strongly site-
attached, rock-loving species known as mbuna, and are
separated by large stretches of unsuitable sandy habi-
tats that create ideal conditions for rapid allopatric spe-
ciation within the lake. Just small changes in body col-
or, which is an important breeding characteristic in
cichlids, can lead to isolated forms (Lowe-McConnell
1993).
Threats
During the colonial period, large areas of mountain for-
est and grassland were converted to commercial es-
tates growing tea, coffee, and pine trees, or to cattle
ranches (Rodgers 1993). More recently, areas of village
forest and Forest Reserve have been cleared to establish
banana, bean, and tree tomato farms to supply lowland
cities with food —for example, in the Uluguru Moun-
tains (Burgess et al. 2002) and the Nguru Mountains (N.
Doggart, pers. comm.). Extensive cardamom planta-
tions are still found beneath the canopy of forests in the
East Usambaras. Natural montane grassland habitats
are increasingly converted to crops such as beans, pota-
toes, and pyrethrum (Chapman and White 1970; Lovett
and Prins 1994). Moreover, large areas of commercial
softwood plantations have been established on former
montane grasslands (for example, the 420-km2 planta-
tion at Sao Hill in Tanzania). The majority of grassland
on Malawi’s second largest plateau, the South Viphya,
has been planted with exotic Pinus spp. (Dowsett-
Lemaire 1989), and other areas of Malawi have been
similarly afforested (Dowsett-Lemaire 1989; McKone
and Walzem 1994).
Outside of protected areas (mainly Forest Reserves) Two young girls inside of what is
and commercial estates, most land is used by villagers left of a forest in the highlands of
for subsistence agriculture. In Tanzania and Malawi, the Usambara Mountains.
© Patricio Robles Gil/Sierra Madre
the mountains support high human population densi-
ties, which are often increasing both through high birth
rates and immigration. Population densities of over
200 people/km2 are common, and in the West Usam-
baras can exceed 400 people/km2. The pressure for
agricultural land from such high human densities is
249
enormous, and occasionally forest areas have been lost.
In one famous example, part of the Shume-Magamba
Forest on the West Usambara Mountains was de-
gazetted from a Forest Reserve soon after indepen-
dence and was then converted to agriculture by land-
hungry residents (Lovett and Stuart 2001). Similarly,
when local people realized in the early 1990s that part
of the northern Uluguru Mountains was not within the
forest reserve, it was rapidly deforested for banana
plantations and subsistence agriculture (Burgess et al.
2002). An analysis of satellite images spanning 20 years
up to 1989 showed that at least 24% of the Kitulo
Plateau had been transformed by cultivation and pas-
ture (Lovett and Prins 1994), a process that continues.
The Eastern Arc Mountains contain commercially
valuable timber species such as Milicia excelsa, Khaya
anthotheca, Beilschmedia kweo, Ocotea usambarensis, and
Podocarpus spp. Many of these species have been
logged on these mountains for more than a century,
and in some parts of the region large specimens are
commercially extinct. An infamous project in the East
Usambara Mountains supported mechanical harvesting
of timber, but caused much environmental damage and
was stopped due to international outcry (Hamilton
and Bensted-Smith 1989; Rodgers 1993). Today, most
timber is extracted using pit-sawing techniques, where
small groups of professional sawyers cut trees into
planks on site and walk out of the forest carrying the
sawn timber. This method is less damaging than mech-
anized logging, but allows all forest areas to be ac-
cessed. Almost all of this harvesting is illegal, as logging
is banned in the Eastern Arc forests, but it supplies the
local market with quality timber. It has proved difficult
for the forestry authorities to eliminate this form of log-
ging, and some claim that forestry authorities are them-
selves involved in the business.
People living on the slopes of the Eastern Arc Moun-
tains and in nearby towns make use of many forest and
grassland resources to augment their subsistence activ-
ities. A particularly important use of the forests is as a
source of firewood for cooking and for heating during
the cold season (Hymas 2000, 2001). Firewood harvest-
ing may be the largest use of the Eastern Arc Mountain
and Southern Rift forests, and one that is particularly
hard to regulate. Other uses of the forests include being
places for hunting, gathering medicinal plants, and tra-
ditional ceremonies —including burials (Mwikomeke et
al. 1998; Shangali et al. 1998).
An important threat to the forest is the many fires
started by people. The enhanced burning regime is be-
lieved to have been the main cause of the replacement
of Afromontane forests with grassland and scrub-grass-
land across large areas (Dowsett-Lemaire 1989). Al-
though there are laws to regulate the burning, there are
also many local beliefs about the benefits of burning,
which are hard to change.
Across different Eastern Arc Mountain blocks there
are variable degrees of artisanal mining for gold, ru-
bies, garnets and, formerly, on the Ulugurus, com-
mercial mining of muscovite mica. Although most of
these activities operate at a low level and have little
impact, ruby mining has destroyed many parts of the
Eastern Arc Mountains-Coastal Forest transition
forests of Ruvu Forest Reserve in the Ulugurus. A re-
cent gold rush has affected the East and West Usam-
baras and the Nguu Mountains. Thousands of people
have flocked to these areas, causing considerable dis-
ruption to local villages and some forest destruction,
mainly along streams.
The above factors have contributed greatly to the
substantial loss of forest in the region. The Eastern Arc
Mountains contain no more than 4 300 km2 of forest
habitat (less than 20% of an original extent of around
23 658 km2 (Newmark 1998). The Udzungwas contain
the largest area of natural forest (just over 1 800 km2). A
number of mountains have lost at least 80% of their
original forest cover, including Taita, Ukaguru, Ma-
henge, and West Usambara. The forests are also highly
fragmented, with mean and median forest patch sizes
estimated at 10 km2 and 58 km2, respectively (New-
mark 1998). At the time that these figures were pub-
lished, there were an estimated 94 forest patches in the
Eastern Arc Mountains, many of which were already
heavily degraded. The habitats of the Southern Rift por-
tion originally covered around 37 465 km2, but at least
70% of this area has been converted to agriculture or
urban areas. In total, then, around 15 539 km2 of origi-
nal habitat remains, or around 25%.
There are also a number of threats that are impact-
ing species directly. For example, a commercial trade
in orchid tubers for food has developed in the South-
ern Rift, with the center of exploitation being northern
Zambia, but now extending across the Southern High-
lands of Tanzania, including Ufipa, the Kipengere
Range, and the Kitulo Plateau. The trade in orchid tu-
bers for consumption in Zambia is threatening as
many as 85 terrestrial orchid species (Davenport and
Ndangalasi 2003). Likewise, a new and recent threat to
the amphibians of the region is the fungal disease
chytridiomycosis, which has led to amphibian extinc- On the opposite page, a steep hill
tions in Central and South America, and Australia. planted with corn in the highlands
This disease has now been confirmed in the Udzung- of the Usambara Mountains.
© Patricio Robles Gil/Sierra Madre
wa Mountains and in other parts of the Eastern Arc
Mountains. Above, a young man bringing down
some planks on the Uluguru
Mountains of Tanzania.
Conservation © Patricio Robles Gil/Sierra Madre
Until now, the main mechanism for conservation in the
Eastern Arc Mountains has been the establishment of
protected areas under the management authority of dif-
ferent parts of the national government, with some
smaller-scale efforts by private enterprise and local
populations. In the Tanzanian portion of the Eastern
251

Clear-cutting and burning in the
Usambara Range.
© Patricio Robles Gil/Sierra Madre
On the opposite page, firewood
harvesting, which is particularly
important for cooking and heating
during the cold season, may be one
of the largest uses of the forests of
the Eastern Arc Mountains and
Southern Rift, and one that is
particularly hard to regulate.
© Patricio Robles Gil/Sierra Madre
252
Arc Mountains, there is one national park (Udzungwa
Mountains National Park, gazetted in 1992; 1 900 km2);
a small area of forest on Malundwe Hill, the only part
of the Eastern Arc Mountains falling within Mikumi Na-
tional Park; one government nature reserve (Amani
Nature Reserve in the East Usambaras); one private na-
ture reserve; a small research reserve owned by the
University of Dar es Salaam at Mazumbai in the West
Usambaras; and some forest within tea estate lands.
The majority of the remaining forest in the region is
within forest reserves. The central government con-
trols the most important water catchment forests as
national forest reserves, most of which are under the
“catchment” project in the following Tanzanian Re-
gions: Kilimanjaro (i.e., North and South Pare), Tanga
(i.e., West and East Usambara, Nguru North), and Mo-
rogoro (i.e., Uluguru, Ukaguru, Nguru South, part of
Rubeho and Udzungwa). National forest reserves in
the Iringa region (Udzungwa Mountains) fall outside the
“catchment” project, but are still managed for water
catchment purposes by central government staff. There
are also some “local authority forest reserves” managed
by the district Natural Resource Offices, but they tend
to be smaller and less important areas. The national
forest reserves, especially those under the catchment
project, are generally better protected than the local au-
thority forests.
There are a few private forests, mainly on tea estates
(e.g., Ambangulu Tea Estate in the West Usambaras and
Brook Bond Estates in the Udzungwas), some of which
are managed for conservation. Many small village for-
ests also exist across the Eastern Arc. Several were tra-
ditionally burial forests, or land that the village was not
using. With the development of the new land laws in
Tanzania (1999) and the new Forest Act (2002), these
areas can now be controlled at the village level, and
many are in the process of being turned into village for-
est reserves.
The Eastern Arc Mountains provide the water sources
for many of the large dams of Tanzania, and for hy-
dropower facilities at Kihansi Gorge and Kidatu in the
Udzungwas, and Pangani River in the Tanga Region. Ki-
hansi Dam was built with World Bank funding, and has
led to the near-extinction of the Kihansi spray toad due
to drying up of its only known habitat (Lovett et al. 1997;
Poynton et al. 1998; Quinn et al., in press). The Ruvu
River, flowing from the Ulugurus, provides the principal
water supply to the 3-4 million people of the capital city
of Dar es Salaam, and for Morogoro. The water supplies
for Mufindi, Iringa, Kilosa, Ifakara, Mpwapwa, Korogwe,
Same, Lushoto, Muheza, and Tanga also originate in
the Eastern Arc Mountains. The same rivers provide the
water for large irrigated rice and sugarcane schemes,
and for sisal estates and other kinds of farms. The value
of these ecological services is not well quantified, but
runs into hundreds of millions of dollars per annum.
The Southern Highlands, meanwhile, contribute to four
of the country’s twelve main drainage basins and pro-
vide water to the majority of people in that part of Tan-
zania.
Tanzania is among the poorest nations in the world
(World Bank 2002), and is being assisted to conserve the
Eastern Arc forests by a number of international agen-
cies. Significant World Bank funding is available to help
build management capacity and develop a Forest
Agency, and the World Bank has also provided $7 mil-
lion to capitalize an Eastern Arc Endowment Fund.
These investments became operational in 2003. The
Global Environmental Facility-United Nations Devel-
opment Programme (GEF-UNDP) is also investing in
the development of a holistic conservation strategy
for the Eastern Arc Mountains, bringing together biodi-
versity, water supply, economics, and poverty-alleviation
elements. One of the aims of the project is to designate
the Eastern Arc Mountains as a World Heritage Site.
Other bilateral donors supporting conservation of the
Arc are the Danish International Development Agency
(DANIDA), Finnish International Development Agen-
cy (Global Finland), and the Norwegian Agency for De-
velopment Cooperation (NORAD), which is providing
long-term support to “catchment” forest reserves. NGOs
also actively pursue conservation in the region, in par-
ticular the Tanzania Forest Conservation Group, World
Wide Fund for Nature, and the Wildlife Conservation
Society of Tanzania. In 2004, the Critical Ecosystem
Partnership Fund approved a $7-million investment
strategy that will provide significant funding for re-
search and conservation within the former Eastern Arc
Mountains and Coastal Forests Hotspot.
Likewise, the governments and populations of the
countries of the Southern Rift are all impoverished,
and hence international efforts are needed to assist
with the conservation of these mountains. Many of the
mountain blocks in Tanzania are heavily populated and
used for agriculture, with most remaining natural veg-
etation confined to government-proclaimed and tradi-
tional reserves (McKone and Walzem 1994; McKone
1995). For example, the Mbeya region of Tanzania’s
Southern Highlands contains 17 forest reserves. A num-
ber of other forest reserves lie within the administra-
tive regions of Rukwa and Iringa. These generally have
low levels of enforcement and are often subject to ille-
gal pit-sawing, fuelwood collection, grazing, agricul-
ture, hunting, and uncontrolled burning. Besides these
official forest reserves, there are numerous smaller tra-
ditional reserves in the Southern Highlands, estab-
lished by local communities for a variety of cultural
reasons. The Government of Tanzania announced in
February 2002 that 13 500 ha of Kitulo Plateau will be
gazetted as a new national park, significantly increas-
ing protection for endemic plants, particularly the
threatened orchid flora (Davenport, 2002).
 The Wildlife Conservation Society (WCS) is assisting
the Tanzania Forest Department and the Tanzanian
National Parks with conservation work in the Tanzan-
ian portion of the Southern Rift. WCS has a long-term
conservation site in the Southern Highlands, working
on research, forest and grassland management, com-
munity conservation, education, and tree planting on
Mt. Rungwe, Kitulo, and other key sites.
The protected area network of Malawi is regarded as
inadequate. Part of the Nyika Plateau is protected in the
Nyika National Park of Malawi and Zambia. Part of
Chipata Mountain is protected in Malawi’s Nkhotakota
Game Reserve (Carter 1987), while Chirobwe Mountain
in the Dedza-Chirobwe Highlands has a forest reserve,
although this is under pressure from wood collectors.
Deforestation is particularly pronounced in Malawi. As
rural population densities are very high, all that re-
mains of the once extensive mid-altitude montane
forests are small relictual groves used as graveyards by
local people (Dowsett-Lemaire 1989). Further south,
Mt. Mulanje is mainly protected as forest reserves with
some forest remaining within tea estates on the moun-
tain. The GEF is working to improve conservation in
Mt. Mulanje, and WCS has recently begun work on bio-
diversity research and monitoring.
In the southern outlier along the border between
Zimbabwe and Mozambique, the higher-altitude veg-
etation types are relatively well protected, in particu-
lar by the Chimanimani National Park (171 km2) and
the Nyanga National Park (440 km2). The Mozam-
bique side of the Chimanimani Mountains is less pro-
tected, and Mt. Gorongosa and Mt. Namuli are pres-
ently unprotected. Zimbabwe was until recently able
to manage well its natural resources, but this has
declined dramatically over the past few years, and
the current situation is unclear. In Mozambique, the
IUCN has assisted the government in re-establishing
the infrastructure of its protected areas following the
end of hostilities in 1992, and progress continues to
be made.
NEIL BURGESS 64, 65
JON LOVETT 84
ALAN RODGERS 79
FELICIAN KILAHAMA 85
EVARIST NASHANDA 85
TIM DAVENPORT 73
TOM BUTYNSKI 86
ALBERTINE RIFT
The Albertine Rift, a series of high mountain chains
that separates the Guineo-Congolian rainforest of Cen-
tral Africa from the forest-savanna mosaic habitats of
East Africa, is an area of exceptional faunal endemism.
This is where the first chimpanzees and gorillas were
studied in the wild, at Gombe Stream and Virunga Vol-
canoes, respectively, and is probably the best place in
the world to see these charismatic flagship species.
While the region is not as rich in plant species as others
elsewhere on the continent, the Albertine Rift is the
most species-rich region for vertebrates in Africa
(Brooks et al. 2001).
Various definitions of the Albertine Rift have been
given by different groups such as BirdLife International
(Stattersfield et al. 1998), the World Wildlife Fund (Ol-
son and Dinerstein 1998), and the Albertine Rift Con-
servation Society (Kanyamibwa and Tumwebaze 1999),
all with a good deal of overlap, but also some differences
in geographical coverage. Since 2001, the protected area
authorities from this region and their NGO partners
have been developing a strategic framework for conser-
vation in the Albertine Rift. As part of this process, it
was agreed that the definition of the Albertine Rift
should be left as wide and as inclusive as possible until
such time as it could be clearly and properly refined.
The current definition adopted by this process, there-
fore, extends from 30 km north of Lake Albert down to
the southern tip of Lake Tanganyika, encompassing the
Rift Valley itself, the lakes in the Rift, and the flanks of
the escarpment and associated protected areas. Accord-
ing to this definition, the area covers about 313 051 km2,
including all natural habitats within 100 km east of the
border of the Democratic Republic of Congo (DRC), and
follows the 900-m contour line in eastern DRC and in-
cludes the protected areas in northern Zambia. The 900-m
contour was selected because there are museum collec-
tions of Albertine Rift endemic birds at the Africa Mu- On the opposite page, the Critically
seum in Tervuren, Belgium, which have been recorded Endangered mountain gorilla
at this altitude (Plumptre et al. 2003a). (Gorilla beringei beringei), an
Albertine Rift primate flagship
The Albertine Rift contains the well-known “Moun-
made famous through the
tains of the Moon” or Rwenzori Massif, made famous by pioneering work of George Schaller
nineteenth-century explorers such as Burton and and Dian Fossey, may benefit
Speke; the Virunga Volcanoes (Virungas), renowned for greatly from plans to expand
their mountain gorillas; Lake Tanganyika, at 1 470 m Virunga Park to include all the
deep, the world’s second deepest lake; and active vol- Virunga Volcanoes region as a
World Heritage Site.
canoes in the Virunga National Park. Volcanic activity
© Anup Shah/naturepl.com
has been pivotal in the formation of the Albertine Rift,
created as it was through a process of upliftment and
volcanism associated with the origins of Africa’s mighty
Rift Valley and famous valley lakes. Large cracks down
the eastern side of Africa, formed by the rotation of the
African continental mass, were filled to form lakes, in-
cluding Lake Tanganyika. This was accompanied by an
upwelling of volcanic material, most spectacularly vis-
ible in the volcanoes within the Parc de Volcans and the
Virunga National Park. Most of the mountain blocks in
255

The tree pangolin (Manis tricuspis)
occurs from Senegal in West Africa
through the lowland rainforest block
of Africa to the Albertine Rift and
western Kenya. Despite its name,
its primary source of food are
termites of the genera Nasutitermes
and Microcerotermes, both found
on the ground.
© Bruce Davidson/naturepl.com
On the opposite page, chimpanzee
(Pan troglodytes) hunting red
colobus monkeys in Mahale
Mountains National Park,
Tanzania. At Gombe Stream, where
hunting behavior was first
described in chimpanzees, it is
estimated that they may kill and eat
more than 150 small- and medium-
sized animals each year.
© Ferrero-Labat/Auscape
256
the Albertine Rift reach altitudes of between 2 000 and
3 500 m, but a number of peaks exceed 4 500 m, with
the highest peak being Mt. Margherita (5 110 m) in the
Rwenzoris, Africa’s third highest peak.
The vegetation of the Albertine Rift is dominated by
montane rainforest, but it undergoes changes with alti-
tude, from the glaciers and rock at the top of the Rwen-
zori Mountains (5 100 m), down through alpine moor-
land (3 400-4 500 m), giant Senecio and lobelia vegetation
(3 100-3 600 m), giant heather (3 000-3 500 m), raised
bogs (3 000-4 000 m), bamboo forest (2 500-3 000 m),
montane forest (1 500-2 500 m), to mid-altitude and low-
land forest (600-1 500 m), savanna woodland (600-
2 000 m), and savanna grassland (600-2 500 m). Papyrus
and Carex wetlands are found around the lakes and
some streams, and the lakes have their own habitat
types varying from rocky and sandy edges to the pelag-
ic zones in their depths. Some particularly unusual habi-
tats include those around the volcanic hot springs and
the peculiar sclerophytic vegetation that colonizes old
lava flows in the Virunga National Park in eastern DRC.
The Albertine Rift experiences a varied climate,
largely as a result of the influence of the high moun-
tains. In the highlands, the climate is temperate, al-
though at lower altitudes it is hot and humid. Average
rainfall throughout the Albertine Rift varies between
1 200 and 2 200 mm per year. There is evidence of rising
temperatures in the region with the retreat of glaciers at
the summit of the Rwenzori Mountains, which has been
documented since the early 1900s (Osmaston 1996).
Humans have been living in the Albertine Rift for
thousands of years. The initial ethnic groups were prob-
ably similar to the current Batwa and Bakonjo, peoples
who lived off the forests like the pygmies in Central
Africa. About 2 000 years ago, “Bantu” peoples moved
into the region from West Africa, and also from the
north, and settled to form various ethnic groups.
Throughout much of the region there is a separation be-
tween pastoralists (Banyamulenge, Hima, Tutsi) and the
many groups of agriculturalists growing crops tradition-
ally. Until the region was colonized by the British, Ger-
mans, and Belgians, there were many kingdoms along
the Rift, a number of which were very aggressive (par-
ticularly in Rwanda and Burundi) and were avoided by
the early explorers as a result. Many of these kingdoms
were divided during colonization, and tribal groups oc-
cur on both sides of the existing international borders.
Biodiversity
Higher plants have been relatively well surveyed in the
forests of Uganda and Rwanda, but elsewhere in the Al-
bertine Rift studies have been patchy. Currently, 5 793
plant species (from 1 537 genera and 233 families) have
been recorded, but this number will rise as surveys are
discovering new species regularly, even within Ugan-
da. The number of plant species is high compared with
regions of similar size, and accounts for 14% of all
mainland Africa’s estimated plant species. The Wildlife
Conservation Society’s Albertine Rift Program has com-
piled a preliminary list of endemic species in the re-
gion, and at present it is estimated that there are 551
endemic species. No plant families appear to be en-
demic to the Albertine Rift, but three genera (Afroligus-
ticum, Micractis, and Rhaesteria) are found nowhere
else. These lists are based on published flora descrip-
tions of plant families, but many families have not been
monographed for the DRC or East Africa or both. As a
result, this list could increase greatly when lower plants
and little-studied growth forms such as climbers, epi-
phytes, lichens, and mosses are included. Virunga Na-
tional Park in eastern DRC and Bwindi Impenetrable
National Park in Uganda have the highest recorded
numbers of plant species, but both sites have been rel-
atively intensively surveyed. Western Tanzania, partic-
ularly around Mahale Mountains National Park, also ap-
pears to be particularly rich in plant species. There are
few records from the Marungu Massif and Itombwe
Massif in eastern DRC, but these regions also could be
relatively species rich.
The Albertine Rift is very rich in vertebrate species,
and is home to more than half of Continental Africa’s
birds and nearly 40% of its mammals (Plumptre et al.
2003a). Indeed, there are more endemic mammals,
birds, and amphibians here than in any other region in
Africa. Only the island of Madagascar has more en-
demic species. A total of 402 mammal species (in 158
genera) have been recorded in the Albertine Rift, of
which 35 are endemic. Most of the endemic mammals
are shrews and rodents, despite the fact that small
mammals have been poorly surveyed throughout much
of the Rift, particularly towards the southern end, and it
is very likely that more species will be added with fur-
ther effort. There are two endemic genera, both mono-
typic, the Ruwenzori shrew (Ruwenzorisorex suncoides,
VU) and Delany’s swamp mouse (Delanymys brooksi).
At least 1 061 bird species (in 368 genera) occur in
the Albertine Rift, of which 4.5% are migratory species
that overwinter in the region and don’t breed here, but
pass through on migrations within the African Conti-
nent. Although this is the most thoroughly surveyed
group of animals, new species for the Rift continue to
be added as migrant species and new range extensions
are recorded. Endemic birds in the region number 41,
with three endemic genera (Pseudocalyptomena, Graue-
ria, and Hemitesia). The currently accepted definition
of the Albertine Rift includes two contiguous Endemic
Bird Areas (EBAs), as defined by BirdLife International,
namely the Albertine Rift and the Eastern Zairean Low-
lands (Stattersfield et al. 1998). It has been argued that
the two EBAs should be merged based on the fact
that old museum collections of Albertine Rift endemic
species show they occur at lower altitudes in eastern
DRC and that they overlap in altitudinal range with the
Eastern Zairean lowland species (M. Herremans, pers.
comm.). The Itombwe Massif, an unprotected area
west of the northern end of Lake Tanganyika, contains
more endemic species than any other site in the Alber-
tine Rift, and is the top priority for conservation in this
hotspot at present.
About 14% of Africa’s reptiles, and 19% of the conti-
nent’s amphibians, occur in the Albertine Rift. As with
other groups, it is likely that these numbers will in-
crease with additional survey work. Sixteen reptile
species (of a total of 175 recorded) and 31 amphibians
(of 146 species) are endemic to the region, with the
Virunga National Park having the highest numbers of
endemic species for both groups. Three amphibian gen-
era are endemic to the Albertine Rift, all represented by
single species: Parker’s tree toad (Laurentophryne par-
keri), the Itombwe golden frog (Chrysobatrachus cupreo-
nitens), and African painted frog (Callixalus pictus, VU).
The former two species are confined to the Itombwe
Massif, although the African painted frog is known from
both the Itombwe Massif and western Rwanda.
The lakes of the Albertine Rift (Albert, George, Ed-
ward, Kivu, and Tanganyika) contain large numbers of
endemic fish species. Although not strictly thought of as
Albertine Rift habitats, these lakes do show a history of
interconnection with one another and also with Lake
Victoria (Snoeks 2000). Lake Tanganyika is home to over
300 fish species, and about 75% of them are endemic.
However, only 10% of Lake Tanganyika’s shore has been
explored and over 1 200 species (vertebrates and inver-
tebrates) have been recorded, making it the second
highest recorded diversity for any lake on Earth (Patter-
son and Makin 1998). Lakes George and Edward have 56
fish species endemic to these two lakes, while Kivu and
Albert have 15 and 6 endemic fishes, respectively. A
conservative estimate of freshwater fish diversity indi-
cates that, together with their surrounding drainages,
the lakes Kivu, Edward, George, Albert, and Tanganyika
harbor over 400 fish species, 274 of which are endemic.
Most of the endemic fishes are cichlids, with 226 en-
demic species and 47 endemic genera. However, a more
recent assessment suggests that these numbers are a
clear underestimate, and that the number of endemics
could be at least 366 species, with around 350 of these
being cichlids (Snoeks 2000). These colorful fish are
very popular with aquarists, and there are many new
species awaiting discovery and scientific description.
Other families that have high levels of endemism in-
clude bagrid catfishes (with two endemic genera), spiny
eels (11 endemic species), and snooks (five endemic
species). The Nile perch (Lates niloticus) is the most in-
famous snook of the six Tanganyika Lates species. Able
to reach a length of two meters, this species is associat-
ed with the extinction of about 200 endemic cichlids fol-
lowing its introduction to Lake Victoria in 1954 (Ogutu-
Ohwayo 1990). The snooks, together with two small
pelagic herrings —Limnothrissa miodon and Stolothrissa
tanganicae—, comprise the bulk of the commercial fish-
eries’ catch in Lake Tanganyika (FAO 2001).
While invertebrate taxa have been poorly studied, this
region is known to have a large number of endemic but-
terflies. The total number of butterfly species found in
the Albertine Rift is unknown, but in Uganda, invento-
ries of the forests in the Albertine Rift show that at least
581 species of butterflies (16% of the estimated 3 630
species in Africa) occur in this part of the hotspot alone
(Howard and Davenport 1996). Given the numbers from
Uganda and Tanzania, it is possible that up to 1 300 but-
terflies might occur in the Rift, about 36% of Africa’s to-
tal. Currently, 117 endemic species from 49 genera are
known from the Albertine Rift (Plumptre et al. 2003a). A
single genus is restricted to the Albertine Rift, namely
Kumothales, while the Ufipa swallowtail (Papilio ufipa) is
found only in Mbizi Forest on the Ufipa Plateau.
The Albertine Rift is important not only for its bio-
diversity, but also for its ecological processes. The sa-
vannas of the Murchison Falls, Virunga, and Queen
Elizabeth National Parks, used to contain the highest
biomasses of large mammals on Earth, at least until the
1960s (Laws et al. 1975; Plumptre and Harris 1995). War
and poaching led to major decreases in the numbers of
large mammals in these parks, but most of these
species are still present and have the potential to re-
cover to their former levels if afforded good protection.
The volcanoes in the Virungas influence the ecology of
a large portion of the Virunga National Park and its sur-
roundings; there are probably unique species associat-
ed with these volcanoes, but few surveys of their fauna
and flora have been conducted.
Flagship Species
The Albertine Rift is steeped in flagship species, but the
mountain gorilla (Gorilla beringei beringei, CR) and
Grauer’s gorilla (G. b. graueri, EN) are the best known.
Made famous through the pioneering work of George
Schaller and Dian Fossey, the mountain gorilla has
been the focus of many wildlife films and Hollywood
movies. Loss of habitat and hunting led to a decline in
the population of mountain gorillas to a low of about On the opposite page, Afroalpine
250 in the Virungas (Sholley 1991), and about 300 goril- flora on Mt. Mikeno in the Virunga
las in Bwindi Impenetrable National Park by the late Volcanoes, Democratic Republic of
the Congo.
1980s (Butynski and Kalina 1998; Butynski 2001). Since
© Gerry Ellis/Minden Pictures
then, better protection has led to increases in both pop-
ulations, with the current estimate at about 380 in the Above, the possibility of expanding
Virungas and 320 in Bwindi. Grauer’s gorilla has not Mahale Mountains National
fared as well over the same time period. As recently as Park on the shore of Lake
1996, it was estimated that there were 16 900 Grauer’s Tanganyika to include other areas
of importance along Lake
gorillas in eastern DRC (Hall et al. 1998), but the civil
Tanganyika and to the east is
war that has raged since 1998 has resulted in major de- currently being investigated. This
clines in some areas (for example, in parts of their area is thought to be particularly
range in the Kahuzi-Biega National Park). Much of the rich in endemic plants, and is also
decline is due to hunting for bushmeat by rebel groups the site of one of the longest-running
and by people mining for gold, diamonds, and coltan research programs on chimpanzee
ecology and behavior.
(columbo-tantalite), a substance used in computer chips
© Ferrero-Labat/Auscape
and cellular phones.
Chimpanzees are found in many of the Albertine
Rift forests, yet quite a number of their populations are
small and unlikely to be viable in the long term unless
corridors among protected areas are maintained. They
259

In the transboundary region of
Uganda, Rwanda, and the
Democratic Republic of Congo, the
volcanic highlands of the Virungas
and Uganda’s Bwindi Impenetrable
National Park shelter the last 670
Critically Endangered mountain
gorillas (Gorilla beringei beringei)
in the world. Chimpanzees, golden
monkeys, forest elephants, giant
forest hogs, African buffaloes, and a
rich variety of avifauna, reptiles,
and amphibians share this
heartland of incredible biodiversity.
© Gerry Ellis/Minden Pictures
260
are probably one of the most important seed dispersers
in a forest and, as such, their loss could seriously affect
forest composition. In addition to chimps, the Albertine
Rift is rich in other primate species, including at least
27 recorded thus far. Fortunately, primate hunting is
rare in Uganda, Rwanda, Burundi, and Tanzania, mean-
ing that these animals are not as threatened here as in
the DRC and West Africa, and are, therefore, usually
easy to see where they occur. As such, they provide
great potential for tourism in the region. Mountain go-
rilla and to some extent chimpanzee tourism have be-
come popular, but the marketing of primate tourism for
other monkeys should be given more emphasis. Both
l’Hoest’s (Cercopithecus lhoesti) and owl-faced guenons
(C. hamlyni) are charismatic species, as is the golden
monkey, (C. mitis kandti, EN), a beautiful guenon with
a soot-black coat and golden-orange mantle across its
back and head, and which is now confined to the Virun-
ga Volcanoes and part of Nyungwe Park.
The Ruwenzori duiker (Cephalophus rubidus, EN),
sometimes considered a subspecies of the black-fronted
duiker (C. nigrifrons), is restricted to the mountains after
which it takes its name, where it occurs at high eleva-
tions, commonly in Hagenia woodland. At the other
end of the size scale is the Ruwenzori otter shrew (Mi-
cropotamogale ruwenzorii, EN), one of only three repre-
sentatives of the family Tenrecidae found on the African
mainland, a family that is otherwise restricted entirely
to Madagascar. The Ruwenzori otter shrew is an aquatic
species, frequenting montane and lowland streams. Its
closest relative, the Mount Nimba otter shrew (M. lamot-
tei), is found in similar habitats in the vicinity of Mt.
Nimba in the Guinean Forests of West Africa Hotspot.
Of the birds present in the Albertine Rift, the Rwen-
zori turaco (Musophaga johnstoni) is probably the most
stunning of the endemic species, with its mantle of iri-
descent green, orange-yellow cheeks, blue back and
tail, and bright red primary feathers. Found in 10 forest
islands of the Albertine Rift, this species is a good flag-
ship for the montane forests. The beautiful, bright
green African green broadbill (Pseudocalyptomena graue-
ri, VU), with its grey-blue throat, is the sole representa-
tive of an endemic genus, and is confined to only three
sites within the Rift. Grauer’s rush warbler (Bradypterus
graueri, EN), while not the most visually exciting, is a
good flagship species for isolated mountain swamps.
This species is confined to small patches of swamp
above 2 000 m and occurs in small populations separat-
ed from each other by large distances. There are sever-
al endemic sunbird species —the regal sunbird (Cin-
nyris regia), Rockefeller’s sunbird (Nectarinia rockefelleri,
VU), Rwenzori double-collared sunbird or Stuhlmann’s
sunbird (Cinnyris stuhlmanni), blue-headed sunbird
(Cyanomitra alinae), and purple-breasted sunbird (Nec-
tarinia purpureiventris)—, and all serve as excellent flag-
ship species, with their brilliant coloration and pres-
ence in most of the Albertine Rift forests.
Of the reptiles, the chameleons are the best flagship
species. Five species are endemic to the Albertine Rift,
including the Rwenzori three-horned chameleon (Cha-
maeleo johnstoni), which reaches a length of 30 cm and
is found in many of the montane forests of the Alber-
tine Rift. The three horns on its head make it look like
a miniature Triceratops, and are used by males to fight
over females. The strange-horned chameleon (Brady-
podion xenorhinus) is similar in size, but has a circular
protuberence on the end of its nose. This species is con-
fined to the Rwenzori Mountains, where it is very rare,
having been overcollected for the wildlife trade.
With a wingspan of 24 cm, the African giant swallow-
tail (Papilio antimachus) is the continent’s largest butter-
fly and is found at many sites in the Albertine Rift.
Three large, conspicuous, yet rare swallowtail butterflies
(P. leucotaenia, P. ufipa, and Graphium gudenusi) serve as
important invertebrate flagships, particularly because
they can help bring attention to those protected areas
where there are fewer of the larger vertebrate flagships.
Finally, the plant genus Impatiens, with 18 endemic
species, comprises important flagships, as species from
this genus have prominent flowers of various shades of
white, pink, and red, and are found in the forest under-
story. The genus is widespread and is much favored by
duikers for food.
Threats
The Albertine Rift has some of the highest human pop-
ulation densities on the African Continent, with up to
750 people per square kilometer in Rwanda and south-
west Uganda. Much of the land in this region has been
converted to agriculture, and the average family size is
6-10 people. As a result, the pressures on the remaining
natural vegetation in this region are enormous, since the
further subdivision of land by families for their children
is impossible and the demand for more land is intense.
Many of the protected areas are fragments or islands in
a sea of humanity, with marked borders between forest
and cultivation. These islands suffer from edge effects
due to the abrupt changes in microclimate and from hu-
man use of the forest edges. People in this region are also
among the poorest in Africa and rely to a great extent on
the environment for their livelihoods. Forests provide
these people with necessary materials such as rope,
bean stakes, firewood, timber, medicines, fruit, bush-
meat, and honey. Fire is also a threat in some regions,
such as on the Ufipa Plateau, where the last remnant of
Congolian forest, Mbizi, is under serious threat from fire.
The peculiar nature of this type of forest-grassland mo-
saic and the geographic relief on which it sits, mean that
there are exposed hill ridges and radiating peninsular ex-
tensions of forest that stretch along the valleys. These
rapidly become isolated by fire and thus more accessible
to human disturbance (Davenport 2002).
A crude estimate can be made of the amount of nat-
ural habitat remaining based on the area remaining as
protected habitat (11.8%; see below) and the authors’
knowledge of zones outside these protected areas.
Based on this, it is likely that no more than 20% of the
habitat can be described as “intact,” while a further
25%-30% is degraded but still relatively natural. An
analysis of forest change over the past 15 years using
satellite images shows that over 1 500 km2 has been
lost to agricultural production in the forested areas of
the Albertine Rift (Plumptre et al. 2003a). This is 0.5%
of the area of the Albertine Rift and 2.2% of its forest-
ed area. Much of this loss has occurred outside pro-
tected areas. In spite of the fact that there is little sup-
port for protection, many forests survive simply
because they have some protected status. However, the
remaining protected areas are experiencing a degrada-
tion of the habitat, and forest use within protected ar-
eas has often increased during times of insecurity. For
instance, the mahogany trees that had been managed
relatively carefully in Uganda since the 1930s have
been felled in the past 20 years because of corrupt
practices and the inability of the forestry staff to con-
trol illegal activities. As a result, many species that re-
quire undisturbed habitat are now confined to a few
small patches within the forest reserves and parks.
The trade in bushmeat is not as developed as in the
Congo Basin, but is probably increasing. There is some
indirect evidence that soldiers are returning from the
Congo to Rwanda and Uganda with a taste for bushmeat.
Poaching in the savanna parks in Uganda and the DRC
is high at present with hippos, buffalos, and the larger
antelope species being targeted. Snares are also set in
forests and lead to death or injury of many species. For
example, in the forests of Uganda, 25% of all chim-
panzees have maimed limbs or are missing hands or
feet as a result of snare injuries (Plumptre et al. 2003b).
Hunters are mainly after duikers and bushpigs in the
forests. Primates tend to be targeted in the DRC and
around Rwenzori Mountains National Park. Most hunt-
ing in Uganda, Rwanda, and Burundi is through the use
of snares or by driving prey into nets with dogs. In the
DRC there is more reliance on the use of guns, espe-
cially with the proliferation of AK-47 assault rifles dur-
ing the civil war there.
Insecurity and civil strife in the Great Lakes region
has led to the degradation and loss of protected areas.
Militia groups have hidden in protected areas and
used them to launch attacks on the nearby inhabi-
tants. Thus, protected areas have not been viewed fa-
vorably by the local communities living adjacent to
them. On the other hand, forests have been places
to which the local people could flee during conflict, so
attitudes vary across the region depending on how
protected areas were used. Several protected areas
have been lost or reduced in size as a result of the
wars, particularly in Rwanda and Uganda. However,
where conservation groups maintained some support
or presence on the ground, protected areas have gen-
erally survived intact. The greatest losses from these
wars have been trained protected-area staff. More than
100 staff members were killed in the protected areas
in eastern DRC in the last six years. Similarly, about a
third of the staff working with the gorillas in Rwanda
were killed between 1990 and 1999. These losses have
rarely been documented, and little support has been
provided to their families because it has always hap-
pened at a time when funding was limited and needs
were great. Sometimes civil wars or insecurity can help
protected areas because they prevent people from en-
tering these areas to carry out illegal activities. It has
been clear that the forests in Uganda suffered most as
the economy started to grow after the wars and the de-
mand for timber for reconstruction grew with it.
Conservation
Currently, about 37 000 km2 of the Albertine Rift is pro-
tected in parks, wildlife-game reserves and forest re-
serves, representing about 11.8% of the total area of the
Albertine Rift. Many of the protected areas are concen-
trated in the north of the region, while fewer in number
and also in total area occur in southeastern DRC and
southwest Tanzania. The largest protected area in the
region is Virunga National Park (8 000 km2), which was
established in 1925; it links the Volcanoes National Park
in Rwanda with Queen Elizabeth, Rwenzori Mountains,
and Semuliki National Parks in Uganda to form a much
larger landscape of protected areas, the Greater Virunga
Landscape, which totals 12 800 km2. A priority-setting
exercise evaluating sites in the Albertine Rift based
upon total species richness and number of endemic
and threatened species identified Virunga National
Park, Itombwe Massif (unprotected as yet), and Kahuzi-
Biega National Park (6 000 km2), in the DRC; Semuliki
(219 km2), Kibale (766 km2), and Bwindi Impenetrable
(331 km2) National Parks in Uganda; and Nyungwe Na-
tional Park (980 km2) in Rwanda as the most important
of the terrestrial sites for conservation. Lake Tanganyi-
ka was also identified for freshwater conservation. Ad-
ditional areas with many endemic and threatened
species but fewer total species were Rwenzori Moun-
tains National Park (996 km2) and Echuya (34 km2) and Gorilla populations in the Rift have
Kasyoha-kitomi (399 km2) Forest Reserves in Uganda; plummeted due to hunting and
Kibira National Park (379 km2) in Burundi; Mt. Kabobo habitat loss. The Critically
Endangered mountain gorillas
in the DRC (unprotected as yet); and Lakes Edward and
plummeted to a low of 250 animals
George (Plumptre et al. 2003a). in the Virungas in the late 1980s
Several protected areas are World Heritage Sites (but have recovered slightly to
—Virunga, Rwenzori Mountains, Bwindi Impenetrable, 380 animals). Reliable estimates for
Kahuzi-Biega National Parks— or Biosphere Reserves the Endangered Grauer’s gorilla
—Queen Elizabeth National Park (2 230 km2), with pro- (Gorilla beringei graueri)
are not available, but it is known
posals currently to expand Virunga Park to include all
that there have been major declines
the Virunga Volcanoes region, namely Mgahinga Goril- in some areas.
la Park (42 km2) in Uganda and Volcanoes National © Bruce Davidson/naturepl.com
Park (150 km2) in Rwanda and their mountain gorillas
under the World Heritage Site listing.
The Albertine Rift has recently become a focus of sev-
eral conservation NGOs, in large part because of its high
vertebrate diversity. Since 2001, a process, supported by
the John D. and Catherine T. MacArthur Foundation, to
develop a strategic framework for conservation and
261

The red-hot pokers of the genus
Kniphofia are a distinctive feature
of the vegetation above 3 000 m.
This one has attracted a Tacazze
sunbird (Nectarinia tacazze) to
feed on the nectar.
© Patricia Rojo
On the opposite page, the
Endangered mountain nyala or
gedemsa (Tragelaphus buxtoni) is
the last large animal to be
discovered in Africa, having been
described in 1910. Endemic to
Ethiopia, it is likely that only 800 to
1 200 mountain nyalas survive
throughout their restricted range.
© Patricio Robles Gil/Sierra Madre
262
joint planning for protected areas has been implement-
ed to bring together NGOs, protected area authorities,
and government ministries in each country. Within this
framework, there is a strong emphasis on involving lo-
cal communities in the management of protected areas.
Uganda, in particular, has been pioneering approaches
to conservation that involve the communities adjacent to
protected areas. Each park or wildlife reserve has a
community conservation warden who holds regular
meetings with the local villagers. There is a process of
sharing revenue from tourists with these communities,
and currently 20% of all gate receipts are put in a fund
for community use (provided it is compatible with con-
servation).
Several protected areas in Uganda and the DRC are
contiguous with one another or linked across interna-
tional boundaries, thus increasing their conservation sig-
nificance. The International Gorilla Conservation Pro-
gramme, a coalition formed in 1991 comprising the
African Wildlife Foundation, Fauna and Flora Interna-
tional, and the World Wide Fund for Nature (WWF), has
been very successful in encouraging coordination and
joint management among Uganda, Rwanda, and the
DRC in the Virunga Volcanoes and Bwindi Impenetrable
National Parks, even when these countries were at war.
Their model has been replicated by the Wildlife Conser-
vation Society (WCS) further north, to include all the
contiguous protected areas linked to the Virunga Na-
tional Park. This “Greater Virunga Landscape” contains
more vertebrates than any other single set of contiguous
protected areas in Africa (Plumptre et al. 2003a). While
many protected areas are islands, the natural habitat in
eastern DRC (forest and woodland) and in western Tan-
zania (woodland) is more intact, and it is still feasible to
manage natural habitat as corridors to link protected ar-
eas. There are possibilities of expanding Mahale Moun-
tains National Park to include other areas of importance
along Lake Tanganyika and to the east. Similarly, it may
be possible to set up linkages between Kahuzi-Biega Na-
tional Park and Maiko to the north and Itombwe to the
south, provided land uses in between are compatible
with nature conservation. Conservation International,
WWF, the Dian Fossey Gorilla Fund (International and
Europe), WCS and the German Agency for Technical Co-
operation (GTZ) are all working in this region to con-
serve biodiversity and to maintain linkages that exist.
Much of the work in this region is supported by the Con-
go Basin Forest Partnership, which was launched in Sep-
tember of 2002, and is an association of 29 governmen-
tal agencies and NGOs working to promote sustainable
management of Congo Basin Forest ecosystems and
wildlife, as well as to improve the lives of people living
there. In Uganda and Tanzania, the World Bank and
UNDP-GEF and the European Union are supporting
much of the conservation of Rift sites. Managing these
areas as larger landscapes will increase the likelihood of
their long-term survival, whereas leaving them as is-
lands means accepting the loss of certain species, par-
ticularly large mammals.
As peace comes to the DRC, the pressures on the nat-
ural habitat will multiply. Logging and mining compa-
nies are already lining up to obtain concessions. There
is a real need to identify which remaining sites deserve
to be protected before they are lost to logging, mining
or agriculture. Itombwe Massif is a clear leading con-
tender for protected area status while, further south,
additional survey work could yet reveal that Mt. Kabo-
bo and the Marungu Massif require protection. The
Tayna Community Reserve to the west of Lake Edward
is also partially established, but needs biological sur-
veys to determine its richness.
What we know already of the Albertine Rift indicates
that it is very important in terms of vertebrate conser-
vation, and yet much of it is still poorly known. Botani-
cally, it may not be as rich as other sites in the world,
but again there is a great need for more survey work,
particularly of herbaceous plants and ferns. Conserva-
tion efforts in this region have to try to balance the pres-
sing human requirements and poverty with the needs of
the wildlife in the ever-shrinking natural habitat of this
region. Many protected areas alone are just too small to
conserve viable populations of the megafauna that still
survive here. There is a need to maintain the linkages
between protected areas where they still exist if the larg-
er vertebrates are to survive in the long term. It will be
harder to lobby for the conservation of these protected
areas if these charismatic megafauna are lost in a region
where human population density and poverty are so
high, and their conservation is a top priority.
ANDREW PLUMPTRE 66
TIM DAVENPORT 73
MATHIAS BEHANGANA 74
ROBERT KITYO 74
GERALD EILU 74
PAUL SSEGAWA 74
CORNEILLE EWANGO 66
CHARLES KAHINDO 74
ETHIOPIAN HIGHLANDS
The Ethiopian Highlands are distinguished from the
rest of Africa by their vast extent of high ground. They
cover an area of some 519 278 km2, almost 95% of
which falls within the political borders of modern
Ethiopia, although also to a lesser degree in neighbor-
ing Eritrea. There are also isolated montane outliers,
including, for example, Jebel Elba and Jebel Hadai
Aweb, parts of which are politically in Egypt but are ad-
ministered by Sudan, and Jebel Ower near Port Sudan.
The geographic and cultural heartland of this region is
a vast plateau, averaging 2 200 m and split into two
halves by the Great Rift Valley. The Ethiopian High-
lands have a lower altitudinal limit of around 1 100 m,
but in many areas the biogeographical boundary be-
tween the Highlands and the neighboring arid zone of
the Horn of Africa is higher, and averages around
1 500 m (Yalden et al. 1996). The cutoff is affected by lo-
cal conditions and variation occurs throughout its
length, although at the northern end of the Highlands it
is lower than at the southern end. The boundary is
somewhat artificial in that there are species that trans-
gress it in both directions, but there is still a clear-cut
separation between highland (Afromontane) flora and
lowland (Somalia-Masai) flora (Friis et al., in press).
The Ethiopian Highlands are thought to have begun
to rise some 75 million years ago. As the Earth’s crust
began to diverge in three plates, volcanoes erupted on
the surface and resulted in an intrusion of trap lavas
that were deposited on the underlying marine Creta-
ceous rock. Between about 45 and 35 million years ago,
the lava was widespread and built up a thick layer of
basalt, up to 3 000 m in some places. During the
Oligocene, the lava deposits folded into an arch or
dome, probably coinciding with the formation of the
Red Sea Rift, and later Oligocene and Miocene lavas
overlaid this arch to produce the high plateau. The
mighty rift that now splits the Ethiopian dome into
the northern and southern massifs began in the Mio-
cene, 13-12 million years ago, and was fully formed by
the Pliocene, 5-4.5 million years ago (Davidson and Rex
1980).
The volcanic activity that dominated the Ethiopian
dome between 45 and 5 million years ago largely pre-
cluded the establishment of a stable fauna and flora.
Thus, it is only in the last 4.5-4 million years that the
Ethiopian Highlands have become habitable. However,
this period of volcanism was followed by severe climat-
ic fluctuations during the Pliocene and Pleistocene; be-
ing a highland area, it was affected by periods of glacia-
tion between 120 000 and 20 000 years ago (although the
Bale Mountains appear to have been glaciated as little
as 14 000 years ago; H. Osmaston and W. Mitchell, pers.
comm.). At this time, the surrounding areas were cov-
ered with open grassland, dry montane forest, and
heath. As the climate warmed, the broad belts of sub-
alpine vegetation contracted, and became restricted to
higher altitudes. Vegetation has only somewhat recent-
ly colonized these areas, although some remain barren,
such as the central peaks area in the Bale Mountains,
where the landscape seen today has resulted from the
lava outpourings modified through a process of erosion
by water, wind, and ice.
Considering this turbulent past, and because the
Ethiopian Highlands are geologically relatively young,
they remain somewhat impoverished in terms of their
fauna and flora. Indeed, for much of their recent histo-
ry, the highlands have been geographically isolated:
the Nile and floodplains of the Sudd, which lie to the
west of the area, were impassable for many potential
colonists from the west, while the majority of the low-
lands that surround the highlands are arid, including
the eastern Sahara to the north, the arid areas of north-
ern Kenya to the south, and the Somali arid zone to the
east. On the other hand, the altitude and isolation of
the highlands have favored speciation of colonists to the
region. These colonists arrived via a number of differ-
ent routes, the most important being the surrounding
dry lowlands, although some tropical species may have
arrived from moist areas in the south and southwest,
passing through the barriers posed by the Kenyan
deserts in the south and the White Nile floodplains in
the west (Kingdon 1990).
Although most species in the region are of Afrotrop-
ical origin, some Palearctic influences are also evi-
dent. During the dry, glacial periods, the jebels and es-
carpments flanking the Red Sea allowed connectivity
with temperate biomes to the north and the Arabian
Peninsula, and a number of Palearctic representatives
achieve their southernmost limit in the highlands.
The Ethiopian Highlands are extremely rugged and
varied, with some regions characterized by steep es-
carpments and deep valleys. Rising to a height of
4 620 m at the summit of Ras Dashen in the scenic
Simien Mountains, the highlands are truly the “Roof of
Africa,” with the majority of land over 3 000 m in Africa
being found in this region (Yalden 1983). Indeed,
around 73% of Sub-Saharan Africa’s Afroalpine ecosys-
tem (which is defined as being over 3 200 m) is found
in Ethiopia.
The altitudinal zonation of the Ethiopian Highlands
is pronounced, so much so that highlanders refer to
each zone in terms of its habitability and the agricul-
ture that can be practiced (see Threats). The foothills
or lower elevations (800-1 500 m) of the Ethiopian
Highlands, known as kolla, support woodland vegeta-
tion which is dominated by Terminalia, Commiphora,
Boswellia, and Acacia species. At slightly higher eleva-
tions (1 500-3 000 m), the vegetation, termed dega or
weyna dega, is dominated by the conifers Podocarpus
falcatus and Juniperus procera. Above 3 000 m, the
Afroalpine ecosystem, known locally as wurch, con-
sists of grassland and moorland with an abundant herb
layer. The Bale Mountains contains the largest patch of
Afroalpine ecosystem (2 067 km2, or 17.5% of all Afro- On the opposite page, the gelada
alpine areas on the continent; S.D. Williams and I. (Theropithecus gelada) lives in the
May, unpubl. data). The treeline is dominated by Ha- high-altitude grasslands of Ethiopia
and is one of the few primate
genia abyssinica and Hypericum revoltum. Above this,
genera endemic to a single country.
the heathland scrub is dominated by heathers such as © Patricia Rojo
Erica arborea. Besides the red-hot pokers of the genus
Kniphofia, a distinctive feature of the vegetation in this Above, another Ethiopian
zone is the giant Lobelia rhynchopetalum, which is par- Highlands endemic is the wattled
ticularly characteristic of Afroalpine vegetation. How- ibis (Bostrychia carunculata). This
gregarious species has been
ever, the flora is not sharply delineated from that of
recorded in flocks of more than a
the ericaceous belt at slightly lower altitudes (Davis et hundred birds in the Bale
al. 1994). Mountains.
At the southern end of the Bale Mountains lies the © Patricio Robles Gil/Sierra Madre
enigmatic Herenna Forest. The altitudinal cline on
which the forest grows has resulted in marked vegeta-
tion belts. The uppermost belt is dominated by Rapanea
and tree heathers, while the moist slopes of the Heren-
na Forest are typified by a shrubby zone of Hagenia and
265

On pp. 266-267, the enigmatic
Herenna Forest, which lies to the
south of the Bale Mountains, shows
a marked altitudinal cline in
vegetation, including an upper belt
dominated by Rapanea and tree
heathers.
© Patricio Robles Gil/Sierra Madre
Above, the spot-breasted lapwing
(Vanellus melanocephalus) is
endemic to the Ethiopian
Highlands; a site of particular
conservation importance for this
species is the Gudo Plain, just west
of Addis Ababa.
© Patricio Robles Gil/Sierra Madre
268
Schefflera growing alongside with the giant lobelias
Lobelia gibberroa. Dense stands of mountain bamboo
(Arundinaria alpina) are also found. Below 2 400 m,
clouds and localized rain support a dense, moist forest,
with trees over 30 m tall, their branches covered with
epiphytes. While the Herenna Forest appears to be rel-
atively impoverished, it does harbor endemic species,
many of which are at the higher altitudes. These in-
clude the Bale monkey (Cercopithecus djamdjamensis), a
little-known endemic primate, and a rich endemic am-
phibian fauna (Largen 2001). The very lowest and dri-
est part of Herenna serves as an example of the sort of
forest that once covered a much larger part of Ethiopia
(Kingdon 1990).
Besides Herenna, the other remaining tract of forest
—the largest within the highlands— is in the Welega,
Illubabor, and Kefa areas of Ethiopia. These forests
share a remarkably small proportion of their species
with similar habitats in East and Central Africa
(Yalden et al. 1996). As noted earlier, the arid and
semiarid belt stretching from southern Sudan to
northern Kenya must, therefore, despite its relative
narrowness (500 km), be an effective barrier to the
forest-dwelling species of the Guineo-Congolian forest
block, and this despite the “stepping stone” provided
by the Imatong Mountains of southeast Sudan (which
contain representatives from the Central African
Forests).
The climate of the area is complicated by influ-
ences from both the Atlantic and Indian Ocean sys-
tems, and at least eight climatic zones are identified
(Gamachu 1977). Rainfall varies from 520 mm in the
north to 2 370 mm in the southwest of the highlands
and occurs in complex uni- or bimodal patterns. Over-
all, the Ethiopian Highlands play a crucial role in cli-
mate control in the entire region of northeast Africa
by attracting large amounts of orographic rainfall
(Hillman 1988). While within the highlands this has
obvious implications for the ecosystems, humans
—numbering in the tens of millions— are dependent
on the water that originates from the Ethiopian High-
lands. Hundreds of streams from the highlands join to
form seven major rivers —the Great Abbai (Blue Nile),
the Tacazze, the Awash, the Wabe Shebelle, the Juba
(in turn, formed from the Web, Genale, Welmel, Du-
mal, and Dawa rivers), the Ghibie and Omo, and the
Sobat (from the Akobo and Baro rivers). The largest of
these rivers have carved out deep gorges, most no-
tably the Tacazze, Great Abbai, and Ghibie that split
the northern dome, and the Wabe Shebelle of the
southern dome.
The Ethiopian Highlands also support a rich and an-
cient cultural diversity; as an example, modern Ethi-
opia harbors some 70 languages. The Ethiopian Ortho-
dox Church was founded in Axum in the fourth cen-
tury; Harar, probably founded in about the eleventh
century, is considered as the fourth holiest Muslim city
in the world.
Biodiversity
The Ethiopian Flora Project was initiated in 1980 (Hed-
berg 1984; Friis and Ryding 2001; and see references in
the Horn of Africa chapter) and has documented the
majority of plant taxa in the greater Horn of Africa re-
gion (the Solanaceae, Lentibulariaceae, Pedaliaceae,
and ferns and fern allies have yet to be included). This
effort has been complemented by an ongoing compila-
tion, review, and assessment of the threatened endem-
ic flowering plants (the Red List Initiative for Plants of
Ethiopia and Eritrea which, to date, has added over 300
taxa to the IUCN Red List; Kelbessa et al. 2003).
The Ethiopian Highlands harbor an estimated 5 200
vascular plant species in an estimated 1 563 genera and
185 families. Of these, 555 species (10.7% of the total)
are endemics, with some groups, the majority of them
associated with the open grasslands, dry woodlands,
and heaths, being very diverse (e.g., the Compositae).
The genus Senecio is particularly diverse, with 12 of the
24 species being endemic. There is only one endemic,
monotypic genus from the area (Nephrophyllum abyssi-
nicum, which is found on heavily grazed pastures, open
ground, and rocky areas on steep slopes between 1 650
and 2 700 m); no plant families are endemic.
Endemism among vertebrates, particularly at the
generic level, is relatively high in this region, especial-
ly when one considers the mammals. Thirty-one of the
193 mammal species in the Ethiopian Highlands are
endemic to the area. Remarkably, there are six endem-
ic genera of mammals, and four are monotypic (three
rodent genera, Megadendromus, Muriculus, Nilopega-
mys; and one primate genus, Theropithecus). The other
endemic genera are Desmomys and Stenocephalemys,
represented by two species each. As with the plants,
these are associated with high-altitude, open grasslands
and dry woodlands.
An estimated 680 species of birds are known to occur
regularly in the highlands and of these, 29 are endem-
ic. Most of the bird species that are endemic to the high-
lands are distributed widely, but five are restricted to
tiny pocket areas in the southern highlands. The latter
region is considered an Endemic Bird Area (EBA) in the
analysis of Stattersfield et al., as is the Central Ethiopian
Highlands, with four species confined to it. There are
four endemic genera, three of which are widespread (Cya-
nochen, Rougetius, and Parophasma), while the fourth
has a very localized distribution in the south of the area
(Zavattariornis). The blue-winged goose (Cyanochen
cyanoptera) is interesting because it seems to have re-
sulted from a chance landfall that has found an amena-
ble environment in the Ethiopian Highlands; the spe-
cies is closely related to the sheldgeese of the alpine
and temperate grasslands of South America. In con-
trast, the Ethiopian bush-crow (Zavattariornis strese-
manni, VU), along with the white-tailed swallow (Hi-
rundo megaensis, VU) and Prince Ruspoli’s turaco
(Tauraco ruspolii, VU), are thought to be relicts caught
at the confluence of four major biogeographic zones at
the southern tip of the highlands.
The amphibian fauna includes six endemic genera
(Sylvacaecilia, Altiphrynoides, Spinophrynoides, Balebre-
viceps, Ericabatrachus, and Paracassina), all of which
comprise single species, with the exception of Paracas-
sina, which is represented by two frog species. There is
also a high level of endemism at the species level (23
species, of a total of 59). The reptilian fauna is less in-
teresting, although of the 80 species present, 15 are en-
demic, including two species of chameleon (Chamaeleo
harennae and C. balebicornutus).
Only 64 fish species occur in Lake Tana and the oth-
er rivers draining the Ethiopian Highlands. Lake Tana
is the source of the Blue Nile, and with a surface area of
over 3 000 km2, is the most prominent freshwater fea-
ture of the Ethiopian Highlands. Nearly a quarter of
fish species are endemic to Lake Tana, including a
loach Nemacheilus abyssinicus and 14 large cyprinid
barbs. Barbus megastoma is one of the largest of a num-
ber of important food fishes and can grow to more than
80 cm, which is unusually large for this genus (Nagel-
kerke and Sibbing 1997).
The number of species in all taxa has been steadily
rising over the past 20 years, meaning that the totals
given here are provisional. The Ethiopian Highlands is
an area where little systematic collecting has taken
place, and many areas, particularly the forests of the
southwest (where expeditions to date have been limit-
ed in duration and poorly equipped), are largely unex-
plored. On the rare occasions that exploratory work is
carried out, it is productive: at least five new species of
small mammals have been described from the Ethiopi-
an Highlands in the last 15 years. As a further example,
the mountain nyala (Tragelaphus buxtoni, EN) was one
of the last large mammals to be described on the
African Continent, in 1910. In conclusion, the final to-
tal of both recorded species and endemics will almost
certainly turn out to be much greater. In addition, the
recognition of the endemic fauna and flora of Ethiopia
requires adequate knowledge of areas of similar ecolo-
gy and history (e.g., the Rwenzori Mountains in the Al-
bertine Rift) to be certain that presumptive Ethiopian
endemics are absent elsewhere (Yalden et al. 1996).
Flagship Species
Almost all the flagship species are confined to the
Afroalpine ecosystem, the open grasslands or the mon-
tane forests. On the high plateau (3 100-4 640 m), the gi-
ant Lobelia rhynchopetalum is instantly recognizable,
reaching a height of nine meters when flowering. Giant
lobelias grow to 2-3 m before sending up a single inflo-
rescence of dark blue-purple flowers. Every few years,
the lobelias have a “musth” year when, for unknown
reasons, the greater proportion of the plants flower.
The inflorescence is hollow and has several thousand
flowers. Each flower produces several thousand tiny
seeds. One inflorescence can, therefore, produce over
seven million seeds! Once the plants have flowered,
they die —although the dead plant “skeletons” last for
several years and are characteristic of this zone.
At these altitudes, plants face two main challenges:
the high levels of solar irradiation, and the extremes in
temperature and wind. The young, sensitive leaves of
giant lobelias are protected from the strong sunlight by
always being vertical. The older leaves, which have a
non-photochemical quenching mechanism for protec-
tion against ultraviolet irradiation, are horizontal.
Young lobelias protect themselves against the extremes
in temperatures by forming a “nightbud,” tightly closing
their leaves at night about the apical meristem. The
young, sensitive leaves are also furry, which insulates
them. The overall anatomy of the leaves —a circular,
rosette form— acts as a parabolic reflector for the apical
meristem to warm it and optimize growth. In older
plants, the old leaves hang down to protect the stem,
which is full of water. Their stems also have thick cork
layers, again insulation to prevent water in the stem
from freezing. The older leaves, in turn, can withstand
temperatures down to –6ºC.
The Ethiopian wolf (Canis simensis, CR) is a rare en-
demic also found in the Afroalpine ecosystem. With
fewer than 450 individuals remaining in seven small
and isolated populations, the Ethiopian wolf is the
rarest canid in the world. Initially considered to be of
Afrotropical origin as a specialized derivative of the
common jackal (Yalden and Largen 1992), it is now re-
solved to be of Paleartic origins. Genetic work has
shown that it is most closely related to the grey wolf
(Canis lupus), from which it diverged an estimated
100 000 years ago (Gottelli et al. 1994). Although
Ethiopian wolves are solitary hunters, specializing on
diurnal rodents, they are social animals, living in packs Rouget’s rail (Rougetius rougetii) is
of up to 13 adults that are dominated by an alpha breed- an Ethiopian Highlands endemic,
ing pair. All pack members actively help to rear the and is particularly characteristic of
the moorlands of Ethiopia. The
young, despite the uncertainty of paternity that may
species mainly lives at elevations of
occur through extra-pack copulations solicited by the up to 4 100 m, where it inhabits
alpha female (Sillero-Zubiri and Gottelli 1995a, b; Sillero- small pockets of grass and wet
Zubiri et al. 1996). hollows with plenty of cover.
The walia ibex (Capra walie, CR) is another Palearctic © Patricio Robles Gil/Sierra Madre
species which, despite the presence of other, charismat-
ic flagships in the region and its close relationship with
the widespread Nubian ibex (C. nubiana), has become a
symbol for wildlife in Ethiopia —mainly because of the
interest of trophy hunters and explorers who arrived
first in the Simien Mountains, which are their only re-
maining refuge.
Three charismatic highland mammal species have
Afrotropical origins. The first, the mountain nyala,
was once widespread two to three million years ago,
but their numbers have declined because of agricul-
tural expansion and killing, and today less than 3 000
individuals remain, the majority of which are found in
the Bale Mountains. The giant mole-rat (Tachyoryctes
269
macrocephalus) is found only in the Bale Mountains and
forms the main prey of the Ethiopian wolf. These mole-
rats are solitary, but their wide-ranging burrow systems
overlap with those of other individuals. Each burrow
system has over 90 m of tunnels, covering an area of up
to 400 m2. In some areas, they reach staggering densi-
ties, with about 6 000 mole-rats/km2 (Sillero-Zubiri et
al. 1995).
Unlike the mountain nyala and giant mole-rats, the
majority of geladas (Theropithecus gelada) are found in
the northern highlands. Gelada is the Amharic name
for this species, which occupies a unique environmen-
tal niche as the only graminivorous primate species.
Like other graminivores, they require a relatively large
intake of leaves, stems, and rhizomes to fulfill their nu-
tritional requirements and consequently spend long
periods of the day grazing. The basic unit of gelada so-
cial structure consists of one reproductive male and up
to ten reproductive females and their young. These
units share a common home range and typically forage
together as a “band.” In turn, bands often aggregate to
form foraging herds that can number as many as 600 in-
dividuals, although the composition of these large for-
aging groups is remarkably flexible.
Prince Ruspoli’s turaco stands out as the prominent
avian flagship of the highlands: it is attractive, threat-
ened, and associated with a declining habitat. The
species is an arboreal frugivore, feeding largely on figs,
as well as Podocarpus and Juniperus fruits, and is usual-
ly observed alone or in groups of up to 11 individuals.
The melodious song of the Abyssinian catbird (Paro-
phasma galinieri) is characteristic of the woodlands
throughout much of the Ethiopian Highlands. Pairs of
birds call at dusk, particularly during the rainy season.
Around wetlands —streams, bogs, and marshes— in the
more open habitats and the Afroalpine ecosystems,
Rouget’s rail (Rougetius rougetii) is a common sighting.
The rail has a conspicuous white tail, which it flicks fre-
quently when disturbed. In contrast with the above
species, which are fairly widely distributed, the charis-
matic Ethiopian bush-crow, first reported in 1938, has a
very confined distribution at the southern end of the
highlands. It is a gregarious species, moving in flocks of
up to 30 birds, and is thought to be a cooperative breed-
er, with three birds attending to a nest.
Threats
The threats to the Ethiopian Highlands are under-
pinned by high human population pressure. Over the
past 60 years, the population of Ethiopia has increased
ten fold (from seven million in 1940 to an estimated 70
million in 2004). Eighty percent (56 million) of the
country’s population live in the highlands. This has put
land, both for agriculture and for livestock husbandry,
at a premium. Of the farmlands in the country, 94% are
operated by seven million smallholders cultivating an
average of less than one hectare. People are pushing
the limits to which agriculture is practiced. In some ar-
eas, land is being tilled for barley production on the
steepest slopes (sometimes in excess of 45º) at alti-
tudes of up to 4 100 m.
The exploitation of the Ethiopian Highlands by hu-
mans is not a modern phenomenon. Indeed, it has
been estimated that it has been ongoing for thousands
of years, particularly to the west of the Rift Valley, and
this has destroyed most of the natural vegetation, in-
cluding a great part of the forests. Indeed, as mentioned
earlier, highlanders even refer to each vegetation zone
in terms of its habitability and the agriculture that can
be practiced there: wurch (Afroalpine; more than
3 000 m; too cold to be habitable; no agriculture); dega
(temperate; 2 300-3 000 m; barley, wheat, potatoes,
pulses); weyna dega (warm temperate; 1 500-2 300 m;
tef, maize, wheat, pulses); and kolla (tropical; 800-
1 500 m; sorghum) (and with bereha being the hot and
dry lower altitudinal areas at less than 800 m; no rain-
fed cultivation).
A suite of plants, whose maximum productivity lies
between 1 800 and 2 100 m, were domesticated histori-
cally in the Ethiopian Highlands, which includes their
centers of diversity and origin. They include khat
(Catha edulis), ensete (Enset ventricosum), noog (Guizo-
tia abyssinica), finger millet (Eleusine coracana for
beer), tef (Eragrostis tef) and coffee (Coffea arabica)
(Harlan 1992). The exact date and location for the do-
mestication of all these plants are unknown. On the ba-
sis of linguistic, historical, geographic and botanical
studies, there is no doubt that, with some variation,
they are very ancient crops and most authors put the
date at between 6 000-3 000 years ago.
Besides agricultural crops, Ethiopia has the largest
national herd of domestic livestock, and cattle in par- On the opposite page, this giant
ticular, in Africa. In part, the number of cattle in the lobelia (Lobelia rhynchopetalum)
country might result from the absence of wood (which is a characteristic species of some
Afromontane regions, reaching a
has been previously removed through human exploita-
height of as much as nine meters
tion for fuel and construction), as the majority of when flowering. These plants have
Ethiopian highlanders use cattle dung as their principal special adaptations to help them
source of fuel. The livestock is increasingly using the survive at such high altitudes, with
more extreme areas to graze. In 2002, the livestock in a older leaves able to withstand
discrete area of the Bale Mountains reached an un- temperatures to –6ºC.
© Patricio Robles Gil/Sierra Madre
precedented density of 314 animals/km2. Besides the
effects of erosion and increasing the abundance of un- Above, the rarest canid in the
palatable or poisonous species, overgrazing also height- world, the Ethiopian wolf (Canis
ens competition between livestock and wildlife species. simensis), which is Critically
In addition, livestock and the domestic dogs that often Endangered, now numbers less
accompany it increase the risk of disease transmission than 450 animals, and is threatened
by loss of habitat to agriculture and
to wildlife species. Two rabies epidemics in the past 14
grazing, diseases such as rabies,
years have occurred among Ethiopian wolves by trans-
hybridization with domestic dogs,
mission from domestic dogs, and this serves as a con- and human persecution.
stant reminder of the seriousness of this threat. Dogs © M. Harvey/DRK PHOTO
also pose a further, insidious threat to wolves through
hybridization.
Finally, humans have hunted and killed birds and
mammals, reducing their populations to a fraction of
271

This local elderly woman in Debre
Libanos relies on whatever bits and
pieces she can carve off old stumps.
© Patricio Robles Gil/Sierra Madre
272
what they were 150 years ago. The killing of animals
has not just been for subsistence use or potentially as a
buffer during famines. During (frequent) political up-
heavals in the region, the infrastructure of the national
parks has been successively used and then destroyed
by armed groups, who also kill animals for food. Fur-
thermore, because the national parks and wildlife pop-
ulations held within them have been largely associated
with repressive regimes (particularly the “dergue,” the
military-Marxist regime of Mengistu Haile Mariam in
Ethiopia), the population vented suppressed anger by
destroying park infrastructure and slaughtering large
mammals (Yalden et al. 1996).
The sum of these factors has resulted in a massive
transformation of the environment, and it is estimated
that as much as 97% of the original vegetation has
been lost. Because human exploitation is linked to it,
altitude has also had a profound effect on the extent of
the original vegetation that remains. The original veg-
etation that remains only does so because it is confined
to the ecosystems that are extreme and defy human
use. These are the steep escarpments of the Rift Valley
and the river gorges, the cold Afroalpine plateaus, and
a few patches of thick forest. Consequently, several key
areas of the remaining original vegetation emerge as
being critically important to the biodiversity of the
Ethiopian Highlands. These are obviously very limited
in size, as they are not only geographic islands above
the surrounding lowlands, but also islands in a human-
transformed environment.
Conservation
The degree to which the natural vegetation and animal
populations have been lost means that the region’s di-
versity is acutely threatened. In Ethiopia, despite hav-
ing a wildlife conservation organization that has been
active for just under 40 years, as well as a succession of
foreign advisors and periodic injections of donor assis-
tance, there has been remarkably little impact on ar-
resting the decline of many habitats and species.
While modern conservation efforts struggle to be
successful, the Ethiopian Highlands contain the oldest
records of conservation efforts on the continent. The
Emperor Zera Yacob (1434-1468) noted the loss of forest
cover on what is now known as Wuchacha Mountain.
The forest was replenished at his orders using seeds
and seedlings of Juniperus procera to create Menagesha
Forest, which stands today (Gilbert 1970).
More remarkably, although not quite as old, in the
Guassa-Menz area of North Shoa, Ethiopia, local com-
munities implemented a sustainable natural resource
management system in the seventeenth century. The
system, known as Qero, allowed equitable use and dis-
tribution of natural resources (thatching grass, fuel-
wood, and grazing) that were, and still are, important
for the livelihood security of the community. By regu-
lating exploitation of the area, the management system
has also effectively protected the biodiversity of the
Afroalpine ecosystem of the Guassa-Menz area.
When the Qero arose, it was supported by the au-
thority of the Ethiopian Orthodox Church, a powerful
component of this ancient society. The system de-
clined in 1975 as a result of the Agrarian Reform of
1975, which was introduced under the socialist regime
that came to power in the revolution of 1974. People
who were previously excluded from resource use
gained uncontrolled access through their constituent
peasant association. When it became apparent that the
resource management system was declining under
the land tenure reform, the community responded by
establishing the Guassa Committee, known locally as
Idir. The committee retained significant community
representation and was still deemed acceptable to the
political and social order of the socialist regime. The re-
markable adaptation and subsequent persistence of the
system suggests that it is stable and resilient in the face
of significant political change (Tefera 2001).
Apart from these noteworthy examples, the realiza-
tion of the conservation significance of the Ethiopian
Highlands has been late in coming; arguably, it has yet
to be fully grasped by the leaders of the countries
spanned by the highlands. While policies are largely in
place (e.g., the National Conservation Strategy, 1994;
the Conservation Strategy of Ethiopia, 1997; the Donor
Coordination Group on the Environment-Contribution
to the Poverty Reduction Strategy Paper (PRSP) Dis-
cussion, 2001; the Ethiopian Sustainable Development
and Poverty Reduction Program, 2002), the strategies
are not being implemented.
In 1909, Ethiopia passed its first wildlife legislation
designed to regulate “sport” hunting —particularly of
elephants. However, prior to 1944, the fauna and flora
of the highlands were still largely viewed as an infinite
source of food and other materials, and as a source of
“sport” for the upper echelons of society and expatri-
ates in the country. The Preservation of Game Procla-
mation of 1944 reinforced earlier legislation to regulate
hunting and to prevent the overhunting of certain
species.
With interest from international conservation or-
ganizations, the Ethiopian Wildlife Conservation Orga-
nization (EWCO) was established in 1964 (Hillman
1993a). Because of a lack of wildlife management expe-
rience (cf. the experience that was built in neighboring
Kenya and Uganda through their colonial past), the ma-
jority of the early work —the production of legislation
and the designation of protected areas— was largely
carried out by expatriates. It has only been since the
first batch of trainees returned from the Mweka
Wildlife College in Tanzania in the early 1970s that
Ethiopian nationals have started taking senior positions
within the EWCO.
The EWCO has been pivotal in the formulation of
legislation to protect the fauna and flora, as well as in
the designation, establishment, and management of na-
tional parks. However, the efforts of the EWCO have
been starved for resources and the legislation designed
to protect wildlife has proved impossible to enforce.
A system of conservation areas has formed the basis
of the wildlife conservation strategy in Ethiopia (Hill-
man 1993b). When they were proposed, they were
based on what was known about the fauna and habitats
at the time, and were primarily directed toward the
more spectacular assemblages of large mammals and
those species considered to be endemic and at risk
(Yalden et al. 1996). However, since this network was
proposed, only two of the 14 “national parks” and “sanc-
tuaries” have been legally constituted, namely Awash
National Park and Simien Mountains National Park, the
latter of which is recognized as a World Heritage Site
(Hillman 1993b). Even these two have never been ade-
quately secured, staffed or equipped. The numerous
“wildlife reserves” and “controlled hunting areas” are
little more than nominal, and provide no protection for
the fauna and flora. Indeed, only 3% of the Ethiopian
Highlands is conserved in protected areas in IUCN cat-
egories I to IV.
Since the mid-1970s, difficulties in wildlife conserva-
tion have been exacerbated by famines, refugee prob-
lems, civil unrest, armed rebellions, and war. This se-
ries of events threatens the livelihoods of the present
generation of Ethiopians. As long as such events con-
tinue and society remains stricken by poverty and food
insecurity, it is unlikely that wise conservation mea-
sures will be implemented.
However, if the region’s current and proposed na-
tional parks were fully established and administered,
they would have the potential to provide some level of
protection for many of the region’s endemic species. It
is notable that the conservation areas were specifically
designed to protect the mammalian fauna. Therefore, a
re-assessment of the conservation areas of the region is
warranted because they may not protect endemic
species across other taxa.
The Bale Mountains National Park is the single most
important conservation area that has been proposed in
the Ethiopian Highlands, harboring the finest and most
intact remnant of the highland’s original vegetation.
The Bale Mountains have 1 321 species of flowering
plants, 163 of which are highland endemics, including
the 27 Bale endemics (e.g., Euryops prostratus, Gladiolus
balensis, Maytenus harennensis, and Solanecio harennen-
sis). These mountains also contain more than half the
global populations of both the Ethiopian wolf and
mountain nyala. Of the mammals that have been
recorded there, 26% are Ethiopian endemics (including
the Bale monkey, Starck’s hare (Lepus starcki), and
eight species of rodents, including the Bale endemics
—the giant mole-rat, unstriped grass rat (Arvicanthis bli-
cki), and brush-furred mouse (Lophuromys melanonyx).
Among several rare endemic amphibians, there are
four species found in Bale alone, including one mono-
typic, endemic genus, the Bale Mountains narrow-
mouthed frog (Balebreviceps hillmani, EN) (Largen
2001), and there are two chameleons that are Bale en-
demics (Largen 1995; M. Largen and S. Spawls, pers.
comm.). The conclusion is that if conservation efforts
in the Bale Mountains are not successful and people
continue to exploit the resources in an unsustainable
way, more species of mammals (and the analysis re-
mains to be done for other taxa) would go extinct there
than in any other area of equivalent size on the globe
(J. Malcolm, pers. comm.).
The Ethiopian Wolf Conservation Program (EWCP),
which has its base in the Bale Mountains, has demon-
strated that working successfully in the difficult cli-
mate of Ethiopia is possible. Information on Ethiopian
wolves was only first collected during the mid-1970s
(Malcolm 1976, 1977, 1988), and through the 1980s and
early 1990s (Hillman 1988; Sillero-Zubiri 1994). These
studies gave the EWCP (based out of the Wildlife Con-
servation Research Unit of the University of Oxford) a
foundation, which was bolstered by the publication of
a conservation action plan (Sillero-Zubiri and Macdon-
ald 1997).
Consequently, the responsibilities of the EWCP have
evolved to ensure the conservation of the Ethiopian
wolf and its Afroalpine ecosystem. This is achieved by
a three-pronged approach: securing the conservation of
areas of Afroalpine ecosystem, their biodiversity and
ecological processes; assessing, addressing, and coun-
teracting threats to the survival of Ethiopian wolves; Local women in village near the
and enhancing the focus on and strength of the envi- Herenna Forest, Bale Mountains,
ronmental sector, and particularly biodiversity conser- Ethiopia.
© Patricio Robles Gil/Sierra Madre
vation, within Ethiopia.
Conservation in the Ethiopian Highlands is far from On pp. 274-275, wood is a scarce
being secure, but there may be room for hope. While its commodity in the Ethiopian
focus has been conspicuously away from natural re- Highlands, having been previously
source and wildlife conservation, the present Govern- removed through human
ment of Ethiopia has been making progress towards exploitation for fuel and
construction. Today, the majority of
creating a climate in which wildlife conservation could,
Ethiopian highlanders use cattle
potentially, play a role. In such a climate, tourism and dung as their principal source of
the revenues generated from it could grow. If this is fuel, possibly one reason why
linked to the unique fauna and flora of the region, then Ethiopia has the largest national
wildlife conservation could receive more attention and herd of domestic livestock in Africa.
political will. Here, people coming to market in
Dinsho village, near the
headquarters of Bale Mountains
STUART WILLIAMS 67, 68
National Park.
JOSÉ LUIS VIVERO POL 81 © Patricio Robles Gil/Sierra Madre
STEVEN SPAWLS 82
ANTENEH SHIMELIS 67
ENSERMU KELBESSA 83
273
 HORN OF AFRICA
The Horn of Africa was already a renowned biologi-
cal hotspot 5 000 years ago, when the ancient Egyp-
tians sent expeditions to the “Land of Punt” to bring
back unique natural commodities such as frankin-
cense and myrrh. During the times of the ancient
Greeks and Romans, these products were brought to
Europe by caravans along the incense route through
the Arabian deserts. Even the isolated island of So-
cotra, with its famous cinnabar (dragon’s blood) and
aloe, was part of this trading system more than 2 000
years ago.
The Horn of Africa is here defined as the arid
Horn and basically covers the area east of the Ethio-
pian Highlands (although it includes the Rift Valley,
which divides the Ethiopian Highlands into two ma-
jor blocks), also including the xeric bushlands of
northeastern Kenya and the southern coastal parts
of the Arabian Peninsula. The arid Horn covers
most of Somalia (including Somaliland and Punt-
land), Djibouti, and parts of Ethiopia, Eritrea, Kenya,
Yemen, Oman and, very marginally, Sudan. The area
covers around 1.65 million km2, but a relatively large
proportion of the land, such as the Danakil Depres-
sion in Ethiopia, has a very depauperate flora, and
the approximately 5 000 species of vascular plants
known from the region actually occupy only a small
percentage of the area. Included in this hotspot is
the Socotra Archipelago, with the main Socotra
Island and three smaller islands off the coast of
northeastern Somalia, covering about 3 636 km2, and
a few hundred small islands in the Red Sea, of which
Dahlak Island is the largest (643 km2).
Phytogeographically, all of the arid Horn as de-
fined here belongs to the Somalia-Masai region of
endemism (White 1983; White and Léonard 1991),
which also extends further south through the Kenyan
lowlands into northern Tanzania. The Horn of
Africa in the present sense can be regarded as the
core part of the Somalia-Masai region.
The most widespread vegetation type of the arid
Horn is Acacia-Commiphora bushland (about 30 spe-
cies of Acacia and 50 species of Commiphora are
endemic to the area), but evergreen bushland, suc-
culent shrubland, dry evergreen forest and wood-
land, semidesert grassland, and low-growing dune
and rock vegetation occupy considerable areas as
well. Succulents are common, including numerous
endemic species of, for example, Euphorbia and Aloe.
Small areas of mangrove are also found, both on the
African and the Arabian sides, as well as riverine
SAUDI OMAN
ARABIA
RED
SUDAN
SEA
YEMEN
ERITREA
GULF
Socotra
DJIBOUTI OF ADEN
ETHIOPIA
SOMALIA
INDIAN
Lake OCEAN
Turkana
KENYA
Lake
Victoria 0 400 km
vegetation along the major rivers: Wabe Shabelle,
Jubba, and Awash.
The altitude within the arid Horn ranges from
155 m below sea level at Lac Assal in Djibouti to
about 2 400 m above sea level in the mountains of
northern Somalia, but most of the area is below
500 m. The Haghier Mountains on Socotra reach just
above 1 500 m altitude and the highest escarpments
in Hadramaut on the Arabian side reach about
2 000 m. An unusual feature of the region is that
land plants, such as the doum palm (Hyphaene the-
baica, Arecaceae), can have an altitudinal range
from about 100 m below sea level to about 1 000 m
above sea level. The altitudinal delimitation between
the arid Horn flora and the highland floras in
Ethiopia, Eritrea, and Yemen is not always clearly On the opposite page, the gerenuk
distinguishable in the field, and also varies between (Litocranius walleri), a species
about 1 500 to 2 000 m according to local climate, ge- confined mainly to the Horn of
Africa, is remarkable in its ability
ology, and topography. Still, there is a clear-cut sepa-
to rise up on its hindlegs to feed
ration between highland (Afromontane) flora and at up to 2 m above ground level.
lowland (Somalia-Masai) flora (Friis et al., in press). In so doing, it reaches a very erect
The arid Horn, including the southern Arabian bipedal posture, facilitated by
Peninsula and Socotra, appears to have been elevated modification of the lumbar
by about the upper Eocene some 40 million years vertebrae.
© Ferrero-Labat/Auscape
ago. Geological evidence indicates that the Arabian
Peninsula and Africa separated by about 10 million
years ago, forming the incipient Red Sea and Gulf of
Aden. When the continental fragment Socotra was
last connected to Africa is less certain, but it may
277

The Endangered Grevy’s zebra
(Equus grevyi) has experienced one
of the most dramatic declines of any
large African mammal, with a 70%
drop in numbers in Kenya recorded
between 1977 and 1988, and a
decline of around 80% in Ethiopia
between 1995 and the present. It is
considered already extirpated in
Eritrea, Djibouti, and Somalia.
© Joe McDonald/DRK PHOTO
278
have been considerably earlier. The massive limestone
series covering most of the island (Beydoun and Bichan
1970) are connected with corresponding series in the
escarpments of northern Somalia and the southern
Arabian Peninsula.
Today, the major part of the arid Horn is covered by
limestone, sandstone or gypsum, whereas Pre-Cambrian
rocks form, for example, the prominent inselbergs in
southern Somalia, as well as the rugged pinnacles of
the mountains of Socotra. Large areas are covered by
deep sand, partly derived from Quaternary coralline
rocks or alluvial soils, and the fossil dune formations
running along the coast of central and southern Soma-
lia are prominent features of the landscape. Lava of
more or less recent origin is found in the Rift Valley
and the Afar Depression, and in parts of the southern
Arabian Peninsula.
The climate of the arid Horn can generally, and not
surprisingly, be described as hot and dry. It is not un-
common for temperatures to reach above 40ºC during
several months of the year. Characteristically, there
are two rainy seasons, one in April-May and one in
September-November, but there are many deviations
from this. The fog oasis of eastern Yemen and Oman
has summer rain during the southwest monsoon,
whereas winter rain occurs along the escarpments of
Eritrea and northern Somalia. Precipitation from mist
plays an important role along the Arabian coast, as well
as in the higher parts of Socotra and Somalia. The rocky
outcrops in southern Somalia are more humid than can
be expected from rainfall data alone.
The arid Horn is a kind of northeastern antipode to
the other arid African hotspot, the Succulent Karoo in
southwestern Africa. These two arid regions are be-
lieved to have been united by an “arid corridor” during
repeatedly drier and colder periods during the Pleis-
tocene, but most probably also earlier during the Ter-
tiary. Some genera of flowering plants are entirely
restricted to these regions, such as Kissenia with one
species in the arid Horn and one in the Succulent
Karoo, and Wellstedia with six species in the arid
Horn and one in the Succulent Karoo. A few species of
plants and animals in the Horn of Africa may also have
their closest relatives in the southern United States and
Central America. For example, Chapmannia, a genus of
mostly woody legumes, has five species in the arid
Horn; one in Mexico, Guatemala, and Venezuela; and
one in Florida (Thulin 1999), and the scorpion genus
Heteronebo has two species in the Socotra Archipelago
and about a dozen species in the Caribbean. This pat-
tern, which may be explained by the Tertiary “North At-
lantic land bridge,” is thought to be common among or-
ganisms that diversified during the Tertiary in xeric
and seasonally dry vegetation (Lavin et al. 2000).
The Horn region is sparsely populated, with general-
ly less than 20 inhabitants per km2, and nomadic pas-
toralism is commonly practiced. Well over half the area
(Somalia, eastern Ethiopia, and northeastern Kenya) is
mainly inhabited by the Somali, a culturally, linguisti-
cally, and religiously homogeneous group. In southern
and northeastern Ethiopia, northern Kenya, as well as
in Djibouti and the lowlands of Eritrea, the situation is
more complex, with many other East Cushitic groups,
such as Borana and Afar, living together. In southwest-
ern Ethiopia, there are also many groups that speak
languages of the Omotic family. The native language
on Socotra, Socotri, is a West Semitic language with ar-
chaic features. It is most closely related to Mahri and
some other languages spoken in southeastern Yemen
and Oman, along with the ubiquitous Arabic.
Biodiversity
Estimating the number of species of vascular plants in
the arid Horn region is not easy. One reason is that the
area falls under five different flora projects: Flora of
Tropical East Africa, Flora of Ethiopia and Eritrea, Flora
of Somalia, Les plantes vasculaires de la République de Dji-
bouti, and Flora of the Arabian Peninsula and Socotra. Of
these, only the Djibouti flora, with a relatively low
number of species, has been completed (Audru et al.
1994), whilst the other projects are ongoing. According
to the admittedly rough, but best possible estimate
available, there are about 5 000 species of vascular
plants in the region, and of these about 2 750 are en-
demic. In terms of plant endemism, this figure is high-
er than that for the Succulent Karoo (2 539 species), but
then the area covered by the arid Horn is more than 16
times as large. Many of the species in the arid Horn
have very restricted areas of distribution, however, with
strong concentrations of endemics in northern Somalia
(Thulin 1994; Friis et al., in press) and in the Socotra
Archipelago (Miller and Bazara’a 1998).
There are nearly 60 endemic genera of vascular
plants in the arid Horn (out of a total of about 970 gen-
era), 13 of which are endemic to the Socotra Archipel-
ago alone. It is also striking that the native flora of the
Socotra Archipelago comprises so few armed plants.
For example, the single endemic Acacia on Socotra, A.
pennivenia (VU), is unarmed, whereas the thorny bush-
lands of the continental Horn abound in armed en-
demic species of Acacia.
Of the 170 families in the region, two are endemic.
These are Barbeyaceae and Dirachmaceae, both woody,
Barbeyaceae with a single species, Barbeya oleoides,
which is relatively widespread in evergreen bushland
and dry evergreen forest, and Dirachmaceae with two
species, Dirachma socotrana (VU) on Socotra and D. so-
malensis (EN) in central Somalia. Barbeyaceae, with
small unisexual flowers without petals, have usually
been associated with Urticales. Dirachmaceae, with rela-
tively large bisexual flowers with prominent petals, have
been associated with Geraniales or Malvales. However,
molecular evidence clearly indicates that Barbeyaceae
and Dirachmaceae, despite their completely different
morphology, are closely related (Thulin et al. 1998).
A total of 219 mammal species in 115 genera are
known from the arid Horn, and of these 20 species
are endemic, the most notable ones being a number of
antelopes, such as the beira (Dorcatragus megalotis,
VU), dibatag (Ammodorcas clarkei, VU), Speke’s gazelle
(Gazella spekei, VU), silver dikdik (Madoqua piacentinii,
VU), and Salt’s dikdik (Madoqua saltiana). In addition,
there is an endemic subspecies of wild ass (the Somali
wild ass, Equus africanus somaliensis, CR). There are
five endemic mammal genera in the Horn, all of them
monotypic, including the aforementioned beira and dib-
atag, and three small mammal genera represented by
single species: the Somali pygmy gerbil (Microdillus
peeli), ammodile (Ammodillus imbellis, VU), and Speke’s
pectinator (Pectinator spekei). Indeed, the arid Horn has
been identified as an important area for rodent conser-
vation (Amori and Gippoliti 2001). No native mammals
are known from the Socotra Archipelago, except possi-
bly some species of bats, and it has been assumed that
this is because Socotra was isolated from the continen-
tal Horn too early for colonization by mammals.
There are 704 bird species regularly recorded from the
arid Horn, and 25 of these are endemic. Four Endemic
Bird Areas (EBAs) fall entirely within the hotspot: Soco-
tra (with six species confined entirely to this EBA); North
Somali Mountains (three species); Central Somali Coast
(two species); and Jubba and Shabelle valleys (four
species). There are seven species of birds found only in
Somalia, including the Bulo Burti boubou (a bushshrike)
(Laniarius liberatus, CR) —known only from a single in-
dividual and described from molecular, photographic,
and vocalization data (Smith et al. 1991)— the Warsangli
linnet (Carduelis johannis, EN), and five species of larks.
Among the six species of birds restricted to Socotra is the
Socotra sunbird (Nectarinia balfouri), a relatively abun-
dant species regularly seen visiting the red flowers of
species of Aloe and Ballochia. Another Socotra endemic,
the golden-winged grosbeak (Rhynchostruthus socotra-
nus), is the only representative of its genus.
There are some 284 reptile species in 94 genera
recorded from the Horn, and at least 93 species are en-
demic. The endemic Somalian spiny-tailed lizard (Uro-
mastyx princeps), for instance, is a diurnal lizard with a
short, robust, spiny tail that frequents the limestone
plateaus along the Indian Ocean coast in northeastern
Somalia. When disturbed, these lizards quickly retreat
into their holes, closing the opening using their armored
tail as a lid. There are six endemic reptile genera, in-
cluding Haackgreerius, a genus of skink represented by a
single species, H. miopus, in Somalia; and Aeluroglena,
comprising a single species of snake, A. cucullata. Three
genera are endemic on Socotra, namely Haemodracon,
with two species of geckos (H. riebeckii and H. trachyrhi-
nus), and Ditypophis and Pachycalamus, each with a sin-
gle species of snake (D. vivax and P. brevis, respectively).
Amphibians are poorly represented in the arid Horn,
with only 53 species recorded, at least seven of which
are endemic. Of the 20 genera represented, only Lan-
zarana, with a single species, Lanza’s frog (L. largeni)
confined to Somalia, is endemic. No amphibian species
are known from Socotra, despite the presence of suit-
able habitats.
It is estimated that there are around 100 species of
freshwater fishes in about 48 genera in the arid Horn,
of which 10 species are endemic. The endemics in-
clude three cave-dwelling species restricted to Somalia,
two of which —the Somalian blind barb (Barbopsis de-
vecchii, VU) and Somalian cavefish (Phreatichthys andruz-
zii, VU)— are blind. The three cave-dwelling species are
each placed in their own endemic genus. No native
freshwater fishes are known with certainty from Soco-
tra, but populations of Aphanius dispar have been in-
troduced to some waters as part of an anti-malaria pro-
gram. However, the endemic freshwater crab Potamon
socotrensis is common in small mountain streams, and
a second endemic species, Socotra pseudocardiosoma,
placed in its own genus, has recently been described
(Cumberlidge and Wranik 2002).
Flagship Species
For thousands of years the Horn of Africa has been fa-
mous as the source of frankincense (mainly from Boswel-
lia sacra in Somalia, Yemen, and Oman, and B. frereana
in Somalia), myrrh (mainly from Commiphora myrrha,
widespread in the Horn, and C. guidottii in Somalia and
eastern Ethiopia), and dragon’s blood or cinnabar (from
Dracaena cinnabari, EN, on Socotra), all of which are
commodities of gum-resins obtained from these trees.
In the Egyptian temple at Deir al-Bahari, near
Thebes, the walls are decorated with colored relief fres-
coes commemorating an expedition sent out to the
Land of Punt by the Queen of Egypt, Hatshepsut, about
1500 B.C. Quantities of incense were brought back by
this expedition, as were living trees in tubs that appar-
ently were then planted in the temple courtyard. The
exact position and extension of Punt is uncertain, but
judging from the commodities obtained from there it
has to have been within the Horn of Africa in the pres-
ent sense. Later, frankincense and myrrh were trans-
ported by huge caravans along the highly organized in-
cense route from southern Arabia to ancient Greece Although Awash National Park
and Rome (Groom 1981). is a stronghold for the
The production of frankincense and myrrh is still of Hamadryas or sacred baboon
(Papio hamadryas), it also has
major economic importance in Somalia, and to some
within its borders a unique hybrid
extent in Ethiopia and northern Kenya. The frankin- zone between this species and the
cense trees in northern Somalia are owned by extend- olive baboon (P. anubis). Hybrids of
ed families and clans. To tap the trees, incisions are the two baboon species have been
made into the inner bark of trunks and branches. The witnessed preying on young Salt’s
resin is left to dry and is later collected. Both Boswellia dikdiks (Madoqua saltiana).
© Patricio Robles Gil/Sierra Madre
frereana and B. sacra are evergreen trees that often
grow on limestone rocks, but B. frereana has the ability
to grow even on vertical cliff faces, and the collecting of
the frankincense from this species, therefore, is a par-
ticularly hard and dangerous occupation.
Dragon’s blood from Socotra (then called Dioscorida)
was in high demand by the ancient Greeks and Romans
for its medicinal properties and as a red dye. However, it
279

On pp. 280-281, many of the plant
species in the Horn of Africa have
very restricted areas of distribution,
but with strong concentrations of
endemics, such as in the Socotra
Archipelago, which has 13 endemic
genera. Croton socotranus,
pictured here in the foreground,
is one of a number of endemic
Socotran species.
© Mats Thulin
Above, one of the larger gazelle
species, the Vulnerable
Soemmerring’s gazelle
(Nanger soemmerringii) probably
numbers fewer than 15 000
animals, with the largest protected
population occurring in Awash
National Park.
© Patricio Robles Gil/Sierra Madre
282
lost its importance long ago and, during the last cen-
turies, dragon’s blood has mainly been obtained from the
related and more easily accessible dragon tree on the Ca-
nary Islands. Dracaena cinnabari is prominent in the
vegetation of Socotra, where the gum-resin is still used as
a medicine and dye. It is a spectacular tree with repeat-
edly ramifying branches forming a dense, umbrella-like
canopy.
Another spectacular tree found only on Socotra is the
cucumber tree (Dendrosicyos socotrana, VU), with a mas-
sive water-storing trunk. Despite being a relatively tall,
free-standing tree when mature, the branches of Den-
drosicyos have tendrils, just as in its herbaceous relatives.
The daban or Bankoualé palm (Livistona carinensis, VU)
is a slender palm reaching well over 20 m in height when
mature. It is known from a few isolated localities in
northeastern Somalia, Djibouti, and southern Yemen.
Phytogeographically, it is of great interest, as the other 30
or so species of Livistona occur in Southeast Asia and
Australia. The largest population is in Yemen, where re-
generation is also satisfactory, but recent reports indicate
that most, if not all, tall mature trees have been felled for
timber. The Djibouti population is smaller, but includes
both mature trees and juveniles in some sites. In Soma-
lia, the daban, as it is called there, is highly threatened
and the total population is now probably less than 40
trees in the two known localities (the author counted 11
mature trees at one locality in 1995 and 28 mature trees
at the other locality in 2000). The trees have been used
for house building, drainage pipes, and so on, and regen-
eration is prevented as the leaves of young plants are
grazed or used for the production of mats or baskets.
The Yeheb nut (Cordeauxia edulis, VU) is a wonderful
evergreen shrub or small tree with yellow flowers
found in the dry bushlands of eastern Ethiopia and cen-
tral Somalia, usually in areas of deep sand. The seeds
are edible and nourishing and highly appreciated as
food, and the foliage provides a red dye. The plant has
attracted considerable interest as a potential food crop
for arid areas, but is not easy to cultivate.
During fieldwork conducted for the Flora of Somalia
project over the last 20 years or so, hundreds of new
species have been discovered, many of them apparently
narrow endemics. Without comparison, the one species
among these that has attracted the most attention is the
Somali cyclamen (Cyclamen somalense). It was first
found in 1986 on one of the misty limestone escarp-
ments in northern Somalia and is still known only from
a very small area. The horticulturally important genus
Cyclamen is otherwise distributed in the Mediterranean
region, extending eastwards to Iran, and the occurrence
of a species of this genus in tropical Africa was most un-
expected, although there is a fairly strong Mediter-
ranean element in the flora of this part of Somalia.
In addition to the plants and plant products that the
ancient Egyptians brought back from the Land of Punt,
they also brought back animals, and the most famous of
these is the Hamadryas or sacred baboon (Papio hama-
dryas). This monkey is today endemic to the arid Horn,
where it lives on hillsides and escarpments bordering
the southern part of the Red Sea and the Gulf of Aden.
The sacred baboon is an omnivore for which desert
dates (Balanites and Ziziphus) are of great seasonal im-
portance, but it raids crops in settled areas and that of-
ten leads to conflict. The animals were held sacred in
ancient Egypt and were also mummified. During the
earliest Egyptian civilization, the distribution of the sa-
cred baboon may have extended to the Egyptian border
area, although they were later imported to the temples,
and the relief frescoes at Deir al-Bahari show baboons
swarming over the ships in Hatshepsut’s expedition.
Besides the baboon, the Horn is an important region
for a number of threatened antelope species, particu-
larly the beira and dibatag antelopes (in both cases, the
Somali name for the species has become the common
name). The beira, with its prominent ears, is confined
to dry and inhospitable hills and mountains of northern
Somalia, eastern Ethiopia, and Djibouti (where its pres-
ence was only confirmed in 1993), in which it is able to
survive without water. The larger, but more slender
dibatag, with its characteristic erect tail and long neck,
is found in the Acacia-Commiphora bushlands of east-
ern Ethiopia and adjoining lowlands of northern and
central Somalia. Both the beira and the dibatag have
suffered from uncontrolled hunting and habitat degra-
dation.
Another important flagship is the desert warthog,
Phacochoerus aethiopicus, recently confirmed to repre-
sent a distinct species from the common warthog, P. afri-
canus, and distributed mainly in eastern Ethiopia, So-
malia, and northern Kenya (D’Huart and Grubb 2001). In
addition, species such as Speke’s gazelle, Salt’s and silver
dik-diks, and beisa oryx (Oryx beisa) have their ranges
entirely or almost entirely within the Horn, while
Swayne’s hartebeest (Alcelaphus buselaphus swaynei,
EN), a subspecies of the common hartebeest, has its
range restricted entirely to the southern Rift Valley of
Ethiopia, despite once occurring throughout the Rift
Valley eastwards into northwestern Somalia (East 1999).
Among the 31 endemic birds in the region, the Warsan-
gli linnet is one of the most notables. This small bird,
with colors in white, black, and various shades of grey
and reddish brown, is locally common in high, steep es-
carpments along the Gulf of Aden in northern Somalia.
Ash and Miskell (1998) reported it as resident in two
sites (Daalo and Mash Caleed) in the 1980s, and the bird
has been seen in both of these sites in recent years. The
species is likely to occur along this whole stretch of es-
carpment, where access is nearly impossible. The Ethi-
opian bush-crow also deserves special mention as a flag-
ship species. This small, starling-like crow is found in
Acacia bushland in a very restricted area of southern
Ethiopia. It is still locally common, but ongoing degra-
dation of the habitat may become a threat in the near
future. Finally, the Djibouti francolin (Francolinus ochro-
pectus, CR) is known from only two sites in Djibouti: Forêt
de Day, which is thought to be the only viable site for this
imperiled species, and the nearby Mabla Mountains.
Threats
The arid Horn is a hotspot under heavy pressure. Close
to 100% of the land is used for grazing, mainly by
camels, goats, and sheep, and the area, therefore, is in a
seminatural state and very little can be said to be pristine
nature. Overgrazing, leading to a gradual degradation of
the vegetation, is a problem over large areas and is par-
ticularly severe near watering points. The stands of
many unique species of trees, such as the dragon tree on
Socotra and the daban palm in Somalia, are increasingly
becoming overmature with little regeneration. One
study has estimated that the region is represented by
around 24.2% undisturbed habitats, 70.3% partially dis-
turbed and 5.5% human-dominated habitats (Hannah et
al. 1995), but this is almost certainly an underestimate of
the disturbance, and perhaps only 5% can be regarded as
undisturbed. However, the category “partially disturbed”
would include all degrees of degradation, from almost
completely destroyed to nearly pristine habitats.
Shifting cultivation, where areas of bushland and
woodland are cut and burned for the cultivation of cas-
sava over a period of a few years, is particularly destruc-
tive in parts of central and southern Somalia. However,
in Somalia the worst threat to the vegetation and the bio-
diversity is the uncontrolled production of charcoal.
Charcoal is now not only produced to cover the domestic
needs within the country, but has become a major export
item to the countries in the Arabian Gulf region. The tree
most sought after for charcoal is Acacia bussei, and the
woodlands formerly dominated by this tree are now
rapidly dwindling. Acacia bussei itself is not threatened as
a species, but the destruction of the big trees changes the
environment completely and adversely affects numer-
ous other species of plants and animals. Agricultural
schemes in the Rift Valley and along the Wabe Shabelle
and Awash rivers in Ethiopia and along the Wabe Sha-
belle and Jubba Rivers in Somalia also threaten the bio-
diversity, particularly of riparian habitats.
On Socotra, there has been considerable infrastruc-
ture development in recent years, including the build-
ing of a new port, an airport with tarmac runway, and
new roads. The potential for sensitively managed eco-
tourism is great, but the ways in which these develop-
ments actually influence the biodiversity of the island
require careful monitoring.
Despite these threats, at least among the plants we
can not point to a single species that is known to be ex-
tinct. Taverniera sericophylla, an endemic on Socotra
earlier reported as extinct (Groombridge 1994), is still
thriving in parts of the island (Thulin, pers. obs.), and
several other endemic species on Socotra reported as
Critical (Lucas and Synge 1978), such as Dirachma soco-
trana and Punica protopunica (VU), also persist in viable
populations. In Somalia, there are a few flowering
plants that have not been seen for more than 100 years,
such as the endemic Somali lupine (Lupinus somalien-
sis), but there is still hope that it survives in some inac-
cessible place. The situation for larger animals, in con-
trast, is much worse. Uncontrolled hunting, particular-
ly of the ungulates, is a real problem in many parts of
the region, with many species classified as threatened.
Conservation
Despite the vast size of this hotspot, less than 9% has
some form of legal protection in 41 protected areas clas-
sified in IUCN categories I to VI, a percentage that falls
to just over 3% when including only those protected ar-
eas in categories I to IV. There are several national
parks and sanctuaries in the Ethiopian part of the re-
gion, such as Awash National Park (750 km2) and the
Chew Bahir Wildlife Reserve (2 730 km2), mainly aimed
at protecting remaining populations of desert-dwelling
ungulates; the problems and challenges faced are to a
large degree similar to those described for the Ethiopi-
an Highlands. In Kenya there is the remote Malka Mari
National Park (870 km2) in the northeastern corner of
the country. In Somalia there are no areas with formal
protection after the breakdown of the federal govern-
ment in 1991, although three protected areas are offi-
cially recognized. In Djibouti, the Forêt de Day Nation-
al Park (100 km2), home to the aforementioned Djibouti
francolin, is the country’s only reserve. The govern-
ments of Somaliland and Puntland have ministries for
the environment, but their resources are totally inade-
quate. The Socotra Archipelago has recently been
added to UNESCO’s World Network of Biosphere Re-
serves, which should help in the implementation of con-
servation management and in the development of eco-
tourism in the area. In Oman, the massive Arabian Oryx
Sanctuary (24 785 km2) is a Natural World Heritage Site
famous for the successful reintroduction of Arabian
oryx (Oryx leucoryx, EN), and Jebel Samhan is a Nation-
al Nature Reserve mainly due to its population of Arabi-
an leopard (Panthera pardus nimr, CR). Confined entirely to northeastern
In general, conservation activities in the Horn of Africa, Salt’s dikdik is a relatively
Africa hotspot are completely inadequate for the long- common antelope typically found in
semidesert scrub, avoiding desert
term preservation of its biodiversity. Today, only Socotra
areas and restricted in its
can be said to receive any serious international atten- distribution by the mountains of the
tion. Recently, for example, a major program for “Sus- Ethiopian Massif to the west.
tainable Development and Biodiversity Conservation for © Patricio Robles Gil/Sierra Madre
the People of the Socotra Islands” was signed by the Unit-
ed Nations Development Program (UNDP). The pro- On pp. 284-285, Afar people with
camels crossing the Awash River in
gram, financed by UNDP, the Government of Italy, and
Awash National Park.
the Yemen Government, will collectively contribute
© Patricio Robles Gil/Sierra Madre
over US$5 million to continue to support the people of
Socotra through conservation and sustainable use of the
islands’ unique biodiversity and natural resources for
the next five years. Much needs to be done before the
arid Horn can regain the kind of international promi-
nence that it had 2 000 years ago.
MATS THULIN 18
283
 IRANO-ANATOLIAN
The Irano-Anatolian Hotspot covers a topographical-
ly complex and extensive system of mountains and
closed basins separating the natural communities
and the indigenous cultures of the Mediterranean
Basin and the dry plateaus of Western Asia. For
many centuries, the historical Silk Road crossing
from east to west through this area was the major
route for trade and cultural exchange between the
two regions, and the area is here recognized as a new
hotspot because of its high endemism and rich flora
and fauna.
The hotspot originally covered an area of
899 773 km2, including major parts of central and
eastern Turkey, Georgia (marginally), the Nahçevan
Province of Azerbaijan, Armenia, northeastern Iraq,
northern and western Iran, as well as the Northern
Kopet Dagh Range in Turkmenistan. The region in-
cludes massive ranges of dry mountains stretching
west to east in the northern part (East Anatolian
Mountains, Southern Elburz Mountains, and the
Kopet Dagh) and north to south in the southern part
(Zagros Mountains). These highlands extend be-
tween the three major remnants of the ancient
Tethys Sea: the Mediterranean, the Black Sea, and the
Caspian.
In terms of its geological make-up, the hotspot has
a mixture of ancient massifs, folded sedimentary
rocks, and recent igneous rocks. Limestone blocks
mostly from the Paleozoic Era predominate in its
southern part. Between the mountain ranges, there
are broken plateaus covered by alluvial soils. Several
parts of the hotspot are seismically active. Here, ele-
vations generally range from 1 000 m to 4 000 m,
with some areas as low as 300 m, such as the foothills
of the Kopet Dagh and western Zagros Mountains. A
number of peaks exceed 5 000 m, including the dor-
mant volcanoes of Mt. Ararat in Turkey (5 165 m)
and Mt. Damavand in Iran (5 671 m). The plateaus of
Anatolia, Armenia, and western Iran range between
800 and 2 000 m, and are characterized by primary
plain steppes, which are climax steppes that have
never been covered by any forest or shrubland (be-
cause these areas are too dry and the soils poor); the
mountain areas around the plateaus are covered by
dry forests or secondary steppes. Three major closed
basins (Konya Closed Basin, the Van Basin and the
Lake Urumiyeh Basin) lie to the north of the hot-
spot, including various small remnant lakes of the
Tethys surrounded by primary plain steppes.
BLACK SEA GEORGIA
ARMENIA
AZERBAIJAN TURKMENISTAN
TURKEY CASPIAN
SEA
SYRIA
MEDITERRANEAN
SEA
IRAQ
IRAN
0 500 km PERSIAN
GULF
The geological history of this hotspot has triggered
the evolution of its biodiversity. The mountains
served both as a refuge and a corridor between the
eastern Mediterranean and western Asia at the time
of movements of the ancient Tethys Sea and its rem-
nants. Roughly 10 million years ago, the Turkish and
Arabian plates moving northwards crashed into Asia
to create the mountains and plateaus of the Irano-
Anatolian region. This cut the Tethys Sea off from
the Indian Ocean, thereby creating the Mediter-
ranean Sea. The remnants of the Tethys in the Irano-
Anatolian plateaus dried up and were refilled sever-
al times. These movements have made the hotspot a
site of constant dispersal and local speciation, and
many endemic species, especially local plant spe-
cies, originated at this time (Ekim et al. 2000). The
hotspot’s centers for endemism of aquatic species,
the Tuz Lake Basin, Van Lake, Urumiyeh Lake, and
many other small lakes in Turkey and Iran, are par-
ticularly important in representing the biodiversity
of the Tethys. On the opposite page, the
During long periods of glaciation, this region was impressive Kopet-Dagh Range is
invaded by refugee species from northern Europe and the border between Turkmenistan
and Iran, and the refuge of many
Asia. Then, periods of aridity disrupted the ranges of
species like the Vulnerable bezoar
many species and effectively isolated the Euro-Siber- or wild goat (Capra aegagrus)
ian forest and meadow species, such that they became and Vulnerable Afghan Urial
restricted to the riparian forests along the mountain (Ovis orientalis cycloceros).
valleys or in alpine lakes of the hotspot. © Patricio Robles Gil/Sierra Madre
The climate is continental, with hot summers and
cold winters. Most of the region has severely cold win-
ters, with extremes dropping to –40ºC. Summers are
dry, with a mean of 20º-30ºC, reaching up to 48ºC.
287

Grey wolf (Canis lupus) in
Badkhyz Nature Reserve,
Turkmenistan.
© Gertrud and Helmut Denzau
On the opposite page, the mountain
seen in the background is the
Erciyes Dagi Volcano. Sultansazligi
and Yay lakes were some of the best
spots for the migration of birds in
the Central Anatolia. Although they
were both protected by the national
law, they were dried up by the
General Directorate of State
Hydraulic Works, which is a
government department.
© Cuneyt Oguztuzun
288
Annual rainfall varies from 100 mm to over 1 000 mm,
most of it falling in winter and spring. Winter rainfall
occurs mainly in the form of snow, which can remain
until June-July at higher altitudes. Rainfall is remark-
ably low in central Turkey and in the southeastern
Zagros Mountains, ranging between 100 and 400 mm
per year.
The Irano-Anatolian Hotspot is confined to the west-
ern part of the Irano-Turanian Floristic Region that also
extends eastwards to include the arid lands bordering
the Caspian Sea, the Hindu Kush, and the spurs of the
western Himalaya (Takhtajan 1986). Some areas, how-
ever, express the mixed character of their biogeograph-
ic connections with Mediterranean and Irano-Turanian
elements, such as the Kopet Dagh and the Zagros
mountains (Davis et al. 1995). The principal habitat in
the hotspot is mountainous forest steppe that supports
oak-dominant (Quercus spp.) deciduous forests in the
west and south (Anatolia and Zagros mountains), and
juniper (Juniperus spp.) forests in the east (southern
slopes of the Elburz Mountains and the Kopet Dagh).
The forest steppe occurs mainly between 700 m and
the timberline (1 900 to 2 200 m). Primary forests are
largely degraded, and mountainous secondary steppes
now dominate the landscape where the dry forest is im-
poverished. A relatively wide zone of subalpine and
alpine vegetation covers the mountain peaks above the
timberline (Zohary 1973). Thorn-cushion (tragacanth-
ic) formations are found in the subalpine zone. The
alpine zone of the Cilo and Hakkâri mountains in
Turkey holds permanent glaciers (Kılıç and Eken 2004).
There is a clear separation between mountainous
flora (dry forest and montane steppe) and steppes of
the plateaus (primary plain steppe). In the Zagros
Mountains, the lower altitudes are dominated by Astra-
galus and Salvia spp., while in Elburz, a number of
flowering plants, such as tulips (Tulipa spp.), irises (Iris
spp.) and crocuses (Crocus spp.) occur. The primary
steppes, characterized by the presence of Artemisia
spp., occupy the fragmented plateaus in the hotspot.
The largest stand of primary steppes extends along the
Konya Closed Basin in Central Turkey, which also in-
cludes a unique complex of salt steppes and halophytic
marshes. The core zone of such plateaus contains a
salty or brackish lake occupying the lowest part of the
closed basin and fed by streams of surrounding moun-
tains (Eken and Magnin 1999). Various types of ripari-
an forests stretch along the river valleys, dominated by
species such as Salix spp. and others.
The human population mainly consists of Turkish,
Kurdish, Persian, and Armenian people. Some other
groups include Assyrians, Azeris, Turkmens, and Yezidis.
Pastoral and truly nomadic cultures still exist in a num-
ber of parts of Turkey and Iran, and traditional animal
farming (goat and sheep) is the main source of income
for many rural communities. Most parts of the hotspot
are not accessible during long winters when high pass-
es are closed. This makes most of the region unsuitable
for large-scale industrial investments or settlement.
Biodiversity
It is difficult to make an accurate estimate of vascular
plant diversity in this hotspot, but there are at least
6 000 species. Of these, an estimated 2 500 are endem-
ic. There are 332 species recorded as local endemics on
the Kopet Dagh, while the Zagros Mountains have at
least 500 local endemics (Davis et al. 1995). The highest
level of endemism is evident in Turkey, where 1 220
species are endemic (Ekim et al. 2000). Several genera
are largely confined or endemic to the hotspot, includ-
ing Astragalus, Centaurea, Acantholimon, Onobrychis,
Acanthocardamum, and Zerdana.
One of the most distinctive biodiversity features of
this hotspot is known as the Anatolian Diagonal (Davis
1965-1985, 1971), a remarkable floristic line crossing In-
ner Anatolia. The Anatolian Diagonal runs from the
southern foothills of the Eastern Black Sea Mountains
in Turkey, through the Turkish part of the hotspot, and
then splits into two branches: one reaches the Mediter-
ranean via the Amanus Mountains and the other, via
the Bolkar Mountains (Davis 1971). Approximately 390
plant species have distributions largely confined to the
Anatolian Diagonal, including the monotypic genus
Neotchihatchewia (N. isatidea) and Graellsia davisiana.
Many of Turkey’s other plant species occur only to the
west or only to the east of this line (Ekim and Güner
1986). The origin of this diagonal is a point of discus-
sion. Davis (1971) and Sonnenfeld (1974) emphasize
paleogeographical factors, while Ekim and Güner (1986)
attribute greater importance to ecological and climatic
factors.
The hotspot also possesses a very diverse fauna, de-
spite the fact that large portions are arid to semiarid.
Two main factors are responsible for this richness and
diversity. Firstly, the great variety of habitats ranging
from alpine tundra and permanent snowfields down to
semiarid plains, and from lush, humid deciduous forest
in the north to sparse juniper scrub in the east and
southeast. Secondly, the hotspot’s position at a cross-
roads between three major faunal realms: the southern
edge of the Palearctic realm, facing a small Afrotropi-
cal, and a strong Indo-Malayan influence in the south.
The Irano-Anatolian Hotspot is most remarkable for
its avian diversity, which includes a substantial West
Palearctic faunal element, including 81 species that
reach the eastern extremity of their range in the hot-
spot, e.g., the blue tit (Parus caeruleus); and a smaller, but
still marked, eastern Palearctic element, with 19 bird
species reaching the westernmost tip of their ranges in
the hotspot, e.g., the chestnut-breasted bunting (Emberi-
za stewarti). The alpine biome of the Elburz and Zagros
mountains, as well as the higher peaks of mountain
ranges in Iran and Turkey, support a montane fauna
typical of all high mountain ranges from the Pyrenees
and Alps in Western Europe to the Himalaya. Charac-
teristic species include the lammergeier (Gypaetus bar-
batus), horned lark (Eremophila alpestris), yellow-billed
chough (Pyrrhocorax graculus), alpine accentor (Prunella

Steppe tortoise (Testudo
horsfieldii) in Badkhyz Nature
Reserve, Turkmenistan. The males
are collected for the pet market, so
females have difficulties finding
mates. Listed on 2003 Red List as
Vulnerable.
© Gertrud and Helmut Denzau
290
collaris), wallcreeper (Tichodroma muraria), and white-
winged snowfinch (Montifringilla nivalis). The unique
enclave breeding population of the white-winged scoter
(Melanitta fusca) in the montane lakes of eastern Turkey
has a Euro-Siberean biogeographic history (Ekim et al.
2000; Eken et al. 2000).
In the drier and more open mixed woodlands of pis-
tachio (Pistacia), maple (Acer), and almond (Amygda-
lus) of the eastern Zagros, only a handful of western
Palearctic species occur. Characteristic birds include a
mixture of Middle Eastern species, e.g., the white-
throated robin (Irania gutturalis) and plain leaf-warbler
(Phylloscopus neglectus); eastern Palearctic species, e.g.,
the rufous-tailed shrike (Lanius isabellinus) and Hume’s
whitethroat (Sylvia althaea); Indo-Malayan species, e.g.,
the bay-backed shrike (Lanius vittatus); and western
Palearctic species at the extreme edge of their ranges,
e.g., the Eurasian blackbird (Turdus merula). Finally,
throughout the remote mountain ranges of Iran there
still exist good stands of juniper (Juniperus) woodland,
with species such as the fire-fronted serin (Serinus pusil-
lus) and, in the northeast, the white-winged grosbeak
(Mycerobas carnipes).
In total, there are at least 363 regularly occurring bird
species in the hotspot, although none are endemic.
However, a number of globally threatened birds have
significant breeding populations in the hotspot, includ-
ing the great bustard (Otis tarda, VU), marbled duck (Mar-
maronetta angustirostris, VU), white-headed duck (Oxyu-
ra leucocephala, EN), and imperial eagle (Aquila heliaca,
VU). In addition, up to 25% of the known world popu-
lation of the sociable plover (Vanellus gregarius, VU)
stop over in the plateaus of Eastern Anatolia in autumn.
The wetlands of the spectacular Tuz, Van, and Uru-
miyeh basins in Turkey and Iran, centered on the very
large and saline lakes, support important breeding
colonies of waterfowl, notably the greater flamingo
(Phoenicopterus ruber) (35 000-40 000 pairs; Lakes Tuz
and Urumiyeh), great white pelican (Pelecanus onocrota-
lus) (1 000-1 600 pairs; Lake Urumiyeh), Eurasian spoon-
bill (Platalea leucorodia), and glossy ibis (Plegadis falci-
nellus). These wetlands are extremely important for
shorebirds migrating along the African-Eurasian Fly-
way, such as the little stint (Calidris minuta) and ruff
(Philomachus pugnax), and in mild winters can hold
over 100 000 wintering ducks and geese.
The Caspian snowcock (Tetraogallus caspius), Ar-
menian gull (Larus armenicus), and Radde’s accentor
(Prunella ocularis) are only found in the Irano-Anatolian
and Caucasus hotspots. Furthermore, a large propor-
tion of the world population of crimson-winged finch
(Rhodopechys sanguinea), Finsch’s wheatear (Oenanthe
finschii), rufous-tailed wheatear (O. xanthoprymna),
Upcher’s warbler (Hippolais languida), white-throated
robin, and eastern rock-nuthatch (Sitta tephronata) oc-
cur in the hotspot.
A total of at least 141 mammals in 77 genera are
known from the hotspot. Of these, 10 are endemic, in-
cluding Dalh’s jird (Meriones dahli, EN), Anatolian vole
(Microtus anatolicus), Dogramaci’s vole (M. dogramacii),
Nasarov’s vole (M. nasarovi), the recently described
M. qazvinensis from northern Iran, Urartsk mouse-like
hamster (Calomyscus urartensis), Brandt’s hamster (Me-
socricetus brandti), Setzer’s mouse-tailed dormouse (My-
omimus setzeri), Schaub’s myotis (Myotis schaubi, EN),
and Hakkari mole (Talpa davidiana). As with birds, the
mammal fauna is composed of species of varied origin,
including the Asian subspecies of the cheetah (Aci-
nonyx jubatus venaticus, CR), jungle cat (Felis chaus),
leopard (Panthera pardus), striped hyaena (Hyaena
hyaena), Asiatic wild ass (Equus hemionus, VU), goi-
tered gazelle (Gazella subgutturosa), and wild goat
(Capra aegagrus, VU).
Reptiles are represented by at least 116 species, in-
cluding 13 endemics, among them the Persian rat snake
(Elaphe persica), Basoglu’s racerunner (Eremias su-
phani), Van Lake lizard (Darevskia bendimahiensis), and
Elburz lizard (Lacerta defilippii). There are four endem-
ic species of the genus Vipera: the mountain viper (V.
albizona, EN), restricted to the Kulmaç Daǧı of central
Anatolia, with a total range estimated to comprise
about 20 km2 of rocky slopes at 2 000 m altitude; Latifi’s
viper (V. latifi, VU), from the Elburz Mountains; Wag-
ner’s viper (V. wagneri, EN), which occurs near Lake
Urumiyeh (Iran) and in Eastern Turkey; and Darev-
sky’s viper (V. darevskii, CR), from the Djavakhk Moun-
tain Range in northern Armenia.
Twenty-one amphibian species are represented in
the region, four of which are endemic, all salamanders
in the endemic genus Neurergus: N. crocatus (VU) from
the vicinity of Beytussebap (Hakkari Province) in
southeast Anatolia, the Kurdish region of northern Iraq,
and also northwestern Iran; N. kaiseri (EN), which is
endemic to Iran; N. microspilotus (VU), restricted to the
Avroman Mountains on the Iraq-Iran-Turkey border,
and N. strauchii (VU), which is known only from Mu,
Bitlis, and Malatya provinces in Eastern Anatolia. An-
other interesting salamander species is the Persian
mountain salamander (Batrachuperus persicus, VU),
which is found in mountain streams on the Caspian
slope of the Talesh and Alborz mountains of Iran.
There is a minimum of 90 freshwater fish species in
the region, of which roughly 30 are endemic; these are
mainly confined to closed-basin lakes and rivers —par-
ticularly the Konya Closed Basin, Van Lake Basin, and
Urumiyeh Lake Basin. Several of these species are glob-
ally threatened, including Salmo platycephalus and
Capoeta pestai.
The invertebrate fauna of this hotspot is not well
known, but it is particularly rich for butterfly species,
with at least 350 species estimated to occur. Information
available for the Turkish part of the hotspot indicates
that there are at least 240 species, of which 18 are
endemic (Hesselbarth et al. 1995). Furthermore, the Ko-
pet Dagh region has at least 20 endemic Noctuidae spe-
cies (Ronkay et al. 1998). Some globally threatened
species are also found in the region, including the
cinquefoil skipper (Pyrgus cirsii, VU), the single-site en-
demic Polyommatus dama (EN), and the Apollo butterfly
(Parnassius apollo, VU). In addition, the hotspot is known
to be the richest part of the Palearctic region for scorpi-
ons, with 42 described species, of which at least 50% are
thought to be endemic to the hotspot (Fet et al. 2000).
Flagship Species
The most obvious flagship species for this hotspot
are the two large cats, namely the cheetah and tiger.
The Caspian subspecies of the tiger (Panthera tigris vir-
gata) was recently reported from the northern Zagros
Mountains range in Turkey (Welch 2003). However,
since the last confirmed individual of this subspecies
was shot in 1970, in the same region, this latest record
requires further investigation. Fortunately, this area
has not developed appreciably since 1970, and the hu-
man population has dramatically decreased due to un-
suitable economic and political conditions in this part
of Turkey. An immediate survey is required in the re-
gion to confirm the presence of the tiger, although this
appears to be extremely difficult, as large parts of the
area are naturally inaccessible and the periphery of
the few roads are covered by land mines.
The situation for the Asiatic cheetah is somewhat
better, although only an estimated 60 Asiatic cheetahs
still survive, all of them to the south of the Kopet Dagh
in Iran’s Great Salt Desert, the Dasht-e-Kavir. This ranks
among the most arid cheetah habitat anywhere, and
these animals mostly occur in the foothills and dry
watercourses of desert massifs, where prey is more
common than on the flats. With the calamitous decline
of gazelles in Iran, the cheetahs now subsist on wild
sheep, Persian wild goats, and perhaps hares. Local
herders believe that cheetahs are rarely responsible for
attacks on livestock, but even so cheetahs are some-
times killed when confused for leopards or other preda-
tors. This animal perhaps ranks as an odd choice for a
flagship species, since the core population is outside
the boundary of the hotspot, but is nonetheless consid-
ered worthy of mention here.
In historical times, the Irano-Anatolian Region be-
tween Turkey in the west and Turkmenistan in the east
was inhabited by several subspecies of Asiatic wild ass.
The onager or Persian wild ass (Equus hemionus onager)
was found east of the Zagros Mountains and south of
the Elburz and the Kopet Dagh. A few hundred onagers
have survived in two populations in Iran, one of which,
Bahram-e-Goor, falls within this hotspot. Another Criti-
cally Endangered subspecies of wild ass, known as the
kulan, Transcaspian or Turkmenian wild ass (E. h. ku-
lan), used to have its southwest extension along the
northern foothills of the Kopet Dagh. However, the sub-
species suffered a drastic decline in range and numbers
due to poaching and, today, a single naturally occurring
population remains in and around the Badkhyz Nature
Reserve in south Turkmenistan, just marginally out-
side the hotspot (Feh et al. 2002).
The key flagship bird species for the large salt lakes
is the greater flamingo, which in some years has ex-
tremely large breeding colonies in Iran and Turkey
—between 35 000 and 40 000 pairs. The main breeding
sites are Tuz Lake in Turkey and Lake Urumiyeh in
Iran. Unfortunately, no breeding has occurred in Lake
Urumiyeh since 2000 due to drought. On the other
hand, Tuz Lake is still considered one of the most im-
portant flamingo breeding areas in the world. Since the
1970s, this lake has been regularly used by flamingos,
and is home to the biggest breeding colony in Turkey.
Recent surveys carried out over the area revealed that
the number of chicks fluctuated between 4 000 in 1991
and 10 000 in 2000. However, there is also a negative
trend (as evidenced by the decreasing numbers of
chicks seen at the main crèche) in Tuz Lake that could
be related to a sharp decrease in water levels. These
lakes are the only major breeding site for this species in
the vast area between the Camargue (France) and the
Rann of Kutch (India). Archaeological evidence shows
that the species has been breeding at Lake Urumiyeh
for several thousand years. Ringing recoveries show
that the birds of Lake Urumiyeh disperse as far west as
Morocco and the Camargue, as far east as central India
and Sri Lanka, and as far south as Somalia and Ethiopia.
The hotspot hosts an endemic genus of amphibians,
Neurergus, with four species. These salamanders are ex-
tremely sensitive to environmental change because they
live in marginal conditions in relation to the general dry
climate of the Zagros Mountains. All four species are
considered threatened, having undergone declines in
abundance due to the effects of habitat destruction, pol-
lution, and drought (Rastegar-Pouyani, in litt. 2003).
An endemic fish species, Chalcalburnus tarichi, has
been used effectively as a flagship species by local con-
servation NGOs in the Van Province to trigger interest
in conservation of the Van Lake Basin in Eastern
Turkey. The rapid decline in the population of the
species, which also has local commercial value, has di-
rected the attention of local and national stakeholders
towards the wider environmental problems of Van
Lake. A number of conservation projects were initiated
since then in the area to protect the basin.
Hundreds of single-locality endemic plants occur in
Turkey, which are also threatened. Orchids are the
most charismatic and the most threatened group, in The great bustard (Otis tarda) is
part because they are still illegally collected in large well distributed over Europe and
quantities from the wild for the production of a popular Asia. Its status is Vulnerable.
The Irano-Anatolian Hotspot is
traditional drink in Turkey called sahlep. Due to the
home to one of the most important
rapid decline of orchids in Turkey, orchid collection has populations in Turkey,
expanded to Iran. with 4 000 birds.
The extremely localized salt plants of Anatolia and © Chris Gomersall/RSPB-images.com
Iran are among the most outstanding plant species in
the hotspot. They occur in the remaining salt steppes
of the Irano-Anatolian closed basins. This vegetation
type is characterized by halophytes of the families
Chenopodiaceae and Plumbaginaceae. Most halophytes
are C4 plants, in which the first products of photosyn-
thesis are 4-carbon organic acids, unlike many other
291

On pp. 292-293, a large group of
Vulnerable Afghan Urials (Ovis
orientalis cycloceros) in Badkhyz
Nature Reserve in Turkmenistan.
© Gertrud and Helmut Denzau
Above, collecting firewood is a
major problem, especially along the
river banks.
© Patricio Robles Gil/Sierra Madre
294
plant species. Most of these are fast-growing species
having a high photosynthesis rate. They have adapted
to extreme conditions of dry, saline soils with high tem-
peratures and radiation, as well as water deficiency. In
saline soils, only these physiologically specialized spe-
cies can survive. The leaves of most of these are succu-
lent or have thick cuticles. The root cells have a high
osmotic pressure value to extract water from saline
soils. Excess salt is either secreted or deposited in small
sacs on the leaves. The following are among the rarest
of Irano-Anatolian halophytes having extremely local-
ized distributions: Gladiolus halophilus, Acantholimon
halophilum, Ferula halophila, Asparagus lyconicus, Alli-
um vuralii, Verbascum pyroliforme, Salvia halophila, Sal-
sola stenoptera, Limonium iconicum, L. anatolicum, L.
tamaricoides, Kalidopsis wagenitzii, Hypericum salsugi-
neum, Onosma halophilum, and Taraxacum mirabile
(Vural et al. 1999).
Threats
The single most important threat to the Turkish part
of the hotspot is the development of irrigation schemes
for agriculture and associated infrastructure, such as
dams. The excessive use of water resources for sugar-
beet agriculture in the Konya Closed Basin has resulted
in the loss of many large steppe areas and closed basin
lakes. Elsewhere, Lake Sevan in Armenia and the
Javakheti mountain wetlands in Georgia have largely
been destroyed. In the Ararat Valley alone, 1 500 km2 of
swamps have been drained and transformed into agri-
cultural land. Other threats that have led to extensive
habitat loss across the hotspot include overgrazing,
overharvesting of woody plants (shrubs and trees) for
fuelwood, and mining. Several highly publicized mili-
tary operations have taken place in Iraq, Iran, and
Turkey during the last few decades, resulting in loss of
forests and wetlands (Evans 1994). In northern Iraq,
the few remaining tracts of natural forests in the more
isolated northern mountains are now probably nearly
depleted, largely because of intensive wood-cutting ex-
acerbated by the political situation there.
More than 90% of natural steppes in the region have
disappeared (Eken and Magnin 1999), whereas the
alpine meadows covering higher parts of mountains are
largely intact. In the most humid parts of the region,
the mountains of Eastern Turkey, the remaining forest
is around 12% of its original coverage (Özhatay et al.
2003), but this seems to be much lower in drier parts;
for instance, in Iraq, only 4% of natural forests remain
(Evans 1994). The only pristine parts of the forested
area are found on the inaccessible mountains of south-
eastern Turkey and in the neighboring territories of
Iran. Recent analyses, including the use of satellite im-
agery, indicate that the overall percentage of vegetation
that remains intact in the hotspot is around 15% (Doǧa
Derneǧi Research Unit in Turkey, in litt. 2004).
The doubling of the human population since the ear-
ly 1970s, coupled with increased deforestation and
overgrazing, has resulted in a decline in many species
present in the hotspot. This has been particularly
noticeable with the bustards. Due to the expansion
of agriculture and crop-improvement projects, many
steppe species, such as the great bustard, appear to have
declined dramatically in Turkey and Iran since the
1970s. All the main sites for this species in Iran in
the 1970s occurred within this hotspot, but many may
no longer have any bustards (Evans 1994). Most of the
large ungulates in the hotspot suffer from illegal hunt-
ing and habitat loss.
Conservation
According to the World Database on Protected Areas,
only 6.2% of this hotspot has some form of legal pro-
tection, a percentage which is more than halved (2.9%)
when one considers only those protected areas in IUCN
categories I to IV. The largest protected area in the
hotspot is the Tuz Lake Specially Protected Area in cen-
tral Turkey, covering approximately 7 000 km2; it was
recently proclaimed and still requires a management
plan. The 4 000-km2 Alborz-e-Markazi Reserve in the
central Elburz Mountains is another large protected
area, which extends towards the north into the Cauca-
sus Hotspot. In Turkey, most protected areas are small,
and not very well managed. In Iran, the largest pro-
tected area within the hotspot is Urumiyeh Lake Na-
tional Park (4 640 km2), although there are a number of
sizable protected areas in the Zagros Mountains, in-
cluding Arjan Biosphere Reserve (528 km2). In Arme-
nia, the principal protected areas within the hotspot are
Sevan National Park, as well as the Khosrov, Dilijan,
and Shikahogh State Reserves. The Kopet Dagh region
is reasonably well covered by a number of protected ar-
eas, of which the Meana-Chaachinskiy Sanctuary is the
largest (600 km2). All countries in the hotspot are party
to the Ramsar Convention (with the exception of Iraq)
and World Heritage Convention. There are nine Ram-
sar sites in the hotspot: two in the Turkish part (the Sul-
tansazligi Marshes and the Seyfe Lake), five in Iran
(Urumiyeh Lake, Kobi Lake, Dorgeh Sangi Lakes, Pari-
shan Lake, and Neriz Lakes) and two in Armenia (Lakes
Sevan and Arpi). Turkey is currently in the process of
declaring at least five new Ramsar sites in the hotspot,
based on recent scientific inventories of the country’s
most important biodiversity sites.
An important new concept being tested in the region
is that of Key Biodiversity Areas, which represent the
most important sites for biodiversity conservation world-
wide. Turkey is one of the few countries, and perhaps
the first in the world, which has identified both Impor-
tant Bird Areas and Key Biodiversity Areas. Many Key
Biodiversity Areas in Turkey are not well protected
(85%), although forests and wetlands are relatively bet-
ter represented in protected areas than the more exten-
sive and homogeneous habitats such as steppes and
alpine. In Turkey, the Turkish Nature Society (Doǧa
Derneǧi) has initiated a national conservation program
for all Key Biodiversity Areas in collaboration with gov-
ernmental institutions and other NGOs. Although there
is a large gap in legal protection, most of the key sites for
conservation have been documented, paving the way
for their protection in the future. Currently, Turkey has
18 different categories of protected areas under nine dif-
ferent laws, but these will be harmonized by the end of
2005 through the new Nature Conservation Law.
During the past five years, a strong grassroots move-
ment for nature conservation has developed in Turkey,
particularly among university students and local NGOs
throughout the country. Furthermore, accession to the
European Union is likely to increase enforcement of
nature conservation laws, as well as help to expand the
protected area network that currently has major gaps in
terms of coverage and management efficiency.
The Environmental Protection and Enhancement
Act (1974) is the main law dealing with nature conser-
vation in Iran. Coverage of all the major habitat types
in the four main types of protected areas recognized in
Iran is considered to be good, reflecting the large
amount of fieldwork that has been carried out in the
last 30 years. The Commission on Environmental, Eco-
nomic, and Social Policy of IUCN is based in Iran, and
there are also a number of other NGOs focusing on the
conservation of flagship species in the country. Since
the discovery of the small population of the Persian
subspecies of the fallow deer (Dama dama mesopotami-
ca) in the western foothills of the Zagros Mountains in
the 1950s, reintroduction and captive-breeding projects
have been initiated in northwestern Iran. However, by
1988, the last wild population was on the verge of ex-
tinction, and individuals were taken to establish a semi-
captive herd of 140 deer there. A captive-breeding
group was also established in Germany, and individuals
from there were sent to a reserve in Israel.
There is apparently no legislation to protect sites for
biodiversity conservation in Iraq. There are a number
of wild animal breeding stations, each of a few hundred
hectares, but with little significance for in situ conser-
vation of wildlife. Responsibility for the management of
the natural environment in Iraq lies with the Ministry
of Agriculture. In the past, that ministry has issued
temporary bans on hunting, but with minimal enforce-
ment. In Iraq, a National Forest Foundation has been
established, partly to protect the remaining forests in the
Zagros region. There are several nature parks in Iraq
that are principally state-owned areas for public recre-
ation, but with no wildlife management.
Turkmenistan is a relatively new, independent state
and is in the process of developing its environmental
laws. Although existing nature reserves are contribut-
ing significantly to the overall protection of the Kopet
Dagh’s diverse woodlands, they often lack effective
management (Pereladova et al. 1997). The Turkme-
nistan Biodiversity Strategy and Action Plan was pub-
lished in 2002 (Ministry of Nature Protection of Turk-
menistan 2002). Relevant laws include “On Nature
Protection in Turkmenistan” (adopted in 1991), “On
State Specially Protected Areas” (1992), and the “Provi-
sion on Hunting and Hunting Regulations” (1995).
Lack of expertise in nature conservation is the main
limiting factor throughout the hotspot, and this prob-
lem has to be addressed immediately to elevate its over-
all conservation status. The protected area networks in
Iraq and Turkey require prompt expansion, building on
sites that are known to be important for global biodi-
versity (e.g., Key Biodiversity Areas and Important Bird
Areas). Furthermore, better management of protected
areas is required throughout the hotspot. Although pro-
tected legally, several globally important sites are se-
verely threatened and could be lost if urgent conserva-
tion action is not taken. Perhaps the most intact and
endemic-rich part of the hotspot is the northern part of
the Zagros Mountains —the area where Turkey, Iraq,
and Iran meet. Unfortunately, this area has hosted sev-
eral military operations for many years. At the same
time, however, the region’s globally important biodi-
versity might offer options for international collabora-
tion in alleviating poverty and achieving a lasting
peace. In most areas of the hotspot, there is great po-
tential for direct participation by local people in nature
conservation, as most of the region’s indigenous com-
munities maintain their traditional lifestyles and keep
practicing what are perhaps the most sustainable ways
of using natural resources in this part of the world.
Agriculture, hydropower energy, and forestry are the
key sectors that need to be tackled on a landscape lev-
el and harmonized with the principles of biodiversity
conservation.
Given that it is situated in such a politically sensitive
area, the Irano-Anatolian Hotspot is perhaps among the
least surveyed hotspots. Yet, even the existing scarce in-
formation indicates the importance of this region for
global biodiversity conservation. If the status of this
hotspot’s biodiversity is to be improved, strong interna-
tional cooperation will be essential.
GÜVEN EKEN 19
MIKE EVANS 19 The hunting of wild game by local
AHMET KARATAŞ 20 people for food is another situation
ÖZGE BALKIZ 22 that is wiping out many species
throughout the region.
EVRIM KARAÇ̧ETIN 21
© Patricio Robles Gil/Sierra Madre
TUBA KILIÇ̧ 22
NURI ÖZBAǦDATLI 22
GERTRUD NEUMANN-DENZAU 23
ENGIN GEM 22
AYŞEGÜL KARATAŞ 20
295
 MOUNTAINS
OF CENTRAL ASIA
The Mountains of Central Asia Hotspot consists of
two of Asia’s major mountain ranges, the Pamir (in-
cluding the Pamir-Alai) and the Tien Shan. These are
situated within Kazakhstan, Kyrgyzstan, Tajikistan,
Uzbekistan, western China, extreme northeastern
Afghanistan, and a small part of Turkmenistan. The
total area covered is about 863 362 km2. The highest
peak, Kongur, in the Chinese Pamir, rises to 7 719 m;
four others are above 7 000 m and many more exceed
6 500 m. Glaciers number well over 20 000 and their
total extent covers around 18 000 km2. The moun-
tains were mainly formed by folding due to tectonic
movements during the Caledonian, Hercynian, and
Alpine orogenic (or mountain-building) periods.
Some features also result from faulting and from vol-
canic activity. The hotspot includes major desert
basins, the largest of which is the Fergana Valley.
The elevated massif of the Pamir was known to ear-
ly Persian geographers as Bam-i-Dunya, or “roof of
the world” and is situated at the center of several
great ranges. The Tien Shan lie adjacent to the north,
the Hindu Kush to the southwest, the Karakoram to
the southeast, and the Kun Lun to the east. The
Pamir extend east to the isolated Muztagh Ata Massif
in western China and south to the Wakhan Valley of
northeastern Afghanistan. The northern rim of the
Pamir is formed by the Trans-Alai Range that drops
steeply to the Alai Valley, a deep fault trench carry-
ing the waters of a major tributary of the Amu Darya,
and dividing the Pamir from the Gissar-Alai moun-
tain system.
The Eastern Pamir have a mean elevation of over
4 000 m and are plateau-like in character, with limit-
ed elevational variation and even relief. The surface
is crossed by broad, shallow, valleys or pamir, that
give the name to the whole range. The Western Pamir,
by contrast, are characterized by sharp ridges and
steep slopes cut by deep valleys and gorges. They have
great elevational variation and typical alpine relief.
The Western Pamir are heavily glaciated, and include
the Fedchenko Glacier, which is over 75 km in length
and one of the two longest glaciers in the world out-
side the polar regions. The highest peaks in the
KAZAKHSTAN
Lake
Balkhash
KYRGYZSTAN
UZBEKISTAN
CHINA
TAJIKISTAN
TURKMENISTAN
AFGHANISTAN PAKISTAN 0 400 km
Pamir are also situated in this part of the range: Gar-
mo, formerly Peak Communism (7 498 m), and Peak
Lenin (7 165 m). To the west and southwest, a series
of ridges radiates out into the desert and these extend
across the Pyanj River into Badakhshan Province of
northern Afghanistan.
The Pamir-Alai Mountains, including the Gissaro-
Alai system, lie between the Pamir and the Fergana
Valley, a deeply downfaulted desert basin, about
300 km long and 150 km wide. The Gissaro-Alai sys-
tem extends roughly northeastwards to join the Tien
Shan. Its ridges are generally around 5 000 m.
To the north of the Fergana Valley, the Tien Shan
—Chinese for “heaven” or “sky”— Mountains extend
for 2 500 km from west to east. The Tien Shan are
made up of a complex series of ranges and are around
300 km wide in the center, narrowing at the eastern
and western ends. The highest peaks are located in a
central cluster on the borders of China, Kyrgyzstan,
and Kazakhstan, and include Peak Pobeda or Tuo-
muer Feng (7 439 m) and Han Tengri (6 995 m). The On the opposite page,
Inylchek Glacier, over 50 km long and the largest in the spectacular Tien Shan
the Tien Shan, is also located in this part of the range. Mountain range of Kyrgyzstan.
© Patricio Robles Gil/Sierra Madre
Across the fertile Ili Valley, the Borohoro Shan links
the Dzungarian Alatau (4 464 m) Range to the Tien
Shan. Glaciers occupy more than 10 000 km2 and
occur along most of the range, east to the Bogda Shan
in the Chinese Tien Shan. The central Tien Shan,
with a mean altitude of over 3 000 m, contains a high,
297

Primula minkwitziae, a rare
species of primrose found in the
western Tien Shan Mountains of
southern Kazakhstan.
© Vladmir Kolbintsev/naturepl.com
298
uplifted massif that shares some of the same landscape
features as the Eastern Pamir. On the northern side of
the Tien Shan, lower arid ranges such as the Nuratau,
Chu-Ili, and Karatau run away northwestwards into the
desert. The Tien Shan drains mainly to the north, and
the many streams plunging down the steep northern
slopes have formed alluvial deposits on the plains be-
low which provide sites for settlements; several major
population centers are located there.
The climate in this region is generally arid. The Tien
Shan acts as a climatic divide and intercepts moist win-
ter air from the north and west, and prevents it from
reaching the hyperarid Tarim Basin to the south. Pre-
cipitation falls mainly in winter and spring, and varies
from over 1 500 mm in the Gissar Range in the west of
the hotspot, to below 100 mm in the Eastern Pamir.
The southwest of the area —the western parts of the
Tien Shan and of the Pamir-Alai— is influenced by sub-
tropical air and also enjoys the mildest winters. Tem-
peratures decrease to the east, although there are con-
siderable variations due to altitude. The high plateaus
of the Eastern Pamir and central Tien Shan are the
coldest areas, having an annual mean temperature be-
low zero and a very short growing season. Winter tem-
peratures there may reach –40ºC, and discontinuous
patches of permafrost occur.
The geological origin of the mountains, the wide al-
titudinal range, and the extreme climatic variation have
combined to produce great landscape and biotic diver-
sity. Altitudinal zonation throughout the hotspot is not
uniform, and there is no continuous forest belt. The
number, extent, and sequence of vegetation zones vary
across the hotspot depending on temperature and mois-
ture gradients, slope aspect, and altitude and latitude.
Desert, semidesert, and arid steppe vegetation types
predominate on all the lower slopes, foothills, and in
some of the outlying ranges, as well as in major basins
such as the Fergana Valley and the basin of Issyk-Kul
Lake. Common plants here include species of wide-
spread genera such as Artemisia, Salsola, and Ephedra,
as well as annual grasses such as Poa and Festuca spp.
In the Ili Valley and a few other places, patches of river-
ine woodland survive, composed of poplar (Populus ta-
lassica), ash (Fraxinus sogdiana), an endemic apple
(Malus sieversii), apricot (Prunus armeniaca), Eleagnus,
tamarisks (Tamarix spp.), and willows (Salix spp.).
Steppe communities, dominated by various species of
grasses and herbs, occur at higher altitudes. A distinc-
tive type of tall-grass steppe, characterized by Elytrigia
trichophora and Hordeum bulbosum, occurs in the west-
ern Tien Shan and Pamir-Alai. Shrub communities are
widespread in the lower steppe zone and may form
dense thickets in gorges. Species present include haw-
thorns (Crataegus pontica, C. turkestanica), Cotoneaster
melanocarpa, Euonymus semenovii, Lonicera spp., Rosa
spp., and Berberis spp., with some pistachio (Pistacia
vera) and iron tree or hackberry (Celtis caucasica). The
area occupied by shrubs has declined markedly due to
cutting for fuelwood.
A type of walnut-fruit forest unique to Central Asia
grows above the steppe zone in warm, sheltered places
in the western Pamir-Alai and Tien Shan. These are
very diverse and are composed of walnut (Juglans regia),
almonds (Amygdalus communis and A. bucharensis),
pears (Pyrus korshinskyi and P. regelii), plums (Prunus
sogdiana and P. ferganica), cherry (Cerasus mahaleb),
and apple, along with maples (Acer turkestanicum and
A. semenovii). A few Chinese walnut (Juglans cathaya-
na) trees survive at one locality in the eastern Tien
Shan. This valuable and ancient forest type contains an-
cestors of domestic fruit varieties and is an important
storehouse of wild genetic diversity. Some of the sur-
viving walnut trees are estimated to be 800 years old.
The area occupied by this habitat has greatly declined,
with around 90% lost during the last 50 years.
Spruce forests occur on moist northern slopes of the
Tien Shan, the only coniferous forest type in the Moun-
tains of Central Asia. These occur sporadically along
most of the range, east as far as the Karlik Tag. They
grow in a broad altitude band between 1 700 m and
2 700 m and are dominated by the endemic Schrenk’s
spruce (Picea schrenkiana). Some silver fir (Abies seme-
novii) occurs and associated species include the en-
demic Tien Shan rowan (Sorbus tianshanica), aspen
(Populus tremula), willow (Salix xerophila), and birches
(Betula spp.). Stands of closed-canopy forest are found
in patches of varying size, with the largest on the Kyr-
gyz Range. More open stands also occur in a forest-
meadow mosaic.
Open juniper or archa forest occurs widely between
900 m and 2 800 m. In the Tien Shan it grows above the
spruce belt and is composed of Juniperus seravschanica,
J. turkestanica, and J. semiglobosa. Juniper woods are
also found at lower altitudes in the Pamir-Alai and
Western Pamir.
Subalpine and alpine meadows occur from 2 000-
4 000 m and above in the western part of the moun-
tains. Plant cover is high, with a tight sward made up of
grasses such as Poa alpina, sedges (Carex and Kobresia
spp.) and carpeted with a rich variety of herbs includ-
ing many endemic species. The fritillary (Rhinopetalum
stenantherum), gentians (Gentiana spp.), globeflower
(Trollius dshungaricus), primulas (Primula spp.), tulips
(Tulipa spp.), anemones (e.g., Anemone protracta), louse-
worts (Pedicularis spp.), and aconites (Aconitum talas-
sicum, A. leucostomum) are prominent among them.
These meadows are at their most attractive in early
summer when the flowers are in full bloom. In drier ar-
eas of the Pamir and central Tien Shan, the montane
meadows are replaced by high-altitude steppes, charac-
terized by grasses such as Festuca valesiaca, Poa attenu-
ata, species of Stipa and Puccinellia, sedges (Carex and
Kobresia spp.), together with a rather sparse cover of xe-
rophytic perennial herbs.
Vegetation cover and plant diversity declines rapid-
ly as one approaches the upper limits of plant cover,
and cushion plants and those with low rosettes that
can withstand the high winds, cold temperatures, and
aridity become more common. Acantholimon diapen-
sioides is the most widespread cushion plant and spe-
cies of Saxifraga, Androsace, Rhodiola, Saussurea, and
Tanacetum are also frequent. At 4 000 to 4 500 m, even
more hardy perennials are found, such as Thyla-
cospermum caespitosum, whose large, tight cushions re-
semble a moss more than a herbaceous plant, and
Dryadanthe tetrandra. Snow patch plants also include
attractive species, such as the alp lily (Lloydia seroti-
na), the large, pale blue and white globeflower (Trollius
lilacinus), and several crucifers (Draba spp.). At 4 200-
4 500 m in the Eastern Pamir and central Tien Shan,
the vegetation has a tundra-like character similar to Ti-
bet, with sedge meadows dominated by species of Ko-
bresia and Carex in mesic areas along valley bottoms.
Above this, there only a few lichens and rare algal
films on some glaciers.
The main ethnic groups in the region are Kazakh,
Kyrgyz, Uzbek, Uighur, and Tajik. The Kazakh, Kyrgyz,
Uzbek, and Uighur live throughout the mountains of
Central Asia and speak a series of Turkic languages.
The other major ethnic groups are the Tajiks and Moun-
tain Tajiks, who inhabit the Pamir and parts of Pamir-
Alai. They are subdivided into a number of groups, one
of which is descended from the ancient Sogdians, and
speak a series of dialects belonging to the eastern Iran-
ian group of languages. Wakhi-speaking communities
live in northeastern Afghanistan and neighboring areas
of China. There are also a small number of Mongols
and there is a Mongol county in the eastern Tien Shan.
Added to these are many Russians in the western part
of the area and Chinese in the eastern Tien Shan, es-
pecially in major towns and oases in the plains sur-
rounding the ranges. Livestock herding (mainly sheep
and goats), with camels at lower and yaks at higher al-
titudes, has traditionally been important among the
peoples of these mountains. The Fergana and Ili valleys
are heavily cultivated, with cotton an important crop in
Fergana. Many people moved out of the mountains dur-
ing the communist era to work in cotton communes.
Biodiversity
The flora of this hotspot is a mix of Boreal, Siberian,
Mongolian, Indo-Himalayan, and Iranian elements.
Over 5 500 species of higher plants are known, around
1 500 (27%) of which are endemic. There are 64
endemic genera, of which 21 are from the Umbellifer-
ae and 12, from the Compositae. Plant species rich-
ness and the number of endemics vary widely through-
out the area covered. The Karatau Mountains, a low
desert range (2 176 m) that runs north from the Tien
Shan, contains 1 700 species and 153 endemics, and
the Dzhungarian Alatau, a relatively small range at the
northern end of the Tien Shan, rising to only 4 464 m,
holds 2 130 species including several of Siberian origin
(Davis et al. 1995). The Pamiro-Alai contain over 4 000
species, while the Eastern Pamir have only 700, most-
ly species with Central Asian affinities (Davis et al.
1995). These include endemic high-altitude species
such as Braya pamirica, Oxytropis bella, and Astragalus
alitshuri.
The endemic flora includes several tree species,
grasses such as Atraphaxis muschketovii and Stipa ka-
ratavica, and many herbs. There are many species of
wild onion, among them Allium pskemense, a very rare,
large onion found only in a small part of the Pskem
Range of the Western Tien Shan. Primula minkwitziae
only grows above 3 000 m in the northern Tien Shan,
and Niedzwedzkia semiretchenskia is restricted to an area
of only 7 ha in the Chu-Ili Mountains.
Among vertebrates, 143 species of mammals occur
in the hotspot, including a rich suite of mountain un-
gulates. The Siberian ibex (Capra sibirica) is the most
numerous and most widespread species, occurring in
all parts of the area above the treeline. Very small pop-
ulations of the Tadjik markhor (C. falconeri heptneri,
CR) occur at lower altitudes in the Kugitang Mountains
at the western end of the Gissar-Alai system, and in a
few parts of the southwest Pamir in Tajikistan. Three
endemic subspecies of argali (Ovis ammon, VU) are
found here, while the blue sheep (Pseudois nayaur), a
typical Tibetan and Trans-Himalayan species, reaches
the southeast corner of the hotspot. The foothills of the
hotspot hold desert species such as goitered gazelle
(Gazella subgutturosa). Mammal endemism is generally
low in this hotspot, with only six species found no-
where else, namely Menzbier’s marmot (Marmota
menzbieri, VU), two susliks or ground squirrels (Sper-
mophilus ralli and S. relictus), the Pamir shrew (Sorex
bucharensis), the Ili pika (Ochotona iliensis, VU), and the
Alai mole vole (Ellobius alaicus, EN). The last men-
tioned species is known only from the Alai Mountains,
in southern Kyrgyzstan.
There are 493 bird species that can be considered to
occur regularly in this hotspot but, incredibly, there are
no endemics. Many species belong to genera typical of The frog-eyed sand gecko
the high ranges of Asia, such as Phoenicurus, Prunella, (Teratoscincus scincus) is a
and Carpodacus. Three species of snowcocks (large game species adapted to the extreme
environment of the desert and arid
birds of the genus Tetraogallus) occur: T. altaicus in the
habitats of Central Asia.
Dzhungarian Alatau, and T. himalayensis and T. thi- These animals are also commonly
betanus in the Pamir and Tien Shan. Coniferous forests kept as pets.
on the northern side of the Tien Shan form the southern © David Northcott/DRK PHOTO
limit of the range of several boreal species in continental
Asia, including the crossbill (Loxia curvirostra), black
grouse (Lyrurus tetrix), and northern hawk owl (Surnia
ulula). Desert birds, including threatened species such as
the great bustard (Otis tarda, VU) and houbara bustard
(Chlamydotis undulata) occur in the low-altitude zones.
299
 The hotspot represents a stronghold for birds of prey.
Twenty-one species breed here and there are important
breeding populations of several species, some of them
globally threatened. These include the golden eagle
(Aquila chrysaetos), imperial eagle (A. heliaca, VU),
steppe eagle (A. rapax), booted eagle (Hieraaetus pen-
natus), lammergeier (Gypaetus barbatus), black vul-
ture (Aegypius monachus), Eurasian griffon (Gyps ful-
vus), Himalayan griffon (G. himalayensis), peregrine
falcon (Falco peregrinus), and saker falcon (F. cherrug).
Several more species pass through the hotspot on mi-
gration. Besides birds of prey, the high lakes in the
Pamir are the breeding grounds for the brown-headed
gull (Larus brunnicephalus) and bar-headed goose (An-
ser indicus), while Lake Bayanbulak in the eastern
Tien Shan holds a breeding colony of whooper swans
(Cygnus cygnus).
Some 59 species of reptiles are known to occur in
this hotspot, only one of which is endemic, a skink,
Asymblepharus alaicus. Not surprisingly, reptile diver-
sity is highest in the lower altitudinal range, in desert
and semidesert areas. There are ten species of Ere-
mias lizards and eight toad-headed agamas (Phryno-
cephalus spp.). Only nine species of amphibians have
been recorded, reflecting the overall aridity, four of
them being endemic. A salamander (Ranodon sibiri-
cus) is found only in the Dzhungarian Alatau Range at
the northern end of the Tien Shan and represents an
endemic genus, while two recently described species,
a toad (Bufo shaartuensis) and a frog (Rana terentievi),
are known only from southern Tajikistan, but may
also occur in adjacent parts of Afghanistan (Borkin
1999).
The hotspot’s high-gradient streams and oligotrophic
lakes have a very limited fish fauna, but a significant
degree of endemism at the species level. Endemism is
centered in the Lake Issyk-Kul Basin of Kyrgyzstan,
which lacks a natural outlet connecting it with any oth-
er body of water. The fish fauna of the hotspot consists
of 27 species, with five endemics. The most notable
hotspot endemic is the Kugitang blind cave fish (Troglo-
cobitis starostini), which is restricted to a small part of
the Kugitang Mountains at the southwestern end of the
hotspot.
The alpine meadows are also rich in insect diversity,
although full inventory data is lacking for many groups.
An estimated 25% of Kyrgyzstan’s 10 000 invertebrates
are endemic (Ministry of Environmental Protection
1998). The large and attractive butterflies of the family
Papilionidae, which comprises swallowtails and apollos,
are one of the best-known invertebrate groups. Twenty-
six species of apollo butterflies occur in the hotspot and
eleven of them are endemic, some having very limited
distributions in parts of individual ranges. There are
also 87 endemic mollusks, including the Kokand fresh-
water clam (Colletopterum kokandicum), which is re-
stricted to one lake in the Fergana Valley. In addition,
Arkhangelskii’s spring snail (Valvatamnicola archangel-
skii) is known only from a few cold-water springs in the
Alai Range; V. schahimardanica occurs in a single popu-
lation in the Alai Range; and Zhadin’s spring snail (Sog-
damnicola shadini) is found only at mid-altitudes in the
Zeravshan Range.
Flagship Species
The Mountains of Central Asia play a key, connecting
role in the distribution of many montane Asian species
through their central location in the continent. Conse-
quently, the hotspot shares many of its flagship species
with the Himalaya Hotspot. Perhaps the best example
of this is the snow leopard (Uncia uncia, EN), univer-
sally regarded as the most charismatic and symbolic an-
imal in the Asian mountain fauna. Its reclusive, wary
nature and mainly crepuscular and nocturnal habits
mean that it is rarely observed in the wild, although its
distinctive field signs can be discerned on paths, ridge-
lines, the base of cliffs, and the bottom of gorges, and
are used to establish its presence in a particular area
and to monitor its movements. The snow leopard is dis-
tributed throughout the Mountains of Central Asia,
which once formed a stronghold for the species, al-
though it has declined in numbers here as elsewhere.
The snow leopard’s population has decreased since the
break-up of the former Soviet Union. Its fur is valued in
some quarters, making it a target for hunters, and the
species is also killed on occasion by livestock owners,
but the main cause is the depletion of its prey base
through illegal hunting. Snow leopards prey on Siberian
ibex, argali, roe deer, red deer, wild boar, smaller prey
such as marmots, and snowcock. They usually inhabit
the alpine and subalpine zones, and have been record-
ed at over 5 000 m in the Pamir, but will follow ibex and
other prey down to lower altitudes at the forest edge. In
some lower ranges it lives all year at 1 500-2 000 m and On the opposite page, the
in spurs of the Dzhungarian Alatau it occurs at 600- Marco Polo sheep (Ovis ammon
1 800 m. The Mountains of Central Asia serve as a key polii) occurs in hilly terrain near
high mountains at elevations of
link in the western part of the range of this species, al-
1 300 to over 6 000 m throughout
lowing movement of individuals and genetic inter- Central Asia. The males are famous
change between populations in the Hindu Kush, for their long horns, which can
Karakoram, Kun Lun, Altai, and Tibet. reach up to 178 cm long.
As noted earlier, this hotspot has a rich array of un- © Patricio Robles Gil/Sierra Madre
gulate species, many of which serve as important flag-
Above, the red deer
ships. Argalis are the largest of the Eurasian wild
(Cervus elaphus) is a widespread
sheep, with a distribution confined to Central Asia, the species in the Northern Hemisphere.
Himalaya, and Tibet. Their enormous horns have long In the Tien Shan Mountains,
made them a prized trophy. Three subspecies are it finds refuge in canyons covered
found in the Mountains of Central Asia, each one en- with deep forest.
demic to a particular range. The Karatau argali (Ovis © Patricio Robles Gil/Sierra Madre
ammon nigrimontana, CR) is confined to the Karatau, a
relatively low desert range stretching northwest from
the Tien Shan, and may now number no more than
200. The Tien Shan argali (O. a. karelini, VU) occurs
301

On pp. 302-303, although the
Siberian ibex (Capra sibirica)
lives at high elevations, often up to
the vegetation line at 5 000 m,
it generally seeks out lower slopes
during the winter. When the snow
cover is heavy, it must find food by
pawing at the snow with its forelegs
to reach the vegetation concealed
below.
© Patricio Robles Gil/Sierra Madre
Above, young wolf trapped by
hunters in Kyrgyzstan. In cold
winters, wolves in packs have
attacked humans.
© Patricio Robles Gil/Sierra Madre
On the opposite page, the fantastic
horns of the male Marco Polo sheep
are highly prized by hunters, and
this has put pressure on some wild
populations, although a more
serious threat is habitat loss from
the grazing of domestic sheep.
© Patricio Robles Gil/Sierra Madre
304
along the Tien Shan in suitable habitat, while the Mar-
co Polo sheep (O. a. polii, VU), whose magnificent curl-
ing horns have long made it a favored target of trophy
hunters, occurs in the Pamir. It is also under great
pressure for its meat, and has disappeared from
around villages and areas easily accessible by road
(Shackleton 1997).
A few of the smaller mammal species deserve men-
tion as flagship species. Menzbier’s marmot is a dis-
tinctive, small species of marmot that is endemic to
western Tien Shan, where it inhabits the alpine zone
and higher parts of the subalpine zones, usually above
2 000 m. Its total range covers only about 1 600 km2,
the smallest distribution of any Old World marmot,
and this is split into two isolated areas, each one occu-
pied by different subspecies. The Ili pika is a small
species of lagomorph restricted entirely to the Chinese
Tien Shan, where it inhabits taluses at altitudes of
2 800 to 4 100 m. Finally, the corsac fox (Vulpes corsac)
is a typical Central Asian species which occurs in lower
steppe and semidesert areas. The corsac is considered
a valuable fur-bearer species, and has been hunted for
centuries, although hunting today is regulated by na-
tional legislation in those countries where it is regard-
ed as a fur-bearer species (such as Kazakhstan and
Uzbekistan).
Tulips (Tulipa spp.) are one of the most attractive
and most valued groups of flowers in the Mountains of
Central Asia. Over 16 endemic species grow in the
steppe and meadow zones. The largest is Greig’s tulip
(Tulipa greigii), sometimes known as the king of the
tulips, a rare plant with a brilliant orange-red flower
and that is only found in western Tien Shan, where it
grows up to about 2 400 m. The late-flowering T. kauf-
manniana is also restricted to western Tien Shan. Its
beautiful white flowers have an almost open shape,
sometimes suffused with orange or gold. Another late-
flowering species is the delicate yellow T. dasyste-
monoides. Two more species from the western Tien
Shan and Pamir-Alai are T. kolpakowskiana and T. os-
trowskiana, which sometimes hybridize, producing
swathes of red, yellow, and orange blooms across the
hillsides. Tulipa zenaidae is endemic to the Kyrgyz
Alatau Range in the Tien Shan, and T. regeli to the Chu-
Ili Mountains. The beauty of these flowers has led di-
rectly to the decline of many species through collecting
for horticulture and decoration.
Threats
The Mountains of Central Asia have long been exploit-
ed for grazing, food, timber, and fuel. The human pop-
ulation of the hotspot numbers around 20 million peo-
ple, with many more living in the adjoining plains.
Population density in the Fergana Valley is around 400-
500 per km2, but fewer than two people per km2 live
in the Pamir. A steady rise in the human population
and domestic livestock, and the associated need for
land and resources, have inexorably increased pressure
on the environment, which has reached unsustainable
levels in many places. This trend has become much
more severe over the last 10-15 years as a result of po-
litical and economic changes in the countries of the for-
mer Soviet Union. The transition to a market economy,
and withdrawal of government subsidies, have created
an extremely difficult and unfavorable economic situa-
tion in which a large proportion of the population cur-
rently lives at subsistence level or below the official
poverty line. This leads directly to intensified use of
natural resources to meet peoples’ needs. Habitat de-
struction, overgrazing, and unregulated hunting of ani-
mals and collection of plants emerge as the three major
and continuing threats.
Virtually all the land in the lowland desert belt and
in many foothill areas has been converted to agricul-
tural use, mainly for cultivation of cotton and other
crops aided by irrigation. The mountain ranges fulfill a
vital water-storage role through their glaciers and win-
ter snow, which enables much of the cultivation on the
plains below to take place. Furthermore, as coal and
other fuel becomes unavailable and unaffordable, the
cutting of trees and shrubs for fuel and building timber
has been exacerbated. This, together with forest fires,
has greatly reduced the area of these habitats, especial-
ly in the case of the steppe shrub communities and the
unique and valuable walnut-fruit forests. As an exam-
ple, between 1995 and 1998, more than 4 500 km2 of
forest in Kazakhstan were lost due to fires. The quality
of remaining forests is affected by overgrazing and oth-
er factors that prevent or slow down natural regenera-
tion, and many of them are dominated by mature trees.
In Kyrgyzstan, the remnant walnut forests are threat-
ened by wind-borne pathogens. Expansion of settle-
ments, construction of roads and other infrastructure,
recreational facilities, mining (deposits of oil, coal, iron,
copper, lead, zinc, and phosphates are all exploited),
and other economic activities also destroy and frag-
ment natural habitats. Much of the hotspot remained
accessible only by foot or on horseback until fairly re-
cently, but roads have opened up wide tracts of the
mountains, and facilitated exploitation and increasing
disturbance.
Numbers of domestic livestock throughout the re-
gion have increased sharply, and overgrazing affects
many areas, especially the foothills and lower slopes,
but also the alpine and subalpine meadows. Severe
degradation is localized around camps and settlements,
but a much wider area is affected in less visible ways.
Overgrazing steadily reduces the fresh grass yield and
causes changes in species composition, with increasing
predominance of less palatable species. This reduces
the productivity of alpine meadows and the number of
wild herbivores they can support, and increases the

Woman milking a cow. When
lacking resources, these mountain
people also use the manure of
domestic sheep and cattle as fuel
for cooking.
© Patricio Robles Gil/Sierra Madre
306
risk of soil erosion. Pastures throughout the hotspot
have been degraded. In one part of eastern Tien Shan,
livestock numbers doubled in the last 50 years, and se-
rious overgrazing and pastureland degradation began as
early as the 1970s (Zhang et al. 2002).
Poaching, especially of larger mammals and birds,
has increased sharply since the break-up of the former
Soviet Union. Mountain ungulates have been increas-
ingly seen as a source of food, and snow leopard num-
bers in Kyrgyzstan are estimated to have fallen by 75%
during the 1990s as a result of heavy hunting pressure
on them and their prey (Koshkarev and Vyrypaev
2000). Other species, too, are killed or captured for
profit; for example, falcons are exported to the Middle
East, where they fetch a high price when sold to fal-
coners. Unregulated collection of plants poses a direct
threat to several species and impoverishes the diver-
sity of meadows. Villagers pick flowers to sell by moun-
tain roadsides, beginning with crocuses and Juno spe-
cies from late February on, then moving on to tulips and
other groups through spring and summer. Crocuses
and tulips have disappeared or become very rare in sev-
eral areas and some endemic species are threatened
with extinction as a result. Collection of plants for me-
dicinal use (of which there are around 200 species) and
horticulture are similarly uncontrolled and operate on
a semi-commercial scale.
A number of other factors also affect the region. Civ-
il conflict in Tajikistan during the 1990s and war in
Afghanistan posed a direct threat to wildlife, while a
more recent phenomenon affecting large animals is
the siting of minefields along some international bor-
ders, with anecdotal reports indicating that snow leop-
ards and other species have been killed as a result.
Damming, reservoir construction, and irrigation have
disrupted water supplies and drainage systems. Many
wetlands have been drained to create cultivable land,
and some watercourses are polluted by mining resi-
dues. Overfishing and the introduction, or accidental
release, of alien species have further damaged fresh-
water ecosystems. Populations of endemic fish in Lake
Issyk-Kul have declined as a result of overfishing and
predation by introduced species.
The recreational load on montane ecosystems is
growing as increasing numbers of local and overseas
tourists visit the region and impact on the environment
through their various activities, especially in the Pamir-
Alai and Tien Shan. Accommodation facilities, camp-
sites, access roads, and infrastructure for skiing and
other activities further encroach on habitats and add to
the disturbance. The region has also become more ac-
cessible to overseas hunters. Many mountain ungulate
species are shot for trophies, and this activity has risen
during the last 15 years. In some cases, hunting permits
have been granted in direct contradiction to existing
protection laws.
In light of the above, natural vegetation communi-
ties in most parts of the hotspot are found only within
protected areas, and even in some of these they are un-
der increasing threat. Almost 30% of the land in the
Mountains of Central Asia has been completely trans-
formed as a result of human activities (Tishkov 2002).
In Kyrgyzstan, many ecosystems have been degraded
and over half of the forest cover has been lost in the
last 50 years (Ministry of Environmental Protection
1998), while 90% of forests in Tajikistan have disap-
peared during the last 100 years (USAID 2001d). Taking
together the analyses in national biodiversity strategy,
as well as action plans and ongoing work on the Mil-
lennium Ecosystem Assessment (GEF/UNDP 1998;
Ministry of Environmental Protection 1998; Regional
Ecological Assessment Centre of Central Asia 2003;
USAID 2001a, b, c, d), it is estimated that no more than
20% of the original vegetation of the hotspot remains
in an intact state.
Finally, a more insidious threat, and one whose con-
sequences may be of particular relevance in both this
and the Himalaya hotspots, is the long-term effect of
global warming, which has the potential to cause radi-
cal changes to the environment and the biota currently
present. Glaciers in the area are estimated to have
shrunk by nearly 20% during the last 30-35 years (Re-
gional Ecological Assessment Centre of Central Asia
2003). In the longer term, the destabilizing effects of
the melting of frozen upper slopes are unknown. Many
montane taxa are predicted to decline and some will
disappear from the area altogether.
Conservation
Overall, protected area coverage in this hotspot is poor.
According to the World Database on Protected Areas,
about 6.9% of the total area has some form of official
protection in nature reserves and other protected areas,
although many reserves are small and isolated. In ad-
dition, some protected areas that have been gazetted
are not fully functioning entities and conservation ef-
fectiveness in many is tenuous. In all the countries of
the former Soviet Union, there has been a breakdown
in former protection regimes such as the dramatic re-
duction in funding staff salaries, patrolling, and other
management activities. This has an effect at all levels,
from strategic planning, day-to-day management, and
combating illegal activities to research.
The size of protected areas varies from 11 ha in the
Chinese Walnut Nature Reserve to Pamir National Park
in Tajikistan and the Taxkorgan Nature Reserve in Chi-
na, in the southeast of the hotspot, both of which cov-
er 15 000 km2. Other large protected areas include Is-
syk-Kul Biosphere Reserve (6 298 km2), in Kyrgyzstan;
and Mount Tomur Nature Reserve (1 000 km2) and
Boghdad Mountain Biosphere Reserve (1 000 km2),
both in the Chinese sector of the Tien Shan. Other
notable reserves in the countries of the former Soviet
Union include Aksu-Dzhebagly (750 km2) in western
Tien Shan, Sary-Chelek (238 km2) in Kyrgyzstan, and
Kugitang Strict Nature Reserve in Turkmenistan. Inter-
national borders often follow mountain ridges, and the
need for transboundary reserves is increasingly being
recognized within the region.
The legislative framework to protect wildlife in the
region is not comprehensive and requires strengthen-
ing. More importantly, enforcement of existing laws is
frequently inadequate or even non-existent in some
cases. National Biodiversity Strategy and Action Plans
have been compiled for Kazakhstan, Uzbekistan, Turk-
menistan, and Kyrgyzstan, and all the countries cov-
ered by the Mountains of Central Asia Hotspot have
compiled national Red Data Books, with the exception
of Afghanistan. Regional cooperation between the
countries of the former Soviet Union has increased
and encompasses a Central Asian Mountain Informa-
tion Network and a Regional Red List program, funded
by a European Union program, which aims to coordi-
nate assessments and set up a database of threatened
species.
Many international donors are involved in the re-
gion. The Global Environment Facility’s Western Tien
Shan Project was established to support conservation of
biodiversity in three counties of western Tien Shan.
Other international NGOs currently active in the re-
gion include WWF (Central Asia program), Fauna and
Flora International (working on community develop-
ment and environment programs, capacity building
and assistance with the development of national biodi-
versity strategy and action plans). The International
Snow Leopard Trust and NABU, a German nature con-
servation organization, are working on snow leopard
conservation in Kyrgyzstan. NABU is also involved
in project work around Lake Issyk-Kul and a project in
Nurata Nature Reserve in Uzbekistan. A coordinated re-
gional snow leopard program has been initiated and
“Asia-Irbis,” a regional working group on conservation
of the snow leopard in the region, has been established.
The governments of Switzerland and Kyrgyzstan have
instituted a forestry project to manage remnant walnut
forests in Kyrgyzstan, and the Central Asia Mountain
Programme (CAMP) has been set up by University of
Berne with financing from the Swiss Agency for Devel-
opment and Cooperation (SDC) to promote sustainable
use of pastures in Central Asia. A Regional Ecological
Centre for Central Asia has been established, and has
begun an assessment of the ecosystems of Central Asia
and the extent of human influence on them at local, na-
tional, and regional levels.
Initiatives have also been undertaken to address
wider issues affecting the environment. A Global Moun-
tain Summit, held in Kyrgyzstan during 2002, the In-
ternational Year of the Mountains, explored united
pproaches for mountain development. The Asian De-
velopment Bank and the Swiss government sponsored
a Regional Strategy for Sustainable Development of the
Mountain Regions of Central Asia. Awareness and edu-
cation programs aimed at the inhabitants of mountain
villages and schools, as well as publicity in the mass
media have been initiated. And, finally, ecotourism and
trekking initiatives have been launched, which can, if
properly planned, contribute to the local economy and
gain support for conservation. Well-managed trophy
hunting programs, while controversial, have the poten-
tial to do the same.
The Mountains of Central Asia contain unique
ecosystems and great landscape and biotic diversity.
They are the traditional home of several different eth-
nic groups, and provide an economic resource and a vi-
tal source of water to the plains below. Wide areas
have been degraded to a greater or lesser extent, and
pressure on all ecosystems is heavy and rising. The
major challenge for the future is to devise and im-
plement measures to reduce human land use to sus-
tainable levels and to conserve the rich biological di-
versity. This will call for imaginative solutions and a
partnership between local communities, conservation-
ists, and governments, as well as considerable finan-
cial resources.
DAVID MALLON 124
ELENA KREUZBERG-MUKHINA 125
LENA BYKOVA 125
ALEXANDER KREUZBERG 125
Sheep herders in their traditional
tent-like yurts, in the high valley of
the Tien Shan Mountains of
Kyrgyzstan.
© Patricio Robles Gil/Sierra Madre
307
 HIMALAYA
The Himalaya include all of the world’s mountain
peaks that exceed 8 000 m in height —including the
world’s tallest, Sagarmatha (Mt. Everest)— and sever-
al of the world’s deepest river gorges. This is the
youngest and the highest mountain chain on Earth,
and stretches in an arc over 3 000 km across northern
Pakistan, Nepal, Bhutan, the northwestern and north-
eastern states of India adjoining Myanmar, and the
southwest China border in the east. This immense
mountain range has been divided into two regions:
the Eastern Himalaya, which covers parts of Nepal,
Bhutan, the northeast Indian states of West Bengal,
Sikkim, Assam, and Arunachal Pradesh, southeast Ti-
bet (China), and northern Myanmar; and the Western
Himalaya, covering the Kumaon-Garhwal, northwest
Kashmir, and northern Pakistan (Mani 1994). While
these divisions are largely artificial (Mani 1994), the
deep defile carved by the antecedent Kali Gandaki
River between the Annapurna and Dhaulagiri moun-
tains has been an effective dispersal barrier to many
species, and provides a biogeographic basis for defin-
ing the distinct Eastern and Western Himalaya along
the length of the mountain range (Wikramanayake et
al. 2002). In total, the area covers some 741 706 km2.
Biogeographically, the Himalayan Mountain Range
straddles a transition zone between the Palearctic and
Indo-Malayan realms, with species from both con-
tributing to its biodiversity. Understanding the distri-
bution of biodiversity requires some knowledge of the
genesis of the Himalaya, and the complex geological
and physical features that influence patterns of bio-
diversity (Molnar 1986). At about 45 million years, the
Himalayan Mountain Range is geologically young.
The mountains were formed as a result of geologic
faulting during the massive collision between Eurasia
and the northward-drifting Deccan Plate, which de-
tached from the southeastern margin of Africa more
than 200 million years ago. The subduction of the
Deccan Plate raised the southern margin of Eurasia
and, because the Deccan Plate is still moving north-
ward, both Tibet and the Inner Himalaya continue to
be pushed upwards even today.
The Himalayan Range now exerts considerable in-
fluence on weather patterns throughout most of South
Asia. During the summer, a zone of low pressure
forms over the Asian landmass, inducing a moisture-
laden monsoon wind from the Bay of Bengal to be fun-
neled through the Ganges River Valley. The winds del-
uge the eastern extent of the mountain range, while
0 500 km
PAKISTAN
CHINA
NEPAL
INDIA
BHUTAN
MYANMAR
(BURMA)
BANGLADESH
the western extent remains drier. The water flows back
into the Indian Ocean along the rivers that drain the
southern slopes, carrying with them sediments eroded
from the unstable, steep mountains. The sediments
are deposited along the foothills to form extensive and
highly productive alluvial plains of unconsolidated
sediment traversed by innumerable braided rivers.
The abrupt rise of the mountains from less than
500 m to over 8 000 m results in a diversity of ecosys-
tems that range over only a couple of hundred kilo-
meters from alluvial grasslands and subtropical
broadleaf forests along the foothills to temperate
broadleaf forests in the mid hills, which transition into
mixed conifer forests, and conifer forests in the high-
er hills and alpine meadows above the treeline. These
ecosystems are layered as narrow bands along the lon-
gitudinal axis of the mountain range, and include sev-
eral ecoregions (Wikramanayake et al. 2002). But the
rugged terrain also creates microenvironments within
these ecoregions that harbor different ecological com-
munities and assemblages. Thus, the biogeographic,
climatic, geological, and altitudinal variations, as well
as the topographic complexity, all combine to con-
tribute to the biological diversity of the Himalaya On the opposite page, the
along their east-west and north-south axes. Phurbichyachu Himal in Nepal,
Below 1 000 m, the forests and open woodland sa- bordering Tibet.
© Gerald Cubitt
vannas are dominated by drought-deciduous floral
communities with affinities to the Indo-Chinese trop-
ical monsoon forests. Characteristic species include
dipterocarps, such as sal (Shorea robusta) on the Nep-
alese terai and Vatica lanceaefolia, Dipterocarpus re-
tusus, D. turbinatus, and Shorea assamica in the Assam
Valley further to the east. Low-lying areas, subject to
floods during the monsoon, support mixed evergreen
309

The bearded vulture or
lammergeier (Gypaetus barbatus)
is a widespread species in Africa,
Europe, and Asia. In the last
century it suffered a massive
decline, and in the Himalaya
only a few small, isolated
populations survive.
© Francisco Márquez
On the opposite page, bharal
(a Hindi name that refers to the
bluish sheen of the coat) or
Himalayan blue sheep (Pseudois
nayaur) occur in montane regions
in the Himalaya and in the Sichuan
region of China between 3 000 and
5 500 m. It is estimated that there
are more than 25 000 individuals in
the Himalaya, although some
populations on the Tibetan Plateau
and in adjacent China, which
probably numbered in excess of a
million in the early 1960s, fell
drastically due to commercial
hunting to obtain meat for
European markets.
© Colin Monteath/Auscape
310
forests. The alluvial grasslands along the foothill valleys
are among the tallest in the world. These grasslands are
rejuvenated by silt deposited when the rivers that de-
scend from the mountains overflow during the monsoon.
As the floodwaters recede, grasses such as Saccharum
spontaneum and pioneer trees such as Trewia nudiflora,
Dalbergia sissoo, Acacia catechu, and Ehretia laevis begin
to colonize the area, and the grasslands regenerate rapid-
ly. These alluvial plains transition into the sal forests
flanking the hills along the lower reaches of river valleys.
The lower hill slopes above 1 000 m tend to be cooler
and less drought-stressed during the spring pre-monsoon
season. Here, subtropical evergreen broadleaf forests are
dominated by tree taxa (Castanopsis, Schima) derived
from subtropical East Asia. This elevation supports a high
diversity of trees, with variation in community composi-
tion dependent on aspect, disturbance regime, and
human-use patterns. Adjacent to and north of the sub-
tropical forest is a band of subtropical pine forest domi-
nated by Chir pine (Pinus roxburghii) and extending across
practically the entire length of the mountain range.
Temperate humid forests equivalent to “cloud forests”
thrive at elevations where moisture tends to condense
and remain in the air during the warm, moist growing
season, from April to November. These forests are dom-
inated by evergreen broadleaf trees (e.g., Quercus, Lau-
raceae) in the lower part (2 000-2 500 m) and a mixture
of evergreen conifers (e.g., Tsuga, Taxus) and winter-
deciduous broadleaf species (e.g., Acer, Betula, Magnolia)
in the upper part (2 500-3 000 m). Drier, south-facing
slopes support extensive stands of arboreal Rhododendron
species that may co-occur with oaks (Quercus semecarpi-
folia) or other ericaceous species like Lyonia ovalifolia. At
higher elevations, nearly monotypic stands of Rhododen-
dron arboreum may give way to forests that include many
a diverse community of shrubby rhododendrons. These
cloud forests in the east are rich in epiphytes, represent-
ed by several dicotyledons and a variety of mosses, ferns,
and orchids, and provide rich nesting habitat for many
bird species. Malingo bamboo (Arundinaria spp.) in the
understory is food for the charismatic red panda (Ailurus
fulgens, EN). Overall, the wetter Eastern Himalaya is rich-
er in species than the Western Himalaya, and is home to
some of the most species-rich temperate forests in the
world (WWF and ICIMOD 2001).
Subalpine conifer forests occur from about 3 000 to
4 000 m, high enough to receive substantial winter snow.
These forests are identified by the dominant conifer
species, which in the Eastern Himalaya are fir (Abies
spectablis), larch (Larix griffithii), hemlock (Tsuga du-
mosa), and junipers. The western forests of the Eastern
Himalaya have extensive stands of blue pine (Pinus wal-
lichiana), chilgoza pine (P. gerardiana), silver fir (Abies
pindrow), and spruce (Picea smithiana). Typically, the
emergent conifers have tall, straight trunks that rise
above the canopy. In the Eastern Himalaya, the sub-
alpine forests also have many deciduous broadleaf trees
tucked in among the taller, more obvious conifers. Typ-
ical broadleaf trees include Betula, Acer, Sorbus, Prunus,
and some high-elevation members of the Lauraceae and
Araliaceae families. Large evergreen Rhododendron
shrubs are also an important component of this zone.
The rhododendrons above the treeline form alpine
scrub communities throughout the wetter parts of the
Himalaya. Dense shrubberies of juniper and Rhododen-
dron clothe hill slopes to elevations of 4 500 m, and pro-
vide cover for many wild mammal species. Compared
to the broadleaf and conifer forests, plant richness in the
alpine shrub and meadow ecoregions is much greater.
Moist alpine scrub is better developed on shady, north-
facing slopes that are protected from extreme winter
cold by an insulating layer of snow. South-facing slopes
support meadows dominated by Kobresia sedge and di-
cotyledonous forbs with scattered shrubs such as Berbe-
ris, Rosa, Lonicera, Cotoneaster, and others to elevations
of 4 500 m, while dwarf rhododendrons, Cassiope, Hip-
pophae, Cotoneaster, and Potentilla fruticosa persist to
4 700 m where microclimatic conditions allow.
Periglacial and subnival communities occur in the
high alpine areas above 4 700 m, where the short grow-
ing season, high winds, and restless soils require spe-
cialization for survival. Soil creep is an important physi-
cal factor influencing the physiognomy of plant
communities at these elevations. Stable soils support
slow-growing rosette or cushion plants. Because the
perennial structures of cushion plants are sequestered
underground, these plants can withstand high winds.
Common genera include Androsace, Arenaria, and Saxi-
fraga. Rosette plants grow in more sheltered locations
and include at least two genera, Meconopsis (poppy) and
Primula, which have their global centers of diversity in
the Eastern Himalaya. Among the numerous other
alpine herbs are Potentilla, Ranunculus, many legume
species, and the spectacular alpine composite Saussurea.
Although the nival zone of permanent rock and ice
begins at 5 500-6 000 m, vascular plants can survive in
microclimates at the highest elevations on Earth. For
example, Arenaria bryophylla, a small, tufted cushion-
forming plant with small, stalkless flowers, was record-
ed at 6 180 m by A.F.R. Wollaston in 1921 (Wollaston, in
Polunin and Stainton 1997). Other records of high-
altitude flora include another cushion plant in the car-
nation family, Stellaria decumbans, at 6 140 m on the
south slope of Makalu, and Ermania himalayensis, a high-
altitude scree plant in the mustard family, at 6 300 m on
the slopes of Mt. Kamet in the northwestern Himalaya.
Paleobotanical evidence indicates that much of the
forest vegetation in the Himalayan Region is derived
from the tropical evergreen forests of the Indian Penin-
sula, which was part of Gondwanaland, while the
alpine flora evolved after the Himalayan uplift (Ram
and Singh 1994). Since this period, Pleistocene glacia-
tion has influenced the flora and fauna, especially the
southward migration of the species from the Central
Asian, Alps, Mediterranean, and Sino-Japanese regions
(Gupta 1994). At the same time, physical and climatic
barriers along the altitudinal transect, temperature
extremes, and high intensity of ultraviolet radiation
have resulted in high endemism, and evolution of mu-
tants, polyploids, and ecotypes among several taxa.
The Himalaya are home to many unique and diverse
human groups. Culturally diverse ethnic groups living
in the river valleys and mountain slopes of the Hi-
malaya differ from each other in terms of language,
culture, tradition, religion, and patterns of resource
use. Over 27 ethnic groups are found in Nepal, either of
Tibetan-Burmese or Indo-Aryan descent, while Bhutan
has three main ethnic groups: the Ngalongs, Sharchog-
pas, and Lhotsampas, all three of which are dominant,
with many smaller groups with their own unique cul-
tural practices. The northeast part of India, however,
has over 500 distinct ethnic groups (Stirn and Van Ham
2001). The people belong to numerous tribal groups
with ancient cultures and traditions, although there is
also a sizeable population of non-tribal groups as well.
Biodiversity
Of the estimated 10 000 species of vascular plants in the
Himalayan region, approximately 3 160 are endemic.
Angiosperms are divided into 192 families and 2 100
genera. The largest families of flowering plants are Or-
chidaceae (750 species), Compositae (734 species), Gra-
mineae (520 species), and Fabaceae (507 species). The
Eastern Himalaya is also a center of diversity for sev-
eral widely distributed plant taxa, such as Rhododen-
dron, Primula, and Pedicularis (Sahni 1979).
The endemic flora are represented by 71 endemic
genera, of which about 40 are confined to the Eastern
Himalaya and about 15 to the Western Himalaya. The
endemic genera include Jaeschkea, Parajaeschkea, Dri-
mycarpus, Parrotiopsis, Listrobanthes, Megacodon, Pseu-
daechmanthera, Pseudostachyum, Pteracanthus, Sympagis,
Catamixis, Physolena, Pottingeria, Roylea, Trachycarpus,
and Triaenanthus. Drimycarpus and Parrotiopsis are
monotypic genera that represent arborescent taxa,
while Listrobanthes, Megacodon, Pseudaechmanthera,
Pseudostachyum, Pteracanthus, Sympagis, and Triaenan-
thus are shrubs. Five families are endemic to the re-
gion, namely Tetracentraceae, Hamamelidaceae, Cir-
caeasteraceae, Butomaceae, and Stachyuraceae, while
over 90% of the species in Berberidaceae and Saxifra-
gaceae are endemic to the Himalaya (Singh et al. 2000).
A large number of orchids, many representing neoen-
demic taxa, have been recently reported from Sikkim
and Arunachal Pradesh, which probably indicates that
further exploration in the Himalaya will reveal a higher
degree of plant endemism. Although many of the other
non-vascular taxonomic groups have yet to be ade-
quately documented, nearly 13 000 species of fungi and
around 1 100 species of lichens have been described.
Overall, the fauna of the Himalaya is not well
known; most of the information available for this re-
gion pertains to larger vertebrates, especially large
mammals and birds that are easily observed. Smaller
mammals, reptiles, amphibians, and fishes have been
undersampled, while the insects have been largely ig-
nored, with the exception of a few studies of the Hi-
malayan Lepidoptera (e.g., Mani 1986; Haribal 1992).
Around 300 mammal species have been recorded
across the Himalayan Range, of which 12 are endemic to
the Himalaya. The endemics include the golden langur
(Trachypithecus geei, EN), which has a very restricted
range in the Eastern Himalaya; the Himalayan tahr
(Hemitragus jemlahicus, VU); and the pygmy hog (Sus sal-
vanius, CR), a species restricted to grasslands in the Terai-
Duar savannah and grasslands, with its stronghold in the
Manas National Park. The Namdapha flying squirrel
(Biswamoyopterus biswasi, CR) also represents the only
endemic genus in the Himalaya, but is a poorly known
species described on the basis of a single specimen taken
from Namdapha National Park. The mammalian fauna in
the lowlands is typically Indo-Malayan, consisting of lan-
gurs (Semnopithecus spp.), Asiatic wild dogs (Cuon alpi-
nus, VU), sloth bears (Melursus ursinus, VU), gaurs (Bos
gaurus, VU), and several species of deer, such as muntjac
(Muntiacus muntjak) and sambar (Cervus unicolor). In the
mountains, the fauna transitions into Palearctic species,
consisting of snow leopard (Uncia uncia, EN), black bear
(Ursus thibetanus, VU), and a diverse ungulate assem-
blage that includes blue sheep (Pseudois nayaur), takin
(Budorcas taxicolor, VU), and argali (Ovis ammon, VU).
Around 979 bird species are recorded from across the
region, with 15 endemics, including one species, the Hi-
malayan quail (Ophrysia superciliosa, CR), which repre-
sents an endemic genus. However, it has not been record-
ed with any certainty since 1876 (although there were
reports of possible sightings around Nainital in 2003).
Four Endemic Bird Areas (EBAs), as defined by BirdLife
International (Stattersfield et al. 1998), overlap entirely or
partly with the Himalaya Hotspot. The Western Hi-
malaya EBA, which is almost entirely contained within
the hotspot, has 11 bird species restricted entirely to it, in-
cluding the aforementioned Himalayan quail, the cheer
pheasant (Catreus wallichii, VU), and the western
tragopan (Tragopan melanocephalus, VU), which is also
endemic to the hotspot. The Central Himalaya EBA has On the opposite page,
two bird species confined entirely to within its bound- Rhododendron arboreum,
aries, the spiny babbler (Turdoides nipalensis) and the from the Pulmuthang Valley Forest
above Chintang, Nepal.
Nepal wren babbler (Pnoepyga immaculata), and both are
Rhododendrons reach a high level of
also endemic to the hotspot. The Eastern Himalaya EBA, diversity in the Himalaya.
which encompasses part of the Chin Hills (here consid- © Gerald Cubitt
ered part of the Indo-Burma Hotspot), has 19 bird species
endemic to it, including the rusty-throated wren babbler Above, Pleione praecox, from
(Spelaeornis badeigularis, VU), chestnut-breasted partridge southern Sikkim, one of many
orchid species occurring in the
(Arborophila mandellii, VU), white-throated tit (Aegithalos
Himalaya, and among the first to
niveogularis), and orange bullfinch (Pyrrhula aurantiaca), flower. The Orchidaceae is the
all four of which are also endemic to the Himalaya largest family of flowering plants
Hotspot. Finally, the Assam Plains EBA is shared with the in the hotspot.
Indo-Burma Hotspot. The black-necked crane (Grus nigri- © Gerald Cubitt
collis, VU), a large bird of cultural and religious signifi-
cance to the people in the Himalaya, spends the winters
in the Himalayan region, but migrates into the wetlands
of the Tibetan Plateau across the Himalayan Crest.
Systematic surveys of reptiles and amphibians are lack-
313

On pp. 314-315, the Bengal tiger
(Panthera tigris tigris) of India,
Nepal, Sikkim, Bhutan,
and Bangladesh, was the first
recognized tiger subspecies.
Kaziranga National Park in
northeastern India, an alluvial
plain grassland, is known to hold
the highest adult tiger density
—almost 17 tigers per 100 km2—,
evidence that tiger densities can still
be high in optimal habitats.
© Patricio Robles Gil/Sierra Madre
Above, the Vulnerable musk deer
(Moschus moschiferus) do not
have antlers, and only the bucks
grow greatly elongated, tusklike
upper canines. There are musk
glands in the male, and their
secretion always is in great demand
as medicine.
© Gerald Cubitt
On the opposite page, the golden
langur (Trachypithecus geei), is
an Endangered primate with a very
restricted range on the Bhutan-
Assam border. These animals were
photographed in Manas
National Park.
© Gerald Cubitt
316
ing for this hotspot, but it is known that at least 177
species of reptiles have been recorded, of which 49 are en-
demic. The lizard Mictopholis austeniana is the only repre-
sentative of the sole endemic genus, but is known only
from the holotype. Other genera are well represented, and
have many endemic species. These include Oligodon
(nine species, six endemic), Cyrtodactylus (10 species, sev-
en endemic), and Japalura (seven species, all endemic). In
terms of amphibians, 124 species are known to occur, and
41 species of these are endemic. The amphibian fauna is
dominated by anurans, although two species of caecilians
occur. One of these (Ichthyophis sikkimensis) is endemic
and occurs in northern India (in the States of Sikkim and
West Bengal) and extreme eastern Nepal (in Dabugaun in
the Ilam District) at elevations of 1 000 to 1 550 m.
The Himalaya Hotspot has fish species from three ma-
jor drainage systems, the Indus, Ganges, and Brahmapu-
tra. However, the cold, steep high-altitude drainages have
fewer fishes than the lowland rivers, and many species
ranges only just reach into this hotspot; as a result, few
species (33 out of 269) are endemic. The three most di-
verse of the 30 different families represented here are
minnows and carps (Cyprinidae; 93 species and 11 en-
demics), river loaches (Balitoridae; 47 species and 14
endemics), and sisorid catfishes (Sisoridae; 34 species
and four endemics). The genus Schizothorax is represent-
ed by at least six endemic species in the high mountain
lakes and streams, while two other genera of these “snow-
trout,” the genus Ptychobarbus and the Ladakh snowtrout
(Gymnocypris biswasi) —a monotypic genus now thought
to be extinct— are also unique to the Himalaya.
Flagship Species
The Himalaya support globally significant populations
of several large mammals, including the tiger (Panthera
tigris), Asian elephant (Elephas maximus), and greater
one-horned rhinoceros (Rhinoceros unicornis, EN), in the
foothill grasslands and forests. The range of the Asian
elephant in the Himalayan Region is marginal; however,
the populations along the north bank of the Brahmapu-
tra River in Assam are one of India’s largest and most im-
portant (Sukumar 1992). The alluvial grasslands support
some of the highest densities of tigers in the world
(Karanth and Nichols 1998), while the greater one-
horned rhinoceros is restricted to several small, isolated
populations within protected areas. The Eastern Hi-
malayan Region is the last bastion for this charismatic
megaherbivore with its armor-plated and prehistoric ap-
pearance. The Brahmaputra and Ganges rivers that flow
along the Himalayan foothills also support globally im-
portant populations of the Gangetic dolphin (Platanista
gangetica), a freshwater dolphin with two Endangered
subspecies. It is endemic to the river system that flows
along the foothills of the Himalayan Mountain Range.
Dolphin populations are threatened by various human
activities, including fishing, dams, and pollution, which
affect both the animals themselves and their prey base.
Other flagships are wild water buffalo (Bubalus arnee)
and swamp deer (Cervus duvaucelii), which are restricted
to protected areas in southern Nepal and northeastern
India, and represent some of the last remaining popula-
tions of these species in the world. The snow leopard has
a wide distribution across the Himalaya, extending into
the high mountains of the Trans-Himalaya, but the pop-
ulations in the Himalayan Mountains are important be-
cause of the low density of this high-altitude predator.
Some of Asia’s largest birds live in the Himalaya,
and are represented by globally significant populations.
Most are threatened by various anthropogenic causes,
such as the vultures (Gyps spp.), which have undergone
dramatic declines after feeding on carcasses of cattle that
have been treated with Diclofenac (Risebrough 2004).
This is a classic example illustrating the effect of drugs
and pesticides along the food chain. The greater and less-
er adjutants (Leptoptilos spp.) in the foothill grasslands and
broadleaf forests, as well as the hornbills in the broadleaf
forests, are threatened by loss of nesting trees and lack of
food sources. Other large birds include the sarus crane
(Grus antigone), which inhabits the wetlands along the
foothills; the black-necked crane, which spends winters in
montane wetland sites; and the lammergeier (Gypaetus
barbatus), which soars among the high mountains.
There are also several smaller mammals and birds
that carry flagship status. The golden langur is a beau-
tiful, golden-yellow primate that lives in the broadleaf for-
ests between the Sankosh and Manas rivers in the East-
ern Himalaya in both Assam and Bhutan, while the red
panda inhabits the old-growth mixed conifer forests.
The white-winged duck (Cairina scutulata, EN), the en-
demic white-bellied heron (Ardea insignis, EN), and the
Bengal florican (Houbaropsis bengalensis, EN) are just a
few of the other avian flagships in the region.
Threats
Despite their apparent remoteness and inaccessibility,
the Himalaya have not been spared human-induced bio-
diversity loss. People have lived in the mountains and
eked out a livelihood there for thousands of years. How-
ever, with better access to global market economies,
both dependence on natural resources and the econom-
ic expectations and aspirations of the people have in-
creased in recent years. Access has also encouraged im-
migration into montane areas from outside in some
regions, such as Arunachal Pradesh, as well as move-
ments within the Himalayan Mountains, such as in
Nepal, where people have migrated from the mountains
to the lowland terai. The consequences of the latter
movements have been to concentrate people in the
more productive ecosystems that are also the richest in
biodiversity, whereas the former have resulted in a
breakdown of cultural and social traditions and ties to
conservation of natural resources among the tribal peo-
ple in the mountains. Moreover, better health care and
disease control have resulted in a net increase of human
populations, placing a greater burden on the sensitive
montane ecosystems.
Today, the distribution of remaining habitat in the Hi-
malaya is patchy. The steadily increasing human popu-
lation has been responsible for extensive clearing of
forests and grasslands for cultivation, and widespread
logging. Cultivation has a soft, upper-elevation limit of
around 2 100 m on slopes exposed to the monsoon, but
many people farm crops like barley, potato, and buck-
wheat at higher elevations in the inner valleys and trans-
montane regions, and in some areas, such as Jumla,
Kashmir, Lahoul, and Ladakh, there are major agricultur-
ally based population centers well above this elevation.
The land is also often cleared by pastoralists for their live-
stock during the summer months. The conversion of
forests and grasslands for agriculture and settlements has
led to large-scale deforestation and habitat fragmentation
in Nepal, and in the Indian States of Sikkim, Darjeeling,
and Assam. In the northeast Indian states, loss of dense
forests is estimated to be as high as 317 km2 every year.
The impact of forest clearing is less intense in Bhutan
compared to the neighboring countries because of the
low density of human population and the government’s
policy of a cautious approach to development and a com-
mitment to the conservation of natural resources.
Large areas of remaining habitat are highly degraded.
Overgrazing by domestic livestock in the resilient low-
lands as well as in the sensitive alpine ecosystems is
widespread. In the former, huge numbers of free-ranging,
unproductive cattle graze the forests and grasslands, de-
stroying all undergrowth and preventing regeneration;
in the alpine ecosystems, virtually all areas are grazed
by increasingly larger herds of domestic yak (Miller
2002). The fragile meadows are also subject to overex-
ploitation of their flora for the traditional medicine
trade. Because the medicinal plant collectors invariably
uproot entire plants, regeneration is retarded. Wood ex-
traction for fuel and fodder also contribute to loss of un-
dergrowth and regeneration, and changes in species
composition. Fuelwood is collected for domestic con-
sumption as well as for export. These activities have in-
flicted severe and sometimes irreversible damage, and
in many areas forests are no longer able to support nat-
ural ecosystems and their associated biodiversity.
It is possible to make an estimate of natural vegeta-
tion remaining intact by considering the state of natural
habitat of the various ecoregions that make up this
hotspot (Wikramanayake et al. 1998). The Himalayan
subtropical broadleaf forests ecoregion has lost more
than 70% of its natural habitat (although most of the hill
forests above 1 000 m still remain uncut because the
shallow, erosion-prone soils are unsuitable for cultiva-
tion), with the remaining forests in scattered fragments.
Over 90% of the adjacent Terai-Duar savanna and grass-
land ecoregion has been converted to agriculture and
settlements, and most of the remaining habitat is now
within protected areas. The temperate broadleaf forests
in the western extent of the Himalayan range have lost
over 70% of their natural vegetation, although several
large patches remain in the extreme western part. The
eastern temperate broadleaf forests have fared better,
with almost 70% of the natural habitat still remaining in
large patches (particularly in northeastern India and
Bhutan), but shifting agriculture has resulted in exten-
sive habitat degradation. Likewise, the adjacent Brahma-
putra Valley, a region characterized by remarkable pro-
ductivity, and hence a long history of cultivation, is
three-quarters cleared, with the largest forest blocks
confined to protected areas in central Assam.
As with the temperate broadleaf forests, habitat loss in
the Western Himalayan Sub-Alpine Conifer Forests is se-
vere, with over 70% of the natural vegetation being lost.
Notwithstanding, this region contains some of the least-
disturbed forests in the Western Himalaya. In the Eastern
Himalaya, the reverse is true, although on gentler slopes
within the northeastern hill states of India most of the
broadleaf forests (over 80%) have been affected due to
slash-and-burn (jhum) agriculture. Most of the Alpine
Shrub and Meadows is remote and inaccessible and, con-
sequently, largely intact thanks to high elevation and
harsh climate. However, all the gentle and accessible
meadows have undergone extensive habitat degradation
due to overgrazing, trampling, and commercial harvest
for medicinal plants. In total, then, nearly 50% of the
alpine vegetation in the region can be said to be intact,
although in Sikkim, Bhutan, and Arunachal, over 60% of
the alpine vegetation is still intact. Since this degradation
can not be estimated from remote-sensed data sources, it
is difficult to assess its extent, as grazing is prevalent all
throughout the alpine regions. In total, we estimate that On the opposite page,
around 25% of the original vegetation of this hotspot, in- the Endangered greater one-horned
cluding alpine areas, remains in intact condition. rhinoceros (Rhinoceros unicornis)
disappeared from much of its range
Besides habitat loss and degradation, poaching is rife
in northwestern India and Pakistan
across the Himalayan Mountains: tigers and rhinoceros due to the loss of alluvial plain
are hunted for their body parts, which are much prized grasslands to agricultural
in traditional Chinese medicines, while the snow leop- development, which destroyed the
ard and red panda are taken for their beautiful pelts. species’ prime habitat and also led
The remoteness of the Himalayan Region and the open to increased conflict with humans.
In addition, many were hunted,
borders have facilitated this illegal trade.
and countless slaughtered as a
Political unrest, usually in the form of insurgencies, result of a government bounty
plague certain sites in the Himalayan region. Protected established to protect tea
areas and forests that harbor wildlife also serve as refuges plantations. By the early 1900s, the
for insurgents, who indulge in indiscriminate poaching species was near extinction.
and felling of trees to obtain funds. In Nepal, the Maoist © Patricio Robles Gil/Sierra Madre
insurgency has severely constrained conservation activ-
Above, the Endangered red panda
ities on the ground since 1996. Similar insurgencies oc- (Ailurus fulgens) is a characteristic
cur in Assam and Nagaland in India, while the dispute flagship species of the Himalaya.
between India and Pakistan over the Kashmir border has Despite its name, the species is not
had implications for wildlife conservation in these areas. considered closely related to the
Endangered giant panda
(Ailuropoda melanoleuca), a
flagship of the Mountains of
Conservation
Southwest China Hotspot.
© Gerald Cubitt
Approximately 15% of the Himalaya Hotspot has some
form of legal protection, although this percentage drops
to 10% when one considers only those in IUCN categories
I to IV. Protected areas have a mixed history in the
Himalaya. In Assam, Manas and Sonai Rupai were first
319

The Vulnerable takin (Budorcas
taxicolor) is found in the
mountainous regions of the
Himalaya and western China,
at altitudes of up to 4 500 m.
Weighing as much as 400 kg, the
flesh of the species is highly prized,
and the takin is heavily hunted
using snares and deadfall traps.
© Pete Oxford/naturepl.com
On the opposite page, much of
Nepal’s original native habitat
has been converted for agriculture
and settlement.
© Gavriel Jecan/Art Wolfe Inc.
320
established as wildlife sanctuaries in 1928 and 1934,
respectively, and are among the earliest contemporary
protected areas in Asia (IUCN 1990). Most other protected
areas are relatively recent, having been established within
the past three to four decades. However, many hill-tribe
communities have traditionally recognized and protected
sacred groves, which have been effective refuges for bio-
diversity for centuries (Gadgil, 1985). Today, several pro-
tected areas —Corbett National Park, Manas National Park,
Kaziranga National Park, Chitwan National Park, and Sagar-
matha National Park— have been distinguished as World
Heritage Sites for their contribution to global biodiversity.
In the 1970s and 1980s, several protected areas were
established or extended in the northeastern Himalayan
states of India, creating a network that showcased
the biodiversity in the area. The protected areas in the
northwestern Indian states include some of the world’s
most renowned, such as Corbett and Rajaji national
parks, which harbor important populations of flagship
species like elephants and tigers.
In Nepal, at least 26 666 km2 of land has been desig-
nated as protected areas, including eight national
parks, four wildlife reserves, one hunting reserve, three
conservation areas, and five buffer zones (WWF-Nepal
2004). Chitwan, the country’s first national park, was
established in 1973. Previously a hunting preserve for
the royal family, the park is well known for its tiger and
greater one-horned rhinoceros populations. Of particu-
lar significance are the Annapurna Conservation Area,
the Kanchenjunga Conservation Area, and the Makalu-
Barun National Park, which have become models for
community-based biodiversity management.
The protected area system of Bhutan includes five
national parks, three wildlife sanctuaries, and one strict
nature reserve, as well as 12 corridors covering almost
16 000 km2 (Biodiversity Action Plan 2002). The current
system was bequeathed as a “Gift to the Earth from the
People of Bhutan” in 1999. Although a protected area
system was established in Bhutan as early as the 1960s,
this system was dominated by the Jigme Dorji Wang-
chuck National Park. The park was mostly confined to
the north of the country, and did little to contribute
towards biodiversity conservation because most of the
park protected vast areas of permanent rock and ice. In
1995, the protected area system was revised to include
all nine of the current protected areas accounting for al-
most 26% of the total land area in Bhutan. In 1999, an-
other 9% was added to the system in the form of bio-
logical corridors, which linked the protected areas to
create a conservation landscape extending across the
country. The landscape is known as the Bhutan Biolog-
ical Conservation Complex (Sherpa et al. 2003).
The many protected areas that lie adjacent to each oth-
er across national borders present promising opportuni-
ties for transboundary conservation activities. The Manas
National Park in Bhutan and Manas Tiger Reserve in As-
sam, India is one such complex. Biological corridors also
provide opportunities to link the protected areas across
international boundaries and create habitat linkages, such
as between Bardia in Nepal and Katerniaghat in India. An-
other important transboundary initiative is Kanchanjunga
Conservation Area (KCA) in the Taplejung District in
Nepal, an area covering some 1 650 km2 named for Mt.
Kanchanjunga (8 586 m) —the third highest mountain
in the world— and planned as a tri-national peace park
with the Tibet Autonomous Region of China to the north,
and Sikkim in India to the east. The Kanchanjunga Con-
servation Area adjoins the Kanchendzoga National Park
in Sikkim, and extension of the Qomolungma Nature Re-
serve in the Tibet Autonomous Region is under way to in-
clude the area bordering Kanchanjunga. The new strate-
gy for creating conservation landscapes in the Himalaya
will not only help to conserve the region’s species and
ecological processes that sustain biodiversity, but also
contribute towards building regional cooperation through
transboundary conservation efforts, thereby paving the
way for a secure future for Himalayan biodiversity.
Despite the efforts to revise the protected area system
across the Himalayan Mountains, about 17% of it, or
over 40 000 km2, still consists of permanent rock and ice
(Alnutt et al. 2002). The protected areas of the alpine re-
gions, in particular, are over-represented by these bio-
logically depauperate habitats. Across the range, 15 pro-
tected areas consist of more than 50% rock and ice.
Further expansion of the protected area network should
consider minimizing representation of these areas.
Investment in biodiversity conservation in the Hi-
malayan Region comes primarily from national govern-
ments, bilateral and multilateral agencies, and internation-
al and regional NGOs. The national governments, backed
by international agencies such as the Global Environmen-
tal Facility (GEF), United Nations Development Program
(UNDP), the World Bank, the European Union (EU), the
Danish International Development Agency (DANIDA),
WWF, and the MacArthur Foundation, are supporting
projects to improve protected area management, sustain-
able natural resources, and livelihoods. All countries in the
Himalaya Hotspot have ratified the Convention on Biolog-
ical Diversity, and have prepared National Biodiversity
Conservation Strategies and Biodiversity Action Plans.
While there have been many successes in establishing
protected areas and more experimental, multiple land-
use conservation areas, much remains to be done to safe-
guard the biological wealth of the Himalayan Region cur-
rently lying outside formally protected reserves. The
protected areas of the Himalaya, particularly in the low-
lands along south-facing slopes, are too small to maintain
viable populations of threatened species. Successful long-
term conservation requires that efforts be made to
expand the conservation benefits beyond existing protect-
ed areas to adjacent habitats, as many species are depen-
dent upon the seasonal use of habitats distributed along
elevational gradients and across national boundaries.
MINGMA SHERPA 64
ERIC WIKRAMANAYAKE 91
GOPAL RAWAT 92
 INDO-BURMA
The redefined Indo-Burma Hotspot has emerged as a
0 600 km
INDIA
result of the exclusion of the Himalayan chain and
associated foothills in Nepal, Bhutan, and India CHINA
(now included in a new Himalaya Hotspot, the addi- BANGLADESH
tion of southern and western Yunnan, and an ex-
pansion of the area of southern subtropical China
MYANMAR
(southern Guangxi and Guangdong). The area cov- (BURMA)
ered by this redefined hotspot is at times referred to as
LAOS
the Indo-Chinese Subregion and can be described Hainan
BAY
as Tropical Asia east of the Ganges-Brahmaputra OF
BENGAL
Lowlands, excluding the Malesian region. The Indo-
Burma Hotspot begins at the evergreen forests in the THAILAND
foothills of Chittagong in Bangladesh and extends SOUTH
CHINA
through the Garo and Khasi Hills of Meghalaya, In- SEA
Andaman CAMBODIA
dia, then eastwards through the States of Manipur, Islands
VIETNAM
Mizoram, and Nagaland to encompass most of Myan-
GULF
mar (except the extreme northern alpine areas), a OF
THAILAND
part of southern and western Yunnan, China, all of
the Lao People’s Democratic Republic (P.D.R.), Viet- Nicobar
Islands
nam, and Cambodia, the coastal lowlands of south- MALAYSIA
ern China, Hainan Island, the vast majority of Thai-
land, a small fraction of Peninsular Malaysia, and
the Andaman Islands of India. As redefined here, the tats that support a high diversity of plant and animal
original extent of the hotspot is 2 373 000 km2. species. This diversity is enhanced by a significant
The transition to the Sundaland Hotspot occurs on endemic element, which may largely derive from
the Thai-Malay Peninsula. The boundary between the habitat isolation caused by periods of high sea level
two hotspots is here taken to be the Kangar-Pattani and vegetation changes during the glacial episodes
Line, which cuts across the Thailand-Malaysia bor- of the Pleistocene. Consequently, the hotspot con-
der, marking the transition between the Indochinese tains many localized centers of endemism, particu-
and Sundaic floras (Van Steenis 1950; Whitmore larly montane isolates, but also areas of lowland wet
1984). However, other analyses indicate that the phy- evergreen forest that were isolated at some stage, and
togeographical and zoogeographical transition be- river basins.
tween the Sundaland and Indo-Burma biotas may lie At present, much of Indo-Burma is characterized
just to the north of the Isthmus of Kra, associated by distinct seasonal weather patterns. During the
with a gradual change from wet seasonal evergreen northern winter months, dry, cool winds blow from
dipterocarp rainforest to mixed moist deciduous for- the stable continental Asian high-pressure system, On the opposite page,
est (Woodruff 2003). For example, more than half of resulting in a dry, cool period under clear skies the white-headed black langur
the 544 forest bird species present on the Thai-Malay across much of the south, center, and west of the (Trachypithecus poliocephalus
leucocephalus) occurs in the
Peninsula have their species limits north of the Isth- hotspot (the dry, northeast monsoon). As the conti-
limestone hills in southwest
mus at 11º-13ºN (Hughes et al. 2003). nental system weakens in summer, the wind direc- Guangxi, China. Critically
Indo-Burma has a complex geological and evolu- tion reverses and air masses forming the southwest Endangered, they are among the
tionary history. The Indian intrusion into the Asian monsoon pick up moisture from the seas to the world’s most threatened primates.
continental landmass has been responsible for the southwest and bring abundant rains as they rise over © Patricio Robles Gil/Sierra Madre
formation of most of the hotspot’s topography, in- the hills and mountains. In northern Vietnam and
cluding the general north-south orientation of the southern coastal China, the dominant weather pat-
mountains and main rivers. The wide variation in tern is the north or northeasterly monsoon during
land form, climate, and latitude within the hotspot the northern winter and east or southeasterly mon-
has led to the development of diverse natural habi- soon in the summer.
323

The grey peacock-pheasant
(Polyplectron bicalcaratum),
also known as the chinquis, is the
national bird of Myanmar
(Burma). It is one of a number of
pheasant species that occur in the
Indo-Burma Hotspot.
© Gerald Cubitt
324
Originally, nearly all of Indo-Burma would have
been dominated by broadleaf forests. The complex
composition and distribution of the hotspot’s principal
vegetation formations are determined by the seasonal-
ity of rainfall, soil characteristics, temperature, and his-
tory. The richest forests —in terms of tree diversity and
overall plant species numbers— are the lowland mixed
wet evergreen forests, which occur in climates with
one to four dry months. The southern mixed wet ever-
green formations comprise the Parashorea stellata asso-
ciation, characteristic of zonal yellow-red clay loams,
and the Dipterocarpus costatus association on yellow
sandy soils, with low nutrient levels and high sensitiv-
ity to erosion, which occurs locally as islands within
the widespread clay loams. The former were once
widespread in Tenasserim south of Tavoy, Peninsular
Thailand from Chumphon to northwestern Malaysia,
and on the lower slopes of granite outcrops in south-
eastern Thailand, as well as south of Da Nang and in
the hills north of Hue in Vietnam, but now survive
only as a few fragments. The forests of D. costatus have
been severely degraded, although significant stands re-
main on the southern slopes of the Cardamom and
Kamchay ranges, and isolated patches in the Thai-
Malay Peninsula. The northern mixed wet evergreen
forest association is at present fragmented by mountain
ranges and habitat conversion, and remnant patches
are restricted to the northern part of the hotspot (north-
ern Myanmar; southern China, including southern
Yunnan and also Hainan; the Lao P.D.R.; and northern
Vietnam).
In lowland areas where rainfall is more limited and
the dry season lasts for five to seven months, the Dipte-
rocarpus turbinatus dry evergreen forest formation is
the natural climax stage in many lowland and hill re-
gions. In the driest areas, it is confined to galleries in
stream and river valleys, in areas otherwise dominated
by deciduous forest types. This dry evergreen forest
type is still widespread across substantial areas of Indo-
Burma.
Deciduous forests occur in areas with five to nine dry
months. The Dalbergia-Lagerstroemia mixed deciduous
forest formation is widespread on yellow clay soils
throughout the hotspot. Teak (Tectona grandis) may be a
major component where soil and climatic conditions
are right. On sandy and shallow lateritic soils, mixed
deciduous forest is replaced by deciduous dipterocarp
forest, which generally forms a low, broken canopy.
Several factors contribute to the formation and mainte-
nance of this forest type, including high frequency of
fire (probably anthropogenic). In the dry, central
Ayeyarwady Plain of Myanmar, a variety of semidesert
thorn communities occur (Stamp 1925), which support
fewer species than most other deciduous forest types
but have relatively high levels of endemism. Some de-
ciduous forests contain highly valuable timber species
(e.g., teak, rosewood). Their leafless period and often
broken canopy permit enough light to reach the ground
and stimulate a dense undergrowth of grasses and
herbs, making these forests excellent feeding areas for
large herbivorous mammals. Historically, these forests
supported some of the most diverse and abundant
mammal and bird megafaunas in Asia.
Throughout the hotspot, montane forests extend on
humus-rich soils from about 800 masl. These forests
are lower in stature with fewer emergents; oaks (Fa-
gaceae) dominate, while laurels (Lauraceae) and mag-
nolias (Magnoliaceae) become notable constituents.
Montane tree species composition is generally less di-
verse than that of lowland forests, but it contains pro-
portionally more endemic species. Diverse edaphic,
topographic, and microclimatic conditions at higher el-
evations give rise to a range of mixed coniferous and
broadleaf evergreen forest formations. On dry hills
and plateaus subject to fire, conifer-dominated savanna
forests occur (typically dominated by Pinus merkusii or
P. kesiya). At the highest elevations, on ridgelines and
ridge crests, stunted, xerophytic formations, character-
ized by the presence of Rhododendron spp., occur. In
general, the diversity and richness of shrubs, herbs,
epiphytes, and acid-loving species increase at higher al-
titudes, as tree diversity declines. Among them are
many endemic species.
Heath forest occurs on some raised beaches in
coastal areas. This evergreen forest type is less species-
rich, but probably contains the highest proportion of
endemics of any regional evergreen forest type. A re-
lated forest occurs on similar soils above 800 masl in as-
sociation with sandstone mountains. Limestone karst
outcrops support distinctive shrub and woodlands on
their summits, as well as rich, mostly herbaceous floras
on sheltered cliffs. Limestone karst formations can sup-
port relatively high levels of endemism, particularly in
groups such as orchids (Orchidaceae). In addition, a wide
variety of distinctive, localized vegetation formations
occur in Indo-Burma, including lowland floodplain
swamps, mangroves, seasonally inundated grasslands,
and successional assemblages, as well as croplands and
plantations.
The hotspot is home to a wide diversity of ethnic
groups, cultures, and languages. Several language
groups, including Mon-Khmer, Austroasiatic, and Tai-
Kadai, originated and developed in the hotspot, and
were later joined by the Hmong-Mien and Sino-Tibetan
language groups. Agriculture developed very early in
the hotspot, and gave rise to different land-use forms,
including both semi-nomadic and rotational swidden
cultivation, and irrigated floodplain rice cultivation.
Among the human population of the hotspot, there is
a broad spectrum in terms of the degree of integration
with the global economy, from the Paleolithic life-
styles of certain ethnic groups in the Thai-Malay
Peninsula and the Jarawa, Onge, Sentinelese, and An-
damanese tribes of the Andaman Archipelago, to the
urban life-styles of the residents of Bangkok, Hong
Kong, and other major cities.
Biodiversity
Attempts to estimate species diversity and endemism
for Indo-Burma are hampered by uneven knowledge of
taxonomy and distribution of species and groups. New,
locally endemic species are regularly discovered, while
advances in taxonomy continue to reveal that single
widespread “species” actually comprise complexes of
separate species, many of which are local or regional
endemics. Socio-political divisions have also complicat-
ed biodiversity assessments because populations in
neighboring countries may be independently described
as locally endemic species, although an independent
revision would consider such pairs to represent just a
single species. Nevertheless, reasonable estimates of
species diversity and endemism can be made for some
groups in the hotspot.
The total plant diversity of the former Indo-Burma
Hotspot was estimated at about 13 500 species of flow-
ering and gymnosperm plants, of which about 7 000
were estimated to be endemic to the hotspot (Dijk et
al. 1999). A reassessment of the revised Indo-Burma
Hotspot results in similar estimates for numbers
of hotspot species and endemics: the modest losses of
species restricted to the Himalayan chain and associ-
ated foothills being compensated for by the species
gained in southern China. These estimates are, how-
ever, quite conservative: Davis et al. (1995) estimated
that there are 12 000-15 000 species of vascular plants
(i.e., including ferns) in Cambodia, Laos, and Vietnam
combined. This flora includes a profusion of orchid
(Orchidaceae) and ginger (Zingiberaceae) species (for
example, there are more than 1 000 orchid species in
Thailand alone), as well as the variety of fine tropical
hardwoods.
The vertebrate fauna of the hotspot is quite diverse.
Mammals number 430 species in 171 genera and 37
families, of which 71 species and seven genera (six
monotypic) are endemic to the hotspot. In addition,
there is a single endemic family, the Craseonycteridae,
which is represented by one species, Kitti’s hog-nosed
bat (Craseonycteris thonglongyai), which is no larger
than a bumblebee, and among the world’s smallest
mammals.
Over the last 12 years, the hotspot has witnessed the
discovery of six large mammal species. Five of these
were discovered in the Annamite Mountains: the saola
(Pseudoryx nghetinhensis, EN), large-antlered muntjac
(Muntiacus vuquangensis), Annamite muntjac (M. tru-
ongsonensis), grey-shanked douc (Pygathrix cinerea),
and Annamite striped rabbit (Nesolagus timminsi) (Do
Tuoc et al. 1994; Vu Van Dung et al. 1994; Nadler 1997;
Giao et al. 1998; Timmins et al. 1998; Averianov et al.
2000). The sixth species, which was discovered in the
mountains of northern Myanmar, is the leaf deer
(Muntiacus putaoensis) (Amato et al. 1999).
The bird fauna is also very diverse, with some 1 277
species, of which 74 are endemic. There are also three
endemic genera, all represented by single species: the
golden-crested myna (Ampeliceps coronatus), short-
tailed scimitar-babbler (Jabouilleia danjoui), and wedge-
billed wren-babbler (Sphenocichla humei). Six Endemic
Bird Areas (EBAs), as identified by BirdLife Interna-
tional (Stattersfield et al. 1998), are found within the
hotspot, namely the Andaman Islands, the Irrawaddy
Plains, Hainan Island, the Annamese Lowlands, the
South Vietnamese Lowlands, and the Da Lat Plateau.
The Eastern Himalayas and the Assam Plains EBAs are
shared with the Himalaya Hotspot.
The non-marine reptiles number at least 519 species
in 151 genera, of which 189 species and 12 genera are
endemic. Nine of the endemic genera are represented
by single species, among them a recently described
form of pit viper from Vietnam (Triceratolepidophis sie-
versorum) (Ziegler et al. 2000). The rich amphibian
fauna contains some 139 endemics among a total of
around 323 species; yet, of the 57 genera represented,
three (Ophryophryne, Bufoides, and Glyphoglossus) are
endemic to the hotspot. Bufoides and Glyphoglossus
comprise single species: the Khasi Hills toad (B. megha-
layanus, EN) is known from only a few sites in north-
eastern India, while the last-mentioned (G. molossus) is
more widespread in the hotspot.
The hotspot’s inland fish fauna is remarkably di-
verse, with 1 262 documented species or 10% of the
world’s fishes that enter fresh water. The total may ul-
timately approach 2 000 species. The 566 fishes that are
restricted to the region amount to more than half of the
hotspot’s endemic vertebrates, and constitute an obvi-
ous priority for conservation efforts. Endemism is also
considerable at higher phyletic levels, with 30 endemic
genera and an endemic family, the Indostomidae, or ar-
mored sticklebacks. This family of strange fishes is an
extraordinary element to be found in tropical fresh wa-
ters, and may be remotely related to the marine
seamoths. Diversity of freshwater fishes is particularly
high on the lower to middle flanks of mountain ranges,
where riffle and pool habitats in small streams have
been the sites for extraordinary diversification, particu-
larly in the loach families, Cobitidae and Balitoridae.
The Siamese fireback
Flagship Species (Lophura diardi), found in
Thailand, Laos, Cambodia, and
Vietnam, is considered Near
First on the list of flagship mammal species is the sao-
Threatened due to hunting and
la. The known distribution of the saola is restricted to habitat loss across the Indo-Burma
the Annamite Mountains, along the border between Hotspot, to which it is endemic.
Vietnam and the Lao P.D.R., and outlying hills to the © Gerald Cubitt
east. Although the ecology of the species is little
known, it is believed to be largely restricted to wet
evergreen forests at elevations below approximately
1 000 masl. This habitat has been extensively degraded,
fragmented, and converted throughout the species’
known range, and most remaining areas are subject to
high levels of human use. While the saola is not a
species in particular demand in the wildlife trade, it is
325

The green magpie (Cissa chinensis)
is a forest bird species found in the
Indo-Burma Hotspot and also in
southern China.
© Gerald Cubitt
On the opposite page, Thi Lan Su
Waterfall in the Umphung Wildlife
Sanctuary, Thailand.
© Gerald Cubitt
326
susceptible to indiscriminate snaring, which may be ex-
pected to increase in some parts of its range with the
ongoing construction of the Ho Chi Minh National
Highway and associated road network. Despite the
global attention that was focused on the species follow-
ing its discovery, none of its populations has been
placed under effective conservation management, and
the very real possibility exists that this enigmatic
species may become extinct within a decade.
Other flagships include the Vietnamese population
of Javan rhinoceros (Rhinoceros sondaicus, CR), esti-
mated to number only a handful of individuals at Cat
Tien National Park, and the kouprey (Bos sauveli, CR),
a large bovid formerly found in forest areas of northern
and eastern Cambodia and adjacent countries, but
which may now be extinct. Also of special significance
are the endemic primates, a number of which are in-
cluded on a list of the world’s top 25 most threatened
primates prepared by Conservation International and
advisors: the eastern black-crested gibbon (Nomascus
nasutus, CR), Tonkin snub-nosed monkey (Rhinopith-
ecus avunculus, CR), grey-shanked douc, white-rumped
black leaf monkey (Trachypithecus delacouri, CR), white-
headed black langur (T. poliocephalus leucocephalus,
CR), and Tonkin hooded black langur (T. p. poliocepha-
lus, CR) (Mittermeier et al. 2002).
Flagship bird species include Gurney’s pitta (Pitta
gurneyi, CR), a lowland evergreen forest specialist en-
demic to Peninsular Thailand and adjacent parts of
southern Myanmar. During the twentieth century, the
species underwent a dramatic decline due to extensive
clearance of its habitat to the point that, by the end of
the century, it was known to persist only at a single lo-
cality: Khao Nor Chuchi in Thailand. However, it has
recently been rediscovered in the Tanintharyi division
in Myanmar, which potentially supports a very signifi-
cant population (BirdLife International 2004). Although
not endemic, the majority of the world population of
green peafowl (Pavo muticus, VU) is found within the
hotspot. This species has undergone a dramatic decline
over the last century as a result of hunting and expan-
sion of human populations into natural landscapes, par-
ticularly the spread of human settlements along per-
manent water sources. Edwards’ pheasant (Lophura
edwardsi, EN), a species endemic to the lowlands of
central Vietnam, is, like the saola, a flagship for the
lowland wet evergreen forests of the Annamite Moun-
tains and foothills. The Bengal florican (Houbaropsis
bengalensis, EN) is a flagship for the threatened grass-
land ecosystems in Cambodia and Vietnam.
Tortoises and freshwater turtles collectively form a
flagship group. The hotspot’s non-marine turtle fauna is
probably the most diverse in the world, with at least 40
(and, depending on taxonomic opinions, up to 52)
species in 31 genera. This represents one-sixth of the
world’s turtle species and over one-quarter of the gen-
era; about two-thirds of the species are endemic to the
hotspot. Noteworthy also are the endemic Siamese
crocodile (Crocodylus siamensis, CR), now greatly re-
duced in the wild with a severely fragmented popula-
tion; and the Chinese crocodile lizard (Shinisaurus cro-
codilurus). Although the latter is not quite endemic
(occurring in northeastern Vietnam and southern Chi-
na), it deserves mention for being the sole member of
the family Shinisauridae, which has its closest relatives
in the lizards of the genus Xenosaurus in southern Mex-
ico and Guatemala.
Flagship amphibians are considered as groups rather
than individual species. Numerous remarkable and en-
demic frog species occur, but several groups such as the
Rhacophorus gliding frogs, the megophryid litter toads,
and various ranid groups stand out for their local evo-
lutionary radiations, conservation concern, and eye-
catching appearance. Salamander diversity is not very
high in the hotspot, but the salamanders contain a high
proportion of species with very restricted ranges and of
high conservation concern, including four endemic
species in the genus Paramesotriton, two of which are
globally threatened: the Vietnamese salamander (P. de-
loustali, VU) from Vietnam and the Guangxi warty newt
(P. guangxiensis, VU) from Guangxi Autonomous Re-
gion, China, and northern Vietnam.
Indo-Burma’s streams and rivers are inhabited by
fish species that are not only of global conservation sig-
nificance, but also include some of the extremes of size
among freshwater fishes. The Tonle Sap Lake and deep
pools of the Mekong River, up to 60 m deep, are critical
habitats for some of the world’s largest freshwater fish-
es: the Mekong giant catfish (Pangasianodon gigas, CR),
giant carp (Catlocarpio siamensis), and giant freshwater
stingray (Himantura chaophraya, proposed CR). Other
flagship fishes include the dragonfish (Scleropages for-
mosus, EN), a relict of a Gondwanan group that is rapid-
ly being depleted by illegal collecting of its juveniles for
the aquarium trade.
Threats
Indo-Burma may have been one of the first places on
the globe where agriculture developed (Solheim 1972;
Diamond 1997), creating a long history of forest burn-
ing and clearance for shifting or permanent small-scale
cultivation. In recent centuries, steadily increasing
trade in agricultural commodities and timber, com-
bined with population growth, have led to widespread
forest destruction. Very little, if any, natural vegetation
has been unaffected by human actions. In particular,
lowland evergreen forests have been extensively
cleared, having been reduced to well under 30% in
Thailand, less than 20% in Vietnam, and only 7% on
Hainan Island by the early 1990s (BirdLife International
2003). Shifting cultivation and logging have also de-
graded large tracts of hill and montane forest, par-
ticularly in Chin State in Myanmar (which, during the
1980s, had one of the highest deforestation rates in
the world), as well as in northern Laos and northern
Vietnam. Tree plantations (particularly teak, rubber,
and oil palm) have replaced large areas of lowland for-
est, while coffee and tea have done the same with large
areas of hill and montane forests. Nevertheless, tracts
of relatively undisturbed forest remain, for example, in
northern Myanmar, western Thailand, and in the An-
namite Mountains.
Due to the long, local dry season in many areas,
what forest remains is highly susceptible to fire. In low-
land mosaic forests, fires are often set deliberately, for
various reasons, including promotion of a grassy un-
derstory suitable for extensively grazed livestock and,
perhaps, wild ungulate populations (for hunting) or to
enhance visibility for hunting. These fires tend to result
in succession from dense forest to more open forest
types. In some montane areas, fires set to clear swid-
dens (usually maize or cassava) can spread into areas of
evergreen forest (especially when this has been de-
graded by selective extraction) and replace it with sa-
vanna grassland and other secondary formations. In
some montane areas, for instance the Da Lat Plateau,
accidental fires from swidden clearance and deliberate-
ly set fires are leading to succession from evergreen for-
est to more open coniferous forest.
Widespread deforestation and forest litter burning
predictably result in soil erosion during the rainy sea-
son. While soil erosion is a natural process in any hill or
montane area, there is no doubt that human activities
have significantly increased erosion throughout the hot-
spot. Mining for ores, gems, jade or cement eliminates
natural habitats in the most extreme way, and pollutes
ecosystems with mining spoils and leached chemicals.
Also very important are the indirect impacts from mining-
quarrying, particularly those resulting from large num-
bers of people living in remote forest areas, contribut-
ing to unsustainable hunting pressure on key wildlife
populations.
Freshwater and coastal habitats have been impacted
just as severely as forest habitats. Floodplain swamps
and other non-flowing wetlands have been drained and
converted to wet rice cultivation on a massive scale
throughout the hotspot (but especially in Thailand,
Myanmar, and Vietnam), while some of these rich agri-
cultural lands are now being lost to urban sprawl. The
lower Mekong River and its major tributaries have
some of the most intact riverine floodplain habitats in
the hotspot (BirdLife International 2003). Rivers have
been dammed in order to store water to generate elec-
tricity for countries’ economic growth, or for export to
neighboring countries so as to generate foreign ex-
change earnings. Damming a river section not only
transforms that section into a large pond; it also re-
duces the temperature and oxygen content, and in-
creases river-bed erosion and water turbidity, while
reservoir operation procedures result in occasional or
regular flooding of sandbars, sandbanks, stretches of
channel mosaic, and other habitats that would normal-
ly be exposed during the dry season, with severe im-
pacts on nesting bird species. The effects on other
wildlife species are widespread. Fish species diversity
in reservoirs is generally less than half that of the orig-
inal river; fish migration routes are invariably cut by
dams only a few meters high; bird and reptile nesting
sites on sandbars or in river banks are destroyed; river-
side forest is replaced by a barren, draw-down zone sur-
rounding the reservoir; and mammalian migration
routes are cut. Estuaries form natural harbors that have
been developed for port facilities and industrial estates
in many places. Mangroves have been converted to
aquacultural (mainly shrimp) ponds, and also degraded
and cleared by charcoal production, while sandy beach-
es have been exploited for intensive tourism. Along
with mangroves, intertidal mudflats are one of the most
threatened natural habitats in the hotspot, having been
extensively afforested with mangrove or intensely
fished by lines of stack nets, which severely impacts
their value as feeding habitat for migratory waterbirds.
Moreover, sand dune ecosystems are severely threat-
ened by afforestation, for instance, with Australian pine
(Casuarina equisetifolia).
Hunting, trapping, fishing, and collecting are inte-
gral to the life-styles of many rural people in the hot-
spot. Previously, much of this was for subsistence use,
although several centuries ago deer hides, antlers, and
rhinoceros horns were already valuable trade com-
modities. In recent decades, the harvest of live exotic
animals for the pet and aquarium trade, of orchids and
other ornamental plants, and of beetles, butterflies,
and seashells has increased enormously. Animal parts
have long been used in traditional medicine, and this
trade continues to flourish. Recently, the demand for
turtles, snakes, pangolins, bears, seahorses, large reef
fishes, and other wildlife for consumption in Southeast On the opposite page, gibbons reach
and East Asia has increased to such levels that wildlife their highest levels of diversity in
collection and export in Southeast Asian countries has Indo-Burma. The white-handed
gibbon (Hylobates lar) seen here
depleted stocks beyond commercial viability and in
occurs in both the Indo-Burma and
many places into local extinction. For example, 90% of Sundaland Hotspots.
the tortoise and freshwater turtle species in the Indo- © Gerald Cubitt
Burma Hotspot are threatened, and mass trade is the
leading cause of threat (Dijk et al. 1999). Harvest of Above, the great hornbill
wildlife has been so intensive in forested areas of the (Buceros bicornis) can be found in
rainforests or evergreen forests.
Lao P.D.R. and Vietnam that the term “empty forest
Most of them live in lowland forests
syndrome” was coined (Redford 1992). Despite some below 1 000 m, although previously
measures to reduce levels of over-exploitation, the they had also been found in high
problems and conservation challenges remain huge. mountains up to 2 000 m. However,
Given all the pressures that this hotspot has faced, it their population is dwindling due to
is not surprising that little habitat remains in pristine hunting and habitat loss. Khao Yai
National Park, Thailand.
condition. While it is difficult to accurately determine
© Art Wolfe
the current extent of natural vegetation, remaining for-
est in pristine condition probably covers less than 5%
of the hotspot, while mildly damaged, yet ecologically
still functional, forest probably covers between a fur-
ther 10% and 25% of the hotspot.
329

Charcoal production is just one of
many threats in the Indo-Burma
Hotspot.
© Patricio Robles Gil/Sierra Madre
On the opposite page, all three
species of douc langur
(Pygathrix spp.) are endemic
to the Indo-Burma Hotspot.
© Anup Shah/naturepl.com
330
Conservation
The total coverage of protected areas in Indo-Burma is
at least 235 758 km2, equivalent to about 10% of the
original extent of natural vegetation, and protecting
examples of many, but not all, vegetation formations
and other wildlife habitats. Certain habitat types re-
main under-represented in the regional protected areas
system, most notably lowland wet evergreen forests,
teak-dominated deciduous forests, riverine habitats
(particularly wide, slow-flowing, lowland rivers), and
intertidal habitats (such as mangroves and intertidal
mudflats). Furthermore, as elsewhere, there are dif-
fering levels of protection among protected areas and,
when only protected areas classified in IUCN cate-
gories I to IV are considered, coverage of protected ar-
eas falls to around 6% of the hotspot.
The countries in the hotspot have a varied history of
protected area establishment. By July 2002, Thailand
had designated 81 terrestrial and 21 marine national
parks, open to the public but safeguarded from distur-
bance or extraction; 55 wildlife sanctuaries closed to
the public and all forms of impacts; and 55 non-hunting
areas, where non-destructive forms of utilization, such
as fishing, are permitted, but hunting is banned
(Carew-Reid 2002). In Myanmar, 38 protected areas had
been declared by 2002, and more are under considera-
tion, making good progress towards the aim of estab-
lishing a protected area system covering about 5% of
the country. Hainan Island contains 26 nature reserves,
totaling about 1 190 km2, that are managed by the
Forestry Department, but there are also a small num-
ber of protected areas run by other government de-
partments (M. Lau, pers. comm.). Cambodia instituted
a National Protected Areas System in 1993, since when
the system has been expanded to currently comprise
seven national parks and 10 wildlife sanctuaries, to-
gether covering 27 658 km2, as well as three protected
landscapes, three multiple-use areas, and seven pro-
tected forests covering 18 469 km2. In Vietnam, there
are currently 95 decreed protected areas, comprising 27
national parks, 40 nature reserves, and 28 cultural and
historical sites, covering a total land area of over
18 000 km2 (Tordoff et al. 2004). The Lao P.D.R. has a
system of 20 national protected areas, covering a total
area of 33 000 km2 or 14% of the nation’s land area
(Southammakoth and Craig 2001). There are also a
number of provincial protected areas in the country, as
well as several sites proposed for protected area status.
While the countries in the hotspot continue to take pos-
itive conservation actions regarding the establishment
and management of protected areas systems, their
achievements are often diminished by roads, reser-
voirs, plantations, and other major developments with-
in and adjacent to protected areas, which undermine
their conservation.
Designation and management of protected areas is
the exclusive domain of governments in the hotspot,
and these governments have increasingly become
aware of environmental damage and are taking pre-
ventive and mitigation measures of varying scope and
effect, comprising legislation (such as wildlife protec-
tion and protected areas legislation), as well as imple-
mentation and enforcement efforts. But conservation
action is not limited to governments. In most countries,
the academic sector is an independent source of
research, support, and initiative regarding local and re-
gional conservation activities. In addition, internation-
al conservation organizations continue to provide ex-
pertise, knowledge, support, and other resources to
responsible government authorities and local NGO
partners. Notable results from recent programs include
identification and prioritization of potential new con-
servation areas, surveys of biodiversity in many sites
and regions, assistance with the development of na-
tional and international conservation legislation, and
raising public awareness of threats and possible solu-
tions for wildlife issues. Other conservation efforts in-
clude varying levels of legal protection and manage-
ment of a wide variety of species, as well as local and
collaborative international research into wildlife man-
agement.
This is a critical time for the biodiversity of Indo-
Burma: much habitat has been lost, yet few known spe-
cies have become extinct to date. Conservation has
been included on the agenda of most national govern-
ments, but effective action is often too little and too
late. Probably the greatest hurdle to progress is the lack
of commitment from national governments to ensure
effective protection of the hotspot’s exceptional biodi-
versity. Recent developments are encouraging, but
must be strengthened urgently to prevent species and
habitat from disappearing in the near future.
PETER PAUL VAN DIJK 71
ANDREW W. “JACK” TORDOFF 19
JOHN FELLOWES 95
MICHAEL LAU 96
MA JINSHUANG 94
 JAPAN
Japan includes more than 3 000 major and minor is-
lands extending from about 22º to 46ºN latitude, or
from the humid subtropics in the south to a temper- CHINA RUSSIA
ate zone (Hokkaido) in the north. It is located at the Hokkaido

intersection of several of the Earth’s crustal plates,

three of which (the Pacific, North American, and NORTH
Philippine plates) plunge into the depths below KOREA SEA
OF
Japan where they generate forces that result in nu- JAPAN
merous volcanoes, hot springs, and mountains. These,
in turn, have a strong influence on the climate and YELLOW SOUTH PACIFIC
SEA KOREA JAPAN OCEAN
vegetation of the Japanese Archipelago. Honshu
In addition to the four main islands of Hokkaido
(78 073 km2), Honshu (227 414 km2), Shikoku EAST Izu-shoto
(18 256 km2), and Kyushu (36 554 km2), the Japan- CHINA
SEA
Shikoku
Kyushu
ese Archipelago is composed of the Ogasawara-shoto
(including the Ogasawara-gunto or Bonin Islands;
and the Kazan-retto, also known as the Iwo or Vol-
cano Islands), Daito-shoto, Nansei-shoto (including
Nansei-shoto Ogasawara-gunto
the Ryukyu Islands and Satsunan Islands), and Izu- TAIWAN (Ryukyu (Bonin Islands)
Islands)
shoto, all with distinctive vegetation that includes
Kazan-retto
numerous endemic species and several endemic gen- (Iwo Islands)

era of plants and animals. The total land area of

Japan is about 373 490 km2. The country is mostly 0 500 km

mountainous (about 73% of the land), and ranges in

elevation from Hachiro-gata at 4 m below sea level
in Akita Prefecture to the highest point, the summit
of Mt. Fuji (Fujiyama or Fuji-san) at 3 776 m in
Shizuoka Prefecture.
Japan was formed by the subduction of the Philip-
pine and Pacific plates under the Eurasian Plate and
by the opening of the Sea of Japan about 15 million
years ago. Prior to those events, the islands that now
make up Japan were part of the Asian Mainland, and
the islands have been regularly connected by land
bridges to the Asian Mainland. It is believed that
Honshu is made up of at least two separate land-
masses that drifted together. The slippage of these
tectonic plates results in frequent and numerous
earthquakes throughout the Japanese Archipelago.
Some, such as the Kobe earthquake of 1995 and the
Great Kanto or Tokyo earthquake of 1923, have re-
sulted in the deaths of more than six thousand
(Kobe) or tens of thousands (Tokyo) of people.
As a result of the wide latitudinal range of Japan,
its position between the Pacific Ocean and the Sea of
Japan, and its numerous high mountains, the cli-
mate varies considerably. During the winter, conti-
nental air masses from the high-pressure zone
formed over eastern Russia generate winds that ab-
sorb large quantities of moisture as they move across
the Sea of Japan. These moist air masses drop
tremendous amounts of snow as they rise and cool
when reaching the central mountains of Honshu,
making it one of the snowiest regions on Earth. At
the same time, the Pacific side of Japan is remark-
ably dry as the now moisture-depleted air masses
cross the mountains (although at other times of the
year, there is abundant rainfall in places). A rainy
season (tsuyu) lasting about two months begins in
early June and moves gradually from south to north,
affecting both sides of Honshu, and eventually On the opposite page, the
reaches Hokkaido, although that island is not as Japanese macaque or snow
strongly influenced. In the winter, the Sea of Ok- monkey (Macaca fuscata) is one
of Japan’s best-known flagship
hotsk is covered with drift ice to such an extent that
species. It frequently bathes in
fishing vessels operate from more southerly ports, al- hot springs in winter.
though in recent years the amount of sea ice each © Jean-Paul Ferrero/Auscape
winter has decreased.
The rainfall pattern reverses in early autumn,
moving from north to south, but is not as severe. By
late autumn most of Japan is generally clear. Annual
precipitation in Tokyo (on the Pacific Ocean side) is
333

The Endangered red-crowned crane
(Grus japonensis) breeds in
southeastern Russia, northeastern
China, Mongolia, and on eastern
Hokkaido in Japan. The wetland
habitat of this species on Hokkaido
continues to be lost to development
and agricultural expansion.
© Jean-Paul Ferrero/Auscape
On the opposite page, the Japanese
serow (Capricornis crispus) occurs
in densely wooded hillsides and
conifer forests in Japan. As with
other species of serow, this one has
been heavily hunted, and declined
to such low levels that it was
declared Endangered in Japan in
the early 1930s. Today, it is
estimated that the Japanese
population numbers about
100 000 individuals.
© Tetsuo Kinoshita/Nature Production
334
1 405 mm; at Owase, on the Kii Peninsula of southern
Honshu, rainfall averages 4 000 mm per year; and at
Takada, on the Sea of Japan side, it is 2 880 mm per
year, about half of which is snow. Yaku-shima, just south
of the southern tip of Kyushu, is among the wettest
places in the world, receiving well over 5 000 mm of
rain per year in places. Japan rarely suffers from a lack
of rain, as can be seen in the lush cover of vegetation
throughout the archipelago. Overall, the climate is gen-
erally mild, being strongly influenced by the warm Kuro
(Kuro-shio) and Tsushima currents from the south. The
cold Chishima (Oya-shio) Current that moves south-
ward from the Kuril Islands sometimes brings heavy fog
to the east coasts of Hokkaido and northern Tohoku
District. Such fog can damage the rice crop, which re-
quires sunny days for successful growth.
Vegetation in Japan ranges from boreal mixed forests
(often called subalpine or subarctic in the Japanese lit-
erature) of Abies (fir), Picea (spruce), and Pinus (pines)
on Hokkaido and at high elevations southward through
Honshu and Shikoku, to subtropical broadleaf ever-
green forests and mangrove swamps in the south. In-
deed, Japan marks the world’s northern limit of man-
groves on the island of Yaku-shima. High elevations on
Honshu and Shikoku support alpine vegetation near
the summits of many mountains. Subalpine vegetation
and natural beech forests, although small in scale, are
distributed throughout the region and are near the south-
ern end of their range. Alpine vegetation occurs near sea
level in several places on Hokkaido and also near and
on the tops of the many mountains and even at lower
elevations in areas with active fumaroles. The soils
around the fumaroles support vegetation that is pecu-
liar to the vicinity where it occurs. The plants in these
areas usually prefer soils of low acidity.
The main species of trees on Hokkaido north of the
Oshima Peninsula (south of Sapporo) are Abies sachali-
nensis, Picea jezoensis, and P. glehnii, which frequently
grow intermixed and in association with birches (Betula
ermanii and B. maximowiczii). On the Oshima Peninsula,
Fagus crenata (Japanese beech) appears and marks the
boundary between the boreal and cool temperate zones.
The conifer forests on Honshu are only on the highest
mountains (generally between 1 600 and 2 500 m eleva-
tion), such as the Japan Alps, but the diversity of
conifers is higher than on Hokkaido and includes Tsuga
diversifolia, Abies veitchii, A. mariesii, Picea koyamai, and
P. maximowiczii. Within the conifer forest zone on Hon-
shu are often pure, conspicuous stands of Betula ermanii.
Honshu is divided throughout its length by high moun-
tains and, at all elevations, there is a clear difference in
vegetation between the Pacific side and Japan Sea side
of the island. The western side of the mountains, facing
the Japan Sea, receives abundant moisture in the form
of snow in the winter. Plants on the Japan Sea side have
adapted to the heavy snow cover, as much as 10 m deep
in some places, and are often distinct at the specific or
infraspecific level from their Pacific side counterparts.
The Pacific side of Honshu, in contrast, is dry and sun-
ny in the winter and receives very little snow.
There are also a few pure and old-growth conifer
forests on Honshu: Thujopsis dolabrata in northern Ao-
mori Prefecture; Cryptomeria japonica forests in Akita
Prefecture; Chamaecyparis obtusa forests in southwest-
ern Nagano Prefecture; and a Sciadopitys verticillata for-
est on Mt. Koya in Wakayama Prefecture. These forests
were preserved because they were the property of fam-
ily clans, Shinto shrines, or Buddhist temples that pro-
tected them from cutting. Often the deciduous trees
were cut, but the conifers were protected so that over
time the composition of these forests became more ho-
mogeneous.
Below the conifer forests are cool temperate broad-
leaf deciduous forests, primarily of Fagus crenata inter-
mixed with Quercus crispula (oak), with a dense under-
growth of Sasa. On the Pacific side of Honshu, Abies
homolepis, often intermixed with Fagus and A. firma,
sometimes forms pure stands adjacent to the Fagus for-
ests. On the Sea of Japan side, broadleaf evergreen
forests are adjacent to the Fagus forests. Sasa is also an
undergrowth plant in the Fagus forests on the Japan
Sea side of Honshu, but the species are different from
those on the Pacific side. Other trees in the Fagus
forests are numerous, and include several species of
Acer (maples), Tilia (linden), Fraxinus (ash), Prunus
(cherries), Magnolia, Kalopanax, Sorbus (mountain ash),
with Aesculus (chestnut), Cercidiphyllum (Katsura), Eup-
telea, and Alnus maximowiczii in more moist places.
Below the Fagus forests are warm temperate broad-
leaf forests primarily made up of various members of
the Fagaceae (beech and oak family), including several
species of Quercus, Castanopsis, and Lithocarpus, Lau-
raceae and Theaceae, with an admixture of numerous
other trees species, both evergreen and deciduous.
Warm temperate vegetation, characterized by broad-
leaf evergreen trees and also trees and plants of more
southern affinities, extends from central Honshu to
Shikoku, Kyushu, and southward. In the northern por-
tion of its range, this vegetation type is restricted most-
ly to the immediate coastal areas, but extends farther
inland towards the south. Characteristic warm temper-
ate broadleaf evergreen trees are various species of oak
(Quercus subgenus Cyclobalanopsis), Lithocarpus, Cas-
tanopsis, Cinnamomum, Machilus, Camellia, and Cleye-
ra. This association marks the northernmost extent of
the Fagaceae-Lauraceae-Theaceae forests that domi-
nate in Southeast Asia. These forests reach 38ºN lati-
tude on both the Japan Sea and Pacific coasts of Hon-
shu, where they give way to temperate broadleaf
deciduous forests.
The subtropical island chains lying to the south of
Japan support a flora and fauna different from the
main Japanese Islands, including a large number of
endemic plants and animals. The Izu-shoto, Oga-
sawara-gunto (or Bonin) and Kazan-retto (Iwo or Vol-
cano) islands comprise 30 or more volcanic islands that
extend southward from Tokyo in the Pacific Ocean. The
Izu-shoto stretches from Oshima in the north, where
the high point is 854 m, to Torishima in the south. The
Ogasawara-gunto can be divided into three groups:
Muko-jima, Chichijima-retto, and Hahajima-retto. The
largest island, Chichi-jima, has an area of 24 km2.
The Ogasawara-gunto have a diverse topography and,
although much of the land has been cleared on the
northernmost islands, some primary forest still remains
in the more inaccessible areas to the south. Broadleaf
evergreen forests cover extensive areas on Haha-jima,
parts of Chichi-jima, and some of the smaller islands,
but most are secondary. Primary native forests are in
the more remote and inaccessible parts. Kazan-retto
comprises three islands, one of which, Iwo-jima (which
means Sulfur Island), is a submerged caldera better
known as a battleground during World War II.
The Nansei-shoto Islands consist of more than 100 is-
lands, from Yaku-shima, south of Kyushu, to Iriomote-
jima, located about 200 km east of Taiwan. The island
chain comprises two main groups: the Satsunan Is-
lands, including the Oosumi Islands, Tokara-retto, and
Amami-shoto, and the Ryukyu Islands, including Oki-
nawa-shoto and Sakishima-shoto. The northern islands
are more temperate in terms of vegetation cover, while
south of Amami the flora is more subtropical. Okinawa-
honto, an island in the Okinawa-shoto and the largest
island in the Nansei-shoto at 1 200 km2, contains the
highest percentage of natural forest in Japan outside of
Hokkaido. Iriomote-jima (293 km2) and northern Oki-
nawa also support especially extensive areas of broad-
leaf evergreen forests. The ecosystems of the Ryukyus,
such as mangroves and coral reefs, are rarely found in
other areas of Japan. The Oriental faunal region, and
the southern limit of the Palaearctic, integrade in the
Nansei-shoto such that these islands harbor fauna from
both regions, as well as their own unique species (Stat-
tersfield et al. 1998).
Despite the relatively small land area and the gener-
al perception of a homogeneous society, Japan is as di-
verse culturally as it is biologically, mainly due to the
same reasons: its geological and geographical complex-
ity. High mountains and ocean, as well as domestic
travel restrictions that were in place until the late
1800s, virtually divided the country into numerous cul-
tural, if not ethnic, groups. The Japanese language is a
distinct language with no linguistic affiliation with oth-
er languages, but has a variety of dialects. There is also
a completely different language, Ainu, which is spoken
by the Ainu people in Hokkaido and the northern parts
of Honshu. However, the Ainu language is said to be at
the verge of extinction, since few people can speak it,
and the Ainu culture was nearly lost until recent inter-
est in it by the remaining Ainu has given it new life
(Fitzhugh and Dubreil 2000). Despite the diversity
within the Japanese language and culture, Japanese
people’s love of nature is uniform and strong, and has a
long history. This may be due to the fact that Japan is
such a small country with a mostly mountainous topol-
ogy and very little arable land. This has forced the
Japanese people to use natural resources in a sustain-
able manner and to live harmoniously with nature.
This spirit is also based on the religious beliefs of Bud-
dhism and the indigenous Shintoism. The word “na-
ture” or shizen in Japanese literally means “being with-
in” or “being part of.” Evidence of such a spirit can be
found in Man’yoshu, an early anthology of poetry and
one of Japan’s earliest recorded documents remaining,
where more than 150 species of plants and over 70 an-
imal species are expressed in the most popular poems
of Japan written before the eighth century. Another
distinct connection between the Japanese people and
nature can be seen in Haiku, the very short 17-syllable
poems, which require as a rule that nature be ex-
pressed.
Biodiversity
There are approximately 5 600 species of vascular
plants in Japan (ferns and fern allies, conifers, and
flowering plants). About 35% of the plants, or some
1 950 species are believed to be endemic (Iwatsuki et al.
1993; 1995a, b; 2001). There are three endemic plant
families, Sciadopityaceae, Glaucidiaceae, and Pterido-
phyllaceae, and around 20 endemic genera, seven of
which are monotypic. The endemic genera include
Sciadopitys (one species, S. verticillata), Anemopsis, Ran-
zania (one species, R. japonica), Glaucidium (one spe-
cies, G. palmatum), Pteridophyllum (one species, P. race-
mosum), Dendrocacalia (one species, D. crepidifolia),
Mallotopus, Kinugasa, Macropodium, Japanolirion (one
species, J. osense), Tsusiophyllum (one species, T. tana-
kae), Boninia (two species, B. grisea and B. glabra), Boni- On the opposite page, misty forest in
nofatsia (two species, B. oligocarpella and B. wilsonii), Yaku-shima, which is situated some
Platycrater, Ancistrocarya, Perillula, Alectorurus, and Dias- 60 km south of the southern tip of
Kyushu, and is covered by various
panthus. There are also a number of interesting near
kinds of trees including Sugi
endemics, such as Trochodendron aralioides, an unusual (Cryptomeria japonica), which is
and primitive vessel-less angiosperm that occurs spo- indigenous to the island.
radically from central Honshu southwards, but other- © Kazuma Anezaki/Nature Production
wise occurs elsewhere only on Taiwan. Interestingly,
though, the genera that are considered characteristic of Above, Japanese giant flying
squirrel (Petaurista leucogenys)
the Japanese flora are rather poorly represented on
in Shirakami-sanchi (Shirakami
Taiwan, with Japan’s flora having come mostly from Mountains), situated in the
east-central China, Korea, and the islands and main- northern part of Honshu, Japan.
land to the north, whereas Taiwan’s came from the Phil- This area of primeval forest of
ippines (southern tip) and southeastern China. mainly Japanese beeches was
On Ogasawara-shoto, there are about 500 native inscribed as a World Heritage Site
in 1993.
plant species, of which about 43% are endemic; of the
© Masayuki Egawa/Nature Production
native tree species, endemism is around 75% (73 of 97
species) (Kitayama 1991). Two genera, Dendrocacalia
and Boninia, are endemic to these islands, and a third,
Dendrocacalia crepidifolia, is endemic to, but widespread
on, Haha-jima.
About 90 genera of the plants in Japan are members
of what has been referred to as the Arcto-Tertiary
337

On pp. 338-339, subtropical
rainforest in the Ogasawara-gunto,
a group of islands some 1 000 km
south of Tokyo, generated as a result
of volcanic activity. Many species
on the Ogasawara-gunto are
threatened by introduced species,
particularly goats.
© Takashi Uzu/Nature Production
Above, the Okinawa woodpecker
(Sapheopipo noguchii), sole
representative of an endemic genus,
is a Critically Endangered species
now found only on Okinawa, where
it is confined to the Yanbaru Forest.
© Kazuo Minato/Nature Production
340
Geoflora (Boufford and Spongberg 1983; Wen 1999). Gen-
era in this group usually represent ancient lineages
that were once distributed around the world in the Ter-
tiary Period. Distinctive genera include: Buckleya, Cau-
lophyllum, Diphylleia, Stewartia, Schisandra, Illicium,
Wisteria, Lespedeza, Penthorum, Itea, Astilbe, Mitella,
Menispermum, and Shortia. Many of them are well
known as fossils at high latitudes worldwide, hence
their designation as “Arcto-Tertiary.” Although the
forests in which the genera occurred were once wide-
spread and continuous, they now have a very frag-
mented distribution. Members of this group have one
or more species in two or three widely separated parts
of the world: southeastern North America, Japan, and
China. Although the species in each region are distinct
from those in the other regions, they are similar to
each other in that their generic placement has never
been questioned. The origins of this biogeographic pat-
tern have been a source of speculation since it was dis-
covered nearly 200 years ago, and even attracted the at-
tention of Charles Darwin as he was preparing his book
The Origin of Species.
Although vertebrate diversity is not especially high
in Japan, levels of endemism are marked. Among
mammals, for instance, only 91 species are recorded,
but half (46) are endemic. Moreover, there are six en-
demic genera, three of which are monotypic: the Ama-
mi rabbit (Pentalagus furnessi, EN), found on Amami-
oshima and Tokuno-shima in the Amami-shoto in the
Nansei-shoto; the Japanese dormouse (Glirulus japoni-
cus, EN), found on Honshu and Shikoku; and the
Ryukyu long-tailed giant rat (Diplothrix legatus, EN), re-
stricted to the Ryukyu Islands of Amami-oshima and
Tokuno-shima, and Yanbaru Forest on northern Oki-
nawa. The other three endemic genera (Phaulomys,
Tokudaia, and Urotrichus) are represented by two spe-
cies each. The spiny rats of the genus Tokudaia have
severely restricted ranges, with the Ryukyu spiny rat
(T. osimensis, EN) found only on Amami-oshima, and
Muennink’s spiny rat (T. muenninki, CR) known only
from Yanbaru Forest. Sado Island (857 km2) off west-
ern Honshu has two endemic mammals, the Sado
shrew (Sorex sadonis, CR) and the Sado mole (Mogera
tokudae, EN).
Japan’s reptile fauna includes 64 species, 28 of which
are endemic, and four are considered threatened: the
Okinawa black-breasted leaf-turtle (Geoemyda japonica,
EN); Kikuzato’s brook snake (Opisthotropis kikuzatoi,
CR), found only on the Kume-jima in the Okinawa-
shoto in the Ryukyus; the Amami takachiho snake
(Achalinus werneri, VU), confined to Amami; and the
Tokashiki ground gecko (Goniurosaurus kuroiwae, VU),
from the Ryukyus. Amphibians include 58 species, of
which 44 are endemic. The amphibian genus Hynobius
is also particularly well represented, with 15 of the 23
known species endemic, one of which, the Oki sala-
mander (H. okiensis, EN), is confined entirely to Dogo,
of the Okino-shima in Shimane Prefecture.
There are 368 bird species known to occur regularly
in the hotspot, although only 15 are endemic. One of
these, the Okinawa woodpecker (Sapheopipo noguchii,
CR), is the only representative of an endemic genus,
and is found in Yanbaru Forest. Three Endemic Bird Ar-
eas (EBAs), as identified by BirdLife International (Stat-
tersfield et al. 1998) are found within Japan: the Izu Is-
lands, with two species confined entirely to this EBA;
the Ogasawara Islands, with a single restricted-range
species confined to it, the Bonin honeyeater (Apalopte-
ron familiare, VU); and Nansei-shoto, with seven species
confined to this EBA, five of which occur only on Ama-
mi and Okinawa.
The freshwater fish fauna is relatively small, with
214 native species of which 52 are endemic. Inland wa-
ters are dominated by fishes entering from the sea, but
they also include a significant representation of strictly
freshwater groups such as minnows (Cyprinidae) and
loaches (Cobitidae and Balitoridae) that have under-
gone diversification within the hotspot. These radia-
tions account for nearly half of the endemic species
and three of its four endemic genera. Species in ancient
lineages (five lampreys and four sturgeons) are an im-
portant component of the fauna, placing it among the
highest-ranking hotspots in terms of phyletic rarity of
fishes.
Some invertebrate groups are fairly well documented
in Japan. For example, around 237 butterfly species are
thought to be native to the country, although this figure
includes naturalized species (Matsuka 2003), while of
24 species of tiger beetles recorded, six are endemic (D.
Pearson, pers. comm.).
Flagship Species
Japan’s flora includes a number of rare, endemic plants
that are well-known favorites in gardens both in Japan
and in temperate areas worldwide. Shirane-aoi (Glau-
cidium palmatum) has perhaps the most extensive
range, occurring from central Honshu to southern
Hokkaido, and is more common on the Sea of Japan
side of Honshu. The large, blue-purple, or rarely white,
flowers make it a favorite horticultural subject, and are
especially noticeable when in bloom in patches of open
ground surrounded by deep snow. Togakushisgouma
(Ranzania japonica) is rarer still, and is found only
in the high mountains from central to northern Hon-
shu. The pale purple flowers of this species are partic-
ularly striking.
Kouya-maki (Sciadopitys verticillata), a conifer tree,
occurs from central Honshu to Shikoku and Kyushu. It
is unusual in having pairs of needle-like leaves com-
pletely fused and then borne in whorls near the apex of
the branches. An excellent Sciadopitys forest, formerly
belonging to a Buddhist temple, occurs on Mt. Koya in
Wakayama Prefecture. Sciadopitys also occurs in the
Sohayaki region of Kyushu, Shikoku, and the Kii Penin-
sula of Honshu. Although Sciadopitys is endemic to
Japan, other plants in the Sohayaki floristic region of-
ten have relatives in western China.
Urahagusa (Hakonechloa macra) is a rather rare and
local grass that grows on wet, rocky cliffs in the Tokai
region of Japan. Because of some forms with variegat-
ed leaves and its graceful arching habit and tendency to
form dense mounds, it is a favorite pot and garden
plant.
Perhaps the best-known flagship in this hotspot is the
Japanese macaque (Macaca fuscata), the most northerly-
living non-human primate, found on Honshu, Shikoku,
and Kyushu (and a few other islands). These animals
play a prominent role in Japanese culture, and are fa-
mous around the world as the “snow monkeys,” which
regularly romp in the snow in winter and also make
use of Japan’s abundant volcanic springs. Japanese
macaques were also very important in the evolution of
the science of primatology, with some of the very first
field studies of wild primates having been conducted
by pioneering Japanese researchers back in the 1950s.
There is also a second macaque endemic to Japan, the
Yaku-shima macaque (M. yakui), found only on the is-
land of Yaku-shima. It used to be considered a sub-
species, but is now considered by many experts to war-
rant full species status.
The Iriomote cat (Prionailurus iriomotensis) is unique
to Iriomote-jima. The species was described in 1967,
when it was noted as being one of the most primitive of
Asian felids. Since then, studies have suggested it may
be a subspecies of the leopard cat (P. bengalensis). The
species is thought to number no more than 100 indi-
viduals. Another Ryukyu endemic is the Amami rabbit.
Following a dramatic decline in numbers in the early
twentieth century, the species was declared a national
monument and given complete legal protection, al-
though this status confers no protection on the species’
habitat. The Amami rabbit habitat is subject to exten-
sive fragmentation, particularly of mature forests
(which make up 10% of Amami and 30% of Tokuno-
shima), and introduced mongooses are also a signifi-
cant threat; it is estimated that perhaps only 2 500 indi-
viduals survive.
The Japanese giant salamander or Osanshouo in
Japanese (Andrias japonicus) is distributed in western
Honshu, Shikoku, and Kyushu, and is one of the world’s
largest amphibians. This extraordinary species can
grow to more than one meter in length (the young tak-
ing five years to reach maturity), and feeds on fish and
crustaceans. The species was put under protection in
1951, after its population was threatened by human
consumption. No longer part of the Japanese diet, its
population has grown and the species is designated as a
special natural monument in Japan. Another flagship
amphibian is Ryukyu spiny newt or Iboimori (Echinotriton
andersoni, EN), which suggests crocodilian features. This
species is distributed on six islands: Amami-oshima,
Tokuno-shima, Yoro-shima, Okinawa-honto, Sesoko-
jima, and Tokashiki-jima. Like its larger relative, it has
been designated as a national monument in Okinawa
and Kagoshima prefectures.
One of the best avian flagships is the Okinawa rail
(Gallirallus okinawae, EN), endemic to Okinawa Island,
where it is confined to Yanbaru in the northern quarter
of the island. The population of this species is estimat-
ed to be around 900 pairs or 1 500-2 100 birds. Yanbaru is
also home to the Okinawa woodpecker, with the main
breeding areas along the mountain ridges between Mt.
Nishime-take and Mt. Iyu-take. This species was consid-
ered close to extinction in the 1930s. In the early 1990s,
the breeding population was estimated to be only 75
birds, and it now stands at between 146 and 584 birds.
Finally, Japan supports some very important water-
bird populations, notably those of the resident red-
crowned cranes (Grus japonensis, EN) on Hokkaido,
and the wintering hooded cranes (G. monacha, VU) and
white-naped cranes (G. vipio, VU) at Izumi on Kyushu,
where around 85% and 40% of the global populations
of these species, respectively, overwinter. In large
part, these unnaturally large concentrations are due to
food provision, as natural habitats and sufficient food
sources no longer exist there, such that the cranes
exist largely on intensive management, which is sup-
ported by the Japanese government. Such unnatural
concentrations put the populations at an elevated risk
of disease or some other such catastrophe (BirdLife In-
ternational 2003).
Threats
Fortunately, in some ways, approximately 70% of
Japan’s 127.5 million people (Population Reference Bu-
reau 2003) are concentrated on 3% of the land area, re-
sulting in a high population density in those areas, but
leaving other areas relatively sparsely populated. That Hokkaido sika deer
said, human settlement in the vicinity of protected ar- (Cervus nippon yesoensis)
eas is considerable, with an estimated 33 million people on the Shiretoko Peninsula on
Hokkaido, a subspecies that
living within 10 km of protected areas in 2002 (L.
apparently has recovered from a
Gorenflo, pers. comm.). Although threats to Japan’s en- population bottleneck about more
vironment from logging interests are not high because than a century ago. Its distribution
of the high cost of Japanese timber compared to cheap has expanded rapidly in the last
imported timber from other parts of the world, the af- three decades.
fluence of Japanese society in general, the high GDP © Masami Goto/Nature Production
per capita achieved within a few decades following
World War II, the reduced work week in recent years,
and the resultant increase in leisure time have placed a
different type of strain on the environment. Forests are
being cleared for ski resorts and golf courses, which
were promoted in the 1980s by government policy, and
more roads have been built to accommodate the steady
increase in automobiles and the desire by many people
to use private rather than public transportation. In re-
cent years, too, fast public transportation has been ex-
panded. It is now possible to travel to such remote parts
of Japan as Yamagata and Akita in the north by bullet
341
train, the Shinkansen, in less than three hours from
Tokyo. Many more vacationers travel to those once re-
mote and sparsely populated areas. With the increase
in tourists has come a greater demand for services and
recreational facilities, which in turn has put more pres-
sure on undeveloped areas.
Since World War II, the country’s high-elevation
conifer forests have been under the administration of
the Forestry Agency of Japan. In response to demands
for timber and pulp following World War II, the agency
promoted clear-cutting of these forests and the re-
planting of deforested areas with Larix leptolepis. Today,
plantations are widespread throughout Japan, with
about 90% of them made up of Cryptomeria japonica
(Sugi), Chamaecyparis obtusa (Hinoki), and Larix lepto-
lepis (Kara-matsu), all native Japanese species. Fagus
was traditionally not used for construction until after
about 1900. From then on, Fagus was used in increas-
ing amounts for furniture, pulp, and construction, and
during World War II for building light aircraft. Clear-cut
areas that supported Fagus have been replaced with
Cryptomeria or Larix plantations, or by pure stands of
Sasa where it has been impossible to establish Cryp-
tomeria or Larix. Natural Fagus crenata vegetation still
occurs in the central mountainous areas of Honshu,
but is replaced by evergreen broadleaf forests in the
southern parts of the island.
As a result of the National Survey on the Natural En-
vironment, Japan maintains vegetation maps (scale
1:50 000) covering the entire national landmass, with
766 plant-community categories. The plant communi-
ties are classified into ten types: natural vegetation of
grasslands; natural vegetation of moorlands; natural
vegetation of forests; secondary forests approximating
the natural vegetation; substitute vegetation of sec-
ondary forests; planted forests; substitute vegetation of
high-profile grasslands; substitute vegetation of low-
profile orchards; substitute vegetation of low-profile
paddies and fields; and urban land. When considering
the percentage of each vegetation type occupying the
overall national land area, forests (natural forests, sec-
ondary forests approximating natural forest, secondary
forests, and planted forests) comprise 67.5% of the na-
tional land. Natural forest vegetation constitutes 18.2%.
Adding the natural vegetation of grasslands and moor-
lands provides a total natural vegetation of 19.3%. Sec-
ondary forests (including secondary forests approxi-
mating natural forest) comprise 24.6% of the national
land, while planted forests occupy 24.7%, secondary
grassland 3.2%, paddies and fields 22.7%, and urban
land 4.0%. The natural forests and natural grasslands
referred to here indicate forests or grasslands that are
climax or are regarded as climax. The most heavily
forested region is the island of Hokkaido, of which 50%
is covered with natural vegetation. Shikoku, however,
has the highest percentage of forest cover in Japan,
while Kyushu has the highest percentage of land in cul-
tivated forests, paddy fields, and croplands in the coun-
try. Overall, deforestation has now ceased to a large de-
gree in the lowlands, and there are still extensive tracts
of forest cover in the higher-lying regions.
Besides forested areas, coastal regions and wetlands
have also been subject to disturbance. On Hokkaido,
the wetlands favored by nesting red-crowned cranes
continue to be lost to development, mainly agricultural
expansion, river channelization, and road building. For
instance, one-third of almost 300 km2 of marshland in
Kushiro has been converted to agricultural, industrial
or residential use since the 1970s (BirdLife Internation-
al 2003).
Unfortunately, as a result of past habitat loss and also
the effects of hunting and pesticide use, several species
have undergone precipitous declines in their popula-
tions, including the oriental stork (Ciconia boyciana,
VU), which no longer breeds in Japan, and the crested
ibis (Nipponia nippon, EN), which is now known to sur-
vive only in the You Prefecture in Shaanxi Province in
central China, where it is a localized breeder. In both
cases, efforts are under way to reintroduce these species
to Japan, with several captive-bred crested ibises from
China having been sent to Japan, and similar plans
to reintroduce captive-bred oriental storks. The popula-
tion of short-tailed albatross (Phoebastria albatrus, VU)
crashed to near extinction during the twentieth century
following massive exploitation for its feathers, although
protection since the 1960s has helped the population to
recover; today, it breeds only on Torishima and the
Senkaku Islands (BirdLife International 2003).
Like much of the rest of Japan, the Ryukyus and Oga-
sawaras have suffered from habitat loss due to the plant-
ing of timber plantations and urban development (and
volcanic eruptions in 2000 resulted in serious damage
to forests on Miyake-jima in the Izu Islands). Almost all
the original subtropical forest on the Ogasawaras has
been cleared, likely one of the factors that led to the ex-
tinction of three avian endemics during the nineteenth
century (Bonin wood-pigeon, Columba versicolor; Bonin
thrush, Zoothera terrestris; and Bonin grosbeak, Chauno-
proctus ferreorostris). In the Ryukyus, only small areas of On the opposite page, ayu fish
forest remain on Amami and Okinawa, mainly in pro- (Plecoglossus altivelis) are a trout-
tected areas, and mature forest now only covers less like species with an interesting
shuttlecock-like dorsal fin. They are
than 5% of Amami (BirdLife International 2003).
an esteemed food item in Japan.
As with other subtropical parts of the world, one of the © Ryu Uchiyama/Nature Production
greatest threats to the native fauna and flora of Japan is
from alien plants and animals, some of which were in- Above, the Iriomote cat
troduced for the purposes of snake control, including (Prionailurus iriomotensis)
the Indian grey mongoose (Herpestes edwardsi), Javan is found only on Iriomote-jima,
and may number no more than
mongoose (H. javanicus), and Siberian weasel (Mustela
100 individuals.
sibirica). For example, the introduction of the Siberian
© Makoto Yokotsuka/Nature
weasel to Miyake-jima in the 1970s and 1980s caused Production
significant declines of Japanese night-herons (Gorsa-
chius goisagi, EN) and Izu thrushes (Turdus celaenops,
VU). On Amami, the Javan mongoose is thought to be
responsible for declines in the Amami woodcock (Sco-
lopax mira, VU) and Amami rabbit. Other introduced
species in Japan include the fish Tilapia zillii and the
343

The Bonin flying fox
(Pteropus pselaphon) is a
Critically Endangered bat species
known only from a handful of
islands in the Ogasawara-gunto
(also known as the Bonin Islands).
© Takashi Uzu/Nature Production
On the opposite page, deforestation
and forest road network in
Shirakami-sanchi
World Heritage Site in the
northern part of Honshu, Japan.
© Kazuma Anezaki/Nature Production
344
toad Bufo gargarizans miyakonis, while introduced goats
are a problem on some islands. Largemouth bass (Mi-
cropterus salmoides), too, pose a serious threat to native
fishes throughout Japan.
Conservation
Although Japan’s 28 national parks have been desig-
nated to preserve “areas of the greatest natural scenic
beauty,” they also protect some of the richest areas of
biodiversity. Together, they cover about 5.5% of Japan’s
land area. In addition to the national parks, there are
numerous quasi-national parks, prefectural natural
parks, and prefectural wildlife protection areas. Taking
into account all protected areas in IUCN categories I to
IV, just under 6% of this hotspot can be considered
to have a high level of protection, although as much as
16% of the hotspot has at least some form of legal pro-
tection when one also includes those protected areas
not classified in these categories. There are also two
Natural World Heritage Sites in Japan, both established
in 1993, namely Shirakami-sanchi, in northern Hon-
shu, and Yaku-shima in the Satsunan-shoto, which con-
tains ancient trees of sugi or Japanese cedar. Many of
the parks, or parts of the parks, may not be completely
preserved areas, and often there are private agricultur-
al lands or other commercial development activities
within the park. Furthermore, there are weaknesses in
reserve management in some areas, such as in the Izu-
shoto, which although declared a national park, contin-
ues to suffer habitat loss.
The main obvious gaps in the protected area system
are in the Ryukyus; small, protected areas exist in
northern Okinawa and on Amami, but most forested ar-
eas are not properly protected. One example is Yan-
baru, which supports important populations of six of
the 32 Critically Endangered and Endangered species
in Japan, including the entire global populations of Oki-
nawa rail and Okinawa woodpecker. Around 25% of
Yanbaru on the eastern slope of the central montane
area is located in the U.S. Marine Corps Training Area,
while the rest of the forest is threatened by clear-cutting
and removal of forest undergrowth: during a 13-year
period, from 1979 to 1991, some 24 km2 of forests were
cut down, more than 60% of which was in the central
part of the forests, while it has been estimated that the
undergrowth has been removed from half of the natural
forests (Ito et al. 2000). It is also worth noting that 19%
of Okinawa-honto’s land area is under exclusive use by
U.S. military bases.
The policy until now regarding protected areas has
been to try to establish the parks with the least nega-
tive impact on the local economy, while at the same
time preserving landscape areas of great national im-
portance. Wildlife within the parks is strictly protect-
ed, and the destruction and collecting of plants and
animals is prohibited. The parks, however, are used
for such activities as hiking, skiing, mountain climb-
ing, camping, boating, swimming, bird watching and
general sightseeing. On the Ogasawaras, more pre-
serves have been established and tourism has become
regulated.
Although pressures on once-remote parts of the
country have increased, in recent years the people of
Japan have returned to embracing the concept of a
green, sustainable world where humans value and care
for their natural resources, and again recognize them-
selves as one part of the environment, just as their an-
cestors did. A significant indication of the value being
placed on renewable resources and green technology
is the Cosmos Prize, one of the world’s top environ-
mental awards, established by the Expo’90 Committee
and presented annually by the Expo’90 Foundation of
Japan. Expo’90 was an event dedicated to the theme
“The Harmonious Coexistence of Nature and Mankind:
How we as human beings can truly respect and live in
harmony with nature.” The purpose of the prize is to
honor those who have, through their work, applied
and realized the ideals promoted at the Expo, but it has
also focused the attention of the public on the need for
conservation and the preservation of nature. Nature
and greenery are now constantly being mentioned in
advertising, by the news media, and in government
policy announcements.
Finally, it is worth mentioning that Japan in recent
years has become an important player in international
biodiversity conservation and an advocate for the hot-
spots. Indeed, the Japanese Government is one of five
partners (along with the World Bank, the Global Envi-
ronment Facility, the MacArthur Foundation, and Con-
servation International) in the Critical Ecosystem Part-
nership Fund (CEPF), which is providing $125 million
over five years to hotspot conservation. Now that Japan
itself has entered the hotspot list, we hope that the
country’s commitment to these critically important re-
gions of our planet will increase even more.
DAVID E. BOUFFORD 94
YASUSHI HIBI 93
HIROMI TADA 93
 EAST MELANESIAN
ISLANDS
Although it was not previously identified as a bio-
0 500 km
diversity hotspot, this region’s accelerating habitat
loss and additional research done there have led to
the identification of the group of East Melanesian PAPUA
Admiralty
islands northeast and east of New Guinea as re- Islands New
Ireland
PACIFIC
OCEAN
quiring hotspot status. This assemblage of great NEW
Bismarck
tropical oceanic islands is without parallel in its Archipelago

combination of insular biodiversity, unique envi- GUINEA Bougainville

ronment, and amazingly rich diversity of tradi- New

Britain SOLOMON
ISLANDS
tional cultures.
The East Melanesian Islands Hotspot encompass-
es the islands of the Bismarck Archipelago, the
Solomon chain, the Santa Cruz Islands (Temotu),
and the islands of Vanuatu. Politically, this includes
the northeastern insular portion of Papua New
Guinea (including the large islands of New Britain,
New Ireland, and Manus); Buka and Bougainville,
CORAL SEA
which are at the northern end of the Solomon VANUATU
chain, but politically part of Papua New Guinea;
and the entirety of the nations of the Solomon Is-
lands and Vanuatu. These islands are little known AUSTRALIA
because of their inaccessibility and because they lie New
Caledonia
within the considerable shadow cast by the subcon- (Fr.)

tinental island of New Guinea —the great biodiver-

sity generator of the tropical Pacific. Being truly
oceanic, the islands of the East Melanesian Islands
Hotspot are relatively species-poor when compared
to the biotic riches of New Guinea. However, when
compared to other tropical Pacific island groups or
any other islands of comparable size, they are the
most diverse physiographically, biotically, and eth-
nically.
Lying directly on the Pacific “Ring of Fire” strings
of active volcanoes erupt periodically, shaking the
earth and sending plumes of ash over the surround-
ing landscape. Rabaul City, capital of East New
Britain Province, was destroyed by the eruption of
Tavurvur in August, 1994. The island of Tanna in
Vanuatu also has an active volcano, which is famed
for being the world’s most accessible. Rough and
rocky coastlines are found along many shore areas,
while in other parts, mangroves or fringing coral
reefs are more typical. This inaccessibility has hin-
dered development and commerce. The region’s reef
ecosystem diversity is high, containing some of the
most pristine reefs on Earth. The region’s rainforests
look much like those found on New Guinea, and
many of New Guinea’s common forest trees are also
found dominating the forests of this hotspot.
The oceanic island arcs of this hotspot are a mix-
ture of young volcanics and very old basement rocks
that date back to the Cretaceous, the same geological
period in which New Zealand and New Caledonia On the opposite page, aerial view
were drifting away from the disintegrating Gond- of rainforest-covered islands and
wanaland. Thus, while the hotspot contains classic coral reefs in the New Georgia
group of the Solomon Islands.
examples of relatively recent adaptive radiation typ-
The Solomons have been heavily
ical of oceanic islands, such as the white-eyes (fami- impacted by logging and
ly Zosteropidae) and monarch flycatchers (family conversion to coconut plantations
Monarchidae), it also carries some odd colonizers over the last decade.
from times past such as the giant monkey-tailed © Michael Pitts/naturepl.com
skink of the genus Corucia, whose closest living rela-
tives are the blue-tongued skinks (genus Tiliqua) of
Australia, New Guinea, and Indonesia; and the giant
Placostylus land snails shared with the Gondwana-
land fragments of Australia, New Caledonia, New
347

The Solomons prehensile-tailed
skink (Corucia zebrata) is a large,
arboreal skink genus that is
endemic to the Solomon
Islands. Although usually
considered a single species,
it is likely that further research will
indicate that a number of species
exist on the different islands
in this country.
© Joe McDonald/Auscape
348
Zealand, and also with Fiji. The East Melanesian Is-
lands Hotspot has affinities with Fiji (included as part
of the Polynesia-Micronesia Hotspot, but which proba-
bly sits better culturally within Melanesia), such as
Platymantis frogs, ancient “monkey-faced” fruit bats of
the genus Pteralopex, and Nesoclopeus rails —all of which
trace back to colonization in a distant archipelagic past
shared with Fiji.
The East Melanesian Islands Hotspot is one of the
most geographically complex areas on Earth. Intricate
tectonic plate movements have produced a mix of col-
liding and subducting plates, which, in turn, have gen-
erated deep oceanic trenches and affiliated strings of
islands of varying age and development. At the top
of the hotspot, at 2ºS latitude, lies the Bismarck Archi-
pelago’s Admiralty group, dominated by Manus Island
(1 834 km2), which is surrounded by a constellation of
small islands, the largest of which are Rambutyo and
Lou. Northeast of the Admiralties lies the Saint
Matthias Group; the largest, Mussau (or St. Matthias;
414 km2) stands as the head of a chain of small island
groups (Tabar, Lihir, Tanga, Feni, and Nissan) that lead
southeastward to Buka and Bougainville, in the
Solomon chain.
The two main islands of the Bismarck Archipelago
—New Ireland (7 174 km2) and New Britain (35 742 km2)—
lie south and southwest of the Tabar-Nissan islands
string. Both islands are complex and mountainous, with
peaks exceeding 2 000 m. Just west of the northern tip of
New Ireland lies New Hanover (1 186 km2), while north
and west of New Britain one finds several island outliers
—the Witu Islands, Siassi Islands (including Umboi), and
Long and Crown in the extreme west, just north of the
coast of mainland Papua New Guinea. The island chain
that includes Long and Umboi arcs northwestward off
the coast of New Guinea and comprises a series of recent
volcanoes, some still active.
Mountainous Bougainville, the largest in the Solo-
mon chain, covers 8 591 km2 and supports several
high massifs (some volcanic), the highest of which,
Mount Balbi, stands 2 685 m above sea level, the high
point of the hotspot. Together with Buka (611 km2),
the northernmost of the Solomon group, these two is-
lands were part of the German New Guinea colonial
administration, and are today part of Papua New
Guinea. The remainder of the Solomons constitutes
the independent nation of the Solomon Islands (a
former British protectorate). Together they comprise
two parallel chains of large islands. The northern
chain includes (from west to east): Choiseul, Santa
Isabel, and Malaita, while the southern chain com-
prises Vella Lavella, Kolombangara, New Georgia and
Rendova, the Russell Islands, Guadalcanal, and Maki-
ra. Isolated Rennell Island (684 km2), one of the most
remarkable and unique islands in the region, lies
some 175 km to the south of Guadalcanal; this island
is home to the largest freshwater lake in the insular
Pacific (Lake Tegano). Most isolated of all in this is-
land nation are the Santa Cruz Islands, 375 km east of
Makira and dominated by Nendo, Tinakula, Utupua,
and Vanikoro. These lie at the top of the hotspot’s
easternmost island chain that leads south to Vanuatu
(another legacy of nineteenth-century colonialism,
the Solomons being a British protectorate and Vanu-
atu being a French-British Condominium of the New
Hebrides, referred to in history texts as the “Pande-
monium”).
The diverse islands of Vanuatu (formerly the New
Hebrides) lie almost 300 km south-southeast from the
Temotu group. Geographically, Vanuatu resembles
the Solomons, but has smaller, lower, younger islands.
To the north lie the Banks and Torres Islands, followed
by a double chain of large islands (Espiritu Santo
and Malakula to the west; Maewo, Pentecost, Am-
brym, and Efate to the east). Below Efate, a single
chain of smaller islands appears: Erromango, Tanna,
and Anatom.
Thus, this insular hotspot supports in excess of 1 600
islands, encompassing a land area of around 99 384 km2.
Although a tiny area in global terms, the East Melane-
sian Islands Hotspot has more land than all of the re-
maining islands in the tropical Pacific, being a little
more than double the land area encompassed in the
Polynesia-Micronesia Hotspot (ca. 47 000 km2).
The geological history of these island groups is very
complex and poorly known. Most of their history seems
to be recent, these groups being of Late Tertiary emer-
gence from the sea. The major formative events, how-
ever, relate to the break-up of Gondwana, the rafting of
the Indo-Australian Plate northward, and its long con-
tact and interaction with the Pacific Plate and several
smaller plates (South Bismarck Plate, Solomon Sea
Plate). The Indo-Australian and Pacific plates are cur-
rently converging at a rate of 12 cm/yr (Honza et al.
1987). One finds the development of island arcs in
association with a single long and sinuous subduction
zone that has created a series of 7 000-m deep-sea
trenches.
In essence, there has been a substantial tectonic in-
teraction between what has been characterized as the
West Melanesia Arc and the New Guinea/Australia
craton, and it is this interaction that has apparently
produced all of the island chains and groups men-
tioned above. A close examination of a bathymetric
map shows a series of trenches and deeps that arc
from the southern verge of New Britain east and
southeastward along the southern verge of the Solo-
mons and southeastward along the western verge of
the islands of Vanuatu. In addition, an examination
of island patterns displayed in relation to this deep
water is that two or even three chains of islands are
arrayed in parallel along this Melanesian arc. It per-
haps shows up most clearly in the vicinity of New Ire-
land and Bougainville, where one can find a line of
high, massive islands, a smaller set of rugged islands
(e.g., Feni, Lihir), and finally a low set of atolls (e.g.,
Lyra Reef, Nuguria Islands, Carteret Islands); one can
also see this in the Solomons (especially the New
Georgia Group, Choiseul, Ontong Java Atoll). The
pattern is only broken by the oddball positioning of
New Britain and its affiliated islands to the west —a
product of little-understood actions of the two Bis-
marck microplates in contact with the Indo-Aus-
tralian Plate.
The islands of this hotspot have never been in land
contact with New Guinea, and the current 100-km-
long Vitiaz Strait separating New Guinea from New
Britain was formerly a much broader ocean channel.
As a result, the biotas of these islands are a mix of
long-distance immigrants and older indigenous lin-
eages derived from ancient Pacific-Gondwanan stock.
Given the uncertain above-sea histories of these
islands, one might presume that the more ancient
lineages survived by hopping from island group to is-
land group as they periodically arose, submerged, and
arose anew.
In terms of climate, the East Melanesian Islands
Hotspot is composed of small and large, low and high
islands that are uniformly tropical and humid. All re-
ceive in excess of 1 500 mm of rain per year, and the
wetter sites receive in excess of 5 000 mm per year.
The Weather Coast of the main Solomon Islands (espe-
cially on Makira and Guadalcanal) receives huge
amounts of rain, as do regions in southern New Britain
(Gasmata receives 6 m; Pomio, 6.5 m). The entire
hotspot lies south of the Equator, from 2ºS to 20ºS and,
therefore, is located wholly within the tropics. In addi-
tion, southernmost Vanuatu exhibits a cool and dry sea-
son from May to October. One major climatic feature
that varies from west to east is susceptibility to cy-
clones during the December to March season. The east-
ern Solomons and all of Vanuatu lie on the cyclone
track, and those forests most exposed have suffered the
repeated devastation of high winds. In the Solomon Is-
lands, there are tall gallery forests in Western Province,
while on Makira Island to the east, which suffers regu-
lar cyclone damage, the forest is more like a natural
mosaic of secondary successions.
This hotspot is culturally and linguistically very rich.
Vanuatu, with 109 living traditional languages, has
more languages per unit area than any nation on Earth.
The Solomon Islands, with 74 languages, is only slight-
ly less diverse. With such a large number of languages
over a small population, it is not surprising that there
are many languages with only a few hundred speakers.
Languages are dying or mixing into “Pijin-Austronesian
Creoles” in the Melanesian Islands. The disappearance
of linguistic diversity results in a rapid loss of tradition-
al knowledge and traditional practice. Typically, this
leads to erosion of traditional links between communi-
ties and the forests that have long served as their pri-
mary source of wealth and subsistence. However, a
shift to the modern cash economy is the underlying
force behind the rise in destructive exploitation of the
region’s natural environments.
Biodiversity
Mueller-Dombois and Fosberg (1998) delineated a se-
ries of natural vegetation types in the Solomon Islands,
and these typify the region as a whole: coastal strand
vegetation; mangrove forest; freshwater swamp forest;
lowland rainforest; seasonally dry forest and grassland;
and montane rainforest. Most are species-poor, with
several tree species dominating (the dominants varying
from site to site and island to island). In certain sites,
single tree species dominate, as with the monodomi-
nant stands of Campnosperma brevipetiolata and Termi-
nalia brassi in swamps (Johns 1993). Recent studies by
Bayliss-Smith et al. (2003) indicate that many “old-
growth” forests in the Solomons are secondary and
show the impacts of past human disturbance. This fact
probably applies to most or all of the lowland and hill
forests of the hotspot.
Based on the analysis of Mueller-Dombois and Fos-
berg (1998), Wikramanayake et al. (2002) summarized
the floristics of forest habitats in the various ecoregions
of the East Melanesian Islands Hotspot. The Bismarck
Archipelago is characterized by a range of typical
species and genera of forest trees that are well known
in New Guinea, and mentioned above, but certain taxa
are conspicuously absent. The genus Araucaria, an
emergent conifer of ancient Gondwanic origins, does
not occur in the hotspot, though its relative Agathis oc-
curs throughout. Nothofagus, another Gondwanan relict
so important in montane New Guinea, is found at rela-
tively low altitudes in the mountains of New Britain,
but absent on New Ireland and the Solomons, where The small Pacific boas
the highlands are locally dominated by Metrosideros (Candoia spp.) are found on a
salmomonis. The Dipterocarpaceae, which dominate in number of islands in the East
Melanesian Islands Hotspot.
Southeast Asia and are rather common in New Guinea,
© Patricio Robles Gil/Sierra Madre
are essentially absent from the hotspot. Overall, the re-
gion’s forest flora is poor compared to that of New On pp. 350-351, aerial view of
Guinea, with dominant trees being Pometia, Dillenia, undisturbed rainforest on the island
Elaeocarpus, Endospermum, Campnosperma, Calophyl- of New Britain, a rare sight on an
lum, Terminalia, Canarium, Agathis, Metrosideros, and island increasingly impacted by
logging and clearance of forest for
Sararanga. All of these genera are found on mainland
oil palm plantations.
New Guinea.
© Patricio Robles Gil/Sierra Madre
In total, the vascular plant flora of the region is esti-
mated to be 8 000 species, and of these some 3 500-
4 000 species are thought to be endemic to the region
(D. Frodin, pers. comm.). The fauna of the East
Melanesian Islands Hotspot is an attenuated sample of
that from New Guinea, plus a minor but distinctive
oceanic element. Certain groups (for example, birds of
paradise, bowerbirds, scrub-wrens, tree-kangaroos,
echidnas, gliders) that are prominent on mainland New
349
Guinea are absent from this insular hotspot. Other taxa
(honeyeaters, white-eyes, monarchs, fantails, flying
foxes and allies, and murid rodents) abound on the is-
lands. Finally, certain sylviine warbler lineages, and
certain gecko and frog lineages are more prominent on
these islands, but rare or absent on mainland New
Guinea. Because of different regimes of oceanic isola-
tion and local adaptive radiation, endemism is very
high on the islands, with some species endemic to the
hotspot, others endemic to subsets of the hotspot, and
yet others confined to single islands. The insular
and fragmented nature of species ranges also leads to
high levels of endangerment.
The hotspot supports 365 regularly occurring bird spe-
cies, 154 of which are endemic. Compositionally, the avi-
fauna is distinct from that of New Guinea, and has seven
endemic genera (Microgoura, Nesasio, Woodfordia, Gua-
dalcanaria, Stresemannia, Mayrornis, and Neolalage). The
hotspot also includes six Endemic Bird Areas (EBAs), as
defined by BirdLife International (Stattersfield et al.
1998): the Admiralty Islands, with six species confined
entirely to this EBA; St. Matthias Island (two species);
New Britain and New Ireland (35 species); the Solomon
group, with a staggering 61 species endemic; Rennell
and Bellona (five species); and Vanuatu and Temotu (15
species).
The region holds a varied mammal fauna, but can be
characterized as poor in nonvolant species and rich in
volant species (bats). Flying foxes and allies (Pteropo-
didae) and Microchiroptera occur in almost equal di-
versity, with 36 and 33 species, respectively. In total, of
the 86 mammal species native to the hotspot, 39 are
endemic. There are three endemic genera: Anthops,
with a single species, the flower-faced bat (A. ornatus,
VU); Melonycteris (three species); and Solomys (three
species).
A total of 114 species of reptiles (54 endemic) and 44
species of amphibians (38 endemic) are known to occur
in this hotspot. The region is home to six endemic gen-
era of reptiles, five of which are represented by single
species, including four species of snake —Bothrochilus
boa, Loveridgelaps elapoides, Parapistocalamus hedigeri,
and Salomonelaps par— and a lizard, Corucia zebrata.
The sixth endemic genus, Geomyersia, consists of two
species of lizards. There are also four endemic genera
of amphibians; two of the amphibian genera are mono-
typic, namely Palmatorappia solomonis (VU), a species
from the Solomon Islands that may actually represent
two species, and Ceratobatrachus guentheri, found on
the Solomon Islands and Bougainville and Buka islands.
It is perhaps not surprising that the herpetofauna of
this hotspot is rather impoverished, given its physical
isolation. However, there is interesting within-region
variation in this pattern. For example, there are 40
species of lizards from Bougainville (many endemic),
but only 26 and 30 species recorded from New Ireland
and New Britain, respectively (Allison 2001). New Ire-
land is indicative of the herpetofauna of a large island
in the region. Allison’s catalog for New Ireland includes
the following families (with species numbers in paren-
theses): Amphibia: Hylidae (2), Ranidae (5); Reptilia:
Crocodylidae (1), Agamidae (2), Gekkonidae (6), Scin-
cidae (19), Varanidae (2), Boidae (2), Pythonidae (2),
Colubridae (4), Elapidae (1), Laticaudidae (1), Typhlo-
pidae (3). Thus, it is a hotspot typified by skinks and
geckos. It is also notable that both the boas and the
pythons co-occur in this hotspot.
The hotspot supports a small but highly unique
freshwater ichthyofauna, and includes families such as
Eleotridae, Gobiidae, and Mugilidae. Until recently,
there has not been a concerted effort to catalogue the
freshwater ichthyofauna inhabiting the South Pacific.
However, conservation groups like the Wildlife Con-
servation Society, Wetlands International, and Conser-
vation International, as well as several academic insti-
tutions, are now beginning to document the region’s
freshwater biota.
It is estimated that there are approximately 52
species of freshwater fishes throughout the hotspot
and, of those, three species are considered endemic:
Stenogobius alleni, found on New Britain Island in
Papua New Guinea; S. hoesei, found throughout the Ad-
miralty and Solomon Islands and the Bismarck Archi-
pelago; and Stiphodon astilbos, found in Vanuatu. The
number of taxa decreases further inland from the ocean
and as altitude increases, with gobioid fishes likely be-
ing the only native taxonomic group observed once a
precipitous obstruction has broken the continuity of
a stream. Because the fishes inhabiting the hotspot are
of marine origin, diadromous behavior is not uncom-
mon and all the freshwater fishes are capable of toler-
ating a wide range of saltwater concentrations. These
factors have helped create the distinct fish faunal as-
semblage for this region.
The invertebrate fauna is poorly known. As with
the other life-forms, the invertebrate faunas of each On the opposite page, impressive
more distant island group are ever more attenuated tree buttresses on a rainforest giant
subsets of that from the source island of New Guinea. on the island of Espíritu Santo,
Vanuatu, part of the East
Butterflies are relatively well known among arthro-
Melanesian Islands Hotspot.
pods. The Ornithoptera (birdwing) butterflies find © Patricio Robles Gil/Sierra Madre
their center of species richness in New Guinea. The
genus ranges eastward through the Bismarcks to Above, the Solomons leaf frog
the Solomons, where three species are found (O. al- (Ceratobatrachus guentheri)
lotae, O. urvillianus, and O. victoriae). By contrast, the is the only representative of a genus
endemic to the Solomon Islands.
papilionid genus Troides (sensu stricto) is widespread
© Mike Tinsley/Auscape
in the Moluccas and New Guinea, but absent from the
East Melanesian Islands Hotspot. The blue emperor
swallowtail (Papilio ulysses), so well known in tropical
Queensland, ranges westward to the Moluccas, north-
ward into New Guinea, and is also widespread in the
Bismarcks and Solomons, but does not reach Vanuatu
(D’Abrera 1990).
353

None of the three species of
cassowary occur naturally in this
hotspot, but the dwarf cassowary
(Casuarius bennetti) occurs on
New Britain. It was presumably
introduced from New Guinea.
© Patricio Robles Gil/Sierra Madre
On the opposite page, the coconut
crab (Birgus latro), the largest
terrestrial invertebrate, is
widespread in the Pacific,
its distribution coinciding with that
of coconut palms. It is thought
that the post-larval stage of these
crabs rafted from island to
island on coconuts.
© Patricio Robles Gil/Sierra Madre
354
Flagship Species
While not necessarily a center for typical flagship
species, the East Melanesian Islands Hotspot has its
share of unusual creatures and plants, all of which mer-
it attention and conservation. The Kauri pine (Agathis
spp.) is perhaps the most logical selection as a plant
flagship. This ancient Gondwanan conifer is the largest
tree in the islands, and is an important timber species.
It grows to a huge girth, and makes for a remarkable
forest where it is common. Foresters, of course, lust af-
ter it, and it tends to be high-graded out of forests wher-
ever it is found.
Besides bird species such as the superb pitta (Pitta su-
perba, VU) of Manus Island, perhaps the most beautiful
species in the hotspot (although one that is very diffi-
cult to see), and the fearful owl (Nesasio solomonenis,
VU), a Solomon Islands endemic and the hotspot’s
largest nightbird, the most majestic avian flagship is the
Solomons sea-eagle (Haliaeetus sanfordi, VU), another
Solomons endemic. This species favors coastal forests,
but pairs also hunt further inland and, at least on the
eastern islands, they appear to have entirely inland
ranges where they prey largely on northern common
cuscus (Phalanger orientalis) and fruit bats. Among flag-
ship mammals, the most remarkable are the flying
foxes, which are known to play an important role in
pollination and seed dispersal of plants. Of the 13
threatened species of pteropid bats recorded from the
hotspot, three are Critically Endangered and poorly
known, but highlighted here to draw attention to their
plight: the Bougainville monkey-faced bat (Pteralopex
anceps) was known mainly from specimens collected in
the 1920s, until six bats were observed during a 1995
survey (and apparently no sign of this bat was found
during fieldwork on Choiseul in 1992 or on Buka in
1997); the montane monkey-faced bat (P. pulchra) is
known from only a single specimen collected on Mt.
Makarakomburu on Guadalcanal; and the Guadalcanal
monkey-faced bat (P. atrata) was last recorded in 1991,
although there has been limited survey work in its like-
ly range. Besides the flying foxes, the Admiralty cuscus
(Spilocuscus kraemeri) is the only endemic cuscus in the
hotspot, being confined to the Admiralty Islands; this
beautifully patterned brown, black, and white species is
a popular game animal on Manus.
Other flagship species in the East Melanesian Islands
Hotspot include the Solomons prehensile-tailed skink
(Corucia zebrata), a very large, arboreal skink that feeds
primarily on the leaves of epiphytes (McCoy 1980), and
the two birdwing butterflies, Ornithoptera allotae and O.
victoriae. The latter two species are large, prominent,
and spectacular, and depend upon specialized food
plants for their reproduction. In addition, the mar-
velous green tree snail of Manus Island, Papustyla pul-
cherrima, is another good flagship. It is collected and
widely appreciated throughout the region and globally.
Threats
Until the 1970s, the region supported relatively intact
lowland forests in abundance. The clearance and degra-
dation of these over the past three decades is a prime
reason why the region is now being classified as a
hotspot. Today, less than 25% of the region’s lowland
forests remain as “old growth,” primarily in the least ac-
cessible areas, especially in places where local commu-
nities have resisted the siren’s song of the foreign log-
ging companies. Upland humid forests remain in better
condition, but with population growth even these are
being reduced, primarily by clearance for subsistence
gardens.
Given the insular environment and absence of
large expanses of contiguous habitat, oceanic island
biodiversity is vulnerable to threats in many ways.
The islands of the East Melanesian Islands Hotspot
have remained pristine longer than more accessible
island groups, but over the last decade virtually all of
the negative impacts that we know for those other is-
land groups have come to the region. Industrial-scale
logging, conversion of forest for monoculture planta-
tions, exotic invasive species, and clearance of habitat
by expanding human populations have all taken their
toll on this region. In addition, it is common for vil-
lage inhabitants to engage in habitual burning of veg-
etation. These burning practices cause gradual ero-
sion of the forest frontier and lead to soil erosion,
which consequently intensifies sedimentation in
streambeds.
By comparing remote sensing analyses with a finer
resolution analysis of logging concessions and oil palm
plantations on New Britain and New Ireland, we calcu-
late that the percentage of remaining habitat for the en-
tire East Melanesian Islands is around 39%. We believe
this figure is perhaps a generous estimate, though prob-
ably close to 2003 conditions, and further it is likely
that it includes a considerable amount of secondary
forest. In light of this, and given the rapid rate of defor-
estation in this region, we feel that a reasonable esti-
mate of the extent of remaining primary habitat in the
region is around 30%.
The Bismarck Islands have been most heavily im-
pacted by extensive logging of lowland and hill forests
and also subsequent clearance of forests for copra and
oil palm monocultures near the coast. The planta-
tion impacts are mainly confined to New Britain,
whereas logging has taken place widely in the hot-
spot, wherever there is accessible forest. By contrast,
the Admiralties have been most affected by rural
agricultural expansion, though logging has had an
impact as well.
The Solomon Islands are much like the Bismarcks,
but even more vulnerable because of the smaller size
of the islands. Logging has had a devastating effect on
lowland forests, and copra plantations are also wide-
spread. Forest conversion is expected to be partic-
ularly important on Makira and New Georgia, and per-
haps elsewhere in the Solomons. The extent of im-
pact of exotic invasive species, especially pigs, cats,
rats, and little red fire ants, is serious, but poorly stud-
ied. Flannery (1995) noted the apparent extirpation
of several species of giant rodents from Guadalcanal,
apparently caused by cat predation. It is unknown
what sort of impact invasives have had in the Bis-
marcks, but it has probably been serious there as
well. An expedition to New Ireland in 1992 found in-
vasives (cane toads and feral cats) in old-growth for-
est. The invasive fire ant Wasmannia sp. invaded the
Solomon Islands in the 1980s and has reached plague
levels on many islands today. It has most recently ar-
rived in Vanuatu. In the Solomons, the area under
cultivation doubled between 1972 and 1992 (This-
tlethwait and Votaw 1992).
The situation in Vanuatu is similar. About 35% of the
land of Vanuatu supports some sort of forest cover and
logging remains an important economic activity, main-
ly on Espiritu Santo. Additionally, clearance for planta-
tions and subsistence agriculture is removing the last
remnant lowland forests of the nation. Grave pressures
on Vanuatu’s natural resources are undoubtedly being
exacerbated by human population growth. Introduced
species, uncontrolled habitual burning of the island’s
native vegetation, and deteriorating social structures
are other threats to biodiversity.
In all of the larger islands of the East Melanesian
Islands Hotspot, the rugged highland forests are large-
ly still intact and in good condition, but it remains un-
clear whether the original native faunas of these
montane forests are comparably intact, especially giv-
en the threat from invasive species such as pigs, cats,
and rats. Mining is a minor threat on the terrestrial
side, quite localized to catchment areas where a par-
ticular mine is operating (as in Bougainville and Li-
hir), but offshore impacts from tailing runoff are a
major concern.
Another broad-scale and diffuse threat is poor gov-
ernance and government instability. This leads to in-
adequate management of resources, poor deals (and
poorly managed deals) with international resource de-
velopment companies (mining and logging in particu-
lar), and social and cultural disruption. The Solomon
Islands have been the most heavily affected, suffering
a constitutional crisis in 2000. Governance troubles that
plague Papua New Guinea also impact the Bismarck
Islands. A nearly decade-long war on Bougainville was
tied to poor management of the large Panguna Copper
Mine —especially the Bougainvillean people’s percep-
tion that they were inflicted with the ill effects of the
mine, while reaping too few of the profits. The im-
plications of this little-known war will be felt in
Bougainville for decades, and its long-term negative
impact on Bougainville’s environment should not be
underestimated. With poor governance there is a typ-
ical tendency to overexploit resources as a quick solu-
tion to budgetary shortfalls, which compromises the
health and stability of forests, fisheries, and water-
sheds.
Conservation
There is little formal protection of land and sea re-
sources in the East Melanesian Islands Hotspot, main-
ly because the three island nations respect local cus-
tomary tenure and have had minimal alienation of land
or coastal marine territory for creation of government-
managed parks and protected areas.
Overall, protected areas coverage in the hotspot is
poor. According to the World Database on Protected Ar-
eas, there are 24 protected areas covering some 6% of
the hotspot, none of which are classified in IUCN cate-
gories I to IV, and only eight are included in categories
V and VI (these eight protected areas cover only 1% of
the hotspot).
What national parks are in place tend to be leftovers
from colonial times. The Queen Elizabeth II National
Park near the capital Honiara in Solomon Islands, the
nation’s only National Park, has been completely de-
graded in the 50 years since it was established in 1954
from primary forest to secondary forest and grassland.
In Vanuatu, many natural areas are protected under
custom law, but these are rarely of sufficient size for
species and habitat protection.
The fundamental constraint on any conservation
initiative in the hotspot, particularly one involving the
establishment of protected areas, is the customary
land tenure system. Unlike the neighboring Melane-
sian nation of Fiji, where customary title is formally
codified, the nations of the East Melanesian Islands
Hotspot recognize customary tenure in broad terms,
and it is generally left up to a system of land dispute
hearings to settle conflicting claims to ownership or
usage rights over land. The first basic step in under- On the opposite page, collection of
taking any conservation initiative, be it protected-area the eggs of the Melanesian
establishment or species-specific actions, is in know- scrubfowl (Megapodius eremita),
a megapode endemic to the island of
ing at that point in time who or where the land-owning
New Britain. These amazing birds
community is. lay enormous eggs in volcanic soils,
The Human Poverty Index for the islands in this and the eggs are regularly harvested
hotspot is the lowest in the Pacific region, and among by local people.
the lowest in the world (UNDP 1999). Given the lack of © Patricio Robles Gil/Sierra Madre
clearly defined legal title over land and widespread ru-
Above, man from Rabaul with
ral poverty, community-managed protection associated
newborn chick of the dusky
with community development activities is the most scrubfowl (Megapodius freycinet).
common strategy followed for establishing conserva- © Patricio Robles Gil/Sierra Madre
tion areas, but the results are mixed and highly depen-
dent on the standard of village-level engagement and
the level of community cohesiveness and collective de-
cision-making capacity. Sustainable incentives and lo-
cally meaningful motivations for rural Melanesian
357

Young girl from Yakel village,
Tanna Island, Vanuatu.
© Patricio Robles Gil/Sierra Madre
On the opposite page, wood harvest
on Espíritu Santo Island,
Vanuatu. This small country, part
of the East Melanesian Islands
Hotspot, is one of the culturally
richest corners of our planet, with
some 109 languages still spoken.
© Patricio Robles Gil/Sierra Madre
358
communities to accept a conservation regime over tane rainforest, while the Vatthe project has focused on
their communally owned resources are necessary, as is conservation of the largest tract of remaining lowland
objective selection of project sites based on national rainforest on Espiritu Santo by means of a community-
and global conservation value. based project (Read 2002). Unfortunately, the project
International and local NGOs have been, and con- was terminated in 2000 due to unrest and ethnic ten-
tinue to be, major players in conservation in the East sion in the region.
Melanesian Islands Hotspot, either in formal or infor- In general, there is very little large-scale conserva-
mal partnerships with the respective government’s tion action currently under way in the hotspot, and the
conservation offices. The regional conservation and region is in urgent need of increased attention and in-
environmental body, SPREP (South Pacific Regional vestment from the conservation community.
Environment Programme) is a multi-governmental or- The habitat protection initiatives to date have just
ganization which works closely with member govern- begun to wrestle with the difficulties of conservation
ments, including the nations of the East Melanesian on uncodified customary land in the Melanesian cul-
Islands Hotspot. tural and social context. There is still no area that
Conservation International and a number of partner could be called a secure protected area. The greatest
institutions have been active in parts of the region for successes have come where relationships are built be-
over a decade. CI fieldwork was initiated in partnership tween communities and conservation agencies over
with the East New Britain Social Action Committee time and collaboration is mutually beneficial. Without
(ENBSEK) in 1993, focusing on coastal communities in conservation agencies valuing community needs and
the Wide Bay region of East New Britain, and has re- aspirations, or the land-owning communities valuing
sulted in the designation of the Klampun Wildlife Man- their biodiversity heritage, such collaboration is not
agement Area (WMA) in 2003, with the neighboring possible. The future of the East Melanesian Island
Tiemtop WMA due for designation in 2004. CI has also Hotspot’s biodiversity lies in developing these success-
been working in the Bauro Highlands of Makira in the ful partnerships.
Solomons since 1994, in partnership with the Maruia
Trust of New Zealand and the Solomon Islands Devel- BRUCE M. BEEHLER 1
opment Trust (and most recently the Makira Commu- ROGER JAMES 30
nity Conservation Foundation), with a goal to establish TODD STEVENSON 1
a formally recognized conservation area in the Bauro GUY DUTSON 31
Highlands (60 000 ha). FRANÇ̧OIS MARTEL 32
The Nature Conservancy (TNC) has been active in
the Kimbe region of New Britain since 1994 developing
a network of locally managed marine protected areas,
promoting sustainable development, and encouraging
community-based conservation and resource manage-
ment. In 1996, a cooperative venture between TNC,
the European Union, and Walindi Plantation Resort
culminated in the establishment of the Mahonia Na
Dari Research and Conservation Center —a locally
managed non-governmental organization—, which has
focused primarily on marine environmental education
and awareness related to logging and fishing practices,
oil palm plantation development, and human popula-
tion growth.
Since 1996, the Worldwide Fund for Nature (WWF),
in partnership with the Tetepare Descendants’ Associa-
tion, has been developing the Tetepare Island Commu-
nity Resource Conservation and Development project
in Western Province, Solomon Islands. This initiative is
located on the largest uninhabited island in the South
Pacific, touted for its natural inheritance and archaeo-
logical values.
Finally, SPREP has developed a community-based
conservation project in Koromandi, on Guadalcanal in
the Solomon Islands and at Vatthe on Espiritu Santo,
Vanuatu. The Koromandi project has led to the cre-
ation of a local conservation area of lowland and mon-
 TAIWAN
Taiwan is situated at the western edge of the Asian
continental shelf. It lies south of Japan via the Ryu-
kyu Island chain, and is separated from the Asian Kyushu

mainland by a narrow strait (130 km at the nar-

rowest point). The Tropic of Cancer cuts across the
middle of Taiwan, and the warm Kuroshiwa Current JAPAN
passes along its eastern shore. As defined here, Tai- CHINA
wan includes 72 small islands, most of them with EAST
limited forest cover, except for Turtle Island CHINA
SEA

o
(2.85 km2) and Lanyu Island (45 km2), which have

t
o
h
substantial remaining tracts of natural forest, and

s
-
i
TAIWAN e
has a total land area of 36 210 km2. Although Tai- n
s
a
wan, happily, has not lost so much of its natural Taiwan
N

habitat to meet the threshold for consideration as a Strait

Iriomote-jima

hotspot, and also does not quite meet the hotspots

threshold for plant endemism, detailed assessment
of the island was necessary to demonstrate this. Lanyu
PACIFIC
Island
This chapter is, therefore, included here for the sake OCEAN
SOUTH
of documentation, and to make sure that this im- CHINA Batan
Islands
portant island is incorporated into global conserva- SEA

tion strategy.
Mountains and geographical location have com- PHILIPPINES
0 400 km
bined to make Taiwan a beautiful place, giving rise to Luzon

the well-known sixteenth-century Portuguese name,

Ilha Formosa. More than two-thirds of Taiwan’s sur-
face is mountainous, with over 200 mountain peaks
above 3 000 m in elevation, and the highest, Mt.
Yushan or Jade Mountain, reaching 3 952 m. Taiwan
is longer than it is wide, running 377 km north to
south, and 142 km west to east. The eastern side of
the island has steep slopes, narrow valleys, and small
plains, while the western side has gentler slopes and
broader plains.
Taiwan was formed around 4 million years ago,
when a small piece of the Eurasian Plate was sub-
ducted under the Philippines Plate, resulting in the
dramatic uplift of Luzon and Taiwan (Teng 1990). In-
deed, the Taiwan Strait is only 60 m at its shallowest
part at the current sea level (Yue-Gau Chen, pers.
comm.), and when sea level was 60 m below its pres-
ent level, it was connected with Fukien Province of
mainland China. A landbridge has existed at four dif-
ferent stages over the course of the last 240 000 years
(Chappell and Shackleton 1986), which has enabled
faunal exchange to take place. However, the rising of
the sea level at the end of the glacial periods rapidly
reestablished the isolation of Taiwan from the main-
land (Ferguson 1993).
The climate on the main island Taiwan is subtrop-
ical in the north and tropical in the south. Mean
monthly temperature ranges from 34.1ºC in July for
Taipei to –5ºC in January on the highest mountain.
The distribution of rainfall on Taiwan is uneven;
mean annual rainfall over the last 30 years has var-
ied between 951 mm and 4 892 mm. In winter, north-
eastern winds bring persistent drizzle to northern
and eastern Taiwan. Typhoons that come mostly in
July to September usually cause abundant rainfall
in eastern and southeastern Taiwan. Weather fronts
that stall above Taiwan every May and June bring On the opposite page,
light rain interspersed with heavy downpours to Dabajiashan Mountain. Nearly
northern Taiwan. Due to the topography, annual 70% of Taiwan is mountainous,
with more than 200 mountain
rainfall on Taiwan is higher in the north, the east,
peaks above 3 000 m in elevation.
and in the mountains than on the southwestern © Jen-Shiu Hsu
plains. Because rivers on Taiwan are short and steep,
the storage capacity of all watersheds is limited.
Rivers typically swell up soon after heavy rain, but
are then reduced to a narrow, shallow flow in a few
days. Rainstorms and frequent earthquakes exert
strong impact on the steep and unstable mountain
slopes. Small-scale natural landslides are common.
361

The Taiwan japalure
(Japalura swinhonis) is a common
lizard in Taiwan’s low-elevation
forests. This species begins
hibernating in November.
© Jen-Shiu Hsu
On the opposite page, the
Vulnerable Taiwan or Formosan
macaque (Macaca cyclopis) is a
gregarious and diurnal species,
spending its time either in trees or
on the ground.
© Jen-Shiu Hsu
362
Typhoons often cause blow-downs, create gaps within
forests, and influence the height of forests.
There are five life zones on the island: tropical rain-
forest in southern Taiwan and Lanyu, evergreen broad-
leaved forest, mixed forest, coniferous forest, and alpine
grassland. The richest and least disturbed lowland ever-
green broad-leaved forests are found in Hengchun
Peninsula at the southern tip of Taiwan, and on Lanyu.
Roughly 52.9% of the 820 species from Lanyu and 42.3%
of the 912 species from Hengchun Peninsula have trop-
ical affinities, while eastern Asian elements are relative-
ly poorly represented. Hengchun Peninsula (280 km2 in
area) supports 49 endemic species, whereas 36 species
are confined to Lanyu. In the global floristic system,
these forests belong to the Philippinean Province of the
Paleotropical Kingdom (Takhtajan 1986; Hsieh 2002),
but are included here with the rest of Taiwan for geo-
graphic proximity and political convenience.
Less than 2% of the population on Taiwan can be
considered indigenous people; the rest are Han Chinese
who immigrated to Taiwan within the last four hundred
years. Taiwan’s indigenous people belong to 10 tribes,
nine of which traditionally resided on Taiwan, while the
Yami tribe lived on Lanyu. Each of these tribes has its
own language and culture. The tribes on Taiwan have a
hunting and slash-and-burn agricultural tradition, and
their folklores contain rich stories of wildlife, hunting,
and harvest. The Yami tradition is tied to fish, sheep, and
yam and taro roots. All the indigenous people are now
heavily influenced by Chinese culture, after several
hundred years of contact, interracial marriages, and
Chinese education. The lowland plain (Pingpu) tribes
have been entirely assimilated into Chinese society. Ex-
cellent hunters with traditional knowledge of wildlife
and nature were still common among Taiwan’s tribes 30
years ago, but modernization and economic incentives
have further diluted these traditional cultures. Only the
Yami people still commonly practice their traditional
knowledge of the sea, although there is a widespread
movement for all the indigenous peoples to re-identify
with their own traditions.
Biodiversity
Taiwan has high levels of endemism in both plants and
animals. About 26% of vascular plants, 25% of mam-
mals, 10% of resident birds, 25% of non-marine rep-
tiles, and 33% of amphibians are endemic. In addition,
a number of Taiwan’s plants, amphibians, and freshwa-
ter fishes are relict species, meaning that Taiwan has
served as a refugium for ancestors of these species dur-
ing glacial periods. Upon the retreat of the ice sheets,
the distribution of their relatives shifted north or up
into the Himalayan Mountains, leaving them isolated
on Taiwan. These populations survived in the moun-
tains of Taiwan, and eventually diverged from their an-
cestors, becoming unique endemic species.
Taiwan has 4 101 species of native vascular plants be-
longing to 1 331 genera and 232 families. Around 1 071
species are known only from Taiwan, and for this reason
it has been considered a Center of Plant Diversity (Davis
et al. 1995). Two plant genera are endemic (Sinopanax
and Kudoacanthus), and endemic species exist in 57.5%
of the families. The ferns (at least 645 species) and or-
chids (337 species) are especially species-rich.
Plant diversity is higher in the more heavily impact-
ed lowlands (below 500 m in elevation) than at upper
elevations, although the proportion of endemism in-
creases with elevation: 17.4% in the lowlands to 59.9%
above 3 000 m. Evergreen broad-leaved forests between
500 and 1 500 m, which for the most part are subject to
frequent human disturbance, contain at least 1 688
species of plants, including 419 species endemic to Tai-
wan. Between 1 500 and 2 500 m, there is mixed forest
in which broad-leaved trees gradually give way to
conifers. These forests boast an abundance of mosses,
epiphytes, lichens, and tree climbers. Conspicuous
emergent conifers tower above the largely evergreen
broad-leaved canopy. Of the 1 324 species of vascular
plants in these forests, 34.6% are endemic. Coniferous
forests occur above 2 500 m and are relatively well pro-
tected, with more than 80% still pristine; of the 539
species of vascular plants occurring in these forests,
277 are endemic to Taiwan. The other species are the
typical boreal-alpine species of Eurasia, distributed dis-
continuously in the Arctic north and the alpine zone in
the south. Alpine grassland appears above 3 500 m. The
dominant species in this habitat is the dwarf bamboo
(Yushania niitakayamensis), which is often less than
one meter tall on exposed areas. This species has a
broad elevation range (ca. 1 400-3 600 m), but attains
maximum development along high mountain ridges
and slopes. This habitat is an important center of en-
demism, with 124 endemic plant species.
Currently, 79 species of mammals in 53 genera are
known to occur in Taiwan (Liang-kung Lin, pers.
comm.). Although no family or genus is endemic to
Taiwan, around 20 species are, including seven bats
and one carnivore, the Taiwan weasel (Mustela for-
mosana). In addition, the Formosan macaque (Macaca
cyclopis, VU) is endemic, although it has since been
introduced to Japan, where it hybridizes with the
Japanese macaque (M. fuscata). Most endemic mam-
mals can be seen in more than one life zone, with the
exception of five species of bats (two restricted to
evergreen broad-leaved forest, two in mixed forest,
and one in coniferous forest), one rodent, Coxing’s
white-bellied rat (Niviventer coxinga) in the mixed for-
est, and the Taiwan vole (Volemys kikuchii, VU) in the
evergreen broad-leaved forest. The mixed forest con-
tains the highest number of endemic mammals (14
species), whereas the coniferous forest and the ever-
green broad-leaved forest each support nine en-
demics. The alpine grassland has six endemic mam-
mals, while the only one found in the lowlands is the

The Endangered Asian yellow pond
turtle (Mauremys mutica) is
widespread in southern China,
Vietnam, Taiwan, Hainan, and
parts of Japan’s Ryukyu Islands.
This Endangered species is thought
to be in decline throughout much of
its range, mainly as a result of
habitat loss and collection for the
food trade.
© Jen-Shiu Hsu
364
Formosan or Taiwan macaque, which is distributed
widely on Taiwan.
Taiwan’s native avifauna is characterized by 147 resi-
dent species, 15 summer breeders, and five suspected or
occasional summer breeders. Fifteen species are en-
demic and, as such, BirdLife International has designat-
ed the island of Taiwan as an Endemic Bird Area (Stat-
tersfield et al. 1998). Most species are found widely
within their preferred altitudinal ranges, although above
the tree line (3 500 m) only five species are found. The
evergreen broad-leaved forests and mixed forests are of
crucial importance to the maintenance of avian biodi-
versity in Taiwan, with the former supporting 58 species,
and some 50 species of birds inhabiting the mixed forests
between 1 500 and 2 500 m. Besides being exceptionally
important for endemic birds, Taiwan lies on the main fly-
way for birds migrating between Japan and the Philip-
pines, and is, therefore, of great importance for migrants
from Japan, mainland China, and areas further north.
Hence, in addition to the species already noted, there are
169 non-breeding migrants, 105 vagrants, and five spe-
cies with unclear status, making a total of 446 species of
birds that have been recorded from Taiwan.
Taiwan’s reptile fauna includes 33 lizards, 45 snakes,
and five turtles, of which 20 species (13 lizards and
seven snakes) are endemic. Most snakes and lizards are
found below 1 500 m, with only seven species of
snakes and three species of lizards above 2 000 m. As
regards amphibians, Taiwan is also home to 30 species
of frogs and three salamander species, of which 13 frog
species and all three salamander species are confined
entirely to Taiwan. The family Ranidae is represented
by 13 species, and four species are endemic; in con-
trast, seven out of 10 species in the family Rhacophori-
dae are endemic. More than 80% of the frogs and toads
have an upper distributional limit of 1 500 m, and only
four species reach 2 500 m; a single species of frog and
three species of salamander are found at 3 000 m or
higher. All the endemic species are at least locally
abundant.
In terms of other animals, Taiwan’s freshwater fish
fauna is characterized by 227 species in 150 genera, of
which 37 species are endemic. The landlocked salmon
(Oncorhynchus masou) in central Taiwan is another
relict species and has the southernmost distribution of
the world’s salmon species. Taiwan is also very rich in
invertebrates. Butterflies are represented by 384
species, of which 12.5% are endemic (Yu-feng Hsu,
pers. comm.), and there are 275 species of land snails
on Taiwan, of which about 70% are endemic (Wen-lung
Wu, pers. comm.).
Flagship Species
Taiwan’s two endemic cypress species (Chamaecyparis
spp.) are very important flagship species. The seven
species of cypress that form ancient forest types now
grow only on the west and east coasts of North Ameri-
ca and in the mountainous regions of Japan and Tai-
wan. The two Taiwanese species found between 1 800
and 2 400 m are the only ones in the subtropical re-
gion. On the best sites, they can reach 50 m in height
and more than 5 m in diameter, and live 3 000 years or
more. Logging between 1900 and 1980 greatly reduced
the virgin stands of these forests. Currently, about
48 500 ha of these forests remain. Taiwan cypress trees
often form mixed stands with other conifers and broad-
leaved species.
Another important plant flagship is Taiwania (Taiwa-
nia cryptomerioides, VU), like Metasequoia and Sequo-
iadendron, one of the world’s classic Tertiary relict gym-
nosperms. Recently, what is probably the world’s
largest population of Taiwania was discovered in south-
ern Taiwan (Yang and Wang 2002). In an area of about
1 300 ha, at least 10 000 mature trees stand shoulder to
shoulder, some of them measuring 60 to 70 m in height,
with trunks more than four meters in diameter. The
Taiwan beech (Fagus hayatae, VU) has a relict distribu-
tion in northern Taiwan and forms one of the rare de-
ciduous forests in Taiwan. It is found in a nearly pure
stand of at least 1 300 ha, located between 1 300 and
2 000 m along the northern ridge of Hsuehshan Range,
with evergreen oak forest below and mixed coniferous
forest above. Strong northeasterly winds in winter stunt
these trees to less than five meters on the summit, al-
though they grow to over 10 m on the adjacent leeward
slopes.
The clouded leopard (Neofelis nebulosa, VU) used to
be found in eastern and southern Taiwan in forests
above 1 000 m, but the last confirmed sighting of this
species was in 1983. Given its extremely secretive
habits and the rugged topography of Taiwan, conserva-
tionists had hoped that some individuals might survive
in deep forests. Unfortunately, recent intensive efforts
aided by automatic infrared photography turned up no
sign of this species and it probably no longer occurs on
Taiwan. The Asiatic black bear (Ursus thibetanus, VU) is
the largest mammal on Taiwan. It lives in the moun-
tains between 1 500 and 3 500 m, foraging mostly on
plant matter mixed with wasp nests and animal car-
casses. The current population is estimated to be be-
tween 200 and 1 000 individuals.
The two endemic pheasants are also obvious flagship
species. The Mikado pheasant (Syrmaticus mikado) was
first discovered when two male tail feathers were found
on an aboriginal headdress. These feathers were so
unique as to form the basis of a description of a new
species named in 1906. The plumage of the Mikado
pheasant is elegant and understated, whereas Swin-
hoe’s pheasant (Lophura swinhoii) is flashy in its color
pattern. The Mikado pheasant lives in forests above
2 000 m on steep slopes, while Swinhoe’s pheasant oc-
curs in lowland forests up to 2 200 m on gentler slopes.
They are both secretive, producing almost no vocaliza-
tion, and are extremely wary of people. They only
come into the open at dawn or dusk or when it is foggy
or raining. Once the sun is out, they withdraw into the
shade of deep forests.
The Taiwan flamecrest (Regulus goodfellowi) is another
endemic bird that was characterized by W.R. Ogilvie-
Grant in 1907 as being more colorful than others in its
genus. In fact, this flamecrest is a relict species once clas-
sified as a subspecies of the European flamecrest (R. ig-
nicapillus). Given that the genus has not yet been record-
ed between Asia Minor and Taiwan, it has been cited as
an excellent example of disjunct species distribution in
birds (Hachisuka and Udagawa 1951).
A few migratory bird species are also important flag-
ships. Indeed, one such species, the black-faced spoon-
bill (Platalea minor, EN), is the best-known bird in Tai-
wan. The species is found only in eastern Asia, and
totals a little more than 1 000 individuals. Two-thirds of
the world population of this species winter in one con-
centrated flock at the Tsengwen Estuary in southern Tai-
wan. Since 1995, international and local conservation
efforts have put this species in newspaper headlines nu-
merous times. Hundreds of tourists go to watch them
rest or bathe in shallow water each day during the time
they are present on Taiwan. A conservation action plan
was formulated for the black-faced spoonbill in 1995, to
be updated in 2003 (Severinghaus et al. 1995).
The grey-faced buzzard eagle (Butastur indicus) is an-
other important migratory species. This bird breeds in
the temperate region and goes south to winter. The
southern tip of Taiwan is a point of concentration for
migratory raptors, and the peak of this species’ south-
ward migration often falls around October 10; conse-
quently, the grey-faced buzzard eagle has been nick-
named the National Day bird. Each year during the
National Day holiday, huge crowds arrive to observe
the National Day bird as it gathers in flocks of thou-
sands. Every time a flock is seen coming to roost at
dusk or departing at dawn, it draws excitement from
the enthusiastic crowd. Many people become bird
watchers subsequent to such an experience.
Among the invertebrates, a beautiful birdwing butter-
fly (Troides magellanus) is an important flagship. Found
only on Lanyu, it has been reduced by overtrapping for
the butterfly trade. Its larvae feed on only one species of
plant (Aristolochia kankaoensis), and grazing by cattle
and other human activities have reduced the abundance
of this key food plant. Now with cattle mostly removed
from Lanyu, planting of A. kankaoensis seedlings and re-
leasing captive-raised butterflies into the wild appear to
have increased the butterfly population, but future mon-
itoring is needed before success can be claimed.
Threats
Habitat loss is the greatest threat to biodiversity on
Taiwan. Being a small island with a high population
density (622 persons per km2 in 2002; Population Ref-
erence Bureau 2003), Taiwan is under intense devel-
opment pressure. Most of the natural habitat below
500 m has been converted to human use, and increas-
ing development pressure and economic aspirations
have already begun to alter habitat at higher eleva-
tions. Improper land-use practices have led to habitat
loss upstream and habitat degradation downstream,
and have accelerated water and soil erosion. Channel-
ization of rivers or the construction of check dams as
measures to control flooding has destroyed riparian
habitat, and often substituted natural river conditions
with cemented river banks or even stream beds which
are foreign and hostile environments for aquatic or-
ganisms. Nevertheless, even under high population
and development pressure, 59% of Taiwan is still
forested, and 73% of the forests are natural forests
(more than 35% are mature forests), suggesting that
some 43% of Taiwan’s original forests remain intact.
However, much of the remaining intact vegetation is
at higher altitudes and, in fact, less than 1% of the
original lowland vegetation is estimated to remain on
Taiwan (Editorial Committee of the Flora of Taiwan
1993-2002).
There was a long history of traditional hunting, trap-
ping, logging, and collecting of forest products on Tai-
wan, but the government began regulating the harvest-
ing of wildlife in the late 1970s. Although illegal hunting
still occurs, public education efforts have reduced the
general demand for wildlife. Logging of primary forest
and other ecologically important forests was stopped in
1991. However, certain plants continue to be harvested
without sufficient control, especially those that are rare,
have showy flowers or foliage, or have supposed medic-
inal qualities.
Invasive species are an issue in Taiwan, with sev-
eral having entered through the pet trade, for agri-
cultural purposes, for private collections or by acci-
dent. Not enough data exists about Taiwan’s natural
conditions prior to the arrival of most of these
species, or about their point of entry and subsequent
expansion, and in most cases they were not noticed
until ecological or commercial damage was already The Nantou flying frog
serious. Examples include the freshwater apple snail (Rhacophorus moltrechti) is a
(Pomacea canaliculata), armored catfish (Hypostomus widely distributed and common
species occurring in hilly areas
spp.), bullfrog (Rana catesbeiana), red-eared slider
below 2 500 m. It lives in forests,
turtle (Trachemys scripta), a crawling plant (Mikania orchards, and tea plantations, and
micrantha), and water hyacinth (Eichhornia cras- breeds in still-water habitats such
sipes). Among the few well-documented examples as ponds, pools, cisterns, and
of the impact of an invasive are hybridizations be- blocked roadside ditches.
tween the endemic Taiwan bulbul (Pycnonotus tai- © Jen-Shiu Hsu
vanus, VU) and the Chinese bulbuls (P. sinensis) or
between the Taiwan hwamei (Garrulax canorus taewa-
nus) and the Chinese hwamei (G. c. canorus), both of
which are reducing the genetic distinctiveness of the
Taiwan endemics such that genetically pure popula-
tions remain in only a few isolated parts of their for-
mer range (Severinghaus and Chi 1999; BirdLife In-
ternational 2003; Liu 2003).
365

The retiring Swinhoe’s pheasant
(Lophura swinhoii) is confined to
the mountains of central Taiwan.
Hunting pressure is no longer
serious, and this species is now
considered only Near Threatened.
© Rod Williams/naturepl.com
On the opposite page,
the Formosan sambar
(Cervus unicolor swinhoei)
is the largest native herbivore in
Taiwan and is an endemic
subspecies to Taiwan that is usually
found in virgin forests at elevations
of 300-1 500 m.
© Sz-Yi Liu
366
Conservation vation actions, including artificial propagation of the
landlocked salmon and the birdwing butterfly. After 20
Many laws provide the legal basis for nature and biodi- years, salmon restoration work remains far from com-
versity protection in Taiwan. The Cultural Property Pro- plete. Inappropriate land use along streams has proven
tection Act covers natural heritage such as important very difficult to change. More research is needed to
endemic species or special landscapes (1982). The Na- identify the exact water and river conditions required
tional Park Law (1983), the Wildlife Conservation Act by the species.
(1989), the Forestry Law (1985), and other laws togeth- Many communities have organized and implemented
er regulate land use, human behavior, and the use of biodiversity conservation programs for their own areas,
biodiversity. In 1994, Taiwan passed the Environmental including issuing community-based fishing permits to
Impact Assessment Act, which requires major develop- earn income, patrolling streams to prevent overfishing,
ment projects to carry out environmental impact eval- and establishing fishing bans to allow stocks to recover.
uations before development permits are granted. These programs have proven highly successful for a
About 15% of Taiwan has some level of protection, number of streams. However, although the benefits of
a figure that drops just slightly when one considers such actions are enormous, the question of whether or
only protected areas classed in IUCN categories I to not privatization of biodiversity resources is consistent
IV. Taiwan’s protected areas include five national with social justice, and to what extent natural resources
parks, 15 nature preserves, 19 wildlife refuges, 29 im- (e.g., water in the river) should be privatized, requires
portant wildlife habitats, and nine national forest pro- further analysis.
tected areas. These areas are set aside to protect rela- Finally, given that a large amount of habitat on Tai-
tively large tracts of natural habitat and the wildlife wan was damaged in the past, it is now important to
within them or to preserve special landscapes. Never- look into restoration, to carry out research to bring back
theless, an increased level of protection is required for critical areas, and to develop plans to return degraded
the lower-altitude forests. areas to ecologically healthy conditions.
Nature education programs are numerous in the
cities. These are either sponsored by government agen- LUCIA LIU SEVERINGHAUS 62
cies, schools or other institutions, or organized by envi- CHANG-FU HSIEH 63
ronmental NGOs with funding from government agen-
cies or the private sector. Researchers are often invited
to give public lectures to introduce their new findings.
Media exposure helps to draw public attention to charis-
matic species or special habitats such as mangroves
with viviparous plants (Kandelia obovata). The story of
the landlocked salmon has been included in high
school textbooks. Satellite tracking of the green sea tur-
tle (Chelonia mydas, EN), which comes on shore to lay
eggs, has greatly increased its prominence with the
general public. News of the recovery of migratory Tai-
wan butterflies (Parantica sita niphonica) in Japan cata-
pulted the already popular sport of butterfly watching
into organized public monitoring activities. With the ex-
ception of the highly popular bird-watching and butter-
fly-watching activities, most existing education pro-
grams focus on flagship species. Future education
needs to increase people’s understanding of ecological
functioning and the significance of all species, includ-
ing the large number of non-charismatic ones.
Biodiversity inventories have been established by
government agencies and academic institutions. The
usefulness and quality of these inventories can only be
strengthened with continued monitoring of species
distribution, population sizes, and ecological interac-
tions, as well as an increased understanding of their
genetic diversity. The critical nature of this task is re-
flected by the fact that a large proportion of Taiwan’s
terrestrial vertebrates have never been studied and
their natural history remains largely unknown. Gov-
ernment agencies have implemented various conser-
 QUEENSLAND WET
TROPICS
At the base of Cape York Peninsula in northeastern
Australia, guarded by the coral ramparts of the Great
Barrier Reef, is an outpost of equatorial splendor in
an otherwise vast, brown land. As its name implies,
the Queensland Wet Tropics is located in the State of
Queensland, abutting the coastline for over 400 km,
between south latitudes 15º40’ and 19º15’, and vary-
ing in width from 20 to 80 km. This region of extra-
ordinary diversity also includes a number of moun-
tainous offshore islands, the largest of which is
Hinchinbrook Island (399 km2). Altogether, the rain-
forests and associated forests and woodlands of the
Queensland Wet Tropics form an assemblage of eco-
logical communities spanning 18 487 km2.
Today, the rainforests of Australia represent almost
an alien presence in a land dominated by eucalypts
and acacias. Occupying less than 0.2% of the land
area of the continent, and about 1% of Queensland,
rainforests are indeed a rarity, but this was not al-
ways the case. When flowering plants first appeared
on Earth, Australia was part of the supercontinent of
Gondwana. At the final breakup of this continent,
about 50 million years ago, Australia took with it a
significant component of the original flora of Gond-
wana. However, despite the fact that broadleaf rain-
forest originally covered much of the continent,
from that time until the present —with many short-
term fluctuations—, Australia has become increas-
ingly drier and today is a largely arid continent.
Out of the ancestral Gondwanan stock arose sclero-
phyllous (thick-leaved) plants adapted to drier cli-
mates, increasingly impoverished soils, and a land-
scape in which fire became a prominent evolutionary
force. It is believed by some that these plants arose
from rainforest progenitors under selection pres-
sure from harsh new environments. It is more likely,
however, that sclerophyllous and rainforest species
co-existed from the earliest days of angiosperm histo-
ry, their proportions being modified as the environ-
ment changed (White 1986). As the Australian Conti-
nent drifted northwards into warmer latitudes,
compensating for a general global cooling, the tropi-
cal flora of Gondwana was preserved in Australia to a
greater extent than anywhere else. Its long isolation
as the island continent finally ended about 15 mil-
lion years ago, when the Australian continental plate
PAPUA
NEW GUINEA
CORAL
SEA
GR
EA T
AUSTRALIA
BA
QUEENSLAND
RR
IE
R
RE
EF
0 300 km
collided with that of Asia, and some interchange of
flora and fauna that had been evolving separately for
over 80 million years occurred.
Today, the Queensland Wet Tropics retains a unique
record of these major events. Within its deep gorges
and on its mountain tops survives an unparalleled
collection of flowering plants with primitive charac-
ters unchanged since the beginnings of angiosperm
evolution. As a result of this unique geological and cli-
matic history, combined with a present-day environ-
ment of great physical diversity, the Queensland Wet
Tropics harbors a vastly disproportionate share of the
biodiversity of Australia, which itself is recognized as
a megadiversity country (Mittermeier et al. 1997).
The Queensland Wet Tropics, as defined here, con-
forms to the Wet Tropics Bioregion (Goosem et al.
1999) and is part of a larger area identified by WWF
as the Queensland Tropical Forests Ecoregion, which
extends south of the Wet Tropics to incorporate the On the opposite page, fan
Central Queensland rainforests. However, the Cen- palm (Licuala ramsayi)
tral Queensland rainforests are separated by a gap of from Tam O’Shanter State Forest.
This beautiful palm occurs along
200 km from the Wet Tropics and enjoy a signifi-
river banks and swamps of the
cantly drier climate (a maximum of 2 000 mm per northeast Queensland rainforest.
annum against a maximum of 4 000 mm per annum © Patricio Robles Gil/Sierra Madre
in the Wet Tropics). Accordingly, we have not in-
cluded the Central Queensland rainforests in the Wet
Tropics for the purposes of this analysis.
The features that set the Wet Tropics apart as a
unique region in the Australian context are related
to its physiography and high rainfall, and also its
369

The tooth-billed catbird
(Ailuroedus dentirostris) is
endemic to higher-elevation
rainforest of the Atherton region in
northeast Queensland.
© Patricio Robles Gil/Sierra Madre
On the opposite page, the southern
cassowary (Casuarius casuarius)
is a Vulnerable species which is
being severely impacted by
development in Australia, and often
suffers from road fatalities.
© Patricio Robles Gil/Sierra Madre
370
situation well within tropical climes. The physiography
of the area is dominated by a north-south spine of high
mountains cut by eight major water courses which
have incised gorges or deep valleys. From north to
south these include the rivers of Daintree, Mossman,
Barron, Mulgrave, Russell, Johnstone, Tully, and Her-
bert. While other parts of eastern tropical Australia
have a rainfall high enough to support well-developed
rainforest, the potential of mountains to strip moisture
from the onshore stream by orographic uplift reaches
its maximum in the Wet Tropics. This process is one of
the four rain-generating mechanisms that influence
tropical regions, including convergence, convection,
and cyclonic phenomena, and which here function
synergistically as they do in few, if any other rain-
forested regions of the world. Its rainfall regime, there-
fore, can not be directly compared to any of them, oc-
cupying the extreme wet end of the spectrum (Bonell
et al. 1991). Unlike that of equatorial regions, however,
it is strongly seasonal, with a marked concentration of
rainfall in the summer months, over 60% of annual
rainfall falling between December and March. The re-
gion is noted for reporting some of the world’s most in-
tense rainfall events, as well as the heaviest individual
events. Rainfalls of up to 1 140 mm in a 24-hour period
have been reported.
The Wet Tropics Bioregion was originally defined “as
the limits of rainforest in the wet tropics, the western
boundary approximating the 1 500 mm rainfall isohyet”
(Stanton and Morgan 1976). In spite of the emphasis on
rainforest in the definition, only 8 816 km2 of its
18 487 km2 is actual rainforest (Wilson et al. 2002) and
it is estimated that the total area of rainforest never ex-
ceeded 10 885 km2. The relationship between the rain-
forest and the vegetation communities surrounding it
is a dramatic one. Rainforest expands under suitable
climatic and soil conditions into adjacent communities
if its scattered invaders can consolidate to change the
microclimate and ground cover before they are de-
stroyed by fire. In the presence of regular fire its
boundaries are sharp, and there are few more striking
contrasts in the Australian environment than the
change along a straight-line boundary from eucalypt
forest to rainforest.
Apart from the frequency of fire, limits to the ex-
pansion of rainforest within the bioregion are largely set
by the depth and drainage of the soil, as well as expo-
sure on steep slopes. The biodiversity of the region re-
sides in its complex interweaving of habitats as diverse
as rainforest, sclerophyll shrubland, sclerophyll wood-
land, tall sclerophyll forest, and melaleuca-dominated
woodlands and swamp forests. Indeed, and particularly
on the coastal plain, where vegetation type is deter-
mined by small changes in soil depth down to a perma-
nent water table, the long-term survival of a wide range
of communities is dependent on preservation of the
landscape and vegetation complexes in which they occur.
A traveler on the highway north from the dry tropics
city of Townsville may see the southern margins of the
Wet Tropics as a line of mountains to the west of the
road, their flanks covered by eucalypt woodland, with
pockets of rainforest in “fire shadows” created by rocky
valleys. However, it is easy to overlook the low wood-
lands of melaleuca and eucalyptus that stand at the por-
tals of one of the most biologically rich and scenically
splendid parts of the continent.
Past this point, one enters the wide valley of the
Herbert River. The ranges now sweep inland to a far-
distant blue scarp rimming the floodplain. Woodlands
that once filled the valley have been replaced by an
endless sea of sugarcane. Gone, too, are most of the pic-
turesque grassy woodlands, lily-covered lagoons and
swamps, and riverine rainforest which, 120 years ago,
graced this widest and most complex floodplain of all
the region’s rivers. Beyond the floodplain, the glory
of the river remains, however, as it cuts a wide gorge
through the region. Tributary streams that have cut
their headwater course from the Herbert River now
provide, at their knick points, some of the highest wa-
terfalls in Australia.
The Herbert River works a 10-km-wide gap in the
north-south distribution of continuous rainforest within
the Wet Tropics. North of the river, rainforest continues
unbroken to the northern end of the region, although
shrinking to a tenuous, narrow band just north of the
city of Cairns. South of the river, rainforest is confined
to two distinct blocks separated by a gap of 30 km in
which it retains only a foothold as scattered patches on
steep slopes and in deep valleys protected from fire.
Between the Herbert River and the Tully River, the
next major river as one heads north, is the largest sur-
viving remnant of the sclerophyll communities of the
coastal sand plain. It is a complex mix of paperbark
(Melaleuca spp.) swamp forests and woodlands, euca-
lypt forests on dunes, and patches of rainforest domi-
nated by the fan palm Licuala ramsayi. As one ap-
proaches the ancestral and current floodplain of the
Tully River, rainfall doubles within a few kilometers
and the “super-wet” belt, with annual average rainfalls
of 4 000 mm or more, is reached. This “super-wet” belt
continues to within 30 km of the city of Cairns, coin-
ciding with an extremely sharp drop in annual rainfall.
Until the 1960s, when massive clearing for a pastoral
development scheme took place, the plains on either
side of the lower Tully River supported a complex veg-
etation mosaic of grassland, sedge swamps, swamp
forests of melaleuca and palm, and eucalypt forests and
woodlands. These are now almost entirely gone and sur-
vive, elsewhere on the coastal plain, largely as dis-
membered fragments of the original mosaic.
The most complex and best-developed rainforests
clothe the deeper soils of the coastal plain footslopes
from the Tully River to the Mulgrave River and occupy
basalt slopes and plateaus in a wide transect of the region
following the catchment of the Johnstone River. More
than half of these forests on basalts have been cleared,
and the largest part of what remains has been se-
lectively logged. However, an outstanding example of

On pp. 372-373, Zillie Falls
in the Atherton Tablelands, one of
many rainforest waterfalls in the
Queensland Wet Tropics.
© Günter Ziesler
Above, a male regent bowerbird
(Sericulus chrysocephalus),
a stunning bird of the
Australian east coast.
© Günter Ziesler
On the opposite page, Victoria’s
riflebird (Ptiloris victoriae)
is one of 11 bird species endemic to
the Wet Tropics region. This bird of
paradise is endemic to the Atherton
Region, where it can be seen
throughout the year.
© Patricio Robles Gil/Sierra Madre
374
virgin complex mesophyll vine forest survives in the
valley of the Russell River, while the rarest type of rain-
forest on basalt survives only as small, scattered rem-
nants on the western side of Atherton Tablelands. Some
of the most extensive remaining tall wet sclerophyll
forests in the Wet Tropics, dominated by Eucalyptus
grandis and E. resinifera, cover the slopes of the Great
Dividing Range west of Atherton, and to the northeast
of it in the Tinaroo Range.
To the east of Cairns, there are communities devel-
oped on wind-formed dunes that are unique to the Wet
Tropics. These are covered with closed sclerophyll
shrublands in which the myrtaceous plant Thryptomene
oligandra is a common canopy species. In depressions
amongst the dunes, extensive open sedge and melaleuca-
dominated swamp communities are formed. The insec-
tivorous pitcher plant (Nepenthes mirabilis) is common
in some of these swamps, which along with Wyvuri
Swamp and adjacent areas to the south, are the only lo-
calities within the region from which it is known.
To the north of Cairns, rainforest shrinks to a narrow
band at the Black Mountain corridor before expanding
northwards into the great wilderness of the mountains
of the Mossman and Daintree river catchments. The
headwaters of the Daintree River are lightly impacted
by human activity. North of the river, Thornton Peak
(1 374 m) dominates the scenery and forms, with the
headwaters of the Bloomfield River, another area of
rugged wilderness almost entirely dominated by rain-
forest. This northern wilderness outpost of the Wet
Tropics provides refuge for many habitat types that
elsewhere have been drastically impacted by humans.
These include complex mesophyll vine forests and tall
wet sclerophyll forests in the headwaters of the Dain-
tree River. This is the only area left where mosaics of
intact lowland forest habitats still have direct connec-
tion with communities of the high mountain peaks.
The area between Cooktown and Cardwell contains
the only existing Australian Aboriginal rainforest cul-
ture. The oral prehistory of the surviving Aboriginal
rainforest culture is the oldest known for any indige-
nous people without a written language (Bottoms 2000).
Aboriginal occupation of the Wet Tropics of Queensland
is thought to date back at least 40 000 years (Sluiter and
Kershaw 1982), and the tribes of the area are consid-
ered to be among the oldest rainforest cultures in the
world (Dasett 1987). Rainforest culture differs marked-
ly from that of most other Australian Aboriginal tribes,
with a heavy dependence on arboreal skills, everyday
use of toxic plants, and unique weapons (Horsfall 1984).
Biodiversity
The Queensland Wet Tropics contains 3 181 vascular
plant species in 224 families representing approximate-
ly 18% of Australia’s vascular flora. Of this total, 576
species and 44 genera are endemic. There are also two
endemic plant families (Austrobaileyaceae and Idio-
spermaceae), the former containing species with pollen
features consistent with primitive flowering plants;
both plant families are represented by single species,
Austrobaileya scandens and Idiospermum australe. Of a
total of 19 of these primitive rainforest angiosperm fam-
ilies in the world, 13 are found in the Wet Tropics.
Vertebrate diversity and endemism are also very
high, with 107 mammal species found in the Wet Trop-
ics, including 11 endemic species and two monotypic
endemic genera, the musky rat-kangaroo (Hypsiprym-
nodon moschatus) and the lemuroid possum (Hemibeli-
deus lemuroides). The mammal endemics include four
ringtail possums, confined to altitudes above 300 m,
along with a native rodent, the Thornton Peak melo-
mys (Melomys hadrourus). Two endemic tree-kangaroo
species, Bennett’s (Dendrolagus bennettianus) and Lum-
holtz’s (D. lumholtzi), are found at all altitudes, although
the latter is rarely encountered in the lowlands. Sever-
al mammal species that extend into New Guinea reach
their southern limits in the Wet Tropics, including the
attractive striped possum (Dactylopsila trivirgata) and
long-tailed pygmy possum (Cercartetus caudatus). At the
same time, other species reach their northern limit in
the Wet Tropics, especially in the open forests and wood-
lands. Two species of mammals, the arboreal yellow-
bellied glider (Petaurus australis) and swamp rat (Rattus
lutreolus), are confined in the Wet Tropics to the tall
open forests adjoining the western edge of the rain-
forests.
In terms of avifauna, there are 368 bird species, of
which 11 species are endemic. The golden bowerbird
(Prionodura newtoniana) is a monotypic endemic genus
restricted to the higher-altitude rainforest. The Queens-
land Wet Tropics as we have defined it overlaps to a
large degree with the Endemic Bird Area of the same
name identified by BirdLife International.
Australia is renowned for its reptile diversity and this
is evident in the Wet Tropics, which contains 113 reptile
species of which 24 species are endemic. The small,
brown, leaf litter skinks of the genera Saproscincus and
Lampropholis and the slightly larger Glaphyromorphus
are particularly well represented, with four, two, and
two endemic species, respectively. There are three en-
demic reptile genera, and all are represented by single
species: the chameleon gecko (Carphodactylus laevis)
and prickly rainforest skink (Gnypetoscincus queenslan-
diae) are located in moist rainforest, mostly above
300 m, while the Mount Bartle Frere skink (Bartleia ji-
gurru) is found only in boulders and wind-swept vegeta-
tion of the highest peak in the region, Mt. Bartle Frere.
The diversity of amphibians (51 species, including 22
endemic species) is also significant, with all Australian
families represented. However, there are no genera re-
stricted to the area. The family Microhylidae and the
rainforest stream frogs of the family Hylidae such as
the Australian lace-lid (Nyctimystes dayi, EN) and the wa-
terfall frogs and mistfrogs (Litoria nannotis, EN), although
having their greatest Australian diversity in the Wet Trop-
ics, exhibit higher species diversity in New Guinea.
 The freshwater fish fauna is characterized by 51 na-
tive species including seven endemics. Furthermore,
there are two endemic genera, both monotypic, and re-
stricted to clear, fast-flowing, perennial streams. The
Cairns rainbowfish (Cairnsichthys rhombosomoides) has
a narrow distribution from Cairns to Tully. There are no
records of the species on the coastal plains, which have
been developed for sugarcane. The recently discovered
Bloomfield River cod (Guyu wujalwujalensis) has a lim-
ited distribution on the Bloomfield River.
Flagship Species
The southern cassowary (Casuarius casuarius, VU) is
usually regarded as the region’s flagship species, al-
though it also ranges into northern Cape York and
southern New Guinea. With an average weight of 60 kg,
the southern cassowary is one of the world’s largest
birds and Australia’s largest land animal; its eggs are the
third largest of any bird species. The southern cas-
sowary can live to 50 years of age, and is an important
seed disperser, helping to spread the seeds of as many
as 150 species of trees and shrubs. The southern cas-
sowary is still regularly encountered throughout the
Wet Tropics, although populations are only estimated to
number between 1 200 and 1 500 individuals. Perhaps
the most striking bird of the rainforests is the golden
bowerbird, which is confined to altitudes above 800 m.
This species engages in elaborate courtship displays,
including the construction of bowers —dens of twigs
and other plant matter that are sometimes decorated
with material such as egg shells— which may be up to
three meters high.
Of particular scientific interest is the musky rat-
kangaroo. Standing only about 25 cm high, and the
only member of its genus, this species is considered to
represent an early stage of evolution of the kangaroos
from an arboreal, possum-like stock. It is one of the
few Australian mammals that is completely diurnal.
Another curious creature is the mahogany glider
(Petaurus gracilis, EN), which was believed extinct un-
til its rediscovery in 1989. It now clings to a precarious
existence in the lowland eucalyptus forests and wood-
lands of the Wet Tropics. Over 80% of its habitat has
been cleared, mainly for sugarcane, and additional
measures are urgently needed to ensure its survival
(Maxwell et al. 1996). The northern bettong (Bettongia
tropica, EN), one of Australia’s rarest kangaroos, sur-
vives only in three small populations in the drier eu-
calyptus woodlands and open forests. Habitat change
in these sclerophyll communities poses some urgent
problems for its conservation.
Among carnivorous mammals, only the Atherton an-
techinus (Antechinus godmani) and rusty antechinus (A.
adustus) are considered endemic to the region. The
Atherton antechinus, weighing less than 100 g, feeds on
a variety of insects, arachnids, frogs, and lizards. One
notable feature of the biology of this species, and indeed
of all members of the genus that have been studied, is
that males die soon after the mating season, when they
are probably only 11-12 months old (Strahan 1995).
The microhylid frog genus Cophixalus is a distinctive
element of the Wet Tropics amphibian fauna, with most
species confined to mountain tops and tablelands at al-
titudes greater than 300 m or in special habitats. The
Black Mountain boulder frog (Cophixalus saxatilis, VU)
is found only between 100 and 300 m altitude in 580 ha
of the granite boulderfields of the Black Trevethan
Range south of Cooktown. Females at night are recog-
nized by their spectacular canary-yellow coloration.
Two mountain-top nursery frogs, the Bellenden Ker
nursery frog (C. neglectus, VU) and the Mt. Elliot nurs-
ery frog (C. mcdonaldi, VU), are restricted to the cloud
forest at high altitudes of mountain tops, where they oc-
cur in leaf litter and lay their eggs on the ground where
the young develop, hatching as fully developed froglets.
The chameleon gecko is an endemic encountered on
the rainforest floor or facing head-down on small twigs
and shrubs. It will readily cast its distinctive black-and-
white banded tail, which can be regenerated. The cast
tail produces a squeaking sound —an attribute that has
not been noted in any other Australian gecko. The oth-
er endemic gecko of the rainforest, the leaf-tailed gecko
(Saltuarius cornutus), has a lichen-colored body with a
spiny-edged, leaf-like tail and lime-green and brown
eyes. The flattened tail and body results in the animal’s
having a very low profile as it clings to the lichen-cov-
ered tree trunks. It has the ability to remain active at
ambient temperatures at which other reptile species
become inactive.
Threats
The European settlement of northeastern Australia has
been a very recent event. One hundred and thirty years On the opposite page,
of development have resulted in a thriving and pros- Boyd’s rainforest dragon
perous region that has been carved out of a wilderness, (Hypsilurus boydii) is an arboreal
agamid species, usually observed
and this has led to the almost total annihilation of an in-
perching on small trees at a height
digenous way of life that had survived for tens of thou- of approximately one to two meters.
sands of years. © Patricio Robles Gil/Sierra Madre
In those 130 years, 23% of all the vegetation of the
area has been totally cleared, mostly in the lowlands Above, the spectacled flying fox
and on the tablelands to the west of the main coastal (Pteropus conspicillatus) occurs in
primary and secondary rainforest
range, for the growing of sugarcane and for pastures. Of
from Halmahera in
the 14 242 km2 remaining uncleared, an estimated Indonesia eastwards through New
3 000 km2 has been subject to selective logging activity Guinea to the Queensland Wet
and, although its essential features remain, can not be Tropics. This species plays an
considered pristine. Some areas of woodland have also important role as a disperser of the
been subject to light grazing activity, although these fruits of many rainforest species,
particularly those drab-colored
ecosystems remain essentially intact. Looking at the re-
fruits not dispersed by birds.
gion as a whole, we estimate that 58% of the Queens-
© Patricia Rojo
land Wet Tropics remains in pristine condition.
Although much of the region is officially protected as
part of a World Heritage Site, clearing of forest for agri-
culture, pastoral activities, and urban infrastructure de-
velopment continue outside the World Heritage Area;
377

Although much of the Queensland
Wet Tropics is protected as part of a
World Heritage Site, clearing of
forest for agriculture, pastoral
activities, and urban infrastructure
development continue outside the
World Heritage Area, and it is not
uncommon to see tracts of rainforest
land for sale.
© Patricio Robles Gil/Sierra Madre
On the opposite page, the Daintree
Rainforest, in Tropical Far North
Queensland, Australia, is between
100 and 135 million years old —the
oldest in the world. Approximately
430 species of birds live among the
trees, including 13 species that are
found nowhere else in the world.
© Patricio Robles Gil/Sierra Madre
378
however, it is increasingly being regulated by legisla-
tion. The greatest threats to the area now arise from al-
tered fire regimes, introduced weeds, feral animals, water
extraction from streams and aquifers, and drainage of
lowland areas.
Global warming poses serious threats to the region,
which have yet to be clearly defined (Williams et al.,
2003). A number of high-altitude species may find that
they are unable to survive and reproduce in a warmer
climate. A one-degree increase in temperature, consid-
ered a certainty, is predicted to decrease the range of
endemic species to an average 63% of their current
range size. A temperature variation of 3.5ºC, consid-
ered a strong possibility within the next century by the
Intergovernmental Panel on Climate Change, will re-
duce range sizes to an average of 11% of their current
area (Williams et al. 2003).
In recent years, the appearance of chytrid fungus has
devastated frog populations (Berger et al. 1998), while
tree deaths in widely scattered patches of rainforest
have been caused by the soil-borne disease Phytophtho-
ra cinnamomi (Gadek and Worboys 2003), such that
there is now a growing awareness of the destructive po-
tential of introduced diseases.
Relatively few invasive plant or animal species are
able to survive in intact rainforest. The most harmful of
these is the feral pig, which has occupied —at various
population levels— all habitat types. Their impact on
the ground cover can look severe, but has not been
quantified. The cane toad (Bufo marinus) can invade
more successfully along roads and trails of disturbed
rainforests, and the toxic venom they exude is poten-
tially lethal for mammalian and reptilian predators. In
recent years, rusa deer (Cervus timorensis), escaped
from deer farms, have begun spreading into disturbed
and open forest habitats. A wide range of plant species
has been recorded invading disturbed rainforest habi-
tats and sclerophyll habitats. The most serious of these,
which has the capacity to prevent the regeneration of
many of the communities it invades, is the pond apple
(Annona glabra), a small tree from southern parts of the
United States which favors wetland environments.
With the removal of regular fire from many areas of
sclerophyll forests and woodlands, particularly within
the last 50 years, changes in the understory have creat-
ed conditions due to which canopy species no longer
regenerate. In particular, the removal of fire has al-
lowed rainforest species to invade sclerophyll habitats,
particularly those dominated by tall eucalypts. These
habitats can not regenerate without the periodic expo-
sure of the soil surface following the removal of ground
cover and understory species by fire. In the absence of
fire, the understory and canopy are progressively re-
placed by rainforest species, and the sclerophyll com-
munity is replaced by a rainforest one. Anywhere in
the wet tropics where soils are fertile enough to support
rainforest growth, the interface between sclerophyll
and rainforest communities is maintained by fire. The
loss of any sclerophyll habitat as a result of this process
represents a significant loss of biodiversity and raises
particularly serious questions for the future of species
such as the yellow-bellied glider.
Conservation
A critical moment in the conservation of the Wet Trop-
ics was the inscription in 1988 of 8 944 km2 of the Wet
Tropics, including most of the state-owned lands in the
area, as the Wet Tropics World Heritage Area. Around
that time, most of the selective logging of rainforest,
which had proceeded at varying levels of intensity for
more than 80 years, was halted by federal legislation.
The cessation of logging activity created enormous con-
troversy in north Queensland. The Wet Tropics became
a major national political issue, with relevant environ-
mental laws being unsuccessfully challenged in Aus-
tralia’s High Court.
Since World Heritage listing, the area of the Wet Trop-
ics with national park status has increased to 3 762 km2.
Major national parks in the region include Barron Gorge
(28 km2), Cedar Bay (56 km2), Daintree (760 km2), Ed-
mund Kennedy (69 km2), Ella Bay (37 km2), Hinchin-
brook Island (399 km2), Lumholtz (1 400 km2), Paluma
Range (106 km2), and Wooroonooran (798 km2). A fur-
ther 4 694 km2 is in the process of transfer to national
park tenure. Recent state and federal legislation provides
additional protection, with groundbreaking new federal
legislation (the Environment Protection and Biodiversity
Conservation Act of 1999) stipulating fines of over $5 mil-
lion for any actions adversely affecting World Heritage
Site values. Notwithstanding, there are still areas of rain-
forest outside the Wet Tropics World Heritage Area and
not under protection that are at risk of being cleared in
the future.
An emerging strategy for conservation in the Wet
Tropics, as it is in other parts of Australia, is the devel-
opment and implementation of private sector (non-
government) initiatives to protect habitat. The largest
non-government conservation area in the Wet Tropics,
the Mount Zero-Taravale Wildlife Sanctuary, is owned
by the non-profit Australian Wildlife Conservancy
(AWC). In addition to protecting critical habitat for sev-
eral threatened species, AWC is working with govern-
ment agencies to conduct management-focused re-
search on key issues such as the ecological role of fire
in maintaining wet sclerophyll communities. The role
of NGOs and other “off-reserve” measures will become
increasingly important in delivering landscape-scale
conservation in the Wet Tropics.
JAMES PETER STANTON 111, 112
PETER D. BOSTOCK 113
KEITH R. MCDONALD 114
GARRY L. WERREN 115
ATTICUS FLEMING 116

Only references specifically cited in the Introduction or in
the individual chapters are included here. For a full list of
references consulted both in this volume and in the origi-
nal Hotspots book (1999), please see the Biodiversity
Hotspots website: www.biodiversityhotspots.org
General References
BIRDLIFE INTERNATIONAL. 2003. Saving Asia’s Threatened
Birds: A Guide for Government and Civil Society. Cam-
bridge.
DODSON, C.H. and A.H. GENTRY. 1991. “Biological Extinc-
tion in Western Ecuador.” Annals Missouri Botanical
Garden 78:273-295.
FLANNERY, T.F. 1995. Mammals of the South-West Pacific and
Moluccan Islands. Reed, Sydney.
FRIIS, I., M. THULIN, H. ADSERSEN, and A.M. BÜRGER. In
press. “Patterns of Plant Diversity and Endemism in
the Horn of Africa.” Biol. Skr.
FWI/GFW. 2002. The State of the Forest: Indonesia. Forest
Watch Indonesia, Bogor, Indonesia, and Global Forest
Watch, Washington, D.C.
GARGOMINY, O. (Ed.). 2003. Biodiversité et conservation dans
les collectivités françaises d’outre-mer. Collection Planè-
te Nature. Comité français pour l’IUCN, Paris.
GENTRY, A.H. 1977. “Endangered Plant Species and Habi-
tats of Ecuador and Amazonian Peru.” In: G.T. PRANCE
and T.S. ELIAS (Eds.). Extinction Is Forever. New York
Botanical Garden, New York.
JØRGENSEN, P.M. and S. LEÓN-YÁÑEZ (Eds.). 1999. Cata-
logue of the vascular plants of Ecuador. Monograph.
Syst. Bot. Missouri Bot. Gard. 75:i-vii, 1-1181.
MACKINNON, J., M. SHA, C. CHEUNG, G. CAREY, X. ZHU, and
D. MELVILLE. 1996. A Biodiversity Review of China.
WWF International, Hong Kong.
MUELLER-DOMBOIS, D. and F.R. FOSBERG. 1998. Vegetation of
tropical Pacific islands. Springer-Verlag, New York.
POPULATION REFERENCE BUREAU. 2003. World Population Data
Sheet. Population Reference Bureau, Washington, D.C.
STEENIS VAN, C.G. 1950. “The Delimitation of Malesia and
its Main Plant Geographical Divisions.” Flora Male-
siana Series 1:70-75.
TAKHTAJAN, A. 1986. Floristic Regions of the World. Univer-
sity of California Press, Berkeley.
VLOK, J.H.J., D.I.W. EUSTON-BROWN, and R.M. COWLING.
2003. “Acocks Valley Bushveld 50 Years on: New Per-
spectives on the Delimitation, Characterisation and
Origin of Thicket Vegetation.” South African Journal of
Botany 69:27-51.
VAN WYK, A.E. and G.F. SMITH. 2001. Regions of Floris-
ticEendemism in Southern Africa. A Review with Em-
phasis on Succulents. Umdaus Press, Pretoria.
WIKRAMANAYAKE, E.D., E. DINERSTEIN, C. LOUCKS, D. OL-
SON, J. MORRISON, J. LAMOREUX, M. MCKNIGHT, and P.
HEDAO (Eds.). 2002. Terrestrial Ecoregions of the Indo-
Pacific: A Conservation Assessment. Island Press, Wash-
ington, D.C.
WORLD BANK. 2002. www.worldbank.org/data/quickrefer-
ence/quickref.html
Introduction
ABE, T.D., E. BIGNELL, and M. HIGASHI. 2000. Termites: Evo-
lution, Sociality, Symbiosis, Ecology. Kluwer Academic
Publishing, Dordrecht.
ÁLVAREZ, L.W., W. ÁLVAREZ, F. ASARO, and H.V. MICHEL.
1980. “Extraterrestrial Cause for the Cretaceous-Ter-
tiary Extinction.” Science 208:1095-1108.
ANDO, A., J. CAMM, S. POLASKY, and A. SOLOW. 1997.
“Species Distributions, Land Values, and Efficient
Conservation.” Science 279:2126-2128.
AVISE, J.C. and G.C. JOHNS. 1999. “Proposal for a Standard
Temporal Scheme of Biological Classification for Ex-
tant Species.” Proceedings of the National Academy of
Sciences of the U.S.A. 96:7358-7363.
BAILEY, R.G. 1994. Descriptions of the Ecoregions of the Unit-
ed States. 2nd edition. Misc. Publ. 1391, revised. U.S.
Department of Agriculture, Washington, D.C.
BAILLIE, J. and B. GROOMBRIDGE (Comps. and Eds.). 1996.
1996 IUCN Red List of Threatened Animals. IUCN,
Gland and Cambridge.
REFERENCES
BAKARR, M.I., G.A. FONSECA, R.A. MITTERMEIER, A.B. RY-
LANDS, and K.W. PAINEMILLA. 2001. “Hunting and Bush-
meat Utilization in the African Rain Forest.” Advances
in Applied Biodiversity. Science 2:1-170.
BALMFORD, A., A. BRUNER, P. COOPER, R. COSTANZA, S. FAR-
BER, R.E. GREEN, M. JENKINS, P. JEFFERISS, V. JESSAMY, J.
MADDEN, K. MUNRO, N. MYERS, S. NAEEM, J. PAAVOLA,
M. RAYMENT, S. ROSENDO, J. ROUGHGARDEN, K. TRUMPER,
and R.K. TURNER. 2002. “Economic Reasons for Con-
serving Wild Nature.” Science 297:950-953.
BALMFORD, A., J. MOORE, T. BROOKS, N. BURGESS, L.A.
HANSEN, P. WILLIAMS, and C. RAHBEK. 2001. “Conser-
vation Conflicts across Africa.” Science 291:2616-2619.
BALMFORD, A., K.J. GASTON, S. BLYTH, A. JAMES, and V. KA-
POS. 2003. “Global Variation in Terrestrial Conserva-
tion Costs, Conservation Benefits, and Unmet Conser-
vation Needs.” Proceedings of the National Academy
of Sciences of the U.S.A. 100:1046-1050.
BIRDLIFE INTERNATIONAL. 2000. Threatened Birds of the
World. Lynx Edicions, Barcelona.
BISBY, F.A., J. SHIMURA, M. RUGGIERO, J. EDWARDS, and C.
HAEUSER. 2002. “Taxonomy, at the Click of a Mouse.”
Nature 418:367.
BOITANI, L., F. CORSI, A. DE BIASE, I.D. CARRANZA, M.
RAVAGLI, G. REGGIANI, I. SINIBALDI, and P. TRAPANESE.
1999. A Databank for the Conservation and Manage-
ment of African Mammals. Institute of Applied Ecolo-
gy, Rome.
BRAMWELL, D. 2002. “How Many Plant Species Are There?”
Plant Talk 28:32-33.
BROOKS, T.M., J.D. PILGRIM, A.S.L. RODRIGUES, and G.A.B.
DA FONSECA. 2004. “Conservation Status and Geo-
graphic Distribution of Avian Evolutionary History.”
In: A. PURVIS, J.L. GITTLEMAN, and T.M. BROOKS (Eds.).
Phylogeny and Conservation. Cambridge University
Press, Cambridge.
BROOKS, T.M., R.A. MITTERMEIER, C.G. MITTERMEIER, G.A.B.
DA FONSECA, A.B. RYLANDS, W.R. KONSTANT, P. FLICK, J.D.
PILGRIM, S. OLDFIELD, G. MAGIN, and C. HILTON-TAYLOR.
2002. “Habitat Loss and Extinction in the Hotspots of
Biodiversity.” Conservation Biology 16:909-923.
BROOKS, T.M., S.L. PIMM, and N.J. COLLAR. 1997. “The Ex-
tent of Deforestation Predicts the Number of Birds
Threatened with Extinction in Insular South-East
Asia.” Conservation Biology 11:382-394.
BRUNER, A.G., R.E. GULLISON, R.E. RICE, and G.A.B. DA
FONSECA. 2001. “Effectiveness of Parks in Protecting
Tropical Biodiversity.” Science 291:125-128.
BURGESS, N. and G.P. CLARKE. 2000. The Coastal Forests of
Eastern Africa. IUCN, Cambridge.
CARROLL, S.S. and D.L. PEARSON. 1998. “Spatial Modeling of
Butterfly Species Richness Using Tiger Beetles (Ciicn-
delidiae) as a Bioindicator Taxon.” Ecological Applica-
tions 8:531-543.
CHOMITZ, K.M., K. ALGER, T.S. THOMAS, H. ORLANDO, and
P. VILA NOVA. 2004. Opportunity Costs of Conservation
in a Biodiversity Hotspot: The Case of Southern Bahia.
Environment and Development Economics.
CINCOTTA, R.P., J. WISNEWSKI, and R. ENGELMAN. 2000. “Hu-
man Population in the Biodiversity Hotspots.” Nature
404:990-992.
CONSERVATION INTERNATIONAL. 2004. Conserving Earth’s Liv-
ing Heritage: A Proposed Framework for Designing Bio-
diversity Conservation Strategies. Conservation Inter-
national, Washington, D.C.
COSTANZA, R., R. D’ARGE, R. DE GROOT, S. FARBER, M. GRAS-
SO, B. HANNON, K. LIMBURG, S. NAEEM, R.V. O’NEILL, J.
PARUELO, R.G. RASKIN, P. SUTTON, and M. VAN DEN BELT.
1997. “The Value of the World’s Ecosystem Services
and Natural Capital.” Nature 387:253-258.
COWLING, R.M., R.L. PRESSEY, A.T. LOMBARD, P.G. DESMET,
and A.G. ELLIS. 1999. “From Representation to Persis-
tence: Requirement for a Sustainable Reserve System
in the Species Rich Mediterranean-Climate Deserts of
Southern Africa.” Diversity and Distributions 5:1-21.
DALTON, R. 2000. “Biodiversity Cash Aimed at Hotspots.”
Nature 406:818.
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1994. Centers of Plant Diversity. A Guide and Strategy
for their Conservation, Vol. 1: Europe, Africa, South
West Asia and the Middle East. The World Wildlife
Fund (WWF) and IUCN-The World Conservation
Union. IUCN Publications Unit, Cambridge.
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1995. Centers of Plant Diversity. A Guide and Strategy
for their Conservation, Vol. 2: Asia, Australasia and the
Pacific. WWF/IUCN-The World Conservation Union.
IUCN Publications Unit, Cambridge.
DAVIS, S.D., V.H. HEYWOOD, O. HERRERA-MACBRYDE, J. VIL-
LA-LOBOS, and A.C. HAMILTON (Eds.). 1997. Centers of
Plant Diversity: A Guide and Strategy for their Conserva-
tion, Vol. 3: The Americas. WWF/IUCN, IUCN Publi-
cations Unit, Cambridge
DIAMOND, J.M. 1976. “Island Biogeography and Conserva-
tion: Strategy and Limitations.” Science 193:1027-1029.
VAN DIJK, P.P., B.L. STUART, and A.G.J. RHODIN. 2000. “Asian
Turtle Trade.” Chelonian Research Monographs 2:1-
164.
DINERSTEIN, E., D.M. OLSON, D.J. GRAHAM, A.L. WEBSTER,
S.A. PRIMM, M.P. BOOKBINDER, and G. LEDEC. 1995. A
Conservation Assessment of the Terrestrial Ecoregions of
Latin America and the Caribbean. World Wildlife Fund,
Washington, D.C.
DUDLEY, J.P., J.R. GINSBERG, A.J. PLUMPTRE, J.A. HART, and
L.C. CAMPOS. 2002. “Effects of War and Civil Strife on
Wildlife and Wildlife Habitats.” Conservation Biology
16:319-329.
ESCHMEYER, W.N. 1998. Catalogue of Fishes. California
Academy of Sciences, San Francisco.
FISHPOOL, L.D.C. and M.I. EVANS. 2001. Important Bird Ar-
eas in Africa and Associated Islands. BirdLife Conser-
vation Series No. 11. BirdLife International, Cam-
bridge.
FITTER, R. and M.S. FITTER. 1987. Road to Extinction: Prob-
lems with Categorizing the Status of Taxa Threatened
with Extinction. IUCN, Gland.
DA FONSECA, G.A.B., A. BALMFORD, C. BIBBY, L. BOITANI, F.
CORSI, T. BROOKS, C. GASCON, S. OLIVIERI, R.A. MITTER-
MEIER, N. BURGESS, E. DINERSTEIN, D. OLSON, L. HAN-
NAH, J. LOVETT, D. MOYER, C. RAHBEK, S. STUART, and P.
WILLIAMS. 2000. “Following Africa’s Lead in Setting
Priorities.” Nature 405:393-394.
FRAZEE, S.R., R.M. COWLING, R.L. PRESSEY, J.K. TURPIE, and
N. LINDENBERG. 2003. “Estimating the Costs of Con-
serving a Biodiversity Hotspot: A Case-Study of the
Cape Floristic Region, South Africa.” Biological Conser-
vation 112:275-290.
FROESE, R. and D. PAULY. 2003. FishBase. WorldFish Center
FishBase Project, Laguna.
GASTON, K.J. 2000. “Global Patterns in Biodiversity.” Na-
ture 405:220-227.
GOVAERTS, R. 2001. “How Many Species of Seed Plants Are
There?” Taxon 50:1085-1090.
GROVES, R.H. and F. DI CASTRI. 1991. Biogeography of
Mediterranean Invasions. Cambridge University Press,
Cambridge.
GUAN, Y., B.J. ZHENG, Y.Q. HE, X.L. LIU, Z.X. ZHUANG, C.L.
CHEUNG, S.W. LUO, P.H. LI, L.J. ZHANG, Y.J. GUAN, K.M.
BUTT, K.L. WONG, K.W. CHAN, W. LIM, K.F. SHORTRIDGE,
K.Y. YUEN, J.S.M. PEIRIS, and L.L.M. POON. 2003. “Iso-
lation and Characterization of Viruses Related to the
SARS Coronavirus from Animals in Southern China.”
Science 302:276-278.
HAUGE, T. and T. ELLINGSEN. 1998. “Beyond Environmental
Scarcity: Causal Pathways to Conflict.” Journal of Peace
Research 35:299-317.
HELLIN, J., M. HAIGH, and F. MARKS. 1999. “Rainfall Char-
acteristics of Hurricane Mitch.” Nature 399:316.
HOLMES, D. 2000. Deforestation in Indonesia: A Review of the
Situation in Sumatra, Kalimantan, and Sulawesi. World
Bank, Jakarta.
HOULAHAN, J.E., C.S. FINDLAY, B.R. SCHMIDT, A.H. MEYER,
and S.L. KUZMIN. 2000. “Quantitative Evidence for
Global Amphibian Declines.” Nature 404:752-755.
HOWARD, P.C., P. VISKANIC, T.R.B. DAVENPORT, F.W. KIGENYI,
M. BALTZER, C.J. DICKINSON, J.S. LWANGA, R.A. MATTHEWS,
and A. BALMFORD. 1998. “Complementarity and the
Use of Indicators for Reserve Selection in Uganda.”
Nature 394:472-475.
ICBP. 1992. Putting Biodiversity on the Map: Priority Areas
for Global Conservation. International Council for Bird
Preservation, Cambridge.
IUCN. 1994. Guidelines for Protected Area Management Cat-
egories. IUCN, Gland.
IUCN. 2001. IUCN Red List Categories and Criteria. Version
3.1. IUCN, Cambridge.
IUCN. 2003. 2003 IUCN Red List of Threatened Species.
IUCN, Cambridge.
VAN JAARSVELD, A.S., S. FREITAG, S.L. CHOWN, C. MULLER, S.
KOCH, H. HULL, C. BELLAMY, M. KRÜGER, S. ENDRÖDY-
YOUNGA, M.W. MANSELL, and C.H. SCHOLTZ. 1998. “Bio-
diversity Assessment and Conservation Strategies.”
Science 279:2106-2108.
JAMES, A., K. GASTON, and A. BALMFORD. 1999. “Balancing
the Earth’s Accounts.” Nature 401:323-324.
KAIMOWITZ, D. and J. SMITH. 2001. “Soybean Technology
and the Loss of Natural Vegetation in Brazil and Bo-
livia.” In: A., ANGELSEN and D. KAIMOWITZ (Eds.). Agri-
cultural Technologies and Tropical Deforestation. CABI
Publishing, Wallingford.
KREVER, V., N. ZAZANASHVILI, H. JUNGIUS, L. WILLIAMS, and
D. PETELIN. 2001. Biodiversity of the Caucasus Ecore-
gion. World Wide Fund for Nature, Moscow.
LAMBSHEAD, P.J.D. 1993. “Recent Developments in Marine
Benthic Biodiversity Research.” Oceanis 19:5-24.
LAMOREUX, J., H.R. AKÇAKAYA, L. BENNUN, N.J. COLLAR, L.
BOITANI, D. BRACKETT, A. BRÄUTIGAM, T.M. BROOKS,
G.A.B. DA FONSECA, R.A. MITTERMEIER, A.B. RYLANDS,
U. GÄRDENFORS, C. HILTON-TAYLOR, G. MACE, B.A.
STEIN, and S. STUART. 2003. “Value of the IUCN Red
List.” Trends in Ecology and Evolution 18:214-215.
LOVETT, J.C. and S.K. WASSER. 1993. Biogeography and Ecol-
ogy of the Rain Forests of Eastern Africa. Cambridge
University Press, Cambridge.
MACE, G.M. and R. LANDE. 1991. “Assessing Extinction
Threats: Towards a Reevaluation of IUCN Threatened
Species Categories.” Conservation Biology 5:148-157.
MARGULES, C.R. and R.L. PRESSEY. 2000. “Systematic Con-
servation Planning.” Nature 405:243-253.
MARTIN, E. and D. STILES. 2002. The South and South-East
Asian Ivory Markets. Save the Elephants, Nairobi.
MAY, R.M. 1992. “How Many Species Inhabit the Earth?”
Scientific American 267:42-48.
MAY, R.M., R.H. LAWTON, and N.E. STORK. 1995. “Assessing
Extinction Rates.” In: J.H. LAWTON and R.M. MAY (Eds.).
Extinction Rates. Oxford University Press, Oxford.
MCALLISTER, D.E., A.L. HAMILTON, and B. HARVEY. 1997.
“Global Freshwater Biodiversity: Striving for the In-
tegrity of Freshwater Ecosystems.” Sea Wind 11(2):1-
106.
MCNEELY, J.A., K.R. MILLER, W.V. REID, R.A. MITTERMEIER,
and T.B. WERNER. 1990. Conserving the World’s Biologi-
cal Diversity. The World Bank, Washington, D.C.
MEIER, R. and T. DIKOW. 2004. “The Significance of Speci-
men Databases from Taxonomic Revisions for Esti-
mating and Mapping the Global Species Diversity of
Invertebrates and Repatriating Reliable and Complete
Specimen Data.” Conservation Biology 18(2):478-488
MIDGLEY, G., L. HANNAH, D. MILLAR, M. RUTHERFORD, and
L. POWRIE. 2002. “Assessing the Vulnerability of Spe-
cies Richness to Anthropogenic Climate Change in a
Biodiversity Hotspot.” Global Ecology and Biogeography
11:445-451.
MITTERMEIER, R.A. 1988. “Primate Diversity and the Trop-
ical Forest.” In: E.O. WILSON (Ed.). BioDiversity. Na-
tional Academy Press, Washington, D.C.
MITTERMEIER, R.A., C.G. MITTERMEIER, P. ROBLES-GIL, J.D.
PILGRIM, W.R. KONSTANT, G.A.B. DA FONSECA, and T.M.
BROOKS (Eds.). 2002. Wilderness. Earth’s Last Wild
Places. CEMEX, Mexico City.
MITTERMEIER, R.A., C.G. MITTERMEIER, T.M. BROOKS, J.D.
PILGRIM, W.R. KONSTANT, G.A.B. DA FONSECA, and C.
KORMOS. 2003b. “Wilderness and Biodiversity Conser-
vation.” Proceedings of the National Academy of Sci-
ences of the U.S.A. 100:10309-10313.
MITTERMEIER, R.A., G.A.B. DA FONSECA, T. BROOKS, J.D.
PILGRIM, and A. RODRIGUES. 2003a. “Hotspots and
Coldspots.” American Scientist 91:384.
MITTERMEIER, R.A., N. MYERS, J.B. THOMSEN, G.A.B. DA
FONSECA, and S. OLIVIERI. 1998. “Biodiversity Hotspots
and Major Tropical Wilderness Areas: Approaches to
Setting Conservation Priorities.” Conservation Biology
12:516-520.
MITTERMEIER, R.A., N. MYERS, P. ROBLES-GIL, and C.G. MIT-
TERMEIER (Eds.). 1999. Hotspots. Earth’s Biologically
Richest and Most Endangered Terrestrial Ecoregions. CE-
MEX/Agrupación Sierra Madre, Mexico City.
MITTERMEIER, R.A., P. ROBLES-GIL, and C.G. MITTERMEIER
(Eds.). 1997. Megadiversity. Earth’s Biologically Wealth-
381
 iest Nations. CEMEX/Agrupación Sierra Madre, Mexi-
co City.
MYERS, N. 1983. A Wealth of Wild Species. Westview, Boulder.
MYERS, N. 1988. “Threatened Biotas: ‘Hotspots’ in Tropical
Forests.” The Environmentalist 8:1-20.
MYERS, N. 1990. “The Biodiversity Challenge: Expanded
Hot-Spots Analysis.” The Environmentalist 10:243-256.
MYERS, N. 1998. “Lifting the Veil on Perverse Subsidies.”
Nature 392:327-328.
MYERS, N. 2003. “Biodiversity Hotspots Revisited.” Bio-
Science 53:916-917.
MYERS, N. and R.A. MITTERMEIER. 2003. “Impact and Ac-
ceptance of the Hotspots Strategy: Response to Ovadia
and to Brummitt and Lughadha.” Conservation Biology
17:1449-1450.
MYERS, N., R.A. MITTERMEIER, C.G. MITTERMEIER, G.A.B.
DA FONSECA, and J. KENT. 2000. “Biodiversity Hotspots
for Conservation Priorities.” Nature 403:853-858.
NOVOTNY, V., Y. BASSET, S.E. MILLER, G.D. WEIBLEN, B. BRE-
MER, L. CIZEK, and P. DROZD. 2002. “Low host Speci-
ficity of Herbivorous Insects in a Tropical Forest.” Na-
ture 416:841-844.
OLSON, D.M. and E. DINERSTEIN. 1998. “The Global 200: A
Representation Approach to Conserving the Earth’s
Most Biologically Valuable Ecoregions.” Conservation
Biology 12:502-515.
OLSON, D.M., E. DINERSTEIN, E.D. WIKRAMANAYAKE, N.D.
BURGESS, G.V.N. POWELL, E.C. UNDERWOOD, J.A. D’AMI-
CO, I. ITOUA, H.E. STRAND, J.C. MORRISON, C.J. LOUCKS,
T.F. ALLNUTT, T.H. RICKETTS, Y. KURA, J.F LAMOREUX,
W.W. WETTENGEL, P. HEDAO, and K.R. KASSEM. 2001.
“Terrestrial Ecoregions of the World: A New Map of
Life on Earth.” BioScience 51:933-938.
PEARSON, D.L. and F. CASSOLA. 1992. “World-Wide Species
Richness Patterns of Tiger Beetles (Coleoptera: Cicin-
delidae): Indicator Taxon for Biodiversity and Conser-
vation Studies.” Conservation Biology 6:376-391.
PEARSON, D.L. and S.S. CARROLL. 1998. “Global Patterns of
Species Richness: Spatial Models for Conservation
Planning Using Bioindicator and Precipitation Data.”
Conservation Biology 12:809-821.
PEARSON, D.L. and S.S. CARROLL. 1999. “The Influence of
Spatial Scale on Cross-Taxon Congruence Patterns and
Prediction Accuracy of Species Richness.” Journal of
Biogeography 26:1079-1090.
PERES, C.A. and I.R. LAKE. 2003. “Extent of Nontimber Re-
source Extraction in Tropical Forests: Accessibility to
Game Vertebrates by Hunters in the Amazon Basin.”
Conservation Biology 17:521-535.
PETERSON, A.T., L.G. BALL, and K.P. COHOON. 2002. “Pre-
dicting Distributions of Mexican Birds Using Ecologi-
cal Niche Modelling Methods.” Ibis 144:27-32.
PIMM, S.L., G.J. RUSSELL, J.L. GITTLEMAN, and T.M. BROOKS.
1995. “The Future of Biodiversity.” Science 269:347-
350.
PIMM, S.L., M. AYRES, A. BALMFORD, G. BRANCH, K. BRAN-
DON, T. BROOKS, R. BUSTAMENTE, R. COSTANZA, R.
COWLING, L.M. CURRAN, A. DOBSON, S. FARBER, G.A.B.
DA FONSECA, C. GASCON, R. KITCHING, J. MCNEELY, T.
LOVEJOY, R.A. MITTERMEIER, N. MYERS, J.A. PATZ, B.
RAFFLE, D. RAPPORT, P. RAVEN, C. ROBERTS, J.P. RO-
DRÍGUEZ, A. RYLANDS, C. TUCKER, C. SAMPER, M.L.J.
STIASSNY, C. SAFINA, J. SUPRIATNA, D.H. WALL, and D.
WILCOVE. 2001. “Can We Defy Nature’s End?” Science
293:2207-2208.
PLUMPTRE, A.J., M. BEHANGANA, T.R.B. DAVENPORT, C.
KAHINDO, R. KITYO, E. NDOMBA, D. NKUUTU, I. OWIUN-
JI, P. SSEGAWA, and G. EILU. 2003. The Biodiversity of the
Albertine Rift. Albertine Rift Technical Reports Series
No. 3. Wildlife Conservation Society, Kampala.
PRANCE G.T., H. BEENTJE, J.D. DRANSFIELD, and R. JOHNS.
2000. “The Tropical Flora Remains Undercollected.”
Annals of the Missouri Botanical Garden 87:67-71.
PRESSEY, R.L. and A.O. NICHOLLS. 1989. “Efficiency in Con-
servation Evaluation: Scoring Versus Iterative Ap-
proaches.” Biological Conservation 50:199-218.
PRESSEY, R.L., I.R. JOHNSON, and P.D. WILSON. 1994.
“Shades of Irreplaceability: Towards a Measure of the
Contribution of Sites to a Reservation Goal.” Biodiver-
sity and Conservation 3:242-262.
RAXWORTHY, C.J., E. MARTÍNEZ-MEYER, N. HORNING, R.A.
NUSSBAUM, G.E. SCHNEIDER, M.A. ORTEGA-HUERTA, and
A.T. PETERSON. 2003. “Predicting Distributions of
Known and Unknown Reptile Species in Madagascar.”
Nature 426:837-841.
REID, W.V. 1998. “Biodiversity hotspots.” Trends in Ecology
and Evolution 13:275-280.
REIS, R.E., S.O. KULLANDER, and C.J. FERRARIS JR. 2003.
Check List of the Freshwater Fishes of South and Central
America. Edipucrs, Porto Alegre.
382
RICKETTS, T.H., E. DINERSTEIN, D.M. OLSON, C.J. LOUCKS,
W. EICHBAUM, D. DELLA SALA, K. KAVANAGH, P. HEDAO,
P.T. HURLEY, K.M. CARNEY, R. ABELL, and S. WALTERS.
1999. Terrestrial Ecoregions of North America: A Con-
servation Assessment. Island Press, Washington, D.C.
ROBERTS, C.M., C.J. MCCLEAN, J.E.N. VERON, J.P. HAWKINS,
G.R. ALLEN, D.E. MCALLISTER, C.G. MITTERMEIER, F.W.
SCHUELER, M. SPALDING, F. WELLS, C. VYNNE, and T.B.
WERNER. 2002. “Marine Biodiversity Hotspots and
Conservation Priorities for Tropical Reefs.” Science
295:1280-1284.
RODRIGUES, A.S.L., S.J. ANDELMAN, M.I. BAKARR, L. BOI-
TANI, T.M. BROOKS, R.M. COWLING, L.D.C. FISHPOOL,
G.A.B. DA FONSECA, K.J. GASTON, M. HOFFMANN, J.S.
LONG, P.A. MARQUET, J.D. PILGRIM, R.L. PRESSEY, J.
SCHIPPER, W. SECHREST, S.N. STUART, L.G. UNDERHILL,
R.W. WALLER, M.E.J. WATTS, and X. YAN. 2004. “Effec-
tiveness of the Global Protected Area Network in Rep-
resenting Species Diversity.” Nature 428:640-643.
SACHS, J. 2002. “A New Global Effort to Control Malaria.”
Science 298:122-124.
SANDERSON, E.W., K.H. REDFORD, A. VEDDER, P.B. COPPOLIL-
LO, and S.E. WARD. 2002. “A Conceptual Model for Con-
servation Planning Based on Landscape Species Re-
quirements.” Landscape and Urban Planning 58:41-56.
SANDERSON, J., K. ALGER, G.A.B. DA FONSECA, C. GALINDO
LEAL, V.H. INCHAUSTY, and K. MORRISON. 2003. Biodi-
versity Conservation Corridors: Planning, Implementing,
and Monitoring Sustainable Landscapes. Conservation
International, Washington, D.C.
SECHREST, W., T.M. BROOKS, G.A.B. DA FONSECA, W.R. KON-
STANT, R.A. MITTERMEIER, A. PURVIS, A.B. RYLANDS, and
J.L. GITTLEMAN. 2002. “Hotspots and the Conservation
of Evolutionary History.” Proceedings of the National
Academy of Sciences of the U.S.A. 99:2067-2071.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. Endemic Bird Areas of the World: Priorities
for Biodiversity Conservation. BirdLife International,
Cambridge.
STEADMAN, D.W. 1995. “Prehistoric Extinctions of Pacific
Island Birds: Biodiversity Meets Zooarcheology.” Sci-
ence 267:1123-1131.
SUGDEN, A. and E. PENNISI. 2000. “Diversity Digitized.” Sci-
ence 289:2305.
TERBORGH, J. 1988. “The Big Things that Run the World.”
Conservation Biology 2:402-403.
THOMAS, C.D., A. CAMERON, R.E. GREEN, M. BAKKENES, L.J.
BEAUMONT, Y.C. COLLINGHAM, B.F.N. ERASMUS, M.F. DE
SIQUEIRA, A. GRAINGER, L. HANNAH, L. HUGHES, B.
HUNTLEY, A.S. VAN JAARSVELD, G.F. MIDGELY, L. MILES,
M.A. ORTEGA HUERTA, A.T. PETERSON, O.L. PHILLIPS,
and S.E. WILLIAMS. 2004. “Extinction Risk from Cli-
mate Change.” Nature 427:125-128.
THORNE, R.F. 2002. “How Many Species of Seed Plants Are
There?” Taxon 51:511-512.
UDVARDY, M.D.F. 1975. A Classification of the Biogeographi-
cal Provinces of the World. IUCN Occassional Paper No.
18. IUCN, Morges.
VITOUSEK, P.M., H.A. MOONEY, J. LUBCHENCO, and J.M.
MELILLO. 1997. “Human Domination of Earth’s
Ecosystems.” Science 277:494-499.
WDPA. 2003. 2003 World Database on Protected Areas.
IUCN-WCPA and UNEP-WCMC, Washington, D.C.
WHEELER, Q.D. and R. MEIER. 2000. Species Concepts and
Phylogenetic Theory: A Debate. Columbia University
Press, New York.
WHITE, F. 1983. “The Vegetation of Africa: A Descriptive
Memoir to Accompany the UNESCO/AETFAT/UNSO
Vegetation Map of Africa.” UNESCO Natural Resources
Research 20:1-356.
VAN WYK, A.E. and G.F. SMITH. 2001. See General References.
WIKRAMANAYAKE, E., D. DINERSTEIN, C. LOUCKS, D. OLSON,
J. MORRISON, J. LAMOREUX, M. MCKNIGHT, and P.
HEDAO. 2002. “Ecoregions in Ascendance.” Conserva-
tion Biology 16:1-6.
WILSON, E.O. 1984. Biophilia. Harvard University Press,
Boston.
WILSON, E.O. 1987. “The Little Things that Run the World
(the Importance and Conservation of Invertebrates).”
Conservation Biology 1:344-346.
WILSON, E.O. 1992. The Diversity of Life. Belknap Harvard,
Boston.
WILSON, E.O. and S.R. KELLERT. 1993. The Biophilia Hy-
pothesis. Island Press, Washington, D.C.
Tropical Andes
DODSON, C.H. and A.H. GENTRY. 1991. See General Refer-
ences.
GENTRY, A.H. 1977. See General References.
GENTRY, A.H. 1990. “La región amazónica.” In: C.C. URIBE
et al. Selva húmeda de Colombia. Villegas, Bogotá.
JØRGENSEN, P.M. and S. LEÓN-YÁÑEZ (Eds.). 1999. See Gen-
eral References.
KESSLER, M. 2000. “Elevational Gradients in Species Rich-
ness and Endemism of Selected Plant Groups in the
Central Bolivian Andes.” Plant Ecology 149:181-193.
KESSLER, M. 2002. “The Elevational Gradient of Andean
Plant Endemism: Varying Influences of Taxon-Specif-
ic Traits and Topography at Different Taxonomic Lev-
els.” J Biogeogr. 29:1159-1165.
LANGENDOEN, D. and A.H. GENTRY. 1991. “The Structure
and Diversity of Rain Forests at Bajo Calima, Chocó
Region, Western Colombia.” Biotropica 23:2-11.
MYERS, N. 1988. “Threatened Biotas: Hotspots in Tropical
Forests.” The Environmentalist 8:118-208.
RON, S.R., W.E. DUELLMAN, L.A. COLOMA, and M.R. BUSTA-
MANTE. 2003. “Population Decline of the Jambato Toad
Atelopus Ignescens (Anura: Bufonidae) in the Andes
of Ecuador.” Journal of Herpetology 37:116-126
STOTZ, D.F., J.W. FITZPATRICK, T.A. PARKER III., and D.K.
MOSKOVITS. 1996. Neotropical Birds: Ecology and Con-
servation. University of Chicago Press, Ltd., London.
VAN DER WERFF, H. and T. CONSIGLIO. In press. “Distribu-
tion and Conservation Significance of Endemic Spe-
cies of Flowering Plants in Peru.” Biodiversity and Con-
servation.
Tumbes-Chocó-Magdalena
GENTRY, A.H. 1977. See General References.
GENTRY, A.H. 1979. “Extinction and conservation of plant
species in tropical America: A phytogeographical per-
spective.” In: I. HEDBERG (Ed.). Systematic Botany,
Plant Utilization, and Biosphere Conservation. Almqvist
and Wiksell, Stockholm.
GENTRY, A.H. 1990. “La Región del Chocó.” In: C.C. URIBE
et al. Selva Húmeda de Colombia. Villegas, Bogotá.
GENTRY, A.H. 1993. “Riqueza de especies y composición
florística.” In: P. LEYVA (Ed.). Colombia Pacífico, Vol. 1.
Fondo Protección del Medio Ambiente José Celestino
Mutis, Publicaciones Financiera Eléctrica Nacional
(FEN), Bogotá.
JØRGENSEN, P.M. and S. LEÓN-YÁÑEZ (Eds.). 1999. See Gen-
eral References.
MITTERMEIER, R.A. 1997. “Colombia.” In: R.A. MITTER-
MEIER, P. ROBLES-GIL, and C.G. MITTERMEIER (Eds.).
1997. See references for the Introduction
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
VALENCIA, R., N. PITMAN, S. LEÓN-YÁÑEZ, and P.M. JØR-
GENSEN. 2000. Libro Rojo de las Plantas Endémicas del
Ecuador 2000. Herbario QCA, Pontificia Universidad
Católica de Ecuador, Quito.
Atlantic Forest
AYRES, J.M., G.A.B. DA FONSECA, A.B. RYLANDS, H.L.
QUEIROZ, L.P. DE S. PINTO, D. MASTERSON, and R. CAV-
ALCANTI. 1997. Abordagens inovadoras para a conser-
vação da biodiversidade no Brasil: Os corredores das flo-
restas neotropicais, Vol. 1: Aspectos Gerais; Vol. 2:
Amazonia; Vol. 3: Mata Atlântica. Versão 2.0. PP/G7
–Programa Piloto para a Proteção das Florestas
Neotropicais: Projeto Parques e Reservas. Ministério
do Meio Ambiente, Recursos Hídricos e da Amazônia
Legal (MMA), Instituto Brasileiro do Meio Ambiente e
dos Recursos Naturais Renováveis (IBAMA), Brasília.
BARBOZA, F., J. PINAZZO, and F. FRACCHIA. 1997. Bosque
Atlántico Interior 1997. Mapa. Proyecto Sistema Am-
biental de la Región Oriental (SARO), Ministério de
Agricultura y Ganadería and World Wildlife Fund,
Asunción.
BENCKE, G.A. and G.M. MAURÍCIO. 2002. Important Bird
Area Program in Brazil. BirdLife International Brasil
Programme, Porto Alegre.
BROOKS, T., J. TOBIAS, and A. BALMFORD. 1999. “Deforesta-
tion and Bird Extinctions in the Atlantic Forest.” Ani-
mal Conservation 2:211-222.
BROWN JR., K.S. and A.V.L. FREITAS. 2000. “Atlantic Forest
Butterflies: Indicators for Landscape Conservation.”
Biotropica 32 (4b):934-956
CARTES, J.L. 2003. “Brief History of Conservation in the In-
terior Atlantic Forest.” In: C. GALINDO-LEAL and I. DE
G. CÂMARA (Eds.). The Atlantic Forest of South America:
Biodiversity Status, Threats and Outlook. Island Press,
Washington, D.C.
CHEBEZ, J.C. and N. HILGERT. 2003. “Brief History of Con-
servation in the Paraná Forest.” In: C. GALINDO-LEAL
and I. DE G. CÂMARA (Eds.). The Atlantic Forest of South
America: Biodiversity Status, Threats and Outlook. Is-
land Press, Washington, D.C.
COIMBRA-FILHO, A.F. and I. DE G. CÂMARA. 1996. Os limites
originais do bioma Mata Atlântica na região nordeste do
Brasil. Fundação Brasileira para a Conservaçao da Na-
tureza, Rio de Janeiro.
DEAN, W. 1995. With Broadax and Firebrand: The Destruc-
tion of the Brazilian Atlantic Forest. University of Cali-
fornia Press, Berkeley.
DA FONSECA, G.A.B. and L.P. DE S. PINTO. 1996. “O papel das
ONGs.” In: I.V. LOPES, G.S. BASTOS FILHO, D. BILLER,
and M. DALE (Eds.). Gestão ambiental no Brasil: Expe-
riencia e successo. Editora Fundação Getulio Vargas,
Rio de Janeiro.
JOLY, C.A., H.F. LEITÃO FILHO, and S.M. SILVA. 1991. “O
patrimônio florístico -The Floristic Heritage.” In: G.I.
CÂMARA (Coord.). Mata Atlântica -Atlantic Rain Forest.
Ed. Index Ltda. and Fundação S.O.S. Mata Atlântica,
São Paulo.
KLEIMAN, D.G. and A.B. RYLANDS (Eds.). 2002. Lion
Tamarins: Biology and Conservation. Smithsonian In-
stitution Press, Washington, D.C.
MORI, S.A., B.M. BOOM, and G.T. PRANCE. 1981. “Distribu-
tion Patterns and Conservation of Eastern Brazilian
Coastal Forest Tree Species.” Brittonia 33:233-245.
OLIVEIRA-FILHO, A.T. and M.A. FONTES. 2000. “Patterns of
Floristic Differentiation among Atlantic Forest in
Southeastern Brazil and the Influence of Climate.”
Biotropica 32:793-810.
RYLANDS, A.B., M.T. DA FONSECA, R.B. MACHADO, and R.B.
CAVALCANTI. 2004. “Brazil.” In: M. SPALDING, S. CHAPE,
and M. JENKINS (Eds.). In press. “The State of the
World’s Protected Areas, United Nations Environment
Programme (UNEP).” World Conservation Monitoring
Centre (WCMC), Cambridge.
STATTERSFIELD, A J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
STRIER, K.B. 1999. Faces in the Forest: The Endangered
Muriqui Monkeys of Brazil (2nd edition). Harvard Uni-
versity Press, Cambridge, U.S.A.
Cerrado
ALHO, C.J.R. and E.S. MARTINS. 1995. De grão em grão o
Cerrado perde espaço. WWF, Pró-Cer, Brasília.
BARBOSA, A.S. and I.V. NASCIMENTO. 1993. “Procesos cul-
turais associados à vegetação.” In: M.N. PINTO (Ed.).
Cerrado: Caracterização, ocupação e perspectivas. Edito-
ra Universidade de Brasília, Brasília.
BARBOSA, A.S., M.B. RIBEIRO, and P.I. SCHMITZ. 1993. “Cul-
tura e ambiente em áreas do sudoeste de Goiás.” In:
M.N. PINTO (Ed.). Cerrado: Caracterização, ocupação e
perspectivas. Editora Universidade de Brasília, Brasília.
COLLI, G.R. and D.R. ZAMBONI. 1999. Ecology of the Worm-
Lizard, Amphisbaena alba, in the Cerrado of central
Brazil. Copeia.
DIANESE, J.C., R.B. MEDEIROS, and L.T.P. SANTOS. 1997.
“Biodiversity of Microfungi Found on Native Plants of
the Brazilian Cerrado.” In: K.D. HYDE (Ed.). Biodiversi-
ty of Tropical Microfungi. Hong Kong University Press,
Hong Kong.
FUNDAÇÃO PRÓ-NATUREZA/FUNATURA, CONSERVATION INTER-
NATIONAL, FUNDAÇÃO BIODIVERSITAS, and UNIVERSIDADE
DE BRASÍLIA. 2000. Avaliação e Ações Prioritárias para a
Conservação da Biodiversidade do Cerrado e Pantanal.
Secretaria de Biodiversidade e Florestas (SBF)/Min-
istério do Meio Ambiente (MMA), Brasília.
GOODLAND, R.J.A. and H.S. IRWIN. 1975. Amazon Jungle;
Green Hell to Red Desert: An Ecological Discussion of the
Environmental Impact of the Highway Construction Pro-
gram of the Amazon Basin. Elsevier Scientific Publish-
ing Co., Amsterdam.
HERINGER, E.P., G.M. BARROSO, J.A. RIZZO, and C.T. RIZZINI.
1977. “A flora do Cerrado.” In: Simpósio sobre o Cerra-
do, Vol. 4. Universidade de São Paulo, Itaiaia e São
Paulo.
HOLANDA, S.B. 1936. Raizes do Brasil. Companhia das Le-
tras, 3rd edition (1997).
MANTOVANI, J.E. and L.A. PEREIRA. 1998. “Estimativa da in-
tegridade da cobertura vegetal do Cerrado/Pantanal
através de dados TM/Landsat.” Unpublished report
for the Workshop “Ações Prioritárias para a Conser-
vação do Cerrado e Pantanal”, Brasília, Brazil. Funatu-
ra, Conservation International, Universidade de
Brasília, Fundação Biodiversitas, Brasília.
MENDONÇA, R.C. DE, J.M. FELFILI, B.M.T. WALTER, M.C.
DA SILVA JÚNIOR, A.V. REZENDE, T.S. FILGUEIRAS, and
P.E. OLIVEIRA. 1997. “Flora Vascular do Cerrado.” In:
S.M. SANO and P. ALMEIDA (Eds.). Cerrado: Flora,
Homem e Ambiente. EMBRAPA - Cerrados, Planalti-
na, Brazil.
RAW, A. 1998. “Número de insetos, a riqueza de espécies e
aspectos zoogeográficos nos cerrados.” Unpublished
report for the Workshop “Ações Prioritárias para a
Conservação do Cerrado e Pantanal”, Brasília, Brazil.
Funatura, Conservation International, Universidade
de Brasília, Fundação Biodiversitas, Brasília.
REDFORD, K.H. and G.A.B. DA FONSECA. 1986. “The Role of
Gallery Forests in the Zoogeography of the Cerrado’s
Non-Volant Mammalian Fauna.” Biotropica 18:125-135.
RYLANDS, A.B., M.T. DA FONSECA, R.B. MACHADO, and R.B.
CAVALCANTI. 2004. “Brazil.” In: M. SPALDING, S. CHAPE
and M. JENKINS (Eds.). The State of the World’s Protect-
ed Areas. United Nations Environment Programme
(UNEP), World Conservation Monitoring Centre
(WCMC), Cambridge.
SARMIENTO, G. 1983. “The Savannas of Tropical America.”
In: F. BOURLIÈRE (Ed.). Tropical Savannas. Ecosystems
of the World Series 13. Elsevier, New York.
SCHMITZ, P.I. 1993. “Caçadores e coletores antigos.” In:
M.N. PINTO (Ed.). Cerrado: Caracterização, ocupação e
perspectivas. Editora Universidade de Brasília, Brasília.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
Chilean Winter Rainfall-Valdivian Forests
ARMESTO, J.J., R. ROZZ, C. SMITH-RAMÍREZ, and M.T.K. AR-
ROYO. 1998. “Conservation Targets in South American
Temperate Forests.” Science 282:1271-1272.
ARROYO, M.T.K., C. MARTICORENA, O. MATTHEI, and L.
CAVIERES. 2000. “Plant Invasions in Chile: Present Pat-
terns and Future Predictions.” In: H.A. MOONEY and
R. HOBBS (Eds.). Invasive Species in a Changing World.
Island Press, New York.
ARROYO, M.T.K., C. MARTICORENA, O. MATTHEI, M. MUÑOZ,
and P. PLISCOFF. 2002. “Analysis of the Contribution
and Efficiency of the Santuario de Naturaleza Yerba
Loca, 33ºS in Protecting the Vascular Flora of the Met-
ropolitan Region of Chile.” Revista Chilena de Historia
Natural 75:767-792.
ARROYO, M.T.K., J.R. ROZZI, J. SIMONETTI, J.A MARQUET,
and M. SALABERRY. 1999. “Central Chile.” In: R.A.
MITTERMEIER, N. MYERS, P. ROBLES-GIL, and C.G. MIT-
TERMEIER (Eds.). 1999. See references for the Intro-
duction.
ARROYO, M.T.K. and L. CAVIERES. 1997. “The Mediter-
ranean-Type Climate Flora of Central Chile: What Do
We Know and How Can We Assure its Protection.”
Noticiero de Biología 5(2):48-56.
ARROYO, M.T.K., L. CAVIERES, C. MARTICORENA, and M.
MUÑOZ. 1995. “Convergence in the Mediterranean
Floras of Central Chile and California: Insights from
Comparative Biogeography.” In: M.T.K. ARROYO, M.D.
FOX, and P. ZEDLER (Eds.). Ecology and Biogeography of
Mediterranean Ecosystems in Chile, California and Aus-
tralia. Springer-Verlag, New York.
ARROYO, M.T.K., M. RIVEROS, A. PEÑALOZA, L. CAVIERES,
and A.M. FAGGI. 1996. “Phytogeographic Relationships
and Regional Richness Patterns of the Cool Temperate
Rainforest Flora of Southern South America.” In: R.G.
LAWFORD, P.B. ALABACK, and E. FUENTES (Eds.). High-
Latitude Rainforests and Associated Ecosystems of the
West Coasts of the Americas. Climate, Hydrology, Ecolo-
gy and Conservation. Springer Verlag, New York.
BREMER, K. and M.H.G. GUSTAFSSON. 1997. “East Gond-
wanaland Ancestry of the Sunflower Alliance of Fam-
ilies.” Proceedings of the National Academy of Sciences
94:9188-9190.
CONAF-CONAMA-BIRF. 1999. Catastro y evaluación de los
recursos vegetacionales nativos de Chile. Informe na-
cional con variables ambientales, Vol. N. Santiago.
FUENTES, E.R., R. AVILES, and A. SEGURO. 1990. “The Nat-
ural Vegetation of a Heavily Man-Transformed Land-
scape: The Savanna of Central Chile.” Interciencia
15:293-295.
HOFFMANN, A., M.T.K. ARROYO, F. LIBERONA, M. MUÑOZ, and
J. WATSON. 1998. Plantas altoandinas en la flora silvestre
de Chile. Ediciones Fundación Claudio Gay, Santiago.
MARTICORENA, C., T.F. STEUSSY, and C.M. BAEZA. 1998. “Cat-
alogo de la flora vascular de las islas Robison Crusoe o
Juan Fernández, Chile.” Gayana, Bot. 55:2-2
MITTERMEIER, R.A., N. MYERS, P. ROBLES-GIL, and C.G. MIT-
TERMEIER (Eds.). 1999. See references for the Intro-
duction.
MUÑOZ, S., M.H. NÚÑEZ, and J. YÁÑEZ (Eds.). 1996. Libro rojo
de los sitios prioritarios para la conservación de la diversi-
dad biológica. Corporación Nacional Forestal, Santiago.
SQUEO, F.A., G. ARANCIO, and J.R. GUTIÉRREZ. 2001. (Eds.)
Libro rojo de los sitios prioritarios para su conservación:
region de coquimbo. Ediciones University La Serena, La
Serena.
VILLAGRÁN, C. and L.F. HINOJOSA. 1997. “Historia de los
bosques del sur de Sudamérica, II: análisis fitogeográ-
fico.” Revista Chilena de Historia Natural 70:241-267.
Mesoamerica
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
Caribbean Islands
BEARD, J.S. 1949. Natural Vegetation of the Windward and
Leeward Islands. Clarendon Press, Oxford.
BEARD, J.S. 1955. “The Classification of Tropical American
Vegetation Types.” Ecology 36:89-100.
DAVIS, S.D., V.H. HEYWOOD, O. HERRERA-MACBRYDE, J. VIL-
LA-LOBOS, and A.C. HAMILTON (Eds.). 1997. See refer-
ences for the Introduction.
DELGADILLO, C., B. BELLO, and A. CARDENAS. 1995. LAT-
MOSS: “A Catalogue of Neotropical Mosses.” Mono-
graphs in Systematic Botany from the Missouri Botan-
ical Garden 56:1-191.
GRAHAM, A. 2003. “Historical Phytogeography of the
Greater Antilles.” Brittonia 55(4):357-383.
RODRIGUES, A.S.L., S.J. ANDELMAN, M.I. BAKARR, L. BOI-
TANI, T.M. BROOKS, R.M. COWLING, L.D.C. FISHPOOL,
G.A.B. DA FONSECA, K.J. GASTON, M. HOFFMANN, J.
LONG, P.A. MARQUET, J.D. PILGRIM, R.L. PRESSEY, J.
SCHIPPER, W. SECHREST, S.N. STUART, L.G. UNDERHILL,
R.W. WALLER, M.E.J. WATTS, and Y. XIE. 2003. “Global
Gap Analysis: Towards a Representative Network of
Protected Areas.” In: Advances in Applied Biodiversity
Science. Conservation International, Washington, D.C.
California Floristic Province
BARBOUR, M., B. PAVLIK, F. DRYSDALE, and S. LINDSTROM.
1993. California’s Changing Landscapes: Diversity and
Conservation of California Vegetation. California Native
Plant Society, Sacramento.
DALLMANN, P.R. 1998. Plant Life in the World’s Mediter-
ranean Climates. University of California Press, Berke-
ley.
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1997. See references for the Introduction.
GROSSMAN, D.H., K.L. GOODIN, and C.L. REUSS (Eds.). 1994.
Rare Plant Communities of the Conterminous United
States: An Initial Survey. The Nature Conservancy, Ar-
lington.
HOLLAND, R.F. and S.K. JAIN. 1990. “Vernal Pools.” In: M.
BARBOUR and J. MAJOR (Eds.). Terrestrial Vegetation of
California. California Native Plant Society, Sacramento.
JENSEN, D.B., M.S. TORN, and J. HARTE. 1993. In Our Own
Hands: A Strategy for Conserving California’s Biodiversi-
ty. University of California Press, Berkeley, California.
KRUCKEBERG, A. 1984. California Serpentines: Flora, Vegeta-
tion, Geology, Soils, and Management Problems. Univer-
sity of California Press, Berkeley.
LOARIE, S.R. and D.D. ACKERLY. 2004. “The Distribution of
California’s Endangered Plants in Multivariable Cli-
mate Space.” 6th Annual Bay Area Conservation Biolo-
gy Symposium.
NOSS, R.F. 1994. “California’s Ecosystem Decline.” Defend-
ers 69(4):34-35.
NOSS, R.F. and R.L. PETERS. 1995. Endangered Ecosystems: A
Status Report on America’s Vanishing Habitat and
Wildlife. Defenders of Wildlife, Washington, D.C.
RAVEN, P.H. 1988. “The California Flora.” In: M.G. BARBOUR
and J. MAJOR (Eds.). Terrestrial Vegetation of California.
California Native Plant Society, Sacramento.
RAVEN, P.H. and D.I. AXELROD. 1978. “Origin and Relation-
ships of the California Flora.” University of California
Publications in Botany 72:1-134.
SEABLOOM, E.W., A.P. DOBSON, and D.M. STOMS. 2002. “Ex-
tinction Rates under Nonrandom Habitat Loss.” Pro-
ceedings of the National Academy of Sciences 99:11229-
11234.
STEBBINS, G.L. 1978. “Why Are There so Many Rare Plants
in California?” I. Environmental factors. Fremontia
5(4):6-10.
STEBBINS, G. L. and J. MAJOR. 1965. “Endemism and Speci-
ation in the California Flora.” Ecol. Monographs 35:1-
35.
STEINHART, P. 1994. “California’s Biodiversity Experiment.”
Defenders 69(4):11-22.
TURNER, T. 1991. Sierra Club: 100 Years of Protecting Nature.
Harry N. Abrams, Inc., New York.
U.S. CENSUS BUREAU. “State and County Quickfacts.”
http://www.census.gov
VANCE-BORLAND, K., R. NOSS, J. STRITTHOLT, P. FROST, C.
CARROLL, and R. NAWA. 1995/96. “A Biodiversity Con-
servation Plan for the Klamath/Siskiyou Region: A
Progress Report on a Case Study for Bioregional Con-
servation.” Wild Earth 5(4):52-59.
WIRKA, J. 1994. “California’s Missing Bruins.” Defenders
69(4):23-25.
Guinean Forests of West Africa
ANONYMOUS. 2003. The Usual Suspects: Liberia’s Weapons
and Mercenaries in Côte d’Ivoire and Sierra Leone. Glob-
al Witness Report, London. www.globalwitness.org
BAKARR, M., B. BAILEY, D. BYLER, R. HAM, S. OLIVIERI, and
M. OMLAND (Eds.). 2001. From the Forest to the Sea: Bio-
diversity Connections from Guinea to Togo. Conserva-
tion International, Washington, D.C.
BRASHARES, J.S. 2003. “Ecological, Behavioral, and Life-
History Correlates of Mammal Extinctions in West
Africa.” Conservation Biology 17:733-743.
BUTYNSKI, T.M. and S.H. KOSTER. 1990. The Status and Con-
servation of Forests and Primates on Bioko Island (Fer-
nando Poo), Equatorial Guinea. WWF, Washington, D.C.
BUTYNSKI, T.M. and S.H. KOSTER. 1994. “Distribution and
Conservation Status of primates in Bioko Island, Equa-
torial Guinea.” Biodiversity and Conservation 3:893-909.
CABLE, S. and M. CHEEK 1998. The Plants of Mount Cameroon:
A Conservation Checklist. Royal Botanic Gardens, Kew,
U.K.
DAVIS, S.C., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1994. See references for the Introduction.
FA, J.E., C. PERES, and J. MEEUWIG. 2002a. “Bushmeat Ex-
ploitation in Tropical Forests: An Intercontinental
Comparison.” Conservation Biology 16(1):232-237
FA, J.E., J.E. GARCIA-YUSTE, and R. CASTELO. 2000. “Bush-
meat Markets on Bioko Island as a Measure of Hunt-
ing Pressure.” Conservation Biology 14:1602-1613.
GRUBB, P., T.S. JONES, A.G. DAVIES, E. EDBERG, E.D. STARIN,
and J.E. HILL. 1998. Mammals of Ghana, Sierra Leone
and the Gambia. The Trendrine Press, Cornwall, U.K.
IUCN. 1996. African Primates: Status Survey and Conserva-
tion Action Plan (revised edition). IUCN/SSC Primate
Specialist Group. IUCN, Gland.
JONES, P.J. 1994. “Biodiversity in the Gulf of Guinea: An
Overview.” Biodiversity and Conservation 3:772-784.
JUSTE, J. and J.E. FA. 1994. “Biodiversity Conservation in
the Gulf of Guinea Islands: Taking Stock and Prepar-
ing Action.” Biodiversity and Conservation 3:759-771.
LAMOTTE, M. and R. ROY (Eds.). 2003. “Le peuplement an-
imal du Mont Nimba (Guinée, Côte d’Ivoire, Liberia).”
Mémoires du Museum National d’Histoire Naturelle
190:1-726.
MARTIN, C. 1991. The Rainforests of West Africa: Ecology,
Threats and Conservation. Birkhäuser Verlag, Basel,
Switzerland.
OATES, J.F. 1986. Action Plan for African Primate Conserva-
tion 1986-1990. IUCN/SSC Primate Specialist Group.
IUCN, Gland.
OATES, J.F. 1999. Myth and Reality in the Rainforest: How
Conservation Strategies are Failing in West Africa. Uni-
versity of California Press, Berkeley.
OATES, J.F., M. ABEDI-LARTEY, W.S. MCGRAW, T.T. STRUHSAK-
ER, and G.H. WHITESIDES. 2000. “Extinction of a West
African Red Colobus Monkey.” Conservation Biology
14:1526-1532.
PARREN, M.P.E., B.M. DE LEEDE, and F. BONGERS. 2002. “A
Proposal for a Transnational Forest Network Area for
Elephants in Côte d’Ivoire and Ghana.” Oryx 36:249-
256.
PARREN, M.P.E. and N.R. DE GRAAF. 1995. The Quest for Nat-
ural Forest Management in Ghana, Côte d’Ivoire and
Liberia. Tropenbos Series 13. Tropenbos Foundation,
Wageningen, The Netherlands.
POORTER, L., F. BONGERS, F.N. KOUAME, and W.D. HAWTHORNE
(Eds.). 2004. Biodiversity of West African Forests: An
Ecological Atlas of Woody Plant Species. CABI Publish-
ing, Oxon, U.K.
RÖDEL, M.O. and A.C. AGYEI. 2003. “Amphibians of the
Togo-Volta Highlands, Eastern Ghana.” Salamandra
39:207-234.
SAYER, J.A., C.S. HARCOURT, and N.M. COLLINS (Eds.) 1992.
The Conservation Atlas of Tropical Forests: Africa.
MacMillan, London.
SMITH, T.B., K. HOLDER, D. GIRMAN, K. O’KEEFE, B. LARISON,
and Y. CHAN. 2000. “Comparative Avian Phylogeogra-
phy of Cameroon and Equatorial Guinea Mountains:
Implications for Conservation.” Molecular Ecology
9:1505-1516.
WHITE, F. 1983. See references for the Introduction.
WILLIAMS, P.H., J.L. MOORE, A.K. TOHAM, T.M. BROOKS, H.
STRAND, J. D’AMICO, M. WISZ, N.D. BURGESS, A. BALM-
FORD, and C. RAHBEK. 2003. “Integrating Biodiversity
Priorities with Conflicting Socio-Economic Values in
the Guinean-Congolian Forest Region.” Biodiversity
and Conservation 12:1297-1320.
Cape Floristic Region
BAARD, E.H.W. 1997. “A Conservation Strategy for the Geo-
metric Tortoise, Psammobates geometricus.” In: Pro-
ceedings: Conservation, Restoration, and Management of
Tortoises and Turtles-An International Conference. The
New York Turtle and Tortoise Society, New York.
COWLING, R.M. and D.M. RICHARDSON. 1995. Fynbos: South
Africa’s Unique Floral Kingdom. Fernwood Press, Cape
Town.
COWLING, R.M., D.M. RICHARDSON, and P.J. MUSTART. 1997.
“Fynbos.” In: R.M. COWLING (Ed.). The Ecology of Fyn-
bos: Nutrients, Fire and Diversity. Oxford University
Press, Cape Town.
COWLING, R.M., D.M. RICHARDSON, and S.M. PIERCE (Eds.).
The Vegetation of Southern Africa. Cambridge Universi-
ty Press, Cambridge.
COWLING, R.M. and R.L. PRESSEY. 2003. “Introduction to
Systematic Conservation Planning in the Cape Floris-
tic Region.” Biological Conservation 122:1-13.
JOHNSON, S.D. 1992. “Plant-Animal Relationships.” In: R.M.
COWLING (Ed.). The Ecology of Fynbos: Nutrients, Fire
and Diversity. Oxford University Press, Cape Town.
KRUGER, F.J. 1979. “South African Heathlands.” In: R.L.
SPECHT (Ed.). Heathlands and related shrublands. Else-
vier, New York.
LINDER, H.P., M.E. MEADOWS, and R.M. COWLING. 1992.
“History of the Cape Flora.” In: R.M. COWLING (Ed.).
The Ecology of Fynbos: Nutrients, Fire and Diversity. Ox-
ford University Press, Cape Town.
MCMAHON, L. and M. FRASER. 1988. A Fynbos Year. David
Phillip, Cape Town and Johannesburg.
PICKER, M.D. and M.J. SAMWAYS. 1996. “Faunal Diversity
and Endemism of the Cape Peninsula, South Africa.”
Biodiversity and Conservation 5(5):591-606.
REBELO, A.G. 1992. “Preservation of biotic diversity.” In:
R.M. COWLING (Ed.). The Ecology of Fynbos: Nutrients,
Fire and Diversity. Oxford University Press, Cape
Town.
ROUGET, M., D.M. RICHARDSON, R.M. COWLING, J.W. LLOYD,
and A.T. LOMBARD. 2003a. “Current Patterns of Habitat
Transformation and Future Threats to Biodiversity in
Terrestrial Ecosystems of the Cape Floristic Region.”
Biological Conservation 112, 63-85.
ROUGET, M., D.M. RICHARDSON, and R.M. COWLING. 2003b.
“The Current Configuration of Protected Areas in the
Cape Floristic Region, South Africa: Reservation Bias
and Representation of Biodiversity Patterns and
Processes.” Biological Conservation 112:129-145.
SKELTON, P. 1993. A Complete Guide to the Freshwater Fish of
Southern Africa. Southern Book Publishers: Halfway
House
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
VLOK, J.H.J., D.I.W. EUSTON-BROWN, and R.M. COWLING.
2003. See General References.
Succulent Karoo
DAVIS, S.C., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1994. See references for the Introduction.
DEAN, W.R.J. and S.J. MILTON. 1999. “Animal Foraging and
Food.” In: W.R.J. DEAN and S.J. MILTON (Eds.). The Ka-
roo: Ecological Patterns and Processes. Cambridge Uni-
versity Press, Cambridge.
DESMET, P., A. ELLIS, and R.M. COWLING. 1998. “Speciation
in the Mesembryanthemaceae.” Aloe 35(2):38-43.
DRIVER, A., P.G. DESMET, M. ROUGET, R.M. COWLING, and
K.E. MAZE. 2003. Succulent Karoo Ecosystem Plan Bio-
diversity Component Technical Report. Cape Conserva-
tion Unit, Botanical Society of South Africa, Cape
Town.
383
HARRISON, J.A., D.G. ALLAN, L.G. UNDERHILL, M. HERRE-
MANS, A.J. TREE, V. PARKER, and C.J. BROWN (Eds.).
1997. The Atlas of Southern African Birds. BirdLife, Jo-
hannesburg.
HOFFMAN, M.T., B. COUSINS, T. MEYER, A. PETERSEN, and H.
HENDRICKS. 1999. “Historical and Contemporary Land
Use and the Desertification of the Karoo.” In: W.R.J.
DEAN and S.J MILTON (Eds.). The Karoo: Ecological Pat-
terns and Processes. Cambridge University Press, Cam-
bridge.
IHLENFELDT, H.D. 1994. “Diversification in an Arid World:
The Mesembryanthemaceae.” Annual Review of Ecolo-
gy and Systematics 25:521-546.
VAN JAARSVELD, E. 1987. “The Succulent Riches of South
Africa and Namibia.” Aloe 24:45-92.
JÜRGENS, N. 1986. “Untersuchungen zur okologie sukku-
lenter pflanzen des sudlichen Afrika.” Mitteilungen
aus dem Institut für Allgemeine Botanik Hamburg
21:139-365.
JÜRGENS, N. 1991. “A New Approach to the Namib Region.”
I: Phytogeographic subdivision. Vegetation 97:21-38.
KLAK, C., G. REEVES, and T. HEDDERSON. 2004. “Unmatched
Tempo of Evolution in Southern African Semi-Desert
Ice Plants.” Nature 427:63-65.
LOMBARD, A.T., C. HILTON-TAYLOR, A.G. REBELO, R.L.
PRESSEY, and R.M. COWLING. 1999. “Reserve Selection
in the Succulent Karoo, South Africa: Coping with
High Compositional Turnover.” Plant Ecology 142(1-
2):35-55.
MILTON, S.J., R. YEATON, W.R.J. DEAN, and J.H.J. VLOK.
1997. “Succulent Karoo.” In: R.M. COWLING, D.M.
RICHARDSON, and S.M. PIERCE (Eds.). Vegetation of
Southern Africa. Cambridge University Press, Cam-
bridge.
RUNDEL, P.W., R.M. COWLING, K.J. ESLER., P.J. MUSTART, E.
VAN JAARSVELD, and H. BEZUIDENHOUT. 1995. “Winter
Growth Phenology and Leaf Orientation in Pachypodi-
um namaquanum (Apocynaceae) in the Succulent Ka-
roo of the Richtersveld, South Africa.” Oecologia
101:472-477.
VERNON, C.J. 1999. “Biogeography, Endemism and Diver-
sity of Animals in the Karoo.” In: W.R.J. DEAN and S.J.
MILTON (Eds.). The Karoo: Ecological Patterns and
Processes. Cambridge University Press, Cambridge.
WHITE, F. 1983. See references for the Introduction.
Madagascar and the Indian Ocean Islands
EGER, J.L. and L. MITCHELL. 2003. “Chiroptera, Bats.” In:
S.M. GOODMAN and J.P. BENSTEAD (Eds.). The Natural
History of Madagascar. The University of Chicago
Press, Chicago.
GOODMAN, S.M. and J.P. BENSTEAD (Eds.). 2003. The Natur-
al History of Madagascar. The University of Chicago
Press, Chicago.
GOODMAN, S.M., J.U. GANZHORN, and D. RAKOTONDRAVONY.
2003. “Introduction to the Mammals.” In: S.M.
GOODMAN and J.P. BENSTEAD (Eds.). The Natural His-
tory of Madagascar. The University of Chicago
Press, Chicago.
HAWKINS, A.F.A. and S.M. GOODMAN. 2003. “Introduction to
the Birds.” In: S.M. GOODMAN and J.P. BENSTEAD
(Eds.). The Natural History of Madagascar. The Uni-
versity of Chicago Press, Chicago.
JOLLY, R. and S. FUKUDA-PARR. 2000. Human Development
Report 2000. United Nations Development Program-
me. Oxford University Press, New York.
LANGRAND, O. and S.M. GOODMAN. 1995. “Monitoring
Madagascar’s Ecosystems: A Look at the Past, Present,
and Future of its Wetlands.” In: T.B. HERMAN, S. BON-
DRUP-NIELSEN, J.H.M. WILLISON, and N.W.P. MUNRO
(Eds.). Ecosystem Monitoring and Protected Areas. Sci-
ence and Management of Protected Areas Associa-
tion, Wolfville, Canada.
LOWRY II, P.P., G.E. SCHATZ, and P.B. PHILLIPSON. 1997. “The
Classification of Natural and Anthropogenic Vegeta-
tion in Madagascar.” In: S.M. GOODMAN and B.D. PAT-
TERSON (Eds.). Natural Change and Human Impact in
Madagascar. Smithsonian Inst. Press, Washington,
D.C.
PASCAL, O. 2002. Plantes et forêts de Mayotte. Muséum Na-
tional d’Histoire Naturelle, Paris.
PERRIER DE LA BATHIE, H. 1936. Biogéographie des plantes de
Madagascar. Société d’Editions Géographiques, Mar-
itimes et Coloniales, Paris.
RANDRIANANDIANINA, B.N., L.R. ANDRIAMAHALY, F.M.
HARISOA, and M.E. NICOLL. 2003. “The Role of the Pro-
tected Areas in the Management of the Island’s Biodi-
versity.” In: S.M. GOODMAN and J.P. BENSTEAD (Eds.).
384
The Natural History of Madagascar. The University of
Chicago Press, Chicago.
RAXWORTHY, C.J. 2003. “Introduction to the Reptiles.” In:
S.M. GOODMAN and J.P. BENSTEAD (Eds.). The Natural
History of Madagascar. The University of Chicago
Press, Chicago.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
STEININGER, M. et al. 2004. “Forest Cover Fragmentation and
Clearance in Madagascar.” Unpublished digital map
available from: ftp://65.205.36.42/Regional_Analysis/
SpatialData/mad_final.img
Mediterranean Basin
BLONDEL, J. and J. ARONSON. 1999. Biology and Wildlife of
the Mediterranean Region. Oxford University Press, Ox-
ford.
DI CASTRI, F. and H.A. MOONEY (Eds.). 1973. “Mediter-
ranean-Type Ecosystems.” In: Ecological Studies, Vol. 7.
Springer-Verlag, Berlin, Heidelberg and New York.
GREUTER, W. 1991. “Botanical Diversity, Endemism, Rarity,
and Extinction in the Mediterranean Area: An Analy-
sis Based on the Published Volumes of Med-Check-
list.” Botanika Chronika 10:63-79
MÉDAIL, F. and P. QUÉZEL. 1997. “Hot-Spots Analysis for
Conservation of Plant Biodiversity in the Mediter-
ranean Basin.” Annals of the Missouri Botanical Garden
84:112-127
MÉDAIL, F. and P. QUÉZEL. 1999. “Biodiversity Hotspots in
the Mediterranean Basin: Setting Global Conservation
Priorities.” Conservation Biology 13:1510-1513.
MÉDAIL, F. and R. VERLAQUE. 1997. “Ecological Character-
istics and Rarity of Endemic Plants from Southeast
France and Corsica: Implications for Biodiversity Con-
servation.” Biological Conservation 80:269-281.
MYERS, N. and R.M. COWLING. 1999. “Mediterranean
Basin.” In: R.A. MITTERMEIER, N. MYERS, P. ROBLES-GIL,
and C.G. MITTERMEIER (Eds.). 1999. See references for
the Introduction.
NAVEH, Z. and P. KUTIEL. 1990. “Changes in the Mediter-
ranean Vegetation of Israel in Response to Human
Habitation and Land Use.” In: G.M. WOODWELL (Ed.).
The Earth in Transition: Patterns and Processes of Biotic
Impoverishment. Cambridge University Press, New
York.
POPULATION REFERENCE BUREAU. 2003. See General Refer-
ences.
QUÉZEL, P. 1985. “Definition of the Mediterranean Region
and the Origin of its Flora.” In: C. GÓMEZ-CAMPO (Ed.).
Plant Conservation in the Mediterranean Area. Dr. W.
Junk Publishers, Dordrecht.
QUÉZEL P. and F. MÉDAIL. 2003. Ecologie et biogéographie des
forêts du bassin Méditerranéen. Elsevier, Collection En-
vironnement, Paris.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
SULAYEM, M. 1994. “North Africa and the Middle East.” In:
J.A. MCNEELY, J. HARRISON, and P. DINGWALL (Eds.).
Protecting Nature: Regional Reviews of Protected Areas.
IUCN, Gland.
VERDÚ M., P. DÁVILA, P. GARCÍA-FAYOS, N. FLORES-HERNÁN-
DEZ, and A. VALIENTE-BANUET. 2003. “‘Convergent’
Traits of Mediterranean Woody Plants Belong to Pre-
Mediterranean Lineages. Biological Journal of the Lin-
nean Society 78:415-427.
Caucasus
BAILLIE, J. and B. GROOMBRIDGE (Eds.). 1996. 1996 IUCN
Red List of Threatened Animals. IUCN-The World Con-
servation Union/Species Survival Commission,
Gland, Switzerland.
BIRSTEIN, V.J. 1993. “Sturgeons and Paddlefishes: Threat-
ened Fishes in Need of Conservation.” Conservation
Biology 7:773-787.
DOLUKHANOV, A.G. 1966. “Vegetation Cover.” In: I.P.
GERASIMOV. The Caucasus. Nauka, Moscow.
FRIMODT, C. 1995. Multilingual Illustrated Guide to the
World’s Commercial Coldwater Fish. Fishing News
Books, Osney Mead, Oxford.
GAGNIDZE, R., T.S. GVINIASHUILI, S.H. SHETEKAURI, and N.
MARGALITADZE. 2002. “Endemic Genera of the Cauca-
sus Flora.” Feddes Repertorium 113(7-8):616-630.
HARDISTY, M.W. 1986. “A General Introduction to Lam-
preys.” In: J. HOLCÍK (Ed.) The Freshwater Fishes of Eu-
rope, Vol. 1 (part 1): Petromyzontiformes.
KREVER, V., N. ZAZANASHVILI, H. JUNGIUS, L. WILLIAMS, and
D. PETELIN. 2001. See references for the Introduction.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
TARKHNISHVILI, D.N, R.S. THORPE, and J.W. ARNTZEN. 2000.
“Pre-Pleistocene Refugia and Differentiation between
Populations of the Caucasian Salamander (Mertensiel-
la caucasica).” Molecular Phylogenetics and Evolution
14:414-422.
WEINBERG, P.I. 1997. “Information about Present Situation
of the Mountain Goats in the CIS Territory.” Abstracts
of International Conference “Rare Species of the
Mammals in Russia and Adjoining Territories” (9-11
April, 1997, Moscow): 238.
WWF. Technical Reports of Project (2002-2004: Leopard Con-
servation in the Caucasus Ecoregion. Prepared by Victor
Lukarevski. Archive of WWF Caucasus Programme
Office.
WWF INTERNATIONAL 1997. “100 European Forests We
Should Protect Now.” Map. WCMC.
ZAZANASHVILI, N., G. SANADIRADZE, and A. BUKHNIKASHVILI.
1999. “Caucasus.” In: R.A. MITTERMEIER, N. MYERS, P.
ROBLES-GIL, and C.G. MITTERMEIER (Eds.). 1999. See
references for the Introduction.
Western Ghats and Sri Lanka
BIJU, S.D. and F. BOSSUYT. 2003. “New Frog Family from In-
dia Reveals an Ancient Biogeographical Link with the
Seychelles.” Nature 425:711-714.
CINCOTTA, R.P., J. WISNEWSKI, and R. ENGELMAN. 2000. “Hu-
man Population in the Biodiversity Hotspots.” Nature
404:990-992.
COLLINS, N.M., J.A. SAYER, and T.C. WHITMORE. 1991. The
Conservation Atlas of Tropical Forest: Asia and the Pacif-
ic. IUCN.
CONTI, E., F. RUTSCHMANN, and T. ERIKSSON. 2003. “Testing
the Out-of-India Origin of Crypteroniaceae with Mol-
ecular Dating Methods” [abstract] p. 37. In: Proceedings
of the Annual Meeting of the Association for Tropical Bi-
ology and Conservation. University of Aberdeen, U.K.
D’ABRERA, B. 1998. The Butterflies of Ceylon. WHT Publica-
tions, Colombo.
DAHANUKAR, N., R. RAUT, and A. BHAT. 2004. “Distribution,
Endemism and Threat Status of Freshwater Fishes in
the Western Ghats of India.” Journal of Biogeography
31:123-136.
DANIELS, R.J.R. 2001. “Endemic Fishes of the Western
Ghats and the Satpura Hypothesis.” Current Science
81:241-244.
DASSANAYAKE, M.D. and F.R. FOSSBERG. 1980-2002. A Re-
vised Handbook to the Flora of Ceylon. Vols. 1-9 edited
by M.D. DASSANAYAKE, F.R. FOSSBERG, and W.D. CLAY-
TON; Vols. 10-14 edited by M.D. DASSANAYAKE and W.D.
CLAYTON. Oxford and IBH Publishing Co., New Delhi.
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1995. See references for the Introduction.
GUNASEKERA, N.S. and S. DAYANANDAN. 2003. “Diptero-
carpaceae as a Model for Understanding Evolution in
Tropical Trees” [abstract] p. 37. In: Proceedings of the
Annual Meeting of the Association for Tropical Biology
and Conservation. University of Aberdeen, U.K.
JAYASURIYA, A.K. and J.F.A. TRANIELLO. 1985. “The Biology
of the Primitive Ant Aneuretus simoni (Emery), 1: Dis-
tribution, Abundance, Colony Structure and Foraging
Ecology.” Insectes Sociaux 32:363-374.
KROMBEIN, K.V. and B.B. NORDEN. 1997. “Bizarre Nesting
Behavior of Krombeinictus nordenae Leclercq (Hy-
menoptera: Sphecidae, Crabroninae).” J. South Asian
Nat. Hist. 2(2):145-154.
KUNTE, K., A. JOGLEKAR, U. GHATE, and P. PRAMOD. 1999.
“Patterns of Butterfly, Birds and Tree Diversity in the
Western Ghats.” Current Science 77(4):577-586
MEEGASKUMBURA, M., F. BOSSUYT, R. PETHIYAGODA, K. MAN-
AMENDRA-ARACHCHI, M.C. MILINKOVITCH, and C.J.
SCHNEIDER. 2002. “Sri Lanka: An Amphibian Hotspot.”
Science 298:379.
PASCAL, J.-P. 1988. Wet Evergreen Forests of the Western
Ghats of India: Ecology, Structure, Floristic Composition
and Succession. French Institute, Pondicherry, India.
PETHIYAGODA, R. 1994. “Threats to the Indigenous Fresh-
water Fishes of Sri Lanka and Remarks on their Con-
servation.” Hydrobiologia 285:189-201.
POPULATION REFERENCE BUREAU. 2003. See General Refer-
ences.
RAMESH, B.R. and J.P. PASCAL. 1991. “Distribution of En-
demic, Arbsorescent Evergreen Species in the West-
ern Ghats.” Proceedings of Symposium on rare and
endangered plants of the Western Ghats Kerala Forest
Department. Trivandrum 20-29.
RODGERS, W.A., H.S. PANWAR, and V.B. MATHUR. 2002.
Wildlife Protected Area Network in India: A review.
Wildlife Institute of India, Dehradun.
ROELANTS, K., J. JIANG, and F. BOSSUYT. In press. “Endemic
Ranid (Amphibia: Anura) Genera in Southern Moun-
tain Ranges of the Indian Subcontinent Represent An-
cient Frog Lineages: Evidence from Molecular Data.”
Molecular Phylogenetics and Evolution.
SIDDALL, M., E.J. ROHLING, A. ALMOGI-LABIN et al. 2003.
“Sea-Level Fluctuations During the Last Glacial Cy-
cle.” Nature 423:853-858.
STATTERSFIELD, A.J., M.J. CROSBY, A.L. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
WIKRAMANAYAKE, E.D., E. DINERSTEIN, C. LOUCKS, D. OL-
SON, J. MORRISON, J. LAMOREUX, M. MCKNIGHT, and P.
HEDAO. 2002. See General References.
Mountains of Southwest China
LEI, F.M., Y.H. QU, J.L. LU, Y. LIU, and Z.H. YIN, 2003.
“Conservation on Diversity and Distribution Patterns
of Endemic Birds in China.” Biodiversity and Conserva-
tion 12:239-254.
MACKINNON, J. 1996. Wild China. New Holland (Publish-
ers), London.
MACKINNON, J., M. SHA, C. CHEUNG, G. CAREY, X. ZHU, and
D. MELVILLE. 1996. See General References.
SICHUAN VEGETATION STUDY GROUP. 1980. Sichuan Zhibei
(Vegetation of Sichuan). Sichuan People’s Press,
Chengdu.
SMIL, V. 1992. “China’s Environment in the 1980’s: Some
Critical Changes.” Ambio 21:431-436.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
WANG, W.T., S.G. WU, K.Y. LANG, P.Q. LI, F.T. PU, and S.K.
CHEN (Eds.). 1993. Vascular Plants of the Hengduan
Mountains, Vol. 1: Pteridophyta, Gymnospermae, Di-
cotyledoneae (Saururacea to Cornaceae). Science
Press, Beijing.
WANG, W.T., S.G. WU, K.Y. LANG, P.Q. LI, F.T. PU, and S.K.
CHEN (Eds.). 1994. Vascular Plants of the Hengduan
Mountains, Vol. 2: Dicotyledoneae (Diapensiacea to
Asteraceae) to Monocotyledoneae (Typhaceae to Or-
chidaceae). Science Press, Beijing.
WU, Z.Y. 1988. “Hengduan Mountain Flora and Her Signif-
icance.” Journal Japanese Botany 63:297-311.
XIONG, J.H. (Ed.). 1989. Flora Sichuanica, Vol. 8: An-
giospermae. Sichuan Government Press, Chengdu.
XU, J.M. (Ed.). 1991. Flora Sichuanica, Vol. 7: Spermato-
phyta. Sichuan Government Press, Chengdu.
ZHAO, E. and K. ADLER. 1993. Herpetology of China. Society
for the Study of Amphibians and Reptiles. Oxford,
U.S.A.
Sundaland
BROOK, B.W., N.S. SODHI, and P.K.L. NG, 2003. “Cata-
strophic Extinctions Follow Deforestation in Singa-
pore.” Nature 424:420-423.
CURRAN, L.M., S. TRIGG, A. MCDONALD, D. ASTIANI, Y.M.
HARDIONO, P. SIREGAR, I. CANIAGO, and E. KASISCHKE.
2004. “Lowland Forest Loss in Protected Areas of In-
donesian Borneo.” Science 303:1000-1003.
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1995. See references for the Introduction.
FWI/GFW. 2002. See General References.
KINNAIRD, M.F., E.W. SANDERSON, T.G. O’BRIEN, H.T.
WIBISONO, and G. WOOLMER. 2003. “Deforestation
Trends in a Tropical Landscape and Implications for
Endangered Large Mammals.” Conservation Biology
17:245-257.
KOTTELAT, M. and T. WHITTEN. 1996. Freshwater Biodiversi-
ty in Asia, with Special Reference to Fish. World Bank
Technical Paper No. 343. The World Bank, Washing-
ton, D.C.
KOTTELAT, M., T. WHITTEN, S.J. KARTIKASARI, and S.
WIRYOATMODJO. 1996. Freshwater Fishes of Western In-
donesia and Sulawesi. Periplus, Singapore.
RIJKSEN, H. D. and E. MEIJAARD. 1999. Our Vanishing Rela-
tive: The Status of Wild Orang-Utans at the Close of the
Twentieth Century. Kluwer Academic Publishers, Dor-
drecht, The Netherlands.
SIEGERT, F., G. RUECKER, A. HINRICHS, and A.A. HOFFMANN.
2001. “Increased Damage from Fires in Logged Forests
during Droughts Caused by El Niño.” Nature 414:437-440.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
STEENIS VAN, C.G. 1950. See General References.
WHITMORE, T.C. (Ed.). 1984. Tropical Rain Forests of the Far
East. 2nd ed. Oxford University Press, Oxford.
WHITTEN, T. 2002. “Limestone Biodiversity: Treasure
Houses of Rare Species.” In: E.D. WIKRAMANAYAKE, E.
DINERSTEIN, C. LOUCKS, D. OLSON, J. MORRISON, J.
LAMOREUX, M. MCKNIGHT, and P. HEDAO (Eds.). 2002.
See General References.
WIKRAMANAYAKE, E.D., E. DINERSTEIN, C. LOUCKS, D. OL-
SON, J. MORRISON, J. LAMOREUX, M. MCKNIGHT, and P.
HEDAO (Eds.). 2002. See General References.
WORLD BANK. 2001. Indonesia: Environment and Natural
Resource Management in a Time of Transition. Wash-
ington, D.C.
Wallacea
BROWN, N. 1998. “Out of Control: Fires and Forestry in In-
donesia.” Trends in Ecology and Evolution 13:41.
DICKERSON, R.E. 1928. Distribution of Life in the Philippines.
Bureau of Printing, Manila.
FWI/GFW. 2002. See General References.
ROMBANG, W.M., C. TRAINOR, and D. LESMANA. 2002. Daer-
ah Penting Bagi Burung (Important Bird Areas): Nusa
Tenggara. Phka/BirdLife Indonesia. Bogor.
ROMBANG, W.M., R. SARYANTHI, and D. LESMANA. In prep.
Daerah Penting Bagi Burung (Important Bird Areas): Su-
lawesi. Phka/BirdLife Indonesia. Bogor.
ROMBANG, W.M., Y. PERSULESSY, and D. PURMIASA. In prep.
Daerah Penting Bagi Burung (Important Bird Areas):
Maluku. Phka/BirdLife Indonesia. Bogor.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
WALLACE, A.R. 1876. The Geographical Distribution of Ani-
mals. Macmillan, London.
Philippines
ASHTON, P.S. 1997. “Before the Memory Fades: Some Notes
on the Indigenous Forests of the Philippines.” San-
dakania 9:1-190.
BROWN, R.M., A.C. DIESMOS, and A.C. ALCALA. 2001. “The
State of Philippine Herpetology and the Challenges
for the Next Decade.” Silliman Journal 42:18-87.
CARPENTER, K.E. and V.E. NIEM. 1999. Species Identification
Guide for Fisheries Purposes: The Living Marine Resources
of the Western Central Pacific. 6 Vols. FAO, Rome.
CURIO, E. 2002. “Prioritisation of Philippine Island Avifau-
nas for Conservation: A New Combinatorial Measure.”
Biological Conservation 106:373-380.
DANIELSEN, F. and C.G. TREADAWAY. 2004. “Priority Conser-
vation Areas for Butterflies (Lepidoptera: Rhopalocera)
in the Philippine Islands.” Animal Conservation 7:79-92.
DAVIES, J., P.M. MAGSALAY, R. RIGOR, A. MAPALO, and H.
GONZALES. 1990. A Directory of Philippines Wetlands. 2
Vols. Asian Wetland Bureau Philippines Founda-
tion/Haribon Foundation.
ENVIRONMENTAL SCIENCE FOR SOCIAL CHANGE. 1999. Decline
of Philippine Forests. ESSC Inc. and Bookmark., Makati,
Philippines.
ESSELSTYN, J.A., P. WIDMANN, and L.R. HEANEY. In press.
“The Mammals of Palawan Island, Philippines.” Pro-
ceedings of the Biological Society of Washington, D.C.
FERNANDO, E.S., D.A. LAGUNZAD, L.L. CO, D.A. MADULID,
J.L. DE LEON, A.B. LAPIS, I.C. PANGGA, L.M. LIAO, C.C.
CUSTODIO, M. MENDOZA, A. MENIADO, N.M. MOLINYAWE,
P.M. ZAMORA, G.I. TEXON, and W.S. POLLISCO. 2003.
Framework for Philippine Plant Conservation: Strategy
and Action Plan. Protected Areas and Wildlife Bureau,
Department of Environment and Natural Resources.
GAPUD, V.P. 2002. “Arthropods.” In: P.S. ONG, L.E. AFUANG,
and R.G. ROSELL-AMBAL (Eds.). Philippine Biodiversity
Conservation Priorities: A Second Iteration of the Nation-
al Biodiversity Strategy and Action Plan. DENR, Biodi-
versity Conservation Program-UP Center for Integra-
tive and Development Studies, and Conservation
International, Manila.
MERRILL, E.D. 1923-1926. An Enumeration of Philippine
Flowering Plants, Vols. I-IV. Bureau of Printing, Manila.
ONG, P.S. 2002. “Current Status and Prospects of Protected
Areas in the Light of the Philippine Biodiversity Con-
servation Priorities.” In. H.M. TSAI (Ed.). Proceedings of
the Fourth Conference on the Protected Areas of East
Asia: Benefits beyond Boundaries in East Asia. Yang-
mingshan National Park, Taipei.
ONG, P.S., L.E. AFUANG, and R.G. ROSELL-AMBAL. 2002.
Philippine Biodiversity Conservation Priorities: A Second
Iteration of the National Biodiversity Strategy and Action
Plan. DENR, Biodiversity Conservation Program-UP
Center for Integrative and Development Studies, and
Conservation International, Manila.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
TREADAWAY, C.G. 1995. “Checklist of the Butterflies of the
Philippine Islands (Lepidoptera: Rhopalocera).”
Nachr. Entomol. Ver. Apollo, Suppl. 14:7-118.
WIKRAMANAYAKE, E.D., E. DINERSTEIN, C. LOUCKS, D. OL-
SON, J. MORRISON, J. LAMOREUX, M. MCKNIGHT, and P.
HEDAO. 2002. See General References.
Southwest Australia
ALLEN, G.R., S.H. MIDGLEY, and M. ALLEN. 2002. Field
Guide to the Freshwater Fishes of Australia. Western
Australian Museum, Perth, Australia.
BEARD, J.S. 1980. “A New Phytogeographic Map of Western
Australia.” Res. Notes. W.A. Herbarium 3:37-58.
BEARD, J.S. 1990. The Plant Life of Western Australia. Kan-
garoo Press, Sydney.
BEARD, J.S., A.R. CHAPMAN, and P. GIOIA. 2000. “Species
Richness and Endemism in the Western Australian
Flora.” Journal of Biogeography 27, 1257-1268.
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1995. See references for the Introduction.
MITTERMEIER, R.A., P. ROBLES-GIL, and C.G. MITTERMEIER
(Eds.). 1997. See General References.
PATE, J.S. and J.S. BEARD. 1984. Kwongan, the Plant Life of
the Sandplain. University of Western Australia Press,
Nedlands.
SHEPHERD, D.P., G.R. BEESTON, and A.J.M. HOPKINS. 2002.
Native Vegetation in Western Australia: Extent, Type and
Status. Resource Management Technical Report 249.
Department of Agriculture, Government of Western
Australia.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
New Zealand
CROMARTY, P.L., K.G. BROOME, A. COX, R.A. EMPSON, W.M.
HUTCHINSON, and I. MCFADDEN. 2002. “Eradication
Planning for Invasive Alien Mammal Species on is-
lands” –the Approach Developed by the New Zealand
Department of Conservation in “Turning the Tide: The
Eradication of Invasive Species.” Proceedings of the in-
ternational conference on eradication of island inva-
sives. In: C.R. VEITCH and M.N. CLOUT (Eds.). IUCN
Species Survival Commission occasional paper No.27.
DAUGHERTY, C.H., G.W. GIBBS, and R.A. HITCHMOUGH.
1993. “Mega-Island or Micro-Continent? New Zealand
and its Fauna.” Trends in Ecology and Evolution 8:437-
442.
MAY, R.M. 1990. “Taxonomy as Destiny.” Nature 347:129-
130.
MINISTRY FOR THE ENVIRONMENT. 1997. The State of New
Zealand’s Environment. Ministry for the Environment,
New Zealand.
SAUNDERS, A. 2000. A Review of Department of Conservation
Mainland Restoration Projects and Recommendations for
Further Action. Department of Conservation, Wellington.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
SUTHERLAND, R. 1999. “Basement Geology and Tectonic
Development of the Greater New Zealand Region: An
Interpretation from Regional Magnetic Data.” Tectono-
physics 308:341-362
VEITCH, C.R. and B.D. BELL. 1990. “Eradication of Intro-
duced Animals from the Islands of New Zealand.” In:
D.R. TOWNS, C.H. DAUGHERTY, and I.A.E. ATKINSON
(Eds.). Ecological Restoration of New Zealand Islands.
Department of Conservation, Wellington.
New Caledonia
BAUER, A.M. and R.A. SADLIER. 2000. The Herpetofauna of
New Caledonia. Society for the Study of Amphibians
and Reptiles, Ithaca.
BOUCHET, P., T. JAFFRÉ, and J.-M. VEILLON. 1995. “Plant Ex-
tinction in New Caledonia: Protection of Sclerophyll
Forests Urgently Needed.” Biodiversity and Conserva-
tion 4:415-428.
CHAZEAU, J. 1993. “Research on New Caledonia Terrestrial
Fauna: Achievements and Prospects.” Biodiversity Let-
ters 1:123-129.
GARGOMINY, O. (Ed.). 2003. See General References.
GILLESPIE, T.W. and T. JAFFRÉ. 2003. “Tropical Dry Forests
in New Caledonia.” Biodiversity and Conservation
12:1687-1697.
HODEL, D.R. and J.C. PINTAUD. 1998. The Palms of New
Caledonia / Les palmiers de la Nouvelle-Calédonie. Kam-
pon Tansacha, Nong Nooch Tropical Garden, Thai-
land.
HOLLOWAY, J.D. and J.V. PETERS. 1976. “The Butterflies of
New Caledonia and the Loyalty Islands.” J. Nat. Hist.
10:273-318.
JAFFRÉ, T., P.H. BOUCHET, and J.M. VEILLON. 1998. “Threat-
ened Plants of New Caledonia: Is the System of Pro-
tected Areas Adequate?” Biodiversity and Conservation
7:109-135.
JAFFRÉ, T., P.H. MORAT, J.-M. VEILLON, F. RIGAULT, and G.
DAGOSTINI. 2001. “Composition and Characterisation
of the Native Flora of New Caledonia.” Centre IRD,
Documents Scientifiques et Techniques II4:1-121.
LOWRY II., P.P. 1998. “Diversity, Endemism, and Extinction
in the Flora of New Caledonia: A Review.” In: C.I.
PENG and P.P. LOWRY II (Eds.). Proc. Intl. Symposium on
Rare, Threatened, and Endangered Floras of Asia and the
Pacific. Academia Sinica, Taipei, Taiwan.
MATHEWS, S. and M.J. DONOGHUE. 1999. “The Root of An-
giosperm Phylogeny Inferred from Duplicate Phy-
tochrome Genes.” Science 286:947-950.
MORAT, P.H. 1993. “Our Knowledge of the Flora of New
Caledonia: Endemism and Diversity in Relation to
Vegetation Types and Substrates.” Biodiversity Letters
1:72-81.
MORAT, P.H., J.-M. VEILLON, and H.S. MACKEE. 1984.
“Floristic Relationships of New Caledonian Rain For-
est Phanerograms.” In: F.J. RADOVSKY, P.H. RAVEN, and
S.H. SOHMER (Eds.). Biogeography of the Tropical Pacif-
ic. Association of Systematics Collections and Bernice
P. Bishop Museum.
MORAT, P.H., T. JAFFRE, and J.M. VEILLON. 1995. “Neocale-
donian Region: Grand Terre, New Caledonia, France.”
In: S.D. DAVIS, V.H. HEYWOOD, and A.C. HAMILTON
(Eds.). Centers of Plant Diversity. A Guide and Strategy
for their Conservation, Vol. 2: Asia, Australasia and the
Pacific. The World Wildlife Fund (WWF) and The
World Conservation Inion (IUCN). IUCN Publications
Unit, Cambridge.
NAJT, J. and P. GRANDCOLAS (Eds.), 2002. Zoologia Neocale-
donica, Vol. 5: Systématique et endémisme en Nou-
velle-Calédonie. Mém. Mus. Natl. Hist. Nat. 187:1-282.
PARNABY, H.E. 2002. “A New Species of Long-Eared Bat
(Nyctophilus: Vespertilionidae) from New Caledonia.”
Australian Mammalogy 23:115-124.
PLATNICK, N.I. 1993. “The Araneomorph Spider Fauna of
New Caledonia.” Biodiversity Letters 1:102-106.
QUI, Y.L, J. LEE, F. BERNASCONI-QUADRONI, D.E. SOLTIS, P.S.
SOLTIS, M. ZANIS, E.A. ZIMMER, Z. CHEN, V. SAVOLAINEN,
and M.W. CHASE. 1999. “The Earliest Angiosperms:
Evidence from Mitochondrial, Plastid, and Nuclear
Genomes.” Nature 402:404-407.
SADLIER, R.A., A.M. BAUER, and A.H. WHITAKER. 2002. “The
Scincid Lizard Genus Nannoscincus Günther from
New Caledonia in the Southwest Pacific: A Review of
the Morphology and Distribution of Species in the
Nannoscincus mariei Species Group, Including the De-
scription of Three New Species from the Province
Nord.” In: Zoologica Neocaledonia, Vol. 5. Mem. Mus.
Natl. Hist. Nat. 187:269-276.
SETOGUCHI, H., T.A. OSAWA, J.C. PINTAUD, T. JAFFRÉ, and J.-
M. VEILLON. 1998. Phylogenetic Relationships within
Araucariaceae Based on rbcL Gene Sequences. Amer-
ican Journal of Botany 85(11):1507-1516.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
TAKHTAJAN, A. 1986. See General References.
WWF. 1997 and 2000. The Living Planet Campaign. World
Wildlife Fund, Washington, D.C.
Polynesia-Micronesia
ALLISON, A. 1996. “Zoogeography of Amphibians and Rep-
tiles of New Guinea and the Pacific Region.” In: A.
KEAST and S.E. MILLER (Eds.). Origin and Evolution of
Pacific Island Biotas, New Guinea to Eastern Polynesia:
Patterns and Processes. SPB Academic Publishing, Am-
sterdam.
ALLISON, A. 2003. “Biological Surveys: New Perspectives in
the Pacific.” Organisms Diversity and Evolution 3:103-110.
ALLISON, A. and S.E. MILLER. 2000. “Hawaii Biological Sur-
vey: Museum Resources in Support of Conservation.”
In: P.H. RAVEN and T. WILLIAMS (Eds.). The Quest for a
Sustainable World. National Academy Press, Washing-
ton, D.C.
BALGOOY, M.M.J. VAN. 1993. Pacific Plant Areas, Vol. 5. Rijk-
sherbarium, Leiden, The Netherlands.
COOTE, T. and E. LOÈVE. 2003. “From 61 Species to Five:
Endemic Tree Snails of the Society Islands Fall Prey to
an Ill-Judged Biological Control Programme.” Oryx
37:91-96.
COWIE, R.H. 1996. “Pacific Island Land Snails: Relation-
ships, Origins, and Determinants of Diversity.” In: A.
KEAST and S.E. MILLER (Eds.). Origin and Evolution of
Pacific Island Biotas, New Guinea to Eastern Polynesia:
Patterns and Processes. SPB Academic Publishing, Am-
sterdam.
COWIE, R.H. and A.C. ROBINSON. 2003. “The Decline of Na-
tive Pacific Island Faunas: Changes in Status of the
Land Snails of Samoa Through the 20th Century.” Bi-
ological Conservation 110:55-65.
COWIE, R.H., N.L. EVENHUIS, and C.C. CHRISTENSEN. 1995.
“Catalog of the Native Land and Freshwater Mollusks
of the Hawaiian Islands.” Backhuys Publishers, Lei-
den.
CROMBIE, R.I. and D.W. STEADMAN. 1986. “The Lizards of
Rarotonga and Mangaia, Cook Islands Group, Ocea-
nia.” Pacific Science 40(1-4):44-57.
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1997. See references for the Introduction.
ELDREDGE, L.G. 2000. “Non-Indigenous Freshwater Fishes,
Amphibians, and Crustaceans of the Pacific and
Hawaiian Islands.” In: Invasive species in the Pacific: a
technical review and draft regional strategy. South Pacif-
ic Environment Programme, Apia.
ELDREDGE, L.G. and N.L. EVENHUIS. 2003. “Hawaii’s Biodi-
versity: A Detailed Assessment of the Numbers of
Species in the Hawaiian Islands.” Bishop Museum Oc-
casional Papers 76:1-28.
FITZSIMMONS, J.M., J.E. PARHAM, and R.T. NISHIMOTO. 2002.
“Similarities in Behavioral Ecology among Amphidro-
mous and Catadromous Fishes on the Oceanic Islands
of Hawaii and Guam.” Environmental Biology of Fishes
654:123-129.
FLANNERY, T.F. 1995. See General References.
GIBBONS, J.R.H. 1981. “The Biogeography of Brachylophus
(Iguanidae) Including the Description of a New
Species, B. vitiensis, from Fiji.” Journal of Herpetology
15(3):255-273.
HADFIELD, M.G. 1986. “Extinction in Hawaiian Achanti-
nelline Snails.” Malacologia 27:67-81.
HOLLAND, B.S. and M.G. HADFIELD. 2002. “Islands within
an Island: Phylogeography and Conservation Genetics
of the Endangered Hawaiian Tree Snail, Achatinella
mustelina.” Molecular Ecology 11:365-375.
JAMES, H.F. and S.L. OLSON. 1991. “Descriptions of Thirty-
Two New Species of Birds from the Hawaiian Islands:
Part II.” Passeriformes, Vol. 46 of Ornithological Mono-
graphs. The American Ornithologist’s Union, Wash-
ington, D.C.
LYDEARD, C., R.H. COWIE, W.F. PONDER, S. CLARK, A.E. BO-
GAN, P. BOUCHET, K.S. CUMMINGS, T.J. FREST, D.G. HER-
BERT, R. HERSHLER, O. GARGOMINY, K. PEREZ, B. ROTH,
M. SEDDON, E.E. STRONG, and F.G. THOMOPSON. In
press. “The Global Decline of Nonmarine Mollusks.”
BioScience, April 2004.
MCDOWALL, R.M. 2003. “Hawaiian Biogeography and the
Islands’ Freshwater Fish Fauna.” Journal of Biogeogra-
phy 30:703-719.
MEYER, J.Y. and J. FLORENCE. 1997. “Tahiti’s Native Flora
Endangered by the Invasion of Miconia calvescens DC
(Melastomataceae).” Journal of Biogeography 23:775-
781.
MUELLER-DOMBOIS, D. and F.R. FOSBERG. 1998. See Gener-
al References.
OLSON, S.L. and H.F. JAMES. 1991. “Descriptions of Thirty-
Two New Species of Birds from the Hawaiian Islands:
Part I.” Non-passeriformes, Vol. 45 of Ornithological
Monographs. The American Ornithologist’s Union,
Washington, D.C.
PRATT, H.D., P.L. BRUNER, and D.G. BERRETT. 1987. The
Birds of Hawaii and the Tropical Pacific. Princeton Uni-
versity Press, Princeton.
RETI, I. 1993. “The South Pacific Biodiversity Conservation
Programme: Its Concept and Scope.” Pacific Science
Association Information Bulletin 45(1-2):9-14.
SPC (SECRETARIAT OF THE PACIFIC COMMUNITY). http://
www.spc.int/demog/demogen/english01-02/recentstats/
2003/03 poster.xls
SPREP (SOUTH PACIFIC REGIONAL ENVIRONMENTAL PRO-
GRAMME). 2003. Action Strategy for Nature Conservation
in the Pacific Islands Region, 2003-2007. South Pacific
Regional Environment Programme, Apia, Samoa.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
385
STEADMAN, D.W. 1995. See references for the Introduc-
tion.
STODDART, D.R. 1992. “Biogeography of the Tropical Pacif-
ic.” Pacific Science 46(2):276-293.
TAKHTADZHËIÌAN, A.L. and A. CRONQUIST. 1986. Floristic Re-
gions of the World. University of California Press,
Berkeley.
TOMICH, P.Q. 1986. Mammals of Hawai’i. Bishop Museum
Press, Honolulu.
Madrean Pine-Oak Woodlands
ALLEN, L.S. 1994. “Fire Management in the Sky Islands.”
In: L.F. DEBANO, G.J. GOTTFRIED, R.H. HAMRE, C.B. ED-
MINSTER, P.F. FFOLLIOTT, and A. ORTEGA-RUBIO (Eds.).
The Sky Islands of Southwestern United States and
Northwestern Mexico. USDA Forest Service: General
Technical Report RM-GTR-264.
BROWER, L.P., D.P. KUST, E. RENDÓN-SALINAS, E. GARCÍA-SER-
RANO, K.R. KUST, J. MILLER, C. FERNÁNDEZ DEL REY, and
K. PAPE. In press. “Catastrophic Winter Storm Mortali-
ty of Monarch Butterflies in Mexico during January
2002.” In: K.S. OBERHAUSER and M.J. SOLENSKY (Eds.).
The Monarch Butterfly. Cornell University Press.
BROWER, L.P., G. CASTILLEJA, A. PERALTA, J. LÓPEZ-GARCÍA,
L. BOJÓRQUEZ-TAPIA, S. DÍAZ, D. MELGAREJO, and M.
MISSRIE. 2002. “Quantitative Changes in Forest Quali-
ty in a Principal Overwintering Area of the Monarch
Butterfly in Mexico, 1971-1999.” Conservation Biology
16:346-359.
BROWN, D.E. (Ed.). 1994. Biotic Communities Southwestern
United States and Northwestern Mexico. University of
Utah Press, Salt Lake City.
CHALLENGER, A. 1998. Utilización y conservación de los eco-
sistemas terrestres de México. Pasado, presente, y futuro.
CONABIO, IBUNAM, and Agrupación Sierra Madre,
Mexico City.
CONANP. 2002. Programa de Trabajo. Comisión Nacional de
Áreas Naturales Protegidas, 2001-2006 (2nd edition).
SEMARNAT, Mexico City.
FARJON, A. and B.T. STYLES. 1997. Pinus (Pinaceae). Flora
Neotropica. Monograph 75. The New York Botanical
Garden.
FARJON, A. and C.N. PAGE (Eds.). 1999. Conifers: Status Sur-
vey and Conservation Action Plan. IUCN/SSC Conifer
Specialist Group. IUCN, Gland and Cambridge.
GONZÁLEZ-MEDRANO, F. 2004. Las comunidades vegetales de
México. INE/SEMARNAT, Mexico City.
LAMMERTINK, J.M., J.A. ROJAS-TOMÉ, F.M. CASILLAS-ORONA,
and R.L. OTTO. 1997. Situación y conservación de los
bosques antiguos de pino-encino de la Sierra Madre Occi-
dental y sus aves endémicas. Consejo Internacional
para la Preservación de las Aves, Sección Mexicana,
Mexico.
MCLAUGHLIN, S.P. 1994. “An Overview of the Flora of the
Sky Islands, Southeastern Arizona: Diversity, Affini-
ties, and Insularity.” In: L.F. DEBANO, G.J. GOTTFRIED,
R.H. HAMRE, C.B. EDMINSTER, P.F. FFOLLIOTT, and A.
ORTEGA-RUBIO (Eds.). Biodiversity and Management of
the Madrean Archipelago: The Sky Islands of Southwest-
ern United States and Northwestern Mexico. USDA For-
est Service: General Technical Report RM-GTR-264.
MILLER, R.R. and M.L. SMITH. 1986. “Origin and Geography
of the Fishes of Central Mexico.” In: C.H. HOCUTT and
E.O. WILEY (Eds.). The Zoogeography of North American
Freshwater Fishes. John Wiley and Sons, New York.
NIXON, K.C. 1993. “The genus Quercus in Mexico.” In: T.P.
RAMAMOORTHY, R. BYE, A. LOT, and J. FA (Eds.). Biolog-
ical Diversity of Mexico: Origins and Distribution. Ox-
ford University Press, Oxford.
PORTALES, G.L., P. REYES, H. RANGEL, A. VELÁZQUEZ, P.
MILLER, S. ELLIS, and A.T. SMITH. 1997. Conservation of
Mexican Lagomorphs in Danger of Extinction. IUCN/SSC
Lagomorph Specialist Group and IUCN/SSC Conser-
vation Breeding Specialist Group.
ROBLES-GIL, P. (Ed.). 1995. Mexico: Diversity of Cultures.
CEMEX, Mexico City.
ROMERO, F.J. and A. VELÁZQUEZ. 1994. El conejo zacatuche: tan
lejos de Dios y tan cerca de la Ciudad de México. Instituto
Nacional de Ecología, Consejo Nacional de la Fauna.
ROMERO, F.J., H. RANGEL-CORDERO, A. ESTÉVEZ-RAMÍREZ, M.
ESCAMILLA, and L. CABRERA-GARCÍA. 1999. “Aspectos
sociodemograficos y actividades productivas rurales
del sur de la cuenca de México.” In: A. VELÁZQUEZ and
F. J. ROMERO (Eds.). Biodiversidad de la region de mon-
taña del sur de la cuenca de México. Universidad
Autónoma Metrooplitana-Unidad Xochimilco.
RZEDOWSKI, J. 1978. Vegetación de México (primera edi-
ción). Limusa, Mexico City.
386
RZEDOWSKI, J. 1993. “Diversity and Origins of the Phanero-
gamic Flora of Mexico.” In: T.P. RAMAMOORTHY, R. BYE,
A. LOT, and J. FA (Eds.). Biological Diversity of Mexico: Ori-
gins and Distribution. Oxford University Press.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
TOLEDO, V.M., P. ALARCÓN-CHAIRES, P. MOGUEL, M. OLIVO,
A. CABRERA, and A. RODRÍGUEZ-ALDABE. 2002. “Biodi-
versidad y pueblos indios en México y Centroaméri-
ca.” Biodiversitas 43:1-8.
VALERO, A., J. SCHIPPER, T. ALLNUTT, and C. BURDETTE.
2001a. “Sierra Madre Occidental Pine-Oak Forests
(NA0302).” Draft Ecoregion Profile.
VALERO, A., J. SCHIPPER, T. ALLNUTT, and C. BURDETTE.
2001b. “Sierra Madre Oriental Pine-Oak Forests
(NA0303).” Draft Ecoregion Profile.
VALERO, A., J. SCHIPPER, T. ALLNUTT, and C. BURDETTE.
2001c. “Trans-Mexican Volcanic Belt Pine-Oak Forests
(NT0310).” Draft Ecoregion Profile.
VALERO, A., J. SCHIPPER, and T. ALLNUTT. 2001d. “Sierra
Madre del Sur Pine-Oak Forests (NT0309).” Draft
Ecoregion Profile.
VALERO, A., J. SCHIPPER, and T. ALLNUTT. 2001e. “Sierra
Madre de Oaxaca Pine-Oak Forests (NT0308).” Draft
Ecoregion Profile.
VELÁZQUEZ, A. 2001. “Teporingo: ¿debería adoptar México
un símbolo de conservación?” Ecosistema 2001 2:22-23.
VELÁZQUEZ, A., F.J. ROMERO, and J. LÓPEZ-PANIAGUA. 1996.
Ecologia y conservacion del conejo zacatuche y su habi-
tat. UNAM.
Maputaland-Pondoland-Albany
ACOCKS, J.P.H. 1953. “Veld Types of South Africa. Memoirs
of the Botanical Survey of South Africa.” 28:1-192. (3rd
edition with updated names and illustrations pub-
lished in 1988 as Memoirs of the Botanical Survey of
South Africa 57:1-146).
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1994. See references for the Introduction.
ESTERHUYSEN, P. (Ed.). 1998a. Africa A-Z: Continental and
Country Profiles. Africa Institute of South Africa, Old
Mill Road, Western Cape.
ESTERHUYSEN, P. 1998b. Africa at a Glance: Facts and Figures
1997/8. 10th edition. Africa Institute of South Africa,
Old Mill Road, Western Cape.
GINN, P.J., W.G. MCILLERON, and P. LE S. MILSTEIN
(Comps.). 1989. The Complete Book of Southern
African Birds. Struik Publishers, Cape Town.
HAMER, M.L., M.J. SAMWAYS, and H. RUHBERG. 1997. “A Re-
view of the Onycophora of South Africa, with a Dis-
cussion of their Conservation.” Ann. Natal. Mus.
38:283-312.
IZIDINE, S.A. 2003. Licuáti Forest Reserve, Mozambique: Flo-
ra, Utilization and Conservation. MSc thesis, Depart-
ment of Botany, University of Pretoria, Pretoria.
LE ROUX, J. (Comp.). 2002. The Biodiversity of South Africa
2002. Indicators, Trends and Human Impacts. Struik
Publishers, Cape Town.
LOW, A.B. and A.G. REBELO. 1998. Vegetation of South Africa,
Lesotho and Swaziland. 2nd edition. Department of En-
vironmental Affairs and Tourism, Pretoria.
SILANDER JR., J.A. 2001. “Temperate Rainforests.” In: Ency-
clopedia of Biodiversity, Vol. 5. Academic Press, San
Diego.
SKELTON, P. 2001. A Complete Guide to the Freshwater Fishes
of Southern Africa. Struik Publishers, Cape Town.
SMITH, G.F., E.J. VAN JAARSVELD, T.H. ARNOLD, F.E. STEF-
FENS, R.D. DIXON, and J.A. RETIEF (Eds.). 1997. List of
Southern African Succulent Plants. Umdaus Press, Pre-
toria.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
VICTOR, J.E. and A.P. DOLD. 2003. “Threatened Plants of
the Albany Centre of Floristic Endemism, South
Africa.” South African Journal of Science 99:437-446.
VLOK, J.H.J., D.I.W. EUSTON-BROWN, and R.M. COWLING.
2003. See General References.
WERGER, M.J.A. (Ed.). 1978. Biogeography and Ecology of
Southern Africa (2 vols). Dr. W. Junk Publications, The
Hague.
WHITE, F. 1983. See references for the Introduction.
VAN WYK, A.E. and G.F. SMITH. 2001. See General References.
Coastal Forests of Eastern Africa
AXELROD, D.I. and P.H. RAVEN. 1978. “Late Cretaceous and
Tertiary Vegetation History of Africa.” In: M.J.A.
WERGER (Ed.). Biogeography and Ecology of Southern
Africa. Dr. W. Junk Publications, The Hague.
BAATVIK, S.T. 1993. “The Genus Saintpaulia: 100 Years of
History, Taxonomy, Ecology, Distribution and Con-
servation.” Fragm. Flos. Geobot. Suppl. 2:97-112.
BEARDER, S.K. 1999. “Physical and Social Diversity among
Nocturnal Primates: A New View Based on Long Term
Research.” Primates 40:267-282.
BEARDER, S.K., L. AMBROSE, C. HARCOURT, P. HONESS, A.
PERKIN, E. PIMLEY, S. PULLEN, and N. SVOBODA. 2003.
“Species-Typical Patterns of Infant Contact, Sleeping
Site Use and Social Cohesion among Noctural Pri-
mates in Africa.” Folia Primatol 74:337-354.
BROADLEY, D. Undated. The Reptiles of the East African
Coastal Mosaic. Biodiversity Foundation for Africa,
BFA Seminar Series No. 19. Bulawayo, Zimbabwe.
BURGESS, N.D. and C. MUIR (Eds.). 1994. “Coastal Forests of
Eastern Africa: Biodiversity and Conservation.” Pro-
ceedings of a workshop held at the University of Dar es
Salaam, 9-11 August 1993. Society for Environmental Ex-
ploration/Royal Society for the Protection of Birds, U.K.
BURGESS, N.D., G.P. CLARKE, and W.A. RODGERS. 1998. “The
Coastal Forests of Eastern Africa: Status, Species En-
demism and its Possible Causes.” Biological Journal of
the Linnean Society 64:337-367.
BURGESS, N.D. and G.P. CLARKE (Eds.). 2000. The Coastal
Forests of Eastern Africa. IUCN, Gland and Cambridge.
BURGESS, N.D., D. KOCK, A. COCKLE, C. FITZGIBBON, P. JENK-
INS, and P. HONESS 2000. “Mammals.” In. N.D. BURGESS
and G.P. CLARKE. The Coastal Forests of Eastern Africa.
IUCN Forest Conservation Programme, Gland and
Cambridge.
BURGESS, N.D. and A. HIPKISS. 2002. “Pande Game Reserve:
Observations on Forest Loss between July 1989 and
November 2001.” The Arc Journal 14:1-3. www.tfcg.org
BURGESS, N.B., G. NEGUSSIE, P. BECHTEL, N.O. MOISES, and
N. DOGGART. 2003a. “Coastal Forests in Northern
Mozambique.” The Arc Journal 15:9-12. www.tfcg.org
BURGESS, N.D., N. DOGGART, K. DOODY, G. NEGUSSIE, P.
SUMBI, and A. PERKIN. 2003b. “New Information on the
Lowland Coastal Forests of Eastern Africa.” Oryx
37:280-281.
BURGESS, N., J. D’AMICO HALES, E. UNDERWOOD, E. DINER-
STEIN, D. OLSON, I. ITOUA, J. SCHIPPER, T. RICKETTS, and
K. NEWMAN. In press. “Terrestrial Ecoregions of Africa
and Madagascar: A Continental Assessment.” Island
Press, Washington, D.C.
CLARKE, G.P. 1995. Status Reports for 6 Coastal Forests in
Lindi Region, Tanzania. Frontier-Tanzania Technical
Report No. 18. The Society for Environmental Explo-
ration/The University of Dar es Salaam, London and
Dar es Salaam.
CLARKE, G.P. 1998a. “Plants in Peril 24: Notes on Lowland
African Violets (Saintpaulia) in the Wild.” Kew Maga-
zine 15:62-67.
CLARKE, G.P. 1998b. “A New Regional Centre of Endemism
in Africa.” In: D.F. CUTLER, C.R. HUXLEY, and J.M LOCK
(Eds.). Aspects of the Ecology, Taxonomy and Chorology
of the Floras of Africa and Madagascar. Kew Bulletin
Additional Series. Royal Botanic Gardens, Kew.
CLARKE, G.P. 2000a. “Defining the Eastern African Coastal
Forests.” In: N.D. BURGESS and G.P. CLARKE (Eds.). The
Coastal Forests of Eastern Africa. IUCN, Gland and
Cambridge.
CLARKE, G.P. 2000b. “Climate and Climatic History.” In:
N.D. BURGESS and G.P. CLARKE (Eds.). The Coastal
Forests of Eastern Africa. IUCN, Gland and Cambridge.
CLARKE, G.P. 2001. “The Lindi Local Centre of Endemism
in SE Tanzania.” Systematics and Geography of Plants
71:1063-1072
CLARKE, G.P. and A. DICKINSON. 1995. Status Reports for 11
Coastal Forests in Coast Region, Tanzania. Frontier-
Tanzania Technical Report No. 17. The Society for En-
vironmental Exploration/The University of Dar es
Salaam, London and Dar es Salaam.
CLARKE, G.P. and L.K. SUBBLEFIELD. 1995. Status Reports
for 7 Coastal Forests in Tanga Region, Tanzania. Fron-
tier-Tanzania Technical Report No. 16. The Society for
Environmental Exploration/The University of Dar es
Salaam, London and Dar es Salaam.
CLARKE, G.P. and N.D. BURGESS. 2000. “Geology and Geo-
morphology.” Chapter 2.1. In: N.D. BURGESS and G.P.
CLARKE (Eds.). The Coastal Forests of Eastern Africa.
IUCN Forest Conservation Series. IUCN, Gland and
Cambridge.
CLARKE, G.P., K. VOLLESEN, and L.B. MWASUMBI. 2000. Vas-
cular Plants. In: N.D. BURGESS and G.P. CLARKE (Eds.).
The Coastal Forests of Eastern Africa. IUCN, Gland and
Cambridge.
CLAUSNITZER, V. 2001. “Notes on the Species Diversity of
East African Odonata, with a Checklist of Species.”
Odonatologica 30:49-66
CEPF. 2003. Ecosystem Profile: Eastern Arc Mountains
and Coastal Forests of Tanzania and Kenya Biodiversi-
ty Hotspot. Critical Ecosystem Partnership Fund,
Washington, D.C. http://www.cepf.net/xp/cepf/where_
we_work/eastern_arc_mountains/full_strategy.xml
GRUBB, P., T.M. BUTYNSKI, J.F. OATES, S.K. BEARDER, T.R.
DISOTELL, C. GROVES, and T. STRUHSAKER. 2003. “An As-
sessment of the Diversity of African Primates.” Inter-
national Journal of Primatology 42:1301-1358.
HOFFMAN, R.L. 2000. “Millipedes.” In: N.D. BURGESS and
G.P. CLARKE (Eds.). The Coastal Forests of Eastern
Africa. IUCN, Gland and Cambridge.
HRBEK, T. and A. LARSON. 1999. “The Evolution of Dia-
pause in the Killifish Family Rivulidae (Atherinomor-
pha, Cyprinodontiformes): A Molecular Phylogenetic
and Biogeographic Perspective.” Evolution 53:1200-
1216.
LOVETT, J.C. and S.K. WASSER (Eds.). 1993. See references
for the Introduction.
LOVETT, J.C., J.R. HANSEN, and V. HØRLYCK. 2000. “Com-
parison with Eastern Arc Forests.” In: N.D. BURGESS
and G.P. CLARKE (Eds.). The Coastal Forests of Eastern
Africa. IUCN, Gland and Cambridge.
MADGWICK, F.J. 1988. “Riverine Forest in the Jubba Valley:
Vegetation Analysis and Comments on Forest Conser-
vation.” The Biogeography of Somalia. Biogeographia
14:67-88.
MILLEDGE, S.A.H. and B.K. KAALE. 2003. Bridging the Gap:
Linking Timber Trade with Infrastructure Development
and Poverty Eradication Efforts in Southern Tanzania.
TRAFFIC East/Southern Africa, Dar es Salaam.
MITTERMEIER, R.A., N. MYERS, J.B. THOMPSEN, G.A.B. DA
FONSECA, and S. OLIVIERI. 1998. See references for the
Introduction.
MYERS, N., J.C. LOVETT, and N.D. BURGESS. 1999. “The East-
ern Arc Mountains and Coastal Lowland Forests
Hotspot.” In: R.A. MITTERMEIER, N. MYERS, P. ROBLES-
GIL, and C.G. MITTERMEIER (Eds.). 1999. See refer-
ences for the Introduction.
MYERS, N., R.A MITTERMEIER, C.G. MITTERMEIER, G.A.B. DA
FONSECA, and J. KENT. 2000. See references for the In-
troduction.
NATURE KENYA and WCST. 2003. Conservation Outcomes for
Eastern Arc and Lowland Coastal Forests Hotspot. Na-
ture Kenya, Nairobi.
OLSON, D.M. and E. DINERSTEIN. 1998. See references for
the Introduction.
OLSON, D.M., E. DINERSTEIN, E.D. WIKRAMANAYAKE, N.D.
BURGESS, G.V.N. POWELL, E. UNDERWOOD, J.A. D’AMICO,
I.I. ITOUA, H.E. STRAND, J.C. MORRISON, C.J. LOUCKS,
T.F. ALLNUTT, J.F. LAMOREUX, T.H. RICKETTS, Y. KURA,
W.W. WETTENGEL, P. HEDAO, and K. KASSEM. 2001. “Ter-
restrial Ecoregions of the World: A New Map of Life on
Earth.” Bioscience 51:933-938.
PARKER, V. 1999. The Atlas of the Birds of Sul do Save, South-
ern Mozambique. Avian Demography Unite and En-
dangered Wildlife Trust, Cape Town, Johannesburg.
ROBERTSON, S.A. and W.R.Q. LUKE. 1993. Kenya Coastal
Forests. WWF, Nairobi.
RODGERS, W.A., J.B. HALL, L.B. MWASUMBI, C.J. GRIFFITHS,
and K. VOLLESEN. 1983. The Conservation Values and
Status of Kimboza Forest Reserve, Tanzania. University
of Dar es Salaam.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
TIMBERLAKE, J., J. GOLDING, and G.P. CLARKE. 2004. “Nias-
sa Botanical Expedition, June 2003.” Report prepared
for the Sociedade para a Gestão e Desenvolvimento da
Reserva do Niassa Moçambique. Biodiversity Founda-
tion for Africa, Zimbabwe.
WHITE, F. 1983. See references for the Introduction.
WORLD BANK. 2002. See General References.
WWF-EARPO. 2002. Coastal Forests of Eastern Africa: Draft
Action Plan. WWF EARPO, Nairobi.
YOUNGE, A., G. NEGUSSIE, and N. BURGESS. 2002. Eastern
Africa Coastal Forest Programme: Regional Workshop
Report Feb 4-7, 2002. WWF Eastern African Programme
Office, Nairobi.
Eastern Afromontane
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1994. See references for the Introduction.
GRAHAM, R.A. 1960. “Rosaceae.” In: C.E. HUBBARD and E.
MILNE-REDHEAD (Eds.). Flora of Tropical East Africa.
Crown Agents for Overseas Governments and Admin-
istrations, London.
HUBBARD, C.E. 1970. “Arundinaria.” In: W.D. CLAYTON.
“Gramineae” (part 1). In: E. MILNE-REDHEAD and R.M.
POLHILL (Eds.). Flora of Tropical East Africa. Crown
Agents for Oversea Governments and Administra-
tions, London.
KINGDON, J. 1989. Island Africa: The Evolution of Africa’s
Rare Animals and Plants. Princeton University Press,
Princeton.
WHITE, F. 1978. “The Afromontane Region.” In: M.J.A.
WERGER (Ed.). Biogeography of Ecology of Southern
Africa. Dr. W. Junk Publications, The Hague.
WHITE, F. 1983. See references for the Introduction.
VAN WYK, A.E. and G.F. SMITH. 2001. See General Refer-
ences.
Eastern Arc Mountains and Southern Rift
BELL-CROSS, G. and J. MINSHULL. 1988. Fishes of Zimbabwe.
National Museums and Monuments of Zimbabwe,
Harare.
BRANCH, B. 1998. Field Guide to Snakes and Other Reptiles of
Southern Africa. 3rd edition. Struik Publishers, Cape
Town.
BROOKS, T.M., R.A. MITTERMEIER, C.G. MITTERMEIER,
G.A.B. DA FONSECA, A.B. RYLANDS, W.R. KONSTANT, P.
FLICK, J. PILGRIM, S. OLDFIELD, G. MAGIN, and C.
HILTON-TAYLOR. 2002. “Habitat Loss and Extinction in
the Hotspots of Biodiversity.” Conservation Biology
16:909-923.
BURGESS, N.D. and S. MBWANA. 2000. “Forestry.” In: N.D.
BURGESS and G.P. CLARKE (Eds.). 2000. The Coastal
Forests of Eastern Africa. IUCN Forest Conservation
Programme, Gland and Cambridge.
BURGESS, N., T.S. ROMDAL, and M. RAHNER. 2001. “Forest
Loss in the Ulugurus, Tanzania and the Status of the
Uluguru Bush Shrike Malconotus Alius.” Bulletin of the
African Bird Club 8:89-90.
BURGESS, N.D., N. DOGGART, and J.C. LOVETT. 2002. “The
Uluguru Mountains of Eastern Tanzania: The Effect of
Forest Loss on Biodiversity.” Oryx 36:140-152.
BURGESS, N., J. D’AMICO HALES, E. UNDERWOOD, E. DINER-
STEIN, D. OLSON, I. ITOUA, J. SCHIPPER, T. RICKETTS, and
K. NEWMAN. In press. Terrestrial Ecoregions of Africa
and Madagascar: A Continental Assessment. Island
Press, Washington, D.C.
CARTER, J. 1987. Malawi: Wildlife Parks and Reserves.
Macmillan Ltd, London and New York.
CHANNING, A. and W.T. STANLEY. 2002. A New Tree Toad
from Tanzania. African Journal of Herpetology 51:121-128.
CHAPMAN, J.D. and F. WHITE. 1970. The Evergreen Forests of
Malawi. Commonwealth Forestry Institute, University
of Oxford, Oxford.
CLAUSNITZER, V. 2001. “Notes on the Species Diversity of
East African Odonata, with a Checklist of Species.”
Odonatologica 30:49-66
CONGDON, T.C.E., A. GARDINER, and I. BAMPTON. 2001.
“Some Endemic Butterflies of Kenya, Tanzania and
Malawi.” In press.
CRIBB, P.J. and G.P. LEEDAL. 1982. The Mountain Flowers of
Southern Tanzania: A Field Guide to the Common Flow-
ers. A.A. Balkema, Rotterdam, Netherlands.
DAVENPORT, T.R.B. 2002. “Tanzania’s New National Park to
Protect Orchids.” Oryx 36:224
DAVENPORT, T.R.B. and H.J. NDANGALASI. 2003. “An Esca-
lating Trade in Orchid Tubers across Tanzania’s South-
ern Highlands: Assessment, Dynamics and Conserva-
tion Implications.” Oryx 37:55-61.
DAVENPORT, T.R.B. and B. BYTEBIER. In press. “Kitulo
Plateau: A New Park for Orchids.” Orchid Review.
DINESEN, L., T. LEHMBERG, M.C. RAHNER, and J. FJELDSÅ.
2000. “Conservation Priorities for the Forests of the
Udzungwa Mountains, Tanzania, Based on Primates,
Duikers and Birds.” Biological Conservation 99:223-
236.
DOWSETT-LEMAIRE, F. 1989. “The Flora and Phytogeography
of the Evergreen Forests of Malawi.” I. Afromontane
and mid-altitude forests. Bulletin du Jardin Botanique
National de Belgique 59:3-131.
DOWSETT-LEMAIRE, F., R.J. DOWSETT, and M. DYER. 2001.
“Malawi.” In: L.D.C. FISHPOOL and M.I. EVANS (Eds.).
Important Bird Areas in Africa and Associated Islands:
Priority Sites for Conservation. Pisces Publications and
BirdLife International (BirdLife Conservation Series
No. 11), Newbury and Cambridge, U.K.
FJELDSÅ, J. and J.C. LOVETT. 1997. “Geographical Pat-
terns of Old and Young Species in African Forest Bio-
ta: The Significance of Specific Montane Areas as
Evolutionary Centers.” Biodiversity and Conservation
6:325-347.
GRIFFITHS, C.J. 1993. “The Geological Evolution of East
Africa.” In: J.C. LOVETT and S.K. WASSER (Eds.). Bio-
geography and Ecology of the Rain Forests of Eastern
Africa. Cambridge University Press, Cambridge.
HAMILTON, A.C. 1982. Environmental History of East Africa:
A Study of the Quaternary. Academic Press, London.
HAMILTON, A.C. and R. BENSTED-SMITH. 1989. Forest Con-
servation in the East Usambara Mountains, Tanzania.
IUCN, Gland and Cambridge.
HOFFMAN, R.L. 1993. “Biogeography of East African Mon-
tane Forest Millipedes.” In: J.C. LOVETT and S.K. WASSER
(Eds.). Biogeography and Ecology of the Rain Forests of
Eastern Africa. Cambridge University Press, Cambridge.
HYMAS, O. 2000. Assessment of the Remaining Forests on the
Uluguru Mountains and the Pressures they Face. Ulugu-
ru Mountains Biodiversity Conservation Project. Tech-
nical Report. Available at
http://www.africanconservation.com/uluguru
HYMAS, O. 2001. People on the Hills: An Assessment of Forest
Disturbance of the Ulugurus in Relation to Access. UM-
BCP, Morogoro. Available at
http://www.africanconservation.com/uluguru
DE JONG, R. and T.C.E. CONGDON. 1993. “The Montane
Butterflies of the Eastern African Forests.” In: J.C.
LOVETT and S.K. WASSER (Eds.). Biogeography and Ecol-
ogy of the Rain Forests of Eastern Africa. Cambridge
University Press, Cambridge.
KERFOOT, O. 1963. “A Preliminary Account of the Vegeta-
tion of the Mbeya Range, Tanganyika.” Kirkia 4:191-
206.
KURZWEIL, H. 2000. “Notes on the Orchids of the Nyika
Plateau, Malawi/Zambia.” Orchids South Africa 31:76-
85.
LA CROIX, I.F., E.A.S. LA CROIX, and T.M. LA CROIX. 1991.
The Orchids of Malawi. A.A. Balkema, Rotterdam.
LOVETT, J.C. 1988. “Endemism and Affinities of the Tan-
zanian Montane Forest Flora.” In: P. GOLDBLATT and
P.P. LOWRY (Eds.). “Proceedings of the Eleventh Plenary
Meeting of the Association for the Taxonomic Study of
Tropical Africa.” Monographs in Systematic Botany
from the Missouri Botanical Garden 25:591-598.
LOVETT, J.C. 1990. “Classification and Status of the Moist
Forests of Tanzania.” Mitteilungen aus dem Institut
für Allgemeine Botanik Hamburg 23a:287-300.
LOVETT, J.C. 1993. “Eastern Arc Moist Forest Flora.” In:
J.C. LOVETT and S.K. WASSER (Eds.). Biogeography and
Ecology of the Rain Forests of Eastern Africa. Cambridge
University Press, Cambridge.
LOVETT, J.C. 1998a. “Continuous Change in Tanzanian
Moist Forest Tree Communities with Elevation.” Jour-
nal of Tropical Ecology 14:719-722.
LOVETT, J.C. 1998b. “Importance of the Eastern Arc Moun-
tains for Vascular Plants.” Journal of East African Nat-
ural History 87:59-74.
LOVETT, J.C. and I. FRIIS. 1996. “Some Patterns of En-
demism in the Tropical North East and Eastern
African Woody Flora.” In: L.J.G. VAN DER MAESEN, X.M.
VAN DER BURGT, and J.M. VAN MEDENBACH DE ROOY
(Eds.). The Biodiversity of African Plants. Proceedings
XIVth AETFAT Congress 22-27 August, 1994, Wa-
geningen, The Netherlands. Kluwer Academic Pub-
lishers, Dordrecht.
LOVETT, J.C. and G.W. NORTON. 1989. “Afromontane Rain
Forest on Malundwe Hill, Mikumi National Park.” Bi-
ological Conservation 48:13-19.
LOVETT, J.C. and E. PRINS. 1994. “Estimation of Land-Use
Changes on Kitulo Plateau, Tanzania, Using Satellite
Imagery.” Oryx 28:173-182.
LOVETT, J.C. and T. PÓCS. 1993. Assessment of the Condition
of the Catchment Forest Reserves, a Botanical Appraisal.
Catchment Forestry Report 93.3, Dar es Salaam.
LOVETT, J.C. and S.N. STUART. 2001. “Avifauna and Vegeta-
tion of the Shume-Juniperus Forest of the West Usam-
bara Mountains, Tanzania.” Scopus 21:1-14
LOVETT, J.C. and S.K. WASSER (Eds.). 1993. Biogeography
and Ecology of the Rain Forests of Eastern Africa. Cam-
bridge University Press, Cambridge.
LOVETT, J.C., J. HATTON, L.B. MWASUMBI, and J. GERSTLE.
1997. “Assessment of the Impact of the Lower Kihansi
Hydropower Project on the Forests of Kihansi Gorge,
Tanzania.” Biodiversity and Conservation 6:915-933.
LOVETT, J.C., R. MARCHANT, J. TAPLIN, and W. KÜPER. 2004.
“The Oldest Rainforests in Africa: Stability or Re-
silience for Survival and Diversity?” In: A. PURVIS, J.L.
GITTLEMAN, and T.M. BROOKS (Eds.). Phylogeny and
Conservation. Cambridge University Press, Cam-
bridge.
LOWE-MCCONNELL, R.H. 1993. “Fish Faunas of the African
Great Lakes: Origins, Diversity, and Vulnerability.”
Conservation Biology 7:634-643.
MCKONE, D. 1995. “A Brief Survey of the Traditional Forest
Reserves of Rungwe District, Mbeya Region, Tanza-
nia.” Unpublished Draft Report. Government of Tan-
zania/EEC Agroforestry, Soil and Water Conservation
Project, Mbeya and District Forestry Office, Rungwe
District. Retrieved (2001) from:
http://www.mckone.org/tfrrung.html
MCKONE, D. and V. WALZEM. 1994. “A Brief Survey of
Mbeya Region Catchment Forest Reserves.” Govern-
ment of Tanzania/EEC Agroforestry, Soil and Water
Conservation Project/Regional Natural Resources Of-
fice, Mbeya. Retrieved (2001) from:
http://www.mckone.org/mbeyacfr.html
MWIKOMEKE, S., T.H. MSANGI, C.K. MABULA, J. YLHÄISI, and
K.C.H. MUNDEME. 1998. “Traditionally Protected For-
ests and Nature Conservation in the North Pare Moun-
tains and Handeni District, Tanzania.” Journal of the
East African Natural History Society 87:279-290.
NEWMARK, W.D. 1998. “Forest Area, Fragmentation, and
Loss in the Eastern Arc Mountains: Implications for
the Conservation of Biological Diversity.” Journal of
the East African Natural History Society 87:29-36.
NEWMARK, W.D. 2002. “Conserving Biodiversity in East
African Forests: A Study of the Eastern Arc Moun-
tains.” In: Ecological Studies, Vol. 155. Springer, Berlin.
OWEN, R.B., R. CROSSLEY, T.C. JOHNSON, D. TWEDDLE, I. KO-
RNFIELD, S. DAVISON, D.H. ECCLES, and D.E. ÁNGSTROM.
1990. “Major Low Levels of Lake Malawi and their Im-
plications for Speciation Rates in Cichlid Fishes.” Proc.
R. Soc. Lond. B 240:519-553.
PERKIN, A., S. BEARDER, T.M. BUTYNSKI, B. AGWANDA, and B.
BYTEBIER. 2003. “The Taita Mountain Dwarf Galago
Galagoides sp.: A New Primate for Kenya.” Journal of
East African Natural History 91:1-13.
PÓCS, T. 1976. “Vegetation Mapping in the Uluguru Moun-
tains (Tanzania, East Africa).” Boissiera 24:477-498.
PÓCS, T. 1998. “Distribution of Bryophytes along the East-
ern Arc Mountains.” Journal of the East African Natur-
al History Society.
POYNTON, J.C., K.M. HOWELL, B.T. CLARKE, and J.C.
LOVETT. 1998. “A Critically Endangered New Species of
Nectophrynoides (Anura: Bufonidae) from the Kihan-
si Gorge, Udzungwa Mountains, Tanzania.” African
Journal of Herpetology 47:59-67.
QUINN, C.H., H.J. NDANGALASI, J. GERSTLE, and J.C.
LOVETT. In press. “Effect of the Lower Kihansi Hy-
dropower Project and Post-Project Mitigation Mea-
sures on Wetland Vegetation in Kihansi Gorge, Tanza-
nia.” Biodiversity and Conservation.
RODGERS, W.A. 1993. “The Conservation of the Forest Re-
sources of Eastern Africa: Past Influences, Present
Practices and Future Needs.” In: J.C. LOVETT and S.
WASSER (Eds.). Biogeography and Ecology of the Rain
Forests of Eastern Africa. Cambridge University Press,
Cambridge.
RODGERS, W.A. and K.M. HOMEWOOD. 1982a. “Biological
Values and Conservation Prospects for the Forests and
Primate Populations of the Udzungwa Mountains,
Tanzania.” Biological Conservation 24:285-304.
RODGERS, W.A. and K.M. HOMEWOOD. 1982b. “Species Rich-
ness and Endemism in the Usambara Mountains
Forests, Tanzania.” Biological Journal of the Linnean So-
ciety 18:197-242.
SCHARFF, N. 1992. “The Linyphiid Fauna of Eastern Africa
(Araneae: Linyphiidae): Distribution Patterns, Diver-
sity and Endemism.” Biological Journal of the Linnean
Society of London 45:117-154.
SHANGALI, C.F., J. YLHAISI, and K.C.H. MUNDEME. 1998.
“Biodiversity and Human Activities in the Udzungwa
Mountain Forest, Tanzania.” I. Ethnobotanical survey
of the Udzungwa Scarp Forest Reserve. Journal of the
East African Natural History Society 87:291-318.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
STUART, S. N., R.J. ADAMS, and M.D. JENKINS (Eds.). 1990.
Biodiversity in Sub-Saharan Africa and its Islands: Con-
servation, Management and Sustainable Use. Occasion-
al papers of the IUCN Species Survival Commission.
IUCN, Gland.
TREWAVAS, E. 1983. Tilapiine Fishes of the Genera
Sarotherodon, Oreochromis and Danakilia. British
Museum of Natural History, London.
WHITE, F. 1983. See references for the Introduction.
WHITE, F., F. DOWSETT-LEMAIRE, and J.D. CHAPMAN. 2001.
Evergreen Forest Flora of Malawi. Royal Botanic Gar-
dens, Kew, London.
WILLIS, C.K., J.E. BURROWS, M. KOEKEMOER, and L. FISH.
2000. Plants of the Nyika, Vol. I: Preliminary Checklist.
National Botanical Institute (SABONET), Pretoria.
WORLD BANK. 2002. See General References.
Albertine Rift
BROOKS, T., A. BALMFORD, N. BURGESS, J. FJELDSA, L.A.
HANSEN, J. MOORE, C. RAHBEK, and P. WILLIAMS. 2001.
“Towards a Blueprint for Conservation in Africa.” Bio-
Science 51:613-624.
BUTYNSKI, T.M. and J. KALINA. 1998. “Gorilla Tourism: A
Critical Review.” In: E.J. MILNER-GULLAND and R.
MACE (Eds.). Conservation of Biological Resources.
Blackwell, Oxford.
BUTYNSKI, T.M. 2001. “Africa’s Great Apes.” In: B.B. BECK,
T.S. STOINSKI, M. HUTCHINS, T.L. MAPLE, B. NORTON, A.
ROWAN, E.F. STEVENS, and A. ARLUKE (Eds.). Great Apes
and Humans. The Ethics of Coexistence. Smithsonian
Institution Press, Washington, D.C.
DAVENPORT, T.R.B. 2002. The Conservation of Mbizi Forest,
Sumbawanga, Rukwa Region: A Discussion Paper.
Wildlife Conservation Society.
FAO, 2001. Research for the Management of Fisheries on Lake
Tanganyika. Terminal Report FI: GCP/RAF/271/FIN.
HALL, J.S., K. SALTONSTALL, B.I. INOGWABINI, and I. OMARI.
1998. “Distribution, Abundance and Conservation Sta-
tus of Grauer’s Gorilla.” Oryx 323:122-130.
HOWARD, P.C. and T.R.B. DAVENPORT. 1996. Biodiversity Re-
ports,Vols 1-33. Uganda Forest Department, Kampala.
KANYAMIBWA, S., and H.J. TUMWEBAZE. 1999. Conservation
Priority-Setting for Albertine Rift Montane Forests. Un-
published workshop report, ARCOS.
LAWS, R.M., I.S.C. PARKER, and R.C.B. JOHNSTONE. 1975.
Elephants and their Habitats. Clarendon Press, Oxford.
OLSON, D.M. and E. DINERSTEIN. 1998. See references for
the Introduction.
OSMASTON, H. 1996. “The Influence of the Quaternary His-
tory and Glaciations of the Rwenzori on the Present
Landscape and Ecology.” In: H. OSMASTON, J. TUKAHIR-
WA, C. BASALIRWA, and J. NYAKAANA (Eds.). The Rwen-
zori Mountains National Park. Geography Department,
Makerere University.
OGUTU-OHWAYO, R. 1990. “The Decline of Native Species of
Lakes Victoria and Kyoga (E. Africa) and the Impact
of Introduced Species, Especially the Nile Perch, Lates
Niloticus and Nile Tilapia, Oreochromis niloticus.” En-
vironmental Biology of Fishes 27:81-96.
PATTERSON, G. and J. MAKIN. 1998. The State of Biodiversity
in Lake Tanganyika: A Literature Review. Natural Re-
sources Institute, Chatham, U.K.
PLUMPTRE, A.J. and S. HARRIS. 1995. “Estimating the Bio-
mass of Large Mammalian Herbivores in a Tropical
Montane Forest: A Method of Faecal Counting that
Avoids Assuming a ‘Steady State’ System.” Journal of
Applied Ecology 32:111-120.
PLUMPTRE, A.J., M. BEHANGANA, T.R.B. DAVENPORT, C.
KAHINDO, R. KITYO, E. NDOMBA, D. NKUUTU, I. OWIUN-
JI, P. SSEGAWA, and G. EILU. 2003a. The Biodiversity of
the Albertine Rift. Albertine Rift Technical Reports No.
3. www.albertinerift.org/publications
PLUMPTRE, A.J., D. COX, and S. MUGUME. 2003b. The Status
of Chimpanzees in Uganda. Albertine Rift Technical Re-
ports Series No. 2. www.albertinerift.org/publications
SHOLLEY, C.R. 1991. “Conserving Gorillas in the Midst of
Guerrillas.” Annual conference Proceedings, American
Association of Zoological Parks and Aquariums: 30-37.
SNOEKS, J. 2000. “How Well Known is the Icthyodiversity
of the Large East African Lakes.” In: A. ROSSITER and
H. KAWANABE (Eds.). “Ancient Lakes: Biodiversity,
Ecology and Evolution.” Advances in Ecological Re-
search 31:17-38. Academic Press.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
Ethiopian Highlands
DAVIDSON, A. and D.C. REX. 1980. “Age of Volcanism and
Rifting in South Western Ethiopia.” Nature 283:657-
658.
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1994. See references for the Introduction.
FRIIS, I. and O. RYDING. 2001. “Biodiversity Research in the
Horn of Africa Region.” In: Proceedings of the Third In-
ternational Symposium on the Flora of Ethiopia and Er-
itrea, August 25-27, 1999, Vol. 54. Biologiske Skrifter,
Carlsberg Academy, Copenhagen.
FRIIS, I., M. THULIN, H. ADSERSEN, and A.M. BÜRGER. In
press. See General References.
GAMACHU, D. 1977. Aspects of Climate and Water Budget in
Ethiopia. Addis Ababa University Press, Addis Ababa.
GILBERT, E.F. 1970. “Mt. Wachacha, a Botanical Commen-
tary.” Walia, Vol. 2.
GOTTELLI, D., C. SILLERO-ZUBIRI, G.D., APPLEBAUM, M.S.
387
 ROY, D.J., GIRMAN, J., GARCÍA-MORENO, E.A. OSTRAN-
DERS, and R.K. WAYNE. 1994. “Molecular Genetics of
the Most Endangered Canid: The Ethiopian Wolf Can-
is simensis.” Molecular Ecology, Vol. 3.
HARLAN, J.R. 1992. “Indigenous African Agriculture.” In:
P.J. WATSON and C.W. COWAN (Eds.). Agricultural Ori-
gins in World Perspective. Smithsonian Institution Pub-
lications in Anthropology, Washington, D.C.
HEDBERG, I. 1984. “Research on the Ethiopian Flora.” In: I.
HEDBERG (Ed.). Proceedings of the First Ethiopian Flora
Symposium, Vol. XXVI. Symb. Bot. Upsal, Uppsala,
Sweden.
HILLMAN, J.C. 1988. “The Bale Mountains National-Park
Area, Southeast Ethiopia, and its Management.”
Mountain Research and Development 8:253-258.
HILLMAN, J.C. 1993a. Ethiopia: Compendium of Wildlife
Conservation Information, Vol. I: Wildlife conservation
in Ethiopia. The NYZS, Wildlife Conservation Society,
Addis Ababa.
HILLMAN, J.C. 1993b. Ethiopia: Compendium of Wildlife
Conservation Information, Vol II: Compendium of in-
formation in individual Wildlife Conservation Ar-
eas. The NYZS, Wildlife Conservation Society, Addis
Ababa.
IUCN. 2003. 2003 IUCN Red List of Threatened Species.
www.redlist.org
KELBESSA, E., J.L. VIVERO, and S. DEMISSEW. 2003. “Progress
on the Red List of Plants of Ethiopia and Eritrea: The
Endemic Taxa, Conservation and Biogeography.” In:
Abstracts of the XVII AETFAT meeting, 21-26 September,
2003. Addis Ababa.
KINGDON, J. 1990. Island Africa: The Evolution of Africa’s
Rare Animals and Plants. Princeton University Press,
Princeton.
LARGEN, M.J. 1995. “A New Species of Chameleon (Reptil-
ia Sauria Chamaeleonidae) from Montane Forest in
Ethiopia.” Tropical Zoology 8:333-339.
LARGEN, M.J. 2001. “Catalogue of the Amphibians of
Ethiopia, Including a Key for their Identification.”
Tropical Zoology 14:307-402.
LARGEN, M.J. and J.B. RASMUSSEN. 1993. “Catalogue of the
Snakes of Ethiopia (Reptilia Serpentes), Including
Identification Keys.” Tropical Zoology 6:313-434.
LARGEN, M.J. and S. SPAWLS. In prep. “Lizards of Ethiopia:
A Checklist, Bibliography, Gazetteer, and Identifica-
tion Key.” Tropical Zoology.
MALCOLM, J.R. 1976. A Survey of the Simien Foxes (Canis
simensis) in the Bale Mountains National Park. Ethiopi-
an Wildlife Conservation Organisation, Addis Ababa.
MALCOLM, J.R. 1977. A Recensus of the Simien Fox in the
Bale Mountains National Park, Ethiopia. Ethiopian
Wildlife Conservation Organisation, Addis Ababa.
MALCOLM, J.R. 1988. A Recensus of Simien Jackals (Canis
simensis) in Bale, Ethiopia. Ethiopian Wildlife Conser-
vation Organisation, Addis Ababa.
NAGELKERKE, L.A.J. and F.A. SIBBING. 1997. “A Revision of
the Large Barbs (Barbus spp., Cyprinidae, Teleostei) of
Lake Tana, Ethiopia, with a Description of Seven New
Species.” In: L.A.J. NAGELKERKE (Ed.). The Barbs of
Lake Tana, Ethiopia: Morphological Diversity and its Im-
plications for Taxonomy, Trophic Resource Partitioning,
and Fisheries. Wageningen University, The Nether-
lands.
SILLERO-ZUBIRI, C. 1994. Behavioural Ecology of the Ethiopi-
an Wolf, Canis simensis. University of Oxford.
SILLERO-ZUBIRI, C. and D. GOTTELLI. 1995a. “Diet and Feed-
ing Behavior of Ethiopian Wolves (Canis simensis).”
Journal of Mammalogy 76:531-541.
SILLERO-ZUBIRI, C. and D. GOTTELLI. 1995b. “Spatial Orga-
nization in the Ethiopian Wolf Canis simensis: Large
Packs and Small Stable Home Ranges.” Journal of Zool-
ogy 237:65-81.
SILLERO-ZUBIRI, C., D. GOTTELLI, and D.W. MACDONALD.
1996. “Male Philopatry, Extra Pack Copulations and
Inbreeding Avoidance in Ethiopian Wolves (Canis
simensis).” Behavioral Ecology and Sociobiology 38:331-
340.
SILLERO-ZUBIRI, C., F.H. TATTERSALL, and D.W. MACDON-
ALD. 1995. “Habitat Selection and Daily Activity of Gi-
ant Molerats Tachyoryctes macrocephalus: Significance
to the Ethiopian Wolf Canis simensis in the Afroalpine
Ecosystem.” Biological Conservation 72:77-84.
SILLERO-ZUBIRI, C. and D.W. MACDONALD. 1997. The
Ethiopian Wolf: Status Survey and Conservation Action
Plan. IUCN, Gland and Cambridge.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
TEFERA, Z. 2001. Common Property Resource Management of
an Afro-Alpine Habitat: Supporting a Population of the
Critically Endangered Ethiopian Wolf, Canis simensis.
388
PhD thesis, Durrell Institute of Conservation and
Ecology, University of Kent at Canterbury.
YALDEN, D.W. 1983. “The Extent of High Ground in
Ethiopia Compared to the Rest of Africa.” Sinet:
Ethiopian Journal of Science 6:35-38.
YALDEN, D.W. and M.J. LARGEN. 1992. The Endemic Mam-
mals of Ethiopia. Mammal Review 22:115-150.
YALDEN, D.W., M.J. LARGEN, D. KOCK, and J.C. HILLMAN.
1996. “Catalogue of the Mammals of Ethiopia and Er-
itrea: Revised Checklist, Zoogeography and Conserva-
tion.” In: Tropical Zoology 9:73-164
Horn of Africa
AMORI, G. and S. GIPPOLITI. 2001. “Identifying Priority
Ecoregions for Rodent Conservation at the Genus Lev-
el.” Oryx 35:158-165.
ASH, J. S. and J.E. MISKELL. 1998. Birds of Somalia. Pica
Press, East Sussex.
AUDRU, J., J. CESAR, and J.-P. LEBRUN (Eds.). 1994. Les
plantes vasculaires de la République de Djibouti, Vols. 1
and 2. CIRAD, Département d’Elevage et de Médecine
Vétérinaire.
BEYDOUN, Z.R. and H.R. BICHAN. 1970. “The Geology of So-
cotra, Gulf of Aden.” The Quarterly Journal of the Geo-
logical Society of London 125:413-446.
CUMBERLIDGE, N. and W. WRANIK. 2002. A New Genus and
New Species of Freshwater Crab (Potamoidea: Potami-
dae) from Socotra Island, Yemen. Journal of Natural
History 36:65-77.
D’HUART, J.-P. and P. GRUBB. 2001. “Distribution of the
Common Warthog (Phacochoerus africanus) and the
Desert Warthog (Phacochoerus aethiopicus) in the Horn
of Africa.” African Journal of Ecology 39:146-155
EAST, R. 1999. African Antelope Database 1998. IUCN,
Gland and Cambridge.
EDWARDS, S., M. TADESSE, and I. HEDBERG (Eds.). 1995. Flo-
ra of Ethiopia and Eritrea, Vol. 2(2). Addis Ababa,
Ethiopia and Uppsala, Sweden.
EDWARDS, S., M. TADESSE, S. DEMISSEW, and I. HEDBERG
(Eds.). 2000. Flora of Ethiopia and Eritrea, Vol. 2(1). Ad-
dis Ababa, Ethiopia and Uppsala, Sweden.
EDWARDS, S., S. DEMISSEW, and I. HEDBERG (Eds.). 1997. Flo-
ra of Ethiopia and Eritrea, Vol. 6. Addis Ababa, Ethiopia
and Uppsala, Sweden.
FRIIS, I., M. THULIN, H. ADSERSEN, and A.M. BÜRGER. In
press. See General References.
GROOM, N. 1981. Frankincense and Myrrh. A study of the Ara-
bian Incense Trade. Longman, London and New York.
GROOMBRIDGE, B. (Ed.). 1994. Biodiversity Data Sourcebook.
Compiled by the World Conservation Monitoring Cen-
tre, IUCN-The World Conservation Union, United Na-
tions Environment Programme, World Wide Fund for
Nature. World Conservation Press, Cambridge.
HANNAH, L., J.L. CARR, and A. LANKERANI. 1995. “Human
Disturbance and Natural Habitat: A Biome Level
Analysis of a Global Data Set.” Biodiversity and Conser-
vation 4:128-155.
HEDBERG, I. and S. EDWARDS (Eds.). 1989. Flora of Ethiopia,
Vol. 3. Addis Ababa and Asmara, Ethiopia, and Upp-
sala, Sweden.
HEDBERG, I. and S. EDWARDS (Eds.). 1995. Flora of Ethiopia
and Eritrea, Vol. 7. Addis Ababa, Ethiopia and Uppsala,
Sweden.
HEDBERG, I., S. EDWARDS, and S. NEMOMISSA (Eds.). 2003.
Flora of Ethiopia and Eritrea, Vol. 4(1). Addis Ababa,
Ethiopia and Uppsala, Sweden.
LAVIN, M., M. THULIN, J.-N. LABAT, and R.T. PENNINGTON.
2000. “Africa, the Odd Man Out: Molecular Biogeogra-
phy of Dalbergioid Legumes (Fabaceae) Suggests Oth-
erwise.” Syst. Bot. 25:449-467.
LUCAS, G. and H. SYNGE. 1978. The IUCN Plant Red Data
Book. IUCN, Switzerland.
MILLER, A.G. and M. BAZARA’A. 1998. “The Conservation
Status of the Flora of the Soqotran Archipelago.” In:
H.J. DUMONT (Ed.). Proceedings of the First Internation-
al Symposium on Soqotra Island: Present and Future,
Vol. 1. United Nations Publications, New York.
MILLER, A.G. and T.A. COPE (Eds.). 1996. Flora of the Ara-
bian Peninsula and Socotra, Vol. 1. Edinburgh Univer-
sity Press.
SMITH, E.F.G., P. ARCTANDER, J. FJELDSÅ, and O. GEDOW
AMIR. 1991. “A New Species of Shrike (Laniidae: La-
niarius) from Somalia, Verified by DNA Sequence Data
from the Only Known Individual.” Ibis 133:227-235.
THULIN, M. (Ed.). 1993. Flora of Somalia, Vol. 1. Royal
Botanic Gardens, Kew.
THULIN, M. 1994. “Aspects of Disjunct Distributions and
Endemism in the Arid Parts of the Horn of Africa.” In:
J.H. SEYANI and A.C. CHIKUNI (Eds.). Proc. XIIIth Ple-
nary Meeting AETFAT, Malawi 2:1105-1119.
THULIN, M. (Ed.). 1995. Flora of Somalia, Vol. 4. Royal
Botanic Gardens, Kew.
THULIN, M. 1999. “Chapmannia (Leguminosae-Stylosan-
thinae) Extended.” Nordic. J. Bot. 19:597-607.
THULIN, M., B. BREMER, J. RICHARDSON, J. NIKLASSON, M.F.
FAY, and M.W. CHASE. 1998. “Family Relationships of the
Enigmatic Rosid Genera Barbeya and Dirachma from
the Horn of Africa Region.” Pl. Syst. Evol. 213:103-119.
WHITE, F. 1983. See references for the Introduction.
WHITE, F. and J. LÉONARD. 1991. “Phytogeographical Links
between Africa and Southwest Asia.” Fl. Veg. Mundi
9:229-246.
Irano-Anatolian
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1994. See references for the Introduction.
DAVIS, P.H. (Ed.). 1965-1985. Flora of Turkey and the East
Aegean Islands, Vol.1-9. Edinburg University Press, Ed-
inburgh.
DAVIS, P. 1971. “Distribution Patterns in Anatolia with Par-
ticular Reference to Endemism.” In: H. DAVIS and
HEDGE (Eds.). Plant Life of South-West Asia. Bot. Soci-
ety of Edinburgh, U.K.
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1995. See references for the Introduction.
EKEN, G. and G. MAGNIN. 1999. A Preliminary Biodiversity
Atlas of the Konya Basin, Central Turkey. Biodiversity
Program Report No. 13. Doǧal Hayatı Koruma
Derneǧi. İstanbul, Turkey.
EKEN, G., G. AYDEMIR, B. KURT, G. YALÇIN, E. BAŞAK, and
O.E. CAN. 2000. Türkiye’nin Biyolojik Çesitlilik Atlasi.
Yesil Atlas 3, Istanbul.
EKIM, T., and A. GÜNER. 1986. “The Anatolian Diagonal:
Fact or Fiction.” In: I.C. HEDGE (Ed.). Plant life of
South-West Asia. Botanical Society of Edinburgh, U.K.
EKIM, T., M. KOYUNCU, M. VURAL, H. DUMAN, Z. AYTAÇ, and
N. ADIGÜZEL. 2000. Türkiye Bitkileri Kırmızı Kitabı.
Türkiye Tabiatını Koruma Derneǧi, Ankara.
EVANS, M.I. 1994. Important Bird Areas in the Middle East.
BirdLife Conservation Series 2. BirdLife International,
Cambridge.
FEH, C., N. SHAH, M. ROWEN, R. READING, and S.P. GOYAL.
2002. “Status and Action Plan for the Asiatic Wild Ass
(Equus hemionus).” In: P.D. MOEHLMAN (Ed.). Equids:
Zebras, Asses and Horses: Status Survey and Conserva-
tion Action Plan. IUCN/SSC Equid Specialist Group.
IUCN, Gland and Cambridge.
FET, V., W.D. SISSOM, G. LOWE, and M.E. BRAUNWALDER.
2000. Catalog of the Scorpions of the World (1758-1998).
The New York Entomological Society, New York.
HESSELBARTH G., G. WAGENER, and H. VAN OORSCHOT,1995.
Die Tagfalter der Türkei, unter Berücksichtigung der an-
grenzenden Lander. Selbstverlag Sigbert Wagener Bo-
cholt Germany.
KILIÇ, T. and G. EKEN. 2004. Türkiye’nin Önemli Kuş Alan-
ları Güncellemesi. Doǧa Derneǧi. BirdLife Internation-
al, Ankara.
MAGNIN, G. and M. YARAR. 1997. Important Bird Areas in
Turkey. Doǧal Hayatı Koruma Derneǧi. İstanbul.
ÖZHATAY, N., A. BYFIELD, and S. ATAY. 2003. Türkiye’nin
Önemli Bitki Alanları. WWF-Türkiye, İstanbul.
www.sifiryokolus.org
PERELADOVA, O., V. KREVER, and M. WILLIAMS. 1997. Biodi-
versity Conservation in Central Asia. Moscow
RONKAY, L., Z. VARGA, and M. HREBLAY. 1998. “Twenty Two
New Species and Six New Subspecies of Noctuidae
from Turkmenistan and Adjacent Regions (Lepi-
doptera).” Acta Zoologica Academiae Scientiarum Hun-
garicae 44(3):205-281.
SONNENFELD, P. 1974. “The Upper Miocene Evaporite
Basins in the Mediterranean Region: A Study in Paleo-
Oceanography.” Geol. Rundschen 63:1133-1172.
VURAL, M., T. EKIM, H. DUMAN, Z. AYTAÇ, and N. ADIGÜZEL.
1999. Vegetation and Flora of the Konya Closed Basin.
Biodiversity Programme Report No 12. Doǧal Hayatı
Koruma Derneǧi-Gazi Universitesi, İstanbul.
WELCH, H. 2003. South East Anatolia Biodiversity Research
Project: Draft Outline Report with Full Recommendations.
Doǧal Hayatı Koruma Derneǧi, İstanbul.
ZOHARY, M. 1973. See references for the Introduction.
Mountains of Central Asia
BORKIN, L.J. 1999. “Distribution of Amphibians in North
Africa, Europe, Western Asia and the Former Soviet
Union.” In: W.E. DUELLMAN (Ed.). Patterns of Distribu-
tion of Amphibians: A Global Perspective. John Jopkins
University Press, Baltimore and London.
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1995. See references for the Introduction.
GEF/UNDP. 1998. Republic of Uzbekistan: Biodiversity Con-
servation, National Strategy and Action Plan. Tashkent.
MINISTRY OF ENVIRONMENTAL PROTECTION. 1998. Kyrgyz Re-
public Biodiversity Strategy and Action Plan. Bishkek.
REGIONAL ECOLOGICAL ASSESSMENT CENTRE OF CENTRAL
ASIA. 2003. Millennium Ecosystem Assessment. Almaty.
SHACKLETON, D.M. (Ed.) and IUCN/SSC CAPRINAE SPECIAL-
IST GROUP. 1997. Wild Sheep and Goats and their Rela-
tives: Status Survey and Conservation Action Plan for
Caprinae. IUCN, Gland and Cambridge.
TISHKOV, A. 2002. “Nature Protection and Conservation.”
In: M. SHAHGEDANOVA (Ed.). The Physical Geography
of Northern Eurasia. Oxford University Press, Ox-
ford.
USAID. 2001a. Biodiversity Assessment for Central Asia: Re-
gional Overview. USAID Central Asian Republics Mis-
sion, Almaty.
USAID. 2001b. Biodiversity Assessment for Kazakhstan. US-
AID Central Asian Republics Mission, Almaty.
USAID. 2001c. Biodiversity Assessment for Kyrgyzstan. US-
AID Central Asian Republics Mission, Almaty.
USAID. 2001d. Biodiversity Assessment for Tajikistan. US-
AID Central Asian Republics Mission, Almaty.
ZHANG, B., Y. YAO, W. CHENG, C. ZHOU, Z. LU, X. CHEN, K.
ALSHIR, I. ERDOWLET, L. ZHANG, and Q. SHI. 2002. “Hu-
man-Induced Changes to Biodiversity and Alpine Pas-
tureland in the Bayanbulak Region of the Eastern
Tien Shan Mountains.” Mountain Research and Devel-
opment 22(4):383-389.
ZLOTIN, R.I. 1999. “Geography and Organization of High-
Mountain Ecosystems in the Former USSR.” In: F.E.
WIEGOLASKI (Ed.). Ecosystems of the World 3: Polar and
Alpine Tundra. Elsevier, Amsterdam.
Himalaya
ALNUTT, T., E. WIKRAMANAYAKE, E. DINERSTEIN, C. LOUCKS,
R. JACKSON, D. HUNTER, and C. CARPENTER. 2002.
“Composition of the Alpine Himalayan Protected Ar-
eas Network and its Contribution to Biodiversity Con-
servation.” In: E.D. WIKRAMANAYAKE, E. DINERSTEIN, C.
LOUCKS, D. OLSON, J. MORRISON, J. LAMOREUX, M.
MCKNIGHT, and P. HEDAO (Eds.). 2002. See General
References.
BIODIVERSITY ACTION PLAN. 2002. Biodiversity Action Plan
for Bhutan 2002. Prepared and published by the Min-
istry of Agriculture, Royal Government of Bhutan.
GADGIL, M. 1985. “Social Restraints on Resource Utiliza-
tion: The Indian Experience.” In: J.A. MCNEELY and D.
PITT (Eds.). Culture and Conservation: The Human Di-
mension in Environmental Planning. Croom Helm,
Dublin.
GUPTA, R.K. 1994. “Arcto-Alpine and Boreal Elements in
the High Altitude Flora of the North West Himalaya.
In: Y.P.S. PANGTEY and R.S. RAWAL (Eds.). High Altitudes
of the Himalaya. (Biogeography, Ecology and Conserva-
tion). Gyanodaya Prakashan, Nainital, Delhi.
HARIBAL, M. 1992. Butterflies of Sikkim Himalaya and their
Natural History. Nature Conservation Foundation,
Gangtok, Sikkim.
IUCN. 1990. IUCN Directory of South Asian Protected Areas.
IUCN, Gland and Cambridge.
KARANTH, K.U. and J.D. NICHOLS. 1998. “Estimating Tiger
Densities in India from Camera Trap Data Using Pho-
tographic Captures and Recaptures.” Ecology 79:2852-
2862.
MANI, M.S. 1986. Butterflies of the Himalaya. Oxford and
IBH Publishing Co., New Delhi.
MANI, M.S. 1994. “The Himalaya, its Ecology and Biogeog-
raphy: A Review.” In: Y.P.S. PANGTEY and R.S. RAWAL
(Eds.). High Altitudes of the Himalaya. (Biogeography,
Ecology and Conservation). Gyanodaya Prakashan,
Nainital, Delhi.
MILLER, D. 2002. “Pastures of Plenty? Livestock Grazing
and Conservation of Himalayan Ranglands.” In: E.D.
WIKRAMANAYAKE, E. DINERSTEIN, C. LOUCKS, D. OLSON,
J. MORRISON, J. LAMOREUX, M. MCKNIGHT, and P.
HEDAO (Eds.). 2002. See General References.
MOLNAR, P. 1986. “The Geologic History and Structure of
the Himalaya.” American Scientist 74:144-154
POLUNIN, O. and A. STAINTON. 1997. Flowers of the Hi-
malaya. Oxford India Paperbacks. Delhi.
RAM, J. and S.P. SINGH. 1994. “Ecology and Conservation of
Alpine Meadows in Central Himalaya, India.” In:
Y.P.S. PANGTEY and R.S. RAWAL (Eds.). High Altitudes of
the Himalaya. (Biogeography, Ecology and Conserva-
tion). Gyanodaya Prakashan, Nainital, Delhi.
RISEBROUGH, R. 2004. “Fatal Medicine for Vultures.” Nature
426:596-598.
SAHNI, K.C. 1979. “Endemic, Relict, Primitive and Spectac-
ular Taxa in Eastern Himalaya and Strategies for their
Conservation.” Indian J. For. 2 (2):181-190.
SHERPA, M., S. WANGCHUK, E.D. WIKRAMANAYAKE. 2003.
“Linking Protected Areas for Ecosystem Conservation:
A Case Study from Bhutan.” A Paper presented at the
World Parks Congress, Durban, South Africa.
SINGH, N.P., D.K. SINGH, P.K. HAJRA, and B.D. SHARMA
(Eds.). 2000. Flora of India, Introductory Volume (part
II). Botanical Survey of India. Calcutta.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
STIRN, A and P. VAN HAM. 2001. The Seven Sisters of India:
Tribal Worlds between Tibet and Burma. Prestal Pub-
lishers, U.S.A.
SUKUMAR, R. 1992. The Asian Elephant. Ecology and Man-
agement. Cambridge University Press. Cambridge.
UNDP. 1998. Ecoregional Co-Operation for Biodiversity Con-
servation in the Himalayas. Proceedings of a regional
meeting organized by UNDP in co-operation with
WWF and ICIMOD.
WIKRAMANAYAKE, E.D., E. DINERSTEIN, T. ALLNUT, C.
LOUCKS, and W. WETTENGEL. 1998. A Biodiversity As-
sessment and Gap Analysis of the Himalayas. World
Wildlife Fund-U.S., Conservation Science Program/
UNDP Rept.
WIKRAMANAYAKE, E.D., E. DINERSTEIN, C. LOUCKS, D. OL-
SON, J. MORRISON, J. LAMOREUX, M. MCKNIGHT, and P.
HEDAO (Eds.). 2002. See General References.
WWF NEPAL. 2004. WWF Nepal annual report published by
WWF in Kathmandu, Nepal.
Indo-Burma
AMATO, G., M. EGAN, and A. RABINOWITZ.1999. “A New
Species of Muntjac Muntiacus putaoensis (Artiodacty-
la: Cervidae) from Northern Myanmar.” Animal Con-
servation 2:1-7.
AVERIANOV, A.O., A.V. ABRAMOV, and A.N. TIKHONOV. 2000.
“A New Species of Nesolagus (Lagomorpha, Lepori-
dae) from Vietnam with Osteological Description.”
Contributions from the Zoological Institute, St. Petersburg
3:1-22.
BIRDLIFE INTERNATIONAL. 2003. See General References.
BIRDLIFE INTERNATIONAL. 2004. Threatened Birds of the
World 2004. CD Rom. BirdLife International, Cam-
bridge.
CAREW-REID, J. (Ed.). 2002. “Lessons Learned from Protected
Areas Management Experience in Thailand: 1992-2001.”
Country lessons paper prepared for the Review of Pro-
tected Areas and their Role in the Socio-economic De-
velopment in the Four Countries of the Lower Mekong
River Region. http://www.mekong-protected-areas.org
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1995. See references for the Introduction.
DIAMOND, J. 1997. Guns, Germs and Steel: A Short History of
Everybody for the Last 13,000 Years. Chatto and Win-
dus, U.K.
DIJK, P.P.V., P. ASHTON and J. MA. 1999. “Indo-Burma.” In:
R.A. MITTERMEIER, N. MYERS, P. ROBLES-GIL, and C.G.
MITTERMEIER (Eds.). 1999. See references for the In-
troduction.
DO T., V.D. VU, S. DAWSON, P. ARCTANDER, and J. MACKIN-
NON. 1994. Introduction of a New Large Mammal Species
in Vietnam. Forest Inventory and Planning Institute,
Hanoi. (In Vietnamese.)
DUNG, V.V., M.G. PHAM, N.C. NGUYEN, T. DO, and J. MACK-
INNON. 1994. “Discovery and Conservation of the Vu
Quang Ox in Vietnam.” Oryx 28:16-21.
GIAO, P.M., D. TUOC, V.V. DUNG, E.D. WIKRAMANAYAKE, G.
AMATO, P. ARCTANDER, and J.R. MACKINNON. 1998. “De-
scription of Muntiacus truongsonensis, a New Species
of Muntjac (Artiodactyla: Muntiacidae) from Central
Vietnam, and Implications for Conservation.” Animal
Conservation 1:61-68.
HUGHES, J.B., P.D. ROUND, and D.S. WOODRUFF. 2003. “The
Indochinese-Sundaic Faunal Transition at the Isthmus
of Kra: An Analysis of Resident Forest Bird Species
Distributions. Journal of Biogeography 30:569-580.
MACKINNON, J., M. SHA, C. CHEUNG, G. CAREY, X. ZHU, and
D. MELVILLE. 1996. See General References.
MITTERMEIER, R.A., W.R. KONSTANT, A.B. RYLANDS, J.
GANZHORN, J.F. OATES, T.M. BUTYNSKI, T. NADLER, J.
SUPRIATNA, C.V. PADUA, and D. RAMBALDI. 2002. Pri-
mates in Peril: The World’s Top 25 Most Endangered Pri-
mates. Conservation International, Margot Marsh Bio-
diversity Foundation, IUCN-SSC Primate Specialist
Group and International Primatological Society,
Washington, D.C.
NADLER, T. 1997. “A New Subspecies of Douc Langur, Py-
gathrix nemaeus cinereus ssp. nov.” Zool. Garten N. F.
67:165-176.
REDFORD, K.H. 1992. “The Empty Forest.” BioScience
42:412-422.
SOLHEIM II., W.G. 1972. “An Earlier Agricultural Revolu-
tion.” Scientific American 266(4):34-41.
STAMP, L.D. 1925. The Vegetation of Burma from an Ecologi-
cal Standpoint. Thacker, Spink and Co., Calcutta.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
VAN STEENIS, C.G. 1950. See General References.
TIMMINS, R.J., T.D. EVANS, K. KHOUNBOLINE, and C. SISOM-
PHONE. 1998. “Status and Conservation of the Giant
Muntjac Megamuntiacus vuquangensis and Notes on
Other Muntjac Species in Lao P.D.R.” Oryx 32:59-67.
TORDOFF, A.W., Q.B. TRAN, D.T. NGUYEN, and H. LE MANH
(Eds.). 2004. Sourcebook of Existing and Proposed Pro-
tected Areas in Vietnam. 2nd edition. BirdLife Interna-
tional in Indochina and the Ministry of Agriculture
and Rural Development, Hanoi.
WHITMORE, T.C. 1984. Tropical Rain Forests of the Far East.
Clarendon, Oxford.
WOODRUFF, D.S. 2003. “Neogene Marine Transgressions,
Palaeogeography and Biogeographic Transitions on
the Thai-Malay Peninsula.” Journal of Biogeography
30:551-567.
ZIEGLER, T., H.-W. HERRMANN, P. DAVID, N.L. ORLOV, and
O.S.G. PAUWELS. 2000. “Triceratolepidophis sieversorum,
a New Genus and Species of Pitviper (Reptilia: Ser-
pentes: Viperidae: Crotalinae) from Vietnam.” Russian
Journal of Herpetology 7:199-214.
Japan
BIRDLIFE INTERNATIONAL. 2003. See General References.
BOUFFORD, D.E. and S.A. SPONGBERG. 1983. “Eastern Asian-
Eastern North American Phytogeographical Relation-
ships: A History from the Time of Linnaeus to the Twen-
tieth Century.” Ann. Missouri Bot. Gard. 70:423-439.
EXPO’90 FOUNDATION. International Cosmos Prize (2 Febru-
ary, 2004). http://www.expo-cosmos.or.jp/index.html
FITZHUGH, W.W. and C.O. DUBREIL, (Eds.). 2000. Ainu: Spirit
of a Northern People. Arctic Studies Center, National Mu-
seum of Natural History, Smithsonian Institution in as-
sociation with University of Washington Press, Seattle.
ITO, Y., K. MIYAGI, and H. OTA. 2000. “Imminent Extinction
Crisis among the Endemic Species of the Forests of
Yanbaru, Okinawa, Japan.” Oryx 34(4):305-316
IWATSUKI, K., T. YAMAZAKI, D.E. BOUFFORD, and H. OHBA
(Eds.). 1993. Flora of Japan, Vol. IIIa: Angiospermae,
Dicotyledoneae, Sympetalae (a). Kodansha, Ltd.,
Tokyo.
IWATSUKI, K., T. YAMAZAKI, D.E. BOUFFORD, and H. OHBA
(Eds.). 1995a. Flora of Japan, Vol. IIIb: Angiospermae,
Dicotyledoneae, Sympetalae (b). Kodansha, Ltd., Tokyo.
IWATSUKI, K., T. YAMAZAKI, D.E. BOUFFORD, and H. OHBA
(Eds.). 1995b. Flora of Japan, Vol. I: Pteridophyta and
Gymnospermae. Kodansha, Ltd., Tokyo.
IWATSUKI, K., D.E. BOUFFORD, and H. OHBA (Eds.). 2001.
Flora of Japan, Vol. IIb: Angiospermae; Dicotyle-
donieae; Archichlamideae(b). Kodansha, Ltd.,
Tokyo.
KITAYAMA, K. 1991. “Threatened Endemic Species of the
Bonin (Ogasawara) Islands.” Pac. Sci. Assoc. Inform.
Bull. 43(3/4):9-10.
MATSUKA, H. 2003. A List of the Butterflies from Japan.
http://www02.so-net.ne.jp/~jamides/jplist/jplist-e.html
POPULATION REFERENCE BUREAU. 2003. See General Refer-
ences.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
WEN, J. 1999. “Evolution of Eastern Asian and Eastern
North American Disjunct Pattern in Flowering
Plants.” Ann. Rev. Ecol. Syst. 30:421-455.
WIKRAMANAYAKE, E.D., E. DINERSTEIN, C. LOUCKS, D. OL-
SON, J. MORRISON, J. LAMOREUX, M. MCKNIGHT, and P.
HEDAO (Eds.). 2002. See General References.
East Melanesian Islands
ALLISON, A. 2001. “The Herpetofauna of Southern New
Ireland.” In: B.M. BEEHLER and L.E. ALONSO (Eds.).
Southern New Ireland, Papua New Guinea: A Biodiver-
sity Assessment. RAP Bulletin of Biological Assess-
ment 21. Conservation International, Washington,
D.C.
BAYLISS-SMITH, T., E. HVIDING, and T. WHITMORE. 2003.
Rainforest Composition and Histories of Human Dis-
turbance in Solomon Islands. Ambio 32:346-352.
D’ABRERA, B. 1990. Butterflies of the Australian Region. Hill
House, Melbourne.
FLANNERY, T.F. 1995. See General References.
HONZA, E., H.L. DAVIES, J.B. KEENE, and D.L. TIFFIN. 1987.
“Plate Boundaries and Evolution of the Solomon Sea
Region.” Geo-Marine Letters 7:161-168.
JOHNS, R.E. 1993. “Biodiversity and Conservation of the
Native Flora of Papua New Guinea.” In B. BEEHLER,
(Ed.). A Biodiversity Analysis for Papua New Guinea.
Conservation Needs Assessment, Vol. 2. Biodiversity
Support Program, Washington, D.C.
MAYR, E. and J.M. DIAMOND. 2001. The Birds of Northern
Melanesia. Harvard, Cambridge, Massachusetts.
MCCOY, M. 1980. Reptiles of the Solomon Islands. Wau Ecolo-
gy Institute Handbook No. 7. Wau, Papua New Guinea.
MUELLER-DOMBOIS, D. and F.R. FOSBERG. 1998. See Gener-
al References.
READ, T. 2002. Navigating a New Course: Stories in Commu-
nity-Based Conservation in the Pacific Islands. United
Nations Development Program, New York.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
THISTLETHWAIT, R. and G. VOTAW. 1992. Environment and
Development: A Pacific Island Perspective. Asian Devel-
opment Bank, Manila and South Pacific Regional En-
vironment Programme, Apia.
UNDP. 1999. Pacific Human Development Report 1999: Cre-
ating Opportunities. UNDP, Suva.
WIKRAMANAYAKE, E.D., E. DINERSTEIN, C. LOUCKS, D. OL-
SON, J. MORRISON, J. LAMOREUX, M. MCKNIGHT, and P.
HEDAO (Eds.). 2002. See General References.
Taiwan
BIRDLIFE INTERNATIONAL. 2003. See General References.
CHAPPELL, J. and N.J. SHACKLETON. 1986. “Oxygen Iso-
topes and Sea Level.” Nature 324:137-140.
CHI, W.L. 1998. The Mammals of Taiwan. Big Tree Publish-
ing Company, Taipei.
DAVIS, S.D., V.H. HEYWOOD, and A.C. HAMILTON (Eds.).
1995. See references for the Introduction.
EDITORIAL COMMITTEE OF THE FLORA OF TAIWAN. 1993-
2002. Flora of Taiwan, 2nd edition, Vol. 1-6. Depart-
ment of Botany, National Taiwan University,
Taipei.
FERGUSON, D.K. 1993. “The Impact of Late Cenozoic Envi-
ronmental Changes in East Asia on the Distribution of
Terrestrial Plants and Animals.” In: N.G. JABLONSKI
(Ed.). Evolving Landscapes and Evolving Biotas of East
Asia since the Mid-Tertiary. Proceedings of the third
conference on the evolution of the East Asian envi-
ronment. Centre of Asian Studies, the University of
Hong Kong, Hong Kong.
HSIEH, C.F. 2002. “Composition, Endemism, and Phygo-
geographcal Affinities of the Taiwan Flora.” In: Flora of
Taiwan, 2nd Edition, Vol. 6. Department of Botany.
National Taiwan University, Taipei.
LIU, H.J. 2003. Investigation of the Hybridization of Styan’s
Bulbul and Chinese Bulbul in Kenting Area. Kenting Na-
tional Park.
POPULATION REFERENCE BUREAU. 2003. See General Refer-
ences.
SEVERINGHAUS, L.L. and L. CHI. 1999. “Prayer Animal
Release in Taiwan.” Biological Conservation 89:301-
304.
SEVERINGHAUS, L.L., K. BROUWER, S. CHAN, J.R. CHONG,
M.C. COULTER, E.P.R. POORTER, and Y. WANG. 1995.
“Task Force to Develop an Action Plan for the Preser-
vation of the Black-Faced Spoonbill.”Action plan for
the black-faced spoonbill platalea minor. The Wild
Bird Society of ROC, Taipei.
STATTERSFIELD, A.J., M.J. CROSBY, A.J. LONG, and D.C.
WEGE. 1998. See references for the Introduction.
TAKHTAJAN, A. 1986. See General References.
TENG, L.S. 1990. “Geotectonic Evolution of Late Cenozoic
Arc-Continent Collision in Taiwan.” Tectonophysics
183:57-76
YANG, K.-C. and C.-N. WANG. 2002. “Largest Extensions
to the Natural Range of Taiwania Cryptomerioides:
Largest Known Population Discovered.” Fitzroya,
the newsletter of the IUCN Conifer Specialist Group
5:5.
Queensland Wet Tropics
BERGER, L., R. SPEARE, P. DASZAK, D.E. GREEN, A.A. CUN-
NINGHAM, C.L. GOGGIN, R. SLOCOMBE, M.A. RAGAN,
A.D. HYATT, K.R. MCDONALD, H.B. HINES, K.R. LIPS, G.
MARANTELLI, and H. PARKES. 1998. “Chytridiomycosis
Causes Amphibian Mortality Associated with Popula-
tion Declines in the Rain Forests of Australia and Cen-
tral America.” Proceedings of the National Academy of
Sciences of the United States of America 95:9031-9036.
BONELL, M., D.A. GILMOUR, and D.S. CASSELLS. 1991. “The
Links between Synoptic Climatology and the Runoff
Response of Rainforest Catchments on the Wet Tropi-
cal Coast of North-Eastern Queensland.” In: G. WER-
REN and P. KERSHAW (Eds.). The Rainforest Legacy, Aus-
tralian National Rainforest Study, Vol. 2. Australian
Heritage Commission, Canberra.
BOTTOMS, T. 2000. “Bama Country: Aboriginal Home-
lands.” In: G. MCDONALD, and M. LANE (Eds.). Securing
the Wet Tropics? Federation Press.
DASETT (DEPARTMENT OF THE ARTS, SPORT, THE ENVIRON-
MENT, TOURISM AND TERRITORIES). 1987. “Nomination
of the Wet Tropical Rainforests of North-East Australia
by the Government of Australia for Inclusion in the
World Heritage List.” Canberra.
GADEK, P.A. and S.WORBOYS. 2003. Rainforest Dieback Map-
ping and Assessment: Phytophthora Species Diversity
and Impacts of Dieback on Rainforest Canopies. Co-op-
erative Research Centre for Tropical Rainforest Ecolo-
gy and Management, Rainforest CRC, Cairns.
GOOSEM, S., G. MORGAN, and J.E. KEMP. 1999. “Wet Trop-
ics.” In: P.S. SATTLER and R.D. WILLIAMS (Eds.). The
Conservation Status of Queensland’s Bioregional Ecosys-
tems. Environmental Protection Agency, Brisbane.
HORSFALL, N. 1984. “The Prehistoric Occupation of Aus-
tralian Rainforests.” In: G.L. WERREN and A.P. KER-
SHAW (Eds.). The Rainforest Legacy: Australian National
Rainforests Study, Vol. 3. Australian Government Pub-
lishing Service, Canberra.
MAXWELL, S., A.A. BURBIDGE, and K. MORRIS (Comps.). 1996.
Action Plan for Australian Marsupials and Monotremes.
IUCN/SSC Australasian Marsupial and Monotreme
Specialist Group.
MITTERMEIER, R.A., P. ROBLES-GIL, and C.G. MITTERMEIER
(Eds.). 1997. See references for the Introduction.
SLUITER, J.R. and A.P. KERSHAW. 1982. “The Nature of Late
Tertiary Vegetation in Australia.” Alcheringa 6.
STANTON, J.P. and M.G. MORGAN. 1976. Criteria and Meth-
ods for the Rapid Selection and Appraisal of Key and En-
dangered Sites: The Queensland Case Study. A report to
the Department of Environment, Housing and Com-
munity Development. Univ. of New England.
STRAHAN, R. (Ed.). 1995. The Mammals of Australia. Aus-
tralian Museum/New Holland.
WHITE, M.E. 1986. The Greening of Gondwana. Reed Books
Pty Ltd NSW.
WILLIAMS, S.E., E.E. BOLITHO, and S. FOX. 2003. “Climate
Change in Australian Tropical Rainforests: An Im-
pending Environmental Catastrophe.” Proceedings of
the Royal Society of London Series B-Biological Sciences
270:1887-1892.
WILSON, B.A. V.J. NELDNER, and A. ACCAD. 2002. “The Ex-
tent and Status of Remnant Vegetation in Qld: Its Im-
plications for Statewide Vegetation Management and
Legislation.” Rangel, J. 24(1):6-35.
389

On p. 380, seaside daisies and
stacks, northern California Coast,
U.S.A.
© Carr Clifton/Minden Pictures
On p. 392, Spanish ibex
(Capra pyrenaica) in the Sierra
de Gredos, Spain. Although not
considered threatened, at least one
subspecies (C. p. pyrenaica)
was declared Extinct as recently as
2000, when a tree fell on the last
individual in the Ordesa
National Park.
© Patricio Robles Gil/Sierra Madre
390
ADDRESSES
1. Conservation International,1919 M St NW, Suite 600, Washington, DC 20036, U.S.A. 69. Environment and Social Development Sector, East Asia and Pacific Region, The World PRODUCTION Agrupación Sierra Madre, S.C. and Redacta, S.A. de C.V.
2. Center for Applied Biodiversity Science-CI, 1919 M St NW, Suite 600, Washington, DC Bank, 1818 H St NW, Washington, DC, 20433, U.S.A.
20036, U.S.A. 70. BirdLife Indonesia, Jalan Dadali 32, Box 310 Boo, Bogor 16161, INDONESIA EDITORIAL DIRECTION Antonio Bolívar EDITORIAL REVISION AND CORRECTION Susan Beth Kapilian
3. Department of Environmental Sciences, University of Virginia, Charlottesville, VA 71. IUCN/SSC Tortoise & Freshwater Turtle Specialist Group, Krimweg 125, 7351 TL
22903, U.S.A. Hoenderloo, THE NETHERLANDS GENERAL COORDINATION Ana Ezcurra, Eugenia Pallares EDITORIAL COORDINATION Oswaldo Barrera
4. Agrupación Sierra Madre, S.C. and Unidos para la Conservación, A.C., Primero de 72. Conservation Biology Building, College of Life Sciences, Peking University, Beijing
Mayo 249, San Pedro de los Pinos, 03800 México, D.F., MEXICO 100871, P.R. CHINA GRAPHIC DESIGN Juan Carlos Burgoa PHOTOGRAPHIC COORDINATION Roxana Vega TECHNICAL SUPPORT Elena León
5. Conservation International-Brazil Getulio Vargas, 1300-7o andar, 30112-021 Belo 73. Wildlife Conservation Society, P.O. Box 1475, Mbeya, TANZANIA
Horizonte, MG, BRAZIL 74. Makerere Institute for Environment and Natural Resources, Makerere University, MAPS AND CARTOGRAPHY Álvaro Couttolenc, Mark Denil TYPESETTING Socorro Gutiérrez
6. Apt. 706, 1616 16th St NW, Washington, DC 20009, U.S.A. Kampala, UGANDA
7. Houston Zoo, 1513 N. MacGregor, Houston, TX 77030, U.S.A. 75. World Wildlife Fund-Eastern Africa Regional Programme, Office International Centre
8. Upper Meadow, Douglas Downes Close, Quarry Road, Headington, Oxford, OX3 8NT, UK of Insect Physiology and Ecology, Box 30772, Nyayo Stadium 00506, Nairobi, KENYA
9. Museum of Comparative Zoology, Harvard University, 26 Oxford Street, Cambridge, MA 76. WWF EARPO, P.O. Box 62440, 00200 Nairobi, KENYA
02138, U.S.A. 77. World Wildlife Fund-Tanzania Programme Office, P.O. Box 63117, Dar es Salaam,
10. American Bird Conservancy, 1834 Jefferson Place NW, Washington, DC 20036, U.S.A. TANZANIA
11. School of Life Sciences, Arizona State University, P.O. Box 874501, Tempe, AZ 85287, 78. Tanzania Forest Conservation Group, P.O. Box 23410, Dar es Salaam, TANZANIA ACKNOWLEDGEMENTS
U.S.A. 79. UNDP-GEF Regional Bureau for Africa, P.O. Box 1041 Arusha, TANZANIA
12. IUCN/SSC-CI/CABS Biodiversity Assessment Unit, Washington, DC 20036, U.S.A. 80. The Hermitage, Crewkerne, Somerset, UK
13. Universidade Federal de Minas Gerais, Belo Horizonte, 31270 MG, BRAZIL 81. Programa Especial de Seguridad Alimentaria, FAO (TCOS), Sololá, GUATEMALA First of all, we would like to thank CEMEX for mak- Kaipu, Sam Kanyamibwa, Mark Keith, Philippe piélago de Cabrera, we particularly thank Jorge
14. Universidade Estadual de Santa Cruz, km 16 Rodovia Illhéus-Itabuna CEP-45662-000 82. 9 Southerwood, Norwich NR 6 6 JN, ENGLAND
Ilhéus, Bahia, BRAZIL 83. The National Herbarium Addis Abeba, University P.O. Box 3434 Addis Ababa, ETHIOPIA ing the production of this book and the many oth- Keith, Susan L. Kelley, Michael Kennedy, Ronan Moreno, Francisco López, and Gaby Coll. We would
15. Departamento de Zoologia, Universidade de Brasília, 70910, Brasília, BRAZIL 84. Centre for Ecology, Law and Policy Environment Department, University of York,
16. National Botanical Institute, Private Bag X101, Pretoria 0001, SOUTH AFRICA York YO10 5DD, UK ers that have come before it a reality. Their com- Kirsch, Sandra Knapp, Cyril Kormos, Maurice also like to acknowledge the help and advice re-
17. Agricultural Research Council, Rangelands & Forestry Institute, Private Bag X05, Lynn 85. MNRT, P.O. Box 9372, Dar es Salaam, TANZANIA
East 0039, SOUTH AFRICA 86. Eastern Africa Biodiversity Hotspots, Conservation International, c/o IUCN, P.O. Box
mitment, their vision, and their support for Kottelat, Chien-Wei Kuan, Gigi Laidler, Pei-Jing ceived from our friends and colleagues at BirdLife
18. Department of Systematic Botany, Evolutionary Biology Centre (EBC), Uppsala 68200, Langata 00200, Nairobi, KENYA biodiversity conservation around the world are Lan, Malcolm Largen, Sebastien Lavergne, Pei-Jen International. Other special thanks are due to the
University, Norbyvägen 18D, SE-752 36 Uppsala, SWEDEN 87. Bishop Museum, 1525 Bernice Street, Honolulu, Hawaii 96817, U.S.A.
19. BirdLife International, Wellbrook Court, Girton Road, Cambridge, CB3 0NA, UK 88. Centre for Wildlife Studies, Jayanagar, Block VII, Bangalore, INDIA 560082 deeply appreciated. In addition, the editors and au- Lee, Keng-Hsien Lin, Frida Lindemalm, Jesús World Wildlife Fund-U.S. for the use of their global
20. Niǧde University, Faculty of Science and Art, Department of Biology, 51200 Niǧde, TURKEY 89. The Wildlife Heritage Trust of Sri Lanka, 95 Cotta Road, Colombo 8, SRI LANKA
21. Oregon State University, Envrionmental Science Program, Cordley 2082 Corvallis, OR 90. Rainforest Research Station, Valparai, INDIA 642127 thors of this book would like to acknowledge the López García, Staci Markos, Vance G. Martin, ecoregion vertebrate species data, to Fabio Cassola,
97330, U.S.A. 91. World Wildlife Fund, 25 Araliya Mawatha, Sirimal Guyana, Ratmalana, SRI LANKA
22. Doǧa Derneǧi, PK 640 06445 Yenişehir Ankara, TURKEY 92. Wildlife Institute of India, Dehradun, INDIA
following individuals and institutions, either for Michael Maunder, Roberto de la Maza, Billy Pat Richard Freitag, Karl Werner, W. Dan Sumlin, and
23. ZSCSP, Zoological Society, D-81377, Munich, GERMANY 93. Center for Conservation and Government CI -Japan, P.O. Box 1502 c/o McDonald’s providing information for use in this book, or for McKinney, Bonnie McKinney, Rodrigo Medellín, Dave Brzoska for the use of data pertaining to tiger
24. Comisión Nacional para el Conocimiento y Uso de la Biodiversidad, Av. Liga Periférico- Company (Japan), Ltd. Shinjuku i-Land Tower 39F 6-5-1 Nishi-Shinjuku, Tokyo 163-
Insurgentes Sur 4903, Parques del Pedregal, Tlalpan, 14010 México, D.F., MEXICO 1339, JAPAN other forms of assistance or support over the last Dolores Mestre de Robles Gil, James Miller, Scott beetles, and, finally, to Peter Raven and colleagues
25. Center for Biodiversity Conservation-Andes, CI-Colombia, Cra. 13 No. 71-41, Bogotá, 94. Harvard University, 22 Divinity Avenue, Cambridge, MA 01238, U.S.A.
COLOMBIA 95. Kadoorie Farm & Botanic Garden in Hong Kong, 44 Hampden Road, Kingston KT1 few years that have helped make it possible. In par- Miller, John Miskell, Dick Moe, Sanjay Molur, María (in particular Peter Stevens) at the Missouri Botani-
26. Missouri Botanical Garden, P.O. Box 299, St. Louis, MO 63166-0299, U.S.A. 3HG, UK
27. Center for Biodiversity Conservation-Andes, CI, Av. Coruña N29-44 y Noboa Caamano, 96. South China Biodiversity Team, Kadoorie Farm & Botanic Garden Corporation, Lam
ticular, we thank Rudi van Aarde, Tony Abbott, Antonieta Morales de Yarrington, Rob Morely, Eli- cal Garden for kindly lending us their botanical ex-
Quito, ECUADOR Kam Road, Tai Po, New Territories, HONG KONG Thomas Akre, Thomas Allnutt, Carlos Álvarez, zabeth Moreno, John Morrison, Federico Mozo, pertise and insights.
28. Terrestrial Ecology Research Unit Department of Botany, University of Port Elizabeth, 97. University of Idaho, Department of Fish and Wildlife, College of Natural Resources,
P.O. Box 1600, Port Elizabeth 6000, SOUTH AFRICA P.O. Box 441136, Moscow, Idaho 83844-1136, U.S.A. Katama Amberbirr, Laura Arriaga, John Ash, Fred Enrique Muñoz, Adolfo Navarro, John Neldner, Al- We are also grateful for the support received
29. PO Box 364, St Francis Bay, 6312, SOUTH AFRICA 98. Department of Animal and Plant Sciences, University of Sheffield, Western Bank,
30. 180 Gladstone Rd, Nth Mosgiel, Otago 9007, NEW ZEALAND Sheffield, S10 2TN, UK R. Barrie, Bruce Bartholomew, Paul Beckman, Afe- fonso Ocampo, Susana Ocegueda, Javier Ortega, from various staff members of Conservation Inter-
31. BirdLife International-Fiji, Suva, FIJI 99. Conservation International-Mesoamérica, Apdo. Postal 2365 - 2050 San Pedro de
32. Conservation International/SPREP, Vailima Headquarters, P.O. Box 240, Apia, WESTERN Montes de Oca, COSTA RICA work Bekelle, Hesiquio Benítez, Jocelyn Bentley, Doris Osuna, Sandra Osuna, Rosa Osuna de Llano, national during the course of this project, includ-
SAMOA 100. 3212 Redwood Drive Apts., CA 95003, U.S.A.
33. Laboratoire de Botanique, Centre IRD, P.B. A5 98848, Nouméa Cedex, NEW CALEDONIA 101. Museum of Vertebrate Zoology, Department of Integrative Biology, University of
Donald Broadley, David E. Brown, Scott Burnett, Guillermo Osuna Sáenz, Guillermo Osuna Villar, ing Gerry Allen, Jake Brunner, Chuck Burg, Janice
34. Taxonomie et Collections, Département de Systématique & Evolution, Case Postale 51, California, Berkeley, CA 94720, U.S.A. James Cambray, Chris Carpenter, Alan Channing, N. Parthasarathy, Raúl Pérez Madero, Brian Perry, Chanson, Shannon Charlton, Philip Chou, Jason
75231 Paris Cedex 05, FRANCE 102. Nicholas School of the Environment and Earth Sciences, Duke University, Durham,
35. Bureau WWF Nouvelle Calédonie, Parc Forestier Michel Corbasson, Rue du Mont Té- NC 27708, U.S.A. Lawrence Chau, Fernando Chiang, Terry Chiu, Otto Pfister, Baz van Pinto, J. Danuta Plisko, Derek Cole, Jessica Donovan, Sarah Frazee, Stuart Gottlieb,
Montravel, BP.692 98853 Nouméa Cédex, NEW CALEDONIA 103. California Natural Diversity Database, Wildlife and Habitat Data Analysis Branch,
36. Department of Biology, Villanova University, 800 Lancaster Avenue, Villanova, PA Department of Fish and Game, 1807-13th Street, Suite 202 Sacramento, CA 95814, U.S.A. Arnaud Collin, Neil Cox, João Crawford-Cabral, Pomeroy, John F. Pruski, Chris Puttock, Richard H. Nicole Guanzon, John Hanks, Bobby Jo Kelso, Jill
19085, U.S.A. 104. University Herbarium, 1001 Valley Life Sciences Building, #2465, University of
37. Leslie Hill Institute for Plant Conservation, University of Cape Town, Private Bag, California, Berkeley, CA 94720-2465, U.S.A.
Thomas B. Croat, Neil Crouch, Jennifer D’Amico, Ree, Berndt J. van Rensburg, Javier Robles Gil, Lucena, Laara Manler, Nina Marshall, Judy Mills,
Rondebosch, 7701 SOUTH AFRICA 105. Center for Biodiveristy Conservation - Andes, Perú- Program, Malecón de la Reserva Thomas F. Daniel, Richard Dean, Sebsebe Demissew, Patricia Rojo, Jerzy Rzedowski, Roberto Saldívar, R. John Musinsky, Ella Outlaw, Ana Rodrigues, James
38. 101 Jeffrey’s Road, Bryndwr, Christchurch, NEW ZEALAND 281, Lima 18, PERU
39. Invasive Species Specialist Group (IUCN), SGES/Tamaki Campus, University of 106. Center for Biodiveristy Conservation -Andes Venezuela Program, Av. San Juan Bosco, Pedro Díaz, Michael Donoghue, Robert L. Dressler, Uma Shaanker, Kuang-Chao Shao, Jevgeni Shergalin, Sanderson, Alexander Shenkin, Chris Stone, Carly
Auckland, Private Bag 92019, Auckland, NEW ZEALAND Edif. San Juan, Piso 8, Ofic. 8A, Altamira, Caracas, VENEZUELA
40. Herpetology, Island Restoration Science and Research Unit, Department of 107. Center for Biodiveristy Conservation- Andes, Bolivia Program, C. Pinilla, 291 San Bob Drewes, Amanda Driver, Sue Edwards, Gerald Lee Kwok Shing, Tatyana Shulkina, Claudio Sillero- Vynne, John Watkin, Michele Zador, and Sterling
Conservation, Private Bag 68-908, Newton, Auckland, NEW ZEALAND Jorge, La Paz, BOLIVIA
41. Museum of New Zealand Te Papa Tongarewa, P.O. Box 467, Wellington, NEW ZEALAND 108. World Wildlife Fund-Caucasus Programme Office, 11 M. Aleksidze St., 0193 Tbilisi, GEORGIA
Eilu, Ernesto Enkerlin, Fan Enyuan, José Manuel Zubiri, Laurence E. Skogg, Hannelie Snyman, Hugh Zumbrunn. In addition, we would also like to pay
42. Department of Conservation, P.O. Box 112, Hamilton, NEW ZEALAND 109. Institute of Zoology of the Georgian Academy of Sciences, 31 Chavchavadze St., 0179 Espinosa, John Fa, Neil Fairall, María Hilda Flores- Spencer (Australian Tropical Research Foundation), tribute to the staff of the GIS department, partic-
43. 6 Fraser Rd, Applecross, Western Australia, AUSTRALIA Tbilisi, GEORGIA
44. Conservation International-Indonesia Jl. Pejaten Barat 16 A, Kemang, Jakarta - 12550, 110. Department of Ecology, Tbilisi State University, University str. 2, Tbilisi, GEORGIA Olvera, Óscar Flores-Villela, Steve Freligh, David Warren Douglas Stevens, Hester Steyn, Daryn ularly Robert Waller, Sonya Krogh, and Joanna
INDONESIA 111. University of Queensland, AUSTRALIA
Frodin, Carlos Galindo, Martín Jon García-Urtiaga, Storch, Holly Strand, Wu Sugong, Charlotte M. Taylor, Seeber, for their help with analyses and in prepar-
45. Tropical Resources Institute, Yale School of Forestry and Environmental Studies, 370 112. Australian Wildlife Conservancy, P.O. Box 154, Redlynch QLD 4870, AUSTRALIA
Prospect St., New Haven, CT 06511, U.S.A. 113. Environmental Protection Agency (Queensland Herbarium), Brisbane Botanic
46. 2 Andrew Street, Stratford QLD 4870, AUSTRALIA Gardens Mt Coot-tha Toowong QLD, 4066, AUSTRALIA Rafael García Zuazua, David Garza Lagüera, Michael Julie L. Tessler, The Society of Wilderness (Taiwan), ing the maps appearing in this book. Special
47. Nature Protection Trust of Seychelles, SEYCHELLES 114. Queensland Parks and Wildlife Service, P.O. Box 834, Atherton QLD 4883, AUSTRALIA Gilbert, Francisco González-Medrano, Janice Golding, Michele Thieme, Lana Troy, Ming-Chang Tu, Jonás thanks also to Shawn Concannon for his long-term
48. University Museum of Zoology, Cambridge, 133 Cherry Hinton Road, Cambridge CB1 115. Australian Centre for Tropical Freshwater Research, James Cook University, P.O. Box
7BX, UK 6811 Cairns QLD 4870, AUSTRALIA Katharine Gotto, Werner Greuter, Robert Gulden- Villalobos, James Vonesh, Wes Wettengel, Daniel commitment and his support for some of the re-
49. WWF Madagascar and Field Museum of Natural History, BP 738, Antananarivo, 116. Australian Wildlife Conservancy, P.O. Box 1897, West Perth WA 6872, AUSTRALIA
MADAGASCAR 117. Department of Biology, Pennsylvania State University, 208 Mueller Lab, University mond, Ron Gutberlet, Sun Hang, James Harrison, Winkler, Derek Yalden, Xie Yan, Jen-Te Yen, Jia Yu, search that went into the writing of this book. Last,
50. Field Museum of Natural History, Lakeshore Drive at Roosevelt Road, Chicago, IL Park, PA 16802-5301, U.S.A.
60605, U.S.A. 118. Institute of Systematic Botany, New York Botanical Garden, Bronx, NY 10458-5126, U.S.A. Billy Hau, Kris Helgen, Rod Henderson, Vernon James L. Zarucchi, Wang Zheng, Yang Zhuliang, but not least, we would like to acknowledge the
51. University of Antananarivo, Department of Biology, University of Antananarivo, 119. Grupo Jaragua El Vergel 33, El Vergel. Santo Domingo, DOMINICAN REPUBLIC
Heywood, David Hibbett, Craig Hilton-Taylor, Carsten the Avian Demography Unit at the University of support of the Critical Ecosystem Partnership Fund
Antananarivo 101, MADAGASCAR 120. Department of Entomology, Montana State University, Bozeman, MT 59717, U.S.A.
52. Conservation International-Madagascar BP 5178, Antananarivo 101, MADAGASCAR 121. Center for Applied Biodiversity Science Fellow, Conservation International, Hobohm, Wayne Hsu, Tseng-Chieng Huang, Luke Cape Town, the KwaZulu-Natal Conservation Ser- and the Global Conservation Fund at Conservation
53. Institute of Biology,University of the Philippines, Diliman, Quezon City 1101, PHILIPPINES Willemstad, CURAÇ̧AO
54. Conservation International-Philippines 5 South Lawin, Philam Homes, Quezon City, 122. Fairchild Tropical Botanic Garden, 10901 Old Cutler Road, Coral Gables, Miami, FL Hunter, Lucas Human, Brian Huntley, Clyde Imada, vice, and the National Science Foundation. Also, of International.
1104, PHILIPPINES 33156-4233, U.S.A.
55. ICRAF, United Nations Avenue, Gigiri, P.O. Box 30677-00100, Nairobi, KENYA 123. Florida International University, U.S.A. Dean Impson, Stoffel de Jager, Raúl Jiménez, Yin the Parque Nacional Marítimo Terrestre del Archi-
56. Department of Anthropology, Hunter College of CUNY, 695 Park Avenue, New York, 124. 3 Acre Street, Glossop, Derbyshire SK13 8JS, UK
NY 10021, U.S.A. 125. Institute of Zoology, Uzbek Academy of Sciences, A. Niyazov str., 1 Tashkent 700095,
57. Department of Experimental Ecology (Bio III), University of Ulm, Albert-Einstein UZBEKISTAN
Allee 11, D – 89069 Ulm, GERMANY 126. CMEB, Departamento de Ciencias Ecológicas, Facultad de Ciencias, Universidad de
58. Department of Nature and Forest Management, University of Professional Education Chile, Casilla 653, Santiago, CHILE
Larenstein P.O.Box 9001, 6880 GB Velp, THE NETHERLANDS 127. CASEB, Facultad de Ciencias Biológicas, Pontificia Universidad Católica, Casilla 114-D.,
59. Department of Animal Ecology and Tropical Biology Biocenter of the University, Am Santiago, CHILE First English edition, 2004
Hubland, D-97074 Würzburg, GERMANY 128. Departamento de Botánica, Facultad de Ciencias Naturales y Oceanográficas,
60. Van der Heimstraat 52 2582 SB Den Haag, THE NETHERLANDS Universidad de Concepción, Casilla 160-C, Concepción, CHILE Copyright of this edition is the property of CEMEX, S.A. de C.V.
61. Institut Méditerranéen d’Écologie et de Paléoécologie (IMEP, CNRS UMR 6116), 129. CMEB and Departamento de Botánica, Facultad de Ciencias Naturales y
Université d’Aix-Marseille III, Europole méditerranéen de l’Arbois, Bâtiment Villemin Oceanográficas, Universidad de Concepción, Casilla, 160-C, Concepción, CHILE
© 2004, Agrupación Sierra Madre, S.C. Primero de Mayo 249, 03800 Mexico City, Mexico
BP 80. F - 13545 Aix-en-Provence Cedex 04, FRANCE 130. CEAZA, Departamento de Biología, Facultad de Ciencias, Universidad de La Serena, asmupc@infosel.net.mx
62. Institute of Zoology Academia Sinica, Taipei, TAIWAN 115 Casilla 599, La Serena, CHILE
63. Institute of Ecology and Evolutionary Biology, National Taiwan University, Taipei, TAIWAN 106 131. CMEB y Facultad de Ciencias. Universidad de Magallanes, Casilla 113-D, Punta All rights reserved
64. WWF-US Conservation Science Programme, 1250 24th Street, NW, Washington DC, U.S.A. Arenas, CHILE
65. UNDP-GEF Eastern Arc Strategy Project, P.O. Box 298, Morogoro, TANZANIA 132. Selmar Schönland Herbarium, Botany Department, Rhodes University, P.O. Box 101,
66. Wildlife Conservation Society, P.O. Box 7487, Kampala, UGANDA Grahamstown, 6140, SOUTH AFRICA ISBN 968-6397-77-9
67. Ethiopian Wolf Conservation Programme, P.O. Box 23400 Addis Ababa, ETHIOPIA 133. Terrestrial Ecology Research Unit, Department of Botany University of Port Elizabeth,
68. Wildlife Conservation Research Unit, University of Oxford, South Parks Road, Oxford P.O. Box 1600, Port Elizabeth 6000, SOUTH AFRICA
OX1 3PS, UK 134. Department of Botany, University of Pretoria, Pretoria 0002, SOUTH AFRICA
Printed in Japan by Toppan Printing Co., on acid-free paper