Content uploaded by Bill Hare
Author content
All content in this area was uploaded by Bill Hare
Content may be subject to copyright.
WISSENSCHAFTLICHER BEIRAT DER BUNDESREGIERUNG
GLOBALE UMWELTVERÄNDERUNGEN
WBGU
materialien
William Hare:
Assessment of Knowledge on Impacts of
Climate Change – Contribution to the
Specification of Art. 2 of the UNFCCC
Externe Expertise für das WBGU-Sondergutachten
"Welt im Wandel: Über Kioto hinausdenken.
Klimaschutzstrategien für das 21. Jahrhundert"
Berlin 2003
Wissenschaftlicher Beirat der Bundesregierung Globale Umweltveränderungen
Geschäftsstelle
Reichpietschufer 60–62, 8. OG.
10785 Berlin
Telefon (030) 263948 0
Fax (030) 263948 50
E-Mail wbgu@wbgu.de
Internet http://www.wbgu.de
Alle WBGU-Gutachten können von der Internetwebsite http://www.wbgu.de in deutscher und englischer Sprache herunter geladen werden.
© 2003, WBGU
Externe Expertise für das WBGU-Sondergutachten
"Welt im Wandel: Über Kioto hinausdenken. Klimaschutzstrategien für das 21. Jahrhundert"
Berlin: WBGU
ISBN 3-936191-03-4
Verfügbar als Volltext im Internet unter http://www.wbgu.de/wbgu_sn2003.html
Autor: William Hare
Titel: Assessment of Knowledge on Impacts of Climate Change – Contribution
to the Specification of Art. 2 of the UNFCCC
Potsdam, Berlin 2003
Veröffentlicht als Volltext im Internet unter http://www.wbgu.de/wbgu_sn2003_ex01.pdf
Assessment of Knowledge on Impacts of Climate Change –
Contribution to the Specification of Article 2 of the UNFCCC: Impacts
on Ecosystems, Food Production, Water and Socio-economic Systems
Mr. William Hare
Visiting Scientist
Potsdam Institute for Climate Impact Research
Potsdam, Germany
November 2003
2
Acknowledgments
Claire Stockwell and Kathrin Gutmann are thanked for the excellent work done helping
to research and edit this report.
3
1. INTRODUCTION.....................................................................................................5
UNFCCC ARTICLE 2 – PREVENTING DANGEROUS ANTHROPOGENIC INTERFERENCE.......5
WHAT MAY CONSTITUTE DANGEROUS ANTHROPOGENIC INTERFERENCE WITH THE
CLIMATE SYSTEM? ...........................................................................................................6
2. ECOSYSTEMS, BIODIVERSITY AND CLIMATE CHANGE .........................9
PROCESSES CAUSING LOSS OF BIODIVERSITY AND ECOSYSTEM DAMAGE........................ 11
CLIMATE CHANGE AND CO
2
EFFECTS ON SPECIES AND ECOSYSTEMS............................. 14
PROJECTED EFFECTS ON SPECIES AND ECOSYSTEMS....................................................... 19
Impacts on coastal wetlands..................................................................................... 21
Impacts on animal species ........................................................................................ 21
Impacts on ecosystems..............................................................................................22
3. IMPACTS ON FOOD PRODUCTION, WATER, AND SOCIO-ECONOMIC
SYSTEMS........................................................................................................................ 53
INTRODUCTION .............................................................................................................. 53
CONTEXT: FINDINGS OF THE SECOND AND THIRD ASSESSMENT REPORTS..................... 54
FOOD PRODUCTION AND AGRICULTURE ........................................................................ 61
Climate change and food security assessments........................................................ 64
Global Agro-Ecological Assessment (GAEZ Study)................................................. 71
Discussion and Summary.......................................................................................... 72
WATER RESOURCES ....................................................................................................... 75
Discussion and Summary.......................................................................................... 76
SOCIO-ECONOMIC DAMAGES .......................................................................................... 79
Discussion and Summary.......................................................................................... 81
4. SUMMARY AND CONCLUSIONS..................................................................... 84
ECOSYSTEMS IMPACTS ................................................................................................... 84
Impacts on coastal wetlands..................................................................................... 84
Impacts on animal species ........................................................................................ 85
Impacts on ecosystems.............................................................................................. 86
AGRICULTURE AND FOOD SECURITY IMPACTS................................................................ 87
WATER IMPACTS ............................................................................................................ 87
SOCIO-ECONOMIC EFFECTS ............................................................................................ 88
CONCLUSIONS................................................................................................................ 89
5. APPENDIX: TEMPERATURE SCALE............................................................. 90
6. REFERENCES........................................................................................................ 92
4
List of Figures
FIGURE 1 - PROPORTION OF THE GLOBAL NUMBER OF BIRDS, MAMMALS, FISH AND
P
LANTS
S
PECIES THAT ARE
C
URRENTLY
T
HREATENED WITH
E
XTINCTION
................ 12
FIGURE 2 - PATHWAYS BY WHICH CLIMATE CHANGE AFFECT S SPECIES AND ECOSYSTEMS15
FIGURE 3 - COMPARISON OF HEMISPHERIC AND LONG-TERM LOCAL TEMPERATURE SERIES
................................................................................................................................... 18
FIGURE 4 - COMPARISON OF MAXIMUM DECADAL RATES OF CHANGE............................. 18
FIGURE 5 - IMPACTS ON COASTAL WETLANDS .................................................................. 24
FIGURE 6 – IMPACTS ON ANIMAL SPECIES ......................................................................... 26
F
IGURE
7 – I
MPACTS ON
E
COSYSTEMS
............................................................................... 29
FIGURE 8 - REGIONAL IMPACTS ON CROP PRODUCTION .................................................... 66
FIGURE 9 - COMPARISON OF POTENTIAL CROP YIELDS PROJECTIONS FOR 2050S AND 2080S
................................................................................................................................... 67
FIGURE 10 - GLOBAL RISK OF HUNGER............................................................................. 69
FIGURE 11 - MILLIONS AT RISK IN 2050S AND 2080S: HUNGER, MALARIA, WATER
SHORTAGE AND FLOODING........................................................................................ 69
F
IGURE
12 - G
AINS AND
L
OSSES IN
P
RODUCTION
P
OTENTIAL UNDER
C
LIMATE
C
HANGE
.. 74
FIGURE 13 - CLIMATE DAMAGES OR BENEFITS AS A FUNCTION OF TEMPERATURE ........... 84
List of Tables
TABLE 1 - ECOSYSTEMS FUNCTION WITH LINKS TO GOOD/SERVICES AND POSSIBLE
SOCIETAL VALUE....................................................................................................... 10
TABLE 2 - PROCESSES DRIVING SPECIES ENDANGERMENT AND EXTINCTION .................... 13
TABLE 3 - RESPONSE AND IMPACTS OF CLIMATE CHANGE ON SPECIES AND ECOSYSTEMS 17
TABLE 4 - ECOSYSTEM EFFECTS OF CLIMATE CHANGE....................................................... 20
TABLE 5 - ECOSYSTEM IMPACTS........................................................................................ 32
TABLE 6 - COMPARISON OF SECOND AND THIRD ASSESSMENT REPORT FINDINGS ............ 55
TABLE 7 - AGRICULTURAL EFFECTS OF CLIMATE CHANGE................................................. 62
TABLE 8 - RISK OF HUNGER - AFRICA ............................................................................... 66
TABLE 9 - SUMMARY OF SCENARIOS USED IN GLOBAL FOOD SECURITY ASSESSMENT...... 70
TABLE 10 - MILLIONS AT RISK .......................................................................................... 70
TABLE 11 - MALNOURISHED COUNTRY GROUP AND CLIMATE CHANGE ........................... 71
TABLE 12 - GLOBAL MEAN TEMPERATURE INCREASE FOR ECHAM4 SCENARIOS............ 72
T
ABLE
13 - D
EVELOPING
C
OUNTRY
C
HANGES IN
R
AIN
F
ED
C
EREAL
P
RODUCTION
POTENTIAL 2080S FOR THREE CLIMATE MODELS...................................................... 72
TABLE 14 - WATER RESOURCE EFFECTS OF CLIMATE CHANGE........................................... 78
TABLE 15 - POPULATION WITH POTENTIAL INCREASE IN WATER STRESS.......................... 78
TABLE 16 - SCENARIO TEMPERATURES ............................................................................. 79
TABLE 17 - OTHER MARKET SECTOR EFFECTS OF CLIMATE CHANGE .................................. 83
TABLE 18 - GLOBAL TEMPERATURE SCALES USED IN THIS REPORT .................................. 91
5
1. Introduction
The purpose of this report is to compile and summarise the present knowledge on impacts
of climate change as a basis for a consideration of what may constitute dangerous
anthropogenic interference with the climate system under Article 2 of the United Nations
Framework Convention on Climate Change (UNFCCC). An attempt will be made to
associate projected global mean surface temperature and/or sea level changes with
specific identified impacts and effects in order to assist a discussion on the
operationalization of Article 2. The main emphasis will be on ecosystem effects, food
production, water resources, and sustainable development. Whilst the starting point for
this work will be the findings of the Intergovernmental Panel on Climate Change Third
Assessment Report (IPCC TAR), it will be heavily supplemented by the underlying
scientific literature used in the TAR as well as more recent studies published since the
conclusion of the TAR in September 2001.
The organization of the report is as follows. In this section the context for the current
assessment is outlined including background information on Article 2 of the UNFCCC,
the WBGU tolerable window and the broad findings of the IPCC TAR. Section 2, on
ecosystems, biodiversity and climate change, will review a range of projected impacts on
ecosystems and species. Section 3 summarizes projected effects on food security, water
supply and economic activities. Section 4 will briefly summarize the information
presented in this report.
UNFCCC Article 2 – preventing dangerous anthropogenic interference
The ultimate objective of the United Nations Framework Convention on Climate Change,
as specified in its Article 2, is the stabilization of greenhouse gas concentrations at levels
that “would prevent dangerous anthropogenic interference with the climate system”.
Such levels should be achieved “within a time frame sufficient to allow ecosystems to
adapt naturally to climate change, to ensure that food production is not threatened and to
enable economic development to proceed in a sustainable manner” (UN 1992). It can be
seen that Article 2 has several interrelated elements, which may be linked to other parts
of the Convention. Article 3.3 is of particular relevance here, relating, as it does, to the
application of the precautionary principle in the face of scientific uncertainty.
Under Article 2, stabilization of greenhouse gas concentrations at some arbitrary level is
not the objective per se, as is sometimes assumed, but rather at a level that would
“prevent dangerous anthropogenic interference with the climate system”. There is no
specific reference to the manner in which this stabilization should be achieved. It is
open, for example, as to whether greenhouse gas concentrations would rise above the
ultimate stabilization level before falling back, provided that in the end interference with
the climate system is prevented. The second part of Article 2, in effect, establishes a set
of criteria and general requirements for the timeframe in which greenhouse gas
concentrations must be stabilized. In other words, one could identify levels of impacts on
the areas mentioned that resulted in, for example, threats to food production and work
6
backwards to compute concentrations of greenhouse gases and/or the time profile of these
concentrations that would prevent these impacts from occurring.
Article 2 requires that greenhouse gases be stabilized in such a way and within a
timeframe that ecosystems can adapt naturally, food production is not threatened and that
economic development is able to proceed in a sustainable manner. Put another way, if
stabilization were achieved in such a way that all of these requirements were met, then it
could be said that dangerous anthropogenic interference with the climate system had been
prevented, provided that no other interference with the climate system was being caused
that could be classified as dangerous. If one or the other element were not met, then there
would be a breach of the Convention’s objective.
It may be useful to note at the outset that Article 2 talks of prevention of “dangerous
anthropogenic interference with the climate system” and is not necessarily limited to
dangerous climate changes per se. In theory at least, dangerous anthropogenic
interference could relate to a variety of human induced changes in the totality of the
climate system, which people and/or governments could consider dangerous. Examples
of such issues could include, for example, the risk of ice sheet instability or irreversible
decay. If, for example, the West Antarctic Ice sheet turned out to be very sensitive to
global warming, it is conceivable that its collapse could be triggered by levels of
greenhouse gases that did not result in immediate threats (within the next decades to
century) to any of the categories of effects cited in Article 2. Nevertheless, such a risk,
with the entailed 6-7 metres of sea level rise over centuries to millennia, would be
considered by many as dangerous (O'Neill and Oppenheimer 2002).
What may constitute dangerous anthropogenic interference with the climate
system?
To date, the UNFCCC itself has not attempted to define what may constitute dangerous
anthropogenic interference with the climate system or what acceptable limits may be to
impacts on ecosystems, food production or economic development.
Nevertheless, over the past decade or so several groups have sought to identify acceptable
limits to climate change. There have been two broad approaches, often combined. One is
based on a “bottom up” assessment of the projected impacts of climate change on
ecosystems, agriculture and other sectors. The other is based on a “top down” approach
which focuses on avoiding greater changes than are thought to have occurred in the
current and the last few interglacial periods. The objective of this approach is, in effect,
to keep the climate system away from situations (greenhouse gas concentrations) where
the projected temperatures are either not known from earlier warm periods or are
associated with past periods of rapid and abrupt change.
Based on a review of estimated impacts on ecosystems, as well as comparison of
projected climate changes with “normal climatic changes” of the past (e.g. over the
7
Holocene and not periods of abrupt damages associated with glacial termination), the
WMO/ICSU/UNEP Advisory Group on Greenhouse Gases (AGGG), in 1990, identified
two main temperature indicators or thresholds with different levels of risk (Rijsberman
and Swart 1990). It was argued that an increase of greater than 1.0°C above pre-
industrial levels “may elicit rapid, unpredictable and non-linear responses that could lead
to extensive ecosystem damage” with warming rates above 0.1°C/decade likely to lead to
rapidly increasing risk of significant ecosystem damage. Furthermore, a 2.0°C increase
was determined to be “an upper limit beyond which the risks of grave damage to
ecosystems, and of non-linear responses, are expected to increase rapidly”.
Corresponding indicators for sea level rise were also developed. It was argued that rates
of sea-level rise of less than 20mm/decade “would permit the vast majority of vulnerable
ecosystems, such as natural wetlands and coral reefs to adapt with rates beyond this
leading to rapidly rising ecosystem damage” (Rijsberman and Swart 1990: viii). The
AGGG felt that limiting total sea level rise to a 50 cm increase above 1990 global mean
sea-level could “prevent the complete destruction of island nations, but would entail large
increases in the societal and ecological damage caused by storms”. This assessment was
based on the scientific knowledge available before the IPCC First Assessment Report was
concluded in 1990.
In 1995, the WBGU used a “top down” approach to determine an upper limit or
“tolerable window” of warming. Adding 0.5°C to the estimated difference between the
recent, pre-industrial Holocene and the warmest period of the last interglacial, the WBGU
arrived at a tolerable warming window (relative to pre-industrial temperatures) of 2°C
(WBGU 1995). This limited additional future warming to around 1.3°C, relative to the
estimated 1995 global mean temperatures. Above this limit, it was argued, was a risk of
“dramatic changes in the composition and function of today’s ecosystems” (WBGU
1995: 7).
At a political level, the European Union’s Environment Council agreed in 1996 that
global temperatures should not be allowed to exceed 2°C above pre-industrial levels
(European Community 1996):
“Given the serious risk of such an increase and particularly the very high rate of
change the Council believes that global average temperatures should not exceed
2 degrees (Celsius) above pre-industrial level and that therefore concentration
levels lower than 550 (parts per million of) CO
2
should guide global limitation
and reduction efforts. This means that the concentrations of all greenhouse gases
should also be stabilised. This is likely to require a reduction of emissions of
greenhouse gases other than CO
2
, in particular CH
4
and N
2
0.”
The Environment Council based this decision on a consideration of the IPCC Second
Assessment Report and the impacts identified therein, which in general were for a
doubling of CO2 above pre-industrial levels.
8
The IPCC itself has not directly addressed the question of what might be dangerous
climate change and has seen its role as limited to providing policy relevant but not policy
prescriptive advice. In the lead up to the Second Assessment Report, the IPCC held a
workshop in Fortaleza, Brazil in 1994 on the issue of Article 2, however the results of
this were inconclusive, except for the reaffirmation by scientists that they did not see a
role for themselves as a group in defining the limits of Article 2.
In its Third Assessment Report the IPCC made several efforts to provide scientific advice
that could be used by policy makers in relation to Article 2. Chapter 19 of the Working
Group II report, which attempted to synthesize the other chapters in this working group
report, identified five “reasons for concern” that could be used to “aid readers in making
their own determination about what is ‘dangerous’ climate change” (Smith et al. 2001:
915):
1) The relationship between global mean temperature increase and damage to or
irreparable loss of unique and threatened systems;
2) The relationship between global mean temperature increase and the distribution of
impacts;
3) The relationship between global mean temperature increase and global aggregate
damages;
4) The relationship between global mean temperature increase and the probability of
extreme weather events;
5) The relationship between global mean temperature increase and the probability of
large-scale singular events such as the breakup of the West Antarctic Ice Sheet or the
collapse of the North Atlantic thermohaline circulation.
The present report will provide information relevant to factors one to three, with the latter
two reasons for concern being beyond the scope of this report.
The Synthesis Report of the IPCC TAR sought to answer nine policy relevant questions
developed in consultation with the UNFCCC, several aspects of which were relevant to
Article 2. The most pertinent to the present work are from questions three and six in the
synthesis report:
Question 3: “What is known about the regional and global climatic, environmental, and socio-
economic consequences in the next 25, 50, and 100 years associated with a range of greenhouse
gas emissions arising from scenarios used in the TAR (projections which involve no climate
policy intervention)? To the extent possible evaluate the ...Projected changes in atmospheric
concentrations, climate, and sea level … impacts and economic costs and benefits of changes in
climate and atmospheric composition on human health, diversity and productivity of ecological
systems, and socio-economic sectors (particularly agriculture and water) ...” (IPCC 2001: 8).
Question 6: “How does the extent and timing of the introduction of a range of emissions reduction
actions determine and affect the rate, magnitude, and impacts of climate change, and affect the
global and regional economy, taking into account the historical and current emissions? What is
known from sensitivity studies about regional and global climatic, environmental and socio-
economic consequences of stabilizing the atmospheric concentrations of greenhouse gases (in
9
carbon dioxide equivalents), at a range of levels from today’s to double that level or more, taking
into account to the extent possible the effects of aerosols?” (IPCC 2001: 19).
Though there were attempts, in various drafts of the IPCC TAR, to associate specific
global mean temperature increases with defined impacts, by the time the report was
finalized most of these examples were reduced to quite general statements in the
summaries for policy makers of Working Group II and the Synthesis Report. However,
the full Synthesis Report does contain several tables outlining identified impacts for
temperature bands in each of the areas relevant to this paper. Whilst there are limitations
to these tables, notably that the temperature bands associated with specific impacts are
often too large and hence lose some precision, such as is possible given all other
uncertainties, they will be used as the starting point for the analysis in each of the
sections of this report. Indeed, this may provide the best and most coherent way of
showing transparently how the analysis presented in this paper builds upon, extends or
diverges from the conclusions of the TAR authors.
2. Ecosystems, Biodiversity and Climate Change
Ecosystems and their species form the fabric of life on the Earth and provide a very wide
range of services to humanity. The IPCC TAR has summarized these and in any event
they are well known (Table 1). Unfortunately, given the large human pressures and
impacts on species and ecosystems, rapid climate change probably could not happen at a
worse time in the history of the biosphere (Soulé 1992). Due to these pressures species
are becoming extinct at a rate 100-1000 times greater than is considered normal over
evolutionary time. As a consequence conservation biologists have labelled the current
epoch the sixth major extinction event in the history of the planet (Chapin et al. 2000;
Novacek and Cleland 2001). The causes of this are anthropogenic in origin, principally
the modification or destruction of habitats, pollution, hunting, resource use, and the
introduction of exotic species. Large fractions of extant species groups are classified as
endangered (see Figure 1).
Species extinction results in loss of biodiversity and often changes in the structure and
function of ecosystems. There is a large risk that many of the ecosystem services
identified in Table 1 could be adversely effected by species loss. However, the ability to
predict which species are the most important is very often quite limited (National
Research Council 1999; Chapin et al. 2000).
10
Table 1 - Ecosystems Function with Links to Good/Services and Possible Societal Value
Function Goods/Service Value
Production – Food Direct
– Fiber (timber and non-wood products)
– Fuel
– Fodder
Biogeochemical cycling – Nutrient cycling (especially N and P
absorption/deposition)
Mostly indirect, although future values have
to be considered
– Carbon sinks
Soil and water
conservation
– Flood and storm control
– Erosion control
Mostly indirect,
although future values have
to be considered
– Clean water
– Clean air
– Water for irrigation
– Organic matter or sediment export
– Pollution control
– Biodiversity
Animal-plant interactions
– Pollination
– Animal migration
Mostly indirect, future, bequest, and
existence values have to be considered
– Biodiversity
Carrier – Landscape connectivity
– Animal migration
Mostly indirect and existence, but bequest
may have to be considered
– Biodiversity
– Aesthetic/spiritual/cultural service
Source: Compiled from information in Figure 5-1 of Gitay et al. (2001).
Although it is clear that climate change is only one of several pressures on ecosystems,
and often not the most immediate (Sala et al. 2000), one must also consider that the
interaction between human activities and their effects on ecosystems and species is likely
to exacerbate the effects of climate change. For a number of ecosystems and species it
seems clear that if non-climatic pressures are successfully relieved but climatic ones
grow, there is still a substantial likelihood of major losses or extinctions in the coming
century (and in some cases several decades).
Significant and systematic effects have been observed on a very wide range of species
and ecosystems globally which have been attributed to climate change (McCarty 2001;
Walther et al. 2002; Parmesan and Yohe 2003; Root et al. 2003). Space does not permit
elaboration of these findings here: it is sufficient to note that a large majority of
observational studies reveal changes consistent with expected effects of climate change.
The rest of this section examines the basic processes leading to climatic impacts on
species and ecosystems followed by a review of the projected effects of climate change
on a range of species and ecosystems. The starting point for this review is the IPCC
Third Assessment Report findings, particularly those of Working Group II, however the
main effort is to attempt to estimate the effects of climate warming on a sample of species
and ecosystems drawn from the literature. Thus a substantial volume of publications and
11
reports not reviewed in the TAR, but which are relevant to an assessment of climate
effects on ecosystems, were sought out and reviewed. Much literature has been
published since the TAR or was not available to the authors at the time of its writing (a
large selection of this is listed in the Appendix to the IPCC Technical Paper on Climate
Change and Biodiversity (Gitay et al. 2002)). This sample will be representative of the
wide range of impact studies in the literature at present, but is by no means
comprehensive.
The IPCC Third Assessment Report reviewed the impacts of climate change on wildlife
and ecosystems in various chapters of the Working Group II Report. Chapter 5 of that
report (Gitay et al. 2001) is the main locus of this review. It covered the effects of global
climate change on the terrestrial biosphere, wildlife in ecosystems, grasslands, savannas,
and deserts, forests and woodlands, lakes and streams, inland wetlands, and arctic and
alpine ecosystems.
1
In addition to the material found in Chapter 5, Price et al. (2000)
2
prepared supplementary information. The impacts of climate change on coastal zones
and marine ecosystems were reviewed in a separate chapter and much additional material
on Arctic and Antarctic ecosystems were reviewed in the polar chapter. In addition, the
regional chapters of this report (Africa, Asia, Small Island States, North America, Latin
America, Australia and New Zealand, and Europe) provide a lot of additional material on
ecosystems and species effects not covered in Chapter 5. Finally, Chapter 19 attempted a
synthesis of the findings of the complete Working Group II Report (Smith et al. 2001).
A huge volume of literature is reflected in the TAR assessment and it is neither desirable
nor feasible to reconstruct this, hence, the effort here has focused on identifying key
findings and studies which can provide the basis for an assessment of the projected
impacts of climate change on species and ecosystems by degrees of projected warming or
sea level rise. Nevertheless, substantial effort has been made here to at least verify the
reviews cited in relevant chapters of the TAR that relate to this objective.
Processes causing loss of biodiversity and ecosystem damage
Climate change is expected to affect ecosystems and species in a variety of different
ways. In this section the general processes, by which increased CO
2
and climate change
affect species and ecosystems, are outlined. Specific examples are discussed in the later
sections that deal with specific classes of species and ecosystem types.
1
See http://www.grida.no/climate/ipcc_tar/wg2/196.htm.
2
See http://www.usgcrp.gov/ipcc/html/ecosystem.pdf.
12
Figure 1 - Proportion of the Global Number of Birds, Mammals, Fish and Plants
Species that are Currently Threatened with Extinction
Source: Figure 2 from Chapin et al. (2000).
The species that are most vulnerable to extinction from whatever cause are those with
restricted ranges, fragmented distribution within their range, low populations, reducing
range, decreasing habitat within the range, and/or which are suffering population declines
(Price et al. 2000). Species with quite restrictive habitat requirements are most
vulnerable to extinction (Pimm et al. 1995). Where climate change is projected to reduce
habitats of such species there is likely to be the greatest extinction risks. Examples from
the IPCC TAR include the Bengal tiger and its habitat in the Sundarbans and several
mountain dwelling species from Africa and Central and South America. In the case of
the Sundarbans, this World Heritage listed mangrove and forested wetland habitat is
projected to be reduced substantially as a consequence of sea level rise. Potential
migration routes for many of the area-dependent species are blocked by human activities
(ADB 1994).
Table 2 summarizes an array of factors known to drive the processes of species
endangerment and extinction. Climate change is one of the pressures that is or is likely to
act to increase species vulnerability now and in the future. However, it will often, if not
usually, act in combination with the other pressures described below. Habitat
fragmentation caused by destruction of habitat, infrastructure or disturbance is likely to
exacerbate the effects of climate change by reducing the migration and dispersal ability
of species (Malcolm et al. 2002b). Pollution may also reduce the ability of species to
cope with the stresses of rapid climate change (Hojer et al. 2001).
13
Table 2 - Processes driving Species Endangerment and Extinction
Process Explanation
Conversion of natural lands to
other uses
This is the main threat to ecosystems and wildlife. 80% of the earth’s
forests are already cleared or degraded and a sizeable fraction of the
remainder is threatened.
Habitat fragmentation This can be caused by agricultural land use and infrastructure such as
roads, railways and urban areas. Habitat fragmentation threatens the long
term viability of wildlife population as:
• Species often require large areas to survive in the long term. At
present, a number of large birds and mammals have range
requirements greater than the remaining habitat area. This
means that in the longer term they are likely to decline due to
the effects of accidents, inbreeding or climate change.
•
Fragmentation is in effect a barrier to the dispersal and
migration of species in response to natural disturbances or
climatic changes.
•
Invasion by exotic species such as new predators is easier.
Habitat degradation Human use of habitats for natural resource extraction or recreation can
introduce exotic predators (e.g. cats, dogs), plant pathogens, disturb water
courses or water quality or disturb breeding environments by noise or
physical disturbance.
Hunting and extraction or use
of natural resources
Hunting, harvesting, culling or inadvertent killing of wildlife is a
substantial threat in many, if not most, regions. Threats arise in a variety
of ways:
• Hunting and harvesting is often not sustainable and has, in the
past, led to extinctions or stock collapses. Well known
historical examples include the extinction of the great Auk and
the passenger pigeon. In recent years, hunting in Europe has led
to a decline in the European Robins populations. In developing
countries wildlife populations adjacent to expanding urban areas
will most likely not be sustained.
• By-catch losses are often significant.
• Culling of wildlife because of actual or perceived competition
with human activities.
• Hunting can result in pollution of wetlands.
Wildlife trade This can place considerable pressure on populations and species and has
caused substantial damage to large mammals such as elephants,
rhinoceros, and tigers.
Pollution Pollutants have been detected in many species throughout the world.
Pollution has been implicated in the decline of a number of species
through:
•
Direct poisoning.
• Indirect effects, due to longer-term exposure to pollutants, on
reproduction, behaviour and survival.
• The elimination or modification of habitat.
Exotic species Introduced species have caused substantial damage to local species and
pose a threat to substantial numbers of mammals and birds.
14
Process Explanation
Climatic change Climate is an important determining factor of the distribution range of
ecosystems and species. Future projected rates of change appear to
exceed previously observed ones, giving rise to concerns as to the ability
of species and ecosystems to adapt to projected changes without
significant loss or disruption.
Synergistic effects of climate
change
Climate change is likely to act synergistically with many of the other
factors mentioned in this table. Habitat fragmentation and loss will
inhibit species abilities to migrate in response to climate changes. Exotic
species invasion may be facilitated by a combination of habitat
degradation and climate change, yielding negative effects on the endemic
species. Human responses to climate change may exacerbate threats to
biodiversity, by, for example, preventing inland movement of coastal
wetlands or as a consequence of increased pesticide use resulting from
enhance pest activity in changed climatic conditions.
Extreme climatic events Extreme climatic events and changes in the pattern of weather and
climate events can cause large-scale losses of species and damages to
ecosystems.
Source: This table has been compiled based on the review of Gitay et al. (2001) and (Hughes 2000).
One of the important processes to bear in mind, when considering biodiversity loss and
ecosystem decay, is the observation that species, or populations of species, that have
survived large scale loss of their habitat in the past may still face extinction (Cowlishaw
1999). Species often require large areas of habitat to be able to weather stochastic events
such as droughts and disease outbreaks, avoid the problems of small gene pools or other
environmental pressures and thus survive in the long-term.
Climate change and CO
2
effects on species and ecosystems
Projected anthropogenic climatic change and increases in CO
2
are expected to result in
large changes in ecosystems globally and to add significantly to the pressure on species
from the human activities outlined in Table 2. In a general sense, species respond to
warming by moving their ranges upwards and polewards. Within this general pattern
however, the range and complexity of responses expected is quite large. Nevertheless,
these can be broken down into a finite list of classes of responses or impacts, which are
summarized in Table 3. Examples of some of the potential impacts and risks are also
given. Hughes (2000) provides a very useful schematic of the main pathways by which
climate change and increases in CO
2
can result in negative impacts on species and
ecosystems (see Figure 2). Increasing CO
2
concentrations impact on plant species
directly affecting growth, nutrient uptake, and water use efficiency.
15
Figure 2 - Pathways by which Climate change affects Species and Ecosystems
Source: Figure 2 of Hughes (2000). Reference numbers in this figure refer to the original publication by
Hughes.
CO
2
(and other greenhouse gases) induced climate changes will result in changes in
temperature, the precipitation regime and the frequency and intensity of extreme events.
Species response can be divided into four groups – changes in physiology, phenology,
distribution and in situ adaptations. The various responses ultimately lead to changes in
species interaction and consequently, to changes in ecosystem structure and composition.
Changes in the frequenc y and intensity of extreme events as a consequence of climate
change, including El Niño cycles, are likely (Easterling et al. 2000) and will have large
effects upon species and ecosystems (Parmesan et al. 2000). Average climate changes
may not be as important as the changes in extremes of weather and climate in triggering
shifts in species and/or major changes in ecosystems. To date, few studies have taken
this into account in projecting the effects of climate change on species.
Beyond the details of what mechanisms and processes will drive species and ecosystem
responses to climate change, is the apparent fact that the rate of global mean surface
temperature change projected over the next century appears quite unprecedented, at least
during the Holocene and perhaps for much longer. The maximum rate of global mean
change consistent with the range of estimates for the transition from the last glacial
16
maximum to the Holocene (also known as Termination I) is around 0.01
o
C/decade
3
. A 3-
5
o
C warming to 2100 is thus about 25-45 times faster than the highest rates of change at
the end of the last glacial over several thousand years.
In relation to century scale changes, it would appear that changes with rates of more than
0.1
o
C/decade are quite unusual. If one compares the maximum trends in temperature
over varying time periods in ice core data and in proxy and instrumental records, it is
apparent that the maximum rates of change drop rapidly with increasing averaging
period. Figure 3 compares a local long-term temperature series with three hemispheric or
global average records for the period 1861-2001. As would be expected the local
temperature series shows much larger variability. In Figure 4 rates of change in
temperature are calculated from the individual time series, over all possible trend periods
in each record and then the maximum rate for each trend period found. For example, the
maximum trend in temperature over all 30-year periods in the Mann et al. (1999) 1000-
year record is 0.2
o
C/decade, whereas for the central England record it is close to
0.5
o
C/decade. For a 100-year trend period, the maximum rate of change observed is less
than 0.1
o
C per decade for all records, excepting projected changes over the next century.
The projected rates of change, in relation to the ability of plants and animals are to move,
migrate or adapt over the next century worry many scientists (Overpeck et al. 1992;
Malcolm et al. 2002b). During the last deglaciation, even widespread and dominant
species became extinct (Jackson and Weng 1999) and there is concern that projected rates
of climate change exceed the observed rates of change in the past (Davis and Shaw 2001;
Malcolm et al. 2002b). Whilst attempts have been made to model migration and
movement of plants under climate change, present methodologies remain problematic
(see discussion in Gitay et al. (2001)). Although there is a general consensus that
projected rates of climate change are very likely to exceed the migrational capacity of
species in at least the mid- and high-latitudes, too little is known to be able to fully
quantify this problem.
3
In the somewhat extreme case that Termination I was associated with an 8
o
C change in global average
temperature over a period of 7,000 years as may be inferred from the Vostok record published by (Petit et
al. 1999).
17
Table 3 - Response and Impacts of Climate Change on Species and Ecosystems
Response or impact Examples of effects and risks
• Changes in distribution of species, ecosystem
boundaries, and biomes
Poleward or upward shift of aquatic and terrestrial biota (McCarty 2001;
Walther et al. 2002; Root et al. 2003). Risk that insufficient altitudinal
range with suitable habitat exists for mountain species to migrate
(Theurillat and Guisan 2001). Risk that rate of change exceeds migratory
capacity of species (Malcolm et al. 2002b).
• Changes in phenology of biotic and abiotic processes
and events
Earlier flowering of plants and budding of trees, earlier egg laying in birds.
Risk of asynchronous timing of events between species with tight
synchronization requirements e.g. late arrival of migratory birds after peak
of food availability has passed (Both and Visser 2001; Visser and
Holleman 2001).
• Changes in structure of plant communities Changes from grassland or savannah to woodlands, or from moist tropical
forest to drier woodlands. Risk of loss of habitat for ungulates with
reduction in savannah and invasion with woody plant species (Bond and
Midgley 2000).
• Changes in species composition and diversity Loss of climatically suitable habitat for species may frequently lead to
range reductions, population fragmentation and reduced genetic diversity.
Risk of major species loss in some regions and risk of ecosystem structural
changes or loss if key species disappear
(Kerr and Packer 1998; Midgley et
al. 2002).
• Changes in animal and plant population dynamics and
structure
Changes in competitive balance between species affecting ecosystem
structure and composition.
• Changes in Net Primary Productivity (NPP), Net
Ecosystem Productivity (NEP), Net Biome Productivity
(NBP)
Increased CO2 and warmer temperatures will lead to changes, often
increases, in NPP, with the balance of ecosystem productivity NEP and
NBP being determined by the precipitation changes (Cramer et al. 2001).
Risk in some ecosystems of reduction in NPP, NEP or NBP with warming
in the coming century (White et al. 2000a; Friedlingstein et al. 2001).
• Changes in carbon and nutrient cycling Changes in NPP, NEP and NBP affect global carbon cycle with increasing
CO
2
likely to enhance the terrestrial uptake of carbon (Lucht et al. 2002).
Risk of positive feedback from climate change to terrestrial carbon cycle
(White et al. 2000a; Friedlingstein et al. 2001).
• Changes in litter, forage and wood quality Increase atmospheric CO
2
, whilst enhancing plant growth may at the same
time results in less nutrient content in leaves (Tuchman et al. 2002)
, forage
(Lenart et al. 2002) and crops (Reyenga et al. 1999). Kanowski (2001)
finds that increased CO
2
will reduce the food quality of rainforest trees for
tree dwelling marsupials, which is likely to reduce their abundance in the
future.
• Changes in water-use efficiency with elevated CO2 Could increase the drought resistance of plant species and with differential
response between species, change the competitive balance between
components of ecosystems (Bond and Midgley 2000).
• Increase in frequency and/or intensity of disturbance
(e.g., fires)
Increased fire frequency in Mediterranean ecosystems as a consequence of
changed drought intensity or frequency leading to shifts in vegetation
structure (Parmesan et al. 2000; White et al. 2000b; Mouillot et al. 2002;
Walther et al. 2002).
• Changes in water flow and level leading to loss of
aquatic habitats, waterfowl, riparian forests, recreational
opportunities, eutrophication
Changes in water regime (flow, duration and extent) can negatively affect
the habitats and breeding possibilities of many species. Risk of loss of
cold freshwater fish species and of major reductions in breeding habitats
for ducks and other waterfowl (Sorenson et al. 1998).
• Increased pests and diseases Changes in climate in the boreal forests could lead to a greater frequency
of pest outbreaks affecting boreal tree species (Ayres and Lombardero
2000; Volney and Fleming 2000).
Note: This table is compiled in part from Figure 5-1 from Gitay et al. (2001), with the examples
drawn from the literature.
18
Figure 3 - Comparison of Hemispheric and Long-Term Local Temperature Series
Comparison of temperature time series
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
1.5
1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000
Year AD
Anomaly oC wrt 1961-1990
Central England Time series 1659-2001
Folland/Jones 1861-2001 global surface temperature
Northern Hemisphere Jones 1856-2001 data
Northern Hemisphere land proxy record Jones (1998)
This graph compares the central England temperature series with a global mean and northern hemisphere
surface instrumental record and a 1000 year proxy record for the northern hemisphere land surface for the
period 1861-2000.
Figure 4 - Comparison of Maximum Decadal Rates of Change
Comparison of maximum decadal rates of temperature change and projected
changes to 2099
0
0.1
0.2
0.3
0.4
0.5
20 30 40 50 60 70 80 90 100 110 120
A
oC/decade
HadCM2 scenario GG2 ensemble member
1990-2099
Jones 1861-2001 Northern Hemisphere record
Mann (1999) AD 1000 - AD 1980 Northern
Hemisphere proxy record
Max Central England before 1950
This graph compares the maximum rates of change observed for different trend periods for three
temperature records with a HadCM2 GCM projection for the period 1990-2099.
19
Projected effects on species and ecosystems
Table 4 from the IPCC TAR Synthesis Report is an attempt to summarize the findings of
the IPCC TAR in relation to the impacts of climate change on ecosystems and species. It
also attempts to place temperature-warming bands on the identified impacts for coral
reefs, coastal wetlands and terrestrial ecosystems. What becomes apparent from
examination of this table is that the risk of significant damages exists at low levels of
warming. A detailed examination of the literature used in the TAR and that has been
published subsequently adds substantial specificity to this picture.
Rather than present the analysis of the literature on the projected effects of climate
change on ecosystems and species in a narrative format the results are presented in a table
format. This facilitates cross comparison with similar systems in different regions as well
as maintaining the compactness of this report. Table 5 details the results of the analysis
here for a large number of projected impacts on species and ecosystems under quite
different climate scenarios. An attempt has been made to reduce all of the scenarios used
in the various studies cited to an estimated change in global mean surface temperatures
that would correspond to the contemporary generation of climate models. This has been
done using the simple climate model MAGICC 4.1 and the downscaling programme,
SCENGEN of Wigley, Raper, Hulme and others (Hulme et al. 1995; Raper et al. 2001;
Wigley and Raper 2001)
4
. Details are given in the table for each case.
Based on the analysis documented in Table 5 an attempt has been made to map the
projected level of impact for different levels of warming graphically in (Figure 5-7).
These figures attempt to associate some level of risk, loss or impact with a range of
temperature increases. Five categories of risk were used in constructing the figures. Less
4
The programmes and references are available at
http://www.cgd.ucar.edu/cas/wigley/magicc/installation.html
20
Table 4 - Ecosystem effects of climate change
No climate policy interventions. Source: IPCC TAR Synthesis Report Technical Summary Table 3-2. *Refer to
footnotes a-d accompanying Table 7 in this report. Note f: These effects have already been observed and are expected
to continue [TAR WGII Sections 5.2.1, 5.4.3, 16.1.3, & 19.2].
5
Using Folland et al.(2001) global temperature data set.
2025 2050 2100
CO
2
concentration
a
405–460 ppm 445–640 ppm 540–970 ppm
Global mean
temperature change
from the year 1990
b
0.4–1.1°C 0.8–2.6°C 1.4–5.8°C
Global mean
temperature change
from the years 1861-
1890 (average)
5
1.0-1.7°C 1.4-3.2°C 2.0-6.4°C
Global mean sea-level
rise from the year
1990
b
3–14 cm 5–32 cm 9–88 cm
Ecosystem Effects
c
Corals [WGII TAR
Sections 6.4.5, 12.4.7,
& 17.2.4]
Increase in frequency of coral
bleaching and death of corals
(high confidence
d
).
More extensive coral bleaching
and death (high confidence
d
).
More extensive coral
bleaching and death (high
confidence
d
).
Reduced species
biodiversity and fish
yields from reefs
(medium confidence
d
).
Coastal wetlands and
shorelines [WGII TAR
Sections 6.4.2 & 6.4.4]
Loss of some coastal wetlands to
sea-level rise (medium
confidence
d
).
Increased erosion of shorelines
(medium confidence
d
).
More extensive loss of coastal
wetlands (medium confidence
d
).
Further erosion of shorelines
(medium confidence
d
).
Further loss of coastal
wetlands (medium
confidence
d
).
Further erosion of
shorelines (medium
confidence
d
).
Terrestrial ecosystems
[WGII TAR Sections
5.2.1, 5.4.1, 5.4.3,
5.6.2, 16.1.3, & 19.2]
Lengthening of growing season
in mid- and high latitudes; shifts
in ranges of plant and animal
species (high confidence
d
).
e,f
Increase in net primary
productivity of many mid- and
high-latitude forests (medium
confidence
d
).
Increase in frequency of
ecosys
tem disturbance by fire and
insect pests (high confidence
d
).
Extinction of some endangered
species; many others pushed
closer to extinction (high
confidence
d
).
Increase in net primary
productivity may or may not
continue.
Increase in frequency of
ecosystem disturbance by fire
and insect pests (high
confidence
d
).
Loss of unique habitats
and their endemic species
(e.g., vegetation of Cape
region of South Africa
and some cloud forests)
(medium confidence
d
).
Increase in frequency of
ecosystem disturbance by
fire and insect pests (high
confidence
d
).
Ice environments
[WGI TAR Sections
2.2.5 & 11.5; WGII
TAR Sections 4.3.11,
11.2.1, 16.1.3, 16.2.1,
16.2.4, & 16.2.7]
Retreat of glaciers, decreased sea-
ice extent, thawing of some
permafrost, longer ice-free
seasons on rivers and lakes (high
confidence
d
).
f
Extensive Arctic sea-ice
reduction, benefiting shipping
but harming wildlife (e.g., seals,
polar bears, walrus) (medium
confidence
d
).
Ground subsidence leading to
infrastructure damage (high
confidence
d
).
Substantial loss of ice
volume from glaciers,
particularly tropical
glaciers (high
confidence
d
).
21
Impacts on coastal wetlands
•
Below a 1
o
C increase the risk of damage is low for most, but not all systems.
•
Between 1-2
o
C moderate to large losses appear likely for a few systems. Of most
concern are threats to the Kakadu wetlands and the Sundarbans of Bangladesh,
both of which may suffer 50% losses at less than 2
o
C:
- Inscribed on the UNESCO World Heritage List for both its outstanding
natural and cultural values, Kakadu is regarded as one of the great
wetlands of the world;
- Also on the World Heritage list and renowned as the largest intact
mangrove wetland system in the world, the Sundarbans is the sole
remaining home of the Royal Bengal tiger (Panthera tigris tigris).
Spanning about 1 million km
2
, 62% of which is in Bangladesh and the
remainder in West Bengal, India, this region is home to a wide variety and
great number of species.
•
Between 2-3
o
C, it is possible that the Mediterranean, Baltic and several migratory
bird habitats in the US experience a 50% loss. In this range it seems likely that
there could be the complete loss of Kakadu and the Sundarbans.
A key issue is the inertia of sea level rise, which makes the assignment of risk to different
temperature levels misleading. Should, for example, sea level rise by 30cm in the
coming decades to a century (threatening Kakadu), the thermal inertia of the ocean is
such that an ultimate sea level rise of 2-4 times this amount may be inevitable even if
temperature stops rising. The prognoses for wetlands in this context is not clear, as many
damages are linked to the rate of sea level rise compared to the accretion and/or
migratory capacity of the system. A major determinant of the latter will be human
activity adjacent to, or in the inland catchments of the wetland system.
Impacts on animal species
Figure 6 summarizes estimated effects on a range of animal species. Along with the
information in Table 5 one could conclude the following:
•
Below 1
o
C warming, there appears to be a risk of extinction for some vulnerable
species in southwestern Australia and to a lesser extent in South Africa. Range
losses for species such as the Golden Bower bird in the highland tropical forests
of North Queensland Australia and for many animal species in South Africa are
likely to become significant and observable.
22
•
Between 1-2
o
C warming, large and sometimes severe impacts appear possible for
some Salmonid fish habitats in the USA, the Collared Lemming in Canada, South
African animals and for Mexico’s fauna. Extinctions in southwestern Australia
seem very likely and possibly South Africa and Mexico for the most vulnerable
species. In general, many endangered species are pushed closer to extinction.
Mid summer ice reduction in the Arctic ocean seems likely to be at a level that
would cause major problems for polar bears at least at a regional level.
•
Between 2-3oC large to severe impacts appear likely for Mexican fauna, many
South African animals, the Collared Lemming in the Arctic (which would have
broad implications for arctic ecosystems), Salmonid fish in Wyoming, with the
likelihood of extinctions in Mexico and South Africa. In Hawaii, extinction of
several Hawaiian Honeycreeper has been predicted for about a 2.8-3.2oC
increase. In this range the Golden Bower bird's range would be reduced by 90%.
•
Above 3oC, large impacts begin to emerge for waterfowl habitat in the Prairie
Pothole region. The collared lemming range is reduced by 80%, very large
reductions are projected for Arctic sea ice cover particularly in summer which is
likely to further endanger polar bears. Extinction of the Golden Bower bird is
predicted in this temperature range. In Mexico very severe range losses for many
animals are projected, as is the case also in South Africa, with Kruger national
park projected to lose two thirds of the animals studied. The likelihood of the
impacts identified above will continue to grow with higher temperatures.
Impacts on ecosystems
Figure 7 shows the impacts projected for a range of ecosystems including tropical forests,
alpine systems in Australia and Europe, the Fynbos and Succulent Karoo in South Africa
and, in the marine domain, coral reefs. With the information in Table 5, one may find the
following conclusions:
•
Between present temperatures and a 1
o
C increase, three ecosystems appear to be
mo ving into a high risk zone - highland tropical forests in Queensland, Australia,
the Succulent Karoo in South Africa and coral reefs. Increased fire frequency and
pest outbreaks may cause disturbance in boreal forests and other ecosystems.
•
Between 1-2
o
C the Australian highland tropical forest, the Succulent Karoo
biodiversity hot spot, coral reef ecosystems and some Arctic and alpine
ecosystems are likely to suffer large or severe damage. The Fynbos will
experience increased losses. Coral reef bleaching will likely become much more
frequent, with slow or no recover, particularly in the Indian Ocean south of the
equator. Australian highland tropical forest types, which are home to many
endemic vertebrates, are projected to halve in area in this range. The Australian
alpine zone is likely to suffer moderate to large losses. The substantial loss of
23
Arctic sea ice likely to occur will harm ice dependent species such as the polar
bears and walrus. Increased frequency of fire and insect pest disturbance is likely
to cause increasing problems for ecoystems and species in the Mediterranean
region. Moderate to large losses of boreal forest in China can be expected.
Moderate shifts in the range of European plants can be expected and in Australia
moderate to large number of Eucalypts may be outside out of their climatic range.
•
Between 2-3
o
C coral reefs are projected to bleach annually in a number of reef
locations. At the upper end of this temperature band, the risk of eliminating the
Succulent Karoo and its 2800 endemic plants is very high. Moderate to large
reductions in the Fynbos can be expected, with the risk of significant extinctions.
In the highland tropical forests of northeastern Australia “catastrophic loss” or
rainforest vertebrates has been predicted Australian mainland alpine ecosystems
are likely to be on the edge of disappearance. European alpine systems will at or
above their anticipated tolerable limits of warming with some vulnerable species
close to extinction. Severe loss of boreal forest in China is projected and large
and adverse changes are also projected for many systems on the Tibetan plateau.
Large shifts in the range of European plants seem likely and a large number of
Eucalypt species may expect to lie outside of their present climatic range.
Moderate to large effects are projected for Arctic ecosystems and boreal forests.
Within this temperature range there is a likelihood of the Amazon forest suffering
potentially irreversible damage leading to its collapse.
24
Figure 5 - Impacts on Coastal Wetlands
0 1 2 3 4 5 6
Global assessment: High - progressive coastal wetland loss with increasing
warming (22.2% for ca. 3.4oC warming) (1a)
Global assessment: Low - progressive coastal wetland loss with increasing
warming (5.7% for ca. 3.4oC warming) (1b)
USA: Southern New England- extensive loss of wetlands if sea level rise greater
than 6mm/yr (2)
USA: Loss of important foraging, migratory and wintering bird habitat at four sites
(20- 70% loss for ca. 2.6oC warming) (3)
USA: Delaware - Loss of 21% ca. 2.5-3.5oC warming - 100 year floods occurring 3-
4 times more frequently (4)
European wetlands: Atlantic coast (0 to 17% loss for 2.6-4.4oC warming in 2080s)
(5)
European wetlands: Baltic coast (84-98% for 2.6-4.4oC warming in 2080s) (6)
European wetlands: Mediterranean coast (31-100% loss for 2.4-4.4oC warming in
2080s) (7)
Bangladesh, Sundarbans: Progressive loss of mangrove forest and wetlands,
including habitat of Bengal tiger (75% loss at 2-3.5oC) (8)
Australia, Kakadu region: Loss of, or serious damage to, Kakadu World Heritage
listed wetlands (30cm,1.7oC - range of 1.2-3.1oC) (9)
No significant effect
(less than 5%)
Small impact (ca 5-
10%)
Moderate loss (ca
10-20%)
Large loss (20-50%
or greater)
Severe loss (50% or
more)
25
Notes: All examples are described in more detail in Table 5 - Ecosystem Impacts.
(1a) Global assessment: Based on the Nicholls et al.(1999) assessment using the high estimate of wetland loss (22.2% in 2100 for around a 3.4
o
C
warming). A linear extrapolation used to calculate 50% loss, which is likely to very much overestimate the temperature at which this would
occur.
(1b) Global assessment: As above but for low estimates (5.7% loss by 2100) with linear extrapolation to 50%, which is likely to radically
underestimate the at which this would occur.
(2) USA, southern New England: Based on Donnelly and Bertness (2001b) with assumption that a 5
o
C increase (3-5oC range) by 2100 is
associated with a 6mm/yr increase in sea level rise and an 80% (extensive) loss of wetlands.
(3) USA, migratory bird habitat: Based on Galbraith et al. (2002). The graph shown is for the average range of losses at the four sites that lose
intertidal habitat for all warming and sea level rise scenarios - Willapa Bay, Humboldt Bay and northern and southern San Francisco Bay.
The average losses at these sites in 2100 for the 2.6oC scenario is 44 % (range 26% to 70%) and for 5.3oC is 79% (range 61% to 91%). The
latter point is used to scale the average losses with temperature, which increases the temperature sligtly for a given loss compared to the
2.6oC scenario. The Delaware bay site loses 57% of intertidal habitat for the 2.6oC (34 cm sea level rise) but gains 20% in the 5.3oC (77cm
sea level rise scenario). Whilst the Bolivar flats site loses significantly by the 2050s for both scenarios (38-81%) it gains by the 2100s for
both scenarios.
(4) USA, Delaware: Based on Najjar et al. (2000) assuming 21% loss at 3.5oC warming with linear extrapolation to 50%. A linear extrapolation
used to calculate 50% loss, which is likely to very much overestimate the temperature at which this would occur.
(5) European wetlands - Atlantic coast: Based on IPCC WGII TAR Table 13-4 which is based new runs using the models described by Nicholls
et al.(1999) with a linear extrapolation of the high range 17% loss with 4.4
o
C warming to higher loss rates. This is likely to very much
overestimate the temperature at which this would occur.
(6) European wetlands- Baltic coast: As above with linear extrapolation of high range 98% loss with 4.4
o
C warming.
(7) European wetlands- Mediterranean coast: As above with a linear extrapolation of high range 100% loss with 4.4
o
C warming.
(8) Bangladesh, Sundarbans: Based on Qureshi and Hobbie (1994) and Smith et al. (1998) with sea level rise and temperature relationship (for
2100) drawn from Hulme et al. (1999b). This produces very similar results to an estimate based on “average” model characteristics. Some
models project higher sea level rise and others lower. Assumed relationship is 15% loss for 1.5oC (range 1-1.5oC) and 75% loss 3.5oC
(range 2-3.5oC).
(9) Australia, Kakadu region: This estimate is highly uncertain. In the WGII TAR report Gitay et al. (2001) assert that the wetlands “could be
all but displaced if predicted sea-level rises of 10–30 cm by 2030 occur and are associated with changes in rainfall in the catchment and
tidal/storm surges” (p308). Here it is assumed that a 30cm sea level rise displaces 80% of the wetlands and that the sea level rise vs.
temperature relationship is drawn from Hulme et al. (1999b) from the HadCM2 and HadCM3. Note that the estimate range from recent
models is 1.2-3.1oC for a 30cm sea level rise.
26
Figure 6 – Impacts on Animal Species
0 1 2 3 4
Arctic: Reduction in range of keystone arctic mammal - Collared
Lemming: 50% reduction in range for 1.7-2.2oC global increase
above 1861-1890 (1)
USA: Waterfowl breeding population reduction (overall reduction
in waterfowl abundance and wetland extent) in the Prairie Pot
Hole region - breeding population reduction 45% for ca 3.3oC
warming (2)
USA: Reduction in range of coolwater, salmonid fish in Rocky
Mountains (3)
USA: Reduction in range of coolwater, salmonid fish in
Wyoming habitat (4)
Mexico: Range reduction for many species with likely severe
ecological perturbations ( 5-19% of species lose 50% or more
of range with 1.9-2.4oC warming) (5)
No significant
effect (<5%) or
very low risk
Small impact (5-
10) or low risk
Moderate impact
(10-20%) or
moderate risk eg
local extinction
Large impacts (20-
50%) or significant
risk of extinction
Severe impacts
(>50%) or high risk
of extinction
See notes below:
27
Figure 6 – Impacts on Animal Species continued:
0 1 2 3 4
South Africa: 78% of 179 animal species studied experience range
contraction - 29 endangered or vulnerable species suffer 50% or more
reduction in range (1.9-3.1oC by 2050s wrt 1861-1890) (6)
South Africa: Predicted extinction of four species (1.9-3.1oC by 2050s
wrt 1861-1890) (7)
Australia: Very large range reduction or elimination of 3 species of
frogs and 15 species of endangered mammals in Dryandra forest of
south western Australia (8)
Australia: Predicted extinction of golden bower bird in highland tropical
rainforests (90% range loss with 3oC warming) (9)
Australia: Catastrophic loss of endemic verterbrates from highland
tropical rainforests for around 3oC warming(10)
Australia: Large range reduction (50%) for majority (>80%) butterfly
species with 2.9oC warming (11)
USA, Hawaii: Predicted extinction of honeycreepers (12)
No significant
effect (<5%) or
very low risk
Small impact (5-
10) or low risk
Moderate impact (10-
20%) or moderate risk
eg local extinction
Large impacts (20-
50%) or significant
risk of extinction
Severe impacts
(>50%) or high risk
of extinction
See notes below:
28
Notes: See Table 5 for more details
(1) Canadian Arctic, collared lemming: Based on data in Kerr and Packer (1998) with conversion of local temperatures to global mean based on a range of the
current AOGCMs; mid-range used. Interpolation is used to estimate range reductions based on data in Kerr and Packer (1998).
(2) USA, waterfowl population Prairie Pot Hole Region: Based on data in Sorenson et al. (1998) with interpolation of data.
(3) USA, reduction of Salmonid fish habitat in Rocky Mountains: Based on data in Keleher and Rahel (1996) with extrapolations to 5% and 10% reductions.
June, July, August temperatures ‘upscaled’ to global by associating projected JJA temperatures from a range of GCMs for the USA with global mean
temperatures using MAGICC/SCENGEN. This is obviously quite uncertain given that temperature changes in the region are likely to be quite different
from the USA average, with mountainous regions likely to experience amplification of trends for the continental averages.
(4) USA, reduction of Salmonid fish habitat in Wyoming: Based on data in Keleher and Rahel (1996) with extrapolations to 50% reduction. Upscaling of
temperatures as in (3).
(5) Mexico: Highly indicative interpretation of results of Peterson et al. (2002) for range reductions. The 50% range reduction level is associated with the
upper end of their warming scenario, which corresponds to 2.4
o
C warming above 1861-1890 and this range reduction applies to up to 19% of the entire
Mexican fauna. Between present temperatures and 2.4
o
C a linear scaling is used here. Note that there is projected to be a severe risk of extinction for up to
several tens of fauna species (0-2.4% of species lose 90% of range for 1.9-2.4
o
C warming).
(6) South Africa, range reductions of large number of animals: Highly indicative only, interpretation of results of Erasmus et al. (2002) for range reductions in
the 29 endangered species projected to experience 50% or more range reductions with a warming of 2.4
o
C (1.9-3.1oC range) (above 1861-1890). The scale
assumes that a 50% reduction in the range of these species occurs with 3.1
o
C. Lower reductions are linearly scaled from 1990 temperatures.
(7) South Africa, predicted extinctions: Highly indicative only interpretation of results of Erasmus et al. (2002) for extinctions projected for a 2.4
o
C increase
(1.9-3.1oC range). The scale used assumes that there is a 100% chance of extinction with a 3.1
o
C increase, zero probability at current temperatures, and the
likelihood of extinction increase linearly.
(8) Australia, south west Dryandra forest: Based on Pouliquen-Young and Newman (1999) as cited by Gitay et al. (2001). Assumed that “very large” range
reduction meant a 90% reduction, that the loss of range scale was linear for the present climate to a warming of 1.1
o
C (above 1861-1890), and that 90%
reduction occurs at 1.1
o
C.
(9) Australia, predicted extinction of Golden Bower bird of highland tropical forests, north east Queensland: Based on (Hilbert et al. 2003) and using range
reduction of 90% with a 3oC warming and linear interpolation for range losses between 1990 (0.6oC and 0% range loss) and this level.
(10) Australia, “catastrophic” loss of endemic vertebrates from rainforest in highland tropical rainforests: Based on (Williams et al. 2003) and with similar
scaling as above.
(11) Australia, large range reduction in range of butterfly species: Based in (Beaumont and Hughes 2002) with risk of large range reductions for large numbers
of species linearly increasing from zero at 0.6oC to 50% loss for 80% of species at 2.9oC.
(12) USA, predicted extinction for honeycreepers in montane forests of Hawaii: Based on (Benning et al. 2002) with risk of extinction to 90% at 3.2oC
29
Figure 7 – Impacts on Ecosystems
0 1 2 3 4 5
Boreal forests, China:
Reduction in extent of boreal
forest (70% reduction for ca.
2.8oC warming) (1)
Arctic, Canadian Low Arctic
Tundra - 77% loss with
3.3oC warming (2)
Arctic/Boreal, Boreal
woodland/Taiga 44% loss
by 3.8oC warming (3)
Arctic, Tundra ecosystem:
global loss of 57% with
3.8oC warming (3)
Alpine ecosystems, Europe:
38% of species losing 90%
of range by 4.5oC (4)
Alpine ecosystems, south
eastern Australia: (ca 90%
loss for 3.8oC warming) (5)
No significant
effect (<5%) or
very low risk
Small impact (5-
10) or low risk
Moderate impact
(10-20%) or
moderate risk eg
local extinction
Large impacts
(20-50%) or
significant risk
of extinction
Severe impacts
(>50%) or high
risk of extinction
30
Figure 7 – Impacts on Ecosystems continued:
0 1 2 3 4
Biodiversity Hot Spot, Succulent Karoo , South Africa: Severe risk of extinction -
projected to lose 80% of range for 1.9-2.4oC warming. Virtual disappearance at 3.4-
4.3oC with likely extinction of its 2800 endemic plants. (6)
Biodiversity Hot Spot, Fynbos , South Africa: Range loss and risk of extinction of
endemic plants in Fynbos biome. Projected to lose 51-61% of area, with 10% of
endemic Proteaceae species suffering complete range loss. (7)
Tropical forests, Highland tropical forests - Australia, Queensland: - 50% loss of area
with about a 1.8oC warming, catastrophic loss of verterbrates 2.6-3oC. (8)
Tropical forests, Amazon: risk of collapse and/or major loss of species (9)
Plant diversity threat, Europe: Changes in plant biodiversity with risk of extinction (32%
of sampled areas in Europe in 2050 no longer have species in them that are present
now for 2.4oC warming) (10)
Plant diversity threat, Australia: Eucalypt species out of climatic range (50% of species
out of current thermal range with 2.7-3.2oC warming) (11)
Coral reefs - Indian Ocean: extinction risk with 1.4oC warming (12)
Coral reefs- global assessment: Projected annual bleaching by 1.7-2.3oC warming (13)
No significant
effect (<5%) or
very low risk
Small impact (5-
10) or low risk
Moderate impact
(10-20%) or
moderate risk eg
local extinction
Large impacts
(20-50%) or
significant risk
of extinction
Severe impacts
(>50%) or high
risk of extinction
31
Notes: Details of each example are to be found in Table 5 - Ecosystem Impacts.
(1) Boreal forests, China: Based on Ni (2001) with linear scaling of loss of boreal forest in China with temperature.
(2) Arctic, Canadian Low Arctic Tundra: Loss of area is 77% with 3.3oC warming based on (Malcolm et al. 2002b) and linearly interpolated from zero at 0.6oC.
(3) Arctic/Boreal, Boreal woodland/Taiga and Arctic Tundra: Loss of ecosystems respectively 44% and 57% with 3.8oC warming and scaled linearly from zero at 0.6oC
warming. Based on (Neilson et al. 1997)
(4) Alpine ecosystems, Europe: Highly indicative measure of risk only. Scale is percentage of alpine species losing 90% of their range with linear scaling of the estimated
38% losing this level with a warming of about 4.7
o
C (range 3.3-4.7°C). This is done only to provide a visual picture of increasing risk with temperature, which is one of
the main findings of the literature for this region (see Table 5 - Ecosystem Impacts).
(5) Alpine ecosystems, south eastern Australia: Assumes 90% reduction with a warming of 3.8
o
C (above 1861-1890) with linear scaling of area loss from present climate.
Busby (1988) found that the alpine zone would be confined to only 6 peaks for a warming of 1.7-3.8
o
C.
(6) Biodiversity Hot Spot, Succulent Karoo , South Africa: Based on Midgley and Rutherford at http://www.nbi.ac.za/frames/researchfram.htm. The scale is likelihood of
extinction of the 2800 plants endemic to the Succulent Karoo ecosystem, where it is assumed that the systems will no longer exist with 100% certainty with an increase of
2.4
o
C and that the likelihood of extinction scales linearly upward from zero at current temperatures.
(7) Biodiversity Hot Spot, Fynbos , South Africa: Based on Midgley et al. (2002) and linear scaling loss of the area of Fynbos with temperature from zero at present up to
61% loss of area with a 2.4
o
C increase (above 1861-1890). Ten percent of endemic Proteaceae species are projected to suffer complete loss of range, and hence are also
very likely to become extinct with a 51-61% area loss in Fynbos.
(8) Tropical forests, Highland tropical forests - Australia, Queensland : Based on results of Ostendorf et al. (2001), Hilbert et al. (2001), Williams et al. (2003) and Hilbert et
al. (2003)with linear scaling of area losses with local temperature increase. Across results from different assessments this produces fairly consistent estimates.
(9) Tropical forests, Amazon: This is speculative drawing on the work of Cowling et al. (2003) and Cox et al. (2003) and assuming that there is a 50% risk of collapse
with a warming of 2.4oC. See discussion in Table 5 and footnote XX and Note (1) at the end of this table.
(10) Plant diversity threat, Europe: Based on Bakkenes et al. (2002) with scale being fraction of plant species occurring at present within a grid cell in Europe that no longer
appear with given level of warming. Assumes linear scaling with temperature increase above the present. As such is indicative only of increasing risk with temperature,
the risk being that of extinction or severe range reduction. The absence of plants from a grid cell in 2050 does not imply that the species is globally extinct, only that it is
no longer climatically suited to that region. The higher the fraction of species displaced in the model is a measure of the ecological dislocation caused by rapid warming
and for some species is indicative of the rising level of extinction risk.
(11) Plant diversity threat, Australia: Based on Hughes et al. (1996). Scaled number of species out of climatic range with temperature above present.
(12) Coral reefs - Indian Ocean: Based on the work of who predicts extinction of reef sites in the southern Indian Ocean for warming in the range 0.9-1.4°C. It is assumed
that there is a 90% chance of extinction at a temperature increase of 1.4oC.
(13) Coral reefs - global assessment: Based on results of Hoegh-Guldberg (1999). For both models used and all reefs studied, annual bleaching occurred by 2040s. Scale is
chance of a major bleaching occurring in a decadal period e.g. 10% corresponds to 1 year per decade, 50% to five year out of 10 and 100% to annual bleaching. Scaling
is from 0.4
o
C above 1861-1890 as unusual bleaching began in the 1980s with annual bleaching occurring at 2.3
o
C above 1861-1890.
32
Table 5 - Ecosystem Impacts
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
Arctic
ecosystems
Arctis Major range reduction
for a keystone arctic
species, the collared
lemming, with likely
large negative impacts
on Arctic ecosystems.
Range reduction
50%
80%
1.7-2.2°C
7
3.3-4.5°C
8
The Collared Lemming (Dicrostonyx groenlandicus) is basic part of the food
chain and a major food source for a number of predators – birds and mammals
(Kerr and Packer 1998). Twenty-five mammals in Canada have their northward
range movement limited by the Arctic ocean (See also TAR WGII Section
5.4.3.2). The temperature scenarios used were for a 2, 4 and 6oC local mean
annual warming
9
. The percentage reductions in range cited are interpolated from
the data in Table II of Kerr and Packer (1998). Temperatures in the Arctic
region are known to be warming rapidly, with the rate of warming appearing to
increase recently
Arctic mammal distribution is closely
correlated with temperature and a
number of mammals are adapted to
survive in colder climates. Warming
is projected to lead to their northward
migration, assuming habitat
availability. The Arctic ocean places
a limit on the extent of this
possibility.
Arctic
ecosystems
Arctis Substantial reduction
of sea ice area and
possible complete loss
of summer sea ice in
the Arctic ocean by
end of 21
st
century, or
earlier depending on
scenario, with major
implications for ice
dependent species.
Reduction in annual
ice cover
15-20%
40-50%
Mid summer ice cover
reduction
50%
2°C
4°C
1.5-2°C
Sea ice area, extent and thickness have been declining in recent decades, with the
perennial cover being lost at a rate of 9% per annum over the period 1978-2000
(Comiso 2002b). A strong correlation has been observed between warming and
ice losses (Comiso 2002a). Record losses were reported by for sea extent and
area in 2002 (Serreze et al. 2003). The HADCM3 model predicts a further 15-
20% (40-50%) reduction in annual ice cover for a 2°C (4°C) increase in global
mean temperature above 1861-1890 (Gregory et al. 2002). A much larger
proportional reduction in summer ice is projected, with a loss of 50% by the
2050s corresponding to a global mean warming of around 1.5-2°C (Gregory et
al. 2002). Johannessen et al. (2002), using ECHAM4 and HADCM3 with a
new sea ice observed data set, predict for summer a “predominantly ice-free
Arctic Ocean” by the end of the 21
st
century. Their mid summer ice loss
projections of 30-60% by the 2050s, depending on scenario, are similar to those
of Gregory et al. (2002). Amongst other effects this could be expected to have
profound implications for arctic and sub arctic marine biodiversity and would
affect, almost certainly negatively, polar bear populations (Stirling et al. 1999;
Stirling 2000).
Arctic sea ice responds rapidly to
warming on a timescale of years
rather than decades. Polar bears are
dependent on sea ice for hunting and a
loss of sea ice is very likely to reduce
the viability of bear populations
(Stirling et al. 1999). An ice free
Arctic ocean in summer would also
lead to very large changes in the
marine biota with negative
consequences for ice dependent
species.
6
Above 1861-1890 average unless otherwise stated. See Appendix on temperature scale.
7
Local temperature increase scenarios are converted to global mean using average of nine recent GCMs upscaled from the Canadian Arctic region using SCENGEN. The scaling used is 1.86°C local increase per degree of
global mean increase calculated with the A1B-MESSAGE scenario, with the range set by the inter-model standard deviation of 0.3°C/°C. Using the full range of models available in SCENGEN produces a lower scaling,
however examination of the scaling for the higher Arctic region of Canada, which is what would apply under the range reductions cited in the table, indicates a higher scaling factor (2.07 °C/°C with inter-model standard
deviation of 0.42°C/°C). This would tend to slightly lower the upper end of the global temperature range (e.g. the range would be 1.5-2.1°C for a 50% loss and 2.9-4.2°C for an 80% loss.
8
The maximum local warming of 6°C produced a range reduction of 78%. The local temperature increase corresponding to an 80% reduction is estimated from a 2nd order polynomial regression on the data in Table II of
Kerr and Packer (1998). A linear extrapolation would produce a slightly lower temperature increase.
9
The baseline for this warming is assumed to be 1961-1990 as the observed distribution of the mammals was regressed against historical means annual temperatures.
33
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
Arctic
ecosystems
Arctis –
global
assessment
Global loss of area of
tundra ecosystem
40-57%
1.3-3.8°C
10
Large losses of tundra ecosystem are projected for a range of future climate
scenarios taking into account the effects of CO
2
increases. Projected ecosystem
area losses are drawn from the assessment of Neilson et al. (1997) prepared for
the IPCC Regional Impacts report
11
. Here only the results from climate models
and scenarios used in the IPCC Second Assessment Report (SAR) are cited.
Results for the scenarios drawn from climate models reviews in the IPCC First
Assessment Report (FAR) are not used here, as the models are older and less
reliable
12
.
Warming causes the northward
migration of Tundra and other high
latitude ecosystems. The tundra in
particular has its migration limited by
the Arctic ocean. The rate of required
migration is found to be higher than
known from past climatic changes.
Arctic
ecosystems
Arctis -
Canada
Large reductions in
tundra and taiga
projected. Estimated
future rate of change
of climate exceeds
known past changes.
Loss of area of
Canadian Low Arctic
Tundra
75-77%
13
(19% loss of species
estimated)
2.2-3.3°C
14
Malcolm et al. (2002b) estimated migration rates for biomes globally in response
to climate change using several GCMs and two vegetation models. They found
that high latitude and Arctic ecosystems (boreal forests, taiga) needed very high
migration rates to keep up with projected rates of climate change. Tundra
systems in particular experienced large area losses in this assessment. In a
related study, Malcolm et al. (2002a) found that several high latitude and arctic
ecosystems were particularly vulnerable to rapid change under each of the
models examined
15
. These systems included the Canadian Low Arctic tundra,
Boreal Taiga, East Siberian Taiga, Russian coastal tundra, as well as several
boreal forests. Species loss in response to the loss of area of ecosystems was
estimated using established species-area relationships. Such estimates may be
conservative (Seabloom et al. 2002). For the Canadian Low Arctic tundra,
where an average 76% area loss was projected, the corresponding species loss
was estimated to be around 19%.
Warming causes the northward
migration of Tundra and other high
latitude ecosystems. The tundra in
particular has its migration limited by
the Arctic ocean. The rate of required
migration is found to be higher than
known from past climatic changes.
10
Based on the transient scenarios used by Neilson et al. (1997), which were with reference to 1961-1990 and are described as having global mean surface temperatures increases in the range 1-3.5°C by the time of CO2
doubling.
11
See Table C-1 in Neilson et al. (1997) for the full range of results.
12 Some literature uses the full range including the IPCC First Assessment Report (FAR) scenarios, which in general produce somewhat different results (see Table 2 of Kittel et al. (2000)) showing a reduction in area of 40
to 67% for the tundra)
13
Malcolm et al. (2002b) for range of BIOME3 and MAPSS projections under the climate scenarios assumed.
14 Malcolm et al. (2002b) base their projections on HadCM2 scenarios with and without sulphur and on the ECHAM4 scenario without sulphur for the period of 2070-2099. These scenarios have a global warming range,
relative to 1961-1990 of 1.9-3.0°C, to which added the warming from 1861-1890 of around 0.3°C to the base period (see their footnote 2 for further details).
15
The global vegetation models MAPSS and BIOME3 were used to model the equilibrium distribution of generalized plant types for the present and future projected climates. At larger spatial scales these kinds of models
perform reasonably well (Pearson and Dawson 2003)
15
Differences between present distributions and projected future distributions were analysed to estimate the rate of migration required for biome types to keep pace with the projected climate changes. Migration rates were
computed taking account physical barriers and human land use. A general pattern observed was a “front” of very rapid migration rates at higher northern latitudes, where climate changes are expected to be most rapid.
BIOME3 and MAPSS use ecological, hydrological and physiological processes to describe the distribution of species.
34
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
Boreal Forests Eurasia and
North
America
Significant losses of
boreal forests and
associated carbon
stocks projected.
Releases of carbon
60-90GtC after 100
years
2.8°C
16
Kirilenko and Solomon (1998) use a transient scenario to assess the effects of
climate changes on a number of major ecosystem types taking into account
different rates of potential plant migration and also agricultural land demand.
They find a large release of carbon from this system due to transient effects, of
the order of 60-90 GtC after 100 years, using average migration rates and taking
into account agriculture. Kirilenko et al. (2000) examine the implications of
changes in the variability of climate for the boreal forests, finding that increased
variability slightly reduces the amount of forest loss.
Whereas tree dieback and loss can
occur very quickly due to
disturbances, regrowth is significantly
slower (Kirilenko and Solomon 1998;
Kirilenko et al. 2000). Several model
projections for changes in high
latitude vegetation and confirm that
these ecosystems will be far from
equilibrium in the future due to the
rapid climate change (Brovkin et al.
2003). The changes are likely to be
abrupt and there is a significant
positive feedback to climate warming
with the changes in vegetation and
snow cover projected.
Boreal Forests Eurasia and
North
America
Losses of boreal forest
and woodlands
Boreal forests
36% - 10% increase
Boreal
woodland/Taiga
36-44%
1.3-3.8°C
10
1.3-3.8°C
10
Using the BIOME3 and MAPSS equilibrium vegetation models large potential
losses of total area of boreal forest and woodland are projected (Neilson et al.
1997). Changes to the boreal forests (not including woodland/Taiga) are in the
range of a 36% decrease to a 16% increase, whereas the boreal woodland/Taiga
has a projected decrease in the range 36% to 44%
17
. Climatic pressure on the
boreal woodland/Taiga is clear also from the work of Malcolm et al. (2002a). In
this latter work, which is based around the same models but a narrow range of
climate scenarios (see footnote 14), a number of Taiga regions are identified as
being particularly and fairly consistently at risk. Using the LPJ dynamic
vegetation model Kittel et al. (2000) find the largest rates of change at the
present southern limits of the boreal forests in central and western Eurasia .
See above
Boreal Forests China Reduction of boreal
forest area in China.
70%
2.8°C
18
Large reductions in the area of boreal forests in China are projected using
BIOME3 (Ni 2002). Ni found “dramatic changes in geographic patterns, with
70% reduction in area and disappearance of almost (sic) boreal forests in
northeast China.” Climate projections from the Hadley model (Johns et al.
1997) for the period 2070-2099
18
were used relative to a 1931-1960 base period,
to estimate changes in ecosystems and carbon storage in China. The atmospheric
CO
2
in the model was increased to 500 ppmv in 2070-2099 from 340 ppmv in
the base period. A reduction in carbon storage in China’s boreal forests is
projected, however other work by (Ni J 2001) and Ni et al. (2000) show that
carbon storage should increase in China as a whole.
Warming causes poleward shift of
many ecosystems and the boreal
forests experiences pronounced
pressure in this direction.
19
16
Kirilenko and Solomon (1998) use projected climate change from a CO2 doubling scenario of Manabe et al. (1992). Table B-1 of the IPCC Special Report on the Regional Impacts of Climate Change (Watson et al. 1998)
indicates that at the time of CO2 doubling, around 2050, the Manabe et al. (1991; 1992)scenario projects a warming of 2.2°C, with respect to 1990.
17
See Table C-1 of Neilson et al. (1997) and using only the IPCC Second Assessment Report scenarios for climate changes at the time of CO2 doubling. The results for the scenarios drawn from First Assessment Report
are not used.
18
Ensemble average of the HadCM2 scenarios forced with IPCC IS92a emissions including the effects of sulphur emissions for the period 2070-2099. Data from the IPCC DDC website.
19
For strengths and weakness of bioclimatic envelope models see Footnote 3.
35
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
Alpine and
mountain
Europe -
Alps
Large range decrease
for alpine species.
Percentage of species
with greater than 90%
range loss:
3.2%
17.7%
38%
1.5-2°C
20
2.4-3.3°C
21
3.3-4.7°C
22
Three local warming scenarios (1.5, 3, and 4.5°C) with respect to the present
climate (assumed to be 1990, as plant distributions are calibrated on the current
climate) show a number of impacts. A few plant extinctions (1-3) are projected
in the study area (Guisan and Theurillat 2000). Perhaps more importantly the
study shows very large range decreases (90%) for 3.2%, 17.7% and 38% of
species for each of the three temperature scenarios respectively (see Table 1 of
Guisan and Theurillat (2000)). Whilst the authors caution that these are not to be
taken as accurate predictions their results do provide a basis for assessing the
major likely direction of changes.
The highest alpine species, whose
ranges are restricted to the alpine
zone, would experience a reduction in
suitable bioclimatic zone due to
warming and topography of
mountains, where suitable physical
habitat area declines rapidly with
altitude.
Alpine and
mountain
Europe –
Alps
Risk of extinction of
high alpine and nival
plant species.
Likely tolerable limit
for most alpine and
nival species but could
be exacerbated by land
use changes in many
areas.
Disappearance of
some categories of
vulnerable plants and
substantial further
range reduction of
many other species.
1.2-2.4°C
23
2.4-4.3°C
24
The IPCC TAR found, based in part on the work of Theurillat and Guisan
(2001), that a local warming of 3-4°C was most likely not to be within the range
species could tolerate.
25
The IPCC TAR also found that in the European Alps
the literature suggested that most alpine and nival species seem likely to able to
cope with a local warming of 1-2°C. Some isolated orophytes living at the tops
of mountains and some nival species are projected to lose area or disappear. By
far the greatest negative ecological impacts appear to be in the upper elevations
or true alpine zone. Theurillat and Guisan (2001) argue that species restricted to
low mountain tops or whose range is limited by soil and other factors to small
areas are likely to be “severely endangered by extinction.” They argue that
whilst the maintenance of traditional land use is essential to reduce the effects of
warming, it is likely that other land uses will reduce the resilience of the alpine
system to climate change.
High Alpine and nival species are
restricted in range and warming will
reduce that range. Where species are
endemic to a mountain or range of
mountains and bioclimatic zone rises
then there is likely to be substantial
pressure on vulnerable species.
Suitable habitat declines with altitude.
Alpine mountain Asia –
Tibetan
Plateau
Large scale changes to
environment of
Tibetan plateau and
acceleration of
desertification.
2.8°C
18
A large reduction in the temperate desert, alpine steppe, desert, and ice/polar
desert are projected using the equilibrium vegetation model BIOME3 driven by a
climate scenario derived with the HadCM2 model (Ni 2000). With the projected
warming it can also be expected that there will be a large increase in the cold-
temperate conifer forest, temperate shrubland/meadows, and temperate steppe,
along with a general north-westward shift of all vegetation zones. Continuous
permafrost would mostly disappear. With the expansion of permafrost free
regions this would accelerate desertification of the Tibetan plateau
(Ni 2000)
.
Warming of the high altitude plateau
of Tibet causes a reduction in the
coldest bioclimatic type and in
permafrost. There are special high
attitude biomes that would be
substantially reduced with warming.
Other ecosystems would expand.
20
A 1.5°C local temperature increase converted to global mean using the average of nine recent GCMs downscaled to the European Alpine region using SCENGEN. The regional to global scaling used is 1.39°C/°C with the
range set by the inter-model standard deviation of 0.3°C/°C. The scaling factors using all 17 models in SCENGEN are not very different from the 9 model estimate. The base period is assumed to be 1990 hence 0.6°C is
added to the global temperature to estimate the increase with respect to 1861-1890.
21
A 3°C local temperature increase converted as in footnote 20 to a global mean increase.
22
A 4.5°C local temperature increase converted as in footnote 20 to a global mean increase.
23
A 1-2°C local temperature increase converted to the global mean as in footnote 20.
24
A 3-4°C local temperature increase converted to the global mean as in footnote 20
25 See IPCC TAR WGII TAR 13.2.1.4. Mountains and Subarctic Environments http://www.grida.no/climate/ipcc_tar/wg2/500.htm (Kundzewicz et al. 2001).
36
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
regions this would accelerate desertification of the Tibetan plateau (Ni 2000).
The scenario used by Ni (2000) is based on the HadCM2 model with an
emissions scenario that includes sulphur and was constructed with a base period
of 1931-1960 compared with 2070-2099. The effects of increased CO
2
concentrations were accounted for, with a CO
2
level of 500 ppmv being used in
the projection period years and 340 ppmv in the base period.
Alpine mountain Australia Major reduction and
ultimate loss of Alpine
zone in southeastern
Australia and
consequent loss of
endemic species.
Confinement of the
alpine bioclimatic
zone to six peaks.
Likely elimination of
northern alpine
bioclimatic zone
Confinement to
isolated mountain tops
in Tasmania
1.7-3.8°C
26
2.1-4.1°C
27
4.0-8.1°C
28
The alpine zone in southeastern Australia appears to be one of the more
vulnerable ecosy stems. The IPCC TAR assessment for Australia (IPCC TAR
WGII Chapter 12 (Pittock et al. 2001)) found that: “Warming of 1°C would
threaten the survival of species currently growing near the upper limit of their
temperature range, notably in some Australian alpine regions that already are
near these limits.” This confirmed the findings of the 1998 IPCC Regional
Impacts assessment report, which found that the Australian alpine region was
one of the most vulnerable systems in the region. This risk was first identified in
1988 by Busby who estimated that a warming of around 2-3°C in southern
Australia would result in the confinement of the alpine zone to only six peaks,
with a “dramatic effect on the survival of the majority of the present alpine
species”. More recent bioclimatic modelling by (Brereton et al. 1995) confirmed
the overall assessment of Busby (1988). Based on lapse rate considerations,
there is a substantial risk that, for a warming above about 3°C over 1990 levels,
the northern Alpine zone would no longer exist and that before this many of the
endemic species to the Australian zone in this region would become extinct in
this region (Hughes 2003).
The geography of this region is such
that there is very limited scope to
attitudinal migration. Using standard
lower troposphere lapse rates the rise
in the Alpine bioclimatic zone with
temperature can be calculated with
increasing mean temperat ure (Peters
and Darling 1985) . The present
estimate is based on the geography of
this Alpine region and its bioclimatic
zonation. Much of the region is
protected as a national park, hence
land use pressures as such are not the
main determinant of the future of this
region.
Alpine mountain Australia Major reductions in
snow area with
negative impacts on
snow dependent
species.
18-66%
39-96%
0.9-1.9°C
29
1.2-4.0°C
30
Projections for the Australian Alps indicate a major loss of snow coverage with
warming (Whetton et al. 1996).
31
The most recent scenarios for southern
Australia are warmer than the 1996 scenarios - 0.6-3.4°C by 2030 as opposed to
0.3-1.3°C and 0.8-5.2°C by 2070 as compared to 0.6-3.4°C (CSIRO 2001),
indicating a larger loss of snow area.
32
Projected climate change results in
warming and changes in circulation
which reduces snow precipitation and
the period in which snow cover can be
maintained,
26
Based on Busby (1988) assuming that the scenario used has a base period of 1975-1984 and that the local temperature increases of 2-3°C is with respect to this period. These regional temperature increases are scaled to an
estimated corresponding global mean temperature increase using 0.985°C/°C with an inter-model standard deviation of 0.194°C/°C obtained using 9 recent models and the SCENGEN programme (REFS). Note that the
regional definition over the Australian Alpine region using SCENGEN is very coarse. Choosing slightly different regions or using the full range of models in the SCENGEN utility does not change the range fundamentally.
27
Estimate based on a lapse rate in the range of 0.6-0.8°C/100m and using Mt Kosciuszko at 2200m as the highest point in the northern alpine zone with the rise from the beginning of the Alpine zone at 1800 m in the 1980s
to 2200m defining the local temperature increase required to eliminate the northern Alpine zone. The same regional to global scaling is used as in footnote 26. Note that using a scaling for a slightly narrower and more
northerly region but still covering the Mt Kosciuszko area would reduce the range to 2-3.9°C.
28
Estimate based on a lapse rate in the range of 0.6-0.8°C/100m and estimating rise of current Alpine zone in Tasmania, which starts at about 800m to 1500 metres, leaving a few peaks above this level. The regional
downscaling used those for southeastern Australia from SCENGEN as these were most consistent with the CSIRO scenarios for southern Australia and Tasmania. The scaling used was 0.874 °C/°C with an inter-model
standard deviation of 0.184 °C/°C. The grid cells available from SCENGEN over Tasmania are mostly ocean and may underestimate the warming locally in Tasmania. If that had been used the scaling used was 0.652°C/°C
with an inter-model standard deviation of 0.192 °C/°C producing a range of warming from 5-12.2°C.
37
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
Alpine mountain Australia Endangerment and
possible extinction of
species.
Likely extinction in
the wild of the
mountain pygmy
possum due to
complete loss of its
bioclimatic zone.
1.2-1.7°C
33
A number of vertebrates and plants are limited to the Alpine zone and require the
seasonal occurrence of snow (Hughes 2003). Three mammals (dusky
antechinus, broad-toothed rates and the mountain pygmy possum), whose
abundance increases with altitude, are adversely affected by loss of snow (Green
and Pickering 2002; Hughes 2003). The bioclimatic zone occupied by the highly
endangered Mountain pygmy-possum (Burramys parvus) is estimated to be lost
with a local warming of 1°C (Brereton et al. 1995). Invasion of the high plains
of the Alpine zone with sub alpine, woody species would lead to a substantial
change in the landscape. Distributions of trees are limited to be zones where the
average temperature of the warmest month is greater than 10°C.
The Australian Alpine zone has a very
limited altitudinal range, being
essentially plateaus, and hence
beyond a certain temperature increase,
upwards altitudinal migration is
impossible. The Pygmy possum
(Burramys parvus) is limited to about
10km
2
of habitat. Given this situation
climate change is clearly a longer-
term pressure on this species,
however there are intensive efforts
being made to maintain this species in
situ. Loss of snow cover would most
likely mean, or at least contribute very
strongly to, the extinction of the
pygmy possum in the wild.
Alpine mountain Australia Large range reduction
of the Alpine tree frog
in Eastern Australia
51-89%
3.1-4.6°C
34
The Alpine tree frog (Litoria verreauxii alpina) is one of the species threatened
by climate change. Using a bioclimatic model BIOCLIM Brereton et al. (1995)
estimated that a 3°C warming would reduce the frog’s range by 51-89%
35
. The
range of the Alpine tree frog is thought to be limited to several national parks
(Kosciusko National Park; Namadgi National Park; Alpine and Buffalo National
Parks) and some public forests
36
. Range losses have already occurred at lower
Alpine plateaus such as Mt Baw Baw. Whilst drought has been linked to these
losses, there is no final assessment of the causes of this range reduction.
Management issues in public forests as well as in the protected areas mentioned
above have a direct bearing on the species vulnerability. It is clear, however that
climate change is likely to have a determining influence in the longer term.
Warming will reduce the frog’s range
according to estimates with a
bioclimatic model. Land use change
pressures occur but most of the
present range lies within protected
areas. Hence climate change is likely
to put very strong adverse pressure on
the species.
29
Original projections for 2030s are with respect to 1990 hence an offset of 0.6°C is added to obtain the range of increase wrt to 1861-1890.
30 Original projections for 2070s are with respect to 1990 hence an offset of 0.6°C is added to obtain the range of increase wrt to 1861-1890.
31
See also CSIRO (1996). Note that the CSIRO has produced new scenarios (CSIRO 2001), which in general predict higher warming than the 1996 scenarios.
32
New snow cover projections have been released recently (Hennessy et al. 2003) which project larger losses of snow cover than those shown here.
33
The regional temperature increases of 1°C is scaled to an estimated corresponding global mean temperature increase using 0.874°C/°C with an inter-model standard deviation of 0.184°C/°C obtained using 9 recent models
and the SCENGEN programme assuming that the baseline climate is 1961-1990. The 1989 CSIRO scenario used for this work was not available. If the baseline was 1990, then the global mean temperature increase above
1861-1890 would be 1.5-2°C.
34
The regional temperature increases of 3°C is scaled to an estimated corresponding global mean temperature increase using 0.874°C/°C with an inter-model standard deviation of 0.184°C/°C obtained using 9 recent models
and the SCENGEN programme assuming that the baseline climat e is 1961-1990. The 1989 CSIRO scenario used for this work was not available. If the baseline were 1990, then the global mean temperature increase above
1861-1890 would be 3.9-4.6°C.
35
Bioclimatic models such as BIOCLIM tend to overestimate species ranges (Hughes 2003).
36 See http://ea.gov.au/biodiversity/threatened/action/frogs/17.html.
38
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
Coastal Wetlands Global Progressive coastal
wetland loss with
increasing warming.
0-2.3%
37
1.8-10.5%
38
5.5-22.2%
39
1.4-1.5°C
40
2.4°C
3.3-3.4°C
The greatest losses of coastal wetlands are projected in the Mediterranean and
Baltic region, with large losses also in the Sundarbans (Nicholls et al. 1999).
Significant losses on the Atlantic coast of Central and North America and the
smaller islands of the Caribbean are also projected. The climate scenarios used in
the study were the result of greenhouse gas only runs with the HadCM2 and the
HadCM3 models forced by greenhouse gas emissions approximating the IS92a
scenario, which produces a warming of around 3-3.1°C by the 2080s. Resulting
sea level rise from these models was 40-41 cm by the 2080s relative to the 1961-
1990 mean sea level (Hulme et al. 1999a).
Model based assessment of the
vulnerability of each region to sea
level rise taking into account local
factors.
Coastal wetlands Australia Loss or serious
damage to Kakadu
World Heritage listed
wetlands.
1.2-3.1°C
41
The topography of the wetlands for the Kakadu regions appears to lead to an
especially vulnerable situation. A sea level rise of 10-30 cm combined with
rainfall changes and increased tidal surges is postulated to severely reduce the
freshwater wetlands of this region. The authors of the ecosystem assessment in
the IPCC TAR argue that these wetlands could “be all but displaced if predicted
sea-level rises of 10–30 cm by 2030 occur and are associated with changes in
rainfall in the catchment and tidal/storm surges” (Gitay et al. 2001) (WGII TAR
Chapter 5, p 308). Associating these sea level increases with a global mean
surface temperature change is difficult (see footnote 41). Sea level will result in
saltwater intrusion and shoreline erosion, with the loss of some coastal
mangroves (with colonization along creeks as the tidal zone expands), extensive
loss of paperbark trees (Melaleuca spp.) in the wetland, and ultimately
replacement of freshwater wetlands by saline mudflats (Eliot et al. 1999). These
vegetation changes would results in adverse changes in the abundance of wildlife
such as Magpie Geese and long-necked turtles, which are hunted by the
aboriginal owners of Kakadu. Gitay et al. (2001) point to the loss of large areas
of freshwater wetlands further to the west, in the Mary river, as a consequence of
salt water intrusion (Mulrennan and Woodroffe 1998). The possibility that the
processes that drive the vulnerability of the Kakadu wetlands to sea level rise
could extend to much larger regions with similar low lying character in the
monsoonal tropic is raised but not explored in the TAR.
The vulnerability of the wetlands of
Kakadu, and of other river systems in
the region, arises as these areas lie
within 0.2-1.2m of high water level.
The coast is largely mangroves with
inland fringing salt flats of low
productivity and diversity. Behind
these lie low ridges that form a barrier
to salt water intrusion onto the low
lying flood plains, below the inland
escarpment some 100km from the
coast. Sea level rise is postulated to
lead to the retreat of the mangrove
zone and the inland spread of the tidal
zones of the creeks of the region,
penetrating into the freshwater zone
(Gitay et al. 2001) and (Bayliss et al.
1997).
37
These results are for a sea level rise of about 12cm in the 2020s with respect to 1961-1990. The range is the highest and lowest estimates taking into account a number of factors and using the sea level rise scenarios from
the HadCM2 and HadCM3 models – see Table 10 of Nicholls et al. (1999).
38
These results are for a sea level rise of about 24-25cm in the 2050s with respect to 1961-1990.
39
These results are for a sea level rise of about 40-41cm in the 2080s with respect to 1961-1990.
40
See Table 1 of Hulme et al. (1999a) for an overview of the results of these scenarios.
41
Estimated warming at the time of sea level rise of 10-30cm (above 1990) being reached based on HadCM2 projections (Hulme et al. 1999a). Note that associating sea level increases of 10-30cm (or any increase) with a
particular global mean surface temperature change at a particular time in the future is difficult and highly problematic due to the long-term character of sea level rise and its response to global warming. Regional sea level
changes are also likely to be different from the global mean changes. More fundamentally sea level rise due to the thermal expansion of the ocean arising from increased heat input due to elevated levels of greenhouse gases
and the response of ice sheets occurs over a long time period. Indeed centuries are required for the ocean to come into full equilibrium with change levels of radiative forcing. As a consequence the sea level rise expressed at
any point in time is a fraction of that which is likely to occur for the full response to elevated greenhouse gas concentrations. One way to characterize the response is estimate the warming, which if held constant, would
result in a commitment to a certain sea level rise. This is also fraught with difficulties not the least of which is the very broad range of uncertainty in sea level rise estimates. From this point of view warming to date, if
39
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
Coastal wetlands Europe Loss of coastal
wetlands in Atlantic
Baltic and
Mediterranean coasts
for the 2080s.
Atlantic coast
0 to 17%
42
Baltic coast
84 to 98%
Mediterranean coast
31 to 100%
2.6-4.4°C
43
Projections for coastal wetland losses for the 2080s in the European region
indicate that the Baltic and Mediterranean coasts are most vulnerable. These
estimates are based on new runs of the model described by (Nicholls et al. 1999),
whose global results are shown below. See IPCC WGII TAR Table 13-4 for the
full results. These were constructed using the four preliminary IPCC SRES
marker scenarios and roughly span a temperature increase relative to 1990 of
2.0-3.8°C with a central estimate of sea level rise of 36-42cm
44
. The range of
losses shown opposite is the result of sea level rise uncertainty, which is larger
than the range mentioned in the preceding sentence, and uncertainty in relation to
the response of wetlands. Nevertheless the WGII TAR Chapter 13 notes that
under the high scenario wetlands in the Baltic and Mediterranean would be
eliminated which “could have serious consequences for biodiversity in Europe,
particularly for wintering shorebird and marine fish populations” (Kundzewicz et
al. 2001).
Model based assessment of the
vulnerability of each region to sea
level rise taking into account local
factors.
Coastal Wetlands USA Extensive loss of
wetlands in southern
New England
3-5°C
45
Recent rates of sea level rise of about 2mm/yr along with local subsidence rates
of about 1mm/yr have led to the displacement of high marsh species with less
rich cordgrass (Donnelly and Bertness 2001a). It is expected that if current rates
of sea level rise continue then high coastal marshes will be further displaced by
cordgrass in the coming century. Higher rates of sea level rise than around
2mm/yr, as projected for next century, will cause the cordgrass to drown and
there will be extensive overall loss of wetlands in southern New England. Local
accretion rates are in the range of 2-6mm/yr. Sea level rise of around or greater
than 6 mm/yr could be anticipated to result in large wetland losses. Warming in
the range of 3-5°C in 2100 (above 1861-1890) could be expected to produce
local sea level rise rates above 6mm/yr. It should be noted that associating a
specific temperature increase with a rate of sea level rise is difficult and
uncertain, nevertheless a warming rate as above would most likely lead to a rate
of sea level sea level rise sufficient to overwhelm the adaptive capacity of the
marshes.
Sea level rise in excess of accretion
rates will results in the loss of
wetlands. High marshes when
invaded frequently with saltwater are
replaced by cordgrass. If the rate of
sea level rise exceeds the accretion
rates possible regionally then total
loss of wetland occurs.
Coastal Wetlands USA Wetland losses in
Delaware.
21%
>2.5-3.5°C
46
Sea level rise projected for 2100 would reduce Delaware’s land area by 1.6% and
likely cause loss of around 21% of the wetlands in the area (Najjar et al. 2000).
47
This loss of wetland area occurs for a loc
al sea level rise of about 70cm of which
about 20cm is due to local effects.
48
As consequence of the projected sea level
Rate of sea level rise exceed capacity
of marshes and wetlands to adapt
given estimates of potential accretion
rates and other factors. As much of
maintained could already mean a commitment to a sea level rise of 30cm or more. At the other extreme one can simply associate the range of sea level rise estimates with the time a which they occur under a range of
scenarios. In this case an estimate of global mean warming at the time of the sea level rise of interest can be made, although it suffers from the inadequacy described above. The IPCC TAR estimates that sea level rise over
the coming century would be in the range of 0.8 cm-8 cm/decade. A 10-30 cm global sea level rise would correspond to a warming in the range of 1-3°C. The range of GCM models available with temperature and sea level
rise projections available on the IPCC Data Distribution Centre website indicate that at the time of a 10 cm sea level rise global mean warming is likely t o be in the range 0.9-1.3°C. For a 30cm sea level rise the range would
be 2-2.8°C (above 1961-1990). The 1961-1990 mean is around 0.3°C above 1861-1890.
42
The higher losses are associated with the higher temperatures and sea level rise. See the general discussion in footnote 41.
43
This the approximate range of global mean temperature increases associated with the losses of coastal wetlands in the European assessment. The temperature increases are for the year 2100 (not the 2080s) and are with
respect to 1861-1890. See footnote 44 for the source.
44
See Table 3-9 of IPCC WGII TAR Chapter 3 (Carter et al. 2001) for a summary of these scenarios, but noting that the SRES A1 scenario was split into three markers in the final SRES scenario set up. This resulted in the
upper end of the A1 range having a higher warming than the A2 scenario but this was not included in the scenarios for European sea level rise and coastal wetland loss.
45 See level rise rates above 2mm/yr may well result from temperature increases less than this.
40
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
This loss of wetland area occurs for a local sea level rise of about 70cm of which
about 20cm is due to local effects.
48
As consequence of the projected sea level
increase, current 100-year floods are projected to occur 3-4 times more
frequently. Under current rates of sea level rise (around 2mm/yr), high coastal
marsh species are being displaced by low-marsh species like cordgrass. Over the
last century, cordgrass has been able to keep pace with or surpass rates of SLR of
2 to 6 mm/year, however local sea level rise of around 6 mm/yr or greater is
possible for the warming range projected for 2100. Rates of change of this
magnitude could lead to the drowning of cordgrass communities and extensive
loss of coastal wetlands in southern New England. Given the inertia of sea level
rise, the warming that actually causes this sea level rise would be less and
correspondingly, a warming at this level would result in much greater sea level
rise in the 22
nd
and following centuries.
given estimates of potential accretion
rates and other factors. As much of
the area affected is undeveloped
inland migration of some of the
wetlands may be possible.
Coastal Wetlands USA Significant loss of
important foraging,
migratory and
wintering bird habitat
at five sites in the
USA.
20-70%
49
2.6°C
50
The effects of two temperature and sea level rise scenarios ( 2°C and 4.7oC
temperature increase above 1990 with a corresponding increase in sea level rise
of 34 cm and 77cm by 2100) on migratory shore bird habitat is estimated by
(Galbraith et al. 2002). The results are complex in that whilst all sites loss
substantially in all scenarios by 2050 and all but one site lose substantially by
2100 for the 2oC scenario, two sites (Bolivar flats and Delaware) gain
significantly by 2100 in the high scenario. Major losses are projected at four
sites - Willapa Bay, Humboldt Bay, San Francisco Bay, and Delaware Bay by
2100 for the 2oC scenario, which could threaten their ability to support current
populations of shorebirds. The worst losses are where existing sea walls
constrain inland migration of the habitats. The 34cm case is for a global
temperature increase of 2oC (above 1990) and is assessed as having a 50%
probability (Titus and Narayanan 1996). The 77cm sea level rise is associated
with a temperature increase of 4.7oC and assessed as having a 5% probability
(Titus and Narayanan 1996).
Most severe losses occur where
coastal topography is steep or where
infrastructure prevents inland
migration of wetlands. The
assessment model accounts for ability
of local sedimentation rates to
preserve intertidal flats in the context
of sea level rise. Local topographic
features including current human
infrastructure are included. In the
Bolivar flats case it is assumed that all
areas above the current extreme high
water mark would be protected by
new infrastructure.
Coral Reefs Global
Annual or almost
annual bleaching by
2040 with negative
implications for coral
reefs and for coral reef
biodiversity, and for
communities
dependent on reef
based resources.
1.7-2.3°C
51
Bleaching
frequency increases with temperature and the crossing of local
bleaching thresholds. Hoegh-Guldberg (1999) investigated the relationship
between seasonal sea surface temperature anomalies and coral reef bleaching
events historically. He found a strong relationship between periods of high
temperature and bleaching events. The temperature thresholds in several reef
locations for bleaching varied by species and location. Using scenarios driven
by IS92a or similar from the ECHAM4 and CSIRO-MkII models and
downscaled to each location he found that the frequency of bleaching is likely to
increase in the future for most reefs. For both models by the 2040s bleaching is
An apparent threshold of seasonal
temperature increases is found to exist
that varies by reef location. When
crossed coral reefs bleach and may
take many years to recover. The
temperature threshold for the same
species varies across its geographic
range, indicating that acclimation has
occurred over the long term.
46 This is the global mean increase in temperature that corresponds to rates of sea level at or above 6mm/yr.
47
Under the scenarios used by Najjar et al. (2000) the sea level rise used here corresponds in time to a global mean temperature increase of around 3.5°C above 1861-1890. Using the HadCM2 model driven by the IPCC
IS92a emissions including sulphur aerosols. The other model used in the Najjar et al. (2000) work, the Canadian Climate Centre (CCC) model, has a warming of about 4°C for the same period (2095).
48 The sea level rise projections, using the IS92a scenario, are drawn from the IPCC SAR WGI Chapter on changes in sea level rise, with a local component of 2mm/yr. See Table 1 of Najjar et al. (2000) for details of the
scenarios used.
49
Range of losses for the 2oC scenario for all sites except Bolivar flats which gains by 1.8%.
50
This is the temperature at the time of the sea level rise assumed in this study. See footnote 41.
41
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
based resources. increase in the future for most reefs. For both models by the 2040s bleaching is
projected to occur (or nearly) annually for all of the reef sites. No account is
taken of changes in El Niño frequency or intensity.
occurred over the long term.
Coral Reefs Indian
Ocean
Risk of local
extinction of coral reef
by 2010-2025 for
many coral reefs in the
Indian Ocean between
10-15°S
0.9-1.4°C
52
Sheppard (2003) estimated bleaching rates based on observed bleaching patterns
from record 1998 bleaching events. Bleaching threshold varies by reef location
for the same species across its range. Local extinction risk is diagnosed when
coral bleaching is estimated to occur every five years. Reefs north and south of
the 10-15°S band have later “extinction dates”. Sheppard notes “the fact that
most sites between 0
o
and 15
o
south will have a 1 in 5 probability annually of
suffering a month as warm as that of 1998 within 10–15 years means that several
of the world’s poorest countries, for which reefs provide essential resources will
be affected soonest”. The HadCM3 scenarios closely match for the 2020s the
scenarios from the HadCM2 model (Hulme et al. 1999a).
The mechanism is very similar to that
described in Hoegh-Guldberg (1999).
Sheppard (2003) defines the
extinction date of coral locally as the
year in which the probability of
bleaching approaches 20%. With
bleaching at frequencies of more than
five yearly, recovery appears unlikely.
The lethal level of temperature during
the warmest month is defined with
respect to those temperatures
observed to be lethal during the 1998
bleaching events. Small increase in
acclimation of the corals would
significantly extent the period before
extinction occurred (raising this to
higher temperatures).
Freshwater
systems
USA –
Prairie Pot
Hole
Region
Major reduction in
waterfowl breeding
population and
wetland extent.
Breeding population
25%
45%
2.5°C
3.6°C
The Prairie Pothole Region is the most important breeding area for waterfowl in
North America (Sorenson et al. 1998). The wetlands appear to be more
sensitive to temperature increase than to precipitation changes. Both of the
climate models used project a major increase in drought conditions. Under the
Hadley model transient scenarios, the drought severity gradually increases from
mild average drought in May in the 2020s, to moderate average drought in the
2050s corresponding to global mean temperature increases of 2.2 and 3.3°C
respectively.
53
Under this model bird breeding numbers are reduced from the
average 5 million in the 1955-1996 period by 25% and around 45%. There is
also projected to be loss of wetland quality, with less open water area preferred
by ducks.
Wetlands are sensitive to an increase
in temperature and summer drought.
Large increases in precipitation would
be necessary to offset the effects of
increased temperature.
Freshwater
systems
USA –
Rocky
Mountains
Large reductions in
habitat for cold water
Salmonid fish
Rocky Mountains.
Habitat changes as a consequence of warming for Salmonid fishes were
estimated for a range of local summer warmings for June, July and August (1-
5°C) (Keleher and Rahel 1996). Suitable habitats were mapped as those with
summer temperature (JJA) less than 22°C, which is known to be suitable for
Salmonid fish species. The corresponding global mean temperature changes are
Increases in stream water temperature
estimated to reduce suitable range.
51
Temperature increase range of the models used for the decade of the 2040s above either 1861-1890 for the ECHAM4 models or 1890-1900 for the CSIRO model. Estimated from the data in the IPCC DDC archived date
set for the CSIRO and ECHAM models forced by IS92a, with and without aerosols.
52 Range of warming from the HadCM3 model used by Sheppard (2003) for the period 2010-2025. See http://ipcc-ddc.cru.uea.ac.uk/cru_data/visualisation/visual_index.html for graphical comparison with other scenarios
and also Table 1 of Hulme et al. (1999a).
53
These temperatures, 2.2°C and 3.3°C are those cited by Sorenson et al. (1998) for the global mean temperature increase for the 2020s and 2050s from the UM Meteorological Office/Hadley model runs cited (Murphy
1995; Murphy and Mitchell 1995). It is assumed here that they are with respect to 1961-1990, although this is not stated in the paper, except in so far as the base period for bird estimates is 1955-1996. More recent HadCM2
and HadCM3 scenarios indicate lower warming levels for these years – 1.5°C and 2.4°C (wrt 1861-1890) respectively (Hulme et al. 1999a).
42
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
17%
50%
72%
Wyoming
14%
43%
0.8-1.0°C
54
1.8-2.4°C
2.7-3.8°C
1.3-1.7°C
2.7-3.8°C
Salmonid fish species. The corresponding global mean temperature changes are
in the range 0.8°C-3.8°C.
54
Salmonid range reductions for this span of
temperature increases were in the range 17% to 72% for the Rocky Mountain
habitats. For the Wyoming habitats the range reductions were smaller - 7 to
43%.
Freshwater
systems
USA-
southern
Appalachian
Mountains
Substantial reduction
in habitat, and a much
smaller reduction in
abundance, of trout
species.
Abundance
10% brook trout
24% rainbow trout
Habitat
80%
1.1-2.3°C
55
Individual model of fish life cycle coupled to GIS database of streams in
southern Appalachian Mountains with scenarios for warming in summer
produced complex pattern of changes. An overall decline in abundance of Brook
trout (10%) and rainbow trout (24%) is projected (Clark et al. 2001). Lower
elevations were projected to experience largest losses. The warming scenarios
applied were for a 1.5-2.5°C increase in summer water temperature above 1977-
1982, which corresponds to a global mean increase of around 1.1-2.3°C above
1861-1890.
55
Warming of freshwater streams
combined with changes in stream
flow reduce suitable habitats, but
abundance changes are linked to a
complex set of negative and positive
effects on the lifecycle of the fish.
Warning water alone increases
abundance.
Biodiversity
Hotspot
South
Africa –
Succulent
Karoo
Very large range
reduction and possible
complete loss of
Succulent Karoo with
likely extinction of
many, if not most, of
the 2500 plants
endemic to the region.
Range Reduction
80%
100%
1.9-2.4°C
56
3.4-4.3°C
57
The Succulent Karoo contains the richest arid flora on the planet and hosts
around 2500 endemic plants. Climate change appears to a first order threat to
these species. Two Climate Models, HadCM2 and CSM model downscaled to
local grid and with bioclimatic model of species at high resolution (Midgley et
al. 2003). (Hannah et al. 2002) estimate that the Succulent Karoo could lose
more than 80% of its range by 2050 with the future bioclimatic region being far
from its present range. A range loss of 80% is likely to lead to very large levels
of extinction in the longer term. The IPCC TAR WGII
reported that Rutherford
et al. (1999a) estimated the complete loss of the Succulent Karoo for a warming
of 3-4°C.
57
Complete loss of range would imply major biodiversity losses.
Projected increase aridification in this
winter rainfall region will reduce the
climatically suitable zone for many of
the endemic species. Land use effects
do not appear to be decisive. It seems
unlikely that the species of this region
would be able to migrate to the
Agulhas plain. This much further to
the south and east and would involve
migration across the Cape Fold
Mountains.
58
54
As before the global mean changes were estimated using SCENGEN for nine recent models downscaled to a broad region covering the Rocky Mountains and Wyoming for the June/July/August period using the SRES
A1B-AIM marker scenario. The model mean for this was 1.73°C JJA increase per degree global mean surface temperature increase, with an inter-model standard deviation of 0.31°C/°C. The latter was used to define the
range for each estimate. Note that using other scenarios produces different scalings, which would increase the range shown here.
55
Estimated using SCENGEN for nine recent models downscaled to a broad region covering the southern Appalachian Mountains for the June/July/August period using the SRES A1B-AIM marker scenario. The model
mean for this was 1.67°C JJA increase per degree global mean surface temperature increase, with an inter-model standard deviation of 0.40°C/°C. The latter was used to define the range for each estimate. The offset from
the 1977-1982 global mean to 1861-1890 was estimated to be 0.37°C using the Folland and Anderson (2002) data set. Note that using other scenarios may produce different scalings, which would increase the range shown
here.
56
As the projected range reduction is based on the HadCM2 scenario for the 2050s, this temperature range is estimated using the HadCM2 range for both sulphate and greenhouse gas only ensemble average in 2050s
accessed from http://ipcc-ddc.cru.uea.ac.uk calculated from model generated increase in 2050s relative to 1961-1990 (1.6-2.1°C) plus the observed increase from 1860s (0.32°C). An alternative approach, which would yield
a larger range is to assume that the range reduction would occur at around the same local temperature increase in any model and to take the range of models or the standard deviation of the inter model estimates for this ratio.
43
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
Biodiversity
Hotspot
South
Africa –
Fynbos or
Cape
Floristic
Province
Very large range
reductions for the
Fynbos biome, which
would threaten many
of its 5600 endemic
species.
Range Reduction
51-61%
1.9-2.4°C
56
The Fynbos is a unique and extremely rich region and forms the smallest of the
six flora kingdoms. It is projected to lose 51-61% of its area for a warming in
the range 1.9-2.4°C. As a consequence about one third of the species suffer
“complete range dislocation” by the 2050s (Midgley et al. 2002).
59
In other
words unless the species can migrate they will become extinct. Around 10% of
the 330 endemic Proteaceae species are projected to suffer complete range loss
.
A high-resolution bioclimatic modeling approach was used driven with three
GCM scenarios (HadCM2) with and without aerosols, and the CSM scenario
without sulphur. HadCM2 produced the lowest global temperature increase
(Midgley et al. 2002). Range dislocation is used as indicator of extinction risk
(Midgley et al. 2003).
Warming moves the suitable
bioclimatic region south and east and
upwards. The effects are expected to
be mitigated by the topographic
complexity of the mountains in the
region which provide more
opportunity for suitable habitat to
remain. For many of the most at risk
Proteaceae species land use change
has less effect than climate change
due to the projected altitudinal shift of
species (Midgley et al. 2003). In the
higher regions over 50% are in
reserves (Rouget et al. 2003).
Mammals and
birds
Mexico Large range losses for
species projected.
90% or more loss
0-45 species
50% or more loss
93-355 species
1.9-2.4°C
56
Large numbers of species appear to be at risk in Mexico. Using an ecological
niche model with three classes of species dispersal abilities the effects of climate
change in Mexico on all of its 1,870 mammal, butterfly and bird species was
estimated for the 2050s (Peterson et al. 2002). The climate scenarios used were
based on the HadCM2 model with two different emissions and corresponded to
global mean warming in the range 1.7-2.4
o
C (above 1861-1890). With range loss
being a powerful predictor of species extinction, these results are quite
concerning for the future of a large number of species in Mexico.
The most serious effects were
projected for the flatlands in the north
of Mexico and the Chihuahuan desert.
This was caused by more drastic
range reductions than in the
mountainous regions (Peterson 2003).
There are different ways to do this. If it is assumed that the range reduction (80% or more) occurs at local temperature increase associated with the HadCM2 range and one takes the standard deviation of the model range
used here, then the minimum global mean temperature at which this would occur is around
57
Baseline is 1961-1990 (Midgley 2003), e.g. 3.3-4.6°C above 1861-1890 average.
58
Rutherford et al. (1999a) and see http://www.nbi.ac.za/climrep/5.htm.
59
See also http://www.nbi.ac.za/climrep/6.htm.
44
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
concerning for the future of a large number of species in Mexico.
Mangroves Bangladesh Losses of forests and
wetlands in
Sundarbans.
Area lost
15% - 10 cm SLR
40% -25cm SLR
75% -45cm SLR
100% -60-100cm SLR
1-1.5°C
60
1.5-2.5°C
2.0-3.5°C
3.0-4°C
The IPCC identified the mangrove forests and wetlands of the Sundarbans as a
unique entity threatened by climate change and sea level rise. Known as the
largest intact mangrove wetland system in the world it is the sole remaining
home of the Royal Bengal tiger (Panthera tigris tigris).
61
A diverse plant flora
grows in the region and the forest is known to support 425 species of wildlife –
49 mammal species, 315 bird species, 53 reptiles and 8 amphibians. Sea level
rise is predicted to result in the progressive loss of the mangrove forest and
wetlands, including habitat of Bengal tiger (Qureshi and Hobbie 1994; Smith et
al. 1998). Estimates of loss for a given level of sea level rise are drawn from the
following sources. An estimate of impacts for a 45cm sea level rise was made in
Chapter 2 of an Asia Development Bank report (Qureshi and Hobbie 1994).
Smith et al. (1998) estimated the loss for 1 metre of sea level rise, which
provided the basis for the estimates made in the IPCC TAR WGII Chapters 11
and 19 (Lal et al. 2001; Smith et al. 2001). Other values are interpolated using
the results from these reports (See also World Bank (2000)).
Freshwater systems and forests would
become inundated, impairing the
growth and reproduction of species
that rely on fresh water. With the
productivity of the system declining,
the closed canopy forests would be
replaced by shrubs and bushes,
leading to loss of species.
Mediterranean
systems
Europe Increased drought risk
is likely to cause
major vegetation
changes.
1.3-3.8°C
62
Recent droughts and associated tree mortality in Spain have indicated that some
tree species that are important to Mediterranean ecology are at present close to
the edge their ability to cope with drought stress (Martinez-Vilalta and Pinol
2002; Ogaya et al. 2003). Projections of the effects of future climate change
indicate a substantially increased risk of tree mortality for some evergreen
species such as the Holm oak (Quercus ilex) due to increased temperature and
extended droughts (Martinez-Vilalta et al. 2002). Holm oak is an important
species to the Mediterranean landscape. Forest currently dominated by it could
be invaded by other species (Pinus latifolia) that are more resistant to drought
and temperature changes. A strong threshold effects is observed in the
modelling, which is supported by observed effects during the severe 1994
drought in the region. If the drought periods extend beyond 3 months there is a
sudden increase in tree mortality.
63
Tree mortality exhibits a strong
dependence in the length of the dry
period or drought rather than
temperature. A strong threshold
effect is observed not far above
present day water stress levels.
60
As pointed out in footnote 41 these estimates are difficult. The range of temperatures here are those corresponding to the global mean surface temperature increase at the time at which the sea level rise is reached under a
range of scenarios taking into account a range of models.
61
It spans about 1 million km
2
, 62% of which is in Bangladesh and the remainder in West Bengal, India.
62
Local summer (June/July/August) temperature increase of 1.5, 3 and 4.5°C relative to 1999-2000 converted to global mean increase using SCENGEN. The scaling factors used were 1.855°C/°C with an inter-model
standard deviation of 0.424°C/°C.
63
The model used treats drying as synonymous with death, however Holm oak is capable of resprouting. However the authors note that Holm oak seems to be very close to it’s water stress limit under present climate
conditions and that trees that are forced to resprout every few years will be at a competitive disadvantage in relation to undamaged trees. See Figure 2 of Martinez-Vilalta et al. (2002) for the threshold response-
45
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
Mediterranean
ecosystems
Europe Increased fire
frequency as a
consequence of
climate change
projected to lead
ecosystem shift to
shrub dominated
landscapes.
1.9-2.4°C
64
Increased drought and water stress predicted for the north western Mediterranean
region are likely to lead to large changes in fire frequency, which in turn is likely
to lead to large changes in vegetation (Mouillot et al. 2002). Using a dynamic
vegetation model SIERRA and climate scenario Mouillot et al. (2002) projected
that there would be increased fire frequency with reduction in return period for
forests from 72 to 62 years and for shrub lands from 20 to 16 years. The
warming scenario was an annual increase locally of about 2.4
o
C relative to 1960-
1997 (the summer period warmed by about 4.0-4.8°C) with precipitation
decreases in the Mediterranean region (Gregory and Mitchell 1995). The
increase fire frequency led to changes in vegetation structure in the model.
Increased frequency of drought
projected under warming scenarios in
the region leads to increase fire
frequency and water stress. This
leads to consequential changes in the
vegetation, shifting it to shrublands
from woodland/forest. Dense forest
and grasslands in particular decline in
favour of low shrub and high shrubs.
65
Montane Cloud
Forests
Hawaii Predicted extinction of
three species of
Hawaiian
Honeycreepers.
2.8-3.2°C
66
Climate change is predicted to act synergistically with past land use changes and
avian malaria risk, to substantially reduce or eliminate the viable habitat of
several Hawaiian honeycreepers (Drepanidae) (Benning et al. 2002). The
honeycreepers live in montane tropical forests, which has been confined to
higher elevations as consequence of past agricultural land clearance. Above this
forest the high elevation area are subject to use for pasture. An introduced
mosquito whose upward range is limited by temperature honeycreeper at lower
altitudes would be subject to attack and mortality from avian malaria. While
past land use change has led to endangerment these pressures are not predicted to
make the species extinct.
Warming of the atmosphere is
predicted to lead to rising cloud base
around the mountains. This would in
principle displace montane forest
upwards, however migration is
limited by the upper elevation land
use. Rising temperatures at the
present elevation range of the
honeycreeper would lead to an
increase risk of contracting avian
malaria.
Plant species Europe Severe risks projected
for biodiversity.
About one third or
more of the species
present in 1990 in
nearly half (44%) of
the European land area
are projected to
disappear from these
areas by 2050 due to
the movement of their
bioclimatic zone.
2.4°C
67
The bioclimatic zones occupied by species are projected to move with climate
change. The IMAGE 2 climate model and the EUROMOVE bioclimatic
envelope model have been used to estimate the changes in biome suitability for
nearly 1400 plant species in Europe. The historical climate envelope for these
species was determined for these species and then the effects of climate change
on their distribution in 2050 projected (Bakkenes et al. 2002). Drier and more
arid regions are found to be the most vulnerable to change – south western
Europe, central European Russia and the Ukraine. Less than 50% of current
species are projected to remain in situ in Spain, southwestern France, the Black
Sea coast and Byelorussia. The lowlands of Germany, Belgium and The
Netherlands are likely to keep 70-80% of their species, however some
endangered species may disappear.
Warming and other climate changes
will lead to the movement of suitable
bioclimatic zones for many species.
The EUROMOVE models establish
the bioclimatic envelope for the
species studied and then estimate how
this will change after climate change.
The actual migration of species in
order to tracking the movement of
their bioclimatic zones is uncertain for
a number of reasons. It is by no
means clear that all of the species
whose bioclimatic zones move away
from current locations will be able to
re-establish successfully.
64
A 2.4°C local warming with respect to 1960-1997 is upscaled to a global mean estimate with SCENGEN using scaling factors of 1.373/°C with an inter-model standard deviation of 0.199°C/°C for the northwest
Mediterranean region. The period 1960-1997 is about 0.38°C warmer than 1861-1890. Whilst the study uses the climate scenario of the UKTR model (and earlier Hadley transient scenario) upscaling the local warming (as
this is given in the paper) to the global mean gives a better idea of the uncertainty range involved.
65 For the scenario (S2) involving changes in both rainfall and temperature. See Figure 6 of Mouillot et al. (2002).
66
A 2°C local warming upscaled to a global mean estimate with SCENGEN using scaling factors of 0.851°C/°C with an inter-model standard deviation of 0.073°C/°C for the equatorial Pacific region assuming the local
increase is with respect to 1990.
46
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
Protected areas Africa Major adverse
consequences
predicted for Malawi’s
Lengwe National Park
(and for Malawi in
general). It is unlikely
that it would be able to
support large
populations of
ungulates if climate
change produces more
drought conditions
and consequent
degradation of habitat.
2.9-3.4°C
68
Projected climate changes in the Malawi region are estimated to have adverse
effects on wildlife (Mkanda 1996, 1999). Recent drought periods were used as
an analogue to future projected changes by comparing the droughts of 1979/80
and 1991/92 with several GCM projections. Mkanda (1996; 1999) found that
there was little difference between the projected effects of the different GCM
scenarios for a doubled CO
2
climate. The increased evapotranspiration caused
by higher temperatures outweighed the benefits of increased precipitation in
these scenarios. Consequently it is predicted that land and vegetation quality is
likely to be much degraded by climate changes in the future, with the possibility
of a “vicious cycle developing” with poor ground caused by climate change
driving further soil degradation and reduced habitat quality.
Increase temperatures lead to more
frequent drought conditions in this
region. Increased precipitation
projected in some scenarios does not
appear sufficient to compensate for
increased evapotranspiration.
Protected areas North
America
High altitude plant
species in
Yellowstone national
park may not cope
with climate change
arising from a
doubling of CO
2
concentration.
4-8°C
69
Complex changes are projected for the vegetation of the Yellowstone national
park as a consequence of projected climate change. The range of high elevation
species is reduced and some species disappear from the region
(Bartlein et al. 1997). Bartlein et al. (1997) argue that the rates of change
projected may exceed the ability of species to migrate as rates of change exceed
those evident from the paleorecord. An early generation GCM from the
Canadian Climate Centre (CCC) was used for this scenario. A global mean
warming of 3.5°C was estimated for doubling of CO2 concentrations (Boer et al.
1992). When downscaled to the Yellowstone region this produced a warming of
about 10°C in January and July. Such warming for this level of global mean
change in this area are not generally found in the most recent generations of
coupled ocean atmosphere GCMs. As these seasonal warming levels were used
to drive the assessment of vegetation effects, the global mean estimates here are
upscaled using the recent generation of AOGCMs. It should also be noted that
there is most of the current generation of models project a decrease in summer
rainfall (model average -9%/°C global warming) in this region whereas the CCC
model used had little change. Such a reduction would exacerbate many of the
effects cited by Bartlein et al. (1997).
Warming and increased summer
drought stress, with consequent
increase in fire frequency, lead to
substantial changes in vegetation.
The later generation of climate
models predict a reduction in summer
rainfall on average, which would
exacerbate the problems identified.
67
The scenario used was computed with the IMAGE 2.0 driven by the IPCC IS92a scenario, which generated a global increase of 1.8°C above 1990 by 2050. As before the 1990 climate is assumed to be about 0.6°C
warmer than the 1861-1890 period.
68
The global mean warming was calculated using SCENGEN. The local temperature increase scenarios used by Mkanda (1999) of 3.1-3.8°C to the global level upscaled using the scaling factors 1.123°C/°C with an inter-
model standard deviation of 0.25°C. Whilst Mkanda used early generation GCM scenarios a check against the projections from the current generation of models indicates that these are not inconsistent. Combined with his
findings that the scenarios he used produced relatively robust results (increased temperatures tended to outweigh the effects of increased rainfall projected from two of the three GCMs he used.
69
This is the range of global mean temperature increases upscaled using SCENGEN from a warming of 10°C in January and in July in this region, as used in the Bartlein et al. (1997) analysis. The CCC model used as the
basis for the scenario of Bartlein et al. (1997) has a global mean warming of 3.5°C, however more recent generation AOGCMs do not produce such pronounced warming in this region.
47
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
Protected areas South
Africa
Loss of two thirds of
animal species studied
in Kruger National
Park.
1.9-3.1°C
70
The bioclimatic range of many animals presently within Kruger National Park
are projected to move outside of the park boundaries (Erasmus et al. 2002).
Migration of these animals in order to track the range shift may be problematic
due to land use pressures in the regions adjacent to the park (Erasmus et al.
2002).
In general there an eastward shift in
ranges are projected with warming.
Large movements of the bioclimatic
zones of many animal species are
projected to occur. Extensive range
shifts are also projected for plant
species in the region (Rutherford et al.
1999b). Substantial and growing land
use and population pressures are very
likely to cause major problems for
migration of animals tracking climate
induced movement of their ranges
(Erasmus et al. 2002).
Protected areas Switzerland Many protected areas
would no longer be
suitable for a large
numbers of their
present forest species.
Proportion of reserves
not suitable for present
forest species
40-50%
70-80%
0.9-1.5°C
71
1.4-2.8°C
72
Based on lapse rate considerations, Kienast et al. (1998), assess the effects of
increasing temperatures on mountain forest communities.
73
Twenty-nine out of
109 reserves have enough altitudinal range to survive a 500-metre change in
effective climatic zone (a 2-2.8
o
C increase) and 12 areas have enough range to
survive a 250 metres change (1-1.4
o
C increase). However 50 reserves (46%)
cannot take a 250m gain and 18 areas have only enough altitude to survive a
250-metre range change. Calculations using degree-days yield similar results.
Authors point to many limitations of the study including no dynamical
assessment of changes, no account taken of land use changes etc.
74
Climatic warming will lead to upward
altitudinal movement of bioclimatic
zones.
70
A 2-3°C temperature increase in South Africa with respect to 1960-1989 is converted to a global mean with respect to 1861-1890 using average of nine recent GCMs downscaled to the European Alpine region using
SCENGEN. The regional to global scaling used is 1.191°C/°C with the range set by the inter-model standard deviation of 0.114°C/°C. The scaling factors using all 17 models in SCENGEN are not very different from the 9
model estimate. Within the paper the climate scenario is not detailed and references are made to it warming South Africa by 2°C and by 2.5-3°C. If 2°C then the global warming range is 1.9-2.2°C and 2.5-3°C then the
range is 2.2-3.1°C. The full range is included here.
71
A 1-1.4°C local temperature increase with respect to 1931-1970 converted to a global mean with respect to 1861-1890 using average of nine recent GCMs downscaled to the European Alpine region using SCENGEN.
The regional to global scaling used is 1.39°C/°C with the range set by the inter-model standard deviation of 0.30°C/°C. The scaling factors using all 17 models in SCENGEN are not very different from the 9 model estimate.
The base period is 1931-1970, which is about 0.26-0.28°C warmer than 1861-1890.
72
A 2-2.8°C local increase above 1931-1970 converted to global mean as in footnote 71.
73
Authors use an adiabatic lapse rate of 0.55°C/100m.
74
The authors also used a spatially explicit forest simulator with four climate scenarios: moderate or strong temperature increases and current levels or a 15% increase in precipitation. The model area includes not only the
reserves but also the entire Swiss forest inventory. For temperature increases only the model supports vegetation shifts along altitudinal lines, however with warmer and wetter conditions, model results indicate that
vegetation shifts may not be as ‘dramatic’. The model did not consider the effects of CO
2
fertilization. For strengths and weakness of bioclimatic envelope models see Footnote 3.
48
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
Species
conservation
Australia Large range reduction
in range of butterfly
species
>20% range decline
54% of species
>50% range decline
83% of species
1.2°C
75
2.9°C
76
Large range reductions are projected for many butterfly species in Australia as a
consequence of projected climate change. Using the bioclimatic envelope
model BIOCLIM the range changes for 24 species of butterflies were examined
for a range of four climate change scenarios for the 2050s (Beaumont and
Hughes 2002). The climate scenarios used were based on the results of seven
climate models running under the IPCC SRES scenarios, with regional estimates
over Australia (Hulme and Sheard 1999) – see footnotes 75 and 76. Changes in
species distribution were estimated using the temperature and precipitation
changes for grid cells over Australia. One of the main findings is that even
species with wide climatic ranges could be very vulnerable to climate change.
The proportion of species suffering large range contractions increases rapidly
with temperature. The larger the warming the smaller is the proportion of a
species current range that lies within the projected future range in the 2050s. For
a small increase of global mean temperature of 1.2°C
75
this proportion is 66%,
whereas for a larger global mean warming of 2.9°C
76
, this falls to less than 22%.
Changes to temperature and
precipitation result in geographic
shifts in the suitable bioclimatic zones
for butterfly species. The model
projects the change in bioclimatic
range from the 1961-1990 period to
the 2050s. The ability of species to
track these geographic changes is not
modelled. It is known that many of
the Australian butterfly species have
limited dispersal ability or cannot
migrate (Beaumont and Hughes
2002). Land clearance and habitat
fragmentation appears likely to
present barriers to dispersal and
migration.
Species
conservation
South
Africa
Predicted local
extinction of four
animal species and
large range reductions
of greater than 50%
for 29 endangered
species. 140 species
(78%) projected to
experience various
levels (4-98%) of
range contraction.
1.9-3.1°C
70
“Profound impacts” are projected for many animal species in South Africa from
climate change (Erasmus et al. 2002). A bioclimatic envelope model approach
was used to study the response of 179 animal species – 34 birds, 19 mammals,
50 reptiles, 19 butterfly and 57 other invertebrates in South Africa – under a
scenario involving a 2-3°C increase above 1960-1989 mean temperature (as well
as precipitation changes) (Erasmus et al. 2002). There were four projected local
extinctions (see Table 2 of Erasmus et al. (2002)). The vast majority of species
are projected to experience range reductions of the order of 4-98%. As a
consequence of land use pressures and habitat fragmentation the ability of
animals to track climate change by moving their range is open to doubt.
Erasmus et al. (2002) point out that “theoretical range shifts into transformed
landscapes may mean local extinction”. The range reductions projected are
conservative and appear likely to underestimate the overall reductions, as
landscape transformation is not accounted for in the model.
Large range shifts are predicted as a
consequence of climate change,
mostly in an easterly direction across
the region. Fragmentation of the
landscape in the region as a
consequence of human activities
means the projected range shifts may
not be realized in practice. Range
reductions projected are likely to
underestimate the actual overall loss
of range for the same reasons.
Reductions in range size are likely to
increase the risk of local extinction.
Species
conservation
Australia Dramatic range
reduction or
disappearance of
frogs, and endangered
mammals and plants
from Dryandra forest
ecosystem in
southwestern
Australia.
“Dramatic d
ecreases in range” (IPCC TAR WGII 12.4.2
(Pittock
et al.
2001)
)
are projected for most species studied in the Dryandra forest ecosystem in
southwestern Australia for quite small warming levels. Effects include the
disappearance of frogs, and endangered mammals and plants (Pouliquen-Young
and Newman 1999). A bioclimatic envelope model was used to estimate the
effects of temperature and rainfall changes using a regional climate model at
125km resolution. The forested studied is part of a larger systems in south
western Australia that has been identified as one of 25 global biodiversity hot
spots (Myers et al. 2000). Three species of frogs, 15 species of endangered or
threatened mammals, 92 vari
eties of the plant genus Dryandra, and 27 varieties
of Acacia were modelled. For 0.5°C warming above 1990 all frogs and mammal
Bioclimatic envelope is estimated
empirically and then climate change
scenario superimposed.
78
Unsuitable
soils and land use patterns several
limit migration potential.
75
The scenario used is the B1-low of Hulme and Sheard (1999), which produces warming over Australia in the range 0.8-1.4°C warming wrt to 1961-1990. This scenario has a global mean warming for the 2050s of 0.9°C
wrt 1961-1990 or 1.2°C wrt the 1861-1890 base period.
76
The scenario used is the A2-high of Hulme and Sheard (1999), which produces warming over Australia in the range 2.1-3.9°C warming wrt to 1961-1990. This scenario has a global mean warming for the 2050s of 2.6°C
wrt 1961-1990 or 2.9°C wrt the 1861-1890 base period.
49
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
All frogs and mammal
species studied would
be restricted to small
areas or disappear
Two thirds of
Drvandra tree species
and all of the Acacia
species projected to
disappear from the
region.
1.1°C
77
2.6°C
of Acacia were modelled. For 0.5°C warming above 1990 all frogs and mammal
species studied would be restricted to small areas or disappear. Under warming
of 2°C two thirds of Dryandra tree species and all of the Acacia species modelled
would disappear from the region.
Temperate
forests and
woodlands
Australia Australian eucalypt
species outside current
thermal range.
25%
40%
>50%
1.1-1.3°C
79
1.9-2.2°C
80
2.7-3.2°C
81
Under the assumed warming scenarios the present bioclimatic zones of eucalypt
species move significantly. As a result “within the next few decades many
eucalypt species will have their entire present day populations exposed to
temperatures and rainfalls under which no individuals currently exist” (Hughes
et al. 1996). Using a bioclimatic model (Hughes et al. 1996) find that of the 819
species of Eucalyptus examined for their climatic range (mean annual
temperature and rainfall), 53% have ranges spanning less than 3°C , 41% having
a range of less than 2°C , and 25% have a range of less than 1°C . In relation to
rainfall, 23% have ranges spanning less than 20% of the variation in mean
annual rainfall. Although actual climatic tolerances of many species are wider
than the climatic envelope they currently occupy, substantial changes in the tree
flora of Australia may be expected (Hughes et al. 1996).
The present distribution of species is
mapped against
Empirical bioclimatic estimates of
species range for temperature rainfall
and other factors with superimposed
temperature and rainfall scenarios.82
Migration of species is not modelled.
Temperate
forests and
woodlands
New
Zealand
Risk of extinction of
New Zealand kauri
tree.
4.8-7.5°C
83
Empirical, isolation and subsequent extinction feared (Mitchell and Williams
1996). A risk of extinction is identified in the TAR: “For example, Mitchell
and Williams (1996) have noted that habitat that is climatically suitable for the
long-lived New Zealand kauri tree (Agathis australis) under a 4°C warming
scenario would be at least 150 km from the nearest extant population. They
suggest that survival of this species may require human intervention and
relocation.”
Warming causes bioclimatic zone of
kauri to move away from existing
locations.
77
Adjusted to 1861-1890 from 0.5°C above 1990.
78
For strengths and weakness of bioclimatic envelope models see Footnote 3.
79
This the global mean temperature range corresponding to a warming over Australia of 1°C upscaled using SCENGEN. The scaling factors used of 1.161°C/°C (with a standard deviation of 0.121°C/°C) is the average of 9
recent AOGCMs computed choosing SCENGEN cells minimizing the area of oceans surrounding Australia as the impact being examined is for land surface. Although it is not clear what the base period is for the climate
data an extensive data resource was used by the authors to map current eucalypt distributions against temperature and precipitation. In this context a conservative assumption is to use the 1961-1990 reference period.
80
As for footnote 79 but for a 2°C local warming.
81
As for footnote 79 but for a 3°C local warming
82
For strengths and weakness of bioclimatic envelope models see Footnote 3.
83
Assuming the 4°C local increase is with respect to 1990 and using SCENGEN as described above to obtain a local to global scaling for the South Island of New Zealand of 0.769°C/°C and an inter-model standard
deviation of 0.193°C/°C.
50
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
Tropical Forests Australia Loss of 50% of the
highland rainforest
habitat in the World
Heritage listed tropical
rain forests in North
Queensland. These
highlands host most of
the more than 60
endemic vertebrates of
this region.
1.6-1.8°C
84
The tropical rainforests of North Queensland, Australia supports 566 species to
terrestrial vertebrates or 28% of the Australian total. Sixty-five are regional
endemics, most of which are hosted by the highland tropical forests within the
region. Using a neural network bioclimatic model to project the effects of
changes in precipitation and climate it has been found that large reductions in
highland rainforest is likely in the wet tropics of North Queensland, Australia
(Hilbert et al. 2001). Lowland mesophyll vine forest is projected to increase in
areas but upland complex notophyll vine forest response depends on
precipitation. Highland rainforest (simple notophyll and simple mesophyll vine
fern forests and thickets) decreases for all rainfall scenarios for a 1°C increase in
temperature. This habitat hosts many of the endemic vertebrates of the region
and severe, adverse consequences have been predicted for many of these (see
below).
Warming causes rise in bioclimatic
zone. No assessment is made of
effects of elevated CO
2
.
Tropical Forests Australia “Predicted extinction”
of Golden Bower bird
and other species
similarly confined to
upland and highland
areas of the wet
tropical forests of
North Queensland.
Range loss
>60%
90%
98%
1.6-1.8°C
84
2.6-3.0°C
85
3.6-4.2°C
86
Climate change is predicted to lead to the extinction of the Golden Bower bird
which is confined to upland and highland areas (Hilbert et al. 2003). Range
losses for this species are projected to be approximately 90% with a 2°C
warming and a 10% decrease in rainfall. This scenario is consistent with recent
model estimates of climate change for the region (Walsh and Ryan 2000).
Whilst the overall change in rainfall is uncertain, it is likely that there will be an
increase in dry season severity and variability in rainfall (Walsh and Ryan 2000).
.
The bioclimatic zone of the Golden
Bower bird is extirpated with
increasing temperature.
Tropical Forests Australia “Catastrophic” loss of
endemic rainforest
vertebrates projected.
>2.6-3.0°C
85
Severe loss of rainforest vertebrate species is projected in the highland tropical
forests of the region as consequence of warming. Williams et al. (2003)
examined a wide range of species and found a risk of catastrophic loss of the
endemic vertebrates of the forest above 300 metres:
“Extinction rates caused by the complete loss of core environments are likely to
be severe, nonlinear, with losses increasing rapidly beyond an increase of 2 °C,
and compounded by other climate-related impacts”.
Most of the rainforest in the region is confined to above 300 metres altitude.
Mountains in the region are no higher than about 1600 metres. Of the 600
vertebrate species in the region 83 are endemic, 72 of these are restricted to the
rainforest and 62 of these confined to the montane forests above 600 metres
altitude.
Verterbrates confined to high altitude
zones are projected to run out of
suitable habitat with increasing
temperature.
Williams et al. (2003) argue that the
results for the wet tropics of Australia
have broad implications for montane
and higland tropical forests. These
often are very diverse with large
numbers of endemic species and may
be “severely threatened by c
limate
change”
(Williams
et al.
2003)
.
84
Assuming the 1°C local increase is with respect to 1990 and using SCENGEN as described above to obtain a local to global scaling for the Wet Tropics area of 0.917°C/°C and with an inter-model standard deviation of
0.072°C/°C.
85
Assuming the 2°C local increase is with respect to 1990 and applying the same scaling factors as in footnote 84.
86
Assuming the 3°C local increase is with respect to 1990 and applying the same scaling factors as in footnote 84.
51
Ecosystem Region Impact
Global mean surface
temperature above pre-
industrial
[°C]
6
Comments
Causal Chain
altitude. change” (Williams et al. 2003).
Tropical Forests Amazon Risk of collapse of
tropical forests in the
Amazon.
1.4-2.4°C
87
Several studies have identified a risk of a climate change induced collapse of the
Amazon rainforests. A firm probability statement cannot yet be made as to the
likelihood of this coming about, however the seriousness of the risk and is large
consequences mean that this needs to be taken seriously.
88
Based in part on a
climate scenario driven by the HadCM3L model Cowling et al. (2003) argue that
the feedbacks that maintained the stability of the Amazon in the past glacial and
interglacial climates
i
cannot be maintained in the future and that there is likely to
be a positive feedback effect which amplifies local drying and warming. As a
consequence, Cowling et al. (2003) argue that there is a threshold “at which
tropical ecosystems exceed their capacity for internal/external feedback effects
compensating of the deleterious effects of warming on tropical plants,” but that
locating this is very difficult. They speculate that the climate system,
temperature, is very close to this threshold at present. Jones et al. (2003) report
on the estimated carbon cycle feedback effects of climatic warming, updating the
earlier work of Cox et al. (2000). An abrupt increase in the land source of CO2
as a consequence of warming and the pattern of climate change in the scenario
occurs, reaching 7GtC/yr in 2001, principally from loss of soil carbon and
Amazon tropical forest dieback. Apart from the drastic biodiversity loss
implications such a feedback would amplify the warming considerably. See Note
1 at the end of this table for a further brief discussion on the Amazon and climate
change issue.
One of the critical mechanisms is the
effect of vegetation feedbacks on
regional climate. Anthropogenic
climate change leads to higher
temperatures and increased
respiration, which leads to a
breakdown of water recycling within
the Amazon basin. As rainfall
declines this contributes to further
vegetation dieback. In addition to the
mechanism identified by Cowling et
al. (in press), it seems likely that the
habitat fragmentation-climate-forest
fire feedback identified by Laurance
and Williamson (2001) will act to
exacerbate any purely climate change
induced propensity of vegetation loss.
Tropical Forests Amazon Projected “dramatic”
loss of species
viability in eastern
Amazonia with
refugial areas
remaining in the
western zone of the
Amazon basin
1.5-2.8°C
89
The Amazon basin and its rainforests host a substantial fraction of the world’s
biodiversity. Climate change is projected to lead to loss of species in parts of the
Amazon (Miles 2002; Miles et al. 2003). Under a “standard” scenario
90
with
warming by the 2080s of 2.5°C wrt 1990 29% of species are projected to have
“no viable distribution”. Under a “reduced impact scenario”, with warming of
1.2°C by this time, 13% had no viable distribution projected. Dispersal or
migration in many of these cases would have to occur over hundreds of
kilometres for species to reach appropriate new bioclimatic zones.
The effects of climate change were
estimated using bioclimatic modelling
of plant species in the Amazon. This
took account of tolerance of plants to
extreme climate values, barriers to
migration and dispersal, and lags in
species response to climate change.
Note [1]: A major caveat on these results is that they are based on the HadCM3 climatic projections for the Amazon region and the TRIFFID
vegetation/terrestrial carbon cycle model. The main driver of the collapse is the increasing El Nino like warming pattern for sea surface
87
This estimate of when an instability threshold may be approached in the Amazon is highly uncertain and most likely model dependent. The range chosen here is global mean warming for the HadCM3 model forced by the
IS92a emissions scenarios for the period 2020s and 2050s (Hulme et al. 1999a). These time periods are chosen as the earliest period in which significant changes can be seen in Amazon rainforest cover and the time at
which a reduction of around 20-25% has occurred in the modeling by Cox et al. (2003). See their Figure 6.
88
Cox et al. (2003) conclude that whilst the mechanisms that could lead to a dieback of the Amazon are qualitatively understood “we are still a long way from being able to estimate the probability of such an ecological
catastrophe occurring.”
89
Scenarios used warm globally be between 1.2 and 2.5°C by 2095.
90
Miles used the HadCM2 model for the assessment. It produces results within the range simulated by both the ECHAM4 and CSIRO MkII models for the Amazon region. The two main scenarios were a) Standard Impact
based on the IPCC IS92a emission scenario. The standard scenario has a 2080s global mean temperature increase of ca 2.5°C w.r.t 1961-1990 and b) a Reduced Impact scenario of half this increase. Both are downscaled to
the region.
52
temperatures projected by the HadCM3 model as greenhouse gases increase (Cox et al. 2003). Whilst some models show this as well it is not a
universal feature. Nevertheless, the HadCM3 model has one of the best associations between current observed and modelled patterns of climate in
this region. It was shown by Cramer et al. (2001) in a comparison of six dynamic vegetation models, and more recently in a review of the carbon
cycle implications of projected climate (Cramer et al. 2003), that the vegetation feedbacks are model dependent. In particular the climate model and
TRIFFID produce larger climate changes and larger, and qualitatively different vegetation responses than other models.
In spite of these uncertainties are several reasons for inclusion of this example. The main mechanism as described in the last column is likely to be
driven, in addition to climate change, by land clearance in the Amazon. There is now a well established link between forest fires, habitat
fragmentation and climate changes and extreme ENSO events in the Amazon (Laurance and Williamson 2001; Laurance et al. 2001; Nepstad et al.
2001; Cochrane and Laurance 2002). In other words there is likely to be a synergistic effect between forest fragmentation and deforestation and
human induced climate change, should the latter lead to more ENSO like climatic conditions in the region. None of the models so far include these
combined effects. Inclusion of such effects and processes would likely reduce the resilience of vegetation to warming and drying. Secondly, it
seems likely that future climate change will produce more El Niño like conditions. It is known that there are substantial releases of carbon from the
Amazon during ENSO years (Tian et al. 1998), which also occurs for the global biota (Jones et al. 2001). It is clear from the work of Tian et al.
(1998) that the Amazon forest can switch from a sink to a source quite. Thirdly, smoke from biomass burning can inhibit rainfall over the Amazon
(Rosenfeld 1999) implying a further and so far unmodelled feedback which would excarbate any tendency to drying and increase fire frequency.
Fourthly, it is sometimes argued that the Amazon forest was substantially reduced in area during the last glacial and expanded with more equable
climates in the early Holocene and hence a global warming reduced reduction would not be much different from what may have happened in the
past. Based on a detailed analysis of available paleorecords Colinvaux et al. (2000) conclude that the Amazon forests retained their integrity
throughout the last glacial period. This is supported by the modelling of Cowling et al. (2001; Cowling et al. 2003) with the indication that there
feedback effects that help the forest cope with cooler and warmer periods. However recent work Cowling et al. (2003) indicates that these feedback
processes could be overwhelmed by the climate changes projected by the models used. Finally, as a risk assessment exercise, the results of Cox et
al. (2000), Cowling et al. (2003) and Cox et al. (2003), present a prima facie risk, that has yet to be eliminated by definitive modelling or other
assessments.
53
3. Impacts on Food Production, Water, and Socio-economic
systems
Introduction
Article 2 is quite specific in relation to the general need to “ensure that food
production is not threatened”. However, it does not make mention of whether this
should be the case regionally as well as globally. As will be seen from the
information presented below this would have a substantial bearing on an
interpretation of Article 2. Whilst current assessments indicate that global aggregate
agricultural production may not be adversely affected up 2-3
o
C warming, this is not
the case for a number of regions. Indeed, the questions of who will be adversely
affected by climate change and who will make the “cruel choices”
91
between the costs
of mitigation and the damages to be borne by climate change are amongst the key
political issues involved in the resolution of the questions embodied in Article 2 and
its implementation.
The other part of Article 2, dealt with here, relates to the need for policy action to be
taken “within a timeframe sufficient to enable economic development to proceed in a
sustainable manner”. Differing interpretations of this requirement have been made,
with a dominant ‘economic’ one relating to the concept that if abatement action is to
be rapid then economic growth would be reduced and resources diverted from other
sustainable development needs. Another interpretation is that rapid climate change
itself may threaten sustainable development in some regions and some sectors. The
focus here is on information relevant to the latter question.
In general, the results presented below account for adaptation possibilities in each of
the sectors considered. Only in cases where there are identified limitations to
adaptation or where adaptation options have not been included, specific reference will
be made to this issue.
With the emphasis in this report on brevity and on salience to a consideration of
impacts at different temperature levels, the information below will be drawn largely
from only a few studies of global impacts and effects based on a few GCMs. Whilst
every attempt will be made to place these in the context of general findings or
qualifications made in the IPCC TAR, the reader should be aware that the overall
literature is rich, complex and sometimes divergent. In general, all of the literature
chosen for use here is consistent with the broad findings of the TAR and where it is
not, the reasons for this are specifically addressed. Space limitations militate against
explanation of processes leading to impacts and effects and hence the discussion will
focus on results only except to the extent necessary for clarity of exposition. The
reader is referred to the underlying literature for an understanding of the processes
mentioned below.
91
Phrase used by a key negotiator from a very large industrial country to describe the process of
deciding upon the ultimate limits to climate change and the trade-off between the economic and
political costs of emission abatement and climate protection targets.
54
Context: Findings of the Second and Third Assessment Reports
Projected climate change effects on the sectors considered here were examined in
detail in the Second Assessment Report (SAR) of the IPCC (1995) and in the IPCC
Regional Impacts report of 1997. The IPCC TAR in many cases explicitly reviewed
the findings of these reports in its assessment of the literature. With few exceptions,
the TAR confirms the general findings of the earlier assessments, however
quantitative assessments have often changed.
92
Table 6 cross compares broad areas
of the findings from the Second and Third Assessment Reports.
93
It gives a clear
picture of the consistency between the findings of the 1995 and 2001 assessments.
Consistency of these assessments, based as they are on quite different literature and
different models at different stages of development, appears to add confidence to the
overall findings of the IPCC TAR.
One of the main conclusions of the IPCC TAR, which strengthens the earlier SAR
assessment, is in relation to the vulnerability of developing countries at low levels of
warming (less than 2
o
C). It is likely that global increases in temperature would
produce net economic losses in many developing countries for all magnitudes of
warming and these losses would be greater the higher the warming. This conclusion
is reflected in each of the sectors discussed below, where many developing countries
are seen to have large projected damages at low levels of warming, even though
global aggregate market impacts are estimated as small or positive, up to a few
o
C
warming. This is particularly true for agriculture and water resources where it is clear
that some regions are particularly vulnerable to the effects of climate change.
92
It is important to bear in mind that more recent impact assessments have used transient scenarios
generated with coupled Atmosphere Ocean General Circulation Models (AOGCMs) rather than, as
was the case in the SAR, doubled CO
2
equilibrium scenarios run with Atmosphere General
Circulation Models (AGCMs) with stylised (slab or mixed layer oceans). In general, the transient
scenarios produce less extreme results at specific time periods in the future than the equilibrium
scenarios.
93
Space does not permit doing this for the Regional Impacts report, which contains much additional
information, however the most salient findings of this report are repeated, in one form or another, in
the IPCC Working Group II TAR Report.
55
Table 6 - Comparison of Second and Third Assessment Report Findings
Category IPCC Second Assessment Report
94
IPCC Third Assessment Report
95
Vulnerability “People who live on arid or semi-arid lands, in low-
lying coastal areas, in water-limited or flood-prone
areas, or on small islands are particularly vulnerable to
climate change” (p. 29).
"The effects of climate change are expected to be greatest in developing countries in terms of
loss of life and relative effects on investment and the economy. For example, the relative
percentage damages to GDP from climate extremes have been substantially greater in
developing countries than in developed countries” (WGII-SPM p. 8).
“The projected distribution of economic impacts is such that it would increase the disparity in
well-being between developed countries and developing countries, with disparity growing for
higher projected temperature increases (medium confidence)” (WGII-SPM p. 8).
Africa: “Increases in droughts, floods, and other extreme events would add to stresses on
water resources, food security, human health, and infrastructures, and would constrain
development in Africa (high confidence).” “Significant extinctions of plant and animal species
are projected and would impact rural livelihoods, tourism, and genetic resources (medium
confidence)” (WGII-SPM p. 14).
Asia: “Extreme events have increased in temperate and tropical Asia, including floods,
droughts, forest fires, and tropical cyclones (high confidence).” “Sea level rise and an
increase in intensity of tropical cyclones would displace tens of millions of people in low-
lying coastal areas of temperate and tropical Asia; increased intensity of rainfall would
increase flood risks in temperate and tropical Asia (high confidence)” (WGII-SPM p. 14).
Latin America: “Loss and retreat of glaciers would adversely impact runoff and water supply
in areas where glacier melt is an important water source (high confidence).” “Floods and
droughts would become more frequent with floods increasing sediment loads and degrade
water supply in some areas (high confidence).” “Increases in intensity of tropical cyclones
would alter the risks to life, property, and ecosystems from heavy rain, flooding, storm surges,
and wind damages” (WGII-SPM p. 15).
94
Conclusions cited are from the Summary for Policy Makers of Working Group II of the Second Assessment Report adopted in Montreal, October 1995 unless otherwise
stated. Where reference is made to WGII Technical Summary or to sections of the report it should be noted that these have not been approved in detail by governments.
95
Conclusions are from Summary for Policy Makers of the Working Group II Report (WGII-SPM) and the Synthesis Report (SR-SPM) of the Third Assessment Report
adopted at Geneva in February 2001 unless otherwise stated. Where reference is made to the full Synthesis Report or other sections of the IPCC TAR it should be noted that
these have not been approved in detail by governments. The confidence levels are those assigned by IPCC WGII under its scale of uncertainties. See footnote 99 d.
56
Small Islands: “Populations that inhabit small islands and/or low-lying coastal areas are at
particular risk of severe social and economic effects from sea-level rise and storm surges.
Many human settlements will face increased risk of coastal flooding and erosion, and tens of
millions of people living in deltas, in low-lying coastal areas, and on small islands will face
risk of displacement. Resources critical to island and coastal populations such as beaches,
freshwater, fisheries, coral reefs and atolls, and wildlife habitat would also be at risk” (SR–
SPM p. 12).
Health “Climate change is likely to have wide ranging and
mostly adverse effect on human health, with significant
loss of life” (p. 35).
“Indirect effects of climate change include increases in
the potential transmission of vector-borne infectious
diseases (e.g., malaria, dengue, yellow fever, and some
viral encephalitis) resulting from extensions of the
geographical range and season for vector organisms”
(pp. 34-35).
“Projections ... indicate that the geographical zone of
potential malaria transmission in response to world
temperature increases at the upper part of the
IPCC-projected range (3-5
o
C by 2100) would increase
from approximately 45% of the world population to
approximately 60% by the latter half of the next
century. This could lead to potential increases in
malaria incidence (on the order of 50-80 million
additional annual cases, relative to an assumed global
background total of 500 million cases), primarily in
tropical, subtropical, and less well-protected
temperate-zone populations” (p. 36).
“Overall, climate change is projected to increase threats to human health, particularly in lower
income populations, predominantly within tropical/subtropical countries” (SR–SPM p. 12).
“Climate change can affect human health directly (e.g., reduced cold stress in temperate
countries but increased heat stress, loss of life in floods and storms) and indirectly through
changes in the ranges of disease vectors (e.g., mosquitoes), water-borne pathogens, water
quality, air quality, and food availability and quality (medium to high confidence)” (SR–SPM
p. 12).
“Climate-related health effects are observed. Many vector-, food-, and water-borne infectious
diseases are known to be sensitive to changes in climatic conditions. Extensive experience
makes clear that any increase in floods will increase the risk of drowning, diarrheal and
respiratory diseases, water-contamination diseases, and—in developing countries—hunger
and malnutrition (high confidence)” (Synthesis Report p. 56).
“Heat waves in Europe and North America are associated with a significant increase in urban
mortality, but warmer wintertime temperatures also result in reduced wintertime mortality. In
some cases health effects are clearly related to recent climate changes, such as in Sweden
where tick-borne encephalitis incidence increased after milder winters and moved northward
following the increased frequency of milder winters over the years 1980 to 1994” (Synthesis
Report p. 56-57).
Latin America: “The geographical distribution of vector-borne infectious diseases would
expand poleward and to higher elevations, and exposures to diseases such as malaria, dengue
fever, and cholera will increase (medium confidence)” (WGII-SPM p. 15).
Africa: “Extension of ranges of infectious disease vectors would adversely affect human
health in Africa (medium confidence)” (WGII-SPM p. 14).
57
Asia: “Human health would be threatened by possible increased exposure to vector- borne
infectious diseases and heat stress in parts of Asia (medium confidence)” (WGII p. 14).
Small Islands: “Many tropical islands are now experiencing high incidences of vector- and
water-borne diseases that are attributed to changes in temperature and rainfall regimes, which
may be linked to events such as ENSO, droughts, and floods. In the Pacific, there is growing
evidence that outbreaks of dengue are becoming more frequent and appear to be strongly
correlated with the ENSO phenomenon” (IPCC WGII Chapter 17 p. 864).
“Climate change will cause some deterioration in air quality in many large urban areas,
assuming that current emission levels continue (medium to high confidence)” (IPCC WGII
Chapter 9 p. 453)
“In areas with limited or deteriorating public health infrastructure, and where temperatures
now or in the future are permissive of disease transmission, an increase in temperatures (along
with adequate rainfall) will cause certain vector-borne diseases (including malaria, dengue,
and leishmaniasis) to extend to higher altitudes (medium to high confidence) and higher
latitudes (medium to low confidence)” (IPCC WGII Chapter 9 p. 453).
“In some settings, the impacts of climate change may cause social disruption, economic
decline, and displacement of populations. The ability of affected communities to adapt to such
disruptive events will depend on the social, political, and economic situation of the country
and its population. The health impacts associated with such social-economic dislocation and
population displacement are substantial [high confidence; well-established]” (IPCC WGII
Chapter 9 p. 454).
Agriculture “Recent studies support evidence in the 1990
assessment that, on the whole, global agricultural
production could be maintained relative to baseline
production in the face of climate change modeled by
GCMs at doubled-equivalent CO
2
equilibrium
conditions. However, more important than global food
production—in terms of the potential for hunger,
malnutrition, and famine—is the access to and
availability of food for specific local and regional
populations” (WGII Technical Summary).
Europe: “There will be some broadly positive effects on agriculture in northern Europe
(medium confidence); productivity will decrease in southern and eastern Europe (medium
confidence)” (Synthesis Report p. 128).
North America: “Some crops would benefit from modest warming accompanied by
increasing CO
2
, but effect would vary among crops and regions (high confidence), including
declines due to drought in some areas of Canada’s Prairies and the U.S. Great Plains, potential
increased food production in areas of Canada north of current production areas, and increased
warm temperate mixed forest production (medium confidence). However, benefits for crops
would decline at an increasing rate and possibly become a net loss with further warming
(medium confidence)” (Synthesis Report p. 128).
58
“At broader regional scales, subtropical and tropical
areas—home to many of the world’s poorest people—
show negative consequences more often than temperate
areas. People dependent on isolated agricultural
systems in semi-arid and arid regions face the greatest
risk of increased hunger due to climate change. Many
of these at-risk populations live in sub-Saharan Africa;
South, East, and Southeast Asia; and tropical areas of
Latin America, as well as some Pacific island nations”
(WGII Technical Summary).
“... many of the world’s poorest people - particularly
those living in the subtropical and tropical areas and
dependent on isolated agricultural systems in semi-arid
and arid regions are most at risk of increased hunger”
(p. 33)
“Increases in atmospheric carbon dioxide
concentrations may raise the carbon-nitrogen ratio of
forage, thus reducing its food value” (p. 30).
Latin America: “Yields of important crops are projected to decrease in many locations even
when the effects of CO
2
are taken into account; subsistence farming in some regions could be
threatened (high confidence)” (Synthesis Report p. 128).
Asia: “Decreases in agricultural productivity and aquaculture due to thermal and water stress,
sea-level rise, floods and droughts, and tropical cyclones would diminish food security in
many countries of arid, tropical, and temperate Asia; agriculture would expand and increase in
productivity in northern areas (medium confidence)” (Synthesis Report p. 128).
Africa: “Grain yields are projected to decrease for many scenarios, diminishing food security,
particularly in small food-importing countries (medium-high confidence)” (WGII-SPM p. 14).
Small Island States: “Limited arable land and soil salinization makes agriculture of small
island states, both for domestic food production and cash crop exports, highly vulnerable to
climate change (high confidence)” (WGII-SPM p. 17).
Australia and New Zealand: “The net impact on some temperate crops of climate and CO
2
changes may initially be beneficial, but this balance is expected to become negative for some
areas and crops with further climate change (medium confidence)” (WGII-SPM p. 15).
“Climate change represents an additional pressure on the world’s food supply system and is
expected to increase yields at higher latitudes and lead to decreases at lower latitudes. These
regional differences in climate impacts on agricultural yield are likely to grow stronger over
time, with net beneficial effects on yields and production in the developed world and net
negative effects in the developing world. This would increase the number of undernourished
people in the developing world (medium confidence).” (IPCC WGII Chapter 9 p. 454).
“Agricultural yields will increase for most crops as a result of increasing atmospheric CO
2
concentration. This effect would be counteracted by the risk of water shortage in southern and
eastern Europe and by shortening of growth duration in many grain crops as a result of
increasing temperature. Northern Europe is likely to experience overall positive effects,
whereas some agricultural production systems in southern Europe may be threatened [medium
confidence, established but incomplete evidence]” (IPCC WGII Chapter 13 p. 643).
Human
infrastructure
“Climate change and resulting sea level rise can have a
number of negative impacts on energy, industry and
transportation infrastructure; human settlements; the
property insurance industry; tourism; and cultural
Asia: “Sea level rise and an increase in intensity of tropical cyclones would displace tens of
millions of people in low-lying coastal areas of temperate and tropical Asia; increased
intensity of rainfall would increase flood risks in temperate and tropical Asia (high
confidence
)” (WGII
-
SPM p. 14).
59
systems and values” (p. 34).
“Protection of many low-lying island states (e.g., the
Marshall Islands, the Maldives) and nations with large
deltaic areas (e.g., Bangladesh, Nigeria, Egypt, China)
is likely to be very costly (High Confidence).”
(Executive Summary of Chapter 9).
“Adaptation to sea-level rise and climate change will
involve important tradeoffs, which could include
environmental, economic, social, and cultural values
(High Confidence)” (Executive Summary of Chapter
9).
“In some societies, resettlement, for example, would
lead to dislocation of social and cultural groups and
might even involve the loss of cultural norms and
values...” (Chapter 9.6.3.3).
confidence)” (WGII-SPM p. 14).
Small Islands: “The projected sea level rise of 5mm yr
-1
for the next 100 years would cause
enhanced coastal erosion, loss of land and property, dislocation of people, increased risk from
storm surges, reduced resilience of coastal ecosystems, saltwater intrusion into freshwater
resources, and high resource costs to respond to and adapt to these changes (high confidence)”
(WGII-SPM p. 17).
Europe: “In coastal areas, the risk of flooding, erosion, and wetland loss will increase
substantially—with implications for human settlement, industry, tourism, agriculture, and
coastal natural habitats. Southern Europe appears to be more vulnerable to these changes,
although the North Sea coast already has high exposure to flooding [high confidence]” (IPCC
WGII Chapter 13 p. 644).
Water resources “Relatively small changes in temperature and
precipitation, together with the non-linear effects on
evapotranspiration and soil moisture, can result in
relatively large changes in runoff, especially in arid and
semi-arid lands. High-latitude regions may experience
increased runoff due to increased precipitation,
whereas runoff may decrease at lower latitudes due to
the combined effects of increased evapotranspiration
and decreased precipitation. Even in areas where
models project a precipitation increase, higher
evaporation rates may lead to reduced runoff” (WGII
Technical Summary).
“Climate change ... can have major impacts on regional
water resources” (p. 32).
“The quantity and quality of water supplies already are
serious problems today in many regions...making
countries in these regions particularly vulnerable to any
“Climate change will exacerbate water shortages in many water-scarce areas of the world.
Demand for water is generally increasing due to population growth and economic
development, but is falling in some countries because of increased efficiency of use. Climate
change is projected to substantially reduce available water (as reflected by projected runoff) in
many of the water-scarce areas of the world, but to increase it in some other areas (medium
confidence) …. Freshwater quality generally would be degraded by higher water temperatures
(high confidence), but this may be offset in some regions by increased flows” (SR-SPM p.
12).
“Projected climate change would exacerbate water shortage and quality problems in many
water-scarce areas of the world, but alleviate it in some other areas. … Climate change is
projected to reduce streamflow and groundwater recharge in many parts of the world but to
increase it in some other areas (medium confidence)” (Synthesis Report p. 72).
Africa: “Changes in rainfall and intensified land use would exacerbate the desertification
processes. Desertification would be exacerbated by reduction in the average annual rainfall,
runoff, and soil moisture in countries of west African Sahel, and northern and southern Africa
(medium confidence). Increases in droughts and other extreme events would add to stresses on
wat
er resources, food security, and human health, and would constrain development in the
60
additional reduction in indigenous water supplies” (p.
32).
“Experts disagree over whether water supply systems
will evolve substantially enough in the future to
compensate for the anticipated negative impacts of
climate change on water resources and for potential
increases in demand” (p. 32).
“… The impacts, however, will depend also on the
actions of water users and managers… In some
cases—particularly in wealthier countries with
integrated water-management systems—these actions
may protect water users from climate change at
minimal cost. In many others however—particularly
those regions that already are water-limited—
substantial economic, social, and environmental costs
could occur. Water resources in arid and semi-arid
zones are particularly sensitive to climate variations
because of low-volume total runoff and infiltration and
because relatively small changes in temperature and
precipitation can have large effects on runoff” (WGII
Technical Summary).
region (high confidence)” (Synthesis Report p. 130).
Asia: “Water shortage— already a limiting factor for ecosystems, food and fiber production,
human settlements, and human health— may be exacerbated by climate change. Runoff and
water availability may decrease in arid and semi-arid Asia but increase in northern Asia
(medium confidence)” (Synthesis Report p. 130).
Europe: “Summer runoff, water availability, and soil moisture are likely to decrease in
southern Europe, and would widen the gap between the north and south (high confidence).
Flood hazards will increase across much of Europe (medium to high confidence); risk would
be substantial for coastal areas where flooding will increase erosion and result in loss of
wetlands” (Synthesis Report p. 130).
Australia and New Zealand: “Water is likely to be a key issue (high confidence) due to
projected drying trends over much of the region and change to a more El Niño-like average
state” (Synthesis Report p. 130).
North America: “Snowmelt-dominated watersheds in western North America will experience
earlier spring peak flows (high confidence) and reduction in summer flow (medium
confidence); adaptive responses may offset some, but not all, of the impacts on water
resources and aquatic ecosystems (medium confidence)” (Synthesis Report p. 130).
Small Islands: “Islands with very limited water supplies are highly vulnerable to the impacts
of climate change on the water balance (high confidence)” (Synthesis Report p. 130).
61
Food Production and Agriculture
Apart from the uncertainty of future climate changes, the impacts of climate change
on food production and agriculture depend on a range of factors. These include the
vulnerability of agricultural activities, regions and populations to changes in climate
and the capacity of these systems to adapt to the changes. Where vulnerability is high
and adaptive capacity low there is likely to be the highest sensitivity to climate
effects.
Relevant factors in determining the response of agricultural systems to climate
change include:
• Rate and magnitude of changes in temperature and extremes of
temperature. In the mid-latitudes increases in temperature, particularly
increases in minimum temperature, can raise crop production providing
water availability is not compromised. In the tropics crops are often close
to thermal optimums, thus reductions rather than increases in production
may result from increased temperatures.
• Changes in precipitation amounts and seasonality, drought, ENSO and
other extreme event frequency, intensity and duration. If increased
temperatures are accompanied by sufficiently increased precipitation,
given that rising temperatures lead to elevated evaporation rates, crop
yields may increase. Otherwise crop production may fall. Changes in
extreme events are likely to influence crop production quite substantially
(Mearns et al. 1997; Phillips et al. 1998; Rosenzweig et al. 2001) either
directly or through changes in pest abundance and prevalence. Few
attempts have been made to model extreme event effects on agricultural
production.
• Effects of CO
2
fertilization on crop and grass production and yield.
Increased CO
2
may increase water use efficiency of crops but is also likely
to reduce the nutrient quality of the crops.
• Socio-economic conditions of rural populations and their access to
markets, technology and resources needed for adaptation or for the
acquisition of replacement food resources. Typically, in poor regions it is
expected that farmers and those directly dependent on rural land activities
will be most vulnerable to climate change.
Taken from the IPCC TAR Synthesis Report (TAR SYR), Table 7 summaries the
findings of the IPCC TAR and attempts to place temperature-warming bands on
impacts and effects. Other findings from the IPCC TAR, for which a temperature
increase may be associated with changes in agricultural production, include:
• Agriculture in mid-latitude countries is expected with medium confidence to
benefit for a warming of “a few degrees”
98
(2.6- 3.6
o
C above 1861-1890
96
).
Over 3-4
o
C warming, there is low to medium confidence in a general decline
in mid-latitude crop production along with quite pronounced drops in
production elsewhere, leading to higher food prices (TAR SYR 4.2).
96
Unless otherwise noted temperature increases in this section will be cited with respect to the 1861-
1890 average – see also Appendix I below.
62
Table 7 - Agricultural effects of climate change
Effect (change) 2025
2050 2100
CO
2
concentration
a
405–460 ppm 445–640 ppm 540–970 ppm
Global mean temperature
change from the year
1990
b
0.4–1.1°C 0.8–2.6°C 1.4–5.8°C
Global mean temperature
change from the years
1861-1890 (average)
97
1.0-1.7°C 1.4-3.2°C 2.0-6.4°C
Global mean sea-level
rise from the year 1990
b
3–14 cm 5–32 cm 9–88 cm
Agricultural Effects
c
Average crop yields
g
[WGII TAR Sections
5.3.6, 10.2.2, 11.2.2, 12.5,
13.2.3, 14.2.2, & 15.2.3]
Cereal crop yields increase in
many mid- and high-latitude
regions (low to medium
confidence
d
).
Cereal crop yields decrease in
most tropical and subtropical
regions (low to medium
confidence
d
).
Mixed effects on cereal yields
in mid-latitude regions. More
pronounced cereal yield
decreases in tropical and
subtropical regions (low to
medium confidence
d
).
General reduction in cereal
yields in most mid-latitude
regions for warming of more
than a few
98
°C (low to
medium confidence
d
).
Extreme low and high
temperatures [WGII TAR
Section 5.3.3]
Reduced frost damage to
some crops (high
confidence
d
).
Increased heat stress damage
to some crops (high
confidence
d
).
Increased heat stress in
livestock (high confidence
d
).
Effects of changes in extreme
temperatures amplified (high
confidence
d
).
Effects of changes in extreme
temperatures amplified (high
confidence
d
).
Incomes and prices
[WGII TAR Sections
5.3.5-6]
Incomes of poor farmers in
developing countries decrease
(low to medium confidence
d
).
Food prices increase relative
to projections that exclude
climate change (low to
medium confidence
d
).
No climate policy interventions. Source: Table 3-3 and references from IPCC TAR Synthesis Report with the addition of the
row headed ‘Global mean temperature change from the years 1861-1890 (average)’. Summarized versions of the original notes
a-d associated with these tables appear below.
99
Note g - these estimates are based on the sensitivity of the present agricultural
practices to climate change, allowing (in most cases) for adaptations based on shifting use of only existing technologies.
97
Using Folland et al. (2001) global temperature data set.
98
This term is not defined in Synthesis Report and nor is it defined in the Working Group II Report
Summary for Policy Makers. Chapter 19 of the WGII TAR does define a range – see Appendix I of
this report. In the Final Government Distribution of the Synthesis Report, which is the version upon
which the final negotiated report is based and is prepared by the IPCC Lead Authors, “few” is defined
as 2-3
o
C above 1990. The removal of this specific definition was initiated and insisted upon by Saudi
Arabia, amongst others, at the IPCC plenary where this report was adopted. Given this context “few”
will be interpreted here in the original sense of the lead authors of the report. In terms of the 1861-
1890 reference period, adopted in this report as surrogate for pre-industrial temperatures 2-3oC above
1990 corresponds to an increase of 2.6-3.6oC.
99 a. The ranges for CO
2
concentration are estimated for the six illustrative SRES scenarios, with the
ranges for minimum and maximum values estimated for the 35 SRES projections of greenhouse gas
emissions. See WGI TAR Section 3.7.3.
b. The reported ranges for global mean temperature change and global mean sea-level rise correspond
to the minimum and maximum values estimated with a simple climate model for the 35 SRES
projections of greenhouse gas and SO
2
emissions. See WGI TAR Sections 9.3.3 and 11.5.1.
63
• In the tropics and some subtropical regions (mostly developing countries),
cereal crop yields are projected to drop with even minimal changes in
temperature (TAR SYR Section 4.2).
• With low to medium confidence, it is expected that income of poor farmers
will decline above a warming of 1.5-2
o
C.
• Above around a 2.5-3
o
C warming, it is estimated, with low confidence, that
there will be a general increase in food prices. One study cited in the TAR
(Parry et al. (1999) – see below) found prices to increase above around 1.6
o
C
in the 2020s, however results in this temperature range are generally mixed.
• A review of the implications for rice production in Asia found that climate
change is likely to seriously threaten sustained food production, with
temperature increases above 2.6
o
C outweighing the positive effects of CO
2
increases (WGII Chapter 11.2.2.1 (Lal et al. 2001)).
• Australian crop yields were estimated to increase up to 1.6-2.6
o
C and then
decline with higher temperatures, with it being noted that drops in rainfall
caused rapid decreases in crop yield. It was reported that the most recent
scenarios show reductions in rainfall over much of Australia. In the case of
Australia, global mean warming in the range of 2.3-2.6
o
C,
100
has been
projected to results in crop yields changing in the range of –3% to +3%, but
significant areas in the west and the south would experience reductions. At
higher temperatures, 4.2
o
C in the 2080s, entire areas are projected to be out of
production, particularly in southwestern Australia (IPCC WGII TAR, Chapter
12 (Pittocket et al. 2001)).
• In general, for a global warming of about 2
o
C European crop production is
expected to increase, with a few exceptions in the south of Portugal and Spain
and in the Ukraine where decreases are estimated (IPCC WGII TAR, Chapter
13 (Kundzewicz et al. 2001)).
• In the USA, it was estimated that agricultural welfare would increase up to
about 2
o
C and then decline at an increasing rate with the magnitude and
direction of changes in rainfall being a decisive factor (IPCC WGII TAR,
Chapter 15.2.3.1. (Cohen et al. 2001)).
• Large drops in the yield of maize and sugarcane are projected for small island
countries for doubled CO
2
conditions (IPCC WGII TAR, Chapter 17.2.8
(Nurse et al. 2001)).
c. Summary statements about effects of climate change in the years 2025, 2050, and 2100 are inferred
from IPCC Working Group II’s assessment of studies that investigate the impacts of scenarios other
than the SRES projections, as studies that use the SRES projections have not been published yet.
Estimates of the impacts of climate change vary by region and are highly sensitive to estimates of
regional and seasonal patterns of temperature and precipitation changes, changes in the frequencies or
intensities of climate extremes and rates of change. Estimates of impacts are also highly sensitive to
assumptions about characteristics of future societies and the extent and effectiveness of future
adaptations to climate change. As a consequence, summary statements about the impacts of climate
change in the years 2025, 2050, and 2100 must necessarily be general and qualitative. The statements
in the table are considered to be valid for a broad range of scenarios. Note, however, that few studies
have investigated the effects of climate changes that would accompany global temperature increases
near the upper end of the range reported for the year 2100.
d. Judgments of confidence use the following scale: very high (95% or greater), high (67–95%),
medium (33–67%), low (5–33%), and very low (5% or less). See WGII TAR Box 1-1.
100
SRES B2 and A1 scenarios; (Hulme and Sheard 1999).
64
More qualitatively a number of general conclusions were reached:
• Climate change is likely to exacerbate degradation of land and water
resources.
• Elevated CO
2
combined with higher temperatures is likely to significantly
reduce the protein and nutrient content of important cereal crops and of
forage.
• Increased pest outbreaks with significant negative impacts on crop production
seem likely for many crops and regions. Very few studies have included
changed pest activity under climate change.
• Africa appears to be particularly at risk of increased hunger due to poverty
and intrinsic vulnerability to climate change.
To look at this picture more closely attention will be focused on the findings of two
recent global assessments. The first, published in 1999 by Parry and co-workers, was
assessed in the IPCC TAR and its results were also presented in a synthesis of a range
of impacts by Parry et al. (2001). Using the methodology and models from this work
Arnell et al. (2002) compared the effects of an unmitigated emission scenario (IS92a)
and concentration stabilization at 550 and 750 ppmv CO
2
scenarios with the HadCM2
GCM on a number of sectors including agriculture. The second is the Global Agro-
Ecological Assessment for Agriculture in the 21st Century (also known as the GAEZ
study) authored by Fischer et al. (2002) under the auspices of IIASA and the FAO.
101
This used a methodology built from a detailed bottom up, national level review of
agricultural systems and was driven by several GCMs, including ECHAM4 and
HadCM2.
Climate change and food security assessments
Several quantitative estimates of likely global impacts of climate change on food
supply and the risk of hunger have been made over the last decade (Parry and Carter
1989; Rosenzweig and Parry 1994). In the 1994, study three early GCMs – the GISS,
GFDL and UKMO models – were driven by an equilibrium CO
2
scenario for an
increase of CO
2
from 330 to 550 ppmv in 2060. As is usual in these studies
continued increases in crop yield and increase arable land availability were assumed.
The results were strongly dependent on whether or not a CO
2
fertilization effect on
crop yield was included and on the assumed level of adaptation. Under the first level
of adaptation only small changes to the existing system were assumed. In this case
the number of additional people at risk of hunger increased by 10-60% (60-350
million extra people at risk). These estimates decreased significantly when the
second level of adaptation was assumed, which represented in the words of the
authors, "a fairly optimistic assessment of the world's response to the changed
climatic conditions tested.”
Parry et al. (1999) used two of the latest generation of AOGCMs from the Hadley
Centre which were driven by the IPCC IS92a emissions scenario to produce time
dependant projections of future climate (see Table 9 below for an outline of the basic
features of these scenarios of relevance here). Two kinds of adaptation were
101
IIASA, International Institute for Applied Systems Analysis; FAO, Food and Agriculture
Organisation.
65
incorporated: farm level measures and economic adjustment effects. Geographically
explicit crop models were used covering over 70% of the world’s current wheat,
maize and soybean production area, however less than half the rice growing regions
were included. The crop models were driven using the effects of increased CO
2
and
the projected climate changes from the model scenarios. Caution was expressed in
relation to the assumed enhanced growth effects of CO
2
on crops included in the
models, which the authors noted had not been verified in field conditions.
Consequently, there is a risk that the positive yield effects assumed were
overestimated (an issue also noted by Darwin and Kennedy (2000)).
Future increases in arable land were based on FAO data and did not account for the
effects of climate change, an issue discussed in detail by Ramankutty et al. (2002)).
The latter study finds that projected climate change is likely to increase the area of
arable land suitable for crop production overall, with increases principally located in
the Northern Hemisphere. However, the tropics (mainly Africa, northern South
America, Mexico and Central America and Oceania) are likely to experience small
reductions in suitable area. This general finding is confirmed in the IIASA study.
Ramankutty et al. (2002) also point out that much of the land, that is at present or
may become climatically suitable for agriculture in the future, is also under valuable
forest cover.
Figure 9 shows a comparison of the results of these scenarios for two different time
periods, the 2050s and 2080s (see explanation beneath figure), when temperatures are
expected to have exceeded 2
o
C by a significant margin. The HadCM3 model
produces more extreme results than the HadCM2 model, notably for large parts of the
Northern Hemisphere. Whereas crop yields increase in these regions by the 2050s
under some of the ensemble members
102
of the HadCM2 scenarios, under the
HadCM3 scenario, which is drier and warmer than the HadCM2 scenarios, large areas
in North America, Russia and eastern Europe experience reductions. By and large,
crop yields are down in developing countries, under both models, with HadCM3
showing the most severe changes. In the 2050s, HadCM2 indicates that India is the
worst affected of the developing countries with reductions of the order of 5-10%,
whereas HadCM3 implies smaller damages in the range 0-2.5%. HadCM3 indicates
that western Africa will experience the worst changes in the 2050s. By the 2080s,
HadCM3 indicates larger damages in India (2.5-5% losses) and large losses in
southern Africa. Figure 8 graphically demonstrates the range of effects projected by
the different models and demonstrates that in general the regions at risk of production
reductions are common to the two models, with the exception of North America.
However, the quantitative scale of the reductions varies significantly.
In interpreting the results of this study it is important to bear in mind the uncertainties
in this work. Apart from climate change itself, the question of whether and to what
extent the CO
2
fertilization effect benefits production, the availability of irrigation
water, trends in demand and the range of adaptation possibilities are all significant
(Parry et al. 1999). Nevertheless, the overall changes and reductions in some regions
translate to additional people at risk of hunger, with increasing temperatures tending
to increase the number at risk (see Figure 10 below). Africa emerges from this study
102
Multiple runs of these complex models driven with the same emissions scenarios produce different
results due to the ‘natural’ variability in the model climate system.
66
as a region particularly at risk from climate change under either of the models and has
the largest share of the additional people at risk of hunger (see Table 8 below).
Figure 8 - Regional Impacts on Crop Production
Source: Figure 7 of Parry et al. (1999: S65). Shaded boxes are the HadCM3 results and the lines are
the range of the HadCM2 ensemble results.
Table 8 - Risk of Hunger - Africa
Model 2080s temperature
increase above 1861-
1890 average
Impact
HadCM2 3.4
o
C Africa: 55-65 million more people at risk of hunger.
Globally: 80 million more at risk of hunger.
HadCM3 3.3
o
C Africa: 70 million more people at risk of hunger.
Globally: 125 million more at risk of hunger.
Note: Compiled from data in Parry et al. (1999).
67
Figure 9 - Comparison of Potential Crop Yields Projections for 2050s and 2080s
Potential change in national cereal yields for the 2050s and 2080s, compared to 1990, for temperature
increases of around 2.4
o
C and 3.3-3.4
o
C respectively compared to 1861-1890 for the HadCM2 and
HadCM3 scenarios. The HadCM2 scenarios are for an ensemble of four members and the HadCM3
for a single scenario. Source: Figure 5 of Parry et al. (1999: S61).
68
Whilst the Parry et al. (1999) assessment provides insight into the risks faced from
future climate change at different time periods into the future, it does not directly
enable a comparison of risks at different levels of warming at different time periods in
the future. Subsequent work by Parry et al. (2001) analysed projected effects at
different temperature levels for the 2050s and 2080s. Known widely as the “Millions
at Risk” paper, it provides a meta-analysis of several impact areas and enables some
rough estimates to be made of impacts at different levels of warming at different
times in the future. Perhaps most significantly, it also illustrates some of the
dynamics of changing vulnerability over time and the interaction of this with
projected climate changes. Levels of adaptive capacity are assumed to vary with
time, with rising economic wealth being associated with higher levels of adaptive
ability and greater resilience to climate change. One of the main drawbacks with this
work, however, is that it is based essentially on one AOGCM, the HadCM2 model.
Where possible, results from other models will be compared with the effects of the
HadCM2 projections in order to at least provide a feel for the uncertainties involved.
For the food security issue, the data embodied in Figure 11 (based on the HadCM2
model climate projection) was used to estimate the levels of additional risk of hunger
at warming of 1
o
C, 2
o
C, 2.5
o
C and 3
o
C, which are tabulated in the third column of
Table 10.
103
For the 2050s, warming of 1-2.5
o
C is estimated to produce an additional
hunger risk of 4-7 million (this can be compared with the HadCM3 based estimate of
close to 40 million people for around a 2.4
o
C increase in 2050). Under the HadCM2
scenario, increasing temperatures in this time period are not projected to change the
number at risk dramatically, as climate change is not projected to affect prices
significantly in the 2050s. During this period, production in North America still
increases (Parry et al. 1999). Over the following 30 years, there is an increase by a
factor of 5-7 in the number at risk, as can be seen from the rapidly rising curve for
hunger on the right hand side of Figure 11. At the maximum of the temperature scale
for the 2080s, around 3.4-3.5
o
C warming, the total number at risk of hunger are in the
range 75-100 millions. The HadCM3 projections are higher for this period, on the
order of 125 million (see Figure 10). This reflects an increasing sensitivity to climate
change during this period and an increased population in vulnerable regions.
103
For convenience, the results of the other impact areas assessed – health, flooding and water shortage are
presented in the graph. However only the water shortage issue is discussed, as the other issues are beyond the
scope of the present paper.
69
Figure 10 - Global Risk of Hunger
Source: Figure 6 (c) of Parry et al. (1999: S64). Shaded boxes are the HadCM3 results and the lines
are the range of the HadCM2 ensemble results.
Figure 11 - Millions at Risk in 2050s and 2080s: Hunger, Malaria, Water Shortage and Flooding
This graph shows the estimated millions at risk associated with global mean warming levels above the
1961-1990 average based on the studies describe in Parry et al. (2001) which were driven by an
ensemble of scenarios from the HadCM2 model. An error band of one standard deviation around the
mean is shown, with the solid lines indicating model results and the dotted lines being inferred from
these results.
Source: This figure is a version of Figure 1 of Parry et al. (2001) taken from the web
document “Defining critical climate change threats and targets: Discussion of the figures from Global
Environment Change 11:3(2001):1-3” by the same authors, downloaded from www.jei.uea.ac.uk,
February 2002, Jackson Environment Institute, School of Environmental Sciences University of East
Anglia, Norwich NR4 7TJ United Kingdom.
70
Table 9 - Summary of Scenarios used in Global Food Security Assessment
Summary of scenarios
1961-90
2020s
2050s
2080s
HadCM2
Temperature change (
o
C) 0 1.2
2.1
3.1
Temperature change wrt
1861-1890
0.3
1.5
2.4
3.4
Precipitation change (%) 0 1.6
2.9
4.5
Sea-level rise (cm) 0 12
25
41
CO
2
(ppmv) 334 441
565
731
HadCM3
Temperature change (
o
C) 0 1.1
2.1
3.0
Temperature change wrt
1861-1890
0.3 1.4
2.4
3.3
Precipitation change (%) 0 1.3
2.4
3.2
Sea-level rise (cm) 0 12
24
40
CO
2
(ppmv 334 433
527
642
Source: Based on Table 1 of Hulme et al. (1999a) describing scenarios with the HadCM2 and
HadCM3 model driven by the IS92a scenario with no aerosol forcing and used by Parry et al. (1999).
The offset from 1961-1990 temperatures to 1861-1890 is with the Folland et al. (2001) global
temperature data set.
Table 10 - Millions at Risk
2050s
Temperature
in 2050s above 1861-1890
(above 1961-1990)
Malaria Hunger Water shortage
Coastal flooding
1
o
C (0.7
o
C)
163
4
1228
12
2
o
C (1.7
o
C)
224
7
2358
26
2.5
o
C (2.2
o
C)
227
7
2675
32
2080s
Temperature in 2080s above
1861-1890
(1961-1990)
Malaria Hunger Water shortage
Coastal flooding
1
o
C (0.7
o
C) 101
10
149
1
1.5
o
C (1.2
o
C) 165
21
562
8
2
o
C (1.7
o
C) 212
33
2427
19
2.5
o
C (2.2
o
C) 250
49
3117
36
3
o
C (2.7
o
C)
277
67
3245
57
3.4
o
C (3.1
o
C) 291
84
3473
79
Data estimated from figures in Parry, et al. (2001) using data-extractor software. Temperatures in
parentheses are relative to 1990, the temperature base year used by Parry et al. (2001). These figures
should be treated as indicative only, as they are based on visual interpolation using a graphical data
digitising programme.
71
Global Agro-Ecological Assessment (GAEZ Study)
The IIASA/FAO assessment of agriculture over the next century (Fischer et al. 2001)
produced qualitatively similar results to that of the Parry et al. (1999) assessment.
Taking into account land suitability, population growth and other factors and a
climate change scenario that brings around a 3
o
C warming in the 2080s, developing
countries as a group suffer production losses. A large group of about 40 developing
countries with a current population of 2 billion people, including around 450 million
undernourished inhabitants, is projected to lose substantially, whilst about half the
developing country group gain. Details of the projections for the group of developing
countries experiencing malnourishment problems are found in Table 11. The 78
countries presently at some level of risk are divided into three groups.
Under the ECHAM4 climate scenario (see Table 12), a 3
o
C warming by the 2080s
results in projected declines in cereal production, although at a world average level
the volume of production is estimated to be sufficient to meet future needs.
Developed countries as a whole are projected to experience a small loss in rain-fed
cereal production. Within this picture 17 countries gain, though only two countries,
Russia and Canada, enjoy 90% of the gain. The majority encompassing 60% of the
population of the developed country group, including Belgium, the Czech Republic,
France, Hungary, Romania, The Netherlands, United Kingdom, Ukraine, and the
USA are projected to lose under this scenario (Fischer et al. 2001).
Table 11 - Malnourished Country Group and Climate Change
Group Population Proportion of
population
undernourished
Number of
countries in group
Number of
countries
negatively
affected
Impact
I
5-20%
undernourished
2.1 billion
12% 28
Includes China
11 -10% decrease
in cereal
production.
China gains
II
20-35%
undernourished
1.5 billion
25% 27
Includes India with
60% of the
undernourished
19 with over
80% of
undernourished
Food deficit
doubled
III
More than 35%
undernourished
440 million 50% 23
Most sub-Saharan
African countries
10 Decrease in
production
6 gain
substantially
Compiled from Fischer et al. (2001: 27).
Within the developing country group, 65 countries are projected to experience
production losses valued at US$56 billion in 1995 terms. These losses equate to 16%
of the agricultural GDP of these countries (Fischer et al. 2001: 26). Africa appears to
be the biggest loser in these scenarios, with 29 countries projected to suffer
production losses. Kenya and South Africa are, however, projected to gain
substantially from climate change. In Asia, China gains substantially whilst India
loses (Fischer et al. 2001: 27). Overall Fischer et al. (2001) identify 40 “losing
countries” with a total population close to 2 billion and an undernourished group of
about 450 million. In these countries the gap between food production and supply is
72
projected to double under climate change, “drastically” increasing the number
suffering from under nourishment (Fischer et al. 2001: 28).
A cross comparison of the projected effects of different climate changes projected by
the ECHAM4, HadCM2 and the CGCM1 models (Carter et al. 2001) is shown in
Table 15. It is clear that there is a wide range of results, with ECHAM4 projecting
losses at the lower end of the other two model estimates. For the developing
countries as a group ECHAM4 projects a potential cereal production increase of 23
million tons per year affecting 3.7 billion people. The other two models project
losses in the range of 63 to 226 million tonnes per year, affecting 3.3 to 5.5 billion
people respectively.
Table 12 - Global Mean Temperature increase for ECHAM4 Scenarios
ECHAM4 2020s
o
C above
1861-1890
2050s
o
C above
1861-1890
2080s
o
C above
1861-1890
Greenhouse gas only
1.5
2.5 3.0
Greenhouse gases plus aerosols
1.3
1.7 NA
Note: Estimated with data from the IPCC DDC web site
104
for the ECHAM 4 scenarios for increases
with respect to 1961-1990 average and converted to the 1861-1890 reference period using the observed
increase in global mean temperature from this period to 1961-1990 (Folland et al. 2001), an offset of
about 0.3
o
C. The scenario with aerosols was used only for the 2020s and not for other time periods in
the GAEZ study.
Table 13 - Developing Country Changes in Rain Fed Cereal Production Potential 2080s for
Three Climate Models
Climate
Model
Number of countries Projected population
2080 (billions)
Change in cereal production
potential (million tons)
G
a
N L G N L G N L Total
ECHAM4 40 34 43 3.1 0.9 3.7 142 –2 –117 23
HADCM2 52 27 38 3.2 1.2 3.3 207 3 –273 –63
CGCM1 25 26 66 1.1 1.1 5.5 39 3 –268 –226
Notes: a. G = countries gaining +5% or more; N = small change of –5 to +5%; L = countries losing –
5% or more. This tables shows the number of developing countries projected to experience gains, no
change or losses in cereal production potential on current cultivated land and potentially cultivatable
land in the 2080s. Source: Table 5.28, Fischer et al. (2002: 105). ECHAM 4 refers to the AOGCM of
the Max Planck Institute for Meteorology, HadCM2 to that for the Hadley Centre in the UK and
CGCM1 to that of the Canadian Climate Modelling Centre.
Discussion and Summary
It seems very likely that the pattern of some regions and countries gaining and others
losing is a robust feature of the impacts of climate change over the next century.
Many studies indicate that developing countries are likely to lose as a whole, relative
to the developed nations. India is projected to experience significant losses, with
quite large areas of current cropland losing significant productivity.
Few estimates have been made of the overall macroeconomic consequences of
projected agricultural effects of climate change for developing countries. In many
104
http://ipcc-ddc.cru.uea.ac.uk/asres/scenario_home.html.
73
countries of Africa and Asia, the agricultural sector represents a large share of the
economic activity. This share is projected to remain large over the next fifty years or
so, thus grounds for concern exist about the possible impacts of macroeconomic
shocks from climate change on vulnerable countries. One of the few computable
general equilibrium assessments published, projected large, even “violent changes in
the economic and social structure,” as a consequence of climate shocks (Winters et al.
1998: 16). Though the basis for the study was the old GCM scenario results of
Rosenzweig and Parry (1994), it does point to some major and apparently under-
examined risks, particularly for African countries.
Within countries there will be regions that gain substantially and regions that lose.
The results of the GAEZ study provide an insight into this issue. Figure 12 shows the
relative change in productivity of cereal cropping regions for the ECHAM4 scenario
in the 2080s (3
o
C) for developed and developing countries. The specific examples of
India, China, the USA and Russia are also given as an illustration of these two
groups.
For the developed countries, it can be seen that whilst gains outweigh losses the
regions in which losses occur represent a large fraction of the total current crop area.
Although the area negatively impacted in Russia is small (the Russian case is
discussed further below). Developed countries will not be immune to large effects of
climate change on their agricultural sectors. As the example of Australia, cited above,
indicates, warming of the order of 4
o
C is likely to put entire regions out of
production, with lesser levels of warming causing substantial declines in the west and
the south.
74
Figure 12 - Gains and Losses in Production Potential under Climate Change
Gains and losses in cereal crop productivity in current cultivated areas under the ECHAM4 model in
the 2080s. Source: Figure 5.4 of GAEZ study (Fischer et al. 2002: 70).
From the point of interpreting these assessments in relation to Article 2, it seems that
the following points might be made:
• Warming of around 1
o
C produces relatively small damages when measured
from the point of increased risk of hunger and/or under nourishment (around
10 million more at risk) over the next century. In this temperature range
nearly all developed countries are projected to benefit, whilst many
developing countries in the tropics are estimated to experience small but
significant crop yield growth declines relative to an unchanged climate.
• At all levels of warming, a large group of the poor, highly vulnerable
developing countries is expected to suffer increasing food deficits. It is
anticipated that this will lead to higher levels of food insecurity and hunger in
these countries.
• Under the Parry et al. (2001) analysis moving from 1
o
C to 2
o
C warming
triples the number of people at risk of hunger in the 2080s.
• A 2
o
C warming and above, is associated with increasing risk. Under the Parry
et al. (2001) analysis, this risk increases 4-5 fold from the 2050s to the 2080s
(for the same temperature). In this temperature range many developed
countries may still gain, although warning signs in the literature caution that
this may not be robust for all regions or even in aggregate terms. It appears
that agricultural production in developed countries is finely balanced in this
75
temperature range between the effects of increased temperature and changes
in precipitation. The effects are very sensitive to the precipitation scenarios,
which vary considerably between the GCMs. Parry et al. (1999) call attention
to the effect of the HadCM3 scenarios, which are ‘drier’ than HadCM2 in
many regions and which indicate production losses in North America, Russia
and Eastern Europe.
• Two recent papers illustrate the uncertainty in this area, which is critical to the
global food balance. Alcamo et al. (2003) find that, in relation to climate
impacts on Russia, uncertainty exists as to whether agricultural production
would increase. This is contrary to previous published estimates. For a global
mean temperature increase of 1.3-1.5
o
C their crop production estimates range
from a 9% reduction to a 12% gain. A new assessment was done for US
agriculture by Reilly et al. (2003). It found a more positive aggregate
response to future climate warming than previous estimates, even with very
high warming levels in 2100. As with most such work, the effects of pests
and of extreme events do not appear to have been evaluated and both models
used in the Reilly et al. (2003) work are relatively ‘wet’ for North America
compared to others.
• For a 2.5
o
C warming by the 2080s, the Parry et al. (1999) analysis indicates
45-55 million extra people at risk of hunger. The number at risk rises very
rapidly with increasing temperature.
• With 3
o
C warming by the 2080s, the GAEZ study projects that a very large
number of people, 3.3-5.5 billion, will be living in countries or regions
expected to experience large losses in crop production potential. Results hold
across a range of climate models. The Parry et al. (2001) work places the
number at risk in this temperature range on the order 65-75 million.
• For a 3-4
o
C warming, the upper end of the Parry et al. (2001) analysis, the
additional number at risk are in the range 80-90 million for the HadCM2
scenario and of the order of 125 million under the HadCM3 scenario.
• Global assessments with a full range of the most recent coupled AOGCMs
have yet to be published and there appear to be no recent transient scenarios
used at the global level to assess warming above 3-3.5
o
C warming. There are
few studies at the global level that have included an estimation of the effect of
changes in extreme events or El Niño frequency or intensity (Rosenzweig et
al. 2001).
Water Resources
The impacts of projected climate change on water resources appear to be significant,
with the general picture from the TAR being that existing water stressed regions are
likely to be more stressed in the future as a consequence of climate change. As
previously done, the summary table from the TAR Synthesis Report is reproduced
below (Table 14) and indicates a wide range of effects. The focus here is on water
stress, as this is a key impact projected to affect large numbers of people in the future.
In addition, threshold behaviour is projected as a consequence of the interplay
between climate change effects, socio-economic trends and limits to adaptation
capacity (Arnell 2000; Jones 2000). From Table 14 it can be seen that for many
76
water-distressed regions global mean temperature increases above around 1.5
o
C lead
to decreases in water supply.
Based on the “Millions at Risk” paper of Parry et al. (2001) and that of Arnell et al.
(2002), which use different stabilization levels, a short analysis of the relationship
between increases in global mean temperature and the risk of water shortage is done.
Table 15, from Arnell et al. (2002), summarizes the risks of water shortage for
unmitigated emissions, and stabilization at 550 and 750 ppmv CO
2
for three different
time period – 2020s, 2050s and 2080s – with the associated increase in global mean
temperature above 1861-1990.
105
One of the main messages from this is that after the
2020s the number at risk rises rapidly with temperature and that reduction of the
increase in temperature, at lower stabilization levels reduces the risk substantially.
One of the very interesting aspects of the result of the Parry et al. (2001) analysis is
the way in which risk changes with the projection period. The shape of the
temperature response curves in the 2050s is quite different from that in the 2080s.
Risk rises rapidly with any temperature increase in the 2050s, whilst in the 2080s, risk
initially rises quite slowly (Figure 11). A 1
o
C increase in the 2050s is associated with
an impact almost ten times larger than in the 2080s, whereas the level of risk are
comparable in both periods for a 2
o
C or higher warming (see Table 10). As
temperature increases in the 2080s period from around 1.0
o
C above 1861-1990
106
to
around 2
o
C, the number at risk increases about five fold. One of the major reasons
for this is the increased water scarcity problem for major mega-cities in Asia
estimated for this time period. Table 15 can also be cross-compared with Table 10
and with Figure 11 from the “Millions at Risk” paper of Parry et al. (2001), with this
figure clearly showing a threshold of major increase in risk in the 2080s.
Discussion and Summary
There are several points that should be mentioned and considered when viewing the
results of these assessments. Firstly, the number of people living in water stressed
countries, defined as those using more than 20% of their available resources, is
expected to increase substantially over the next decades irrespective of climate
change. Particularly in the next few decades population and other pressures are likely
to outweigh the effects of climate change (see, for example, the discussion of
Vorosmarty et al. (2000) for the period to 2025), although some regions may be badly
affected during this period (see, for example, the analysis for China for 2030 by
Aiwen (2000)). In the longer term, however, climate change becomes much more
important. Secondly, exacerbating factors such as the link between land degradation,
climate change and water availability are in general not yet accounted for in the
105
The stabilization scenarios are those of Mitchell et al. (2000) with the estimated scenario
temperatures for the three periods concerned tabulated in Table 16. Note that in the paper of Arnell et
al. (2002) these scenarios are reproduced in Figure 3, however the figure caption may be in error
where it states that the temperature increase is with respect to 1990. Examination of the Mitchell et al.
(2000) paper indicates that the increases in Figure 3 of Arnell et al. (2002) are with respect to the
1961-1990 mean, a difference of about 0.3
o
C. In relation to water shortage such a difference in the
2080s corresponds to large changes in affected populations. For example, at 1.5
o
C warming (above
1861-1990) the number at risk is around 600 million. At 1.8
o
C warming this number increases to
around 1,500 million.
106
Note the base period of temperatures for the graph is 1961-1990 and that the offset to 1861-1990 is
around 0.3
o
C based on Folland et al. (2001).
77
global assessments. Available studies on this issue indicate substantial negative
effects for Africa (Feddema 1998; Feddema 1999; Feddema and Freire 2001).
Thirdly, one should be aware that regional impacts in arid and semi-arid areas are
likely to be much larger, relatively, than the aggregate estimates of global
assessments may imply (see for example Ragab and Prudhomme (2002)). Finally, in
relation to the results described, it must be borne in mind that the HadCM2 scenarios,
the primary scenarios used in the Arnell et al. (2002) and Parry et al. (2001) work,
generate much larger impacts in the 2050s than other comparable models (see Table
4-6 of section 4.5.2 of Chapter 4 of the WGII TAR: 413). Hence, although the shape
of the damage functions to be seen in Figure 11 might be correct, the scale of
numbers at risk could be significantly lower.
Interpreting the results discussed above from the point of view of Article 2 may imply
the following:
• 1
o
C of warming or below may still yield high levels of additional risk,
particularly in the period to the 2020s and 2050s, with this risk decreasing due
to the increased economic wealth and higher adaptive capacity projected for
the coming century. For the 2020s, most of the current GCMs imply a level
of risk of additional number of people in water shortage regions in the range
400-800 million for around a 1
o
C warming.
107
• 1.5
o
C of warming produces quite different but nevertheless substantial levels
of risk in the different time periods under the Parry et al. (2001) analysis, with
a peak in the 2050s at over 1,500 million, declining to around 500 million in
the 2080s.
• A major threshold change in risk occurs in the Parry et al. (2001) analysis in
moving from 1.5
o
C to 2-2.5
o
C, with the numbers rising from close to 600
million to between 2.4-3.1 billion. As explained earlier, this is driven by the
water demand of megacities in Indian and China in their model.
• 2
o
C warming and above produces consistently very high levels of additional
risk at all time periods under the HadCM2 scenarios. The range of risk for the
current array of models in the 2050s is in the range 662 million to around 3
billion.
• Above 2.5
o
C warming the level of risk begins to saturate in the range of 3.1-
3.5 billion additional persons at risk.
Clearly one of the major future risks identified in the Parry et al. (2001) and
Arnell et al. (1999; 2002) work is that of increased water demand from mega-
cities in India and China in the 2080s. It is not clear whether or to what extent
additional water resource options would be available for these cities and hence, to
what extent this finding is robust. If such a threshold does exist in reality then its
resolution, absent a reduction in warming, may have broad implications for
environmental flows of water in major rivers of China, India and Tibet should the
mega-cities of India and China seek large scale diversion and impoundments of
flows in the region.
107
See Table 4.6 of the WGII TAR Chapter 4: 213.
78
Table 14 - Water resource effects of climate change
No climate policy interventions. Note: Refer to notes a-d accompanying Table 7 above. Source: Table 3.4
of IPCC TAR SYR: 72.
Table 15 - Population with Potential Increase in Water Stress
Year
or
period
No climate
change
a
(Millions)
Unmitiga
ted
emissions
(Millions)
Temperature
above
1861-1890
S750
(Millions)
Temperature
above
1861-1890
S550
(Millions)
Temperature
above
1861-1890
1990 1710 0.6
o
C 0.6
o
C 0.6
o
C
2020s
5022 338–623 1.2
o
C 242 1.0°C 175 0.8
o
C
2050s
5914 2209–3195 2.2
o
C 2108 1.5
o
C 1705 1.2
o
C
2080s
6405 2831–3436 3.2
o
C 2925 2.0
o
C 762 1.5
o
C
Source: Table II of Arnell et al. (2002): 424 with scenario temperatures added (see Table 16).
a
Number of people in countries using more than 20% of their resources. Increase in stress means a
reduction in resource availability by more than 10%. The range in estimates for the unmitigated
scenario reflects the range between the four ensemble partners.
Effect 2025 2050 2100
CO
2
concentration
a
405–460 ppm 445–640 ppm 540–970 ppm
Global mean temperature
change from the year
1990
b
0.4–1.1°C 0.8–2.6°C 1.4–5.8°C
Global mean temperature
change from the years
1861-1890 (average)
5
1.0-1.7°C
1.4-3.2°C
2.0-6.4°C
Global mean sea-level
rise from the year 1990
b
3–14 cm 5–32 cm 9–88 cm
Water Resource Effects
c
Water supply [WGII
TAR Sections 4.3.6 &
4.5.2]
Peak river flow shifts from
spring toward winter in
basins where snowfall is an
important source of water
(high confidence
d
).
Water supply decreased in
many water-stressed
countries, increased in some
other water- stressed
countries (high
confidence
d
).
Water supply effects
amplified (high
confidence
d
).
Water quality [WGII
TAR Section 4.3.10]
Water quality degraded by
higher temperatures. Water
quality changes modified by
changes in water flow
volume. Increase in
saltwater intrusion into
coastal aquifers due to sea-
level rise (medium
confidence
d
).
Water quality degraded by
higher temperatures (high
confidence
d
).
Water quality changes
modified by changes in
water flow volume (high
confidence
d
).
Water quality effects
amplified (high
confidence
d
).
Water demand [WGII
TAR Section 4.4.3]
Water demand for irrigation
will respond to changes in
climate; higher temperatures
will tend to increase demand
(high confidence
d
).
Water demand effects
amplified (high
confidence
d
).
Water demand effects
amplified (high
confidence
d
).
Extreme events [WGI
TAR SPM; WGII TAR
SPM]
Increased flood damage due
to more intense precipitation
events (high confidence
d
).
Increased drought frequency
(high confidence
d
).
Further increase in flood
damage (high confidence
d
).
Further increase in drought
events and their impacts.
Flood damage several-fold
higher than “no climate
change scenarios.”
79
Table 16 - Scenario Temperatures
Scenario 2020s
o
C
above
1861-1890 average
2050s
o
C
above
1861-1890
average
2080s
o
C
above
1861-1890
average
IS92a 1.2 2.2 3.2
S750 1.0 1.5 2.0
S550 0.8 1.2 1.5
Note: Estimated temperatures for the IS92a, stabilization at 550 and 750 ppmv CO
2
scenarios from
HadCM2 model Mitchell et al. (2000) used in the study by Arnell et al. (2002). See footnote 105.
Socio-economic damages
Owing to the large range of results in the literature, as well methodological issues
such as accounting for risk aversion and distributional issues, the IPCC TAR found it
difficult to reach very firm conclusions on the quantitative estimation of the socio-
economic damages of climate change.
Table 17 reproduces the relevant summary table from the IPCC TAR SYR. Some of
the key, heavily negotiated, agreed
108
conclusions are repeated verbatim below:
109
“The effects of climate change are expected to be greatest in developing countries in
terms of loss of life and relative effects on investment and the economy. For
example, the relative percentage damages to GDP from climate extremes have been
substantially greater in developing countries than in developed countries.” (WGII
TAR SPM Section 2.8)
“More people are projected to be harmed than benefited by climate change, even for
global mean temperature increases of less than a few degrees (low confidence).”
“Notwithstanding the limitations expressed above, based on a few published
estimates, increases in global mean temperatures would produce net economic losses
in many developing countries for all magnitudes of warming studied (low
confidence), and losses would be greater in magnitude the higher the level of
warming (medium confidence).”
“In contrast an increase in global mean temperature of up to a few degrees Celsius
would produce a mixture of economic gains and losses in developed countries (low
confidence), with economic losses for larger temperature increases (medium
confidence).”
110
108
The texts of Summaries for Policy Makers are negotiated line by line by Governments and, in effect, usually reflect a
consensus between different governmental views and those of the IPCC Convening Lead Authors present. It has rarely
happened that the IPCC Chair has permitted a conclusion to be changed by governments to the extent that CLAs present can no
longer associate themselves with it. More commonly contested conclusions are reduced in specificity, generalized or plain
‘watered down’ under pressure from governments who disagree with the drafts prepared by the lead authors and reviewed three
times by governments and experts. In the case of the conclusions cited here some governments vigorously contested the drafts
prepared by the Lead Authors. After lengthy negotiations the final text is different from that proposed by the Lead Authors, with
the final emphasis on the mixture of economic losses and gains reflecting a feeling that presenting net aggregate figures was
misleading as it did not say who would benefit and who would lose. From the studies cited, it was clear than even for low levels
of warming there were developed countries that would suffer net losses and within countries significant sectors would lose
whilst others gained.
109
Quotes are from the TAR SYR unless noted otherwise.
110
The proposed text from the lead authors originally attempted to include temperatures in this statement, however as noted
earlier such references were deleted. Originally the text said: “In many developed countries, net economic gains are projected
for global mean temperature increases up to roughly 2
o
C (medium confidence). Mixed or neutral net effects are projected in
80
“The projected distribution of economic impacts is such that it would increase the
disparity in well-being between developed countries and developing countries, with
disparity growing for higher projected temperature increases (medium confidence).
The more damaging impacts estimated for developing countries reflects, in part their
lesser adaptive capacity relative to developed countries [7.2.3]” (WGII TAR SPM
Section 2.8).
Chapter 19 of IPCC TAR WGII has a very useful discussion on the estimation of
economic damages and benefits to which the reader is referred,
111
some of the key
points are:
• Impact estimates are highly sensitive to inequity aversion or risk aversion
assumptions, with the greater the aversion to risk or inequity the higher the
estimated damage costs (see, for example, Tol (2001)).
• Current market estimates of damage are lower than in the Second Assessment
Report due to the inclusions of better adaptation estimates.
• Non-market damages are likely to be quite high.
• Global aggregate estimates are very sensitive to the weights given to different
regions (see, for example, the discussions in Fankhauser and Tol (1997; 1998)
and Azar (1999)).
• The shape of the climate damage function in relation to future temperature
change is quite uncertain. Whether or not the damages from climate change
rise rapidly or slowly with increasing temperature is a quite fundamental
concern for policy.
Figure 13 below (reproduced from IPCC WGII TAR Chapter 19) shows a range of
estimates for global aggregate damages and gives an impression of the uncertainty
and divergence in global aggregate damage estimates. The curve of Mendelsohn et
al. (2000) is essentially flat up to a global mean warming of 6
o
C warming, which is
quite interesting given that this a change in magnitude the same as, but 60-120 times
faster than, the transition from a full glacial to interglacial climate. Tol’s curves show
quite different shapes depending on the equity assumptions underlying the global
aggregation of regional damage estimates (Tol 2002). Equity weighted estimates
almost completely eliminate the substantial benefits for a warming in the range of 1-
3
o
C, with net damages above about 1.5
o
C warming and a roughly linear increase in
damages thereafter. The Nordhaus and Boyer (2000) curves are quite non-linear and
show steepening damages as temperature rises.
Another way to look at this issue is with respect to the question of what are the ranges
of temperature associated with reductions in world GDP. From the examination of
the Nordhaus and Tol model outcomes (depicted in Figure 13), it is estimated that a
1% reduction may occur within a 2.5-3.6
o
C warming or possibly never, a 2%
reduction for temperatures between 3.2-6.5
o
C or possibly never and a 5% reduction
for temperature increases between 4.6-5.6
o
C.
developed countries for temperature increases in the approximate range of 2 to 3
o
C, and net losses for larger temperature
increases (medium confidence).” Whilst the confidence interval placed on the first conclusion was changed after referral back to
the original literature, the temperature references were changed only to reflect the decision described in footnote 98.
111
See IPCC TAR WGII Chapter 19: 941-5.
81
The pattern of regional impacts seems to have greater consistency than the global
aggregates. Most of these studies show that Africa and India lose significantly from a
small warming. Tol (2002) estimates an African loss of about 4% of GDP for a 1
o
C
warming (over 1990) and about a 1.7% loss for India.
112
Similarly Nordhaus and
Boyer (2000) indicate substantial losses (4-5%) for their benchmark warming of
2.5
o
C (inferred to be above 1900). In one out of two broad scenarios, Mendelsohn et
al. (2000) estimate smaller losses for a 2
o
C warming (0.25-0.5%) for India, and in the
range 0.5-1% loss, or greater for much of Africa. Their other scenarios find benefits
in most places. It is worth noting that under their model, the very large benefits to
Russian agriculture computed for a 2
o
C warming tend to be decisive in determining
the global aggregate effects at this level of warming.
Finally, it appears that very few analyses have been conducted across the full range of
uncertainties in the parameters and assumptions underlying model based estimates of
climate damages and benefits. Recently Tol (2003) explored this issue, publishing a
Monte Carlo simulation for his FUND model, which was used to compute the damage
curve described above. He found a non-zero probability of very negative effects in
some regions. It is worth repeating some of his own discussion, coming, as it does,
from one of the more prominent modellers in this area:
“Suppose that climate change is worse than expected. Suppose that the impacts of
climate change are worse than expected. Suppose that a lot of money needs to be
spent on building sea walls and curing malaria. Suppose that agricultural yields are
disappointing and storms and floods damage roads and houses. In a fragile
economy, this means that economic growth is halted. It means that investment and
past savings are diverted from enhancing productivity and preventing further
havoc to restoring damage. It means that the economy grows more fragile. It
means that climate change can do even more damage, making the economy yet
more fragile” (281).
“Can climate change cause a poverty trap? Recurring natural disasters can
definitely contribute to poverty trap … Estimates of the impact of climate change
suggest that they can be worth a couple of percent of GDP, particularly in poor
regions. Climate change seems likely to cause poverty traps in some places, and
with some non-negligible change at a regional scale” (281).
Discussion and Summary
From the point of view of Article 2 and its interpretation, it seems that the tentative
conclusions one might draw from the above would include:
• For a 1
o
C warming a significant number of developing countries appear likely
to experience net losses, which range as high as a few % of GDP, whilst most
developed countries are likely to experience a mix of damages and benefits,
with net benefits predicted by a number of models.
• For a 2
o
C warming the net adverse effects projected for developing countries
appear to be more consistent and of the order of a few to several percentage
points of GDP depending upon the model. Regional damages for some
112
See Table VII of Tol (2002).
82
developing countries and regions, particularly in Africa, may exceed several
percentage points of GDP.
• Above 2
o
C the likelihood of global net damages increases but at a rate that is
quite uncertain. Apart from the results of Mendelsohn et al (2000), the effects
on several developing regions in the literature appear to be in the range of 3-
5% for a 2.5-3
o
C warming, if there are no adverse climate surprises. Global
damage estimates are in the range of 1-2% for 2.5-3
o
C warming, with some
estimates increasing substantially with increasing temperature. If major
identified risks such as thermohaline shutdown or non-linear feedbacks in the
carbon cycle eventuate, then the damages could be very high. Regionally,
there is very little evidence that the pattern of increasing damages to many
developing countries would reverse and most indicates a continuing increase
in net damages. Africa seems to be consistently amongst the regions with
high to very high projected damages.
83
Table 17 - Other market sector effects of climate change
Effects
2025 2050 2100
CO
2
concentration
a
405–460 ppm 445–640 ppm 540–970 ppm
Global mean temperature
change from the year 1990
b
0.4–1.1°C 0.8–2.6°C 1.4–5.8°C
Global mean temperature
change from the years 1861-
1890 (average)
5
1.0-1.7°C
1.4-3.2°C
2.0-6.4°C
Global mean sea-level rise
from the year 1990
b
3–14 cm 5–32 cm 9–88 cm
Other Market Sector
Effects
c
Energy [WGII TAR Section
7.3]
Decreased energy
demand for heating
buildings (high
confidence
d
).
Increased energy demand
for cooling buildings
(high confidence
d
).
Energy demand effects
amplified (high
confidence
d
).
Energy demand effects
amplified (high
confidence
d
).
Financial sector [WGII
TAR´Section 8.3]
Increased insurance
prices and reduced
insurance availability
(high confidence
d
).
Effects on financial
sector amplified.
Aggregate market effects
e
[WGII TAR Sections 19.4-
5]
Net market sector losses
in many developing
countries (low
confidence
d
).
Mixture of market gains
and losses in developed
countries (low
confidence
d
).
Losses in developing
countries amplified
(medium confidence
d
).
Gains diminished and
losses amplified in
developed countries
(medium confidence
d
).
Losses in developing
countries amplified
(medium confidence
d
).
Net market sector
losses in developed
countries from warming
of more than a few °C
(medium confidence
d
).
No climate policy interventions. Refer to footnotes a-d accompanying Table 7 and footnote e. Aggregate market
effects represent the net effects of estimated economic gains and losses summed across market sectors such as
agriculture, commercial forestry, energy, water, and construction. The estimates generally exclude the effects of
changes in variability and extremes, do not account for the effects of different rates of change, and only partially
account for impacts on and services that are not traded in markets. These omissions are likely to result in
underestimates of economic losses and overestimates of economic gains. Estimates of aggregate impacts are
controversial because they treat gains for some as cancelling losses for others and because the weights that are
used to aggregate across individuals are necessarily subjective. Source: Table 3-5 of IPCC TAR SYR: 74.
84
Figure 13 - Climate Damages or Benefits as a Function of Temperature
Note: This figure is drawn from figure 19-4 IPCC WGII TAR Chapter 19 and shows several examples
of aggregated global monetary damage functions as a percentage of world GDP from three prominent
economists. ‘Output’, ‘population’ and ‘equity’ refer to the weightings used in making the aggregate
assessments. See text for references. Note that temperature increase is with respect to 1990. One
needs to add 0.6
o
C to convert to an estimate of increases above 1861-1890.
4. Summary and Conclusions
The results of the foregoing analysis are difficult to synthesize into a simple picture
without losing many of the caveats and qualifications required. Nevertheless an
attempt will be made here but the reader is urged to also examine the underlying
arguments in the preceding sections. The summary here will focus on conclusions
that can be drawn in relation to projected impacts at different levels of global mean
temperature increase above 1861-1890, which is here used as the proxy for the pre-
industrial climate. Bear in mind that there has been a global mean warming of around
0.6oC since that time.
Ecosystems impacts
Impacts on coastal wetlands
• Below a 1
o
C increase the risk of damage is low for most systems.
• Between 1 and 2
o
C warming moderate to large losses appear likely for a few
vulnerable systems. Of most concern are threats to the Kakadu wetlands of
85
northern Australia and the Sundarbans of Bangladesh, both of which may
suffer 50% losses at less than 2
o
C and are both on the he UNESCO World
Heritage List.
• Between 2-3
o
C, it is possible that the Mediterranean, Baltic and several
migratory bird habitats in the US experience a 50% or more loss. In this
range it seems likely that there could be the complete loss of Kakadu and the
Sundarbans.
A key issue is the inertia of sea level rise, which makes the assignment of risk to
different temperature levels misleading. Should, for example, sea level rise by 30cm
in the coming decades to a century (threatening Kakadu for example), the thermal
inertia of the ocean is such that an ultimate sea level rise of 2-4 times this amount
may be inevitable even if temperature stops rising. The prognoses for wetlands in
this context is not clear, as many damages are linked to the rate of sea level rise
compared to the accretion and/or migratory capacity of the system. A major
determinant of the latter will be human activity adjacent to, or in the inland
catchments of the wetland system.
Impacts on animal species
• Below 1
o
C warming, there appears to be a risk of extinction for some highly
vulnerable species in south-western Australia and to a lesser extent in South
Africa. Range losses for species such as the Golden Bower bird in the
highland tropical forests of North Queensland Australia and for many animal
species in South Africa are likely to become significant and observable.
• Between 1 and 2
o
C warming, large and sometimes severe impacts appear
possible for some Salmonid fish habitats in the USA, the Collared Lemming
in Canada, many South African animals and for Mexico’s fauna. Extinctions
in the Drayandra forest of south-western Australia seem very likely. There is
an increasing risk of this in South Africa, Mexico for the most vulnerable
species and for especially vulnerable highland rainforest vertebrates in North
Queensland, Australia. Mid summer ice reduction in the Arctic ocean seems
likely to be at a level that would cause major problems for polar bears at least
at a regional level.
• Between 2-3
o
C large to severe impacts appear likely and there are several
predicted extinctions in the literature. These include towards the upper end of
this temperature increase range several Hawaiian Honeycreepers, the Golden
Bower bird of the highland tropical rainforest of North Queensland Australia
and four species in South Africa. In general large impacts are projected for
Mexican fauna, many South African animals, the Collared Lemming in the
Arctic (which would have broad implications for arctic ecosystems), Salmonid
fish in Wyoming. Perhaps the most worrying projection is for the
“catastrophic loss” of endemic rainforest vertebrates of the highland
rainforests of North Queensland, Australia. These would be experiencing
rapidly increasing losses of their core environments in the upper end of this
warming range.
86
• Above 3oC, large impacts begin to emerge for waterfowl populations in the
Prairie Pothole region in the USA. In the Arctic the collared lemming range
is reduced by 80%, very large reductions are projected for Arctic sea ice cover
particularly in summer that is likely to further endanger polar bears.
Extinction of the Golden Bower bird is predicted in this temperature range and
there seems to be a very high likelihood that large numbers of extinctions
would occur amongst the 65 endemic vertebrates of the highland rainforests of
North Queensland, Australia. In Mexico very severe range losses for many
animals are projected, as is the case also in South Africa, with Kruger national
park projected to lose two thirds of the animals studied.
Impacts on ecosystems
• Between present temperatures and a 1
o
C increase, three ecosystems appear to
be moving into a high risk zone - coral reefs, the highland tropical forests in
Queensland, Australia, the Succulent Karoo in South Africa. Increased fire
frequency and pest outbreaks may cause disturbance in boreal forests and
other ecosystems.
• Between 1-2
o
C the Australian highland tropical forest, the Succulent Karoo
biodiversity hot spot, coral reef ecosystems and some Arctic and alpine
ecosystems are likely to suffer large or even severe damage. The Fynbos of
South Africa will experience increased losses. Coral reef bleaching will likely
become much more frequent, with slow or no recovery, particularly in the
Indian Ocean south of the equator. Australian highland tropical forest types,
which are home to many endemic vertebrates, are projected to halve in area in
this range. The Australian alpine zone is likely to suffer moderate to large
losses and the European Alpine may be experiencing increasing stress. The
substantial loss of Arctic sea ice likely to occur will harm ice dependent
species such as the polar bears and walrus. Increased frequency of fire and
insect pest disturbance is likely to cause increasing problems for ecosystems
and species in the Mediterranean region. Moderate to large losses of boreal
forest in China can be expected. Moderate shifts in the range of European
plants can be expected and in Australia moderate to large number of Eucalypts
may be outside out of their climatic range.
• Between 2-3
o
C coral reefs are projected to bleach annually in many regions.
At the upper end of this temperature band, the risk of eliminating the
Succulent Karoo and its 2800 endemic plants is very high. Moderate to large
reductions in the Fynbos can be expected, with the risk of significant
extinctions. Australian mainland alpine ecosystems are likely to be on the
edge of disappearance. European alpine systems will at or above their
anticipated tolerable limits of warming with some vulnerable species close to
extinction. Severe loss of boreal forest in China is projected and large and
adverse changes are also projected for many systems on the Tibetan plateau.
Large shifts in the range of European plants seem likely and a large number of
Eucalypt species may expect to lie outside of their present climatic range.
Moderate to large effects are projected for Arctic ecosystems and boreal
87
forests. Within this temperature range there is a likelihood of the Amazon
forest suffering potentially irreversible damage leading to its collapse.
Agriculture and food security impacts
Many studies indicate that developing countries are likely to lose as a whole, relative
to the developed nations. India, for example, is projected to experience significant
losses, with quite large areas of current cropland losing significant productivity. At
all levels of warming, a large group of the poor, highly vulnerable developing
countries is expected to suffer increasing food deficits. It is anticipated that this will
lead to higher levels of food insecurity and hunger in these countries. Developed
countries will not be immune to large effects of climate change on their agricultural
sectors.
• Warming of around 1
o
C produces relatively small damages when measured
from the point of increased risk of hunger and/or under nourishment (around
10 million more at risk) over the next century. In this temperature range
nearly all developed countries are projected to benefit, whilst many
developing countries in the tropics are estimated to experience small but
significant crop yield growth declines relative to an unchanged climate.
• Warming from 1
o
C to 2
o
C warming triples the number of people at risk of
hunger in the 2080s.
• Between 2-3oC warming the risk of damage begins to increase significantly.
Whilst developing countries may still gain in this temperature range the
literature indicates that production is finely balanced in this temperature range
between the effects of increased temperature and changes in precipitation.
‘Drier’ models show losses in North America, Russia and Eastern Europe
whereas ‘wetter’ models show increases. One study shows rapidly rising
hunger risk in this temperature range with 45-55 million extra people at risk of
hunger by the 2080s for 2.5oC warming which rises to 65-75 million for a
3oC warming. Another study shows that a very large number of people, 3.3-
5.5 billion, may be living in countries or regions expected to experience large
losses in crop production potential at 3oC warming.
• For a 3-4
o
C warming, in one study the additional number at risk of hunger is
estimated to be in the range 80-125 million depending on the climate model.
In Australia a warming of the order of 4
o
C is likely to put entire regions out of
production, with lesser levels of warming causing substantial declines in the
west and the south.
Water impacts
The number of people living in water stressed countries, defined as those using more
than 20% of their available resources, and is expected to increase substantially over
the next decades irrespective of climate change. Particularly in the next few decades
population and other pressures are likely to outweigh the effects of climate change,
although some regions may be badly affected during this period. In the longer term,
however, climate change becomes much more important. Exacerbating factors such
88
as the link between land degradation, climate change and water availability are in
general not yet accounted for in the global assessments.
• Around 1
o
C of warming may entail high levels of additional risk in some
regions, particularly in the period to the 2020s and 2050s, with this risk
decreasing due to the increased economic wealth and higher adaptive capacity
projected for the coming century. For the 2020s the additional number of
people in water shortage regions is estimate to be in the range 400-800.
• Between 1-2
o
C warming the level of risk appears to depend on the time frame
and assumed levels of economic development in the future. One study for the
middle of this temperature range has a peak risk in the 2050s at over 1,500
million, which declines to around 500 million in the 2080s.
• Over 2oC warming appears to involve a major threshold increase in risk. One
study shows risk increasing for close to 600 million people at 1.5oC to 2.4-3.1
billion at around 2.5
o
C. This is driven by the water demand of mega-cities in
India and China in their model. In this study the level of risk begins to
saturate in the range of 3.1-3.5 billion additional persons at risk at 2.5-3oC
warming.
One of the major future risks identified by two studies is that of increased water
demand from mega-cities in India and China. It is not clear whether or to what
extent additional water resource options would be available for these cities and
hence, to what extent this finding is robust. This may have broad implications for
environmental flows of water in major rivers of China, India and Tibet should the
mega-cities of India and China seek large-scale diversion and impoundments of
flows in the region.
Socio-economic effects
• For a 1
o
C warming a significant number of developing countries appear likely
to experience net losses, which range as high as a few % of GDP. Most
developed countries are likely to experience a mix of damages and benefits,
with net benefits predicted by a number of models.
• For a 2
o
C warming the net adverse effects projected for developing countries
appear to be more consistent and of the order of a few to several percentage
points of GDP depending upon the model. Regional damages for some
developing countries and regions, particularly in Africa, may exceed several
percentage points of GDP.
• Above 2
o
C the likelihood of global net damages increases but at a rate that is
quite uncertain. The effects on several developing regions appear to be in
the range of 3-5% for a 2.5-3
o
C warming, if there are no adverse climate
surprises. Global damage estimates are in the range of 1-2% for 2.5-3
o
C
warming, with some estimates increasing substantially with increasing
temperature.
If major identified risks such as thermohaline shutdown or non-linear feedbacks
in the carbon cycle eventuate, then the damages could be very high. Regionally,
there is very little evidence that the pattern of increasing damages to many
developing countries would reverse and most indicates a continuing increase in
89
net damages. Africa seems to be consistently amongst the regions with high to
very high projected damages.
Conclusions
It seems clear from this partial review of the available literature that the risks arising
from projected human induced climate change increase significantly with increasing
temperature. Below a 1oC increase the level of risk are low but in some case not
insignificant particularly for highly vulnerable ecosystems. In the 1-2oC-increase
range risks across the board increase significantly and at a regional level are often
substantial. Above 2oC the risks increase very substantially involving potentially
large extinctions or even ecosystem collapses, major increases in hunger and water
shortage risks as well as socio-economic damages, particularly in developing
countries.
90
5. Appendix: Temperature scale
The analysis in this paper focuses on determining the projected impacts of climate
change associated with increases above the pre-industrial average global temperature
(approximated by 1851-1880 average). One of the main reasons for using this
baseline is that the temperature limit of the WBGU tolerable window is specified with
respect to the pre-industrial atmosphere.
In most cases the IPCC has used the base year of 1990 for its analyses of projected
effects, whereas the literature uses a variety of different base periods or years (for
example, the 1961-1990 climatology is often used). By and large the projected
temperature increases presented in the TAR are with respect to 1990, which is
thought to be about 0.6°C warmer than the pre-industrial average (Folland et al.
2001). Most contemporary GCM scenarios start from the pre-industrial period and
hence their changes in climate, in effect, are with respect to the assumed state of the
pre-industrial atmosphere and climate system. Often, the changes in climate statistics
computed by these models are reported with the respect to a standard 30-year mean
climatology of 1961-1990. A thirty-year averaging period is used as this eliminates
most of the year-to-year variability in global mean temperature. Many recent impact
studies are based on the change in the GCM climatology between a future period e.g.
2050s (30 year average around 2050) and the 1961-1990 average climatology from
the model, applied to the observational mean of the 1961-1990 period. The 1961-
1990 average temperature is about 0.3°C less than the 1990 global average
temperature.
To further complicate matters, rather than specify a temperature range for an impact
many of the IPCC TAR assessments from Working Group II are reported with respect
to a temperature range classification of “small” (less than 2°C above 1990), medium
(2-3°C above 1990) and “large” (greater than 3°C above 1990).
113
It should be noted
with respect to the latter that the IPCC states that a “2°C warming from 1990 to 2100
would be a magnitude of warming greater than any that human civilization has ever
experienced. Thus, “small” does not necessarily mean negligible.”
114
As a consequence of these and other factors, there often needs to be a conversion
from the reported impact or effect (which may also be estimated with respect to a
local temperature increase against a different base period) to a scale with respect to
pre-industrial. Table 18 outlines the overall scales used to convert the different base
periods or classifications mentioned above to a pre-industrial temperature increase.
113
It should be noted that this was in response to pressure principally from the Saudi Arabian
delegation at the IPCC Working Group II Plenary in Geneva in February of 2001 and was resisted by
most of the lead authors of the relevant chapters of Working Group II. In the end, after many hours of
negotiation, the contact group chair, then IPCC Chair Bob Watson, concluded that a consensus could
only be reached with the adoption of the above classification.
114
Chapter 19 of the IPCC WGII TAR: 957.
91
Table 18 - Global Temperature Scales used in this Report
Increase above
pre-industrial
temperature
°C
Increase
above 1990
temperature
°C
Above 1961-
1990 average
temperature
°C
IPCC TAR
classification
0.0 -0.6 -0.3 Small
1.0 0.4 0.7 Small
1.6 1.0 1.3 Small
2.0 1.4 1.7 Small
2.6 2.0 2.3 Medium
3.0 2.4 2.7 Medium
3.6 3.0 3.3 Medium
4.0 3.4 3.7 Large
4.6 4.0 4.3 Large
5.0 4.4 4.7 Large
5.6 5.0 5.3 Large
6.0 5.4 5.7 Large
6.6 6.0 6.3 Large
7.0 6.4 6.7 Large
7.6 7.0 7.3 Large
8.0 7.4 7.7 Large
Placing sea level rise on a common scale is an altogether more complex task and
there has been no attempt here to do this in a general sense. Each example where an
impact is linked to a specific sea level rise is taken on its own and where possible
converted to global mean temperature range using available information as described
in a footnote or textual explanation.
92
6. References
ADB (1994). Climate change in Asia: Bangladesh country report. Vol 2: ADB
regional study on global environmental issues. Manila, Philippines, Asian
Development Bank:
Aiwen, Y. (2000). "Impact of Global Climate Change on China's Water Resources."
Environmental Monitoring and Assessment 61(1): 187-191.
Alcamo, J., N. Dronin, M. Endejan, G. Golubev, and A. Kirilenko (2003). Will
climate change affect food and water security in Russia? Summary report of
the international project on global environmental change and its threat to food
and water security in russia. Draft. Kassel, Germany, University of Kassel,
Center for Environmental Systems Research: 20.
Arnell, N. W. (1999). "Climate change and global water resources - a new
assessment." Global Environmental Change 9(1001): 31-49.
— (2000). "Thresholds and response to climate change forcing: The water sector."
Climatic Change 46(3): 305-316.
Arnell, N. W., M. G. R. Cannell, M. Hulme, R. S. Kovats, J. F. B. Mitchell, R. J.
Nicholls, M. L. Parry et al. (2002). "The consequences of CO2 stabilisation
for the impacts of climate change." Climatic Change 53(4): 413-446.
Ayres, M. P., and M. J. Lombardero (2000). "Assessing the consequences of global
change for forest disturbance from herbivores and pathogens." Science of the
Total Environment 262(3): 263-286.
Azar, C. (1999). "Weight factors in cost-benefit analysis of climate change."
Environmental & Resource Economics 13(3): 249-268.
Bakkenes, M., J. R. M. Alkemade, F. Ihle, R. Leemans, and J. B. Latour (2002).
"Assessing effects of forecasted climate change on the diversity and
distribution of European higher plants for 2050." Global Change Biology 8(4):
390-407.
Bank, W. (2000). Bangladesh: Climate Change & Sustainable Development. Dhaka,
World Bank Office, South Asia Rural Development Team:
Bartlein, P. J., C. Whitlock, and S. L. Shafter (1997). "Future climate in the
Yellowstone National Park region and its potential impact on vegetation."
Conservation Biology 11(3): 782-792.
Bayliss, B. L., K. G. Brennan, I. Eliot, C. M. Finlayson, R. N. Hall, T. House, R. W.
J. Pidgeon et al. (1997). Vulnerability Assessment of Predicted Climate
Change and Sea Level Rise in the Alligator Rivers Region, Northern
Territory, Australia. Canberra, Australia, Supervising Scientist Report 123:
134.
Beaumont, L. J., and L. Hughes (2002). "Potential changes in the distributions of
latitudinally restricted Australian butterfly species in response to climate
change." Global Change Biology 8(10): 954-971.
Benning, T. L., D. LaPointe, C. T. Atkinson, and P. M. Vitousek (2002). "Interactions
of climate change with biological invasions and land use in the Hawaiian
Islands: Modeling the fate of endemic birds using a geographic information
system." PNAS 99(22): 14246-14249.
Boer, G. J., N. A. McFarlane, and M. Lazare (1992). "Greenhouse Gas Induced
Climate Change Simulated with the CCC 2nd-Generation General-Circulation
Model." Journal of Climate 5(10): 1045-1077.
93
Bond, W. J., and G. F. Midgley (2000). "A proposed CO2-controlled mechanism of
woody plant invasion in grasslands and savannas." Global Change Biology
6(8): 865-869.
Both, C., and M. E. Visser (2001). "Adjustment to climate change is constrained by
arrival date in a long-distance migrant bird." Nature 411(6835): 296-298.
Brereton, R., S. Bennett, and I. Mansergh (1995). "Enhanced greenhouse climate
change and its potential effect on selected fauna of South-Eastern Australia: A
trend analysis." Biological Conservation 72(3): 339-354.
Brovkin, V., S. Levis, M. F. Loutre, M. Crucifix, M. Claussen, A. Ganopolski, C.
Kubatzki et al. (2003). "Stability analysis of the climate-vegetation system in
the northern high latitudes." Climatic Change 57(1-2): 119-138.
Busby, J. R. (1988). "Potential implications of climate change on Australia's flora and
fauna" in Pearman, G. I., ed. Greenhouse: Planning for Climate Change.
Melbourne, CSIRO Division of Atmospheric Research: 387-388.
Carter, T. R., E. L. La Rovere, R. N. Jones, R. Leemans, L. O. Mearns, N.
Nakicenovic, A. B. Pittock et al. (2001). "Chapter 3: Developing and
Applying Scenarios" in Climate Change 2001: Impacts, adaptation and
vulnerability. Working Group II of the Intergovernmental Panel on Climate
Change. Cambridge, UK, Cambridge University Press: 145-190.
Chapin, F. S., E. S. Zavaleta, V. T. Eviner, R. L. Naylor, P. M. Vitousek, H. L.
Reynolds, D. U. Hooper et al. (2000). "Consequences of changing
biodiversity." Nature 405(6783): 234-242.
Clark, M. E., K. A. Rose, D. A. Levine, and W. W. Hargrove (2001). "Predicting
climate change effects on Appalachian trout: Combining GIS and individual-
based modeling." Ecological Applications 11(1): 161-178.
Cochrane, M. A., and W. F. Laurance (2002). "Fire as a large-scale edge effect in
Amazonian forests." Journal of Tropical Ecology 18: 311-325.
Cohen, S., K. Miller, K. Duncan, E. Gregorich, P. Groffman, P. Kovacs, V. Magaña
et al. (2001). "Chapter 15: North America" in Climate Change 2001: Impacts,
adaptation and vulnerability. Working Group II of the Intergovernmental
Panel on Climate Change. Cambridge, UK, Cambridge University Press: 735-
800.
Colinvaux, P. A., P. E. DeOliveira, and M. B. Bush (2000). "Amazonian and
neotropical plant communities on glacial time-scales: The failure of the aridity
and refuge hypotheses." Quaternary Science Reviews 19(1/5): 141.
Comiso, J. C. (2002a). "Correlation and trend studies of the sea-ice cover and surface
temperatures in the Arctic" in Annals of Glaciology, Vol 34, 2002. Annals of
Glaciology. Cambridge, INT GLACIOLOGICAL SOC: 420-428.
— (2002b). "A rapidly declining perennial sea ice cover in the Arctic." Geophysical
Research Letters 29(20): art. no.-1956.
Cowling, S. A., P. M. Cox, R. A. Betts, V. J. Ettwein, C. D. Jones, M. A. Maslin, and
S. A. Spall (2003). "Contrasting simulated past and future responses of the
Amazon rainforest to atmospheric change." Philosophical Transactions of the
Royal Society of London: in press.
Cowling, S. A., M. A. Maslin, and M. T. Sykes (2001). "Paleovegetation simulations
of lowland Amazonia and implications for neotropical allopatry and
speciation." Quaternary Research 55(2): 140-149.
Cowlishaw, G. (1999). "Predicting the pattern of decline of African primate diversity:
an extinction debt from historical deforestation." Conservation Biology 13(5):
1183-1193.
94
Cox, P. M., R. A. Betts, M. Collins, P. Harris, C. Huntingford, and C. D. Jones
(2003). Amazon dieback under climate-carbon cycle projections for the 21st
century. UK, Hadley Centre:
Cox, P. M., R. A. Betts, C. D. Jones, S. A. Spall, and I. J. Totterdell (2000).
"Acceleration of global warming due to carbon-cycle feedbacks in a coupled
climate model." Nature(6809): 184-186.
Cramer, W., A. Bondeau, W. Lucht, and S. Schaphoff (2003). "Tropical forests and
the global carbon cycle: Simulations for 20th century reconstruction and
future scenarios." Philosophical Transactions of the Royal Society of London:
in press.
Cramer, W., A. Bondeau, F. I. Woodward, I. C. Prentice, R. A. Betts, V. Brovkin, P.
M. Cox et al. (2001). "Global response of terrestrial ecosystem structure and
function to CO2 and climate change: results from six dynamic global
vegetation models." Global Change Biology 7(4): 357-373.
CSIRO (1996). Climate change scenarios for the Australian region. Melbourne,
Climate Impact Group, CSIRO Division of Atmospheric Research: 8.
— (2001). Climate Change: Projections for Australia. Melbourne, Climate Impact
Group, CSIRO Division of Atmospheric Research: 8.
Darwin, R., and D. Kennedy (2000). "Economic effects of CO2 fertilization of crops:
transforming changes in yield into changes in supply." Environmental
Modeling & Assessment 5(3): 157-168.
Davis, M. B., and R. G. Shaw (2001). "Range shifts and adaptive responses to
Quaternary climate change." Science 292(5517): 673-679.
Donnelly, J. P., and M. D. Bertness (2001a). "Rapid shoreward encroachment of salt
marsh cordgrass in response to accelerated sea-level rise." PNAS 98(25):
14218-14789.
— (2001b). "Rapid shoreward encroachment of salt marsh cordgrass in response to
accelerated sea-level rise." Proceedings of the National Academy of Sciences
of the United States of America 98(25): 14218-14223.
Easterling, D. R., G. A. Meehl, C. Parmesan, S. A. Changnon, T. R. Karl, and L. O.
Mearns (2000). "Climate Extremes: Observations, Modeling, and Impacts."
Science 289(5487): 2068-2074.
Eliot, I., C. M. Finalyson, and P. Waterman (1999). "Predicted climate change, sea-
level rise and wetland management in the Australian wet-dry tropics."
Wetlands ecology and management.- 7(1/2): 63.
Erasmus, B. F. N., A. S. Van Jaarsveld, S. L. Chown, M. Kshatriya, and K. J. Wessels
(2002). "Vulnerability of South African animal taxa to climate change."
Global Change Biology 8(7): 679-693.
European Community (1996). Climate Change - Council conclusions 8518/96 (Presse
188-G) 25/26. VI.96, pers. comm.
Fahnestock, M. A., W. Abdalati, and C. A. Shuman (2002). "Long melt seasons on
ice shelves of the Antarctic Peninsula: an analysis using satellite-based
microwave emission measurements" in Annals of Glaciology, Vol 34, 2002.
Annals of Glaciology. Cambridge, INT GLACIOLOGICAL SOC: 127-133.
Fankhauser, S., and R. S. J. Tol (1997). "The Social Costs of Climate Change: The
IPCC Second Assessment Report and Beyond." Mitigation and Adaptation
Strategies for Global Change 1(4): 385-403.
— (1998). "The Value of Human Life in Global Warming Impacts - a Comment."
Mitigation and Adaptation Strategies for Global Change 3(1): 87-88.
95
Feddema, J. J. (1998). "Estimated impacts of soil degradation on the African water
balance and climate." Climate Research 10(2): 127-141.
— (1999). "Future African water resources: Interactions between soil degradation and
global warming." Climatic Change 42(3): 561-596.
Feddema, J. J., and S. Freire (2001). "Soil degradation, global warming and climate
impacts." Climate Research 17(2): 209-216.
Fischer, G., M. Shah, H. van Velthuizen, and F. Nachtergaele (2001). Global agro-
ecological assessment for agriculture in the 21st century. Laxenburg, Austria,
IIASA: 33.
Fischer, G., H. van Velthuizen, M. Shah, and F. Nachtergaele (2002). Global agro-
ecological assessment for agriculture in the 21st century: methodology and
results. Laxenburg, Austria, IIASA: 119.
Folland, C., and C. Anderson (2002). "Estimating changing extremes using empirical
ranking methods." Journal of Climate 15(20): 2954-2960.
Folland, C. K., N. A. Rayner, S. J. Brown, T. M. Smith, S. S. P. Shen, D. E. Parker, I.
Macadam et al. (2001). "Global temperature change and its uncertainties since
1861." Geophysical Research Letters 28(13): 2621-2624.
Friedlingstein, P., L. Bopp, P. Ciais, J. L. Dufresne, L. Fairhead, H. LeTreut, P.
Monfray et al. (2001). "Positive feedback between future climate change and
the carbon cycle." Geophysical Research Letters 28(8): 1543-1546.
Galbraith, H., R. Jones, R. Park, J. Clough, S. Herrod-Julius, B. Harrington, and G.
Page (2002). "Global climate change and sea level rise: Potential losses of
intertidal habitat for shorebirds." Waterbirds 25(2): 173-183.
Gitay, H., S. Brown, W. Easterlin, and B. Jallow (2001). "Chapter 5: Ecosystems and
Their Goods and Services" in Climate Change 2001: Impacts, Adaptation and
Vulnerability. Working Group II of the Intergovernmental Panel on Climate
Change. Cambridge, UK, Cambridge University Press: 237-342.
Gitay, H., A. Suárez, R. Watson, and D. J. Dokken, eds (2002). Climate change and
biodiversity. IPCC Technical Paper V. Geneva, IPCC: 77.
Green, K., and C. M. Pickering (2002). "A potential scenario for mammal and bird
diversity in the Snowy Mountains of Australia in relation to climate change"
in Körner, C., and E. M. Spehn, eds. Mountain Biodiversity: A Global
Assessment. London, Parthenon Publishing: 241-249.
Gregory, J. M., and J. F. B. Mitchell (1995). "Simulation of Daily Variability of
Surface-Temperature and Precipitation over Europe in the Current and
2xco(2) Climates Using the Ukmo Climate Model." Quarterly journal of the
Royal Meteorological Society 121(526): 1451-1476.
Gregory, J. M., P. A. Stott, D. J. Cresswell, N. A. Rayner, C. Gordon, and D. M. H.
Sexton (2002). "Recent and future changes in Arctic sea ice simulated by the
HadCM3 AOGCM." Geophysical Research Letters 29(24).
Guisan, A., and J.-P. Theurillat (2000). "Assessing alpine plant vulnerability to
climate change: a modeling perspective." Integrated Assessment 1(4): 307-
320(314).
Hannah, L., G. F. Midgley, T. Lovejoy, W. J. Bond, M. Bush, J. C. Lovett, D. Scott et
al. (2002). "Conservation of Biodiversity in a Changing Climate."
Conservation Biology 16(1): 264-268.
Hennessy, K., P. Whetton, I. Smith, J. Bathols, M. Hutchinson, and J. Sharples
(2003). The impact of climate change on snow conditions in mainland
Australia. Victoria, Australia, CSIRO:
96
Hilbert, D. W., M. Bradford, T. Parker, and D. A. Westcott (2003). "Golden
bowerbird (Prionodura newtonia) habitat in past, present and future climates:
predicted extinction of a vertebrate in tropical highlands due to global
warming." Biological Conservation In Press, Corrected Proof.
Hilbert, D. W., B. Ostendorf, and M. S. Hopkins (2001). "Sensitivity of tropical
forests to climate change in the humid tropics of north Queensland." Austral
Ecology 26(6): 590-603.
Hoegh-Guldberg, O. (1999). "Climate change, coral bleaching and the future of the
world's coral reefs." Marine and Freshwater Research 50(8): 839-866.
Hojer, R., M. Bayley, C. F. Damgaard, and M. Holmstrup (2001). "Stress synergy
between drought and a common environmental contaminant: studies with the
collembolan Folsomia candida." Global Change Biology 7(4): 485-494.
Hughes, L. (2000). "Biological consequences of global warming: is the signal already
apparent?" Trends in Ecology & Evolution 15(2): 56-61.
— (2003). "Climate change and Australia: Trends, projections and impacts." Austral
Ecology 28(4): 423-443.
Hughes, L., E. M. Cawsey, and M. Westoby (1996). "Climatic range sizes of
Eucalyptus species in relation to future climate change." Global Ecology and
Biogeography Letters 5(1): 23-29.
Hulme, M., J. Mitchell, W. Ingram, J. Lowe, T. Johns, M. New, and D. Viner
(1999a). "Climate change scenarios for global impacts studies." Global
Environmental Change-Human and Policy Dimensions 9: S3-S19.
Hulme, M., S. C. B. Raper, and T. M. L. Wigley (1995). "An Integrated Framework
to Address Climate-Change (Escape) and Further Developments of the Global
and Regional Climate Modules (Magicc)." Energy Policy 23(4-5): 347-355.
Hulme, M., and N. Sheard (1999). Climate change scenarios for Australia Climatic
Research Unit. Norwich, UK,
Hulme, M., N. Sheard, and A. Markham (1999b). Global Climate Change Scenarios.
Norwich, Climatic Research Unit: 2.
IPCC (2001). Synthesis Report. Summary for policy makers. An assessment of the
Intergovernmental Panel on Climate Change. Cambridge, UK, Cambridge
University Press.
Jackson, S. T., and C. Y. Weng (1999). "Late quaternary extinction of a tree species
in eastern North America." Proceedings of the National Academy of Sciences
of the United States of America 96(24): 13847-13852.
Johannessen, O. M., L. Bengtsson, M. W. Miles, S. I. Kuzmina, V. A. Semenov, G.
V. Alekseev, A. P. Nagurnyi et al. (2002). Arctic climate change - Observed
and modeled temperature and sea ice variability. Bergen, Nansen
Environmental and Remote Sensing Center:
Johns, T. C., R. E. Carnell, J. F. Crossley, J. M. Gregory, J. F. B. Mitchell, C. A.
Senior, S. F. B. Tett et al. (1997). "The second Hadley Centre coupled ocean-
atmosphere GCM: Model description, spinup and validation." Climate
Dynamics 13(2): 103-134.
Jones, C. D., M. Collins, P. M. Cox, and S. A. Spall (2001). "The carbon cycle
response to ENSO: A coupled climate-carbon cycle model study." Journal of
Climate 14(21): 4113-4129.
Jones, C. D., P. M. Cox, R. L. H. Essery, D. L. Roberts, and M. J. Woodage (2003).
"Strong carbon cycle feedbacks in a climate model with interactive CO2 and
sulphate aerosols." Geophysical Research Letters 30(9): art. no.-1479.
97
Jones, R. N. (2000). "Analysing the risk of climate change using an irrigation demand
model." Climate Research 14(2): 89-100.
Kanowski, J. (2001). "Effects of elevated CO2 on the foliar chemistry of seedlings of
two rainforest trees from north-east Australia: Implications for folivorous
marsupials." Austral Ecology 26(2): 165-172.
Keleher, C. J., and F. J. Rahel (1996). "Thermal limits to salmonid distributions in the
rocky mountain region and potential habitat loss due to global warming: A
geographic information system (GIS) approach." Transactions of the
American Fisheries Society 125(1): 1-13.
Kerr, J., and L. Packer (1998). "The impact of climate change on mammal diversity in
Canada." Environmental Monitoring and Assessment 49(2-3): 263-270.
Kienast, F., O. Wildi, and B. Brzeziecki (1998). "Potential Impacts of climate change
on species richness in mountain forests--an ecological risk assessment."
Biological Conservation 83(3): 291-305.
Kirilenko, A. P., N. V. Belotelov, and B. G. Bogatyrev (2000). "Global model of
vegetation migration: incorporation of climatic variability." Ecological
Modelling 132(1-2): 125-133.
Kirilenko, A. P., and A. M. Solomon (1998). "Modeling Dynamic Vegetation
Response to Rapid Climate Change Using Bioclimatic Classification."
Climatic change 38(1): 15.
Kittel, T. G. F., W. L. Steffen, and F. S. Chapin (2000). "Global and regional
modelling of Arctic-boreal vegetation distribution and its sensitivity to altered
forcing." Global Change Biology 6: 1-18.
Kundzewicz, W., P. Martin, W. Cramer, J. Holten, Z. Kaczmarek, P. Martens, R.
Nicholls et al. (2001). "Chapter 13: Europe" in Climate Change 2001:
Impacts, adaptation and vulnerability. Working Group II of the
Intergovernmental Panel on Climate Change. Cambridge, UK, Cambridge
University Press: 641-692.
Lal, M., H. Harasawa, D. Murdiyarso, W. N. Adger, S. Adkhikary, M. Ando, Y.
Anokhin et al. (2001). "Chapter 11: Asia" in Climate Change 2001: Impacts,
adaptation and vulnerability. Working Group II of the Intergovernmental
Panel on Climate Change. Cambridge, UK, Cambridge University Press: 533-
590.
Laurance, W. F., and G. B. Williamson (2001). "Positive feedbacks among forest
fragmentation, drought, and climate change in the Amazon." Conservation
Biology 15(6): 1529-1535.
Laurance, W. F., G. B. Williamson, P. Delamonica, A. Oliveira, T. E. Lovejoy, C.
Gascon, and L. Pohl (2001). "Effects of a strong drought on Amazonian forest
fragments and edges." Journal of Tropical Ecology 17: 771-785.
Lenart, E. A., R. T. Bowyer, J. V. Hoef, and R. W. Ruess (2002). "Climate change
and caribou: effects of summer weather on forage." Canadian Journal of
Zoology-Revue Canadienne De Zoologie 80(4): 664-678.
Lucht, W., I. C. Prentice, R. B. Myneni, S. Sitch, P. Friedlingstein, W. Cramer, P.
Bousquet et al. (2002). "Climatic Control of the High-Latitude Vegetation
Greening Trend and Pinatubo Effect." Science 296(5573): 1687-1689.
Malcolm, J., C. Liu, L. B. Miller, T. Allnutt, and L. Hansen (2002a). Global warming
and species loss in global significant terrestrial ecosystems. Gland,
Switzerland, WWF:
98
Malcolm, J. R., A. Markham, R. P. Neilson, and M. Garaci (2002b). "Estimated
migration rates under scenarios of global climate change." Journal of
Biogeography 29(7): 835-849.
Manabe, S., M. J. Spelman, and R. J. Stouffer (1992). "Transient Responses of a
Coupled Ocean Atmosphere Model to Gradual Changes of Atmospheric Co2
.2. Seasonal Response." Journal of Climate 5(2): 105-126.
Manabe, S., R. J. Stouffer, M. J. Spelman, and K. Bryan (1991). "Transient responses
of a coupled-ocean atmosphere model to gradual changes of atmospheric
CO2. Part I: Annual mean response." Journal of Climate 4: 785-818.
Mann, M. E., R. S. Bradley, and M. K. Hughes (1999). "Northern hemisphere
temperatures during the past millennium: Inferences, uncertainties, and
limitations." Geophysical Research Letters 26(6): 759-762.
Martinez-Vilalta, J., and J. Pinol (2002). "Drought-induced mortality and hydraulic
architecture in pine populations of the NE Iberian Peninsula." Forest Ecology
and Management 161(1-3): 247-256.
Martinez-Vilalta, J., J. Pinol, and K. Beven (2002). "A hydraulic model to predict
drought-induced mortality in woody plants: an application to climate change
in the Mediterranean." Ecological Modelling 155(2-3): 127-147.
McCarty, J. P. (2001). "Ecological consequences of recent climate change."
Conservation Biology 15(2): 320-331.
Mearns, L. O., C. Rosenzweig, and R. Goldberg (1997). "Mean and Variance Change
in Climate Scenarios: Methods, Agricultural Applications, and Measures of
Uncertainty." Climatic Change 35(4): 367-396.
Mendelsohn, R., W. Morrison, M. E. Schlesinger, and N. G. Andronova (2000).
"Country-Specific Market Impacts of Climate Change." Climatic Change
45(3/4): 553-569.
Midgley, G. F. (2003). Oct. 13 pers. comm.
Midgley, G. F., L. Hannah, D. Millar, M. C. Rutherford, and L. W. Powrie (2002).
"Assessing the vulnerability of species richness to anthropogenic climate
change in a biodiversity hotspot." Global Ecology and Biogeography 11(6):
445-452.
Midgley, G. F., L. Hannah, D. Millar, W. Thuiller, and A. Booth (2003). "Developing
regional and species-level assessments of climate change impacts on
biodiversity in the Cape Floristic Region." Biological Conservation 112(1-2):
87-97.
Miles, L. 2002. The impact of global climate change on tropical forest biodiversity in
Amazonia. PhD thesis, University of Leeds.
Miles, L., O. Phillips, A. Grainger, and S. Carver (2003). Climate change and
Amazonian forest biodiversity, April 2003 pers. comm.
Mitchell, J. F. B., T. C. Johns, W. J. Ingram, and J. A. Lowe (2000). "The effect of
stabilising atmospheric carbon dioxide concentrations on global and regional
climate change." Geophys. Res. Lett. 27(18): 2977-2980.
Mitchell, N. D. a., and J. E. Williams (1996). "The consequences for native biota of
anthropogenic-induced climate change." in Bouma, W. J., G. I. Pearman, and
M. R. Manning, eds. Greenhouse: Coping with Climate Change. Collingwood,
Victoria, Australia, CSIRO Publishing,: 308-324.
Mkanda, F. X. (1996). "Potential impacts of future climate change on nyala
Tragelaphus angasi in Lengwe National Park, Malawi." Climate Research
6(2): 157-164.
99
— (1999). "Drought as an analogue climate change scenario for prediction of
potential impacts on Malawi's wildlife habitats." Climate Research 12(2-3):
215-222.
Mouillot, F., S. Rambal, and R. Joffre (2002). "Simulating climate change impacts on
fire frequency and vegetation dynamics in a Mediterranean-type ecosystem."
Global Change Biology 8(5): 423-437.
Mulrennan, M. E., and C. D. Woodroffe (1998). "Holocene development of the lower
Mary River plains, Northern Territory, Australia." Holocene 8(5): 565-579.
Murphy, J. M. (1995). "Transient-Response of the Hadley-Center Coupled Ocean-
Atmosphere Model to Increasing Carbon-Dioxide .1. Control Climate and
Flux Adjustment." Journal of Climate 8(1): 36-56.
Murphy, J. M., and J. F. B. Mitchell (1995). "Transient-Response of the Hadley-
Center Coupled Ocean- Atmosphere Model to Increasing Carbon-Dioxide .2.
Spatial and Temporal Structure of Response." Journal of Climate 8(1): 57-80.
Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. da Fonseca, and J. Kent
(2000). "Biodiversity hotspots for conservation priorities." Nature 403(6772):
853-858.
Najjar, R. G., H. A. Walker, P. J. Anderson, E. J. Barron, R. J. Bord, J. R. Gibson, V.
S. Kennedy et al. (2000). "The potential impacts of climate change on the
mid-Atlantic coastal region." Climate Research 14(3): 219-233.
National Research Council (1999). Perspectives on Biodiversity: Valuing its Role in
an Everchanging World. Washington, DC, USA, National Academy Press.
Neilson, R. P., I. C. Prentice, B. Smith, T. Kittel, and D. Viner (1997). "Simulated
changes in vegetation distribution under global warming" in Waston, R. T.,
M. C. Zinyowera, R. H. Moss, and D. J. Dokken, eds. The Regional
Impactions of Climate Change. An Assessment of Vulnerability. Special
Report of the IPCC Working Group II. New York, Cambridge University
Press: 439-456.
Nepstad, D., G. Carvalho, A. C. Barros, A. Alencar, J. P. Capobianco, J. Bishop, P.
Moutinho et al. (2001). "Road paving, fire regime feedbacks, and the future of
Amazon forests." Forest Ecology and Management 154(3): 395-407.
Ni, J. (2000). "A simulation of biomes on the Tibetan Plateau and their responses to
global climate change." Mountain Research and Development 20(1): 80-89.
— (2001). "Carbon Storage in Terrestrial Ecosystems of China: Estimates at
Different Spatial Resolutions and Their Responses to Climate Change."
Climatic Change 49(3): 339-358.
— (2002). "Effects of climate change on carbon storage in boreal forests of China: A
local perspective." Climatic Change 55(1-2): 61-75.
Ni J (2001). "Carbon Storage in Terrestrial Ecosystems of China: Estimates at
Different Spatial Resolutions and Their Responses to Climate Change."
Climatic Change 49(3): 339-358(320).
Ni, J., M. T. Sykes, I. C. Prentice, and W. Cramer (2000). "Modelling the vegetation
of China using the process-based equilibrium terrestrial biosphere model
BIOME3." Global Ecology and Biogeography 9(6): 463-479.
Nicholls, R. J., F. M. J. Hoozemans, and M. Marchand (1999). "Increasing flood risk
and wetland losses due to global sea- level rise: regional and global analyses."
Global Environmental Change-Human and Policy Dimensions 9: S69-S87.
Nordhaus, W., and J. Boyer (2000). Warming the World: Economic Models of
Climate Change. Cambridge, MA, MIT Press.
100
Novacek, M. J., and E. E. Cleland (2001). "The current biodiversity extinction event:
Scenarios for mitigation and recovery." Proceedings of the National Academy
of Sciences of the United States of America 98(10): 5466-5470.
Nurse, L. A., G. Sem, J. E. Hay, A. G. Suarez, P. P. Wong, L. Briguglio, S.
Ragoonaden et al. (2001). "Chapter 17: Small Island States" in Climate
Change 2001: Impacts, adaptation and vulnerability. Working Group II of the
Intergovernmental Panel on Climate Change. Cambridge, UK, Cambridge
University Press: 843-875.
Ogaya, R., J. Penuelas, J. Martinez-Vilalta, and M. Mangiron (2003). "Effect of
drought on diameter increment of Quercus ilex, Phillyrea latifolia, and
Arbutus unedo in a holm oak forest of NE Spain." Forest Ecology and
Management 180(1-3): 175-184.
O'Neill, B. C., and M. Oppenheimer (2002). "CLIMATE CHANGE: Dangerous
Climate Impacts and the Kyoto Protocol." Science 296(5575): 1971-1972.
Ostendorf, B., D. W. Hilbert, and M. S. Hopkins (2001). "The effect of climate
change on tropical rainforest vegetation pattern." Ecological Modelling 145(2-
3): 211-224.
Overpeck, J. T., R. S. Webb, and T. Webb (1992). "Mapping Eastern North-American
Vegetation Change of the Past 18 Ka - No-Analogs and the Future." Geology
20(12): 1071-1074.
Parkinson, C. L., and D. J. Cavalieri (2002). "A 21 year record of Arctic sea-ice
extents and their regional, seasonal and monthly variability and trends" in
Annals of Glaciology, Vol 34, 2002. Annals of Glaciology. Cambridge, INT
GLACIOLOGICAL SOC: 441-446.
Parmesan, C., T. L. Root, and M. R. Willig (2000). "Impacts of extreme weather and
climate on terrestrial biota." Bulletin of the American Meteorological Society
81(3): 443-450.
Parmesan, C., and G. Yohe (2003). "A globally coherent fingerprint of climate
change impacts across natural systems." Nature 421(6918): 37-42.
Parry, M., N. Arnell, T. McMichael, R. Nicholls, P. Martens, S. Kovats, M.
Livermore et al. (2001). "Millions at risk: defining critical climate change
threats and targets." Global Environmental Change 11(ER3): 181-183.
Parry, M., C. Rosenzweig, A. Iglesias, G. Fischer, and M. Livermore (1999).
"Climate change and world food security: a new assessment." Global
Environmental Change-Human and Policy Dimensions 9: S51-S67.
Parry, M. L., and T. R. Carter (1989). "An Assessment of the Effects of Climatic-
Change on Agriculture." Climatic Change 15(1-2): 95-116.
Pearson, R. G., and T. P. Dawson (2003). "Predicting the impacts of climate change
on the distribution of species: are bioclimate envelope models useful?" Global
Ecology and Biogeography 12(5): 361-371.
Peters, R. L., and J. D. Darling (1985). "The greenhouse effect and nature reserves."
BioScience 35(707).
Peterson, A. T. (2003). "Projected climate change effects on Rocky Mountain and
Great Plains birds: generalities of biodiversity consequences." Global Change
Biology 9(5): 647-655.
Peterson, A. T., M. A. Ortega-Huerta, J. Bartley, V. Sanchez-Cordero, J. Soberon, R.
H. Buddemeier, and D. R. B. Stockwell (2002). "Future projections for
Mexican faunas under global climate change scenarios." Nature 416(6881):
626-629.
101
Petit, J. R., J. Jouzel, D. Raynaud, N. I. Barkov, J.-M. Barnola, I. Basile, M. Bender et
al. (1999). "Climate and atmospheric history of the past 420,000 years from
the Vostok ice core, Antarctica." Nature 399(6735): 429-436.
Phillips, J. G., M. A. Cane, and C. Rosenzweig (1998). "ENSO, seasonal rainfall
patterns and simulated maize yield variability in Zimbabwe." Agricultural and
Forest Meteorology 90(1): 39-50(12).
Pimm, S. L., G. J. Russell, and T. M. Brooks (1995). "The Future of Biodiversity."
Science 269(5222): 347.
Pittock, B., D. Wratt, R. Basher, B. Bates, M. Finalyson, H. Gitay, A. Woodward et
al. (2001). "Chapter 12: Australia and New Zealand" in Climate Change 2001:
Impacts, adaptation and vulnerability. Working Group II of the
Intergovernmental Panel on Climate Change. Cambridge, UK, Cambridge
University Press: 591-639.
Pittocket, B., D. Wratt, R. Basher, B. Bates, M. Finalyson, H. Gitay, A. Woodward et
al. (2001). "Chapter 12: Australia and New Zealand" in Climate Change 2001:
Impacts, adaptation and vulnerability. Working Group II of the
Intergovernmental Panel on Climate Change. Cambridge, UK, Cambridge
University Press: 591-639.
Pouliquen-Young, O., and P. Newman (1999). The Implications of Climate Change
for Land-Based Nature Conservation Strategies. Perth, Australia, Australian
Greenhouse Office, Environment Australia, Canberra, and Institute for
Sustainability and Technology Policy, Murdoch University: 91.
Price, J. T., T. L. Root, K. R. Hall, and e. al (2000). Climate change, wildlife and
ecosystems. Supplemental information prepared for the Intergovernmental
Panel on Climate Change Working Group II. IPCC:
Qureshi, A., and D. Hobbie (1994). Climate change in Asia. Manila, Asian
Development Bank. Cited by World Bank 2000. Chapter 2 Potential Impacts
of climate change in Bangladesh:
Ragab, R., and C. Prudhomme (2002). "Climate Change and Water Resources
Management in Arid and Semi-arid Regions: Prospective and Challenges for
the 21st Century." Biosystems Engineering 81(1): 3-34.
Ramankutty, N., J. A. Foley, J. Norman, and K. McSweeney (2002). "The global
distribution of cultivable lands: current patterns and sensitivity to possible
climate change." Global Ecology and Biogeography 11(5): 377-392.
Raper, S. C. B., J. M. Gregory, and T. J. Osborn (2001). "Use of an upwelling-
diffusion energy balance climate model to simulate and diagnose A/OGCM
results." Climate Dynamics 17(8): 601-613.
Reilly, J., F. Tubiello, B. McCarl, D. Abler, R. Darwin, K. Fuglie, S. Hollinger et al.
(2003). "U.S. Agriculture and Climate Change: New Results." Climatic
Change 57(1): 43-69.
Reyenga, P. J., S. M. Howden, H. Meinke, and G. M. McKeon (1999). "Modelling
global change impacts on wheat cropping in south-east Queensland,
Australia." Environmental Modelling and Software 14(4): 297-306.
Rijsberman, F. J., and R. J. Swart eds. (1990). Targets and Indicators of Climate
Change, Stockholm Environment Institute.
Root, T. L., J. T. Price, K. R. Hall, S. H. Schneider, C. Rosenzweig, and J. A. Pounds
(2003). "Fingerprints of global warming on wild animals and plants."
Nature(6918): 57-59.
Rosenfeld, D. (1999). "TRMM observed first direct evidence of smoke from forest
fires inhibiting rainfall." Geophysical Research Letters 26(20): 3105-3108.
102
Rosenzweig, C., A. Iglesias, X. B. Yang, P. R. Epstein, and E. Chivian (2001).
"Climate Change and Extreme Weather Events; Implications for Food
Production, Plant Diseases, and Pests." Global Change and Human Health
2(2): 90-104.
Rosenzweig, C., and M. Parry (1994). "Potential impact of climate change on world
food supply." Nature 367(6459): 133-138.
Rouget, M., D. M. Richardson, and R. M. Cowling (2003). "The current configuration
of protected areas in the Cape Floristic Region, South Africa--reservation bias
and representation of biodiversity patterns and processes." Biological
Conservation 112(1-2): 129-145.
Rutherford, M. C., G. F. Midgley, W. J. Bond, L. W. Powrie, C. F. Musil, R. Roberts,
and J. Allsopp (1999a). South African Country Study on Climate Change.
Pretoria, South Africa, Terrestrial Plant Diversity Section, Vulnerability and
Adaptation, Department of Environmental Affairs and Tourism:
Rutherford, M. C., L. W. Powrie, and R. E. Schulze (1999b). "Climate change in
conservation areas of South Africa and its potential impact on floristic
composition: a first assessment." Divers Distrib 5(6): 253-262.
Sala, O. E., F. S. Chapin, J. J. Armesto, E. Berlow, J. Bloomfield, R. Dirzo, E. Huber-
Sanwald et al. (2000). "Biodiversity - Global biodiversity scenarios for the
year 2100." Science 287(5459): 1770-1774.
Seabloom, E. W., A. P. Dobson, and D. M. Stoms (2002). "Extinction rates under
nonrandom patterns of habitat loss." PNAS 99(17): 11229-11234.
Serreze, M. C., J. A. Maslanik, T. A. Scambos, F. Fetterer, J. Stroeve, K. Knowles, C.
Fowler et al. (2003). "A record minimum arctic sea ice extent and area in
2002." Geophysical Research Letters 30(3).
Sheppard, C. R. C. (2003). "Predicted recurrences of mass coral mortality in the
Indian Ocean." Nature 425(6955): 294-297.
Smith, J. B., A. Rahman, S. Haq, and M. Q. Mirza (1998). Considering Adaptation to
Climate change in the sustainable development of Bangladesh. World Bank
Report. Washington, DC, World Bank: 103.
Smith, J. B., H.-J. Schellnhuber, M. M. Q. Mirza, S. Fankhauser, R. Leemans, L.
Erda, L. Ogallo et al. (2001). "Chapter 19: Vulnerability to climate change
and reasons for concern: A synthesis" in Climate Change 2001: Impacts,
adaptation and vulnerability. Working Group II of the Intergovernmental
Panel on Climate Change. Cambridge, UK, Cambridge University Press: 915-
967.
Sorenson, L. G., R. Goldberg, T. L. Root, and M. G. Anderson (1998). "Potential
effects of global warming on waterfowl populations breeding in the Northern
Great Plains." Climatic Change 40(2): 343-369.
Soulé, M. E. (1992). "Foreword: The Wrong Time for Climate Change" in Peters, R.
L., and T. E. Lovejoy, eds. Global Warming and Biological Diversity. New
Haven, Conneticut, Yale University Press:
Stirling, I. (2000). "Running out of ice? Polar bears need plenty of it." Natural history
109(2): 92-92.
Stirling, I., N. J. Lunn, and J. Iacozza (1999). "Long-term trends in the population
ecology of polar bears in western Hudson Bay in relation to climatic change."
Arctic 52(3): 294-306.
Theurillat, J.-P., and A. Guisan (2001). "Potential Impact of Climate Change on
Vegetation in the European Alps: A Review." Climatic Change 50(1/2): 77-
109.
103
Tian, H. Q., J. M. Melillo, D. W. Kicklighter, A. D. McGuire, J. V. K. Helfrich, B.
Moore, and C. J. Vorosmarty (1998). "Effect of interannual climate variability
on carbon storage in Amazonian ecosystems." Nature 396(6712): 664-667.
Titus, J. G., and V. Narayanan (1996). "The risk of sea level rise." Climatic Change
33(2): 151-212.
Tol, R. S. J. (2001). "Equitable cost-benefit analysis of climate change policies."
Ecological Economics 36(1): 71-85.
— (2002). "Estimates of the damage costs of climate change. Part 1: Benchmark
estimates." Environmental & Resource Economics 21(1): 47-73.
— (2003). "Is the uncertainty about climate change too large for expected cost-benefit
analysis?" Climatic Change 56(3): 265-289.
Tuchman, N. C., R. G. Wetzel, S. T. Rier, K. A. Wahtera, and J. A. Teeri (2002).
"Elevated atmospheric CO2 lowers leaf litter nutritional quality for stream
ecosystem food webs." Global Change Biology 8(2): 163-170.
UN. 1992. United Nations Framework Convention on Climate Change. United
Nations, New York.
Visser, M. E., and L. J. Holleman (2001). "Warmer springs disrupt the synchrony of
oak and winter moth phenology." Proceedings of the Royal Society of London
Series B Biological Sciences 268(1464): 289-294.
Volney, W. J. A., and R. A. Fleming (2000). "Climate change and impacts of boreal
forest insects." Agriculture Ecosystems & Environment 82(1-3): 283-294.
Vorosmarty, C. J., P. Green, and R. B. Lammers (2000). "Global Water Resources:
Vulnerability from Climate Change and Population Growth." Science
289(5477): 284.
Walsh, K. J. E., and B. F. Ryan (2000). "Tropical cyclone intensity increase near
Australia as a result of climate change." Journal of Climate 13(16): 3029-
3036.
Walther, G. R., E. Post, P. Convey, A. Menzel, C. Parmesan, T. J. C. Beebee, J. M.
Fromentin et al. (2002). "Ecological responses to recent climate change."
Nature 416(6879): 389-395.
Watson, R. T., M. C. Zinyowera, R. H. Moss, and D. J. Dokken eds. (1998). The
Regional Impacts of Climate Change: An Assessment of Vulnerability, A
Special Report of IPCC Working Group II. Cambridge, UK, Cambridge
University Press.
WBGU (1995). Scenario for the derivation of global CO2 reduction targets and
implementation strategies. Berlin, Germany, WBGU:
Whetton, P. H., M. R. Haylock, and R. Galloway (1996). "Climate change and snow-
cover duration in the Australian Alps." Climatic Change 32(4): 447-479.
White, A., M. G. R. Cannell, and A. D. Friend (2000a). "CO2 stabilization, climate
change and the terrestrial carbon sink." Global Change Biology 6(7): 817-833.
White, T. A., B. D. Campbell, P. D. Kemp, and C. L. Hunt (2000b). "Sensitivity of
three grassland communities to simulated extreme temperature and rainfall
events." Global Change Biology 6(6): 671-684.
Wigley, T. M., and S. C. Raper (2001). "Interpretation of high projections for global-
mean warming." Science 293(5529): 451-454.
Williams, S. E., E. E. Bolitho, and S. Fox (2003). "Climate change in Australian
tropical rainforests: an impending environmental catastrophe." Proceedings of
the Royal Society of London Series B-Biological Sciences 270(1527): 1887-
1892.
104
Winters, P., R. Murgai, E. Sadoulet, A. De Janvry, and G. Frisvold (1998).
"Economic and welfare impacts of climate change on developing countries."
Environmental & Resource Economics 12(1): 1-24.
i
