Ecology and Biodiversity

ECOLOGY AND BIODIVERSITY
"This page is Intentionally Left Blank"

ECOLOGY AND
BIODIVERSITY
By
Rahul Srivastava
PRAGUN PUBLICATIONS
(A O.K. PUBLISHERS DISTRIBUTORS ENTERPRISE)
4224/1, Ansari Road, Darya Ganj, New Delhi-110002
Ph. : 011-41562573-77 Fax: 011-41562578
E-mail .. pragun@dkpd.com urI,' www.dkpd.com
Published by:
Pragun Publications
(A D.K. Publishers Distributors Enterprise)
4224/1, Ansari Road, Darya Ganj, New Delhi-11 0002
Ph . : 011-41562573-77 Fax: 011-41562578
E-mail : pragun@dkpd.com uri: www.dkpd.com
Distributed by:
O.K. Publishers Distributors Pvt. Ltd.
4834/24, Ansari Road, Darya Ganj, New Delhi - 110002
Ph . : 011-41562573 (5-lines) Fax: 011-41562578
E-mail .' info@dkpd.com urI: www.dkpd.com
© Pragun Publications, 2006
All rights reserved . No part of this publication may be reproduced , stored in a retrieval system,
or transmitted, in any form or by any means, without the prior permission in writing of
Pragun Publications, or as expressly permitted by law, or under terms agreed
with the appropriate reprographics rights organization . Enquiries concerning
reproduction outside the scope of the above should be sent to the
Publishers , at the address above.
You must not circulate this book in any other binding or cover and you must impose
this same condition on any acquiror
ISBN:978-81-89645-06-9
ISBN: 81-89645-06-4
Published in India and Printed at :
liP Printers, New Delhi -110015 Mobile: 9810271526

Preface
The term "ecology" was introduced by Haeckel in 1869.

His purpose was to focus attention on relationships,
especially relationships with the environment, rather than
on organisms and species. The coinage was taken from
the Greek for household (oikos) and suggested a broader
interdisciplinary perspective on phenomena in context. In
practice, it has proved very difficult to cover the structure
of the "house," as well as the relationships of all the
occupants with it and with each other, in one analysis.
Ecology has, by and large, been natural ecology at its
broadest. Where human activities have been included in
the subject matter of ecological studies (for the most part
a recent development), they have been studied
naturalistically, or as though they were a function of
natural processes, rather than an integral part of a larger
universe.The variation of the living nature on the planet
earth-the biodiversity or diversity of life-is still
overwhelming.
Our planet supports between 3 and 30 million species
of plants, animals, fungi, single-celled prokaryotes such
as bacteria, and single-celled eukaryotes such as
protozoans. Of this total, only about 1.4 million species
have been named so far, and fewer than 1 percent have
been studied for their ecological relationships and their
role in ecosystems. A little more than half the named
vi Ecology and Biodiversity
species are insects, which dominate terrestrial and

freshwater communities worldwide; the laboratories of
systematists are filled with insect species yet to be named
and described. Hence, the relationships of organisms to
their environments and the roles that species play in the
biosphere are only beginning to be understood. This book
gives some answers to these questions.
Editor
Contents
Preface v
1. Function of Biodiversity in the Ecology 1
2. Natural Ecosystems 22
3. Impacts of Climate Change on Biodiversity 45
4. Biodiversity and Ecosystem Functioning 52
5. Ecosystem Functioning at
Local and Regional Scales 61
6. Freshwater Ecosystem 81
7. Aquatic Systems 101
8. Impact of Dams on Biodiversity 107
9. Agricultural Ecosystem 151
10. Bacterial Biodiversity 174
11. Ecological Forecasting and the
Urbanisation of Stream Ecosystems 189
12. Land Use Appraisal 212
13. Population Diversity 228
14. Model Systems for Ecology 240
15. Biodiversity, Ecosystem Services
and the UN Millennium Declaration 250
Bibliography 262
Index 264
"This page is Intentionally Left Blank"
1
Function of Biodiversity in the
Ecology
All species provide some kind of function to an ecosystem.

They can capture and store energy, produce organic
material, decompose organic material, help to cycle water
and nutrients throughout the ecosystem, control erosion
or pests, fix atmospheric gases, or help regulate climate.
Ecosystems also provide various supports of production
(soil fertility, pollinators of plants, predators,
decomposition of wastes ... ) and services such as
purification of the air and water, stabilisation and
moderation of the climate, decrease of flooding, drought,
and other environmental disasters.
These functions are important for ecosystem function
and human survival. Research suggests that a more diverse
ecosystem is better able to withstand environmental stress
and consequently is more productive. The loss of a species
is thus likely to decrease the ability of the system to
maintain itself or to recover from damage or disturbance.
Just like a species with high genetic diversity, an ecosystem
with high biodiversity may have a greater chance of
adapting to environmental change.
In other words, the more species comprising an
ecosystem, the more stable the ecosystem is likely to be.
2 Ecology and Biodiversity
The mechanisms underlying these effects are complex and

hotly contested. In recent years, however, it has become
clear that there are real ecological effects of biodiversity.
The importance of species diversity in the
performance of ecosystem functions, such as primary
production and resource extraction, is widely debated. The
primary experimental method used for addressing this
issue is to assemble communities of, for instance,
herbaceous plants or protists that differ in the number of
species which are drawn from the pool of possible species.
A key issue in this controversy is whether any measured
variation in ecosystem processes among the assembled
communities is due to differences in species richness per
se, or to differences in the specific identity or functional
roles of the species added.
Resolving this issue is important for policy and
management actions that would impact biodiversity. If
species diversity per se consistently has an impact on
ecosystem processes, then the loss of species, irrespective
of their particular functional roles, would be expected to
affect ecosystem services such as efficient capture of energy
from sunlight or filtering of pollutants. On the other hand,
if particular functional groups or individual species have
a disproportionately large effect on ecosystem functions,·
the loss of other components of the ecological community
would likely have little impact on ecosystem services.
Consensus seems to be emerging that greater diversity
in the functional roles of species in experimental
communities leads to greater efficiency in the use of scarce
resources and increases the resilience of the community
in the face of environmental variability. Similarly, high
diversity within an ecological community may interfere
with the ability of competitively superior species to become
numerically dominant. Under such conditions, diversity
begets diversity.
Function of Biodiversity in the Ecology 3
ECOLOGY OF INFEcrIOUS DISEASES
Vector-borne Zoonotic Diseases

Species diversity may also be important in the ecology of
infectious diseases, particularly vector-borne zoonoses,
diseases of humans in which the disease agent resides
predominantly in a non-human animal host and is
transmitted among hosts (including humans) via the bite
of a vector, typically an arthropod. On one hand, high
species diversity in vertebrate hosts of vectors may playa
beneficial role by impeding dominance by particular
species that act as key reservoirs of the pathogen. On the
other hand, high diversity of vertebrate hosts may provide
generalist vectors or pathogens with a hedge against local
extinction that would accompany the extirpation of a
primary host, and therefore may play a role in increasing
the disease risk to humans.
Lyme disease and the dilution effect

Lyme disease was first described in the 1970s, when a
cluster of cases of childhood arthritis in Lyme, Connecticut,
U.S.A., was linked to tick bites and a spirochetal pathogen
subsequently named Borrelia burgdorferi. The discovery
of this spirochetal pathogen and the link to disease
symptoms associated with it led to a flurry of research
designed to uncover the enzootic cycle both in North
America and in Europe, where similar symptoms had been
linked to tick bites. Within 10 years of the initial
description, a basic understanding of the natural history
of Lyme disease had developed for foci in both eastern
North America and northern Europe.
Lyme disease is transmitted via the bite of a member
of the Ixodes ricinus complex (Acari: Ixodidae), which
includes Ixodes scapularis in eastern and central North
America, Ixodes pacificus in western North America,

Ixodes ricinus in Europe, and Ixodes persulcatus in Asia.
Each of these tick species has a life cycle that includes
three active stages, larva, nymph, and adult, each of which
takes a single blood meal from a single individual before
dropping off the host and either molting to the next stage
(in the case of larvae and nymphs) or reproducing and
dying (in the case of adults). Because transovarial
transmission of B.burgdorferi is highly inefficient, the vast
majority of larval ticks hatch uninfected with the
spirochete and therefore unable to infect their host during
feeding.
The larval meal represents an opportunity for the tick
to acquire an infection, which is maintained through all
subsequent molts. Larval ticks in the 1. ricinus complex
tend to be highly generalised in their host selection,
feeding from a wide variety of mammalian, avian, and
reptilian hosts. The specific identity of the host of larval
ticks is important in the enzootiology of Lyme disease,
because host species vary dramatically in their probability
of infecting a feeding larval tick. In eastern North America,
the white-footed mouse (Per~myscus leucopus) is highly
efficient at infecting feeding ticks, and is considered the
principal natural reservoir of Lyme disease. Eastern
chipmunks (Tamias striatus) also appear to be competent
reservoirs in North America.
In Europe and Asia, voles (Clethrionomys glareolus),
mice (e.g., Apodemus spp.), introduced gray squirrels
(Sciurus carolinensis), red squirrels (Sciurus vulgaris), and
blackbirds (Turdus merula) are competent reservoirs. Other
hosts, such as deer (Odocoileus virginianus and Capreolus
capreolus), lizards (Sceloporus occidentalis and Lacerta
vivipara), and ovenbirds (Seiurus aurocapillus) do not
transmit B. burgdorferi to feeding ticks, and are considered
incompetent reservoirs. Ticks infected during their larval
Function of Biodiversity in the f,cology 5
meal become active about 1 year later, after they molt into
infected nymphs capable of transmitting the pathogen to
their hosts. Those not infected during their larval meal
have a second opportunity to acquire the Lyme disease
pathogen during their nymphal meal.
Ticks that become infected during either their larval
or their nymphal meal will molt into an infected adult,
which becomes active between several months and >1 year
later. Thus, both nymphs and adults are capable of
transmitting Lyme disease, as well as perpetuating the
enzootic cycle, when they bite a reservoir host. The
synchrony between annual peaks in activity of nymphs
and human cases of Lyme disease suggests that most cases
of Lyme disease result from transmission of the pathogen
by nymphs rather than by adults. Given the small size of
nymphs of Ixodes spp. and their tendency to reach a
seasonal activity peak in summer, when humans are most
likely to enter tick habitat, it is not surprising that the
nymphal stage is most dangerous to people.
Two parameters describing the tick population are
crucial in determining the probability of human exposure
to Lyme disease within specific localities that people use
domestically or recreationally. The first is nymphal
infection prevalence (NIP) within the local population,
defined as the proportion of nymphal ticks that are
infected with B. burgdorferi. NIP will determine the
probability that a given bite from a nymphal tick will
transmit Lyme disease to the human host. NIP will be a
function of the distribution of larval ticks among vertebrate
hosts. The larger the proportion of larvae that feed from
highly competent reservoirs, the higher will be the
infection prevalence in the nymphal generation. The
second parameter is the local density of infected nymphs
(DIN), which will strongly influ~nce the probability that
a person will encounter a tick capable of transmitting
Lyme disease.
DIN will be a function of both NIP and the local

population density of ticks. The density of nymphs, in
turn, will be influenced by both the distribution of ticks
among vertebrate hosts, which vary in their ability to
support successful feeding by ticks, and by biotic and
abiotic conditions affecting tick survival and reproduction.
Recent research has suggested that variation in the
diversity of vertebrate hosts of ticks might influence the
risk of human exposure to Lyme disease as measured by
either NIP or DIN. This assertion is based on the following
observations about the ecology of Lyme disease in eastern
and central North America. First, the main vector, I.
scapularis, shows little or no transovarial transmission of
B. burgdorferi, therefore larval ticks typically hatch free
of Lyme disease bacteria. Second, I. scapularis is a
generalist vector: larvae and nymphs feed from dozens of
species of vertebrate hosts, including mammals, birds, and
reptiles. Third, I. scapularis larvae acquire Lyme disease
bacteria more efficiently from white-footed mice than from
other hosts.
A few other hosts, such as chipmunks and American
robins (Turdus migratorius), are competent reservoirs, but
most hosts show a low probability of infecting feeding
ticks. Fourth, the white-footed mouse is one of the most
abundant and widespread of all possible hosts for ticks,
being present in both species-rich and species-poor
vertebrate communities. From these observations, experts
argued that species-poor vertebrate communities should
be characterised by a high relative abundance of white-
footed mice, and that speciesrich communities should
include a higher relative abundance of non-mouse species
that are poorer disease reservoirs. As a result, high
vertebrate diversity should dilute the impact of white-
footed mice on tick infection prevalence and consequently
reduce the risk of human exposure to Lyme disease.
Some of the experts further analyzed the dilution

effect using both empirical and modeling approaches. First,
they found that at sites in southeastern New York State,
eastern chipmunks are highly competent reservoirs of
Lyme disease, infecting almost 70% of larval ticks that fed
to repletion from them. Chipmunks were only moderately
less competent than white-footed mice, which infected
>90% of larval ticks that fed to repletion. Second, they
demonstrated that the infection prevalence of questing
nymphal ticks was dramatically lower than would be
expected if all larval meals were taken from either mice
or chipmunks. Third, using a simple mathematical model,
they postulated that at their study sites, about 60% of
larval meals must be taken from non-mouse, non-
chipmunk hosts, in order to account for the observed
infection prevalence of questing nymphs (38%).
They interpreted this result as demonstrating the
dilution effect at the scale of local vertebrate assemblages.
The presence of a diverse assemblage of relatively
inefficient reservoir hosts reduces NIP. However, experts
noted that species diversity in a community can be
measured as the number of species (species richness), the
relative abundance of the various species (species
evenness), or a combination of richness and evenness, as
represented by diversity indices such as the Shannon
Index. These metrics of diversity have different
implications for disease risk.
Adding species to a species-poor community without
changing the absolute abundance of the dominant species
in the community may provide the tick population with
more feeding opportunities than it would have in the
absence of the added hosts. If the added species are
incompetent reservoirs, the additional diversity would
decrease NIP, but the added feeding opportunities might
simultaneously increase the total density of nymphs. To
investigate the net effects of variation in species richness
and evenness on disease risk, Schmidt and Ostfeld used

compu ter simulation modeling to assess the effects of both
host ~pecies richness and evenness on the density of
infected nymphs.
Schmidt and Ostfeld assembled simulated vertebrate
communities by adding between 6 and 12 species drawn
randomly from a pool of potential community members
to an initial community consisting of white-footed mice
and eastern chipmunks. These species were assumed to
have several attributes.
First, each species was assumed to have a low to
moderate fixed reservoir competence, which is the ability
of a particular host species to infect a vector; this was
assigned randomly from a uniform distribution between
o and 0.20. This assumption was based on characteristics
of host communities in eastern North America in which
most hosts are weakly capable of infecting feeding ticks.
Second, each species was assumed to have relative
dominance in the community, which is the product of its
characteristic population size and average tick burden. This
assumption allowed us to represent species that exhibit
covariation in population density and average tick burden.
For example, for some larger-bodied species (e.g., raccoons,
Procyon lotor), population densities are likely to be
relatively low but the tick burden per individual is likely
to be high.
Third, each species was assigned an interaction
coefficient with white-footed mice (mean = 0.90, SD = 0.15),
so each added species modified the relative dominance of
mice. Although any given species could have either a net
antagonistic or net mutualistic effect on mice, we assumed
that ,m ,\Iltdgnni~hc efft.'ct is more likely, therefore the
mean interaction coefficient was <1.0.
The effect of these simulated species can be either a
direct reduction in mouse abundance (e.g., predation,
competition) or a siphoning of tick meals away from mice.

Finally, to reflect the observation that numerically
dominant members of host communities are often, but
perhaps not always, the most competent reservoirs, we
allowed the correlation between host dominance and
reservoir competence to vary. Running the simulations
with both high and low correlations between host
dominance and reservoir competence allowed us to assess
outcomes when dominance and competence are not linked.
The model showed that the effect of species richness on
disease risk is contingent on the community composition
of hosts.
When neither interaction coefficients among hosts nor
correlations between community dominance and reservoir
competence were incorporated, increasing species richness
had a positive effect on DIN simply by providing more
feeding opportunities for ticks. However, when we
allowed additional hosts to have a net negative impact on
the abundance of, or tick burdens on, highly competent
reservoirs, increasing species richness dramatically
decreased DIN and therefore disease risk.
Unlocking reservoir competence

Often, reservoir competence is determined using
xenodiagnosis, a process in which uninfected vectors are
fed on hosts that have been either artificially (via injection)
or naturally (via vector) infected with the pathogen. The
reservoir competence of field-caught hosts can be
measured by determining the infection prevalence of
previously uninfected vectors that are known to have
parasitised the host. When the infection status of the host
is known, field determinations of reservoir competence can
provide an ecologically realistic estimation of the
probability that a particular host species will infect a vector
in nature.
Laboratory measurements of reservoir competence, on

the other hand, typically assess the maximum value for
hosts that have recently been infected. Because the actual
reservoir competence of an individual host tends to decline
with time following inoculation with the pathogen, these
maximum values may not accurately represent reservoir
competence under natural conditions. The reservoir
competence of particular species can be characterised by
three different parameters:
a maximum value achieved shortly after inoculation;
a characteristic rate of decay with time since
inoculation; and
a characteristic probability that repeated inoculations
will return competence values to the maximum~
The wide variation in estimates of the reservoir
competence of particular host species may be due to failure
to incorporate these three parameters when determining
competence. By specifying values for these parameters, the
reservoir competence of a host over the course of a season
of varying vector activity, or its effective reservoir
competence, can be estimated. Effective reservoir
competence is useful in assessing the net effects of
particular species on the infection prevalence of vectors.
Models of E.M. Schauber and R.S. Ostfeld suggest
that species diversity in host communities can influence
disease risk via an additional pathway: reducing the
effective reservoir competence of each host. Imagine a host
community that is dominated by a species with a high
maximum reservoir competence, such as one with many
whitefc>oted mice and few alternative hosts. Because the
population of larval ticks will feed predominantly on this
competent reservoir, many will become infected and molt
the next year into infected nymphs. As a result of high
NIP, inoculation rates of both the highly competent and
the less competent reservoirs will tend to be rapid, pushing

the effective reservoir competence of all hosts toward their
maxima.
High effective reservoir competence will reinforce
high NIP, resulting in a positive feedback loop. Now,
imagine another host-species-rich community with low
maximum reservoir competence. The same initial
population of larval ticks will have a lower average
probability of feeding on a competent reservoir, and fewer
will molt the next year into infected nymphs. As a result
of the lower NIP, inoculation rates of all hosts will be low,
and larvae will be more likely to feed after reservoir
competence has decayed, resulting in low effective
reservoir competence and a positive feedback loop toward
low NIP. The model suggests that high diversity in the
vertebrate community may reduce disease risk not only
by diluting the effects of the most competent reservoirs,
but also by reducing the reservoir competence of hosts.
Evaluating the veracity of this model requires further
research into the various parameters that influence
effective reservoir competence.
Dilution Effect in Vectorborne Zoonoses

Although Lyme disease is the most common vector-borne
disease in parts of North America and Europe, many other
vector-borne zoonoses plague humans through-out the
world. Often these diseases are debilitating and even
lethal. It is therefore useful to assess the degree to which
the dilution effect may apply to diseases other than Lyme
disease. Four attributes are necessary for the dilution effect
to operate in a disease system: a generalist vector, a
significant role for oral acquisition of the pathogen by
vectors, variation in reservoir competence among hosts,
and a positive correlation between reservoir competence
and numerical dominance of hosts in the community.
Arthropod parasites vary widely in their degree of

specialisation on hosts. For instance, for a vector-borne
zoonosis to exist, the vector must at least parasitise both
a non-human animal and a human. For zoonoses in which
vectors are extreme specialists, i.e., parasitise one or a few
host species, diversity in the host community will be
relatively unimportant in determining vector abundance
and infection prevalence because most hosts will be
irrelevant to the enzootic cycle. Only in those cases in
which variation in host diversity represents a change in
feeding opportunities for the vector will the dilution effect
apply.
Oral Acquisition of the Pathogen: Some pathogens are
transmitted across vector generations via transovarial
passage from mother to offspring. For these zoonoses,
vectors are often infected when they hatch, which liberates
them from immediate reliance on a blood meal from a host
for pathogen acquisition. Only in those cases in which a
significant proportion of pathogen acquisition by vectors
is from host meals will the dilution effect apply.
Variation in reservoir competence among hosts: If the
probability of transmitting the disease agent to a feeding
vector is similar for all hosts, then variation in host
diversity is unlikely to influence the infection prevalence
of vectors because a meal from each host species will have
a similar probability of transmitting the pathogen. Only
when different host species vary in their reservoir
competence will changes in the composition of the host
community influence the disease risk.
A positive correlation between reservoir competence and
numerical dominance in the community: If the most competent
reservoirs of a pathogen tend to be rare members of their
communities, they will probably be absent from species-
poor communities and present only in species-rich
communities. In such cases, the disease risk will be higher
in more diverse communities, which is the opposite of the

dilution effect. Only in situations in which high species
richness is accompanied by the inclusion of species with
low average reservoir competence will the dilution effect
apply.
Although comprehensive assessments of the degree
to which zoonotic vector species are selective of hosts are
rare, it appears that extreme specialists are few. For
instance, some species of mosquitoes feed largely on birds
and others largely on mammals, but most or all of these
mosquitoes feed on at least several species within each
vertebrate class. The widespread use of domestic chickens
and livestock as sentinels to detect mosquito-borne
diseases such as malaria, dengue fever, and various
mosquito-borne encephalitis viruses capitalises on the
tendency for vector mosquitoes to feed on both human
and non-human hosts.
Indeed, zooprophylaxis, which is related to the
dilution effect, consists of using domestic animals as
alternative hosts that may deflect host-seeking mosquitoes
from human hosts. However, whereas the dilution effect
is concerned with the impact of diversity of native host
communities on both abundance and infection prevalence
of vectors, zooprophylaxis is concerned with the role of
domestic animals in vector use of human hosts.
Ixodid ticks appear to vary greatly in their degree of
host generalisation, from species such as Ixodes neotomae,
which attacks relatively few mammalian hosts, and not
humans, to I.scapularis and I. ricinus, which parasitise at
least several dozen host species in three vertebrate classes.
Tick vectors of the Crimean-Congo hemorrhagic fever
(CCHF) virus (genera Hyalomma, Rhipicephalus, and
Amblyomma) infest numerous species of wild and
domestic ungulates, rodents, and ground-dwelling birds.
Rhipicephalus sanguineus, a tick vector of tick typhus,
Rocky Mountain spotted fever, Qfever, Lyme disease,
tularemia, and CCHF is known to parasitise domestic

dogs, livestock, wild carnivores, ungulates, and other
mammals.
Similarly, in the Neotropics, several syntopic species
of sand fly (genus Lutzonyia) vectors of zoonotic
cutaneous leishmaniasis (ZCL) vary dramatically in their
host specificity, but few take more than 50% of their blood
meals from any single host species. Old World sand fly
(genus Phlebotomus) vectors of ZCL are known to be
attracted to and feed on several species of native rodents
and rabbits, as well as domestic stock and pets. Triatomine
bug vectors of Chaga's disease feed on a wide variety of
mammalian and avian hosts.
Zeled6n et al. reported a single adult Triatoma
dimidiata as having fed on six species of hosts belonging
to five mammalian orders. A popUlation of mosquitoes
(Aedes albopictus) from Missouri, U.S.A., fed from at least
nine species of mammals and members of four orders of
birds. Tsetse fly (Glossina spp.) vectors of African
trypanosomiasis are known to feed on dozens of species
of both wild and domestic mammals.
Host generalisation by disease vectors appears to be
widespread. However, with few exceptions, the absolute
degree of host generalisation by arthropod vectors, as
represented by the distribution of blood meals among
hosts, is poorly understood.
Oral versus transovarial transmission of pathogens: Many
pathogens have multiple routes of transmission to vectors,
including transovarial, venereal, and oral. For some vector-
borne zoonoses, especially mosquito-borne viral diseases,
transovarial transmission is relatively effi-cient, leading to
moderate or high infection prevalences in newly hatched
larvae produced by infected mothers. For other vector-
borne zoonoses, including many viral, bacterial, and
protozoal diseases, transovarial transmission is highly
inefficient or nonexistent, leading to low to zero infection

prevalences in F} generations produced by infected
mothers. When transovarial or venereal transmission is the
main route of pathogen transmission, infection prevalence
of vectors may be largely independent of the source of
blood meals, therefore the dilution effect is less likely to
be a strong determinant of disease risk.
Nevertheless, even for zoonoses with efficient
transovarial transmission, efficiency rarely exceeds 80%,
and is often <50%. Infection prevalence in the F2 generation
often declines dramatically from that in the F} generation.
If transovarial transmission were the only means of
pathogen acquisition, infection prevalence of vectors would
decline exponentially with succeeding generations, leading
to the virtual loss of the pathogen in relatively few
generations. Therefore, even for zoonoses with significant
vertical transmission, the distribution of blood meals
among hosts is likely to influence the infection prevalence
of vectors, and therefore disease risk.
Further studies are warranted to compare the
infection prevalence expected if transmission is exclusively
transovarial with the actual infection prevalence of vectors.
This difference, the effective degree of oral transmission,
should be correlated with the potential for the dilution
effect to influence disease risk.
Variation in reservoir competence among hosts: The
reservoir competence (or infectivity) of hosts is typically
assessed by means of one of three methods. First, the
proportion of vectors acquiring an infection from a blood
meal is determined using xenodiagnosis, as described
above. Second, hosts are sampled to determine whether
they possess circulating antibodies to specific pathogens,
under the assumption that demonstrated recent or current
exposure to a pathogen reflects a host's state of infectivity
to vectors. Third, hosts are sampled for the presence of
the pathogen in the bloodstream or other tissues, under

the same assumption as for antibody tests.
The latter two methods by themselves are often
unreliable because the presence of either particular
antibodies or pathogens may not accurately reflect the
probability of transmission from host to vector. For
example, circulating levels of B. burgdorferi-specific
immunoglobulins in three species of wild rodents were
found to be negatively correlated with spirochete
infectivity to ticks. Similarly, in the case of visceral
lei&hmaniasis, both culture methods and polymerase chain
reaction techniques demonstrated the presence of
Leishmania chagasi in individual opossums that were not
infective, as determined by xenodiagnosis. Whether the
presence of pathogens or antibodies specific to those
pathogens in hosts is correlated with infectivity to vectors
is rarely examined for vector-borne zoonoses. For nearly
all vector-borne zoonoses of which we are aware, reservoir
competence appears to vary substantially among hosts.
However, accurate determination of reservoir
competence in specific disease foci is problematic for
several reasons. First, intraspecific variation in reservoir
competence appears to be commonplace but the causes of
this variation are typically unknown. Second, the burden
of vectors on hosts can influence reservoir competence,
but the presence of heavy vector burdens in some cases
increases, but in other cases decreases, reservoir
competence. Third, reservoir competence is likely to be
determined by a combination of characteristics that are
intrinsic to the host and extrinsic to the host. Fourth, some
hosts that are incompetent reservoirs may nevertheless
infect vectors when they are simultaneously parasitised by
infected and uninfected vectors.
In such cases, pathogens are transmitted from vector
to vector without being maintained or amplified in the
host. Further studies to assess the mechanisms behind both

inter- and intra-specific variation in infectivity are
warranted. Such studies would facilitate accurate
determination of whether host species have modal values
of reservoir competence, and if so, whether modal values
vary substantially among species.
A positive correlation between reservoir competence and
numerical dominance in the community: Few data exist to
allow this condition to be explored. Although many
studies of vector-borne zoonoses involve assessment of the
reservoir competence of various hosts, few simultaneously
assess the relative abundance of each host species, or
average parasite burdens on those hosts. Searches for
reservoir hosts of zoonotic pathogens typically focus
exclusively on the most abundant host species. In fact, if
a host species is not both abundant and highly infective
to vectors, it is usually ignored, under the assumption that
only species having both characteristics can be important
epidemiologically.
In marked contrast, the dilution effect argues that,
because of their potential to divert meals away from highly
infective hosts, both less common species and those with
low infectivity are likely to be more important than was
previously recognised. In many endemic foci of Lyme
disease in North America, the white-footed mouse is both
the most competent reservoir and the numerically
dominant member of the community of tick hosts. In Lyme
disease foci in Europe and Asia, the bank vole
(Clethrionomys glareolus) tends to be the numerically
dominant tick host; in some studies it has been found to
be also the most competent reservoir, whereas others have
found higher reservoir competence in syntopic mice (genus
Apodemus). It is noteworthy that some European studies
of Lyme borreliosis have shown positive correlations
between average tick burden on hosts and reservoir
competence of those hosts. This finding is consistent with
the model of Schmidt and Ostfeld, which defines host

community dominance in terms of the number of tick
meals supplied by a given host species, rather than host
density per se.
For ZCL in Jordan, Saliba et al. found that the fat
sand rat or jird, Psammomys obesus, was both the most
abundant rodent and the most competent reservoir for
Leishmania spp. Analyzing ZCL in Iran, Yaghoobi-Ershadi
et al. found that the numerically dominant species and
most competent reservoir at one site was the Libyan jird
(Meriones libycus), but at another site was the great gerbil.
At a Saudi Arabian site where Meriones crassus was
captured more frequently than were Meriones libycus or
three other co-occurring rodents, M. crassus was the most
competent reservoir for ZCL.
Although the evidence for a link between abundance
and reservoir competence of hosts is suggestive for a few
zoonoses, data are insufficient for a thorough evaluation
of this condition more generally. Pathogens that are
transmitted by generalist vectors to have had opportunities
to interact, and possibly to coevolve, with multiple host
species across many ecological communities. Under such
conditions, we expect that selection might favor pathogen
genotypes that are able to exploit the dominant members
of ecological communities, which would provide them
with the most stable "habitats" promoting persistence. We
also urge that more attention be paid to host species that
are poor or incompetent reservoirs for zoonotic agents, and
to rarer members of vertebrate communities. Such species
may provide vector populations with abundant
opportunities to feed but little opportunity for infection.
According to the dilution-effect model, increasing the
diversity of the community of hosts of vectors will lead
to a greater proportion of blood meals being taken from
rarer, less competent reservoirs, resulting in lower infection
prevalence in the vector population. The suppressive

effects of high diversity on disease risk will be reinforced
when species added to the host community either diminish
the population density of the primary reservoir host, e.g.,
via predation or competition, or reduce the absolute vector
burden on the reservoir host, e.g., by diverting vector
meals away from the reservoir host to incompetent
reservoirs.
Moreover, the presence of alternative host species
with low reservoir competence may, by reducing
encounter rates between infected vectors and hosts, reduce
the effective reservoir competence of other host species.
Although the dilution-effect model was developed for
Lyme disease, it may apply, in principle, to many other
vector-borne zoonoses. The four attributes necessary for
the dilution effect to operate seem to apply broadly, but
data are insufficient for a rigorous analysis of its general
applicability.
If disease risk were related solely to the population
abundance of a single reservoir host, then the dilution
effect would not operate. On the other hand, overall
diversity (e.g., the Shannon index) within a community
of hosts will decrease whenever the most abundant species
increases in absolute abundance, if the remaining species
remain at constant density. Therefore, we expect the
dynamics of particular, common hosts to covary with
cOmr:r\unity diversity. Experimental studies that separately
manipulate absolute or relative abundance of particular
species and community diversity per se will facilitate
obtaining an answer to this question. In addition, multiple
regression models in which abundances of single host
species and diversity metrics for the entire community are
used as independent variables should prove useful.
Species diversity can be estimated using metrics such
as species richness (number of species) or diversity indices,
which incorporate species richness and evenness
(proportional representation of each species). Which metric

seems most appropriate will vary according to specific
features of models linking diversity to disease risk. In a
model in which ticks passively encounter hosts in
proportion to host abundance, diversity indices that
incorporate evenness would seem most appropriate,
because evenness, not richness, would capture the
probability of encounter between a passive, nonselective
vector and each species of host. In a model in which
vectors are selective of hosts, or in which host species
interact (via predation or competition), species richness
would seem more appropriate for capturing the impacts
of the addition or deletion of particular species.
Claims that habitat destruction and alteration affect
disease transmission are frequent but rarely demonstrated.
For some diseases, such as yellow fever, dengue fever, and
African trypanosomiasis, disturbance of natural habitats
appears to alter the behavior of vectors so that encounter
rates between vectors and humans are elevated. The
dilution effect suggests an alternative mechanism by which
habitat alteration can influence disease risk. Habitat
destruction and the fragmentation of landscapes into small
isolated units are known to cause reduction or elimination
of some vertebrate species and therefore of diversity.
Often, the species most sensitive to s1;lch habitat
destruction are large species that occupy high trophic
levels, such as raptors and carnivorous mammals.
Loss of these species, which are rarely found to be
competent reservoirs for vector-borne zoonoses, may
increase disease risk both via reduction in feeding
opportunities from these incompetent hosts and via the
loss of a regulatory effect of such predators on typically
more reservoir-competent rodents.
If disease risk, as measured by vector infection
prevalence or density of infected vectors, decreases more
or less linearly with increasing vertebrate diversity, a
critical implication would be that any loss of diversity will

result in a measurable increase in disease risk. On the other
hand, analyses of Lyme disease by Ostfeld and Keesing
and Schmidt and Ostfeld suggest that the association
between diversity and disease risk is better described by
a saturating function, so changes in diversity have a strong
effect only on speciespoor communities. Similar, saturating
curves describing the relationship between diversity and
ecosystem function appear to be commonplace. In such
cases, the system may be more resilient to the loss of
diversity; highly diverse vertebrate communities should
be affected only modestly, if at all, by modest losses in
diversity, whereas already depauperate communities
should be affected dramatically.
Disease systems inconsistent with the dilution effect
should be sought. One likely exception to the beneficial
impact of host diversity on disease risk i~ the plague
epizootic, in which the pathogen (Yersinia pestis) is
transmitted from nonhuman mammal reservoirs to
humans by several species of fleas. Unlike most zoonotic
pathogens, Y. pestis is highly pathogenic to virtually all
hosts, often causing mortality in ca. 99% of infected
individuals across a wide range of species. As a result of
high mortality rates, plague tends to occur as a rapidly
developing epizootic that decimates host populations and
then goes locally extinct. In this case, low diversity in the
host community appears to facilitate rapid extinction of
the plague epizootic, whereas more diverse mammal
communities provide the pathogen with refugia from rapid
local extinction, maintaining its potential to infect people.
2
Natural Ecosystems
The biological underpinnings are encompassed in the

phrase ecosystem services, which refers to a wide range
of conditions and processes through which natural
ecosystems, and the species that are part of them, help
sustain and fulfill human life. The biodiversity services
maintain biodiversity and the production of ecosystem
goods, such as seafood, wild game, forage, timber, biomass
fuels, natural fibers, and many pharmaceuticals, industrial
products, and their precursors.
Many societies today have technological capabilities
undreamed of in centuries past. Their citizens have such
a global command of resources that even foods flown in
fresh from allover the planet are taken for granted, and
daily menus are decoupled from the limitations of regional
growing seasons and soils.
These developments have focused so much attention
upon human-engineered and exotic sources of fulfillment
that they divert attention from the local biological
underpinnings that remain essential to economic
prosperity and other aspects of our well-being. The harvest
and trade of these goods represent important and familiar
parts of the human economy. In addition to the production
of goods, ecosystem services support life through:
Natural Ecosystems 23
purification of air and water.

mitigation of droughts and floods.
generation and preservation of soils and renewal of
their fertility.
detoxification and decomposition of wastes.
pollination of crops and natural vegetation.
dispersal of seeds.
cycling and movement of nutrients.
control of the vast majority of potential agricultural
pests.
maintenance of biodiversity.
protection of coastal shores from erosion by waves.
protection from the sun's harmful ultraviolet rays.
partial stabilisa,tion of climate.
moderation of weather extremes and their impacts.
provision of aesthetic beauty and intellectual
stimulation that lift the human spirit.
Although the distinction between inaturali and ihuman-
dominatedi ecosystems is becoming increasingly blurred,
here emphasise the natural end of the spectrum, for three
related reasons. First, the services flowing from natural
ecosystems are greatly undervalued by society. For the
most part, they are not traded in formal markets and so
do not send price signals that warn of changes in their
supply or condition.
Furthermore, few people are conscious of the role
natural ecosystem services play in generating those
ecosystem goods that are traded in the marketplace. As a
result, this lack of awareness helps drive the conversion
of natural ecosystems to human-dominated systems (e.g.,
wheatlands or oil palm fields), whose economic value can

be expressed, at least in part, in standard currency. The
second reason to focus on natural ecosystems is that many
human-initiated disruptions of these systems - such as
introductions of exotic species, extinctions of native
species, and alteration of the gaseous composition of the
atmosphere through fossil fuel burning - are difficult or
impossible to reverse on any time scale relevant to society.
Third, if awareness is not increased and current trends
continue, humanity will dramatically alter Earth's
remaining natural ecosystems within a few decades. The
lack of attention to the vital role of natural ecosystem
services is easy to understand. Humanity came into being
after most ecosystem services had been in operation for
hundreds of millions to billions of years. These services
are so fundamental to life that they are easy to take for
granted, and so large in scale that it is hard to imagine
that human activities could irreparably disrupt them.
Perhaps a thought experiment that removes these
services from the familiar backdrop of the Earth is the best
way to illustrate both the importance and complexity of
ecosystem services, as well as how ill-equipped humans
are to recreate them. Imagine, for example, human beings
trying to colonise the moon. Assume for the sake of
argument that the moon had already miraculously
acquired some of the basic conditions for supporting
human life, such as an atmosphere, a climate, and a
physical soil structure similar to those on Earth.
The big question facing human colonists would then
be, which of Earth's millions of species would need to be
transported to the moon to make that sterile surface
habitable? One could tackle that question systematically
by first choosing from among all the species exploited
directly for food, drink, spices, fiber, timber,
pharmaceuticals, and industrial products such as waxes,
rubber, and oils. Even if one were highly selective, the
list could amount to hundreds or even thousands of

species.
And that would only be a start, since one would then
need to consider which species are crucial to supporting
those used directly: the bacteria, fungi, and invertebrates
that help make soil fertile and break down wastes and
organic matter; the insects, bats, and birds that pollinate
flowers; and the grasses, herbs, and trees that hold soil in
place, regulate the water cycle, and supply food for
animals. The clear message of this exercise is that no one
knows which combinations of species-or even
approximately how many-are required to sustain human
life.
Rather than selecting species directly, one might try
another approach: Listing the ecosystem services needed
by a lunar colony and then guessing at the types and
numbers of species required to perform each. Yet
determining which species are critical to the functioning
of a particular ecosystem service is no simple task. Soil
organisms are crucial to the chemical conversion and
physical transfer of essential nutrients to higher plants. But
the abundance of soil organisms is absolutely staggering.
Under a square-yard of pasture in Denmark, for
instance, the soil is inhabited by roughly 50,000 small
earthworms and their relatives, 50,000 insects and mites,
and nearly 12 million roundworms. And that tally is only
the beginning. The number of soil animals is tiny
compared to the number of soil microorganisms: a pinch
of fertile soil may contain over 30,000 protozoa, 50,000
algae, 400,000 fungi, and billions of individual bacteria.
Most of these soil-dwelling species have never been
subjected to even cursory inspection: no human eye has
ever blinked at them through a microscope, no human
hand has ever typed out a name or description of them,
and most human minds have never spent a moment
reflecting on them.
ECOSYSTEM SERVICES
Ecosystem services and the systems that supply them are

so interconnected that any classification of them is
necessarily rather arbitrary. Here briefly explore a suite of
overarching services that operate in ecosystems worldwide.
Ecosystem Goods
Humanity obtains from natural ecosystems an array of
ecosystem goods - organisms and their parts and
products that grow in the wild and that are used directly
for human benefit. Many of these, such as fishes and
animal products, are commonly traded in economic
markets. The annual world fish catch, for example,
amounts to about 100 million metric tons and is valued at
between $50 billion and $100 billion; it is the leading
source of animal protein, with over 20% of the population
in Africa and Asia dependent on fish as their primary
source of protein.
The commercial harvest of freshwater fish worldwide
in 1990 totaled approximately 14 million tons and was
valued at about $8.2 billion. Interestingly, the value of the
freshwater sport fishery in the U.S. alone greatly exceeds
that of the global commercial harvest, with direct
expenditures in 1991 totaling about $16 billion. When this
is added to the value of the employment generated by
sport fishing activities, it raises the total to $46 billion.
The future of these fisheries is in question, however,
because fish harvests have approached or exceeded
sustainable levels virtually everywhere. Nine of the
world's major marine fishing areas are in decline due to
overfishing, pollution, and habitat destruction. Turning
attention to the land, grasslands are an important source
of marketable goods, including animals used for labor
(horses, mules, asses, camels, bullocks, etc.) and those
whose parts or products are consumed (as meat, milk,

wool, and leather).
Grasslands were also important as the original source
habitat for most domestic animals such as cattle, goats,
sheep, and horses, as well as many crops, such as wheat,
barley, rye, oats, and other grasses. In a wide variety of
terrestrial habitats, people hunt game animals such as
waterfowl, deer, moose, elk, fox, boar and other wild pigs,
rabbits, and even snakes and monkeys. In many countries,
game meat forms an important part of local diets and, in
many places, hunting is an economically and culturally
important sport.
Natural ecosystems also produce vegetation used
directly by humans as food, timber, fuelwood, fiber,
pharmaceuticals and industrial products. Fruits, nuts,
mushrooms, honey, other foods, and spices are extracted
from many forest species. Wood and other plant materials
are used in the construction of homes and other buildings,
as well as for the manufacture of furniture, farming
implements, paper, cloth, thatching, rope, and so on.
About 15 percent of the world's energy consumption
is supplied by fuelwood and other plant material; in
developing countries, such ibiomassi supplies nearly 40
percent of energy consumption, although the portion of
this derived from natural rather than human-dominated
ecosystems is undocumented.
In addition, natural products extracted from many
hundreds of species contribute diverse inputs to industry:
gums and exudates, essential oils and flavorings, resins
and oleoresins, dyes, tannins, vegetable fats and waxes,
insecticides, and multitudes of other compounds. The
availability of most of these natural products is in decline
due to ongoing habitat conversion.
Biological diversity, or biodiversity for short, refers
to the variety of life forms at all levels of organisation,
from the molecular to the landscape level. Biodiversity is

generated and maintained in natural ecosystems, where
organisms encounter a wide variety of living conditions
and chance events that shape their evolution in unique
ways. Out of convenience or necessity, biodiversity is
usually quantified in terms of numbers of species, and this
perspective has greatly influenced conservation goals.
It is important to remember, however, that the
benefits that biodiversity supplies to humanity are
delivered through populations of species residing in living
communities within specific physical settings - in other
words, through coinplex ecological systems, or ecosystems.
For human beings to realise most of the aesthetic, spiritual,
and economic benefits of biodiversity, natural ecosystems
must therefore be accessible. The continued existence of
coniferous tree species somewhere in the world would not
help the inhabitants of a town inundated by flooding
because of the clearing of a pine forest upstream.
Generally, the flow of ecosystem goods and services
in a region is determined by the type, spatial layout, extent,
and proximity of the ecosystems supplying them. Because
of this, the preservation of only one minimum viable
popUlation of each non-human species on Earth in zoos,
botanical gardens, and the world's legally protected areas
would not sustain life as we know it. Indeed, such a
strategy, taken to extreme, would lead to collapse of the
biosphere, along with its life support services.
Biodiversity is a direct source of ecosystem goods. It
also supplies the genetic and biochemical resources that
underpin current agricultural and pharmaceutical
enterprises and may allow to adapt these vital enterprises
to global change. To increase crop productivity in the face
of new pests, diseases, and other stresses has depended
heavily upon the transfer to our crops of genes from wild
crop r~latives that confer resistance to these challenges.
Such extractions from biodiversity's igenetic libraryi

account for annual increases in crop productivity of about
1 percent, currently valued at $1 billion. Biotechnology
now makes possible even greater use of this natural
storehouse of genetic diversity via the transfer to crops of
genes from any kind of organism - not simply crop
relatives - and it promises to playa major role in future
yield increases. By the tum of the century, farm-level sales
of the products of agricultural biotechnology, just now
entering the marketplace, are expected to reach at least
$10 billion per year.
In addition to sustaining the production of
conventional crops, the biodiversity in natural ecosystems
may include many potential new foods. Human beings
have utilized around 7,000 plant species for food over the
course of history and another 70,000 plants are known to
have edible parts. Only about 150 food plants have ever
been cultivated on a large scale, however. Currently, 82
plant species contribute 90 percent of national per-capita
supplies of food plants, although a much smaller number
of these supply the bulk of the calories humans consume.
Many other species, however, appear more nutritious
or better suited to the growing conditions that prevail in
important regions than the standard crops that dominate
world food supply today. Because of increasing salinisation
of irrigated croplands and the potential for rapid climate
change, for instance, future food security may corne to
depend on drought-and salt-tolerant varieties that now
play comparatively minor roles in agriculture. Turning to
medicinal resources, a recent survey showed that of the
top 150 prescription drugs used in the United States, 118
are based on natural sources: 74% on plants, 18'X. on fungi,
5% on bacteria, and 3% on one vertebrate (snake) specle~.
The commercial value of pharmaceuticals in the
developed nations exceeds $40 billion per year. Looking
at the global picture, approximately 80% of the human
population relies on traditional medical systems, and about

85'1<, of traditional medicine involves the use of plant
extracts. Saving only a single population of each species
could have another cost.
Different populations of the same species may
produce different types or quantities of defensive
chemicals that have potential use as pharmaceuticals or
pesticides; and they may exhibit different tolerances to
environmental stresses such as drought or soil salinity. For
example, the development of penicillin as a therapeutic
antibiotic took a full 15 years after Alexander Fleming's
famous discovery of it in common bread mold.
In part, this was because scientists had great difficulty
producing, extracting, and purifying the substance in
needed quantities. One key to obtaining such quantities
was the discovery, after a worldwide search, of a
population of Fleming's mold that produced more
penicillin than the original. Similarly, plant populations
vary in their ability to resist pests and disease, traits
important in agriculture. Many thousands of varieties of
rice from different locations were screened to find one with
resistance to grassy stunt virus, a disease that posed a
serious threat to the world's rice crop. Despite numerous
examples like these, many of the localities that harbor wild
relatives of crops remain unprotected and heavily
threatened.
Climate and Life

Earth's climate has fluctuated tremendously since
humanity came into being. At the peak of the last ice age
20,Ollll years ago, for example, much of Europe and North
America were covered by mile-thick ice sheets. While the
global climate has been relatively stable since the invention
of agriculture around 10,000 years ago, periodic shifts in
climate have affected human activities and settlement
patterns. Even relatively recently, from 1550-1850, Europe

was significantly cooler during a period known as the
Little Ice Age.
Many of these changes in climate are thought to be
caused by alterations in Earth's orbital rotation or in the
energy output of the sun, or even by events on the Earth
itself - sudden perturbations such as violent volcanic
eruptions and asteroid impacts or more gradual tectonic
events such as the uplift of the Himalayas. Remarkably,
clima te has been buffered enough through all these
changes to sustain life for at least 3.5 billion years. And
life itself has played a role in this buffering.
Climate, of course, plays a major role in the evolution
and distribution of life over the planet. Yet most scientists
would agree that life itself is a principal factor in the
regulation of global climate, helping to offset the effects
of episodic climate oscillations by responding in ways that
alter the greenhouse gas concentrations in the atmosphere.
For instance, natural ecosystems may have helped to
stabilise climate and prevent overheating of the Earth by
removing more of the greenhouse gas carbon dioxide from
the atmosphere as the sun grew brighter over millions of
years.
Life may also exert a destabilising or positive
feedback that reinforces climate change, particularly during
transitions between interglacial periods and ice ages. One
example: When climatic cooling leads to drops in sea level,
continental shelves are exposed to wind and rain, causing
greater nutrient runoff to the oceans. These nutrients may
fertilize the growth of phytoplankton, many of which form
calcium carbonate shells. Increasing their populations
would remove more carbon dioxide from the oceans and
the atmosphere, a mechanism that should further cool the
planet.
Living things may also enhance warming trends

through such activities as speeding up microbial
decomposition of dead organic matter, thus releasing
carbon dioxide to the atmosphere. The relative influence
of life's stabilising and destabilising feedbacks remains
uncertain; what is clear is that climate and natural
ecosystems are tightly coupled, and the stability of that
coupled system is an important ecosystemserv~ce.
Besides their impact on the atmosphere, ecosystems
also exert direct physical influences that help to moderate
regional and local weather. For instance, transpiration
(release of water vapor from the leaves) of plants in the
morning causes thunderstorms in the afternoon, limiting
both moisture loss from the region and the rise in surface
temperature. In the Amazon, for example, 50% of the mean
annual rainfall is recycled by the forest itself via
evapotranspiration - that is, evaporation from wet leaves
and soil combined with transpiration.
Amazon deforestation could so dramatically reduce
total precipitation that the forest might be unable to
reestablish itself following complete destruction.
Temperature extremes are also moderated by forests,
which provide shade and surface cooling and also act as
insulators, blocking searing winds and trapping warmth
by acting as a local greenhouse agent.
Mitigation of Floods and Drougltts

An enormous amount of water, about 119,000 cubic
kilometers, is rained annually onto the Earth's land surface
- enough to cover the land to an average depth of 1
meter. Much of this water is soaked up by soils and
gradually meted out to plant roots or into aquifers and
surface streams. Thus, the soil itself slows the rush of
water off the land in flash floods. Yet bare soil is
vulnerable. Plants and plant litter shield the soil from the
full, destructive force of raindrops and hold it in place.
When landscapes are denuded, rain compacts the
surface and rapidly turns soil to mud (especially if it has
been loosened by tillage); mud clogs surface cavities in
the soil, reduces infiltration of water, increases runoff, and
further enhances clogging. Detached soil particles are
splashed downslope and carried off by running water.
Erosion causes costs not only at the site where soil is
lost but also in aquatic systems, natural and human-made,
where the material accumulates. Local costs of erosion
include losses of production potential, diminished
infiltration and water availability, and losses of nutrients.
Downstream costs may include disrupted or lower quality
water supplies; siltation that impairs drainage and
maintenance of navigable river channels, harbors, and
irrigation systems; increased frequency and severity of
flo.ods; and decreased potential for hydroelectric power as
reservoirs fill with silt.
Worldwide, the replacement cost of reservoir capaci ty
lost to siltation is estimated at $6 billion per year. In
addition to protecting soil from erosion, living vegetation
- with its deep roots and above-ground evaporating
surface - also serves as a giant pump, returning water
from the ground into the atmosphere. Clearing of plant
cover disrupts this link in the water cycle and leads to
potentially large increases in surface runoff, along with
nutrient and soil loss. A classic example comes from the
experimental clearing of a New Hampshire forest, where
herbicide was applied to prevent regrowth for a 3-year
period after the clearing. The r~sult was a 40 percent
increase in average stream flow. During one four-month
period of the experiment, runoff was more than 5 times
greater than before the clearing.
On a much larger scale, extensive deforestation in the

Himalayan highlands appears to have exacerbated recent
flooding in Bangladesh, although the relative roles of
human and natural forces remain debatable. In addition,
some regions of the world, such as parts of Africa, are
experiencing an increased frequency and severity of
drought, possibly associated with extensive deforestation.
Wetlands are particularly well-known for their role
in flood control and can often reduce the need to construct
flood control structures. Floodplain forests and high salt
marshes, for example, slow the flow of floodwaters and
allow sediments to be deposited within the floodplain
rather than washed into downstream bays or oceans. In
addition, isolated wetlands such as prairie potholes in the
Midwest and cypress ponds in the Southeast, serve as
detention areas during times of high rainfall, delaying
saturation of upland soils and overland flows into rivers
and thereby damping peak flows.
Retaining the integrity of these wetlands by leaving
vegetation, soils, and natural water regimes intact can
reduce the severity and duration of flooding along rivers.
A relatively small area of retained wetland, for example,
could have largely prevented the severe flooding along
the Mississippi River in 1993.
Soil
Soil represents an important component of a nation's
assets, one that takes hundreds to hundreds of thousands
of years to build up and yet very few years to be lost.
Some civilizations have drawn great strength from fertile
soil; conversely, the loss of productivity through
mismanagement is thought to have ushered many once
flourishing societies to their ruin. Today, soil degradation
induced by human activities afflicts nearly 20 percent of
the Earth's vegetated land surface.
In addition to moderating the water cycle, as

described above, soil provides five other interrelated
services. First, soil shelters seeds and provides physical
support as they sprout and mature into adult plants. The
cost of packaging and storing seeds and of anchoring plant
roots would be enormous without soil. Human-engineered
hydroponic systems can grow plants in the absence of soil,
and their cost provides a lower bound to help assess the
value of this service. The costs of physical support trays
and stands used in such operations total about US$55,OOO
per hectare.
Second, soil retains and delivers nutrients to plants.
Tiny soil particles (less than 2 microns in diameter), which
are primarily bits of humus and clays, carry a surface
electrical charge that is generally negative. This property
holds positively charged nutrients - cations such as
calcium and magnesium - near the surface, in proximity
to plant roots, allowing them to be taken up gradually.
Otherwise, these nutrients would quickly be leached away.
Soil also acts as a buffer in the application of fertilizers,
holding onto the fertilizer ions until they are required by
plants.
Hydroponic systems supply water and nutrients to
plants without need of soil, but the margin for error is
much smaller - even small excesses of nutrients applied
hydroponically can be lethal to plants. Indeed, it is a
complex undertaking to regulate the nutrient
concentrations, pH, and salinity of the nutrient solution
in hydroponic systems, as well as the air and solution
temperature, humidity, light, pests, and plant diseases.
Worldwide, the area under hydroponic culture is only a
few thousand hectares and is unlikely to grow significantly
in the foreseeable future; by contrast, global cropped area
is about 1.4 billion hectares.
Third, soil plays a central role in the decompOSition
of dead organic matter and wastes, and this decompOSition
process also renders harmless many potential human

pathogens. People generate a tremendous amount of waste,
including household garbage, industrial waste, crop and
forestry residues, and sewage from their own populations
and their billions of domesticated animals. A rough
approximation of the amount of dead organic matter and
waste (mostly agricultural residues) processed each year
is 130 billion metric tons, about 30 percent of which is
associated with human activities.
Fortunately, there is a wide array of decomposing
organisms - ranging from vultures to tiny bacteria - that
extract energy from the large, complex organic molecules
found in many types of waste. Like assembly-line workers,
diverse microbial species process the particular compounds
whose chemical bonds they can cleave and pass along to
other species the end products of their specialised
reactions. Many industrial wastes, including soaps,
detergents, pesticides, oil, acids, and paper, are detoxified
and decomposed by organisms in natural ecosystems if
the concentration of waste does not exceed the system's
capacity to transform it. Some modern wastes, however,
are virtually indestructible, such as some plastics and the
breakdown products of the pesticide DDT.
The simple inorganic chemicals that result from
natural decomposition are eventually returned to plants
as nutrients. Thus, the decomposition of wastes and the
recycling of nutrients - the fourth service soils provide
- are two aspects of the same process. The fertility of
soils - that is, their ability to supply nutrients to plants
- is largely the result of the activities of diverse species
of bacteria, fungi, algae, crustacea, mites, termites,
springtails, millipedes, and worms, all of which, as groups,
play important roles. Some bacteria are responsible for
ifixingi nitrogen, a key element in proteins, by drawing it
out of the atmosphere and converting it to forms usable
by plants and, ultimately, human beings and other
animals. Certain types of fungi play extremely important

roles in supplying nutrients to many kinds of trees.
Earthworms and ants act as Imechanical blenders,i
breaking up and mixing plant and microbial material and
other matter. For example, as much as 10 metric tonnes of
material may pass through the bodies of earthworms on
a hectare of land each year, resulting in nutrient rich kastsi
that enhance soil stability, aeration, and drainage.
Finally, soils are a key factor in regulating the Earth's
major element cycles - those of carbon, nitrogen, and
sulfur. The amount of carbon and nitrogen stored in soils
dwarfs that in vegetation, for example. Carbon in soils is
nearly double (1.8 times) that in plant matter, and nitrogen
in soils is about 18 times greater.
Alterations in the carbon and nitrogen cycles may be
costly over the long term, and in many cases, irreversible
on a time scale of interest to society. Increased fluxes of
carbon to the atmosphere, such as occur when land is
converted to agriculture or when wetlands are drained,
contribute to the buildup of key greenhouse gases, namely
carbon dioxide and methane, in the atmosphere.
Changes in nitrogen fluxes caused by production and
use of fertilizer, burning of wood and other biomass fuels,
and clearing of tropical land lead to increasing atmospheric
concentrations of nitrous oxide, another potent greenhouse
gas that is also involved in the destruction of the
stratospheric ozone shield. These and other changes in the
nitrogen cycle also result in acid rain and excess nutrient
inputs to freshwater systems, estuaries, and coastal marine
waters. This nutrient influx causes eutrophication of
aquatic ecosystems and contamination of drinking water
sources - both surface and ground water - by high levels
of nitrate-nitrogen.
Animal Pollination
Animal pollination is required for the successful
reproduction of most flowering plants. About 220,000 out
of an estimated 240,000 species of plants for which the
mode of pollination has been recorded require an animal
such as a bee or hummingbird to accomplish this vital
task. This includes both wild plants and about 70 percent
of the agricultural crop species that feed the world. Over
100,000 different animal species - including bats, bees,
beetles, birds, butterflies, and flies - are known to provide
these free pollination services that assure the perpetuation
of plants in our croplands, backyard gardens, rangelands,
meadows and forests.
In tum, the continued availability of these pollinators
depends on the existence of a wide variety of habitat types
needed for their feeding, successful breeding, and
completion of their life cycles. One third of human food
is derived from plants pollinated by wild pollinators.
Without natural pollination services, yields of important
crops would decline precipitously and many wild plant
species would become extinct.
In the United States alone, the agricultural value of
wild, native pollinators - those sustained by natural
habitats adjacent to farmlands - is estimated in the
billions of dollars per year. Pollination by honey bees,
originally imported from Europe, is extremely important
as well, but these bees are presently in decline, enhancing
the importance of pollinators from natural ecosystems.
Management of the honey bee in the New World is
currently threatened by the movement of, and
hybridization with, an aggressive African strain of honey
bee that was accidentally released in Brazil in 1956.
Diseases of honey bee colonies are also causing a
marked decline in the number of managed colonies.
Meanwhile, the diversity of natural pollinators available
to both wild and domesticated plants is diminishing: more

than 60 genera of pollinators include species now
considered to be threatened, endangered or extinct.
Natural Pest Control Services

Humanity's competitors for food, timber, cotton, and other
fibers are called pests, and they include numerous
herbivorous insects, rodents, fungi, snails, nematodes, and
viruses. These pests destroy an estimated 25 to 50 percent
of the world's crops, either before or after harvest. In
addition, numerous weeds compete directly with crops for
water, light, and soil nutrients, further limiting yields.
Chemical pesticides, and the strategies by which they
are applied to fight crop pests, can have harmful
unintended consequences. First, pests can develop
resistance, which means that higher and higher doses of
pesticides must be applied or new chemicals developed
periodically to achieve the same level of control. Resistance
is now found in more than 500 insect and mite pests, over
100 weeds, and in about 150 plant pathogens. Second,
populations of the natural enemies of pests are decimated
by heavy pesticide use.
Natural predators are often more susceptible to
synthetic poisons than are the pests because they have not
had the same evolutionary experience with overcoming
plant chemicals that the pests themselves have had. And
natural predators also typically have much smaller
population sizes than those of their prey. Destruction of
predator populations leads to explosions in prey numbers,
not only freeing target pests from natural controls but
often ipromotingi other non-pest species to pest status.
In California in the 1970s, for instance, 24 of the 25
most important agricultural pests had been elevated to that
status by the overuse of pesticides. Third, exposure to
pesticides and herbicides may pose serious health risks to
humans and many other types of organisms; the recently

discovered declines in human sperm counts may be
attributable in part to such exposure.
Fortunately, an estimated 99 percent of potential crop
pests are controlled by natural enemies, including many
birds, spiders, parasitic wasps and flies, lady bugs, fungi,
viral diseases, and numerous other types of organisms.
These natural biological control agents save farmers
billions of dollars annually by protecting crops and
reducing the need for chemical control.
Seed Dispersal
Once a seed germinates, the resulting plant is usually
rooted in place for the rest of its life. For plants, then,
movement to new sites beyond the shadow of the parent
is usually achieved through seed dispersal. Many seeds,
such as those of the dandelion, are dispersed by wind.
Some are dispersed by water, the most famous being the
seafaring coconut. Many other seeds have evolved ways
of getting around by using animals as their dispersal
agents.
These seeds may be packaged in sweet fruit to reward
an animal for its dispersal services; some of these seeds
even require passage through the gut of a bird or mammal
before they can germinate. Others require burial - by,
say, a forgetful jay or a squirrel which later leaves its cache
uneaten - for eventual germination. Still others are
equipped with sticky or sharp, spiny surfaces designed to
catch onto a passing animal and go for a long ride before
dropping or being rubbed off. Without thousands of
animal species acting as seed dispersers, many plants
would fail to reproduce successfully.
For instance, the whitebark pine (Pinus albicaulis), a
tree found in the Rockies and Sierra Nevada - Cascade
Mountains, cannot reproduce successfully without a bird
called Clark's Nutcracker (Nucifraga columbiana), which

chisels pine seeds out of the tightly closed cones and
disperses and buries them; without this service, the cones
do not open far enough to let the seeds fall out on their
own. Animal seed dispersers play a central role in the
structure and regeneration of many pine forests.
Disruption of these complex services may leave large areas
of forest devoid of seedlings and younger age classes of
trees, and thus unable to recover swiftly from human
impacts such as land clearing.
Many human beings have a deep appreciation of
natural ecosystems. That is apparent in the art, religions,
and traditions of diverse cultures, as well as in activities
such as gardening and pet-keeping, nature photography
and film-making, bird feeding and watching, hiking and
camping, ecotouring and mountaineering, river-rafting and
boating, fishing and hunting, and in a wide range of other
activities. For many, nature is an unparalleled source of
wonderment and inspiration, peace and beauty, fulfillment
and rejuvenation.
THREATS TO ECOSYSTEM SERVICES
Ecosystem services are being impaired and destroyed by

a wide variety of human activities. Foremost among the
immediate threats are the continuing destruction of natural
habitats and the invasion of non-native species that often
accompanies such disruption; in marine systems,
overfishing is a major threat. The most irreversible of
human impacts on ecosystems is the loss of native
biodiversity. A conservative estimate of the rate of species
loss is about one per hour, which unfortunately exceeds
the rate of evolution of new species by a factor of 10,000
or more.
But complete extinction of species is only the finai
act in the process. The rate of loss of local populations of
species - the populations that generate ecosystem services

in specific localities and regions - is orders of magnitude
higher. Destroying other life forms also disrupts the web
of interactions that could help to discover the potential
usefulness of specific plants and animals. Once a pollinator
or a predacious insect is on the brink of extinction, for
instance, it would be difficult to discover its potential
utility to farmers.
Other imminent threats include the alteration of the
Earth's carbon, nitrogen, and other biogeochemical cycles
through the burning of fossil fuels and heavy use of
nitrogen fertilizer; degradation of farmland through
unsustainable agricultural practices; squandering of
freshwater resources; toxification of land and waterways;
and overharvesting of fisheries, managed forests, and other
theoretically renewable systems.
These threats to ecosystem services are driven
ultimately by two broad underlying forces. One is rapid,
unsustainable growth in the scale of the human enter-prise:
in population size, in per-capita consumption, and also in
the environmental impacts that technologies and
institutions generate as they produce and supply those
consumables. The other underlying driver is the frequent
mismatch between short-term, individual economic
incentives and long-term, societal well-being.
Ecosystem services are generally greatly undervalued,
for a number of reasons: many are not traded or valued
in the marketplace; many serve the public good rather than
provide direct benefits to individual landowners; private
property owners often have no way to benefit financially
from the ecosystem services supplied to society by their
land; and, in fact, economic subsidies often encourage the
conversion of such lands to other, market-valued activities.
Thus, people whose activities disrupt ecosystem
services often do not pay directly for the cost of those lost
services. Moreover, society often does not compensate

landowners and others who do safeguard ecosystem
services for the economic benefits they lose by foregoing
more lucrative but destructive land uses. There is a critical
need for policy measures that address these driving forces
and embed the value of ecosystem services into decision
making frameworks.
VALUATION OF ECOSYSTEM SERVICES
Human society would cease to exist in the absence of

ecosystem services. Thus, their immense value to humanity
is unquestionable. Yet quantifying th~ value of ecosystem
services in specific localities, and measuring their worth
against that of competing land uses is no simple task.
When tradeoffs must be made in the allocation of land
and other resources to competing human activities, the
resolution often requires a measure of what is known as
the marginal value.
In the case of ecosystem services, for example, the
question that might be posed would be: By how much
would the flow of ecosystem services be augmented (or
diminished) with the preservation (or destruction) of the
next hectare of forest or wetland? Estimation of marginal
values is complex. Often a qualitative comparison of
relative values is sufficient - that is, which is greater, the
economic benefits of a particular development project or
the benefits supplied by the ecosystem that would be
destroyed, measured over a time period of interest to
people concerned about the well-being of their
grandchildren?
There are, and will remain, many cases in which
ecosystem service values are highly uncertain. Yet the pace
of destruction of natural ecosystems, and the irreversibility
of most such destruction on a time scale of interest to
humanity, warrants substantial caution. Valuing a natural
ecosystem, like valuing a human life, is fraught with

difficulties. Just as societies have recognised fundamental
human rights, however, it may be prudent to establish
fundamental ecosystem protections even though
uncertainty over economic values remains.
New institutions and agreements at the international
and subnational level will be needed to encourage fair
participation in such protections. The tremendous expense
and difficulty of replicating lost ecosystem services is
perhaps best illustrated by the results of the first Biosphere
2 imissionJ in which eight people lived inside a 3.1S-acre
closed ecosystem for two years. The system featured
agricultural land and replicas of several natural ecosystems
such as forests and even a miniature ocean.
In spite of an investment of more than $200 million
in the design, construction, and operation of this model
earth, it proved impossible to supply the material and
physical needs of the eight Biospherians for the intended
2 years. Many unpleasant and unexpected problems arose,
including a drop in atmospheric oxygen concentration to
14% (the level normally found at an elevation of 17,500
feet), high spikes in carbon dioxide concentrations, nitrous
oxide concentrations high enough to impair the brain, an
extremely high level of extinctions, overgrowth of
aggressive vines and algal mats, and population explosions
of crazy ants, cockroaches, and katydids.
Even heroic personal efforts on the part of the
Biospherians did not suffice to make the system viable and
sustainable for either humans or many nonhuman species.
3
Impact of Climate Change on
Biodiversity
Human activities have led to changes in ecosystems and

attendant loss of biodiversity in many regions. Regional
changes in climate and temperature, have affected a
diverse set of physical and biological systems in many
parts of the world. Examples of observed changes include
shrinking glaciers, earlier break up of ice on rivers and
lakes, lengthening of mid to high latitude growing seasons,
poleward and altitudinal shifts in the ranges of some plant
and animal species, declines in some plant and animal
populations and earlier flowering of trees and emergence
of insects and egg-laying of birds. In fact, from the
collective evidence there is high confidence that recent
regional changes in temperature have had discernible
impacts on many physical and biological systems.
As a result of changing climate, ecosystem services
that are crucial to human survival may be affected, or may
have already been affected. Of particular concern is the
critical role of ecosystems in global biogeochemical
processes that underline the functioning of the earth's
systems. Of further concern is the potential that indigenous
and rural communities may lose access to services like
food, fibre, fuel, energy, fodder and medicines, and the
consequences of biodiversity loss to cultural, spiritual,
aesthetic and recreational values. The effects of climate

change and biodiversity loss will not operate in isolation,
but rather must be understood in the context of global
change, incorporating pressures such as increased demand
for natural resources (including water), exploitation or
destruction of biodiversity through human activities, land
use change, pollution, and the destruction of the ozone
layer.
Approximately 1.7 billion people, one third of the
world's population, presently live in countries that are
water stressed (defined as using more than 20% of their
renewable water supply). Demand for water is generally
increasing due to population growth and economic
development. In addition, climate change may
substantially affect irrigation withdrawals. Higher
temperatures will lead to higher crop evaporative
demands, but overall effects on withdrawal will depend
on how increases in evaporation are offset by changes in
precipitation. However, the general tendency will be
towards an increase in irrigation demands. As a result of
increased temperature and reduced flow, water quality
within streams and rivers will generally be degraded.
In addition, changes in hydrological processes will
affect many wetland habitats. For example, about 20% of
coastal wetlands could be lost by 2030 as a result of sea
level rise. And, increased temperatures are expected to
lead to widespread coral bleaching. Associated with these
changes are likely to be major losses of biodiversity.
MITIGATION AND ADAPTATION
Broadly, the two tools to address to climate change are

mitigation and adaptation. Mitigation is defined as a
human intervention to reduce net greenhouse gas
emissions, or to increase sequestration of CO2 through
sinks such as forests. Adaptation to climate change,
Impact of Climate Change on Biodiversity 47
particularly at the community level, can supplement

mitigation meaSUl:es, and reduce the adverse impacts of
climate change on human systems and biodiversity.
Clean Development Mechanism

The Clean Development Mechanism (CDM) of the Kyoto
Protocol is one of three flexibility mechanisms for cost-
effective mitigation of climate change. CDM can use either
technologies such as carbon neutral technologies or
sequestration of CO 2 through afforestation and
reforestation project activities.
The Article 12 of the Kyoto Protocol defines the objectives
of CDM as:
to assist developing countries in achieving sustainable
development;
to contribute to the ultimate objective of the
Convention i.e. stabilisation of greenhouse gas
concentrations in the atmosphere at levels that would
prevent dangerous anthropogenic interference with the
climate system;
to assist developed countries in achieving compliance
with their quantified emission limitation and reduction
commitments (QERCs).
The opportunities of the industrialised countries to reduce
emissions through CDM project activities in developing
countries are enormous and have fairly low associated
costs, particularly in the sectors of energy, transport,
building materials (brick, cement and steel etc.), municipal
wastes, and animal husbandry.
It is expected that CDM project activities will
contribute to progress in achieving sustainable
development in India, particularly under the four pillars
of sustainable development: social, economic,
environmental and technological well-being. In doing so,

this will help to address India's main agenda of
development-poverty eradication and improving the
quality of life of people. In addition, CDM projects will
attract additional foreign investment and India may be
entitled to a share of the certified emission reductions
(CERs) that accrue from CDM project activities. Finally,
the development and implementation of CDM projects will
contribute to capacity building within the country. To
participate in a CDM activity:
participation must be voluntary;
the country participating in the CDM activities shall
designate a national authority (DNA);
a country not included in the Annex I of the
Convention may participate in a CDM project activity
if the country has ratified the Kyoto Protocol.
The readiness of India to join the CDM process is
demonstrated by India's recent ratification of the Kyoto
Protocol, as well as by hosting the Eighth Conference of
the Parties to the United Nations Framework Convention
(COP 8) at New Delhi, and by the Prime Minister's address
at COP 8. The Government of India has already set up an
enabling environment to speed up the process of
endorsement and approval of CDM project proposals. It
is also in the process of establishing a designated national
authority (DNA) as required under the Marrakesh Accord.
Further evidence of India's readiness to join the CDM
process is clear in the Government of India's endorsement
of two CDM projects in response to the Certified Emission
Reduction Unit Procurement Tender (CERUPT) of the
Netherlands.
In addition, industrial associations and NGOs such
as ClI, Development Alternatives and other institutions
like IIM-Ahmedabad, TERI, Winrock are taking keen
interests, and during the last couple of years a number of
CDM projects have been developed in close collaboration

with the Indian business sector.
India conducted a study to produce a least-cost
greenhouse gas abatement strategy (ALGAS-India) and
identified potential sectors for greenhouse gas abatement.
The focal points of the abatement strategy are the
reduction of C02 emissions in the energy and forestry
sectors and the reduction of methane (CH4) emissions in
the agriculture sector. The mitigation options for the
energy sector are improvements in energy efficiency
through upgrading the currently employed technologies,
the introduction of advanced technologies that are more
efficient, and the use of renewable energy sources
wherever feasible.
The recently declared 'Energy Policy' of the
Government of India aims to increase the contribution of
energy from renewable sources to at least 10% of the total
supply (at least 10,000 Megawatt) by 2012. India has
considerable renewable energy potential and much has yet
to be exploited (Table 1).
The IPCC Third Assessment Report (IPCC 2002) on
Climate Change has clearly stated that adaptation at all
scales is a necessary strategy to complement climate
change mitigation efforts. Achieving sustainable
development, particularly through CDM processes, will
enhance country's capacity to adapt to climate change. To
strengthen the framework of adaptation, the world
community has to work as one to eradicate poverty, as
articulated in the United Nations Millennium Development
Goals. These goals are fundamental in developing the
coping capacity and resilience of vulnerable communities
in developing countries.
Table 1: Renewable Energy Potential in Illdia
Source I Technology Potential Potential

ExplOIted (MW)
Biogas Plants 12 million MW 2.7 million

Biomass Based Power 17,000 MW 769.5
Efficient Woodstoves 120 million MW 20 million
Solar Energy 5 x 1015 W hr/yr 25
Small Hydro 10,000 MW 250
Wind Energy 20,000 MW 1,000
Ocean Thermal 50,000 MW
Sea Wave Power 20,000 MW
Tidal Power 9,000 MW
Adaptation to Climate Change

Tools for adaptation to climate change that are considered
suitable for the developing countries such as India are:
empowering communities to reduce their
vulnerability;
education, training and public awareness;
sustainable livelihood practices (e.g. providing
electricity through renewable energy technologies
using local resources);
cooperative/participatory activities such as cooperative
banks;
insurance against natural disasters such as floods,
drought, cyclones and crop damage;
operational research on adaptation to increase
resilience and coping capacity of vulnerable
communities.
The success of adaptation tools can be measured using
indica tors such as:
- percentage of poverty reduction;
percentage improvement in public awareness (could

be achieved by introducing curriculum on climate
change in middle and high schools);
changes in per capita emission of CO2 per year;
GOP per unit of energy use;
yearly incidences of malaria / dengue fever;
percentage of people having access to clean water and
sanitation.
',,'
4
Biodiversity and
Ecosystem Functioning
One of the most striking features of the earth's biota is its

extraordinary diversity, estimated to include about 10
million different species. One of the most conspicuous
aspects of contemporary global change is the rapid decline
of this diversity in many ecosystems. The decline is not
limited to increased rates of species extinction, but includes
losses in genetic and functional diversity across population,
community, ecosystem, landscape, and global scales.
The wide-ranging decline in biodiversity results
largely from habitat modifications and destruction,
increased rates of invasions by deliberately or accidentally
introduced non-native species, over-exploitation and other
human-caused impacts. On a global scale, even at the
lowest estimated current extinction rate, about half of all
species could be extinct within 100 years. Such an event
would be similar in magnitude to the five mass extinction
events in the 3.5 billion year history of life on earth.
On local and regional scales, biodiversity declines are
already pronounced in many areas, especially where
natural ecosystems have been converted to croplands,
timber plantations, aquaculture and other managed
ecosystems. The diversity of these managed ecosystems is
Biodiversity and Ecosystem Functioning 53
often low, and species composition very different,

compared with those of the natural systems they have
replaced.
The earth's living organisms contribute to human
welfare in a variety of ways. First, humans derive from
them goods and products essential to life, including food,
medicine, and industrial products, genetic resources for
crop breeding, and natural pest control services. Such
benefits can be viewed as the market values of biodiversity
because they are are readily tied to our economy and often
can be assigned a dollar value in the marketplace. Second,
biodiversity has nonmarket values that can be expressed
in terms such as knowledge, aesthetic, existence and other
values. These non-market values of biodiversity are
difficult to quantify, but are, for many, sufficient
justification for preserving biodiversity independent of
market values.
The organisms that live, grow, reproduce, and interact
within ecosystems help to mediate local and regional flows
of energy and materials. Energy flow refers to the capture
of light energy by green plant or algal photosynthesis and
its dispersal as chemical energy throughout the food web
to plant- or algal-feeding animals, predators, and
eventually decomposers. The flow of materials involves
the recycling of carbon, nitrogen, phosphorus and other
elements between living organisms and the air, water, and
soil. These biologically mediated energy and materials
flows contribute to many ecological or life support services
that benefit human welfare such as greenhouse gas
regulation, water treatment, erosion control, soil quality
control, and plant growth.
Ecosystem services can also include cultural benefits,
such as religious, aesthetic, recreational, or inspirational
values that humans derive from ecosystems. Determining
whether biodiversity per se is important to ecosystem
functioning has been difficult, partly because many of the
factors such as habitat conversion that reduce local

biodiversity also directly affect many ecological processes,
masking the more subtle impacts of species loss on
functioning. Recent studies, however, have begun to shed
considerable light on the issue. Studies have shown" that
ecosystems are indeed sensitive to changes in the numbers
and kinds of species found in their communities.
ECOSYSTEM FUNcnONING
Ecosystem functioning reflects the collective life activities

of plants, animals, and microbes and the effects these
activities "feeding, growing, moving, excreting waste, etc."
have on the physical and chemical conditions of their
environment. Functioning' means 'showing activity' and
does not imply that organisms perform purposeful roles
in ecosystem-level processes. A' functioning ecosystem is
one that exhibits biological and chemical activities
characteristic for its type. A functioning forest ecosystem,
for example, exhibits rates of plant production, carbon
storage, and nutrient cycling that are characteristic of most
forests.
If the forest is converted to an agroecosystem, its
functioning changes. Ecologists abstract the essential
features of an ecosystem into two compartments, the biotic
and the abiotic. The biotic compartment consists of the
community of species, which can be divided functionally
into plant producers, the consumers that feed on producers
and on each other, and the decomposers Figure (1). The
abiotic compartment consists of organic and inorganic
nutrient pools. Energy and materials move between these
two compartments, as well as into and out of the system.
Ecosystem processes are quantified by measuring
rates of these movements (e.g., plant production,
decomposition, nutrient leaching or other measures of
material production, transport or loss). Ecosystem
functioning, in turn, is quantified by measuring the

magnitudes and dynamics of ecosystem processes.
Ecosystem functioning results from interactions among and
within different levels of the biota, which ecologists
describe as a 'nested' hierarchy. For example, green plant
production on land is the end product of interactions of
individual plants nested within populations; interactions
among populations nested within a single species;
interactions among a variety of species nested within a
group of functionally similar species; and so on up to the
level of interactions between different types of ecosystems
nested within landscapes.
ECOSYSTEM
_&
.~
-- "-&
. ... ....... '.

~
-
Figure 1. Basic ecosystem functioning
Figure 1 discribes, producers acquire energy through

photosynthesis and take up inorganic nutrients to produce
living biomass, forming the food base for consumer species
such as herbivores and their predators. Mortality leads to
accumulation of organic nutrients which are transformed
by decomposers into living biomass, forming the food base
for consumers. Decomposers and consumers contribute to
formation of inorganic nutrients by mineralisation,
completing the cycling of nutrients between organic and
inorganic forms. Energy flows (wavy, dashed lines) begin
with acquisition by producers and end in loss due to the

respiration activities of all organisms.
Although every organism contributes to ecosystem
processes, the nature and magnitude of individual
contributions vary considerably. Research in biodiversity
places much emphasis on the uniqueness of individual
species and their singular contributions to ecosystem
services. Yet most ecosystem processes are driven by the
combined biological activities of many species, and it is
often not possible to determine the relative contributions
of individual species to ecosystem processes. Species
within groups such as grazing mammals, large predators,
perennial grasses, or nitrogen-fixing microbes may
therefore be functionally similar despite their uniqueness
in genes, life history, and other traits.
Groups of species that perform similar roles in an
ecosystem process are known as functional types or
functional groups. Species may also be divided into
functional types based on what they consume or by trophic
status (e.g., their place in the food web as producers,
decomposers, predators). Within trophic groups, species
may be further divided according to life history, climatic
or nutrient needs, physiology or other biological traits.
Researchers may place a species into several different
functional categories depending on the ecosystem process
they are studying. Because species can vary dramatically
in their contributions to ecosystem functioning, the specific
composition or identity of species in a community is
important.
The fact that some species matter more than others
becomes especially clear in the case of 'keystone species'
or 'ecosystem engineers' or organisms with high
'community importance values.' These terms differ in
usage, but all refer to species whose loss has a
disproportionate impact on the community when
compared to the loss of other species. For example, a
species of nitrogen-fixing tree, Myrica faya, introduced to

the Hawaiian islands has had large-scale effects on
nitrogen cycling, greatly increasing the amount of this
essential plant nutrient in soils where the tree invades.
The nitrogenfixing lupine Lupinus arboreus also enriches
soils and, as a consequence, encourages invasions of weedy
grasses.
Among animals, moose through their dietary
preferences greatly reduce soil nitrogen levels and also
influence the succession of trees in the forest. Beavers, too,
through their feeding and dam-building not only alter soil
fertility and forest succession but increase the diversity of
ecosystems in a landscape. Even termites play critical roles
in soil fertility and other ecological processes in many arid
grasslands. On the other hand, there are some examples
where additions or losses of particular species have had
little effect on ecosystem processes.
Since Darwin, prominent biologists have hypothesised
about the relationship between biodiversity and ecosystem
functioning. More recently, concerns about increasing loss
of biodiversity and questions about resulting degradation
of ecosystem services have stimulated unprecedented
observational, theoretical, and experimental studies.
It might seem that observational studies comparing
one ecosystem. type with another, or comparing similar
ecosystems at different locations, could provide ready
answers to questions about the impacts of species richness
on ecosystem processes. But these studies have invariably
proven problematic. For example, an ecosystem such as a
tropical forest or a coastal wetland may vary from one
site to another not only in species number and
composition, but also in physical and chemical conditions
such as soil type, slope, rainfall, or nutrient levels.
Comparing different ecosystems is likely to yield an
unclear result because the response to variations in
biodiversity cannot easily be distinguished from responses

caused by variations in environmental and other factors.
It is possible, though difficult, to control statistically· for
such potentially confounding factors.
Experimental studies, if well-designed, can minimise
the confounding factors that plague observational studies.
Experiments can provide insights not only into the
relationships between biodiversity and ecosystem
functioning but also into the possible mechanisms behind
the relationships. Studies to date have ranged from large
outdoor experiments and trials in large controlled
environment facilities to modest-sized pot experiments and
tests in small laboratory microcosms.
Biodiversity and Levels of Ecosystem Functioning
Results from many recent experimental studies conducted
in North America and Europe demonstrate that ecosystem
productivity increases with species richness. These studies
range from large outdoor experiments to controlled
laboratory experiments conducted in growth chambers,
greenhouses, or small containers. Outdoor experiments
such as those conducted in grasslands on nutrient-poor
serpentine soils at Stanford, California and on prairie
grasslands at Cedar Creek Natural History Area,
Minnesota, work with plant communities similar to those
found in nature, but researchers vary the number of plant
species from one experimental plot to another. This
approach is also used in the BIODEPTH experiments, in
which seven European countries have established outdoor
plots that range in plant diversity from low species
numbers to the average numbers typically found at each
site.
More precise experiments using growth chambers
have been conducted by researchers at Imperial College
of London, Silwood Park, England and Centre d'Ecologie
Fonctionnelle et Evolutive, Montpellier, France. More
recent laboratory experiments in Europe and North
America have begun to examine the impact of other

components of biodiversity, such as the diversity of soil
microorganisms, on plant production and the role of
bacteria, predators, and herbivores in freshwater microbial
communities.
Studies show that ecosystem functioning is decreased
as the number of species in a community decreases.
Declines in functioning can be particularly acute when the
number of species is low, such as in most managed
ecosystems including croplands or timber plantations. In
addition, recent experimental studies in grasslands indicate
that the effects of biodiversity on production can depend
on both the number of functional groups present and the
identity of the plant species (Le., on community
composition). Other studies have shown that loss of
functional groups from a food web, or reductions in the
number of species per trophic group (producers,
consumers, decomposers) can also cause declines in
ecosystem functioning. Finally, another study has shown
that some species of plants may be more or less productive
or show no response at all to changes in the diversity of
their communities, even though total community
productivity is, on average, lower at lower diversity.
Studies on plants have been particularly revealing and
support results from recent theoretical models which
predict that decreasing plant diversity leads to lower plant
productivity. These models predict that diversity and
composition are approximately equal in importance as
determinants of ecosystem functioning. Two possible
mechanisms have been identified to explain why levels of
ecosystem functioning increase with increasing
biodiversity.
First is the 'sampling effect'; When the pool of species
available in a region contains individual species that vary
in productivity and other contributions to ecosystem
functioning, then species-rich ecosystems have a higher

probability of containing species with high levels of
functioning. Second is the 'complementarity effect': This
occurs when increasing diversity results in increasing
numbers of species that are complementary rather than
competitive in their use of resources, exploiting different
niches, such as rooting depths, and allowing more effective
use of available resources.
STABILITY, PREDICTABILITY AND RELIABILITY
Few experimental studies of the impact of biodiversity on

stability have been attempted, largely because stability is
a long term attribute of a system and testing for it requires
either long-running experiments or experiments with
short-lived organisms. In the one available long-term
ecological field study, however, reductions in plant species
richness also lowered the resistance of grassland
production to drought. Predictably "lower year-to-year
fluctuations in community productivity" was also
significantly lower at lower diversity. In addition, studies
of microbial communities in small experimental chambers
have also shown that fluctuations in ecosystem functions
such as productivity can be greater when species richness
is reduced. Thus, the loss of diversity causes a loss of
ecosystem stability.
Several mechanisms could account for these results.
One mechanism comes from the ability of competing
species to replace or compensate for one another and thus
minimise, at higher diversity, the ups and downs in
functioning. Another mechanism is the 'portfolio effect,' a
theory which suggests that cumulative properties such as
ecosystem functioning show less severe fluctuations in
systems with many species, much the way investment
portfolios of varied stocks have lower long term variance
than portfolios of one or a few kinds of stocks.
5
Ecosystem Functioning .at
Local and Regional Scales
The relationship between biodiversity and ecosystem

functioning has emerged as a central issue in ecological
and environmental sciences. Increasing domination of
ecosystems by humans is steadily transforming them into
depauperate systems. Because ecosystems collectively
determine the biogeochemical processes that regulate the
Earth system, the potential ecological consequences of
biodiversity loss have aroused considerable interest.
In recent years, several investigators have
demonstrated that species richness within local
communities can influence ecosystem functioning, such as
productivity and stability. A wide variety of experiments
conducted in disparate ecosystems have shown that greater
species diversity positively affects ecosystem functioning.
These studies have been used to make an important
argument for the conservation of species.
However, other studies have found neutral or
sometimes negative results. Thus, although there is often
an overall effect of diversity on ecosystem functioning, the
shape of the relationship is not always predictable and it
is unclear why this variation among studies occurs. Here
use theoretical models as a tool for understanding the

variation in studies of diversity and ecosystem functioning.
Two recent models have provided a framework for
understanding the relationship between biodiversity and
ecosystem functioning; however, they predict opposing
patterns.
The first model is based on local niche
complementarity and assumes that each species possesses
certain traits that allow species to utilize available
resources differently. As species diversity increases, each
species utilizes a different component of the resource base.
Thus, diversity positively contributes to ecosystem
functioning in the local community. However, as species
diversity continues to increase, the probability that species
will overlap in their resource use increases, thus creating
a decelerating relationship.
The second model is based on regional processes and
source-sink dynamics and suggests that, when immigration
from a region is high, local diversity increases, but
ecosystem functioning decreases. This can be due to
interspecific competition between the superior competitor
in the patch and the inferior competitors that are
maintained locally as sink populations when immigration
is high. The decline in ecosystem functioning may occur
when interspecific competition between the competitors
is greater than intraspecific competition within a species.
Here connect these two theories to provide a synthetic
view of the diversity-functioning relationship when
community assembly controls local species diversity.
Further, ecosystem functioning can be viewed from the
regional scale when environmental heterogeneity allows
species to exist in different patch types and thus, coexist
regionally. This regional complementarity among species
may then cause ecosystem functioning to increase as
regional species diversity increases. Link these two models
by assuming that when species diversity is low, the
Ecosystem Functioning at Local and Regional Scales 63
addition of new species complement one another (local

niche complementarity).
Alternatively, when species diversity exceeds the
number of local limiting factors, competition for these
factors may cause a decline in local ecosystem functioning
(Fig. la). Below provide a simple verbal model describing
how this relationship might come about based on
community assembly; but, note that the shape of this
relationship does not necessarily rely on these specific
assumptions.
First, there are several local patches within a regional
landscape. Within each local patch, there are several
functional roles that species could fill. For example, these
functional roles could be based on resource utilization (i.e.
different nutrients). For each functional role there is a
single species that can maximise functioning and
outcompete all other species.
This superior competitor may maximise functioning
by most efficiently transferring their resources to biomass
(net primary productivity) or by providing resources to
other biota through their tissues (nutrient cycling). Under
the assumption of community assembly, as species
diversity increases each species falls into a particular
functional role in the community. Thus, ecosystem
functioning will increase until the number of species
equals the number of functional roles in the local
community (ascending part of Fig. la).
A single species can outcompete all others for a
particular functional role, the only way for species
diversity to exceed the available functional roles in a local
community is through source-sink dynamics. In order to
explore the consequences of species diversity at these
higher levels, there is environmental heterogeneity among
patches, such that local patches vary in some other
environmental factor (Le. pH or temperature) in addition
to resource availability. Therefore, a species that is a

superior competitor in one patch type may be an inferior
competitor in another. That is, a species cannot exist in
sinks throughout the entire region or it would become
regionally extinct.
(a) Local
!lc
R
!
:I
lL
E
J ........DlMiIII;r
... tCjl
I.-l
he'"
~ .. C 2 stdIII
..
(b) Regio...
ICII
Ic
If
I
w
..... ~
........... ~
Figure 1. Species diversity and ecosystem functioning in local and

regional spatial scales.
Based on source-sink dynamics, both superior and inferior

competitors may be present in the local patch. As a result,
if immigration is high and if the inferior competitor can
detract from the overall functioning of the superior
competitor (through source-sink dynamics), ecosystem
functioning may decline (descending part of Fig. la).
By assuming that community assembly controls local
species diversity, here propose a hump-shaped relationship
between local species diversity and local ecosystem
functioning when immigration rates are high among
patches in the region (Fig. la). The peak of this hump is
dependent on the number of available functional roles in
a local community and may be predicted a priori for the
ecosystem of interest.
Also, the descent of the curve may depend on the
rate of immigration and strength of competition from the
inferior competitor. If immigration is low or the inferior
competitor is not able to detract from the superior
competitor's functioning, then this decline may be weak
or nonexistent. So far, ecosystem functioning has primarily
been considered on the local scale. However, the average
functioning of an entire region may not necessarily be
additive across all local patches.
In this verbal model, local ecosystem functioning is
reduced when species diversity· increases through source-
sink dynamics, but this effect might not be seen when
consider environmentally heterogeneous patches in a
region. On the regional scale, different species are superior
competitors in different patch types. When pool across all
patches in the region, all species may now coexist and
complement one another regionally. This can be considered
niche complementarity at the regional spatial scale.
Thus, even though local functioning is not maximised
at high levels of local diversity (within patches), as regional
complementarity (among patches) and regional species
diversity increases, there may be a linear increase in

regional ecosystem functioning (Fig. Ib). As a result,
ecosystem functioning can be highest when all species are
maintained in the region, whereas, within any local patch,
ecosystem functioning might actually be lower when all
species are present.
The effects of biodiversity on ecosystem functioning,
such as primary productivity, here ignored a possible
feedback of the effect of primary productivity on
biodiversity. The relationship between primary
productivity (independent variable) and diversity
(dependant variable) shows a scale-dependent pattern that
is superficially similar to the one predict here. However,
there is a problem with cause and effect in the relationship
between biodiversity and productivity and the relationship
between productivity and biodiversity.
Environmental variables which influence primary
productivity (e.g. nutrients) other than species diversity
are held constant. The productivity gradient was driven
primarily by variation in environmental factors (Le.
nutrients), and we suspect that the feedback of diversity
on productivity was probably much weaker than the
influence of environmental variation on productivity.
Nevertheless, the complexity of cause and effect in the
relationship between biodiversity and productivity
illustrates an important issue in need of further
exploration.
As environmental conditions vary through time,
higher species diversity may be important in maintaining
ecosystem functioning. Indeed, empirical evidence already
supports the idea that greater species diversity may have
different contributions to ecosystem functioning as
environmental conditions shift through time. Thus, while
empirical evidence is accruing to support the notion that
increased species diversity is important through time, her
suggest that spatial scale is also an important component
to consider empirically when investigating the relationship

between diversity and ecosystem functioning.
The ideas presented here suggest that species
diversity may become increasingly important to ecosystem
functioning at higher spatial scales. As environmental
conditions vary across space, a variety of species with
different environmental tolerances would be required to
maintain ecosystem functioning across the landscape (Le.
regional complementarity).
A variety of results between species diversity and
ecosystem functioning. Through synthesising two previous
models and increasing the spatial scale under
consideration, these ideas may be used to describe the
combination of processes (local and regional) which may
influence the relationship found in empirical
investigations.
CHANGES IN ECOSYSTEM FUNcrIONING
Human impacts on the environment from local to global

scales cause not only a general decline in diversity, but
also predictable functional shifts as sets of species with
particular traits are replaced by other sets with different
traits. The relative importance of functional substitutions
and declining species richness as determinants of changes
in ecosystem functioning. The extent to which functional
substitutions alter ecosystem properties such as
productivity, decomposition rates, nutrient cycling, and
resistance and resilience to perturbations.
On the other hand, a new wave of experimental
studies has manipulated species richness by using
synthesised model ecosystems in both terrestrial and
aquatic environments. Both approaches suggest that a large
pool of species is required to sustain the assembly and
functioning of ecosystems in landscapes subject to
increasingly intensive land use. It is not yet clear, however,
whether this dependence on diversity arises from the need

for recruitment of a few key species from within the
regional species pool or is due to the need for a rich
assortment of complementary species within particular
ecosystems.
Experimentally Altered Diversity

Experimentally manipulated diversity did so across several
trophic levels, mainly on effects of plant taxonomic
diversity and plant functional-group diversity on primary
production in grassland ecosystems. Because plants, as
primary producers, represent the basal component of most
ecosystems, they represented the logical place to begin
detailed studies. Several, although not all, experiments
using randomly assembled communities found that
primary production exhibits a positive relationship with
plant species and functional-group diversity (Fig. 2).
These results attracted a great deal of interest, not
only because they were novel, but also because they
seemed counter to patterns often observed in nature,
where the most productive ecosystems are typically
characterised by low species diversity. The controversy
over the interpretation of these results started with the
realisation that they can be generated by different
mechanisms.
The mechanisms discussed so far may be grouped
into two main classes. First are local deterministic
processes, such as niche differentiation and facilitation,
which increase the performance of communities above that
expected from the performance of individual species
grown alone. The term "complementarity" for
convenience's sake. Second are local and regional
stochastic processes involved in community assembly,
which are mimicked in experiments by random sampling
from a species pool.
,. A
•
seD • •
•
1&00 •
•• •• ••I •• •
•
I-
I-
1000 •
2 4 41 8
IN....... til ....... (lflii .....
12 24
c
I 3DD fp
200 IlP
NIL ELij EP tElP
fELPH
'10 IE
........ .,....... ,....

B
D I 2 3 I
"500 fB
• ••
. 2:'0· •
..-E .•
•
" t
~
•
I
..
-.
C
III
1000
~
]I
750
I 500
250
•
a 1 2 4 8 16 32
Number of tpecilt,log lealt)
Figure 2. Responses of total (A) or ai1f.)vegrolllld (B and C) plant biomass
(in grams per meter squared) to experimental //Iall/pulations of plant
species richness (A and B) or functional-group ric/mess (0 in grasslallds
in Minnesota (A), across Europe (B), and in California (C). Points in (A)
and (B) are data for individual plots.
Random sampling coupled with local dominance of highly

productive species can also lead to increased average
primary production with increasing diversity, because
plots that include many species have a higher probability
of containing highly productive species.
Two issues are involved in this controversy: Are
stochastic community assembly processes relevant? And
what is the relative importance of the two classes of
mechanisms? There are diverging views on the relevance
of the sampling component of biodiversity effects. As
sampling processes were not an explicit part of the initial
hypotheses, they have been viewed by some as "hidden
treatments" whereas others have viewed them as the
I
simplest possible mechanism linking diversity and

ecosystem functioning.
If dominant species control ecosystem processes and
mostly rare species go extinct, the vagaries of community
assembly or disassembly may have little relevance. But
environmental changes and landscape fragmentation could
prevent recruitment of appropriate dominants. Also,
climate change could lead to gradual losses of species as
abiotic conditions begin to exceed species' tolerance limits.
Such losses could be random with respect to species effects
on any given ecosystem process, leading to patterns of
process response to changes in diversity similar to those
observed in randomly assembled communities.
It should be emphasised that recent experiments were
not intended to reproduce any particular sequence of
species loss; they reflect potential patterns, unaffected by
correlations between diversity loss and compositional
changes, rather than actual predictions of functional
consequences of biodiversity loss under specific global
change scenarios. Assessing the relative importance of
complementarity and sampling effects has been done so
far indirectly, by using comparisons between the
performances of mixtures and monocultures. Furthermore,
it is becoming clear that complementarity and sampling

are not mutually exclusive mechanisms as previously
thought.
Communities with more species have a greater
probability of containing a higher phenotypic trait
diversity. Dominance that is brought about by ecological
"selection" of species with particular traits and
complementarity among species with different traits are
two ways by which this phenotypic diversity maps onto
ecosystem processes. These two mechanisms, however,
may be viewed as two poles on a continuum from pure
dominance to pure complementarity.
Intermediate scenarios involve complementarity
among particular sets of species or functional groups, or
dominance of particular subsets of complementary species
(Fig. 3). Any bias in community assembly that leads to
correlations between diversity and community composition
may involve both dominance and complementarity.
Rigorously testing the hypothesis that there is a minimum
subset of complementary species that is sufficient to
explain diversity effects will often be difficult because it
would ideally require testing, with replication, the
performance of all species combinations at all diversity
levels.
Re-analysis of data from previously published
experiments suggests significant effects of species richness
on plant biomass even after controlling for the strong
effects of certain species, such as legumes. Although these
new results presented at the Paris conference will need to
be critically evaluated, they suggest that complementarity
does occur among at least several species belonging to
different functional groups in these experiments. No clear
evidence, however, has been provided so far for
complementarity among a large number of species,
although complementarity among rare species would be
difficult to detect.
• OAO.
~pooI V .*~
.*.6,- D
&rftPIrv ~
lManbilldoonnllnlJ •
A
O. 0. ,..,.,.........
U ,.".,.,...,
~7~~
Figure 3. Hypothesized mechanisms involved in biodiversity experiments

using synthetic communities.
Cannot reject the hypothesis that a few dominant species

suffice to provide the functional diversity that is necessary
to explain the level of primary production observed in
grassland ecosystems at the small spatial and temporal
scales considered in recent experiments.
Greater attention should be paid to what individual
species do in these experiments. One option for assembly
experiments is to have carefully balanced designs to allow
contrasts between plots with and without particular
species or subsets of species. Another option is to include
manipulations of evenness within a level of species
richness, which could provide an alternative to methods
based on comparisons with monocultures, to separate

dominance and complementarity effects. There is also a
great need for other approaches based on "natural"
ecosystems, such as removal experiments and comparative
approaches that control for variation of factors other than
diversity.
•
•
0 t5 10
0 5 10
1•
1_ ••
I.
f•
•84QC1
IDa
0
0 I
..... ......
10
",
111 2D
Figure 4. Observed decreases in variability of ecosystem processes as

•
spedes richness increases.
In figure 4, Interpretation of these patterns, however, is

complicated by the correlation of additional factors with
species richness. (A) Adjusted coefficient of temporal
variation of annual total plant biomass (in grams per meter
squared) over 11 years for plots differing in number of
species in experimental and natural grasslands in
Minnesota. The correlation of variations in soil nitrogen
with species richness in these plots precludes the
interpretation of increased stability as a pure diversity

effect, although the diversity effect remained significant
even after controlling for potentially confounding
variables. (B) Standard deviation of CO2 flux (in microliters
per 18 hours) from microbial microcosms. In these data,
temporal variability in response to diversity is confounded
with between-replicate variability.
Biodiversity as Insurance
Even when high diversity is not critical for maintaining
ecosystem processes under constant or benign
environmental conditions, it might nevertheless be
important for maintaining them under changing
conditions. The insurance hypothesis and related
hypotheses propose that biodiversity provides an
"insurance" or a buffer, against environmental fluctuations,
because different species respond differently to these
fluctuations, leading to more predictable aggregate
community or ecosystem properties.
In this hypothesis, species that are functionally
redundant for an ecosystem process at a given time are
no longer redundant through time. In a way, this is the
old stability-versus complexity debate resurfacing in a new
form. Several problems, however, have confused this
historical controversy:
The general concept of "stability" actually covers a
wide array of different properties;
the relationship between these properties and diversity
may change across ecological levels of organisation
such that large variability at the population level may
not imply large variability of ecosystem processes; and
stability has been approached mainly within a
deterministic, equilibrium theoretical framework.
Theoretical work has attempted to remove these obstacles

and has provided support for the insurance hypothesis.
As diversity increases, the variability of individual
populations may increase as a result of the destabilising
influence of strong species interactions internal to the
system, but the variability of aggregate ecosystem
properties often decreases because of the stabilising
influence of asynchronous species responses to intrinsic
or extrinsic environmental fluctuations. What remains
unclear, however, is whether this stabilising effect saturates
at low or high diversity, which depends on model
condi tions.
Whereas experimental work has played a leading role
regarding short-term effects of biodiversity on ecosystem
functioning, theory has been prominent in the
diversitystability debate, both historically and recently. A
number of empirical and experimental studies have shown
decreased variability of ecosystem processes as diversity
increases (Fig. 4). However, have been based either on
diversity gradients established naturally or after other
treatments, or on microcosm experiments in which
variability among replicates was also considered, which
does not fully preclude alternative interpretations.
Experiments in which both diversity and
environmental fluctuations are controlled are now needed
to perform rigorous tests of the insurance hypothesis.
Theory too should evolve to provide better guidance for
experiments. Most of the classical equilibrium approaches
may be inadequate to understand stability properties such
as resilience and resistance at the ecosystem level. New
approaches should be developed that take into account
the dynamics of diversity and the potential for adaptation
through phenotypic plasticity, evolutionary changes, and
species replacement.
The relationship between productivity and diversity
has long been studied from an angle different from that
in recent experimental studies. It is often, although not

always, described by a hump-shaped curve, in which
diversity is considered a function of productivity (Fig. SA).
These curves have typically been obtained by using
correlations across different sites or nutrient addition
treatments. Some comparative approaches have also
suggested negative relationships between plant species
evenness and rates of various ecosystem processes. The
differences between these large-scale, observational
approaches and the small-scale, experimental approaches
have also generated debate.
The two approaches examine different causal
relationships under different sets of conditions. The
classical approach attempts to identify the causes of spatial
variation in diversity across environmental gradients.
Variation in diversity is often correlated with productivity,
but also with many other factors that influence
productivity, such as soil fertility, climate, disturbance
regime, or herbivory. The recent experimental approach
examines whether diversity alone has a local effect on
productivity within each site, when all these other factors
are held constant.
The two approaches can be reconciled by considering
that spatial patterns reveal correlations between diversity
and productivity driven by environmental factors, whereas
small-scale experiments reveal the effects of species
properties and diversity on productivity that are detected
after the effects of other environmental factors have been
removed (Fig. SB).
Whether biodiversity loss will affect large-scale
patterns of productivity hinges on the shape and steepness
of the local dependence of productiv1ty---on diversity.
Generally speaking, the relative effects of individual
species and species richness may be expected to be greatest
at small-to-intermediate spatial scales, but these biological
factors should be less important as predictors of ecosystem
processes at regional scales, where environmental

heterogeneity is greater.
~
A ..:!... B
:>. f8VOtab1e
6011 and climate
.i
0
~
e
Q.
soli and citmate

Productlvity Diversity
Sotl and climate effects
Figure 5. Hypothesized relationships between (A) diversityproductivity

patterns driven by environmental conditions across sites, and (B) the local
effect of species diversity on productivity.
Whereas diversity was manipulated as the independent

variable in recent experiments, at large scales species
diversity itself is a dynamical variable and adjusts to
changes in environmental conditions. Abiotic factors then
tend to be the main drivers of variations in ecosystem
processes across environmental gradients. Diversity loss
at regional scales and dispersal limitations due to
landscape fragmentation, however, will very likely feed
back and reduce the pool of potential colonists at local
scales and hence the potential for local compositional
adjustments to environmental changes.
Species-area relations imply that the long-term
maintenance of a given level of diversity at local scales
requires a much higher diversity at regional scales. One
of the most potent effects of declining diversity could be
the decline in the rate at which appropriate potential
dominants are recruited during ecosystem assembly.
To understand and predict changes in biodiversity

and ecosystem processes at large scales, therefore, we need
to move beyond unidirectional causality approaches in
which diversity is either cause or effect, and address
feedbacks among biodiversity changes, ecosystem
functioning, and environmental factors. Relationships
between local, landscape, and regional scales also require
particular attention.
Generalising Ecosystems
Most of the recent experiments that found significant
effects of species diversity have concerned effects of plant
diversity on primary production and nutrient retention in
temperate grasslands, both of which are under direct plant
control. These and other experiments have often failed to
detect significant effects on below-ground decomposition
processes, perhaps because these processes are under
microbial control.
Primary production in grasslands can be generalised
to other processes and ecosystems. Plants can affect soil
processes either directly, by stimulating or inhibiting
decomposition rates, or indirectly, through increased
primary production, by enhancing decomposition fluxes.
Although some experiments found positive effects of plant
diversity on soil microbial processes, experiments using
litter addition, cotton strips, or litter mixing often showed
variable and weak effects of plant diversity on
decomposition rates.
Current evidence suggests that properties of
individual plant species are more important than plant
diversity in governing soil process rates. Theoretical work
predicting that plant chemical quality diversity should
decrease or not affect long-term nutrient recycling
efficiency and productivity. In contrast, increased primary
production generated by higher plant diversity is expected
to stimulate secondary productivity. More generally,

diversity changes at one trophic level may lead to a variety
of potential responses for processes at higher trophic
levels.
Species diversity in consumer trophic levels can also
have complex effects on production at these and lower
levels. Complementarity and sampling effects should tend
to improve resource exploitation just as in plants. This
should lead to higher secondary productivity if bottomup
control prevails, as in plant-decomposer interactions.
Enhanced resource exploitation, however, can lead to
overexploitation, and thus decreased productivity, if top-
down control is important, as might be the case with
herbivores and predators. There have been few
experiments to test these hypotheses.
Recent microcosm experiments found significant
effects of bacterial diversity on bacterial and algal
biomasses and of diversity of leaf-eating insects on
decomposition rates, but others suggested that individual
species and functional composition were the most
important factors. The functional role of diversity in
mutualistic interactions has also been poorly studied
despite their importance in the maintenance of ecosystem
processes, as shown by one experiment on mycorrhisal
fungal diversity.
Although there is a clear case for incorporating
multiple trophic levels into studies of biodiversity-
ecosystem functioning relationships, logistical constraints,
such as the high mobility of herbivores and carnivores and
the difficulty of taxonomic identification of decomposers,
partly explains why so few studies have done so as yet.
Of particular importance are the vast areas of biodiversity
that involve small organisms such as viruses, bacteria,
archae a, protists, and microarthropods, which drive the
bulk of ecosystem processes.
For example, chemical transformations in the nitrogen

cycle are predominantly driven by prokaryotic organisms,
as is decomposition of organic matter. Modem molecular
tools are beginning to make possible the integration of
microbial diversity into studies of ecosystem processes.
There is also a need to extend these knowledge to
ecosystem types other than temperate grasslands, such as
forest, freshwater, and marine ecosystems.
Top-down control is often thought to be more
common in freshwater than in terrestrial ecosystems;
significant differences might then be expected between
ecosystem types just as between trophic levels. Generally
speaking, differences in coexistence mechanisms may lead
to differences in biodiversity effects on ecosystem
functioning. For example, in disturbance-driven systems,
the colonisation ability and growth rate of individual
species, rather than niche complementarity, might drive
ecosystem processes.
6
Freshvvater Ecosystenn
Fresh water is vital to human life and economic wellbeing,

and societies draw heavily on rivers, lakes, wetlands, and
underground aquifers to supply water for drinking,
irrigating crops, and running industrial processes. The
benefits of these extractive uses of fresh water have
traditionally overshadowed the equally vital benefits of
water that remains in stream to sustain healthy aquatic
ecosystems. There is growing recognition that functionally
intact and biologically complex freshwater ecosystems
provide many economically valuable commodities and
services to society. The services supplied by freshwater
ecosystems include flood control, transportation, recreation,
purification of human and industrial wastes, habitat for
plants and animals, and production of fish and other foods
and marketable goods.
These human benefits are what ecologists call
ecological services, defined as lithe conditions and
processes through which natural ecosystems, and the
species that make them up, sustain and fulfill human life."
Over the long term, healthy freshwater ecosystems are
likely to retain the adaptive capacity to sustain production
of these ecological services in the face of future
environmental disruptions such as climate change.
Ecological services are costly and often impossible to
replace when aquatic ecosystems are degraded. Yet today,

aquatic ecosystems are being severely altered or destroyed
at a greater rate than at any other time in human history,
and far faster than they are being restored.
Debates involving sustainable allocation of water
resources should recognise that maintenance of freshwater
ecosystem integrity is a legitimate goal that must be
considered among the competing demands for fresh water.
Coherent policies are required that more equitably allocate
water resources between natural ecosystem functioning
and society's extractive needs. Current water management
policies in the United States are clearly unable to meet
this goal. Literally dozens of different government entities
have a say in what wastes can be discharged into water
or how water is used and redistributed, and the goals of
one agency are often at crosspurposes with those of others.
Water are implemented in a management context that
focuses primarily on maintaining the lowest acceptable
water quality and minimal flows, and protecting single
species rather than aquatic communities. A fundamental
change in water management policies is needed, one that
embraces a much broader view of the dynamic nature of
freshwater resources and the short- and long-term benefits
they provide.
Hydrologists, engineers, and water managers, the
people who design and manage the nation's water
resource systems, are rarely taught about the ecological
consequences of management policies. Likewise, ecologists
are rarely trained to consider the critical role of water in
human society or to understand the institutions that
manage water. Economists, developers, and politicians
seldom project far enough into the future to fully account
for the potential ecological costs of short-term plans. Few
Americans are aware of the infrastructure that brings them
pure tap water or carries their wastes away, and fewer
Freshwater Ecosystem 83
still understand the ecological tradeoffs that are made to

allow these conveniences.
Although the requirements of healthy freshwater
ecosystems are often at odds with human activity, this
conflict need not be inevitable. The challenge is to
determine how society can extract the water resources it
needs while protecting the important natural complexity
and adaptive capacity of freshwater ecosystems. Current
scientific understanding makes it possible to outline here
in general terms the requirements for adequate quantity,
quality, and timing of water flow to sustain the functioning
of freshwater ecosystems. A critical next step will be
communication of these requirements to a broader
community.
Some studies that have addressed the overall
condition of freshwater resources have recognised that
water movement through the biosphere is highly
altered by human activities;
water is intensively used by humans;
poor water quality is pervasive;
and freshwater plant and animal species are at greater
risk of extinction from human activities compared
with all other species.
These and other analyses indicate that freshwater
ecosystems are under stress and at risk. Clearly, new
management approaches are needed.
REQUIREMENTS FOR FRESHWATER ECOSYSTEMS
Freshwater ecosystems differ greatly from oneanother

depending on type, location, and climate, but they
nevertheless share important features. For one, lakes,
wetlands, rivers, and their connected ground waters share
a common need for water within a certain range of

quantity and quality. In addition, because freshwater
ecosystems are dynamic, all require a range of natural
variation or disturbance to maintain viability or resilience.
Water flows that vary both season to season and year to
year, for example, are needed to support plant and animal
communities and maintain natural habitat dynamics that
support production and survival of species.
Variability in the timing and rate of water flow
strongly influence the sizes of native plant and animal
popUlations and their age structures, the presence of rare
or highly specialised species, the interactions of species
with each other and with their environments, and many
ecosystem processes.
Periodic and episodic water flow patterns also
influence water quality, physical habitat conditions and
connections, and energy sources in aquatic ecosystems.
Freshwater ecosystems, therefore, have evolved to the
rhythms of natural hydrologic variability. The structure
and functioning of freshwater ecosystems are also tightly
linked to the watersheds, or catchments, of which they
are a part.
Water flowing through the landscape on its way to
the sea moves in three dimensions, linking upstream to
downstream, stream channels to floodplains and riparian
wetlands, and surface waters to ground water. Materials
generated across the landscape ultimately make their way
into rivers, lakes, and other freshwater ecosystems. Thus
these systems are greatly influenced by what happens on
the land, including human activities. There are five
dynamic environmental factors that regulate much of the
structure and functioning of any aquatic ecosystem,
although their relative importance varies among aquatic
ecosystem types (Figure 1).
WATER QUALITY
Figure 1. Conceptual model of major forces that influence freshwater

ecosystems.
The interaction of these drivers in space and time defines

the dynamic nature of freshwater ecosystems:
The flow pattern defines the rates and pathways by
which rainfall and snowmelt enter and circulate within
river channels, lakes, wetlands, and connecting ground
waters, and also determines how long water is stored
in these ecosystems.
Sediment and organic matter inputs provide raw
materials that create physical habitat structure, refugia,
substrates, and spawning grounds and supply and
store nutrients that sustain aquatic plants and animals.
Temperature and light characteristics regulate the
metabolic processes, activity levels, and productivity
of aquatic organisms.
Chemical and nutrient conditions regulate pH, plant
and animal productivity, and water quality.
The plant and animal assemblage influences ecosystem

process rates and community structure.
In naturally functioning freshwater ecosystems, all five of
these factors vary within defined ranges throughout the
year, tracking seasonal changes in climate and day length.
Species have evolved and ecosystems have adjusted to
accommodate these annual cycles. They have also
developed strategies for surviving - and often requiring -
periodic hydrologic extremes caused by floods and
droughts that exceed the normal annual highs or lows in
flows, temperature, and other factors. Focusing on one
factor at a time will not yield a true picture of ecosystem
functioning . Evaluating freshwater ecosystem integrity
requires that all five of these dynamic environmental
factors be integrated and considered jointly.
Flow Patterns
An evaluation of the characteristics required for healthy
functioning can begin with a description of the natural or
historical flow patterns for streams, rivers, wetlands and
lakes. Certain aspects of these patterns are critical for
regulating biological productivity (that is, the growth of
algae or phytoplankton that form the base of aquatic food
webs) and biological diversity, particularly for rivers. These
aspects include base flow, annual or frequent floods, rare
and extreme flood events, seasonality of flows, and annual
variability. Such factors are also relevant for evaluating
the integrity of lakes and wetlands because flow patterns
and hydroperiod (that is, seasonal fluctuations in water
levels) influence water circulation patterns and renewal
rates, as well as types and abunaances of aquatic
vegetation such as reeds, grasses, and flowering plants.
Furthermore, the characteristic flow pattern of a lake,
. wetland, or stream critically influences algal productivity
and is an important factor to be considered when
determining acceptable levels of nutrient (nitrogen and

phosphorus) runoff from the surrounding landscape.
Human alterations of river flow have seldom taken into
account the ecological consequences. "Many rivers now
resemble elaborate plumbing works, with the timing and
amount of flow completely controlled, like water from a
faucet, so as to maximise the rivers' benefits for humans,"
wrote water policy expert Sandra L. Postel. "But while
modern engineering has been remarkably successful at
getting water to people and farms when and where they
need it, it has failed to protect the fundamental ecological
function of rivers and aquatic systems."
Rivers in the U.S. West are prime examples of how
human manipulation of water flows can lead to multiple
damages to riverbank and floodplain processes and
communities. Damming rivers and dampening natural
variations in flow rates by maintaining minimum flows
year round have contributed to widespread loss of native
fish species and regeneration failure of native cottonwood
trees, which used to support diverse riparian communities.
Sediment and Organic Matter Inputs

In river systems, the movement of sediments and influxes
of organic matter are important components of habitat
structure and dynamics. Natural organic matter inputs
include seasonal runoff and debris such as leaves and
decaying plant material from land-based communities in
the watershed. Especially in smaller rivers and streams,
the organic matter that arrives from the land is a
particularly important source of energy and nutrients, and
tree trunks and other woody materials that fall into the
water provide important substrates and habitats for aquatic
organisms.
Natural sediment movements are those that
accompany natural variations in water flows. In lakes and
wetlands, all but the finest inflowing sediment falls

permanently to the bottom, so that over time these systems
fill. The invertebrates, algae, bryophytes, vascular plants,
and bacteria that populate the bottoms of freshwater
systems are highly adapted to the specific sediment and
organic matter conditions of their environment, as are
many fish species, and do not persist if changes in the
type, size, or frequency of sediment inputs occur. The fate
of these organisms is critical to sustaining freshwater
ecosystems since they are responsible for much of the work
of water purification, decomposition, and nutrient cycling.
Humans have severely altered the natural rates of
sediment and organic matter supply to aquatic systems,
increasing some inputs while decreasing others. Poor
agricultural, logging, or construction practices, for example,
promote high rates of soil erosion. In many areas small
streams or wetlands have even been completely eliminated
through filling, paving, or re-routing into artificial
channels. The U.S. Environmental Protection Agency (EPA)
reports that in one quarter of all lakes with sub-standard
water quality, the cause of impairment is silt entering from
agricultural, urban, construction, and other non-point
sources. Dams alter sediment flows both for the reservoirs
behind them and the streams below, silting up the former
while starving the latter. Channel straightening,
overgrazing of river and stream banks, and clearing of
streamside vegetation reduce organic matter inputs and
often increase erosion.
Temperature and Light

The light and heat properties of a body of water are
influenced by climate and topography as well as by the
characteristics of the water body itself: its chemical
composition, suspended sediments, and algal productivity.
Water temperature directly regulates oxygen
concentrations, the metabolic rate of aquatic organisms,

and associated life processes such as growth, maturation,
and reproduction. The temperature cycle greatly influences
the fitness of aquatic plants and animals and, by extension,
where species are distributed in the system and how the
living community in a body of water varies from season
to season.
In lakes particularly, the absorption of solar energy
and its dissipation as heat are critical to development of
temperature gradients between the surface and deeper
water layers and also to water circulation patterns.
Circulation patterns and temperature gradients in turn
influence nutrient cycling, distribution of dissolved oxygen,
and both the distribution and behavior of organisms,
including game fishes. Water temperature can change
dramatically downstream of dams. In Utah's Green River,
mean monthly water temperatures ranged between 2
degrees Celsius (C) in winter and 18 degrees C in summer
before completion of the Flaming Gorge Dam in 1962.
After dam closure, the annual range of mean monthly
water temperatures below the dam was greatly narrowed,
to t.etween 4 C and 9 C. As a result, species richness
declined and 18 genera (that is, groups of related species)
of insects were lost; other species, notably freshwater
shrimp, came to dominate the ranks of invertebrate
animals. Aquatic insects have not recovered despite 20
years of partial temperature restoration achieved by
releasing water from warmer reservoir water layers. Water
temperature also dropped in the Colorado River after
closure of the Glen Canyon Dam in 1963, and there was a
dramatic increase in water clarity.
Water clarity now routinely allows visibility to greater
than 7 meters, whereas prior to dam closure, the water
column was opaque with suspended sediments. The
colder, clearer waters have allowed a nonnative trout
population to flourish, at the top of an unusual food web

more commonly found much further north.
Nutrient and Chemical Conditions

Natural nutrient and chemical conditions are those that
reflect local climate, bedrock, soil, vegetation type, and
topography. Natural water conditions can range from clear,
nutrient-poor rivers and lakes on crystalline bedrock to
much more chemically enriched and algae-producing
freshwaters in catchments with organic matter-rich soils
or limestone bedrock. This natural regional diversity in
watershed characteristics, in turn, sustains high
biodiversity. A condition known as cultural eutrophication
occurs when additional nutrients, chiefly nitrogen and
phosphorus, from human activities enter freshwater
ecosystems. The result is a decrease in biodiversity,
although productivity of certain algal species can increase
well beyond original levels.
Midwestern and Eastern lakes such as Lakes
Michigan, Huron, Erie, and from irrigation return flows.
Geological Survey, South Platte Assessment Programme
(NAWQA). Ontario demonstrate the consequences of
excess inputs of nutrients and toxic contaminants, as well
as non-native species introductions and over-fishing.
Onondaga Lake, New York, which was polluted with salt
brine effluent from a soda ash industry, likewise
responded with marked changes in the plankton and fish
communities, including invasions by non-native fish
species. Among U.S. lakes identified by the EPA as
impaired in 1996, excess nutrients contributed to more than
half of the water quality problems. More than half of
agricultural and urban streams sampled by the U. S.
Geological Survey were found to have pesticide
concentrations that exceed guidelines for the protection of
aquatic life.
Plant and Animal Assemblages

The community of species that lives in any given aquatic
ecosystem reflects both the pool of species available in the
region and the abilities of individual species to colonise
and survive in that water body. The suitability of a
freshwater ecosystem for any particular species is dictated
by the environmental conditions - that is, water flow,
sediment, temperature, light, and nutrient patterns - and
the presence of, and interactions among, other species in
the system. Thus, both the habitat and the biotic
community provide controls and feedbacks that maintain
a diverse range of species.
The high degree of natural variation in environmental
conditions in fresh waters across the United States
promotes high biological diversity. In fact, North American
freshwater habitats are virtually unrivaled in diversity of
fish, mussel, crayfish, amphibian, and aquatic reptile
species compared with anywhere else in the world. The
biota, in turn, are involved in shaping the critical ecological
processes of primary production, decomposition, and
nutrient cycling. Within a body of water, species often
perform overlapping, apparently redundant roles in these
processes, a factor that helps provide local ecosystems with
a greater capacity to adapt to future environmental
variation.
High apparent redundancy (that is, species richness
or biodiversity) affords a kind of insurance that ecological
functions will continue during environmental stress.
Critical to this is connectivity among water bodies, which
allows species to move to more suitable habitat as
environmental conditions change. Human activities that
alter freshwater environmental conditions can greatly
change both the identity of the species in the community
and the functioning of the ecosystem. Excessive stress or
simplification of natural complexity has the potential to
push functionally intact freshwater ecosystems beyond the

bounds of resilience or sustainability, threatening their
ability to provide important goods and services on both
short and long time scales.
Further, introduction of non-native species that can
thrive under the existing or altered range of environmental
variation can contribute to the extinction of native species,
severely modify food webs, and alter ecological processes
such as nutrient cycling. Exotic species are often successful
in modified systems, where they can be difficult to
eradicate.
TOOLS FOR RESTORATION
Despite widespread degradation of freshwater ecosystems,

management techniques are available that can restore these
systems to a more natural and sustainable state and
prevent continued loss of biodiversity, ecosystem
functioning, and ecological integrity. One technique, for
example, involves restoring some of the natural variations
in stream flow, based on the understanding that river
systems are naturally dynamic. New statistical approaches
to setting management targets for streamflow variability
over time have been applied to or proposed for several
rivers, including the Flathead River in Montana, the
Roanoke River in North Carolina, and the vast Colorado
River system in the West.
These variable streamflow techniques seek a balance
between water delivery needs for power generation or
irrigation, and instream ecological needs for flow
variability that displays a certain timing, frequency,
duration, and rate of change characteristic of the natural
system. Restoring this flow variability helps to reconnect
dynamic riparian and groundwater systems with surface
flows, enabling water to move more naturally through all
the spatial dimensions that are essential to fully functional

ecosystems.
Other restoration efforts target pollution, both from
point sources such as effluent from industrial or sewage
pipes and nonpoint sources such as fertiliser runoff from
urban lawns and rural croplands. Point sources of water
pollution are readily identified, and many have been
controlled, thanks in large part to the federal Clean Water
Act and Safe Drinking Water Act. Nonpoint sources of
nutrients and toxins now supply the majority of pollutants
to freshwater ecosystems. In some situations, best
management practices have succeeded in reducing runoff
of agricultural pollutants. These practices include erosion
control and moderate applications of fertilisers, pesticides
and herbicides. Best management practices require willing
farmers, however, and willingness is often a response
either to economic incentives or to stringent regulations.
To help in determining best management practices,
the EPA has recently published guidelines for establishing
acceptable nutrient runoff criteria for different regions of
the United States, recognising the inherent natural
variability in local and regional availability of nutrients.
The guidelines are based on Total Maximum Daily Load
(TMDL), a calculation of the maximum amount of a
pollutant that a water body can receive and still meet
water quality standards. To allow for natural variation,
water quality standards for a pollutant are established
within each ecoregion based on comparison with relatively
unpolluted waters or - if few or no unpolluted waters
remain in a region - on waters with the lowest pollution
levels (Figure 2).
Once a standard is set, management practices can be
enacted to reduce inputs of unwanted pollutants. Another
large source of nonpoint pollution is atmospheric
deposition of nitrogen and other contaminants that fall as
acid rain or dry pollutants. These could be reduced

through more stringent controls on emissions of sulfur,
nitrogen, metals, and organic toxins, and through
development and application of more efficient
transportation and energy production technologies.
Higher water quality ---lit.. lo_r water quality

Lower nutrients _ _ _... Higher nutJienta
Figure 2. Two different approaches for establishing a standard or "reference

condition value" for freshwaters
Reference condition values can be selected from waters

that are representative of the most pristine, or least
disturbed condition. If this goal is unrealistic, or if
undisturbed water bodies no longer exist in the region,
the reference condition value can be selected from among
the least disturbed and polluted water bodies found in
the region. Surveys of existing water quality from a broad
range of water bodies are necessary in order to establish
realistic water quality goals.
FRESHWATER ECOSYSTEM AND BALANCING HUMAN USE
The sustainability of aquatic ecosystems can best be

ensured by maintaining naturally variable flows, adequate
sediment and organic matter inputs, natural fluctuations
in heat and light, clean water, and a naturally diverse plant
and animal community. Failure to provide for these

essential requirements results in loss of species and
ecosystem services in wetlands, rivers, and lakes. Aquatic
ecosystems can be protected or restored by recognising the
following:
Aquatic ecosystems are not simply isolated bodies or
conduits but are tightly connected to terrestrial
environments. Further, aquatic ecosystems are
connected to each other and provide essential
migration routes for species.
Dynamic patterns of flow that are maintained within
the historical range of variation will promote the
integrity and sustainability of freshwater systems.
Aquatic ecosystems additionally require that sediment
loads, heat and light conditions, chemical and nutrient
inputs, and plant and animal populations fluctuate
within natural ranges, neither experiencing excessive
swings beyond their natural ranges nor being held at
constant, and therefore unnatural, levels.
Stating these requirements for maintaining aquatic
ecosystem integrity, of course, is not the same as
implementing them in the context of today's complicated
society. Policies for maintenance of water quality and f).ow
are primarily based on human health needs. The age of
ever-increasing exploitation is over, however. We must
begin to redefine water use based on the recognition that
supplies are finite and that healthy freshwater ecosystems
must be sustained or restored. Some of the
recommendations are following for how water is viewed
and managed:
Incorporate freshwater ecosystem needs, particularly
naturally variable flow patterns, into national and
regional water management policies along with
concerns about water quality and quantity. Because
most land and water use decisions are made locally,
we recommend empowering local groups and

communities to implement sustainable water policies.
A large and growing number of watershed groups is
already moving in this direction with the support and
guidance of state and federal agencies. Flexibility,
innovation, and incentives such as tax breaks,
development permits, conservation easements, and
pollution credits are effective tools for achieving
freshwater ecosystem sustainability goals.
Define water resources to include watersheds so that
fresh waters are viewed within a landscape or systems
context. Many of the problems facing freshwater
ecosystems come from outside the lakes, rivei's, or
wetlands themselves. Laws and agency regulations lag
in their recognition of this fact. One place to initiate a
change is through existing governmental permitting
processes.
Requests to the Federal Energy Regulatory
Commission for hydropower dam renewals, permit
requests to the Army Corps of Engineers for dredge
and fill operations under the Clean Water Act Section
404, and land use and effluent discharge permit
requests to state, county, and local entities present
ideal opportunities to integrate ecosystem needs with
traditional water uses. The EPA's TMDL Programme
is an effort to address both point and nonpoint
pollution from a watershed to a water body, although
the programme has not yet been fully implemented.
It should also be refined to consider how flow
variability influences the transport of pollutants.
Increase communication and education across
disciplines. Interdisciplinary training and experience,
particularly for engineers, hydrologists, economists,
and ecologists, can foster a new generation of water
managers and users who think about fresh waters as
systems with ecological purposes as well as water

supply functions.
Increase restoration efforts for wetlands, lakes, and
rivers using ecological principles as guidelines. While
some restoration has occurred, a greater effort is
required to restore the ecological integrity of the
nation's water resources. The goal of restoration
should be to reinstate natural variations in the
fundamental environmental factors identified above.
Yet many restoration projects, especially for wetlands,
have focused only on replanting vegetation while
ignoring underlying hydrologic, geomorphic,
biological, and chemical processes. Highly visible yet
ecologically incomplete restoration efforts such as
these wetland revegetation projects may even foster
complacency among the public. In any given
freshwater system, the extent of restoration and
protection that is eventually undertaken will be widely
debated because active management is inherently a
social process, although one ideally informed by
science. Restoration efforts can encompass a spectrum
of goals, from nearly full recovery of native species
and environmental conditions to the management of
dynamic, biologically diverse communities that do not
necessarily resemble native ecosystems.
Maintain and protl'ct remaining minimally impaired
freshwater ecosystems. Many restoration projects fail
to reestablish ecosystem functioning once major
processes have been disturbed. It is far wiser and
cheaper to conserve what we have. Moreover,
remaining functionally intact freshwater ecosystems
can provide a source of plant and animal colonists
for restoration projects elsewhere.
Bring the ecosystem concept home. Achieving
ecological sustainability requires that we come to
recognise the interdependence of people and the
environments of which they are a part. For fresh

waters, this will require broad recognition of the
sources and uses of water for societal and ecological
needs. It will also require taking a much longer view
of water processes. Water delivery systems and even
dams are developed with life spans and management
guidelines of decades to, at most, a century.
Freshwater ecosystems have evolved over aeons, and
their sustainability must be considered from a long-
term perspective. Governmental policies, mass media,
and a market-driven economy all focus on much
shorter-term benefits. Educational programmes at the
kindergarten through high school level, individual
initiatives to become informed, and efforts by local
watershed groups interested in protecting their natural
resources can provide good first steps toward
enduring stewardship. These steps must be matched
1::y state and national acknowledgment that
fundamental human needs for water can only be met
in the future through policies that preserve the
integrity and functioning of freshwater ecosystems
today.
CHALLENGES
The problems confronting freshwater ecosystems will be

intractable if they continue to be approached piecemeal.
Several government programmes, such as the EPA Clean
Lakes Programme, the Wetlands Restoration Act, and even
the Endangered Species Act, mandate actions to prevent
specific aspects of ecosystem degradation. But these
programmes are narrow in focus, effectively addressing
symptoms rather than root causes of aquatic ecosystem
decline.
Control of pollution is necessary, for instance, but
insufficient for maintaining a native species community if
adequate water flows are not available at the right time,

if the channel has been severely degraded, or if invasive
species have been allowed to take hold. The needs of
aquatic ecosystems and the needs of society for water
supplies must be addressed collectively if freshwater
ecological integrity is to be maintained or restored.
Politically, this requires that broad coalitions of water
users must work together towards a mutually acceptable
future. The best time to develop such coalitions is before
water is allocated and before ecological crises occur. In
many parts of the world, this opportunity was missed long
ago. The potential for full or partial restoration remains,
however. An ambitious example is taking place in south
Florida, where water control structures are being
physically removed and nutrient inputs curtailed in an
attempt to encourage a more natural system.
The ecological consequences that arise when
freshwater ecosystems are deprived of adequate water,
proper timing of flows, and suitable water quality often
become apparent to people only after the degradation
begins to interfere with societal uses of fresh water.
Nuisance algal blooms and loss of commercial or sport
fisheries are examples of failures in ecosystem processes
that were often years in the making. Some ecosystems
naturally experience wide swings in environmental and
ecological conditions from one year to the next that can
mask gradual changes in physical and chemical factors.
Most systems are inherently resilient to a particular
pattern of disturbance, and their plant and animal
communities will persist as long as conditions fluctuate
within a certain range. Once a threshold is reached,
however, these ecosystems may change rapidly to a new
stable state that is very difficult to reverse. The collapse
of a fishery and permanent cultural eutrophication from
nutrient inputs are two examples of conditions that, once
reached, make it difficult to restore the integrity of a

freshwater system. Detecting such trends before problems
become critical requires both monitoring the biological and
physical conditions in freshwater ecosystems and
understanding the natural ecological dynamics of these
systems.
7
Aquatic Systems
The consequences of species loss for the functioning of

ecoystems has been addressed through several major
research programmes in recent years, mostly in terrestrial
environments. The effects of species loss in aquatic
environments, in contrast, have received much less
attention, yet the nature of these ecosystems and their biota
differ markedly from those on land. This raises the
question as to whether, and how far, insights from
terrestrial research programmes can be extrapolated to
lakes and open oceans, rivers and freshwater wetlands,
coastal marine and deep-sea ecosystems.
Teasing apart the polarity of causation is important,
as the failure to do so has clouded the debate on the
biodiversity-ecosystem functioning issue. Claims that
changes in biodiversity can influence rates of ecosystem
processes, such as primary production, are based on
evidence from experiments involving deliberate
manipulation of biodiversity levels (Fig. 1A).
Observational evidence, in contrast, has typically
shown a hump-shaped relationship between these two
variables (Fig. 1B). These seemingly opposing findings can
re reconciled by recognising that they represent in fact
two dimensional projections of a three-dimensional
relationship among site fertility, primary productivity and
species diversity. Site fertility governs both plant species

diversity and primary productivity. This results in a
correlation between the last two variables although a
mechanism that would convincingly explain a dependence
of diversity on productivity is lacking.
6icdiwnily
'ProdLlClivIty"
Figure 1. Schematic of two seemingly opposing views of the relationship

between biodiversity and rates of ecosystem processes such as primary
production
In figure 1, (A) epicts the dependence of a process rate

(e.g. primary productivity) on biodiversity (e.g. species
richness) as established in manipulative experiments
conducted primarily in terrestrial grasslands. Primary
productivity is defined as the average rate of total
autotrophic biomass production in an ecosystem over an
extended period such as a year. (8) shows the correlation
between plant productivity and biodiversity typically
observed over broad spatial scales. The seeming
contradiction between the relationships shown in (A) and
(B) can be reconciled by acknowledging that 'productivity'
Aqllatic Systems 103
is a convenient and often used, but incorrect substitute

for site fertility, on which both biodiversity and
productivity depend while there is no causal dependence
of biodiversity on productivity.
Consequently, abandoning the incorrect 'productivity'
shorthand for site fertility would put to rest part of the
ongoing controversy and allow gathering forces to
scrutinise the implications of biodiversity loss on
ecosystem processes and properties. The unique features
of marine and freshwater systems provide ample
opportunities for testing whether the principles emerging
from research in some terrestrial ecosystems hold generally
true and for addressing currently unresolved questions of
the biodiversity-ecosystem functioning issue. To this end,
four novel experimental designs are proposed. These
include
the functional consequences of non-random as
opposed to random species loss, to be investigated by
manipulating the sequence and magnitude of species
loss by means of dilution experiments;
the significance of spatial heterogeneity by
manipulating biodiversity in interconnected habitat
patches;
the relative importance of local demographic processes
and species exchanges across ecosystem boundaries by
manipulating recruitment rate and diversity both
within patches and within the supply propagule pool;
and
the effects of species extinctions following exposure
to multiple stressors.
Much criticism on previous biodiversity-ecosystem
functioning experiments relates to the limited spatial and
temporal scales at which experiments have been
conducted. Some would argue this limitation precludes
inferences about phenomena in natural ecosystems and

thus the experimental findings would bear little relevance
to the real world. Extending Giller et al.' s proposal to
consider habitat diversity in the context of biodiversity-
ecosystem functioning relationships, Cardinale et al.
develop a patch-dynamics model that assesses biodiversity
effects when accounting for large spatial and temporal
scales. Their modelling results suggest that the effect of
biodiversity on an ecosystem process, plant production,
grows stronger with successional time. Significantly,
biodiversity effects do not necessarily change across spatial
scales, even though the mechanism underlying the effects
can indeed vary with scale of observation. When built into
the model, regional processes such as dispersal and
disturbance could amplify the biodiversity effect on
ecosystem functioning.
A new conceptual framework to analyse the impact
of multiple stresses on the functioning of ecosystems is
proposed by Vinebrooke et al. It is based on the idea that
the sensitivities of species to any two stressors can be
correlated. The direction and strength of this correlation
would determine what proportion of the species persists
when both stressors have an effect and hence determine
the potential of the persisting, tolerant species to
compensate for the loss of sensitive taxa and thus to
maintain ecosystem functioning. Evidence from planktonic
communities in temperate lakes affected by multiple
stressors provides initial support for the proposed
framework. It appears, therefore, that predicting the
impacts of global environmental change on biodiversity
loss and ecosystem functioning is enhanced by a
consideration of the possible interactive effects among
multiple stressors as mediated by correlated species
sensitivities to different stressors.
A key problem to demonstrate biodiversity effects on
ecosystem functioning is the large number of experimental
Aquatic Systems 105
units typically required for experimentation. Hence the

manipulation of microbial communities in microcosms has
been instrumental in investigating relations between
various aspects of aquatic ecosystem functioning and
changes in biodiversity. Even the use of simple
microcosms, however, may entail problems with the
analysis of data. Some researchers present a remedy by
proposing novel analytical methods based on resampling
. statistics to separate the contributions of temporal and
spatial variation to (}verall variation in ecosystem
functioning. Their ~,e.analysis of published data confirms
the neg~tj.v:e- relatiOnship found previously between species
ri~hness and the temporal variability of an ecosystem
process, the flux of carbon dioxide measured in a
microcosm. This negative relationship reflects high
variation among communities of low species richness,
rather than high temporal variation within communities
of low richness.
Most experiments and theory that dominate
investigations into the functional consequences of
biodiversity loss are based on model communities
comprising a single trophic level. This simplification
represents a major weakness in the current attempts to
understand effects of biodiversity loss on ecosystems,
especially for some aquatic habitats where strong trophic
interactions tend to impinge on the fluxes of energy and
matter. Based on the distribution of species richness and
the patterns of species extinction of different trophic levels
observed in a range of case studies, Petchey et al. develop
a Lotka_/Volterra predation model to assess whether the
trophic level from which species go extinct matters for the
functional consequences of biodiversity loss.
Their modelling results indicate that species belonging
to higher trophic levels are more likely to be lost than
species at lower trophic levels and, more importantly, that
the impact of species loss on an important ecosystem
property (total biomass) depends on both food-web

structure (the occurrence of omnivory and distribution of
species richness among trophic levels) and the trophic level
from which species are eliminated.
A comprehensive synthesis of evidence linking
changes in biodiversity to ecosystem functioning and the
mechanisms underlying the relationships is not attempted,
since assessing the ecosystem-level consequences of species
loss has just begun in aquatic systems and pertinent
information is still limited at present. However, aquatic
biota and ecosystems may serve as excellent testing
grounds for exploring the biodiversity-ecosystem
functioning issue with both experimental and theoretical
approaches, offering significant potential for the results of
studies in aquatic environments to contribute to general
theory building.
Biodiversity effects on ecosystem functioning
improves, predictions on the consequences of species loss
will enable scientifically informed, integrative management
of aquatic organisms and ecosystems. This ultimate goal
will conceivably continue to be the key driver for research
on the functional implications of biodiversity change,
overshadowing the scientifically motivated interest to
grasp how the structure of communities in general affects
ecosystem processes and properties. In face of the current
record rates of species losses and invasions, clearly a
thorough understanding of the functional consequences of
changing community structure at large, and the
mechanisms underlying them, is critically needed.
8
Impact of Dams on Biodiversity
Conserving habitats and ecosystems is the key to species

conservation. Here habitat may be defined as the place
where an organism lives or living space and an ecosystem
as the interaction or functioning between a community of
organisms and their nonliving environment. The earth can
be divided into a series of biogeographical regions, or
biomes, ecological copununities where certain species of
organisms co-exist within particular climatic conditions.
Within a biome there are several local factors which affect
the distribution of species.
Degradation of habitat leads to lowered population
size and loss of habitat to extirpation. Humans mainly
induce extinctions by causing habitat loss. The importance
of all components of the ecosystem including primary
producers, herbivores, carnivores, detritivores and recyclers
and their ecological function should be considered in the
design of dams. In addition there are certain special
linkages between species, e.g. freshwater mussel larvae
(glochidia) are parasitic on the gills of fishes for part of
their life cycle. In many cases specific fish hosts are
required.
WORLD'S FRESHWATERS
Watersheds of the world, their ecological value and their

vulnerability. The impacts to the freshwater environment

by various activities or sectors are summarised in Table 1.
Table 1. Sector threats to freshwater environments
Sector ~easure of threat Impacts (Add

to each, biodiversity loss)
Agnculture 11 % of land in crops, Runoff of toxic

26% in pasture. 3,4 of pesticides (fish kills);
human water withdrawals, fertilisers and manure
250 million hectares under (eutrophication); soil
irrigation. (turbidity and siltation).
Overgrazing (loss plant
cover, bank stability).
Deforestation 50% of world's forests lost; Soil erosion (turbidity
widespread clear-cut instead and sedimentation.
of selective harvesting. Rapid runoff. Loss
stream food/habitat
(leaves, wood, insects).
Changed hydrological
cycles.
Dams 60% world's river flow Fish migrations blocked;
regulated. 15% world's stocks lost. Seasonal
precipitation held in flows changed; flows
500,000 km2 of reservoirs. reduced. 25 million
Blocking of movement of km river habitat
localand long-distance modified. Flood plains
migrations in neighbour & deltas lost. Lowered
-hood of dam. fish production.
Sediment/turbidity /
nutrient changes.
Running to still water.
Industry Release toxic substances, Fish kills and advisories.
and urban hormone blockers, Impaired reproduction.
areas untreated sewage. Eutrophication.
IA of human water Reduced flows.
withdrawals.
Aquaculture Escape of alien Competition with
and intro- species. and loss of native
ductions Pollution. biota. Spread alien pests
and diseases. Loss of
native habitats. Genetic
Impact of Dams on Biodiversity 109
pollution.
Eutrophication.
Channelisation Simplification of Loss of habitats,
and levee river structure. flood plains and
construction 500,000 km of wetlands.
river altered for
shipping.
Fishing Over-harvesting. Reduced populations,
Gear damage. loss of stocks, changed
food webs, and
habitat loss.
Acid rain Reduction of Reduction of
pH (increase in populations or
acidity) of lakes extirpation of species
and streams of molluscs,
down to 4.5 or lower amphibians, fishes,
in thousands of water etc. in water
bodies in North bodies. Development
America and Europe. of skeletal abnormalities.
Deposition of aluminium
on fish gills.
Human population Doubled to 6 billion
and per capita since 1975.
consumption Per capita consumption
doubled since 1950.
Population/ consumption
rate increases magnify
each sector impact
above. Humans use 54%
of geographically &
temporally accessible
water.
This table shows that indirect landscape and direct

waterscape changes have had profound impacts on the
freshwater environment including:
River seasonal flow patterns (levelling)
River flow volume (reduction)
Accessibility of species to river segments (blockage of
migrations)
Input of organic matter (leaves, wood, insects)

(reduced)
Toxicity (increased)
Turbidity & sediments (both increases & decreases to
natural levels)
Nutrient levels (increased).
Freshwaters, and especially rivers and wetlands, are
amongst the world's most severely impacted and have
received many of the direct effects of human activities.
Long-term and quality data on river discharge patterns
are poor. Historical hydrological data are rare for many
rivers. However data from old air photos and satellite
images, including the new Radarsat images, which can be
made through clouds, offer a promising source of
information in particular the extent of flooding in the river
floodplains.
An analysis indicates that humans presently use 54%
of the geographically and seasonally accessible runoff.
Demands by the year 2025 may increase to more than 70%
of accessible runoff. The existing and growing friction
between countries over shared waterways shows that
humans are already facing water shortages that are
international in scope. Problems already exist at the
national and local level between and within sectors such
as agriculture, domestic users and industry. The fixed
supply of water poses profound problems for how it will
be shared by aquatic life and increasing demands of
humans.
BIODIVERSITY STATUS
Assessing the status of biodiversity at the global level is

difficult because the evaluation is incomplete and uneven.
Birds are quite well assessed, oligochaetes poorly, and
many micro biota not at all. The IUCN Red Lists ease the
task by bringing together what is known and applying

uniform criteria. However, summaries do not separate
freshwater biota from biota in other environments. Table
2 summarises terrestrial and freshwater vertebrate data at
the global lev~l.
Table 2. Proportion of terrestrial and freshwater

vertebrates globally threatened.
Group Proportion
Threatened (%)
Mammals - all 25
Land birds 11
Waterfowl (freshwater) 13
Turtles, tortoises and terrapins 38
Crocodiles 43
Amphibians 25 (estimated)
Freshwater fishes 33 (estimated)
Freshwater mammals
(freshwater dolphins and otters) 65
The level of threat for dominantly terrestrial vertebrates

is 11 to 25%, while the remaining values for groups
occurring more frequently or uniquely in freshwater range
from 13 to 65%. This gives a sense that, globally,
freshwater species are more at risk than terrestrial species.
Waterfowl are more threatened, 12.7%, than land birds,
10.8%. Similarly, freshwater mammals are more threatened,
65%, than all mammals, 25%. Data for North American
animals is more complete and the proportion of selected
groups of animals at 'risk is given in Table 3. These data
indicate that freshwater animals are much more at risk,
39 to 68%, than predominantly terrestrial ones, 15 to 17%.
Table 3. The proportion of selected North American animals at risk.
Group Proportion at risk (%)
Birds 15
Mammals 17
Freshwater fishes 39
Amphibians 40
Freshwater crayfishes 51
Freshwater mussels 68
Molluscs
Extinctions are a significant problem in terrestrial,
freshwater and marine molluscs. Of all the species that
became extinct since 1600 AD, 37% were mollusc species,
which is more than any other group evaluated (birds 17%,
mammals 14%, fish 14%, reptiles 3%, and all others 15%).
These percentages refer to the total known globally. It
should be made clear that the percentages are affected by
the degree to which the group has been studied and the
number of species in the group, e.g. the status of birds is
better studied than that of molluscs, but there are more
mollusc than bird species. The 1996 IUCN Red List of
Threatened Animals lists 12 bivalves and 216 gastropods
as extinct, and 114 bivalves and 806 gastropods as
threatened, for a total of 228 extinct and. 920 threatened
terrestrial, freshwater and marine molluscs.
About 18% or 145 of the threatened molluscs are
spring molluscs. Data on threatened freshwater molluscs
are given in Table 4).
Table 4. Freshwater mol/uses in the IUCN 1996 Red List of threatened

animals.
Risk category Bivalves Gastropods Total molluscs

Extinct 12 14 26
Critically endangered 85 60 145
Endangered 24 86 110
Vulnerable 8 194 202
Near threatened 66 35 101
Data deficient 4" 104" 108"
Total listed 199 493 692
Spring snails in the critically endangered category in

Austria, Australia and United States are threatened by
over-abstraction of water from their habitat or by
pollution. African freshwater molluscs are threatened by
decline in quality of water, pollution, damming, and
increased sediment load.
Researchers reviewed the threats to imperilled species
in the United States. They found that the chief threats to
freshwater molluscs were habitat degradation and loss,
97%, pollution, 90%, and alien species, 17%. The chief
causes of habitat loss were water development, 99%,
pollution, 97%, dams and other barriers to flow, 96%, and
agriculture 64%. More recently in North America, alien
zebra mussel invasions have become the major factor in
loss of native mussel diversity.
Birds
According to the 1996 Red List, 1,107 species or 11% of
all bird species are threatened and 104 are extinct.
Amongst the more threatened of bird groups are the
aquatic Gruiformes (rails and cranes) with 54 species, and
the partially aquatic Coraciiformes (kingfishers and bee-
eaters) with 11.5% threatened, while 18% of the
Podicipediformes (grebes) are threatened. Extinct aquatic
birds include the Colombian grebe (Podiceps andinus) and
the Atitlan grebe (Podilymbus gigas). Thirteen per cent of
globally threatened birds are (freshwater) water-birds.

There are 90 species of critically endangered, endangered
and vulnerable water-birds. This suggests that water-birds
are slightly more threatened than land birds.
Habitat loss and degradation are the key factors
affecting threatened species. This is generally due to the
loss and change through drainage and land reclaiming,
converting natural wetlands into urban and industrial
lands in Europe and into agricultural land in North
America.
Plants
There are about 270,000 scientifically described species of
vascular plants, but the true number may be in the order
of 300,000-350,000 species. Two important documents on
global plant biodiversity have been published in recent
years, Centres of plant diversity, in three volumes, and
the 1997 IUCN Red list of threatened plants. About 78%
of the world's plants are tropical (the zone between the
Tropic of Cancer and the Tropic of Capricorn). More than
40,000 plant species, about a quarter of the world's tropical
plant diversity, occurs in Colombia, Ecuador and Peru,
while Brazil has between 40,000 and 80,000 species.
Tropical and subtropical (areas north and south of
the tropics but outside of the temperate zone) Asia has at
least 50,000 species and Southern Africa has 21,000 species
of plants, of which 80% are endemic. The Red list of
threatened plants demonstrated that 32,112 species or
11.9% of the world's 270,000 vascular plant species, are
threatened, and 374 or 14% are extinct. The counts are for
terrestrial and aquatic species combined. Of these, 6,522
species are classed as endangered. Thirty-two countries
have at least 5% of their native species threatened. The
main countries (excluding small islands) having high
proportions of threatened species, 11-29%, include the
USA, Jamaica, Turkey, Spain, Australia, Sri Lanka, Cuba,

Panama, Japan and Greece. In the United States about half
of the potentially extirpated species are either obligate or
facultative native wetland species.
The Ramsar Convention in 1971 adopted a wide
definition for wetlands: areas of marsh, fen, peatland or
water, whether natural or artificial, permanent or
temporary, with water that is static or flowing, fresh,
brackish or salt, including areas of marine water the depth
of which at low tide does not exceed six metres. These
wetland plant species may be affected by changes in the
aquatic environment mediated by dams. Terrestrial species
may be affected by reservoir flooding, de-watering
downstream of dams; construction of transmission lines,
access roads or canals or through lowered water tables.
Walter and Gillett list the numbers of species, the
number globally threatened by category and the threatened
percentage for each family of vascular plants. The status
of selected freshwater plants based on their Red List is
presented in Table 5.
Table 5. Status of selected families of freshwater plants
Family % 'of threatened species
Water lilies, Ceratophyllaceae 16.7
Water lilies, Nymphaeaceae 8.0
Water poppies, Limnocharitaceae 31.6
Water plantains, Alismataceae 13.3
Reeds and sedges, Cyperaceae 7.0
Frog-bits, Hydrocharitaceae 14.0
Duckweeds, Lemnaceae 9.7

This small sample illustrates the level of threats, 7.0-31.6%

in 'water-loving' plants is about as high as in all vascular
plants (terrestrial and aquatic) 11.7%. The mean, 14.3, and
the midpoint, 19.1, of this sample are both higher than
the mean of the family values for terrestrial and aquatic
plants combined. This suggests that freshwater plants are
more threatened than land plants. About half of the
world's wetlands have been lost over the past century. In
Asia more than 5,000 km2 of wetland is lost every year
due to agriculture, irrigation, dam construction; etc.
LARGE DAMS
Dams, including large dams, are constructed because of

the potential benefits that they bring:
Water for increased food production - 250 million
hectares of agricultural land are under irrigation and
use three-quarters of the water supply
Generation of electric power without releasing
atmospheric pollutants or greenhouse gases -
hydropower contributes 20% of electricity production
Control of floods.
Drinking water. Of the Earth's 6 billion people, 1.5
billion are without access to reliable sources of
drinking water
The large expenditures involved with the construction and
operation of large dams and the benefits for agriculture
and power generation, are of considerable long-term
economic importance.
Large dams are usually >15 m from foundation to
crest. Dams of 10-15 m can also be defined as large dams
if they meet the following criteria: crest length 500 m or
more; reservoir capacity of at least one million cubic
metres; maximum flood discharge of at least 2,000 m 3s·!;
'specially difficult' foundation problems, or 'unus~al

design.' Major dams meet one or more of the followmg
criteria: at least 150 m high; having a volume of at least
15 million m 3; reservoir capacity of at least 25 km3; or
generation capacity of at least one gigawatt.
There are 306 major dams in the world and 57 are
planned in the near future. Table 6 lists watersheds with
more than five major dams.
Table 6. Watersheds with more tllIlII fi!'(' rna/or dal/l~
Watershed Number of dams
Parana 14
Colombia 13
Colorado 12
Mississippi 9
Volga 9
Tigris and Euphrates 7
Nelson 7
Danube 7
Yenisey 6
Yangtze 6
Construction of large dams includes the creation of access

roads, preparation of the reservoir, excavation, construction
of buildings and dams within and between river
diversions, digging of canals and erection of power lines.
Forested reservoir basins provide a particular challenge.
Leaving some trees may provide fish habitat although
leaving trees in any quantity may pose problems for future
fishing, water quality and turbine safety. While most
reservoirs tend to trap sediments, e.g. a new delta is being
formed within Lake Nasser, in some cases e.g. South
Indian Lake Manitoba, the exposure of clay soils to shifting
reservoir water levels increases erosion and downstream

sediment discharge.
The quantity of water discharged, when it is
discharged during diel and seasonal cycles relative to the
river's natural flow pattern and abiotic characteristics of
the discharge such as temperature, oxygen, turbidity, and
water quality significantly affect downstream biodiversity.
V ALUE OF BIODIVERSITY
Globally terrestrial and aquatic ecological functions have
been calculated to be minimally worth US$33 trillion per
year, almost twice the value of the global gross national
product, some $18 trillion, although the figure contains the
value of some biological resources as well as functions.
The annual per hectare total global flow value of inland
water systems, US$6,579 x 109 exceeded that from all other
non-marine ecosystems combined - US$5,740 x 109.
Ecological functions, although not ordinarily included
in gross global or national! domestic products nevertheless
make significant contributions to economies. Freshwater
ecosystems are economically more valuable than terrestrial
ones. In many developing countries, fishes, including those
from freshwater make a notable contribution in animal
proteins to an otherwise carbohydrate-based diet. In the
Amazon, the per capita consumption rate is 67 kgyr- 1
higher than in many areas. In Tonie Sap, Cambodia,
100,000 tonnes of freshwater fish are caught annually,
which source alone would provide a per capita 10 kgyr 1 •
Biodiversity has many kinds of values and potential
benefits for humans and the world as a whole. Before it
is diminished, those responsible may well wish to consider
the Precautionary Principle and take action to conserve it
before components of it are permanently lost, even when
the evidence for loss is not as strong as might be desired.
That approach is advocated by the Convention on

Biological Diversity.
FRESHWATER BIODIVERSITY PATTERNS
Although freshwaters comprise only 0.8% of the surface

area of the world and they have fewer species than other
systems, Table 7 shows that freshwaters contain more
species per unit area than terrestrial and marine
environments. This is particularly notable for fishes. The
total number of species (species richness) is only one
measure of biodiversity. Other measures include species
abundance. For example molluscs can form significant
proportions of the benthos; 80% of the biomass of the River
Thames at Reading is composed of freshwater unionid
mussels.
Table 7. Species richness of the world's major environments
Environment Area of world No. living Richness:

surface % species % %species /%area
Freshwater 0.8 2.4 3.0

Terrestrial 28.4 77.5 2.7
Marine 70.8 14.7 0.2
Symbiotic N.A. 5.3 N.A.
Long term investigations in the Czech and Slovak

Republics show a large biomass of fishes per unit area. In
the mountains, the mean fish biomass varies from 27-80
kg ha-1, in foothill streams from 90-500 kg ha-1 and in the
lowland streams from 300-600 kg ha-t, while in human-
made lakes it varies from 65-200 kg ha-1• Diversity indices,
which measure the relative richness, evenness, rarity and
abundance, require quantitative sampling and at present
there are little data for aquatic ecosystems.
Longitudinal Gradients
Species richness can change from the headwaters to the
river mouth. This may be related to changes in stream
order, water temperature, oxygen, current, turbidity and
available nutrients. The small headwater streams may have
low numbers of fishes that increase downstream as the
number of available habitats increases. Further and larger
downstream segments of the river may have moderate
numbers of species because lower habitat variety, and river
estuaries, where salinity varies, may also have moderate
species numbers.
Higher numbers of fish species are found in the
Tennessee-Cumberland plateau drainage of the Mississippi,
USA, than in the adjacent mainstream Mississippi; this
may reflect more numerous habitats in the varied
topography of the former, compared to the more constant
gradient of the latter. Different species may characterise
different river segments. Trout (Salmo) and sculpins
(Cottus) may occupy headwaters, minnows and catfishes
(Cyprinidae and Ictaluridae) midriver sections, and
euryhaline (salinity-level tolerant) species are found in the
estuaries. Freshwater molluscs generally increase from the
headwaters to the river mouth. This again is related to an
increase in habitats in the floodplain areas in the middle
or lower reaches.
Latitudinal Gradients
Latitude-longitude grids have the disadvantage that the
size of their grid cells shrinks towards the poles, a
disadvantage in making comparisons of numbers of
species if the study area spans several degrees of latitude.
It is generally accepted that the number of species tends
to increase from the poles to the tropics. Arctic waters can
therefore be expected to contain fewer species than ones
in the tropics. In many freshwater lakes in the Arctic there
is only one fish species, the Arctic charr, Salvelinus

alpinus. The greatest freshwater fish diversity is found in
Southeast Asia, tropical South America and central Africa
although many fishes have not been scientifically
described. The latitudinal trends in African freshwater
molluscs which decreased in diversity from the tropics
towards the Mediterranean and southern Africa.
Moist-arid Gradients
Species diversity of North American fishes was related to
a measure of 'aridity'. Data indicated that more species
were found in moist compared to arid areas. The study
showed at what critical level of moisture fish diversity
began to increase rapidly. Although arid areas may be
poor in species, desert springs or other water bodies may
be rich in endemic organisms, for example, there are vast
numbers of endemic spring-snails and fairy shrimps in the
arid west regions of the USA.
Hotspots
'Hotspots' are geographic areas rich in species. 'Hotspots'
often dominate over latitudinal and other gradients for
plants and animals. McAllister et al. listed countries that
are rich in absolute numbers of aquatic species and in the
number of species per unit area. There is some evidence
that 'hotspots' for different groups of freshwater organisms
are correlated, for example McAllister et al. found that the
numbers of amphibian and fish species in 12 'mega-
diversity' countries were highly correlated (r=O.937).
Revenga et al. catalogued the number of fish species,
endemic bird areas and percentage area of wetlands for
most of the world's primary watersheds. From this was
determined the twenty richest basins (Table 8).
Table 8. River basins and sub-basins with the highest number of native
species per unit area (in descending order) and the number of associated
large and major dams.
River or sub-basin Number of large or

major darns
Kapuas, Indonesia, Borneo o

Rio Negro-Amazon, northern o
South America
Chao Phrysa, Thailand 3
Hong, China 3
Xing Jiang (Hsi Chiang), China 7
Lake Victoria-Nile, Africa 1
Susquehanna, United States 124
Ohio-Mississippi, United States 711
Mekong, Cambodia, Laos, etc. 4
Alabama-Tombigbee, United States 103
Orinoco, northern South America 10
Lake Ontario, United States and Canada 2
Madeira-Amazon, South America 0
Magdalena, Colombia 5
Uruguay, South America 2
Hudson, United States 53
Fly, New Guinea 0
Yalu-Jiang, North Korea and China 3
Yangtze, China 17
Parana-Paraguay, South America 0
The number of species varies with size of basin; generally

larger basins have more species. Global 'hotspots' of
freshwater mollusc species include the Mobile Bay and
Tennessee River basin faunas in the United States (North
America has the richest freshwater mollusc faunas in the

world); the lower Mekong River of southeast Asia (160
species of which 72% are endemic; the upper Mekong has
been studied by the Chinese although data are not
available); the northern Western Ghats, India (71 species,
18% endemic); the Lower Uruguay River and Rio de la
Plata (93 species, 37% endemic); lower Zaire (96 species,
25% endemic); Lake Tanganyika (83 species, 64% endemic);
Balkans region (190 species, 95% endemic); Lake Baikal
(nearly 180 species, about 67% endemic). Note that the
freshwater molluscs display very high levels of endemism.
McAllister used an equal-area grid to show
geographic patterns of freshwater molluscs, marine fishes
and terrestrial mammals in the region surrounding the
Great Whale River Hydroelectric Project. Their approach
helped to identify gradients in species numbers and
determine any 'hotspots' in species including endemics.
Global, regional and national 'hotspots' are often the
dominating feature in geographic patterns of biodiversity.
Hence it is vital that they should be taken into account in
evaluating prospective dam sites. However more accurate
identification of freshwater 'hotspots' is needed. If a
standardised method was used for different groups it
would assist in identifying common underlying factors and
assist in application of the data to environmental impact
analyses. From data provided by Revenga et aI, the 11
most species-rich watersheds were determined (Table 9).
Table 9. Number of fish species in the world's 11 most species-rich

primary watersheds.
Watershed/ Number of fish Number of species/

continent species 100,000 km 2
Amazon, South America 3,000 49

Congo, Africa 700 13
Nile-Lake Victoria, Africa 432 12
Mississippi, N. America 375 12

Parana, South America 355 14
Yangtze, Asia 322 19
Kapuas, Indonesia, Asia 320 360
Orinoco, South America 318 33
Xi Jiang (Pearl), Asia 290 71
Mekong, Asia 244 30

Chao Phrya, Thailand 222 124
In Table 9 the Nile River and Lake Victoria watersheds

were combined. This results in a value inflated by a few
percent for those species shared between the river and the
lake. The Lake Victoria sub-basin has 343 species and 121
species/100,OOO km2, while the rest of the Nile watershed
has only 129 species and 4 species/100,OOO km2. The
difference in diversity is due to the rich flock of endemic
cichlids in Lake Victoria. The Amazon, Congo, Nile, Parana
and Yangtze watersheds are the most species rich, with
the Amazon far ahead of the others. When area is taken
into account, the Indonesian Kapuas and Thailand Chao
Phryas watersheds are significantly richer.
However area correction will not cover 'hotspots'
lying partly within a basin any more than they will within
a country. Using Revenga et al.'s data for sub-watersheds
shows this quite clearly. The Rio Negro, sub-watershed of
the Amazon has 600 species with 83 per 100,000 km2, the
Ohio sub-watershed of the Mississippi has 281 species and
57 per 100,000 km2, while Lake Victoria, a sub-watershed
of the Nile, has 343 species and 121 species per 100,000
km2• Using as closely-spaced geographic grid as the data
will permit, would provide the most precise localisation
of species or endemic species 'hotspots'. 'Hotspot' analysis
can be a useful tool in evaluating potential impacts of
different dam sites.
Phylogenetic Diversity: Species are the most common

units used in ~valuating biodiversity. Yet species are only
one unit in a hierarchy: Since families are more distinct
genetically, ecologically and behaviourally, and generally
have a more ancient origin than species, many taxonomists
and conservationists would give a higher pri(nity to
conservation of the sole species representing a family than
to one of a family with numerous species.
Alien (exotic) Species: Disturbed environments created
by dams can foster populations of alien species and that
diversions associated with dam projects may enable the
invasion of such species. In the environmental assessments
of dams, alien species should be separated from
indigenous ones, and not tabulated in measurements of
local indigenous biodiversity. In general, dam projects
should not foster the introduction of alien species.
BIODIVERSITY IMPACfS OF DAMS
There are a number of different migratory patterns of river-

dwelling species. These include the well-known
anadromous fishes e.g. salmon and hilsa and the
catadromous fishes such as eels. Adults of anadromous
species migrate up rivers to spawn and the young descend,
while the reverse occurs with catadromous species. But
many other freshwater fishes move up rivers or their
tributaries to spawn, while the glochidia larvae of
freshwater mussels 'hitch rides' on host fishes. To help
counteract the drift downstream of their larvae, some
aquatic insect adults such as mayflies and stoneflies fly
upstream to lay their eggs.
Dams block these migrations to varying degrees.
However most waterfowl are able to fly over dams.
Reservoirs provide waterfowl habitat and may aid longer
migrations by providing 'stopover' sites. Waterfowl is used
in the present study to cover all wetland bird families (e.g.
in the Ramsar Convention), including divers, grebes,

cormorants, Anatidae (swans, geese, ducks), coots and
rails; "shorebirds" (synonymous with waders); and some
other wetland bird families notably gulls, terns, herons and
egrets.
The blockage of fish movements upstream can have
a very significant and negative impact on fish biodiversity.
Many stocks of Salmonidae and Clupeidae have been lost
as a consequence. In the Columbia River, USA, more than
200 stocks of anadromous, Pacific salmonids became
extinct. Sturgeon populations in the Caspian Sea rely on
hatcheries, mainly in Iran, since Russian dams block
natural spawning migrations. Hydroelectric dams in the
Amazon basin have halted the long distance upstream
migration of several species of catfishes and interrupted
the downstream migration of their larvae. On the
Araguaia-Tocantins River basin, Brazil, several species of
migrating catfish have been drastically reduced in
abundance as a result of dams; catches in the downstream
fisheries have been reduced by 70%.
McDowall noted that diadromous fishes (those that
migrate at regular phases of their life history between
freshwater and the sea, comprise about 250 species, <1.2%
of all fishes species but form 3% of those classed as
endangered. He observes that amongst them are species
of great importance to fisheries, out of proportion to their
number. Due to their occupation of connected habitats
through which passes are needed at two or more life
history phases, they pose special problems for
conservation. In particular the diversity of habitats used,
the extensive areas occupied, the spatial separation of the
habitats and the need for fish passage between them.
Fishways must be designed to assist not only upstream
and downstream migrations of large, fast-swimming
migrants such as salmon that can pass substantial barriers,
but also those of the lesser-known climbers like eels and
gobies that require continuous dampened surfaces on

which to move.
While dramatic declines in migratory species such as
lampreys, sturgeons, salmons and clupeids were well
known in European rivers, other fishes, the socalled
resident or non-migratory fishes which perform in-stream
movements require attention. These include the European
minnow, Phoxinus phoxinus, the Japanese sculpin, Cottus
pollux, the grayling, Thymallus thymallus and Balon's
ruffe, Gymnocephalus baloni. Even smallsized species such
as the white bream, Abramis bjoerkna, were found to
migrate up to 60 km from the place they were tagged.
Some of these small-sized species are among the most
endangered. Damming can contribute towards their
. decline ,as in the unique and critically endangered percid,
the asprete, Romanichthys valsanicola, endemic to the
Arges River in Romania.
Artificial barriers also lead to the dramatic decline of
the endangered cyprinid fish, Anaecypris hispanica in
Iberia. In general, a river is a one-way system for molluscs,
as many molluscs can only move downstream by drifting
or being dislodged by flood events and moved
downstream. But some species with a larval form can
move significant distances upstream with the aid of a third
party, e.g. host fish during the larval stage.
Nutrients
Conventionally, rivers have been regarded as the one-way
transfer of matter downstream. Experiments by the
Fisheries Research Board of Canada several decades ago
showed that removal of spawned out sockeye salmon
carcasses from streams reduced the growth of fry in the
following year. Recently there has been greater
appreciation that migrating species carry nutrients
upstream. Researchers proposed that intensive coastal
12X Ecology and Biodiversity
catches of Atlantic salmon in the Queen Charlottes, Canada

reduce the nutrients in adjacent stream riparian zones and
estuaries. The contribution of nutrients from both Atlantic
and Pacific salmon carcasses has been linked to riparian
tree growth.
Data from 45 watersheds in British Columbia suggests
that up to half the nitrogen stored in giant oldgrowth trees
originates from sockeye salmon, using the nitrogen isotope
N 15 as a tag. The largest source of N 15 is in the ocean. Some
of the researchers reviews the contribution of Pacific
salmon carcasses to the flow of nutrients and energy for
aquatic and terrestrial ecosystems. Anadromous fishes
carry nutrients in their bodies as well as gametes up river.
The fishes and their eggs are used for food by a variety
of aquatic and terrestrial predators, scavengers and
detritivores. Decaying bodies, eggs and faeces of the
consumers provide nutrients for the algae and other plants.
Researchers calculated that in the Columbia River,
USA prior to dam construction, spawning salmon
contributed to 45,150 metric tonnes of fish bodies to the
aquatic and terrestrial ecosystems. By 1997, following
construction of multiple dams and impacts of other human
activities, only 3,400 metric tonnes were contributed, 8%
of the pre-dam level. The decreased production could be
self-perpetuating, since small stocks produce lower
amounts of in-stream nutrients for themselves, as well as
other species. To a degree fertilisation from the lower
riparian vegetation will also affect fish productivity. The
fall of insects, leaves and twigs into streams, which serves
as direct or indirect food can be expected to decline, as
riparian vegetation growth diminishes.
There is no reason to suppose that the upstream
transport of nutrients is restricted to anadromous
salmonids. It can be expected that andromous hilsa in
southeast Asia, Arctic whitefish of the genera Coregonus
and Stenodus, lampreys such as Petromyzon marinus, and
New Zealand retropinnids such as Stokellia anisodon

would also transport nutrients upstream during their
spawning migrations.
Turbidity
Reservoirs trap suspended particles, reducing turbidity
downstream. Many species are adapted to natural
turbidity. For example turbid water catfishes have small
eyes, refined senses of smell and touch in their sensitive
barbels. The turbid water helps conceal the fish and other
biota from visual predators like birds. When normally
turbid water becomes clear below dams, the indigenous
species may find themselves at a disadvantage. Other
animal species may move in, filter feeders and aquatic
vegetation may flourish. Sediment burrowing species may
find their habitat has diminished. Flood plain ecosystems
and deltas may no longer be replenished by the annual
transport of sediment. Silt and increased turbity, above
natural levels, can interfere with primary production. In
the Mekong River system, silt levels increased following
deforestation. This resulted in siltation of the river, lakes
and swamps threatening the river fisheries.
Large Organic Debris (LOD)

This consists of branches and tree trunks that fall into the
river because of age, storms, beaver activity and eroded
banks. Numerous organisms feed on LOD which is often
the first link in the food chain. Trees can also playa
complex role in creating habitats e.g. they divert, slow and
speed up current flow, they shelter a variety of biota from
currents and predators and create feeding stations. On
land, LOD helps stabilise slopes and reduces erosion, and
is converted into humus which helps hold water and
moderates the runoff. In estuaries, along shores of lakes
and coastal areas LOD functions as a source of food,

energy and habitat.
In the Santilla River, Georgia wood represented 4%
of the total habitat, yet supplied 60% of the invertebrate
biomass and 78% of the drifting invertebrates. Dams tend
to 'sieve out' LOD. The logs and branches may become
waterlogged and sink, drift onto the shore or are removed
by booms or other systems designed to protect turbines.
If the integrity of downstream ecosystems is to be
maintained, then LOD input must be sustained.
Species and materials may move laterally away from
the river, extending the effect of river changes to a band
of varying width, parallel to the river.
Watering wildlife
As long as there is sufficient river flow below the dam,
wildlife such as deer, antelope and elephants wiII come
to the water, especially in the dry and hot season for
drinking. Hippopotamuses will use water of sufficient
depth as a day-time refuge, emerging to forage at night.
Many birds may fly in to drink. These lateral movements
can extend to several kilometres from the river. The
reservoir itself, however, may serve as a source of water
during the dry season or droughts, to wildlife living within
range. Watering terrestrial vegetation. Water release
protocols can lower water tables lateral to the rivers which
may affect vegetation there.
Riparian ecosystems along most major western rivers
of the U. S. have changed as a result of water development
and flood control. Losses of riparian forest downstream
of dams have been reported throughout western North
America. Cottonwood-willow stands are being replaced by
nonnative woody species such as Russian olive and
tamarisk. This may result in diminished LOD input.
Loss of riverbank forests

The extinction of species may be related to the loss of
gallery forests adjacent to rivers which became submerged
following dam construction. The landsnail, Anthinus
albolabiatus, was formerly endemic to gallery forest
adjacent to the Uruguay River but became extinct after
the formation of the SaIto Grande Dam, Uruguay. This
factor is of concern with regard to proposed dam projects
in South Africa, where much of the remaining forest is
preserved on the steep sides of valleys, which are also
suitable sites for dam construction.
Maintaining populations and gene flow

The main stem of a river performs two related ecological
functions to biota in tributaries. Periodically, populations
of a tributary stream species, particularly fish, may go
extinct and may be restocked from nearby rivers. Secondly
individuals may ascend a non-home tributary and
contribute to the resident population's genetic diversity.
For example salmonids are known to home to their natal
streams for spawning, using celestial and olfactory cues.
A small percentage of migrants are known, through
tagging studies, to stray into adjacent tributaries or rivers.
Straying can serve to repopulate streams and also
contribute to the genetic diversity of populations through
gene flow.
A single dam and more significantly multiple dams
along a given river interfere with the genetic bridging
function of the mainstem. Thus dams on the main stem
can influence species diversity in lateral tributaries, even
though there may be no changes in water flow or quality
characteristics. In the Murray River, Australia, river
regulation has lead to the separation of billabongs from
the main channel habitat and thus for many molluscs in
particular freshwater gastropods. Since river regulation

was introduced some mollusc species have become extinct.
Oxbows, wetlands and springs

Oxbows, ponds, lakes and wetlands are often isolated in
the floodpain from the river's main stem. These may be
replenished with water, biota, sediment and nutrients
during natural, seasonal floods. Levelling out of river
discharge has been known to prevent these periodic
linkages with the mainstem. Most of the 50 species of
fishes (many endangered) in the Austrian Danube, depend
on the connection between the river and its backwaters.
Transmission lines. Transmission lines affect biodiversity
on land. The use of herbicides to control plant growth
under power lines probably reduces native plant diversity
in favour of weed species, which are often exotic.
Power lines corridors can serve as refuges for rare
species. Tree trimming may provide a haven for native
'sun-loving plants'. Shrub communities may flourish under
power lines and provide habitat for nesting and migratory
birds. Only nine snoutbean plants were thought to exist
in all of Kentucky, USA. However an East Kentucky Power
Plant Corporation (EKPC) study showed that about two
thousand specimens survived under their power line.
EKPC adjusted their mowing schedule and removal of
woody plants and educated other utilities about protecting
native plants. Utility rights of way may harbour rare birds,
amphibians, reptiles, tree snails, mammals and other
species.
Upstream
Above the reservoir: Water quality, flow and seasonality of
flow are not normally disrupted in the upstream area
above the reservoir so impacts are generally less than for
the reservoir and downstream areas. Nevertheless, the dam
and the reservoir affect migratory movements of species

into and out of this upstream area. The genetic exchange
with downstream segments is reduced or prevented. A
study was made of molluscs upstream in a braided river
that enters a reservoir on the River Inn in Austria. Data
shows that there was a decline of 10 species upstream of
the reservoir. This was due to channelisation of the braided
area, an increase in the overall river gradient and a
consequent reduction in the active floodplain area.
The extirpation of freshwater mussel populations
upstream of the dam construction at Lake Pepin on the
Mississippi River, USA, was due to the lost migratory fish
host species, skipjack herring, Alosa chrysochloris.
Reservoirs. In the construction of reservoirs, the clearing
of vegetation, movement of earth and rock, the presence
of humans and machinery, bringing in construction
materials, use of explosives, noise, and reducing or cutting
off river flow and increasing turbidity, will affect
biodiversity. Removal of forests or other vegetation over
a wide area, excavation, earth and rock movement and
reduction in river flow are the most significant.
Some of the on-site activities are mirrored in off-site
disturbances such as the mass displacement of earth and
rocks and road building. During reservoir filling the river
and any associated wetland areas become inundated.
Riffles, runs and pools of the river are lost beneath the
rising waters, leading to the extirpation (or extinction) of
habitat sensitive riverine species with tightly defined niche
requirements. Fishes in rivers are generally well adapted
to flowing water. Similarly molluscs are often restricted
to specific habitats within the river system, e.g. some
species are bottom-dwelling filter feeders, others live in
weeds at the edge of the channel.
The transformation of a river to a reservoir therefore
poses a problem for the resident, mainly riverine species
that are not adapted to the new conditions. Lacustrine
fishes have been introduced into reservoirs in a number

of cases e.g. Lake Kariba although this may pose new
problems. In eastern Canada, lakes and streams, which
have emerged fairly recently from glaciation, contain a
number of fish species able to dwell in both habitats.
However some species like the longnose dace, Rhinichthys
cataractae, and rainbow darter, Etheostoma caeruleum, are
adapted to running water only. Molluscs generally show
a drop in species richness from pre- to post-impoundment.
Unionid mussels are exposed to a number of changes
in the reservoir impoundment. The most important ones
include: changes in the fish host population size (needed
for the glochidia stage of the mussel larvae), reduction in
the mussel population size (fertilisation success),
eutrophication (affecting adult and larvae) and dissolved
oxygen levels (low levels of oxygen in the profundal zone
would eliminate mussels from that zone). Eutrophication
can limit the interstitial habitat of post-glochidial juveniles.
Raised nitrate and phosphate levels are especially
deleterious to juveniles.
Organic debris can clog benthic interstitial spaces.
Some of these effects can extend downstream of the dam.
Table 10 shows the environmental and biotic factors which
influence the various stages in the life cycle of unionid
mussels. Extinction of 38 out of 42 taxa of molluscs in the
Mobile Bay, Alabama, USA, basin occurred when the big
river shoal fauna were covered by deep standing water in'
the impoundments and subsequently buried under
increased siltation. However molluscs may comprise an
important part of the benthic fauna of some reservoirs. In
the man-made Lake Kariba, molluscs made up nearly the
entire biomass of the benthic animals.
In the Tennessee River Basin, U.S., several molluscs
are under threat of global extinction following the
construction of dams and the subsequent regulation of
flow. A number of gastropods of the family Pleurocercidae
are under threat as they persist on clean-swept shoal areas

below dams on the river and three other species have been
extirpated from the river. Over 85 mussel species were
known in the Cumberland River of the upper Ohio-
Tennessee River basin prior to the construction of
impoundments and locks between 1916 and 1923. In
Kentucky portion of the lower Cumberland, for example,
there were 25 species in 1911, 15 in 1981 and only 4 in
1994, for a total of 21 extirpations.
Table 10. Factors affecting the life cycle stages of IInionid molluscs.
Factor Life cycle stages affected
Exploitation (pearl fishing, shells Adults

for buttons or seeding
Fish host stock size Glochidium
Mussel stock size Fertilisation,
glochidium numbers
River bed Adults, juveniles
Flow regime Fertilisation,
glochidium infection,
settlement, juvenile
and adult
Suspended solids Adults, breeding and
brooding
Eutrophication Adults, juveniles
Nitrogen Adults, juveniles
Phosphate Adults, juveniles
Dissolved oxygen All stages
Conductivity Adults, juveniles
Calcium Ad ults, juveniles
pH (acidity) Adults, juveniles
Interstitial particulates Juveniles
Interstitial water chemistry Juveniles

Industrial pollutants All stages
Pesticides All stages
Extensive mussel beds contribute to the health of rivers

by their filtration power. Reductions in those beds reduces
this ecological function. However replacement mollusc
faunas have developed in some African reservoirs.
River sections with steep gradients or escarpments
sometimes offer optimal conditions for locating
hydroelectric and other dams. However, those locations
may provide special fastwater habitats for species with
only scattered distribution, or local endemics. In some
cases species may multiply following construction of a
dam. The status of a freshwater pulmonate, Bulinus
truneatus, changed from rare to common in Lake Volta,
Ghana. This led to an increase in the level of urinary
schistomiasis infections in the region.
Some of the researchers listed 12 darter species
(family Percidae) as endangered or threatened, and 9 of
special concern in Tennessee, USA, a state which has many
dams due to the activities of the Tennessee Valley
Authority. The species of darter found in streams, now
flooded by reservoirs in the Tennessee River system, had
well defined river run and pool niche requirements. In
Texas, USA, the filling of a reservoir was involved in the
extinction of the spring-dwelling Amistad gambusia,
Gambusia amistatensis. In some tropical reservoirs the
overall number of fish species has increased, although
several riverine species have disappeared, e.g. Lakes
Kariba, Zambia and Zimbabwe, and Ayame, Cote d' Ivoire.
Tilapias of the family Cichlidae are usually the most
successful in these lakes.
The extent of entrainment of larval and juvenile fishes
in hydroelectric turbines varies according to flushing
Impact of Dams on Biodivel sHy 137
duration, depth of extraction and species present. Passage

through turbines of young anadromous salmonids, en
route downstream, is a well-known source of mortality.
In the catadromous eels, family Anguillidae, it is the
downstream migrating adults which are killed by the
turbines. Reservoirs promote waterfowl and many dams
have substantial populations. The type of shoreline,
shallow, with fringing vegetation, supports greater species
diversity and larger numbers compared to steeply shelving
mostly deep water sites. A substantial number of dammed
sites support nationally or internationally important
waterfowl assemblages. In more arid areas, creating dams
increases numbers of birds able to remain all year round
in otherwise largely seasonally dry places. The presence
of a dam has substantively altered the migration
phenology and distribution of some species.
A study of natural lakes and dammed reservoirs in
Switzerland (Table 11) showed that waterfowl species
diversity was considerably higher on natural lakes than
on dammed lakes. Nevertheless there is considerable
overlap in the number of species on the two types of water
bodies. The most common five species were the same on
natural lakes and dammed lakes, common coot, tufted
duck, mallard, pochard and great crested grebe. Damming
of rivers has increased the number of open water sites
available to wintering waterfowl.
Table 11. Waterfowl species diversity in Switzerland.
Natural lakes Dams
Number of sites 8 6
Number of species per site 14-30 11-20
Total number of species 33 23
The United Kingdom is of major importance for wintering

waterfowl which use the Eurasian-African flyways.
Because of its relatively mild winters, its wetlands seldom

freeze over, and, in severe winter weather in continental
Europe, it additionally serves as a cold weather refuge for
waterfowl species. Estuaries and coasts tend to have higher
numbers of species because they attract waders as well as
waterfowl, while inland systems generally support only
wildfowl in internationally important numbers.
A much smaller number of artificial wetlands are of
international importance for wintering waterfowl. Only 11
support one or more internationally important populations
and only three have large overall wintering populations
>20,000 waterfowl. For several species a large proportion
of nationally important lakes are artificial ones: 40% or
more of important sites are artificial for eight waterfowl:
little grebe, great crested grebe, great cormorant, gadwall,
northern shoveller, pochard, tufted duck and common
coot. A further analysis showed 27 species on five natural
lakes in the UK, as compared with 33 species on artificial
lakes, in contrast to the situation in Switzerland.
On the negative side was the fact that two exotic
species, the Canada goose and the ruddy duck were
present on dammed lakes. Those species have expanding
populations and are of conservation concern. Much of
South Africa is arid and has few natural permanent water
bodies. Almost all permanent water bodies are dammed
sites constructed for water storage. There are at least 517
major reservoirs and numerous small, farm dams. The
. presence of these new wetlands has had several
consequences. At least 12 impoundments support major
and important concentrations of waterfowl. Suitable
conditions have been provided for the Pelecaniformes
(pelicans, darters and cormorants), 70% of the global
population of the South African shelduck during moulting,
and refuges for species of national conservation concern
such as the pink-backed pelican, Pelecanus rufescens, and
the Caspian tern, Hydroprogne caspia.
Negative impacts in overall waterfowl assemblages

in southern Africa are due to the loss of many of the
former natural marshes and riverine habitats through
reduction in river flow, removal of seasonal flow
variability and consequent changes in sediment movement
and channel stability. In addition poor dam management
may cause sudden major releases of water, causing major
downstream floods in areas that have had little or no flood
activity for years. That can affect species that use
unvegetated river banks and sand banks between river
channels.
Running compared to still water impacts: The
construction of reservoirs converts lotic (running) into
lentic (still water) habitats. Species dependent on running
water will diminish or disappear. In almost all cases, the
diversity of fish species will drop. Reservoir fisheries are
one of the frequently claimed benefits of impoundments.
The changes in catches following impoundment are
variable. However the catches in new reservoirs frequently
go through a "boom and bust" cycle, with catches initially
increasing following filling of the reservoir and then
declining. Therefore impact assessments of dams should
be based on the long term catches. The shore-edge
(marginal) ecosystem also changes.
A study of the Upper Mississippi River, USA, an area
with many dams, showed that open water and marsh
habitats generally increased between 1891 and 1989,
although at the expense of grass-herb, woody terrestrial,
and agricultural habitats. For example, in 'Pool 8', open
water and marsh increased from 3,600 hectares in 1891 to
9,500 hectares in 1989. The edges of new reservoirs are
often exposed to erosion, while deeper areas are
sedimented. In the Upper Mississippi River sedimentation
rates of one to three cm per year have been measured.
Erosion was more prevalent in shallow areas and
sedimentation at deeper depths, the processes converging

between 0.9-1.5 m.
Creation of new sublittoral and profundal zones: Water
oscillations in the littoral zone reduce its suitability for
species requiring stable conditions. Some mobile species,
such as shorebirds, may find this habitat suitable for
feeding. The nature of the deeper or profundal zone will
depend on the climate, preparation of the reservoir prior
to filling and other factors. In boreal and arctic areas the
deeper waters are normally cooler than the surface waters
and this provide habitats for both warm- and coolloving
species. If the trees and other vegetation are not cleared
from the reservoir, then decomposition commonly leads
to low oxygen levels in the profundal zone usually only
suitable for anoxic microorganisms.
Weeds, exotics and diseases: The changed or fluctuating
conditions in the reservoir may lead to opportunities for
weed or exotic species e.g. the water hyacinth, Eichhornia
crassipe. Increases in the number of mollusc-borne diseases
following dam construction in various countries. For
example at least four genera of mollusc-borne human
diseases have increased as a result of impoundments in
Thailand. Following the construction of Lake Volta Ghana,
the gastropod, Bulinus truneatus colonised the lake,
replacing Bulinus globosus and other species which were
not able to withstand the lake-level fluctuations. This lead
to an increase in the level of urinary schistomiasis in
villages around the lake.
Mercury: Mercury is in a harmless inorganic form in
many soils. Bacteria which feed on decomposing matter
in c;. new reservoir transform the mercury into methyl
mercury which passes up the food chain from plankton
to fishes and those species that feed on fishes, including
humans. Biomagnification results in higher mercury levels
as mercury ascends the food chain. In the La Grande Phase
of the James Bay, Canada, hydro project it was found that
reservoir fish became contaminated with mercury at levels

exceeding World Health Organisation (WHO) standards.
Sixty-four percent of the Cree living in the La Grande
estuary had blood mercury levels far exceeding WHO
standards. Mercury can also negatively affect wildlife that
prey on mercury-contaminated fishes. Loons are severely
affected by mercury pollution. Sport fishing also plays a
role in the local economy, and mercury contamination
creates unfavourable publicity. Sedimentation. Reservoirs
tend to serve as sediment traps since river velocities and
carrying capacity for particles decrease in reservoirs.
However sometimes fluctuating water levels in reservoirs
erode the shores and add to the turbidity of the reservoir
discharge. Sedimentation can degrade habitat both in the
reservoir and below the dam, as well as reduce storage
capacity.
Many of the molluscan extinctions in the Mobile Bay,
USA drainage, following multiple impoundments, were
due to siltation. The degree of tolerance to silt cover
depends on the species of mollusc. Suspended silt may
reduce the feeding efficiency of filter-feeding bivalves and
other species. About 50 km3 of sediment, nearly 1% of
global reservoir capacity, was estimated in 1997 to be
trapped behind dams. Keeping sediments flowing through
reservoirs will benefit both reservoir life and the life of
downstream ecosystems such as flood plains and deltas.
Many of the existing potential sohtions, such as dredging
and sluicing, have economic or environmental limitations.
Downstream
In the downstream segment, most of the impacts of a dam
are negative. In a preliminary assessment of 66 case studies
of the impact of dam construction on fishes, based on
qualitative information, 73% of the impacts were negative
and only 27% were positive. About 55% of the impacts
were below the dam and linked to fish migrations and to

floodplain access. In the distribution of molluscs along a
240 km stretch of the Little River in Oklahoma, USA there
were mussel extinction gradients downstream from large
impoundments. With increasing distance from the dam
there was a relative increase in mussel species richness
and species abundance. Only those stretches furthest from
the dam contained the relatively rare species. Richness
declined below each successive dam, with a multiplier
effect.
Overall volume of flow

Some reservoirs have been filled by cutting off all or
almost all flow downstream of the dam, e.g. Cabora Bassa,
Mozambique with the consequent loss of organisms. Large
aquatic species such as sturgeons, crocodiles and dolphins
require minimal flows in which to navigate and feed. Such
,;pecies may be affected by reduced flows including a
reduction in the area of habitat utilised. This may lead to
smaller populations, reduced growth rates and, where
populations are already at risk, extirpation or extinction.
A certain level of downstream flow is needed to maintain
a minimum volume and area of habitat, oxygen
concentration and other 'desirable' in-stream conditions
and avoid lethal temperatures.
Normal seasonal flow patterns are a key to
maintaining river biodiverstiy. Balancing reservoirs may
help avoid pulse discharges, delay peak discharges and
reduce them to an ecologically acceptable level and
guarantee a certain minimum discharge. Constructing
dams just above large tributaries can moderate changes
to downstream flow patterns. If the tributary has a similar
seasonal flow cycle to the mains tern, then the downstream
flow pattern and seasonal cues will be less impacted than
if the mainstem dam had been sited below the tributary.
Seasonal variability of flow and flood plains

The pattern of flow of a river undergoes a regular series
of changes with the seasons. The patterns can differ
profoundly from region to region e.g. in an Indian river
the peak flow may be during the monsoon, in an Arctic
river during snow-melt or ice break-up. The expansion and
contraction of the river controls living space and access to
particular habitats. It is to these profound seasonal patterns
that the species in a drainage basin adapt. It IS from river
flow events that species take cues to migrate, spawn, etc.
The rhythm of the river is thus tied intimately to the life
of river species. Dams alter the natural flow of rivers as is
shown in the Colorado River, USA.
Some species are adapted to strongly seasonal flow
regimes with flash flooding, as in certain river systems in
Australia. Drastic declines in the molluscs of the Murray-
Darling River system have been attributed to: predation
and sediment disturbance by introduced fish such as the
common carp, Cyprinus carpio, changes in flow patterns
through intensive flow regulation after impoundments and
possible changes in algae, bacteria and fungi, which form
potential mollusc food sources.
Studies identified about 72 species in the Lower Nile.
It was deduced that approximately 35 species have
disappeared from the Lower Nile or have become very
rare, since the construction of the Aswan High Dam and
other dams and barrages in the Nile River.
Water table changes

Water diversion for irrigation may lower the downstream
water table adjacent to the river. Further, the 'levelling out'
of floods, may reduce seasonal recharging of the water
table. The supply to springs, artesian flows and cave
streams may also be affected. Cave streams and some
aquifers sometime have species characterised by reduced
or absent eyes, loss of pigmentation and enhanced non-

visual sensory systems. According to David Culver
(personal communication) there are 1,300 described
obligatory cave species in the United States alone, while
the undescribed species probably number several times
that value. Many cave organisms are restricted to a single
cave or cave complex.
Cutting off the cave water supply, either temporarily
or permanently, may lead to extinctions. When selecting
potential dam sites, planners should be alert to known
cave dwelling speciesor check for the presence of unknown
species. Of the world's 110 described species of cave fishes,
a high proportion are threatened. The percentage of
threatened hypogean fish species has risen from 18% in
1977 to 87% in 1996. Most of the species have very limited
distribution. The primary threats are habitat degradation
(e.g. quarrying and siltation due to dams and logging),
hydrological manipulations (e.g. water removal for human
consumption or irrigation and damming), environmental
pollution (e.g. eutrophication and poisoning from
agricultural and industrial runoff), overexploitation
(capture for the aquarium trade) and introduced exotic
animals.
Abiotic changes
In summer, in temperate lakes, solar radiation heats the
eplimnion but not the hypolimnion. The hypolimnion is
often anoxic. In the autumn the lake undergoes mixing
and some heat is transferred to the bottom layers. Water
discharge from the dam is usually below the epilimnion.
Therefore in summer the water discharged into the river
below the dam is colder and has less oxygen than normal
and in winter it is warmer. These physical changes can
effect the biota for long distances down the river.
Discharges from the reservoir are variable usually resulting
from the requirements for hydroelectric power and not

related to natural cycles.
Flow below the dam can rapidly alter from almost
standing water to torrential flows and water depth, water
velocity, oxygen concentration, temperature, suspended
solids, pollutants and chemical composition can change
in a very short period of time. Many of the effects
experienced downstream of a dam are in reverse of those
produced in the reservoir above them. Heat, silt, inorganic
and organic nutrients retained in the lake are lost to the
stream below. Large annual variations in water level in
the reservoir results in a decrease in the annual variation
in water level in the efferent river.
Inland deltas
Dams trap sediments, diminishing the downstream supply.
An inland delta and flood plain, including a network of
oxbow lakes, in the middle Danube, was supplied by
sediment during natural seasonal floods. The area
produced an impressive harvest of trees, fishes and cereals.
Biota included 65 species of fishes, 11 of amphibians, nine
reptiles, 41 mammals and 242 birds. Construction of a
series of dams, dredging the river channel and construction
of a canal deprived the delta and flood plain of the annual
supply of silt and resulted in severe alterations in the
benthos and zooplankton communities as well as the
change and decrease of species diversity and biomass. The
inland delta was lost eliminating spawning, feeding and
overwintering grounds for fishes.
Amongst fish species lost from the affected area were
the percids, Zingel streber and Zingel zingel, both classed
in the IUCN Red List as vulnerable, and the salmonids,
Hucho hucho, classed by the IUCN Red List as
endangered, and Salmo labrax m. fario. The European
beaver, Castor fiber, has left the territories influenced by
the dam. The mean annual fish catch has dropped by 87%.
Anti-gradient, thermal transport. North-flowing rivers such
as the McKenzie River, Canada, transport warmer water
into the Arctic than that from local tributaries. This enables
some species to range further north. .The same
phenomenon would occur in reverse in the Southern
Hemisphere. Rivers flowing towards the tropics or down
from higher cooler altitudes would permit cool water
fishes to extend their ranges.
Dams or series of dams may affect anti-gradient
thermal transport and the ranges of aquatic species which
depend on them. However the thermal transport affects
more than just the species in the river. In the Arctic the
northern limits of the tree-line, appears to be extend
adjacent to north flowing rivers, as well as next to large
lakes which act as reservoirs of summer heat. It is not clear
whether reducing the flow of anti-gradient thermally
transporting rivers, or the effects of reservoirs and
discharge from either epilimnion or hypolimnion will
affect the downstream and lateral distribution of species.
Analysis of data from Russian rivers with older dams
might be useful.
Water basin connections

When waters of one basin are diverted into another,
impacts can be expected from changes in volume and
seasonality of flow. New biota from the source basin may
invade the recipient basin and compete with the native
species. If all the water is diverted from the source basin,
this will obviously have serious impacts on any unique
species or genetically different stocks. Individuals washed
down irrigation canals, especially if there is a drop from
the impoundment, may no longer be able to return and
may not be able to maintain viable populations in the new
habitat. In the Nicola River, western Canada, considerable
numbers of Pacific salmon fry pass down into irrigation
ditches, depleting the river populations. Screen systems

have been installed in Canada and the USA.
Estuarine and marine impacts

Many of the effects in estuaries are similar to those
upstream, e.g. loss of habitat and changes in seasonal flow,
turbidity and productivity. Water withdrawal on the North
Caspian had the following effects:
the mean salinity increased from 8-11 ppt;
the estuarine mixing zone was compressed and moved
up to the delta;
the nutrient yield, especially phosphorus, and
sediment load were reduced by as much as x2.5 and
x3, respectively;
biomass of phytoplankton, zooplankton, and benthic
organisms were decreased by as much as x2.5; and
a substantial part of the Volga flood plains that served
as a nursery ground for many valuable fishes was
transformed into drying swamps or desserts.
This led to a progressive deterioration and significant
decline in natural recruitment. Commercial catches fell by
as much as three to five times for three sturgeon species,
xlO for bream (Abramis brama), p*eperch (Stizostedion
lucioperca), Caspian roach (Rutilus rutilus), and carp
(Cyprinus carpio), and nearly xl00 for the commercial
fishery of Caspian herrings (Alosa kessleri volgensis). The
sevruga, Acipenser stellatus, has been saved from
extinction by release of fry reared in hatcheries over the
last two decades. Estuarine deltas. Deltas below
impoundments tend to shrink, reducing habitats, because
of the capture of sediments by impoundments.
The Nile Delta, Egypt, has shrunk at a rate of 125 to
175 m yrl, and more saline water has invaded inland. The
Danube Delta, central Europe, shoreline is receding at a

rate of up to 17 m yr-l , threatening benefits from tourism
to birdlife_ The delta support large populations of bird
species that are generally widespread over Europe; some
170 species of birds breed in the delta, including pelicans,
herons, ibises and terns. Impoundments upstream,
including 7 major dams, -on the 2,860 km long river,
channelisation and the loss of the nutrient absorption
capacity of upstream floodlands has meant nutrients and
other pollutants are affecting delta water quality. Bird
populations are at a fraction of their historical numbers.
So although reservoirs may provide new habitat upstream
their impacts on birds may be negative in the long-term.
Entrainment
The surface outflow of freshwaters in estuaries, results in
a return current of deeper, nutrient-rich waters. These
nutrients contribute to the high productivity of estuaries.
Reduction of flow may therefore reduce import of
nutrients. There are numerous impoundments in the North
American Great Lakes and St. Lawrence River basin. It is
estimated that the spring and summer runoff at the
entrance to the Gulf of St. Lawrence has been reduced by
between one third and one half.
Many decades of anthropogenic activity have altered
the Laurentian Great Lakes ecosystem and the devastating
changes that took place in the northwestern Atlantic
ground fisheries can be related to this. Coastal fish catches
adjacent to deltas with large upstream volumes of
impoundments have declined seriously from 1950 to 1990,
e.g. the Egyptian Mediterranean to 18% and the western
Black Sea, Sea of Azov and Caspian Sea to <3% of the
original catches. Based on world-wide experience, no more
than 25-30% of the historical river flow to the estuary can
be diverted without disastrous ecological consequences to
the receiving estuary. Economic losses in the Black, Azov

and Caspian seas total about $3 billion per year.
Dams often interact with the effects of other human
activities. McAllister classes the potential interaction of
human impacts as either neutralising, additive or
synergistic He postulated that synergistic interactions,
those where harmful affects were amplified, were more
likely, considering that genomes by selection are adapted
to a certain set of undisturbed natural environmental
conditions. Species classed as endangered are more likely
to become extinct. Factors to be considered include climate
change agriculture, forestry, industry and municipal
effects.
Upstream impacts are generally less than those in the
reservoir or downstream. The exception to this
generalisation is the migratory species that move up and
downstream and use such movements to maintain genetic
diversity (in a stock or the survival of a stock) or the
survival of a species in that part of the basin, or in their
entire global range. For molluscs and fishes the change
from running to still waters (and other related conditions)
in a newly established reservoir is profound. Most species
highly adapted to currents will be extirpated and the
reservoir diversity will drop. The reservoir becomes
stocked with ecologically flexible native species or with
exotics. One exception is in African reservoirs where fish
numbers may increase. Reservoir fishery harvests usually
increase after impoundment, but then drop.
For waterfowl the situation is different. Reservoirs
provide new habitat for over-wintering in cool regions and
for residence in warm arid regions which have few natural
water bodies. Thus new reservoirs can increase waterfowl
populations, though their diversity will not be as high as
that in natural lakes. The effects of impoundment on birds
that existed in the reservoir basin and downstream of the
dam have not been well studied. The impounding of rivers

has terrestrial impacts on biodiversity.
The biodiversity of land flooded by reservoirs, and
floodplains, wetlands, oxbows and other river valley
aquatic ecosystems deprived of normal flooding may be
diminished or lost. River ecology is tied to that of estuaries
in the transport of silt, nutrients and seasonally different
volumes of river discharge. This is important in the
physical maintenance of delta and coastal habitats and the
nutrient-based estuarine food chains. The nutrient plume
of rivers can extend far out to sea. So regulation of rivers
can influence even ocean species and ecosystems.
Cumulative Effects
The addition of each new dam in a river contributes to
the fragmentation of habitat and separation of populations.
Gene flow, hitherto bidirectional, becomes unidirectional,
downstream, reducing genetic diversity. Each new dam
also prevents natural restoration of upstream populations
lost through natural or anthropogenic causes. One of the
biggest cumulative impacts may be that a greater
proportion of running water is converted to still reservoirs
habitat. Table 6 shows that there are 10 basins with 6 to
14 major dams. The Itaipu Reservoir, Brazil, is sited below
a floodplain and hence enhances migratory fishes. The
species inhabit the floodplain, then, when mature, migrate
down into the reservoir. However, the floodplain will
disappear when a new dam being built will cause it to go
underwater.
Canada, river basin impoundments discharge 50%
more water in winter than in the pre-dam era. This has a
number of ecological effects in the estuaries and seawards.
If the dams reservoirs are used for irrigation water supply,
then the volume of flow will become progressively
attenuated, as in the Colorado River, USA, where the
mouth is virtuallv waterless.
9
Agricultural Ecosystem
Population Action International recently overlaid global

population data with maps delineating Conservation
International's 'biodiversity hotspots' - areas holding 44
per cent of the world's vascular plant species and 35 per
cent of its bird, mammals, reptiles and amphibians. Their
analysis showed that 1.1 billion people live in the 25
biodiversity hot spots, most of which have higher
population growth rates than the global average. FAD and
other data on rates of malnutrition indicate that at least a
fifth of all malnourished people live in these biodiversity
hotspots.
Farming is the principal livelihood of most of these
people, and in low-income biodiversity-rich countries, it
is a major engine of economic development. While
protected areas are necessary in these biodiversity
hotspots, and elsewhere, they are not sufficient. Additional
approaches are needed.
Over a third of the global agricultural extent is in
high-intensity systems that generally use high levels of
agrochemicals for continuous cropping, and often reshape
land and waterways. The rest of the agricultural extent is
under extensive farming systems that use far fewer inputs,
but require relatively large expanses of land to produce

relatively low crop and livestock yields.
Agriculture is necessary to feed people, but both
broad types of agriculture have had notable negative
impacts on wild biodiversity:
Nearly half of all temperate broadleaf forest and
tropical and subtropical dry forest, and a third of
temperate grass and shrubland, have been lost as
wildlife habitat, through conversion to agricultural use;
conversion rates are especially high in Asia and
Europe.
Irrigation is practiced on over 250 million hectares, and
uses over 70 per cent of all freshwater - 89 per cent
in low-income countries, often diverting water
resources needed by land-based and aquatic wildlife.
Globally, over half of wetlands-among the planet's
most valuable wildlife habitats-have been converted
to agriculture.
Farming has led to significant soil degradation on 16
per cent of all crop, pasture and forestland worldwide,
and half of all land within the agricultural extent,
thereby affecting the diversity of soil microorganisms.
Excessive use and poor management of crop nutrients,
pesticides, and penned livestock wastes are a· major
cause of habitat pollution that can kill wildlife directly
or impair reproduction.
Such evidence suggests a need to redouble efforts to
establish protected areas 'off limits' to agriculture. But this
is not enough. Of over 17,000 major sites already devoted
to conserving wild biodiversity, 45 per cent (accounting
for 20 per cent of total protected land area) have at least
30 per cent of their land used for agriculture. Most of the
rest are islands within a 'sea' of agriculture.
Agricultural Ecosystem 153
Some ecologists calculate that even if the existing

protected areas do continue as wildlife habitat, 30-50 per
cent of their species may still be lost because such isolated
protected areas do not contain large enough populations,
especially of large species with relatively low populations,
to be viable. FAO statistics indicate that only 12 per cent
of global land area is in agriculture, and previous analyses
from remote sensing was consistent with this, as these
defined land units as 'agricultural' only if crops or planted
pasture covered 60 per cent of the land unit analysed.
Another 40 per cent is in grasslands, of which much is
used for grazing of domestic livestock. Thus the scale of
agricultural impact is much greater than had previously
been recognised. In many countries, as much as 70 per
cent of land area is in agricultural use.
In Europe and North America, wealthy urban
populations are able to transfer large financial payments
to their small farming popUlations to take land out of
(surplus) production to preserve as wildlife habitat or to
provide financial incentives for conservation farming. But
in poor countries with large rural populations, this
approach is viable only in a few selected areas receiving
generous foreign assistance, or where protected areas also
provide highly valued environmental services (such as
water and tourism) to urban populations.
Elsewhere, environmental planners must rely upon
local support for conservation efforts. While protected
areas are still required, and need to be expanded, they
ultimately will be successful only if surrounded by
production systems of high environmental value that are
also economically viable. Ecoagriculture encompasses two
sets of strategies for land and resource management. First,
it increases wildlife habitat in non-farmed patches in
agricultural landscapes, creating mosaics of wild and
cultivated land uses, by:
1) Creating new protected areas that also directly benefit

local farming communities (by increasing the flow of
wild or cultivated products, enhancing locally valued
environmental services, or increasing agricultural
sustainability),
2) Establishing habitat networks and corridors in
nonfarmed areas (such as hedgerows or windbreaks
that are compatible with farming).
3) Raising the productivity of existing farmland to
prevent or reverse conversion of wild lands (where
that is possible, given tenure, labour and price
conditions; and efforts to protect or restore the
biodiversity value of uncultivated lands are also
undertaken). Second, ecoagriculture enhances the
habitat quality of productive farmlands, by:
4) Reducing agricultural pollution through new methods
of nutrient and pest management, and farm and
waterway filters;
5) Modifying the management of soil, water and natural
vegetation to enhance habitat quality; and
6) Modifying the mix and configuration of agricultural
species to mimic the structure and function of natural
vegetation.
A joint study by IUCN and Future Harvest identified at
least 36 examples of ecoagriculture, from diverse regions
of the world and types of farming systems, have been
documented to have significant positive impacts on
wildlife populations, farm yields and farmer income. A
quarter of these are already being practiced on millions
of hectares (including wildlands re-established as a result
of crop intensification on a smaller area; integrated pest
management and organic production to reduce pesticide
pollution; minimum tillage in mechanised cropping; trees
grown in pastures; and species-rich agroforests). The rest

are being used on a smaller or pilot scale.
To have a meaningful impact on biodiversity
conservation at global or regional scales, ecoagriculture
must be developed and promoted for many more systems,
on far larger areas. Agricultural and environmental policies
need to be modified to encourage these new approaches.
In some cases, ecoagriculture systems can be developed
by using available components and information from
scientific and local knowledge, and by improving these
through trial and error to design landscapes that address
both local livelihood and conservation objectives.
But in most cases major scientific initiatives will also
be required, using sophisticated methods and tools from
various disciplines. Indeed, ecoagriculture is feasible now
in large part because of our greater capacity to find
synergies through scientific management. Advances in
conservation biology, agricultural ecology, plant breeding,
ecosystem monitoring systems and modelling are
revolutionising our ability to understand and manipulate
wildlife-habitat-agriculture interactions.
For example, recent reseatch on cotton, maize and
tobacco has demonstrated the potential for farmers to assist
plants in manipulating predator-prey interactions through
allelochemicals that activate plant defence genes that
attract the predators of their insect pests. Completely new,
low-cost and environmentally benign pest control systems
could be developed based on this basic research. New
technologies, supported by needed policy changes, are
enabling the design of farming systems and landscapes
supporting ecoagriculture. For example:
Using new methods to monitor wildlife and· analyse
patterns of 'countryside biogeography', conservation
biologists have been able to determine spatial and
temporal movement patterns and territorial
requirements for wildlife. These are enabling the

design and placement of corridors and habitat patches
in farmlands, and spatial configuration of wild and
domesticated plant species within farms, for cost-
effective wildlife conservation. Local farmers can
organise themselves effectively to play a lead role in
designing landscape and farm interventions.
Promising examples include LandCare groups in
Australia, farmer federations in the Phlippines, and
forest user groups in Nepal.
The use of analytical spectrometry with remote
sensing has enabled scientists to identify sources of
nitrogen- and phosphorus-rich agricultural sediments
in Lake Victoria, that feed water hyacinth (an invasive
alien species) and cause turbidity and loss of native
aquatic biodiversity. These data are being used by
public agencies and farmer groups to target
revegetation and conservation programmes.
Scientists working in West Africa developed a natural
biocide, from a strain of an environmentally friendly
fungus (Metarhizium anisopliae), that was successful
in controlling grasshopper and desert locust pests that
were devastating grain crops in West Africa, and
greatly reduced the need for insecticides that had been
threatening stork and songbird populations. Field-
based research and monitoring programmes are
essential elements for success of such efforts.
Veterinary research to develop a livestock vaccine
against rinderpest, a viral disease, has not only greatly
protected domestic cattle in East Africa, but also
protected millions of wild buffalo, eland, kudu,
wildebeest, giraffe and warthog that share rangelands
and reserves, and that are also susceptible to the
disease. New park zoning and use regulations, as well
as communications systems with local herders, are
needed for successful co-management.
Crop breeders in the U.S. are developing native

perennial grains (such as bundleflower, leymus,
eastern gamagrass, Maximilian sunflower) that can be
grown more sustainably with much less
environmental damage in dryland farming regions.
The systems are not yet econo~cally competitive, but
yields have reached 70 per cent of annual wheat
varieties, while production costs are lower; habitat
value for wildlife is many times higher than in
conventional wheat fields. (Pimm, 2000) Promoting
these species will require changes in agricultural
subsidy policies.
In the humid tropics, research has demonstrated the
benefits for both sustainability of, production and
biodiversity conservation of farming systems that
'mimic' the structure of the natural forest ecosystems.
Millions of hectares of multi-strata 'agroforests' in
Indonesia produce commercial rubber, fruits, spices
and timber, often in a mosaic with rice fields and rice
fallows. The number of wild plant and animal species
in these agroforests are often nearly as high as in
natural forests. Maintaining these systems involves
policy reforms to strengthen farmers' tenure claims,
and 'level the playing field' with subsidised rice
production.
In Central America, researchers are developing
modified systems of shaded coffee with domesticated
native shade tree species, that maintain coffee yields
while also diversifying income sources and conserving
wild biodiversity. Farmer adoption of these systems
has been promoted through changes in public coffee
policy to favour shade systems, technical assistance,
and in some cases price premiums in international
markets for certified 'biodiversityfriendly' coffee.
Private food processing companies that obtain a large
share of their raw material from smallholder farmers
located near protected areas may be motivated to

encourage ecoagriculture (e.g., current trends to reduce
agrochemical use in cocoa production). Private agricultural
service companies might, for example, sell pest control
services to farmers based on ecoagriculture principles
rather than simply selling them products. Private tourism
industry that benefits from wild biodiversity may be
willing to help support ecoagriculture.
But public sector institutions and civic organisations
will have to play a leading role in ecoagriculture
development, simply because so much of the necessary
research and investment is in support of providing 'public
goods'. In many parts of the world, wildlife conservation
organisations will need to take the lead in developing
ecoagriculture strategies and contracting for targeted
research to support those strategies, as is already being
done in many U.S. organisations.
The Consultative Group on International Agricultural
Research (CGIAR), Global Environment Facility, the United
Nations Foundation and other international donors can
lead in funding ecoagriculture research and development
in and around globally important protected areas, such as
World Heritage Sites and Biosphere Reserves; such an
effort is already under development in Kenya.
However, this new challenge emerges just at a time
when public resources for agriculture are declining.
International aid for developing-country agriculture has
declined almost 50 per cent in real terms between 1986
and 1996. In relation to their agricultural production,
developing countries spend, on average, only a fifth as
much as more developed countries on agricultural research
and development. The Future Harvest Centres - a network
of 16 international food and environment research
institutions supported by the CGIAR - is well positioned
to work on these issues, but faces stpgnating resources
even as their mandate has expanded to address

agricultural sustainability and biodiversity conservation.
To accelerate ecoagriculture development in 'hotspots'
for both biodiversity and rural poverty, several important
steps are needed:
1) Develop and fund (from sources not already being
used for agricultural research or biodiversity
conservation), a Global Programme for Ecoagriculture
Research and Development, in selected biodiversity
hotspots. These could focus on collabourative efforts
by the Future Harvest Centres, public agricultural
institutions, conservation organisations, and farmer
organisations to develop operational systems, backed
by socioeconomic analysis and supportive policy
interventions (US$50 million per year over 5 years);
2) Undertake international and national policy research
to determine cost-effective market, legislative and
institutional interventions to promote ecoagriculture
on a large scale ($10 million per year over 5 years);
3) Develop networks that link farmer organisations with
technical specialists in agriculture and environment
who work in particular habitat types, through
websites, e-workshops, and field tours in biodiversity
hotspots of mutual interest ($10 million per year over
5 years);
4) Establish a budget line item in national scientific
institutions in industrialised countries and in
'megadiversity' countries such as Brazil and Indonesia
that have strong agricultural research capacity, to fund
basic research in biodiversity hotspots, on interactions
between agricultural systems and wildlife habitat and
species in landscape ecology, agricultural ecology, and
wildlife behaviour etc ($20 million per year over 5
years);
5) Develop programmes to educate farmers,

agriculturalists and policymakers in key elements of
ecosystem management, and to educate wildlife
biologists, ecologists and conservation policymakers in
key elements of agricultural resource management ($10
million per year over 5 years).
In a world where human population may reach 9 billion
by mid-century, it is not enough to 'leave wildlife alone';
'wild lands' must be actively managed as we already do
our agricultural lands. With such compelling evidence on
the vulnerability of wildlife to agricultural expansion and
intensification, and the dependence of much of the world's
poor on agricultural development, ecoagriculture has
become a pressing policy and research priority.
AGRO-BIODIVERSITY: FUTURE OF AGRICULTURE SYSTEM IN

INDIA
India is one of the world's largest and oldest agricultural

societies, one which as remained predominantly rural
despite decades of modernisation. Even today, every aspect
of the country's economy and polity, and thE day-to-day
lives of the majority of its 900-million population, are
governed by what happens in the agricultural sector. The
stability and sustainability of its agriculture is therefore
of paramount importance.
Like many large tropical countries, India is
characterised by a complex mosaic of distinct agro-
ecosystems, differentiated by their climatic, soil, geological,
vegetational, crop-growing, and other features. A recent
classification by the National Bureau of Soil Survey and
Land Use Planning distinguishes 20 broad agro-ecological
zones, separated by natural features and crop growing
periods. Each of these agro-ecological zones is in tum
comprised of myriad micro-habitats. It is within this
diversity of habitats that an amazing variety of crops and
livestock has been developed over the millenia, by Indian

farmers.
Species which may have originated exclusively in
India include mango, taro, cucumber, pigeonpea, pepper,
eggplant, and cardamom. While the species diversity
among Indian crops is significant, what is truly mind-
boggling is the genetic diversity within each of these
species. To give some examples, one species of rice has
been diversified into at least 50,000 distinct varieties. One
species of mango has yielded over 1000 varieties, ranging
from the size of a peanut to a musk melon; a similar figure
is estimated in the case of taro.
India also has amongst the world's largest diversity
of domesticated animals, with some 26 breeds of cattle,
40 of sheep, 20 of goats, 8 of camels, 6 of horses, and 18
of poultry, apart from the yak, the mithun, and several
species and breeds of birds including geese, ducks,
pigeons, and doves. It is noteworthy that the
characterisation of Indian livestock breeds was last done
in the first half of this century. Since no recent estimates
are available, and surveys in some regions are far from
complete, some scientists feel that the diversity may be
even greater.
Over centuries, Indian farmers have continuously
adapted and modified the rich genetic material available
to them from nature. The diversity of crops and livestock
is not only accidental, nor is it purely natural; it is more
the outcome of thousands of years of deliberate selection,
planned exposure to a range of natural conditions, field-
level cross-breeding, and other manipulations which
farmers have tried out. In other words, a single species of
rice collected from the wild some time in the distant past,
has diversified into 50,000 varieties as a result of a
combination of evolutionary/habitat influences and the
ingenuity and innovative skills of farming communities.
This latter contribution to genetic diversity is a fact that
the modern seed industry always conveniently sidesteps,

and that the non-discerning consumer is ignorant of.
There are sheep which are adapted to the harsh
summers of the west Indian desert, and others which can
survive the equally harsh winters of the Himalayan tracts.
There are cow breeds which thrive in the humid hills of
the Western Ghats, while other breeds produce well in
the driest regions of western India. In the Garhwal hills
of the Himalaya, over 40 crop species and numerous
varieties are grown, a diversity which is maintained
through diverse cropping patterns, and which has evolved
in the context of wide variations in edaphic, topographic,
and climatic conditions, coupled with careful selection by
farmers.
This process of adaptation continues even today.
Livestock scientists recently found that migratory
pastoralists in Rajasthan had selected for, and helped
develop, a new breed of sheep, called kheri, in response
to the increasing drought incidence and declining pasture
availability. Adaptation to localised environments has been
only one mechanism or reason for diversification. What is
even more striking is the use of a large diversity of the
same crop within a single village, and sometimes within
the same field.
Many tribal villages in the hills of north-east India
have been known to grow over 20 rice varieties within a
single year in their terraced fields. In one region of
Koraput district of Orissa alone, scientists identified over
1500 rice varieties. Within a single village, domesticated
diversity can be spread over time (seasonal) and space
(geographical), over vertical and horizontal layers within
the same field, and within and between species of plants
and animals.
Many traditional communities practiced a form of
agroforestry which combined trees with crops and animal
husbandry. In the high-rainfall areas of Kerala and

Meghalaya, tribal communities maintained tiny home
gardens which gave them wood, medicinal plants, spices,
and ornamentals, apart from food (Santhakumar 1996);
systems of shifting cultivation (jhum) in north-east India
encouraged the use of a large diversity of crops, upto 35
species within a single cultivation cycle in some cases.
Apart from physical and biological adaptation, a host
of economic, cultural, religious, and survival factors have
played a role in this diversification. For instance, the late
scientist Winin Perreira notes, amongst the Warli tribals
of the west Indian state of Maharashtra, a great diversity
of rice grown for different water and soil needs, varying
maturity periods, resistance to different diseases, and
various cultural events.
Several varieties of rice and other crops were grown
in many parts of India just for their use during festivals,
marriages, or other auspicious occasions; several others
were grown for their taste, colour, or smell; yet others for
their pesticidal or soil-fertilisation characteristics.
Diversification also provided buffer food output in times
of drought, flood, or pest attack, when the main crop
might fail.
The stability of a biodiverse agriculture is perhaps its
most important characteristic, as recorded from many parts
of the world. This is wonderfully illustrated by a once-
common practice of the Garhwal Himalaya, the baranaja.
Literally meaning '12 grains', this practice involves the
sowing of a mixture of crops into a single plot of land.
Rajma (beans, Phaseolus vulgaris), urad (black gram, Vigna
mungo), mung (green gram, Vigna radiata), kulth or gahat
(horsegram, Macrotyloma uJ;liflorum), marsha or ramdana
(Amaranthus frumentaceous), mandua (finger millet,
Eleusine coracana), jhangora (barnyard millet, Echinochloa
frumentacea), bhat (soyabean, Glycine soja), lobiya (Vigna
catiang), and other crops are grown in a jumbled profusion
which at first glance would appear a mess, but which

probably a carefully considered way of obtaining optimal
and sustained yields.
Since maturity periods of these crops vary, different
crops are harvested at different times, helping to retain
soil moisture, and providing a constant supply of food.
Fertility is continuously recharged by the use of
leguminous plants like pulses. In addition, bunds along
the fields support trees like bhimal, used for making rope,
soap, baskets, and for fodder. According to some
assessments, baranaja gives a higher overall productivity
(apart from meeting diverse needs) than if the field was
to be converted into a soyabean monoculture, which is
being propagated by official agricultural agencies in the
region.
Apart from cultivated crops, all communities also
used a variety of wild plants (and often, animals) for food
and other uses. Not many studies have been carried out
of the use of uncultivated plants, but where conducted,
they have revealed that these are a major source of
nutrition, medicine, and other requirements. Most
important, they provided critical inputs when cultivation
failed.
Erosion of Agro-Biodiversify
However, since traditional agricultural systems were finely
intervowen with the social and cultural fabric of villages,
as also with the forests and other ecological features within
which the villages existed, they could not withstand the
far-reaching changes in land-use, taxation, forest policy,
and administrative structures brought about by the colonial
government in the 19th and 20th centuries. These changes
severely disrupted traditional agriculture. But even more
dramatic changes in Indian agriculture have come in the
last few decades.
With the advent Qf the Green Revolution in the mid-

1960s, a handful of laboratory generated varieties have
been promoted over vast areas, particularly in the plains
of northern India. Given certain inputs such as irrigation
and chemical fertilisers and pesticides, these varieties
produce high yields (thus the somewhat misleading term
High Yielding Varieties, or HYVs). It is understandable
for farmers who can afford such inputs, or who have
access to bank loans or moneylenders, to take
enthusiastically to these varieties.
Agricultural schemes have also resulted in
homogenising growing conditions, for example by surface
irrigation, so that where there was earlier a complex
mosaic of diverse micro-habitats, there are now immense
stretches of uniform agricultural landscape. Inter-cropping
is replaced by monocropping, a wide diversity of species
is replaced by a handful of profitable ones, and genetic
diversity within the same crop species is replaced by a
narrow genetic range of financially lucrative varieties. The
net effect of these and other practices has been a massive
displacement of indigenous crop diversity, such that in
the case of most crops now, the majority of indigenous
cultivars are no longer grown.
The erosion in agro-biodiversity in the hills is due to
a number of factors: (a) degradation of natural forests,
which sustained traditional agriculture; (b) changing
attitudes towards coarse and fine grains, the latter being
considered more "progressive" to produce and consume,
(c) large-scale migration for employment, causing fields
to be abandoned or neglected, (d) supply of HVY seeds
and other inputs at subsidized cost by the government,
(e) attraction to maximise profits through cash crop
monocultures, and (f) lack of incentives for marketing of
traditional crops.
Livestock diversity has also faced a serious threat. It
is estimated that 10 (50%) of the goat breeds, five (almost
20%) of the cattle breeds, 12 (30%) of the sheep breeds,

and all the 18 breeds of poultry, are today threatened. The
greatest factor in the loss of domesticated animal diversity
has been deliberate cross-breeding with exotics, carried out
extensively by the government in order to increase the
yields of milk or other animal products. Semen banks have
generally stored the semen of exotics. The policy thrust
towards maximising financial profits has severely
threatened species and breeds which were bred for a
variety of domestic uses: the Deoni cattle of the semi-arid
hills of Maharashtra, ideally suited for the climatic
conditions and fulfilling local requirements of draft, dung,
and milk, has been pushed out by 'high' milk producing
cross-breeds and exotics; the Aseel poultry, bred for meat,
cock-fighting, and some egg production, and easily
managed by poor rural women, has been displaced by
exotics which are better suited for large-scale commercial
farming; and so on.
While all kinds of livestock are affected, perhaps the
worst off is poultry; exotics now make up 80% of the total
poultry population, with disastrous effects on indigenous
breeds. The current thrust towards export-oriented poultry
production is likely to intensify the loss. Other factors
which have caused an erosion in agricultural biodiversity
include:
1. The destruction or conversion of habitats to which
breeds or varieties were specially adapted, and the
disruption of traditional lifestyles, through urban
migration and through displacement by development
projects.
2. Changing social and religious norms, and cultivation
methods, which threaten the genetic diversity of crops,
especially cereals, pulses, vegetables, and plants used
for religious and social purposes.
3. Intense grazing activity by cattle, which has depleted

wild cereal grasses, vital sources of genes for the
improvement of existing crops.
4. The clearing, in modem agricultural practice, of bunds
and hedgerows, which once served as respositories of
wild and semi-wild genetic diversity of crop and
animal species.
5. The subtle (and not-so-subtle) changes in food habits;
everywhere, people have been brainwashed into
believing that wheat and rice are the only two cereals
worth eating.
This erosion of agricultural biodiversity threatens the long-
term stability and sustainability of Indian agriculture itself,
in many ways:
1. It erodes the genetic base on which scientists are
depending for continuous improvement of crops and
livestock. The majority of HYVs themselves have been
developed from genetic material taken from traditional
varieties and wild relatives of crops. These HYVs, in
particular hybrids, are not very long-living: they tend
to lose their viability and productivity, or become
increasingly susceptible to pest/disease attacks, within
a few years. This necessitates the infusion of fresh
genetic material, which is again obtained from existing
traditional varieties or from wild plants. But then the
introduction of these HYVs is itself a major cause of
the erosion of traditional crop diversity. Some of this
diversity can be stored in gene banks and accessed
when needed (see below), but there is an inevitable
loss even in such storage, and the continuous
evolution of new varieties which was taking place on
the farm is no longer happening. As has been said,
modern agriculturists are somewhat like masons
building the roof of a house by taking the bricks from
the walls!
2. 'The failure of a single HYV crop due to any natural

calamity is a crippling blow for a farmer who has no
other crop to fall back on, unlike traditional agriculture
where some fall-back crops were also grown. Some
degree of security against such eventualities can be
artificially achieved by measures like protective
irrigation, subsidies, and credit schemes, but such
measures are expensive and prone to failun', In the
Garhwal Himalaya, for instance, data for the periods
1970-74 and 1990-94 show that yields of most
traditional food crops remained stable, and that the
recent food insecurity or shortage problem is largely
due to the decline of these crops. For the country as
a whole too, the increasing reliance on a narrow
genetic range of crops represents a high-risk
proposition.
3. Both the above features result in an increasing
dependence of the farmer on the industry-dominated
market and the government. Virtually all inputs for
farming, except land and family labour, is now
obtained from outside the village: seeds, irrigation,
fertilisers, pesticides, credit. And despite huge
subsidies on these inputs, as also support prices and
the like, an increasing number of farmers are facing
the economic treadmill, spending more and more to
achieve the same output. Some commentators have
observed, somewhat controversially, that at least part
of the unrest in places like Punjab and eastern Uttar
Pradesh is because of the frustration of farmers
trapped by the short-term lure of the Green
Revolution.
4. Several other effects of modern farming have brought
insecurity in the lives of farmers. For instance, the
traditional paddy field in north-eastern, south-western,
and central India provided not only rice but also fish,
frogs, and other elements of biodiversity which were
an important part of the diet of several communities,

especially tribals. Modem paddy fields, which require
large amounts of chemical fertilisers and pesticides,
are devoid of much of this biodiversity, with a
resultant loss of nutrition for farmers. Similarly, in the
Western Ghats, e.g. of Kerala, farmers grew a profuse
mix of fruit trees and food crops on slopes, along with
paddy in the valleys; the former is now increasingly
being replaced by plantations of single cash crops like
tea, so that there is a heavy dependence on the market
for food requirements.
Economy and Globalisation

The 1990s have seen a major shift in India's economic
policy, away from goals of self-sufficiency and public
sector towards an 'open' economy and privatisation. While
there was undoubtedly a need to reduce bureaucratic
hurdles and top-heavy administration, the new economic
policies appear to be throwing out the baby with the
bathwater. In the field of agriculture, there has been an
aggressive thrust towards commercialisation, especially to
feed an insatiable export market. Cash cropping, already
a threat to the small-scale biodiverse farm, has been given
a major boost.
New trends include floriculture, industrial
aquaculture, and other forms of intensive farming which
leave little scope for biologically diverse production
systems. Perhaps most devastating are recent moves by
the Indian government and some state governments to
relax land ceiling and other regulations which restricted
the conversion of agricultural into non-agricultural land.
The intent is clear: make it easier for industrial level
agriculture, or even non-agricultural land uses such as
industry, to acquire land.
The parallel move towards more high-technology

agriculture makes the country as a whole also more
insecure, as it increases its dependence on biotechnologies
controlled by industrial countries and multinational
corporations. The entry of Cargill, Cieba- Geigy, Monsanto,
McGain and other globally powerful companies into
India's seed and agro-products sector is a major step
towards this crippling dependence, and a direct reversal
of policies which had all this while tried to take us towards
self-reliance.
The benefits and risks of genetic manipulation and
other biotechnologies are currently being debated all over
the world. Acute focus has been on the controversy
surrounding the so-called "control of plant gene
expression" technology which will make second generation
seeds sterile, and thereby force farmers to buy seeds every
time from the manufacturer. Apart from the ecological
risks, one aspect which has been less focused on is the
fact that most such biotechnologies are likely to remain
beyond the reach of the small farmer (and for that matter
the small seed manufacturer, who along with farmers
themselves still serve the majority of the seed needs of
Indian agriculture). These developments can only deal a
further blow to farmer self-sufficiency.
This process of homogenisation of agriculture, and
increasing dependence on alien agencies and technologies,
is likely to be greatly intensified with the implementation
of the General Agreement on Tariffs and Trade (GATT),
concluded in 1993. This requires countries to greatly "open
up" their borders to both imports and exports, and
substantially reduce governmental controls and
interventions.
Furthermore, the provisions in the Trade-Related
Intellectual Property Rights (TRIPs) part of GATT,
especially those seeking to harmonise IPR regimes across
the globe and to enforce patentability of life forms, are
forcing changes in the Indian Patent Act and related

legislation. This could have severe implications for
biodiversity and farmers' rights. IPRs are expensive, and
corporations would try to push their protected seeds over
as wide an area as possible to recover costs and make
profits. Further displacement of traditional diversity and
homogenisation would result. Additionally, innovations by
farmers, which result in expanding diversity, may be
hindered if IPR regimes favour the formal sector breeder
at the cost of the farmer.
Conservation Opportunities
A considerable amount of the genetic material which has
been grown or bred by farmers may no longer be available
in the field, but has been collected and stored in gene
banks and breeding stations. The National Bureau of Plant
Genetic Resources and the Indian Council of Agricultural
Research, in their network of gene banks, have several
hundred thousand accessions. Such ex-situ collections are
important, as they are able to store material which may
no longer be possible to grow in the field, and as they
make available the base material for genetic up.sradation
of agriculture.
But such collections also suffer from severe
limitations: they are very expensive, lack adequate space
to store the complete genetic diversity found in agriculture,
and suffer loss of viability of stored germplasm. They also
freeze evolution, since the environmental conditions which
crops are constantly adapting to cannot be recreated in
the icy chills of the gene bank. Finally, they have various
political problems associated with them; on the one hand,
farmers experience considerable difficulty in accessing the
genetic material, and on the other, there is relatively easy
access to formal sector breeders and corporations who use
the material for commercial benefit.
For this and other reasons cited above, there is no

alternative to the conservation and continued use of crop
and livestock diversity in- situ, i.e. on the farmers' fields
and the pastoralists' rangelands. Unfortunately this aspect
has been almost completely ignored in governmental
programmes, with the exception of some efforts to
encourage continued use of traditional livestock breeds.
However, in- situ conservation of crops is finding
increasing attention in the work of community
organisations and NGOs. Farmers in many regions are
beginning to compare their indigenous biodiverse forms
of agriculture with the modern monocultures, and at least
some of them are coming to the answer that a revival of
the former, with appropriate modern inputs where
necessary, is preferable to running on the economic
treadmill of the latter.
Actually, a revival or development of a biologically
diverse agriculture is eminently possible in India, since
the destruction of traditional diversity is not yet
irreversible. Consider the following:
1. The Green Revolution technology has not spread to
many parts of the country, for several reasons,
including its exhorbitant costs, and lack of appropriate
packages for so-called 'marginal' areas (mountains,
flood-prone areas, arid zones .... ). This means that a
lot of traditional agriculture still survives, retaining
with it considerable diversity of crops and livestock,
and the knowledge and practices associated with
them. The example of Tehri Garhwal given below is
illustrative.
2. Even where new HYV crops and cross-bred or exotic
livestock breeds have been introduced, in many areas
they have failed to produce the necessary results, or
have not performed to the satisfaction of farmers. This
is especially true of 'marginal' areas. In many cases,
therefore, farmers have reverted back to their
indigenous varieties, or continue to grow these

varieties along with the HYV ones, as insurance
against the failure of the latter.
3. There is a certain resilience to change (what
agricultural scientists call "stubbornness" or
"backward mentality") amongst Indian farmers, which
has helped to retain elements of traditional diversity
and practices even in areas where the Green
Revolution has been aggressively pushed.
4. There is a tendency of many farmers to grow HYVs
for the market, but their traditional varieties for home
consumption. This has been found in areas which are
converting to intensive modem farming in Rajasthan,
Andhra Pradesh, the Himalayan foothills in Uttar
Pradesh, and elsewhere. Agricultural planners would
call this "double standards", but the farmer is simply
combining the possibilities of earning good
renumeration (made possible more by an economic
system which subsidises and favours these HYVs than
by any inherent characteristic of HYVs), with the
personal desire to eat healthy food. Again,
considerable on-farm diversity may exist because of
this.
10
Bacterial Biodiversity
Humans have long been fascinated by the extraordinary

diversity of life on Earth. Not only is the sheer diversity
of living creatures intriguing, but there are also striking
patterns in their distribution over space and time.
However, most of what we know about the origin,
maintenance and distribution of biodiversity stems from
research on plants and animals. Although there may be
millions of species of bacteria, we are only beginning to
investigate patterns in their diversity and the forces that
govern these patterns.
Understanding patterns of bacterial biodiversity is of
particular importance because bacteria may well comprise
the majority of the Earth's species diversity, they mediate
many of the environmental processes that sustain life on
Earth and their diversity is of great applied importance in
bioremediation (the biological degradation of pollutants)
and bioprospecting (the search for novel biochemicals for
use in medicine and industry).
There are many unique aspects to microbial life (e.g.
parasexuality), micro-organisms and macroorganisms also
share many fundamental aspects of their biology.
Ecological patterns and principles that may be common
to micro-organisms and macro-organisms. Microbiologists
Bacterial Biodiversity 175
and ecologists alike have touted the benefits of such a

unifying approach. For example, the microbiologists
argued that a synthetic view of microbial, plant and animal
ecology was not just possible, but crucial to the
development of microbial ecology and the general study
of ecology.
Similarly, although the details of individual organisms
and ecological systems matter, ecologists would profit most
from trying to uncover underlying patterns, rules and
laws. Furthermore, they has argued that such
generalisations would be most likely to be discovered at
very small scales (e.g. populations) and at very large scales
(e.g. aggregate patterns in the distribution of biodiversity).
BACfERIA: UBIQUITOUS AND DIVERSE
One of the most striking ecological aspects of bacteria is

their ubiquity. Bacteria inhabit an extraordinary array of
habitats, from those that offer ideal conditions for most
living creatures to those too extreme to support most life
forms. They inhabit the relatively benign and nutrient-rich
environments of soils, lakes, oceans and other organisms,
but they are also found in extreme environments such as
hot springs, nearly saturated salt brines, acid mine waters
at pHs near zero, deep in Antarctic ice and kilometres
below the Earth's surface.
In fact, bacteria (and their fellow prokaryotes, the
archaea) define the outer environmental limits of life. For
example, Thermus aquaticus, the bacterium renowned for
its contribution of Taq polymerase to molecular biology,
has an optimum growth temperature between 70°C and
79°C, a temperature lethal to most plant and animal life.
Not only are bacteria everywhere, but they are also
incredibly abundant (table 1). The total number of bacteria
on Earth may be as high as 4-6x1030, with the largest
proportion of bacterial cells possibly residing in the oceanic
and terrestrial subsurfaces (3.5 x 1Q30 and 0.25-25 x 1030,

respectively;). Perhaps most notably, bacterial cells are
estimated to contain, in total, 350-550 Pg of carbon, up to
60-100% of the total carbon found in plants, as well as
large amounts of nitrogen and phosphorous. Despite their
modest size as individuals, as a group these organisms
not only contribute to the flow of nutrients worldwide,
but may also constitute a significant proportion of the
nutrients in living biomass.
Table 1. Prokaryotic abundance and diversity.
habitat abundance diversity

(cells cm-3) (genome equivalents)
Forest soil 4.8 x 1(JJ 6000
Forest soil
(cultivated isolates) 1.4 x 1(Y' 35
Pasture soil 1.8 x 1010 3500-8800
Arable soil 2.1 x 1010 140-350
Pristine marine
Sediment 3.1 x 1(JJ 11400
Marine fish-farm
Sediment 7.7 x 1(JJ 50
Salt-crystallizing
Pond, 22% salinity 6.0 x UY 7
The diversity of bacteria is also unparalleled in the

biological world. Until recently, it was impossible to detect
the vast majority of bacterial species. Owing to the
limitations of traditional detection techniques that require
growth of organisms in the laboratory, less than 1% of all
bacterial species have been described to date. However,
recent biomarker-based techniques, such as DNA analyses,
have enabled the detection of non-culturable species and
have allowed a more complete and detailed picture of
bacterial communities to be developed.
The most common of these techniques uses ribosomal

gene sequences as indicators of bacterial diversity,
although other genes, including protein-coding genes, have
also been used. The use of these molecular techniques has
been reviewed in detail elsewhere. Current culture-
independent estimates of the number of prokaryotic
species in a gram of soil range from several hundred to
nearly nine thousand, orders of magnitude greater than
estimates based on culture-dependent approaches.
Dykhuizen has speculated, based on these estimates,
that there could be 10 billion species of bacteria on Earth.
These bacteria represent a significant proportion of all
evolutionary diversity. The metabolic diversity of bacteria
is perhaps as remarkable as their taxonomic and
evolutionary diversity. Although culture-based studies are
limited in their ability to estimate bacterial diversity, they
have demonstrated that bacteria have a variety of modes
of energy conversion, wide ranges of substrate use and
unique metabolic pathways.
For example, bacteria can use kerogen and longchain
alkanes, compounds thought to be refractory, as growth
substrates under anaerobic conditions. The flexibility of
bacterial metabolism is perhaps best illustrated by the
ability of these organisms to degrade xenobiotic
compounds such as malathion (an insecticide) and 2,4,5-
trichlorophenoxyacetic acid (a herbicide), which are toxic
to many other organisms.
DISTRIBUTION OF BACTERIAL BIODIVERSITY
The advent of biomarker-based techniques has enabled

microbial ecologists to ask fundamental questions about
the distribution and abundance of bacterial diversity.
Microbial ecologists have begun to examine whether
bacterial communities are distinct in different
environments, whether bacterial diversity changes with
habitat heterogeneity and how patterns in the diversity of

bacteria compare to those of plants and animals.
Habitat Type
It has long been assumed that major environmental
variables, such as vegetation type and temperature,
influence bacterial-community composition. Recent studies
using culture-independent techniques, a select few of
which we cite here, have confirmed this assumption. The
composition of bacterial communities has been shown to
vary with land-use type, plant species, agricultural
growing practice, temperature, nutrient status, salinity,
contamination with pollutants, predation and other
environmental variables.
We understand less about how these environmental
: variables influence bacterial dive~sity, although, like the
diversity of plants and animals, bacterial diversity is
thought to be heterogeneously distributed across the Earth.
For example, there is a growing consensus that aquatic
environments support fewer bacterial taxa than do soil and
sediment environments. It is not clear which factors are
responsible for these differences, but the high
heterogeneity of the soil environment relative to that of
aquatic environments may playa role.
Habitat Heterogeneity
Habitat heterogeneity has long been posited as one of the
main determinants of biological diversity and is thought
to underlie the seemingly universal species-area
relationship, which states that the number of plant and
animal species observed increases with an increase in area
sampled. This heterogeneity can be thought of as taking
two forms:
structural heterogeneity, such as discontinuities in

space and time (i.e. 'patchiness'); and
complexity in resources, conditions and/or interacting
populations.
There are many reasons to expect that bacterial diversity
will increase in response to increases in heterogeneity;
however, observations of such patterns have been rare.
Environmental patchiness may play a role in the
maintenance of bacterial diversity.
(U!
50 8611
.'5 40
A 886
30 17331
.a5
"
20
. 10 730§ I
891
0
665
8
.2
" eh)
,...,
ill
'a
tl
"5::>
CI
:1i0
..
.~ 40
65
...
"0
'i!
30 CiIJ
.... 20
(:.5
~ 10 mJ
0
85
operational taxonomic units
Figure 1. The effect of environmental patchiness on bacterial diversity in
(a) surface and (b) saturated soils.
Soils which were saturated with water had both fewer

bacterial taxa and a more uneven distribution of bacterial
taxa than unsaturated soils (figure 1). They suggested that
this difference was the result of patchiness in the
unsaturated soil (i.e. the lack of water reduced migration
between soil particles), which allowed the coexistence of
a larger number of bacterial taxa and produced a more
uniform community composition.
Complexity in environmental conditions can vary at
multiple spatial scales and could potentially influence
bacterial diversity. Microscale heterogeneity (e.g. at the
scale of a soil particle) can be very high and potentially
allows for high microbial diversity in a relatively small
area. For example, within a single soil particle, oxygen
concentrations can range from 21% to 0% over only a few
millimetres. Such steep environmental gradients can also
be present in aquatic and marine systems, for example at
the interface of oxic and anoxic layers in stratified lakes
or at the surface of the lake sediments.
Heterogeneity in soil characteristics, bacterial density
and the distributions of some bacterial taxa have also been
observed at the microscale (from micrometers and up to
65 cm) in soils. However, to our knowledge, despite these
observations, relationships between spatial distance and
bacterial genetic or taxonomic diversity have not been
observed at this scale. At larger spatial scales, the bacterial-
community composition in soil has been observed to be
heterogeneous in some studies and homogeneous in
others.
Soil bacterial communities showed significant spatial
heterogeneity in composition at the scale of ca. 1 m, the
spatial scale over which both pH and ammonium
concentration varied. However, other studies have
observed homogeneity in community composition at this
spatial scale. Observed homogeneity may be real (for
example, the result of high rates of dispersal), or it may
be an artefact of undersampling (less common species that

vary at this spatial scale may not have been sampled).
Although community composition has been observed to
vary at this spatial scale in some studies, the relationship
between bacterial diversity and heterogeneity at this spatial
scale is unknown.
Despite observations that bacteria are heterogeneously
distributed in space and that environmental heterogeneity
can alter their distribution, the species-area relationship
has not been examined explicitly for bacteria. It is possible
that differences in characteristics such as dispersal ability
and survival under harsh conditions may result in a
qualitatively different species-area relationship for bacteria
from that for plants and animals. Alternatively, bacteria
may exhibit a similar relationship to that of plants and
animals but over different spatial scales.
Primary Productivity
Primary productivity (the rate of energy capture and
carbon fixation by primary producers) is thought to be a
key determinant of plant and animal biodiversity. Many
studies of plants and animals have reported a positive
quadratic or hump-shaped relationship between
productivity and diversity, where diversity peaks at
intermediate productivities, although other patterns have
also been observed.
There is evidence from laboratory studies that
productivity can influence bacterial diversity; however, we
are aware of only four studies that attempt to document
the relationship between primary productivity and
bacterial diversity in field systems. Two lagoons with
different levels of primary productivity. More unique
bacterial ribosomal gene sequences were retrieved from
the more productive lagoon than from the less productive
lagoon, suggesting that bacterial taxonomic richness
increased with primary productivity. In contrast, Pristine

aquatic sediments had much higher bacterial genetic
diversity than sediments below fish farms, suggesting that
increased energy decreased diversity.
Over a period of 13 days, nutrient addition both
decreased and increased bacterial diversity. Increasing
productivity both increased and decreased taxonomic
diversity of bacteria in aquatic mesocosms and that the
shape of the relationship between productivity and
diversity differed for different taxonomic groups of
bacteria. For example, they observed that the diversity of
the Cytophaga-Flavobacteria-Bacteroides group, the most
common taxon in their study system, exhibited a
significant hump-shaped relationship with primary
productivity (figure 2).
20
18
16
14
'"
~ 12
..c:
~ 10
.1.. 8
III
6
!iii
2
o~--~--~--~--~~~~~~
1.2 1.4 1.6 1.8 2.0 2.2 2,4 2.6
log chlorophyll "
Figure 2. The relationship between primary productivity and taxonomic
diversity of bacteria in five aquatic mesocosms.
By contrast, they observed a significant U-shaped

relationship between primary productivity and diversity
for the a-proteobacteria, the second most common group,

and no discernible relationship between primary
productivity and diversity for the ~-proteobacteria, the
third most common group. These initial results suggest
that bacterial diversity can vary with primary productivity
and that the nature of the relationship can, in some
instances, resemble that of plants and animals.
Disturbance and Bacterial Diversity

Disturbance, commonly defined as 'any relatively discrete
event in time that disrupts ecosystem, community, or
population structure and changes resources, substrate
availability, or the physical environment', is known to be
an important influence on the diversity of plants and
animals. It has been suggested that diversity may peak at
intermediate intensities or frequencies of small-scale
disturbance.
Ecologists have tested this hypothesis for a variety
of taxa in a number of different environments and have
found that, in general, this intermediate-disturbance
hypothesis holds. Laboratory studies suggest that the
hypothesis can, in fact, be applied to bacteria. Several field
studies have also examined the relationship between
disturbance and bacterial diversity, most of which
addressed the effect of either chemical or physical
disturbance of soil on bacteria.
For example, bacterial diversity along a gradient of
mercury contamination in fields and found that sequence
diversity decreased with increasing exposure to mercury.
The diversity of mercury-resistance genes in sedimentary
bacteria at sites with varying levels and histories of
mercury contamination. They found that the sites exposed
to intermediate levels of mercury had the highest genetic
diversities, while the most heavily contaminated site and
pristine sites had low diversity. Although these studies

seem to suggest a relationship between disturbance and
bacterial diversity, disturbance in these studies is always
confounded with other factors (e.g. vegetative cover, land
history, soil structure) that may influence bacterial
composition and diversity.
Furthermore, few studies systematically examine how
the frequency, intensity or size of disturbance affect
bacterial diversity. Bacterial-community composition and
diversity may respond to gradients in disturbance, but it
is currently difficult to assess the applicability of the
intermediate-disturbance hypothesis to bacteria outside the
laboratory.
1.0 4
0' .I
ON (0 m JBN ISO m,I JD ( 80 Lml rogion(t),19 ;:,2 x W'kml

geographical distance
Figure 3. Relationship between geographical distance and genetic distance
for fluorescent Pseudomonas isolates.
Geographical distances are based on the distance from a

reference site (BN or JD). Error bars indicate the range of
the values of genetic distance. The question mark indicates
that genetic distance may have continued to diverge if the
method's resolution did not saturate. Grey bars, observed
values; dotted line, suggested values.
Biogeographical Patterns
Biogeographical patterns (i.e. the geographical distributions
of organisms over the Earth in both space and time) have
been relatively well documented for plants and animals.
Whether bacteria exhibit biogeographical patterns has been
controversial for much of the last century. The traditional
view among microbiologists, dating back to at least the
1930s, is that high dispersal rates result in cosmopolitan
distributions and a lack of biogeographical patterns for
free-living micro-organisms. This view has been reinforced
by some field studies of eukaryotes.
However, recent research suggests that at least some
bacterial taxa can exhibit biogeographical patterns. For
example, 258 isolates of the bacterial genus Pseudomonas
from 10 sites on four continents. They analysed the isolates
using three molecular methods that provided different
levels of resolution. At the finest level of resolution, there
was a statistically significant association between genotype
and geographical distance, with evidence of migration
within sites but not between sites.
Similarly, a recent survey of the genetic diversity of
cyanobacterial communities from thermal springs in North
America, Japan, New Zealand and Italy showed that
geographical isolation was present at both global and local
(km) spatial scales. The strongest evidence to date for
biogeographical patterns in prokaryotic organisms comes
from a recent study of the hotspring archaean Sulfolobus.
The genetic diversity of Sulfolobus isolates from five
geographically distinct regions and found a significant
correlation between genetic distance and geographical
distance.
The rates of speciation, extinction and dispersal-the
three fundamental processes responsible for producing
biogeographical patterns-relative to the other factors are
likely candidates. A preliminary understanding of these
three processes for bacteria.
Dispersal
Some bacteria have the potential for very high rates of
dispersal. It is known that bacteria can be dispersed
passively in the atmosphere and through water owing to
their small size. In addition, some bacteria, such as
members of the genus Bacillus, can form hardy life stages
that are highly resistant to environmental stresses such as
desiccation; such stages could allow these organisms to
disperse widely. However, few studies have been able to
quantify even relatively small-scale dispersal and
colonisation rates. There is some evidence that some
bacterial taxa may have wide distributions, suggesting that
rates of dispersal are high.
However, most of these studies use taxon definitions
based on sequence similarity of the 16S ribosomal gene, a
very conservative definition; sequences from protein-
coding genes may provide greater resolution and be more
appropriate for inferring rates of dispersal.Only one study
that used protein-coding genes to infer dispersal rates in
free-living bacteria. The populations were sampled at
geographical scales ranging from 30 to 10 000 km.
The magnitude of the migration rate was generally
associated with geographical scale (migration was highest
between the closest sites). However, even at the largest
scale, where migration rates were lowest, the rate of
exchange was sufficient to prevent neutral geographical
evolutionary divergence (Le. the most distant populations
were not isolated enough to demonstrate genetic drift).
Speciation
The influx of new species in a given area is the product
of both dispersal and local speciation. Speciation requires
variation in both ecologically relevant traits and ecological
opportunity, and the rate of speciation may be
substantially higher in bacteria than in larger organisms
because of differences in the generation of both variation

and opportunity.
Some bacteria exist in large populations and have
rapid growth rates under favourable conditions. As the
production of novel mutations scales with population size
and growth rate, this can generate significant genetic
variation. In addition, the breadth of physiologies present
in bacteria, coupled with their small size, enables them to
exploit a wide range of ecological oppnrtunities for
diversification.
It has also been argued that speciation rates may be
high in bacteria because of the relatively low rate of
genetic exchange between individual organisms. For
example, recombination rates at or below the rate of
mutation (rates typical of many bacteria) enable genetic
variation associated with niche diversification to lead to
speciation, and thus ecological divergence is sufficient for
speciation to occur. Although rates of genetic exchange
may be low overall in bacteria relative to sexual organisms,
some ecologically important traits, such as antibiotic
resistance and mercury resistance, are acquired through
gene transfer.
It is possible for bacteria to acquire genes from
distantly related organisms, which also has the potential
to accelerate the rate of speciation. For all of these reasons,
we might expect speciation rates of bacteria to be high
relative to those of eukaryotes. Although there is some
evidence for rapid diversification in the laboratory, there
is no evidence for rapid speciation in bacteria in the field.
Measuring rates of speciation is difficult and is
usually done through studies of the fossil record. Whether
rapid speciation can alter the magnitude and distribution
of bacterial diversity, however, depends not only on how
widespread rapid speciation rates are among bacteria, but
also on the rates of extinction and dispersal.
Extinction
Bacterial diversity may also be high owing to potentially
low extinction rates. The large population sizes assumed
for many micro-organisms may make extinction less likely.
Extinction rates may' also be relatively low for bacteria
because some bacteria have traits that allow them to
reduce the risk of catastrophic losses typical of extinction
events in plants and animals. For example, some bacteria
are known to form life stages that can survive harsh
environmental extremes, reducing the probability that
chance fluctuations in environmental conditions will drive
them to extinction.
High dispersal rates over large distances may also
reduce the chance that local environmental change will
result in extinction. Some bacteria are also able to avoid
the negative effects of competitive interactions; for
example, resistance to starvation has been documented for
some species in the laboratory. It is not clear how
widespread the traits discussed in the previous paragraph
are among bacterial taxa (and thus how likely it is that
extinction rates are actually lower for bacteria than for
other organisms such as plants and animals), and there
are no direct measures of extinction rates for bacteria in
the field.
However, it should be possible to determine whether
the relative magnitude of speciation and extinction is
different for bacteria relative to other organisms. For
example, the difference between speciation and extinction
can be estimated from gene phylogenies via lineage-
through-time plots. This comparison has not yet been
attempted.
11
Ecological Forecasting and the
Urbanisation of Stream Ecosystems
The environment, as well as the human societies that exist

within it, depend on fresh water and its associated
resources. Freshwater capital and services are rapidly
becoming depleted; their quality has been severely
degraded on a worldwide basis. There are several causes
for these trends, but land-use change associated with
changing economic activities and sociodemographics is one
of the most conspicuous drivers.
Although a good deal of attention is now being paid
to the effects of land use change on terrestrial systems, far
less work has been devoted to aquatic systems-in
particular, rivers and streams. Yet, because they act as
topographic sinks, these aquatic environments may be the
areas that are most seriously affected by alterations to the
landscape.
The science underlying projections of how the
ongoing development of land will influence running water
ecosystems is still in its infancy. Nevertheless, scientists
are increasingly being called upon to provide policy
makers and managers with such predictions. Forecasts will
depend on how well understand the complex relationships
among the behaviour of economic agents, subsequent land-

use changes, and ecological processes.
EEconomiclS ... liMld Uu ,..

"
,J EeologiUI
Change Chango ' i Change:
.-"
Figure 1. Ecological effects in a river,
Some of these relationships are direct and relatively

simple, such as when population growth in a developing
country leads to agricultural expansion, the clearing of
riparian forests, and the consequent loss of species habitat.
But other relationships are indirect, mediated by changes
in geomorphology and hydrology.
For example, in regions where economic growth
stimulates home building, the rapid rate of land conversion
may lead to flash floods and suspended sediment loads
that restructure aquatic and riparian communities and alter
ecosystem processes. Forecasting the effects of land-use
conversion on stream ecosystems is an enormous
challenge, not only because of its importance for future
ecosystem management but also because it requires an
integration of knowledge from diverse disciplines,
including economics, hydrology, fluvial geomorphology,
and ecology.
The best approach was to begin with economic
forecasts of land-use change, and then to link them in a
stepwise fashion to hydrologic forecasts (that is, how the
expected land use will affect flows), geomorphic forecasts
Ecological Forecasting and the Urbanisation of· ... 191
(how land use and hydrologic changes will affect

channels), and finally ecological forecasts.
ECOLOGICAL FORECASTS
A process that would enable predictions of spatial pattern,

timing, and amount of land-use change to be generated
from economic models of development, which in turn
would be linked to predictions of future flow regimes,
channel form, and finally ecological change (Figure 1).
Because of their broad ecological importance, and because
both attributes are sensitive to land-use changes, here
focused on species diversity and nutrient cycling as
dependent ecological variables.
Economics, hydrology, geomorphology, and ecology
differ fundamentally in the types of knowledge they
produce, the space and time scales over which that
knowledge can be applied, and the format in which that
knowledge is communicated.
Gaps in knowledge and data exist within each
discipline, and interestingly, many of them only arose as
gaps in the context of what was required to complete the
integrated forecasting exercise. Some of these gaps can be
filled by assuming that certain variables will experience
some average conditions and hopefully show little
variation; others relate to mechanisms that researchers
from particular diSciplines found too important to permit
any compromise.
To provide an idea of the variety of gaps focus on
five examples of such gaps that differ in size, importance,
and status, including limitations in
economic forecasts,
land-cover descriptions,
low-flow models and measurements,
geomorphic models and measurements, and

ecological data and models.
Economic Forecasts
In understanding the impact of urbanisation on watershed
ecology, economics can provide a translation betwe~n
policy actions and regional economic forces on the one
hand and local land-use outcomes on the other. To forecast
future outcomes, here explain why firms and households
make the land-use decisions they do. Because they both
contribute to and respond to market signals, part (It the
economist's task is to understand how lanu mClrkd~ and
the markets for related goods and services operate.
Economists must also understand how regulation and
other types of intervention by government entities alter
these signals. Currently, economic forecasting is limited
by difficulties in projecting supply and demand functions
into the future, particularly in the face of new public
policies (for example, minimum lot sizes), and by
difficulties in producing spatially explicit predictions at
scales relevant to ecologists. Consider the task of
explaining "simply" the change in the aggregate amount
of residential land use likely to occur in a small, specific
watershed over the next 20 years.
Here choose a mid-Atlantic landscape where
residential land use represents over 90% of the developed
land in the watershed. The horizontal axis measures the
amount of new land demanded annually by the residential
sector; the price of that land is measured along the vertical
axis. The actual amount converted is depicted by the
intersection of the supply and demand functions.
To forecast how much land will be converted in a
year, the location of these two functions needs to be
forecast. Estimates of these functions exist for many
Ecological Forecasting and the Urbanisation of . . . . 193
regions. However, as forecasts are made further and

further into the future, it cannot be assumed that the
locations of these functions in the graph will remain stable.
Land
PriCe
Amount of UiIId
e,,-.-..
Figure 2. The supply of new land and the demand for it are r(,glliated by
land prices. The balance point between supply and demand is g(lvt'rned by
the market price.
Shifts backward in supply will cause prices to increase and

the amount of converted land to decrease; shifts backward
in demand will cause prices to decrease and the amount
of converted land to decrease. To forecast the aggregate
amount of conversion of open-space uses into residential
use, preferences, technology, the adoption of policies that
will affect demand and supply functions, and the
behavioural response to these policies also must be
forecast.
These factors include, but are not limited to, funds
made available to purchase development easements from
undeveloped land holders, minimum lot size zoning, the
provision of public services, road construction, agricultural
support policies, inner city restoration, tax incentives for
home ownership, and gasoline taxes. The task of
forecasting aggregate land-use conversion in a region is
difficult, but it is a task for which some economic tools
exist. However, the aggregate amount of land-use change
for a region is not sufficient input for hydrologic,

geomorphic, or ecological models. Here also need a
forecast of the explicit spatial pattern of that development,
because the spatial domain of the land market will not
align with any ecological "domain", since the boundaries
of watersheds rarely figure into the economic decisions
related to land conversion unless they are forced to do so
by some policy instrument.
Until recently, almost all spatial economic models of
land use were quite abstract and were descended from
the bid-rent model (or monocentric city model) of urban
economics. More recently, a few economists have
attempted to treat land-use change as the result of the
cumulative interactions among many economic agents
distributed in space. A few spatially explicit simulation
models of urban growth patterns have recently emerged.
Tht'~l' models estimate the probabilities of land-use
tran!->ltion using discrete choice methods that are based on
the behaviour of the individual agents making land-use
decisions.
The micro-level, spatially explicit model of Landis for
the San Francisco Bay and Sacramento areas is an example
that uses spatially articulated data from a geographic
information system (GIS) to generate spatially disaggregate
predictions of landuse change. Researchers have also
sought to develop economic models of land-use change
that are both spatially explicit and disaggregate, so that
scenario forecasts can be linked with ecological models of
landscape changes.
Such modeling efforts are examples of the progress
that is made possible by geographic data. However,
although these models can produce probabilistic forecasts
at a spatially explicit scale, they are still in their infancy.
They are very data-intensive, requiring significant
geospatial economic data; and they have not yet produced
generalisable results.
Ecological Forecasting and the Urbanisation of . ... 195
Land-cover Descriptions
The rapid development of Geographic Information System
(GIS), which accompanied the emergence of landscape
ecology, has enabled ecological research that was
unimaginable only 2 decades ago. These techniques have
been particularly useful at large spatial scales, but their
linkage to smaller-scale pattern and process-such as those
of interest in single, small streams-remains problematic.
Many of the limitations of using GIS-based land-cover
descriptions to make ecological predictions stem from
resolution and interpretation; they include difficulties in
estimating the size of riparian zones and wetlands,
underestimating the heterogeneity of land-use types, and
dealing with the occurrence of software errors. If the data
are in raster format, the resolution is defined by the pixel
size; whereas if the data are in vector format, it is defined
by some minimum mapping unit.
Unfortunately, the linear nature of most stream
channels makes it difficult, in terms of resolution, to
document the land changes that directly affect the fate of
streams and rivers. The riparian zone is an extremely
diverse habitat and an important determinant of the biota
in the river channel. The width of these zones directly
affects species richness; the amount of leaf litter
production; the stream's buffering capacity, particularly
with respect to nutrient removal; and organism dispersal.
For all these reasons, quantification of the width of
the riparian zone is a high priority in forecasting changes
related to land use. Unfortunately, unless this zone is very
wide, the resolution of remotely sensed data is usually too
coarse to aid in the determination of the width of the
riparian zone. One of the basic steps in documenting
landscape change is to assign land-cover classes to patches
on remotely sensed images. The definition of landcover
categories and the determination of their spatial extent
within an image are fraught with logistical problems. One

of the most common land-cover classification schemes is
that of Anderson and others.
This scheme is a hierarchical description of land use
across a number of broad classes, such as urban,
agricultural, and forest. But no matter how refined these
classifications may be, they still collapse a continuously
varying landscape into a finite number of land-cover
classes. Further, these classifications often include
categories that, while descriptive of the land cover (for
example, institutional, commercial), are inherently
heterogeneous.
An institutional area of land may correspond to a
highly impervious area, such as the parking lot or rooftop
associated with a school. Alternatively, an institutional area
may correspond to land holdings that are quite sizable,
are largely forested, and contain little or no impervious
surface. Particularly relevant to the challenge of relating
landscape change to the fate of aquatic ecosystems is the
confusing nature of the available maps and databases
dealing with wetlands.
Because of the inadequate policies now in place for
defining and surveying wetlands, wetlands have been
defined in different and often inaccurate ways. Finally,
confusion in land-use descriptions can result from
automated interpretation software that falsely attributes
detected land features to the incorrect land cover, or even
from human error in interpreting and digitising images.
For example, the multi-resolution land characteristics
(MRLC) data have a tendency to attribute forest land cover
to areas where the actual land cover is residential but with
a welldeveloped tree canopy.
The land appears to be forested, but beneath the
canopy are roads, curbs, gutters, and rooftops that behave
much differently from a forest. Fortunately, technological
advances in remote sensing are proceeding at a prodigious

rate, particularly with respect to improved resolution and
the real-time availability of images. Furthermore, digital
maps and digital elevation models are also of growing
importance and will be processed together with remote
sensing data and data from landscape ecological analyses
within GIS of future generations. Recent developments in
satellite imagery hold particular promise for river
characterisation and may be useful for elucidating the
extent of floodplain inundation or suspended sediment
transport processes.
The importance of the flow regime in shaping aquatic
communities is well recognised in stream ecology. Flow
regulation has had major consequences for stream
ecosystems. In all streams, regulated or not, both flood
flows and low flows have dramatic effects on the structure
of biotic communities and the rates of ecological processes.
Estimating the magnitude, duration, and future changes
in flows is an important challenge that will need to be
met before, make accurate forecasts of the future of
running water ecosystems.
Two widely used models in such hydroecological
modeling are the physical habitat simulation (PHABSIM)
model and the Instream Flow Incremental Methodology
(IFIM). These were both developed in the United States
in the late 1970s by the staff of the Fish and Wildlife
Service and are related in that the PHABSIM is a major
component of the IFIM. The models are used to determine
the instream flows needed in rivers and streams to support
populations of aquatic organisms, particularly fish. They
assume that the physical habitat needs of an aquatic
animal are related to attributes of the physical
environment.
It is further assumed that the most important
attributes of this environment are the current velocity,
water depth, and substrate characteristics. Whereas the
PHABSIM model focuses on the areal extent of habitat that

is usable by any target species under different flow
regimes, the IFIM model takes a broader view and also
includes a legal-institutional analysis. The IFIM also
attempts to achieve an acceptable balance between
biological and economic values. This negotiation
component of the IFIM makes it an open-ended model
without a single, best solution.
In fact, the result is very much dependent on the skills
of the people involved, and the model may need to be
run several times before an acceptable solution (a
compromise) is reached. A drawback of the models is that
they are designed for use in rivers that are regulated by
dams and that they use a point source discharge of water
from the dam and route it through the river. Diffuse,
multiple source inputs of water from throughout a
watershed - that is, the situation met in urbanising
watersheds - are not specifically addressed.
Clearly, ecologists depend on accurate predictions of
both peak flows and low flows. However, relating the
magnitude and duration of low-flow events to changes in
land use is often more difficult than it is for peak flows.
The resulting uncertainty or lack of models for estimating
low flows is particularly problematic for forecasting efforts.
This is because the timing, duration, and spatial extent of
low-and no-flow conditions can dramatically alter
ecosystem dynamics, particularly when such drought
condi tions are novel to the system and the biota lack
adaptations for resisting or recovering from desiccation.
Low flows are difficult to estimate for two reasons.
First, hydrologic modeling has its origins largely in
engineering applications. Historically, hydrologic engineers
have designed structures that must withstand high flows;
thus, they have concentrated on describing and predicting
the dynamics of floods. But low flows generally do not
threaten such structures. Second, funding for hydrologic
data collection, at least in the United States, has focused

on utilitarian purposes, such as the measurement of high
flows, rather than the improvement of basic hydrologic
understanding.
Accurate modeling of low-flow conditions requires a
high level of understanding of system hydrology because
these flows are subject to a wide range of influences -
for example, deep groundwater discharge, hyporheic
processes, evapotranspiration, and local effects due to
complex, small-scale riverbed topography. So far, most
low-flow studies have focused on natural catchment
conditions. However, each new combination of natural
proCl'~~l'~ clnd various anthropogenic impacts complicates
the understanding of low-flow mechanisms.
Groundwater is typically the single biggest
contributor fostreamflow during low-flow conditions.
Estimating groundwater contributions to stream-flow
depends on accurate estimates of the state of the
groundwater table, which in tum is strongly influenced
by recharge rates and the nature of the underlying porous
media. The connectivity and existence of preferential
flowpaths within the undenlying aquifer are rarely, if ever,
well known and are generally handled stochastically.
As the water table moves up and down in response
to storm-generated recharge, these flowpaths move in and
out of operation, and their ultimate contribution to
streamflow therefore varies. Precise predictions of how the
groundwater table will change would require a detailed
understanding of infiltration, exfiltration, and
evapotranspiration processes, each of which is the focus
of ongoing research within the hydrologic community.
Streamflow gauges capture not only watershed
response to storm events but also low-flow conditions,
including groundwater inputs. However, accurate low-
flow gauge data are difficult to obtain and are, only
available at some sites. Many gauges are not calibrated to

distinguish between a few centimeters of water and a dry
streambed (the latter has far different ecological
implications than the former).
Even when adequate low-flow data are available, they
typically only allow for point estimates of processes that
vary continuously throughout the stream network. The
stream network integrates the low-flow runoff just as it
does the storm runoff, obscuring variability within the
gauged system. Ecologists often need information not at a
single point but throughout the watershed; however, they
are currently limited by the lack of spatial resolution
provided by the stream gauge network. Increases in the
impervious surfaces associated with urbanisation limit
infiltration and should therefore lead to reduced
groundwater levels and low flow (Figure 3).
Figure 3. Idiosyncratic movement of the water table.
However, the growth of the water supply network

necessitated by urban expansion adds substantial
complexity to runoff dynamics. Water supply networks-
particularly older distribution systems that contain pipes
more than 100 years old - can leak and thus supply
recharge to groundwater. An additional source of recharge
to groundwater may come from irrigation, particularly in
wealthy urban areas.
The dominant paradigm of reduced groundwater

levels and reduced low flows is confounded by the leaky
water supply and irrigation associated with urbanisation.
Quantifying these behaviours and processes represents a
further challenge to the hydrologic estimation of low-flow
conditions.
Geomorpllic Models and Measurements

Geomorphic variables influence a wide range of
populations, ~munities, and ecosystem dynamic~ as
well as fIOwdynarnics in running water systems. Further,
theintetacnon between flow and bed composition can
exert significant control over biological processes in
streams. The three most ecologically important geomorphic
factors are
substrate size and bed mobility,
suspended sediment loads to and in the channel, and
channel form.
Because each of these three variables is so ecologically
important, here focus separately on the problems with
effective quantification and modeling (particularly as a
function of land-use changes) associated with each of
them.
Substrate size and bed mobility

Land-use change is frequently associated with altered flood
regimes and sediment input to the channel, which in turn
affect particle size distribution and bed mobility. Because
of the well-known role that patches of stable streambed
play in mediating the effects of floods on stream biota,
the most appropriate measure of bed mobility ecologically
is the percentage of the streambed that is disturbed by a
particular discharge.
Sediment transport theorists currently use two

methods for determining the initiation of sediment motion,
both of which provide measures fundamentally different
from the areal measures required by ecologists. The
"threshold of sediment motion method" relies on
specifying the conditions required to move "some" of the
grains of a given size on the bed; if these conditions are
met, then the sediment is considered to be "in motion."
In practice, the threshold is determined subjectively
by observing the bed and "deciding" if movement is
occurring or not. The "reference transport rate method"
defines the beginning of sediment motion as occurring
when the transport rate exceeds a small, essentially
arbitrary threshold value. Both methods define the initial
disturbance of the bed in terms of a threshold transport
rate with units of either number of grains in motion per
unit time (threshold approach), or volume or mass fluxes.
Near the threshold of sediment motion, most
sedimentary particles are at rest - a condition referred to
as partial transport. Thus, determining the threshold of
motion using criteria appropriate for sediment transport
studies will not provide the information required by
ecologists: When sediment transport criteria suggest that
particles of a given size are "in motion," most of the
particles of that size actually remain at rest. Clearly,
sediment transport criteria provide very poor predictions
of the areal extent of bed disturbance.
If additional information were available, then the
reference transport approach could be used to determine
the areal extent of bed disturbance. One of these
componen ts is the fraction of grains of a particular size
actually in motion for any particular flow. Wilcock's
approach could be used to determine the areal extent of
bed disturbance, but the published method relies on
coefficients determined from only one laboratory
experiment.
Suspended sediment loads

Land-use conversion often causes an increase in suspended
sediments in the water column, a result that can have
numerous ecological effects. When a land development
project is initiated, it is therefore important to know how
much fine material will be in the water and for how long.
Most of the sediment transported in suspension is silt -
or day-sized ("washload"), and its concentration in the
water is controlled by supply from the watershed, not by
local flow.
To predict how land-use changes will influence
suspended sediment concentrations, the delivery of upland
sediment to river channels must be specified as a function
of land-use category. Methods exist for determining
sediment production from agricultural landscapes, but
corresponding approaches for other land uses (for example,
urban and suburban landscapes) do not exist. Furthermore,
available methods are designed to assess erosion from
upland plots, but they may not accurately predict the
delivery of sediment to stream channels.
If the upland supply of suspended sediment cannot
be predicted, another promising approach might involve
developing empirical relationships between land use and
observed sediment fluxes or concentrations measured at
gauging stations. This would require a network of
suspended sediment gauging stations with long records
(several decades) in watersheds with varying land uses.
Unless such data become available, the ability of
geomorphologists to quantify changes in sediment
concentrations caused by land-use changes is likely to
remain rudimentary.
Channel Form
Many studies define how changes in land use influence

stream morphology based on field observations. Most of
these studies, however, focus on reach - averaged channel

properties such as width, depth, slope, or planform. These
variables provide no information regarding the spatial
variability in morphology that is so important for
ecologists. Studies are needed that view morphologic
variance as a dependent variable to be explained, rather
than as noise to be minimised.
These include bed forms such as alternate bars or
pools and riffles, step pool sequences, pebble clusters, bed
load sheets, and "patches" of different grain size. In many
cases, sophisticated theories have been proposed to explain
these features. Despite the considerable progress that has
been made in explaining these subreach-scale features, it
is still not possible to determine how they will change
under conditions of varying sediment and water discharge
- precisely those conditions that are most influenced by
altered land use. Focused studies are needed at sub reach
scales so that meaningful forecasts of ecological responses
can be made.
Further studies are also needed to define the time
scales at which channels adjust to changing land use.
Observations of channel morphology rarely span the broad
time scales required to define the temporal evolution of
channel change. Instead, it is frequently assumed that
channels rapidly reach a quasi-steady state even though
this assumption has not been supported by the few long-
term observations that are available.
Ecological Models
To forecast land-use effects on species diversity and
nutrient dynamics in streams, it is critical that some
knowledge of the relationship between land use and the
dependent variables of interest. Because so much land-use
change has already occurred, here should be able to draw
on existing information, such as time series data for species
distributions or water quality, to define the ecological

responses to land-use change. Unfortunately, predicting,
or even documenting, the effects of landuse change on
aquatic systems is constrained by the interrelated problems
of limited data availability and the spatial and temporal
extent of research relevant to this problem.
Species diversity
Current knowledge of species distributions in running
waters is quite uneven because of biases in the selection
of which organisms are sampled, where the sampling is
done, and the spatial and temporal extent of the sampling
effort. For example, fish (especially game fish) are widely
surveyed by resource management agencies and therefore
probably represent the best source of information about
species distributions and abundances.
Yet data for many other groups are rare or
nonexistent (for example, there is a paucity of data for
invertebrates and even for riparian plants). Further, the
data that are collected may not be disseminated because
of security reasons (for example, information about
threatened species). In other cases, data may be hard to
find because the existence of many databases is not
publicised. For example, in some countries hydrologic data
are collected by many different agencies that rarely
communicate with each other.
Limitations due to the paucity of accessible data are
even further constrained by the spatial and temporal extent
over which most data are collected. Many ecological
studies are snapshots in time (1-2 years) conducted at small
spatial scales-often 1 m 2 and consider only one or two
species at a time. Studies that are spatially extensive
usually consist of a single, temporally unreplicated
inventory, even for conspicuous aquatic taxa such as fish
or riparian vegetation.
The limited scope of data collection is particularly

precarious because the scale of the cause (that is, land-use
change) and the effect (species diversity) do not match.
When data collected at limited spatial scales are used,
extrapolation to the watershed scale is required as part of
the forecasting exercise, and the direct extrapolation of data
from short to long periods, or from small areas to whole
ecosystems or landscapes, is likely to yield erroneous
conclusions about community and ecosystem processes.
These limitations of scope can be overcome by
instituting long-term sampling regimes, expanding the
spatial extent of the sampling, and developing robust
approaches to extrapolation such that models developed
in one place or at a small spatial scale can be reasonably
applied elsewhere. Some agencies and groups are working
toward the development of species-monitoring networks
that will help to assess the effects of land-use change.
Notable examples include the US Geological Survey's
National Water-Quality Assessment Programme
(NAWQA) and the National Biological Information
Infrastructure (NBII) in the United States. NAWQA, which
was designed to monitor the status and trends in ground
-- and surface-water quality and to provide a sound
understanding of the natural and human factors affecting
these resources, now includes 59 separate study regions.
Fish and invertebrate populations are routinely
monitored in concert with a wealth of additional aquatic
and terrestrial state variables, providing a strong empirical
base for understanding the relationships between land use
and the ecological attributes of lotic ecosystems. The NBII
serves as a point of access for biological data collected by
several government agencies and should help to alleviate
some of the limitations of data access.
Nonetheless, lack of information about species
distributions - particularly at a scale relevant to land use
change - is one of the greatest obstacles to forecasting

efforts. The urgent need for extensive and coordinated
ecological data collection networks and for sustained
interdisciplinary work groups focused on the collection of
information that would enable ecological forecasting has
been emphasised in many recent venues. Coordinated
biodiversity inventories are largely lacking, yet they are
fundamental to forecasting environmental change.
The National Science Foundation's (NSF) proposed
National Environmental Observatory Network (NEON)
could fill this data void and similar types of projects are
currently being discussed in other countries (for example,
France). Such site-based facilities for biological and
physical data collection, in conjunction with natural history
archive facilities, would allow ecologists to begin the
arduous process of documenting biodiversity patterns and
processes in a spatially explicit fashion.
Nutrient dynamics. The data needed to investigate the
relationships between land use and the nutrient
characteristics of streams and rivers appear to be less
severely limited than those needed to study species
diversity. Indeed, there are several studies that have
examined possible relationships between land use and
water quality.
However, these efforts have not led to a clear
consensus on the effects of land use on streams and rivers.
In some cases, land cover is a strong predictor of nitrogen
and phosphorus concentrations in stream water; but in
others, land use in adjacent buffer zones is the best
predictor of stream nutrient status.
In response to the gaps in researchers knowledge of
land-water linkages, a dizzying array of watershed models
have been developed to predict, manage, and understand
the effects of land-use change on hydrology and nutrient
losses. These watershed nutrient models may represent the
greatest advance in the realm of forecasting. However, in

order to use these models effectively, it is important to be
aware of their inherent limitations or assumptions and to
understand the robustness of the model's output. As with
species diversity, there are important limitations to
watershed modeling that derive from issues of data
availability and the scope of the model (extent).
First, many watershed models - particularly process-
based models - have intensive data requirements and
thus can only be used in specific locations where the
needed information is available. Data aggregation is
commonly used to overcome problems of data limitation,
but reduced data resolution can have substantial effects
on model predictions. Second, watershed models typically
describe only upland processes and do not include
instream dynamics, despite growing evidence that
instream processes can have strong effects on the form and
amount of nutrients exported from a watershed.
Finally, watershed models have met with the greatest
success when used in either very small or very large
catchments. In contrast, land-use change, and the factors
causing land-use change, may be best charact~rised at a
spatial scale that falls between these two extremes - a
spatial extent for which the successful development of
hydrologic and biogeochemical models has been elusive.
The task of adapting existing models or developing new
ones that are relevant to this intermediate scale represents
a substantial challenge for hydrologists and ecologists
alike.
MECHANISTIC MODELS
Certain forms of knowledge and kinds of data only take

on significance in the context of solving a complex
forecasting problem. Therefore, to accommodate the
demands of managers and concerned citizens,
environmental scientists need to communicate with

researchers in other disciplines. If ecologists and physical
scientists do not articulate the specific nature of their
knowledge requirements and spell out the limitations of
their methods to specialists in other fields, progress in
environmental forecasting will surely proceed very slowly.
Unfortunately, the subject of disciplinary limitations
is not often addressed, perhaps because of a fear that such
openness might call the researchers' own intellectual
capacity into question. This attitude will, of course, hamper
scientific progress. This is a particularly regrettable
situation, given that many of the requirements for one
discipline may be easily produced or obtained from
another one, yet these contacts are not generated unless a
compelling common interest can be established. Other
variables, such as some of the geomorphic ones, will
continue to present daunting technical problems even if
research priorities are changed.
As of now, workers in virtually all environmental
disciplines - and in particular environmental managers
- are faced with the dilemma of having to implement
decisions or make predictions about the future in spite of
these severe constraints on the information available to
them. Thus, the development of methods for forecasting
under constrained conditions is a critical area in ecology
today. Some of the more commonly used approaches for
forecasting include the use of
empirical regression models,
Bayesian or dynamic linear models, and
adaptive environmental assessment.
Empirical regression models provide one means of
circumventing the difficulties of linking mechanistic
models from different disciplines. These models are
relatively straightforward and have been widely used to
relate changes in biodiversity (for example, species

richness) and ecosystem function (for instance, nutrient
export from a watershed) to landuse variables (for
example, land cover, population density, road density,
number of dams, and degree of flow regulation) in the
surrounding watershed.
These models can be used to forecast the outcome of
a specific land-use change, such as deforestation, they can
also be used to predict the consequences of the addition
or removal of dams, or to evaluate the best management
practices for a given set of conditions.
Bayesian modeling is a means of coping with the
problems of temporal variability in controlling mechanisms
and the limited availability of observations to parameterise
models. Bayesian statistical inference is used to calculate
the probability of the value of a parameter given the data,
in contrast to traditional frequentist statistical methods,
where one calculates the probability of observing data
given a value for a parameter.
As a consequence, the Bayesian approach enables
beliefs about the state of a system to be incorporated prior
to data collection. In addition, models can be constructed
from small sample sizes, and the analysis can be updated
as new data are collected. Bayesian methods are becoming
increasingly popular in many fields of applied science,
such as fisheries biology. Bayesian dynamic moaeling is
also used in forecasting-for example, predictions of climate
change. However, the use of Bayesian modeling in ecology
also has its opponents.
There are strong differences in opinion regarding the
validity of explicitly including subjective information about
the state of the system under study prior to the collection
of experimental data.
Adaptive management is a practical, applied approach
to environmental management that can be applied when
Ecological Forecasting and the Urbanisation of . ... 211
the development of accurate combined models is difficult

or unfeasible. With this method, people with a varying
types of expertise and experience come together to identify
a range of possible management actions and to develop
computer simulation models that test the potential
outcomes of different actions.
The management option that appears most likely to
succeed can then be chosen and tested in the field.
Monitoring and reassessment of the field experiment is
essential so that policy can be changed and improved as
new information becomes available. This approach allows
managers and scientists to "learn as they go". In this way,
a reasoned response to the circumstances becomes possible
despite limited baseline data. It minimises the risk of
management actions taken under uncertain conditions and
improves successive management decisions by gathering
more data as the experiment progresses.
From a more academic standpoint, the adaptive
management approach offers ecologists an opportunity to
field-test linked interdisciplinary models that currently are
very uncertain. Thus, it would enable better models to be
developed as knowledge and data gaps are filled.
However, there is a risk that extreme experimental
manipulations may harm sensitive species or disrupt
critical ecosystem processes.
Although empirical regression models, Bayesian
modeling, and adaptive management offer ways to
circumvent researchers current inability to link mechanistic
models from different disciplines, they should be viewed
more as a surrogate than as a cure - all for the
uncertainties that plague the environmental forecasting of
stream and riverine ecosystems.
12
Land Use Appraisal
Land transformation affects many of the planet'S physical,

chemical, and biological systems and directly impacts the
ability of the Earth to continue providing the goods and
services upon which humans depend. Unfortunately,
potential ecological consequences are not always
considered in making decisions regarding land use.
A critical challenge for land use and management
involves reconciling conflicting goals and uses of the land.
The diverse goals for use of the land include:
resource-extractive activities (e.g., forestry, agriculture,
grazing, and mining);
infrastructure for human settlement (housing,
transportation, and industrial centers);
recreational activities;
services provided by ecological systems (e.g, flood
control and water supply and filtration);
support of aesthetic, cultural, and religious values; and
sustaining the compositional and structural complexity
of ecological systems.
Land Use Appraisal 213
These goals often conflict with one another, and difficult

land-use decisions may develop as stakeholders pursue
different land-use goals. Local versus broad-scale
perspectives on the benefits and costs of land management
also provide different views. To meet the challenge of
sustaining ecological systems, an ecological perspective
should be incorporated into landuse and land-management
decisions. Specifying ecological principles and
understanding their implications for land use and land-
management decisions are essential steps on the path
toward ecologically based land use. Key ecological
principles deal with time, species, place, disturbance, and
the landscape. While they are presented as separate
entities, the principles interact in many ways.
TIME PRINCIPLE
Ecological processes function at many time scales, some

long, some short; and ecosystems change through time.
Metabolic processes occur on the scale of seconds to
minutes, decomposition occurs over hours to decades, and
soil formation occurs at the scale of decades to centuries.
Additionally, ecosystems change from season to season
and year to year in response to variations in weather as
well as showing long-term successional changes.
Human activities can alter what makes up an
ecosystem or how biological, chemical, and geological
materials flow through an ecosystem. These in turn can
change the pace or direction of succession and have effects
lasting decades to centuries. The time principle has several
important implications for land use:
The current composition, structure, and function of an
ecological system are, in part, a consequence of
historical events or conditions and current land uses
may limit land use options that are available in the
future.
The full ecological effects of human activities often are

not seen for many years.
The imprint of a land use may persist on the
landscape for a long time, constraining future land use
for decades or centuries even after it ceases.
Long-term effects of land use or management may be
difficult to predict due to variation and change in
ecosystem structure and process.
This problem is exacerbated by the tendency to overlook
lowfrequency ecological disturbances, such as lOO-year
flooding, or processes that operate over periods longer
than human life spans (e.g., forest succession).
SPECIES PRINCIPLE
Particular species and networks of interacting species have

key, broad-scale ecosystem-level effects. These focal species
affect ecological systems in diverse ways:
Indicator species tell us about the status of other
species and key habitats or the impacts of a stressor.
Keystone species have greater effects on ecological
processes than would be predicted from their
abundance or biomass alone.
Ecological engineers [e.g., the gopher tortoise or
beaver] alter the habitat and, in doing so, modify the
fates and opportunities of other species.
Umbrella species either have large area requirements
or use multiple habitats and thus overlap the habitat
requirements of many other species.
Link species exert critical roles in the transfer of matter
and energy across trophic levels (of a food web) or
provide critical links for energy transfer within
complex food webs.
The impacts of changes in the abundance and distribution

of focal species are diverse. For example, keystone species
affect ecosystems through such processes as competition,
mutualism, dispersal, pollination, and disease and by
modifying habitats and abiotic factors. An introduced,
nonnative species can assume a focal-species role and
produce numerous effects, including altering community
composition and ecosystem processes via their roles as
predators, competitors, pathogens, or vectors of disease
and, through effects on water balance, productivity, and
habitat structure.
The impacts of land use changes on keystone and
invasive species can spread well beyond the boundaries
of the land-use unit and are difficult to predict prior to
changes in their abundance. Changes in the pattern of land
cover can affect, and even promote, the establishment of
nonnative species. Trophic levels refer to the stages in food
chains such as producers, herbivores, consumers, and
decomposers.
Changes in the abundance of a focal species or group
of organisms at one trophic level can cascade across other
trophic levels and result in dramatic changes in biological
diversity, community composition, or total productivity.
Changes in species composition and diversity can result
from land use through alterations to such ecosystem
properties as stream flow or sediment load, nutrient
cycling, or productivity. The effects of land use on species
composition have implications for the future productivity
of ecological systems.
PLACE PRINCIPLE
Local climatic, hydrologic, soil, and geomorphologic factors

as well as biotic interactions strongly affect ecological
processes and the abundance and distribution of species
at anyone place. Local environmental conditions reflect

location along gradients of elevation, longitude, and
latitude and the multitude of microscale physical, chemical,
and edaphic factors that vary within these gradients. These
factors constrain the suitability of various land uses, as
well as defining resident species and processes.
Rates of key ecosystem processes, such as primary
production and decomposition, are limited by soil
nutrients, temperature, water availability, and the temporal
pattern of these factors controlled by climate and weather.
Thus, only certain ranges of ecological-process rates can
persist in a locale without continued management inputs
(e.g., irrigation of crops growing in a desert). Chronic
human intervention may broaden these ranges but cannot
entirely evade the limitations of place without a cost.
Naturally occurring patterns of ecosystem structure
and function provide models that can guide sustainable
and ecologically sound land use. Only those species
adapted to the environmental constraints of an area will
thrive there. Precipitation limits which species are
appropriate for landscape plantings as well as for managed
agricultural, forestry, or grazing systems. Further, some
places with unique conditions may be more important than
others for conservation of the species and ecosystems they
support.
Land uses that cannot be maintained within the
constraints of place will be costly when viewed from
longterm and broad-scale perspectives. For example,
establishing croplands and ornamental lawns in arid areas
is possible, but draws down fossil groundwater at a rate
unsustainable by natural recharge. Only certain patterns
of land use, settlement and development, building
construction, or landscape design are compatible with local
and regional conditions. In terrestrial systems, land-use
and land-management practices that lead to soil loss or
degradation reduce the longterm potential productivity of

a site and can affect species composition.
Land use practices can also influence local climate
(e.g., the urban heat island concept). Sustainable settlement
is limited to suitable places on the landscape. For instance,
houses or communities built on transient lake shore dunes,
major flood plains, eroding seashores, or sites prone to
fires are highly vulnerable to loss over the long term.
Ideally, the land should be used for the purpose to which
it is best suited.
DISTURBANCE PRINCIPLE
The type, intensity, and duration of disturbance shape the

characteristics of populations, communities, and
ecosystems. Disturbances are events that disrupt ecological
systems; they may occur naturally [e.g., wildfires, storms,
or floods] or be induced by human actions, such as
clearing for agriculture, clearcutting in forests, building
roads, or altering stream channels. The effects of
disturbances are controlled in large part by their intensity,
duration, frequency, timing, and the size and shape of the
area affected.
Disturbances may affect both above- and below-
ground processes and can impact communities and
ecosystems by changing the number and kinds of species
present; causing inputs or losses of dead organic matter
and nutrients that affect productivity and habitat structure;
and creating landscape patterns that influence many
ecological factors, from movements and densities of
organisms to functional attributes of ecosystems. Land-use
changes that alter natural-disturbance regimes or initiate
new disturbances are likely to cause changes in species'
abundance and distribution, community composition, and
ecosystem function. In addition, the susceptibility of an
ecosystem to other disturbances may be altered.
Land managers and planners should be aware of the

prevalence of disturbance in nature. Disturbances that are
both intense and infrequent, such as hurricanes or lOO-yr
floods, will continue to produce "surprises." Ecosystems
change, with or without disturbance; thus, attempts to
maintain landscape conditions in a particular state will be
futile over the long term. Attempts to control disturbances
are generally ineffectual and suppression of a natural
disturbance may have the opposite effect of that intended.
For example, suppression of fire in fire-adapted systems
results in the buildup of fuels and increases the likelihood
of severe, uncontrollable fires.
Similarly, floodcontrol efforts have facilitated
development in areas that are still subject to infrequent
large events, resulting in tremendous economic and
ecological impacts. Land-use policy that is based on the
understanding that ecosystems are dynamic in both time
and space can often deal with changes induced by
disturbances. Understanding natural disturbances can help
guide land-use decisions, but the differences between
natural and human-made disturbances must be recognised.
Continued expansion of human settlement into
disturbanceprone landscapes is likely to result in increased
conflicts between human values and the maintenance of
natural-disturbance regimes necessary to sustain such
landscapes.
LANDSCAPE PRINCIPLE
The size, shape, and spatial relationships of landcover

types influence the dynamics of populations, communities,
and ecosystems. The spatial arrangement of ecosystems
comprises the landscape and all ecological processes
respond, at least in part, to this landscape template. The
kinds of organisms that can exjst are limited by the siz~s,
shapes, and patterns of habitat across a landscape. Human-
Lan4 Use Appraisal 219
settlement patterns and landuse decisions often fragment

the landscape or otherwise alter land-cover patterns.
Decreases in the size of habitat patches or increases
in the distance between habitat patches of the same type
can greatly reduce or eliminate populations of organisms,
as well as alter ecosystem processes. However, landscape
fragmentation is not always necessarily destructive of
ecological function or biodiversity because a patchwork
of habitat types often maintains more types of organisms
and more diversity of ecosystem process than does a large
area of uniform habitat. Making a naturally patchy
landscape less patchy may also have adverse affects.
Larger patches of habitat generally contain more
species (and often a greater number of individuals) than
smaller patches of the same habitat. Larger patches also
frequently contain more local environmental variability.
This variability provides more opportunities for organisms
with different requirements and tolerances to find suitable
sites within the patch. In addition, the edges and interiors
of patches may have quite different conditions, favoring
some species over others. The abundance of edge and
interior habitat varies with patch size; large patches are
likely to contain both edge and interior species, whereas
small patches will contain only edge species.
The extent and pattern of habitat connectivity can
affect the distribution of species by making some areas
accessible and others inaccessible. The amount of
connectivity needed varies among species and depends on
two factors: the abundance and spatial arrangement of the
habitat and the movement capabilities of the organism.
While gradual reduction in habitat may have gradual
effects, once a certain threshold is reached, the effects
become dramatic. Land-cover changes are most likely to
have substantial effects when habitat is low to intermediate
in abundance and small changes may cause large impacts.
The ecological importance of a habitat patch may be

much greater than is suggested by its size and distribution
across the landscape. Some habitats, such as bodies of
water or riparian corridors, are small and discontinuous,
but nevertheless have ecological impacts that greatly
exceed their spatial extent. For example, the presence of
riparian vegetation, which may occur as relatively narrow
bands along a stream or as small patches of wetland,
generally reduces the amount of nutrients being
transported to the stream. This filtering by the vegetation
is an ecologically important function because excess
nutrients that unintentionally end up in lakes, streams, and
coastal waters are a major cause of eutrophication,
acidification, and other water quality problems. Thus, the
presence and location of particular vegetation types can
strongly affect the movement of materials across the
landscape and can contribute to the maintenance of
desirable water quality.
GUIDELINES FOR LAND USE
Ecologically based guidelines are proposed here as a way

to assist land managers and others considering the
ecological ramifications of land-use decisions. These
guidelines are meant to be flexible and to apply to diverse
land-use situations. The guidelines recognise that the same
parcel of land can be used to accomplish multiple goals
and emphasise that decisions should be made within an
appropriate spatial and temporal context. For example, the
ecological implications of a decision may last for decades
or even centuries. All aspects of a decision need to be
considered in setting the time frame and spatial scale for
impact analysis. In specific cases, the relevant guidelines
can be developed into prescriptions for action. These
guidelines are best seen as a checklist of factors to be
considered in making a land-use decision.
Examine Impacts of Local Decisions in a Regional Context

The spatial array of habitats and ecosystems shapes local
conditions and responses and local changes can have
broad-scale impacts over the landscape. Therefore, it is
critical to examine both the limitations placed on a location
by the regional conditions and the implications of decisions
for the larger area. This guideline suggests two
considerations for planning land use:
identify the surrounding region that is likely to affect
and be affected by the local project and
examine how adjoining jurisdictions are using and
managing their lands.
Once the regional context is identified, examine regional
data (e.g., land-cover classes, soils, patterns of water
movement, historical disturbance regimes, and habitats of
focal species and other species of special concern). Recent
technological advances [such as the development of
geographic information systems (GIS) and the general
availability of databases for soils, roads, and land cover
on the Internet] make regional analysis a possibility even
for small projects. Some land uses offer more flexibility in
terms of future or adjoining uses than do others.
The ecological constraints of an area determine the
flexibility of diverse land uses for that area. This guideline
calls for examining local decisions within the regional
context of ecological concerns as well as in relation to the
social, economic, and political perspectives that are
typically considered.
Plan for Long-Term Change and Unexpected Events

Impacts of land-use decisions can, and often do, vary over
time. Long-term changes that occur as a response to land-
use decisions can be classified into two categories:
Delayed impacts may not be observed for years or

decades. An example is the composition of forest
communities in New England; today, those forests
differ substantially depending on whether they were
previously woodlots, pasture, or croplands.
Cumulative effects are illustrated by events that
together determine a unique trajectory of effects that
could not be predicted from anyone event.
Land-use decisions should be implemented with some
consideration as to the physical, biological, aesthetic or
economic constraints they will place on future uses of the
land. External effects, which can extend beyond the
boundaries of individual ownership and affect surrounding
owners, must also be considered. Planning for the long
term requires consideration of the potential for unexpected
events, such as variations in temperature or precipitation
patterns or disturbances.
Estimating the occurrence and implications of these
unanticipated events is difficult, but land-use plans should
attempt to include them and estimate likely changes. Long-
term planning should also recognise that one cannot
simply extrapolate historical land-use impacts forward to
predict future consequences of land use. Transitions of
land from one use or cover type to another often are not
predictable because of changes in demographics, public
policy, market economies, and technological and ecological
factors.
Preservation of Landscape Elements

Rare landscape elements provide critical habitats or
ecological processes. For example, in the Southern
Appalachian Mountains, 84% of the federally-listed
terrestrial plant and animal species occur in rare
communities. While these communities occupy a small
area of land, they contain features important for the
region's biological diversity. Therefore, rare landscape

elements need to be identified, usually via an inventory
and analysis of vegetation types, physical features,
hydrology, soils, and associated species.
Once the inventory is complete, effects of land-use
decisions on these landscape elements and species can be
routinely estimated. These effects can then be considered
in view of the overall goal for the project, the distribution
of elements and species across the landscape, and their
susceptibility, given likely future land changes in the
vicinity and region. Strategies to avoid or mitigate serious
impacts can then be developed and implemented.
Depleteion of Natural Resources

Depletion of natural resf)Urces disrupts natural processes
in ways that often are irreversible over long periods of
time. The loss of soil via erosion that occurs during
agriculture and the loss of wetlands and their associated
ecological processes and species are two examples. To
prevent diminishment, those resources at risk must first
be determined. For example, in the southwestern United
States, water might be the most important resource.
Evaluation of whether a resource is at risk is an ongoing
process as the abundance and distribution of resources
change. Ways to avoid actions that would jeopardize
natural resources should be identified and considered.
Some land actions are inappropriate in a particular
setting or time, and they should be avoided. Examples are
farming on steep slopes, which might produce soil loss;
logging, grazing, or farming too close to streambanks,
which may jeopardize water quality and aquatic habitats;
and growing plants with high water demands in arid
areas.
Retain Large Contiguous or Connected Areas

Large areas are often important to maintaining key
organisms and ecosystem processes. Habitat patch size is
critical to the survival of a species or population when it
is rare or disconnected. In most parts of the nation, large
areas of natural habitats are becoming less common as they
are fragmented into smaller habitat patches suitable for
fewer species. A useful management approach generally
favors protecting large areas and smaller areas that are
well-connected to other habitats.
Habitat connectivity is not always a positive attribute
for species and ecosystems. Land uses that serve as
connectors to species' movement can have long-term
positive effects on populations; but, at the same time,
corridors can facilitate the spread of nonnative species or
diseases. The importance of spatial connections depends
on the priorities and elements of a situation. A first step
in implementing this guideline is to examine the spatial
connectivity of key habitats in an area, determining which
patches are connected and whether the connectivity varies
with time. Second, opportunities for connectivity must be
promoted. Sometimes, thos.e opportunities can complement
other planning needs.
Minimising the Spread of Nonnative Species

Nonnative organisms often have negative effects on native
species and the structure and functioning of ecological
systems. Land-use decisions must consider the potential
for the introduction and spread of nonnative species. For
example, kudzu, first used for erosion control, is now
overwhelming and killing native trees. Land planning
should consider vehicle movement along transportation
routes, the planting of native species, and control of pets.
For example, the spread of the gypsy moth is correlated
with overseas transportation of eggs, larvae, and adults

in the cargo holds of ships or along roads when egg sacs
are attached to vehicles or outdoor furniture.
The introduction of aquatic organisms (e.g., the zebra
mussel) transported incidentally with shipping traffic is a
comparable example for aquatic ecosystems. Many of these
introductions have had devastating effects. Growing native
species can reduce the need for planting nonnative species,
particularly in urban, suburban, or other developed areas.
The planted native species can then reseed themselves.
Native species are also adapted to long-term variations in
climate or disturbance regimes to which nonnative species
often succumb. Environmental conditions associated with
native vegetation may also deter the spread of nonnatives.
Compensate for Effects of Development on Ecological

Processes
Negative impacts of development might be avoided or
mitigated by some forethought. To do so, potential impacts
need to be examined at the appropriate scale. At a fine
scale, the design of a structure may interrupt ecoregional
processes, while at a broad scale, patterns of watershed
processes may be altered, for example, by changing
drainage patterns as a result of development. Eow
proposed actions might affect other systems (or lands)
should be examined.
Human uses of the land should avoid structures and
uses that might have a negative impact on other systems;
at the very least, ways to compensate for those anticipated
effects should be determined. It is useful to look for
opportunities to design land use to benefit or enhance the
ecological attributes of a region. For example, parts of golf
courses can be designed to serve as wildlife habitat, or
traffic in rural areas can be concentrated on fewer and
more strategically placed roads, resulting in decreased
226 Ecology and Bioz:tfversity
traffic volumes and flows within the region as a whole

and less impact on wildlife.
Implementing Land-Use and Management Practices

Because local physical and biotic conditions affect
ecological processes, the natural potential for productivity
and for nutrient and water cycling partially determine the
appropriate land-use and -management practices for a site.
Land-use practices that fall within these place limits are
usually cost-effective in terms of human resources and
future costs caused by unwarranted changes on the land.
Implementing land-use and -management practices that
are compatible with the natural potential of the area
requires that land managers have an understanding of the
site potential.
Much progress has been made in managing land in
ecologically sustainable ways. However, more actions are
needed before ecologically based land management is
broadly implemented. These guidelines must be translated
to particular land uses. This translation can be done, for
example, by using these principles and guidelines to shape
municipal ordinances for land use practices. In addition,
these guidelines can provide the basis for specifying and
understanding ecological concerns relevant to the needs
of specific types of land users, such as farmers or foresters.
These guidelines do not address the environmental,
social, economic, and political tradeoffs that often occur
in setting land policy. Tradeoffs are often based on
subjective value judgments reflecting economic, social,
cultural, and aesthetic preferences accorded by a society
to different objectives and variation in local circumstances.
For example, consideration of such tradeoffs are central
components of land-use agreements in the restoration of
the Everglades and in developing options for fisheries and
ancient forests in the Pacific Northwest.
However, society and the ecological community have

not yet converged on a consistently applicable mechanism
for incorporating science into land-use policy. Positive
steps in integrating scientific ideas and land-use
management are being taken at both the international and
national scales. At the local scale, an unprecedented
increase in numbers of watershed alliances and other types
of nongovernmental organisations has occured in response
to the perception that government agencies are not doing
enough to manage the land sustainably. In each case,
science is only a part of the solution, although an essential
part.
13
Population Diversity
The relationship between biodiversity and human well

being is primarily a function of populations of species.
Whether the benefits of biodiversity are conveyed directly
(e.g. food) or indirectly (e.g. pollination), their supply is
generally determined by the diversity of popUlations
producing them. Hence, population change can have a
substantial impact on an ecosystem that is independent
of changes in species diversity.
Species loss dramatically represents the mass
extinction currently underway. Yet, species extinctions are
an inadequate measure of biodiversity loss and do not
provide information about changes in the capacity of
particular species to contribute to the functioning of
ecosystems. For focusing on changes in population size
and habitat extent when measuring the state of nature.
The consequences of population loss for species
conservation arewell recognised, but have been little
addressed from the viewpoint of the functioning of
ecosystems and the provision of ecosystem services. To
explore the link between species populations and 'the
services that they provide, even to the point where species-
specific services have been identified. This approach
should be greatly expanded for two reasons:
Population Diversity 229
to assess the theoretical and practical implications of

population change for the functioning of humanity's
life-support systems; and
to reflect more accurately the state of global
biodiversity.
Population change can occur through variation in the
number of populations for a given species, the number of
individuals per population, the spatial distribution of
populations, and the genetic differentiation within and
among populations. Historically, the term 'population
,~4iversity' has largely been used by geneticists when
·~discussing genetic differentiation, or as a measure of the
number of Mendelian populations in a given area.
These approaches are limited because they ignore
important demographic characteristics of populations that
can influence the provision of ecosystem services.
Assessments of population diversity should consider both
the demographic (e.g. size, number and distribution) and
genetic nature of populations.
The terms 'ecosystem service' and 'ecosystem
function' are largely interchangeable, although ecosystem
services can be defined as ecological processes that benefit
people, whereas ecosystem functions can be considered as
all ecological processes regardless of wQether they are
beneficial to humanity.
There are many approaches to defining species
populations. Consistently problematic is defining
population boundaries so that the number of popUlations
can be clearly determined. Geneticists use measures of
gene flow and genetic differentiation to distinguish one
population from another. In a demographic sense, this can
be achieved by careful measures of individual movement,
which enables the delineation of populations that are
sufficiently isolated from each other to have independent
dynamics.
Populations can also be distinguished with the use

of some arbitrarily defined spatial and/or temporal context
(e.g. linear distance between groups, or the presence of
geographical barriers or other spatial disjunctions) or
differences in phenology, morphology or physiology.
The perspective of species conservation, but they
might be inadequate when assessing the contribution that
a group of individuals from a given species makes to an
ecosystem service. To address this issue, it is essential to
recognise the link between a group of individuals and the
ecosystem service(s) that it provides, and use this as a
further criterion to delineate populations. This new
population categorisation as a 'service-providing unit'.
Imagine a single species occupying a heterogeneous
environment such that groups of individuals can be
distinguished on the basis of habitat suitability or the
spatial characteristics of the landscape (e.g. a
rainforestdependent species occupying forest fragments in
an agricultural landscape). The groups might be linked by
sufficient dispersal to classify them as belonging to a single
evolutionary unit (EU) or demographic unit.
However, each group provides nonoverlapping,
localised services to areas of the surrounding landscape,
suggesting that they could be categorised as different
SPUs. Pollination services provided by native bee
populations to watermelon crops are reduced on farms that
are distant from native vegetation «1% of natural habitat
within a 1-km radius), which provides nesting and
foraging resources for the bees.
In agricultural landscapes with scattered patches of
native vegetation that support bee populations (equivalent
to a collection of individuals in this context), the distance
between patches (and, hence, populations) appears to be
crucial to the provision of pollination services. In the
example above, distances >2 km between populations
would mean that pollination services are nonoverlapping.
TAXONOMY OF POPULATIONS
Evolutionary Unit
Evolutionary units (EUs) are populations with independent
evolutionary dynamics. Thus, a classic Mendelian
population (a reproductive community of sexual and cross-
fertilizing individuals which share a common gene pool)
is an EU. Using neutral loci will generally circumscribe
larger EUs than would using loci under strong,
geographically varying selection. An example of EUs is
the island populations of various plant species (from the
Family Asteraceae) that evolved different diaspore
morphology and reduced dispersal ability, compared with
the mainland populations from which they originated.
Demographic Unit
Demographic units (DUs) are populations with
independent demographic dynamics. In general,
populations that fluctuate in size asynchronously, or are
shown to have only a few (or no) migrants pass between
them, should be considered DUs. A classic example of a
set of DUs is the now-extinct three populations of the
checkerspot butterfly Euphydryas editha in the Jasper
Ridge Biological Reserve of Stanford University. They had
asynchronous dynamics and delineating the DUs was
crucial to understanding those dynamics. Local
populations in metapopulations are ordinarily DUs, but
DUs are not necessarily elements of metapopulations.
Conservation Unit
The designation of conservation units (CUs) will depend
on the associated conservation goals, which vary
tremendously. Goals oriented around evolution typically
involve maintaining the genetic diversity of a species, or
the potential for future genetic divergence or speciation.
Pursuit of these goals has been formalised using concepts

such as minimum viable populations, evolutionarily
significant units, and so-called management units. Here,
focus on the goal of maintaining ecological functions and
their associated ecosystem services.
Service-Providing Unit (SPU)

The new population category that we propose is a service-
providing unit (SPU). SPUs provide, or might provide in
the future, a recognised ecosystem service at some
temporal or spatial scale. All populations are potentially
SPUs and the population categories described here can
overlap (e.g. a single population might be an EU, DU, CU
and SPU). This will not always be the case, making it
crucial to delineate SPUs when assessing the consequences
of population change for the provision of ecosystem
services.
For example, population categories might occur in a
nested hierarchy (Fig. Ia) and the loss of a SPU might not
result in the loss of a DU or CU, but still might have
consequences for the functioning of local ecosystems.
Alternatively, multiple DUs, CUs or EUs might comprise
a single SPU (Fig. Ib). For example, multiple DUs of a
plant species might provide a water purification service
to a given area and loss of a few DUs might not adversely
disrupt the provision of the service.
Service-Providing Unit
The delineation of a service-providing unit (SPU) will vary
depending on the ecosystem service being considered, and
any temporal or spatial variation inherent in the species
of interest and the service itself. For example, the entire
population of a given tree speciesmight provide the global
service of carbon sequestration. Regional populations of
the same tree species might provide a water filtration
service that benefits local communities. Very localised

populations might provide livestock protection for
individual farms.
\1>'1
Figure 1. Taxonomy of Populations
Another example is generalist pollinators that might be

frequent visitors to particular plant species only during
certain seasonal conditions, such aswhen other flowering
resources are scarce, or certain pollinator populations that
are only important to crops during crucial flowering
periods. Service provision and the delineation of SPUs can
occur at multiple levels and will probably depend as much
on our social institutions e.g. what they acknowledge as a
service and the geographical scope of theiroperation, as

theydoonscientific issues.
In cases such as these, populations should be
considered SPUs because the loss of particular populations
(e.g. through native habitat clearance) has implications for
service provision. Moreover, it is realistic to suggest that
multiple, spatially discrete SPUs can comprise a single EU,
DU or conservation unit (CU), and just focusing on the
loss of these latter categories as a measure of population
decline might not adequately assess the consequences of
population change for the provision of pollination services.
This example raises two important issues.
First, the consequences of the loss of a SPU (of a
single species) will be dependent, in part, on the capacity
of individuals from neighboring extant SPUs (of the same
species) to cover the loss (e.g. through re-colonisation of
habitat, assuming that habitat clearance is not the cause
of the loss). The extinction and colonisation dynamics of
networks of localised populations are the focus of the
burgeoning research on metapopulation theory. Under
some circumstances, a SPU might be analogous to a local
population in a metapopulation context if there is sufficient
movement among SPUs.
The important issue is the time taken to recolonise a
site and to reinstate a given service, and the time
constraints associated with that service. In the example
given above, even short-term (e.g.,six months) loss of bee
populations can have significant consequences for
pollination success if it occurs during a crucial period, and
can impact greatly on farm productivity in a given year.
COMPONENTS OF POPULATION DIVERSITY
There are four key components of population diversity:

richness, the size of each population, spatial distribution
and differentiation.
Population Richness
Population richness is the number of populations of a
species in a given area, which depends on the criteria used
to delineate population boundaries. When the focus is on
ecosystem services, we propose using as a criterion the
spatial disjunctions in the services provided by conspecific
individuals occupying a heterogeneous environment (i.e.
SPUs). If the required data are unavailable, standard
genetic or demographic approaches would be the next best
option.
Population Size
Data about the number of individuals per population
provide an indication of the frequency distribution of
population sizes. Absolute numbers would be the most
useful, but, owing to the difficulty of obtaining such
information, orders of magnitude or some other
representative measure might be more appropriate. It is
important to document the distribution of population sizes
to determine whether a species is characterised by, for
example, a single large population and many small
populations, or several similarly sized populations.
This distribution has implications, not only for species
conservation, but also for the contribution that each
population makes to the functioning of ecosystems, and
raises the important issue of how variability in the number
of individuals in a given population affects functi6nality.
For ubiquitous species for which discrete populations
cannot be identified, variability in population density
might be a useful surrogate.
Examples in which pollination services are disrupted
when the population size falls below a certain level.
Moreover, they also show how changes in population
density (for plants or pollinators) can alter pollination
success. For example, which found that seed production

in Lesquerella fendleri was greater for plants with a high
density of conspecifics, possibly as a consequence of the
behaviour of the associated pollinators. In detail, a concept
that they refer to as 'ecologically effective population
densities', for which dramatic changes in the densities of
interactive species (e.g. predators and their prey) can have
substantial consequences for the functioning of ecosystems.
Hence, not only is the number of individuals in each
population important, but it also appears that the number
d individuals in a given area requires careful
consideration. A SPU can become 'functionally extinct' at
very low densities or abundance.
Population Distribution
The third component of population diversity is the spatial
distribution of the populations under study. The important
measures here are the extent of the populations relative
to their maximum possible extent in a defined area, and
population dispersion. Focusing on maximum possible
extent facilitates comparisons between species with large
geographical ranges and those with relatively restricted
ranges.
Geographical range is a measure of the maximum
area in which a species can provide a given service. An
assessment of population extent might consider, for
example, the number of sites that populations occupy
relative to all possible sites available for occupancy. This
would be applicable in landscapes in which populations
are tied to spatially discrete habitat patches and it is
relatively easy to identify suitable, but currently
unoccupied patches.
If data are available, the current extent of populations
could also be compared with the historic geographical
range of the target species (e.g. before a major perturbation,

such as the European colonisation of Australia). Population
dispersion is a measure of the spatial aggregation of
populations and how this can affect the delivery of
services over a given area. For example, the services
provided by fish populations might be extremely localised
owing to natural (e.g. shorelines) or anthropogenic
boundaries (e.g. regional fisheries) and the adaptation of
popUlations to a narrow range of environmental
variability.
Deep-rooted tree species, which can control water-
table levels and thus reduce the adverse effects of dryland
salinity in agricultural regions of southern Australia, might
be most effective in controlling salinisation if dispersed
among various recharge areas (where the water table is
replenished), rather than clustered in a single area or only
among discharge areas.
Genetic Differentiation of Populations

The final component of population diversity is genetic
differentiation within and among populations. From both
conservation and ecosystem service perspectives, more
genetic variation within populations might confer greater
resilience in the face of environmental change (such as the
possible stabilising influence of genetic diversity in key
tree species enabling forests to track climate change;.
Genetic differentiation among populations can be
associated with different types or levels of a given
ecosystem service.
For instance, some grass populations (e.g. Agrostis
tenuis) have evolved genetic characteristics that enable
them to colonise mine tailings and other sites with soil
mineral concentrations that are lethal to other populations
of the same species, giving them value as soil stabilisers.
Perhaps the most important ecosystem service in which

study of genetic differentiation of populations is crucial is
in maintaining or augmenting crop yields, which depends
on genetic material that often is only available from
populations of wild relatives of crops.
BIODIVERSITY DECLINE
Focusing on changes in population diversity is a more

comprehensive assessment of biodiversity decline than is
a narrow emphasis on the loss of species alone. Rough
estimates of large-scale population loss have already been
attempted. Consideration of population size, distribution,
genetic differentiation and density are also needed to
reflect accurately the consequences of population change.
This focus requires a meaningful definition of 'population'.
Measurmg changes in populations defined by certain
genetic or demographic criteria, although a useful first
step, might not adequately account for the consequences
of population change for the functioning of ecosystems and
the provision of ecosystem services to humanity. Hence,
the second crucial issue we raise is the value of tying
groups of individuals to the services that they provide and
using this to define a new population category (SPUs).
Ttraditional species-based approaches should be
abandoned when assessing biodiversity decline, but argue
that the consequences of changes in population diversity
merit much more attention than they currently receive. A
comprehensive measure of biodiversity loss should include
both speciesand population-based approaches.
Categorising groups of individuals from the same species
into SPUs, and focusing on changes in SPU diversity, is a
meaningful population-based approach, and one that is
tractable in many crucial economic situations (e.g. crop
pollination).
Reduced diversity has implications for species

conservation and the contribution that a particular species
can make to the provision of ecosystem services. Adequate
population diversity provides insurance against change in
environmental conditions and confers greater flexibility on
ecological communities for coping with anthropogenic or
nonanthropogenic stress.
14
Model Systems for Ecology
Ecology, has only a few putative model systems. Ecologists

have not taken full advantage of the power of model
systems, and that natural MICROCOSMS are worth
considering as such models. Model systems have three
useful features: tractability, generality and realism, which.
enable future experiments to build on previous results.
Ecologists have adopted many systems that meet
some but not all of these requirements. For example,
laboratory-assembled communities of protozoa enable
quick, precise and ~ghly replicated experiments (i.e. have
high tractability), but have been criticised for their
artificiality (i.e. have potentially low generality).
Whole-ecosystem experiments represent the opposite
extreme: entire natural communities (Le. are highly
realistic) whose large size usually necessitates poorly
replicated, long duration and mechanistically simple
studies (Le. have potentially low tractability). There has
been heated debate about the apparentmerits of 'replication
versus realism' in ecology. Natural microcosms potentially
offer a way to circumvent this tradeoff between artificiality
and tractability.
Natural microcosms are small contained habitats that
are naturally populated by minute organisms. Examples
Model Systems for Ecology 241
include the protozoan and metazoan communities of

aquatic PHYTOTELMATA (pitcher plants, bromeliads,
treeholes, etc.), microarthropod communities of moss
patches, micro-crustacean communities in rock pools,
invertebrate communities on mollusk shells, and beetles
in fungal sporocarps.
NATURAL MICROCOSMS
Ecologists are often first drawn to natural microcosms for

entirely practical reasons. These advantages can be
summarised as small size, restricted movement and fast
temporal dynamics. Natural microcosms are all, by
definition, small habitats. The small habitat size enables
high replication in experiments, and thus sufficient power
in observational studies to remove covariate effects
statistically. Natural microcosms also tend to be contained
habitats, that is, they have a clearly delineated arena for
species interactions.
Describing such local (or localised) communities is
relatively simple. By contrast, food webs and communities
in more continuous habitats tend to have 'fuzzy'
boundaries, creating one of the oldest problems in ecology:
the precise delineation of communities. Finally,
experimental manipulations of animal communities in
continuous habitat are often complicated either by the
unwanted movement of individuals into and out of the
study arena, or by fence effects when movement is
artificially constrained.
By contrast, the physical boundaries of natural
microcosms (typically air-water, but also interfaces such
as soil-rock) represent a natural constraint for biota, which
facilitates the addition or removal of species or even the
reassembly of an entire community from scratch. These
manipulations in natural microcosms are usually
maintained long enough for the purposes of many

experiments.
However, natural microcosms are not closed systems.
Eventually, all manipulations will be altered by processes
such as the emergence of adult insects, oviposition of eggs
and colonisation by microorganisms. The organisms in
natural microcosms also tend to be small, typically
including: insects and smaller arthropods (amphipods,
mites, collembola, etc.), annelids, microcrustaceans (e.g.
ostracods and Daphnia), metazoa and protozoa (e.g.
rotifers, ciliates and flagellates) and bacteria.
In the case of most microscopic organisms «1 mm),
the fast generation time enables experiments to run for
many generations, allowing ecologists to test theory about
both short- and long-term effects of manipulations with
experiments that last only several weeks or months.
Experiments with larger organisms, by contrast, are
sometimes criticised for only capturing short, transient
dynamics, even when these experiments are conducted
over several years.
The usefulness of natural microcosms will depend not
only on their properties, but also on the theoretical
questions being addressed. All theory makes assumptions
about the scale and importance of ecological processes.
These assumptions might be met by only some natural
microcosms, or by only some taxa in a natural microcosm,
or not at all. For example, both rotifers and mosquitoes
are found in pitcher plants.
However, rotifers complete their lifespan within an
individual pitcher, whereas only the larval stages of
mosquitoes occur in pitchers (the adults emerge and utilize
the entire bog within which the pitcher is located). The
drying of a pitcher can thus be used to study the effects
of disturbance on METAPOPULA TION dynamics for
rotifers, but only local variation in survivorship for
mosqui toes.
Recently, natural microcosms have proved to be

ideally suited, in terms of scale and process, to testing two
new and very active areas of ecological theory: the effect
of declining diversity on ecosystem function, and the
effects of neighbouring communities (the
'metacommunity') on species richness.
BIODIVERSITY-ECOSYSTEM FUNcnON
Community ecologists have recently focused on

understanding the effect of species loss on the rates and
stability of ecosystem functions. To date, most experiments
have involved assembling random subsets of monotrophic
communities, such as grassland plants in I-m2 plots. Such
manipulations of synthetic communities have been useful
for the development of theory, but the current challenge
is to extend these results to multitrophic food webs of
coevolved species experiencing real patterns of species loss.
Natural microcosms have the potential to playa
particularly important role at this stage of the diversity-
function research programme, because real local extinctions
can be easily (and ethically) induced by changes in the
habitat, and because responses can be tracked over
multiple generations and through multiple trophic levels.
Local extinction has been experimentally induced by
means of fragmenting moss patches, and the effects
tracked in terms of microarthropod biomass. Declines in
community biomass were temporally decoupled from
species extinctions because of two patterns:
(i) rare microarthropod species were lost first with
minimal impact on community biomass; and
(ii) biomass was most affected by the declines in
abundance of common species that foreshadowed their
eventual extinction. This demonstrates that real
patterns in extinctions might have a lag in affecting
ecosystem function, causing a 'functioning debt'.
Adding corridors between the habitat fragments

reduces both extinctions and the loss of function A second
example is offered by diversity-stability studies in rock
pools. Recent debate has focused on whether increased
diversity can reduce variability in ecosystems, whether this
reduction in variability is expected at population or
community levels and under which conditions such an
effect can occur.
By using natural variation between rock pools in
faunal diversity, positive diversity-stability relationships
have now been shown at both population and community
levels, but only when the confounding effects of
environmental variation or habitat specialisation are
removed. In all of the above examples, natural microcosms
have provided important 'real world' tests of diversity-
function theory developed frdm synthetic communities,
and suggest how theory can be modified to incorporate
real patterns in extinctions or community assembly.
Natural microcosms enable tests of metacommunity
theory Natural microcosms are often embedded in a
hierarchical spatial structure, which is ideally suited to test
metacommunity theory. The crucial insight of
meta community theory is that a single ('local') community
can be significantly affected by dispersal and extinction
events in the surrounding region. Metacommunties are
often, but not necessarily, modeled as arrays of patches,
with one community per patch. Such models are well
approximated in nature by clusters of natural microcosms.
It is also relatively easy to manipulate crucial
parameters in natural microcosms, such as species richness
per patch, dispersal rate and spatial structure. For example,
dispersal between natural microcosms has been modified
by adding corridors (e.g. between patches of moss) or by
experimental additions (e.g. rotifers and protozoa pipetted
between pitcher plants or seeds added to riverine
tussocks). Researchers have laid out groups of marine pen

shells and fungal-sporocarps in different patterns to
examine the effects of spatial structure on the invertebrates
inhabiting these natural microcosms.
Several metacommunity models assume that the
extinction rate in a patch increases with patch species
richness. This was first demonstrated experimentally using
Daphnia communities of coastal rock pools.
Metacommunity models also often assume that
colonisation rate increases with the number of
neighbouring patches; there is also evidence from coastal
rockpools for this effect.
Interestingly, in desert rockpools , local richness can
be adequately modeled without requiring competition to
cause extinctions; rather, local richness is a dynamic
equilibrium between colonisation from the surrounding
region and extinction resulting from desiccation. In a
second metacommunity model, local diversity is predicted
to be maximised at intermediate levels of dispersal.
At low dispersal rates, dominant competitors exclude
other species, whereas, at high dispersal rates, the species
with low dispersal abilities are unable to persist regionally,
reducing local richness. The first test of this model was
carried out in pitcher plants, and involved manipulating
the dispersal rate of microorganisms by pipetting specific
water volumes between pitchers.
Local diversity was highest at intermediate dispersal
rates, as predicted by the model, but this pattern
disappeared in the presence of predators. A related model
predicts that the strongest correlation between local and
regional richness will occur at intermediate assembly times.
With marine organisms assembling on pen shells supports
this prediction, but only for the motile organisms,
suggesting that dispersal mode must be considered.
A third group of models have shown that species co-

existence at the metacommunity level can be facilitated
not only by local effects, such as resource heterogeneity,
but also by larger scale effects, such as patterns of spatial
aggregation. As aggregation of the superior competitor
increases, it occupies fewer patches, creating a probabilistic
refuge for inferior competitors.
Some of the best evidence for the aggregation
hypothesis has come recently from natural microcosms,
specifically insect larvae in fungal sporocarps, carrion and
fallen fruit. Many of the practical advantages that for
natural microcosms apply just as well to artificial
microcosms (e.g. beaker and bottle experiments). Indeed,
it could be argued that artificial microcosms have further
advantages, as the even greater control of habitat enables
very subtle effects to be detected with reasonable power.
However, the main criticism of beaker experiments
has never been tractability, but rather the general
applicability of the results. Naeem describes artificial
microcosms as having high 'internal validity' (defined as
transparency of mechanisms, and similarity to theoretical
constructs) but low applicability to natural ecosystems, or
'external validity'.
For example, artificially assembled communities do
not represent coevolved or co-occurring taxa with natural
abundance distributions. One could argue that the very
features that make natural microcosms unusually tractable,
such as small size and discrete boundaries, also make them
ecological oddballs. For example, it could be argued that
the small size of natural microcosms could make them
particularly prone or sensitive to disturbance. Freezing,
drought and treefall can have catastrophic effects on the
fauna of individual treeholes but minor effects on the
larger forest ecosystem. In a rock pool, diurnal variation
in temperature exceeds that in a large body of water.
Organisms often cannot flee these disturbances

because of the discrete boundaries; larval aquatic insects,
for example, are trapped in a drying treehole. One could
argue, therefore, that the ecology of natural microcosms
will be dominated by non-equilibrium dynamics, in which
dispersal, local extinction and spatial structure play key
roles. There are two counterarguments to this proposition.
First, such processes are probably important in all
communities, just more evident in a contained and rapidly
changing system. Second, most taxa in natural microcosms
have short generation times. Once dispersal and
disturbances are scaled by generation times, there might
be little difference in their frequency or intensity between
natural microcosms and any other system, including larger
systems. Spatial effects could be similarly scaled by body
size of the organisms involved.
Unfortunately, this type of biological scaling has
rarely been investigated by ecologists, but is a fertile area
for future research. Such information would determine to
what degree results from natural microcosms could be
compared to those from larger areas, such as human-
dominated landscapes. These effects of system size are a
double-edged sword, not only limiting the scaling-up of
results from small systems to large systems, but also the
applicability of large system studies to smaller systems.
The onus is not only on ecologists who work with
small systems to prove the relevance of their results to
larger systems, but also the reverse. If a general ecological
theory does not apply to natural microcosms, or to lakes,
or to any common type of ecosystem, then it is, by
definition, not general. The challenge is to develop theory
applicable to a range of ecosystem sizes.
Each pitcher, or water-filled leaf, of a pitcher plant
contains a discrete food web consisting of aquatic insects
andmicroscopic organisms. The basal energy for this food
web is provided by the carcasses of drowned terrestrial

insects. The pitcher plant illustrated, Sarracenia purpurea,
contains ca. 3-12 such pitchers per plant. Sarracenia pitcher
plants occur near other pitcher plants within the same bog,
and bogs often occur near other bogs in the landscape.
Thus, the processes that affect a pitcher plant food web
occur in a hierarchical order of spatial scale, from pitchers
to plants to bogs to region.
Food web within ~ pitcher
Clusler of
pitcher plants
Bog
Figure 1. Pitcher Plant Communities and Hierarchical Spatial Structure
At the local (pitcher) scale, species diversity of many

microscopic organisms (bacteria, rotifers and protozoa) is
strongly affected by both TOP-DOWN (mosquito
occurrence) and BOTTOM-UP (carcass provisioning)
processes. Mosquito occurrence at the pitcher scale is
affected by bottom-up effects, such as carcass provisioning,
carcass processing by midges and pitcher size. Importantly,
these local effects on mosquito occurrence are modified
by landscape-scale features, such as pitcher density and
location, probably because of the behaviour of adults

during oviposition.
Going one scale larger, differences between bogs in
pitcher plant arthropods are related to features such as
bog size, isolation and exposure, and reflect processes of
colonisation and population extinction. Finally, at the
continental scale, highest protozoan and bacterial richness
occur at the highest latitudes. This unusual patternis linked
to climatic limits to the northern range of the mosquito
and, thus, ultimately to the same top-down processes seen
at the local level.
15
Biodiversity, Ecosystem Services and
the UN Millennium Declaration
The strong international consensus behind the Millennium

Declaration makes it a unique and powerful basis for
development cooperation. There is dawning recognition
among decision-makers that biodiversity and ecosystem
services are central to achieving the Millennium
Development Goals (MDGs). This has created extremely
important opportunities: biodiversity specialists can make
a decisive contribution to the UN strategy for the MDGs,
while the MDGs offer a gateway opportunity to advance
conservation and sustainable use of biodiversity.
The 'WEHAB' framework launched by the Secretary-
General in May 2002 reinforced recognition of biodiversity, ,
creating further opportunities. The UN strategy for the
MDGs has four main elements: the Millennium Project;
the Millennium Campaign; Country-Level Monitoring; and
Country-Level Operations of the UN agencies. The
Secretary-General's Road Map, which presents a plan for
achieving the goals set out in the Millennium Declaration,
focuses on the MDGs, among other things providing an
indicator framework.
Ecosystem services such as soil protection, pollution
control and water purification have great economic value.
Ecosystems proyide food, health care and income
Biodivfrsity, Ecosystem Services and the UN ... 251
generation opportunities. The sharpest interface between

conservation and sustainable use of biodiversity and
poverty reduction is at the local community level, which
is also where ingenuity, learning and innovation can be
found, as exemplified by the projects selected as finalists
for the Equator Initiative awards in 2002. Of course,
biodiversity also has an intrinsic value, as the Preamble
to the Convention on Biological Diversity (CBD)
recognises, and this is interwoven with cultural, spiritual
and other human values.
A key challenge now is how to 'scale up': how to
capture the lessons generated by projects such as those
recognised by the Equator Initiative and translate them
into wider application at the policy-making level.
NEW PARADIGM FOR BIODIVERSITY AND ECOSYSTEM SERVICES
A powerful new understanding is emerging about the

central role biodiversity and ecosystem services have to
play in the achievement of the MDGs. Initiatives such as
the Millennium Ecosystem Assessment and the Equator
Initiative signalled the emerging change. The Equator
Initiative has published a series of papers, in cooperation
with the International Institute for Environment and
Development (IEED) and others, that have helped set the
agenda for the new debate. Other developments include,
for example, a major paper on 'Linking Poverty Reduction
and Environmental Management', prepared for the World
Summit on Sustainable Development (WSSD) by the UK
Department for International Development (DFID), the
European Commission, UNDP and the World Bank.
In a ground-breaking speech in May 2002, UN
Secretary-General Kofi Annan set out proposed priorities
for the WSSD under the innovative 'WEHAB' framework,
recognising biodiversity and sustainable ecosystem
management as one of the five priorities: water, energy,

health, agriculture and biodiversity. The WEHAB concept,
which provided a basis for major plenary discussions at
the WSSD, is proposed here to be of fundamental
importance to the UN strategy for the MDGs.
All five WEHAB priorities, including biodiversity, are
essential for achieving the MDGs. It is important to note
that the WEHAB concept does not represent a return to a
'sectoral' approach: integration of the five themes is a
defining feature. The biodiversity area has lacked the
targets and timetables that are usually needed to redirect
policymaking. This has begun to change. The Plan of
Implementation adopted at the WSSD contains a target of
achieving a significant reduction in the currmt rate of
biodiversity loss by 2010, which echoes a similar call by
Ministers at the Sixth Conference of the Parties (COP 6)
of the CBD.
In follow-up of the call made at COP 6, the WEHAB
Working Group published A Framework for Action on
Biodiversity and Ecosystem Management in August 2002.
Crucially, the WEHAB working group paper on
biodiversity highlights the need to shift focus from the
proximate causes of biodiversity loss to the underlying
causes. It focuses on two key Action Areas: integration of
biodiversity in country development programmes and
economic sectors; and halting the loss of biodiversity and
restoring, if possible, biodiversity in degraded areas, as
part of reversing loss of environmental resources.
Importantly, reflecting the close linkage between the
WEHAB framework and the MDGs, the two Action Areas
in this WEHAB paper are built upon and consistent with
targets of MDG 7 to 'ensure environmental sustainability.'
The action frameworks provide indicative targets or
milestones, with ~xamples of activities - a 'menu' for
further development of activities.
Biodiversity, Ecosystem Services and the UN ... 253
The linkages between biodiversity and the MDGs are

far from fully understood. Papers such as those produced
by the Equator Initiative and lIED, and the recently
released Interim Executive Summary of a project of the
Center for Health and the Global Environment at Harvard
Medical School, with WHO and UNEP, entitled:
Biodiversity: Its Importance to Human Health, have begun
to explore the issues. However, there is an urgent need
for further analytical work, in particular relating to the
indicator framework for the MDGs. The framework was
developed in consultations involving the UN, the
International Monetary Fund (IMF), the Organisation for
Economic and Cooperation and Development (OECD) and
the World Bank.
The framework is crucial for making the linkages
between biodiversity and the MDGs. However, it is widely
recognised that the existing framework needs further
development, and it is important that biodiversity
specialists contribute to this.
THE UN MILLENNIUM DECLARATION
The UN Millennium Declaration, adopted in September

2000 by 189 UN Member States, reaffirmed the
international community's commitment to the fundamental
values of the UN. In the Millennium Declaration,
governments confirmed that people have a right to a future
free from poverty, hunger, violence, oppression and
injustice. Governments made a high-level commitment to
achieving specific goals in the following areas:
Peace, security and disarmament;
Development and poverty eradication;
Protecting our common environment;
Human rights, democracy and good governance;
Protecting the vulnerable;

Meeting the special needs of Africa; and
Strengthening the UN.
In addition to development and poverty eradication, the
goals of the Millennium Declaration address issues such
as the world drug problem; human rights; protection of
civilians in complex emergencies; helping Africa tackle the
spread of the HIV / AIDS pandemic and other infectious
diseases; and giving greater opportunities to non-
governmental organisations (NGOs) and other civil society
actors and the private sector to contribute to the realisation
of the UN's goals and programmes. Partnerships and
stakeholder involvement are important elements of the UN
strategy for the MDGs. A key challenge in the area of
biodiversity will be to expand current efforts beyond the
UN and intergovernmental processes.
SECRETARy-GENERAL'S ROAD MAP AND THE MILLENNIUM

DEVELOPMENT GOALS
In a resolution following the Millennium Declaration, the

UN General Assembly requested the Secretary-General to
prepare a 'Road Map', setting out in detail how the
commitments made in the Millennium Declaration could
be achieved. From the outset, the Secretary-General's Road
Map recognised certain key factors:
Most of the targets set by the Millennium Declaration
are not new. They derive from the global conferences
of the 1990s and from the body of international norms
and laws that have been codified over the past half-
century.
The action plans for reaching the targets have for the
most part already been agreed and adopted.
The missing factor is political will to implement the
commitments that governments have made.

National reporting is a key part of the UN strategy for
the MDGs, a responsibility entrusted to UNDP by the UN
Secretary-General. User-friendly MDG reports measure
progress and raise awareness. Several countries have
already submitted reports. The Secretary-General will
,report on progress on the Millennium Declaration to the
'UN General Assembly, focusing on selected themes each
year, with a comprehensive progress report every five
years. In 2002 the themes are preventing armed conflict
and the treatment and prevention of diseases, including
HIV / AIDS and malaria.
In 2003, the Secretary-General will report on the
themes of financing for development and strategies for
sustainable development. All the goals in the Millennium
Declaration and the Secretary-General's Road Map are
meant to support and reinforce each other, so achievement
of the MDGs is linked to achievement of the other goals.
With the UN strategy for the MDGs, a plan for
achievement of the MDGs is advancing, with targets and
indicators to measure progress against. There could be
scope to develop similar targets and/or indicators for the
other goals of the Millennium Declaration.
THE UN STRATEGY FOR THE MILLENNIUM DEVELOPMENT

GOALS (MDGs)
The UN strategy for the MDGs strikes at the heart of the

development challenge, with a coordinated approach on
issues such as hunger, health and environmental
sustainability. The strategy comprises four main elements:
The Millennium Project is a global research project
which analyses policy options, with the aim of
identifying the best strategies for achieving the MDGs.
Ten thematically oriented Task Forces, which include
members from the public and private sectors, and

from civil society organisations, are undertaking
research. The UN Experts Group for the Millennium
Project facilitates collaboration with the UN system
to ensure the project builds on the best work carried
out by the UN agencies. There is strong support for
the Millennium Project's intellectual autonomy to put
forward optimal strategies to the Secretary General.
Dr. Jeffrey Sachs, Special Advisor to the UN Secretary-
General Kofi Annan, leads the UN Millennium Project.
The Millennium Campaign, led by Eveline Herfkens,
the Secretary-General's Executive Coordinator, and
former Minister of International Cooperation,
Netherlands, is aimed at mobilising political support
for the MDGs. The campaign will initiate and
coordinate national campaigns around the world to
raise awareness on the MDGs.
Country-Level Monitoring is led by the UN and
coordinated by Jan Vandemoortele from UNDP. Key
outputs are national MDG reports, which support
advocacy and building awareness on MDGs.
Country-Level Operations of the UN agencies are led
and coordinated by the UN Development Group
(UNDG).
As highlighted by a recent article by Professor Sachs in
The Economist magazine, donor countries need to engage
fully in the UN strategy. Millennium Development Goal
8, to 'develop a global partnership for development', is
crucial.
BIODIVERSITY AND THE MILLENNIUM DECLARATION
Biodiversity is linked to many of the goals of the

Millennium Declaration. For example, the section on
human rights, democracy and good governance in the
Secretary-General's Road Map addresses issues of concern

to indigenous communities, recognised as having a unique
role in relation to biodiversity. The section on 'Protecting
our common environment', as set out in the Secretary-
General's Road Map, is particularly relevant. It includes
the following goals:
To make every effort to ensure the entry into force of
the Kyoto Protocol, preferably by the tenth
anniversary of the UN Conference on Environment
and Development in 2002, and to embark on the
required reduction in emissions of greenhouse ga~es;
To intensify our collective efforts for the management,
conservation and sustainable development of all types
of forests;
To press for the full implementation of the Convention
on Biological Diversity and the Convention to Combat
Desertification in those Countries Experiencing Serious
Drought and/ or Desertification, particularly in Africa;
To stop the unsustainable exploitation of water
resources by developing water management strategies
at the regional, national and local levels which
promote both equitable access and adequate supplies;
To intensify our collective efforts to reduce the number
and effects of natural and man-made disasters.
The goal to ' .... press for full implementation of the
Convention on Biological Diversity (CBD) and the
Convention to Combat Desertification in those Countries
Experiencing Serious Drought and/or Desertification,
particularly in Africa' is a key commitment on biodiversity.
This paper focuses on the CBD, but it should be noted
that full implementation of the Desertification Convention
and the UN Framework Convention on Climate Change
(UNFCCC) must be a high international priority.
Governments have recognised that implementation of

the CBD and the Desertification Convention (and the
UNFCCC and its Kyoto Protocol) need to be coordinated
closely. The CBD provides a mechanism for cooperation
across many issue areas which can support achievement
of the MDGs. Indicators, the Convention on Biological
Diversity (CBD) and the Millennium Declaration Goal 7
of the Millennium Development Goals is to 'ensure
environmental sustainability,' and the targets are:
Target 9. Integrate the principles of sustainable
development into country policies and programmes
and reverse the loss of environmental resources
Target 10. Halve, by 2015, the proportion of people
without sustainable access to safe drinking water
Target 11. By 2020, to have achieved a significant
improvement in the lives of at least 100 million slum
dwellers.
The base year for the targets is 1990. The targets are
accompanied by seven indicators. There are multiple links
between the indicators and biodiversity. For example,
Indicator # 29 addresses the proportion of population with
sustainable access to an improved water source - protected
areas and functioning ecosystems can help maintain water
supplies. Other indicators relate to health - as noted, close
links exist between biodiversity and human health.
Here addresses two indicators for Millennium
Development Goal 7, which are also priority issues for the
Convention on Biological Diversity:
Indicator # 25. Proportion of land area covered by
forest
Indicator # 26. Land area protected to maintain
biological diversity
The issue areas covered by these indicators (forests and
protected land area) are mission-critical to the CBD. The

CBD, which covers a broad range of issues at the centre
of the North-South agenda, from biosafety and forests to
endangered species and benefit-sharing, is a mega-
convention that has not yet come into its own. Critics have
focused on its tendency to process, rather than results, and
its difficulties in setting priorities that could focus the work
of the vast, unwieldy convention. In April 2002, the Sixth
Conference of the Parties to the CBD (COP 6) marked the
tenth anniversary of the CBD.
The dramatic decline in world forest cover and
degradation of forest ecosystems has been a focus of
international concern. There have been hopes that the CBD
could make progress where other international bodies have
failed. The adoption of a revised work programme on
forests was one of the priority agenda items at COP 6.
While this reflected progress, it does not appear to have
caused a breakthrough. The forest issue is related closely
to protected areas. The CBD is' due to consider protected
areas or 'areas where special measures need to be taken
to conserve biological diversity' as one of the main themes
at the next, Seventh Conference of the Parties (COP 7) in
March 2004. Prior to this, the CBD's Subsidiary Body on
Scientific, Technical and Technological Advice (SBSTTA)
will consider protected areas.
The World Parks Congress in September 2003 will be
another important preparatory event. A successful COP 7
outcome on protected areas could bolster the revised work
programme on forests. With protected areas as a main
theme, CBD COP 7 will be in a position to take decisions
which contribute directly to achievement of Goal 7 of the
MDGs, through indicators #25 and #26. This could be an
opportunity for biodiversity to step into the international
decision-making arena and for the CBD to move from
special-interest convention into the international
mainstream. The CBD may also wish to consider how it
can ensure that the Millennium Declaration goal of

achieving full implementation of the CBD is reached.
Presumably, the ideal scenario is satisfactory progress
towards the objectives of the CBD: conservation and
sustainable use of biological diversity, and fair and
equitable sharing of benefits from the utilisation of genetic
resources, including access and transfer of technology. A
reality-check of implementation to date and continuing
human-caused loss of species and habitats suggests that
this is still far off.
The CBD could consider setting specific criteria for
full implementation for the purposes of the Millennium
Declaration, by identifying high-priority actions that are
closely and directly linked to achievement of the MDGs
and other goals. Recent efforts to adopt a Strategic Plan
for the CBD suggest that this may be difficult, but the
goals of the Millennium Declaration could provide an
opportunity to refine further priorities on the basis of the
Strategic Plan. The UN General Assembly considers the
CBD annually. With full respect for the CBD's Conference
of the Parties, as the decision-making body of the CBD,
the General Assembly could address some of the links
between the CBD and the goals of the Millennium
Declaration.
The future may view the MDGs as extremely low
in aspiration, for a world as rich as ours. For example,
Target 2 of the goal to 'Eradicate extreme poverty and
hunger' (Goal 1) accepts that hundreds of millions of
women, children and men will still suffer from hunger in
the year 2015. Here and now, the political and practical
challenges of halving the proportion of people who suffer
from hunger by that date are daunting.
Biodiversity specialists, who tend to be creative
pioneers of on-the-ground solutions, can bring enormous
commitment and expertise to the UN strategy for the
Biodiversity, Ecosystem Services and the UN . 261
MDGs. In tum, the UN strategy can help raise biodiversity

higher on the agendas of key decision-makers. Biodiversity
specialists have, started to succeed in making their voice
heard beyond their own community in 2002, as evidenced
by the recognition of the role biodiversity has to play in
the achievement of the MDGs and the inclusion of
biodiversity as one of the five WEHAB priorities. As many
observers have recognised, if this is to tum into a real
breakthrough, the momentum has to be maintained. It
should be noted that while the MDGs and the CBD are
important routes for moving the biodiversity agenda of
WEHAB forward, they are not the only ones.
The response to the key challenges of 'scaling up'
experiences from the local community level and taking the
'B' in WEHAB forward can draw on the great knowledge
-both formal and informal-which can be found in the
biodiversity community. It can build on the strong, diverse
partnerships and networks that bring together
implementers of local projects, campaigners, scientists and
international policy specialists in adaptive learning
processes within this community. Events such as the
World Parks Congress in 2003 and the Seventh Conference
of the Parties (COP 7) of the CBD in 2004 will be critical
-if they succeed in reaching out beyond their specialist
audiences. The main test may be whether the biodiversity
community can mobilise its formidable resources and take
full advantage of the new appreciation of biodiversity and
the enthusiasm generated by recent developments.
Bibliography
Abe, Takuya, Simon A. Levin, and Masahiko Higashi, eds.

Biodiversity: An Ecological Perspective. New York: Springer,
c1997. Biodiv QH541.15 .B56 B573 1997.
Cook, Laurence Martin. Genetic and Ecological Diversity: The
Sport of Nature. London: Chapman and Hall, 1991. Main
QH75 .C77 1991.
Cropper, Simon C. Management of Endangered Plants. E.
Melbourne, Australia: CSIRO Publications, 1993. Biodiv
REF QH77 .A8 C74 1993.
Freese, Curtis H. Harvesting Wild Species: Implications for
Biodiversity Conservation. Baltimore, MD: Johns Hopkins
University Press, 1997. Biodiv QH75 .H38 1997.
Groombridge, Brian and Martin Jenkins. Global Biodiversity:
Earth's Living Resources in the 21st Century. Cambridge,
UK: World Conservation Press, c2000. Biodiv REF
QH541.15 .B56 G76 2000.
Johnson, Nels. Biodiversity In the Balance: Approaches to Setting
Geographic Conservation Priorities. Washington, D.C.:
Biodiversity Support Program, 1995. Main QH75 .J64 1995.
Kim, Ke Chung and Robert D. Weavers, eds. Biodiversity and
Landscapes: A Paradox of Humanity. NY: Cambridge
University Press, 1994. Main QH75 .B53 1994.
Kramer, Randall A., Carel van Schaik, and Julie Johnson, eds.
Last Stand: Protected Areas and the Defense of Tropical
Biodiversity. New York: Oxford University Press, 1997.
Main QH75 .L365 1997.
Bibliography 263
Magurran, Anne E. Ecological Diversity and Its Measurement.

Princeton, NJ: Princeton University Press, 1988. Biodiv
QH75 .M32 1988.
Master, Lawrence L., Stephanie R. Flack, and Bruce A. Stein,
eds. Rivers of Life: Critical Watersheds for Protecting
Freshwater Biodiversity. Arlington, VA: Nature Conservancy
in cooperation with natural heritage programs and
Association for Biodiversity Information, c1998. Biodiv
REF H541.5 .F7 R58 1998.
Norton, Bryan.G. The Preservation of Species: The Value of
Biological Diversity. Princeton, NJ: Princeton University
Press, 1986. Main QH75 .P74 1986.
Orians, Gordan H., et al. Preservation and Valuation of Biological
Resources. Seattle, WA: University of Washington Press,
1990. Main QH75 .P73 1989.
Perrings, Charles. Biodiversity Loss: Economic And Ecological
Issues. New York, NY: Cambridge University Press, 1997.
Biodiv QH541.15.B56 B584 1997.
Pickett, Steward T. The Ecological Basis of Conservation:
Heterogeneity, Ecosystems, and Biodiversity. New York:
Chapman & Hall, 1997. Biodiv QH75 .E25 1997.
Van der Ryn, Sim. Ecological Design. Washington, D.C.: Island
Press, 1996. Pub GE170 .V36 1996.
Index
Agricultural ecology, 155 Designate a National Authority

Agricultural support policies 193 (DNA) 48
Agro-Biodiversity 160 Dilution-effect model 18
Agro-ecological zones 160 Disturbance-driven systems 80
Alternative mechanism 20 Diversity-functioning relationship
Analytical spectrometry 156 62
Antagonistic effect 8
Anthropogenic impacts 199 Ecological Forecasts 191
Autotrophic biomass production Ecological Models 204
102 Ecosystem service 229
Effective reservoir competence 10
Biodiverse agriculture 163 Environmental change 1
Environmental fluctuations 74
Certified Emission Reductions Environmental heterogeneity 63
(CERs) 48 Environmental variability 2
Classical equilibrium approaches Equilibrium theoretical frame-
75 work 74
Clean Development Mechanism Evolutionary Units (EUs) 231
(CDM)47 Extrinsic environmental fluctua-
Community organisations 172 tions 75
Conservation Opportunities 171
Conservation Units (CUs) 231 Fluvial geomorphology 190
Consultative Group on Interna-
tional Agricultural R 158 General Agreement on Tariffs and
Crimean-Congo Hemorrhagic Trade (GATT) 170
Fever (CCHF) 13 Geographic information system
(GIS) 194
Deep-sea ecosystems 101 Globalisation 169
Demographic Units (DUs) 231
Density of infected nymphs (DIN) High Yielding Varieties (HYVs)
5 165
Index 265
High-intensity systems 151 Predator-prey interactions 155

Host-species-rich community 11
Reservoir competence 9,16
Instream Flow Incremental
Methodology (IFIM) 197 Service-Providing Unit (SPU) 232
Interdisciplinary models 211 Soil microorganisms 152
Species diversity 67
Livestock diversity 165 Species-area relations 77
Local ecosystem functioning 65 Species-poor vertebrate communi-
Low-flow mechanisms 199 ties 6
Multi-Resolution Land Character- Three-dimensional relationship

istics (MRLC) 196 101
Trade-Related Intellectual Prop-
National Environmental Observa- erty Rights (TRIPs) 170
tory Network (NEON) 207 Traditional diversity 173
National Science Foundation's Transovarial transmission 6
(NSF) 207
Non-agricuIturalland uses 169 United Nations Millennium
Nymphal Infection Prevalence Development Goals 49
(NIP) 5
Vector-borne Zoonotic Diseases 3
Older distribution systems 200
Zoonotic cutaneous leishmaniasis
Population Diversity 234 (ZCL) 14
Potential ecological consequences Zoonotic pathogens 17
61

Ecology and Biodiversity

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ecology and Biodiversity

Uploaded by

Copyright:

Available Formats

ECOLOGY AND BIODIVERSITY

"This page is Intentionally Left Blank"

© Pragun Publications, 2006

Published in India and Printed at :

liP Printers, New Delhi -110015 Mobile: 9810271526

The term "ecology" was introduced by Haeckel in 1869.

species are insects, which dominate terrestrial and

All species provide some kind of function to an ecosystem.

The mechanisms underlying these effects are complex and

ECOLOGY OF INFEcrIOUS DISEASES

Vector-borne Zoonotic Diseases

Lyme disease and the dilution effect

America, Ixodes pacificus in western North America,

DIN will be a function of both NIP and the local

Some of the experts further analyzed the dilution

and evenness on disease risk, Schmidt and Ostfeld used

competition) or a siphoning of tick meals away from mice.

Unlocking reservoir competence

Laboratory measurements of reservoir competence, on

the less competent reservoirs will tend to be rapid, pushing

Dilution Effect in Vectorborne Zoonoses

Arthropod parasites vary widely in their degree of

in more diverse communities, which is the opposite of the

tularemia, and CCHF is known to parasitise domestic

inefficient or nonexistent, leading to low to zero infection

the pathogen in the bloodstream or other tissues, under

host. Further studies to assess the mechanisms behind both

the model of Schmidt and Ostfeld, which defines host

prevalence in the vector population. The suppressive

(proportional representation of each species). Which metric

critical implication would be that any loss of diversity will

The biological underpinnings are encompassed in the

purification of air and water.

wheatlands or oil palm fields), whose economic value can

list could amount to hundreds or even thousands of

Ecosystem services and the systems that supply them are

whose parts or products are consumed (as meat, milk,

from the molecular to the landscape level. Biodiversity is

Such extractions from biodiversity's igenetic libraryi

population relies on traditional medical systems, and about

Climate and Life

patterns. Even relatively recently, from 1550-1850, Europe

Living things may also enhance warming trends

Mitigation of Floods and Drougltts

On a much larger scale, extensive deforestation in the

In addition to moderating the water cycle, as

process also renders harmless many potential human

animals. Certain types of fungi play extremely important

to both wild and domesticated plants is diminishing: more

Natural Pest Control Services

humans and many other types of organisms; the recently

called Clark's Nutcracker (Nucifraga columbiana), which

THREATS TO ECOSYSTEM SERVICES

Ecosystem services are being impaired and destroyed by

species - the populations that generate ecosystem services

services. Moreover, society often does not compensate

VALUATION OF ECOSYSTEM SERVICES

Human society would cease to exist in the absence of

ecosystem, like valuing a human life, is fraught with

Human activities have led to changes in ecosystems and

aesthetic and recreational values. The effects of climate

MITIGATION AND ADAPTATION

Broadly, the two tools to address to climate change are

particularly at the community level, can supplement