FINAL DRAFT                                                          Chapter 2                               IPCC WGII Sixth Assessment Report

 2                        Chapter 2: Terrestrial and Freshwater Ecosystems and their Services
 4   Coordinating Lead Authors: Camille Parmesan (France/USA/United Kingdom), Mike D. Morecroft
 5   (United Kingdom), Yongyut Trisurat (Thailand)
 7   Lead Authors: Rita Adrian (Germany), Gusti Zakaria Anshari (Indonesia), Almut Arneth (Germany),
 8   Qingzhu Gao (China), Patrick Gonzalez (USA), Rebecca Harris (Australia), Jeff Price (United Kingdom),
 9   Nicola Stevens (South Africa), Gautam Hirak Talukdar (India)
11   Contributing Authors: David D Ackerly (USA), Elizabeth Anderson (USA), Vanessa Bremerich
12   (Germany), Lluís Brotons (Spain), Yu-Yun Chen (Taiwan, Province of China), Meghnath Dhimal (Nepal),
13   Sami Domisch (Germany), Errol Douwes (South Africa), Alexander Flecker (USA), Wendy Foden (South
14   Africa), Rachael V. Gallagher (Australia), Michael Goulding (USA), Aurora Gaxiola (Chile), Kerry-Anne
15   Grey (South Africa), Susan Harrison (USA), David A. Keith (Australia), Benjamin M. Kraemer
16   (USA/Germany), Simone Langhans (Switzerland), Andrew Latimer (USA), Julie Loisel (Canada, USA),
17   James Pearce-Higgins (United Kingdom), Guy Midgley (South Africa), Erin Mordecai (USA), Francisco
18   Moreira (Portugal), Isla Myers-Smith (United Kingdom/USA), A. Townsend Peterson (USA), Julio C.
19   Postigo (Peru/USA), Joacim Rocklöv (Sweden), Angela Gallego-Sala (Spain/United Kingdom), Nathalie
20   Seddon (United Kingdom), Michael C. Singer (France/United Kingdom), Jasper Slingsby (South Africa),
21   Stavana E. Strutz (USA), Merritt Turetsky (USA), Beth Turner (Canada), Kenneth Young (USA).
23   Review Editors: Carol Franco Billini (Dominican Republic/USA), Yakiv Didukh (Ukraine), Andreas
24   Fischlin (Switzerland)
26   Chapter Scientists: Stavana E. Strutz (USA), Dalila Mezzi-Booth (France)
28   Date of Draft: 1 October 2021
30   Notes: TSU Compiled Version
33   Table of Contents
35   Executive Summary.......................................................................................................................................... 3
36   2.1 Introduction .............................................................................................................................................. 9
37       2.1.1 Overview ......................................................................................................................................... 9
38       2.1.2 Points of Departure......................................................................................................................... 9
39       2.1.3 Guide to Attribution and Traceability of Uncertainty Assessments.............................................. 10
40   2.2 Connections of Ecosystem Services to Climate Change...................................................................... 11
41   2.3 Hazards and Exposure ........................................................................................................................... 13
42       2.3.1 Observed Changes to Hazards and Extreme Events..................................................................... 13
43       2.3.2 Projected Impacts of Extreme Events ........................................................................................... 15
44       2.3.3 Biologically Important Physical Changes in Freshwater Systems ............................................... 16
45   Cross-Chapter Box EXTREMES: Ramifications of Climatic Extremes for Marine, Terrestrial,
46       Freshwater and Polar Natural Systems................................................................................................ 22
47   2.4 Observed Impacts of Climate Change on Species, Communities, Biomes, Key Ecosystems and
48       their Services ........................................................................................................................................... 26
49       2.4.1 Overview ....................................................................................................................................... 26
50       2.4.2 Observed Responses to Climate Change by Species and Communities (Freshwater and
51                   Terrestrial) ............................................................................................................................ 27
52   FAQ2.1: Will species go extinct with climate change and is there anything we can do to prevent it? ... 29
53   FAQ2.2: How does climate change increase the risk of diseases?.............................................................. 40
54       2.4.3 Observed Changes in Key Biomes, Ecosystems and their Services .............................................. 44
55       2.4.4 Observed Changes in Ecosystem Processes and Services ............................................................ 52
56   FAQ2.3: Is climate change increasing wildfire? .......................................................................................... 57
57       2.4.5 Conclusions on Observed Impacts ................................................................................................ 64
     Do Not Cite, Quote or Distribute                                              2-1                                                         Total pages: 237
     FINAL DRAFT                                                            Chapter 2                                IPCC WGII Sixth Assessment Report

 1   2.5 Projected Impacts and Risk for Species, Communities, Biomes, Key Ecosystems and their
 2       Services .................................................................................................................................................... 68
 3       2.5.1 Projected Changes at Species and Community Levels ................................................................. 68
 4       2.5.2 Projected Changes at Level of Biomes and Whole Ecosystems .................................................... 75
 5   Box 2.1: Assessing Past Projections of Ecosystem Change Against Observations ................................... 84
 6       2.5.3 Risk Assessment of Ecosystems and Related Services .................................................................. 86
 7       2.5.4 Key Risks to Terrestrial and Freshwater Ecosystems from Climate Change .............................. 98
 8   FAQ2.4: How does nature benefit human health and well-being and how does climate change affect
 9       this? ........................................................................................................................................................ 103
10   2.6 Climate Change Adaptation for Terrestrial and Freshwater Ecosystems ..................................... 105
11       2.6.1 Limits to Autonomous (Natural) Adaptation............................................................................... 106
12       2.6.2 Adaptation for Biodiversity Conservation .................................................................................. 107
13       2.6.3 Nature-based Solutions: Ecosystem-based Adaptation .............................................................. 111
14       2.6.4 Adaptation for Increased Risk of Disease ................................................................................... 115
15   Cross-Chapter Box ILLNESS: Infectious Diseases, Biodiversity and Climate: Serious Risks Posed by
16       Vector- and Water-borne Diseases ..................................................................................................... 116
17       2.6.5 Adaptation in Practice: Case Studies and Lessons Learned ...................................................... 121
18       2.6.6 Limits to Adaptation Actions by People ...................................................................................... 129
19       2.6.7 Climate Resilient Development ................................................................................................... 130
20   Cross-Chapter Box NATURAL: Nature-Based Solutions for Climate Change Mitigation and
21       Adaptation ............................................................................................................................................. 132
22   FAQ2.5: How can we reduce the risk of climate change to people by protecting and managing nature
23       better? .................................................................................................................................................... 140
24       2.6.8 Feasibility of Adaptation Options ............................................................................................... 142
25   Box 2.2: Risks of Maladaptive Mitigation .................................................................................................. 145
26   FAQ2.6: Can tree planting tackle climate change? ................................................................................... 146
27   2.7 Reducing Scientific Uncertainties to Inform Policy and Management Decisions .......................... 148
28       2.7.1 Observed impacts ........................................................................................................................ 148
29       2.7.2 Projected risks ............................................................................................................................ 149
30       2.7.3 Adaptation and Climate Resilient Development ......................................................................... 149
31   References...................................................................................................................................................... 150
32   Large Tables.................................................................................................................................................. 232

     Do Not Cite, Quote or Distribute                                                2-2                                                           Total pages: 237
     FINAL DRAFT                                         Chapter 2                     IPCC WGII Sixth Assessment Report

 1   Executive Summary
 3   Chapter 2, building upon prior assessments1 provides a global assessment of observed impacts and projected
 4   risks of climate change to terrestrial and freshwater ecosystems, including their component species and the
 5   services they provide to people. Where possible, differences among regions, taxonomic groups and
 6   ecosystems types are presented. Adaptation options to reduce risks to ecosystems and people are assessed.
 8   Observed Impacts
10   Multiple lines of evidence, combined with strong and consistent trends observed on every continent,
11   make it very likely2 that many observed changes in ranges, phenology, physiology and morphology of
12   terrestrial and freshwater species can be attributed to regional and global climate changes,
13   particularly increases in frequency and severity of extreme events (very high confidence3) {2.3.1;
14; 2.4.2; 2.4.5; Table 2.2; Table 2.3; Table 2.S.1; Cross-Chapter Box EXTREMES this Chapter}. The
15   most severe impacts are occurring in the most vulnerable species and ecosystems, characterized by inherent
16   physiological, ecological or behavioural traits that limit their abilities to adapt and those most exposed to
17   climatic hazards (high confidence) {;;; 2.4.5; 2.6.1; Cross-Chapter Box EXTREMES
18   this Chapter}.
20   New studies since AR5 and SR1.5 (now >12,000 species globally) show changes consistent with climate-
21   change. Where attribution was assessed (>4,000 species globally) approximately half of species had
22   shifted their ranges to higher latitudes or elevations and two-thirds of spring phenology had advanced,
23   driven by regional climate changes (very high confidence). Shifts in species ranges are altering
24   community make-up, with exotic species exhibiting greater ability to adapt to climate change than natives,
25   especially in more northern latitudes, potentially leading to newly invasive species {}. New
26   analyses demonstrate that prior reports underestimated impacts due to complex biological responses to
27   climate change (high confidence). {;;;; 2.4.5; Table 2.2; Table SM2.1; Table
28   2.3}
30   Responses in freshwater species are strongly related to changes in the physical environment (high
31   confidence){2.3.3;}. Global coverage of quantitative observations in freshwater ecosystems has
32   increased since AR5. Water temperature has increased in rivers (up to 1°C decade-1) and lakes (up to 0.45°C
33   decade-1) {; Figure 2.2}. Extent of ice cover has declined by 25% and duration by >2 weeks {;
34   Figure 2.4}. Changes in flow have led to reduced connectivity in rivers (high confidence) {; Figure
35   2.3}. Indirect changes include alterations in river morphology, substrate composition, oxygen concentrations
36   and thermal regime in lakes (very high confidence) {;}. Dissolved oxygen concentrations have
37   typically declined and primary productivity increased with warming. Warming and browning (increase in
38   organic matter) have occurred in boreal freshwaters with both positive and negative repercussions on water
39   temperature profiles (lower vs. upper water)(high confidence) and primary productivity (medium confidence)
40   and reduced water quality (high confidence) {; Figure 2.5}.

       Previous IPCC assessments include the IPCC Fifth Assessment Report (AR5) (IPCC, 2013; IPCC, 2014c; IPCC,
     2014d; IPCC, 2014a), the Special Report on Global Warming of 1.5°C (SR1.5) (IPCC, 2014b), the Special Report on
     Ocean and Cryosphere in a Changing Climate (SROCC) (IPCC, 2019b) and the IPCC Sixth Assessment Report
     Working Group I (AR6 WGI).
       In this Report, the following terms have been used to indicate the assessed likelihood of an outcome or a result:
     Virtually certain 99–100% probability, Very likely 90–100%, Likely 66–100%, About as likely as not 33–66%,
     Unlikely 0–33%, Very unlikely 0–10%, and Exceptionally unlikely 0–1%. Additional terms (Extremely likely: 95–
     100%, More likely than not >50–100%, and Extremely unlikely 0–5%) may also be used when appropriate. Assessed
     likelihood is typeset in italics, e.g., very likely). This Report also uses the term ‘likely range’ to indicate that the
     assessed likelihood of an outcome lies within the 17-83% probability range.
       In this Report, the following summary terms are used to describe the available evidence: limited, medium, or robust;
     and for the degree of agreement: low, medium, or high. A level of confidence is expressed using five qualifiers: very
     low, low, medium, high, and very high, and typeset in italics, e.g., medium confidence. For a given evidence and
     agreement statement, different confidence levels can be assigned, but increasing levels of evidence and degrees of
     agreement are correlated with increasing confidence.

     Do Not Cite, Quote or Distribute                           2-3                                           Total pages: 237
     FINAL DRAFT                                    Chapter 2                  IPCC WGII Sixth Assessment Report

 1   Climate change has increased wildlife diseases (high confidence). Experimental studies provide high
 2   confidence in attribution of observed increased disease severity, outbreak frequency and emergence of novel
 3   vectors and their diseases into new areas to recent trends in climate and extreme events. Many vector-borne
 4   diseases, and those caused by ticks, helminth worms and Bd (chytrid) fungus, have shifted poleward,
 5   upward, and are emerging in new regions (high confidence).In the high Arctic and high elevations of Nepal,
 6   there is high confidence that climate change has driven expansion of vector-borne diseases that infect
 7   humans. {,,,,,,, Cross-Chapter Box ILLNESS this
 8   Chapter}
10   Forest insect pests have expanded northward and severity and outbreak extent has increased in
11   northern North America, northern Eurasia, due to warmer winters reducing mortality and longer
12   growing seasons favouring more generations per year (high confidence) {}
14   Climate-caused local population extinctions have been widespread among plants and animals, detected
15   in 47% of 976 species examined and associated with increases in hottest yearly temperatures (very high
16   confidence) {}. Climate-driven population extinctions have been higher in tropical (55%), than
17   temperate habitats (39%), higher in freshwater (74%), than in marine (51%) or terrestrial (46%) habitats and
18   higher in animals (50%) than in plants (39%). Extreme heat waves has led to local fish kills in lakes
19   {}. Intensification of droughts contributes to disappearance of small or ephemeral ponds, which often
20   hold rare and endemic species. {; Cross-Chapter Box EXTREMES this Chapter}
22   Global extinctions or near-extinctions have been linked to regional climate change in three
23   documented cases {}. The cloud-forest-restricted Golden toad (Incilius periglenes) was extinct by
24   1990 in a nature preserve in Costa Rica following successive extreme droughts (medium confidence). The
25   white sub-species of the lemuroid ringtail possum (Hemibelideus lemuroides) in Queensland, Australia,
26   disappeared after heatwaves in 2005 (high confidence): intensive censuses found only 2 individuals in 2009.
27   The Bramble Cays Melomys (Melomys rubicola), was not seen after 2009 and declared extinct in 2016, with
28   SLR and increased storm surge, associated with climate change, the most probable drivers (high confidence).
29   The interaction of climate change and chytrid fungus (Bd) has driven many of the observed global
30   amphibian declines and species' extinctions (robust evidence, high agreement) {}.
32   A growing number of studies document genetic evolution within populations in response to recent climate
33   change (very high confidence). To date, genetic changes remain within the limits of known variation for
34   species (high confidence). Controlled selection experiments and field observations indicate that
35   evolution would not prevent a species becoming extinct if its climate space disappears globally (high
36   confidence). Climate hazards outside those to which species are adapted are occurring on all continents
37   (high confidence). More frequent and intense extreme events, superimposed on longer-term climate trends,
38   have pushed sensitive species and ecosystems towards tipping points, beyond ecological and evolutionary
39   capacity to adapt, causing abrupt and possibly irreversible changes (medium confidence). {2.3.1; 2.3.3;
40;; 2.6.1; Cross-Chapter Box ILNESS this Chapter, Cross-Chapter Box EXTREMES this
41   Chapter}
43   Since AR5, biome shifts and structural changes within ecosystems have been detected at an increasing
44   number of locations, consistent with climate change and increasing atmospheric CO2 (high
45   confidence). New studies document changes that were projected in prior reports, including upward shifts in
46   the forest/alpine tundra ecotone, northward shifts in the deciduous/boreal forest ecotones, increased woody
47   vegetation in sub-Arctic tundra, and shifts in thermal habitat in lakes. A combination of changes in grazing,
48   browsing, fire, climate, and atmospheric CO2 are leading to observed woody encroachment into grasslands
49   and savannas, consistent with projections from process-based models driven by precipitation, atmospheric
50   CO2 and wildfire (high confidence). {2.4.3; Table 2.3; Table 2.S.1; Box 2.1; Figure Box 2.1.1; Table Box
51   2.1.1} There is high agreement between projected changes in earlier reports and recent observed trends for
52   areas of increased tree death in temperate and boreal forests and of woody encroachment in savannas,
53   grasslands and tundra {2.5.4; Box 2.1; Figure Box 2.1.1; Table Box 2.1.1}. Observed changes impact
54   structure, functioning and resilience of ecosystems, and ecosystem services such as climate regulation (high
55   confidence). {2.3; 2.4.2; 2.4.3; 2.4.4, 2.5.4, Figure 2.11, Table 2.5, Box 2.1; Figure Box 2.1.1; Table Box
56   2.1.1}

     Do Not Cite, Quote or Distribute                     2-4                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   Regional increases in area burned by wildfires (up to double natural levels), tree mortality up to 20%,
 2   and biome shifts up to 20 km latitudinally and 300 m upslope, have been attributed to anthropogenic
 3   climate change in tropical, temperate and boreal ecosystems around the world (high confidence),
 4   damaging key aspects of ecological integrity. This degrades vegetation survival, habitat for biodiversity,
 5   water supplies, carbon sequestration, and other key aspects of the integrity of ecosystems and their ability to
 6   provide services for people (high confidence). {,,,; Table 2.3; Table 2.S.1}
 8   B.11 Fire seasons have lengthened on one-quarter of vegetated area since 1979 as a result of increasing
 9   temperature, aridity, and drought (medium confidence). Field evidence shows that anthropogenic climate
10   change has increased the area burned by wildfire above natural levels in western North America from
11   1984-2017 by double for the Western USA and 11 time higher than natural in one extreme year in
12   British Columbia (high confidence). Burned area has increased in the Amazon, the Arctic, Australia, and
13   parts of Africa and Asia, consistent with, but not formally attributed to anthropogenic climate change.
14   Wildfires generate up to one-third of global ecosystem carbon emissions, a feedback that exacerbates climate
15   change (high confidence). Deforestation, peat draining, agricultural expansion or abandonment, fire
16   suppression, and inter-decadal cycles such as the El Niño-Southern Oscillation, can exert a stronger
17   influence than climate change on increasing or decreasing wildfire. {; Table 2.3; Table 2.S.1;
18   FAQ2.1}. Increases in wildfire from levels to which ecosystems are adapted degrades vegetation, habitat for
19   biodiversity, water supplies, and other key aspects of the integrity of ecosystems and their ability to provide
20   services for people (high confidence). {,,,; Table 2.3; Table 2.S.1}
22   Anthropogenic climate change has caused drought-induced tree mortality of up to 20% in the period
23   1945-2007 in three regions in Africa and North America. It has also potentially contributed to over 100
24   other cases of drought-induced tree mortality across Africa, Asia, Australia, Europe, and North and South
25   America (high confidence). Field observations document post-mortality vegetation shifts (high confidence).
26   Timber cutting, agricultural expansion, air pollution, and other non-climate factors also contribute to tree
27   death. Climate-change driven increases in forest insect pests have contributed to mortality and changes in
28   carbon dynamics in many temperate and boreal forest areas (very high confidence). The direction of changes
29   in carbon balance and wildfires following insect outbreaks depends on the local forest-insect communities
30   (medium confidence). {; Table 2.3; Table 2.S.1}.
32   Terrestrial ecosystems currently remove more carbon from the atmosphere, 2.5-4.3 Gt y-1, than they
33   emit. Intact tropical rainforests, Arctic permafrost, and other healthy high carbon ecosystems provide
34   a vital global ecosystem service of preventing release of stored carbon (high confidence). Terrestrial
35   ecosystems contain stocks of ~3500 GtC in vegetation, permafrost, and soils, three to five times the amount
36   of carbon in unextracted fossil fuels (high confidence), and >4 times the carbon currently in the atmosphere
37   (high confidence). Tropical forests and Arctic permafrost contain the highest ecosystem carbon, with
38   peatlands following (high confidence). Deforestation, draining and burning of peatlands, and thawing of
39   Arctic permafrost due to climate change shifts these ecosystems from carbon-sinks to carbon-sources (high
40   confidence). {;;}
42   Evidence indicates that climate change is affecting many species, ecosystems, and ecological processes
43   that provide ecosystem services connected to human health, livelihoods, and well-being (medium
44   confidence). These services include climate regulation, water and food provisioning, pollination of crops,
45   tourism and recreation. It is difficult establish end-to-end attribution from climatic changes to changes in a
46   given ecosystem service and to identify the location and timing of impacts. This limits specific adaptation
47   planning, but protection and restoration of ecosystems could build resilience of service provision.{2.2; 2.3;
48; 2.4.5; 2.6.3; 2.6.4; 2.6.5; 2.6.6; 2.6.7; Cross-Chapter Box NATURAL this Chapter; Cross-Chapter
49   Box ILLNESS this Chapter; Cross-Chapter Box EXTREMES this Chapter; Cross-Chapter Box COVID in
50   Chapter 7; Cross-Chapter Box MOVING PLATE in Chapter 5}
52   Projected Risks
54   Climate change increases risks to fundamental aspects of terrestrial and freshwater ecosystems, with
55   the potential for species' extinctions to reach 60% at 5°C GSAT warming (high confidence), biome
56   shifts (changes in the major vegetation form of an ecosystem) on 15% (at 2ºC warming) to 35% (at
57   4ºC warming) of global land (medium confidence), and increases in the area burned by wildfire of 35%

     Do Not Cite, Quote or Distribute                      2-5                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   (at 2ºC warming) to 40% (at 4ºC warming) of global land (medium confidence). {2.5.1; 2.5.2; 2.5.3;
 2   2.5.4; Figure 2.6; Figure 2.7; Figure 2.8; Figure 2.9; Figure 2.11; Table 2.5; Table 2.S.2; Table 2.S.4; Cross-
 3   Chapter Box DEEP in Chapter 1; Cross-Chapter Paper 1}
 5   Extinction of species is an irreversible impact of climate change, the risk of which increases steeply
 6   with rises in global temperature. It is likely that the percentage of species at high risk of extinction (median
 7   and maximum estimates) will be 9% (max 14%) at 1.5°C, 10% (max 18%) at 2°C, 12% (max 29%) at 3.0°C,
 8   13% (max 39%) at 4°C and 15% (max 48%) at 5°C (Figure 2.7). Among groups containing largest numbers
 9   of species at high risk of extinctions for mid-levels of warming (3.2°C) are: invertebrates (15%), specifically
10   pollinators (12%), amphibians (11%, but salamanders are at 24%) and flowering plants (10%). All groups
11   fare substantially better at 2°C, with extinction projections reducing to <3% for all groups, except
12   salamanders at 7% (medium confidence) (Figure 2.8a). Even the lowest estimates of species' extinctions
13   (9%) are 1000x natural background rates. Projected species' extinctions at future global warming levels are
14   consistent with projections from AR4, but assessed on many more species with much greater geographic
15   coverage and a broader range of climate models. {; Figure 2.6; Figure 2.7; Figure 2.8; Cross-Chapter
16   Box DEEP in Chapter1; Cross-Chapter Paper 1}
18   Species are the fundamental unit of ecosystems, and increasing risk to species increases risk to
19   ecosystem integrity, functioning and resilience with increasing warming (high confidence). As species
20   become rare, their roles in the functioning of the ecosystem diminishes (high confidence). Loss of species
21   reduces the ability of an ecosystem to provide services and lowers its resilience to climate change (high
22   confidence). At 1.58°C (median estimate), >10% of species are projected to become endangered (sensu
23   IUCN); at 2.07°C (median) >20% of species are projected to become endangered, representing high and very
24   high biodiversity risk, respectively (medium confidence){2.5.4; Figure 2.8b, Figure 2.11; Table 2.5, Table
25   2.S.4}. Biodiversity loss is projected for more regions with increasing warming, and to be worst in northern
26   South America, southern Africa, most of Australia, and northern high latitudes (medium confidence){;
27   Figure 2.6}.
29   Climate change increases risks of biome shifts on up to 35% of global land at ≥4ºC warming, that
30   emissions reductions could limit to <15% for <2ºC warming (medium confidence). Under high
31   warming scenarios, models indicate shifts of extensive parts of the Amazon rainforest to drier and
32   lower-biomass vegetation (medium confidence), poleward shifts of boreal forest into treeless tundra
33   across the Arctic, and upslope shifts of montane forests into alpine grassland (high confidence). Area at
34   high risk of biome shifts from climate change and land use change combined can double or triple compared
35   to climate change alone (medium confidence). Novel ecosystems, with no historical analogue, are expected
36   to become increasingly common in future (medium confidence). {2.3,, 2.5.2; 2.5.4, Figure 2.11;
37   Table 2.5; Table 2.S.4}
39   Risk of wildfires increases with global temperature (high confidence). With 4ºC warming by 2100
40   wildfire frequency is projected to have a net increase of ~30% (medium confidence). Increased wildfire,
41   combined with soil erosion due to deforestation, could degrade water supplies (medium confidence). For
42   ecosystems with historically low fire frequencies, a projected 4ºC global temperature rise increases risks of
43   fire, with potential increases in tree mortality and conversion of extensive parts of Amazon rainforest to drier
44   and lower-biomass vegetation (medium confidence). {;}
46   Continued climate change substantially increases risk of carbon stored in the biosphere being released
47   into the atmosphere due to increases in processes such as wildfires, tree mortality, insect pest
48   outbreaks, peatland drying and permafrost thaw (high confidence). These phenomena exacerbate self-
49   reinforcing feedbacks between emissions from high-carbon ecosystems (that currently store ~3030–4090
50   GtC) and increasing global temperatures. Complex interactions of climate change, land use change, carbon
51   dioxide fluxes, and vegetation changes, combined with insect outbreaks and other disturbances, will regulate
52   the future carbon balance of the biosphere, processes incompletely represented in current earth system
53   models. The exact timing and magnitude of climate-biosphere feedbacks and potential tipping points of
54   carbon loss are characterized by large uncertainty, but studies of feedbacks indicate increased ecosystem
55   carbon losses can cause large future temperature increases (medium confidence). {AR6 WGI 5.4, Table 5.4,
56   Figure 5.29;;;;;;;; Figure 2.10; Figure 2.11; Table 2.4;
57   Table 2.5; Table 2.S.2; Table 2.S.4}

     Do Not Cite, Quote or Distribute                      2-6                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 2   Contributions of Adaptation Measures to Solutions
 4   The resilience of biodiversity and ecosystem services to climate change can be increased by human
 5   adaptation actions including ecosystem protection and restoration (high confidence). Ecological theory
 6   and observations show that a wide range of actions can reduce risks to species and ecosystem integrity. This
 7   includes minimising additional stresses or disturbances, reducing fragmentation, increasing natural habitat
 8   extent, connectivity and heterogeneity, maintaining taxonomic, phylogenetic and functional diversity and
 9   redundancy; and protecting small-scale refugia where microclimate conditions can allow species to persist
10   (high confidence). Adaptation also includes actions to aid the recovery of ecosystems following extreme
11   events. Understanding the characteristics of vulnerable species can assist in early warning systems to
12   minimise negative impacts and inform management intervention. {2.3; Figure 2.1;, 2.6.2, Table 2.6,
13   2.6.5, 2.6.7, 2.6.8}.
15   There is new evidence that species can persist in refugia where conditions are locally cooler, when they
16   are declining elsewhere (high confidence) {2.6.2}. Protecting refugia, for example where soils remain wet
17   during drought or fire risk is reduced, and in some cases creating cooler microclimates, are promising
18   adaptation measures {2.6.3; 2.6.5; CCP1; CCP5.2.1}. There is also new evidence that species can persist
19   locally because of plasticity, including changes in phenology or behavioural changes that move an individual
20   into cooler micro-climates, and genetic adaptation may allow species to persist for longer than might be
21   expected from local climatic changes (high confidence) {;, 2.6.1}. There is no evidence to
22   indicate that these mechanisms will prevent global extinctions of rare, very localised species at their climatic
23   limits or species inhabiting climate/habitat zones that are disappearing (high confidence). {, 2.5.1,
24, 2.5.4, 2.6.1, 2.6.2, 2.6.5}
26   Since AR5, many adaptation plans and strategies have been developed to protect ecosystems and
27   biodiversity but there is limited evidence of the extent to which adaptation is taking place and virtually
28   no evaluation of the effectiveness of adaptation measures in the scientific literature (medium
29   confidence). This is an important evidence gap that needs to be addressed to ensure a baseline is available
30   against which to judge effectiveness and develop and refine adaptation in future. Many proposed adaptation
31   measures have not been implemented (low confidence) {2.6.2; 2.6.3; 2.6.4; 2.6.5; 2.6.6; 2.6.8; 2.7}
33   Ecosystem restoration and resilience building cannot prevent all impacts of climate change, and
34   adaptation planning needs to manage inevitable changes to species distributions, ecosystem structure
35   and processes (very high confidence). Actions to manage inevitable change include local modification of
36   microclimate or hydrology, adjustment of site management plans and facilitating the dispersal of vulnerable
37   species to new locations, both by increasing habitat connectivity or by active translocations of species.
38   Adaptation can reduce risks but cannot prevent all damaging impacts so is not a substitute for reductions in
39   greenhouse gas emissions (high confidence). {2.2; 2.3; 2.3.1; 2.3.2; 2.4.5;;; 2.5.2;;
40; 2.5.4; 2.6.1; 2.6.2; 2.6.3; 2.6.4; 2.6.5; 2.6.6; 2.6.8; Cross-Chapter Box NATURAL this Chapter}.
42   Ecosystem-base Adaptation (EbA)can deliver climate change adaptation for people with multiple
43   additional benefits, including for biodiversity (high confidence). An increasing body of evidence
44   demonstrates that climatic risks to people, including from flood, drought, fire and over-heating, can be
45   lowered by a range of Ecosystem-based Adaptation techniques in urban and rural areas (medium confidence).
46   EbA forms part of a wider range Nature-based Solutions (NbS) actions and some have mitigation co-
47   benefits, including the protection and restoration of forests and other high-carbon ecosystems, as well as
48   agroecological farming practices {2.6.3; 2.6.5; Cross-Chapter Box NATURAL this Chapter}. However, EbA
49   and other NbS are still not widely implemented. {2.2;; 2.6.2; 2.6.3; 2.6.4; 2.6.5; 2.6.6, 2.6.7; Table
50   2.7; Cross-Chapter Box NATURAL this Chapter; Cross-Chapter Paper 1}.
52   To realise potential benefits and avoid harm, it is essential that EbA is deployed in the right places and
53   with the right approaches for that area, with inclusive governance (high confidence). Interdisciplinary
54   scientific information and practical expertise, including local and Indigenous knowledge, are essential to
55   effectiveness (high confidence). There is a large risk of maladaptation where this does not happen (high
56   confidence). {1.4.2; 2.2; 2.6; Table 2.7; Box 2.2; Figure Box 2.2.1; Cross-Chapter Box NATURAL this
57   Chapter; Cross-Chapter Paper1; 5.14.2}.

     Do Not Cite, Quote or Distribute                      2-7                                       Total pages: 237
     FINAL DRAFT                                    Chapter 2                  IPCC WGII Sixth Assessment Report

 2   Ecosystem-based Adaptation and other Nature-based Solutions are themselves vulnerable to climate
 3   change impacts. They need to take account of climate change adaptation if they are to remain effective and
 4   will increasingly be under threat at higher warming levels. Nature-based Solutions cannot be regarded as an
 5   alternative to, or a reason to delay, deep cuts in greenhouse gas emissions (high confidence). {2.6.3, 2.6.5;
 6   2.6.7; Cross-Chapter Box NATURAL this Chapter}
 8   Climate Resilient Development
10   Protection and restoration of natural and semi-natural ecosystems are key adaptation measures in
11   light of clear evidence that damage and degradation of ecosystems exacerbates the impacts of climate
12   change on biodiversity and people (high confidence). Ecosystem services that are at threat from a
13   combination of climate change and other anthropogenic pressures include climate change mitigation, flood
14   risk management, food provisioning and water supply (high confidence). Adaptation strategies that treat
15   climate, biodiversity and human society as coupled systems will be most effective. {2.3; Figure 2.1; 2.5.4;
16   2.6.2; 2.6.3; 2.6.7; Cross-Chapter Box NATURAL and Cross-Chapter Box ILLNESS this Chapter}
18   A range of analyses have concluded that ~30% of Earth's surface needs to be effectively conserved to
19   maintain biodiversity and ecosystem services (medium confidence). Climate change places additional
20   stress on ecosystem integrity and functioning, adding urgency for taking action. Low intensity,
21   sustainable management, including by Indigenous peoples, is an integral part of some protected areas and
22   can support effective adaptation and maintain ecosystem health. Food and fibre production in other areas
23   will need to be efficient, sustainable and adapted to climate change to meet the needs of the human
24   population (high confidence). {Figure 2.1; 2.5.4; 2.6.2; 2.6.3; 2.6.7}
26   Natural ecosystems can provide carbon storage and sequestration at the same time as providing
27   multiple other ecosystem services, including EbA (high confidence) but there are risks of
28   maladaptation and environmental damage from some approaches to land-based mitigation (high
29   confidence). Plantation forests in areas which would not naturally support forest, including savannas, natural
30   grasslands and temperate peatlands, or replacing native tropical forests on peat soils, have destroyed local
31   biodiversity and created a range of problems, including for water supply, food supply, fire risk and
32   greenhouse gas emissions. Large scale deployment of bioenergy, including Bioenergy with Carbon Capture
33   and Storage (BECCS) through dedicated herbaceous or woody bioenergy crops and non-native production
34   forests can damage ecosystems directly or through increasing competition for land. {2.6.3, 2.6.5, 2.6.6,
35   2.6.7; Box 2.2; Cross-Chapter Box NATURAL this Chapter; CCP7.3.2; Cross-Working Group Box
36   BIOECONOMY in Chapter 5}.
38   Terrestrial and aquatic ecosystems and species are often less degraded in lands managed by
39   Indigenous Peoples and local communities than in other lands (medium confidence). Including
40   indigenous and local institutions is a key element in developing successful adaptation strategies. Indigenous
41   and local knowledge contain a wide variety of resource-use practices and ecosystem stewardship strategies
42   that conserve and enhance both wild and domestic biodiversity. {2.6.5; 2.6.7; Cross-Chapter Box
43   NATURAL this Chapter}
45   Extreme events are compressing the timeline available for natural systems to adapt and impeding our
46   ability to identify, develop and implement solutions (medium confidence). There is now an urgent need to
47   build resilience and assist recovery following extreme events. This, combined with long-term changes in
48   baseline conditions means that implementation of adaptation and mitigation measures cannot be delayed if
49   they are to be fully effective. {2.3; Cross-Chapter Box EXTREMES this Chapter}

     Do Not Cite, Quote or Distribute                     2-8                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   2.1 Introduction
 3   2.1.1   Overview
 5   Chapter 2: We provide assessments of observed and projected impacts of climate change across species,
 6   biomes (vegetation types), ecosystems and ecosystem services, highlighting processes emerging on a global
 7   scale. Where sufficient evidence exists, differences in biological responses among regions, taxonomic groups
 8   or types of ecosystems are presented, particularly when such differences provide meaningful insights into
 9   current or potential future autonomous or human-mediated adaptations. Human interventions that might
10   build resilience of ecosystems and minimise negative impacts of climate change on biodiversity and
11   ecosystem functioning are assessed. Such interventions include adaptation strategies and programmes to
12   support biodiversity conservation and Ecosystem-based Adaptation (EbA). The assessments were done in the
13   context of the Convention on Biological Diversity (CBD) and Sustainable Development Goals (SDGs),
14   whose contributions to climate-resilient development pathways are assessed. This chapter highlights both
15   successes and failures of adaptation attempts and considers potential synergies and conflicts with land-based
16   climate change mitigation. Knowledge gaps and sources of uncertainty are included to encourage additional
17   research.
19   The Working Group II Summary for Policy Makers of the 5th Assessment Report (WGII AR5 SPM) stated
20   that “many terrestrial and freshwater species have shifted their geographic ranges, seasonal activities,
21   migration patterns, abundances, and species interactions in response to ongoing climate change” (IPCC,
22   2014e). Based on long-term observed changes across the regions, it was estimated that approximately 20–
23   30% of plant and animal species are at risk of extinction when global mean temperatures rise 2‒3°C above
24   preindustrial levels (Fischlin et al., 2007). In addition, WGII AR5 (IPCC, 2014f) broadly suggested that
25   autonomous adaptation by ecosystems and wild species might occur, and proposed human-assisted
26   adaptation to minimise negative climate change impacts.
28   Risk assessments for species, communities, key ecosystems and their services were based on the Risk
29   Assessment Framework introduced in the IPCC AR5 (IPCC, 2014). Assessments of observed changes in
30   biological systems emphasise detection and attribution of climate change on ecological and evolutionary
31   processes with an emphasis on freshwater ecosystems, and assess ecosystem processes that were lightly
32   assessed in previous reports, such as wildfire. Where appropriate, assessment of interactions between climate
33   change and other human activities is provided.
35   Land-use and land cover change (LULCC), and unsustainable exploitation of resources from terrestrial and
36   freshwater systems continue to be a major factor of natural ecosystem and biodiversity loss (high
37   confidence). Fertiliser input, pollution of waterways, dam construction and extraction of freshwater for
38   irrigation put additional pressure on biodiversity and alter ecosystem function (Shin et al., 2019). Likewise,
39   for biodiversity, invasive alien species have been identified as a major threat, especially in freshwater
40   systems, islands and coastal regions (high confidence) (IPBES, 2018b; IPBES, 2018e; IPBES, 2018c;
41   IPBES, 2018d; IPBES, 2019). Climate change and CO2 are expected to become increasingly important as
42   drivers of change over the coming decades (Ciais et al., 2013; Settele et al., 2014; IPBES, 2019; IPCC,
43   2019c).
45   2.1.2   Points of Departure
47   Species diversity and ecosystem function influence each other reciprocally, while ecosystem function forms
48   the necessary basis for ecosystem services (Hooper et al., 2012; Mokany et al., 2016). Drivers of impacts on
49   biodiversity, ecosystem function and ecosystem services have been assessed in reports from IPCC, Food and
50   Agriculture Organization (FAO), IPBES and the Global Environmental Outlook (Settele et al., 2014; FAO,
51   2018; IPBES, 2018b; IPBES, 2018e; IPBES, 2018c; IPBES, 2018d; IPBES, 2019; UNEP, 2019; Diversity,
52   2020). Most recently, the IPCC SRCCL provided an assessment on land degradation and desertification,
53   greenhouse gas emissions and food security in the context of global warming (IPCC, 2019c), and the IPBES-
54   IPCC joint report on Biodiversity and Climate Change provided a synthesis of current understanding of the
55   interactions, synergies and feedbacks between biodiversity and climate change (Pörtner et al., 2021). This
56   chapter builds on and expands the results from these assessments.

     Do Not Cite, Quote or Distribute                      2-9                                       Total pages: 237
     FINAL DRAFT                                    Chapter 2                   IPCC WGII Sixth Assessment Report

 1   Assessment of impacts of climate change on freshwater systems has been limited in previous assessments,
 2   and interlinkages between terrestrial and freshwater processes have not been fully explored (Settele et al.,
 3   2014; IPBES, 2019). Improved treatment of impacts on terrestrial and freshwater systems is critical
 4   considering the revisions of international sustainability Goals and Targets, especially the conclusion that
 5   many of the proposed post-2020 Biodiversity-targets of the Convention on Biological Diversity (CBD)
 6   cannot be met due to climate change impacts (Arneth et al., 2020).
 8   Previous reports highlighted the possibility of new ecosystem states stemming from shifts in thermal
 9   regimes, species composition, and energy and matter flows (Settele et al., 2014; Shin et al., 2019). Projecting
10   such “tipping points” (see glossary) has been identified in previous reports as a challenge since neither
11   monitoring programmes nor field studies, nor ecosystem and biodiversity modelling tools capture the
12   underlying species-species and species-climate interactions sufficiently well to identify how biological
13   interactions within and across trophic levels may amplify or dampen shifts in ecosystem states (Settele et al.,
14   2014; Shin et al., 2019). Building on these previous analyses and recent literature, Chapter 2 in this AR6
15   provides new insights compared to previous assessments by (i) emphasising freshwater aspects, and the
16   interlinkages between freshwater and terrestrial systems, (ii) assessing more clearly the link between
17   biodiversity and ecosystem functioning, (iii) assessing impacts associated with climate change mitigation
18   scenarios versus impacts of climate change, including interactions with adaptation, and (iv) where possible,
19   places findings in context of the United Nations Sustainable Development Goals (SDGs) 2030, and services
20   for human societies.
22   2.1.3   Guide to Attribution and Traceability of Uncertainty Assessments
24   For biological systems we use the framework for detection and attribution outlined in AR5 in which
25   attribution of observed biological changes is made not to global, but to local or regional climate changes,
26   (Parmesan et al., 2013; Cramer et al., 2014). However, global distribution of regional responses is desirable
27   to achieve generality, and data in prior reports were concentrated from the northern hemisphere. The critique
28   of "global" studies by (Feeley et al., 2017) argues that their naming is misleading, that most of them are far
29   from global and that considerable geographic and taxonomic bias remains. This bias is diminishing, as data
30   from southern hemisphere regions are added and there is now representation from every continent.
32   Overall confidence in climate change attribution of a biological change can be increased in multiple ways
33   (Parmesan et al., 2013), of which we list four here. First, confidence rises when the time span of biological
34   records is long, such that decadal trends in climate can be compared with decadal trends in biological
35   response and long-term trends can be statistically distinguished from natural variability. Secondly,
36   confidence can be increased by examining a large geographic area, which tends to diminish the effects of
37   local confounding factors (Parmesan et al., 2013; Daskalova et al., 2020). Third, confidence is increased
38   when there is experimental or empirical evidence of a mechanistic link between particular climate metrics
39   and biological response. Fourth, confidence is increased when particular fingerprints of climate change are
40   documented that uniquely implicate climate change as the causal driver of the biological change (Parmesan
41   and Yohe, 2003). These conditions constitute multiple lines of evidence which, when they converge, can
42   provide very high confidence that climate change is the causal driver of an observed change in a particular
43   biological species or system (Parmesan et al., 2013).
45   Important factors that may confound or obscure effects of climate change are presence of invasive species,
46   changes in land use (LULCC), and, in freshwater systems, eutrophication (IPCC, 2019a). Temporal and
47   spatial scale of the study also affect estimates of impacts. The most extreme published estimates of
48   biological change tend to be derived from smaller areas and/or shorter timeframes (Daskalova et al., 2020),
49   and a recent large global analysis of data for 12,415 species found that differences in study methodology
50   accounted for most of the explained variance in reported range shifts (Lenoir et al., 2020). The importance of
51   land-use change is frequently stressed, but there is a paucity of studies that actually quantify the relative
52   effects of climate change and land-use change on species and communities. Sirami et al. (2017) found only
53   13 such studies, among which four concluded that effects of land-use change over-rode those of climate
54   change, four found that the two drivers independently affected different species and five found that they
55   acted in synergy.

     Do Not Cite, Quote or Distribute                     2-10                                      Total pages: 237
     FINAL DRAFT                                        Chapter 2                       IPCC WGII Sixth Assessment Report

 1   2.2    Connections of Ecosystem Services to Climate Change
 3   Ecosystems provide services essential for human survival and well-being. The Millennium Ecosystem
 4   Assessment defined ecosystem services as “the benefits people obtain from ecosystems” including
 5   “provisioning services such as food and water; regulating services such as regulation of floods, drought, land
 6   degradation, and disease; supporting services such as soil formation and nutrient cycling; and cultural
 7   services such as recreational, spiritual, religious, and other nonmaterial benefits” (Assessment, 2005).
 9   The Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) re-named
10   the concept “nature’s contributions to people” and broadened the definition to “the contributions, both
11   positive and negative, of living nature (i.e. diversity of organisms, ecosystems, and their associated
12   ecological and evolutionary processes) to the quality of life for people. Beneficial contributions from nature
13   include such things as food provision, water purification, flood control, and artistic inspiration, whereas
14   detrimental contributions include disease transmission and predation that damages people or their assets”
15   (IPBES, 2019). IPBES modified the concept to include more social viewpoints and broaden analyses beyond
16   narrow economic stock-and-flow valuation approaches (Díaz et al., 2018). IPBES developed a classification
17   of 18 categories of ecosystem services (see Table 2.1).
19   When anthropogenic climate change affects ecosystems, it can also affect ecosystem services for people.
20   Climate change connects to ecosystem services through three links: climate change—species—ecosystems—
21   ecosystem services. This IPCC chapter assesses these connections through all three links when end-to-end
22   published scientific analyses are available for terrestrial and freshwater ecosystems. This type of robust
23   evidence exists for some key ecosystem services (Section 2.5.4), and is assessed in specific report sections:
24   biodiversity habitat creation and maintenance (Sections 2.4, 2.5), regulation of detrimental organisms and
25   biological processes (Sections,, 2.4.4, 2.5.3, 2.6.4, Cross-Chapter Box ILLNESS this
26   Chapter), regulation of climate through ecosystem feedbacks in terms of carbon storage (Sections,
27,, and albedo (Section, and provision of freshwater from ecosystems to
28   people (Section
30   For ecosystem services that do not have published scientific information to establish unambiguous links to
31   climate change, the climate—species—ecosystem links are assessed. Global ecological assessments,
32   including the Global Biodiversity Assessment (Programme, 1995), the Millennium Ecosystem Assessment
33   (Assessment, 2005), and the IPBES Global Assessment Report (IPBES, 2019) have synthesised scientific
34   information on the ecosystem—ecosystem services link, but full assessment from climate change to
35   ecosystem services is often impeded by limited quantitative studies that span this entire spectrum, see
36   (Mengist et al., 2020) for a review of this gap in montane regions.
38   IPCC and IPBES are collaborating to address gaps in knowledge on the effects of climate change on
39   ecosystem services (Services and Ecosystem, 2021). Table 2.1 provides a guide for finding information on
40   climate change and individual ecosystem services in the IPCC Sixth Assessment Report.
43   Table 2.1: Connections of ecosystem services to climate change, indicating the 18 categories of nature’s contributions
44   to people of the IPBES (IPBES, 2019), the most relevant sections in the IPCC Sixth Assessment Report, and the level
45   of evidence in this report for attribution to anthropogenic climate change of observed impacts on ecosystem services.
46   The order of services in the table follows the order presented by IPBES and does not denote importance or priority.
47   Connections denote observed impacts, future risks, and adaptation. The order of connections follows the relevance or
48   the order of sections.
      Ecosystem service                    Connections to climate change
      Habitat creation and                 Species extinctions (,, Species range shifts (,,
      maintenance                          Ecological changes in freshwater ecosystems (2.3.3,,,
                                 ,,, 2.5.4,,, Vegetation changes (2.4.3,
                                 ,,,, 2.5.2,, Biome shifts (,
                                           2.5.4), Wildfire (,, Tree mortality (,
                                           (robust evidence)

     Do Not Cite, Quote or Distribute                          2-11                                            Total pages: 237
FINAL DRAFT                                        Chapter 2                       IPCC WGII Sixth Assessment Report

 Pollination and dispersal of seeds   Species extinctions (,, Species range shifts (,,
 and other propagules                 Phenology changes (,
                                      (medium evidence)

 Regulation of air quality            Wildfire (,, Chapter 7), Tree mortality (,
                                      (medium evidence)

 Regulation of climate                Ecosystem carbon stocks, emissions, and removals (,, IPCC AR6
                                      Working Group I, Chapter 5), Amazon rainforest dieback (,,
                            ,,,, Tundra permafrost thaw (,,
                            , 2.5.4), biome shifts (2.4.3, 2.5.2,, Wildfire (,,
                                      Tree mortality (,, Primary productivity changes (,
                                      (robust evidence)

 Regulation of ocean acidification    Ocean acidification (IPCC AR6 Working Group I, Chapter 5), Changes in
                                      marine species distribution and abundance (Chapter 3)
                                      (robust evidence)

 Regulation of freshwater             Physical changes in freshwater systems (2.3.3), Ecological changes in
 quantity, location, and timing       freshwater ecosystems (,,,,, Tree
                                      mortality (,, Freshwater supply from ecosystems (
                                      (medium evidence)

 Regulation of freshwater and         Coastal ecosystem changes (Chapter 3), Physical changes in freshwater systems
 coastal water quality                (2.3.3), Ecological changes in freshwater ecosystems (;,
                                      (robust evidence)

 Formation, protection, and           Agricultural ecosystem changes (Chapter 5), Physical changes in freshwater
 decontamination of soils and         systems (2.3.1), Vegetation changes (2.4.3, 2.5.4), Wildfire (,
 sediments                            (medium evidence)

 Regulation of hazards and            Coastal ecosystem changes (Chapter 3), Vegetation changes (2.4.3, 2.5.2),
 extreme events                       Wildfire (,, Summary of hazards (2.3), Cross-Chapter Box
                                      EXTREMES this Chapter
                                      (medium evidence)

 Regulation of detrimental            Inter-species interactions (2.4.2), Control of disease vectors (, 2.5.1,
 organisms and biological             2.6.4), Insect pest infestations (, Cross-Chapter Box ILLNESS this
 processes                            Chapter
                                      (medium evidence)

 Energy                               Forestry plantation changes (Chapter 5), Biomass changes in natural
                                      ecosystems (, Bioeconomy (Cross-Working Group Box
                                      BIOECONOMY in Chapter 5), Tree mortality (,
                                      (limited evidence)

 Food and feed                        Agricultural ecosystem changes (Chapter 5), Species extinctions (,
                            , Species range shifts (, Nature-based services from natural
                                      ecosystems (Cross-Chapter Box NATURAL this chapter), shifts in commercial
                                      food species Cross-Chapter Box Moving Plate in Chapter 5)
                                      (medium evidence)

 Materials, companionship, and        Forestry plantation changes (Chapter 5), Species extinctions (,,
 labour                               Species range shifts (, Tree mortality (,
                                      (limited evidence)

 Medicinal, biochemical, and          Species extinctions (,, Species range shifts (
 genetic resources                    (limited evidence)

 Learning and inspiration             All observed impacts (2.4) and future risks (2.5) in terrestrial and freshwater
                                      (limited evidence)

Do Not Cite, Quote or Distribute                          2-12                                             Total pages: 237
     FINAL DRAFT                                      Chapter 2                      IPCC WGII Sixth Assessment Report

      Physical and psychological         All observed impacts (2.4) and future risks (2.5) in terrestrial and freshwater
      experiences                        ecosystems
                                         (limited evidence)

      Supporting identities              All observed impacts (2.4) and future risks (2.5) in terrestrial and freshwater
                                         (limited evidence)

      Maintenance of options             All observed impacts (2.4) and future risks (2.5) in terrestrial and freshwater
                                         ecosystems, Nature-based services from natural ecosystems (Cross-Chapter
                                         Box NATURAL this Chapter, Cross-Chapter Box DEEP in Chapter 17
                                         (limited evidence)
 3   2.3     Hazards and Exposure
 5   In AR6, WGI describes changes in physical climate systems using the term ‘climatic impact-drivers’ (CIDs),
 6   which can have detrimental, beneficial or neutral effects on a system. In contrast, the literature on natural
 7   systems tends to focus on hazards, which include natural or human-induced physical events, impacts or
 8   trends with the potential to cause negative effects on ecosystems and environmental resources. Hazards are
 9   affected by current and future changes in climate, including altered climate variability and extreme events
10   (WGI Chapter 12). Hazards can occur suddenly (e.g., a heat wave or heavy rain event), or more slowly (e.g.,
11   land loss, degradation and erosion linked to multiple climate hazards compounding). Observed exposure and
12   risks to protected areas is assessed in Section See also Cross-Chapter Box EXTREMES this
13   Chapter.
15   Non-climatic hazards such as land use change, habitat fragmentation, pollution and invasive species have
16   been the primary drivers of change in terrestrial and freshwater ecosystems in the past (high confidence)
17   (Figure 2.1). These impacts have been extensively documented in reports by the Intergovernmental Science-
18   Policy Platform on Biodiversity and Ecosystem Services (IPBES, 2021). However, whilst climate change has
19   not been the predominant influence to date, its relative impact is increasing (IPCC Special Report on Climate
20   Change and Land (SRCCL)), with greater interactive effects of non-climate and climate hazards now
21   occurring (Birk et al., 2020).
23   2.3.1    Observed Changes to Hazards and Extreme Events
25   The major climate hazards at the global level are generally well understood (WGI AR6 Chapter 12; WGI
26   AR6 Interactive Atlas). Increased temperatures and changes to rainfall and runoff patterns; greater variability
27   in temperature, rainfall, river flow and water levels; rising sea-levels and increased frequency of extreme
28   events means that greater areas of the world are being exposed to climate hazards outside those to which
29   they are adapted (high confidence) (Lange et al., 2020).
31   Extreme events are a natural and important part of many ecosystems and many organisms have adapted to
32   cope with long-term and short-term climate variability, within the disturbance regime experienced during
33   their evolutionary history (high confidence). However, climate changes, disturbance regimes change and the
34   magnitude and frequency of extreme events such as floods, droughts, cyclones, heatwaves and fire have
35   increased in many regions (high confidence). These disturbances affect ecosystem functioning, biodiversity
36   and ecosystem services (high confidence) but are, in general, poorly captured in impact models (Albrich et
37   al., 2020b), although this should improve as higher-resolution climate models that better capture smaller-
38   scale processes and extreme events become available (WGI AR6, Chapter 11). Extreme events pose large
39   challenges for Ecosystem-based Adaptation (IPCC Special Reports on Extremes, Section 2.6.3). Ecosystem
40   functionality on which such adaptation measures rely may be altered or destroyed by extreme episodic
41   events (Handmer et al., 2012; Lal et al., 2012; Pol et al., 2017).
43   There is high confidence that the combination of internal variability, superimposed on longer-term climate
44   trends, is pushing ecosystems to tipping points, beyond which abrupt and possibly irreversible changes are
45   occurring (Harris et al., 2018a; Jones et al., 2018; Hoffmann et al., 2019b; Prober et al., 2019; Berdugo et al.,

     Do Not Cite, Quote or Distribute                       2-13                                           Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1   2020; Bergstrom et al., 2021). Increases in the frequency and severity of heatwaves, droughts and aridity,
 2   floods, fires and extreme storms have been observed in many regions (Seneviratne et al., 2012; Ummenhofer
 3   and Meehl, 2017) and these trends are projected to continue (high confidence) (Section, Cross-
 4   Chapter Box EXTREMES this Chapter, AR6 WGI Chapter 11;, SR1.5, Hoegh-Guldberg et al., 2018b).
 6   While the major climate hazards at the global level are generally well described with high confidence, there
 7   is less understanding about the importance of hazards on ecosystems when they are superimposed (Allen et
 8   al., 2010; Anderegg et al., 2015; Seidl et al., 2017; Dean et al., 2018), and the outcomes are difficult to
 9   quantify in future projections (Handmer et al., 2012). Simultaneous or sequential events (coincident or
10   compounding events) can lead to an extreme event or impact, even if each event is not in themselves extreme
11   (Denny et al., 2009; Hinojosa et al., 2019). For example, the compounding effects of sea-level rise, extreme
12   coastal high tide, storm surge, and river flow can substantially increase flooding hazard and impacts on
13   freshwater systems (Moftakhari et al., 2017). On land, changing rainfall patterns and repeated heat waves
14   may interact with biological factors such as altered plant growth and nutrient allocation under elevated CO2,
15   affecting herbivory rates and insect outbreaks leading to widespread dieback of some forests (e.g. in
16   Australian Eucalypt forests) (Gherlenda et al., 2016; Hoffmann et al., 2019a). Risk assessments typically
17   only consider a single climate hazard without changing variability, potentially underestimating actual risk (
18   Milly et al., 2008; Sadegh et al., 2018; Zscheischler et al., 2018; Terzi et al., 2019; Stockwell et al., 2020; ).
20   Understanding impacts associated with the rapid rate of climate change is less developed and more uncertain
21   than changes in mean climate. High climate velocity (Loarie et al., 2009) is expected to be associated with
22   distribution shifts, incomplete range filling and species extinctions (high confidence) (Sandel et al., 2011;
23   Burrows et al., 2014), although not all species are equally at risk from high velocity (see Sections,
24 It is generally assumed that the more rapid the rate of change, the greater the impact on species and
25   ecosystems, but responses are taxonomically and geographically variable (high confidence) (Kling et al.,
26   2020).
28   For example, strong dispersers are less at risk, while species with low dispersal ability, small ranges and long
29   lifespans (e.g. many plants, especially trees, many amphibians and some small mammals) are more at risk
30   (Hamann et al., 2015) (IPCC, 2014). This is likely to favour generalist and invasive species, altering species
31   composition, ecosystem structure and function (Clavel et al., 2011; Büchi and Vuilleumier, 2014). The
32   ability to track suitable climates is substantially reduced by habitat fragmentation and human modifications
33   of the landscape such as dams on rivers and urbanisation (high confidence). Freshwater systems are
34   particularly at risk of rapid warming given their naturally fragmented distribution. Velocity of change in
35   surface temperature of inland standing waters globally has been estimated as 3.5 ± 2.3 km per decade from
36   1861 to 2005. This is projected to increase from 2006 to 2099 from between 8.7 ± 5.5 km per decade (RCP
37   2.6) to 57.0 ± 17.0 km per decade (RCP 8.5) (Woolway and Maberly, 2020). Although the dispersal of aerial
38   adult stage of some aquatic insects can surpass these climate velocities, rates of change under mid- and high
39   emissions scenarios (RCP4.5, RCP6.0, RCP8.5) are substantially higher than known rates of active dispersal
40   of many species (Woolway and Maberly, 2020). Many species, both terrestrial and freshwater, are not
41   expected to be able to disperse fast enough to track suitable climates under mid- and high emissions
42   scenarios (medium confidence) (RCP4.5, RCP6.0, RCP8.5; Brito-Morales et al., 2018).

     Do Not Cite, Quote or Distribute                      2-14                                        Total pages: 237
     FINAL DRAFT                                         Chapter 2                     IPCC WGII Sixth Assessment Report

 2   Figure 2.1: Map of global land use change from 1982-2016. Based on satellite records of global tree canopy (TC),
 3   short vegetation (SV) and bare ground (BG) cover (from Song et al., 2018). a) Mean annual estimates of cover (% of
 4   pixel area at 0.05° resolution). b) Long-term change estimates (% of pixel area at 1.5° resolution), with pixels showing
 5   a statistically significant trend (n = 35 years, two-sided Mann–Kendall test, P < 0.05) in TC, SV or BG. The dominant
 6   changes are Tree canopy gain with Short vegetation loss; Bare Ground gain with Short vegetation loss; Tree canopy
 7   gain with Bare Ground loss; Bare Ground gain with Tree canopy loss; 5, Short vegetation gain with Bare Ground loss;
 8   and Short vegetation gain with TC loss. Grey indicates areas with no significant change between 1982-2016.
11   2.3.2   Projected Impacts of Extreme Events
13   Understanding of the large-scale drivers and the local to regional feedback processes that lead to extreme
14   events is still limited and projections of extremes and coincident or compounding events remain uncertain
15   (Prudhomme et al., 2014; Sillmann et al., 2017; Hao et al., 2018; Miralles et al., 2019). Extreme events are
16   challenging to model because they are by definition rare and often occur at spatial and temporal scales much
17   finer than the resolution of climate models (Sillmann et al., 2017; Zscheischler et al., 2018). Additionally,
18   the processes that cause extreme events often interact, as is the case for drought and heat events, and are
19   spatially and temporally dependent, for example, as is the case in soil moisture and temperature (Vogel et al.,
20   2017). Understanding feedbacks between land and atmosphere also remains limited. For example, positive
21   feedbacks between soil and vegetation, or between evaporation, radiation and precipitation are important in
22   the preconditioning of extreme events such as heatwaves and droughts, increasing the severity and impact of
23   extreme events (Miralles et al., 2019).
25   Despite recent improvements in observational studies and climate modelling (Santanello et al., 2015;
26   Stegehuis et al., 2015; PaiMazumder and Done, 2016; Basara and Christian, 2018; Knelman et al., 2019), the
27   potential to quantify or infer formal causal relationships between multiple drivers and/or hazards remains

     Do Not Cite, Quote or Distribute                          2-15                                          Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1   limited, for several reasons (Zscheischler and Seneviratne, 2017; Kleinman et al., 2019; Miralles et al., 2019;
 2   Yokohata et al., 2019; Harris et al., 2020). The mechanisms underlying the response are difficult to identify
 3   (e.g., response to heat stress, drought, insects), effects vary among species and at different life stages, and an
 4   initial stress may influence the response to further stress (Nolet and Kneeshaw, 2018). Additionally, hazards
 5   such as drought are often exacerbated by societal, industrial and agricultural water demands, requiring more
 6   sophisticated modelling of the physical and human systems (Mehran et al., 2017; Wan et al., 2017).
 7   Observations of past compound events may not provide reliable guides to how future events may evolve,
 8   because human activity and recent climate change continue to interact to influence both system functioning
 9   and the climate state that have not been experienced in the past (see Chapter 11, WGI AR6).
11   2.3.3   Biologically Important Physical Changes in Freshwater Systems
13   Physical changes are fundamental drivers of change at all levels of biological organisation, from individual
14   species to communities to whole ecosystems and climate hazards specific to freshwater systems which are
15   not documented elsewhere in AR6 are summarised here.
17 Observed Change in Thermal Habitat and Oxygen Availability
19   Since AR5, evidence for changes in temperatures of lakes and rivers has continued to increase. Global
20   warming rates for lake surface waters were estimated as 0.21°C to 0.45°C per decade between 1970-2010,
21   exceeding SST trends of 0.09°C per decade between 1980-2017 (robust evidence, high agreement) (Figure
22   2.2; Schneider and Hook, 2010; Kraemer et al., 2015; O'Reilly et al., 2015; Woolway et al., 2020b).
23   Warming of lake surface water temperatures was variable within regions (O'Reilly et al., 2015) but more
24   homogeneous than changes of deep water temperature (Pilla et al., 2020). Because temperature trends in
25   lakes can vary vertically, horizontally, and seasonally, complex changes have occurred in the amounts of
26   habitat available to aquatic organisms at particular depths and temperatures (Kraemer et al., 2021).
28   Changes in river water temperatures ranged from -1.21 to +1.076 °C per decade between 1901-2010
29   (medium evidence, medium agreement) (Figure 2.2; Hari et al., 2006; Kaushal et al., 2010; Jurgelėnaitė et al.,
30   2012; Li et al., 2012; Latkovska and Apsīte, 2016; Marszelewski and Pius, 2016). The more rapid increase in
31   surface water temperature in lakes and rivers in regions with cold winters (O'Reilly et al., 2015) can in part
32   be attributed to the amplified warming in polar and high latitude regions (robust evidence, high agreement)
33   (Figure 2.2b; Screen and Simmonds, 2010; Stuecker et al., 2018).
35   Shifts in thermal regime: Since AR5 the trend that lake waters mix less frequently continues (Butcher et al.,
36   2015; Adrian et al., 2016; Richardson et al., 2017; Woolway et al., 2017). This results from greater warming
37   of surface temperatures relative to deep water temperatures and the loss of ice during winter which prevents
38   inverse thermal stratification in north temperate lakes (robust evidence, high agreement) (Adrian et al., 2009;
39   Winslow et al., 2015; Adrian et al., 2016; Schwefel et al., 2016; Richardson et al., 2017).
41   Oxygen availability: Increased water temperature and reduced mixing cause a decrease in dissolved oxygen.
42   In 400 lakes, dissolved oxygen in surface and deep waters declined by 4.1 and 16.8%, respectively between
43   1980 and 2017 (Jane et al., 2021). The deepest water layers are expected to experience an increase in
44   hypoxic conditions by more than 25% due to reduced winter mixing and fewer complete mixing events, with
45   strong repercussions on nutrient dynamics and loss of thermal habitat (robust evidence, high agreement)
46   (Straile et al., 2010; Zhang et al., 2015; Schwefel et al., 2016; Adrian et al., 2016; Kraemer et al., 2021).

     Do Not Cite, Quote or Distribute                       2-16                                       Total pages: 237
     FINAL DRAFT                                          Chapter 2                     IPCC WGII Sixth Assessment Report

 2   Figure 2.2: Observed global trends in lake and river surface water temperature. a) Left panel: Map of
 3   temperatures of lakes (1970-2010). b) Left panel: Map of temperatures of rivers (1901-2010). Note that the trends of
 4   river water temperatures are not directly comparable within rivers or directly comparable to lakes since time periods are
 5   not consistent across river studies. Right panels in a) and b) depict water temperature trends along a latitudinal gradient
 6   highlighting the above average warming rates in northern Polar Regions (polar amplification). Data sources for lakes:
 7   (O'Reilly et al., 2015; Carrea and Merchant, 2019; Woolway et al., 2020a; Woolway et al., 2020b). Data sources for
 8   rivers: (Webb and Walling, 1992; Langan et al., 2001; Daufresne et al., 2004; Moatar and Gailhard, 2006; Lammers et
 9   al., 2007; Patterson et al., 2007; Webb and Nobilis, 2007; Durance and Ormerod, 2009; Kaushal et al., 2010; Pekárová
10   et al., 2011; Jurgelėnaitė et al., 2012; Markovic et al., 2013; Arora et al., 2016; Latkovska and Apsīte, 2016;
11   Marszelewski and Pius, 2016; Jurgelėnaitė et al., 2017).
14   Observed Changes in Water Level
16   Depending on how the intensification of the global water cycle affects individual lake water budgets, the
17   amount of water stored in specific lakes may increase, decrease, or have no substantial cumulative effect
18   (Notaro et al., 2015; Pekel et al., 2016; Rodell et al., 2018; Busker et al., 2019; Woolway et al., 2020b). The
19   magnitude of hydrological changes that can be assuredly attributed to climate change remains uncertain
20   (Hegerl et al., 2015; Gronewold and Rood, 2019; Kraemer et al., 2020). Attribution of water storage
21   variation in lakes due to climate change is facilitated when such variations occur coherently across broad
22   geographic regions and long timescales, preferably absent other anthropogenic hydrological influences
23   (Watras et al., 2014; Kraemer et al., 2020). There is increasing awareness that climate change contributes to
24   the loss of small temporary ponds, which cover a greater global area than lakes (Bagella et al., 2016).
26   Lakes fed by glacial meltwater are growing in response to climate change and glacier retreat (robust
27   evidence, high agreement) (Shugar et al., 2020). Water storage increases on the Tibetan Plateau (Figure 2.3a)
28   have been attributed to changes in glacier melt, permafrost thaw, precipitation and runoff, in part as a result
29   of climate change (Huang et al., 2011; Meng et al., 2019; Wang et al., 2020a). High confidence in attribution
30   of these trends to climate change is supported by long-term ground survey data and observations from the
31   GRACE satellite mission (Ma et al., 2010; Rodell et al., 2018; Kraemer et al., 2020).
33   In the Arctic, lake area has increased in regions with continuous permafrost and decreased in regions where
34   permafrost is thinner and discontinuous (robust evidence, high agreement) (see Chapter 4; Smith et al., 2005;
35   Andresen and Lougheed, 2015; Nitze et al., 2018; Mekonnen et al., 2021).

     Do Not Cite, Quote or Distribute                           2-17                                           Total pages: 237
     FINAL DRAFT                                       Chapter 2                     IPCC WGII Sixth Assessment Report


 4   Figure 2.3: Change in water extent in the Tibetan Plateau and annual mean global river flow. a) Changes in
 5   water storage on the Tibetan Plateau. Map of the Qinghai-Tibetan Plateau, Asia showing the percent change in surface
 6   water extent from 1984-2019 based on LANDSAT imagery. Increases in surface water extent in this region are mainly
 7   caused by climate-change mediated increases in precipitation and glacial melt (Source: EC JRC/Google; (Pekel et al.,
 8   2016). b) Global map of the median trend in annual mean river flow derived from 7250 observatories around the world
 9   (period 1971-2010). Some regions are drying (northeast Brazil, southern Australia, and the Mediterranean) and others
10   are wetting (northern Europe) mainly caused by large-scale shifts in precipitation, changes in factors that influence
11   evapotranspiration and alterations of the timing of snow accumulation and melt driven by rising temperatures (Source:
12   Gudmundsson et al., 2021).
15   Observed Changes in Discharge
17   Analysis of river flows from 7,250 observatories around the world covering the years 1971 to 2010 and
18   identified spatially complex patterns, with reductions in northeastern Brazil, southern Australia and the

     Do Not Cite, Quote or Distribute                         2-18                                        Total pages: 237
     FINAL DRAFT                                      Chapter 2                    IPCC WGII Sixth Assessment Report

 1   Mediterranean and increases in northern Europe (medium evidence, medium agreement) (Figure 2.3b;
 2   Gudmundsson et al., 2021). More than half of global rivers undergo periodic drying that reduces river
 3   connectivity (medium evidence, medium agreement). Increased frequency and intensity of droughts may
 4   cause perennial rivers to become intermittent and intermittent rivers to disappear (medium evidence, medium
 5   agreement), threatening freshwater fish in habitats already characterised by heat and droughts (Datry et al.,
 6   2016; Schneider et al., 2017; Jaric et al., 2019). In high altitude/latitude streams, reduced glacier and
 7   snowpack extent, earlier snowmelt and altered precipitation patterns, attributed to climate change, have
 8   increased flow intermittency (Vorosmarty et al., 2010; Siebers et al., 2019; Gudmundsson et al., 2021)
 9   Patterns in flow regimes can be directly linked to a variety of processes shaping freshwater biodiversity,
10   hence any climate-change induced changes on flow regimes and river connectivity are expected to alter
11   species composition, as well as having societal impacts (See Chapter 3 of IPCC SR1.5; Bunn and
12   Arthington, 2002; Thomson et al., 2012; Chessman, 2015; Kakouei et al., 2018)
14   Observed Loss of Ice
16   Studies since AR5 have confirmed ongoing and accelerating loss of lake and river ice in the northern
17   hemisphere (robust evidence, high agreement) (Figure 2.4). In recent decades, systems have been freezing
18   later in winter and thawing earlier in spring, reducing ice duration by >2 weeks per year and leading to
19   increasing numbers of years with loss of perennial ice cover, leading to intermittent ice-cover or even
20   absence of ice (Adrian et al., 2009; Kirillin et al., 2012; Paquette et al., 2015; Adrian et al., 2016; Park et al.,
21   2016; Roberts et al., 2017; Sharma et al., 2019). The global extent of river ice declined by 25% between
22   1984 and 2018 (Yang et al., 2020). This trend has been more pronounced at higher latitudes, consistent with
23   enhanced polar warming (large geographic coverage) (Du et al., 2017). Empirical long-term and remote
24   sensing data gathered in an increasingly large number of freshwater systems supports very high confidence in
25   attribution of these trends to climate change. For declines of glaciers, snow and permafrost see AR6 WGII
26   Chapter 4 and SROCC report.

     Do Not Cite, Quote or Distribute                       2-19                                        Total pages: 237
     FINAL DRAFT                                         Chapter 2                     IPCC WGII Sixth Assessment Report

 2   Figure 2.4: Global ice cover trends of lakes and rivers. a) Spatial distribution of current (light grey areas) and future
 3   (coloured areas) Northern Hemisphere lakes that may experience intermittent winter ice cover with climate warming.
 4   Projections were based on current conditions (1970-2010) and four established air temperature projections (Data
 5   source: (Sharma et al., 2019). b) Spatial distribution of projected change in Northern Hemisphere river ice duration
 6   under the RCP 4.5 emission scenario by 2080-2100 relative to the period 2009-2029. White areas refer to rivers without
 7   ice cover in the period 2009-2029 (zero days). Reference period isolines indicate river ice duration in the period 2009-
 8   2029. Coloured areas depict loss of ice duration in days. Blue areas depict a projected increase in river ice duration.
 9   Grey land areas indicate a lack of Landsat-observable rivers (Data source: Yang et al., 2020).
12   Extreme Weather Events and Freshwater Systems
14   Since AR5, numerous drastic short-term responses have been observed in lakes and rivers, both to expected
15   seasonal extreme events and to unexpected supra-seasonal extremes extending over multiple seasons.
16   Consequences for ecosystem functioning are not well understood (Bogan et al., 2015; Death et al., 2015;
17   Stockwell et al., 2020). Increasing frequencies of severe floods and droughts attributed to climate change are
18   major threats for river ecosystems (Peters et al., 2016; Alfieri et al., 2017). While extreme floods cause

     Do Not Cite, Quote or Distribute                          2-20                                          Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   massive physical disturbance, moderate floods can have positive effects, providing woody debris that
 2   contributes to habitat complexity and diversity, flush fine sediments, dissolving organic carbon and
 3   providing important food sources from terrestrial origins (Peters et al., 2016; Talbot et al., 2018). Droughts
 4   reduce river habitat diversity and connectivity, threatening aquatic species, especially in deserts and arid
 5   regions (Bogan et al., 2015; Death et al., 2015; Ledger and Milner, 2015; Jaric et al., 2019).
 7   Rivers already stressed by human activities such as urban development and farming on floodplains are prone
 8   to reduced resilience to future extreme events (medium confidence) (Woodward et al., 2016; Talbot et al.,
 9   2018). Thus, potential for floods to become catastrophic for ecosystem services are exacerbated by land-use
10   changes (Peters et al., 2016; Talbot et al., 2018). However, biota can recover rapidly from extreme flood
11   events if river geomorphology is not reformed. If instream habitat is strongly affected, recovery, if it occurs,
12   takes much longer, resulting in decline in biodiversity (medium confidence) (Thorp et al., 2010; Death et al.,
13   2015; Poff et al., 2018).
15   However, not all extreme events will have a biological impact, depending in particular on the timing,
16   magnitude, frequency of events and the antecedent conditions (Bailey and van de Pol, 2016; Stockwell et al.,
17   2020; Jennings et al., 2021; Thayne et al., 2021). For instance, an extreme wind event may have little impact
18   on phytoplankton in a lake, which was fully mixed prior to the event. Conversely, storm effects on
19   phytoplankton communities may compound when lakes are not yet recovered from a previous storm or if
20   periods of drought alternate with periods of intense precipitation (limited evidence) (Leonard, 2014;
21   Stockwell et al., 2020).
23   In summary, extreme events (heat waves, storms, loss of ice) affect lakes in terms of water temperature,
24   water level, light, oxygen concentrations and nutrient dynamics, that in turn affect primary production, fish
25   communities and greenhouse gas emissions (high confidence). These impacts are modified by levels of solar
26   radiation, wind speed and precipitation (Woolway et al., 2020a). Droughts have negative impact on water
27   quality in streams and lakes by increasing water temperature, salinity, the frequency of algal blooms and
28   contaminant concentrations, and reducing concentrations of nutrients and dissolved oxygen (medium
29   confidence) (Peters et al., 2016; Alfieri et al., 2017) (Woolway et al., 2020a). Understanding how these
30   pressures subsequently cascade through freshwater ecosystems will be essential for future projections of
31   their resistance and resilience towards extreme events (Leonard, 2014; Stockwell et al., 2020). See Table
32   SM2.1 for specific examples of observed changes.
34   Projected Changes in Physical Characteristics of Lakes and Rivers
36   Given the strength of relationship between past GAST and warming trends at lake surfaces (Figure 2.2;
37   section, and projected increases in heatwaves, surface water temperatures are projected to continue
38   to increase (Woolway et al., 2021). Mean May to October lake surface temperatures in 46,557 European
39   lakes were projected to be 2.9°, 4.5°, and 6.5°C warmer by 2081-2099 compared to historic (1981-1999)
40   under RCPs 2.0, 6.0, and 8.5, respectively (Woolway et al., 2020a). Under RCP 2.6, average lake heatwave
41   intensity increases from 3.7° to 4.0°C and average duration from 7.7 to 27.0 days, relative to the historical
42   period (1970-1999). For RCP 8.5, warming increases to 5.4°C and duration increases dramatically to 95.5
43   days (medium confidence) (Woolway et al., 2021).
45   Worldwide alteration of lake mixing regimes in response to climate change are projected (Kirillin, 2010).
46   Most prominently, monomictic lakes–undergoing one mixing event in most years–will become permanently
47   stratified, while lakes that are currently dimictic–mixing twice per year–will become monomictic by 2080-
48   2100 (medium confidence) (Woolway and Merchant, 2019). Nevertheless, predicting mixing behavior
49   remains an important challenge and attribution to climate change remains difficult (Schwefel et al., 2016;
50   Bruce et al., 2018).
52   Under climate projections of 3.2°C warming, 4.6% of the ice covered lakes in the northern hemisphere could
53   switch to intermittent winter ice cover (Figure 2.4a; Sharma et al., 2019). Unfrozen and warmer lakes lose
54   more water to evaporation (Wang et al., 2018b). By 2100, global annual lake evaporation will increase by
55   16%, relative to 2006-2015, under RCP 8.5 (Woolway et al., 2020b). Moreover, melting of ice decreases the
56   ratio of sensible to latent heat flux, thus channelling more energy into evaporation (medium confidence)

     Do Not Cite, Quote or Distribute                      2-21                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   (Wang et al., 2018b). Between 2009-2029 and 2080-2100, average river ice duration is projected to decline
 2   by 7.3 and 16.7 days under the RCP 4.5 and RCP 8.5 (Figure 2.4b; Yang et al., 2020).
 4   Projections of lake water are limited by the absence of reliable, long-term, homogenous, and spatially
 5   resolved hydrologic observations (Hegerl et al., 2015). This uncertainty is reflected in the widely divergent
 6   projections for lake water storage in response to future climate changes in individual lakes (Angel and
 7   Kunkel, 2010; Malsy et al., 2012) (MacKay and Seglenieks, 2012; Notaro et al., 2015). Selecting models that
 8   perform well when comparing hindcasted to observed past water storage variation often does little to reduce
 9   water storage projection uncertainty (Angel and Kunkel, 2010). This wide range of potential changes
10   complicates lake management. For information on observed and projected changes in the global water cycle
11   and hydrological regimes for streams, lakes, wetland, groundwater and their implications on water quality
12   and societies, see Chapter 4 of WGII and Chapter 8 of WGI. For the role of weather and climate extremes on
13   the global water cycle, see Chapter 11 of WGI.
15   In summary, with ongoing climate warming and an increase in the frequency and intensity of extreme
16   events, observed increases in water temperature, losses of ice and shifts in thermal regime, are projected to
17   continue (high confidence).
22   Cross-Chapter Box EXTREMES: Ramifications of Climatic Extremes for Marine, Terrestrial,
23        Freshwater and Polar Natural Systems
25   Authors: Rebecca Harris (Australia, Chapter 2, CCP3), Philip Boyd (Australia, Chapter 3), Rita Adrian
26   (Germany, Chapter 2), Jörn Birkmann (Germany, Chapter 8), Sarah Cooley (USA, Chapter 3), Simon
27   Donner (Canada, Chapter 3), Mette Mauritzen (Norway, Chapter 3), Guy Midgley (South Africa, Chapter
28   16); Camille Parmesan (France/USA/United Kingdom, Chapter 2), Dieter Piepenburg (Germany, Chapter
29   13, CCP6), Marie-Fanny Racault (United Kingdom/France, Chapter 3), Björn Rost (Germany, Chapter 3,
30   CCP6), David Schoeman (Australia, Chapter 3), Maarten van Aalst (The Netherlands, Chapter 16).
32   Introduction
34   Extreme events are now causing profound negative effects across all realms of the world (marine, terrestrial,
35   freshwater and polar) (medium confidence) (WGI, Chapter 9, 11; WGII AR6 Section 2.3.1, 2.3.2,,
36   Chapter 3, Chapters 9–12). Changes to population abundance, species distributions, local extirpations and
37   extinctions are leading to long-term, potentially irreversible shifts in the composition, structure and function
38   of natural systems (medium confidence) (Frolicher and Laufkotter, 2018; Harris et al., 2018a; Maxwell et al.,
39   2019; Smale et al., 2019). These effects have widespread ramifications for ecosystems and the services they
40   provide – physical habitat, erosion control, carbon storage, nutrient cycling and water quality, with knock-on
41   effects on tourism, fisheries, forestry and other natural resources (Kaushal et al., 2018; Heinze et al., 2021;
42   Pörtner et al., 2021).
44   Increasingly, the magnitude of extreme events is exceeding values projected for mean conditions for 2100,
45   regardless of emissions scenario (Figure Cross-Chapter Box EXTREMES.1). This has collapsed the timeline
46   organisms and natural communities have to acclimate or adapt to climate change (medium confidence).
47   Consequently, rather than having decades to identify, develop and adopt solutions, there is now an urgent
48   need to build resilience and assist recovery following extreme events.
50   Recent extremes highlight characteristics that enable natural systems to resist or recover from events, helping
51   natural resource managers to develop solutions to improve resilience of natural communities and identify
52   limits to adaptation (Bergstrom et al., 2021).
54   Marine Heatwaves
56   Consensus is emerging that anthropogenic climate change has significantly increased the likelihood of recent
57   marine heat waves (MHWs) (medium confidence) (WGI AR6, Chapter 9; Oliver et al., 2018). A widespread

     Do Not Cite, Quote or Distribute                      2-22                                      Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1   MHW occurred in the NE Pacific between 2013–2015, with upper ocean temperature anomalies of up to
 2   6.2°C relative to 2002-2012 (Gentemann et al., 2017). This event, termed the “Blob”, enhanced surface
 3   water stratification, decreasing nutrient supply, primary and community production, leading to widespread
 4   changes to open ocean and coastal ecosystems, with geographical shifts of key species across trophic levels,
 5   mass strandings of marine mammals, seabird mortalities and closures of commercially important fisheries
 6   (Cavole et al., 2016; Piatt et al., 2020). The heatwave reappeared in 2019 (“Blob 2.0”) (Amaya et al., 2020),
 7   with similarly high temperature anomalies extending from Alaska to California, but the ecological effects of
 8   this event are expected to differ because the Blob originated in winter, and Blob2.0 intensified in summer
 9   (Amaya et al., 2020). Modelling suggests rapid shifts in the geographic distributions of important fish
10   species in response to MHWs (Cheung & Frölicher, 2020), with projected decreased biomass and
11   distributional shifts of fish at least four times faster and larger than the effects of decadal-scale mean changes
12   throughout the 21st century under RCP8.5 (high confidence) (Cheung & Frölicher, 2020). Marine heatwaves
13   can also dramatically increase CH4 emissions from oceans, a significant positive feedback on global
14   warming (See also Chapter 3; Borges et al., 2019).
16   The Arctic region is warming more than twice as fast as the global mean, and polar organisms and
17   ecosystems are likely to be particularly vulnerable to heatwaves due to their specific thermal niches and
18   physiological thresholds and the lack of poleward ‘refugia’ (high confidence). The consequences of MHWs
19   are exacerbated by concomitant sea-ice melting and freshening of surface waters, leading to secondary
20   effects due to osmotic stress and failing pH homeostasis. Since sea-ice associated organisms are often critical
21   components of polar food chains, cascading effects up to top predators are expected. In 2015–2016 a MHW
22   occurred in the Gulf of Alaska/Bering Sea (Walsh et al., 2018) which was unprecedented in terms of surface
23   temperatures and ocean heat content, geographical extent, depth range and persistence, impacting the entire
24   marine food web. Persistent warming favoured some phytoplankton species and triggered one of the largest
25   algal blooms recorded in this region, with concomitant oyster farm closures due to uncommon paralytic
26   shellfish poisoning events (Walsh et al., 2018). There were also massive die-offs of common murres (Uria
27   aalge) and puffins (Fratercula cirrhata), attributed to starvation resulting from warming-induced effects on
28   food supply (Jones et al., 2019). A 2017 survey found a 71% decline in abundance of Pacific cod (Gadus
29   macrocephalus) since 2015, likely due to an increase in metabolic demand and reduced prey supply during
30   the MHWs (Barbeaux et al., 2020).
32   Terrestrial Heatwaves
34   Heatwaves are now regularly occurring that exceed the physiological thresholds of some species, including
35   birds and other small endotherms such as flying-foxes (high confidence) (Sections,
36   Heatwaves in Australia, North America and southern Africa have caused mass mortality events due to lethal
37   hyperthermia and dehydration (Saunders et al., 2011; Conradie et al., 2020; McKechnie et al., 2021),
38   reducing fitness (du Plessis et al., 2012; Andrew et al., 2017; Sharpe et al., 2019; van de Ven et al., 2019;
39   van de Ven et al., 2020), breeding success and recruitment (Kennedy et al., 2013; Wiley and Ridley, 2016;
40   Ratnayake et al., 2019) and affecting daily activity and geographic distributions (Albright et al., 2017). They
41   also place enormous demands on wildlife management agencies and pose human health risks (Welbergen et
42   al., 2008).
44   Recent mortality events affected 14 species of bird and fruit bats (Epomophorus wahlbergi) in South Africa
45   when maximum air temperatures exceeded 43–45°C in 2020 (McKechnie et al., 2021). Passerine birds seem
46   more vulnerable to lethal hyperthermia due to the relative inefficiency of panting to lose heat (McKechnie et
47   al., 2021) and their small size, as heat tolerance generally increases with body mass (McKechnie et al.,
48   2017). Several mass mortality events of flying-foxes (Pteropus poliocephalus, P. alecto) have occurred in
49   eastern Australia when maximum air temperatures exceeded 42℃ (Welbergen et al., 2008). Nineteen such
50   events occurred between 1994 and 2008, compared to three events prior to 1994. In January 2002, maximum
51   temperatures exceeded the 30-year average mean daily maximum by up to 16.5℃ and killed more than 3500
52   individuals (Welbergen et al., 2008). In 2014, an estimated 45,500 flying-foxes died in a single day, when
53   average maximum temperatures were 8°C or more above average (Meteorology, 2014). Drought compounds
54   the impacts, as mortality increases when water availability is low (Welbergen et al., 2008; Mo and Roache,
55   2020; McKechnie et al., 2021).

     Do Not Cite, Quote or Distribute                       2-23                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   Antarctica encountered its first recorded heatwave in 2020. Record-high temperatures occurred in East
 2   Antarctica (Robinson et al., 2020), with a maximum (9.2℃) temperature ~7℃ above the mean maximum,
 3   and minimum temperatures > 0℃. Record-high temperatures (18.3℃) were also recorded in West Antarctica
 4   (Robinson et al., 2020). It is too soon to know the impacts on polar life, but such abrupt heating is expected
 5   to have wide-ranging effects on biota, from flash-flooding and dislodgement of plants, to excess melt waters
 6   supplying moisture to arid polar ecosystems (CCP Polar). Heatwaves in Siberia in 2016, 2018 and 2020,
 7   with air temperature anomalies >6℃, were associated with extensive wildfires, pest infestations and melting
 8   permafrost (Overland and Wang, 2021).
10   Freshwater Extremes
12   Heatwaves, storms and floods affect the thermal regime and biogeochemical functioning of lakes and rivers
13   (Woolway and Merchant, 2017; Vicente-Serrano et al., 2020). Extreme heatwaves lead to abnormally high
14   water temperatures (Till et al., 2019) and reduce mixing of lakes (Woolway et al., 2021), causing a decrease
15   in oxygen and deep-water oxygen renewal (Zhang et al., 2015). Ectotherms such as fish and invertebrates are
16   particularly susceptible to such temperature and oxygen stress (Stoks et al., 2014). Their metabolic demands
17   increase with rising temperature and suitable habitat is eroded due to both high temperatures and lower
18   oxygen concentrations in lakes and rivers. Till et al. (2019) attributed 502 fish kill events in Wisconsin lakes
19   (USA) to warmer summers in lakes that experienced abnormally high water temperatures. Such events are
20   predicted to double by 2041–2059 and increase fourfold by 2081–2099 compared to historical levels (Till et
21   al., 2019). This anticipated increase in die-offs may facilitate warm-water fish species displacing cool-water
22   species (Hansen et al., 2017; Jennings et al., 2021). Floods mobilise nutrients and sediment, and aid dispersal
23   of invasive species in rivers (Death et al., 2015), while drought extremes reduce river connectivity,
24   threatening biodiversity in rivers ( Section; Tickner et al., 2020).
26   Learnings from Recent Extremes
28   These examples show that the impact of an extreme event is a function of its characteristics and those of the
29   exposed ecosystem. The timing, frequency, absolute magnitude and geographic extent of the extreme event,
30   relative to antecedent conditions, the life-cycle, resistance and resilience of the natural community, all
31   determine the biological response (Figure Cross-Chapter Box EXTREMES.2; Hillebrand et al., 2018; Gruber
32   et al., 2020). Impacts appear to be greater when extreme events occur more frequently, particularly when the
33   interval between events is insufficient to allow recovery to previous population sizes (e.g. frequent fire, coral
34   bleaching), or coincides with vulnerable life cycle stages, even when populations are adapted to cope with
35   such disturbances. Events occurring over large spatial areas reduce the potential for recolonisation from
36   nearby populations (e.g. regional droughts causing widespread declines). Often the magnitude of extreme
37   events exceeds historical levels, so organisms are less likely to be adapted to them, particularly when several
38   extremes coincide (e.g. high water temperatures, drought) (Duke et al., 2017). When hazards occur
39   simultaneously (compound events), impacts of extremes can be substantially aggravated, triggering
40   cascading effects in ecosystems (Gruber et al., 2020).

     Do Not Cite, Quote or Distribute                      2-24                                       Total pages: 237
     FINAL DRAFT                                        Chapter 2                     IPCC WGII Sixth Assessment Report

 2   Figure Cross-Chapter Box EXTRMES.1: A conceptual illustration of how extinction risk is affected by changes in
 3   the frequency, duration and magnitude of extreme weather or climate events (e.g,. drought, fire, flood, heatwaves).
 4   Many organisms have adapted to cope with long-term and short-term climate variability, but as the magnitude and
 5   frequency of extreme events increases, superimposed on the long-term climate trend, the threshold between survivable
 6   extreme weather events (yellow) and extremes that have high risk of causing population or species extinctions (red) is
 7   crossed more frequently. This can lead to local extinction events with insufficient time between to enable recovery,
 8   resulting in long-term, irreversible changes to the composition, structure and function of natural systems. When the
 9   extreme event occurs over a large area relative to the distribution of a species (e.g., a hurricane impacting an island
10   which is the only place a given species occurs), a single extreme event can drive the global extinction of a species.
13   Several characteristics of natural systems are associated with greater vulnerability to extreme events (Figure
14   Cross-Chapter Box EXTREMES.2), knowledge of which can inform solutions to build resilience and aid
15   recovery (Robinson et al., 2020). Resilience can be built prior to an event by minimising additional
16   disturbances, such as water extraction from river systems, pollution of aquatic systems, land-use change and
17   fragmentation. Managing landscapes to reduce fragmentation and increase habitat extent, connectivity and
18   heterogeneity, by increasing the number and extent of reserves, may provide local refugia from extreme
19   events and enhance post-event recolonisation, but may be less effective for marine systems (Chapter 3,
20   Section 3.6). Maintaining taxonomic, phylogenetic and functional diversity is important, as more diverse
21   systems may be more stable in the face of disturbance (Pimm, 1984; García-Palacios et al., 2018).
23   Several characteristics increase vulnerability: low or narrow thermal tolerances, high habitat specificity, low
24   dispersal ability, long generation times, low competitive ability and lifecycle constraints that limit recovery
25   or recolonisation. Populations living close to one or more limiting factors near range edges are also
26   vulnerable (Arafeh-Dalmau et al., 2019). Understanding these characteristics can inform management
27   intervention to aid recovery following an extreme event. For instance, knowledge of the flying-fox’s
28   physiological temperature threshold led to successful interventions, including misting of populations to
29   reduce mortality (Mo and Roache, 2020) and the development of a ‘heat stress forecaster’, an online tool
30   which uses weather forecasts to identify roosts at risk of extreme heat events (Ratnayake et al., 2019). This
31   early-warning system increases the preparedness of wildlife management and conservation agencies,
32   enabling efficient allocation of management resources towards locations that are likely to be most affected.
33   Monitoring following extreme events can help identify immediate impacts and the potential for cascading
34   interactions, such as changes to competitive interactions following range shifts, impacts on freshwater
35   ecosystems following wildfires and the spread of invasive species. Ongoing monitoring of recovery and
36   effectiveness of management intervention is important, focussing on habitat-forming species (eg. kelp,
     Do Not Cite, Quote or Distribute                          2-25                                         Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   corals, dominant tree species) and keystone species (eg. filter feeders, macrophytes, top predators), as the
 2   loss of these species can lead to ecosystem tipping points, beyond which the system may not recover
 3   (Section 2.5.3; SROCC Chapter 6, Section; AR6 sectoral chapters, Chapter 3 section
 4   The acute impacts of extreme events, in addition to the chronic stress of changing mean conditions, are
 5   accelerating and amplifying the biological effects of climate change. This amplification is being observed
 6   globally and in all realms where life exists. Extreme events are compressing the timeline available for natural
 7   systems to adapt and impeding our ability to identify, develop and adopt solutions. Recent events highlight
 8   the urgent need to mitigate global greenhouse gas emissions and identify solutions to halt accelerating
 9   impacts on natural systems (Díaz et al., 2020).

13   Figure Cross-Chapter Box EXTREMES.2: Characteristics of natural systems that affect vulnerability and help
14   identify solutions – both prior to and after extreme events - to build resistance, resilience and recovery.
20   2.4     Observed Impacts of Climate Change on Species, Communities, Biomes, Key Ecosystems and
21           their Services
23   2.4.1    Overview
25   Global meta-analyses of terrestrial systems in AR3 and AR4 concentrated on long time frames (>20 years)
26   and findings from relatively undisturbed areas, where confidence in attributing observed changes to climate
27   change is high. Recent global and regional meta-analyses (AR5 and later) have been broader, including data
28   from degraded and disturbed areas and studies with shorter time frames (Tables 2.2a,b).
30   By the time of AR5, >4000 species with long-term observational data had been studied in the context of
31   climate change (Parmesan, 2006; Parmesan and Hanley, 2015). Since then, hundreds of new studies have
32   been added, leading to higher confidence in climate change attribution (Table 2.2; Scheffers et al., 2016;
33   Wiens, 2016; Cohen et al., 2018; Feeley et al., 2020). Freshwater habitats have been under-represented in
34   prior reports, but new long-term data sets, coupled with laboratory and field experiments, are improving our
35   understanding and this assessment stresses observations from lakes and streams. As numbers of studies
36   increase and data is increasingly extracted from areas with high LULCC, attribution is more difficult as

     Do Not Cite, Quote or Distribute                      2-26                                       Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1   habitat loss and fragmentation are known major drivers of changes in terrestrial and freshwater species
 2   (IPBES) (Gardner and Finlayson, 2018; Grill et al., 2019; Zarfl et al., 2019; Tickner et al., 2020).
 4   2.4.2     Observed Responses to Climate Change by Species and Communities (Freshwater and Terrestrial)
 6    Observed Range Shifts Driven by Climate Change
 8   Poleward and upward range shifts were already attributable to climate warming with high confidence in
 9   AR5. Publication of observed range shifts in accord with climate change have accelerated since AR5 and
10   strengthened attribution. Ongoing latitudinal and elevational range shifts driven by regional climate trends
11   are now well-established globally across many groups of organisms, and attributable to climate change with
12   very high confidence due to very high consistency across a now very large body of species and studies and
13   in-depth understanding of mechanisms underlying physiological and ecological responses to climate drivers
14   (Table 2.2; Table 2.3, Table SM2.1;Pöyry et al., 2009; Chen et al., 2011; Grewe et al., 2013; Gibson-
15   Reinemer and Rahel, 2015; MacLean and Beissinger, 2017; Pacifici et al., 2017; Anderegg et al., 2019b).
16   Range shifts stem from local extinctions along warm-range-boundaries (Anderegg et al., 2019b), as well as
17   from colonisation of new regions at cold-range-boundaries (Ralston et al., 2017).
19   Many studies since AR4 have tended not to be designed as attribution studies, particularly recent large scale,
20   multispecies meta-analyses. That is, all data available were included in such studies (from both undisturbed
21   and from highly degraded lands, and including very short term datasets of <20 years) with little attempt to
22   design the studies to differentiate effects of climate change from effects of other potential confounding
23   variables. These studies tended to find greater lag and lower proportion of species changing in directions
24   expected from climate change, with authors concluding that LULCC, particularly habitat loss and
25   fragmentation, was impeding wild species from effectively tracking climate change (Lenoir and Svenning,
26   2015; Rumpf et al., 2019; Lenoir et al., 2020).
28   Unprecedented outbreaks of spruce beetles occurring from Alaska to Utah in the 1990s were attributed to
29   warm weather that, in Alaska, facilitated a halving of the insect’s life cycle from two years to one (Logan et
30   al., 2003). Milder winters and warmer growing seasons were likewise implicated in poleward range
31   expansions and increasing outbreaks of several forest pests (Weed et al., 2013), leading to the current
32   prediction that 41% of major insect pest species will increase their damage further as climate warms, and
33   only 4% will reduce their impacts, while the rest will show mixed responses (Lehmann et al., 2020).
35   During their range shifts, forest pests remain climate-sensitive. For example, the distribution of Western
36   Spruce Budworm is limited at its warm range edges by adverse effects of mild winters on overwinter
37   survival, and at its cool range limits by ability to arrive at a cold-resistant stage before winter arrives
38   (Régnière and Nealis, 2019). We might therefore expect tree mortality from insect outbreaks to be most
39   severe in sites climatically less suitable for the plants, where plants would be under more stress. However,
40   (Jaime et al., 2019), using separate SDMs (MaxEnt) for the insects and plants, found that mortality of Scots
41   Pine from bark beetles was highest in sites most climatically suitable for the trees as well as for the insects.
42   In a study of tree mortality in California, bark beetles selectively killed highly-stressed fir trees but killed
43   pines according to their size, irrespective of stress status (Stephenson et al., 2019).
45   Range shifts in a poleward and upward direction, following expected trajectories given the local and regional
46   climate trends, are strongly occurring in freshwater fish populations in North America (Lynch et al., 2016b),
47   Europe (Comte and Grenouillet, 2013; Gozlan et al., 2019) and Central Asia (Gozlan et al., 2019). Cold
48   water fish, such as coregonids and smelt have been negatively affected at the equatorial borders of their
49   distributions (Jeppesen et al., 2012). Upward elevational range shifts in rivers and streams have been
50   observed. Systematic shifts towards higher elevation and upstream were found for 32 stream fish species in
51   France following regional variation in climate change (Comte and Grenouillet, 2013). Bull trout (Salvelinus
52   confluentus) in Idaho (USA), were estimated to have lost 11–20% (8–16% decade-1) of the headwater stream
53   lengths necessary for cold water spawning and early juvenile rearing, with the largest losses occurring in the
54   coldest habitats (Isaak et al., 2010). Range contractions of the same species have been found in the Rocky
55   Mountain watershed (Eby et al., 2014). Likewise, the distribution of the stonefly Zapada glacier, endemic to
56   alpine streams of Glacier National Park in Montana (USA), has been reduced over several decades by

     Do Not Cite, Quote or Distribute                      2-27                                        Total pages: 237
     FINAL DRAFT                                    Chapter 2                  IPCC WGII Sixth Assessment Report

 1   upstream retreat to higher, cooler sites as water temperatures have increased and glacial masses decreased
 2   (Giersch et al., 2015).
 4   The melting of glaciers has led to a change in water discharge associated with community turnover in
 5   glacier-fed streams (Cauvy-Fraunié and Dangles, 2019). For instance, glacier‐obligate macroinvertebrates
 6   have started disappearing when glacial cover drops below approximately 50% (robust agreement, high
 7   confidence), reviewed in (Hotaling et al., 2017). For freshwater invertebrates, no meaningful trends have
 8   been detected in geographic extent or population size for most species (Gozlan et al., 2019).
10   An invasive freshwater cyanobacterium in lakes, Cylindrospermopsis raciborskii, originating from the
11   tropics, has spread to temperate zones over the last few decades due to climate change-induced earlier
12   increase of water temperature in spring (Wiedner et al., 2007), aided by a competitive advantage in eutrophic
13   systems (Ekvall et al., 2013; Urrutia-Cordero et al., 2016).
15   Observed Local Population and Global Species' Extinctions Driven by Climate Change
17   Disappearances of local populations within a species range are more frequent and better documented than
18   whole species' extinctions, and attribution to climate change is possible for sites with minimal confounding
19   non-climatic stressors. Changes of temperature extremes are often more important to these local extinction
20   rates than changes of mean annual temperature (see Sections 2.3.1, 2.3.2,,, Cross-chapter Box
21   EXTREMES this Chapter; Parmesan et al., 2013). In a study of 538 plant and animal species, sites with local
22   extinctions were associated with smaller changes of mean annual temperature but larger and faster changes
23   of hottest yearly temperatures than sites where populations persisted (Román-Palacios and Wiens, 2020).
24   Near warm range limits, 44% of species had suffered local extinctions. In both temperate and tropical
25   regions, sites with local extinction had greater increases in maximum temperatures than those without (Tmax
26   increased 0.456°C and 0.316°C vs. Tmean increase of 0.153 °C and 0.061 °C for temperate (n=505 sites) and
27   tropical (n=76 sites), respectively, P < 0.001) (Román-Palacios and Wiens, 2020).
29   Wiens (2016) assumed that population extinctions were primarily driven by climate change when they
30   occurred at elevational or latitudinal "warm edge" range limits, and were in relatively undisturbed sites that
31   were stated by authors to be under increasing climatic stress. By this criterion, climate-caused local
32   extinctions were widespread among plants and animals, detected in 47% of 976 species examined. The
33   percentage of species suffering these extinctions was higher in the tropics (55%), than in temperate habitats
34   (39%), higher in freshwater (74%), than in marine (51%) or terrestrial (46%) habitats and higher in animals
35   (50%) than in plants (39%). The difference between plants and animals varied with latitude: in the temperate
36   zone a much higher proportion of animals than plants suffered range-limit extinctions (38.6% of 207 animal
37   species versus 8.6% of 105 plants, p < 0.0001) while at tropical sites local extinction rates were
38   (nonsignificantly) higher in plants (59% of 155 species) than in animals (52% of 349 species), the reverse of
39   their temperate zone relationship. Rates varied among animal groups, from 35% in mammals through 43% in
40   birds to 56% in insects and 59% in fish (Wiens, 2016).
42   Freshwater population extinctions are mainly due to habitat loss, introduction of alien species, pollution,
43   over-harvesting (Gozlan et al., 2019; IPBES, 2019) and climate change induced epidemic diseases (Pounds
44   et al., 2006)(see Section Climate warming particularly through intensification and severity of
45   droughts, contributes to the disappearance of small ponds, which hold rare and endemic species (Bagella et
46   al., 2016). Systematic data on the extent and biology of small ponds is, however, lacking at a global scale.
47   Extreme heat waves can lead to large local fish kills in lakes (see Section, when water temperature
48   and oxygen concentrations surpass critical thresholds, threatening cold water fish and amphibians
49   (Thompson et al., 2012). Evidence for a local extinction of some invertebrate species with a 1.4°–1.7°C rise
50   in mean annual stream winter temperature from 1981-2005 was reported in (Abrahams et al., 2013).
51   Population declines of specialist species in glacier-fed streams, such as the non-biting midge Diamesa davisi
52   (Chironomidae), can be attributed to glacier retreat given climate change (Cauvy-Fraunié and Dangles,
53   2019), and the flatworm Crenobia alpina (Planariidae) has been reported as locally extinct in the Welsh Llyn
54   Brianne river (Durance and Ormerod, 2010; Larsen et al., 2018).
56   Many high montane possums in Australia have low physiological tolerance to heatwaves, with death
57   occuring due to heat-driven dehydration at temperatures exceeding 29°–30°C for >4–5 hours over several

     Do Not Cite, Quote or Distribute                     2-28                                     Total pages: 237
     FINAL DRAFT                                    Chapter 2                  IPCC WGII Sixth Assessment Report

 1   days (Meade et al., 2018; Turner, 2020), with major declines recorded for several species, and population
 2   extinctions at lower elevations, since the early 2000s (Chandler, 2014; Weber et al., 2021).
 4   Two terrestrial and freshwater species have gone extinct, with climate change implicated as a key driver. The
 5   cloud-forest-restricted Golden toad (Incilius periglenes) was extinct by 1990 in a nature preserve in Costa
 6   Rica, driven by successive extreme droughts. This occurred in the absence of chytridiomycosis infection,
 7   caused by the fungal pathogen Batrachochytrium dendrobatidis (BD), verified during field censuses of
 8   golden toad populations in the process of extinction and through genetic analyses of museum specimens,
 9   although Bd was present in other frog species in the region (medium evidence, high agreement) (Pounds et
10   al., 1999; Pounds et al., 2006; Puschendorf et al., 2006; Richards-Hrdlicka, 2013). The interaction between
11   expansion of chytrid fungus globally and local climate change is implicated in the extinction of a wide range
12   of tropical amphibians (see Section Case study 2 Chytrid fungus and climate change).
14   The Bramble Cay Melomys (Melomys rubicola), the only mammal endemic to the Great Barrier Reef,
15   inhabited a small (five hectare) low-lying (<3m high) cay in the Torres Strait Islands, Australia. Recorded
16   having a population size of several hundred in 1978, this mammal has not been seen since 2009 and was
17   declared extinct in 2016 (Gynther et al., 2016). SLR, documented increases in storm surge and in tropical
18   cyclones, driven by climate change, led to multiple inundations of the island in the 2000s. Between 1998 and
19   2014, herbacious vegetation, the food resource for the BC Melomys, declined by 97% in area (from 2.2 ha
20   down to 0.065 ha), and from 11 plant species down to two (Gynther et al., 2016; Watson, 2016; Woinarski,
21   2016; Woinarski et al., 2017). The island was unihabited with few non-climatic threats, providing high
22   confidence in attribution of extinction of the BC Melomys to climate change-driven increases in frequency
23   and duration of island inundation (Turner and Batianoff, 2007; Woinarski et al., 2014; Gynther et al., 2016;
24   Watson, 2016; Woinarski et al., 2017).
26   In the IUCN Red List (IUCN, 2019), 16.2% of terrestrial and freshwater species (n=3,777 species) that are
27   listed as endangered, critically endangered or extinct in the wild (n=23,251 species) list climate change or
28   severe weather as one of their threats.
31   [START FAQ2.1 HERE]
33   FAQ2.1: Will species go extinct with climate change and is there anything we can do to prevent it?
35   Climate change is already posing major threats to biodiversity and the most vulnerable plants and animals
36   are likely to go extinct. If climate change continues to worsen, it is expected to cause many more species to
37   go extinct unless we take actions to improve the resilience of natural areas, through protection, connection
38   and restoration. We can also help individual species that we care most about by reducing the stress they are
39   under from other human activities, and even helping them move to new places as their climate space shifts
40   and they need to shift to keep up.
42   Climate change has already caused some species to go extinct, and is likely to drive more species to
43   extinction. Species have always gone extinct in the history of our planet but human activities causing climate
44   change are accelerating this process. For instance, recent research predicts that one-third of all plant and
45   animal species could be extinct by 2070 if climate change continues as it is. Species can adapt to some extent
46   to these rapidly changing climate patterns. We are seeing changes in behaviour, dispersal to new areas as the
47   climate becomes more suitable, and genetic evolution. However, these changes are small, and adaptations
48   are limited. Species that cannot adapt beyond their basic climate tolerances (ability to survive extremes of
49   temperature or rainfall) or successfully reproduce in a different climate environment from what they have
50   evolved in, will simply disappear. In the Arctic for example, sea ice is melting and will likely disappear in
51   summer time within a century. This means that the animals that have evolved to live on sea ice - polar bears
52   and some seals and sea lions - will go extinct.
54   Fortunately, there are some things we can do to help. We can take actions to assist, protect and conserve
55   natural ecosystems and prevent the loss of our planet’s endangered wildlife, such as:

     Do Not Cite, Quote or Distribute                     2-29                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   “Assisting” species’ migration: This has many names, “assisted colonisation”, "assisted translocation",
 2   "assisted migration", "assisted movement". In effect, it is about helping endangered species to move to a new
 3   area with a good habitat for them to survive. "Passive" assisted colonisation focuses on helping species move
 4   themselves, whilst the most “active” form implies picking up individuals and transporting them to a new
 5   location. This is different from re-introductions that are already a normal part of conservation programs.
 6   Climate-driven translocations are moving plants or animals to an area where they have never lived
 7   historically, a new location that is now suitable for them due to climate change.
 9   This active form of “assisted colonisation” has been controversial, because exotic species can become
10   invasive when they are moved between continents or oceans. For example, no one would advocate moving
11   polar bears to Antarctica, as they would likely fest on native penguins, thus causing another conservation
12   problem. However, moving species only a few hundred kilometers avoids most adverse outcomes, and that
13   is often all that is needed to help a wild plant or animal cope with climate change. In extreme cases, another
14   type of assisted adaptation is to preserve species until we get climate change under control and can
15   reintroduce them to the wild. This might include moving them into zoos or into seed or frozen embryo
16   banks.
18   Extending protected zones and their connectivity: Species' ability to move to new locations and track climate
19   change are very limited – in particular, when a habitat has been turned into a crop field or a city. To help
20   them move between their natural habitats, we can increase the connectedness of protected areas or simply
21   create small patches or corridors of semi-wild nature within a largely agricultural or inhabited region, that
22   encourages wildlife to move through an area, and in which they are protected from hunting and poisons.
23   Those semi-wild protected areas can be very small, like the hedgerows between fields in England, that
24   provide both habitats for many flowers, birds, insects and corridors to move between larger protected areas.
25   Alternatively, it can just be an abandoned field that is now growing "weeds" without pesticides, hunting or
26   farming. For instance, in the United States of America, private landowners get a tax break by making their
27   land a "wildlife conservation" area using no pesticide, not cutting weeds too often, putting up brush piles and
28   bird boxes for nesting, and providing a water source.
30   Assisting, protecting and conserving natural ecosystems would help enhance biodiversity overall as well as
31   already endangered species. Diverse plant and animal communities are more resilient to disturbances,
32   including climate change. A healthy ecosystem also recovers more quickly from extreme events, such as
33   floods, droughts and heat waves that are a part of human-driven climate change. Healthy ecosystems are
34   critical to prevent species’ extinctions from climate change, but are also important for human health and
35   well-being, providing clean, plentiful water, cleaning the air, providing recreation and holiday adventures,
36   and making people feel happier, calmer and more contented.


     Do Not Cite, Quote or Distribute                     2-30                                       Total pages: 237
     FINAL DRAFT                                        Chapter 2                    IPCC WGII Sixth Assessment Report

 1   Figure FAQ2.1.1: Possible actions to assist protect and conserve natural ecosystems and prevent the loss of our
 2   planet’s endangered wildlife in the face of continued climate change. (Inspired by Natural Alliance website © Chris
 3   Heward/GWCT)
 5   [END FAQ2.1 HERE]
 8   Observed Changes in Community Composition Driven by Climate Change
10 Overall patterns of community change
11   The most common type of community change takes the form of in situ decreases of cold-adapted species and
12   increases of warm-adapted species (Bowler et al., 2017; Hughes et al., 2018; Kuhn and Gégout, 2019; Feeley
13   et al., 2020). This process has lead to increases of species richness on mountaintops and decreased richness
14   at adjacent lower elevations (medium evidence, high agreement) (Forister et al., 2010; Steinbauer et al.,
15   2018). Observed shifts in community composition have consequences for species' interactions. Such indirect
16   effects of climate change have been shown to often have greater impacts on species than direct effects of
17   climate itself, particularly for higher level consumers (Ockendon et al., 2014).
19   Like other responses, analyses indicated responses were lagging behind the change expected from regional
20   warming, and thereby accumulating 'climate debt.' Examples of climate debt, measured from community
21   composition changes, come from birds and butterflies in Europe (Devictor et al., 2012) and from lowland
22   forest herbaceous plants in France (Bertrand et al., 2011). The French study found that larger debts occurred
23   in communities with warmer baseline conditions and that some of the apparent debt stemmed from species'
24   ability to tolerate warming in situ. Geothermal streams have provided evidence about community structure
25   and ecosystem function in high temperatures. A study of 14 such habitats reported simplified the food-web
26   structures and shortened pathways of energy flux between consumers and resources (high confidence)
27   (O’Gorman et al., 2019).
29   Prominent changes in freshwater community composition, such as increases in cyanobacteria and warm
30   tolerant zooplankton species, loss of cold water fish, gain in thermo-tolerant fish and macroinvertebrates and
31   gain in floating macrophytes, are occurring (medium evidence, high agreement, medium confidence) (Adrian
32   et al., 2016; Hossain et al., 2016; Short et al., 2016; Huisman et al., 2018; Gozlan et al., 2019). Changes in
33   relative species abundances, species composition and biodiversity due to warming trends and non-climate
34   driven changes are to be expected in lakes and rivers globally. However, thus far empirical evidence and
35   mechanistic understanding to inform modelling is too limited to draw general conclusions about the nature of
36   current and future climate change driven changes within entire food webs on a global scale (Urban et al.,
37   2016).
39 Freshwater mechanistic drivers and responses
40   Physical changes in lakes (see Section 2.3.3) have affected primary production (see Section, algal
41   bloom formation and composition, zooplankton and fish size distribution and species composition (Urrutia-
42   Cordero et al., 2017; Gozlan et al., 2019; Seltmann et al., 2019). Declines in abundance of cold-stenothermal
43   species (particularly Arctic charr, Salvelinus alpinus, coregonids and smelt) and increases in eurythermal
44   fish (e.g. the thermo-tolerant carp Cyprinus carpio, common bream, pike perch, roach and shad) have been
45   observed in northern temperate lakes associated with warming trends (high agreement, medium confidence)
46   (Jeppesen et al., 2012; Jeppesen et al., 2014). These changes increase predation pressure on zooplankton and
47   reduce grazing pressure on phytoplankton, which may result in higher phytoplankton biomass (De Senerpont
48   Domis et al., 2013; Jeppesen et al., 2014; Adrian et al., 2016). Reduction in lake mixing lowers the
49   concentration of nutrients in the epilimnion and may lead to higher silicon to phosphorous ratios negatively
50   affecting diatom growth (Yankova et al., 2017) or overall primary productivity (see Section

51   In as study of 1,567 lakes across Europe and North America, (Kakouei, 2021) identified climate change as
52   the major driver of increases in phytoplankton biomass in remote areas with minimal LULCC. Greater
53   temperature variability can be more important than long-term temperature trends as a driver of zooplankton
54   biodiversity (Shurin et al. (2010). Reductions of winter severity attributed to anthropogenic climate change
55   are increasing winter algal biomass, and motile and phototropic species at the expense of mixotrophic
56   species (Özkundakci et al., 2016; Hampton et al., 2017).

     Do Not Cite, Quote or Distribute                         2-31                                         Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 2   Tropical lakes are prone to loss of deep-water oxygen due to lake warming with negative consequences for
 3   their fisheries and their biodiversity (Lewis Jr, 2000; Van Bocxlaer et al., 2012). Many ancient tropical lakes
 4   (Malawi, Tanganyika, Victoria, Titicaca, Towuti and Matano) hold thousands of endemic animal species
 5   (Vadeboncoeur et al., 2011).
 7   Observed climate-change effects on freshwater invertebrates are variable (Knouft and Ficklin, 2017). In
 8   glacier-fed streams globally, climate change has caused community turnover and changes in abundances in
 9   terms of increased generalist and decreased specialist species abundances (Lencioni, 2018; Cauvy-Fraunié
10   and Dangles, 2019). In turn, dragonflies in flowing waters, monitored during the warming period from 1988
11   through 2006 in Europe, did not show consistent changes in their distribution (Grewe et al., 2013), reviewed
12   in (Knouft and Ficklin, 2017). Long-term trends in species composition and community structure of stream
13   macroinvertebrates, specifically a general trend for decreases in species characteristic of cold, fast-flowing
14   waters and increases of thermophilic species typical of stagnant or slow-moving waters, have been attributed
15   to climate change (high agreement, high confidence) (Daufresne et al., 2007; Chessman, 2015). A study of
16   14 geothermal streams reported simplified food-web structures and shortened pathways of energy flux
17   between consumers and resources (O’Gorman et al., 2019). Macrophytes benefit from rising water
18   temperatures, but increased shading from increased phytoplankton biomass could offset this (see
19   for projections; Hossain et al., 2016; Short et al., 2016;Zhang et al., 2017a).
22 Emergence of novel communities and invasive species
23   As climate change is increasing the movements of species into new areas, there is concern about how exotic
24   species are being impacted, either by becoming invasive or by already invasive species gaining even more
25   advantage over native species. Modeling predicts that effects of climate warming on food web structure and
26   stability favour success of invading species (Sentis et al., 2021). Both simulated warming experiments
27   (Zettlemoyer et al., 2019) and long-term observations (Losos et al., 2010) have found phenologies of exotic
28   species to respond more adaptively to warming than those of natives, and in the long-term observations the
29   success of exotics was attributed to their greater phenological responsiveness. In an expert assessment of the
30   future relative importance of different drivers of the impacts of biological invasions, climate change was
31   named as the most important driver in polar regions, the second most important in temperate regions (after
32   trade/transport) and the third most important in the tropics (after trade/transport and human
33   demography/migration) (Essl et al., 2020).
34   However, not all exotic species become invasive. As novel climate conditions develop, novel communities
35   made up of new combinations of species are emerging as populations and species adapt and shift ranges
36   differentially, not always with negative consequences (high confidence) (Dornelas et al., 2014; Evers et al.,
37   2018; Teixeira and Fernandes, 2020). Novel communities differ in composition, structure, function and
38   evolutionary trajectories, as the proportion of specialists and generalists, native, introduced and range
39   shifting species changes and species interactions are altered, ultimately affecting ecosystem dynamics and
40   functioning (Lurgi et al., 2012; Hobbs et al., 2014; Heger and van Andel, 2019 Towards an Integrative). The
41   exact nature of novel communities is difficult to predict because species-level uncertainties propagate at the
42   community level due to ecological interactions (Williams and Jackson, 2007), but observations, experimental
43   mesocosms (Bastazini et al., 2021); and theoretical models (Lurgi et al., 2012; Sentis et al., 2021) provide
44   support that they will continue to emerge with climate change.
46   Observed Phenological Responses to Climate Change
48   With advances in remote sensing, quality and quantity of phenological data are rapidly increasing (Piao et
49   al., 2019). Since AR5, numbers of studies have increased substantially with consistent conclusions in
50   response to warming, including advancement of spring events and lengthening of growing seasons in
51   temperate regions (through a combination of advancement of spring events and to a lesser extent, retardation
52   of autumn events) (robust evidence, high agreement) (Table 2.2, Table 2.3, Table SM2.1; Menzel et al.,
53   2020). In the tropics, by contrast, precipitation changes have more strongly influenced phenology than
54   temperature changes (Cohen et al., 2018). A meta-analysis comparing observed phenological advances in
55   birds with expectations from warming of local climates concluded that the advances fell short of expectation
56   and that substantial phenological climate debt had been generated (Radchuk et al., 2019).

     Do Not Cite, Quote or Distribute                     2-32                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   Taxonomic groups have differed in their responses (Parmesan, 2007; Thackeray et al., 2010), and a few have
 2   completely abstained from the general trends—for example, seabirds continue to breed with their pre-
 3   climate-change phenologies (Keogan et al., 2018). Newer reviews and analyses reveal differences in
 4   responses among continents and across time intervals (Piao et al., 2019). Mean advance in days per decade
 5   was 5.5 in China, and 3.0-4.2 in Europe but only 0.9 in North America (Piao et al., 2019). Mean values for
 6   retardation of autumn leaf fall, which can be more influenced by photoperiod and less by temperature than
 7   spring leaf-out, were 0.36 days per decade in Europe (Menzel et al., 2020), 2.6 days per decade in China and
 8   around 3 days per decade in the USA (medium evidence, high agreement) (Piao et al., 2019).
10   The rapid rates of advance of spring events in the 1990s slowed down in the 2000s and stalled or even
11   reversed in some regions (Menzel et al., 2020). Wang et al. (2019) noted, from remote sensing, that during
12   the 'global warming hiatus' from 1998-2012, there were no global trends in either Spring green-up or
13   autumn colouring. Annual crops, for which timing is determined by farmers, were an exception. When
14   natural systems were advancing fast prior to 1998, farmers advanced more slowly, but during the natural
15   'hiatus', farmed crops advanced faster than wild plants and cultivated trees (Menzel et al., 2020). In a long
16   (67 year) European time series (Menzel et al., 2020), autumn leaf colouring showed delays attributed to
17   winter & spring warming in 57% of observations (mean delay 0.36 days per decade); spring & summer
18   phenologies advanced in 89% of wild plants despite decreased winter chilling, with c.60% of trends
19   significant and 'strongly attributable' to winter & spring warming; and growing season length increased in
20   84% of cases (mean lengthening 0.26 days yr-1) (Table 2.2).
22   Changes in freshwater systems are consistent with changes in terrestrial systems: earlier timing of spring
23   phytoplankton and zooplankton development and earlier spawning by fish, as well as extension of the
24   growing season are occurring (robust evidence, high agreement) (Adrian et al., 2009; De Senerpont Domis et
25   al., 2013; Adrian et al., 2016; Thackeray et al., 2016). Phenological changes in lakes have been related to
26   rising water temperatures, reductions of ice cover and prolongation of thermal stratification (increasing
27   evidence and agreement since AR5; very high confidence). Crozier and Hutchings (2014) reviewed
28   phenological changes in fish and documented that changes in the timing of migration and reproduction, age
29   at maturity, age at juvenile migration, growth, survival and fecundity were associated primarily with changes
30   in temperature. The median return time of Atlantic salmon among rivers in Newfoundland and Labrador
31   advanced by 12 to 21 days over the past decades, associated with overall warmer conditions (Dempson et al.,
32   2017).
34   Observed Complex Phenological and Range Shift Responses
36   Early meta-analyses tested the straightforward hypotheses that warming should shift timing earlier and
37   ranges poleward. Once these trends had been established, exceptions to them became foci of study. For
38   example, some plants in northern regions of the northern hemisphere were retarding their spring flowering
39   instead of advancing it as expected with warming. These turned out to be species requiring vernalisation
40   (winter chilling) to speed spring development. For these plants, phenological changes result from combined
41   effects of advancement caused by spring warming and retardation caused by winter warming. Incorporating
42   this level of complexity into analyses revealed that a greater proportion of species were responding to
43   climate change than estimated under the simple expectation that warming should always cause advancement
44   (92% responding vs 72% from earlier analyses) (Cook et al., 2012).
46   Animal species can show vernalisation equivalent to that in plants (Stålhandske et al., 2017). However, a
47   semi-global meta-analysis across terrestrial animals failed to detect delaying effects of warming winters
48   (Cohen et al., 2018). The same animal-based meta-analysis contrasted phenological changes in temperate-
49   zone animals, which are principally explained by changes of temperature, with those at lower latitudes,
50   which follow changes of precipitation (Cohen et al., 2018).
52   Vitasse et al. (2018), working with Alpine trees, found that phenological delay with increasing elevation had
53   declined from 34 days per 1,000 m in the 1960s to 22 days per 1,000 m, greatly reducing the differences in
54   timing between trees growing at different elevations. This reduction was greatest after warmer winters,
55   suggesting winter warming as a principal cause of the overall trend.

     Do Not Cite, Quote or Distribute                     2-33                                       Total pages: 237
     FINAL DRAFT                                         Chapter 2                     IPCC WGII Sixth Assessment Report

 1   Lian et al. (2020) observed that earlier spring leaf-out in the Northern Hemisphere is causing increases in
 2   evapotranspiration that are not fully compensated by increased precipitation. The consequence is increased
 3   soil moisture deficit in summer, expected to exacerbate impacts of heatwaves as well as drought stress. In
 4   Arctic freshwater ecosystems, Heim et al. (2015) demonstrated the importance of seasonal cues for fish
 5   migration, which can be impacted by climate change due to reduced stream connectivity and fragmentation,
 6   earlier peak flows, and increased evapotranspiration.
 8   Precipitation has also been implicated in exceptions to the rule that ranges should be shifting to higher
 9   elevations. In dry climates, increases of precipitation accompanying climate warming can facilitate
10   downslope range shifts (Tingley et al., 2012).
12   Multiple responses can co-occur. Hällfors et al (Hällfors et al., 2021), in a study of 289 lepidoptera in
13   Finland, found 45% had either shifted their ranges northward or advanced their flight season with warming.
14   The 15% of species that did both (shifting northward by 113.1 km and advancing flight period by 2.7 days
15   per decade, on average, over a 20 year period) had the largest population increases, and the 40% of species
16   that showed no response had the largest population declines.
19   Table 2.2: Global Fingerprints of Climate Change Impacts across Wild Species. Updated from (Parmesan and Hanley,
20   2015). For each dataset, a response for an individual species or functional group was classified as (1) no response (no
21   significant change in the measured trait over time), (2) if a significant change was found, the response was classified as
22   either consistent or not consistent with expectations from local or regional climate trends. Percentages are approximate
23   and estimated for the studies as a whole. Individual analyses within the studies may differ. The specific metrics of
24   climate change analysed for associations with biological change vary somewhat across studies, but most use changes in
25   local or regional temperatures (e.g. mean monthly T or mean annual T), with some using precipitation metrics (e.g. total
26   annual rainfall). For example, a consistent response would be poleward range shifts in areas that are warming.
27   Probability (P) of getting the observed ratio of consistent: not consistent responses by chance was <10-13 for (Parmesan
28   and Yohe, 2003; Root et al., 2003; Root et al., 2005; Poloczanska et al., 2013) and was <0.001 for Rosenzweig 2008
29   (source=publication) (Parmesan and Yohe, 2003; Root et al., 2003; Root et al., 2005; Rosenzweig et al., 2008;
30   Poloczanska et al., 2013). Test were all binomial tests against p=0.5, performed by Parmesan.
34   Observed Changes to Physiology and Morphology Driven by Climate Change
36   Impacts on species physiology in terrestrial and freshwater systems have been observed and attributed to
37   climate change (medium confidence), including changes in tolerances to high temperatures (Healy and
38   Schulte, 2012; Gunderson and Stillman, 2015; Deery et al., 2021), increased metabolic costs of living under
39   elevated temperatures (Scheffers et al., 2016) and shifts in sex ratios in species with temperature-dependent
40   sex determination (e.g. masculinisation of lizard populations (Schwanz and Janzen, 2008; Schwanz, 2016;
41   Edmands, 2021) and feminisation of turtle populations (Telemeco et al., 2009)). Skewed sex ratios can lead
42   to mate shortages, reduced population growth and adaptive potential, and increased extinction risk, because
43   genetic diversity decreases as fewer individuals mate and heterozygosity is lost (Mitchell and Janzen, 2010;
44   Edmands, 2021).
46   Behavioural plasticity such as nest-site selection can provide a partial buffer from the effects of increasing
47   temperature, but there are environmental and physical limits to this plasticity (medium confidence)
48   (Refsnider and Janzen, 2016; Telemeco et al., 2017). Plasticity in heat tolerance (e.g. due to reversible
49   acclimation or acclimatisation) can also potentially compensate for rising temperatures (Angilletta Jr, 2009),
50   but ectotherms have relatively low acclimation in thermal tolerance and acclimation is expected to only
51   slightly reduce overheating risk in even the most plastic taxa (low confidence) (Gunderson and Stillman,
52   2015).
54   Geographic variation in thermal tolerance plasticity is expected to influence species vulnerability and range
55   shifts in response to climate change (Gunderson and Stillman, 2015; Sun et al., 2021). In many ectotherms,
56   plasticity in thermal tolerance increases towards the poles, as thermal seasonality increases (Chown et al.,
57   2004), contributing to higher vulnerability to warming in tropical organisms (low confidence) (Huey et al.,
58   2009; Campos et al., 2021). Some species have evolved extreme upper thermal limits at the expense of

     Do Not Cite, Quote or Distribute                          2-34                                           Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   plasticity, reflecting an evolutionary trade-off between these traits (Angilletta et al., 2003; Stillman, 2003).
 2   The most heat-tolerant species, such as those from extreme environments, may therefore be at greater risk of
 3   warming because of an inability to physiologically adjust to thermal change (low confidence) (Bozinovic et
 4   al., 2011; Overgaard et al., 2014; Magozzi and Calosi, 2015).
 6   Physiological changes have observable impacts on morphology, such as changes to body size (and length of
 7   appendages) and colour changes in butterflies, dragonflies, and birds (medium confidence) (Galeotti et al.,
 8   2009; Karell et al., 2011), but trends are not always linear or consistent across realms, taxonomic groups or
 9   geographic regions (Gotanda et al., 2015). Some morphological changes arise in response to environmental
10   changes rather than as the result of genetic adaptation or selection for an optimum body type. For example,
11   dietary changes associated with climate change have led to changes in chipmunk skull morphology (Walsh et
12   al., 2016).
14   Decreased body size has been suggested as a general response of species to climate change in freshwater
15   species given the temperature related constraints of metabolism with increasing body size. Reduced body
16   size in response to global warming has been documented for freshwater bacteria, plankton and fish, as well
17   as a shift towards smaller species (low confidence) (Daufresne et al., 2009; Winder et al., 2009; Jeppesen et
18   al., 2010; Crozier and Hutchings, 2014; Jeppesen et al., 2014; Farmer et al., 2015; Rasconi et al., 2015;
19   Woodward et al., 2016). However, the lack of systematic empirical evidence in freshwaters and confounding
20   effects such as interactions between temperature, nutrient availability and predation limit generalisations
21   about body size effects (Pomati et al., 2020 Nutrients).
23   Evidence is weak for a consistent reduction in body size across taxonomic groups in terrestrial animals (low
24   confidence) (Siepielski et al., 2019). Decreased body size in warmer climates (as higher surface area to
25   volume ratios maximise heat loss) is expected based on biogeographic patterns such as Bergmann’s Rule,
26   but both increases and decreases have been documented in mammals, birds, lizards and invertebrates and
27   attributed to climate change (Teplitsky and Millien, 2014; Gotanda et al., 2015; Gardner et al., 2019; Hill et
28   al., 2021). Contrasting patterns (increased body size) may be due to short-term modifications in selection
29   pressures (e.g. changes to predation and competition), variation in life histories or a result of interactions
30   with climate variables other than temperature (e.g. changes to food availability with rainfall changes) and
31   other disturbances (Yom-Tov and Yom-Tov, 2004; Gardner et al., 2019; Wilson et al., 2019) or body size
32   measurements (linear vs. volumetric dimensions) (Salewski et al., 2014).
34   Several lines of evidence suggest evolution of melanism in response to climate change (low confidence),
35   with colour changes associated with thermoregulation being demonstrated in butterflies (Zeuss et al., 2014;
36   MacLean et al., 2016; MacLean et al., 2019a), beetles (de Jong and Brakefield, 1998; Brakefield and de
37   Jong, 2011; Zvereva et al., 2019), dragonflies (Zeuss et al., 2014) and phasmids (Nosil et al., 2018). Such
38   changes may represent decreased phenotypic diversity and, potentially, genetic diversity (low confidence),
39   but the consequences of climate change on the genetic structure and diversity of populations have not been
40   widely assessed (Pauls et al., 2013). Simplistically, the thermal melanism hypothesis suggests that lighter
41   (higher reflectance) individuals should have increased fitness and therefore be selected for in a warmer
42   climate (Clusella-Trullas et al., 2007). However, several biotic (e.g thermoregulatory requirements, predator
43   avoidance, signalling) and abiotic (e.g. UV, moisture, interannual variability) factors interact to influence
44   changes in colour, making attribution to climate change across species and broad geographic regions difficult
45   (Kingsolver and Buckley, 2015; Stuart-Fox et al., 2017; Clusella-Trullas and Nielsen, 2020).
47   Interactions between morphological changes and changes to phenology may facilitate or constrain adaptation
48   to climate change (medium confidence) (Hedrick et al., 2021). For example, advancing phenology in
49   migratory species may impose selection on morphological traits (e.g. wing length) to increase migration
50   speed. If advancing spring phenology results in earlier breeding, this may offset the effect of rising
51   temperatures in the breeding range and reduce the effect of increasing temperature on body size (Zimova et
52   al., 2021). A study of 52 species of North American migratory birds, based on more than 70,000 specimens,
53   showed that spring migration phenology has advanced over the past 40 years, concurrent with widespread
54   shifts in morphology (reduced body size and increased wing length), perhaps to compensate for the increased
55   metabolic cost of flight as body size decreases (Weeks et al., 2020).

     Do Not Cite, Quote or Distribute                      2-35                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   A lack of understanding of physiological constraints and mechanisms remains a barrier to predicting many of
 2   the ecological effects of climate change (Bozinovic et al., 2011; Vázquez et al., 2017; González-Tokman et
 3   al., 2020). Many behavioural, morphological and physiological responses are highly species and context
 4   specific, making generalisations difficult (Bodensteiner et al., 2021). Recent advances in mechanistic
 5   understanding (from experiments), in process-based modeling that includes microclimates and
 6   developmental processes (Carter and Janzen, 2021) and in sophistication of niche models (Kearney et al.,
 7   2009) have improved projections, but comprehensive tests of geographic patterns and processes in thermal
 8   tolerance and plasticity are still lacking, with studies limited to a few phylogenetically restricted analyses
 9   showing mixed results (Gunderson and Stillman, 2015). Improved understanding of the mechanistic basis for
10   observed geographic patterns in thermal tolerance and plasticity is needed to identify species’ physiological
11   limits, the potential for adaptation and the presence of evolutionary trade-offs, which will strongly influence
12   population declines, species range shifts, invasive interactions and success of conservation interventions
13   (Cooke et al., 2021; Ryan and Gunderson, 2021).
15     Observed Impacts of Climate Change on Diseases of Wildlife and Associated Impacts on Humans
17   Assessment of changes in diseases of terrestrial and freshwater wild organisms was scarce in WGII AR4,
18   AR5, IPCC SR1.5 and IPCC SRCCL. Most emerging infectious diseases (EIDs) are zoonoses, that is,
19   transmissible between humans and animals, and climate sensitive (Woolhouse et al., 2001; Woolhouse and
20   Gowtage-Sequeria, 2005; McIntyre et al., 2017; Salyer et al., 2017). WGII AR4 found weak to moderate
21   evidence that disease vectors and their diseases had changed their distributions in concert with climate
22   change, but attribution studies were lacking. In WGII AR5 Chapter 11, geographic expansion of a few vector
23   borne diseases (VBDs) to higher latitudes and elevations had been detected and associated with regional
24   climate trends, but non-climatic drivers were not assessed well, leading to only a medium confidence in
25   attribution) (IPCC, 2014). Here we build upon previous assessments by focusing on changes in population
26   dynamics and geographic distributions of diseases of wildlife, and those of humans and domestic animals
27   that are also harbored, amplified, and transmitted by wild animal reservoir hosts and vectors.
29   Increased disease incidence is correlated with regional climatic changes and is expected from underlying
30   biology of relationships between temperature, precipitation, and disease ecology (robust evidence, high
31   agreement) (Norwegian Polar Institute, 2009; Tersago et al., 2009; Tabachnick, 2010; Paz, 2015; Dewage et
32   al., 2019; Deksne et al., 2020; Shocket et al., 2020; Couper et al., 2021). Whether increases in diseases in
33   wild and domestic animals correspond to increased disease risk in nearby human populations is complicated
34   by potential buffering effects of the local medical system, healthcare access, socio-economic status,
35   education, behaviours and general health of the human population (see also Chapter 7 and Cross-Chapter
36   Box ILLNESS this Chapter).
38    Direct effects of climate on reproduction, seasonality, growing season length and transmission of
39                pathogens, vectors, and hosts
40   VBDs require arthropod vector hosts (e.g., insects or ticks), while other infectious diseases (e.g., fungi,
41   bacteria, and helminths) have free-living life stages and/or complex life cycles that require intermediate hosts
42   (e.g., snails), all of which have temperature-driven rates of development and replication/reproduction (robust
43   evidence, high agreement) (Mordecai et al., 2013; Liu-Helmersson et al., 2014; Moran and Alexander, 2014;
44   Bernstein, 2015; Marcogliese, 2016; Ogden and Lindsay, 2016; Mordecai et al., 2017; Short et al., 2017;
45   Caminade et al., 2019; Cavicchioli et al., 2019; Mordecai et al., 2019; Liu et al., 2020; Rocklöv and Dubrow,
46   2020). Additionally, microbes such as bacteria thermally adapt to temperature changes through multiple
47   mechanisms, indicating that warming will not reduce antibiotic resistance (MacFadden et al., 2018; Pärnänen
48   et al., 2019; Shukla, 2019; McGough et al., 2020; Rodriguez-Verdugo et al., 2020).
50   There is increasing evidence for a role of extreme events in disease outbreaks (Tjaden et al., 2018; Bryson et
51   al., 2020). Heat waves have been associated with outbreaks of helminth pathogens, especially in subarctic
52   and Arctic areas. For example, a severe outbreak of microfilaremia, a vector-borne disease spread by
53   mosquitoes and flies, plagued reindeer in northern Europe following extreme high temperatures (Laaksonen
54   et al., 2010). More frequent and severe extreme events such as floods, droughts, heat waves, and storms can
55   either increase or decrease outbreaks, depending upon the region and disease (robust evidence, high
56   agreement) (Anyamba et al., 2001; Marcheggiani et al., 2010; Brown and Murray, 2013; Paz, 2015; Boyce et
57   al., 2016; Wu et al., 2016b; Wilcox et al., 2019; Nosrat et al., 2021). Heavy precipitation events have been

     Do Not Cite, Quote or Distribute                      2-36                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   shown to increase some infectious diseases with aquatic life cycle components such as mosquito-borne,
 2   helminth, and rodent-borne diseases (robust evidence, high agreement) (Anyamba et al., 2001; Zhou et al.,
 3   2005; Wu et al., 2008; Brown and Murray, 2013; Anyamba et al., 2014; Boyce et al., 2016). Conversely,
 4   flooding also increases flow rate and decreases parasite load and diversity in other aquatic wildlife (Hallett
 5   and Bartholomew, 2008; Bjork and Bartholomew, 2009; Marcogliese, 2016; Marcogliese et al., 2016) and
 6   can reduce mosquito abundance by flushing them out of the system (Paaijmans et al., 2007; Paz, 2015).
 8   Droughts reduce aquatic habitat of some mosquito species while simultaneously increasing the availability of
 9   stagnant standing pools of water that are ideal breeding habitats for other species, such as dengue vector
10   Aedes mosquitos (medium evidence, medium agreement) (Chareonviriyaphap et al., 2003; Chretien et al.,
11   2007; Padmanabha et al., 2010; Trewin et al., 2013; Paz, 2015). Extreme drought has been associated with
12   an increase in bluetongue virus haemorrhagic disease in wildlife in eastern North America, though
13   mechanisms were not identified (Christensen et al., 2020). Heatwaves in some regions, especially coastal
14   regions, increased parasitism and decreased host richness and abundance leading to population crashes
15   (Larsen and Mouritsen, 2014; Mouritsen et al., 2018). Changes in temperature and precipitation, especially
16   extreme events, can alter community structure (Larsen et al., 2011) by increasing or decreasing parasites and
17   their host organisms and even altering host behavior in ways advantageous to parasites (Macnab and Barber,
18   2012).
20   Climate change not only affects the occurrence of pathogens and their hosts in geographic space but also the
21   temporal patterns of disease transmission. Warmer winters allow greater overwinter survival of arthropod
22   vectors which, coupled with lengthened tranmission seasons, drive increases in vector population sizes,
23   pathogen prevalence, and hence proportion of vectors infected (robust evidence, high agreement)
24   (Laaksonen et al., 2009; Molnár et al., 2013; Waits et al., 2018). For example, a parasitic nematode lung
25   worm (Umingmakstrongylus pallikuukensis) has shortened its larval development time in half (from two
26   years to one year), which has increased infection rates in North American muskoxen (Norwegian Polar
27   Institute, 2009).
29   Case study 1: Climate change impacts on pathogenic helminths in Europe
31   Parasitic helminth worms can reduce growth and yield, or kill livestock, and infect humans and wildlife,
32   leading to health, agricultural and economic losses (Fairweather, 2011; Charlier et al., 2016; Charlier, 2020).
33   Attribution of increased helminth disease incidence and risk to climate change is stronger than for most
34   human diseases because of long-term records and careful analysis of other anthropogenic drivers (e.g. land
35   use change, agricultural/livestock intensification, and antihelminthic intervention and resistance) (van Dijk et
36   al., 2008; van Dijk et al., 2010; Fox et al., 2011b; Martínez-Valladares et al., 2013; Charlier et al., 2016;
37   Innocent et al., 2017; Mehmood et al., 2017).
39   In Europe, evidence from laboratory studies, long term surveillance, statistical analyses, and modelling
40   shows that multiple helminth pathogens and their host snails have extended their transmission windows and
41   have increased survival, fecundity, growth and abundances (robust evidence, high agreement). Furthermore,
42   they have expanded or shifted their ranges poleward due to increases in temperature, precipitation and
43   humidity (robust evidence, high agreement ) (Lee et al., 1995; Pritchard et al., 2005; Poulin, 2006; van Dijk
44   et al., 2008; van Dijk et al., 2010; Fairweather, 2011; Fox et al., 2011b; Martínez-Valladares et al., 2013;
45   Bosco et al., 2015; Caminade et al., 2015; Caminade et al., 2019). These documented changes in climate,
46   hosts and pathogens have been linked to higher disease incidence and more frequent outbreaks in livestock
47   across Europe (high confidence) (Bosco et al., 2015).
49   Case study 2: Chytrid fungus and climate change
51   Infection by the chytrid fungus, Batrachochytrium dendrobatidis (Bd), can cause chytridiomycosis in
52   amphibians. Bd is widely distributed globally and has caused catastrophic disease in amphibians, associated
53   with declines of 501 species and extinctions of a further 90 species, primarily in tropical regions of the
54   Americas and Australia (Scheele et al., 2019; Fisher and Garner, 2020). Bd successfully travelled with high-
55   elevation Andean frog species as they expanded their elevational ranges upward, driven by regional
56   warming, to > 5200 m (Seimon et al., 2017).

     Do Not Cite, Quote or Distribute                      2-37                                       Total pages: 237
     FINAL DRAFT                                    Chapter 2                   IPCC WGII Sixth Assessment Report

 1   New findings since AR5 from controlled laboratory experiments (manipulating temperature, humidity and
 2   water availability), intensive analyses of observed patterns of infection and disease in nature, and modeling
 3   studies have led to an emerging consensus that interactions between chytrids and amphibians are climate-
 4   sensitive, and that the interaction of climate change and Bd has driven many of the observed global
 5   amphibian declines and species' extinctions (robust evidence, high agreement) (Rohr and Raffel, 2010;
 6   Puschendorf et al., 2011; Rowley and Alford, 2013; Raffel et al., 2015; Sauer et al., 2018; Cohen et al.,
 7   2019a; Sauer et al., 2020; Turner et al., 2021).
 9   The "thermal mismatch hypothesis" posits that vulnerability to disease should be higher at warm
10   temperatures in cool-adapted species and higher at cool temperatures in warm-adapted species and is
11   generally supported. However, the most recent studies reveal more complex mechanisms underlying
12   amphibian-disease-climate change dynamics, including variation in thermal preferences among individuals
13   in a single amphibian population (robust evidence, high agreement) (Zumbado-Ulate et al., 2014; Sauer et
14   al., 2018; Cohen et al., 2019b; Neely et al., 2020; Sauer et al., 2020).
16   Bd is not universally harmful—it has been recorded as endemic in frog populations that did not suffer
17   disease, where it may be commensal rather than parasitic (Puschendorf et al., 2006; Puschendorf et al., 2011;
18   Rowley and Alford, 2013). Projections of future impacts are difficult, as the virulence of Bd is variable
19   across Bd populations and dependent upon the evolutionary and ecological histories, and evolutionary
20   potentials, of both the local amphibian populations and the endemic or invading Bd (robust evidence, high
21   agreement) (Retallick et al., 2004; Daskin et al., 2011; Puschendorf et al., 2011; Phillips and Puschendorf,
22   2013; Rowley and Alford, 2013; Zumbado-Ulate et al., 2014; Sapsford et al., 2015; Voyles et al., 2018;
23   Bradley et al., 2019; Fisher and Garner, 2020; McMillan et al., 2020). Further, specific local habitats might
24   serve as regional climate refugia from chytrid infection (e.g. hot and dry) (medium evidence, high
25   agreement) (Zumbado-Ulate et al., 2014; Cohen et al., 2019b; Neely et al., 2020; Turner et al., 2021).
28 Effects on geographic distribution and connectivity patterns of pathogens
29   As species’ geographic ranges and migration patterns are modified by climate change (Section, Table
30   2.2), pathogens accompany them. Diverse vectors and associated parasites, pests, and pathogens of plants
31   and animals are being recorded at higher latitudes and elevations in conjunction with regional temperature
32   increases and precipitation changes (robust evidence, high agreement), although analysis of realized disease
33   incidence often lacks inclusion of non-climatic vs climate drivers, compromising attribution (Ollerenshaw
34   and Rowlands, 1959; Purse et al., 2005; Laaksonen et al., 2010; van Dijk et al., 2010; Alonso et al., 2011;
35   Genchi et al., 2011; Pinault and Hunter, 2011; Jaenson et al., 2012; Loiseau et al., 2012; Kweka et al., 2013;
36   Medlock et al., 2013; Dhimal et al., 2014a; Dhimal et al., 2014b seasonal; Siraj et al., 2014; Khatchikian et
37   al., 2015; Hotez, 2016a; Hotez, 2016b; Bett et al., 2017; Mallory and Boyce, 2017; Strutz, 2017; Booth,
38   2018; Dumic and Severnini, 2018; Carignan et al., 2019; Gorris et al., 2019; Le et al., 2019; Stensgaard et
39   al., 2019b snails and; Brugueras et al., 2020; Gilbert, 2021).
41   At least six major VBDs affected by climate drivers have recently emerged in Nepal and are now considered
42   endemic, with climate change implicated as a primary driver as LULCC has been assessed to have a minimal
43   influence on these diseases (high confidence) (Table SM2.1). There is increasing evidence that climate
44   warming has extended the elevational distribution of Anopheles, Culex and Aedes mosquito vectors above
45   2,000 m in Nepal (limited evidence, high agreement) (Dahal, 2008; Dhimal et al., 2014a; Dhimal et al.,
46   2014b; Dhimal et al., 2015) with similar trends being recorded in neighboring Himalayan regions (medium
47   evidence, high agreement) (Phuyal et al., 2020; Dhimal et al., 2021). Host animals in novel areas may be
48   immunologically naive, and therefore more vulnerable to severe illness (Bradley et al., 2005; Hall et al.,
49   2016).
51   Case study 3: Arctic and subarctic disease expansion and intensification
53   High Arctic regions have warmed by more than double the global average, >2°C in most areas (see Sections
54, Figure 2.11, and Atlas, in WGI). Experimental, field ecology studies and computational
55   models in Arctic and subarctic regions indicate that milder winters have reduced mortality of vectors and
56   reservoir hosts and increased their habitat as forested taiga expands into previously treeless tundra (Table
57   SM2.1; Parkinson et al., 2014). Warmer temperatures and longer seasonal windows have allowed faster

     Do Not Cite, Quote or Distribute                     2-38                                      Total pages: 237
     FINAL DRAFT                                      Chapter 2                    IPCC WGII Sixth Assessment Report

 1   reproduction/replication, accelerated development, and increased the number of generations per year of
 2   pathogens, vectors and some host animals, that in turn increase the populations of disease organisms and
 3   disease transmission (Sections, Ticks, mosquitos, culicoides biting midges, deer flies,
 4   horseflies, and simuliid black flies that transmit a variety of pathogens are being documented in high-latitude
 5   regions at higher numbers or where they have been historically absent (robust evidence, high agreement)
 6   (Waits et al., 2018; Caminade et al., 2019; Gilbert, 2021). In concert with these poleward shifts of hosts and
 7   vectors, pathogens, particularly tick-borne pathogens and helminth infections, have increased dramatically in
 8   incidence and severity from once rare occurrences and have appeared in new regions (robust evidence, high
 9   agreement, very high confidence) (Caminade et al., 2019; Gilbert, 2021).
11   Zoonoses and VBDs that have been historically rare or never documented in Arctic and subarctic regions of
12   Europe, Asia, and North America, such as anthrax, cryptosporidiosis, elaphostrongylosis, filariasis (Huber et
13   al., 2020), tick-borne encephalitis, and tularemia (Evander and Ahlm, 2009; Parkinson et al., 2014; Pauchard
14   et al., 2016), are spreading poleward and increasing in incidence (robust evidence, high agreement, very high
15   confidence) (Table SM2.1; Omazic et al., 2019). Recent anthrax outbreaks and mass mortality events among
16   humans and reindeer, respectively, have been linked to abnormally hot summer temperatures that caused
17   permafrost to melt and exposed diseased animal carcasses, releasing thawed, highly infectious, Bacillus
18   anthracis spores (medium evidence, medium agreement) (Ezhova et al., 2019; Hueffer et al., 2020; Ezhova et
19   al., 2021). Multiple contributing factors conspired over different time scales to compound a 2016 anthrax
20   outbreak occurring on the Yamal peninsula: (i) rapid permafrost thawing for 5 years preceding the outbreak;
21   (ii) thick snow cover the year before the outbreak insulated the warmed permafrost and kept it from re-
22   freezing; and (iii) anthrax vaccination rates had decreased or ceased in the region (Ezhova et al., 2019;
23   Ezhova et al., 2021). These precursors converged with an unusually dry and hot summer that: (i) melted
24   permafrost, creating an anthrax exposure hazard; (ii) increased the vector insect population; and (iii)
25   weakened the immune systems of reindeer thus increasing their susceptibility (Waits et al., 2018; Hueffer et
26   al., 2020).
28   Warmer temperatures have increased blood-feeding insect harassment of reindeer with compounding
29   consequences: (1) increased insect bite rates lead to higher parasite loads, (2) time spent by raindeer in trying
30   to escape biting flies reduces foraging while simulataneously increasing energy expenditure, (3) the
31   combination of (1) and (2) lead to poor body condition, that subsequently leads to (4) reduced winter
32   survival and fecundity (Mallory and Boyce, 2017). As temperatures warm and connectivity increases
33   between the Arctic and the rest of the world, tourism, resource extraction, and increased commercial
34   transport will create additional risks of biological invasion by infectious agents and their hosts (Pauchard et
35   al., 2016). These increases in introduction risk compounded with climate change have already begun to harm
36   indigenous peoples dependent on hunting and herding livestock (horses and reindeer) that are suffering
37   increased pathogen infection (Deksne et al., 2020; Stammler and Ivanova, 2020).
39 Biodiversity-disease links
40   Anthropogenic impacts, such as disturbances caused by climate change, can reduce biodiversity through
41   multiple mechanisms and increase disease risk to humans (limited evidence, low agreement) but more
42   research is needed to understand the underlying mechanisms (Civitello et al., 2015; Young et al., 2017b;
43   Halliday et al., 2020; Rohr et al., 2020; Glidden et al., 2021). Known wildlife hosts of human-shared
44   pathogens and parasites overall comprise a greater proportion of local species richness (18–72% higher) and
45   abundance (21–144% higher) in sites under substantial human use (agricultural and urban lands) compared
46   with nearby undisturbed habitats (Gibb et al., 2020).
48   Exploitation of wildlife and degradation of natural habitats have increased opportunities for ‘spill over’ of
49   pathogens from wildlife to human populations and increased emergence of zoonotic disease epidemics and
50   pandemics (robust evidence, high agreement); animal and human migrations driven by climate change have
51   added to this increased risk (medium evidence, medium agreement) (see Section, Chapter 8, Cross-
52   Chapter Box MOVING PLATE in Chapter 5; Patz et al., 2004; Cleaveland et al., 2007; Karesh et al., 2012;
53   Altizer et al., 2013; Allen et al., 2017; Plowright et al., 2017; Olivero et al., 2017; Faust et al., 2018; Carlson
54   et al., 2020; Gibb et al., 2020; Hockings et al., 2020; IPBES, 2020; Volpato et al., 2020; Glidden et al.,
55   2021). Agricultural losses and subsequent food scarcity, that is increasing due to climate change, can also
56   lead to an increase in the use of bushmeat, and, hence increase risk of diseases jumping from wild animals to
57   humans (medium evidence, high agreement) (Brashares et al., 2004; Leroy et al., 2004; Wolfe et al., 2004;

     Do Not Cite, Quote or Distribute                       2-39                                        Total pages: 237
     FINAL DRAFT                                      Chapter 2                 IPCC WGII Sixth Assessment Report

 1   Rosen and Smith, 2010; Kurpiers et al., 2016).
 3 Implications for humans of changes in diseases in wild animals
 4   Changes in temperature, precipitation, humidity, and extreme events have been associated with more
 5   frequent disease outbreaks, increases in disease incidence and severity, and novel disease and vector
 6   emergence into new areas for wild animals, with a mechanistic understanding of the roles of these drivers
 7   from experimental studies providing high confidence for the role of climate change. However, attribution of
 8   how this has impacted human infectious diseases remains difficult, and definitive attribution studies are
 9   lacking. The specific role of recent climate change is difficult to examine in isolation for most regions where
10   human disease incidence has also been affected by land use change (particularly agricultural and urban
11   expansion), changes in public health access and measures, socio-economic changes, increased global
12   movements of people, and changes in vector and rodent control programs, supporting medium confidence in
13   the role of climate change driving observed changes in human diseases globally. Exceptions are in areas
14   noted above (Arctic, subarctic, and high elevation regions), in which climate change fingerprints are strong
15   and/or concurrent changes in non-climatic drivers are less pronounced than in other regions (high confidence
16   for climate change attribution) (see Table SM2.1, Sections,, Cross-Chapter Box ILLNESS
17   this Chapter; Harvell et al., 2002; Norwegian Polar Institute, 2009; Tersago et al., 2009; Tabachnick, 2010;
18   Altizer et al., 2013; Garrett et al., 2013; Paz, 2015; Wu et al., 2016b; Caminade et al., 2019; Dewage et al.,
19   2019; Coates and Norton, 2020; Deksne et al., 2020; Shocket et al., 2020; Couper et al., 2021; Gilbert,
20   2021).
23   [START FAQ2.2 HERE]
25   FAQ2.2: How does climate change increase the risk of diseases?
27   Climate change is contributing to the spread of diseases in both wildlife and humans. Increased contact
28   between wildlife and human populations increases disease risk and climate change is altering where
29   pathogens that cause diseases and the animals that carry them live. Disease risk can often be reduced by
30   improving health care and sanitation systems, training the medical community to recognize and treat
31   potential new diseases in their region, limiting human encroachment into natural areas, limiting wildlife
32   trade, and promoting sustainable and equitable socioeconomic development.
34   Diseases spread between humans and animals are called zoonoses. Zoonoses comprise nearly two-thirds of
35   known human infectious diseases and the majority of newly emerging infectious diseases (EIDs). COVID-19
36   is the most recent zoonosis and has killed millions of people globally while devastating economies. The risk
37   posed by EIDs has increased because of: (1) movement of wild animals and their parasites into new areas via
38   climate change, global trade, and travel; (2) human intrusion into and conversion of natural areas for
39   agriculture, livestock, industrial/raw materials extraction, and housing; (3) increased wildlife trade and
40   consumption; (4) increased human mobility resulting from global trade, war/conflicts, and migration made
41   faster and farther by fossil fuel powered travel; and (5) widespread antimicrobial use, which can promote
42   antibiotic resistant infections (Figure FAQ2.3.1).

     Do Not Cite, Quote or Distribute                      2-40                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 2   Figure FAQ2.2.1: How diseases move from the wild into human populations. Climate change may increase diseases in
 3   nature, but whether or not this leads to an increase in disease risk for humans depends upon a range of societal,
 4   infrastructure and medical buffers that form a shield protecting humans.
 7   Climate change further increases risk by altering pathogen and host animal (1) geographic ranges and
 8   habitats; (2) survival, growth, and development; (3) reproduction and replication; (4) transmission and
 9   exposure (5) behavior; and (6) access to immunologically naïve animals and people who lack infection
10   resistance. This can lead to novel disease emergence in new places, more frequent and larger outbreaks, and
11   longer or shifted seasons of transmission. Climate change is making it possible for many EIDs to colonize
12   historically colder areas that are becoming warmer and wetter in temperate and polar regions and in
13   mountains. Vector-borne diseases (VBDs) are diseases spread by vectors such as mosquitoes, sand flies,
14   kissing bugs, and ticks. For example, ticks that carry the virus that causes tick-borne encephalitis have
15   moved into northern subarctic regions of Asia and Europe. Viruses like dengue, chikungunya, and Japanese
16   encephalitis are emerging in Nepal in hilly and mountainous areas. Novel outbreaks of Vibrio bacteria
17   seafood poisoning are being traved to the the Baltic States and Alaska where they were never documented
18   before. Many scientific studies show that infectious disease transmission and the number of individuals
19   infected depends on rainfall and temperature; climate change often makes these conditions more favourable
20   for disease transmission.
22   Climate change can also have complicated, compounding, and contradictory effects on pathogens and
23   vectors. Increased rainfall creates more habitat for mosquitoes that transmit diseases like malaria but too
24   much rain washes away the habitat. Decreased rainfall also increases disease risk when people without
25   reliable water access use containers to store water in that mosquitos, such as the vectors of dengue fever,
26   Aedes aegypti and Ae. albopictus, use for egg laying. Hotter temperatures also increase mosquito bite rate,

     Do Not Cite, Quote or Distribute                      2-41                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   parasite development, and viral replication! Certain species of snails are intermediate hosts for many
 2   helminth worm parasites that make humans, livestock, and wild animals sick. When it gets hot, the snails can
 3   produce two to three times as many infective larvae but if it becomes too hot many pathogens and their
 4   vectors cannot survive or reproduce.
 6   Humans also contract zoonoses directly through their skin, mucus membranes, and lungs when eating or
 7   butchering animals, or coming into contact with pathogen shed in the air, urine, or faeces that contaminates
 8   water, food, clothing, and other surfaces. Any activity that increases contact with wildlife, especially in high
 9   biodiversity regions like the tropics and subtropics, increases disease risk. Climate change-related disease
10   emergence events are often rare but may become more frequent. Fortunately, there are ways to reduce risks
11   and protect our health, as described below.
13   Habitat and biodiversity protection: Encroachment of humans into natural areas, due to expansion of
14   agriculture and livestock, timber harvests, resource extraction, and urban development has increased human
15   contact with wild animals, and creates more opportunities for disease spillover (transmission from an animal
16   to a new species, including humans). By conserving, protecting, and restoring wild habitats, we can build
17   healthier ecosystems that provide other services, such as clean air, clean and abundant water, recreation,
18   spiritual value, and well-being, as well as reduced disease spillover. If humans must go into wild areas or
19   hunt, they should take appropriate precautions such as wearing protective clothing, using insect repellant,
20   performing body checks for vectors like ticks, and washing hands and clothing well.
22   Food resilience: Investing in sustainable agroecological farming will alleviate the pressure to hunt wild
23   animals and reduce the conversion of more land to agriculture/livestock use. Stopping illegal animal trading
24   and poaching and decreasing reliance on wild meats and products made from animal parts will reduce direct
25   contact with potentially infected animals. This has the added benefit of increasing food security, nutrition,
26   improving soil, reducing erosion, preserving biodiversity, and mitigating climate change.
28   Disease prevention and response: The level of protection against infection is linked directly to the level of
29   development and wealth of a country. Improved education, high-quality medical and veterinary systems,
30   high food security, proper sanitation of water and waste, high housing quality, and disease surveillance and
31   alarm systems dramatically reduce disease risk and improve health. Utilizing a One Biosecurity or One
32   Health framework further improves resilience. Sharing knowledge within communities, municipalities,
33   regional, and between national health authorities globally is important to assessing, preventing and
34   responding to outbreaks and pandemics more efficiently and economically.
36   Humans are facing many direct or indirect challenges because of climate change. Increasing EIDs is one of
37   our greatest challenges, due to our ever-growing interactions with wildlife and the climatic changes creating
38   new disease transmission patterns. COVID-19 is a current crisis, and follows other recent EIDs: SARS,
39   HIV/AIDS, H1N1 influenza, Ebola, Zika, and West Nile fever. EIDs have accelerated in recent decades,
40   making it clear that new societal and environmental approaches to wildlife interactions, climate change, and
41   health are urgently needed to protect our current and future well-being as a species.
43   [END FAQ2.2 HERE]
46   Observed Evolutionary Responses to Climate Change
48   Prior sections document species' tendencies to retain their climate envelopes by some combination of range
49   shift and phenological change. However, this tracking of climate change can be incomplete, causing species
50   or populations to experience hotter conditions than those to which they are adapted and thereby incur
51   'climate debts' (Devictor et al., 2012). The importance of population-level debt is illustrated by a study in
52   which estimated debt values were correlated with population dynamic trends in a North American migratory
53   songbird, the Yellow Warbler, Setophaga petechia. Populations that were genetic outliers for their local
54   climate space had larger population declines (greater debt) than populations with genotypes closer to the
55   average values for that particular climate space. Debt values were estimated from genomic analyses
56   independent of the population trends, and were distributed across the species' range in a mosaic, not simply

     Do Not Cite, Quote or Distribute                      2-42                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   concentrated at range margins, rendering the results robust to being confounded by broad-scale geographical
 2   trends (Bay et al., 2018).
 4   In the absence of evolutionary constraints, climate debts can be cancelled by genetically-based increases in
 5   thermal tolerance and ability to perform in high ambient temperatures. In species already showing local
 6   adaptation to climate, populations currently living at relatively cool sites should be able to evolve to adopt
 7   traits of populations currently at warmer sites, as their local experience of climate changes (Singer, 2017;
 8   Socolar et al., 2017).
10   An increasing number of studies document evolutionary responses to climate change in populations not at
11   warm range limits (Franks and Hoffmann, 2012). Organisms with short generation times should have higher
12   capacity to genetically track climate change than species with long generation times, such as mammals
13   (Boutin and Lane, 2014). Indeed, observed evolutionary impacts have been mainly documented in insects,
14   especially at expanding range margins (Chuang and Peterson, 2016) where evolutionary changes have been
15   documented of dispersal ability (Thomas et al., 2001) and host specialisation (Bridle et al., 2014; Lancaster,
16   2020).
18   Away from range margins, individual populations experiencing regional warming have evolved diverse traits
19   related to climate adaptation. For example, pitcher-plant mosquitos (Wyeomyia smithii) in Pacific NW
20   America have evolved to wait for shorter day lengths before initiating diapause. This adaptation to
21   lengthening summers enables them to delay overwintering until later and add an extra generation each year
22   (Bradshaw and Holzapfel, 2001). Among 26 populations of Drosophila subobscura studied on three
23   continents, 22 experienced climate warming across two or more decades, and 21 of those 22 showed
24   increasing frequencies of chromosome inversions characteristic of populations adapted to hot climates
25   (Balanya et al., 2006).
27   However, for populations already at their warm range limits, their ability to track climate change in situ
28   would require evolving to survive and reproduce outside their species' historical climate envelope, which is
29   not supported by experimental or observational evidence (medium evidence, high agreement) (Singer, 2017).
30   Whether or not they can do so depends on the level of 'niche conservatism' operating at the species level
31   (Lavergne et al., 2010). If a species' whose range limits are determined by climate finds itself completely
32   outside of its traditional climate envelope, extinction is expected in the absence of 'evolutionary rescue' (Bell
33   and Gonzalez, 2009; Bell et al., 2019). To investigate the evolutionary potential enabling a species to survive
34   in a novel climate entirely outside its traditional climate envelope, experiments have been carried out on
35   ectotherms testing thermal performances, thermal tolerances, and their evolvabilities (Castaneda et al., 2019;
36   Xue et al., 2019). Tests of thermal performance have been complicated as both long-term acclimation and
37   transgenerational effects occur (Sgro et al., 2016). However, the results to date have been consistent: despite
38   widespread local adaptation to climate across species' ranges, substantial constraints exist to the evolution of
39   greater stress tolerance (e.g. high temperatures and drought) at warm range limits (medium evidence, high
40   agreement) (Hoffmann and Sgro, 2011; MacLean et al., 2019b). For example, as temperature was
41   experimentally increased, the amount of genetic variance in fitness of Drosophila melanogaster decreased:
42   in hot environments, flies had low evolvability (Kristensen et al., 2015). The hypothesis that heat stress
43   tolerance is evolutionarily constrained is further supported by experiments in which 22 Drosophila species
44   drawn from tropical and temperate climes were subjected to extremes of heat and cold. They differed as
45   expected in cold tolerances, but not in heat tolerances nor in temperatures at which optimal performances
46   were observed (MacLean et al., 2019b).
48   Plasticity in acclimating to thermal regimes helps organisms adapt to environmental change. The form and
49   extent of plasticity can vary among populations experiencing different climates (Kelly, 2019) and generate
50   phenotypic values outside the prior range for the species, but plasticity itself has not yet been observed to
51   evolve in response to climate change (Kelly, 2019). Relevant genetic changes in nature (e.g. affecting heat
52   tolerance) have not yet been shown to alter the boundaries of existing genetic variation for any species.
53   Evolutionary rescue of entire species has not yet been observed in nature, nor is it expected based upon
54   experimental and theoretical studies (medium evidence, high agreement).
56   Hybridisation between closely related species has increased in recent decades as one species shifts its range
57   boundaries and positions itself more closely to the other—hybrids between polar bears and brown bears have

     Do Not Cite, Quote or Distribute                      2-43                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   been documented in northern Canada (Kelly et al., 2010). In North American rivers, hybridisation between
 2   invasive rainbow trout and native cutthroat trout has increased in frequency as the rainbow trout expanded
 3   into warming waters (Muhlfeld et al., 2014). Whether climate-changed induced hybridisations can generate
 4   novel climate adaptations remains to be seen.
 6   In summary, with present knowledge, evolution is not expected to be sufficient to prevent whole species'
 7   extinctions if a species' climate space disappears (high confidence).
 9   2.4.3   Observed Changes in Key Biomes, Ecosystems and their Services
11 Detection and Attribution for Observed Biome Shifts
13   Attribution for biome (major vegetation form of an ecosystem) shifts is complex because of their extensive,
14   sometimes continental, spatial scale (Whittaker, 1975; Olson et al., 2001; Woodward et al., 2004); and
15   therefore, non-climatic factors strongly influence biome spatial distributions (Ellis and Ramankutty, 2008).
17   The most robust attribution studies use data from individual locations with minimal confounding factors,
18   particularly recent land use change, and scale up by analysing multiple locations across a long zone between
19   biomes. As with individual species, multiple lines of evidence increase confidence (Hegerl et al., 2010;
20   Parmesan et al., 2013). Multivariate statistical analyses aid attribution studies by allowing the assessment of
21   relative weights among multiple factors, including variables related to climate change (Gonzalez et al.,
22   2012). However, drivers often have strong, significant interactions with one another, complicating
23   quantitative assessment of the strength of individual drivers (Parmesan et al., 2013). In these cases,
24   manipulative experiments are critical in assessing attribution to climate change drivers.
26   Certain biomes exhibit a relatively stronger relationship to climate; for example, Arctic tundra generally has
27   a distinct ecotone with boreal conifer forest (Whittaker, 1975). In these areas, attribution of biome shifts to
28   climate change are relatively straightforward, if human land use change is minimal. However, other biomes,
29   such as many grassland systems, are not at equilibrium with climate (Bond et al., 2005). In these systems
30   their evolutionary history (Keeley et al., 2011; Strömberg, 2011; Charles-Dominique et al., 2016),
31   distribution, structure and function have been shaped by climate and natural disturbances, such as fire and
32   herbivory (Staver et al., 2011; Lehmann et al., 2014; Pausas, 2015; Bakker et al., 2016; Malhi et al., 2016).
33   Disturbance variability is an inherent characteristic of grassland systems and suitable “control” conditions
34   are seldom available in nature. Furthermore, due to the integral role of disturbance, these biomes have been
35   widely affected by long-term and widespread shifts in grazing regimes, large-scale losses of mega-
36   herbivores and fire suppression policies (Archibald et al., 2013; Malhi et al., 2016; Hempson et al., 2017). It
37   is necessary to conduct climate change attribution on a case-by-case basis for grasslands; such assessments
38   are complex as direct climate change impacts from either inherent variation within disturbance regimes or
39   directional changes in background disturbances are difficult to separate (detailed in Sections;
40; Confidence in assessments is increased when observed trends are supported by
41   mechanistic understanding of responses identified by physiological studies, manipulative field experiments,
42   greenhouse studies and lab experiments (Table SM2.1).
44 Global Patterns of Observed Biome Shifts Driven by Climate Change
46 Observed biome shifts predominantly driven by climate change
47   Th IPCC Fifth Assessment Report and a meta-analysis found that vegetation at the biome level shifted
48   poleward latitudinally and upward altitudinally due to anthropogenic climate change at 19 sites in boreal,
49   temperate, and tropical ecosystems from 1700 to 2007 (Gonzalez et al., 2010a; Settele et al., 2014). In these
50   areas, temperature increased 0.4° to 1.6ºC above the pre-industrial period (Gonzalez et al., 2010a; Settele et
51   al., 2014). Field research since the IPCC Fifth Assessment Report detected additional poleward and upslope
52   biome shifts over periods of 24 to 210 years at numerous sites (described below) but were not directly
53   attributed to anthropogenic climate change as the studies were not designed nor conducted properly for
54   attribution.
56   Many of the recently detected shifts were nevertheless consistent with climate change temperature increases
57   and observed in areas lacking agriculture, livestock grazing, timber harvesting, or other anthropogenic land

     Do Not Cite, Quote or Distribute                      2-44                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   uses. For example, in the Andes Mountains in Ecuador, a biome shift was detected by comparing a survey by
 2   Alexander von Humboldt in 1802 to a re-survey in 2012, making this the longest time span in the world for
 3   this type of data (Morueta-Holme et al., 2015) and 2017 (Moret et al., 2019). During 210 years, temperature
 4   increased 1.7ºC (Morueta-Holme et al., 2015) and the upper edge of alpine grassland shifted upslope 100‒
 5   450 m (Moret et al., 2019).
 7   Other biome shifts consistent with climate change and not substantially affected by local land use include
 8   northward shifts of deciduous forest into boreal conifer forest in Canada (5 km between 1970-2012, (Sittaro
 9   et al., 2017) and 20 km between 1970-2014, (Boisvert-Marsh et al., 2019)) and northward shifts of
10   temperate conifer into boreal conifer forest in Canada (21 km between 1970-2015, (Boisvert-Marsh and de
11   Blois, 2021)). Research detected upslope shifts of boreal and sub-alpine conifer forest into alpine grassland
12   at 143 sites on four continents (41 m, 1901-2018, (Lu et al., 2021)) and individual sites in Canada (54 m,
13   1900-2010, (Davis et al., 2020)), China (300 m, 1910-2000 (Liang et al., 2016); 33 m, 1985-2014, (Du et
14   al., 2018)), Nepal (50 m, 1860-2000, (Sigdel et al., 2018)), Russia (150 m, 1954-2006, (Gatti et al., 2019))
15   and the United States (19 m, 1950-2016, (Smithers et al., 2018); 38 m, 1953-2015, (Terskaia et al., 2020)).
16   Other upslope cases include shifts of temperate conifer forest in Canada (Jackson et al., 2016) and the United
17   States (Lubetkin et al., 2017), temperate deciduous forest in Switzerland (Rigling et al., 2013) and temperate
18   shrubland in the United States (Donato et al., 2016).
20   In summary, anthropogenic climate change has caused latitudinal and elevational biome shifts in at least 19
21   sites in boreal, temperate, and tropical ecosystems between 1700 and 2007, where temperature increased 0.4°
22   to 1.6ºC above the pre-industrial period (robust evidence, high agreement). Additional cases of 5 to 20 km
23   northward and 20 to 300 m upslope biome shifts between 1860 and 2016, under approximately 0.9ºC mean
24   global temperature increase above the pre-industrial period, are consistent with climate change (medium
25   evidence, high agreement).
28 Observed biome shifts from combined land use change and climate change
29   Research has detected biome shifts in areas where agriculture, fire use or suppression, livestock grazing,
30   timber and fuelwood harvesting, or other local land use actions substantially altered vegetation, in addition to
31   changes in climatic factors and CO2 fertilisation. These studies were not designed or conducted in a manner
32   to make climate change attribution possible, although vegetation changes are consistent with climate change:
33   for example, a global review of observed changes in treelines found that 2/3 of treelines globally have
34   shifted upslope in elevation over the past 50 years or more (Hansson, 2021, a review of).
36   Upslope and poleward forest shifts have occurred where timber harvesting or livestock grazing was
37   abandoned, allowing regeneration of trees at sites in Canada (Brice et al., 2019; Wang et al., 2020b), France
38   (Feuillet et al., 2020), Italy (Vitali et al., 2017), Spain (Ameztegui et al., 2016), the United States (Wang et
39   al., 2020b) and mountain areas across Europe (Cudlin et al., 2017). Intentional use of fire drove an upslope
40   forest shift in Peru (Bush et al., 2015) while mainly human-ignited fires drove conversion of shrubland to
41   grassland in a drought-affected area of the United States (Syphard et al., 2019b). In eastern Canada, timber
42   harvesting and wildfire drove conversion of mixed conifer-broadleaf forests to broadleaf-dominated forests
43   (Brice et al., 2020; Wang et al., 2020b).
45   Shrub encroachment onto savanna has occurred at numerous sites, particularly across the Southern
46   Hemisphere, mainly between 1992 and 2010 (Criado et al., 2020). Globally, overgrazing initiates shrub
47   encroachment by reducing grasses more than woody plants, while fire exclusion maintains the shrub cover
48   (D'Odorico et al., 2012; Caracciolo et al., 2016; Bestelmeyer et al., 2018). The magnitude of woody cover
49   change in savannas is not correlated to mean annual temperature change (Criado et al., 2020), however,
50   higher atmospheric CO2 increases shrub growth in savannas (Nackley et al., 2018; Manea and Leishman,
51   2019). A global remote sensing analysis of biome changes from all causes, including agricultural and grazing
52   expansion and deforestation, estimated that 14% of pixels changed between 1981 and 2012, although this
53   approach can overestimate global changes since it uses a new biome classification system, which doubles the
54   conventional biome classifications (Higgins et al., 2016). In addition to climate change, land use change
55   causes vegetation changes at the biome level (robust evidence, high agreement).

     Do Not Cite, Quote or Distribute                      2-45                                       Total pages: 237
     FINAL DRAFT                                    Chapter 2                   IPCC WGII Sixth Assessment Report

 1 Observed Changes in Deserts and Arid Shrublands
 3   Divergent responses to anthropogenic climate change are occurring within and across arid regions,
 4   depending on time period, location, detection methodology and vegetation type (see Cross-Chapter Paper 3).
 5   Emerging shifts in ecosystem structure, functioning and biodiversity are supported by evidence from
 6   modelled impacts of projected climate and CO2 levels. While observed responsiveness of arid vegetation
 7   productivity to rising atmospheric CO2 (Fensholt et al., 2012b) may offset risks from reduced water
 8   availability (Fang et al., 2017), climate- and CO2 driven changes are key risks in arid regions, interacting
 9   with habitat degradation, wildfire, and invasive species (Hurlbert et al., 2019).
11   Widespread vegetation greening, as projected in AR4, is occurring in arid shrublands (Zhang et al., 2019a;
12   Maestre et al., 2021) as a result of increases in leaf area, woody cover and herbaceous production at desert-
13   grassland interfaces (Gonsamo et al., 2021). Plant productivity in arid regions has increased (Fensholt et al.,
14   2012b) because of improved water use efficiency associated with elevated CO2 (Norby and Zak, 2011;
15   Donohue et al., 2013; Burrell et al., 2020; Gonsamo et al., 2021) (medium evidence, high agreement), altered
16   rainfall seasonality and amount (Rohde et al., 2019; Zhang et al., 2019a ) (robust evidence, high agreement),
17   increases in temperature (Ratajczak et al., 2014; Wilcox et al., 2018)(robust evidence, high agreement) and
18   heavy grazing (robust evidence, high agreement) with relative importance differing among locations
19   (Donohue et al., 2013; Caracciolo et al., 2016; Archer et al., 2017; Hoffmann et al., 2019b; Rohde et al.,
20   2019). Woody plant encroachment into arid shrublands is occurring in North America (Caracciolo et al.,
21   2016; Archer et al., 2017), southern Africa (du Toit and O’Connor, 2014; Ward et al., 2014; Masubelele et
22   al., 2015a; Hoffman et al., 2019; Rohde et al., 2019) (high confidence) and Central Asia (Li et al., 2015) (low
23   confidence). In North America, sagebrush steppe changes have been attributed to increases in temperature
24   and earlier snowpack melt (Wuebbles et al., 2017; Mote et al., 2018; Snyder et al., 2019).
26   Non-native grasses are invading the sagebrush steppes (cold deserts) in North America (Chambers et al.,
27   2014) attributed to warming (Bradley et al., 2016; Hufft and Zelikova, 2016). In the eastern semi-desert
28   (Karoo) of South Africa, annual rainfall increases and a rainfall seasonality shift (du Toit and O’Connor,
29   2014) are increasing grassiness as arid grasslands expand into semi-desert shrublands (du Toit et al., 2015;
30   Masubelele et al., 2015b; Masubelele et al., 2015a) causing fire in areas seldom burned (Coates et al., 2016).
32   Drought, warming, and land management interactions have caused vegetation mortality (see section
33   and reduced vegetation cover in shrublands as projected by AR4 (Burrell et al., 2020). Increased heat and
34   drought are causing succulent species health and abundance to decline (Musil et al., 2009; Schmiedel et al.,
35   2012; Aragón-Gastélum et al., 2014; Koźmińska et al., 2019). Hot droughts especially reduce population
36   resilience (medium confidence) (Koźmińska et al., 2019).
38 Observed Changes in Mediterranean-Type Ecosystems
40   Since AR5, Settele et al. (2014) found that all five Mediterranean-Type Ecosystems (MTEs) of the world
41   experienced extreme droughts within the past decade, with South Africa and California reporting the worst
42   on record (robust evidence, high agreement) (Diffenbaugh et al., 2015; Williams et al., 2015a; Garreaud et
43   al., 2017; Otto et al., 2018; Sousa et al., 2018). Climate change is causing these droughts to become more
44   frequent and severe (medium evidence, medium agreement) (AghaKouchak et al., 2014, Garreaud et al 2017
45   The 2010-2015 megadrought , AR6 WGI Chpt 11; Otto et al., 2018).
47   MTEs show a range of direct responses to various forms of water deficit, but have also been affected by
48   increasing fire activity linked to drought (Abatzoglou and Williams, 2016), and interactions between drought
49   or extreme weather and fire, affecting post-fire ecosystem recovery (Slingsby et al., 2017). Responses
50   include shifts in functional composition (Acácio et al., 2017; Syphard et al., 2019a), decline in vegetation
51   health (Hope et al., 2014; Asner et al., 2016a), decline or loss of characteristic species (White et al., 2016;
52   Stephenson et al., 2019), shifts in composition towards more drought- or heat-adapted species and declining
53   diversity (also see Section; Slingsby et al., 2017.; Harrison et al., 2018).
55   Declines in plant health and increased mortality in MTEs associated with drought have been widely
56   documented (robust evidence, high agreement) (Section Remote sensing studies show drought
57   associated mortality in postfire vegetation regrowth in the Fynbos of South Africa (Slingsby et al., 2020b),

     Do Not Cite, Quote or Distribute                     2-46                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   reduced canopy health in forests within MTE zones of South Africa (Hope et al., 2014), and declines in
 2   canopy water content in forests of California (Asner et al., 2016a). Several studies reported climate-
 3   associated responses of dominant or charismatic species. High mortality in the Clanwilliam Cedar between
 4   1931 to 2013 occurred at lower, hotter elevations in the Fynbos of South Africa (White et al., 2016). Drought
 5   reduced growth and increased mortality of the holm oak, Quercus ilex, in the Iberian Peninsula of Spain,
 6   Natalini et al. (2016). Portuguese shrublands experienced losses of many deciduous and evergreen oak
 7   species, and increasing dominance of to pyrophytic xeric trees (Acácio et al., 2017). The 2012–2015 drought
 8   in California caused: high canopy foliage die-back of the Giant Sequoia (Sequoiadendron giganteum)
 9   (Stephenson et al., 2019), increased the dominance of oaks relative to pines resulting from increased water
10   deficit, and large-scale mortality from drought and insect pest outbreak interactions (McIntyre et al., 2015;
11   Fettig et al., 2019).
13   Species distribution or community composition changes have contributed to declines in diversity and/or
14   shifts towards more drought- or heat-adapted species (medium evidence, high agreement). Two conifer
15   species (Pinus longaeva, P. flexilis) shifted upslope 19 m from 1950 to 2016 in the Great Basin, USA,
16   (Smithers et al., 2018). Reduced winter precipitation caused native annual forbs to recede resulting in long-
17   lasting and potentially unidirectional reductions in diversity in a Californian grassland (Harrison et al.,
18   2018). More frequent extreme hot and dry weather between 1966 and 2010 caused declines in diversity
19   during the post-fire regeneration phase in the Fynbos of South Africa (Slingsby et al., 2017) resulting in
20   shifts towards species with higher temperature preferences (Slingsby et al., 2017). In Italy, Del Vecchio et al.
21   (2015) observed increases in plant cover and thermophilic species in coastal foredune habitats between 1989
22   and 2012.
24   In southern California, USA, areas of forest and woody shrublands are shifting to grasslands, driven by a
25   combination of climate and land use factors such as increased drought, fire ignition frequency and increases
26   in nitrogen deposition (robust evidence, high agreement) (Jacobsen and Pratt, 2018; Park et al., 2018; Park
27   and Jenerette, 2019; Syphard et al., 2019b).
29   The effects of climate change on heat, fuel, and wildfire ignition limitations show spatial and temporal
30   variation globally (see Section, but there have been a number of observed impacts in MTEs (medium
31   evidence, high agreement). Climate change has caused increases in fuel aridity and area burned by wildfire
32   across the western United States from 1985 to 2015 (Abatzoglou and Williams, 2016). Local and global
33   climatic variability led to a 4 year decrease in the average fire return time in Fynbos, South Africa when
34   comparing fires recorded between 1951 to 1975 and 1976 to 2000 (Wilson et al., 2010). For Chile, González
35   et al. (2018) reported a significant increase in the number, size, duration and simultaneity of large fires
36   during the 2010 to 2015 “megadrought” when compared to the 1990 to 2009 baseline.
38 Observed Changes in Savanna and Grasslands
40   Savannas consist of coexisting trees and grasses in the tropics and temperate regions (Archibald et al., 2019).
41   The global trend of woody encroachment reported in AR5 (Settele et al., 2014) is continuing (robust
42   evidence, high agreement, very high confidence) (see Table 2.S.1), with increases occurring in: temperate
43   savannas in North America (10-20% per decade) (Archer et al., 2017), tropical savannas in South America
44   (8% per decade), Africa ( 2.4% per decade) and Australia (1% per decade) (O'Connor et al., 2014; Espírito-
45   Santo et al., 2016; Skowno et al., 2017; Stevens et al., 2017; McNicol et al., 2018; Venter et al., 2018; Rosan
46   et al., 2019). Additionally, forest expansion into mesic savannas reported in AR5 (Settele et al., 2014), is
47   continuing in Africa, South America and Southeastern Asia (Marimon et al., 2014; Keenan et al., 2015;
48   Baccini et al., 2017; Ondei et al., 2017; Stevens et al., 2017; Aleman et al., 2018; Rosan et al., 2019).
49   Extreme high rainfall anomalies also contributed to an increase in herbaceous and foliar production in the
50   Sahel (Brandt et al., 2019; Zhang et al., 2019a).
52   New studies since AR5, using multiple study designs (experimental manipulations in lab and field, meta-
53   analyses and modelling), attribute climate change increases in woody cover to elevated atmospheric CO2
54   (Donohue et al., 2013; Nackley et al., 2018; Quirk et al., 2019) and increased rainfall amount and intensity
55   (robust evidence, high agreement) (Venter et al., 2018; Xu et al., 2018b; Zhang et al., 2019a). Direct
56   quantification of climate change drivers is confounded with local land use changes such as fire suppression
57   (Archibald, 2016; Venter et al., 2018), heavy grazing (du Toit and O’Connor, 2014; Archer et al., 2017),

     Do Not Cite, Quote or Distribute                      2-47                                      Total pages: 237
     FINAL DRAFT                                       Chapter 2                   IPCC WGII Sixth Assessment Report

 1   removal of native browsers, and specifically loss of mega-herbivores in Africa (medium evidence, medium
 2   agreement) (Asner et al., 2016b; Daskin et al., 2016; Stevens et al., 2016; Davies et al., 2018). The relative
 3   importance of the climate- and non-climate-related causes of woody plant vary between regions, but there is
 4   general agreement that climate change impacts, specifically, increasing rainfall and rising CO2, are frequent
 5   and strong contributing factors of woody cover increase (robust evidence, high agreement).
 7   Extensive woody cover increases in non-forested biomes is reducing grazing potential (Smit and Prins,
 8   2015), and changing the carbon stored per unit land area (González-Roglich et al., 2014; Puttock et al., 2014;
 9   Pellegrini et al., 2016; Mureva et al., 2018) and hydrological characteristics (Honda and Durigan, 2016;
10   Schreiner-McGraw et al., 2020). Woody cover encroachment also reduces biodiversity by threatening fauna
11   and flora adapted to open ecosystems (Ratajczak et al., 2012; Smit and Prins, 2015; Pellegrini et al., 2016;
12   Andersen and Steidl, 2019).
14   The global extent of grasslands is declining significantly because of climate change (medium confidence). In
15   temperate and boreal zones, where about half of treelines are shifting, they are overwhelmingly expanding
16   poleward and upward, with accompanying loss of montane grassland (robust evidence, high agreement);
17   whereas tropical treelines have been generally stable (medium evidence, medium agreement) (Harsch et al.,
18   2009; Rehm & Feeley 2015; Silva et al., 2016; Andela et al., 2017; Song et al., 2018; Aide et al., 2019;
19   Gibson and Newman, 2019). The Eurasian steppes experienced a 1% increase in woody cover per decade
20   since 2000 (Liu et al., 2021) and Inner Mongolian grasslands in China experienced broad encroachment as
21   well (Chen et al., 2015). Climatic drivers of woody expansion in temperature limited grasslands, particularly
22   alpine grasslands, are most frequently attributed to warming (robust evidence, high agreement, high
23   confidence) (D'Odorico et al., 2012; Hagedorn et al., 2014), increases in water and nutrient availability from
24   thawing permafrost (medium evidence, high agreement) (Zhou et al., 2015b; Silva et al., 2016) and rising
25   CO2 (medium evidence, medium agreement) (Frank et al., 2015; Aide et al., 2019). Interactions between land
26   use changes: land abandonment, grazing management shifts, and fire suppression, and climate change are
27   contributing factors (Liu et al., 2021)
29   Remote sensing shows overall increasing trends in both the annual maximum NDVI and annual mean NDVI
30   in global grasslands ecosystems between 1982 and 2011 (Gao et al., 2016). Multiple lines of evidence
31   indicate that changes in grassland productivity are positively correlated with increases in mean annual
32   precipitation (Hoover et al., 2014; Brookshire and Weaver, 2015; Gang et al., 2015; Gao et al., 2016; Wilcox
33   et al., 2017; Wan et al., 2018). Increasing temperatures positively impact grassland production and biomass,
34   especially in temperature limited regions (Piao et al., 2014; Gao et al., 2016). However, grasslands in hot
35   areas are expected to decrease production with increases in temperature (limited evidence, low agreement)
36   (Gang et al., 2015). Nevertheless, grassland responses to warming and drought are being ameliorated by
37   increasing CO2 and associated improved water use efficiency (Roy et al., 2016). For example, in a cool
38   temperate grassland experiment, warming led to a longer growing season and elevated CO2 further extended
39   growing by conserving water, which enabled most species to remain active longer (medium evidence,
40   medium agreement) (Reyes-Fox et al., 2014).
42 Observed Changes in Tropical Forest
44   Overall declines of tropical forest cover (Kohl et al., 2015; Liu et al., 2015; Baccini et al., 2017; Harris et al.,
45   2021), with declines more than triple the gains (Harris et al., 2021) have been driven primarily by
46   deforestation and land conversion (robust evidence, high agreement) (Lewis et al., 2015; Curtis et al., 2018;
47   Espaciais, 2021). In opposition to this general trend, expansion of tropical forest cover into savannas and
48   grasslands has occurred in Africa, South America, and Australia (Baccini et al., 2017; Aleman et al., 2018;
49   Staver, 2018) (Marimon et al., 2014; Ondei et al., 2017; Stevens et al., 2017; Rosan et al., 2019).
51   Specific examples of climate-change driven range shifts of tropical deciduous forests upslope into alpine
52   grasslands have been documented in the Americas (Chacón-Moreno et al., 2021; Jiménez-García et al.,
53   2021) and in Asia (Sigdel et al., 2018). However, treeline behaviours are diverse. A study in Nepal recorded
54   that treeline fomed by Abies spectabilis had been stable for more than a century, while the upper limit of
55   large shrubs (Rhododendron campanulatum) had been advancing (Mainali et al., 2020). In both the Andes
56   (Harsch et al., 2009) and Himalaya (Singh et al., 2021) most treelines have been stable, leading (Rehm &
57   Feeley 2015) to postulate a "grass ceiling" that has been difficult for trees to penetrate. The treeline shifts

     Do Not Cite, Quote or Distribute                        2-48                                        Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   that have occurred are probably driven by interactions between changing land-use, such as fire suppression,
 2   and climate changes such as increased rainfall, warming and elevated CO2 either through CO2 fertilisation or
 3   increases in water-use efficiency (medium evidence, medium agreement) (Cernusak et al., 2013; Huang et al.,
 4   2013; Van Der Sleen et al., 2015; Yang et al., 2016).
 6   Increases in productivity of tropical forests (Gatti et al., 2014; Brienen et al., 2015; Baccini et al., 2017),
 7   Africa and SE Asia (Qie et al., 2017) have been attributed to elevated CO2 (robust evidence, medium
 8   agreement) (Ballantyne et al., 2012; Brienen et al., 2015; Sitch et al., 2015; Yang et al., 2016; Mitchard,
 9   2018). The rates of these increases have been slowing down in the central Amazon (Brienen et al., 2015; de
10   Meira Junior et al., 2020) and SE Asia (Qie et al., 2017). In contrast, the carbon sink (and hence rate of
11   biomass gain) in intact African forests was stable until 2010 and has only recently started to decline,
12   indicating asynchronous carbon sink saturation in Amazonia and Africa, the difference driven by rates of
13   tree mortality (Hubau et al. 2020). At a global level (Hubau et al 2020) argue that the carbon sink associated
14   with intact tropical forests peaked in the 1990s and is now in decline.
16   Declines in productivity are most strongly associated with warming (Sullivan et al., 2020), reduced growth
17   rates during droughts (Bennett et al., 2015; Bonai et al., 2016; Corlett, 2016), drought related mortality
18   (Brando et al., 2014; Zhou et al., 2014; Brienen et al., 2015; Corlett, 2016; McDowell et al., 2018), fire (Liu
19   et al., 2017), and cloud-induced radiation-limitation (robust evidence, high agreement) (Deb Burman et al.,
20   2020). Increases in frequency and severity of droughts and shorter tree residence times due to increases in
21   growth rates caused by elevated CO2 may be additional interactive factors increasing tree mortality (Malhi et
22   al., 2014; Brienen et al., 2015). Vulnerability to drought varies between tree species and sizes with large,
23   long-lived trees at highest risk of mortality (McDowell et al., 2018; Meakem et al., 2018). Mortality risk also
24   varies between forest types with seasonal rainforests appearing most vulnerable to drought (Corlett, 2016).
26   Lianas (long-stemmed woody vines) generally negatively impact trees, significantly reducing the growth of
27   heavily infested trees (Reis et al., 2020). They would benefit from climate change and disturbance (LingZi et
28   al., 2014; Hodgkins et al., 2018). The extent of their suitable niche can increase (Taylor and Kumar, 2016),
29   thereby decreasing forest biomass accumulation (robust evidence, high agreement) (van der Heijden et al.,
30   2013; Fauset et al., 2015; Estrada-Villegas et al., 2020).
32   Climate change continues to degrade forests by reducing resilience to pests and diseases, increasing species
33   invasion, facilitating pathogen spread (Malhi et al., 2014; Deb et al., 2018) and intensifying fire risk and
34   potential die-back (Lapola et al., 2018; Marengo et al., 2018). Drought, temperature increases and forest
35   fragmentation interact to increase the prevalence of fires in tropical forests (robust evidence, high
36   agreement). Warming increases water stress in trees (Corlett, 2016) and together with forest fragmentation,
37   dramatically increases desiccation of forest canopies—resulting in deforestation that then leads to even
38   hotter and drier regional climates (Malhi et al., 2014; Lewis et al., 2015). Warming and drought increase
39   invasion of grasses into forest edges and increase fire risk (robust evidence, high agreement) (Brando et al.,
40   2014; Balch et al., 2015; Lewis et al., 2015). Droughts and fires additively increase mortality and,
41   consequently, reduce canopy cover and aboveground biomass ( Cross-Chapter Paper 7; Brando et al., 2014,
42   2020; Balch et al., 2015; Lewis et al., 2015).
44 Observed Changes in Boreal and Temperate Forests
46   The IPCC Fifth Assessment Report found increased tree mortality, wildfire and plant phenology changes in
47   boreal and temperate forests (Settele et al., 2014). Expanding on those conclusions, this Assessment, using
48   analyses of causal factors, attributes to anthropogenic climate change the following observed changes in
49   boreal and temperate forests in the 20th and 21st centuries: upslope and poleward biome shifts at sites in Asia,
50   Europe, and North America (Section; range shifts of plants (Section; earlier blooming and
51   leafing of plants (Section; poleward shifts in tree-feeding insects (Section; increases in
52   insect pest outbreaks (Section; increases in area burned by wildfire in western North America
53   (Section; increased drought-induced tree mortality in western North America (Section;
54   and thawing of permafrost that underlies extensive areas of boreal forest (IPCC Sixth Assessment Report,
55   Working Group I, Chapter 2, Section Atmospheric CO2 from anthropogenic sources has also
56   increased net primary productivity (Section In summary, anthropogenic climate change has
57   caused substantial changes to temperate and boreal forest ecosystems, including biome shifts and increases

     Do Not Cite, Quote or Distribute                      2-49                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   in wildfire, insect pest outbreaks, and tree mortality, at a global mean surface temperature increase of 0.9º C
 2   above the pre-industrial period (robust evidence, high agreement).
 4   Other changes detected in boreal forests and consistent with, but not formally attributed to climate change,
 5   include increased wildfire in Siberia (Section, long-lasting smoldering belowground fires in
 6   Canada and the United States (Scholten et al., 2021), tree mortality in Europe (Section, and post-
 7   fire shifts of boreal conifer to deciduous broadleaf tree species in Alaska (Mack et al., 2021). From 1930 to
 8   1960, boreal forest growth became limited more by precipitation than temperature in the Northern
 9   Hemisphere (Babst et al., 2019).
11   For some vegetation changes, land use and land management changes have exerted more influence than
12   climate change. These include upslope and poleward forest shifts in Europe following abandonment of
13   timber harvesting or livestock grazing (Section, changes in wildfire in Europe affected by fire
14   suppression, fire prevention, and agricultural abandonment (Section, and forest species
15   composition changes in Scotland due to nitrogen deposition from air pollution (Hester et al., 2019). Remote
16   sensing suggests that the area of temperate and boreal forests increased in Asia and Europe between 1982
17   and 2016 (Song et al., 2018) and in Canada between 1984 and 2015 (Guindon et al., 2018), but forest
18   plantations and regrowth are probable drivers (Song et al., 2018).
20 Observed Changes in Peatlands
22   Globally, peatland ecosystems store approximately 25% (600±100 GtC) of the world’s soil organic carbon
23   (Yu et al., 2010; Page et al., 2011; Hugelius et al., 2020) and 10% of the world’s freshwater resources
24   (Joosten and Clarke, 2002), despite only occupying 3% of the global land area (Xu et al., 2018a). The long-
25   term role of northern peatlands in the carbon cycle was mentioned for the first time in IPCC AR4 (IPCC,
26   2007b), while SR1.5 briefly mentioned the combined effects of climate and land-use change on peatlands
27   (IPCC, 2018b). New evidence confirms that climate change, including extreme weather events (e.g.,
28   droughts; Section, permafrost degradation (Section, sea-level rise (Section, and fire
29   (Section (Henman and Poulter, 2008; Kirwan and Mudd, 2012; Turetsky et al., 2015; Page and
30   Hooijer, 2016; Swindles et al., 2019; Hoyt et al., 2020; Hugelius et al., 2020; Jovani‐Sancho et al., 2021;
31   Veraverbeke et al., 2021), superimposed on anthropogenic disturbances (for example, draining for
32   agriculture or mining; Section, has led to rapid losses of peatland carbon across the world (robust
33   evidence, high agreement) (Page et al., 2011; Leifeld et al., 2019; Hoyt et al., 2020; Turetsky et al., 2020;
34   Loisel et al., 2021). Other essential peatland ecosystem services, such as water storage and biodiversity, are
35   also being lost worldwide (robust evidence, high agreement) (Bonn et al., 2014; Martin-Ortega et al., 2014;
36   Tiemeyer et al., 2017).
38   The switch from carbon sink to source in peatlands globally is mainly attributable to changes in water table
39   depth, regardless of management or status (robust evidence, high agreement) (Lafleur et al., 2005; Dommain
40   et al., 2011; Lund et al., 2012; Cobb et al., 2017; Evans et al., 2021; Novita et al., 2021). Across the
41   temperate and tropical biomes, extensive drainage and deforestation have caused widespread water table
42   drawdowns and/or peat subsidence, as well as large CO2 emissions (medium evidence, high agreement).
43   Climate change is compounding these impacts (medium evidence, medium agreement). For example, in
44   Indonesia, the highest emissions from drained tropical peatlands were reported in the extremely dry year of
45   the 1997 El Niño (810-2570 TgC yr-1) (Page et al., 2002) and the 2015 fire season (380 TgC yr-1) (Field et
46   al., 2016). These prolonged dry seasons have also led to tree die-offs and fires, which are relatively new
47   phenomena in these latitudes (medium evidence, high agreement) (Cole et al., 2015; Mezbahuddin et al.,
48   2015; Fanin and van der Werf, 2017; Taufik et al., 2017; Cole et al., 2019). Low soil moisture contributes to
49   increased fire propagation (see Cross-Chapter Box 5, and Section; Dadap et al., 2019), causing
50   long-lasting fires responsible for smoke and haze pollution (robust evidence, high agreement) (Ballhorn et
51   al., 2009; Page et al., 2009; Gaveau et al., 2014; Huijnen et al., 2016; Page and Hooijer, 2016; Hu et al.,
52   2018; Vadrevu et al., 2019; Niwa et al., 2021). Increases in fires and smoke lead to habitat loss and
53   negatively impact regional faunal populations (limited evidence, high agreement) (Neoh et al., 2015; Erb et
54   al., 2018b; Thornton et al., 2018).
56   In large lowland tropical peatland basins that are less impacted by anthropogenic activities (i.e., Amazon and
57   Congo river basins), the direct impact of climate change is that of a decreased carbon sink (limited evidence,

     Do Not Cite, Quote or Distribute                      2-50                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   medium agreement) (Roucoux et al., 2013; Gallego-Sala et al., 2018; Wang et al., 2018a; Dargie et al., 2019;
 2   Ribeiro et al., 2021). As for the temperate and boreal regions, climatic drying also tends to promote peat
 3   oxidation and carbon loss to the atmosphere (medium evidence, medium agreement) (section
 4   (Helbig et al., 2020; Zhang et al., 2020). In Europe, increasing mean annual temperatures in the Baltic,
 5   Scandinavia, and Continental Europe (Section have led to widespread lowering of peatland water
 6   tables at intact sites (Swindles et al., 2019), Sphagnum moss desiccation and die off (Bragazza, 2008; Lees et
 7   al., 2019), and increased fire intensity and frequency resulting in rapid carbon loss (Davies et al., 2013;
 8   Veraverbeke et al., 2021). Nevertheless, longer growing seasons and warmer, wetter climates have increased
 9   carbon accumulation and promoted thick deposits regionally, as reported for some North American sites
10   (limited evidence, medium agreement) (Cai and Yu, 2011; Shiller et al., 2014; Ott and Chimner, 2016).
12   In high-latitude peatlands, the net effect of climate change on the permafrost peatland carbon sink capacity
13   remains uncertain (Abbott et al., 2016; McGuire et al., 2018b; Laamrani et al., 2020; Loisel et al., 2021; Sim
14   et al., 2021; Väliranta et al., 2021). Increasing air temperatures have been linked to permafrost degradation
15   and altered hydrological regimes (Section, Figure 2.4a,, and Box 5.1), which have led to rapid
16   changes in plant communities and biogeochemical cycling (robust evidence, high agreement) (Liljedahl et
17   al., 2016; Swindles et al., 2016; Voigt et al., 2017; Zhang et al., 2017b; Voigt et al., 2020; Sim et al., 2021).
18   In many instances, permafrost degradation triggers thermokarst land subsidence associated with local
19   wetting (robust evidence, high agreement) (Jones et al., 2013; Borge et al., 2017; Olvmo et al., 2020;
20   Olefeldt et al., 2021). Permafrost thaw in peatland-rich landscapes can also cause local drying through
21   increased hydrological connectivity and runoff (Connon et al., 2014). In the first decades following thaw,
22   increases in methane, CO2, and nitrous oxide emissions have been recorded from peatland sites, depending
23   on surface moisture conditions (Schuur et al., 2009; O’Donnell et al., 2012; Elberling et al., 2013; Matveev
24   et al., 2016; Euskirchen et al., 2020; Hugelius et al., 2020). Conversely, some evidence suggests increased
25   peat accumulation after thaw (Jones et al., 2013; Estop-Aragonés et al., 2018; Väliranta et al., 2021). There
26   is also a need to consider the impact of wildfire on permafrost thaw, due to its effect on soil temperature
27   regime (Gibson et al., 2018), wildfire as a}, as fire intensity and frequency have increased across the boreal
28   and Arctic biomes (limited evidence, high agreement) (Kasischke et al., 2010; Scholten et al., 2021).
30   Unfortunately, the CO2 emissions from degrading peatlands is contributing to climate change in a positive
31   feedback loop (robust evidence, high agreement). In the midlatitudes, widespread anthropogenic disturbance
32   led to large historical GHG emissions and current legacy emissions of 0.15 PgC yr-1 between 1990 and 2000
33   (limited evidence, high agreement) (Maljanen et al., 2010; Tiemeyer et al., 2016; Drexler et al., 2018; Qiu et
34   al., 2021). About 80 million ha of peatlands have been converted to agriculture, equivalent to 72 PgC
35   emissions between 850–2010 CE (Leifeld et al., 2019; Qiu et al., 2021). In southeast Asia, an estimated 20–
36   25 Mha of peatlands have been converted to agriculture with carbon currently being lost at a rate of
37   ~155 ± 30 MtC yr−1 (Miettinen et al., 2016; Leifeld et al., 2019; Hoyt et al., 2020). Extensive deforestation
38   and drainage have caused widespread peat subsidence and large CO2 emissions at a current average of ~10 ±
39   2 t ha-1 yr-1 (excluding fires, (Hoyt et al., 2020)), with values estimated from point subsidence measurements
40   being as high as 30–90 t CO2 ha−1 yr−1 locally (robust evidence, high agreement) (Wösten et al., 1997;
41   Matysek et al., 2018; Swails et al., 2018; Evans et al., 2019; Conchedda and Tubiello, 2020; Anshari et al.,
42   2021). On balance, at the global scale, increases in GHG emissions from peatlands have primarily come
43   from the compounded effects of land-use change, drought, and fire, with emissions from some thawing
44   permafrost peatlands (robust evidence, high agreement).
46   Observed Changes in Polar Tundra
48   Warming at high latitudes, documented in both AR4 and AR5, is leading to earlier snow and sea ice melt and
49   longer growing seasons (WGI AR6) which are continuing to alter tundra plant communities (medium
50   evidence, high agreement) (Post et al., 2009; Gauthier et al., 2013). Woody encroachment and increases in
51   vegetation productivity observed in both AR4 and AR5 are widespread and continuing. Both experiments
52   and monitoring indicate that climate warming is causing increases in shrub, grass and sedge abundance,
53   density, frequency and height, with decreases in mosses and/or lichens (robust evidence, high agreement)
54   (Myers-Smith et al., 2011; Bjorkman et al., 2018; Bjorkman et al., 2019). Shrub growth is climate-sensitive
55   and greater in years with warmer growing seasons (Myers-Smith et al., 2015). Plant species that prefer
56   warmer conditions are increasing (Elmendorf et al., 2015; Bjorkman et al., 2018), plant cover is increasing
57   and bare ground is decreasing in long-term monitoring plots (Bjorkman et al., 2019; Myers-Smith et al.,

     Do Not Cite, Quote or Distribute                      2-51                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   2019). Animals such as moose, beavers and songbirds may already be responding to these vegetation
 2   changes by expanding their ranges northward or upslope into shrub tundra (Boelman et al., 2015; Tape et al.,
 3   2016a; Tape et al., 2016b; Tape et al., 2018).
 5   In addition to direct warming, indirect effects of climate change such as thawed permafrost, altered
 6   hydrology and enhanced nutrient cycling (as observed in AR4 and AR5) continue and are causing
 7   pronounced vegetation changes (medium evidence, medium agreement) (Schuur et al., 2009; Natali et al.,
 8   2012). Soil moisture status influences temperature sensitivity of plant growth and canopy heights (Myers-
 9   Smith et al., 2015; Ackerman et al., 2017; Bjorkman et al., 2018). In tundra ecosystems permafrost thawing
10   can decouple below-ground plant growth dynamics from above-ground dynamics, with below-ground root
11   growth continuing until soils refreeze in autumn ( Cross-Chapter Paper 6; Iversen et al., 2015; Blume-Werry
12   et al., 2016; Radville et al., 2016).
14   2.4.4   Observed Changes in Ecosystem Processes and Services
16 Observed Browning of Rivers and Lakes
18   In boreal coniferous areas there has been an increase in terrestrial derived dissolved organic carbon (DOC)
19   transport into rivers and lakes, which has caused increased opacity and shift toward a brown colour
20   (browning). This process was not given much attention in AR5 even though it is a consequence of climate
21   change: hydrological intensification, greening of the Northern Hemisphere, and degradation of carbon sinks
22   in peatlands (robust evidence, high agreement) (Solomon et al., 2015; Catalán et al., 2016; Crowther et al.,
23   2016; de Wit et al., 2016; Finstad et al., 2016; Creed et al., 2018; Hayden et al., 2019) factors that enhance
24   terrestrial productivity, alter vegetation communities, and affect the hydrological control on production and
25   transport of DOC (Weyhenmeyer et al., 2016). Non climate-related drivers of browning are: declining
26   atmospheric sulphur deposition, forestry practices and land-use changes (see Table 2.S.1 for detail).
28   Browning creates a positive feedback by absorbing photosynthetically active radiation accelerating upper
29   water (epilimnetic) warming (Solomon et al., 2015). Browning of lakes leads to shallower and more stable
30   thermoclines and thus, overall deep water cooling (Solomon et al., 2015; Williamson et al., 2015) and can
31   provoke a transition of the seasonal mixing regime from a mixed lake (polymictic) to one that is seasonally
32   stratified (Kirillin and Shatwell, 2016).
34   The ecological responses of browning need to be considered as concomitant effects of climate change and
35   nutrient status. Results from long-term, large-scale lake experiments were variable, showing both strong
36   synergistic effects (Urrutia-Cordero et al., 2016) and no significant effects of browning on plankton
37   community food webs (Rasconi et al., 2015). Browning has driven a shift from auto- to
38   heterotrophic/mixotrophic-based production (Wilken et al., 2013; Urrutia-Cordero et al., 2017) and supports
39   heterotrophic metabolism of the bacterial community (Zwart et al., 2016). Browning may also accelerate
40   primary production through input of nutrients associated with DOM in nutrient poor lakes and increase
41   cyanobacteria, which better cope with low light intensities (Huisman et al., 2018) and toxin levels (Urrutia-
42   Cordero et al., 2016). However, the synergistic impacts of browning and climate change on aquatic
43   communities depends on regional precipitation patterns (Weyhenmeyer et al., 2016), watershed type (de Wit
44   et al., 2016), and food chain length (Hansson et al., 2013). Quantitative attribution of browning to climate
45   change remains difficult (medium evidence, medium confidence).
47   In summary, new studies since AR5 have explicitly estimated the effects of warming and browning on
48   freshwaters in boreal areas with complex positive and negative repercussions on water temperature profiles
49   (lower vs upper water) (high confidence) and primary production (medium confidence).

     Do Not Cite, Quote or Distribute                     2-52                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 2   Figure 2.5: Large scale observed changes in freshwater ecosystems attributed to climate change over more than
 3   four decades. For description and references, see 2.3.3, 2.4.2, and
 6   Observed Changes in Wildfire
 8 Detection and attribution of observed changes in wildfire
 9   Wildfire is a natural and essential component of many forest and other terrestrial ecosystems. Excessive
10   wildfire, however, can kill people, cause respiratory disease, destroy houses, emit carbon dioxide, and
11   damage ecosystem integrity (see Section, Anthropogenic climate change increases wildfire
12   by exacerbating its three principal driving factors—heat, fuel, and ignition (Moritz et al., 2012; Jolly et al.,
13   2015). Non-climatic factors also contribute to wildfires—in tropical areas fires are set intentionally to clear
14   forest for agricultural fields and livestock pastures (Bowman et al., 2020b). Urban areas and roads create
15   ignition hazards. Governments in many temperate zone countries implement policies to suppress fires, even
16   natural ones, producing unnatural accumulations of fuel in the form of coarse woody debris and high
17   densities of small trees (Ruffault and Mouillot, 2015; Hessburg et al., 2016; Andela et al., 2017; Balch et al.,
18   2017; Lasslop and Kloster, 2017; Aragao et al., 2018, (Kelley et al., 2019). Globally, 4.2 million km2 of land
19   per year burned on average from 2002 to 2016 (Giglio et al., 2018) with the highest fire frequencies in the
20   Amazon rainforest, deciduous forests and savannas in Africa, and deciduous forests in northern Australia
21   (Earl and Simmonds, 2018; Andela et al., 2019).
23   Since the IPCC Fifth Assessment Report and the IPCC Special Report on Land, published research has
24   detected increases in the area burned by wildfire, analysed relative contributions of climate and non-climate
25   factors, and attributed burned area increases above natural levels to anthropogenic climate change in one part
26   of the world – western North America (robust evidence, high agreement) (Abatzoglou and Williams, 2016;
27   Partain et al., 2016; Kirchmeier‐Young et al., 2019; Mansuy et al., 2019; Bowman et al., 2020b). Across the
28   western United States, increases in vegetation aridity due to higher temperatures from anthropogenic climate
29   change doubled burned area from 1984 to 2015 over what would have burned due to non-climate factors,
30   including unnatural fuel accumulation from fire suppression, with the burned area attributed to climate
31   change accounting for 49% (32-76%, 95% confidence interval) of cumulative burned area (Abatzoglou and
32   Williams, 2016). Anthropogenic climate change has doubled the severity of a southwest North American
33   drought from 2000 to 2020 that has reduced soil moisture to its lowest levels since the 1500s (Williams et
34   al., 2020), driving half of the increase in burned area (Abatzoglou and Williams, 2016; Holden et al., 2018;
35   Williams et al., 2019). In British Columbia, Canada, the increased maximum temperatures due to
36   anthropogenic climate change increased burned area in 2017 to its highest extent in the 1950-2017 record,
37   seven to eleven times the area that would have burned without climate change (Kirchmeier-Young et al.,
38   2019). In Alaska, USA, the high maximum temperatures and extremely low relative humidity due to
39   anthropogenic climate change accounted for 33‒60% of the probability of wildfire in 2015, when the area
40   burned was the second highest in the 1940-2015 record (Partain et al., 2016). In protected areas of Canada
41   and the United States, climate factors (temperature, precipitation, relative humidity, evapotranspiration)

     Do Not Cite, Quote or Distribute                      2-53                                      Total pages: 237
     FINAL DRAFT                                      Chapter 2                    IPCC WGII Sixth Assessment Report

 1   accounted for 60% of burned area from local human and natural ignitions from 1984 to 2014, outweighing
 2   local human factors (population density, roads, and built area) (Mansuy et al., 2019).
 4   In summary, field evidence shows that anthropogenic climate change has increased the area burned by
 5   wildfire above natural levels across western North America in the period 1984-2017, at global mean surface
 6   temperature increases of 0.6ºC -0.9ºC, increasing burned area up to 11 times in one extreme year and
 7   doubling burned area over natural levels in a 32 year period (high confidence).
 9 Observed changes in wildfire globally
10   For global terrestrial area as a whole, wildfire trends vary depending on the time period of analysis. From
11   1900 to 2000, global average fire frequency, based on field data, increased 0.4% but the change was not
12   statistically significant, (Mouillot and Field, 2005; Gonzalez et al., 2010b). Fire frequency increased on one-
13   third of global land, mainly from burning for agricultural clearing in Africa, Asia, and South America,
14   slightly less than the area of fire frequency decrease, mainly from fire suppression across Australia, North
15   America, and Russia (Gonzalez et al., 2010b). Analyses of the Global Fire Emissions Database document
16   that from 1996 to 2015, global burned area decreased at a rate of –0.7% y-1 (Forkel et al., 2019) but the
17   change was not statistically significant, (Giglio et al., 2013). From 1998 to 2015, global burned area
18   decreased at a rate of −1.4 ± 0.5% y-1 (Giglio et al., 2013; Andela et al., 2017). The area of fire increases was
19   a third of the area of decreases, due to reduction of vegetation cover from agricultural expansion and
20   intensification (Andela et al., 2017) and increased precipitation (Forkel et al., 2019). Furthermore, much of
21   the decreasing trend derives from two years: 1998 with high burned area and 2013 with low burned area
22   (Forkel et al., 2019). Wildfire does not show a clear long-term trend for the world as a whole because of
23   increases and decreases in different regions (medium evidence, medium agreement).
25   Where global average burned area has decreased in the past two decades, higher correlations of rates of
26   change in burning to human population density, cropland area, and livestock density than to precipitation
27   indicate that agricultural expansion and intensification were main causes (Andela et al., 2017). The global
28   decrease of fire frequency from 2000 to 2010 is correlated to increasing human population density (Knorr et
29   al., 2014). The fire-reduction effect of reduced vegetation cover following expansion of agriculture and
30   livestock herding can counteract the fire-increasing effect of increased heat of climate change (Lasslop and
31   Kloster, 2017; Arora and Melton, 2018; Forkel et al., 2019). The reduction of burning needed after the initial
32   clearing for agricultural expansion drives much of the decline in fire in the tropics (Andela et al., 2017; Earl
33   and Simmonds, 2018; Forkel et al., 2019). The human influence on fire ignition can be seen through the
34   decrease documented on holy days (Sundays and Fridays), traditional religious days of rest (Earl et al.,
35   2015). Overall, human land use exerts an influence on wildfire trends for global terrestrial area as a whole
36   that can be stronger than climate change (medium confidence).
38 Observed changes in wildfire in individual regions
39   While burned area has increased in parts of Asia, Australia, Europe, and South America, published research
40   has not yet attributed the increases to anthropogenic climate change (medium evidence, high agreement).
42   In the Amazon, deforestation for agricultural expansion and the degradation of forests adjacent to deforested
43   areas cause wildfire in moist humid tropical forests not adapted to fire (robust evidence, high agreement)
44   (Fonseca et al., 2017; van Marle et al., 2017; da Silva et al., 2018; da Silva et al., 2021; dos Reis et al., 2021;
45   Libonati et al., 2021). Roads facilitate deforestation, fragmenting the rainforest and increasing the dryness
46   and flammability of vegetation (Alencar et al., 2015). Extreme droughts that occur during warm phases of
47   the El Niño-Southern Oscillation (ENSO) and the Atlantic Multidecadal Oscillation combine with the
48   degradation of vegetation to cause extreme fire events (robust evidence, high agreement) (Fonseca et al.,
49   2017; Aragao et al., 2018; da Silva et al., 2018; Burton et al., 2020; dos Reis et al., 2021; Libonati et al.,
50   2021). In the State of Roraima, Brazil, distance to roads, infrastructure that enables deforestation, and ENSO
51   were the two factors most explaining fire occurrence in the extreme 2015-2016 fire season (Fonseca et al.,
52   2017). From 1973 to 2014, burned area increased in the Amazon, coinciding with increased deforestation
53   (van Marle et al., 2017). In the State of Acre, Brazil, burned area increased 36-fold from 1984 to 2016, with
54   43% burned area in agricultural and livestock settlement areas (da Silva et al., 2018). In 2019, the extreme
55   fire year 2019, 85% of the area burned in the Amazon occurred in areas deforested in 2018 (Cardil et al.,
56   2020). Even though relatively higher moisture in 2019 led to burning below the 2002-2019 average across
57   most of South America, burning in areas of recent deforestation in the Amazon were above the 2002-2019

     Do Not Cite, Quote or Distribute                       2-54                                        Total pages: 237
     FINAL DRAFT                                    Chapter 2                   IPCC WGII Sixth Assessment Report

 1   average, indicating that deforestation, not meteorological conditions, triggered the 2019 fires (Kelley et al.,
 2   2021; Libonati et al., 2021). Furthermore, from 1981 to 2018, deforestation in the Amazon reduced moisture
 3   inputs to the lower atmosphere, increasing drought and fire in a self-reinforcing feedback (Xu et al., 2020).
 4   In the Amazon, deforestation exerts an influence on wildfire that can be stronger than climate change (robust
 5   evidence, high agreement).
 7   In Australia, burned area increased significantly between the periods 1950-2002 and 2003-2020 in the
 8   southeast state of Victoria, with the area burned in the 2019-2020 bushfires the highest on record
 9   (Lindenmayer and Taylor, 2020). In addition to the deaths of dozens of people and destruction of thousands
10   of houses, the 2019-2020 Australia bushfires burned almost half of the area protected for conservation in
11   Victoria and two-thirds of forests allocated for timber harvesting (Lindenmayer and Taylor, 2020), wildlife,
12   and extensive areas of habitat for threatened plant and animal species (Geary et al., 2021). Generally, past
13   timber harvesting did not lead to more severe fire canopy damage (Bowman et al., 2021). Across
14   Southeastern Australia, the fraction of vegetated area that burned increased significantly in eight of the 32
15   bioregions from 1975 to 2009 but decreased significantly in three bioregions (Bradstock et al., 2014).
16   Increases in four bioregions were correlated to increasing temperature and decreasing precipitation.
17   Decreases in burned area occurred despite increased temperature and decreased precipitation. Analyses of
18   climate across Australia from 1950 to 2017 (Dowdy, 2018; Harris and Lucas, 2019) and during periods with
19   extensive fires in 2017 in eastern Australia (Hope et al., 2019), in 2018 in Northeastern Australia (Lewis et
20   al., 2020), and in period 2019-2020 in Southeastern Australia (Abram et al., 2021; van Oldenborgh et al.,
21   2021) indicate that temperature and drought extremes due to the El Niño-Southern Oscillation, Southern
22   Annular Mode, and other natural interdecadal cycles drive interannual variability of fire weather. While the
23   effects of interdecadal climate cycles on fire are superimposed on long-term climate change, the relative
24   importance of anthropogenic climate change in explaining changes in burned area in Australia remains
25   unquantified (medium evidence, high agreement).
27   In Africa, the rate of change of burned area for the continent as a whole ranged from a non-statistically
28   significant –0.45% y-1 from 2002 to 2016 (Zubkova et al., 2019) to a significant –1.9% y-1 from 2001 to
29   2016 (Wei et al., 2020). Burned area decreases coincided with areas of agricultural expansion or areas where
30   drought reduced fuel loads (Zubkova et al., 2019; Wei et al., 2020). It is possible, however, that the 500 m
31   spatial resolution of Modis remote sensing fire data underestimates burned area in Africa by half by missing
32   small fires (Ramo et al., 2021). In the Serengeti-Mara savanna of east Africa, burned area showed no
33   significant change from 2001 to 2014, although an increase in domestic livestock would tend to reduce the
34   grass cover that fuels savanna fires (Probert et al., 2019).
36   In Mediterranean Europe, burned area for the region as a whole decreased from 1985 to 2011 (Turco et al.,
37   2016), although burned area for Spain did not show a significant long-term increase from 1968 to 2010
38   (Moreno et al., 2014) while burned area for Portugal in 2017 was the highest in the period 1980-2017
39   (Turco et al., 2019). Increased summer maximum temperature and decreased soil moisture explained most of
40   observed burned area, suggesting a contribution of climate change, but fire suppression, fire prevention,
41   agricultural abandonment, and reforestation, and reduction of forest area exerted even stronger influences on
42   burned area than climate across Mediterranean Europe (robust evidence, high agreement) (Moreno et al.,
43   2014; Turco et al., 2017; Viedma et al., 2018; Turco et al., 2019).
45   In the Arctic tundra and boreal forest, where wildfire has naturally been infrequent, burned area showed
46   statistically significant increases of ~50% y-1 across Siberia, Russia, from 1996 to 2015 (Ponomarev et al.,
47   2016) and 2% y-1 across Canada from 1959 to 2015 (Hanes et al., 2019). Wildfire burned ~ 6% of the area of
48   four extensive Arctic permafrost regions in Alaska, USA, eastern Canada, and Siberia from 1999 to 2014
49   (Nitze et al., 2018). In boreal forest in the Northwest Territories, Canada, and Alaska, USA, the area burned
50   by wildfire increased at a statistically significant rate of 6.8% y-1 in the period 1975-2015, (Veraverbeke et
51   al., 2017), with smouldering belowground fires that lasted through the winter covering ~1% of burned area
52   in the period 2002-2016 (Scholten et al., 2021). While burned area was correlated to temperature and
53   reduced precipitation in Siberia (Ponomarev et al., 2016; Masrur et al., 2018) and to lightning, correlated
54   with temperature and precipitation in the Northwest Territories and Alaska (Veraverbeke et al., 2017), no
55   attribution analyses have examined relative influences of climate and non-climate factors.

     Do Not Cite, Quote or Distribute                     2-55                                      Total pages: 237
     FINAL DRAFT                                    Chapter 2                   IPCC WGII Sixth Assessment Report

 1   In Indonesia, deforestation and draining of peat swamp forests dries out the peat, providing substantial fuel
 2   for fires (Page and Hooijer, 2016). Extreme fire years in Indonesia, including 1997, 2006, and 2015, coincide
 3   with extreme heat and aridity during the warm phase of the El Niño-Southern Oscillation (Field et al., 2016).
 4   Fire-resistant forest in 2019 covered only 3% of peatlands and 4.5% of non-peatlands on Sumatra and
 5   Kalimantan (Nikonovas et al., 2020).
 7   In Chile, burned area in the summer of 2016–2017 was 14 times the mean for the period 1985–2016 and the
 8   highest on record (Bowman et al., 2019). While that extreme fire year coincided with the highest daily mean
 9   maximum temperature in the period 1979–2017 (Bowman et al., 2019), in central Chile, the area of highest
10   fire activity, burned area from 1976 to 2013 showed highest correlation to precipitation cycles of the El
11   Niño-Southern Oscillation and temperature cycles of the Antarctic Oscillation (Urrutia-Jalabert et al., 2018).
13   Overall, burned area has increased in the Amazon, the Arctic, Australia, and parts of Africa and Asia,
14   consistent with, but not formally attributed to anthropogenic climate change (medium evidence, high
15   agreement). Deforestation, peat draining, agricultural expansion or abandonment, fire suppression, and
16   interdecadal cycles such as the El Niño-Southern Oscillation exert a stronger influence than climate change
17   on wildfire trends in numerous regions outside of North America (high confidence).
19 Observed changes in fire seasons globally
20   IPCC AR6 Working Group 1, Chapter 12, has assessed fire weather, while this chapter assesses burned area
21   and fire frequency. The global increases in temperature of anthropogenic climate change have increased
22   aridity and drought, lengthening the fire weather season (annual period with a heat and aridity index greater
23   than half of its annual range) on one-quarter of global vegetated area and increasing average fire season
24   length by one-fifth, from 1979 to 2013 (Jolly et al., 2015). Climate change has contributed to increases in the
25   fire weather season or the probability of fire weather conditions in the Amazon (Jolly et al., 2015), Australia
26   (Dowdy, 2018; Abram et al., 2021; van Oldenborgh et al., 2021), Canada (Hanes et al., 2019), central Asia
27   (Jolly et al., 2015), East Africa (Jolly et al., 2015), and North America (Jain et al., 2017; Williams et al.,
28   2019; Goss et al., 2020). In forest areas, the burned area is correlated to fuel aridity, a function of
29   temperature; in non-forest areas, the burned area is correlated to high precipitation in the previous year,
30   which can produce high grass fuel loads (Abatzoglou et al., 2018). Fire use in agriculture and livestock
31   raising or other factors have generated a second fire season on approximately one-quarter of global land
32   where fire is present, despite sub-optimal fire weather in the second fire season (Benali et al., 2017). In
33   summary, anthropogenic climate change, through a 0.9ºC surface temperature increase since the pre-
34   industrial period, has lengthened or increased the frequency of periods with heat and aridity that favour
35   wildfire on up to one-quarter of vegetated area, since 1979 (robust evidence, high agreement).
37 Observed changes in post-fire vegetation
38   Globally, fire has contributed to biome shifts (Section and tree mortality (Section,
39   attributed to anthropogenic climate change. Research since the IPCC Fifth Assessment Report has also found
40   vegetation changes from wildfire due to climate change. Through increased temperature and aridity,
41   anthropogenic climate change has driven post-fire changes in plant regeneration and species composition in
42   South Africa (Slingsby et al., 2017) and tree regeneration in the western United States (Davis et al., 2019b).
43   In the Fynbos vegetation of the Cape Floristic Region, South Africa, post-fire heat and drought and legacy
44   effects of exotic plant species reduced native plant species regeneration, decreasing species richness 12%
45   from 1966 to 2010 and shifting the average temperature tolerance of species upward by 0.5ºC (Slingsby et
46   al., 2017). In burned areas across the western United States, the increasing heat and aridity of anthropogenic
47   climate change from 1979 to 2015 pushed low-elevation ponderosa pine (Pinus ponderosa) and Douglas-fir
48   (Pseudotsuga menziesii) forests across critical thresholds of heat and aridity that reduced post-fire tree
49   regeneration by half (Davis et al., 2019b). In the Southwestern United States of America, where
50   anthropogenic climate change has caused drought (Williams et al., 2019) and increased wildfire (Abatzoglou
51   and Williams, 2016), high-severity fires have converted some forest patches to shrublands (Barton and
52   Poulos, 2018). Field evidence shows that anthropogenic climate change and wildfire together have altered
53   vegetation species composition in the Southwestern USA and in the Cape floristic region, South Africa,
54   reducing post-fire natural regeneration and species richness of tree and other plant species, between 1966
55   and 2015, at global mean surface temperature increases of 0.3-0.9ºC (medium evidence, high agreement).

     Do Not Cite, Quote or Distribute                     2-56                                      Total pages: 237
     FINAL DRAFT                                       Chapter 2                    IPCC WGII Sixth Assessment Report

 1   [START FAQ2.3 HERE]
 3   FAQ2.3: Is climate change increasing wildfire?
 5   In the Amazon, Australia, North America, Siberia, and other regions, wildfires are burning wider areas than
 6   in the past. Analyses show that human-caused climate change has driven the increases in burned area in the
 7   forests of western North America. Elsewhere, deforestation, fire suppression, agricultural burning, and
 8   short-term cycles like El Niño can exert a stronger influence than climate change. Many forests and
 9   grasslands naturally require fire for ecosystem health but excessive wildfire can kill people, destroy homes,
10   and damage ecosystems.

14   Figure FAQ2.3.1: (a) Springs Fire, May 2, 2013, Thousand Oaks, California, USA (photo by Michael Robinson
15   Chávez, Los Angeles Times). (b) Cumulative area burned by wildfire in the western U.S., with (orange) and without
16   (yellow) the increased heat and aridity of climate change (Abatzoglou and Williams, 2016).
18   Wildfire is a natural and essential part of many forest, woodland, and grassland ecosystems, killing pests,
19   releasing plant seeds to sprout, thinning out small trees, and serving other functions essential for ecosystem
20   health. Excessive wildfire, however, can kill people, cause breathing illnesses from the smoke, destroy
21   homes (Figure FAQ2.1a), and damage ecosystems.
23   Human-caused climate change increases wildfire by intensifying its principal driving factor – heat. The heat
24   of climate change dries out vegetation and accelerates burning. Non-climate factors also cause wildfires.
25   Agricultural companies, small farmers, and livestock herders in many tropical areas cut down forests and
26   intentionally set fires to clear fields and pastures. Cities, towns, and roads increase the number of fires that
27   people ignite. Governments in many countries suppress fires, even natural ones, producing unnatural
28   accumulations of fuel in the form of coarse woody debris and dense stands of small trees. The fuel
29   accumulations cause particularly severe fires that burn into tree crowns.
31   Evidence shows that human-caused climate change has driven increases in the area burned by wildfire in the
32   forests of western North America. Across the western U.S., the higher temperatures of human-caused
33   climate change doubled burned area from 1984 to 2015, compared with what would have burned without
34   climate change (Figure FAQ2.1b). The additional area burned, 4.9 million hectares, is greater than the land
35   area of Switzerland. In this region, human-caused climate change has driven a drought from 2000 to 2020
36   that is the most severe since the 1500s, severely increasing the aridity of vegetation. In British Columbia,
37   Canada, the higher maximum temperatures of human-caused climate change increased burned area in 2017
38   to its widest extent in the 1950-2017 record, seven to eleven times the area that would have burned without
39   climate change. Moreover, in national parks and other protected areas of Canada and the U.S., climate
40   factors explained the majority of burned area from 1984 to 2014, with climate factors (temperature, rainfall,
41   aridity) outweighing local human factors (population density, roads, and urban area).
43   In other regions, wildfires are also burning wider areas and occurring more often. This is consistent with
44   climate change but analyses have not yet shown if climate change is more important than other factors. In the
45   Amazon, deforestation by companies, farmers, and herders who cut down and intentionally burn rainforests
46   to expand agricultural fields and pastures causes wildfires even in relatively moister years. Drought
47   exacerbates these fires. In Australia, much of the southeastern part of the continent has experienced extreme

     Do Not Cite, Quote or Distribute                        2-57                                        Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   wildfire years, but analyses suggest that El Niño, a heat phenomenon that cycles up and down periodically, is
 2   more important than long-term climate change. In Indonesia, intentional burning of rainforests for oil palm
 3   plantations and El Niño seem to be more important than long-term climate change. In Mediterranean Europe,
 4   fire suppression seems to have prevented any increasing trend in burned area but suppression and
 5   abandonment of agricultural lands have allowed fuel to build up in some areas and contribute to major fires
 6   in years of extreme heat. In Canada and Siberia, wildfires are now burning more often in permafrost areas
 7   where fire had been rare, but analyses are lacking on the relative influence of climate change. For the world
 8   as a whole, satellite data indicate that the vast amount of land that converted from forest to farmland from
 9   1998 to 2015 actually decreased total burned area. Nevertheless, the evidence from the forests of western
10   North America shows that human-caused climate change has, on one continent, clearly driven increases in
11   wildfire.
13   [END FAQ2.3 HERE]-
16   Observed Changes in Tree Mortality
18 Observed tree mortality globally
19   Anthropogenic climate change can cause tree mortality directly through increased aridity or drought (Section
20 or indirectly through wildfire (Section and insect pests (Section Catastrophic
21   failure of the plant hydraulic system, in which a lack of water causes the xylem to lose hydraulic
22   conductance, is the principal mechanism of drought-induced tree death (Anderegg et al., 2016; Adams et al.,
23   2017; Anderegg et al., 2018; Choat et al., 2018; Menezes-Silva et al., 2019; Brodribb et al., 2020).
25   Up through the IPCC Fifth Assessment Report (Settele et al., 2014), detection and attribution analyses had
26   found that anthropogenic climate change, with global temperature increases of 0.3°-0.9ºC above the pre-
27   industrial period and increases in aridity exceeding the effects of local non-climate change factors, caused
28   three cases of drought-induced tree mortality of up to 20% in the period 1945-2007, in western North
29   America (van Mantgem et al., 2009), the African Sahel (Gonzalez et al., 2012), and North Africa (le Polain
30   de Waroux and Lambin, 2012). Increased wildfire and pest infestations, driven by climate change, also
31   contributed to the North American tree mortality (van Mantgem et al., 2009). In addition, a meta-analysis of
32   published cases found that drought consistent with, but not formally attributed, to climate change, had
33   caused tree mortality at 88 sites in boreal, temperate and tropical ecosystems (Allen et al., 2010), with 49
34   additional cases found by the IPCC Fifth Assessment report (Settele et al., 2014).
36   Since the IPCC Fifth Assessment Report (Settele et al., 2014), global meta-analyses have found at least 15
37   (Allen et al., 2015) and 25 (Hartmann et al., 2018) additional sites of drought-induced tree mortality around
38   the world. These and other global analyses found more rapid mortality than previously (Allen et al., 2015),
39   rising background mortality (Allen et al., 2015), mortality increasing with drought severity (Greenwood et
40   al., 2017), mortality of tropical trees increasing with temperature (Locosselli et al., 2020), mortality
41   increasing with tree size for many species (Bennett et al., 2015), mortality predominantly at the dry edge of
42   species ranges (Anderegg et al., 2019a), and three-fourths of drought-induced mortality cases leading to a
43   change in the dominant species (Batllori et al., 2020). Multiple non-climate factors contribute to tree
44   mortality, including timber cutting, livestock grazing, and air pollution (Martinez-Vilalta and Lloret, 2016).
45   Globally, tropical dry forests lost, from all causes, 95,000 km2, 8% of their total area, from 1982 to 2016, the
46   most extensive area of mortality of any biome (Song et al., 2018).
48   In summary, anthropogenic climate change has caused drought-induced tree mortality up to 20% in the
49   period 1945-2007 in western North America, the African Sahel, and North Africa, through global
50   temperature increases of 0.3°-0.9ºC above the pre-industrial period and increases in aridity, and contributed
51   to over 100 other cases of drought-induced tree mortality in Africa, Asia, Australia, Europe, and North and
52   South America (high confidence). Field observations document accelerating mortality rates, rising
53   background mortality, and post-mortality vegetation shifts (high confidence). Water stress, leading to plant
54   hydraulic failure, is the principal mechanism of drought-induced tree mortality. Timber cutting, agricultural
55   expansion, air pollution, and other non-climate factors also contribute to tree death.

     Do Not Cite, Quote or Distribute                      2-58                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1 Observed tree mortality in tropical ecosystems
 2   In the Brazilian Amazon, deforestation to clear agricultural land comprises the principal cause of tree
 3   mortality, reducing forest cover an average of 13,900 km2 y-1 from 1988 to 2020 (Espaciais, 2021). In
 4   addition, an annual average temperature increase of 1.2ºC from 1950 to 2018 (Marengo et al., 2018)
 5   contributed to mortality in a set of 310 Amazon field plots of ~40% from 1983 to 2011 (Brienen et al.,
 6   2015). In another set of plots, mortality among newly recruited trees of mesic genera increased and drought-
 7   tolerant genera became more abundant from 1985 to 2015 (Esquivel-Muelbert et al., 2019). In other plots,
 8   tree mortality did not show a statistically significant change from 1965 to 2016 but rose abruptly in severe
 9   drought years, mainly during warm phases of the El Niño-Southern Oscillation (ENSO) (Aleixo et al., 2019).
10   Nearly half the area of the Amazon has experienced extremely dry conditions during ENSO warm phases,
11   which can cause extensive wildfire (Section Wildfire can increase tree mortality rates by >600%
12   above rates in non-burned areas, with the higher mortality persisting up to a decade after a fire (Silva et al.,
13   2018; Berenguer et al., 2021). Climate change has contributed to tree mortality in Amazon rainforest
14   (medium evidence, medium agreement).
16   In the African Sahel field research has continued to detect tree mortality, ranging from 20% to 90% in the
17   period 1965-2018 (Kusserow, 2017; Trichon et al., 2018; Dendoncker et al., 2020), and declines in tree
18   biodiversity, with local losses of tree species up to 80% in the period 1970-2014 (Hanke et al., 2016;
19   Kusserow, 2017; Ibrahim et al., 2018; Dendoncker et al., 2020), consistent with, but not formally attributed
20   to climate change. In Algeria, mortality of Atlas cedar (Cedrus atlantica) increased from 1980 to 2006,
21   coinciding with a ~1ºC spring temperature increase, but non-climate factors were not examined (Navarro-
22   Cerrillo et al., 2019). Across southern Africa, nine of the 13 oldest known baobab trees (Adansonia digitata),
23   1100‒2500 years old, have died since 2005, although the causes are unknown (Patrut et al., 2018). In South
24   Africa, savanna trees experienced an order of magnitude increase in mortality, related, but not formally
25   attributed to decreased rainfall (Case et al., 2019). In Tunisia, insect infestations related, but not formally
26   attributed to, hotter temperatures led to mortality of cork oaks (Quercus suber) (Bellahirech et al., 2019).
28 Observed tree mortality in boreal and temperate ecosystems
29   The most extensive research on tree mortality since the IPCC Fifth Assessment Report has occurred in the
30   western United States, where anthropogenic climate change accounts for half the magnitude of a drought
31   from 2000 to 2020 that has been the most severe since the 1500s (Williams et al., 2020) and for one-tenth to
32   one-quarter of the magnitude of the 2012-2014 period of the severe 2012-2016 drought in California
33   (Williams et al., 2015a). Across the western United States, anthropogenic climate change doubled tree
34   mortality between 1955 and 2007 (van Mantgem et al., 2009). Lodgepole pine (Pinus contorta) mortality
35   increased 700% from 2000 to 2013 (Anderegg et al., 2015) and piñon pine (Pinus edulis) experienced over
36   50% mortality from 2002 to 2014 (Redmond et al., 2018). In California montane conifer forest,
37   anthropogenic climate change increased tree mortality one-quarter (Goulden and Bales, 2019). One-quarter
38   of trees died in some areas with mortality rates of ponderosa pine (Pinus ponderosa) and sugar pine (Pinus
39   lambertiana) increasing up to 700% of pre-drought rates (Stephenson et al., 2019; Stovall et al., 2019).
40   Substantial field evidence shows that anthropogenic climate change has caused extensive tree mortality in
41   North America (robust evidence, high agreement).
43   In western North America, increased infestations of bark beetles and other tree-feeding insects that benefit
44   from increased winter temperatures (IPCC AR6 WGI and longer growing seasons (IPCC AR6 WGI
45 have killed drought-stressed trees (Section; Anderegg et al., 2015; Kolb et al., 2016; Lloret
46   and Kitzberger, 2018; Redmond et al., 2018; Stephens et al., 2018; Fettig et al., 2019; Restaino et al., 2019;
47   Stephenson et al., 2019). Increasing temperatures have allowed bark beetles to move further north and higher
48   in elevation, survive through the winter at sites where they would previously have died, and reproduce more
49   often (Raffa et al., 2008; Bentz et al., 2010; Jewett et al., 2011; Macfarlane et al., 2013; Raffa et al., 2013;
50   Hart et al., 2017; Stephenson et al., 2019; Teshome et al., 2020; Koontz et al., 2021). Under warmer
51   conditions, some insects that were previously innocuous have become important agents of tree mortality
52   (Stephenson et al., 2019; Trugman et al., 2021). Field observations show mixed effects of bark beetle-
53   induced tree mortality on subsequent fire-caused tree mortality (Andrus et al., 2016; Meigs et al., 2016;
54   Candau et al., 2018; Lucash et al., 2018; Talucci and Krawchuk, 2019; Wayman and Safford, 2021). From
55   1997 to 2018, ~5% of western U.S. forest area died from bark beetle infestations (Hicke et al., 2020). In
56   most circumstances, trees that have been weakened by drought are more vulnerable to being killed by bark
57   beetles (Anderegg et al., 2015; Kolb et al., 2016; Lloret and Kitzberger, 2018; Redmond et al., 2018;

     Do Not Cite, Quote or Distribute                      2-59                                      Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1   Stephens et al., 2018; Fettig et al., 2019; Restaino et al., 2019; Stephenson et al., 2019; Koontz et al., 2021).
 2   Climate change has contributed to bark beetle infestations that have caused much of the tree mortality
 3   in North America (robust evidence, high agreement) (Section
 5   Across Europe, rates of tree mortality in field inventories from 2000 to 2012 were highest in Spain, Bulgaria,
 6   Sweden, and Finland, positively correlated to maximum winter temperature and inversely correlated to
 7   spring precipitation (Neumann et al., 2017). Tree mortality in Austria, the Czech Republic, Germany,
 8   Poland, Slovakia, and Switzerland doubled from 1984 to 2016, correlated to intensified logging and
 9   increased temperatures (Senf et al., 2018). Drought-related tree mortality rates from 1987 to 2016 were
10   highest in Ukraine, Moldova, southern France, and Spain (Senf et al., 2020). Climate contributed to tree
11   mortality across Europe from 1958 to 2001 (Seidl et al., 2011). In addition, insect infestations related to
12   higher temperatures (Okland et al., 2019) have caused extensive mortality of Norway spruce (Picea abies)
13   across nine European countries (Marini et al., 2017; Mezei et al., 2017). Across the Mediterranean Basin, a
14   combination of drought, wildfire, pest infestations, and livestock grazing has driven tree mortality (Penuelas
15   and Sardans, 2021). Climate change has contributed to tree mortality in Europe (high agreement, medium
16   confidence). (Section
18 Tree mortality and fauna
19   A global meta-analysis of 631 cases of bird and mammal abundance changes in areas of tree mortality found
20   increasing abundance in a set of 186 bird species with increasing mortality and no trend in mammal
21   abundance (Fleming et al., 2021). Ground-nesting, ground foraging, tree hole nesting, and bark foraging
22   increased most, while nectar-feeding and foliage-gleaning birds declined. Invertebrates, especially ground-
23   foraging predators and detritivores, decreased.
25   Observed Terrestrial Ecosystem Carbon
27 Observed terrestrial ecosystem carbon globally
28   Terrestrial ecosystems contain stocks of 450 Gt (380‒540 Gt) carbon in vegetation, 1700 Gt ± 250 Gt carbon
29   in soils, and 1400 Gt ± 200 Gt carbon in permafrost (Hugelius et al., 2014; Batjes, 2016; Jackson et al.,
30   2017; Strauss et al., 2017; Erb et al., 2018a; Xu et al., 2021). Ecosystem carbon stocks, totaling 3030–4090
31   GtC (from lowest and highest estimates above) substantially exceed the ~900 Gt carbon in unextracted fossil
32   fuels (see Chapter 5 of WGI).
34   Deforestation, draining of peatlands, expansion of agricultural fields, livestock pastures, and human
35   settlements, and other land use changes emitted carbon at a rate of 1.6 ± 0.7 Gt y-1 from 2010 to 2019,
36   (Friedlingstein et al., 2020), of which wildfires and peat burning emitted 0.4 ± 0.2 Gt y-1 from 1997 to 2016
37   (van der Werf et al., 2017). Anthropogenic climate change has caused a portion of these emissions through
38   increases in wildfire (Section and tree mortality (Section but the fraction of the total
39   remains unquantified. Land use change produced ~15% of global anthropogenic emissions, from fossil fuels
40   and land (Friedlingstein et al., 2020). Terrestrial ecosystems removed carbon from the atmosphere through
41   plant growth at a rate of -3.4 ± 0.9 Gt y-1 from 2010 to 2019 (Friedlingstein et al., 2020).
43   Tropical deforestation and draining and burning of peatlands produce almost all of the carbon emissions
44   from land use change (Houghton and Nassikas, 2017; Friedlingstein et al., 2020), while forest growth
45   accounts for two-thirds of ecosystem carbon removals from the atmosphere (Pugh et al., 2019b). Global
46   terrestrial ecosystems comprised a net sink of -1.9 ± 1.1 Gt y-1 from 2010 to 2019 (Friedlingstein et al.,
47   2020), mainly due to growth in forests (Harris et al., 2021; Xu et al., 2021), mitigating ~31% of global
48   emissions from fossil-fuel burning and land use change (Friedlingstein et al., 2020).
50   In summary, terrestrial ecosystems contain 3000-4000 Gt carbon in vegetation, permafrost, and soils, three
51   to five times the amount of carbon in unextracted fossil fuels, and 4.4 times the carbon currently in the
52   atmosphere (robust evidence, high agreement). Tropical deforestation, draining and burning of peatlands and
53   other land use changes emit 0.9-2.3 Gt y-1 of carbon, ~15% of global emissions from fossil fuel and
54   ecosystems (robust evidence, high agreement). Terrestrial ecosystems currently remove more carbon from
55   the atmosphere, 2.5-4.3 Gt y-1, than they emit, so tropical rainforests, Arctic permafrost, and other
56   ecosystems provide the global ecosystem service of naturally preventing carbon from contributing to climate
57   change (high confidence).

     Do Not Cite, Quote or Distribute                       2-60                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 2 Observed stocks in high-carbon terrestrial ecosystems
 3   The ecosystem that attains the highest aboveground carbon density in the world is coast redwood (Sequoia
 4   sempervirens) forest, in California, USA, with 2600 ± 100 t ha-1 carbon (Van Pelt et al., 2016). The
 5   ecosystem with the second highest documented carbon density in the world is mountain ash (Eucalyptus
 6   regnans) forest in Victoria, Australia, with ~1900 t ha-1 (Keith et al., 2009). Within the tropics, tropical
 7   evergreen broadleaf forests (rainforests) in the Amazon, the Congo, and Indonesia attain the highest carbon
 8   densities, reaching a maximum of 230 t ha-1 in the Amazon (Mitchard et al., 2014) and the Congo (Xu et al.,
 9   2017). Temperature increases reduce tropical rainforest aboveground carbon density 9.1 t ha-1 per degree
10   Celsius, through reduced growth and increased mortality (Sullivan et al., 2020).
12   Tropical forests contain the largest vegetation carbon stock in the world, with 180‒250 Gt of above- and
13   belowground carbon (Saatchi et al., 2011; Baccini et al., 2012; Avitabile et al., 2016). The Amazon contains
14   a carbon stock of 45‒60 Gt (Baccini et al., 2012; Mitchard et al., 2014; Englund et al., 2017).
16   Ecosystems with high soil carbon densities include peat bogs in Ireland, with up to 3000 t ha-1 (Tomlinson,
17   2005), Cuvette Centrale swamp forest peatlands in Congo, with an average of ~2200 t ha-1 (Dargie et al.,
18   2017), Arctic tundra, with an average of ~900 t ha-1 (Tarnocai et al., 2009), and mangrove peatlands in
19   Kalimantan, Indonesia, with an average of 850 ± 320 t ha-1 (Murdiyarso et al., 2015). Arctic permafrost
20   contains 1400 Gt ± 200 Gt to 3 m depth, the largest soil carbon stock in the world (Hugelius et al., 2014).
21   Globally, peatlands contain 470‒620 Gt carbon (Page et al., 2011; Hodgkins et al., 2018) of which boreal
22   and temperate peatlands contain 415 ± 150 Gt (Hugelius et al., 2020) and tropical peatlands 80‒350 Gt (Page
23   et al., 2011; Dargie et al., 2017; Gumbricht et al., 2017; Ribeiro et al., 2021). Other analyses increase the
24   upper estimates for boreal and temperate peatlands to 800-1200 Gt (Nichols and Peteet, 2019; Mishra et al.,
25   2021b).
27   Tropical forests and Arctic permafrost contain the highest ecosystem carbon stocks in aboveground
28   vegetation and soil, respectively, in the world (robust evidence, high agreement). These ecosystems form
29   natural sinks that prevent the emission to the atmosphere of 1400-1800 Gt carbon that would otherwise
30   increase the magnitude of climate change (high confidence).
32 Biodiversity and observed terrestrial ecosystem carbon
33   High biodiversity and ecosystem carbon generally occur together, with rainforests in the Amazon, the
34   Congo, and Indonesia containing the largest aboveground vegetation carbon stocks (Saatchi et al., 2011;
35   Baccini et al., 2012; Avitabile et al., 2016) and the highest vascular plant species richness (Kreft and Jetz,
36   2007) in the world. Aboveground ecosystem carbon and animal species richness show high correlation but
37   also high spatial variability (Strassburg et al., 2010). Aboveground carbon is correlated to genus richness
38   globally (Cavanaugh et al., 2014), but to species richness only in local areas (Poorter et al., 2015; Sullivan et
39   al., 2017). Species richness generally increases vegetation productivity in the humid tropics while tree
40   abundance increases productivity in drier conditions (Madrigal-Gonzalez et al., 2020). Across the Amazon,
41   ~1% of tree species contain 50% of the aboveground carbon, due to abundance and maximum height (Fauset
42   et al., 2015). Aboveground carbon in tropical forest shows positive correlations to vertebrate species richness
43   (probability values not reported) (Deere et al., 2018; Di Marco et al., 2018). In logged and burned tropical
44   forest in Brazil, species richness of plants, birds, and beetles increased with carbon density up to ~100 t ha-1
45   (Ferreira et al., 2018).
47   National parks and other protected areas, which, in June 2021, covered 15.7% of global terrestrial area
48   (UNEP-WCMC, 2021), contain ~90 Gt carbon in vegetation and ~150 Gt carbon in soil (one-fifth and one-
49   tenth, respectively, of global stocks) and remove carbon from the atmosphere at a rate of ~0.5 Gt y-1 (one-
50   sixth of global removals) (Melillo et al., 2016). The most strictly protected areas contain carbon at higher
51   densities, but illegal deforestation and fires in some protected areas emit 38 ± 17 Mt y-1 globally (Collins and
52   Mitchard, 2017). In the Amazon, protected areas store more than half of the aboveground vegetation carbon
53   stock of the region but account for only one-tenth of net emissions (Walker et al., 2020). Conservation of
54   high biodiversity areas, particularly in protected areas, protects ecosystem carbon, prevents emissions to the
55   atmosphere, and reduces the magnitude of climate change (high confidence).

     Do Not Cite, Quote or Distribute                      2-61                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1 Observed emissions and removals from high-carbon terrestrial ecosystems
 2   Most global deforestation is occurring in tropical forests (Pan et al., 2011; Liu et al., 2015; Houghton and
 3   Nassikas, 2017; Erb et al., 2018a; Li et al., 2018; Harris et al., 2021), primarily for clearing of agricultural
 4   land (Hong et al., 2021), causing primary tropical forest to comprise a net source of carbon to the atmosphere
 5   from 2001 to 2019 (emissions to the atmosphere 0.6 Gt y-1, removals from the atmosphere -0.5 Gt y-1, net 0.1
 6   Gt y-1) (Harris et al., 2021). While wildfires emitted an average of 0.4 ± 0.2 Gt y-1 carbon from 1997 to 2016
 7   (van der Werf et al., 2017), individual fire seasons can emit the same magnitude, such as the 0.4 Gt carbon
 8   from the Amazon fires of 2007 (Aragao et al., 2018), 0.5 Gt carbon from the Amazon fires of 2015-2016
 9   (Berenguer et al., 2021) and 0.2 Gt from the Australia fires of 2019-2020 (Shiraishi and Hirata, 2021). So,
10   wildfires account for up to one-third of annual average ecosystem carbon emissions, while major fire seasons
11   can emit up to two-thirds of global ecosystem carbon emissions (medium evidence, medium agreement).
13   Primary boreal and temperate forests also comprised net sources in the period 2001-2019, but, when
14   including all tree age classes, boreal, temperate, and tropical forests were net sinks, as growth exceeded
15   permanent forest cover losses (Harris et al., 2021), though boreal and temperate forests are much stronger
16   sinks (Pan et al., 2011; Liu et al., 2015; Houghton and Nassikas, 2017). Estimates of carbon removals from
17   remote sensing may provide more accurate estimates of boreal forest carbon balances than earth system
18   models, which overestimate regrowth after forest and timber (Wang et al., 2021a). Mortality of boreal forest
19   in British Columbia from mountain pine beetle infestations converted 374 000 km2 from a net carbon sink to
20   a net carbon source (Kurz et al., 2008). Modeling suggests that a potential increase in water-use efficiency
21   and regrowth could offset the losses in part of the forest mortality area (Giles-Hansen et al., 2021).
23   The Amazon as a whole was a net carbon emitter from 2003 to 2008 (Exbrayat and Williams, 2015; Yang et
24   al., 2018b), primarily due to expansion of agricultural and livestock areas, which caused over two-thirds of
25   deforestation from 1990 to 2005 (De Sy et al., 2015; De Sy et al., 2019). Four sites in the Amazon also
26   showed net carbon emissions from 2010 to 2018, from deforestation and fire (Gatti et al., 2021). In the
27   Amazon, deforestation emitted 0.17 ± 0.05 Gt y-1 carbon from 2001 to 2015 (Silva Junior et al., 2020) while
28   fires emitted 0.12 ± 0.14 Gt y-1 carbon from 2003 to 2015 (Aragao et al., 2018). An analysis of the Amazon
29   carbon loss from deforestation and degradation estimated a loss of 0.5 Gt y-1 from 2010 to 2019, with
30   degradation accounting for three-fourths (Qin et al., 2021). Intact old-growth Amazon rainforest has been a
31   net carbon sink (Hubau et al., 2020) but may have become a net carbon source from 2010 to 2019 (Qin et al.,
32   2021).
34   In Indonesia and Malaysia, draining and burning of peat swamp forests for oil palm plantations emitted 60 ‒
35   260 Mt y-1 carbon from 1990 to 2015, converting peatlands in that period from a carbon sink to a source
36   (Miettinen et al., 2017; Wijedasa et al., 2018; Cooper et al., 2020). Deforestation of mangrove forests
37   emitted 10‒30% of deforestation emissions in Indonesia from 1980 to 2005 (Donato et al., 2011; Murdiyarso
38   et al., 2015), even though mangroves comprised only 3% of Indonesia primary forest area in 2000 (Margono
39   et al., 2014; Murdiyarso et al., 2015).
41   In North America, wildfire emitted 0.1 ± 0.02 Gt y-1 of carbon from 1990 to 2012, but regrowth was slightly
42   greater to produce a net sink (Chen et al., 2017). In California, USA, two-thirds of the 70 Mt carbon
43   emissions from natural ecosystems from 2001 to 2010 came from the 6% of the area that burned (Gonzalez
44   et al., 2015). Anthropogenic climate change caused up to half of the burned area (Section
46   In the Arctic, anthropogenic climate change has thawed permafrost (Guo et al., 2020), leading to carbon
47   emissions of 1.7 ± 0.8 Gt y-1 in the winter from 2003 to 2017 (Natali et al., 2019). Wildfires in Arctic tundra
48   in Alaska from ~1930 to 2010 caused up to 0.5 m of permafrost thaw (Brown et al., 2015), exposing
49   peatland carbon (Brown et al., 2015; Gibson et al., 2018), including soil carbon deposits up to 1600 years old
50   (Walker et al., 2019).
52   Tropical deforestation, draining and burning of peatlands, and thawing of Arctic permafrost due to climate
53   change have caused those ecosystems to emit more carbon to the atmosphere than they naturally remove
54   through vegetation growth (high confidence).
56 Observed Changes in Primary Productivity

     Do Not Cite, Quote or Distribute                      2-62                                      Total pages: 237
     FINAL DRAFT                                    Chapter 2                   IPCC WGII Sixth Assessment Report

 1 Observed changes in terrestrial primary productivity
 2   The difference between photosynthesis by plants (gross primary productivity [GPP]) and plant energy use
 3   through respiration is the net growth of plants (net primary productivity [NPP]), which removes CO2 from
 4   the atmosphere and mitigates emissions from deforestation and other land use changes (Section
 5   Global terrestrial NPP has exceeded land use emissions since the early 2000s, making terrestrial ecosystems
 6   a net carbon sink (Friedlingstein et al., 2020).
 8   Global terrestrial NPP increased 6% from 1982 to 1999, through increased temperature and increased solar
 9   radiation in the Amazon from decreased cloud cover (Nemani et al., 2003), then decreased 1% from 2000 to
10   2009, because of extensive droughts in the southern hemisphere (Zhao and Running, 2010). From 1999 to
11   2015, increased aridity caused extensive declines in the Normalized Difference Vegetation Index (NDVI)
12   globally, particularly semi-arid ecosystems (Huang et al., 2016), indicating widespread decreases in NPP
13   (Yuan et al., 2019).
15   Global terrestrial GPP increased 2% from 1951 to 2010 and continued increasing at least through 2016, with
16   increased atmospheric CO2 showing a greater influence than natural factors (Li et al., 2017; Fernandez-
17   Martinez et al., 2019; Liu et al., 2019a; Cai and Prentice, 2020; Melnikova and Sasai, 2020). Global forest
18   area increased 7% from 1982 to 2016, mainly from forest plantations and regrowth in boreal and temperate
19   forests in Asia and Europe (Song et al., 2018), while regrowth in secondary forests > 20 years old, mainly in
20   boreal, temperate, and sub-tropical regions, generated a net removal of 7.7 Gt y-1 CO2 from the atmosphere
21   from 2001 to 2019 (Harris et al., 2021). Vegetation growth that exceeds the modelled CO2 fertilisation, gaps
22   in field data, and incomplete knowledge of plant mortality and soil carbon responses introduce uncertainties
23   into quantifying the magnitude of CO2 fertilisation (Walker et al., 2021). A combination of CO2 fertilisation
24   of global vegetation and secondary forest regrowth has increased global vegetation productivity (medium
25   evidence, medium agreement).
27   The relative increase in GPP per unit of atmospheric CO2 increase declined from 1982 to 2015, indicating a
28   weakening of any CO2 fertilisation effect (Wang et al., 2020c). Increased growth from CO2 fertilisation has
29   begun to shorten the life span of trees due to a trade-off between growth rate and longevity, based on
30   analyses of tree rings of 110 species around the world (Brienen et al., 2020). Furthermore, water availability
31   controls the magnitude of NPP (Beer et al., 2010; Jung et al., 2017; Yu et al., 2017), including water from
32   precipitation (Beer et al., 2010), soil moisture (Stocker et al., 2019), groundwater storage (Humphrey et al.,
33   2018; Madani et al., 2020a), and atmospheric vapour (Novick et al., 2016; Madani et al., 2020b). Drought
34   stress reduced NPP across tropical forests from 2000 to 2015 (Zhang et al., 2019b) and GPP in the tropics
35   from 1982 to 2016 (Madani et al., 2020b). Drought stress has also reduced GPP in some semi-arid and arid
36   lands (Huang et al., 2016; Liu et al., 2019a). In addition, nitrogen and phosphorus constrain CO2 fertilisation
37   (Terrer et al., 2019), though phosphorus limitation of tropical tree growth is species-specific (Alvarez-Clare
38   et al., 2013; Thompson et al., 2019). NPP has decreased during some time periods and in some regions
39   where drought stress has exerted a greater influence than increased atmospheric CO2 (medium evidence, high
40   agreement).
42 Observed changes in freshwater ecosystem productivity
43   Temperature affects primary productivity through moderating phytoplankton growth rates, ice cover, thermal
44   stratification and growing season length (Rühland et al., 2015; Richardson et al., 2018). Global warming has
45   reinforced eutrophication, especially cyanobacteria blooms (Wagner and Adrian, 2009; Kosten et al., 2012;
46   O’Neil et al., 2012; De Senerpont Domis et al., 2013; Adrian et al., 2016; Visser et al., 2016; Huisman et al.,
47   2018) (very high confidence). Conversely, warming can reduce cyanobacteria in hypertrophic lakes
48   (Richardson et al., 2019). Freshwater cyanobacteria may benefit directly from elevated CO2 concentrations
49   (Visser et al., 2016; Ji et al., 2017; Huisman et al., 2018; Richardson et al., 2019).
51   Macrophyte growth in freshwaters is likely to increase with rising water temperatures, atmospheric CO2 and
52   precipitation (robust evidence, high agreement) (Dhir, 2015; Hossain et al., 2016; Short et al., 2016;
53   Reitsema et al., 2018). Nonetheless, primary productivity in rivers is variable and unpredictable (Bernhardt
54   et al., 2018) because seasonal variations in temperature and light are uncorrelated, frequent high flow events
55   reduce biomass of autotrophs and droughts can strand and desiccate autotrophs.

     Do Not Cite, Quote or Distribute                     2-63                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   In large, nutrient-poor lakes warming-induced prolonged thermal stratification can reduce primary
 2   production (medium evidence) (Kraemer et al., 2017). Warming may reduce phytoplankton concentrations
 3   when temperature-induced increases in consumption of phytoplankton outpace increases in phytoplankton
 4   production (De Senerpont Domis et al., 2013). These decreases in productivity may be under-recognised
 5   responses to climate change.
 7   Summary: Evidence is robust for increase in primary production with warming trends, but increases or
 8   declines of algae cannot entirely be attributed to climate change; they are lake specific and modulated
 9   through weather conditions, lake morphology, salinity, land-use and restoration, and biotic interactions
10   (medium evidence, medium confidence) (O’Beirne et al., 2017; Velthuis et al., 2017; Rusak, 2018; Ho et al.,
11   2019).
13   2.4.5   Conclusions on Observed Impacts
15   The consistency of patterns of biological change with expectations from regional or global warming
16   processes, coupled with an understanding of underlying processes, the coherence of these patterns at both
17   regional and global scales, all form multiple lines of evidence (Parmesan et al., 2013) that it is very likely
18   that observed range shifts and phenological changes in individual species can be attributed to regional and
19   global climate changes (very high confidence) (Section 2.4.2, Table 2.2; Table 2.3; Table SM2.1; Parmesan
20   et al., 2013).
22   Global and regional meta-analyses of diverse systems, habitats and taxonomic groupings document that
23   approximately half of all species with long-term records have shifted their ranges poleward and/or upward in
24   elevation and ~2/3 have advanced their timing of spring events (phenology) (Section 2.4.2, Table 2.2;
25   Parmesan and Hanley, 2015; Parmesan, 2019). Changes in abundance tend to match predictions from climate
26   warming, with warm-adapted species significantly out-performing cold-adapted species in warming habitats
27   (Feeley et al., 2020) and the composition of local communities becoming more 'thermophilised' i.e.,
28   experiencing 'increase in relative abundance of heat-loving or heat-tolerant species' (Section; Cline et
29   al., 2013; Feeley et al., 2020).
31   New studies since AR5, with more sophisticated analyses designed to capture complex responses, indicate
32   that past estimates of the proportion of species impacted by recent climate change have been underestimates
33   due to their unspoken assumptions that local or regional warming should lead solely to poleward/upward
34   range shifts and advancements of spring timing (Duffy et al., 2019). More complex analyses have
35   documented cases of winter warming driving delayed spring timing of northern temperate species due to
36   chilling requirements, and increased precipitation driving species' range shifts downslope in elevation, and
37   eastward and westward in arid regions (high confidence). Further new studies have shown that phenological
38   changes have, in some cases, successfully compensated for local climate change and reduced degree of range
39   shifts (medium confidence). Limited number of studies of this type make it difficult to estimate the generality
40   of these effects globally (Section, Table 2.2).
42   Responses in freshwater species are consistent with responses in terrestrial species, including poleward and
43   upward ranges shifts, earlier timing of spring plankton development, earlier spawning in fish, and extension
44   of the growing season. Observed changes in freshwater species are strongly related to anthropogenic climate
45   change driven changes in the physical environment (e.g. increased water temperature, reduced ice cover,
46   reduced mixing in lakes, loss of oxygen, reduction in river connectivity). While evidence is high for an
47   increase in primary production in nutrient rich lakes with warming trends (high confidence), increasing or
48   declining algal formations are lake specific and modulated through variability in weather conditions, lake
49   morphology, changes in salinity, stoichiometry, land-use and restoration measures, and food web
50   interactions. In boreal coniferous forest, there has been an increase in terrestrial derived dissolved organic
51   matter transported into rivers and lakes as a consequence of climate change (that has induced increases in
52   run-off and greening of the northern hemisphere), as well as to changes in forestry practices. This has caused
53   waters to become brown resulting in an acceleration of upper water warming and an overall cooling of deep
54   water (high confidence). Browning may accelerate primary production through input of nutrients associated
55   with DOM in nutrient poor lakes and increase cyanobacteria growth, which better cope with low light
56   intensities (medium confidence) (Sections,,,

     Do Not Cite, Quote or Distribute                      2-64                                      Total pages: 237
     FINAL DRAFT                                         Chapter 2                    IPCC WGII Sixth Assessment Report

 1   Field research since the IPCC Fifth Assessment Report has detected biome shifts at numerous sites,
 2   poleward and upslope, that are consistent with increased temperatures and altered precipitation patterns
 3   driven by climate change, and support prior studies that attributed such shifts to anthropogenic climate
 4   change (high confidence). These new studies help fill prior geographic and habitat gaps, for example
 5   documenting upward shifts in the forest/alpine tundra ecotone in the Andes, Tibet and Nepal, and northward
 6   shifts in the deciduous/boreal forest ecotones in Canada. Globally, woody encroachment into open areas
 7   (grasslands, arid regions and tundra) is likely being driven by climate change and increased CO2 in concert
 8   with changes in grazing and fire regime (medium confidence) (Section 2.4.3).
10   Climate change has driven or is contributing to increased tree mortality directly through increased aridity or
11   drought and indirectly through increased wildfire and insect pests in many locations (high confidence).
12   Analyses of causal factors have attributed increasing tree mortality at sites in Africa and North America to
13   anthropogenic climate change and field evidence has detected tree mortality from drought, wildfire, and
14   insect pests in temperate and tropical forests around the world (high confidence). Water stress, leading to
15   plant hydraulic failure, is the a principal mechanism of drought-induced tree mortality, along with indirect
16   effects of climate change mediated through community interactions (high confidence) ( Section
18   Terrestrial ecosystems sequester and store globally critical stocks of carbon but these stocks are at risk from
19   deforestation and climate change (high confidence). Tropical deforestation, draining, and burning of
20   peatlands produce almost all of the carbon emissions from land use change. In the Arctic, increased
21   temperatures have thawed permafrost at numerous sites, dried some areas, and increased fire, causing net
22   emissions of carbon from soils (high confidence) (Sections,
24   Globally, increases in temperature, aridity, and drought have increased the length of fire seasons and doubled
25   potentially burnable area (medium confidence). Increases in burnt area have been attributed to anthropogenic
26   climate change in North America (high confidence). In parts of Africa, Asia, Australia, and South America,
27   area burned have also increased, consistent with anthropogenic climate change. Deforestation, peat burning,
28   agricultural expansion or abandonment, fire suppression, and inter-decadal cycles, strongly influence fire
29   occurrence. Areas with the greatest increases in fire season length include the Amazon, western North
30   America, western Asia, and East Africa. (Section
32   The changes we have observed, and project to continue, in biodiversity and ecosystem health pose a risk of
33   declines in human health and well-being: e.g. tourism, recreation, food, livelihoods and quality of life
34   (medium confidence). Clear attribution of these impacts is often not possible, but inference can be made by
35   comparison of observed changes in biodiversity / ecosystem health and known services from those particular
36   ecosystems.
39   Table 2.3: Confidence in detecting and attributing observed changes in terrestrial and freshwater species and systems to
40   climate change. Lines of evidence for attribution of observed changes to climate change and increased CO2 are used to
41   support stated confidence in attribution of key statements on observed biological changes to climate change and
42   increased atmospheric CO2. Icons represent lines of evidence. This is a summary table that is fully detailed in Table
43   SM2.1 in Supplementary Material.
44   Lines of evidence:
       Paleo data                             Experiment                   Long-term observations

      Fingerprint of climate                Models                         Complex statistical
      change response                                                      analysis
      Key statement                             Region       Period            Lines of evidence        Climate change

     Do Not Cite, Quote or Distribute                         2-65                                          Total pages: 237
FINAL DRAFT                                               Chapter 2              IPCC WGII Sixth Assessment Report

 About half of all species where land use       Global        Range 20 - 260
 change has been minimal have shifted their                   years                             robust evidence
 ranges, with 80‒90% of movements being                                                         high agreement
 in the direction expected from regional                                                        very high
 warming trends ‒ i.e. poleward and                                                             confidence

 Downslope elevational shifts and east-west     USA           ~ 40‒60 years
 shifts (shown for trees and birds) have been                                                   limited evidence,
 associated with regional increases in                                                          high agreement,
 precipitation where precipitation has been                                                     medium confidence
 shown to be the principal driver of a range

 About two- thirds of all species with long-    Global        Varies by study.
 term (>20 years) records have shifted the                    Range: 20‒400                     robust evidence
 timing of spring events in directions                        years                             high agreement
 expected from regional winter and spring                                                       very high
 warming.                                                                                       confidence

 Winter chilling-depending species have         Northern      Varies by study.
 delayed or not changed spring events           Europe        Range = 26‒46                     medium evidence
 despite spring warming countered by            and USA       years                             high agreement
 winter warming. When these species are                                                         high confidence
 taken into account, it is estimated that 92%
 of species in these studies have responded
 to regional warming trends

 Anthropogenic climate change, acting           western       1984-2017
 through increased heat and aridity at global   north                                           robust evidence
 mean surface temperature increases of 0.6-     America                                         high agreement
 0.9ºC, has increased the area burned by                                                        high confidence
 wildfire over natural levels, increasing
 burned area up to 11 times in one extreme
 year and doubling over natural levels in a
 32-year period

 Anthropogenic climate change has caused        North         ca. 1945-2007
 drought-induced tree mortality of up to        America                                         medium evidence
 20% in three regions, through global mean      and Africa                                      high agreement
 surface temperature increases of 0.3-0.9ºC                                                     medium confidence
 above the pre-industrial period and
 increases in aridity, more than non-climate
 change factors

 Anthropogenic climate change has caused        Global        1500-2007
 latitudinal and elevational vegetation biome
 shifts in at least 19 sites in boreal,                                                         robust evidence
 temperate, and tropical ecosystems,                                                            high agreement
 between 1700 and 2007, through local                                                           high confidence
 temperature increases of 0.4 to 1.6ºC above
 the pre-industrial period more than non-
 climate change factors

 Anthropogenic climate change and wildfire      western       1966-2015                         medium evidence
 together have altered vegetation species       North                                           high agreement
 composition in at least two regions,           America,                                        medium confidence
 reducing post-fire natural regeneration and    Africa
 species richness of tree and other plant
 species, at global mean surface temperature
 increases of 0.3-0.9ºC

Do Not Cite, Quote or Distribute                               2-66                                 Total pages: 237
FINAL DRAFT                                              Chapter 2             IPCC WGII Sixth Assessment Report

 Beetles & moths shifting poleward and         North         Varies by study
 upward has brought new pest species into      America,                                       medium-high
 some forests; warming winters and longer      Europe                                         confidence
 growing season has increased destructive
 outbreaks of beetles and moths in temperate
 and boreal forests
 Exotic species are responding differently     North
 from native species in both abundance         America                                        low/medium
 changes and phenological changes, but not                                                    evidence
 in a consistent fashion                                                                      low agreement

 The most cold-adapted species are             Arctic,
 generally declining in population             Himalayas
 abundances and contracting their ranges       ,                                              medium confidence
 poleward and upward: (e.g. sea-ice            Antarctic,
 dependent, mountain-top restricted, upper     Alps
 headwaters, coldest lakes)

 Diseases of both wildlife and humans have     Global
 emerged into new areas they have not been                   past 20-100                      medium confidence
 in historically                                             years

 Warming has amplified the trophic state       Global        Past 20-50
 lakes are already in. Eutrophic lakes have                  years                            robust evidence
 become more productive while oligotrophic                                                    high agreement
 lakes tend to become more nutrient limited                                                   high/medium

 Woody encroachment into open (grassland,      Global
 desert) systems has occurred, with climate
 change as one of the drivers, along with                                                     medium confidence
 changes in grazing and other land uses

 In boreal, coniferous areas changes in        Boreal        Past decades
 forestry practices and climate change have
 caused an increase in terrestrial derived                                                    robust evidence
 dissolved organic matter (DOM) transport                                                     high agreement
 into rivers and lakes leading to their                                                       high confidence

 Climate change induced warming leads to       Global        Past decades
 shifts in thermal regime of lakes                                                            robust evidence
                                                                                              high agreement
                                                                                              high confidence

 Climate change causes gains and losses in     Global        Past decades                     limited evidence
 freshwater water level                                                                       low confidence

 Greenhouse gas emissions from freshwater      Global        Past decades                     medium evidence
 ecosystems are equivalent to around 20% of                                                   medium agreement
 global burning fossil-fuel CO2 emission                                                      medium/low

 In lakes weather extremes in wind,            North         Varies by study                  medium/limited
 temperature, precipitation and loss of ice    America,                                       evidence
 foremost affect the thermal regime with       Europe                                         high agreement
 repercussions on water temperature,                                                          medium/low
 transparency, oxygen and nutrient                                                            confidence
 dynamics, affecting ecosystem functionality

Do Not Cite, Quote or Distribute                              2-67                                Total pages: 237
     FINAL DRAFT                                             Chapter 2           IPCC WGII Sixth Assessment Report

      Climate change induced warming leads to      North         Past decades
      shifts in thermal regime of rivers and       America,                                      robust evidence
      streams; lowland rivers show a stronger      Europe                                        medium agreement
      thermal response than high-altitude, cold-                                                 high confidence
      water receiving streams

      Loss of biodiversity in streams can be       Global        Past decades
      directly attributed to climate change                                                      high agreement
      through increased water temperatures,                                                      very high
      hydrological changes such as increased                                                     confidence
      peak discharges, flow alteration and

      Climate change is causing range shifts of    North         Past decades                    high agreement,
      freshwater fish                              America                                       high confidence
 3   2.5 Projected Impacts and Risk for Species, Communities, Biomes, Key Ecosystems and their
 4   Services
 6   Under the Risk Assessment Framework that was introduced in AR5 (2014), risk means the probability of
 7   harmful consequences resulting from climate change. It results from the interaction of vulnerability,
 8   exposure, and hazard (see Chapter 1) and can be represented as the probability of occurrence of hazardous
 9   events or trends multiplied by the impacts if these events or trends occur (IPCC, 2014). The Framework
10   defines vulnerability as a pre-existing condition, incorporating the extent to which species or ecosystems are
11   susceptible to, or unable to cope with, adverse effects of climate change. Vulnerable species have limited
12   adaptive capacity stemming from physiological and behavioural constraints, limited dispersal abilities and
13   restricted resource requirements or capacities for distributional and genetic changes (Foden et al., 2019)
14   (Foden et al., 2013; Cizauskas et al., 2017). Traits that render entire ecosystems vulnerable are harder to
15   define, but it is clear that vulnerabilities are high in the coldest habitats, in those with limited geographic
16   ranges such as low-lying islands, and in specialized, restricted habitats such as serpentine outcrops in
17   California (Anacker and Harrison, 2012) and dry meadows in Fennoscandia and Tibet (Yang et al., 2018a).
18   Ecosystem vulnerability can depend critically on the fates of plants that function as 'foundation species,'
19   providing community biomass aboveground and below, structuring habitat for fauna and providing
20   ecosystem services such as erosion control (Camac et al., 2021).
22   2.5.1     Projected Changes at Species and Community Levels
24    Assessment of Models and Sources of Uncertainties
26   Methods for projecting impacts of climate change on biodiversity can be classified into three types: 1)
27   statistical models such as species distribution models (Elith and Leathwick, 2009); 2) mechanistic or
28   process–based models (Chuine and Régnière, 2017) and 3) trait–based models (Pacifici et al., 2015). It is
29   only recently that models have been developed looking at smaller levels of warming. such as 1.5°C (Hoegh-
30   Guldberg et al., 2018a; Warren et al., 2018).
32   Species distribution models (SDMs) or niche-based models assess potential geographic areas of suitable
33   climate for the species in current conditions and then project them into future conditions (Trisurat, 2018;
34   Vieira et al., 2018). There are limitations in all models and it is critical that modellers understand the
35   assumptions, proper parameterization, and limitations of each model technique, including differences among
36   climate models, emission scenarios or representative concentration pathways, and baselines (Araujo et al.,
37   2019). Several systems automate development of SDMs, including R-packages (e.g., (Beaumont et al., 2016;
38   Hallgren et al., 2016), development of other model types (Foden et al., 2019) and aid in use of climate model
39   data (Suggitt et al., 2017), including allowing for connectivity constraints (Peterson et al., 2013). Buisson et
40   al. (2010) found most variation in model outputs stem from differences in design, followed by GCMs.
42   Mechanistic approaches, also known as process-based models, project species’ responses to climate changes
43   by explicitly incorporating known biological processes, thresholds and interactions (Morin and Thuiller,

     Do Not Cite, Quote or Distribute                             2-68                               Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1   2009; Maino et al., 2016). Mechanistic models are able to accommodate a broad range of climate change
 2   impact mechanisms and include species-specific characteristics such as dispersal distances, longevity,
 3   fecundity, genetic evolution, phenotypic plasticity. However, sufficient knowledge is available for only a
 4   few well-studied species. Species traits have been used to more broadly estimate potential climate change
 5   impacts (Foden et al., 2013; Cizauskas et al., 2017).
 7   Most models are at large scales (20 km–50 km), and so cannot capture micro-climatic refugia generated by
 8   diversities of slope aspect, elevation or shade (Suggitt et al., 2015; Suggitt et al., 2018). In analyzing records
 9   of 430 climate-threatened and range-declining species in England, (Suggitt et al., 2015; Suggitt et al., 2018)
10   showed that topographic diversity reduced population declines most strongly in areas experiencing the most
11   local warming and in species most sensitive to warming. In these circumstances topographic diversity
12   reduced population extinction risk by 22% for plants and by 9% for insects.
14   None of the modelling techniques are predictions of the future; they are projections of possible futures. To
15   date, few studies have validated model performance against observations, and many of these have been on
16   islands, reducing ability to generalize (Fordham et al., 2018). Species' models should be considered as
17   hypotheses of what a future world might look like if the climate projections came to pass. Suggestions have
18   been made on how to start bringing more biotic interactions into species distribution models (Early and
19   Keith, 2019), but limited basic ecological understandings of interactions, along with limits on computation
20   and funding, constrain how far and how fast these modelling techniques can advance.
22   Risk Assessment and non-modelling approaches
24   In order to add realism and reliability to risk assessments at the species and community levels, non-
25   modelling approaches based on known biological traits or processes, as well as on expert opinion (Camac et
26   al., 2021), are used to temper model outputs with ground-based validation. Trait-based assessment
27   approaches use species’ biological characteristics as predictors of sensitivity, adaptive capacity and
28   extinction risk due to climate change. Climate exposure can be estimated using GIS-based modelling,
29   statistical programs or expert judgment (Chin et al., 2010). These trait-based approaches are widely applied
30   to predict responses of biodiversity to climate change because they do not require modelling expertise nor
31   detailed distibutional data (Pacifici et al., 2015; Willis et al., 2015). Most of these methods have not been
32   independently validated and do not allow direct comparison of vulnerability and risk among taxonomic
33   groups.
35   Some studies have combined two or three approaches for climate change risk assessment of biodiversity in
36   order to capture the advantages of each and avoid their limitations. Warren et al. (2013) used combinations
37   of SDMs and trait-based approaches to estimate the proportions of species losing their climatically suitable
38   ranges under the various future scenarios of climate and dispersal rate. Similarly, spatial projections of
39   climate change exposure were combined with traits to assess vulnerability of sub-Saharan amphibians
40   (Garcia et al., 2014). Laurance et al. (2012) combined 31 functional groups of species and 21 potential
41   drivers of environmental change to assess both the ecological integrity and threats for tropical protected
42   areas on a global scale. Keith et al. (2014) used a combination of three approaches (SDMs-trait-mechanistic)
43   to determine how long before extinction a species would become eligible for listing as threatened based on
44   the IUCN Red List criteria.
46   Risk of Species’ Extinctions
48 Overview
49   This assessment of current findings is of studies across a range of taxa and modelling techniques. Extinction
50   risk estimates whether or not a particular species may be at risk of extinction over the coming decades if
51   climatic trends continue and usually does not take into account other human-induced stressors (e.g. invasive
52   species or pollution). It is not a prediction that a species will definitely go extinct, because even complete
53   loss of a species’ range is projected, the scale of the model cannot estimate persistence in micro-climatic
54   refugia (Suggitt et al., 2015; Suggitt et al., 2018). Individuals and populations can survive after conditions
55   for successful reproduction are gone, leading to a lagged decline, called 'extinction debt' (Alexander et al.,
56   2018). Therefore, range loss is an established criterion for assessing endangerment status and risk of
57   extinction. As a species range becomes smaller and occupied habitats become more isolated, the likelihood

     Do Not Cite, Quote or Distribute                       2-69                                        Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   of a single stochastic event causing extinction increases. It is this combination of projected loss of
 2   climatically suitable space and additional stressors (especially LULCC of critical habitat) that is expected to
 3   drive future extinctions.
 5   The IUCN Red List Criteria (IUCN, 2019) classifies a species as 'critically endangered' if it has suffered a
 6   range loss of ≥80%, with a resulting likelihood of extinction of >50% in the near term (10-100 years,
 7   depending upon generation time). A species is classified as 'endangered' if it has suffered a range loss of
 8   ≥50%, with a resulting likelihood of extinction of >20% in the near term (10-100 years).
10 Projections for freshwater biodiversity
11   Lakes, rivers and freshwater wetlands cover approximately 7.7–9.1 % of global land surface area; (Lehner et
12   al., 2008; Fluet-Chouinard et al., 2015; Allen and Pavelsky, 2018), and hold 9.5% of the Earth’s described
13   animals (Balian et al., 2008), with climate change indicated as a threat for 50–75% of fish (Xenopoulos et
14   al., 2005; Darwall and Freyhof, 2015 who). Climate change is cited as a primary factor in species' extinction
15   risk, through changes in water temperatures, stream flow, loss of cold water habitat, increased variability of
16   precipitation, and increased disease risk from warming temperatures (high evidence, high agreement, high
17   confidence) (Knouft and Ficklin, 2017; Pletterbauer et al., 2018; Reid et al., 2019; Jaric et al., 2019), adding
18   to stress from overexploitation and LULCC (Craig et al., 2017; IPBES, 2019 Global assessment report).
20   Increased frequency of stream drying events, reducing hydrologic connectivity and limiting access of native
21   fishes to spawning habitats is projected for RCP 8.5 in Colorado, USA (medium evidence, medium
22   agreement) (Jaeger et al., 2014). Cold-water habitats and associated obligate species are particularly
23   vulnerable and losses in these habitats have been both documented and projected, for example in salmonids
24   (Santiago et al., 2016; Fullerton et al., 2017; Merriam et al., 2017). River networks are projected to lose
25   connections to cold tributary refugia, that are important thermal refuges for cold water species (robust
26   evidence, high agreement) (Isaak et al., 2016) during low flows (Merriam et al., 2017).
28   Community turnovers are expected in freshwaters as cold-adapted species lose and warm-adapted species
29   gain climatically suitable habitat (Domisch et al., 2011; Domisch et al., 2013; Shah et al., 2014). While a
30   number of warm-adapted species may experience range expansions, the majority of species are predicted to
31   lose climatically suitable areas by on average 38–44%, depending on the emission scenario (A2a and B2a)
32   (medium evidence) (Domisch et al., 2013).
34   Molluscs are projected to be the most at risk group, given their limited dispersal capability (Woodward et al.,
35   2010). Mediterranean freshwater fish are especially susceptible to climate change due to increasing flood and
36   drought events and risks of surpassing critical temperature thresholds (Santiago et al., 2016; Jaric et al.,
37   2019). In southern Europe, aquatic insects (Ephemeroptera, Plecoptera, and Trichoptera) are endangered by
38   climate change (Conti et al., 2014). European protected areas are not expected to be sufficient under
39   warming to provide habitat for the majority of rare molluscs and fish (Markovic et al., 2014). Observed
40   trends agree with model projections in direction, but magnitude remains uncertain (medium evidence,
41   medium agreement, medium confidence). (see Figure 2.8 for extinction risk globally for dragonflies,
42   amphibians and turtles).
44   Regional threats from climate change have been reported for 40% of amphibians in China, (Wu, 2020), 33%
45   of European freshwater fish species (Janssen et al., 2016) and 56-69% of Odonates in Australia, (Bush et al.,
46   2014b). Assessment of site-specific extirpation likelihoods for 88 aquatic insect taxa projected that climate-
47   change induced hydrological alteration would result in a 30–40% loss of taxa in warmer, drier ecoregions
48   and 10–20% loss in cooler, wetter ecoregions (medium evidence) (Pyne and Poff, 2017). In Africa’s
49   Albertine Rift, 51% (n=551) of fish are expected to be impacted by climate change, with 5.5% at high risk
50   due to their sensitivity and poor adaptative capability (high agreement) (Carr et al., 2013).
52   The GLOBIO-Aquatic model (Janse et al., 2015 a) links models for demography, economy, land use
53   changes, climate change, nutrient emissions, a global hydrological model and a global map of water bodies.
54   It projects that changes in both water quality (eutrophication ) and quantity (flow) will generate negative
55   relations in freshwater ecosystems between persistence of species originally present in each community and
56   a constellation of stressors, including harmful algal blooms. Under 4°C rise by 2050, mean abundance of

     Do Not Cite, Quote or Distribute                      2-70                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   species is projected to decline by 70% in running water and 80% in standing water (medium evidence, high
 2   agreement, medium confidence) (Janse et al., 2015 a).
 4 Global projections of extinction risk
 5   In prior reports, risk assessed from the literature was generally based on estimates of overall range
 6   contractions with climate change. In AR4, extinction risk was carefully quantified: 'There is medium
 7   confidence that approximately 20–30% of species assessed so far are likely to be at increased risk of
 8   extinction if increases in global average warming exceed 1.5–2.5°C (relative to 1980 to 1999). As global
 9   average temperature increase exceeds about 3.5°C, model projections suggest significant extinctions (40–
10   70% of species assessed) around the globe.' These estimates approximately correspond to 50% reductions in
11   range size (IPCC, 2007a). AR5 stated 'a large fraction of terrestrial and freshwater species face increased
12   extinction risk under projected climate change during and beyond the 21st century, especially as climate
13   change interacts with other pressures ... (high confidence)' (Field et al., 2014). A series of multi-species and
14   global analyses have been published since AR5, using both statistical models and trait-based approaches.
16   In this Chapter, risk to species, with implications for ecosystems, is assessed using three different
17   approaches. First is an assessment of the geographic distributions of species' losses at different levels of
18   GAST warming, termed 'biodiversity loss, measured as the proportion of species within a given location
19   becoming classified as 'endangered' or 'critically endangered' (sensu IUCN). This measure provides estimates
20   of which sites are at most risk of losing substantial numbers of species locally, leading to degradation of that
21   ecosystems' ability to function. Second is an assessment of risk of proportions of species' becoming extinct
22   globally at different levels of GAST warming, measured using the IUCN criteria for 'critically endangered',
23   and termed 'species' extinction risk'. This measure is closest to assessing the complete loss of a species in the
24   wild, and can be used to compare to past (paleo) extinction rates. Third is an assessment of proportions of
25   species becoming rare or endangered globally (not just locally), and is the foundation for the Burning
26   Embers on biodiversity risk in Figure 2.11. These three approaches provide complementary information of
27   the overall risks to biodiversity and ecosystem integrity under different warming levels.
29   Biodiversity risk, estimated as the proportion of species in a given area projected to become endangered
30   (sensu IUCN), is projected to affect a greater number of regions with increasing warming, with about one-
31   third of land area risking loss of >50% of species currently inhabiting those ecosystems. Species' losses are
32   projected to be worst in the northern South America, southern Africa, most of Australia, and northern high
33   latitudes (medium confidence) (Figure 2.6).

     Do Not Cite, Quote or Distribute                      2-71                                       Total pages: 237
     FINAL DRAFT                                         Chapter 2                     IPCC WGII Sixth Assessment Report

 2   Figure 2.6: Biodiversity loss at increasing levels of climate change. The higher the percent of species projected to
 3   lose suitable climate in a given area, the higher the risk to ecosystem integrity, functioning and resilience to climate
 4   change. Warming levels are based on global levels (GSAT) above pre-industrial temperatures. Colour shading represent
 5   proportion of species for which the climate is projected to become unsuitable within a given pixel across their current
 6   distributions at a given GSAT warming level, based on the data underpinning (Warren et al., 2018) (modelled
 7   n=119,813 species globally, with no dispersal, averaged over 21 CMIP5 climate models). Areas shaded in green are
 8   above the 50% biodiversity loss threshold, meaning that <50% of species in that area are projected to go locally extinct.
 9   Areas shaded in pink and purple represent significant risk of biodiversity loss (areas where climates become unsuitable,
10   rendering them locally extinct, for >50% and >75% of species, respectively). The maps of species richness remaining
11   have been overlaid with a landcover layer (2015) from the European Space Agency Climate Change Initiative. This
12   landcover layer leaves habitats classified by the ESA as natural as being transparent, cities as black, water as blue,
13   permanent snow/ice as white and bare/rock as dark brown. Areas with a landcover identified as agriculture are 5%
14   transparent, such that potential species richness remaining if the land had not been converted to agricultural shows as
15   pale shading of the legend colours (very pale pink or very pale green). These paler areas represent biodiversity loss due
16   to habitat destruction, but with a potential to be restored, with green shading having potential for restoration to higher
17   species richness than pink and purple shadings.

     Do Not Cite, Quote or Distribute                          2-72                                           Total pages: 237
     FINAL DRAFT                                          Chapter 2                     IPCC WGII Sixth Assessment Report

 1   Risk of species' becoming extinct globally was estimated as the probability of loss of suitable climate space
 2   rendering it critically endangered (sensu IUCN). It is likely that the percentage of species at high risk of
 3   extinction (median and maximum estimates) will be 9% (max 14%) at 1.5°C, 10% (max 18%) at 2°C, 12%
 4   (max 29%) at 3.0°C, 13% (max 39%) at 4°C and 15% (max 48%) at 5°C (Figure 2.7). Among groups
 5   containing largest numbers of species at high risk of extinctions for mid-levels of projected warming (3.2°C)
 6   are: invertebrates (15%), specifically pollinators (12%), amphibians (11%, but salamanders are at 24%) and
 7   flowering plants (10%) (Figure 2.8a). All groups fare substantially better at 2°C, with extinction
 8   projectionsreducing to <3% for all groups, except salamanders at 7% (medium confidence) (Figure 2.8a).
 9   Even the lowest estimates of species' extinctions (9%) are 1000x natural background rates (Section 2.5.4; De
10   Vos et al., 2015). Projected species' extinctions at future global warming levels are in accord with
11   projections from AR4, assessed on much larger numbers of species with much greater geographic coverage
12   and a broader range of climate models. (Figure 2.7; Figure 2.8)

16   Figure 2.7: Synthesis of modelled climate-driven extinction risk studies. The relationship between modelled
17   projections of extinction (expressed as a proportion of species at risk of extinction assessed in individual studies) and
18   GSAT increase above the pre-industrial average. Data (global sample size n = 178 modelled estimates) were sourced
19   from a number of sources, including digitization of data points in Figure 2 in the synthetic analysis of (Urban, 2015),
20   note: unweighted for sample size, n = 126); Table 4.1 of the AR4 WG2 chapter 2 (Fischlin et al., 2007), n = 40;
21   (Hannah et al., 2020) n = 6); and (Warren et al., 2018) n = 6). The. Quantile regression (which is robust to the non-
22   normal distribution of the response variable, and less sensitive to data outliers) was used to fit quantile estimates for
23   levels relevant to informing “likely” (between the 0.17 and 0.83 quantiles, shaded in orange) and “very likely” ranges
24   (between the 0.05 and 0.95 quantiles, shaded in green) relating extinction risk to GSAT increase (Quantile regression
25   implemented using the Barrodale and Roberts algorithm in XLSTAT). The roughly equivalent estimate of this risk as
26   expressed in AR4 (Fischlin et al., 2007) is indicated by the dotted block indicating the medium confidence statement
27   “Approximately 20 to 30% of plant and animal species assessed so far (in an unbiased sample) are likely to be at
28   increasingly high risk of extinction as global mean temperatures exceed a warming of 2 to 3°C above preindustrial
29   levels (medium confidence).” This box is open on the right-side because AR4 estimates stipulated temperatures at or
30   exceeding given levels.
33   Projections of extinction risk by taxa are presented both for risk of becoming critically endangered (losing
34   ≥80% of suitable climate habitat, Figure 2.8a) and endangered (losing ≥50% of suitable climate habitat,
35   Figure 2.8b). The percentages of species projected at risk of becoming endangered (or worse) was 49% for
36   insects, 44% for plants, and 26% for vertebrates at ~3°C global rise in temperature (Warren et al., 2018).
37   Those estimates dropped considerably at lower levels of warming, down to 18%, 16%, and 8% at 2°C; and
38   6%, 8% and 4% at 1.5°C (Warren et al., 2018); thus not entirely dis-similar to the numbers in AR4 (Figure
39   2.7). Figure 2.8 shows the benefits of dispersal in offsetting extinction risk in birds, mammals, butterflies,
     Do Not Cite, Quote or Distribute                           2-73                                           Total pages: 237
     FINAL DRAFT                                           Chapter 2                      IPCC WGII Sixth Assessment Report

 1   moths and dragonflies. While dispersal may benefit individual species, it poses additional risks to
 2   communities and ecosystems as interactions between species are changed or eliminated.

 6   Figure 2.8: Percent of species of different groups classified as being under risk of extinction. a) Percent of the species
 7   group listed projected to be at very high risk of extinction, corresponding to the IUCN Red List criteria for a species
 8   classified as "critically endangered" (version 3.1) through losing >80% of its climatically suitable range area. b) Percent
 9   of the species group listed projected to be at high risk of extinction, corresponding to the IUCN Red List criteria for a
10   species classified as "endangered" (version 3.1) through losing >50% of its climatically suitable range area. For a) and
11   b), values calculated from the underlying data underpinning (Warren et al., 2018). Values for each temperature are the
12   mean values across 21 CMIP5 models. The grey band represents the high-end of extinction risk from the 10th
13   percentile of the climate models to show the maximum range of values while the low end (90th percentile, 1.5°C) is not
14   shown as it is too small to appear on the plots. Taxa marked with * represent potential benefits from adaptation,
15   specifically dispersal at realistic rates (Warren et al., 2018); those with no * have dispersal rates that are essentially not
16   detected in the spatial resolution of the models (20 km). See (Warren et al., 2018) for caveats and more details. Sample
17   size for each group is as follows: Invertebrates (33949), Annelid Worms (155), Butterflies (1684), Moths (6910),
18   Dragonflies (599), Pollinators (1755), Spiders (2212), Beetles (7630), True Bugs (1728), Bees/Ants/Wasps (5914),
19   Flies (4809), Plants (72399), Flowering Plants (52310), Conifers (340), Timber spp (1328), Grasses (3389), Fungi
20   (16187), Vertebrates (12642), Mammals (1769), Carnivores (107), Ungulates (80), Bats (500), Birds (7968),
21   Passeriformes (4744), Non-passeriformes (3224), Amphibians (1055), Frogs (887), Salamanders (163), Reptiles (1850),
22   Snakes (1741), Turtles (94).
25   For local biodiversity loss, at 1.58°C (median estimate), >10% of species are projected to become
26   endangered (sensu IUCN); at 2.07°C (median) >20% of species are projected to become endangered,
27   representing high and very high biodiversity risk, respectively (medium confidence) (see Section 2.5.4;
28   Figure 2.11; Table 2.5, Table 2.S.4).
30   Using data from geological time scales, Song et al. (2021) predicted that a warming of 5.2 °C above pre-
31   industrial would result in mass extinction comparable to that of the five mass extinctions over the past 540
32   My, on the order of 70–85% of species going extinct, in the absence of non-climatic stressor. Mathes et al.
33   (2021) found evidence in the geological record that short-term rapid warming, on top of long-term warming
34   trends, increases extinction risk by up to 40% over that expected from the long-term trend alone, with a
35   biodiversity 'memory' up to 60 Myr, indicating an additonal risk of multi-decadal overshoot.

     Do Not Cite, Quote or Distribute                            2-74                                            Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 2   Most of the large-scale studies that have been performed are for losses based on climate alone (Figures 2.6,
 3   2.7, 2.8). However, climate is rarely the only stressor affecting species survival. Habitat loss is currently the
 4   largest driver of range loss and extinction risk for most species (IUCN). Communities in different regions are
 5   becoming more similar to each other as species tolerant of human activities prosper and spread, with many
 6   rare and endemic species already having been driven extinct, primarily by LULCC (Pimm et al., 2006).
 7   Thus, it will likely be the interaction of climate change and habitat conversion (often also being driven by
 8   climate change) that will ultimately determine the risk and ability to survive of many species.
10    Changing Risks of Diseases
12   Multiple studies predict increases in disease incidence or geographic and phenological changes of pathogens,
13   vectors, and reservoir host species due to climate change with or without other non-climatic variables
14   (González et al., 2010; Moo-Llanes et al., 2013; Roy-Dufresne et al., 2013; Liu-Helmersson et al., 2014;
15   Laporta et al., 2015; Ryan et al., 2015; Haydock et al., 2016; Hoover and Barker, 2016; Prist et al., 2017;
16   Blum and Hotez, 2018; Dumic and Severnini, 2018; Hundessa et al., 2018; Ryan et al., 2019; Ryan et al.,
17   2021). However, models predicting changes in infectious disease risk are complex and sometimes produce
18   conflicting results and lack consensus (Caminade et al., 2014; Giesen et al., 2020). For example, malaria is
19   projected to increase in some regions of Africa, Asia, and South America by the end of the 21st century if
20   public health interventions are not sufficient, but malaria is also forecasted to decrease in some of the higher
21   risk areas (Peterson, 2009; Caminade et al., 2014; Ryan et al., 2015; Khormi and Kumar, 2016; Leedale et
22   al., 2016; Murdock et al., 2016; Endo and Eltahir, 2020; Mordecai et al., 2020).
24   While malaria risk is predicted to decrease in some lowland tropical areas as temperatures become too hot
25   for vector or parasite development, other, warmer-adapted diseases like dengue and Zika, transmitted by
26   Aedes aegypti, are predicted to increase (Ryan et al., 2019; Ryan et al., 2021). In more temperate regions,
27   arboviruses and other vector-borne diseases with wider thermal breadths, such as West Nile fever, Ross
28   River fever, and Lyme disease, are predicted to increase with climate warming (Ogden et al., 2008; Leighton
29   et al., 2012; Shocket et al., 2018; Shocket et al., 2020; Couper et al., 2021), and drought can exacerbate these
30   effects of temperature (Paull et al., 2017).
32   A global analysis of 7346 wildlife populations and 2021 host-parasite combinations found that organisms
33   adapted to cool and mild climates are likely to experience increased risks of outbreaks with climate warming
34   while warm-adapted organisms may experience lower disease risk, providing further support for predictions
35   that climate change will increase infectious disease transmission in higher latitude regions across a
36   taxonomically diverse array of pathogens (robust evidence, high confidence) (Cohen et al., 2020). A study
37   examining the future risk of arboviruses (chikungunya, dengue, yellow fever, and Zika viruses) spread by
38   Aedes aegypti and Ae. albopictus projected increased disease risk due to interactions of multiple variables,
39   including increased human connectivity, urbanisation and climate change (Kraemer et al., 2019), although
40   vector species’ ranges broaden only slightly (Campbell et al., 2015).
42   In sum, climate change is expected to expand and redistribute the burden of vector-borne and other
43   environmentally-transmitted diseases by shifting many regions toward the thermal optima of vector-borne
44   disease transmission for multiple parasites, increasing transmission, while pushing temperatures above
45   optima and toward upper thermal limits for other vectors and pathogens, decreasing transmission (Mordecai
46   et al., 2019; Mordecai et al., 2020). These effects are mediated by other human impacts such as land use
47   change, mobility, socio-economic conditions, and vector and pathogen control measures (Parham et al.,
48   2015; Tjaden et al., 2018).
50   2.5.2     Projected Changes at Level of Biomes and Whole Ecosystems
52     Global Overview, Assessment of Ecosystem-level Models, and Sources of Uncertainties
54   Shifts in terrestrial biome and changes in ecosystem processes in response to climate change are most
55   frequently projected with dynamic global vegetation models (DGVMs), or land-surface models that form
56   part of Earth System Models, which use gridded climate variables, atmospheric CO2 concentration and
57   information on soil properties as input variables. Since AR5, most of DGVMs have been upgraded to capture

     Do Not Cite, Quote or Distribute                      2-75                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   carbon-nitrogen cycle interactions (e.g. Le Quéré et al., 2018), many also include a representation of
 2   wildfire, and fire-vegetation interactions (Rabin et al., 2017), and a small number now also accounts for land
 3   management (such as wood removal from forests, crop fertilisation harvest of irrigation (Arneth et al., 2017).
 4   Other forms of disturbance, such as tree mortality in response to, for example, episodic weather extremes or
 5   insect pest outbreaks, are relatively poorly represented, or not at all, although they demonstrably impact
 6   calculated carbon cycling (Pugh et al., 2019a). Simulated biome shifts are generally in agreement in
 7   projecting broad patterns at the global scales, but vary greatly regarding the simulated trends in historical and
 8   future carbon uptake or losses, both regionally and globally (WGI, Chapter 5 AR6; Chang et al., 2017).
10   Similar to other models, models to project large-scale changes in vegetation and ecosystem processes have to
11   deal with structural uncertainty (associated with the choice and the representation of processes in models),
12   input-data uncertainty (associated with variability in initial conditions and parameter values) and error
13   propagation (associated with coupling models) (Rounsevell et al., 2019). The IPBES methodological
14   assessment report on scenarios and models of biodiversity and ecosystem services provides a comprehensive
15   overview over the relevant issues (Ferrier et al., 2016).
17   In order to assess the model’s performance, most models have been individually evaluated against a range of
18   observations. Moreover, in the annual updates of the global carbon budgets a model has to meet a small set
19   of basis criteria to have its output included (Le Quéré et al., 2018). More systematic benchmarking
20   approaches have also been proposed that utilise a range of different data sets (Kelley et al., 2013; Chang et
21   al., 2017), in order to assess multiple simulated processes. These methods in principle allow to assign quality
22   scores to models based on their overall performance (Kelley et al., 2013). So far, this scoring does not yet
23   allow a clear quality ranking of models since the individual DGVMs tend to score well for some variables
24   and badly for others. A recent comparison of global fire-vegetation model outputs was also able to clearly
25   identify outliers when using a formalised benchmarking and scoring approach (Hantson et al., 2020).
26   However, benchmarking does not address sources of uncertainty and it would be advisable to perform
27   “perturbed-physics” experiments, in which multiple model parameters are varied in parallel more frequently,
28   as a means to test parameter-value uncertainty (Wramneby et al., 2008; Booth et al., 2012; Lienert and Joos,
29   2018).
31   Species diversity impact ecosystem functioning and hence ecosystem services (Hooper et al., 2012; Mokany
32   et al., 2016). So far, however, integrated modelling of ecosystem processes and biodiversity across multiple
33   trophic levels and food webs is in its infancy (Harfoot et al., 2014). Whether or not enhanced integration of
34   state, function, and functional diversity across multiple trophic levels in models will markedly alter
35   projections of how ecosystems respond to climate change thus remains an open research question.
37   Beyond simulating dynamically biome shifts and carbon cycling, which are important aspects of climate
38   regulation, DGVMs can also provide information on a number of variables closely linked to other ecosystem
39   services, such as water availability, air quality or food provisioning (Krause et al., 2017; Rabin et al., 2020).
40   However, they are not intended to provide a comprehensive assessment of ecosystem services. For these,
41   other approaches applied, but to date these are mostly applied on regional scales and are only weakly
42   dynamic (Ferrier et al., 2016).
44 Projected Changes Globally at the Biome Level
46   Climate change and the associated change in atmospheric CO2 levels already exacerbate other human-caused
47   impacts on structure and composition of land and freshwater ecosystems, such as land-use change, nitrogen
48   deposition, of pollution. The relative importance of these drivers for ecosystems over the coming decades
49   will likely differ between biomes, but climate change and atmospheric CO2 will be pervasive unless we
50   manage to rapidly limit fossil-fuel emissions and warming (high confidence) (Pereira et al., 2010; Warren,
51   2011; Ostberg et al., 2013; Davies-Barnard et al., 2015; Pecl et al., 2017; Ostberg et al., 2018). Global
52   vegetation and Earth system models agree on climate-change driven shifts of biome boundaries potentially
53   of hundreds of km over this century, combined with several substantial alterations that take place within
54   biomes (e.g., changes in phenology, canopy structure and functional diversity, etc.). Large discrepancies
55   exist between models and between scenarios regarding the region and the speed of change (Gonzalez et al.,
56   2010b; Pereira et al., 2010; Pecl et al., 2017), but robust understanding is emerging in that the degree of
57   impact increases in high emission and warming scenarios (high confidence) (Figure 2.9).

     Do Not Cite, Quote or Distribute                      2-76                                       Total pages: 237
     FINAL DRAFT                                          Chapter 2                      IPCC WGII Sixth Assessment Report


 4   Figure 2.9: Projected fraction of global terrestrial area that could experience a biome shift by 2100, due to climate
 5   change (filled symbols) or a combination of climate change and land use change (outline symbols), from publications in
 6   Supplementary Table 2.S.3 (Projected vulnerabilities and risks of ecosystems to biome shifts). Circle filled (Bergengren
 7   et al., 2011), square filled (Alo and Wang, 2008), diamond filled (Gonzalez et al., 2010b), triangle up filled (Scholze et
 8   al., 2006), triangle down filled (Sitch et al., 2008), triangle on side filled (Li et al., 2018), cross filled (Warszawski et
 9   al., 2013), circle outline (Ostberg et al., 2018), diamond outline (Eigenbrod et al., 2015).
12   Substantial changes in vegetation structure and ecosystem processes are already happening (see section 2.4).
13   Many of these observations have already been projected to take place as early as at least IPCC AR3
14   (Rosenzweig et al., 2007), and can they now be increasingly tested for their robustness with observational
15   evidence. These multiple changes in response to warming (and changes in precipitation and increasing
16   atmospheric CO2 levels that go hand-in-hand with the warming) are further expected for already relatively
17   small additional temperature increases, in particular in cold (boreal, tundra) regions, as well as in dry regions
18   (high confidence): alterations of 2‒47% of the areal extent of terrestrial ecosystems in scenarios of <2°C
19   warming above pre-industrial have been projected, increasing drastically with higher-warming scenarios
20   (Warren, 2011; Wårlind et al., 2014). More recent work, applying also probabilistic methods confirm the risk
21   of drastic changes in vegetation cover (e.g. forest to non-forest or vice versa) at the end of the 21st century
22   even for ca. 2°C warming scenarios, especially in tundra, and in tropical forest and savannah regions, with
23   more subtle changes (within a given biome types) likely to occur in all regions (Ostberg et al., 2013; Ostberg
24   et al., 2018). Model studies have found 5‒20% of terrestrial ecosystems affected by warming ca. 2‒3oC,
25   increasing to above one-third at a warming of 4‒5oC (Ostberg et al., 2013; Warszawski et al., 2013).
27   In general, vegetation types are projected to be moving into their 'neighbouring' climates, depending on
28   whether temperature or precipitation is expected to be the predominant factor, and how vegetation interacts
29   with the increasing CO2 levels in the atmosphere (Wårlind et al., 2014; Scheiter et al., 2015; Schimel et al.,
30   2015; Huntzinger et al., 2017). For instance, boreal or temperate forest vegetation is simulated to migrate
31   polewards, closed tropical (moist) forest is expected to transition towards dry tropical forest types, while
32   climate-driven degradation might expand arid vegetation cover (Sections However,
33   'novel ecosystems', that is, communities with no current or historical equivalent because of the novel
34   combinations of abiotic conditions under climate change, are expected to be increasingly common in the
35   future (medium confidence) although the regions where these novel ecosystems might emerge are still
36   disputed (Reu et al., 2014; Radeloff et al., 2015; Ordonez et al., 2016). The possibility of these novel
37   ecosystems and the communities that live within them poses challenges to current modelling of ecosystem
38   shifts, and will require new approaches to conservation that are designed to adapt to rapid changes in species
39   composition and ensuing conservation challenges.

     Do Not Cite, Quote or Distribute                            2-77                                           Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 2 Risk to Arid Regions
 4   Shifts in arid system structure and functioning that have been observed to date (section are projected
 5   to continue and include widespread woody plant encroachment, notably in savanna systems in Africa,
 6   Australia and South America, and attributed to interacting land use change, climate change, and CO2
 7   fertilisation effects (Fensholt et al., 2012a; Fang et al., 2017; Stevens et al., 2017). Arid Mongolian Steppe
 8   grassland did not respond to experimentally elevated CO2 (Song et al., 2019). Woody encroachment is
 9   projected to continue or not reverse in North American drylands (Caracciolo et al., 2016), and in southern
10   African arid ecosystems (Moncrieff et al., 2014b). Dryland woody encroachment may increase carbon
11   stocks, depending on emissions scenario (Martens et al., 2021), but reduce soil water and biodiversity of
12   grassland-dependent species diversity (Archer et al., 2017). Warm season (C4) grass expansion into arid
13   shrublands risks sudden ecosystem transformation due to introduced wildfire (Bradley et al., 2016), a risk
14   anticipated for grass-invaded desert ecosystems of Australia and south-western United States (Horn and St.
15   Clair, 2017). Novel fire regimes in grassy shrublands have enhanced grass cover locally in southern African
16   Nama-Karoo (du Toit et al., 2015).
18   Range retractions are projected for endemic plants in southern Africa (Young et al., 2016) and dry
19   woodlands in Morocco (Alba-Sánchez et al., 2015). Increasing thermal stress is projected to increase woody
20   plant mortality in Sonoran Desert ecosystems (Munson et al., 2016), and facilitate perennial grass
21   replacement by xeric shrubs in the south-western USA (Bestelmeyer et al., 2018). Ecological effects may
22   occur rapidly when extreme events compound long-term trends (Hoover et al., 2015), but evolve more
23   slowly as opportunity costs accumulate due to warming (Cross-Chapter Paper 3; Cunningham et al., 2021).
25   Risk to Mediterranean-type Systems
27   All Mediterranean Type Ecosystems (MTEs) show high confidence in projected increases in the intensity
28   and frequency of hot extremes and decreases in the intensity and frequency of cold extremes and medium
29   confidence in increasing ecological drought due to increased evapotranspiration (all regions) and reduced
30   rainfall (excluding California, where model agreement is low) (see Chapter 11 of WG1). Projections also
31   show a robust increase in the intensity and frequency of heavy precipitation in the event of 2° C warming or
32   more for MTEs in South Africa, the Mediterranean Basin and California, but are less clear for Australia and
33   Chile (see Chapter 11 of WG1).
35   MTEs are characterised by the distinctive seasonal timing of precipitation and temperature and disruption of
36   this regime is likely to be critical for their maintenance. Unfortunately, projections of changes in rainfall
37   seasonality have received less attention and are far more uncertain than many other aspects of climate
38   change (Pascale et al., 2016; Breinl et al., 2020), limiting our ability to predict the ecological consequences
39   of climate change in MTEs. Responses to experimental manipulation of rainfall seasonality show potential
40   for shifts in plant functional composition and diversity loss, but results vary with soil type (van Blerk et al.,
41   2021).
43   Unfortunately, global and regional scale dynamic vegetation models show poor performance for large areas
44   of MTEs, because they do not characterise shrub and CAM-photosynthetic plant functional types well
45   (Moncrieff et al., 2015). Furthermore, the grain of these models are too coarse for quantifying impacts to
46   many vegetation formations which are patchy or of limited extent (e.g. forests). There is high confidence that
47   observations of high mortality in trees and other growth forms, reduced reproductive and recruitment
48   success, range shifts, community shifts towards more thermophilic species, and type conversions are set to
49   continue, either due to direct climate impacts through drought and other extreme weather events, or their
50   interaction with factors like fire and pathogens (Sections;;;;
52   Fire is a key driver across most MTEs due to summer-dry conditions. Climate projections for the MTEs
53   translate into high confidence that periods of low fuel moisture will become more severe and prolonged and
54   that episodes of extreme fire weather will become more frequent and severe (see Chapter 11, Section 8.3
55   WGI). This will lead to the birth of novel fire regimes in MTEs characterised by an increase in the
56   probability of greater burned area and extreme wildfire events (e.g. megafires), with associated loss of
57   human life and property, and long-term impacts on ecosystems and accelerating the possible loss of

     Do Not Cite, Quote or Distribute                       2-78                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   resilience and capacity to recover (Abatzoglou and Williams, 2016; González et al., 2018; Boer et al., 2020;
 2   Moreira et al., 2020; Nolan et al., 2020; Duane et al., 2021; Gallagher et al., 2021).
 4   Fire is virtually certain to have additional impacts through compound events (AR6 WGI Chapter 11.8).
 5   Extreme postfire weather is extremely likely to continue to impact diversity (Slingsby et al., 2017), retard
 6   vegetation regrowth (Slingsby et al., 2020a) and accelerate vegetation shifts (Batllori et al., 2019). Any
 7   increases in the intensity and frequency of heavy precipitation are highly likely to compromise soil stability
 8   in recently burnt areas (Morán-Ordóñez et al., 2020). The impacts of fire often depend on interactions with
 9   non-climatic factors such as habitat fragmentation (Slingsby et al., 2020b), management (Steel et al., 2015)
10   or the spread of flammable exotic plantation forestry and invasive species (Kraaij et al., 2018; McWethy et
11   al., 2018). Managing these factors provides opportunities for adaptation and mitigation (Moreira et al.,
12   2020).
14   Human adaptation and mitigation responses to climate change may create additional threats to MTEs. MTEs
15   have dry summers by definition, posing a challenge for the year-round supply of water to growing human
16   populations and agriculture. With recent major droughts in all MTEs (Section, there is increasing
17   reliance on groundwater for bulk water supply (Kaiser and Macleod, 2018). The majority of groundwater
18   systems have exceeded or are rapidly approaching their environmental flow limits (de Graaf et al., 2019),
19   threatening human populations and ecosystems that depend on these systems for their persistence through
20   unfavourable climatic conditions (McLaughlin et al., 2017 Plants). Similarly, much of the MTEs are open
21   shrublands and grasslands and proposed extensive tree-planting to sequester atmospheric CO2 could result in
22   loss of biodiversity and threaten water security (Doblas-Miranda et al., 2017; Bond et al., 2019).
24 Risk to Grasslands and Savannas
26   Worldwide, woody cover is increasing in savannas (Buitenwerf et al., 2012; Donohue et al., 2013; Stevens et
27   al., 2017), as a result interactions of elevated CO2 combined with altered fire and herbivory impacts (i.e.
28   from land-use change; see Section; CCP3.2; Venter et al., 2018; Wu et al., 2021). In some regions,
29   altered climate may also contribute (CCP3.2). Elevated CO2 benefits plants with C3 photosynthesis (often
30   woody plants), more than C4 species (Moncrieff et al., 2014a; Scheiter et al., 2015; Knorr et al., 2016a).
31   Increases in woody vegetation in grassy ecosystems could provide some carbon increase (medium
32   confidence) (Zhou et al., 2017; Mureva et al., 2018), but is expected to decrease biodiversity (Smit and Prins,
33   2015; Abreu et al., 2017; Andersen and Steidl, 2019), decrease water availability (Honda and Durigan, 2016;
34   Stafford et al., 2017) and alter ecosystem services like grazing and wood provision (high confidence)
35   (Anadon et al., 2014).
37   The relative importance of climate, disturbance (e.g. fire/herbivory) and plant feedbacks in shaping present
38   and future savanna distribution vary between continents (Lehmann et al., 2014), which makes projections of
39   changing biome extent challenging (Moncrieff et al., 2016). It has been shown that simulation studies that do
40   not account of CO2 interactions and only consider climate change impacts do not realistically capture the
41   future distribution of savannas (high confidence) (Higgins and Scheiter, 2012; Moncrieff et al., 2016;
42   Scheiter et al., 2020). Due to the continued strong effect of CO2 on tree (and shrub) to grass ratios in future,
43   models suggest a loss of savanna extent and conversion into closed canopy forest/thicket and an expansion
44   of savanna-type vegetation into arid grasslands (Wårlind et al., 2014; Moncrieff et al., 2016). In arid
45   savannas and their interface to grasslands, survival of woody vegetation (which may be stimulated to grow
46   by increasing CO2) will depend l on their capacity to survive potentially more severe and frequent droughts
47   (Sankaran and Staver, 2019). Across a range of models, for RCP4.5 future climate change and CO2
48   concentrations, savanna expanse declined by around 50% (converting to closed canopy systems) by 2070 in
49   Africa and South America, 25% in Asia with small changes in Australia (Moncrieff et al., 2016; Kumar et
50   al., 2021). Future fire spread is expected to be reduced with increased woody-dominance (Scheiter et al.,
51   2015; Knorr et al., 2016b; Scheiter et al., 2020), feeding back to further increase tree to grass ratios (high
52   confidence).
54   Like the tropical forest biome, savannas are at large risk, given the projected climate changes in combination
55   with land-use change (see Cross Chapter Paper 3). About 50% of Brazilian Cerrado has been transformed to
56   agriculture and pastures (Lehman and Parr, 2016), and African savannas have been proposed to follow a
57   similar tropical agricultural revolution pathway in order to enhance agronomical prosperity (Ryan et al.,

     Do Not Cite, Quote or Distribute                      2-79                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   2016). In fact, indirect climate change impacts arising from mitigation efforts on land may be particularly
 2   perilous to savannas: extensive tree-planting to restore ecosystems and remove CO2 from the atmosphere, as
 3   pledged, for example, under the African Forest Restoration Initiative, could lead to carbon losses, loss of
 4   biodiversity and damage ecosystem’s water balance if trees are planted in what naturally are grasslands or
 5   savannas (Box 2.2, FAQ2.6; Bond et al., 2019).
 7 Risk to Tropical Forest
 9   Key factors affecting the future distribution of tropical humid and dry forests are amounts and seasonalities
10   of precipitation, increased temperatures, prolonged droughts and droughted-moderated fires (robust
11   evidence, high agreement) (Bonai et al., 2016; Corlett, 2016; Lyra et al., 2017; Anderson et al., 2018; da
12   Silva et al., 2018; Fontes et al., 2018; O'Connell et al., 2018; Aguirre-Gutiérrez et al., 2019; Bartlett et al.,
13   2019; Brando et al., 2019; Stan and Sanchez-Azofeifa, 2019). Probability of severe drought is projected to
14   quadruple in natural areas in Brazil with above 2°C warming (Barbosa and Lakshmi Kumar, 2016; Marengo
15   et al., 2020). Most multi-model studies assuming rapid economic growth/business-as-usual scenarios (A2,
16   A1B, RCP8.5) show an increase in future woody biomass and areas of woody cover towards the end of the
17   21st century in the temperate regions (Boit et al., 2016; Nabuurs et al., 2017) and in tropical forests in East
18   Africa (Ross et al., 2021) but decrease in the remaining tropical regions (Anadón et al., 2014; Boit et al.,
19   2016; Lyra et al., 2017; Nabuurs et al., 2017; Maia et al., 2020). Terrestrial species are predicted to shift to
20   cooler temperatures and higher elevations (Pecl et al., 2017). Tropical species are more susceptible to
21   climate warming than temperate species (Rehm and Feeley, 2016; Sentinella et al., 2020). This susceptibility
22   will be exacerbated by road-building increasing ease of access into forests (Brinck et al., 2017; Taubert et al.,
23   2018; Bovendorp et al., 2019; Senior et al., 2019). Furthermore, most tropical cloud forest species are unable
24   to invade grasslands and this will increase risk of extinctions in tropical cloud forests (Rehm and Feeley,
25   2015).
27   Sea level rise as the result of climate change is likely to influence mangroves in all regions, with greater
28   impact on North and Central America, Asia, Australia, and East Africa than West Africa and South America
29   (robust evidence, high agreement) (Alongi, 2015; Ward et al., 2016). On a small scale, mangroves are
30   potentially moving landward (Di Nitto et al., 2014), while on a large scale they will continue to expand
31   poleward (Alongi, 2015).
33   Most simulations predict a significant geographical shifts of transition areas between tropical forests and
34   savanna in the tropical and subtropical Americas and Himalayas (Anadón et al., 2014) (Rashid et al., 2015).
35   Forest die-back, as postulated for the Amazon region, does not occur in the majority of simulations (Malhi et
36   al., 2009; Poulter et al., 2010; Rammig et al., 2010; Higgins and Scheiter, 2012; Huntingford et al., 2013;
37   Davies-Barnard et al., 2015; Sakschewski et al., 2016; Wu et al., 2016a). Model projections of future
38   biodiversity in tropical forests are rare. Arguably, species are most vulnerable to climate change effects in
39   higher altitudes or at the dry end of tropical forest occurrence (medium evidence, medium agreement)
40   (Krupnick, 2013; Nobre et al., 2016; Trisurat, 2018). Tropical lowlands are expected to lose plant species as
41   temperatures rise above species’ heat tolerance but could also generate novel communities of heat tolerant
42   species (robust evidence, high agreement) (Colwell et al., 2008; Trisurat et al., 2009; Trisurat et al., 2011;
43   Krupnick, 2013; Zomer et al., 2014a; Zomer et al., 2014b; Sullivan et al., 2020; Pomoim et al., 2021).
45   Statistical models that correlate data on species abundance with information on human pressures (such as
46   land-use change (Srichaichana et al., 2019), population density (Leclère et al., 2020) hunting (Mockrin et al.,
47   2011) found for tropical and sub-tropical forests that birds, invertebrates, mammals and reptiles show a
48   decline in their probability of presence with declining forest cover, which is particularly pronounced in forest
49   specialists or narrow-ranged species (Newbold et al., 2014). Different soil fauna groups showed different
50   responses in abundance and diversity to climate change conditions (Coyle et al., 2017; Facey et al., 2017) but
51   these changes can impact decomposition rates and biogeochemical cycles (medium evidence, low
52   agreement).
54   Invasive plant species are predicted to expand upward by 500-1,500m in the Western Himalaya (Thapa et al.,
55   2018), and by 6-35% per year from the current extent in South America (robust evidence, high agreement)
56   (Bhattarai and Cronin, 2014). Global assessment (Wang et al., 2017) also revealed that ecoregions of high
57   elevation tropical forests and sub-tropical coniferous forests have high risk of invasive plant expansion in the

     Do Not Cite, Quote or Distribute                      2-80                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   low CO2 emission scenarios, with negative impacts on ecosystem functioning and local livelihoods (Shrestha
 2   et al., 2019).
 4   The impact of unsustainable land use on tropical forests continues in all regions (see Cross-Chapter Paper 7).
 5   Projected climate changes will not only cause impacts on biodiversity but also on the livelihoods of affected
 6   people (robust evidence, high agreement). Increased drought drives crop failures that cause local
 7   communities to expand agricultural area by further clearing native forests (Desbureaux and Damania, 2018).
 8   Climate change is projected to enlarge the area of suitability for booming tree crops such as oil palm, acacia,
 9   Eucalyptus, and rubber (Koninck et al., 2011; Cramb et al., 2015; Nath, 2016; Hurni et al., 2017; Li et al.,
10   2017; Varkkey et al., 2018). An increase of 8% in area of rubber plantations in Yunnan province, China,
11   between 2002–2010 to higher altitude due to decreased environmental limits, potentially increases pressure
12   on remaining biodiversity both within and outside of protected areas (Zomer et al., 2014a). As a
13   consequence, the suitable area for mammals is projected to be reduced by 47.7% (RCP 2.6) and 67.7%
14   (RCP8.5) by 2070, with large variability depending on the different species (See also Cross-Chapter Paper 7;
15   Brodie, 2016).
17 Risk to Boreal and Temperate Forests
19   As in the Arctic, warming substantially exceeding the global average has already been observed for the
20   northern parts of the temperate and boreal forest zone (Gauthier et al., 2015), and is projected to continue
21   (see Chapter 4 of WGI, see Cross-Chapter Paper 6). As a consequence, boreal tree species are expected to
22   move northwards (or in mountain regions: upwards) into regions dominated by tundra, unless constrained by
23   edaphic features, and temperate species are projected to grow in regions currently occupied by southern
24   boreal forest (high confidence). In both biomes, deciduous trees are simulated to increasingly grow in
25   regions currently dominated by conifers (Wårlind et al., 2014; Boulanger et al., 2017). These simulation
26   results have been supported by observational examples. In Eastern Siberia, fire disturbance of larch-
27   dominated forest was followed by recovery to birch-dominated forest (Stuenzi and Schaepman-Strub, 2020).
28   In Alberta Lodgepole Pine (Pinus contorta) lost its dominant status after attacks by Mountain Pine Beetles
29   (Dendroctonus ponderosae) caused the canopy to switch to non-pine conifers and broadleaved trees
30   (Axelson et al., 2018). In contrast to the examples above, some boreal forests have proven resilient to
31   disturbances, including to recent unprecedented spates of insect attacks (Campbell et al., 2019a; Prendin et
32   al., 2020).
34   Reforestation, either natural or anthropogenic, leads to summer cooling and winter warming of the ground,
35   while forest thinning or removal by fire has the reverse effects and deepens the upper layer free of
36   permafrost (Stuenzi et al., 2021a). Interactions between permafrost and vegetation are important. For
37   example, trees in East Siberian taiga obtained water mostly from rain in wet summers and mostly from
38   permafrost meltwater in dry summers (Sugimoto et al., 2002), suggesting that these forests will be
39   particularly vulnerable to combination of drought with retraction further underground of permafrost under
40   climate warming.
42   Risk to Peatland Systems
44   The overall effect of climate change on the extent of northern peatlands is still debated (limited evidence, low
45   agreement). It is expected that climate change will drive high-latitude peatland expansion poleward of their
46   present distribution due to warming, permafrost degradation, and glacier retreat, which could provide new
47   land and conditions favourable for peat development (limited evidence, medium agreement) (Zhang et al.,
48   2017b), as seen during the last deglacial warming (robust evidence, high agreement) (MacDonald et al.,
49   2006; Jones and Yu, 2010; Ratcliffe et al., 2018). Peatland area loss (shrinking) near the southern limit of
50   their current distribution or in areas where the climate becomes unsuitable is also expected (medium
51   evidence, medium agreement) (Section; Finkelstein and Cowling, 2011 temperature, and; Gallego-
52   Sala and Colin Prentice, 2013; Schneider et al., 2016; Müller and Joos, 2020), though these peatlands could
53   persist if moisture is maintained via peatlands' self-regulating capacity. In Western Canada, a study suggests
54   that peatlands may persist until 2100, even though the climate will be less suitable (Schneider et al., 2016).
55   Simulations suggest that climate change driven increases in temperature and atmospheric CO2 could drive
56   reductions in the northern peatland area up to 18% (SSP1-2.6), 41% (SSP2-4.5), and 61% (SSP5-8.5) by
57   2300 (Müller and Joos, 2020). This is in contrast with findings of northern peatland persistence and

     Do Not Cite, Quote or Distribute                      2-81                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   expansion under RCP2.6 and RCP6.0 scenarios during 1861–2099 by another modelling study (Qiu et al.,
 2   2020). In the tropics, the only available study suggests peatland area will increase until 2300, mainly due to
 3   the CO2 fertilisation effect (Müller and Joos, 2020).
 5   The combination of climate and land-use change represents a substantial risk to peatland carbon stocks, but
 6   full assessment is impeded because peatlands are yet to be included in Earth System models (limited
 7   evidence, high agreement) (Loisel et al., 2021). It is expected that the carbon balance of peatlands globally
 8   will switch from sink to source in the near future (2020-2100), mainly because tropical peatland emissions,
 9   together with those from climate change -driven permafrost thaw, will likely surpass the carbon gain
10   expected from climate change-driven enhanced plant productivity in the northern high latitudes (Gallego-
11   Sala et al., 2018; Chaudhary et al., 2020; Turetsky et al., 2020; Loisel et al., 2021), mainly because of
12   groundwater drawdown (robust evidence, medium agreement) (Hirano et al., 2014; Brouns et al., 2015; Cobb
13   et al., 2017; Itoh et al., 2017; Evans et al., 2021). The overall northern peatland carbon sink has been
14   simulated to persist for at least 300 years under RCP2.6, but not under RCP8.5 (Qiu et al., 2020).
16   Increases in fire extent, severity, and duration are expected in all peatland regions in the future due to
17   temperature increases (Section, changes in precipitation patterns (section, and increases in
18   ignition sources (such as lightning) (Section, with associated rapid carbon losses to the atmosphere
19   (medium evidence, high agreement) (Dadap et al., 2019; Chen et al., 2021a; Nelson et al., 2021). For
20   example, drought has been linked to fires in SE Asian peatlands (Field et al., 2009) and there are predicted
21   decreases in mean summer precipitation (10-30%) for high and low RCPs, particularly over the Indonesian
22   region by mid and late twenty-first century (Section; Tangang et al., 2020; Taufik et al., 2020).
23   During wet years, the fire probability in Indonesian peatlands also significantly increases (+15-40 %) when
24   July-October temperatures surpass 0.5°C anomalies compared to a 1995-2015 baseline (Fernandes et al.,
25   2017). Overall, current evidence suggests that peat carbon losses via fire have the potential to be equal to, or
26   greater than, losses due to human peatland drainage and disturbance (limited evidence, high agreement)
27   (Turetsky et al., 2015).
29   In permafrost peatlands, studies differ, with some projecting net loss or and others net gain of carbon
30   (medium evidence, low agreement) (Estop-Aragonés et al., 2018; Hugelius et al., 2020; Loisel et al., 2021;
31   Väliranta et al., 2021). In some permafrost peatlands, prolonged and warmer growing seasons due to climate
32   change (section, along with increases in nitrogen deposition since 1850 (Lamarque et al., 2013),
33   are promoting plant primary productivity. Other studies indicate increased nitrogen-mediated sequestration
34   could be exceeded by increased decomposition due to climate-change-driven warming and fire (medium
35   evidence, low agreement) (Natali et al., 2012; Vonk et al., 2015; Keuper et al., 2017; Burd et al., 2018;
36   Estop-Aragonés et al., 2018; Gallego-Sala et al., 2018; Serikova et al., 2018; Wild et al., 2019; Chaudhary et
37   al., 2020; Hugelius et al., 2020).
39   Any climate change or human-driven degradation of peatlands will also entail losses in water storage
40   (limited evidence, high agreement) (Wooster et al., 2012 drought and; Hirano et al., 2015; Cole et al., 2019;
41   Taufik et al., 2019) and biodiversity (Harrison, 2013; Lampela et al., 2017; Renou-Wilson et al., 2019). The
42   environmental archive contained in peat that preserves records of vegetation, hydrology, climate change,
43   pollution and/or human disturbances is also lost as the peatlands degrade (Greiser and Joosten, 2018).
44   (Kasischke and Turetsky, 2006; MacDonald et al., 2006; Turunen, 2008; Field et al., 2009; Flannigan et al.,
45   2009; Jones and Yu, 2010; Kasischke et al., 2010; Peterson et al., 2010; Finkelstein and Cowling, 2011
46   temperature, and; Rooney et al., 2012; Gallego-Sala and Colin Prentice, 2013; Lamarque et al., 2013; Hirano
47   et al., 2014; Brouns et al., 2015; Turetsky et al., 2015; Miettinen et al., 2016; Schneider et al., 2016; Cobb et
48   al., 2017; Fernandes et al., 2017; Itoh et al., 2017; Gallego-Sala et al., 2018; Greiser and Joosten, 2018;
49   Ratcliffe et al., 2018; Dadap et al., 2019; Leifeld et al., 2019; Chaudhary et al., 2020; Hoyt et al., 2020;
50   Müller and Joos, 2020; Qiu et al., 2020; Tangang et al., 2020; Taufik et al., 2020; Turetsky et al., 2020; Chen
51   et al., 2021a; Evans et al., 2021; Loisel et al., 2021; Nelson et al., 2021; Qiu et al., 2021)
53 Risk to Polar Tundra Ecosystems
55   For boreal-tundra systems, AR5 projected transformation of species composition, land cover and permafrost
56   extent, decreasing albedo and increasing greenhouse gases emission (medium confidence). The Special
57   Report on Global Warming of 1.5°C classified tundra and boreal forests as particularly vulnerable to

     Do Not Cite, Quote or Distribute                      2-82                                       Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1   degradation and encroachment of woody shrubs (high confidence). The Special Report on Oceans and
 2   Cryosphere (SROCC) projected climate-related changes to arctic hydrology, wildfire and abrupt thaw, (high
 3   confidence) and broad disappearance of arctic near-surface permafrost in this century, with important
 4   consequences for global climate (very high confidence). Chapter 2 of AR6 has focused on new key findings
 5   about observed and projected changes in tundra vegetation and related hydrology, with implications for
 6   feedbacks to the climate system.
 8   Due to the rapid warming in high northern latitudes, Arctic tundra is one of the terrestrial biomes where
 9   climate change impacts are already clearly visible (Settele et al., 2014; Uboni et al., 2016). Climate models
10   project that warming for the Arctic is likely to continue at more than double the global rate. Compared to the
11   period 1995-2014, mean annual surface air temperatures in Arctic tundra are projected to increase by 7.9°‒
12   10°C by the end of the century for scenarios of high greenhouse gas emissions (RCP 7.0 and 8.5). For
13   scenarios of low greenhouse gas emissions (RCP 1.9 and 2.6), the projected increase is 2.6°‒3.2°C (see
14   Chapter 4 pf WGI). The Arctic is also projected to have among the largest increases in precipitation globally,
15   although there is high uncertainty in these projections. In contrast to climate change, land use change is
16   projected to be very low in Arctic tundra systems (van Asselen and Verburg, 2013).
18   Models of vegetation response to climate project acceleration in coming decades of observed increases in
19   shrub dominance and boreal forest encroachment that have been driven by recent warming (Settele et al.,
20   2014), leading to a shrinking of the area of tundra globally (medium confidence) (Mod and Luoto, 2016;
21   Gang et al., 2017). Simulating changes in tundra vegetation is complicated by permafrost dynamics (e.g.
22   formation of thaw ponds), changes in precipitation, or low nutrient availability, which may promote
23   abundance of graminoids (van der Kolk et al., 2016). The changes in vegetation, when combined with
24   warming and increased precipitation effects on soil thawing and carbon cycling, are projected to modify
25   greenhouse gas emissions and have biophysical feedbacks to regional and global climate. Large uncertainty
26   in modelled carbon cycle changes arises from differences between the vegetation models (Nishina et al.,
27   2015; Ito et al., 2016). In addition, climate change is expected to strongly interact with other factors, such as
28   fire, to further increase uncertainty in projections of tundra ecosystem function (Jiang et al., 2017).
30 Committed Impacts of Climate Change on Terrestrial Ecosystems and Implications of Overshoot
32   Projections point to potentially large changes of canopy structure and composition within and across the
33   terrestrial biomes in response to climate change and changes in atmospheric CO2. These changes will
34   contribute to altered ecosystem carbon uptake and losses, biophysical climate feedbacks (Sections 2.3.2;
35   2.4.4;;,, Figure 2.10, Table 2.4), and multiple other ecosystem services
36   (Sections 2.5.3, 2.5.4), as well as for biodiversity (Sections 2.4.2, 2.4.3, 2.4.4, 2.4.5,,, 2.5.2,
37   Figure Box 2.1.1, Table Box 2.1.1, Table SM2.4). Until now, most studies project changes over next decades
38   until the end of this century.
40   However, there is an increasing body of literature that has found continued, longer-term responses of
41   ecosystems to climate change, so-called 'committed changes,' that arise from lags that exist in many systems.
42   Many processes in ecosystems take more than a few decades to quasi-equilibrate to environmental changes.
43   Therefore, trends of changing vegetation cover identified in simulations of transient warming continue to
44   show up in simulations that hold climate change at low levels of warming (medium confidence) (Boulton et
45   al., 2017; Pugh et al., 2018; Scheiter et al., 2020). Such changes, which could tip ecosystems into an
46   alternative state, could also be triggered by a ‘warming overshoot’ – if global warming were to exceed a
47   certain threshold, even if mean temperatures afterwards decline again (Albrich et al., 2020a).
49   For instance, even if warming achieved by 2100 remained constant after 2100, such committed responses
50   continue to occur. These include: (1) continued Amazon forest loss (Boulton et al., 2017), consistent with
51   results in (Pugh et al., 2018) that found continued tropical forest cover loss across a range of models and
52   simulation set-ups, and (2) across Africa, an increased shift towards woody C3 vegetation was found in
53   equilibrium state, the overall response depending on the atmospheric CO2 concentration (Scheiter et al.,
54   2020). In Pugh et al. (2018), the opposite was found for boreal forest cover, which showed a strong
55   committed increase. The committed changes in vegetation composition correspond to large committed
56   changes in terrestrial carbon uptake and losses (Boulton et al., 2017; Pugh et al., 2018; Scheiter et al., 2020),
57   and would plausibly also appear in other ecosystem functioning and services. These studies point to the

     Do Not Cite, Quote or Distribute                       2-83                                        Total pages: 237
     FINAL DRAFT                                    Chapter 2                  IPCC WGII Sixth Assessment Report

 1   importance of having not only a multi-decadal but also a multi-century perspective when exploring the
 2   impacts of political decisions on climate change mitigation taken now. Even if climate-warming targets are
 3   met, published evidence so far suggests that fundamental changes in some ecosystems are likely as these
 4   correspond to well-understood ecosystem physiological responses that trigger long-term changes in
 5   composition.
 8   [START BOX 2.1 HERE]
10   Box 2.1: Assessing Past Projections of Ecosystem Change Against Observations
12   To assess future climate change impacts on ecosystems we use models to project their future distribution.
13   Comparing the trends in the observed changes against the projections can help assess the strength of the
14   model projections. In this box, we compare observed trends of changes in ecosystem structure to projections
15   highlighted in previous IPCC reports (specifically AR3 (IPCC, 2001), AR4 (Fischlin et al., 2007) and AR5
16   (Settele et al., 2014). We use this to assess how well the projections are matching up with observed changes.
17   The map represents studies documenting observed changes in common plant functional groups (e.g. trees,
18   grasses, shrubs). Studies, documenting changes in plant functional groups, were collated from published
19   papers in natural and semi-natural areas. Studies were included if climate change, or interactions between
20   climate change and land use showed a causal link to the observed change. Studies were excluded if the
21   changes only from landscape/land use transformation (e.g. deforestation). In each paper, we recorded the
22   geographical location, the type of functional change and noted the causes. Observed changes are plotted onto
23   a biome map derived from the WWF ecoregions database (Olson et al., 2001). Trends in changing plant
24   functional types are good indicators of potential biome shifts and are used to assess how observations match
25   up with projections.


     Do Not Cite, Quote or Distribute                    2-84                                      Total pages: 237
    FINAL DRAFT                                             Chapter 2                            IPCC WGII Sixth Assessment Report

1   Figure Box 2.1.1: Observed changes in the distribution of plant functional types that are caused by climate change or
2   combination of land use and climate change. Shifts in plant functional types are indicative of shift in biome function
3   and structure
6   Table Box 2.1.1: Comparison of projections on biome change from the Third, Fourth and Fifth assessment report
7   (IPCC, 2001; Fischlin et al., 2007; Settele et al., 2014) with observed changes in ecosystems as assessed in this current
8   report (see section 2.4 and Fig Box 2.1.1). Observed changes marked in bold show good agreement with past
9   projections, those in red show mismatch with observations and projections.
        Biome                 AR3                        AR4                            AR5                 Observed trends 1990-2021
    Mediterranean Increased disturbance      Loss of 65% of area due to   Range contractions of            Increased in water deficit and fire
    Type ecosystems by fire and warming      warming. Increased fire      all species                      activity (sections,
                    will cause a loss of     frequencies will favor                                        causing declines in diversity, tree
                    unique habitats          resprouting plants. An                                        mortality (Fig Box 2.1.1) with
                                             increase in grass dominance.                                  resprouting trees worst affected.
                                             Forest expansion within
                                             MTE’s systems due to                                          Increasing dominance of grasses
                                             elevated CO2.                                                 (often alien). Increasing
                                                                                                           dominance of deciduous over
                                                                                                           evergreen species (Fig Box 2.1.1).
    Tundra            Tree and shrub         Increased woody plant growth     Continued woody              Increase in woody shrub cover in
                      encroachment into      due to longer and warmer         expansion in tundra          tundra and expansion of boreal
                      tundra.                growing seasons and              regions with reduced         forest into tundra. (Fig Box 2.1.1,
                                             replacement of dwarf tundra      surface albedo due to
                                             by shrub tundra                  less snow and more
                                                                              woody cover
                                             Poleward expansion of tundra
                                             into polar desert and
                                             encroachment of coniferous
                                             trees into tundra

    Boreal forest     Reduced productivity Extensive boreal tree spread                                    Expansion into Tundra and
                      due to weather related into tundra.                                                  upslope tree line advance
                      disturbances (e.g.                                                                   (Section and Fig Box
                      increase fire risk).   Boreal forest dieback within                                  2.1.1).
                                             boreal zone and contraction
                      Deciduous broadleaf of boreal forest at southern                                     Increased mortality due to
                      tree encroachment      ecotone with continental                                      drought, fire, beetle infestations
                      into boreal forest     grasslands                                                    (Sections,,

    Tropical forest   Increasing CO2         Increases in forest              Shift in the climate         Expansion of tropical forest into
                      concentration would    productivity and biomass         envelope of moist            savannas in Africa, Asia, South
                      increase net primary   through increased CO2 with       tropical forests but         America (Section, Fig Box
                      productivity           localised decreases in the       forests are less likely to   2.1.1)
                                             Amazon. Shift in forest          undergo major
                                             species composition.             retractions or               Forest biomass increases (though
                                             Expansion of forest area into    expansions than              slowing). (Section
                                             mesic savanna.                   suggested in AR4
                                                                                                           Forest degradation from drought,
                                                                                                           warming x fire and shorter
                                                                                                           residence time of trees. (Section

                                                                                                           Shift in species composition
                                                                                                           towards species with more arid
                                                                                                           adapted trait (Section
    Temperate forest Forest decline and      Increase in tree mortality                                    Map indicates a shift towards
                     increased mortality     from drought related declines.                                deciduous species in W N America
                                                                                                           (Fig Box 2.1.1)
                                             A general increase of
                                             deciduous at the expense of                                   Tree death due to interactions
                                             evergreen vegetation is                                       with drought x pest outbreaks x
                                             predicted at all latitudes                                    fire (,,

    Do Not Cite, Quote or Distribute                               2-85                                                      Total pages: 237
     FINAL DRAFT                                           Chapter 2                         IPCC WGII Sixth Assessment Report

     Grasslands and Increasing CO2         Increased tree dominance in      Rising CO2 will           Greening and encroachment
     savannas       concentration will     savannas and grasslands          increase the likelihood   across tropical and temperate
                    increase net primary   (from elevated CO2). With C3     woodier states (but the   savannas in Africa, Asia,
                    productivity           plants benefiting more than      transition will vary in   Australia and N America
                                           C4 plants                        different                 (Section
                                                                                                      Expansion of trees into
                                                                                                      grasslands and advancement of
                                                                                                      tree lines

                                                                                                      Signs of increased C4 grass
                                                                                                      productivity in drought conditions.
                                                                                                      Increased C3 grass productivity

     Desert/ arid                          An increase in desert                                      Greening (increased LAI, woody
     shrublands                            vegetation productivity was                                cover) and increased herbaceous
                                           projected in southern Africa,                              production is occurring at desert
                                           the Sahel, central Australia,                              grassland interfaces (Chapter –
                                           the Arabian Peninsula and                                  Cross-Chapter Paper 3).
                                           parts of central Asia due to a
                                           positive impact of rising
                                           atmospheric CO2.
 3   Assessment: There is a high agreement between observations and projections of tree death in temperate and
 4   boreal forests, with current projections (AR6) indicating this trend will continue (Section 2.5.4). Forest death
 5   is most widely recorded in central Europe and Westerns North America (Fig Box 2.1.1). There is also very
 6   high agreement between observations and projections of woody encroachment in savannas, grasslands and
 7   tundra, with projections also indicating that this trend is likely to continue (Section 2.5.4). Observations of
 8   desert greening show good agreement with earlier projections. Patterns of desertification are also occurring
 9   although the geographical match between projections and observations shows moderate agreement, likely
10   due to the strong role of land use in this process. Projections of tropical forest expansion into mesic savannas
11   and boreal forest expansion into tundra also shows agreement with observations.
13   Projections on the future of Mediterranean shrublands, deserts, xeric shrublands and temperate grassy
14   systems are limited making assessment of this relationship less clear. It is also unclear, due to limited
15   observations, how widespread a shift from deciduous forest species to evergreen forest species is. Some
16   observations suggest this is occurring although it is not clear how widespread this change is and if the
17   geographical pattern is as projected.
19   [END BOX 2.1 HERE]
22   2.5.3     Risk Assessment of Ecosystems and Related Services
24    Risks in Protected Areas
26   National parks and other protected areas, which, in June 2021, covered 15.7% of global terrestrial area
27   (UNEP-WCMC, 2021), conserve higher biodiversity than adjacent unprotected areas (Gray et al., 2016), and
28   protect one-fifth of global vegetation carbon stocks and one-tenth of global soil carbon stocks (Section
29 This section assesses climate change specifically in protected areas. Even though it is in a part of
30   the chapter on projected risks, this section includes both observed exposure and projected risks to gather the
31   chapter information on protected areas in one place.
33 Observed exposure of protected areas
34   Deforestation, agricultural expansion, overgrazing, and urbanisation exposed to intense human pressure one-
35   third of global protected area (6 million km2) in 2009, a 6% increase from 1993 (Venter et al., 2016; Jones et
36   al., 2018). The observed change in exposure to climate change has not yet been quantified for protected areas
37   globally but research has analysed spatial patterns and magnitudes of observed changes for the 360 000 km2

     Do Not Cite, Quote or Distribute                             2-86                                                 Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   system of U.S. national parks (Gonzalez et al., 2018), including the first national park in the world. From
 2   1895 to 2010, mean annual temperature of the U.S. national park area increased at a rate of 1 ± 0.2ºC
 3   century-1, double the rate of the U.S. as a whole, and precipitation decreased on 12% of the national park
 4   area, compared with 4% for the U.S. as a whole, due to a high fraction of U.S. national park area in the
 5   Arctic, at high elevations, and the arid Southwestern USA. (Gonzalez et al., 2018). In addition, analyses of
 6   weather station measurements in and near six South African National Parks found that maximum
 7   temperature increased at a rate of 0.024 ± 0.003ºC y-1 from 1960 to 2010 (Van Wilgen et al., 2016). While a
 8   substantial fraction of global protected area has been exposed to observed human land cover change, the
 9   global exposure to observed climate change is unquantified.
11 Projected risks in protected areas
12   Under a climate change scenario of ~3.5ºC temperature increase by 2070, current climate could disappear
13   from individual protected areas that comprise half of global protected area and novel climates (climate
14   conditions that are currently not present within an individual protected area) could expose half of global
15   protected area (Hoffmann et al., 2019b). A lower emissions scenario of ~1.5ºC could reduce the climate
16   disappearance to 40% and the exposure to novel climates to 41% (Hoffmann et al., 2019b). Models project
17   the highest projected exposure in subtropical projected areas (Hoffmann and Beierkuhnlein, 2020). Projected
18   disappearance of current climate conditions from protected areas is most extensive in Africa, Oceania, and
19   North and South America (Elsen et al., 2020).
21   Projections indicate higher exposure to novel climates of tropical rainforests, shrublands, and grasslands,
22   temperate conifer forests and grasslands, and tundra (Hoffmann et al., 2019b; Elsen et al., 2020). A climate
23   change scenario of ~3.5ºC temperature increase by 2100 could expose 32% of protected area in humid
24   tropical forests, 1.6 million km2 in 2000, to climate that would be novel to humid tropical forest protected
25   areas, while climate currently present in humid tropical forest protected areas could disappear from 0.6
26   million km2, 12% of current total area by 2050 (Tabor et al., 2018). High deforestation and climate change
27   combined could expose 2% of the humid tropical forest protected area (Tabor et al., 2018). Regional
28   analyses under RCP8.5 also project substantial disappearance of current climate from protected areas in
29   Bolivia, Chile, and Peru (Fuentes-Castillo et al., 2020), Canada, Mexico, and the U.S. (Batllori et al., 2017;
30   Holsinger et al., 2019), China (Zomer et al., 2015), Europe (Nila et al., 2019), and Indonesia (Scriven et al.,
31   2015). Projected climate change could expose an extensive part of global protected area to disappearing and
32   novel climate conditions (high confidence) (Cross-Chapter Paper 1).
34   Continued climate change increases risks to individual species and vegetation types in protected areas. Under
35   a climate change scenario of 4ºC temperature increase by 2100, suitable climate for two species of baobab
36   trees (Adansonia perrieri, A. suarezensis) in Madagascar could shift entirely out of the protected areas
37   network (Vieilledent et al., 2013). Other species and vegetation types at risk of partial disappearance of
38   suitable climate from protected areas include Atlantic Forest amphibians in Brazil (Lemes et al., 2014), birds
39   in Finland (Virkkala et al., 2013), birds and trees in Canada and Mexico (Stralberg et al., 2020), bog
40   woodlands in Germany (Steinacker et al., 2019), butterflies and mammals in Egypt (Leach et al., 2013), and
41   tropical dry forests in Mexico (Prieto-Torres et al., 2016). Projected disappearance of suitable climate
42   conditions from protected areas increase risks to the survival of species and vegetation types of conservation
43   concern in tropical, temperate, and boreal ecosystems (high confidence) (Cross-Chapter Paper 1).
45   Protected rivers, lakes, and other freshwater protected areas require inter-catchment connectivity to maintain
46   species and population movements (Bush et al., 2014a; Hermoso et al., 2016; Thieme et al., 2016), but dams
47   and other barriers interrupt connectivity (Grill et al., 2019). Climate change could also reduce freshwater
48   connectivity (Section Globally, over two-thirds of river reaches (by length) lack protected areas in
49   their upstream catchments and nine-tenths of river reaches (by length) do not achieve full integrated
50   protection (Abell et al., 2017).
52   Terrestrial and freshwater protected areas can also serve as climate change refugia, locations where suitable
53   conditions may persist for the species into the future (e.g. Section In Canada, Mexico, and the U.S.,
54   only a fraction of protected area is located in potential climate change refugia under a 4ºC temperature
55   increase, estimated at 4% (Michalak et al., 2018) to 7% (Batllori et al., 2017). Potential refugia from biome
56   shifts due to climate change under temperature increases of 1.8-3.4ºC cover <1% of the U.S. national park
57   area (Gonzalez et al., 2010b), a fraction that reduces to near zero when climate change is combined with

     Do Not Cite, Quote or Distribute                      2-87                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   habitat fragmentation due to land use change (Eigenbrod et al., 2015). Protected areas in boreal ecosystems
 2   could serve as refugia for species shifting north in Canada (Berteaux et al., 2018) and Finland (Lehikoinen et
 3   al., 2019). Invasive species, habitat loss, and other disturbances in protected areas could be lower than in
 4   unprotected areas across Europe (Gallardo et al., 2017) and specifically in Spain (Regos et al., 2016) and in
 5   Sri Lanka (Kariyawasam et al., 2020). Protected areas conserve refugia from climate change under a
 6   temperature increase of 4ºC, important for biodiversity conservation but limited to <10% of current protected
 7   area (medium confidence).
 9   Risks to Ecosystems and Services from Wildfire
11 Future projections of wildfire globally
12   Continued climate change under high emissions scenarios that increase global temperature ~4ºC by 2100
13   could increase global burned area 50% (Knorr et al., 2016b) to 70% (Kloster and Lasslop, 2017) and global
14   mean fire frequency ~30% (Gonzalez et al., 2010b), with increases on one-third (Gonzalez et al., 2010b) to
15   two-thirds (Moritz et al., 2012) of global land and decreases on one-fifth (Gonzalez et al., 2010b; Moritz et
16   al., 2012). Lower emissions that would limit the global temperature increase to <2ºC would reduce projected
17   increases of global burned area to 30% (Lange et al., 2020) to 35% (Kloster and Lasslop, 2017) and
18   projected increases of fire frequency to ~20% (Gonzalez et al., 2010b; Huang et al., 2015). Continued
19   climate change could further lengthen fire weather seasons (IPCC AR6 WGI Chapter 12). Models that
20   combine projected climate change with potential agricultural expansion project decreases in total burned area
21   (Huang et al., 2015; Knorr et al., 2016b; Park et al., 2021). The area of projected increases in burned area
22   and fire frequency due solely to continued climate change is higher for the world as a whole than the area of
23   projected decreases (medium evidence, medium agreement).
25   Increased wildfire due to continued climate change increases risks of tree mortality (Sections,,
26, biome shifts (Section, and carbon emissions (Sections, Wildfire and
27   biome shifts under projected climate change of 4º C above the pre-industrial period, combined with
28   international trade and transport, cause high risks of invasive species across one-sixth of global area,
29   including extensive high-biodiversity regions (Early et al., 2016).
31   Wildfire risks to people include death and destruction of homes, respiratory illnesses from smoke (Ford et
32   al., 2018; Machado-Silva et al., 2020), post-fire flooding from areas exposed by vegetation loss, and
33   degraded water quality through increases in sediment flows (Dahm et al., 2015) and chemical precursors of
34   carcinogenic trihalomethanes when water is later chlorinated for drinking (Section; Uzun et al., 2020)
35   . Under RCP8.5 and shared socio-economic pathway SSP3 (high population growth, slow urbanisation), the
36   number of people living in fire-prone areas could increase by three-quarters, to 720 million people in 2100,
37   in a projected global population of 12.4 billion people (Knorr et al., 2016b). Lower emissions under RCP4.5
38   could reduce the number of people at risk by 70 million people. In these projections, human population
39   growth increases human exposure to wildfires more than increase in burned area (Knorr et al., 2016c). A
40   global temperature increase <2ºC could increase global population exposure to wildfire by ~30% (Lange et
41   al., 2020). Increased wildfire under continued climate change increases probabilities of human exposure to
42   fire and risks to public health (medium evidence, high agreement).
44 Future projections of wildfire in high-risk areas
45   Regions identified at high risk of increased burned area, fire frequency, or fire weather by multiple global
46   analyses include: Amazon (Gonzalez et al., 2010b; Huang et al., 2015; Knorr et al., 2016c; Burton et al.,
47   2018; Abatzoglou et al., 2019), Mediterranean Europe (Gonzalez et al., 2010b; Burton et al., 2018;
48   Abatzoglou et al., 2019), Arctic tundra (Moritz et al., 2012; Flannigan et al., 2013), western Australia
49   (Gonzalez et al., 2010b; Burton et al., 2018; Abatzoglou et al., 2019), western United States (Gonzalez et al.,
50   2010b; Moritz et al., 2012; Knorr et al., 2016c). Higher-resolution spatial projections indicate high risks of
51   increased wildfire in the Amazon, Australia, boreal ecosystems, Mediterranean Europe, and the United
52   States under climate change (medium evidence, medium agreement).
54   In the Amazon, climate change under RCP8.5, combined with high deforestation, could double the area of
55   high fire probability (Fonseca et al., 2019), double burned area by 2050 (Brando et al., 2020) increase burned
56   area 400‒2800% by 2100 (Le Page et al., 2017), and increase fire intensity 90% (De Faria et al., 2017).
57   Lower greenhouse gas emissions (RCP4.5) and reduced deforestation could reduce fire risk to a one-fifth

     Do Not Cite, Quote or Distribute                      2-88                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   increase in the area of high fire probability (Fonseca et al., 2019) and a 100‒500% increase in burned area by
 2   2100 (Le Page et al., 2017). Moreover, increased fire, deforestation, and drought, acting through vegetation-
 3   atmosphere feedbacks, increase risks of extensive forest dieback and potential biome shifts of up to half of
 4   Amazon rainforest to grassland, a tipping point that could release an amount of carbon that would
 5   substantially increase global emissions (Oyama and Nobre, 2003; Sampaio et al., 2007; Lenton et al., 2008;
 6   Nepstad et al., 2008; Malhi et al., 2009; Settele et al., 2014; Lyra et al., 2016; Zemp et al., 2017a; Zemp et
 7   al., 2017b; Brando et al., 2020). Continued climate change, combined with deforestation, increases risks of
 8   wildfire and extensive forest dieback in the Amazon rainforest (robust evidence, high agreement).
10   In Australia, climate change under RCP8.5 increases risks of pyroconvective fire by 20 to 40 days in
11   rangelands of Western Australia, South Australia, and the Northern Territory (Dowdy et al., 2019).
12   Pyroconvective fire conditions could reach more frequently into the more populated areas of New South
13   Wales, particularly at the start of austral summer (Di Virgilio et al., 2019). General circulation models do not
14   agree, however, on projected areas of fire increase in New South Wales (Clarke and Evans, 2019). Increases
15   in heat and potential increases in wildfire threaten the existence of temperature montane rainforest in
16   Tasmania, Australia (Mariani et al., 2019).
18   In Mediterranean Europe, climate change of 3ºC could double or triple burned area, while keeping the
19   temperature increase to 1.5ºC could limit burned area increase to 40‒50% (Turco et al., 2018). Under
20   RCP8.5, the frequency of heat-induced fire weather could increase 30% (Ruffault et al., 2020). Severe fire
21   followed by drought could cause biome shifts of forest to non-forest (Batllori et al., 2019) and tree mortality
22   >50% (Dupire et al., 2019).
24   In Arctic tundra, boreal forests, northern peatlands, including permafrost areas, climate change under
25   scenarios of 4ºC temperature increase could triple burned area in Canada (Boulanger et al., 2014), double the
26   number of fires in Finland (Lehtonen et al., 2016), increase lightning-driven burned area 30 to 250%
27   (Veraverbeke et al., 2017; Chen et al., 2021a), push half of the area of tundra and boreal forest in Alaska
28   above the burning threshold temperature, and double burned area in Alaska (Young et al., 2017a). Thawing
29   of Arctic permafrost from a projected temperature of 4ºC and resulting wildfire could release 11-200 Gt
30   carbon that could substantially exacerbate climate change (Section
32   In the United States, climate change under RCP8.5 could increase burned area 60-80% by 2049 (Buotte et
33   al., 2019) and the number of fires with an area >50 km2 by 300‒400% by 2070 (Barbero et al., 2015). In
34   montane forests in the U.S., climate change under RCP8.5 increases the risk of fire-facilitated conversion of
35   ~7% of forest to non-forest by 2050 (Parks et al., 2019). In California, climate change under a scenario of
36   4ºC temperature increase could double fire frequency in some areas (Mann et al., 2016), but emissions
37   reductions that limit the temperature increase to ~2ºC could keep fire frequency from increasing (Westerling
38   et al., 2011). Carbon dioxide fertilisation and increased temperature under climate change could increase
39   invasive grasses and wildfire in desert ecosystems of the Southwestern United States, where wildfire has
40   historically been absent or infrequent, and increase mortality of the sparse tree cover (Horn and St. Clair,
41   2017; Klinger and Brooks, 2017; Syphard et al., 2017; Moloney et al., 2019; Sweet et al., 2019).
43   In summary, under a high emissions scenario that increases global temperature 4ºC by 2100, climate change
44   could increase global burned area 50-70% and global mean fire frequency ~30% with increases on one-third
45   to two-thirds of global land and decreases on one-fifth of global land (medium confidence). Lower emissions
46   that would limit the global temperature increase to <2ºC would reduce projected increases of burned area to
47   ~35% and projected increases of fire frequency to ~20% (medium confidence). Increased wildfire, combined
48   with erosion due to deforestation, could degrade water supplies (high confidence). For ecosystems with
49   historically low fire frequencies, a projected 4ºC global temperature increases risks of fire, contributing to
50   potential tree mortality and conversion of over half of Amazon rainforest to grassland and thawing of Arctic
51   permafrost that could release 11-200 Gt carbon that could substantially exacerbate climate change (medium
52   confidence).
54   Risks to Ecosystems and Services from Tree Mortality

     Do Not Cite, Quote or Distribute                      2-89                                      Total pages: 237
     FINAL DRAFT                                    Chapter 2                   IPCC WGII Sixth Assessment Report

 1   Under continued climate change, increased temperature, aridity, drought, wildfire (Section, and
 2   insect infestations (Section will tend to increase tree mortality across wide parts of the world
 3   (McDowell et al., 2020). Boreal and temperate forest loss to fire, wind, and bark beetles could cause more
 4   negative than positive effects for most ecosystem services, including carbon storage to regulate climate
 5   change (Sections,,,, water supply for people (Section, timber
 6   production (Chapter 5), and hazard protection (Thom and Seidl, 2016). In addition, deforestation in tropical
 7   and temperate forests can increase local temperatures 0.3° to 2ºC (Hesslerová et al., 2018; Lejeune et al.,
 8   2018; Zeppetello et al., 2020) and this effect can extend up to 50 km (Cohn et al., 2019).
10   In Amazon rainforests, the relatively lower buffering capacity for plant moisture during drought increases
11   the risk of tree mortality and, combined with increased heat from climate change and fire from deforestation,
12   the possibility of a tipping point of extensive forest dieback and a biome shift to grassland (Oyama and
13   Nobre, 2003; Sampaio et al., 2007; Lenton et al., 2008; Nepstad et al., 2008; Malhi et al., 2009; Salazar and
14   Nobre, 2010; Settele et al., 2014; Lyra et al., 2016; Zemp et al., 2017b; Brando et al., 2020). This could
15   occur at a 4-5ºC temperature increase above the pre-industrial period (Salazar and Nobre, 2010). Under
16   RCP8.5, half of Amazon tropical evergreen forest could shift to grassland through drought-induced tree
17   mortality and wildfire, but lower emissions (RCP4.5) could limit the loss to ~5% (Lyra et al., 2016).
18   Precipitation declines from reduced evapotranspiration inputs after forest loss could cause additional
19   Amazon forest loss of one-quarter to one-third (Zemp et al., 2017a). Similarly, in Guinean tropical deciduous
20   forest in Africa, climate change under RCP8.5 could increase mortality 700% by 2100 or 400% under lower
21   emissions (RCP4.5; Claeys et al., 2019). These projections indicate risks of climate change-induced tree
22   mortality reducing tropical forest areas in Africa and South America up to half under a 4ºC increase above
23   the pre-industrial period, but a lower projection of a 2ºC increase could limit the projected increases in tree
24   mortality (robust evidence, high agreement).
26   Temperate and boreal forests possess greater diversity of physiological traits related to plant hydraulics, so
27   they are more buffered against drought than tropical forests (Anderegg et al., 2018). Nevertheless, in
28   temperate forests, drought-induced tree mortality under RCP8.5 could cause the loss of half of northern
29   hemisphere conifer forest area by 2100 (McDowell et al., 2016). In the western United States, one-tenth of
30   forest area is highly vulnerable to drought-induced mortality under RCP8.5 by 2050 (Buotte et al., 2019). In
31   California, increased evapotranspiration in Sierra Nevada conifer forests increases the potential fraction of
32   the area at risk of tree mortality 15‒20% per degree Celsius (Goulden and Bales, 2019). In Alaska, fire-
33   induced tree mortality from climate change under RCP8.5 could reduce the extent of spruce forest (Picea
34   sp.) 8‒44% by 2100 (Pastick et al., 2017). Under RCP8.5, tree mortality from drought, wildfire, and bark
35   beetles could reduce timber productivity of boreal forests in Canada by 2100 below current levels (Boucher
36   et al., 2018; Chaste et al., 2019; Brecka et al., 2020). In Tasmania, projected increases in wildfire (Fox-
37   Hughes et al., 2014) increase risks of mortality in mesic vegetation (Harris et al., 2018b) and threaten the
38   disappearance of the long-lived endemic pencil pine (Athrotaxis cupressoides) (Holz et al., 2015; Worth et
39   al., 2016) and temperate montane rainforest (Mariani et al., 2019). These projections indicate risks of climate
40   change-induced tree mortality reducing some temperate forest areas by half under emissions scenarios of
41   2.5-4ºC above the pre-industrial period (medium evidence, high agreement).
43   Risk to Terrestrial Ecosystem Carbon Stocks
45   Globally, increasing atmospheric CO2 enhances the terrestrial sink but temperature increases constrain it,
46   reflecting biological process understanding, highlighted in previous IPCC reports (high confidence).
47   Analyses of atmospheric inversion model output and spatial climate data indicate a sensitivity of net
48   ecosystem productivity to CO2 fertilisation of 3.1 ± 0.1 to 8.1 ± 0.3 Gt per 100 ppm CO2 (~1°C increase) and
49   a sensitivity to temperature of -0.5 ± 0.2, to -1.1 ± 0.1 Gt per degree Celsius (Fernandez-Martinez et al.,
50   2019). The future of the global land carbon sink (Section nevertheless remains highly uncertain
51   because (i) of regionally complex interactions of climate change and changes in atmospheric CO2 with
52   vegetation, soil and aquatic processes, (ii) episodic events such as heat-waves or droughts (and related
53   impacts through mortality, wildfire or insects, pests and diseases, (Section, so far are only
54   incompletely captured in carbon cycle models, and (iii) legacy effects from historic land-use change and
55   environmental changes are incompletely captured but likely to decline in future, and (iv) lateral carbon
56   transport processes such as export of inland waters or erosion are incompletely understood and modelled (
57   AR6 WGI Chapter 5; Pugh et al., 2019a; Friedlingstein et al., 2020; Krause et al., 2020). Enhanced carbon

     Do Not Cite, Quote or Distribute                     2-90                                      Total pages: 237
     FINAL DRAFT                                    Chapter 2                   IPCC WGII Sixth Assessment Report

 1   losses from terrestrial systems further limit the available carbon budget for global warming staying below
 2   1.5oC (Rogelj et al., 2018). Analyses of satellite remote sensing and ground-based observations has indicated
 3   that between 1982 and 2015 the CO2 fertilisation effect has already declined, implying a negative climate
 4   system feedback (Wang et al., 2020c). Peatlands, permafrost regions and tropical ecosystems are particularly
 5   vulnerable due to their large carbon stocks in combination with over-proportional warming, increases in
 6   heatwaves and droughts and/or a complex interplay of climate change and increasing atmospheric CO2
 7   (Section,,
 9   Model projections suggest under all warming scenarios a reduction of permafrost extent and potentially large
10   carbon losses (AR6 WGI Chapter 5). Already a mean temperature increase of 2ºC could reduce the total
11   permafrost area extent by ca 5-20% by 2100 (Comyn-Platt et al., 2018; Yokohata et al., 2020). Associated
12   CO2 losses of order of 15 up to nearly 70 GtC by 2100 have been projected across a number of modelling
13   studies (Schneider von Deimling et al., 2015; Comyn-Platt et al., 2018; Yokohata et al., 2020). Limiting the
14   global temperature increase to 1.5oC, compared to 2ºC could reduce projected permafrost CO2 losses by
15   2100 by (median) 24.2 GtC (calculated for 3m depth (Comyn-Platt et al., 2018). Losses are possibly
16   underestimated in those studies that consider only upper permafrost layers. Likewise, the actual committed
17   carbon loss may well be larger (e.g., eventually a loss of ca. 40% of today’s permafrost area extent if climate
18   is stabilised at 2oC above pre-industrial levels) due to the long time-scale of warming in deep permafrost
19   layers (Chadburn et al., 2017). It is unknown at which level of global warming abrupt permafrost collapse
20   estimated to enhance CO2 emission by 40% in 2300 in a high emissions scenario, compared to gradual thaw
21   emissions (Turetsky et al., 2020) would have to be considered an important additional risk. Large
22   uncertainties arise also from interactions with changes in surface hydrology and/or northward migrating
23   woody vegetation as climate warms, which could dampen or even reverse projected net carbon losses in
24   some regions (McGuire et al., 2018a; Mekonnen et al., 2018; Pugh et al., 2018) so overall there is low
25   confidence on how carbon-permafrost interactions will affect future carbon cycle and climate, although net
26   carbon losses and thus positive (amplifying feedbacks) are likely (Sections,; Shukla et al.,
27   2019). See also AR6 WGI (Chapter 5) for discussion of impacts of higher emission and warming scenarios.
29   Peatland carbon is estimated as ca. 550‒1000 GtC in northern latitudes (many of these peatlands would be
30   found in permafrost regions) (Turetsky et al., 2015; Nichols and Peteet, 2019) and > 100 GtC in tropical
31   regions (Turetsky et al., 2015; Dargie et al., 2017). Both for northern mid- and high-latitude and for tropical
32   peatlands a shift from contemporary CO2 sinks to sources were simulated in high warming scenarios (Wang
33   et al., 2018a; Qiu et al., 2020). Due to the lack of large-scale modelling studies, the confidence on climate
34   change impacts on peat carbon uptake and emissions is low. The largest risk to tropical peatlands is expected
35   to arise from drainage and conversion to forestry or agriculture, outpacing impacts of climate change (Page
36   and Baird, 2016; Leifeld et al., 2019; Cooper et al., 2020) although the magnitude of possible carbon losses
37   are uncertain and depend strongly on socio-economic scenarios. (Sections,;,
39   For tropical and sub-tropical regions the interplay of atmospheric CO2 with precipitation and temperature
40   becomes of particular importance for future carbon uptake, since in warm and dry environments, elevated
41   CO2 fosters plants with C3 photosynthesis and enhances their water use efficiency relative to C4 species
42   (Moncrieff et al., 2014a; Midgley and Bond, 2015; Knorr et al., 2016a). As a consequence, enhanced woody
43   cover is expected to occur in future especially in mesic savannas, while in xeric savannas an increase in
44   woody cover would occur in regions with enhanced precipitation (Criado et al., 2020). Even though semi-
45   arid regions have dominated the recent decades’ global trend in land CO2 uptake (Ahlström et al., 2015), so
46   far most studies that investigated future climate change impacts on savanna ecosystems have concentrated on
47   changes in areal extent (, rather than on carbon cycling, with medium confidence on increasing
48   woody:grass ratios (Moncrieff et al., 2014a; Midgley and Bond, 2015; Moncrieff et al., 2016; Criado et al.,
49   2020). Increases in woody vegetation in what is now grass-dominated would possibly come with a carbon
50   benefit, for instance a broad range of future climate and CO2 changes were found to enhance vegetation C
51   storage in Australian savannas (Scheiter et al., 2015). Results from a number of field experiments indicate
52   however, that impacts on total ecosystem carbon storage may be smaller, due to a loss in belowground
53   carbon (Coetsee et al., 2013; Wigley et al., 2020). Nunez et al., 2021) critique existing incentives to promote
54   invasion of non-native trees into treeless areas as a means of carbon sequestration, raising doubts about the
55   effects on fire, albedo, biodiversity and water yield (see Box 2.2).

     Do Not Cite, Quote or Distribute                     2-91                                      Total pages: 237
     FINAL DRAFT                                       Chapter 2                    IPCC WGII Sixth Assessment Report

 1   Substantial climate-change driven impacts on tropical tree cover and vegetation type are projected in all
 2   studies, irrespective of whether or not the degree amounts to a forest “dieback” (Sections,;
 3,; AR6 WGI Chapter 5) (Davies-Barnard et al., 2015; Wu et al., 2016a; Zemp et al., 2017a).
 4   Accordingly, models also suggest a continuation of tropical forests acting as carbon sinks (Huntingford et
 5   al., 2013; Mercado et al., 2018). A recent study, combining field plot data with statistical models, (Hubau et
 6   al., 2020) indicates that in the Amazonian and possibly also in the African forest the carbon sink in
 7   aboveground biomass has already declined over the three decades to 2015. This trend is distinct in the
 8   Amazon, whereas data from Africa suggest a possible decline after 2010. the authors estimate the vegetation
 9   carbon sink in 2030-2040 to decline to zero (-0.5–0-46) Pg C a-1 in the Amazon and 0.26 (0.04–0.47, a loss
10   of 14% compared to present) Pg C a-1 in Africa. Their results suggest that CO2 fertilisation is over time
11   outweighed by impacts of higher temperatures and drought, enhancing tree mortality and diminishing
12   growth. The degree of thermal resilience of tropical forest still remains uncertain, however (Sullivan et al.,
13   2020).
15   The lack of simulation studies that seek to quantify all important interacting factors (CO2, drought and fire)
16   for future carbon-cycling in savannas and tropical forests, and the apparent disagreement between trends
17   projected in models compared to data-driven estimates results in low confidence regarding the direction or
18   magnitude of carbon flux and pool size changes. Similar to tropical peatlands, given projected human
19   population growth and socio-economic changes. the continued conversion of forests and savannas into
20   agricultural or pasture systems very likely poses a significant risk of rapid carbon loss which will amplify
21   climate change induced risks substantially (high confidence) (Sections,; Aragao et al., 2014;
22   Searchinger et al., 2015; Aleman et al., 2016; Nobre et al., 2016).
24   The impacts of climate-induced altered animal composition and trophic cascades on global land ecosystem
25   carbon cycling are as yet unquantified (Schmitz et al., 2018) even though climate change is expected to lead
26   to shifts in consumer-resource interactions that also contribute to losses of top-predators or top herbivores
27   (Sections,, 2.5.4; Lurgi et al., 2012; Damien and Tougeron, 2019). Cascading trophic effects
28   triggered by top predators or the largest herbivores propagate through food webs and reverberate through to
29   the functioning of whole ecosystems, altering notably productivity, carbon and nutrient turnover and net
30   carbon storage (medium confidence) (Wilmers and Schmitz, 2016; Sobral et al., 2017; Stoner et al., 2018).
31   Across different field experiments, the ecosystem consequences of the presence or absence of herbivores and
32   carnivores have been found to be quantitatively as large as the effects of other environmental change drivers
33   such as warming, enhanced CO2, fire or variable nitrogen deposition (medium confidence) (Hooper et al.,
34   2012; Smith et al., 2015). Some local and regional-scale modelling experiments have begun to explore
35   animal impacts on vegetation dynamics and carbon and nutrient cycling (Pachzelt et al., 2015; Dangal et al.,
36   2017; Berzaghi et al., 2019). Given that turnover rate is a chief factor that determines future land ecosystem
37   carbon dynamics and hence carbon-climate feedbacks (Friend et al., 2014). To improve projections, it is
38   imperative to better quantify the broader role of carnivores, grazers, browsers, and the way these interact in
39   global studies of how ecosystems respond to climate change.
41   Feedbacks between Ecosystems and Climate
43   The possibility of feedbacks and interactions between climate drivers and biological systems or ecological
44   processes was identified as a significant emerging issue in AR5, and has since also been highlighted in the
45   CRCCL and the Special Report on 1.5oC. It is virtually certain that land cover changes affect regional and
46   global climate through changes to albedo, evapotranspiration and roughness (very high confidence) (Perugini
47   et al., 2017). There is growing evidence that biosphere-related climate processes are being affected by
48   climate change in combination with disturbance and land use change (high confidence) (Jia et al., 2019). It is
49   virtually certain that land surface change caused by disturbances such as forest fire, hurricanes, phenological
50   changes, insect outbreaks and deforestation affect carbon, water, and energy exchanges, thereby influencing
51   weather and climate (very high confidence) (Table 2.4; Figure 2.10; Bright et al., 2013; Brovkin et al., 2013;
52   Naudts et al., 2016; Prăvălie, 2018).
56   Table 2.4: Terrestrial and freshwater ecosystem feedbacks which affect the Earth's climate system dynamics, following
57   (Prăvălie, 2018).

     Do Not Cite, Quote or Distribute                        2-92                                         Total pages: 237
FINAL DRAFT                                            Chapter 2                        IPCC WGII Sixth Assessment Report

        Perturbation                             Implications for Warming/Feedback Mechanism
                                                      the Earth's Climate System Dynamics
 Phenological change        Increased primary productivity and plant growth with CO2 fertilisation (Mao et al., 2016; Wang et
                            al., 2018a); Increasing growing season length (Peñuelas et al., 2009; Barichivich et al., 2013);
                            reduced diurnal temperature range through evapotranspiration (mid-latitudes) and albedo (high
                            latitudes) caused by vegetation greening (Jeong et al., 2011); increased CO2 storage in biomass
                            (cooling) (Keenan et al., 2014); Reduced albedo in snow-covered regions as canopies become taller
                            and darker (warming); increased evapotranspiration, a key component of the global water cycle and
                            energy balance which influences global rainfall, temperature, and atmospheric motion (Zeng et al.,
 Insect outbreaks           Reduced carbon uptake and storage (warming); Increased surface albedo (cooling) (Landry et al.,
                            2016); increased CO2 emissions (warming); decreased leaf area index and gross primary
                            productivity (Ghimire et al., 2015), leading to reduced evapotranspiration and increased land
                            surface temperature (Bright et al., 2013).
 Range shifts               Reduced albedo in snow-covered regions as trees expand poleward (warming) (Chae et al., 2015);
                            enhanced permafrost thawing; expansion of insect outbreak range, increasing forest impact
                            (Pureswaran et al., 2018); biome dependent changes in albedo and evapotranspiration regimes
                            (Naudts et al., 2016). Reduction in snow and ice albedo in freshwater due to loss of ice (warming)
                            (Lang et al., 2018).
 Die-off and large-scale    Decreased Gross primary productivity (GPP); decline in carbon storage (warming); increased CO2
 mortality events           emissions; increased solar radiation, reduced soil moisture, higher surface runoff; albedo effects
                            (Lewis et al., 2011; Prăvălie, 2018)
 Deforestation              Reduced carbon storage (warming) (Pugh et al., 2019a); increase in (regional) surface air
                            temperature due to reduced evaporation (less cooling); increased albedo in high-latitude systems
                            (regional radiative cooling) (Loranty et al., 2014); increased air temperature and diurnal
                            temperature variation (Alkama and Cescatti, 2016), locally and globally (Winckler et al., 2019);
                            reduced precipitation (Perugini et al., 2017); decreased biogenic volatile organic compounds
                            (BVOC) and aerosol emissions (warming through direct and indirect aerosol effects; cooling
                            associated with reduction in atmospheric methane (Jia et al., 2019)
 Forest degradation         Reduced carbon storage (warming) (de Paula et al., 2015; Bustamante et al., 2016; de Andrade et
                            al., 2017; Mitchard, 2018)
 Fragmentation              Carbon losses because biomass is less developed in forest edges (Pütz et al., 2014; Chaplin-Kramer
                            et al., 2015; Haddad et al., 2015)
 Air pollution              Decreased plant productivity, transpiration and carbon sequestration in forest with lower biomass
                            due to ozone toxicity (Sitch et al., 2007; Ainsworth et al., 2012); increased (regional) productivity
                            due to increase in diffuse solar radiation caused by terrestrial aerosols (Xie et al., 2021)

 Declining populations of   Changes to physical and chemical properties of organic matter, soils and sediments influence
 megafauna                  carbon uptake and storage (Schmitz et al., 2018); increased or decreased carbon storage biomass
                            and carbon storage, with differences across biomes determined by floristic structure and animal size
                            (Bello et al., 2015; Osuri et al., 2016; Peres et al., 2016; He et al., 2017; Berzaghi et al., 2018;
                            Schmitz et al., 2018; He et al., 2019)
 Fire                       Increased carbon and aerosol emissions(van der Werf et al., 2017); surface warming (Liu et al.,
                            2019b); albedo effect dependent on ecosystem and species-level traits (Rogers et al., 2015; Chen et
                            al., 2018a) (initial albedo decrease post-fire; increased albedo where snow exposure is increased by
                            canopy removal and species composition changes during recovery); black carbon deposition on
                            snow and sea ice (short-term) (Randerson et al., 2006); indirect increases in carbon emissions due
                            to soil erosion (Caon et al., 2014)
 Change in forest           Reduced carbon storage due to decline in biomass (warming) (McIntyre et al., 2015)
 Woody encroachment in      Reduced production, increased water use, reduced albedo and altered land-atmosphere feedbacks;
 non-forested ecosystems    increased carbon storage in woody savannas (Zhou et al., 2017; Mureva et al., 2018); Uncertain
                            feedbacks to C cycle (some suggest an increase, others a decrease)
 Net Primary Productivity   Reduced albedo following high-latitude expansion of trees caused by photosynthetic enhancement
 (NPP) shifts               of growth (cooling); increased photosynthesis and net ecosystem production (NEP) (Fernandez-
                            Martinez et al., 2019); increased NPP in N‐limited ecosystems due to increased nitrogen deposition
                            from agriculture and combustion (Du and de Vries, 2018; Schulte-Uebbing and de Vries, 2018);
                            Nutrient limited lakes are likely to become less productive, while nutrient rich lakes are likely to
                            become more productive due to warming induced prolongation of stable stratification (Adrian et
                            al., 2016; Kraemer et al., 2017).
 Biogeochemical shifts      Decline in carbon storage due to nitrogen limitation in N limited systems (warming)(Reich et al.,
                            2014; Wieder et al., 2015); increased carbon storage on land(Peñuelas et al., 2013) and in lakes
                            (Heathcote et al., 2015; Mendonça et al., 2017); Increase in CO2 and CH4 emissions from
                            freshwater ecosystems due to increased eutrophication (DelSontro et al., 2018), the imbalance

Do Not Cite, Quote or Distribute                              2-93                                               Total pages: 237
     FINAL DRAFT                                          Chapter 2                         IPCC WGII Sixth Assessment Report
                               between losses and gains of CO2 by photosynthesis and respiration (metabolic theory of ecology),
                               enhanced emissions from exposed river and lake sediments during droughts and re-wetting (Marcé
                               et al., 2019; Keller et al., 2020), enhanced CH4 ebullition of seasonally hypoxic lakes (Aben et al.,
                               2017; DelSontro et al., 2018; Bartosiewicz et al., 2019; Beaulieu et al., 2019; Sanches et al., 2019),
                               increased transfer of organic carbon from land to water (particularly in permafrost areas) (Wauthy
                               et al., 2018)

 4   Figure 2.10: Terrestrial ecosystem feedbacks, which affect the Earth's climate system dynamics. Perturbations and
 5   implications for climate system dynamics (warming/cooling) are shown for the three global forest biomes (adapted
 6   from Figure 5, Prăvălie, 2018). The strength of the mechanism is estimated in general terms based on the magnitude of
 7   carbon storage and evaporative cooling processes that characterise each forest biome (Bonan, 2008). Carbon storage
 8   includes forest biomass, without accounting for carbon dynamics in soil, peat and underlying permafrost deposits.
 9   Implications of biogeochemical shifts were only estimated in relation to the intensification of the carbon cycle and
10   increase in biomass at high latitudes, assuming N availability for the stoichiometric demands of forest vegetation).
13   Feedbacks can be positive or negative (i.e., amplify or dampen the original forcing), vary spatially and
14   seasonally, and act over large geographic areas and long time periods (>decades), making them difficult to
15   observe and quantify directly (AR6 WGI Chapter 5; Schimel et al., 2015). Due to the positive impacts of
16   CO2 on vegetation growth and ecosystem carbon storage (high confidence) (Sections;; AR6
17   WGI chapter 5), the associated climate feedback is negative (increased removal of atmospheric CO2 and
18   dampened warming, compared to absence of the feedback). By contrast, projected global losses of carbon in
19   warmer climates (AR6 WGI Chapter 5) imply a positive climate feedback. Chapter 5 of WGI assesses an
20   overall increase in land carbon uptake through the 21st century. However, the overall strength of the carbon
21   cycle-climate feedback remains very uncertain. One of the underlying reasons may be complex interactions
22   with ecosystem water balance and nitrogen and phosphorous availability, which are poorly constrained by
23   observational evidence and incompletely captured in Earth System Models (AR6 WGI Chapter 5, Section
24; Huntzinger et al., 2017).

     Do Not Cite, Quote or Distribute                             2-94                                                Total pages: 237
     FINAL DRAFT                                      Chapter 2                    IPCC WGII Sixth Assessment Report

 1   Land ecosystems contribute substantially to global emissions of nitrous-oxides and methane. As with CO2,
 2   these emissions respond both directly and indirectly to atmospheric CO2 concentration and climate change,
 3   which gives rise to potential additional biogeochemical feedbacks in the climate system. A large part of these
 4   emissions stem from land and water management, such as fertiliser application, rice production, aquaculture
 5   or animal husbandry (Jia et al., 2019). However, nearly 60% of total nitrous-oxide emissions (2007-2016) is
 6   estimated to stem from natural ecosystems, especially in the tropics (AR6 WGI Chapter 5; Tian et al., 2019),
 7   while freshwater wetlands and peatlands are estimated to contribute between 83% (top-down estimates) and
 8   40% (bottom up estimates) of total natural CH4.(and 31% or 20% of total methane emissions, respectively)
 9   for the period 2008-2017 (AR6 WGI Chapter 5). Median CH4 emissions from northern latitude wetlands in
10   2100 were estimated to be 12.1 and 13.5 Pg C in emission scenarios leading to 1.5oC and 2oC warming,
11   respectively (Comyn-Platt et al., 2018). Likewise, global warming was attributed to soil N2O emission
12   increases since the pre-industrial period of 0.8 (0.3‒1.3) Tg N a-1 (Tian et al., 2020). Overall, climate
13   feedbacks from future altered land ecosystem emissions of CH4 or N2O are uncertain, but expected to be
14   small (AR6 WGI Chapter 5).
16   Changes in regional biodiversity are integral parts of ecosystem-climate feedback loops, including and
17   beyond carbon-cycle processes (Figure 2.10; Table 2.4). For instance, the impacts of climate-induced altered
18   animal composition and trophic cascades on ecosystem carbon turnover (see Sections; could
19   be a substantive contribution to carbon-climate feedbacks (low confidence). Additional surface-atmosphere
20   feedbacks that arise from changes in vegetation cover and subsequently altered albedo, evapotranspiration or
21   roughness (often summarised as biophysical feedbacks) can be regionally relevant and could amplify or
22   dampen vegetation cover changes (Jia et al., 2019).
24   Climate-induced shifts towards forests in what is currently tundra would be expected to reduce regional
25   albedo especially in spring but also during parts of winter when trees are snow-free (whereas tundra
26   vegetation would be covered in snow), which amplifies warming regionally (high confidence) (Perugini et
27   al., 2017; Jia et al., 2019). Trees would also enhance momentum absorption compared to low tundra
28   vegetation thus impacting surface-atmosphere mixing of latent and sensible heat fluxes (Jia et al., 2019).
29   Boreal forests insulate and stabilize permafrost and reduce fluctuations of ground temperature: the amplitude
30   of variation of ground surface temperatures was 28°C in a forested site, compared to 60°C in nearby
31   grassland (Section; Bonan, 1989; Stuenzi et al., 2021a; Stuenzi et al., 2021b). Likewise, a shift in
32   moist tropical forests towards vegetation with drought-tolerant traits could possibly reduce
33   evapotranspiration, increase albedo, alter heat transfer at the surface and lead to a negative feedback to
34   precipitation (Section; Jia et al., 2019). In savannas, restoration of woody vegetation has been shown
35   to enhance cloud formation and precipitation in response to enhanced transpiration and turbulent mixing,
36   leading to a positive feedback on woody cover (Syktus and McAlpine, 2016). While this has not yet been
37   systematically explored, similar feedbacks might also emerge from a CO2-induced woody cover increase in
38   savannas (low confidence) (Section
40   Since biophysical feedbacks can contribute to both surface temperature warming or cooling, analyses so far
41   suggest that, at global scale, the net impact on climate change is small (Perugini et al., 2017; Jia et al., 2019),
42   unless these feedbacks also accelerate vegetation mortality and lead to substantive carbon losses (Zemp et
43   al., 2017a; Lemordant and Gentine, 2019). More than one third of the Earth’s land surface has at least 50%
44   of its evapotranspiration regulated by vegetation, and in some regions between 40 and >80% of the land’s
45   evaporated water is returned to land as precipitation. Locally, both, direct human-mediated as well as climate
46   change-mediated vegetation cover change can therefore notably affect annual average freshwater availability
47   to human societies, especially if negative feedbacks amplify vegetation cover reduction, reduced
48   evapotranspiration and reduced precipitation (medium confidence) (Keys et al., 2016; Keys and Wang-
49   Erlandsson, 2018).
51   Since AR5, freshwater ecosystems (lakes, reservoirs, rivers, ponds) have been increasingly recognised as
52   important sources of greenhouse gas emissions (CO2, CH4, N2O) into the atmosphere. Key mechanisms
53   which contribute to rising GHG emissions from freshwater ecosystems are the temperature imbalance
54   between photosynthesis and respiration (respiration increases more than photosynthesis with rising
55   temperature), CO2 and CH4 emissions from exposed sediments during droughts, increased matter transport
56   from land to water, changes in water retention time in rivers and lakes, and temperature effects on lake
57   stratification and anoxia, favouring CH4 emissions.

     Do Not Cite, Quote or Distribute                       2-95                                        Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 2   DelSontro et al. (2018) assembled the largest global dataset to date on emission rates from lakes of CO2, CH4
 3   and N2O and found that they co-vary with lake size and trophic state. They estimated that moderate global
 4   increases in eutrophication of lakes could translate to 5–40% increases in the GHG effects in the atmosphere.
 5   Moreover, they estimated that greenhouse gas emissions from lakes and impoundments in past decades
 6   accounted for 1.25-2.30 Pg C-CO2 yr-1 (DelSontro et al., 2018), thus around 20% of global burning fossil-
 7   fuel CO2 emission (9.4 PgC-CO2 yr-1 (Friedlingstein et al., 2020).
 9   Global warming will strongly enhance freshwater CH4 emissions through a disproportionate increase in
10   ebullition (gas flux) by 6–20% per 1°C increase in water temperature (Aben et al., 2017). It can be expected
11   that ongoing eutrophication enhanced by climate change-related increases in sediment nutrient release and
12   organic carbon and nutrient loading from catchments will enhance CH4 ebullition at a global scale (Aben et
13   al., 2017; DelSontro et al., 2018; Bartosiewicz et al., 2019; Beaulieu et al., 2019; Sanches et al., 2019). The
14   strongest increase in ebullition is expected in shallow waters where sediment temperatures are strongly
15   related to atmospheric temperature (Aben et al., 2017). Given that small ponds and shallow lakes are the
16   most abundant freshwater ecosystems globally they may become hot spots of CH4 ebullition in the future
17   (Aben et al., 2017). On average CH4, CO2, N2O account for 75%, 23, and 2% of the total CO2 equivalent
18   emissions, respectively in lakes (DelSontro et al., 2018).
20   Further, the exposure of lake and river sediments during droughts activates decomposition of buried organic
21   carbon. In dry river beds, mineralisation of buried organic matter is likely to increase with climate change as
22   anoxic sediments are oxygenated downwards during drying along with pulses of microbial activity following
23   rewetting of desiccated sediment. Conservative estimates indicate that adding emissions from exposed
24   sediments of dry inland waters across diverse ecosystem types and climate zones to current global estimates
25   of CO2 emissions could result in a 6% (~0.12 Pg C y−1) increase of total inland water CO2 emission rates
26   covering streams and rivers (334 mmol m-2 day-1), lakes and reservoirs (320 mmol m-2 day-1) and small
27   ponds (148 mmol m-2 day-1) (Marcé et al., 2019; Keller et al., 2020).
29   Overall, uncertainty in the quantity of carbon fluxes within freshwater ecosystems and between terrestrial
30   and freshwater systems and subsequent emissions to the atmosphere remain very high (Raymond et al.,
31   2013; Catalán et al., 2016; Stanley et al., 2016; Evans et al., 2017; Drake et al., 2018; Seekell et al., 2018;
32   Sanches et al., 2019; Bodmer et al., 2020; Keller et al., 2020) (see Table 2.SM., see also Chapter 5 of WGI).
33   Projections of carbon fluxes are e.g. challenged by the complex interaction between rising water
34   temperature, loss of ice, changes in hydrology, ecosystem productivity, increase in extreme events, and
35   variation in terrestrial matter transport. While we are still short in empirical data, particularly in the tropics
36   (DelSontro et al., 2018), improvements in sensor technology (Eugster et al., 2011; Gonzalez-Valencia et al.,
37   2014; Maeck et al., 2014; Delwiche et al., 2015) and the use of statistically robust survey designs (Beaulieu
38   et al., 2016; Wik et al., 2016) have improved the accuracy of GHG emission rate measurements in freshwater
39   ecosystems. Global networks such as GLEON (Global Lakes Ecological Observatory Network) increasingly
40   allow a global view of carbon fluxes improving estimates of the contribution of freshwater ecosystems to
41   global GHG emissions to the atmosphere.
43   In summary (Drake et al., 2018) aggregated contemporary estimates of CO2 and CH4 emissions from
44   freshwater ecosystems with global estimates made by (Raymond et al., 2013) and arrived at an estimate of
45   3.9 Pg C yr-1. Rivers and streams accounted for 85% of the emissions and lakes and reservoirs for 15%
46   (Raymond et al., 2013). This trend will continue under scenarios of nutrient loading to inland waters over the
47   next century where inland water increased CH4 emission has an atmospheric impact of 1.7–2.6 Pg C-CO2-eq
48   y−1, which is equivalent to 18–33% of annual CO2 emissions from burning fossil fuels (medium evidence,
49   medium agreement, medium confidence) (Beaulieu et al., 2019). For comparison, annual uptake of CO2 in
50   land ecosystems is estimated as 3.4 (± 0.9) PgC/yr (Friedlingstein et al., 2020). The freshwater numbers
51   combine CO2 and CH4 and are thus not directly comparable. However, they are indicative for the importance
52   to account better for freshwater systems in global carbon budgets.
54   Risks to Freshwater Ecosystem Services: Drinking Water, Fisheries and Hydropower
56   AR5 named water supply and biodiversity as freshwater ecosystem services vulnerable to climate change.
57   We discuss risks to these and to additional services identified by model projections based both on climate-

     Do Not Cite, Quote or Distribute                      2-96                                        Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   change scenarios (Schröter et al., 2005; Boithias et al., 2014; Huang et al., 2019; Jorda-Capdevila et al.,
 2   2019) and on the Common International Classification of Ecosystem Services (high agreement, high
 3   confidence) (CICES, 2018). Effects of floods, droughts, permafrost and glacier melting on global changes in
 4   water quality, particularly with respect to contamination with pollutants, are described in Section 4.2.6.
 6 Risks to quantity and quality of drinking water
 7   Forests and other vegetated ecosystems assist production of drinkable water by facilitating infiltration of
 8   rainfall and snowfall into the ground, where water either moves through the soil saturated zone to supply
 9   streams and other surface waters or infiltrates further to recharge groundwater aquifers (Ellison et al., 2012;
10   Bonnesoeur et al., 2019). Globally, 4 billion people depend on forested watersheds for drinking water
11   (Mekonnen and Hoekstra, 2016). Chapter 4 assesses the physical science of water supply, including
12   precipitation, runoff, and hydrology, and social aspects of human water use. This section assesses ecological
13   aspects of risks to freshwater supplies for people.
15   Reduction of vegetation cover following wildfires (Section and tree mortality (Section can
16   reduce long-term water infiltration, increase soil erosion and flash floods, and release sediment that degrades
17   drinking water quality. Widlfires increase impacts of extreme precipitation events due to climate change,
18   which contribute to increased surface runoff and hence to increased risks of land erosion, landslides and
19   flooding (Ebel et al., 2012; Robinne et al., 2020). Under current conditions, nearly half of global land area is
20   at moderate to high risk of water scarcity due to wildfires (Robinne et al., 2018; Robinne et al., 2020). From
21   1984 to 2014 wildfires in the western USA affected 6-11% of stream and river length (Ball et al., 2021).
22   Under a high emissions scenario of a 3.5ºC temperature increase post-fire erosion across the western USA
23   could double sedimentation and degrade drinking water quality in one-third of watersheds by 2050 (Sankey
24   et al., 2017). In Brazil, post-fire vegetation loss tends to increase runoff, reduce infiltration, and reduce
25   groundwater recharge and flow of springs (Rodrigues et al., 2019). Runoff from wildfires can contain
26   dissolved organic carbon precursors for the formation of carcinogenic trihalomethanes during water
27   chlorination for drinking (Uzun et al., 2020), plus chromium, mercury, selenium, and other toxic trace metals
28   (Burton et al., 2016; Burton et al., 2019).
30   Net effects of deforestation and afforestation on runoff and water supply depend on local factors, leading to
31   conflicting evidence for effects of land cover change (Ellison et al., 2012; Chen et al., 2021b), but
32   combinations of climate change and deforestation are projected to reduce water flows (Olivares et al., 2019).
33   In southern Thailand, the combination of conversion of forest to rubber plantations and a one-third increase
34   in rainfall could increase erosion and sediment load 15% (Trisurat et al., 2016). In the watershed that
35   supplies São Paulo, Brazil, afforestation could increase water quantity and quality (Ferreira et al., 2019). In
36   most regions with dry or Mediterranean subtropical climates, climate change reduces renewable surface
37   water and groundwater resources significantly (Doell et al., 2015). In northeast Spain, reduced precipitation
38   and vegetation cover under a high emissions scenario of a 3.5ºC temperature increase could reduce drinking
39   water supplies by half by 2100 (Bangash et al., 2013).
41   Changes in algal biomass development and spread of cyanobacteria blooms, related to global warming,
42   resemble those triggered by eutrophication with well-known negative effects on the services lakes provide,
43   particularly for drinking water provision and recreation (robust empirical evidence, high agreement, high
44   confidence) (Carvalho et al., 2013; Adrian et al., 2016; Gozlan et al., 2019).
46   Based on a 10% increase in precipitation, (de Wit et al., 2016) estimated increased mobilisation of organic
47   carbon from soils to freshwaters by at least 30%, demonstrating the importance of climate wetting for the
48   carbon cycle. Browning negatively affects the taste of drinking water and may be difficult to address
49   (Kothawala et al., 2015; Mineau et al., 2016; Kritzberg et al., 2020). It also often reduces attractives for
50   recreational purposes, especially swimming (Arthington and Hadwen, 2003; Keeler et al., 2015). Based on a
51   worst-case climate scenario until 2030 (Weyhenmeyer et al., 2016) projected an increase in browning of
52   lakes and rivers in boreal Sweden by a factor of 1.3. The chemical character of dissolved organic matter, as
53   modified by climate change (Kellerman et al., 2014), determines its amenability to removal by water
54   treatment (Ritson et al., 2014). Therefore, in order to provide safe and acceptable drinking water, more
55   advanced, more expensive and more energy/resource intensive technical solutions may be required
56   (Matilainen et al., 2010).

     Do Not Cite, Quote or Distribute                      2-97                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   In summary, climate change increases risks to the integrity of watersheds and provision of safe, acceptable
 2   freshwater to people (medium evidence, medium agreement).
 4 Risks to freshwater fisheries and biodiversity
 5   Climate change will increase water temperatures and decrease dissolved oxygen levels (Section 2.3.1)
 6   impacting freshwater fisheries which form an important ecosystem service (Vári et al., 2021). People living
 7   in the vicinity of cold lakes will be affected by projected losses of ice. In a worst-case scenario (air
 8   temperatures increase of 8°C), 230,400 lakes and 656 million people in 50 countries will be impacted (Reid
 9   et al., 2019; Sharma et al., 2019). Winter ice fishing (Orru et al., 2014), transportation via ice roads (Prowse
10   et al., 2011) and cultural activities (Magnuson and Lathrop, 2014) are ecosystem services at stake with
11   ongoing loss of lake ice.
13   Eutrophication of central European lakes has wiped out a significant proportion of the endemic fish fauna
14   (Vonlanthen et al., 2012), so climate-induced further eutrophication is expected to represent an additional
15   threat to fish fauna and commercial fisheries (Ficke et al., 2007). Given that the ecological consequences of
16   lake warming may be especially strong in the tropics ( Section, ecosystem services may be most
17   affected there. Tropical lakes support important fisheries (Lynch et al., 2016a economic; McIntyre et al.,
18   2016) providing critical sources of nutrition to adjacent human populations. These lakes are especially prone
19   to loss of deep-water oxygen due to warming, with adverse consequences for fisheries productivity and
20   biodiversity (medium evidence, medium confidence) (Lewis Jr, 2000; Van Bocxlaer et al., 2012).
22   Tropical lakes tend to be hotspots of freshwater biodiversity (Vadeboncoeur et al., 2011; Brawand et al.,
23   2014; Sterner et al., 2020); ancient tropical lakes such as Malawi, Tanganyika, Victoria, Titicaca, Towuti
24   and Matano hold thousands of animal species found nowhere else (Vadeboncoeur et al., 2011). While
25   biodiversity and several ecosystem services can be considered synergistic (food webs, tourism, aesthetical
26   and spiritual value (Langhans et al., 2019), others can be considered antagonistic in case of a strong
27   ecosystem service demand (such as water abstraction, water use, food security in terms of over-exploitation).
28   Here the balance between biodiversity and ecosystem services is key (Langhans et al., 2019), where
29   biodiversity can be integrated into water policy through Integrated Water Resource Management (IWRM)
30   towards nature-based solutions (Ligtvoet et al., 2017)
32 Risks to hydro power and erosion control
33   River banks, riparian vegetation and macrophyte beds play important roles in erosion control through
34   reducing current velocities, increasing sedimentation and reducing turbidity (Madsen et al., 2001). Rates of
35   flow in rivers affect and inland navigation (Vári et al., 2021). Changing seasonality in snow-dominated
36   basins is expected to enhance hydropower production in winter, but decrease it during summer (Doell et al.,
37   2015). Glacier melt changes hydrological regimes, sediment transport, and biogeochemical and contaminant
38   fluxes from rivers to oceans, profoundly influencing ecosystem services that glacier-fed rivers provide,
39   particularly provision of water for agriculture, hydropower, and consumption (Milner et al., 2017). Loss of
40   glacial mass and snowpack has already impacted flow rates, quantities and seasonality (Hock et al., 2019);
41   see AR6 WGII, Chapter 4 Water). Meltwater yields from glacier ice are likely to increase in many regions
42   during the next decades, but decrease thereafter as glaciers become smaller and smaller and finally disappear
43   (Hock et al., 2019).
45   2.5.4    Key Risks to Terrestrial and Freshwater Ecosystems from Climate Change
47   Among numerous risks to terrestrial and freshwater ecosystems from climate change, this IPCC chapter
48   identified five phenomena as the most fundamental risks of climate change to ecosystem integrity and the
49   ecosystem services that support human well-being that are also quantified: species losses to ecosystems,
50   increased wildfire, increased tree mortality, ecosystem carbon losses, and ecosystem structure change (Table
51   2.5, Table 2.S.4, Figure 2.11). These key risks form part of the overall assessment of key risks in Chapter 16.
52   The IPCC Fifth Assessment Report chapter on terrestrial ecosystems (Settele et al., 2014) had also identified
53   three of these key risks – species extinctions, tree mortality, ecosystem carbon losses – and a fourth –
54   invasion by non-native species. This IPCC chater assesses impacts of climate change on invasive species in
55   multiple sections with respect to different processes or systems (e.g. in Section, and here includes
56   this aspect in a new broader key risk of ecosystem structure change. The IPCC Fifth Assessment Report had
57   included wildfire as a mechanism of the ecosystem carbon loss key risk. Based on additional research and

     Do Not Cite, Quote or Distribute                      2-98                                       Total pages: 237
     FINAL DRAFT                                          Chapter 2                     IPCC WGII Sixth Assessment Report

 1   field experience with major wildfires since then, this IPCC chapter sets wildfire apart as a specific key risk to
 2   ecosystem integrity and human well-being. These different measures of risk are interconnected but approach
 3   assessment of risk to terrestrial and freshwater ecosytems from different angles, using complementary
 4   metrics.
 6   Species are the fundamental unit of ecosystems. As species become rare, their roles in the functioning of the
 7   ecosystem diminishes, and disappears altogether if they go locally extinct (high confidence) (Isbell et al.,
 8   2015; Chen et al., 2018b; van der Plas, 2019; Wang et al., 2021b). Loss of species and functional groups
 9   reduces the ability of an ecosystem to provide services, and lowers its resilience to climate change (high
10   confidence) (Section 2.6.7; Elmqvist et al., 2003; Cadotte et al., 2011; Harrison et al., 2014; Carlucci et al.,
11   2020). For example, among crop systems, a key factor to succesful pollination is the phylogenetic diversity
12   of bee species available, more than total abundances (Drossart and Gérard, 2020). Because many species
13   have obligate interactions with, or are resources for, other species (e.g. predators and their prey, insects and
14   their host plants, plants and their mycorrhizae symbionts), loss of one species affects risk to another species
15   and, ultimately, ecosystem functioning (Mahoney and Bishop, 2017)
17   Global rates of species extinction are accelerating dramatically (Barnosky et al., 2011), with approximately
18   10% of species having been driven extinct by humans since the late Pleistocene, principally by over-
19   exploitation and habitat destruction, a rate estimated to be 1000 times higher than pre-Anthropocene
20   (natural) background extinction rates (De Vos et al., 2015). Therefore, this level—10%—of species
21   becoming endangered (sensu IUCN) due to loss of suitable climate space (Figure 2.8b), is used here as a
22   threshold moving risk to biodiversity from moderate to high, and twice that (20%) as the threshold from high
23   to very high.
25   Key risks assessed here are interconnected. Extinction of species is an irreversible impact of climate change,
26   has negative consequences on ecosystem integrity and functioning, and risk increases steeply with even
27   small rises in global temperature (Section; Figure 2.6; Figure 2.7; Figure 2.8). Continued climate
28   change substantially increases risks of carbon losses due to wildfires, tree mortality from drought and insect
29   pest outbreaks, peatland drying, permafrost thaw, and ecosystem structure change which could exacerbate
30   self-reinforcing feedbacks between emissions from high-carbon ecosystems and increasing global
31   temperatures (medium confidence). Thawing of Arctic permafrost alone could release 11-200 Gt carbon
32   (medium confidence). Complex interactions of climate change, land use change, carbon dioxide fluxes, and
33   vegetation changes will regulate the future carbon balance of the biosphere, processes incompletely
34   represented in earth system models. The exact timing and magnitude of climate-biosphere feedbacks and
35   potential tipping points of carbon loss are characterized by high ranges of the estimates, yet studies indicate
36   increased ecosystem carbon losses could cause extreme future temperature increases (medium confidence)
37   (Sections;;;;;;; Figure 2.10; Figure 2.11; Table 2.4;
38   Table 2.5; Table 2.S.2; Table 2.S.4)
41   Table 2.5: Key risks to terrestrial and freshwater ecosystems from climate change. This IPCC chapter assesses these as
42   the most fundamental risks of climate change to ecosystem integrity and the ecosystem services that support human
43   well-being. Climate factors include the primary variables governing the risk. Non-climate factors include other
44   phenomena that can dominate or contribute to the risk. Detection and attribution comprise cases of observed changes
45   attributed predominantly or in part to anthropogenic climate change (Section 2.4.2, 2.4.3, 2.4.4, 2.4.5, Table 2.2, Table
46   2.3, Table 2.S.1). Adaptation includes options to address the risk (Section 2.6). Risk transitions (defined in Figure 2.11)
47   indicate an approximate global mean surface temperature increase, relative to the pre-industrial period (1850-1900), to
48   move from one level of risk to the other and confidence in the assessment. Table 2.S.4 provides details of the
49   temperature levels for risk transitions. Both tables provides details for the key risk burning embers diagram (Figure
50   2.11).
      Biodiversity Risk: Increasing high extinction risk (species projected loss of >50% of range) among increasing
      number of plant and animal species. The transition from non-detectable to moderate was based on the number of
      local population extinctions, major declines of sub-species and two global species extinctions and that are
      attributable to climate change. The transition from moderate to high is centred around 1.5°C based on a few taxa
      that are known from their basic biology and habitat requirements to be at risk of extinction (endangered) at 1.5°C
      and on the increasing number of taxa that are projected to have high extinction risk (losing >50% of their suitable
      climate space) affecting >10% of the species in that taxa (1000x natural background rates of extinction). The

     Do Not Cite, Quote or Distribute                           2-99                                           Total pages: 237
FINAL DRAFT                                         Chapter 2                      IPCC WGII Sixth Assessment Report

 transition to very high comes from the increasing number of taxa projected to have >20% of species at high risk
 of extinction. In the worst-case scenario (10th percentile of the models), some of the taxa show >50% of the
 species at high risk of extinction. These assessments are also weighted by role the species in the taxa play in
 performing ecosystem services (both to the ecosystems and to humans, e.g. pollinators, detritivores). Confidence
 for the moderate threshold is high because it is based on observed trends attributed to climate change. Confidence
 for future projections are is medium as these are based on one large study (covering more than 135,000 species)
 and primarily based on loss of suitable climate. Based on Sections 2.4.2, 2.5.1, 2.6.1, 2.6.6, Table 2.3, Figure 2.6,
 Table SM2.1 and Table SM2.2.

 Climate factors               Non-climate        Detection and        Adaptation                Risk transitions
                               factors            attribution                                    (confidence)

 Shifts in geographic          land use           Observed D&A:        Habitat restoration,      0.8ºC undetectable-
 placements of climate         change, habitat    many cases of        habitat creation,         moderate (high
 space; loss of climate        degradation        population           increased connectivity    confidence)
 space globally; emergence     from pollution,    extinctions; two     of habitats and           1.58°C moderate-
 of non-analogue climates,     fertilisation,     cases of species     protected areas,          high (medium
 increases in extreme          invasive           extinctions          increase in protected     confidence)
 climate events                species            (; species   areas, assisted           2.07°C high-very
                                                  have tracked their   colonisation              high (medium
                                                  climate niches                                 confidence)
                                                  confidence in
                                                  SDM projections

 Wildfire: Increasing risk of wildfire that exceeds natural levels, damaging ecosystems, increasing illnesses and
 death of people, and increasing carbon emissions. Field evidence shows that anthropogenic climate change has
 increased the area burned by wildfire above natural levels across western North America in the period
 1984-2017, increasing burned area up to 11 times in one extreme year and doubling burned area over natural
 levels in a 32-year period. Burned area has increased in the Amazon, the Arctic, Australia, and parts of Africa and
 Asia, consistent with, but not formally attributed to anthropogenic climate change. These changes have occurred
 at global mean surface temperature increases of 0.6-0.9ºC. Empirical and dynamic global vegetation models
 project increases in burned area and fire frequency above natural levels on all continents under continued climate
 change, emergence of an anthropogenic signal from natural variation in fire weather for a third of global area, and
 increases of burned area in regions where fire had been rare or absent, particularly Arctic tundra and Amazon
 rainforest, at global temperature increases of 1.5-2.5ºC. Models project up to a doubling of burned area globally
 and wildfire-induced conversion of up to half the area of Amazon rainforest to grassland at temperature increases
 of 3-4.5ºC. (Sections,

 Climate factors               Non-climate        Detection and        Adaptation                Risk transitions
                               factors            attribution                                    (confidence)

 Increase in magnitude and     Deforestation,     Increased burned     Reduce deforestation,     0.75°C
 duration of high              agricultural       area in western      reduce use of fire in     undetectable-
 temperatures, decrease in     burning,           North America        tropical forests, use     moderate (high)
 precipitation, decrease in    peatland           above natural        prescribed burning and    2.0 °C moderate-
 relative humidity             burning            levels               allow naturally ignited   high (medium)
                                                                       fires to burn in          4.0°C high-very
                                                                       targeted areas to         high (medium)
                                                                       reduce fuel loads,
                                                                       encourage settlement
                                                                       in non-fire-prone areas

Do Not Cite, Quote or Distribute                          2-100                                           Total pages: 237
FINAL DRAFT                                         Chapter 2                     IPCC WGII Sixth Assessment Report

 Tree mortality: Tree mortality that exceeds natural levels degrades habitat for plant and animal species, increases
 carbon emissions, and reduces water supplies for people. Anthropogenic climate change caused three cases of
 drought-induced tree mortality in the period 1945-2007 in western North America, the African Sahel, and North
 Africa in temperate and tropical ecosystems. Pest infestations and wildfire due to climate change also caused
 much of the tree mortality in North America. These changes occurred at global mean surface temperature
 increases of 0.3-0.9ºC above the pre-industrial period. Models project increasingly extensive drought-induced
 tree mortality at continued temperature increases of 1-2ºC. Models project risks of mortality of up to half of
 forest area in different biomes at temperature increases of 2.5-4.5ºC. In Amazon rainforests, insufficient plant
 moisture reserves during drought increase the risk of tree mortality and, combined with increased fire from
 climate change and deforestation, the risk of a tipping point of massive forest dieback and a biome shift to
 grassland. (Sections,,,

 Climate factors              Non-climate        Detection and         Adaptation               Risk transitions
                              factors            attribution                                    (confidence)

 Increase in temperature,     Deforestation,     Tree mortality up     Reduce deforestation,    0.6°C undetectable-
 decrease in precipitation,   land-use           to 20% in three       reduce habitat           moderate (high)
 increase in aridity,         change             regions in Africa     fragmentation,           1.5°C: moderate-
 increase in frequency and                       and North             encourage natural        high (medium)
 severity of drought                             America               regeneration, restore    3.0°C high-very
                                                                       fragmented habitats      high (medium)

 Ecosystem carbon loss: Increasing risk of ecosystem carbon losses that could substantially raise the atmospheric
 carbon dioxide level. Measurements have detected emissions of carbon from boreal, temperate, and tropical
 ecosystems in places where increases in wildfire and tree mortality have been attributed to anthropogenic climate
 change, at global mean surface temperature increases of 0.6-0.9ºC above the pre-industrial period. Many factors
 govern the carbon balance of ecosystems, so changes have not been attributed to climate change. Tropical forests
 and Arctic permafrost contain the highest ecosystem stocks of aboveground and belowground carbon,
 respectively. Primary tropical forests currently emit more carbon to the atmosphere than they remove due to
 deforestation and forest degradation. Wildfires in the Arctic are contributing to permafrost thaw and soil carbon
 release. An emissions scenario of 2ºC increase could thaw ~15% of permafrost area and emit 20-100 Gt carbon by
 2100, Under emissions scenarios of a 4ºC global temperature increase, models project possible tipping points of
 conversion of half of Amazon rainforest to grasslands and thawing of Arctic permafrost that could release 11-200
 Gt carbon that could substantially exacerbate climate change. (Sections 2.4.3,,,–10;–
 5; Figure 2.9; Figure 2.10; Figure 2.11; Table 2.4; Table 2.5; Table 2.S.2; Table 2.S.3; Table 2.S.4)

 Climate factors              Non-climate        Detection and         Adaptation               Risk transitions
                              factors            attribution                                    (confidence)

 Increase in temperature,     Deforestation,     Losses of carbon      Reduce deforestation,    0.75°C
 increase in aridity,         road and           detected in           especially in tropical   undetectable-
 increase in frequency and    infrastructure     boreal, temperate,    forests, reduce road     moderate (medium)
 severity of drought          expansion,         and tropical          and infrastructure       2°C: moderate-high
                              agricultural       ecosystems, due       expansion, especially    (medium)
                              expansion          to wildfire and       in the Arctic, reduce    4°C high-very high
                                                 tree mortality, not   use of fire to clear     (low)
                                                 formally              agricultural land,
                                                 attributed to         increase protected
                                                 climate change        areas

Do Not Cite, Quote or Distribute                          2-101                                          Total pages: 237
    FINAL DRAFT                                          Chapter 2                     IPCC WGII Sixth Assessment Report

     Ecosystem Structure Change: Increasing risk of large-scale changes in ecosystem structure. Ecosystem
     structural change with most information derived for tropical forest, boreal forest, savannas, and tundra for both
     observations and future projections. The transition from non-detectable to moderate is based on detected changes
     attributable to climate change, or interactions between changing disturbance regime, climate and rising CO2,
     already observed at 0.5°C above pre-industrial, with shifts initially detected in boreal forests, tundra, and tropical
     grassy ecosystems. Transition from moderate to high is centred around 1.5°C based on widespread global
     observations (at current GSAT of 1.09°C above pre-industrial) that agree with projected future impacts with at
     least 10% area of key ecosystems being affected (Box 2.1). Overall confidence in projections is medium, based on
     existing observations and projections giving high confidence of risk for several ecosystems but because data and
     projections are not available for all biomes, overall confidence lowers to medium. Transition from high to very
     high occurs when more than 50% of multiple ecosystems are projected to experience shifts in structure. (Sections, 2.4.3, 2.4.5, 2.5.2, Box 2.1, Figure Box 2.1.1, Table Box 2.1.1, Table 2.S.2, Table 2.S.2, Table 2.S.3,
     Table 2.S.5)

     Climate factors               Non-climate        Detection and        Adaptation                 Risk transitions
                                   factors            attribution                                     (confidence)

     Increases in average and      Land use           Individual           Conservation of            0.5°C undetectable-
     extreme temperatures,         change,            species ranges       potential refugia,         moderate (high)
     changes in precipitation      livestock          shifts, biome        habitat restoration,       1.5°C moderate-
     amount and timing,            grazing,           shifts               increasing                 high (medium)
     increased atmospheric         deforestation,                          connectivity of            2.5°C high-very
     CO2                           fire                                    habitats and protected     high (medium)
                                   suppression,                            areas, increase in
                                   loss of native                          protected areas,
                                   herbivores,                             changes in grazing and
                                   food, fiber, or                         fire management

    Do Not Cite, Quote or Distribute                          2-102                                           Total pages: 237
     FINAL DRAFT                                         Chapter 2                     IPCC WGII Sixth Assessment Report

 2   Figure 2.11: Key risks to terrestrial and freshwater ecosystems from climate change. This IPCC chapter assesses these
 3   as the most fundamental risks of climate change to ecosystem integrity and the ecosystem services that support human
 4   well-being, based on observed impacts and future risks of: (far left) Losses of animal and plant species from different
 5   ecosystems globally with resulting declines in ecosystem integrity, functioning and resilience (Section,,
 6; (middle left) wildfire exceeding natural levels (Section,; (middle) tree mortality exceeding
 7   natural levels (Chapter,, (middle right) ecosystem carbon losses that could occur abruptly and
 8   substantially raise atmospheric carbon dioxide (Sections—,,—,,;
 9   (far right) major changes occurring in ecosystem structure (Sections 2.4.3, Box 2.1, 2.5.2, Figure 2.9, Figure Box 2.1.1,
10   Table Box 2.1.1, Table 2.S.5). This burning embers diagram shows impacts and risks in relation to changes in global
11   mean surface temperature, relative to the pre-industrial period (1850-1900). Risk levels reflect current levels of
12   adaptation and do not include more interventions that could lower risk. The compound effects of climate change
13   combined with deforestation, agricultural expansion, urbanisation, air, water, and soil pollution, and other non-climate
14   hazards could increase risks. Tables 2.5 and 2.S.4 provide details of the key risks and temperature levels for the risk
15   transitions.
18   [START FAQ2.4 HERE]
20   FAQ2.4: How does nature benefit human health and well-being and how does climate change affect
21       this?
23   Human health and well-being are highly dependent on the “health” of nature. Nature provides material and
24   economic services that are essential for human health and productive livelihoods, but studies also show that
25   being in “direct contact with natural environments” has direct positive effects on well-being, health and
26   socio-cognitive abilities. Therefore, the loss of species and biodiversity under climate change will reduce
27   natural space, decrease biodiversity and in turn, decrease human-well-being and health worldwide.

     Do Not Cite, Quote or Distribute                          2-103                                         Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   Human health and well-being are highly dependent on the “health” of nature. Biodiversity – the variety of
 2   genes, species, communities and ecosystems – provides services that are essential for human health and
 3   productive livelihoods, such as breathable air, drinkable water, productive oceans and fertile soils for
 4   growing food and fuels. Natural ecosystems also help store carbon and regulate climate, floods, disease,
 5   pollution and water quality. The loss of species, leading to reduced biodiversity, has direct and measurable
 6   negative effects on all of these essential services, and therefore, on humankind. A recent demonstration of
 7   this is the decline of pollinator species, with potential negative effects on crop pollination, a fundamental
 8   ecosystem function crucial for agriculture. The loss of wild relatives of the domesticated varieties humans
 9   rely on for agriculture reduces the genetic variability that may be needed to support the adaptation of crops
10   to future environmental and social challenges.
12   The number of species that can be lost before negative impacts occur is not known and is likely to differ in
13   different systems. However, in general, more diverse systems are more resilient to disturbances and able to
14   recover from extreme events more quickly. Biodiversity loss means there are fewer connections within an
15   ecosystem. A simpler food web with fewer interactions means less redundancy in the system, reducing the
16   stability and ability of plants and animal communities to recover from disturbances and extreme weather
17   events such as floods and drought.
19   In addition to “material” and economic services such as eco-tourism, nature also provides cultural services
20   such as recreation, spirituality and well-being. Specifically, being in “direct contact with natural
21   environments” (versus urban environment) has a high positive and causal impact on human well-being (e.g.
22   mood, happiness), psychological and physical health (energy, vitality, heart rate, depression) and socio-
23   cognitive abilities (attention, memory, hyperactivity, altruism, cooperation). Therefore, the loss of species
24   and biodiversity under climate change and urbanisation will reduce natural space, decrease biodiversity and
25   in turn, decrease human-well-being and health worldwide.
27   Finally, the extent to which humans consider themselves as part of the natural world – known as human-
28   nature connectedness – has been demonstrated to be closely associated with human health and well-being.
29   Individuals who are more connected to nature are not only happier and healthier but also tend to engage
30   more in pro-nature behaviours, making the enhancement of human-nature connectedness worldwide a
31   valuable win-win solution for humans and nature to face environmental challenges.


     Do Not Cite, Quote or Distribute                     2-104                                      Total pages: 237
     FINAL DRAFT                                        Chapter 2                    IPCC WGII Sixth Assessment Report

 1   Figure FAQ2.4.1: Positive relationship between human health and well-being and nature conservation. Nature
 2   provides essential services to humans including material and economic services (i.e. ecosystem services) as well as
 3   cultural, experiential and recreational services, which, in turn enhance human psychological and physical health, and
 4   well-being. People who are more connected to nature are not only happier and healthier but are also more likely to
 5   engage in pro-nature behaviours, making the enhancement of human-nature connectedness worldwide a valuable win-
 6   win solution for humans and nature to face environmental challenges.
 8   [END FAQ2.4 HERE]
11   2.6    Climate Change Adaptation for Terrestrial and Freshwater Ecosystems
13   Adaptation to reduce vulnerability of ecosystems and their services to climate change has been addressed in
14   previous IPCC Reports, with AR4 and AR5 recognising both autonomous adaptation and human assisted
15   adaptation to protect natural species and ecosystems. In AR5, Ecosystem-based Adaptation (EbA),
16   adaptation for people, based on better protection, restoration and management of the natural environment,
17   was identified as an area of emerging opportunity, with a dedicated Cross Chapter Box on the topic. In the
18   SRCCL report, conservation, EbA and related concepts were integrated throughout the report; SR1.5 also
19   noted the role of EbA. Since the last assessment report the scientific literature has expanded considerably,
20   with growing interest in the concept of Nature-based Solutions (NbS). This section assesses this new
21   literature and its implications for the implementation of climate change adaptation.
23   Previous sections of this chapter have set out the vulnerability of natural and semi-natural ecosystems to
24   climate change and the risks this poses both to biodiversity and ecosystem services (also sometimes
25   described as ‘Nature’s Contributions to People’). Natural systems respond to climatic and other
26   environmental changes in variety of ways. Individual organisms can respond through growth, movement
27   and developmental processes. Species and populations genetically adapt to changing conditions and evolve
28   over successive generations. Geomorphological features, such as the path of watercourses, can also change
29   naturally in response to climate change. However, there is a limit to which these natural processes can
30   maintain biodiversity and the benefits people derive from nature, partly because of intrinsic limits, but also
31   because of the pressures that people exert on the natural environment. Most of this section therefore focuses
32   on human interventions to build the resilience of ecosystems, enable species to survive or to adjust
33   management to climate change. Vulnerability is in many cases exacerbated by the degraded state of many
34   ecosystems as a result of human exploitation and land use change, leading to fragmentation of habitats, loss
35   of species and impaired ecosystem function. This interaction between climate change and environmental
36   degradation means that protecting ecosystems in a natural or near-natural state will be an important pre-
37   requisite for maintaining resilience and give many species the best chance of persisting in a changed climate
38   (Belote et al., 2017; Arneth et al., 2020; Ferrier et al., 2020; França et al., 2020). Protection from
39   degradation, deforestation and exploitation is also essential to maintain critical ecosystem services, including
40   carbon storage and sequestration and water supply (Dinerstein et al., 2020; Pörtner, 2021).
42   It is worth briefly considering some key concepts that are relevant to adaptation in ecosystems. Adaptation
43   for biodiversity and ecosystems can encompass both managing change and building resilience. We use the
44   definition of ‘resilience’ set out Chapter 1: ‘the capacity of social, economic and environmental systems to
45   cope with a hazardous event or trend or disturbance, responding or reorganising in ways that maintain their
46   essential function, identity and structure while also maintaining the capacity for adaptation, learning and
47   transformation’, It includes the concept of ‘resistance’, which is used in some ecological literature to
48   distinguish systems which a resistant to change from those that recover quickly from change. We consider
49   both interventions designed primarily to protect biodiversity and those intended to reduce the risks of climate
50   change to people.
52   A variety of terms are used to describe using environmental management reduce the impacts of climate
53   change on people in ways that also benefit biodiversity in the scientific literature, particularly Ecosystem
54   based Adaptation (EbA) and Nature-based Solutions (NbS) (see also Section 1.4). EbA is the use of
55   biodiversity and ecosystem services as part of an overall adaptation strategy to help people to adapt to
56   climate change (CBD, 2009). EbA aims to maintain and increase the resilience and reduce the vulnerability
57   of ecosystems and people in the face of the adverse effects of climate change (Vignola et al., 2009). NbS is a
58   broader term which is not restricted to climate change and is also often used to refer to climate change

     Do Not Cite, Quote or Distribute                        2-105                                         Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   mitigation; it has been defined by the International Union for the Conservation of Nature (IUCN) as ‘Actions
 2   to protect, sustainably manage and restore natural or modified ecosystems that address societal challenges
 3   effectively and adaptively, simultaneously providing human well-being and biodiversity benefits’ (Cohen-
 4   Shacham et al., 2016). This widely accepted definition excludes actions, which use the natural environment
 5   to solve human problems but do not provide benefits for biodiversity and is closely linked to the concept of
 6   the Ecosystem Approach. NbS is not a universally accepted term but it is increasingly used in the scientific
 7   literature. It is a concept which recognises the importance of biodiversity in ecosystem service provision and
 8   offers the opportunity to address climate change and loss of biodiversity together in an efficient integrated
 9   way (Chong, 2014; Seddon et al., 2020a; Ortiz et al., 2021). Given the focus of this chapter is on adaptation
10   we primary use the term EbA as it is more specific, but we do so understanding that it can be regarded as a
11   subset of NbS. The wider concept of NbS for climate change adaptation and mitigation is covered in a Cross-
12   Chapter Box on the topic (see Cross-Chapter Box NATURAL this Chapter).
14   Whilst we distinguish between adaptation for biodiversity and EbA, it is important to recognise that the two
15   are linked in that if ecosystems themselves are not resilient to climate change, they will not be able to
16   provide adaptation benefits for people. The case for resourcing biodiversity conservation and building the
17   resilience of ecosystems is also strengthened when there are direct benefits for people in addition to the more
18   general benefits of biodiversity.
20   Ecosystems are specifically included in the adaptation goals set out in the Paris Agreement and are addressed
21   in most national adaptation plans (Seddon et al., 2020b). There is also now a large number of adaptation
22   programmes and plans for local governments and governmental and non-governmental conservation
23   organisations. Adaptation for and by ecosystems needs to be understood and developed in the wider contexts
24   of conservation, Climate Resilient Development and Sustainable Development: there is significant potential
25   synergies, but also conflicts between different objectives, which require an integrated approach (covered
26   further in 2.6.7).
28   2.6.1   Limits to Autonomous (Natural) Adaptation
30   Natural ecosystems often have a high degree of resilience and can to some extent adjust to change. Species
31   can adjust through evolutionary adaptation, distribution change, behavioural change, developmental
32   plasticity and ecophysiological adjustment. There are, however, limits to autonomous adaptation, because of
33   intrinsic limitations, the rate at which the climate is changing and the degraded state of many ecosystems.
35   None of the evolutionary changes either documented or theorised would enable a species to survive and
36   reproduce in climate spaces that it does not already inhabit. Evolutionary responses are very unlikely to
37   prevent species extinctions in the case of that species losing its climate space entirely on a regional or global
38   scale (Parmesan and Hanley, 2015). At highest risk are the world's most cold-adapted species (whose
39   habitats are restricted to polar and high mountaintop areas). Examples include the polar bear (Regehr et al.,
40   2016), "sky-island" plants in the tropics (Kidane et al., 2019), mountain-top amphibians in Spain (Enriquez-
41   Urzelai et al., 2019), mountain-top lichens in the Appalachians (USA) (Allen and Lendemer, 2016), and
42   silverswords in Hawaii (Krushelnycky et al., 2013). However, there is potential for using evolutionary
43   changes to enhance the adaptive capacity of target species, as is being done in the Great Barrier reef by
44   translocating symbionts and corals that have survived recent intense heat-induced bleaching events into areas
45   that have had large die-off (Rinkevich, 2019). Multiple studies assessed when and how evolution might be
46   able to help wild species adapt to climate change (Ratnam et al., 2011; Sgro et al., 2011).
48   Some of the reasons cited in the literature as limits to autonomous adaptation are:
50   1) Genetic changes in populations require many generations and for many species operate on longer time
51   scales than those, on which the climate is currently changing.
53   2) Many species are moving to higher latitudes as the climate warms, but not all are keeping pace with
54   changes in suitable climate space (Valladares et al., 2014; Mason et al., 2015). Such climate debt indicates
55   an inability for non-genetic autonomous adaptation (e.g. evidence-limited ability for plastic responses, such
56   as stemming from dispersal limitations, or behavioural restrictions, or physiological constraints).

     Do Not Cite, Quote or Distribute                     2-106                                       Total pages: 237
     FINAL DRAFT                                         Chapter 2                    IPCC WGII Sixth Assessment Report

 1   3) Some species have low capacity for dispersal, which, combined with increased fragmentation of habitats,
 2   creates barriers to range shifts to match climate warming. Studies have shown that changes in distribution of
 3   species and composition of communities are limited by the presence of intensively managed agricultural land
 4   fragmenting natural habitats (Oliver et al., 2017).
 6   There are a variety of mechanisms which promote the resilience of ecosystems through persistence,
 7   recovery, and reorganisation (Falk et al., 2019). Changes in the balance of different plant species within a
 8   community can maintain the persistence of the community itself, maintaining its value as a habitat for other
 9   species and providing ecosystem services (add reference?). In some cases there are negative feedback
10   mechanisms between biological and physical processes, for example in peatlands, lowered water tables
11   resulting from drier conditions can lead to reduced permeability of peat, increasing rates of water loss (Page
12   and Baird, 2016). There are limits to this resilience and the concept of tipping points beyond which
13   ecosystems change state and returning to the original state has been subject of much recent research (van Nes
14   et al., 2016). There is clear evidence that the degradation of ecosystems has reduced their resilience and
15   restoration can help to reduce risks to biodiversity and ecosystem services, discussed below (see Section
16   2.6.2, 2.6.3). However, as rates of climate change increase, the limits of this approach will start to be
17   reached and losses, including some with potentially catastrophic consequences, cannot be prevented; this is
18   discussed further in Section 2.6.6.
20   2.6.2   Adaptation for Biodiversity Conservation
22   A variety of approaches have been identified as potential adaptation measures which people can take to
23   reduce the risks of climate change to biodiversity. (Heller and Zavaleta, 2009 ) (quoted in AR5) identified
24   113 categories of recommendation for adaptation from a survey of 112 papers and reports. Since this time
25   the literature has greatly expanded, with thousands of relevant publications. Whilst there is increasing
26   interest in adaptation for biodiversity conservation and a wide range of plans and strategies, there is less
27   evidence of these plans being implemented. Since AR5 a number of studies, predominantly from Europe and
28   North America, have investigated the extent to which adaptation has been integrated into conservation
29   planning and is being implemented at site and regional scale (Macgregor and van Dijk, 2014; Delach et al.,
30   2019; Prober et al., 2019; Clifford et al., 2020; Barr et al., 2021; Duffield et al., 2021). A common pattern in
31   these studies is that vulnerability has been assessed and potential adaptation actions identified, but
32   implementation has been limited beyond actions to improve ecological condition, which may increase
33   resilience at a local scale.
35   To date most scientific literature on adaptation to reduce risk to biodiversity from climate change has been
36   based on ecological theory rather than observations or practical experience. A recent review (Prober et al.,
37   2019) concluded that out of 473 papers on adaptation, only 16% presented new empirical evidence and very
38   few assessed the effectiveness of actual adaptation actions. It is also the case that relatively little research is
39   focussed on local-scale management interventions rather than larger scale strategies (Ledee et al., 2021),
40   although there are some exceptions (Duffield et al., 2021).
42   Although direct assessments of the effectiveness of adaptation actions are rare, since AR5 there has been an
43   increasing number of empirical analyses of how different land use and management influences the
44   vulnerability of species and habitats. As climate change often interacts with other factors including
45   ecosystem degradation and fragmentation (Oliver et al., 2015a), actions to address these other interacting
46   factors is expected to build resilience to climate change. Table 2.6 summarises evidence that supports the
47   main categories of proposed adaptation measures. We have taken an inclusive approach and included studies
48   that address extreme weather events such as droughts, which may be exacerbated by climate change as well
49   as long-term changes in climate variables. We have not distinguished between studies in which climate
50   change adaptation was an explicit focus and those in which lessons for adaptation can be learnt from studies
51   conducted for other reasons but inform the assessment of the impacts of actions identified as potential
52   adaptation measures.
55   Table 2.6: Evidence to support proposed climate change adaptation measures for biodiversity. Highlights that
56   adaptation for biodiversity is a broad concept, encompassing a wide range of actions. It includes targeted interventions
57   to change the microclimate for particular species (for example by shading); through to changing national conservation
58   objectives to take account of changing distributions of species and communities. It includes targeted actions addressing

     Do Not Cite, Quote or Distribute                         2-107                                          Total pages: 237
    FINAL DRAFT                                          Chapter 2                      IPCC WGII Sixth Assessment Report

1   both climate change and protection and restoration of ecosystems, with multiple additional benefits including reducing
2   vulnerability to climate change. Most of the studies are not direct tests of the impacts of adaptation actions, which, as
3   noted above, is an important evidence gap. There is also a major limitation in that reported studies are predominantly
4   from Europe, North America and Australasia, with little research in other regions.
       Proposed Adaptation        Confidence                         Comment                           Selected References
            Measures for          Assessment
      Protect large areas of     Robust           Considerable evidence that intact systems          Dinerstein et al. (2019);
      natural and semi-          evidence, high provide better quality and quantity of               Woodley et al. (2019);
      natural habitat            agreement        ecosystem services; that larger intact areas       Brooks et al. (2020);
                                                  provide better ecosystem services; that            Zhao et al. (2020); Sala
                                                  species' extinction risk with disturbances,        et al. (2021); Pimm et al.
                                                  including climate change, are reduced by           (2018); Hannah et al.
                                                  having large, connected populations; that          (2020); Luther et al.
                                                  more biodiverse systems provide higher
                                                  levels of ecosystem services and are more
                                                  resilient to climate change than degraded
                                                  systems that have lost species
      Increase connectivity in Medium             Good evidence that some species move more Keeley et al. (2018);
      terrestrial habitats –     evidence,        quickly in more connected landscapes.              Stralberg et al. (2019);
      corridors, stepping        Medium           However, not all species do and some               von Holle et al. (2020)
      stones                     agreement        species that benefit are invasive / pest /
                                                  disease species and there is limited empirical
                                                  evidence showing connectivity has reduced
                                                  climate change impacts on species to date.
      Increase connectivity in Limited            Connectivity is needed to maintain species         Hermoso et al. (2016);
      river networks             evidence,        and population movements, but river reaches Thieme et al. (2016);
                                 High             and catchments lack integrated protection.         Abell et al. (2017);
                                 agreement                                                           Brooks et al. (2018)
      Increase habitat patch     Limited          Generally increase resilience because of           Eigenbrod et al. (2015);
      size site and expand       evidence         functioning natural processes, large species       Oliver et al. (2015a)
      protected areas            High             populations and refugial areas
      Increase replication       Limited          Various benefits inferred, including, wider        Mawdsley et al. (2009);
      and representation of      evidence,        range of climatic and other conditions, less       Thomas et al. (2012a);
      protected areas            High             risk of extreme events affecting many rather Virkkala et al. (2014);
                                 agreement        than few areas. More sites available for           Gillingham et al. (2015);
                                                  colonisation by range expanding species and Pavón-Jordán et al.
                                                  better conditions to maintain species in situ      (2020b)
                                                  under range contraction.
      Protect microclimatic      Medium           Locally cool areas can be identified and           Haslem et al. (2015);
      refugia                    evidence,        there is evidence species can survive better       Suggitt et al. (2015);
                                 High             in such areas.                                     Isaak et al. (2016);
                                 agreement                                                           Morelli et al. (2016);
                                                                                                     Merriam et al. (2017);
                                                                                                     Bramer et al. (2018);
                                                                                                     Suggitt et al. (2018);
                                                                                                     Massimino et al. (2020)
      Creating shade to          Limited          Creating shade has been used as an                 Broadmeadow et al.
      lower temperatures for     evidence,        adaptation strategy, for example by                (2011); Lagarde et al.
      vulnerable species         High             watercourses but improvements in species           (2012); Patino-Martinez
                                 agreement        survival under warming conditions have yet         et al. (2012); Thomas et
                                                  to be demonstrated.                                al. (2016)
      Restoring hydrological Medium               Wetland restoration is well established as a       Carroll et al. (2011);
      processes of wetlands,     evidence,        conservation measure in some countries. Can Hossack et al. (2013);
      rivers and catchments,     High             reduce vulnerability to drought with climate       Dokulil (2016);
      including by raising       agreement        change but evidence to demonstrate                 Timpane-Padgham et al.
      water tables and                            effectiveness as an adaptation measure is          (2017); Moomaw et al.
      restoring original                          limited and requires long-term monitoring of (2018)
      channels of                                 a range of sites. Little restoration of
      watercourses,                               degraded tropical peatlands to date
      Restoration of natural     Medium           Includes reintroduction of native herbivores       Coffman et al. (2014);
      vegetation dynamics        evidence,        and reversing woody encroachment of                Valkó et al. (2014);

    Do Not Cite, Quote or Distribute                           2-108                                           Total pages: 237
     FINAL DRAFT                                      Chapter 2                     IPCC WGII Sixth Assessment Report

                                 Medium         savannas. Benefits for biodiversity are well     Batáry et al. (2015);
                                 agreement      established in a wide range of different         Smit et al. (2016);
                                                regions                                          Stevens et al. (2016);
                                                                                                 Hempson et al. (2017);
                                                                                                 Bakker and Svenning
                                                                                                 (2018); Cromsigt et al.
                                                                                                 (2018); Fulbright et al.
                                                                                                 (2018); Olofsson and
                                                                                                 Post (2018)
      Reduce non-climatic        Limited        As a general principal climate change is         Oliver et al. (2017);
      stressors to increase      evidence,      recognised as a ‘threat multiplier’ but          Pearce-Higgins et al.
      resilience of ecosystems   Medium         specific details are often unclear               (2019)
      Assisted translocation     Limited        Assisted translocation has been commonly         Willis et al. (2009);
      and migration of           evidence,      suggested as an adaptation measure, but          Brooker et al. (2018);
      species                    Medium         there have been very few examples of this        Skikne et al. (2020)
                                 agreement      being trialed. Translocations have been
                                                carried out for other reasons and lessons for
                                                climate change adaptation have been
      Intensive management       Medium         A variety of approaches including                Angerbjörn et al. (2013);
      for specific species       Evidence,      manipulating microclimate and competition        Greenwood et al. (2016);
                                 Medium         between species to improve chances of            Pearce-Higgins et al.
                                 Agreement      survival under climate change.                   (2019)
      Ex-situ conservation       Not possible   Seed banks have been established but long-       Christmas et al. (2016)
      (seedbanks/genetic         to assess at   term effectiveness could only be evaluated at
      stores, etc.)              present time   a later point.
      Adjusting conservation     Robust         Conservation management will need to take        Stein et al. (2013);
      strategies and site        evidence,      account of changes that cannot be prevented,     Rannow et al. (2014);
      objectives to reflect      High           for example in the distribution of species and   Oliver et al. (2016);
      changing species           Agreement      composition of communities, in order to          Stralberg et al. (2019);
      distributions and                         protect and manage biodiversity as               Duffield et al. (2021)
      habitat characteristics                   effectively as possible in a changing climate.

      Softening the matrix of    Limited        Potential for agri-environment schemes to do     Donald and Evans
      unsuitable habitats        evidence       this in hostile farmed landscapes.               (2006); Stouffer et al.
      between patches to                                                                         (2011)
      increase permeability
      for species movement in
      response to climate
 3   Many climate adaptation actions for biodiversity operate at the landscape scale (von Holle et al., 2020). The
 4   total area of habitat, how fragmented it is, the size of habitat patches and the connectivity between them are
 5   interlinked properties at this scale. A growing number of studies have investigated how these properties
 6   affect species ability to persist in situ and colonise new areas. Overall, larger areas of semi-natural habitat are
 7   associated with increased resilience to ongoing climate change and extreme events and the capacity to
 8   colonise new areas (Haslem et al., 2015; Oliver et al., 2017; Papanikolaou et al., 2017). Larger habitat
 9   patches can support larger populations, which are more likely to maintain themselves and recover from
10   periods of adverse conditions. A large patch size has been found to increase resilience of some populations
11   of species to extreme events such as droughts (Oliver et al., 2015b). They are also more likely to provide a
12   range of different resources and microclimate conditions, which may increase chances of species persistence
13   under climate change. A larger area of habitat may also enables greater connectivity between patches and
14   increases the chances of species colonising new areas as they track climate change. (Oliver et al.,
15   2015b)Protecting and restoring natural processes is a general principle for maintain and building resilience to
16   climate change for biodiversity (Timpane-Padgham et al., 2017). One element of this is ensuring naturally
17   functioning hydrology for wetlands and river systems (Table 6), which is particularly important in a context
18   of changing rainfall patterns and increased evapo-transpiration. An important development in approaches to

     Do Not Cite, Quote or Distribute                      2-109                                           Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1   conservation over recent decades has been the concept of re-wilding (Schulte To Bühne et al., 2021); this
 2   encompasses a number of elements of restoring natural processes, including the reintroduction of top
 3   predators, larger conservation areas and less prescriptive outcomes than much previous conservation. There
 4   are elements of re-wilding which may well contribute to building resilience to climate change but it will be
 5   increasingly important to factor climate change adaptation into the planning of rewilding schemes (Carroll
 6   and Noss, 2021).
 8   The most consistently cited climate change adaptation measure for species is increasing connectivity to
 9   facilitate colonisation of new areas. This reflects the fact that many species’ habitats are highly fragmented
10   in areas with more intensive land management, which prevents them naturally changing their range to track
11   changing climatic conditions. Advances and innovations in modelling techniques can support decision
12   making on connectivity (Littlefield et al., 2019). There is evidence from empirical as well as modelling
13   studies that species can disperse more effectively in better connected areas in terrestrial habitats (Keeley et
14   al., 2018). The issues are different in more natural landscapes—species may still be threatened in
15   intrinsically isolated habitats, such as mountain top, but connectivity cannot be created in the same way.
16   Evidence suggests that increased connectivity will only benefit a subset of species, probably those which are
17   intermediate habitat specialists that are able to disperse (Pearce-Higgins et al., 2014). Generalists do not
18   require corridors or stepping stones whilst many corridors or stepping stones will not be of sufficient quality
19   to be used by the most habitat specialists. There should also be a caveat to the general principle that
20   increasing connectivity is a benefit for climate change adaptation. It can increase the spread of invasive, pest
21   and disease-causing species into newly suitable regions. In some places isolated refugia may better allow
22   vulnerable species and biological communities to survive.
24   There are many different approaches to increasing connectivity, ranging from increasing overall area of
25   suitable habitat through to ‘corridors’ and ‘stepping stones’, with different strategies likely to be more
26   effective for different species and circumstances (Keeley et al., 2018). Connectivity can also be important in
27   increasing resilience of populations to extreme climatic events (Newson et al., 2014; Oliver et al., 2015b).
28   Within freshwater environments, connectivity of watercourses is essential. Fluvial corridors are necessary to
29   ensure migrating fish population survival, even without climate change; with climate change, connectivity
30   becomes crucial for relatively cold-adapted organisms to migrate upstream to colder areas. Connectivity is
31   also important for the larvae of benthic invertebrates to be able to drift downstream and hence to disperse
32   (Brooks et al., 2018); for adult benthic invertebrates, riparian and terrestrial habitat features can potentially
33   affect dispersal. Connectivity within river and wetland systems for some species can also mediated by more
34   mobile animal species such as fish and birds (Martín-Vélez et al., 2020) Which factors are the most
35   important in either promoting their colonisation of new sites or persisting in situ will differ between species
36   and locations. Some general principle have been recognised and can guide conservation policy and practice
37   (England and RSPB, 2020; Stralberg et al., 2020) but this will often require additional investigation and
38   planning based on understanding individual the niche of specific species.
40   Managed, translocation by moving species from areas where the climate is becoming unsuitable to places
41   where there persistence under climate change is more likely has been discussed as an adaptation option for
42   many years. So far there have been very few examples of this and it is likely to be a last resort in most cases
43   as in many cases it requires a large investment of resources, the outcome is uncertain and there may be
44   adverse impacts on receiving sites. Nevertheless there are cases where it may be a be a viable option
45   (Stralberg et al., 2019). This is discussed in more detail as a case study in section
47   The evidence that species can persist in microclimatic refugia where suitable conditions for them are
48   maintained locally (for example, because of variations in topography) has increased in recent years. This has
49   opened up the potential to include refugia in conservation plans and strategies to facilitate local survival of
50   species (Jones et al., 2016; Morelli et al., 2016; Morelli et al., 2020). For example, in targeting management
51   actions (Sweet et al., 2019) aimed at supporting populations of species and is likely to become an important
52   aspect of climate change adaptation for biodiversity conservation in future. It is also possible to manipulate
53   microclimate for example by creating shelters for birds’ nests (see case study of African Penguins
54   (Patino-Martinez et al., 2012). One specific approach of this sort is planting or retention of trees and wooded
55   corridors to shade water courses (see also case study below; Thomas et al., 2016). In the latter case,
56   riparian shading can also possibly help to reduce phytoplankton and benthic diatom growth in smaller
57   streams and rivers (Halliday et al., 2016).

     Do Not Cite, Quote or Distribute                      2-110                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 2   Refugia often refer to locally places in a landscape, such as on shaded slopes or high elevations, but they can
 3   also include places where water supply may continue during dry periods (Morelli et al., 2016) Monitoring
 4   can reveal which streams, wetlands, springs, and other aquatic resources retain suitable discharges, water
 5   quality, wetland area, and ecological integrity, especially during dry years (Cartwright et al., 2020).
 6   Measures to conserve drought refugia may include protecting springs and other groundwater-fed systems
 7   from groundwater extraction, contamination, salinisation, surface-water diversion, channelisation of
 8   streams, livestock trampling, recreation and invasive species, as well as effects from surrounding landscape
 9   disturbances (Cartwright et al., 2020; Krawchuk et al., 2020). Restoration of degraded aquatic ecosystems
10   can include removal of flow-diversion infrastructure, exclusion of livestock, reduction of other human
11   impacts, geomorphic restructuring, invasive species removal, and planting of native riparian vegetation.
13   In fire prone areas, fire suppression and management is a key element of protecting refugia ( Section
14   below). In ecosystems in which a natural fire regime has been suppressed restoration practices such as
15   prescribed fires, thinning trees, and allowing some wildfires where it benefits the ecosystem can be
16   introduced to reduce increasing risks from severe wildfires (Meigs et al., 2020).
18   Protected areas—areas of land set aside for species and habitat protection with legal protection from
19   development or exploitation—have been a cornerstone of nature conservation for many years. Their
20   effectiveness under a changing climate has been the subject of debate and investigation. There is now a large
21   body of evidence demonstrating that colonisations by range shifting species are more likely to occur on
22   protected sites compared to non-protected sites for a wide range of taxa (e.g. Thomas et al., 2012b;
23   Gillingham et al., 2015), including across continents (Pavón-Jordán et al., 2020a). This is probably because
24   by protecting large areas of natural and semi-natural habitats they provide suitable places for colonising
25   species (Hiley et al., 2013) which may not be available in the surrounding landscape. Although the evidence
26   for protected areas being associated with reduced extinctions is weaker, the finding in Gillingham et al.
27   (2015) that protected sites were associated with reduced extinction rates at low latitudes and elevations is
28   strongly suggestive that they can help species' persistence in the face of climate change.
30   It is intrinsically difficult to assess the effectiveness of climate change adaptation measures, the benefit of
31   which will be realised in years and decades ahead (Morecroft, 2019, Measuring success). Nevertheless,
32   taking account of the wide range of evidence reported above, including theory, modelling and observations
33   of climate change impacts in contrasting circumstances, it is possible to make an overarching assessment that
34   appropriate adaptation measures can reduce the vulnerability of many aspects of biodiversity to climate
35   change (robust evidence, high agreement). It is also however clear that to be most effective and avoid
36   unintended consequences, measures need to be carefully implemented taking account of specific local
37   circumstances (robust evidence, high agreement) and include the management of inevitable changes (robust
38   evidence, high agreement). It is also clear that whilst there are now many plans and strategies for adapting
39   biodiversity conservation to climate change, many have yet to be implemented fully (medium evidence, high
40   agreement).
42   2.6.3   Nature-based Solutions: Ecosystem-based Adaptation
44   Ecosystem-based Adaptation is an increasingly important element of Nature-based Solutions (see 2.6 above).
45   A study published in 2020 found that out of 162 Intended Nationally Determined Contributions (covering
46   189 countries) submitted to the United Nations Framework Convention on Climate Change, as commitments
47   to action under the Paris Agreement, 109 indicated ‘ecosystem-orientated visions’ for adaptation, although
48   only 23 use the term ‘Ecosystem-based Adaptation’ (Seddon et al., 2020b).
50   EbA includes a range of different approaches. Examples include restoring coastal and river systems to
51   reduce flood risk and improve water quality and the creation of natural areas within urban areas to reduce
52   temperatures through shading and evaporative cooling. EbA is closely linked with a variety of other concepts
53   such as ecosystem services, natural capital and Disaster Risk Reduction (DRR). EbA was becoming a well-
54   recognised concept at the time of AR5 but implementation was still at an early stage in many cases. Since
55   then pilot studies have been assessed and EbA projects have been initiated around the world. The evidence
56   base continues to grow (Table 2.7) and this has led to increasing confidence in approaches which have been
57   shown to work leading to further expansion in some countries (Table 2.7). However, this is not uniform and

     Do Not Cite, Quote or Distribute                     2-111                                      Total pages: 237
     FINAL DRAFT                                        Chapter 2                       IPCC WGII Sixth Assessment Report

 1   there is relatively little synthesis across disciplines and regions (Seddon et al., 2020a). Chausson et al. (2020)
 2   used a systematic mapping methodology to characterise 386 published studies. They found that interventions
 3   in natural or semi-natural ecosystems ameliorated adverse climate change impacts in 66% of cases, with
 4   fewer trade-offs than for more artificial systems such as plantation forest. However, the evidence base has
 5   substantial gaps. Most of the evidence has been collected in the Global North and there is a lack of robust,
 6   site-specific investigations of the effectiveness of interventions compared to alternatives and of more holistic
 7   appraisals accounting for broader social and ecological outcomes.
10   Table 2.7: Examples of key Ecosystem- based Adaptation measures with assessments of confidence. Note only
11   adaptation related services are shown – many measures also provide a range of other benefits to people. All also
12   provide benefits for biodiversity.
                                                               Climate         Social    Relevant
                                     Services for
       Ecosystem Based    Confidence                           Change         Benefits  Ecosystems           Selected
      Adaptation Measures Assessment                           Impact          from        and              References
                                                              addressed      adaptation  contexts
     Natural Flood risk         Medium       Flood          Increased        Reduction of Multiple      Iacob et al. (2014);
     management in river        evidence     regulation;    rainfall         flood                      Meli et al. (2014);
     systems –restoring         Medium       sediment       intensity        damage                     Burgess-Gamble et
     natural river courses      agreement    retention;                      Increased                  al. (2017); Dadson
     (removing                               water storage;                  water                      et al. (2017);
     canalisation), restoring                water                           security                   Rowiński et al.
     and protecting                          purification                    (quality and               (2018)
     wetlands and riparian                                                   supply)
     Shade rivers and           Medium       Provision of    Warmer       Food              Multiple    Broadmeadow et
     streams by restoration     evidence     fish stocks     water        security                      al. (2011); Isaak et
     of riparian vegetation     High                         temperatures income                        al. (2015);
     or trees.                  agreement                                 benefits                      Williams et al.
                                                                                                        (2015b); Thomas
                                                                                                        et al. (2016)
     Managed realignment        Robust       Coastal storm   Rising sea      Protection of Coastal      Høye et al. (2013);
     of coastlines; re-         evidence     and flood       level           life, property             Spalding et al.
     establishing and           High         protection      Increasing      and                        (2014); Narayan et
     protecting coastal         agreement    Coastal         storm energy    livelihoods                al. (2016); Morris
     habitats including                      erosion                         Water                      et al. (2018);
     mangroves, salt marsh,                  control                         security                   Chowdhury et al.
                                             Salt water                                                 (2019); Powell et
                                             intrusion                                                  al. (2019)
     Agroforestry and other     Medium       Local climate   High            Food           Multiple    Vignola et al.
     agro-                      evidence     regulation;     temperature     security                   (2015); Torralba et
     ecological/conservation    Medium       soil            or changing     income                     al. (2016); Paul et
     agricultural practices     agreement    conservation;   temperature     benefits                   al. (2017); Blaser
     on agricultural land                    soil nutrient   regimes                                    et al. (2018);
                                             regulation;     Changing                                   Nesper et al.
                                             water           precipitation                              (2019); Verburg et
                                             conservation;   regimes                                    al. (2019);
                                             pest control;                                              Aguilera et al.
                                             food                                                       (2020); Tamburini
                                             provisioning                                               et al. (2020)
     Restore and maintain       Robust       Local climate   Higher          Cooler        Urban        Norton et al.
     urban and peri-urban       evidence     regulation      temperatures    microclimate areas         (2015); Liquete et
     green space – trees,       High         Flood           and             Reduced                    al. (2016); Liu
     parks, local nature        agreement    regulation      heatwaves       flood                      (2016); Bowler et
     reserves, created                       Water           Increased       damage,                    al. (2017); Aram et
     wetlands                                purification    rainfall        water                      al. (2019);
                                             Water storage   intensity or    security

     Do Not Cite, Quote or Distribute                         2-112                                         Total pages: 237
FINAL DRAFT                                        Chapter 2                       IPCC WGII Sixth Assessment Report

                                      Erosion          reduced                                       Stefanakis (2019);
                                      control          rainfall                                      Ziter et al. (2019)
Ecological restoration    Medium      Regulation of Mega-fires         Reduce         Fire-          Waldram et al.
for fire risk reduction   evidence    wildfires     from               deaths and     adapted        (2008); Stephens et
through restoration of    High                      increases in       infrastructure ecosystems     al. (2010); van
natural vegetation and    agreement                 drought and        damage from                   Mantgem et al.
herbivory and by re-                                heat               fires                         (2016); Boisramé
instating natural fire                                                                               et al. (2017);
regimes                                                                                              Johnson et al.
                                                                                                     (2018); Parisien et
                                                                                                     al. (2020a);
                                                                                                     Parisien et al.
                                                                                                     (2020b); Stephens
                                                                                                     et al. (2020)
Invasive non-native      Robust       Water            Increasing      water          Water       van Wilgen and
aquatic plant control to evidence     provision        droughts        security       scarce      Wannenburgh
improve water security High                                                           regions     (2016)
                         agreement                                                    prone to an
                                                                                      increase in
Woody plant control       Medium      Fodder           Elevated        income        Savanna         Haussmann et al.
(of encroaching           evidence    biomass          CO2             through bush and              (2016)
biomass) in open          Medium      production       increasing      clearing,    grasslands
grassy ecosystems to      agreement                    tree growth/    fuelwood
restore and maintain                                   increases in    supplies,
grassy vegetation (see                                 rainfall        restore                                               promoting       grazing
                                                       tree growth
Rangeland                 Medium      Fodder           Changing        Food           Rangelands Descheemaeker et
rehabilitation and        evidence    biomass          precipitation   security                  al. (2010); Wairore
management such as        Medium      production;      and             Water                     et al. (2016);
through livestock         agreement   soil erosion     temperature     security,                 Kimiti et al. (2017)
enclosures, appropriate               control; soil    regimes         income
grazing management,                   formation;       including       benefits
re-introducing native                 nutrient         prolonged
grassland species                     cycling; water   dry seasons
                                      retention        and
Sustainable forestry of   Medium      Timber           Increased     Livelihood       Boreal,        Gyenge et al.
biodiverse managed        evidence    production       frequency     and income       temperate,     (2011); Barsoum et
forests, maintaining      Medium                       and severity benefits          subtropical,   al. (2016); Jactel et
forest cover and          agreement                    of storms                      tropical       al. (2017); Cabon
protecting soils                                       Higher                         forests        et al. (2018)
                                                       intensive wet
                                                       and dry
                                                       incidents of
                                                       pest, and

Do Not Cite, Quote or Distribute                        2-113                                            Total pages: 237
     FINAL DRAFT                                      Chapter 2                        IPCC WGII Sixth Assessment Report

     Watershed                 Medium      Flood control; Changing        Food            Boreal,        Filoso et al.
     reforestation and         evidence    erosion        precipitation   security;       temperate,     (2017);
     conservation for          Medium      control; water regimes         Water           subtropical,   Bonnesoeur et al.
     hydrological services     agreement   provisioning;                  security;       tropical       (2019)
                                           water                          Flood           forests
                                           purification                   Protection
     Multifunctional forest    Medium      Timber and      Multiple       Food            Boreal,        Lunga and
     management and            evidence    non-timber                     security;       temperate,     Musarurwa (2016);
     conservation to provide   Medium      forest                         Water           subtropical,   Strauch et al.
     climate resilient         agreement   production;                    security;       tropical       (2016); Adhikari et
     sources of food and                   fuel wood                      income          forests        al. (2018)
     livelihoods and protect               production;                    benefits
     water sources                         water
     Slope revegetation for    Robust      Soil retention; Increased      Reduced         Montane        Fox et al. (2011a);
     landslide prevention      evidence    slope           rain           landslide       and other      Krautzer et al.
     and erosion control       High        stabilisation   frequency      damage;         steep          (2011); Leal Filho
                               agreement                                  prevention of   sloped         et al. (2013);
                                                                          loss of life    regions        Bedelian and
                                                                                                         Ogutu (2017);
                                                                                                         Getzner et al.
                                                                                                         (2017); de Jesús
                                                                                                         Arce-Mojica et al.
 3   Restoring coasts, rivers and wetlands to reduce flood risk have probably seen the largest investment in EbA
 4   and it is becoming an increasingly accepted approach in some places (e.g. case studies in Sections,
 5 although significant social, economic and technical barriers remain (Wells et al., 2020; Bark et al.,
 6   2021; Hagedoorn et al., 2021). Natural flood management encompasses a wide range of techniques in river
 7   systems and at the coast and have been used in varied locations around the world. In tropical and sub-
 8   tropical areas, the restoration of mangroves to reduce the risk of coastal flooding is widely advocated,
 9   evidence-based approach (for example (Høye et al., 2013; Sierra-Correa and Kintz, 2015; Powell et al.,
10   2019)). In temperate regions salt marsh is a similarly important habitat (Spalding et al., 2014). Both provide
11   buffering against rising sea levels and storm surges. Managed realignment of the coast, by creating new
12   habitats can lead to a loss of terrestrial and freshwater ecosystems, but it can protect them and the services
13   they provide by reducing the risks of catastrophic failure from hard-engineered sea defences. In river
14   systems (Iacob et al., 2014) management of both catchments and the channel itself is important: restoring
15   natural meanders in canalised water courses and allowing the build-up of woody debris can slow flows rates;
16   restoring upstream wetlands or creating them in urban and peri-urban situations can store water during flood
17   events if they are in the right place in a catchment (Acreman and Holden, 2013; Ameli and Creed, 2019; Wu
18   et al., 2020). There is less data on the potential for natural flood management in tropical compared to
19   temperate catchments, however (Ogden et al., 2013) showed that flooding was reduced from a secondary
20   forested catchment compared to those which were pasture or a mosaic of forest, pasture and subsistence
21   agriculture. EbA approaches to reduce flooding can be applied within urban areas, as well as in rural
22   catchments, as in Durban (Section, although its effectiveness will depend on its being implemented
23   at a sufficient scale and in the right locations (Hobbie and Grimm, 2020; Costa et al., 2021). which may in
24   turn provide protection to downstream urban communities.
26   Protecting and restoring natural river systems, natural vegetation cover within catchments and integrating
27   agro-ecological techniques into agricultural systems can also help to maintain and manage water supplies for
28   human use, under climate change, including during drought periods, by storing water in catchments and
29   improving water quality (Taffarello et al., 2018; Agol et al., 2021; Khaniya et al., 2021) (Lara et al., 2021)
30   showed that replacing a non-native Eucalyptus plantation in Chile with native forest caused base flow
31   increased by 28% to 87% during the restoration period compared to pre-treatment, and found it remained
32   during periods with low summer precipitation

     Do Not Cite, Quote or Distribute                      2-114                                            Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 2   EbA can operate at a range of different scales, from local to catchment to region. At the local scale, there is
 3   a variety of circumstances in which microclimates can be managed and local temperatures lowered by the
 4   presence of vegetation (Table 2.7), and these EbA techniques are now being used more widely. In both urban
 5   and agricultural situations, shade trees are a traditional technique, which can be applied to contemporary
 6   climate change adaptation. Shading of water courses can lower temperatures, as reported in Section 2.6.2,
 7   above, which can allow species to survive locally, as well as supporting diversity it can help to maintain are
 8   important fisheries, including of Salmonid fish (O'Briain et al., 2020). Within cities, green spaces, including
 9   parks, local nature reserves and green roofs and walls can also provide cooling as a result of evapo-
10   transpiration (Bowler et al., 2010a; Aram et al., 2019; Hobbie and Grimm, 2020), although this may be
11   reduced in drought conditions.
13   Wildfire is an increasing risk for people as well as to ecosystems, in many parts of the world. As discussed
14   in Section, this is the result not just of climate change but also past management practices, including
15   fire suppression. Better fire management including reinstating more natural fire regimes can reduce risks.
17   EbA is usually a place-specific approach and a number of studies have documented how attempts to
18   implement it without an understanding of local circumstances and full engagement of local communities
19   have been unsuccessful (Nalau et al., 2018). Since AR5, a number of studies have considered the factors that
20   are important for environment adaptation programmes and projects (UNFCCC, 2015; Nalau et al., 2018;
21   Duncan et al., 2020; Network and ENCA, 2020; Townsend et al., 2020). Considering these sources, others
22   described above and the case studies presented, in 2.6.5, a number of requirements for effective
23   implementation of EbA can be identified, including the following:
25     • Targeting of the right EbA measure in the right location
26     • Decision-making at the appropriate level of governance with participation from all affected communities
27     • Integration of Local Knowledge and Indigenous Knowledge & capacity into decision-making and
28       management of project
29     • Involvement of government and non-government stakeholders
30     • Full integration of EbA with other policy areas, including agriculture, water resources and natural
31       resource protection
32     • Protection and if possible improvement of incomes of local people.
33     • Effective institutional support to manage finances and implementation of projects and programmes.
34     • Time -many EbA interventions take time to establish e.g. trees to grow, wetlands recover
35     • Monitoring of intended outcomes and other impacts and communication of results
37   Whilst it is essential to develop place-specific EbA measures, with full engagement of local communities, it
38   is worth noting that new opportunities may emerge that would not have been possible in the past. As the
39   climate changes, novel ecosystems may emerge with no present day analogue which have the potential to
40   provide different adaptation benefits and societies may be more willing to adopt transformative approaches
41   (Colloff et al., 2017; Lavorel et al., 2020).
43   Increasingly it is essential to integrate adaptation and the protection of biodiversity with land based climate
44   change mitigation initiatives; this is discussed in more detail in Cross-Chapter Box NATURAL, this
45   Chapter. The new IUCN standard (IUCN, 2020, Global standard) offers a basis for assessing whether actions
46   are true Nature-based Solutions and take account of the wider factors necessary for success.
48   Whilst policy interest is growing and there is an increasing deployment of EbA there is still a long way to go
49   in delivering it full potential (Huq and Stubbings, 2015) and significant institutional and cultural barriers
50   remain(Huq et al., 2017; Nalau et al., 2018). Nevertheless it is increasingly clear that EbA can offer a
51   portfolio of effective measures to reduce risks from climate change to people at the same time as benefiting
52   biodiversity (robust evidence, high agreement), providing they are deployed with careful planning in a way
53   that is appropriate local ecological and societal contexts (robust evidence, high agreement).
55   2.6.4   Adaptation for Increased Risk of Disease

     Do Not Cite, Quote or Distribute                     2-115                                       Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   Low-probability events can be very high impact (for example, the transmission of the SARS-CoV-2 from
 2   wild animals to humans). A robust disease risk reduction policy would include utilizing a One Biosecurity
 3   approach (Hulme, 2020) with actions to reduce disease risk across multiple sectors and from a variety of
 4   anthropogenic drivers, including climate change, even if there is high uncertainty in projected risk (see
 5   Cross-Chapter Boxes: ILLNESS, COVID, DEEP). Kraemer et al. (2019) found that vector importation was a
 6   key risk factor and focus should be on preventing invasive species introductions. Further, many neglected
 7   tropical diseases (NTDs) are also VBDs, and the UN SDG of good health and well-being explicitly calls for
 8   increased control and intervention with a focus on emergency preparedness and response (Stensgaard et al.,
 9   2019a). Online tools are being developed to warn conservation biologists when species of conservation
10   concern are at greater risk of disease outbreaks due to environmental changes (e.g., for Hawaiian
11   Honeycreepers and avian malaria (Berio Fortini et al., 2020) and for coral diseases (Caldwell et al., 2016)).
12   Forecasting models to warn of human disease outbreaks like malaria and dengue are also now available, with
13   findings that multiple model ensemble forecasts outperform individual models (Lowe et al., 2013; Lowe et
14   al., 2014; Lowe et al., 2018; Zhai et al., 2018; Johansson et al., 2019; Tompkins et al., 2019; Muñoz et al.,
15   2020; Colón-González et al., 2021; Petrova et al., 2021). Improving vector-borne disease and NTD public
16   health responses will require multi-disciplinary teams capable of interpreting, analyzing, and synthesizing
17   diverse components of complex ecosystem-based studies for effective intervention (Mills James et al., 2010;
18   Rubin et al., 2014; Valenzuela and Aksoy, 2018), broad epidemiological and entomological surveillance
19   (Depaquit et al., 2010; Lindgren et al., 2012; Springer et al., 2016), and community-based disease control
20   programs that build local capacity (Andersson et al., 2015; Jones et al., 2020b).
25   Cross-Chapter Box ILLNESS: Infectious Diseases, Biodiversity and Climate: Serious Risks Posed by
26        Vector- and Water-borne Diseases
28   Authors: Marie-Fanny Racault (United Kingdom/France, Chapter 3), Stavana E. Strutz (USA, Chapter 2),
29   Camille Parmesan (France/United Kingdom /USA, Chapter 2), Rita Adrian (Germany, Chapter 2), Guéladio
30   Cissé (Mauritania/Switzerland/France, Chapter 7), Sarah Cooley (USA, Chapter 3), Adugna Gemeda
31   (Ethiopia, Chapter 9), Nathalie Jeanne Marie Hilmi (Monaco/France, Chapter 18), Salvador E. Lluch-Cota
32   (Mexico, Chapter 5), Gretta Pecl (Australia, Chapter 11), David Schoeman (Australia, Chapter 3), Jan C.
33   Semenza (Italy, Chapter 7), Maria Cristina Tirado (USA/Spain, Chapter 7), Gautam Hirak Talukdar (India,
34   Chapter 2), Yongyut Trisurat (Thailand, Chapter 2), Meghnath Dhimal (Nepal, CA), Luis E. Escobar
35   (Guatemala/Columbia/USA, CA), Erin Mordecai (USA, CA), A. Townsend Peterson (USA, CA), Joacim
36   Rocklöv (Sweden, CA), Marina Romanello (United Kingdom, CA)
38   Climate change is altering the life cycles of many pathogenic organisms and changing the risk of vector- and
39   water-borne infectious diseases transmission to humans (high confidence). Re-arrangement and emergence
40   of some diseases are already observed in temperate-zone and high-elevation areas, and coastal areas
41   (medium to high confidence). Shifts in the geographic and seasonal range suitability of pathogens and
42   vectors are related to climatic-impact drivers (warming, extreme events, precipitation, humidity) (high to
43   very high confidence), but there are substantial non-climatic drivers (land use change, wildlife exploitation,
44   habitat degradation, public health and socio-economic conditions) that affects the attribution of the overall
45   impacts on prevalence or severity of some vector- and water-borne infectious diseases over recent decades
46   (high confidence). Adaptation options that involve sustained and rapid surveillance systems, and the
47   preservation and restoration of natural habitats, with their associated higher levels of biodiversity, both
48   marine and terrestrial, will be key to reducing risk of epidemics and large-scale disease transmissions
49   (medium confidence).
51   Since AR5, further evidence is showing that climate-related changes in the geographic and seasonal range
52   suitability of pathogens and vectors and in the prevalence or new emergence of vector- and water-borne
53   infectious diseases have continued across many regions worldwide and are sustained over decadal time
54   scales (medium to high confidence) (WGII Sections,, 7.2, 7.3,; Harvell et al., 2009;
55   Garrett et al., 2013; Burge et al., 2014; Guzman and Harris, 2015; Baker-Austin et al., 2018; Watts et al.,
56   2019; Semenza, 2020; Watts et al., 2021) . Ecosystem-mediated infectious diseases at risk of increase from
57   climate change include water-borne diseases associated with pathogenic Vibrio species (e.g., those causing

     Do Not Cite, Quote or Distribute                     2-116                                      Total pages: 237
     FINAL DRAFT                                           Chapter 2                        IPCC WGII Sixth Assessment Report

 1   cholera and vibriosis), and harmful algal blooms (e.g., ciguatera-fish poisoning) (SROCC, Box 5.4,
 2   AR6 WGII 3.5, Table 3.S.3, 5.12; Baker-Austin et al., 2013; Levy, 2015; Trtanj et al., 2016; Ebi Kristie et
 3   al., 2017; Mantzouki et al., 2018; Nichols et al., 2018), and vector-borne diseases associated with arthropods
 4   (e.g., malaria, dengue, chikungunya, Zika virus, West Nile virus, and Lyme disease), helminths (e.g.,
 5   schistosomiasis) and zoonotic diseases associated with cattle and wildlife (e.g., leptospirosis) (medium to
 6   high confidence) (Sections, 3.5, 7.2, 7.3,,, 14.4.6, Cross-Chapter Box COVID in
 7   Chapter 7; Table Cross-Chapter Box ILLNESS.1; SR1.5; Ebi et al., 2021) .
 9   The attribution of observed changes in disease incidence partly or fully to climatic-impact drivers remains
10   challenging because of the difficulty of accurately capturing the contributions of multiple, interacting, and
11   often nonlinear underlying responses of host, pathogen, and vector, which can be influenced further by non-
12   climate stressors and the long history of anthropogenic disturbance. Disease emergence in new areas requires
13   independent drivers to coincide (i.e., increasing climate suitability for pathogen or vector survival and
14   competence/capacity, and introduction of the pathogen via mobility of human populations). Further, the
15   extent to which changes in ecosystem-mediated diseases impact human health is highly dependent upon
16   local socio-economic status, sanitation, medical systems, and practices (Section; Figure FAQ2.3.1;
17   Gething et al., 2010; Lindgren et al., 2012; Mordecai et al., 2013; Liu-Helmersson et al., 2014; Bhatt et al.,
18   2015; Morin et al., 2015; Ryan et al., 2015; Wesolowski et al., 2015; Stanaway et al., 2016; Yamana et al.,
19   2016; Mordecai et al., 2017; Tesla et al., 2018; Ryan et al., 2019; Shah et al., 2019; Iwamura et al., 2020;
20   Mordecai et al., 2020; Colón-González et al., 2021; Ryan et al., 2021).Thus, the links between climate
21   change, ecosystem change, health and adaptation need to be considered concurrently (AR6 WGII 2.4, 3.5.3,
22   7.2, 7.3, 4.3.3,, Table 2.S.1).
25   Table Cross-Chapter Box ILLNESS.1: Observed climate change impacts on cholera, dengue, and malaria. 1)
26   Cholera: Endemicity based on Ali et al., 2015. Changes (2003-2018) in suitability for coastal Vibrio cholerae estimated
27   from model observations driven by sea-surface temperature (SST) and chlorophyll-a (CHL) concentration (Escobar et
28   al., 2015; Watts et al., 2019). Vulnerabilities based on Sigudu et al., 2015, Agtini et al., 2005, and Sack et al., 2003. 2)
29   Dengue: Endemicity based on Guzmen et al., 2015. 3) Malaria: Endemicity based on Phillips et al. 2017, and WHO
30   Global Malaria Programme. Impacts of climate change on diseases and their vectors are most evident at the margins of
31   current distributions. However, climate change is difficult to implicate in areas with extensive existing transmission and
32   vector/pathogen abundance, and in particular is difficult to separate from concurrent directional trends in disease
33   control, changes in land use, water access, socioeconomic and public health conditions. As a result, while many studies
34   indicate increasing climate suitability of some areas for cholera, dengue, and malaria, the degree to which these changes
35   can be attributed to climate change remains challenging. For these cases, confidence statements of low, medium, or high
36   reflect confidence that variation in the disease and/or vector/pathogen is associated with variation in climate drivers,
37   rather than with directional climate change per se. Acronyms: ONI (Oceanic Niño Index), Tmin (minimum
38   temperature), SPI (Standardised Precipitation Index), LST Land-surface temperature. Full references for this table can
39   be seen in supplemental table 6 (see Table 2.S.6).
                                        Cholera                           Dengue                             Malaria
         Endemicity        Endemic                             Endemic in sub-Saharan Africa       Endemic
                                                               but not S South Africa
         Climate drivers   Disease incidence: NE Africa,                                           W Africa: Temp ( medium to
                           Central Africa & Madagascar:                                            high confidence)
                           Rainfall ( medium confidence)                                           E Africa: Temp medium to high
                           SE Africa: Rainfall, LST, SST,                                          confidence)
                           Plankton (medium to high
                           confidence )
                           ES Africa: SST, CHL (low
                           confidence due to limited
                           W Africa: Rainfall (floods), LST,
                           SST (medium to high confidence)
         Change and        Area of coastline suitable for      Potentially expanding ( low         E Africa: Upward shift and
         Confidence        outbreak: N&W Africa: Increase      confidence)                         increase in malaria & Anopheles
                           (low to medium confidence)          Dengue and Ae. aegypti present      spp. in highland areas (medium to
                           C & E Africa: No change (low to     but underdetected in climatically   high confidence )
                           medium confidence )                 suitable areas.                     Widespread decreases due to
                           S Africa: Decrease (low to                                              malaria control ( medium
                           medium confidence)                                                      confidence) and warming climate
                                                                                                   (low confidence)

     Do Not Cite, Quote or Distribute                            2-117                                             Total pages: 237
FINAL DRAFT                                             Chapter 2                         IPCC WGII Sixth Assessment Report
   Vulnerabilities   ES Africa: women of all ages
                     more affected than men by
   Endemicity        Endemic                               Endemic in S Asia, SE Asia, and      Endemic in S Asia, SE Asia,
                                                           E Asia                               Partially endemic in E Asia
   Climate drivers   Disease incidence: E Asia: SST,       S Asia: Rainfall, Temp,              S Asia: Rainfall, Temp (medium
                     CHL, Sea Level (medium to high        Humidity (medium confidence)         to high confidence)
                     confidence)                           SE Asia: Rainfall, Temp medium       SE Asia: Rainfall, Temp
                     S Asia: SST, CHL, LST,                confidence)                          (medium confidence)
                     Rainfall(floods) (high to very        E Asia: Rainfall, Temp,
                     high conficende)                      Typhoons (low confidence)

   Change and        Area of coastline suitable for        SE Asia: Increase (low               S Asia: Increase (medium
   Confidence        outbreak: Increase (low to            confidence)                          confidence)
                     medium confidence)                    S Asia: Increase (medium
                                                           E Asia: Increase (low confidence)
   Vulnerabilities   SE Asia: infants (<9 years) with
                     highest incidences of cholera
                     S Asia: older children and young
                     adults (16-20 years old) more
                     frequently reported with cholera
                     than non-cholera diarrhoea
   Endemicity        Not endemic                           Partially endemic in N Australia     Not endemic
   Climate drivers   No evidence for disease               Rainfall, Temp (low confidence)
   Change and        Area of coastline suitable for        Increase in sporadic outbreaks       No change
   Confidence        outbreak: No change (low to           due to climate change (low
                     medium confidence)                    confidence)
 Central America
    Endemicity       Not endemic                           Endemic                              Partially endemic
   Climate drivers   No evidence for disease                ONI, SST, Tmin, Temp,
                     incidence                             Rainfall, Drought (low
   Change and        Areas of coastline suitable for       Increasing due to climate (low       Overall decrease not linked to
   Confidence        outbreak: Decrease (low to            confidence)                          climate change. Focal increases
                     medium confidence)                    Upward expansion of Ae. aegypti      due to human activities.
                                                           (low confidence)
 South America
   Endemicity        Epidemic                              Endemic in all regions except S      Endemic
                                                           South America
   Climate drivers   Abundance of coastal V.               Temp, Prec, Drought                  N South America: Temp (low
                     cholerae: NW South America:                                                confidence)
                     SST, Plankton (low to medium                                               N SE South America: Tmax,
                     confidence)                                                                Tmin, humidity (low confidence)
   Change and        Area of coastline suitable for        Increasing due to urbanization       Higher elevation regions:
   Confidence        outbreak: No change (low to           and decreased vector control         Increase (low confidence)
                     medium confidence)                    programmes, not strongly linked
                                                           to climate

   Endemicity        Not endemic                           S Europe: focal outbreaks            Not endemic
   Climate drivers   No evidence for disease
                     Abundance of coastal V.
                     cholerae: N Europe: SST,
                     Plankton (medium confidence)
  Change and         Area of coastline suitable for        Mediterranean                        No change
  Confidence         outbreak: Increase (low to            regions of S Europe: Outbreaks
                     medium confidence)                    (low confidence)
 North America
   Endemicity        Not endemic                           Partially endemic in S North         Not endemic
   Climate drivers   Area of coastline suitable for        Winter minT (Low)
                     outbreak: Increase (low to
                     medium confidence)

Do Not Cite, Quote or Distribute                             2-118                                              Total pages: 237
     FINAL DRAFT                                           Chapter 2                        IPCC WGII Sixth Assessment Report
       Change and         No evidence for disease             Declining                           No change
       Confidence         incidence
                          Abundance of coastal V.
                          cholerae: EN America: SST
                          (Low due to limited evidence)
      Small Islands
        Endemicity        Epidemic                            Endemic in many small islands in    Endemic in many small islands in
                                                              the Tropics                         the Tropics
       Climate drivers    Disease incidence: Caribbean:       Caribbean: SPI, Tmin (low
                          SST, LST, Rainfall (low to          confidence)
                          medium confidence)
       Change and         Area of coastline suitable for      Increasing (low confidence)         Decrease in Caribbean not linked
       Confidence         outbreak: Caribbean & Pacific                                           to climate
                          Small Island: Decrease (low to
                          medium confidence)
 3   Observed and projected changes
 5   In aquatic systems, at least 30 human pathogens with water infection-routes (freshwater and marine) are
 6   affected by climate change (Section 3.5.3 Table SM3.G; Nichols et al., 2018) . Warming, acidification,
 7   hypoxia, sea-level rise, and increases in extreme weather and climate events (e.g. marine heatwaves, storm
 8   surges, flooding, and drought), which are projected to intensify in the 21st century (high confidence) (AR6
 9   WGI SPM B2.2, B.3.2), are driving species’ geographic range shifts and global rearrangements in the
10   location and extent of areas with suitable conditions for many harmful pathogens, including viruses, bacteria,
11   algae, protozoa, and, helminths (high confidence) (Sections 2.3,,, SROCC, Box 5.4;
12   Trtanj et al., 2016; Ebi Kristie et al., 2017; Manning and Nobles, 2017; Pecl et al., 2017; Mantzouki et al.,
13   2018; Nichols et al., 2018; Bindoff et al., 2019; Kubickova et al., 2019; Watts et al., 2019; Watts et al., 2020;
14   Watts et al., 2021).
16   Incidence of cholera and Vibrio-related disease outbreaks has been shown to originate primarily in coastal
17   regions, and then spread inland via human transportation. Our understanding of impacts of climate drivers on
18   the dynamics of Vibrio-related infections have been strengthened through improved observations from long-
19   term monitoring programs (e.g., (Vezzulli et al., 2016)), and statistical modelling supported by large-scale
20   and high-resolution satellite observations of climate drivers (high confidence) (e.g., Baker-Austin et al.,
21   2013; Escobar et al., 2015; Jutla et al., 2015; Martinez et al., 2017; Semenza et al., 2017; Racault et al.,
22   2019; Campbell et al., 2020).
24   The coastal area suitable for V. cholerae (the causative agent for cholera) has increased by 9.9% globally
25   compared to a 2000s baseline (Escobar et al., 2015; Watts et al., 2019). The poleward expansion of the
26   distribution of Vibrio spp. has increased the risk of vibriosis outbreaks in northern latitudes. Specifically, the
27   coastal area suitable for Vibrio infections in the past 5 years has increased by 50.6% compared with a 1980s
28   baseline at latitudes of 40–70°N; in the Baltic region, the highest-risk season has been extended by 6.5
29   weeks over the same periods (Watts et al., 2021). Already, studies have noted greater numbers of Vibrio-
30   related human infections, and most notably disease outbreaks linked to extreme weather events such as heat
31   waves in temperate regions such as Northern Europe (Baker-Austin et al., 2013; Baker-Austin et al., 2017;
32   Baker-Austin et al., 2018) (high confidence). By the end of the 21st century, under RCP6.0, the number of
33   months of risk of Vibrio illness is projected to increase in Chesapeake Bay by 10.4±2.4%, with largest
34   increases during May and September, which are the months of strong recreational and occupational use,
35   compared to a 1985-2000 baseline (Jacobs et al., 2015; Davis et al., 2019a). In the Gulf of Alaska, the
36   coastal area suitable for Vibrio spp. is projected to increase on average by 58%±17.2% in summer under
37   RCP6.0 by the 2090s, compared to a 1971-2000 baseline (low to medium confidence) (Jacobs et al., 2015).
39   On land, increased global connectivity and mobility, unsustainable exploitation of wild areas and species,
40   land conversion (agricultural expansion, intensification of farming, deforestation, infrastructure
41   development), together with climate-change-driven range shifts of species and human migration (Cross-
42   Chapter Box MOVING PLATE in Chapter 5), have modified interfaces between people and natural systems
43   (IPBES, 2018a). Climate-driven increase in temperature, frequency and intensity of extreme events, and
44   changes in precipitation and relative humidity, have provided opportunities for re-arrangements of disease
45   geography and seasonality, and emergence into new areas (high confidence) (Section In particular,

     Do Not Cite, Quote or Distribute                           2-119                                             Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   malaria has expanded into higher elevations in recent decades and although climate change attribute remains
 2   challenging (Hay et al., 2002; Pascual et al., 2006; Alonso et al., 2011; Campbell et al., 2019c), evidence that
 3   the elevational distribution of malaria has tracked warmer temperatures is compelling for some regions (Siraj
 4   et al., 2014). Models based on both empirical relationships between temperature and the Anopheles mosquito
 5   and Plasmodium parasite traits that drive transmission (Mordecai et al., 2013; Yamana Teresa and Eltahir
 6   Elfatih, 2013; Johnson et al., 2015) and existing mosquito distributions (Peterson, 2009) predict that
 7   warming will increase the risk of malaria in highland East Africa and Southern Africa, while decreasing the
 8   risk in some lowland areas of Africa, as temperatures exceed the thermal optimum and upper thermal limit
 9   for transmission (Peterson, 2009; Yamana Teresa and Eltahir Elfatih, 2013; Ryan et al., 2015; Watts et al.,
10   2021).
12   In contrast to malaria, dengue has expanded globally since 1990, particularly in Latin America and the
13   Caribbean, South Asia, and sub-Saharan Africa (Stanaway et al., 2016). While urbanization, changes in
14   vector control, and human mobility play roles in this expansion (Gubler, 2002; Åström et al., 2012;
15   Wesolowski et al., 2015), the physiological suitability of temperatures for dengue transmission is also
16   expected to have increased as climates have warmed (Colón-González et al., 2013; Liu-Helmersson et al.,
17   2014; Mordecai et al., 2017; Rocklöv and Tozan, 2019). Models predict that dengue transmission risk will
18   expand across many tropical, subtropical, and seasonal temperate environments with future warming
19   (Åström et al., 2012; Colón-González et al., 2013; Ryan et al., 2019; Iwamura et al., 2020; Watts et al.,
20   2021)).
22   Adaptation options
24   During the 21st century, public health adaptation measures (Figure Cross-Chapter Box ILLNESS.2) have
25   been put in place in attempts to control or eradicate a variety of infectious diseases by improving
26   surveillance and early detection systems; constraining pathogen, vector, and/or reservoir host distributions
27   and abundances; reducing likelihood of transmission to humans; and improving treatment and vaccination
28   programs and strategies (medium to robust evidence, medium to high agreement) (Chinain et al., 2014;
29   Adrian et al., 2016; Friedman et al., 2017; Konrad et al., 2017; Semenza et al., 2017; Borbor-Córdova et al.,
30   2018; Rocklöv and Dubrow, 2020). In addition, effective management and treatment of domestic and
31   wastewater effluent, through better infrastructure and preservation of aquatic systems acting as natural water
32   purifiers, have been key to securing the integrity of the surrounding water bodies, such as groundwater,
33   reservoirs and lakes, and agricultural watersheds, as well as protecting public health (high confidence)
34   (Okeyo et al., 2018; Guerrero-Latorre et al., 2020; Kitajima et al., 2020; Sunkari et al., 2021). The
35   preservation and restoration of natural ecosystems, with their associated higher levels of biodiversity, have
36   been reported as significant buffers against epidemics and large-scale pathogen transmission (medium
37   confidence) (Johnson and Thieltges, 2010; Ostfeld and Keesing, 2017; Keesing and Ostfeld, 2021). Further,
38   the timely allocation of financial resources and sufficient political will in support of a “One Health”
39   scientific research approach, recognising the health of humans, animals and ecosystems as interconnected
40   (Rubin et al., 2014; Whitmee et al., 2015; Zinsstag et al., 2018), holds potential for improving surveillance
41   and prevention strategies that may help to reduce the risks of further spread, and new emergence of
42   pathogens and vectors (medium confidence) (Destoumieux-Garzón et al., 2018; Hockings et al., 2020;
43   Volpato et al., 2020; Hopkins et al., 2021; Services and Ecosystem, 2021).

     Do Not Cite, Quote or Distribute                     2-120                                      Total pages: 237
     FINAL DRAFT                                       Chapter 2                    IPCC WGII Sixth Assessment Report

 2   Figure Cross-Capter Box ILLNESS.1: Adaptation measures to reduce risks of climate change impact on water- and
 3   vector-borne diseases. Impacts are identified at three levels: 1) impact on pathogen, host/vector distributions and
 4   abundance; 2) impact on pathogen-host transmission cycle occurrence and efficiency; and 3) impact on likelihood of
 5   transmission to humans. Adaptation typology is based on (Biagini et al., 2014; Pecl et al., 2019). For each type of
 6   adaptation, examples are provided with their level of evidence and agreement.
11   2.6.5   Adaptation in Practice: Case Studies and Lessons Learned
13   Adaptation plans for biodiversity and EbA have been adopted in many places and different scales but it is
14   difficult to get a systematic overview of adaptation in practice. We have therefore reviewed a series of
15   contrasting case studies to illustrate the some key issues. There is a pressing need for more thorough
16   monitoring and evaluation of adaptation to assess effectiveness. Climate change adaptation is conceptually
17   difficult to measure but it is possible to test which techniques work in reducing vulnerability and monitor
18   their deployment (Morecroft et al., 2019).
20   Adaptation can take place at a range of scales with specific projects nested within overarching national
21   strategies. Small scale projects can be adaptation focused, but in larger scale adaptation is often integrated
22   with wider objectives. Within an urban or peri-urban context, the benefits of natural and semi-natural areas
23   for health and well- help to justify support for EbA. Economic wellbeing is also an important factor in many
24   cases, whether as, in Durban (Section, by providing new job opportunities or, in the Andes (Section
25 by supporting long-established agricultural practice. Action on the ground often depends on factors

     Do Not Cite, Quote or Distribute                       2-121                                         Total pages: 237
     FINAL DRAFT                                     Chapter 2                  IPCC WGII Sixth Assessment Report

 1   at a range of scales, for example, a local plan, a national strategy and international funding. Within Durban,
 2   partnership between local communities, local authorities and the academic community were essential,
 3   together with an international context. Nevertheless, there are examples of communities using traditional or
 4   local knowledge to adapt to changing circumstances, with little or no external input, (Section They
 5   are, however, limited in their scope to adapt by factors beyond their direct control.
 7   Specific interventions to protect species from climate change, such as the case of South African
 8   penguins(Section and the Tasmanian Wilderness World Heritage Area (Section, are rare.
 9   However, in countries where nature reserves are actively managed or where ecosystem restoration projects
10   are progressing, local practitioners may use local knowledge to adapt to weather conditions and their
11   associated effects (fire or water shortage for example). This is good practice, but it may not be sufficiently
12   to address likely future changes in climate (Duffield et al., 2021). Training and resources to support
13   conservation practitioners are becoming available to help address this. Examples include the Climate Change
14   Adaptation Manual in England (Section, and The Alliance for Freshwater Life
15   (, that provide expertise for the sustainable management of freshwater
16   biodiversity (Darwall et al., 2018).
18   Adaptation is widely recognised as important for national conservation policy and is being considered in a
19   variety of countries (Section, Adaptation in this strategic context includes decisions about
20   the selection and objectives for protected areas, for example identifying places which can act as refugia. It
21   can also mean recognising where protected areas remains important but will support a changing range of
22   species and ecosystems. This is important for directing resources effectively and ensuring that site
23   management remains appropriate. There are however often major uncertainties and the extent to which there
24   will be a need for more radical measures will depend on success in reducing greenhouse gas emissions
25   globally. A global rise of 1.5–2°C would require relatively incremental adjustments to conservation
26   management in many parts of the world, but a 3‒4°C rise would require radical, transformational changes to
27   maintain many species and ecosystem services (Morecroft et al., 2012).
29   Whilst adaptation strategies for conservation are relatively common, at least at an outline level,
30   implementation is slow in most places. This may partly reflect lack of resources for conservation in many
31   parts of the world; however, another barrier is that people often value protected sites in their present form.
32   Actions, which might jeopardise this, are inevitably a last resort. Initiatives to engage wider communities in
33   discussions are likely to be essential in gaining support for such changing approaches.
35   EbA and adaptation for biodiversity are intrinsically linked and the largest scale interventions for adaptation
36   in ecosystem have tended to bring together both elements. For example adaptation to reduce flood risk by
37   habitat creation and using natural processes, (Section, Cross-Chapter Box SLR in Chapter 3), such as
38   by re-naturalising straightened river systems or creating wetlands for water storage, offers the potential to
39   meet multiple objectivesand has increased overall funding available for ecosystem restoration.
41 Case study: Assisted Colonisation / Managed Relocation in Practice
43   Scale: Global
44   Issue: Helping species move to track shifting climate space
46   Managed relocation (assisted migration, assisted colonisation) is the movement of species, populations, or
47   genotypes to places outside the areas of their historical distributions (Hoegh-Guldberg et al., 2008) and may
48   be an option where they are not able to disperse and colonise naturally. It requires careful consideration of
49   scientific, ethical, economic and legal issues between the object of relocation and the receiving ecosystem
50   (Hoegh-Guldberg et al., 2008; Richardson et al., 2009; Schwartz et al., 2012).
52   Individual cases show that assisted migration can be successful. Anich & Ward (2017) extended the
53   geographic breeding range of a rare bird, Kirtland's warbler, Setophaga kirtlandii, by 225km by using song
54   playbacks to attract migrating individuals. Wadgymar (2015) successfully transplanted an annual legume,
55   Chamaecrista fasciculata, to sites beyond its current poleward range limit, while Liu (2012) found that all
56   but one of 20 orchid species survived when transplanted to higher elevations than their current range limits.
57   After introducing two British butterfly species to sites ∼65 and ∼35 km beyond their poleward range

     Do Not Cite, Quote or Distribute                     2-122                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   margins, Willis (2009) observed that both introduced populations grew, expanded their ranges and survived
 2   for at least 8 years.
 4   Butterflies have been favoured subjects for assisted migration in response to regional climate warming, since
 5   they are easy to move and their range dynamics have been extensively studied. The Chequered Skipper
 6   butterfly, Carterocephalus palaemon, became locally extinct in England in the 1970's, in an area not close to
 7   either the species' poleward or equatorial range limits. Nonetheless, Maes (2019) consider climate a crucial
 8   parameter for re-introduction, using SDMs both for choosing the source population in Belgium and
 9   introduction site.
11   Success of assisted migration for conservation purposes has been variable. Bellis (2019), identified 56
12   successes and 33 failures among 107 translocations of insects that had been undertaken explicitly for
13   conservation purposes. They concluded failure was most strongly associated with low numbers of
14   individuals released. Another potential source of failure is local adaptation: there is good evidence that
15   adaptive differences among potential source populations can be important. For example, the transplants of C.
16   fasciculata were more successful when sourced from the most poleward existing sites, while individuals
17   from more equatorial habitats performed poorly even when artificially warmed (Wadgymar et al., 2015).
19 Case study: Adaptation for conservation and Natural Flood Management in England, United
20            Kingdom
22   Scale: National
23   Issue: National approach to adaptation in the natural environment
25   Climate change threats to biodiversity in England include range retraction of cold adapted species and effects
26   of more frequent extreme weather events such as droughts. These threats are exacerbated by land use and
27   management: with habitats fragmented, land often drained and rivers straightened. There are also risks to
28   people, which are exacerbated by environmental factors, including flooding and over-heating in urban areas.
29   A National Adaptation Programme, provides a broad policy framework for England and includes a chapter
30   on the natural environment. There are also adaptation plans produced by public bodies such as Natural
31   England, the conservation agency and the Environment Agency, with a wide range of responsibilities
32   including flood defence. The principles of climate change adaptation are well established in the UK
33   conservation community and resources are available. Natural England has published a Climate Change
34   Adaptation Manual jointly with the Royal Society for the Protection of Birds–a major conservation NGO
35   (England and RSPB, 2020) and spatial mapping tool for climate change vulnerability (Taylor et al., 2014).
37   (Duffield et al., 2021) found that awareness of the need for adaptation was common amongst nature reserve
38   managers and that they were implementing actions that might building resilience to climate change, such as
39   restoring ecosystem processes and reducing fragmentation. . There is a recognition that it will be necessary
40   to change management objectives of protected sites to adjust to changing circumstances but there was little
41   implementation of such changes (Duffield et al., 2021). The main examples of managing change, was at the
42   coast where rising sea level is causing transitions from terrestrial and freshwater systems to coastal and
43   marine ones.
45   A range of EbA approaches are starting to contribute to adaptation in England but the best developed is
46   Natural Flood Management (NFM): restoring natural processes and natural habitats to reduce flood risk
47   (Wingfield et al., 2019). Over the last decade, a series of NFM projects have been established in local areas.
48   The Environment Agency collated the evidence base on NFM (Burgess-Gamble et al., 2017) and was able to
49   draw on 65 case studies (Ngai et al., 2017), covering river and floodplain management, woodland
50   management, run-off management and coast and estuary management.
52   NFM includes a broad range of techniques, some of which, deliver real benefits for biodiversity and allow
53   natural ecological process to re-establish. Others, such as creating ‘woody debris dams’ – barriers artificially
54   constructed from tree trunks and branches in water courses to slow flow of water ‒will have fewer benefits,
55   although they may benefit some species. Dadson et al. (2017) concluded that ‘the hazard associated with
56   small floods in small catchments may be significantly reduced’ by natural flood management techniques.

     Do Not Cite, Quote or Distribute                     2-123                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   However, they noted that the most extreme flood events may overwhelm any risk management measures and
 2   failed to find clear evidence of NFM in reducing flood risk downstream in large catchments.
 4   There remain challenges in deploying NFM at larger scales, partly reflecting the time necessary to
 5   demonstrate the effectiveness of pilot studies and build confidence and building stakeholder support is
 6   important (Huq et al., 2017). There are now a number of examples of where collaborative initiatives between
 7   local communities, land owners and government agencies have been successful in establishing effective
 8   NFM schemes (Short et al., 2019).
10   Case Study: Protected Areas Planning in Response to Climate Change in Thailand
12   Scale: National
13   Issue: Protected area network planning
15   Many countries in the Association of South East Asian Nations (ASEAN) are expanding protected area
16   networks to meet the Aichi target 11 of at least 17% of terrestrial area protected and it is important to take
17   the effects of climate change into account. Existing protected areas in Thailand cover approximately 21% of
18   the land area, and it is one of the few tropical countries that passes the Aichi Target 11. Most protected areas
19   in Thailand were established on an ad hoc basis to protect remaining forest cover, and as a result do not
20   represent diverse habitats and their associated species (Chutipong et al., 2014; Tantipisanuh, 2016) and may
21   not be resilient to the interacting impacts of future land use and climate change (Klorvuttimontara et al.,
22   2011; Trisurat, 2018).
24   Recent research conducted in northern Thailand indicated that the existing protected areas (31% of the
25   region area) cannot secure viability of many medium- size and large mammals. Most species climate space
26   would substantially shift, bringing a risk of extinction. The model results based on the spatial distribution
27   model and network flow determined there was a need for expansion areas of 5,200 km2 or 3% of the region
28   to substantially minimise the high-risk level and increase the average coping capacity of the protection of
29   suitable habitats from 82% as the current plan to 90%. These results were adopted by the Thailand’s
30   Department of National Parks, Wildlife and Plant Conservation and included in the National Wildlife
31   Administration and Conservation Plan (2021-2031).
33   Case Study: Effects of Climate Change on Tropical High Andean Social Ecological Systems
35   Scale: Regional
36   Issue: Complex ramifications of glacial retreat on vegetation, animals, herders and urban populations
38   Accelerated warming is shrinking tropical glaciers at rates unseen since the middle of the Little Ice Age
39   (Rabatel et al., 2013; Zemp et al., 2015). Climate-driven upward migration of species associated with
40   warming and glacier retreat has modified species distribution and richness, and community composition
41   along the Andes altitudinal gradient (Seimon et al., 2017; Carilla et al., 2018; Zimmer et al., 2018; Moret et
42   al., 2019). Climate-driven glacier retreat alters hydrological regimes, impacting Andean pastoralists directly
43   (López-i-Gelats et al., 2016; Postigo, 2020; Thompson et al., 2021), and water provisioning to lowland
44   regions (Vuille et al., 2018; Hock et al., 2019; Orlove et al., 2019; Rasul and Molden, 2019). Drying
45   wetlands has modified alpine plant communities, which are relevant to storing carbon, regulating water, and
46   providing food for local livestock, leading to negative impacts on herders’ livelihoods (Dangles et al., 2017;
47   Polk et al., 2017; Postigo, 2020) and differently affecting the wild vicuña and the domesticated alpaca and
48   llama. Vicuña (Vicugna vicugna) and alpaca (Vicugna pacos) wool are important income sources for
49   indigenous communities and Llama (Lama glama) is the main source of meat. Vicuña is adjusting its feeding
50   behaviour and spatial distribution as vegetation migrates upwards (Reider and Schmidt, 2020), causing them
51   to roam outside protected areas and become vulnerable to illegal poaching.
53   Andean herders have responded to drying of grasslands by increasing livestock mobility, accessing new
54   grazing areas through kinship and leases, creating and expanding wetlands through building long irrigation
55   canals (of several km), limiting allocation of wetlands to new households, and sometimes cultivating grasses
56   (Postigo, 2013; López-i-Gelats et al., 2015; Postigo, 2020). These adaptive responses to regional climate
57   change are enabled by deeply-embedded indigenous institutions that have traditionally governed Andean

     Do Not Cite, Quote or Distribute                     2-124                                       Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1   pastoralists, but have become severely compromised by national socio-economic pressures (Valdivia et al.,
 2   2010; Postigo, 2019; Postigo, 2020). For instance, water quality, access and control by local pastoralists has
 3   declined due to new mining concessions in headwaters of Andean watersheds (Bebbington and Bury, 2009)
 4   and diversion of water to lowland coastal desert for agricultural irrigation (Mark et al., 2017).
 6   Glacier mass and runoff in the tropics are projected to reduce >70% and >10%, respectively, by 2100 under
 7   RCP 2.6, RCP 4.5 and RCP 8.5 (Huss and Hock, 2018; Hock et al., 2019). In Peru, montane ice-field melt-
 8   water provides 80% of the water resources for the arid coast where half the population lives (Thompson et
 9   al., 2021). Increasing variability of precipitation has compromised rain-fed agriculture and power generation,
10   particularly in the dry season, exacerbating pressures for new water sources (Bradley et al., 2006; Bury et al.,
11   2013; Buytaert et al., 2017). Thus, there is risk of increasing conflicts between climate change adaptation to
12   benefit high Andean human and natural communities and adaptation to maintain water provisioning for
13   lowland agricultural and urban areas.
15   Case Study: Helping African Penguins Adapt to Climate Change
17   Scale: Regional / local
18   Issue: Adaptation for a threatened species
20   The African penguin, Spheniscus demersus, is the only resident penguin species on mainland Africa and
21   breeds in a handful of colonies in South Africa and Namibia. In 2017, penguins in Cape Town’s Boulders
22   Beach colony attracted almost one million visitors, providing 885 jobs and $18.9m revenue (Van Zyl and
23   Kinghorn, 2018).Ninety-six percent of the population has been lost since 1900, with a 77% decline in the last
24   two decades (Sherley et al., 2018) and by 2019 only 17,700 pairs remained (Sherley et al., 2020). The
25   species is listed as Endangered on the IUCN Red List (IUCN, 2018) and if this trajectory persists the African
26   penguin will become functionally extinct in the near future (Sherley et al., 2018).
28   Historically, hunting, egg and guano collection were the species’ main threats, but three aspects of climate
29   change now predominate. Firstly, a several-hundred-kilometre eastward shift in distributions of their main
30   prey species, anchovies and sardines, has reduced food availability (Roy et al., 2007; Crawford et al., 2011).
31   While adult penguins typically forage up to 400 km from their colonies, they are restricted to a ~20 km
32   radius from their colonies during breeding months (Ludynia et al., 2012; Pichegru et al., 2012). The resulting
33   food shortage at this critical time is compounded by competition with commercial fisheries and
34   environmental fluctuations (Crawford et al., 2011; Pichegru et al., 2012; Sherley et al., 2018). This has
35   impacted adults’ survival and their ability to raise high-quality offspring (Crawford et al., 2006; Crawford et
36   al., 2011; Sherley et al., 2013; Sherley et al., 2014).
38   Increasing heat wave frequency and intensity recorded in recent decades presents a second threat (van
39   Wilgen and Wannenburgh, 2016; Van Wilgen et al., 2016; Mbokodo et al., 2020). Nests were historically
40   built in insulated guano burrows, but are now frequently sited on open ground (Kemper et al., 2007;
41   Pichegru et al., 2012; Sherley et al., 2012). High temperatures frequently expose the birds to severe heat
42   stress, causing adults to abandon nests and resulting in mortality of eggs and chicks (Frost et al., 1976;
43   Shannon and Crawford, 1999; Pichegru et al., 2012).Intensifying storm surges and greater wave heights can
44   cause nest flooding (Randall et al., 1986; de Villiers, 2002).
46   The African penguin’s survival in the wild is dependent on the success of adaptation action. Increasing
47   access to food resources is a management priority (IUCN, 2018). One approach is to reduce fishing pressure
48   immediately around breeding colonies. An experiment excluding fishing around colonies since 2008 has
49   demonstrated positive effects (Pichegru et al., 2010; Pichegru et al., 2012; Sherley et al., 2015; Sherley et al.,
50   2018; Campbell et al., 2019b). A second approach is to establish breeding colonies closer to their prey. An
51   ongoing translocation initiative aims to entice birds eastwards to recolonise an extinct breeding colony and
52   potentially to establish a new one (Schwitzer et al., 2013; Sherley et al., 2014; International, 2018). Penguin
53   “look-alikes” or decoys, constructed from rubber and concrete, have been placed at the extinct colony site
54   and, along with call play-backs, give the illusion of an established penguin colony (Morris and Hagen,
55   2018). This approach has not yet proven successful.

     Do Not Cite, Quote or Distribute                      2-125                                       Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1   To promote on-site adaptation to heat extremes and flooding, initiatives are underway to provide cooler
 2   nesting sites that also provide storm protection and are sufficiently above the high water level (Extinction,
 3   2018; International, 2018). Artificial nest boxes of various designs and constructed from a range of materials
 4   have been explored in combination with use of natural vegetation. Some designs have proven successful,
 5   increasing breeding success (Kemper et al., 2007; Sherley et al., 2012), but the same designs have had less
 6   success at other locations (Pichegru, 2013; Lei et al., 2014).
 8   Hand-rearing and releasing African penguin chicks, including from eggs, has long proven valuable because
 9   moulting parents, being shore-bound, are unable to feed late-hatching chicks. Since 2006, over 7,000-
10   orphaned chicks have been released into the wild as part of the Chick Bolstering Project with a success rate
11   of 77% (Schwitzer et al., 2013; Sherley et al., 2014; Klusener et al., 2018; SANCCOB, 2018). A new project
12   at Boulders Beach aims to use real-time weather station data, within-nest temperatures and known thresholds
13   of penguin heat stress as triggers for implementing a Heat Wave Response Plan. Drawing on well-
14   established chick-rearing facilities and a large body of expertise, this includes removing heat-stressed eggs
15   and birds, hand rearing and/or rehabilitation and release. It is hoped that such birds may be released at the
16   proposed new colony site.
18   Case study: Conserving Climate Change Refugia for the Joshua tree in Joshua Tree National Park,
19             California, United States of America
21   Scale: Local
22   Issue: Possible extirpation of a plant species from a national park
24   Joshua Tree National Park conserves 3200 km2 of Mojave and Sonoran Desert ecosystems. The climate of
25   the national park is arid, with a 1971-2000 average summer temperature of 27.3ºC ± 0.7ºC and average
26   annual precipitation of 170 ± 80 mm y-1 (Gonzalez et al., 2018). From 1895 to 2017, average annual
27   temperature increased at a significant (p < 0.0001) rate of 1.5 ± 0.1ºC century-1 and average annual
28   precipitation decreased at a significant (p = 0.0174) rate of -32 ± 12% century-1 (Gonzalez et al., 2018).
29   Anthropogenic climate change accounts for half the magnitude of a 2000–2020 drought in the Southwestern
30   USA, the most severe since the 1500s (Williams et al., 2020).
32   The national park was established to protect ecosystems and cultural features unique to the region,
33   particularly the Joshua tree (Yucca brevifolia), a tall, tree-like yucca that provides habitat for birds and other
34   small animals and holds cultural significance. The national park protects the southernmost populations of the
35   Joshua tree. Paleobiological data from packrat (Neotoma spp.) middens and fossil dung of the extinct Shasta
36   ground sloth (Nothrotheriops shastensis) show that Joshua trees grew 13 000‒22 000 thousand years before
37   present a across a wider range, extending as far as 300 km south into what is now México (Holmgren et al.,
38   2010; Cole et al., 2011). A major retraction of the range began ~11 700 thousand years before present,
39   coinciding with a warming in the region of 4ºC, caused by Milankovitch cycles, that marked the end of the
40   Pleistocene and beginning of the Holocene (Cole et al., 2011), suggesting a sensitivity of Joshua trees of 300
41   km latitude per 4ºC.
43   Under an emissions scenario that could increase park temperatures over 4ºC by 2100, suitable climate for the
44   Joshua tree could shift north and the species become extirpated from the park (Sweet et al., 2019). Plant
45   mortality would increase from drought stress and wildfires, which have been rare or absent in the Mojave,
46   but which invasive grasses have and may continue to fuel (Brooks and Matchett, 2006; DeFalco et al., 2010;
47   Abatzoglou and Kolden, 2011; Hegeman et al., 2014).
49   The national park had been trying to conserve the species wherever in the park it was found. The future risk
50   of extirpation prompted them to adapt conservation to focus on protecting potential refugia, where suitable
51   conditions may persist for the species into the future (Barrows et al., 2020). The national park used spatial
52   analyses of suitable climate to identify potential refugia under all emissions scenarios except for the highest
53   (Barrows and Murphy-Mariscal, 2012; Sweet et al., 2019). The park prioritises the refugia for removal of
54   invasive grasses and fire control (Barrows et al., 2020)and works to restore refugia that have burned in fires,
55   using native plants, including nursery-grown Joshua tree seedlings. The park and its partners are monitoring
56   plant species composition and abundance in the refugia for early warning of any changes (Barrows et al.,
57   2014).

     Do Not Cite, Quote or Distribute                      2-126                                       Total pages: 237
     FINAL DRAFT                                    Chapter 2                   IPCC WGII Sixth Assessment Report

 2   Case Study: Ecosystem based Adaptation in Durban, South Africa
 4   Scale: Local
 5   Issue: Ecosystem based adaptation (EbA) in a city and surrounding area
 7   Durban was an early pioneer of EbA in a city context, establishing a Municipal Climate Protection
 8   Programme (MCPP) in 2004 (Roberts et al., 2012). The City, situated in a global biodiversity hotspot
 9   (Bank, 2016), has a rapidly growing population (approximately 3.5 million) and is highly fragmented
10   (Roberts et al., 2013). High levels of development, particularly in peri-urban areas, have encroached into
11   natural habitats (Bank, 2016). Degradation of the natural resource base in this way has direct economic and
12   financial costs, is threatening the City’s long-term sustainability, and is exacerbated by climate change
13   (Bank, 2016; Municipality, 2020). The impacts of climate change are anticipated to increase unless
14   appropriate mitigation and adaptation interventions are prioritised (Municipality, 2020). High rates of
15   poverty, unemployment and health problems have pushed Durban to explore a climate change adaptation
16   work stream within its MCPP (Roberts et al., 2013; Roberts et al., 2020b).
18    A single approach to adaptation is likely to be insufficient (Archer et al., 2014), and community-based
19   adaptation should be integrated as part of a package of tools applied at the city level. Durban’s climate
20   change adaptation work stream is composed of three separate components: municipal adaptation (adaptation
21   activities linked to the key functions of local government); community-based adaptation (focused on
22   improving the adaptive capacity of local communities); and a series of urban management interventions
23   (addressing specific challenges such as the urban heat island, increased storm water runoff, water
24   conservation and sea level rise) (Roberts et al., 2013).
26   Lessons learnt from Duban’s experience include the importance of meaningful partnerships, long-term
27   financial commitments (Douwes et al., 2015) and significant political and administrative will (Roberts et al.,
28   2012; Roberts et al., 2020b). Securing these requires strong leadership (Douwes et al., 2015), including from
29   local champions (Archer et al., 2014), even if EbA is considered cost-effective (Roberts et al., 2012). Natural
30   habitat restoration projects are seen as an ideal tool, as they combine mitigation outcomes with increased
31   adaptation capacity that not only reduces vulnerability of ecosystems and communities (Douwes et al.,
32   2016), but creates economic opportunities. These include direct job creation (Diederichs and Roberts, 2016;
33   Douwes and Buthelezi, 2016) with various spinoffs such as better education for schoolchildren (Douwes et
34   al., 2015). Indirect benefits, include better water quality and reduced flooding, are generated as a result of
35   improved ecosystem service delivery (Douwes and Buthelezi, 2016). In areas that are already developed,
36   opportunities for green roof infrastructure can yield reductions in roof storm water run-off (by approximately
37   60 ml/m2/minute during a rainfall event), slow release of water over time, and reduced temperatures on roof
38   surfaces (Roberts et al., 2012).
40   Case Study: Protecting Gondwanan refugia against fire in Tasmania, Australia
42   Scale: Local
43   Issue: Protection of rare endemic species
45   The Tasmanian Wilderness World Heritage Area (TWWHA) has a high concentration of ‘paleo-endemic’
46   plant species restricted to cool, wet climates and fire free environments, but recent wildfires have burnt
47   substantial stands, which are unlikely to recover (Harris et al., 2018b, Bowman et al., 2021, The 2016
48   Tasmanian). The fires led to government inquiries and a fire-fighting review, which have suggested changes
49   to management as that climate change will make such fires more likely in the future (Council, 2016; Press,
50   2016; Council, 2019).
52   The majority of the TWWHA is managed as a Wilderness Zone, where management is currently carried out
53   in a manner that allows natural processes to predominate. The exclusion of fire from stands of fire-sensitive
54   trees such as the Pencil pine, Athrotaxis cupressoides, is part of this management strategy, possible in the
55   past due to the moisture differential and lower flammability of these areas. However, in recent years, the
56   threat posed by extensive and repeated wildfires, and an increasing awareness that fire risk is likely to
57   increase (Fox-Hughes et al., 2014; Love et al., 2017; Love et al., 2019) have meant that more direct

     Do Not Cite, Quote or Distribute                     2-127                                     Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1   management intervention has been implemented. There has been a realisation that a “hands off” approach to
 2   managing the threat will not be sufficient to protect the paleo-endemics. Not only is fire-fighting difficult in
 3   the remote wilderness area, but limited resources mean that fire managers must prioritise where fires will be
 4   fought when many fires are threatening towns and lives across the state simultaneously.
 6   After wildfires in 2016 caused extensive damage (Bowman et al., 2020a), significant effort and resources
 7   were spent trying to protect the remaining stands of Pencil pine during the 2019 fires, using new approaches
 8   including the strategic application of long-term fire retardant and the installation of kilometres of sprinkler
 9   lines (Council, 2019). These approaches are thought to have been effective at halting the fire and protecting
10   the high value vegetation in some situations. Impact reports are currently being finalised to quantify the
11   extent of fire-sensitive vegetation communities that have been affected. However, there is concern that these
12   interventions may have adverse effects on the values of the TWWHA if applied widely, so while research is
13   ongoing, these will only be applied in strategic areas (e.g., fire retardant is not being applied to some areas).
15   The TWWHA Management Plan (2016) emphasises Aboriginal fire management as an important value of
16   the area, along with their knowledge of plants, animals, marine resources, minerals (ochre and rock sources),
17   and their connection with the area as a living and dynamic landscape. Fire management planning aims to
18   protect important sites from fire and ensure that management does not impact Aboriginal cultural values
19   (DPIPWE, 2016). Increasingly, there is an acknowledgment that the cessation of traditional fire uses has led
20   to changes in vegetation and calls to incorporate Aboriginal burning knowledge into fire management of the
21   TWWHA.
23   Case Study: Bhojtal Lake, Bhopal, India
25   Scale: Local
26   Issue: Protection of water resources and biodiversity
28   The city of Bhopal, the capital of Madhya Pradesh state in central India, is dependent on Bhojtal, a large
29   man-made lake bordering the city, for its water supply (Everard et al., 2020). It is also an important
30   conservation site with wetlands protected under the Ramsar convention and diverse flora and fauna (WWF,
31   2006). Bhojtal also provides a wide range of other benefits to people, including tourism, recreation,
32   navigation, and subsistence and commercial fisheries, supporting the livelihoods of many families (Verma,
33   2001).
35   Climate change in Bhopal may pose ecological and socio-economic stresses due to changes in rainfall and
36   weather patterns (Anonymous, 2019), exacerbated by a series of problems such as wastewater discharge,
37   illegal digging of bore wells and unsustainable water extraction/exploitation (Everard et al., 2020).
38   Ecosystem service provision at Bhojtal was assessed using the Rapid Assessment of Wetland Ecosystem
39   Services (RAWES) approach, including water quality analysis from the lake. Information on the geology,
40   hydrology and catchment ecology of the lake was collected and a Baseline Biodiversity Assessment was
41   conducted.
43   The Lake Bhopal Conservation and Management Project (JICA, 2007) was developed with the following
44   actions:
46   1. Desilting and dredging; deepening and widening of spill channel; prevention of pollution (sewerage
47      scheme); management of shoreline and fringe area; improvement and management of water quality
49   2. Soil and water conservation measures using vegetative and engineering structures particularly at upper
50      ridges of watersheds; construction of small check dams or percolation tanks for recharge purposes in
51      areas marked for ‘drainage line recharge measures’.
53   3. Afforestation initiatives.
55   Implementation of these measures with the help of local communities improved the lake’s health. Nature-
56   based solutions are more resilient adaptation measures towards climate change. Restoration not only reduced
57   water stress but also provides multiple societal benefits in the urban area (Kabisch et al., 2016).

     Do Not Cite, Quote or Distribute                      2-128                                       Total pages: 237
     FINAL DRAFT                                      Chapter 2                    IPCC WGII Sixth Assessment Report

 2 Case Study: Addressing Vulnerability of Peat Swamp Forests in South East Asia (SEA)
 4   Scale: Regional
 5   Issue: Protecting peatland biodiversity, carbon and ecosystem services from climate change and land
 6   degradation
 8   SEA peatlands have undergone extensive logging, drainage and land-use conversion that have caused habitat
 9   loss for endemic species, i.e., orangutan (Pongo spp) (Gregory et al., 2012; Struebig et al., 2015). Prolonged
10   droughts associated with El Niño (Section compound the effects of drainage, leading to large
11   recurrent fires (Langner and Siegert, 2009; Gaveau et al., 2014; Putra et al., 2019). Under RCP 8.5, it is
12   projected that by the end of the century, the annual rainfall will significantly decrease (30%) over SEA, and
13   the number of consecutive dry days will significantly increase (60%) over Indonesia and Malaysia (Supari et
14   al., 2020). Peat degradation and losses to fire result in large GHG emissions (Miettinen et al., 2016) as well
15   as haze pollution that is a trans-boundary problem in the region (Heil et al., 2007).
17   Improving resilience to fire and climate change in SEA peatlands through restoration is extremely difficult
18   and presents many challenges. The Indonesian government has tasked the Badan Restorasi Gambut
19   (Peatland Restoration Agency) to restore peatlands (Darusman et al., 2021; Giesen, 2021). Other local
20   initiatives exist, such as fire management programmes and restoration projects (Puspitaloka et al., 2020).
21   Since 2016, the Government of Indonesia has rewetted ~380,000 hectares of degraded peatlands mainly
22   through canal blocking and flooding, but less than 2000 hectares were successfully restored to native plant
23   species common to peat swamp forests (Giesen, 2021).Replanting native trees has had relatively low success
24   (Lampela et al., 2017) because they have a low tolerance to prolonged inundation and a lack of fire
25   adaptation strategies (Page et al., 2009; Roucoux et al., 2013; Dohong et al., 2018; Cole et al., 2019; Luom,
26   2020; Giesen, 2021). Barriers to successful management are complex, and include the disparity in
27   timeframes between ecological restoration and political/socioeconomic needs (Harrison et al., 2020) and an
28   over-focus on fire-fighting rather than fire prevention (Mishra et al., 2021a). Early protection of peat forests
29   has been highlighted as a more effective management strategy than restoration, not only in insular SEA but
30   also in areas like Papua New Guinea, which may be targeted for expansion of estate crop plantations (Neuzil
31   et al., 1997; Dennis, 1999; Anshari et al., 2001; Anshari et al., 2004; Hooijer et al., 2006 Assessment of; Heil
32   et al., 2007; Page et al., 2009; Page et al., 2011; Posa et al., 2011; International, 2012; Miettinen et al., 2012;
33   Biagioni et al., 2015; Miettinen et al., 2016; Rieley and Page, 2016; Adila et al., 2017; Cole et al., 2019;
34   Vetrita and Cochrane, 2019; Harrison et al., 2020; Hoyt et al., 2020; Ruwaimana et al., 2020; Ward et al.,
35   2020; Cole et al., 2021).
37   2.6.6   Limits to Adaptation Actions by People
39   The evidence summarised above (Sections 2.6.2 - 2.6.4) shows that by restoring ecosystems it is possible to
40   increase their resilience to climate change, including the resilience of the populations of species they support
41   and of human communities. However, changes to healthy ecosystems and biodiversity are already happening
42   as described in this chapter (robust evidence, high agreement) and further changes are inevitable even under
43   low greenhouse gas emissions scenarios (robust evidence, high agreement). Planning to manage the
44   consequences of inevitable changes and prioritise investments in conservation actions where they have best
45   chance of succeeding (e.g. Section will be an increasingly necessary component of adaptation
46   (robust evidence, high agreement) (Table 2.6).
48   It is possible to help species survive by active interventions such as translocation but as described above
49   (Section it is not a straightforward process, is not suitable for all species and is resource intensive.
50   Modifying local microclimate or hydrological conditions can work for some species (Sections 2.6.2,,
51   but is likely to be less effective at higher levels of climate change. It will also be less successful for larger
52   species and more mobile ones. The microclimate of a tree is much more closely coupled with wider
53   atmospheric conditions than that of a small plant or animal in the boundary layer and mobile species such
54   birds and large mammals range over large areas rather than being confined to discrete locations where
55   conditions can be manipulated. There is potential for using evolutionary changes to enhance the adaptive
56   capacity for target species, such as is being done in the Great Barrier reef by translocating symbionts and
57   corals that have survived recent intense heat-induced bleaching events into areas that have had large die-off.

     Do Not Cite, Quote or Distribute                      2-129                                        Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   However, known limitations to genetic adaptation preclude species-level adaptation to climates beyond their
 2   ecological and evolutionary history (Sections; 2.6.1). All of these interventionist approaches are
 3   constrained by requiring significant financial resources and expertise, they also require a high level of
 4   understanding of individual species autecology, which can take years to acquire, even if resources were
 5   available. Ex situ conservation (for example in seed banks may be the only option to conserve some species,
 6   especially as levels of warming increase and this will not be feasible for all.
 8   While the science of restoration has generated many successes, some habitats are very difficult to restore,
 9   making certain decisions effectively irreversible. For example, Acacia nilotica was introduced into Indonesia
10   in the 1850s for gum arabic with planting expanded for fire breaks in the 1960s. This tree became invasive
11   and has already replaced >50% of savanna habitat in Baluran National Park, with complete replacement
12   expected in the near future. This shift from savanna to Acacia forest is causing large declines in native
13   species, including the charismatic wild banteng, Bos javanicus and wild dog (dhole, Cuon alpinus),
14   (Caesariantika et al., 2011; Padmanaba et al., 2017; Zahra et al., 2020). Multiple approaches to controlling
15   the spread of this Acacia have been ineffective, highlighting the difficulty of reversing the decision to plant
16   this tree (Zahra et al., 2020). Another example is the difficulties in restoring tropical peat forests of South
17   East Asia (Section 6.5.10).
19   EbA when implemented well can reduce risks to people but there are limits, for example, an extreme flood
20   event may exceed the capacity of natural catchments to hold water or slow its flow (Dadson et al., 2017),
21   urban shade trees and green space can make a few degrees difference to temperatures experienced by people
22   but that may not be enough in the hottest conditions.
24   In general adaptation measures can substantially reduce the adverse impacts of 1-2°C of global temperature
25   rise, but beyond this losses will increase (IPCC, 2018b), including species extinctions and changes, such as
26   major biome shifts which are cannot be reversed on human timescales. Some adaptation measures will also
27   become less effective at higher temperatures. Whilst adaptation is essential to reduce risks, it cannot be
28   regarded as a substitute for effective climate change mitigation (robust evidence, high agreement).
30   2.6.7   Climate Resilient Development
32   Climate Resilient Development (CRD) is the subject of Chapter 18. This section briefly assesses some of the
33   fundamental issues for CRD relating to ecosystems; an overview of the importance of specific ecosystem
34   services for CRD is presented in Box 18.7 in Chapter 18. A large body of evidence has demonstrated the
35   extent to which human life, well-being and economies are dependent on healthy ecosystems and the range of
36   threats they are under (high confidence) (IPBES, 2019; Dasgupta, 2021; Pörtner et al., 2021). An analysis of
37   64 studies found a strong positive synergy among eight critical regulating services of healthy ecosystems,
38   including climate regulation, water provisioning, pest and disease control and adjacent crop pollination (Lee
39   and Lautenbach, 2016). Health of ecosystems is, in turn, reliant upon maintenance of natural levels of
40   species' richness and functional diversity (high confidence) (Lavorel et al., 2020) (see Section 2.5.4). A
41   meta-analysis of 74 studies documented the mechanism for increased ecosystem stability was increased
42   asynchrony among species, that itself was a product of higher species diversity (Xu, 2021, consistently
43   positive effect). Responding to these threats requires the protection and restoration of natural and semi-
44   natural ecosystems, together with sustainable management of other areas.
46   The Convention on Biological Diversity set the Aichi 2020 target of 17% of each country to be protected for
47   biodiversity. Analyses suggest that 30% or even 50% of land and sea needs to be protected or restored to
48   confer adequate protection for species and ecosystem services (high confidence) (Pimm et al., 2018;
49   Dinerstein et al., 2019) (Woodley et al., 2019; Brooks et al., 2020; Hannah et al., 2020; Luther et al., 2020;
50   Zhao et al., 2020; Sala et al., 2021). Hannah (2020) estimated that limiting warming to 2°C and protecting
51   30% of high biodiversity regions (Africa, Asia and Latin America) reduced risk of species' extinctions in
52   half (medium confidence). Placement of protected areas is as important as total area (Pimm et al., 2018), and
53   quality of protection (strictness and enforcement) is as important as the official land designation (Shah et al.,
54   2021). Pimm et al. (Pimm et al., 2018) found that many small protected areas are successful because they are
55   in areas of very high biodiversity containing species of small ranges size, while many large regions identified
56   as wild often are of low biodiversity value, though they may have high mitigation value (e.g. high Arctic
57   tundra). Finally, a global meta-analysis of 89 restoration projects, biodiversity increased by 44% and

     Do Not Cite, Quote or Distribute                     2-130                                       Total pages: 237
     FINAL DRAFT                                      Chapter 2                    IPCC WGII Sixth Assessment Report

 1   ecosystem services by 25% after restoration, but values remained lower than in intact reference systems
 2   (Benayas José et al., 2009).
 4   There is also increasing evidence, reported in this chapter, that the loss and degradation of natural and semi-
 5   natural habitats exacerbates the impacts of climate change and climatic extreme events on biodiversity and
 6   ecosystem services (high confidence); example references include: (Ogden et al., 2013; Eigenbrod et al.,
 7   2015; Struebig et al., 2015; Stevens et al., 2016; Oliver et al., 2017; McAlpine et al., 2018; Taffarello et al.,
 8   2018; Lehikoinen et al., 2019; Birk et al., 2020; Chapman et al., 2020; Agol et al., 2021; Khaniya et al.,
 9   2021; Lara et al., 2021; Lehikoinen et al., 2021). Considering these two sets of evidence together, it is clear
10   that climate change adaptation and ecosystem degradation both need to be addressed if either is to be tackled
11   successfully (robust evidence, high agreement) as a number of recent publications have concluded (Haddad
12   et al., 2015; Hannah et al., 2020; Arneth et al., 2021; Pörtner, 2021). Taking this combined body of evidence
13   together, this assessment is that the protection and restoration of natural and semi-natural ecosystems are key
14   adaptation measures (robust evidence, high agreement) (Section 2.5.4).
16   Large scale protection and restoration of ecosystems can also make a significant contribution to climate
17   change mitigation (Dinerstein et al., 2020; Roberts et al., 2020a; Soto-Navarro et al., 2020). Globally, there
18   is a 38% overlap between areas of high carbon storage and high intact biodiversity (mainly in the peatland
19   tropical forests of Asia, Western Amazon and the high Arctic), but only 12% of that is protected (high
20   confidence) (Soto-Navarro et al., 2020). Peatlands are particularly important carbon stores but are threatened
21   human disturbance, land use change (Leifeld et al., 2019) and fire (Turetsky et al., 2015). Restoration of
22   peatlands is not only an efficient nature climate solution in terms of GHG (Nugent et al., 2019), but it may
23   also increase ecosystem resilience (Glenk et al., 2021). Global restoration efforts are ongoing to target
24   degraded temperate peatlands in the American and Europe (Chimner et al., 2017), as a result of their
25   importance for climate change mitigation being recognised (Paustian et al., 2016; Bossio et al., 2020;
26   Humpenöder et al., 2020; Drever et al., 2021; Tanneberger et al., 2021). It has been estimated that the global
27   GHG saving potential of peatland restoration is similar to the most optimistic sequestration potential from all
28   agricultural soils (Leifeld and Menichetti, 2018). However the pressure on peatlands from human activity
29   remains high in many parts of the world (Humpenöder et al., 2020; Tanneberger et al., 2021). Currently, the
30   rapid destruction of tropical peatlands overshadows any current restoration efforts in temperate peatlands or
31   any potential carbon gain from natural high-latitude peatlands (Roucoux et al., 2017; Wijedasa et al., 2017;
32   Leifeld et al., 2019). (Sections,,,,
34   Recent studies have highlighted the importance of ensuring that ecosystem protection is not implemented in
35   a way which disadvantages those who live in or depend on the most intact ecosystems (Mehrabi et al., 2018;
36   Schleicher et al., 2019) or risk food security. The actual area of land to be protected and the balance between
37   sustainable use and protection will need careful planning and targeting to where it can have most benefit
38   (Pimm et al., 2018); it will also be important to ensure that protection measures are effective in preventing
39   damage (Shah et al., 2021).
41   At a local level, EbA can often provide a wide range of additional benefits for sustainable development in
42   both rural and urban areas (Wilbanks, 2003; Nelson et al., 2007; Cohen-Shacham et al., 2016; Hobbie and
43   Grimm, 2020; Martín et al., 2020). A number of the case studies above, such as those in Durban and at
44   Bhojtal Lake illustrate this (2.6.5). A key element of Climate Resilient Development is ensuring that actions
45   taken to mitigate climate change do not compromise adaptation, biodiversity and human needs. This depends
46   on choosing appropriate actions for different locations (Box 2.2, Cross-Chapter Box NATURAL this
47   Chapter). A particularly notable case of this is woodland creation as described in Box 2.2: re-afforestation of
48   previously forested areas can provide multiple benefits (Lee et al., 2018; Lee et al., 2020) including for
49   climate change mitigation, adaptation and biodiversity. However planting trees where they would not
50   naturally grow can create multiple problems include the loss of native biodiversity and disruption of
51   hydrology (Box 2.2). It is also the case that protection of existing natural forest ecosystems is the highest
52   priority for reducing greenhouse gas emissions (Moomaw et al., 2019) and restoration may not always be
53   practical (see Section (Sections,,,,,,, Box 2.2,
54   Cross-Chapter Box NATURAL this Chapter)
56   In some cases actions supported by international donors and presented as addressing climate change
57   adaptation and mitigation in the natural environment can have damaging consequences for people and nature

     Do Not Cite, Quote or Distribute                       2-131                                       Total pages: 237
     FINAL DRAFT                                    Chapter 2                   IPCC WGII Sixth Assessment Report

 1   as well as failing to deliver adaptation and mitigation. One example of this was presented by (Work et al.,
 2   2019) who reviewed three climate change mitigation and adaptation projects in Cambodia: an irrigation
 3   project, a protected-area forest management project, and a reforestation project. In each case they found
 4   evidence of local communities rights being violated, maladaptation and destruction of biodiverse habitats.
 5   They concluded that the potential for maladaptation and adverse social and environmental impacts had been
 6   ignored by international donors as well as national authorities and that there was a need for much strict
 7   accountability mechanisms. Moyo et al. (Moyo et al., 2021), using case studies from South Africa,
 8   documented higher success of ecosystem restoration projects when they embraced broader SDGs,
 9   particularly enhancement of people's livelihoods. Better assessment of the impacts of adaptation and
10   mitigation measures on people and ecosystems, before they are implemented, will be increasingly necessary
11   to avoid unintended and damaging consequences as their deployment is scaled up (Larsen, 2014; Enríquez-
12   de-Salamanca et al., 2017; Pour et al., 2017). This applies to ostensibly nature-based approaches as well as
13   more engineering-based ones.
15   Another aspect of the benefits to people from ecosystems that needs to be taken into account in Climate
16   Resilient Development is increasingly strong evidence of the benefits of natural environments for human
17   health and well-being beyond the provision of basic necessities, such as food and water (Bratman et al.,
18   2019; Marselle et al., 2021). Meta-analyses of 162 studies across 51,738 people documented that individuals
19   with high levels of contact with nature through their lives felt significantly happier, healthier and more
20   satisfied with their lives, and engage in more pro-nature behaviours, than those with little or no contact with
21   nature (high confidence) (Capaldi et al., 2014; Mackay and Schmitt, 2019; Pritchard et al., 2020; Whitburn et
22   al., 2020). Meta-analyses of manipulative human trials across 65 studies document a significant increase in
23   positive feelings and attitudes, and declines in negative feelings, after experimental treatments involving
24   nature (medium confidence) (Bowler et al., 2010b; McMahan and Estes, 2015; Soga et al., 2017). Within the
25   context of CRD improving the extent to which humans see themselves as part of the natural world – known
26   as human-nature connectedness (HNC) – by increasing access to natural areas, particularly within urban
27   areas, can provide additional health, cultural and recreation benefits of NbS as well as increasing public
28   engagement and support (robust evidence, high agreement) (Wilbanks, 2003; Nelson et al., 2007; Bowler et
29   al., 2010b; Capaldi et al., 2014; McMahan and Estes, 2015; Cohen-Shacham et al., 2016; Soga et al., 2017;
30   Mackay and Schmitt, 2019; Work et al., 2019; Hobbie and Grimm, 2020; Pritchard et al., 2020; Whitburn et
31   al., 2020).
36   Cross-Chapter Box NATURAL: Nature-Based Solutions for Climate Change Mitigation and
37        Adaptation
39   Authors: Camille Parmesan (Chapter 2), Gusti Anshari (Chapter 2, CCP7), Polly Buotte (Chapter 4),
40   Donovan Campbell (Chapter 15), Edwin Castellanos (Chapter 12), Annette Cowie (WGIII Chapter 12),
41   Marta Rivera Ferre (Chapter 8), Patrick Gonzalez (Chapter 2, CCP3), Elena López Gunn (ch4), Rebecca
42   Harris (ch2, CCP3), Jeff Hicke (Chapter 14), Rachel Bezner Kerr (Chapter 5), Rodel Lasco (Chapter 5),
43   Robert Lempert (Chapter 1), Brendan Mackey (Chapter 11), Paulina Martinetto (Chapter 3), Robert
44   Matthews (WGIII, Chapter 3), Timon McPhearson (Chapter 6), Mike Morecroft (Chapter 2, CCP5), Aditi
45   Mukherji (Chapter 4), Gert-Jan Nabuurs (WGIII Chapter 7), Henry Neufeldt (Chapter 5), Roque Pedace
46   (WGIII Chapter 3), Julio Postigo (Chapter 12), Jeff Price (Chapter 2, CCP1), Juan Pulhin (Chapter 10), Joeri
47   Rogelj (WGI Chapter 5), Daniela Schmidt (Chapter 13), Dave Schoeman (Chapter 3), Pramod Kumar Singh
48   (Chapter 18), Pete Smith (WGIII Chapter 12), Nicola Stevens (Chapter 2, CCP3), Stavana E. Strutz (ch2),
49   Raman Sukumar (Chapter 1), Gautam Talukdar (Chapter 2, CCP1), Maria Cristina Tirado (Chapter 7),
50   Chapter ris Trisos (Chapter 9)
52   Nature-based Solutions provide adaptation and mitigation benefits for climate change as well as
53   contributing to other sustainable development goals (high confidence). Effective nature-based climate
54   mitigation stems from inclusive decision-making and adaptive management pathways that deliver climate-
55   resilient systems serving multiple sustainable development goals. Robust decision-making adjusts
56   management pathways as systems are impacted by on-going climate change. Poorly conceived and
57   designed nature-based mitigation efforts have the potential for multiple negative impacts, including

     Do Not Cite, Quote or Distribute                     2-132                                     Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   competing for land and water with other sectors, reducing human well-being and failing to provide
 2   mitigation that is sustainable in the long-term (high confidence).
 4   The concept of Nature-based Solutions (NbS) is broad and debated but has become prominent in both the
 5   scientific literature and policy since AR5, including earlier concepts, including Ecosystem based Adaptation
 6   (EbA). The key point is that these are actions benefiting both people and biodiversity (IUCN 2020; WGII
 7   Glossary). In the context of climate change, NbS provide adaptation and mitigation benefits for climate
 8   change in ways that support wild species and habitats, often contributing to other sustainable development
 9   goals (robust evidence, high agreement) (Keesstra et al., 2018; Lavorel et al., 2020; Malhi et al., 2020)
10   (Griscom et al., 2017; Hoegh-Guldberg et al., 2019; IPCC, 2019a; Lewis et al., 2019; Seddon et al.,
11   2020b)(AR6 WGIII Chapter 12, see Sections 2.2, 2.5.4, 2.6.3, 2.6.5, 2.6.7). Well-designed and implemented
12   NbS mitigation schemes can increase carbon uptake or reduce greenhouse gases emissions at the same time
13   as protecting or restoring biodiversity and incorporating elements of food provisioning (Mehrabi et al.,
14   2018). A variety of measures can be part of NbS, ranging from the protection of natural terrestrial,
15   freshwater and marine ecosystems, the restoration of degraded ones (this Cross-Chapter Box; Section 13.3),
16   to more sustainable management of naturally regenerating ecosystems used for food, fibre and energy
17   production (Figure Cross-Chapter Box NATURAL 1, Chapter 5, Cross-Working Group Box
18   BIOECONOMY in Chapter 5). Agroecological practices mitigate and adapt to climate change and can
19   promote native biodiversity (high confidence) (Sinclair et al., 2019; Snapp et al., 2021).
21   The role of restoration in NbS
23   Where natural ecosystems have been degraded or destroyed, re-establishing them and restoring natural
24   processes can be a key action for adaptation and mitigation, and the science of restoration is well-established
25   (de los Santos et al., 2019; Duarte et al., 2020) (Section 13.4.1). Such restoration activities need to adapt to
26   on-going climate change risks for the landscape and ocean scape and the species composition of biological
27   communities. Indeed, climate-change impacts may overwhelm attempts at restoration/conservation of
28   previous or existing ecosystems, particularly when the ecosystem is already near its tipping point, as are
29   tropical coral reefs (Bates et al., 2019; Bruno et al., 2019).
31   Lands (e.g.forests) and oceans (e.g. fisheries) managed for products using sustainable practices (whether
32   applied by private, state, or indigenous peoples) can also be carbon- and biodiversity-rich, and so effective
33   NbS (Paneque-Gálvez et al., 2018; Soto-Navarro et al., 2020). Indigenous people and private forest owners
34   manage, use or occupy at least one-quarter of global land area, over one-third of which overlaps with
35   protected areas, thus combining both protection and production (Jepsen et al., 2015; Garnett et al., 2018;
36   IPBES, 2019; Santopuoli et al., 2019).
38   Protection/restoration of natural systems, including reducing non-climate stressors, and sustainable
39   management of semi-natural areas emerge as necessary actions for adaptation to minimise extinctions of
40   species, reaching tipping points that cause regime shifts in natural systems, loss of whole ecosystems, and
41   their associated benefits for humans (Scheffer et al., 2001; Folke et al., 2005; Luther et al., 2020) (Chapters 2
42   and 3, AR6 WGIII Chapter 7). Such measures are critical for conservation of biodiversity and the provision
43   of ecosystem goods and services in the face of projected climate change (Duarte et al., 2020). Supporting
44   local livelihoods and providing benefits to indigenous, local communities and millions of private
45   landowners, together with their active engagement in decision-making, is critical to ensuring support for
46   NbS and its successfully delivery (high confidence) (Chapter 05; Figure Cross-Chapter Box NATURAL 1;
47   Ceddia et al., 2015; Blackman et al., 2017; Nabuurs et al., 2017; Smith et al., 2019a; Smith et al., 2019b;
48   Jones et al., 2020a; McElwee et al., 2020; Cao et al., 2021).
50   Forests
52   Intact natural forest ecosystems are major stores of carbon and support large numbers of species that cannot
53   survive in degraded habitats (very high confidence). Extensive areas of natural forest ecosystems remain in
54   tropical, boreal and (to a lesser extent) temperate biomes regions, but in many regions, they are managed
55   (sustainably and unsustainably) or have been degraded or cleared. Deforestation and degradation continue to
56   be a source of global greenhouse gas emissions (very high confidence) (Friedlingstein et al., 2019).
57   Protecting existing natural forests and sustainable management of semi natural forests providing goods and

     Do Not Cite, Quote or Distribute                      2-133                                      Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   services is a highly effective NbS (Bauhus et al., 2009) (high confidence). Natural forests and sustainably
 2   managed diverse forests play important roles for climate change mitigation and adaptation while providing
 3   many other ecosystem goods and services (very high confidence) (Bradshaw and Warkentin, 2015; Favero et
 4   al., 2020; Mackey et al., 2020). Contributions to climate change mitigation are estimated at medians of 5-7
 5   Gt CO2/y (Roe et al., 2019). Forests influence the water cycle at local, regional and global scales (Creed and
 6   van Noordwijk, 2018) reducing surface runoff, increasing infiltration to groundwater and improving water
 7   quality (Bruijnzeel, 2004; Zhou et al., 2015a; Ellison et al., 2017; Alvarez-Garreton et al., 2019). Recent
 8   evidence shows that downwind precipitation is also influenced by evapo-transpiration from forests (Keys et
 9   al., 2016; Ellison et al., 2017). Protecting existing natural forest and sustainably managing production
10   forests, in a holistic manner, can optimise the provision of the many functions forests fulfil for owners,
11   conservation, mitigation and for society as a whole (Bauhus et al., 2009; Nabuurs et al., 2013).
13   Reforestation of formerly forested land can help to protect and recover biodiversity and can be one of the
14   most practical and cost effective ways of sequestering and storing carbon (high confidence) (Nabuurs et al.,
15   2017; Hoegh-Guldberg et al., 2018b; Paneque-Gálvez et al., 2018; Smith et al., 2018; Cook-Patton et al.,
16   2020; Cowie et al., 2021; Drever et al., 2021). This can be achieved through planting or by allowing natural
17   colonisation by tree and shrub species. The most effective method to employ depends upon local
18   circumstances (such as the presence of remnant forest cover) or socio-cultural and management objectives.
19   Reforestation with climate-resilient native or geographically near species restores biodiversity at the same
20   time as sequestering large amounts of carbon (Lewis et al., 2019; Rozendaal et al., 2019). It can also restore
21   hydrological processes, improving water supply and quality (Ellison et al., 2017) and reducing risks of soil
22   erosion and floods (high confidence) (Locatelli et al., 2015).
24   Climate change may mean that in any given location, different species will be able to survive and become
25   dominant, and restoring the former composition of forests may not be possible (Sections 2.4; 2.5). Severe
26   disturbances such as insect/pathogen outbreaks, wildfires, and droughts, which are an increasing risk, can
27   cause widespread tree mortality resulting in sequestered forest carbon being returned to the atmosphere
28   (Anderegg et al., 2020; Senf and Seidl, 2021) thus suggesting we need to adapt (Sections 2.4, 2.5, 13.3
29   14.4.1, Box 14.1). Adaptation measures, such as increasing the diversity of forest stands through ecological
30   restoration rather than monoculture plantations can help to reduce these risks (confidence). When plantations
31   are established without effective landscape planning and meaningful engagement including free prior and
32   informed consent, they can present risks to biodiversity and the rights, well-being and livelihoods of
33   Indigenous and local communities, as well as being less climate-resilient than natural forests (very high
34   confidence) (Section 5.6; Corbera et al., 2017; Mori et al., 2021).
36   Afforesting areas such as savannahs and many temperate peatlands, which would not naturally be forested,
37   damages biodiversity and increase vulnerability to climate change (high confidence) so is not a Nature-based
38   Solution and can exacerbate greenhouse gas emissions (Sections,, Box 2.2 this Chapter).
39   Remote sensing based assessments of suitability for tree planting can over-estimate potential, due to failure
40   to adequately distinguish between degraded forest and naturally open areas (Bastin et al., 2019; Veldman et
41   al., 2019; Bastin et al., 2020; Sullivan et al., 2020).
43   Peatlands
45   Peatlands are naturally high-carbon ecosystems, which have built up over millennia. Draining, cutting and
46   burning peat lead to oxidation and the release of CO2 (very high confidence). Rewetting by blocking
47   drainage, preventing cutting and burning can reverse this process on temperate peatlands (medium
48   confidence) although can take many years (Bonn et al., 2016). Trees are naturally found on many tropical
49   peatlands and restoration can involve removing non-native species such as oil palm and re-establishing
50   natural forest. However, peatland tropical forest is difficult to fully restore, and native pond fish, that are
51   vital as a local food, often do not return. Protection of intact peat forests, rather than attempting to restore
52   cleared forest, is by far the more effective pathway both in terms of cost, CO2 mitigation, and protection of
53   food sources (Kreft and Jetz, 2007). Naturally treeless temperate and boreal peatlands have in some cases
54   been drained to enable trees to be planted, which leads to CO2 emissions, and restoration requires removal of
55   trees as well as re-blocking drainage. (high confidence) (Sections;;
57   Blue Carbon

     Do Not Cite, Quote or Distribute                     2-134                                      Total pages: 237
     FINAL DRAFT                                    Chapter 2                   IPCC WGII Sixth Assessment Report

 2   Blue Carbon ecosystems (mangroves, saltmarshes and seagrass meadows; see glossary) often have high rates
 3   of carbon accumulation and sequestration (Section; Macreadie et al., 2019). However, quantification
 4   of their overall mitigation value is difficult due to variable production of CH4 and N2O (Adams et al., 2012;
 5   Rosentreter et al., 2018; MacLean et al., 2019b), uncertainties regarding the provenance of carbon
 6   accumulated (Macreadie et al., 2019), and the release of CO2 by biogenic carbonate formation in seagrass
 7   ecosystems (Saderne et al., 2019). Therefore, blue carbon strategies, referring to climate change mitigation
 8   and adaptation actions based on conservation and restoration of blue carbon ecosystems, can be effective
 9   NbS, with evidence of recovery in carbon stocks following restoration, although their global or regional
10   carbon sequestration potential and net mitigation potential may be limited (medium confidence) (Sections
11; 13.4.3, AR6 WGI; Duarte et al., 2020). They can also significantly attenuate wave
12   energy, raise the seafloor thus counteracting sea level rise effects, and buffer storm surges and flooding
13   erosion (high confidence) (Sections 13.2.2; 13.10.2). Additionally, they provide a suite of cultural (for
14   example, tourism, livelihood and well-being for native and local communities), provision (e.g. mangrove
15   woods, edible fish and shellfish) and regulation (e.g. nutrient cycling) services (high confidence) (Section
16 These services have motivated the implementation of management and conservation strategies of
17   these ecosystems (Sections; 13.4.2). Blue carbon strategies are relatively new, with many of them
18   experimental and small scale; therefore there is limited evidence of their long-term effectiveness. There is
19   also limited information on the potential emission of other GHGs from restored blue carbon ecosystems,
20   although reconnecting hydrological flow in mangroves and saltmarsh restoration are effective interventions
21   to reduce CH4 and CO2 (limited evidence, medium agreement) (Kroeger et al., 2017; Al-Haj and Fulweiler,
22   2020).
24   Urban NbS
26   Nature-based Solutions can be a key part of urban climate adaptation efforts. Direct human adaptation
27   benefit may stem from the cooling effects of urban forests and green spaces (parks and green roofs), from
28   coastal wetlands and mangroves reducing storm surge and flooding, and from sustainable drainage systems
29   designed to reduce surface flooding from extreme rainfall, as well as general benefits to human health and
30   well-being (high confidence) (Sections 2.2; 2.6; Chapter 6; Frantzeskaki et al., 2019; Keeler et al., 2019)
31   (Kowarik, 2011). Not all green schemes are considered "Nature Based Solutions" if they do not benefit
32   biodiversity, but carefully designed urban greening can be effective NbS. Careful planning also helps limit
33   negative equity consequences, benefiting wealthy neighbourhoods more than poor (Geneletti et al., 2016;
34   Pasimeni et al., 2019; Grafakos et al., 2020). Effective planning should also consider what is appropriate for
35   the climate and conditions of each city. For example, some trees emit volatiles (e.g. isoprene) that, in the
36   presence of certain atmospheric pollutants, can increase surface ozone that in turn can cause human
37   respiratory problems (Kreft and Jetz, 2007). Wetlands restoration close to human settlements needs to be
38   paired with mosquito control to prevent negative impacts on human health and well-being (Stewart-Sinclair
39   et al., 2020), but has been shown to provide better filtration and toxicity reduction with lower environmental
40   impact than other forms of waste-water treatment (Vymazal et al., 2021), including "green roofs" and "green
41   walls" (Chapter 06; Addo-Bankas et al., 2021).
43   Agroecological Farming
45   Agroecological farming (AF) is a holistic approach that incorporates ecologic and socio-economic
46   principles. It strives to enhance biodiversity, soil health and synergies between agroecosystem components,
47   reduce reliance on synthetic inputs (e.g., pesticides), builds on Indigenous knowledge and local knowledge,
48   and fosters social equity (e.g., supporting fair, local markets (HLPE, 2019; Wezel et al., 2020). AF practices
49   include intercropping, mobility of livestock grazing across landscapes, organic agriculture, integration of
50   livestock, fish and cropping, cover crops and agroforestry. (Sections 5.14; FAQ 12.5, 13.5.)
52   Agroforestry, cover crops and other practices that increase vegetation cover and enhance soil organic matter,
53   carefully management and varying by agroecosystem, mitigate climate change (high confidence) (Zomer et
54   al., 2016; Aryal et al., 2019; Nadège et al., 2019). Global meta-analyses demonstrate agroforestry storing 20
55   -33% more soil carbon than conventional agriculture (De Stefano and Jacobson, 2018; Shi et al., 2018) and
56   reducing the spread of fire (Sections 5.6, 13.5.2, 7.4.3, Box 7.7). Minimising synthetic inputs such as N-
57   based fertilisers reduces emissions (Gerber et al., 2016). Cover crops can reduce N2O emissions and increase

     Do Not Cite, Quote or Distribute                     2-135                                     Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1   soil organic carbon (Abdalla et al., 2019). Conservation farming (no-till with residue retention and crop
 2   rotation) increases soil organic carbon particularly in arid regions (Sun et al., 2020). Silvo-pastoral systems
 3   (pastures with trees), and other practices that increase vegetation cover and enhance soil organic matter
 4   increase sequestered carbon in vegetation and soils (Zomer et al., 2016; Aryal et al., 2019; Nadège et al.,
 5   2019; Ryan et al. 2019). Agroecologically improved cropland and grazing land management have significant
 6   mitigation potential, estimated at 2.8- 4.1 GtCO2e per year (Smith et al., 2020). (Sectiond 5.10, 5.14, Box
 7   5.10, Cross Working-Group Box BIOECONOMY in Chapter 5; WGIII 7.4.3; Box 7.7).
 9   AF enhances adaptation to climate change, including resilience to extreme events. Building organic matter
10   improves soil water-holding capacity and buffers against drought; increased perenniality and high levels of
11   ground cover reduce soil erosion during storms; agroforestry shelters stock and crops in heat waves;
12   landscape complexity and agrobiodiversity increase resilience to disease and pests and stabilise livestock
13   production and restoration of oyster reefs provides thermal refugia and storm surge protection (Allred et al.,
14   2014; Henry et al., 2018; Kremen and Merenlender, 2018; Beillouin et al. 2019; ; Kuyah et al. 2019; Gilby
15   et al., 2020; Niether et al. 2020; Richard et al. 2020; Howie and Bishop, 2021; Snapp et al. 2021) (;) .
16   Livestock mobility enables adjusting to increased climatic variability while maintaining pastoral systems’
17   productivity (Turner and Schlecht, 2019; Scoones, 2020). Thus, adoption of agroecology principles and
18   practices will be highly beneficial to maintaining healthy, productive food systems under climate change
19   (high confidence) (Sections 5.4.4; 13.5.2; FAQ 12.4).
21   AF practices such as hedgerows and polycultures maintain habitat and connectivity for biodiversity and
22   support ecosystem functioning under climate stress compared to conventional agriculture (high confidence),
23   Section; Buechley et al., 2015; Kremen and Merenlender, 2018; Albrecht et al., 2020). Increasing
24   farm biodiversity benefits pollination, pest control, nutrient cycling, water regulation and soil fertility (Snapp
25   et al. 2021; Beillouin et al. 2019; Tamburini et al. 2020). Biodiverse agroforestry systems increase ecosystem
26   services and biodiversity benefits compared to simple agroforestry and conventional agriculture (high
27   confidence); up to 45% more biodiversity and 65% more ecosystem services compared to conventional
28   timber, crop or livestock production in the Brazilian Atlantic Forest (Santos et al., 2019), including for birds
29   (M. Greenler and Ebersole, 2015; Karp et al., 2019), local tree species (Braga et al., 2019) and fewer
30   invasive exotic plants species (Cordeiro et al., 2018). AF includes conservation of semi-natural woodlands,
31   which can conserve bird predators of insect pests (Gonthier et al., 2019). Organic production increases insect
32   species richness and abundance in and around the farm, including essential pollinators (Sections 5.10, 12.6;
33   Kennedy et al., 2013; Haggar et al., 2015; Lichtenberg et al., 2017).
35   AF significantly improves food security and nutrition by increasing access to healthy, diverse diets and
36   rising incomes for food producers, through increased biodiversity of crops, animals, and landscapes (high
37   confidence) (Garibaldi et al., 2016; D'Annolfo et al., 2017; Isbell et al., 2017; Dainese et al., 2019; Bezner
38   Kerr et al. 2021). Livestock mobility improves the site-specific matching of animals’ needs with food
39   availability (Damonte et al., 2019; Mijiddorj et al., 2020; Postigo, 2021), and can generate a form of
40   rewilding that restores lost ecosystem functioning (Gordon et al., 2021). Conservation of crop wild relatives
41   in situ supports genetic diversity in crops for the range of future climate scenarios (Redden et al., 2015).
43   System-level agroecological transitions require policy support for farmer experimentation and knowledge
44   exchange, community-based participatory methodologies and market and policy measures e.g. public
45   procurement, local and regional market support, regulation or payments for environmental services (HLPE
46   2019; Snapp et al. 2021; Mier y Teran et al. 2018). Scientific consensus about the food security and
47   environmental implications of agroecological transitions at a global scale is lacking. Yields in agroforestry
48   and organic can be lower than high-input agricultural systems, but conversely, AF can boost productivity and
49   profit, varying by timeframe, socio-economic, political and ecosystem context (medium confidence) (Section
50   5.14; Muller et al., 2017; LaCanne, 2018; Barbieri et al., 2019; Rosa-Schleich et al. 2019; Smith et al.,
51   2019b; Smith et al., 2020). ;. Such contrasting results and the limited investment in agroecological research
52   to date make paramount assessing the global and regional impacts of agroecological transitions on food
53   production, ecosystems and economy (Section 5.14; DeLonge et al., 2016; Muller et al., 2017; Barbieri et al.,
54   2019).
56   Conclusions

     Do Not Cite, Quote or Distribute                      2-136                                       Total pages: 237
     FINAL DRAFT                                        Chapter 2                         IPCC WGII Sixth Assessment Report

 1   Nature–based Solutions provide adaptation and mitigation benefits for climate change as well as contributing
 2   to other sustainable development goals (high confidence). NbS avoid further emissions and promote CO2
 3   removal using approaches that yield long-lasting mitigation benefits and avoid negative outcomes for other
 4   sustainable development goals. Poorly conceived and designed mitigation efforts have the potential for
 5   multiple negative impacts: (1) They can have cascading negative effects on long-term mitigation by
 6   promoting short term sequestration over existing long-term accumulated carbon stocks, (2) They can be
 7   detrimental to biodiversity, undermining conservation adaptation, and (3) They can erode other ecosystem
 8   services important for human health and well-being (high confidence). Conversely, well-designed and
 9   implemented mitigation efforts have the potential to provide co-benefits in terms of climate-change
10   adaptation, as well as multiple goods and services, including conservation of biodiversity, clean and
11   abundant water resources, flood mitigation, sustainable livelihoods, food and fibre security, and human
12   health and well-being (high confidence). A key aspect to such 'smart' climate mitigation is implementation of
13   inclusive and adaptive management pathways (Section 1.4.2). These entail acceptance of the inherent
14   uncertainty in projections of future climate change, especially at a regional or local level, and using decision
15   making processes that keep open as many options as possible, for as long as possible, with periodic re-
16   evaluation to aid in choosing pathways forward even as systems are being impacted by on-going climate
17   change (Figure Cross-Chapter Box NATURAL 1; Cross-Chapter Box DEEP in Chapter 17; Section 1.4.2).

21   Figure Cross-Chapter Box NATURAL.1: Decision-making framework to co-maximise adaptation and mitigation
22   benefits from natural systems. Decision-making pathways are designed to add robustness in the face of uncertainties in
23   future climate change and its impacts. Emphasis is on keeping open as many options as possible, for as long as possible,
24   with periodic re-evaluation to aid in choosing pathways forward even as systems are being impacted by on-going
25   climate change
28   Table Cross-Chapter Box NATURAL.1: Assessment of benefits and tradeoffs between mitigation and strategies for
29   both biodiversity and human adaptation to future climate change. Best practices highlight approaches that lead to
30   maximal positive synergies between mitigation and adaptation; worst practices are those most likely to lead to negative
31   tradeoffs for adaptation. Many best practices have additional societal benefits beyond adaptation, such as food
32   provisioning, recreation and improved water quality. Mitigation Potential (Mit. Pot.) and Restoration Potential (Rest.
33   Pot.) are considered.
                                                                         Worst practices and
                                           Best practices and                                  Additional societal
        System      Mit. Pot. Rest. Pot.                                 negative adaptation                         References
                                           adaptation benefits                                      benefits

     Do Not Cite, Quote or Distribute                            2-137                                         Total pages: 237
FINAL DRAFT                                          Chapter 2                            IPCC WGII Sixth Assessment Report

Boreal         Medium   Medium     Maintain or restore            Very large scale clear cuts,   Providing goods and Drever et al.
Forests                            species and structural         aiming for one or few tree     services, improved (2021)
                                   diversity, reduce fire risk,   species, although boreal is    air quality,
                                   spatially separate wood        characterised by few tree      improved
                                   production, and                species and a natural fire     hydrology, jobs
                                   sustainably intensify          risk
                                   management in some
Temperate      Very high High      Maintain or restore natural    Planting large scale non-      Providing goods and   Sections 2.4.3;
forests                            species and structural         native monocultures which      services, jobs,       2.5; Box 2.2 ;
                                   diversity, leading to more     would lead to loss of          improved hydrology    Nabuurs et al.
                                   biodiverse and resilient       biodiversity and poor          and biodiversity      (2017); Roe et
                                   system                         climate change resilience                            al. (2019);
                                                                                                                       Favero et al.
Tropical wet   High     Moderate Maintain or restore natural      Planting non-native            Indiginous foods,     Section 2.4.3
forests                          species and structural           monocultures, loss of          medicines and other Edwards et al.
                                 diversity, high                  biodiversity, poor climate     forest products,      (2014)
                                 biodiversity, more               change resilience, soil        including
                                 resilient to climate change      erosion                        sustainable selective
Tropical dry   High     Moderate Integrated landscape             Planting non-native                                  Foli et al.
forests                          management                       monocultures, Loss of                                (2018)
                                                                  biodiversity, poor climate
                                                                  change resilience, soil
Tropical       Very high Low       Integrated landscape           Cutting native rainforest and Forest pond fish are Section 2.4.3;
peatland                           management                     planting palm oil for         major food for local 2.5; Smith et
forests                                                           biodiesel results in very     communities          al. (2019b)
                                                                  high carbon emissions from
                                                                  exposed peat soils
Blue Carbon                                                                                                            AR6 WGI
Mangroves      Moderate High       Conservation, restoration Potential NH4 emissions             Improved fisheries Sections
                                   of hydrological flows,                                        and biodiversity,;
                                   revegetation with native                                      coastal protection;
                                   plants, livelihoods                                           against SLR and ;
                                   diversification, landscape                                    storm surges,         Macreadie et
                                   planning for landward and                                     recreation and        al. (2019);
                                   upstream migration                                            cultural benefits     Duarte et al.
                                                                                                                       Sasmito et al.
Saltmarshes    Moderate High       Conservation, reduce        Potential NH4 emissions           Improved fisheries Sections
                                   nutrient loads, restoration                                   and biodiversity,;
                                   of hydrological flows and                                     protection against;
                                   sediment delivery,                                            SLR and storm ;
                                   revegetation with native                                      surges, recreational Macreadie et
                                   plants, landscape planning                                    and cultural benefits al. (2019);
                                   for landward and                                                                    Duarte et al.
                                   upstream migration                                                                  (2020)
Seagrasses     Moderate High       Conservation; restoration; Potential NH4 emissions            Improved fisheries Section
                                   improve water quality and                                     and biodiversity;;
                                   reduce local stressors                                        protection from;
                                   (reduction of industrial                                      shoreline erosion; ; de los
                                   sewage, anchoring and                                         recreational benefits Santos et al.
                                   trawling regulation)                                                                (2019);
                                                                                                                       Macreadie et
                                                                                                                       al. (2019);
                                                                                                                       Duarte et al.
Urban Ecosystems

Urban forests Moderate Moderate Integrated landscape              monoculture of an exotic       Recreation &
              to High*          management. Species               tree lowers resilience and     aesthetics;          WGII Chapter
                                richness (including               reduces biodiversity           stormwater           06
                                exotics) can be high                                             absorption benefits;

Do Not Cite, Quote or Distribute                            2-138                                                   Total pages: 237
    FINAL DRAFT                                           Chapter 2                           IPCC WGII Sixth Assessment Report

                                                                                                       heat mitigation, air
    Urban        Moderate* Moderate Integrated landscape                                          Recreation &                  WGII Chapter
    wetlands                        management.                                                   aesthetics;                   06
                                                                                                  absorption; heat
                                                                                                  mitigation; coastal
                                                                                                  flood protection
    Urban        Moderate* Moderate Integrated landscape              fertilized commercial grass Recreation &                  WGII Chapter
    grasslands                      management                        monocultures often require aesthetics;                    06
                                                                      irrigation and are less     stormwater
                                                                      resilient to droughts than  absorption; heat
                                                                      native, mixed grasses and mitigation
    Open grasslands &
    Boreal &      High    Moderate Blocking drainage                  Inappropriate hydrological       Improved water           Sections 2.4.3;
    Temperate                      channels; Raise water              restoration, e.g., flood         quality in some          2.5 ; Bonn et
    Peatlands                      level to natural condition;        surface depth greater than       conditions.              al. (2016);
                                   remove planted trees;              natural depth leading to                                  Nugent
                                   revegetation of bare peat;         methane emissions                                         (2019);
                                   No burns; Increases                                                                          Taillardat et
                                   biodiversity resilience;                                                                     al. (2020)
                                   Reduce flood risk
    Tropical     Moderate High     Control of feral                   Afforestation,over-              Improved grazing         Sections 2.4.3;
    savannas and                   herbivores; Reintroduce            grazing/stocking; No burns;      potential for            2.5; Box 2.1 ;
    grasslands                     indigenous burns;                  inappropriate placement and      livestock and dairy      Stafford et al.
    (including                     reintroduce native                 design of watering points.       production,              (2017); Moura
    rangelands)                    Herbivores, controlled             All leads to loss of             sustainable wildlife     et al. (2019);
                                   grazing; strategic design          biodiversity, and resilience;    harvests, Increased      Shackelford et
                                   of water-holes;                    soil erosion; water              water security,          al. (2021);
                                   Community-based natural            insecurity                       income from eco-         Stringer et al.
                                   resource management,                                                tourism, medicanal       (2021);
                                   grass reseeding, clearing                                           plants, fuel wood.       Wilsey (2021)
                                   of invasive and                                                     Enhanced food
                                   encroaching woody plants                                            security.
    Temperate    Moderate Moderate Integrated landscape               Monoculture of introduced        Sustainable harvest      Sections 2.4.3;
    Grasslands   to High  to High  management; sustainable            species; over-fertilising with   of wildlife, livestock   2.5, Box 2.1;
    and                            grazing; Community-                chemical or organic              and dairy                Farai, (2017);
    rangelands                     based natural resource             amendments; Failure to           production, wild         Baker et al.
                                   management; Native                 manage human-wildlife            fruits, medicinal        (2018);
                                   grassland species more             clashes; Failure to distribute   plants, construction     Homewood et
                                   resistant to drought than          income equitably;                material, fuelwood;      al. (2020)
                                   introduced species                 inadequate enabling policy       income from              Wilsey (2021)
                                                                      to facilitate integrated         ecotourism
                                                                      landscape management
    Agroecologial High      High        Biodiverse systems at         Poorly chosen species,           Food security,           Sections 5.4,
    farming and             (context    landscape scale;              practices and amendments         human health,            5.10, 5.12,
    aquaculture             specific)   participatory adaptation to   can lead to low yields;          livelihoods, socio-      5.14 ;
                                        context; Short value          Simplified agroforestry          cultural benefits e.g.   Coulibaly et
                                        chains; Farmer incentives;    systems or industrial scale      culturally-              al. (2017);
                                        Biodiversity synergies;       organic agriculture lacks        appropriate foods.       HLPE (2019);
                                        reduced climate risk          holistic system-wide                                      Quandt et al.
                                                                      approach. Over-fertilising                                (2019);
                                                                      with organic amendments                                   Sinclair et al.
                                                                                                                                (2019); Smith
                                                                                                                                et al. (2019b);
                                                                                                                                Muchane et al.
                                                                                                                                Reppin et al.

    Do Not Cite, Quote or Distribute                            2-139                                                     Total pages: 237
     FINAL DRAFT                                    Chapter 2                   IPCC WGII Sixth Assessment Report

 3   FAQ2.5: How can we reduce the risk of climate change to people by protecting and managing nature
 4       better?
 6   Damage to our natural environment can increase the risk climate change poses to people. Protecting and
 7   restoring nature can be a way to adapt to climate change, with benefits for both humans and biodiversity.
 8   Examples include reducing flood risk by restoring catchments and coastal habitats, the cooling effects of
 9   natural vegetation and shade from trees and reducing the risk of extreme wildfires by better managing of
10   natural fires.
12   Protecting and restoring natural environments, such as forests and wetlands can reduce the risks climate
13   change poses to people, as well as supporting biodiversity, storing carbon and providing many other benefits
14   for human health and wellbeing. Climate change is bringing an increasing number of threats to people,
15   including flooding, droughts, wildfire, heat waves and rising sea levels. These threats can however be
16   reduced or aggravated, depending on how land, sea and freshwater are managed or protected. There is now
17   clear evidence that ‘Nature-based Solutions’ (NbS) can reduce the risks that climate change presents to
18   people. This is sometimes called ‘Ecosystem-based Adaptation (EbA) and includes::
20       •   Natural flood management: As warm air holds more water, and in some places, because of changing
21           seasonal rainfall patterns, we are seeing more heavy downpours in many parts of the world. This can
22           create serious flooding problems, with loss of life, homes and livelihoods. The risk of flooding is
23           higher where natural vegetation has been removed, wetlands drained or channels straightened. In
24           these circumstances, water flows quicker and the risk of flood defenses being breached is increased.
25           Restoring the natural hydrology of upstream catchments, including by restoring vegetation, creating
26           wetlands and re-naturalising watercourse channels and reinstating connections with the flood plain
27           can reduce this risk. In a natural catchment with trees or other vegetation, water flows slowly
28           overland and much of it soaks into the soil. When the water reaches a watercourse, it moves slowly
29           down the channel, both because of the longer distance it travels when the channel bends and because
30           vegetation and fallen trees slow the flow. Wetlands, ponds and lakes can also hold water back and
31           slowly release it into river systems.
33       •   Restoring natural coastal defences: Rising sea levels as a result of climate change, mean that coasts
34           are eroding at a fast rate and storm surges are more likely to cause damaging coastal flooding.
35           Natural coastal vegetation, such as saltmarsh and mangrove swamps can, in the right places, stabilise
36           the shoreline and act as a buffer, absorbing the force of waves. On a natural coast, the shoreline will
37           move inland and as sea level rises, the coastal vegetation will gradually move inland with it. This
38           contrasts with hard coastal defences such as sea walls and banks, which can be overwhelmed and
39           fail. In many places however, coastal habitats have been cleared and where there are hard sea
40           defences behind the coastal zone, the vegetation disappears as the coast erodes rather than moving
41           inland. This is often referred to as ‘coastal squeeze’ as the vegetation is squeezed between the sea
42           and the sea wall. Restoring coastal habitats and removing hard sea defences, can help reduce the
43           risks of catastrophic flooding.
45       •   Providing local cooling: Climate change is bringing higher temperatures globally, which can result
46           in heatwaves affecting people’s health, comfort and agriculture. In cities, this can be a particular
47           problem for health as temperatures are typically higher than in the countryside. Trees give shade,
48           which people, in both rural and urban areas, have long used to provide cool places for themselves
49           and for crops such as coffee and livestock. Planting trees in the right place can be a valuable, low-
50           cost Natural-based Solution to reduce the effects of increasing heat, including in reducing water
51           temperatures in streams and rivers, which can help to maintain fisheries. Trees and other vegetation
52           also have a cooling effect as a result of water being lost from their leaves through evaporation and
53           transpiration (loss of water through pores in the leaves, known as stomata). Natural areas, parks,
54           gardens in urban areas can help reduce air temperatures by up to a few degrees.
56       •   Restoring natural fire regimes: Some natural ecosystems are adapted to burning, such as savannas
57           and boreal forests. Where fire has been suppressed or non-native species of trees planted in more

     Do Not Cite, Quote or Distribute                    2-140                                      Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1           open habitats, there is a risk that potential fuel accumulates, which can result in larger and hotter
 2           fires. Solutions can include restoring natural fire regimes and removing non-native species to
 3           decrease people and ecosystems’ vulnerability to the exacerbated fire risk climate change is bringing
 4           through higher temperatures and in some places changing rainfall patterns.
 6   Nature-based Solutions, including protecting and restoring mangroves, forests and peatlands, also play an
 7   important part in reducing greenhouse gas emissions and taking carbon dioxide out of the atmosphere. They
 8   can also help people in a wide range of other ways, including through providing food, materials and
 9   providing opportunities for recreation. There is increasing evidence that spending time in natural
10   surroundings is good for physical and mental health.
12   It is important that the right adaptation actions are carried out in the right place and that local communities
13   play an active part in making decisions about their local environment if Nature-based Solutions are to be
14   effective. When they are not part of the process, conflicts can emerge and benefits can be lost.
16   Whilst Nature-based Solutions help us to adapt to climate change and reduce the amount of greenhouse
17   gases in the atmosphere, it is important to note that there are limits to what they can do. To provide a safe
18   environment for both people and nature, it will be essential to radically reduce greenhouse gas emissions,
19   especially from fossil-fuel burning in the near future.

     Do Not Cite, Quote or Distribute                      2-141                                       Total pages: 237
     FINAL DRAFT                                      Chapter 2                IPCC WGII Sixth Assessment Report

 2   Figure FAQ2.5.1: Different Nature-based Solutions strategies.
 4   [END FAQ2.5 HERE]
 7   2.6.8   Feasibility of Adaptation Options
 9   IPCC (2018a) defined feasibility as “the degree to which climate goals and response options are considered
10   possible and/or desirable” (IPCC 2018) and set out an approach to assessing feasibility of pathways to limit
11   global temperature rise to 1.5 °C. (Singh et al., 2020) have developed this approach for adaptation,
12   recognising 6 different dimensions of feasibility: economic, technological, institutional, socio-cultural,
13   environmental / ecological and geophysical (Table 2.9). Feasibility is considered more fully in other chapters
14   of this report, including Cross-Chapter Box FEASIB in Chapter 18. Adaptation for biodiversity conservation

     Do Not Cite, Quote or Distribute                       2-142                                  Total pages: 237
     FINAL DRAFT                                         Chapter 2                     IPCC WGII Sixth Assessment Report

 1   and EbA encompasses a large range of approaches and techniques (Sections 2.6.2, 2.6.3) and will vary in
 2   different contexts globally, as illustrated by the range of case studies (Section 2.6.5). It is important to take
 3   account of specific regional and local circumstances as well as the type of adaptation action that envisaged
 4   before making a feasibility assessment. It is also important to note that what is a feasible adaptation response
 5   may change with the level of warming experienced – some techniques will be become less effective at higher
 6   levels of warming. With global temperature rises of less than 2°C, in many cases it will be realistic to build
 7   resilience and maintain species and ecosystems in situ, but at higher levels of warming, this will become
 8   increasingly difficult and managing inevitable change, including the consequences of loss and damage will
 9   be important (Prober et al., 2019). Similarly to be effective at higher levels of warming may require the
10   adaptation of EbA approaches themselves (Calliari et al., 2019; Martín et al., 2021; Ossola and Lin, 2021).
11   We have therefore not attempted a global scale assessment of the feasibility of adaptation options, but rather
12   present some key cross cutting considerations in assessing feasibility for adaptation for and through
13   ecosystems.
15   Many of the necessary techniques for climate change adaptation for biodiversity and EbA have been
16   demonstrated and shown to provide a wide range of additional benefits. This does however depend on
17   deploying the right techniques in the right place (Box 2.2) and the engagement of local communities (see
18   Section 2.6.6). There is also a challenge where there is high demand for land for other purposes, especially
19   for agriculture and urban developments. Table 2.8 summarise the main feasibility considerations, drawing on
20   previous sections. An assessment of constraints on EbA by Nalau et al. (2018) addressed similar issues.
23   Table 2.8: Considerations in assessing feasibility of ecosystem restoration for climate change adaptation, following
24   Singh et al. (2020)
           Feasibility                    Feasibility indicators                Factors relevant to ecosystem restoration
      Economic              Micro-economic viability                           Costs are highly variable, depending on
                            Macro-economic viability                           techniques and whether land purchase is
                                                                               required. Costs will depend on local rates for
                            Socio-economic vulnerability reduction             labour and materials.
                            potential                                          Economic benefits to local communities
                            Employment & productivity enhancement              where employment is created and where loss
                            potential                                          from extreme events are avoided. (Section
                                                                               2.6.4; De Groot et al., 2013)
     Technological           Technical resource availability                   Techniques are available for restoration of
                             Risks mitigation potential (stranded Assets,      most ecosystems (Sections 2.6.2; 2.6.3)
                             unforeseen Impacts)                               although it can be very difficult to achieve in
                                                                               some circumstances and take a long time, e.g.
                                                                               the restoration of peat swamp forests (Section
                                                                               Successful implementation may also require
                                                                               skills which are in short supply and training
                                                                               may be required.
     Institutional           Political acceptability                           This will vary according to local factors. It
                             Legal, Regulatory feasibility                     should however be noted that EbA and
                                                                               adaptation for conservation has been
                             Institutional capacity & Administrative           implemented in wide range of different
                             feasibility                                       countries, including ones (see case studies in
                             Transparency & accountability potential           Section 2.6.5). In many cases EbA can meet
                                                                               multiple policy objectives but fall between
                                                                               different decision makers responsibilities.
     Socio-cultural          Social co-benefits (Health, education)            Multiple benefits to local communities are
                             Socio-Cultural acceptability                      possible but full engagement and/or
                                                                               leadership of affected members of
                             Social & Regional Inclusiveness                   communities has been shown to be critical.
                             Benefits for gender equity                        Local Knowledge and Indigenous Knowledge
                             Intergenerational equity                          can provide important insights. (Section

                             Ecological capacity

     Do Not Cite, Quote or Distribute                          2-143                                         Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

     Environmental/         Adaptive capacity/potential                   It is important to assess the benefits for
     ecological                                                           ecosystems in relation to other potential
                                                                          options. In particular for some EbA
                                                                          approaches, it may be possible to achieve a
                                                                          range of different outcomes for biodiversity.
     Geophysical            Physical Feasibility                          Appropriate measures need to be designed to
                            Land use change enhancement potential         take account of local geophysical conditions,
                                                                          for example catchment characteristics which
                            Hazard risk reduction potential               define where some habitats can occur. This is
                                                                          also critical for ensuring the effectiveness of
                                                                          EbA in reducing natural hazards.
 3   A key element of economic feasibility is the cost of adaptation options. Costs of adaptation vary greatly
 4   depending on the actions taken, the location, the methods used, the need for ongoing maintenance and
 5   whether land purchase is necessary. At its simplest adaptation may be a matter of taking account of actual or
 6   potential climate change impacts in the course of conservation planning and have little or no additional cost.
 7   For example, if a species of conservation concern colonises or starts to use a new area as a result of climate
 8   change (for example, migrant waterfowl shifting the locations where they overwinter (Pavón-Jordán et al.,
 9   2020b), protection or habitat management may be re-directed there. At the other extreme large scale
10   restoration can incur significant costs, for example between 1993 and 2015, the EU‐LIFE nature programme
11   invested 167.6M € in 80 projects, which aim to restore over 913 km2 of peatland habitats in Western
12   European countries (Andersen et al., 2017). This is equivalent to less than 2% of the remaining peatland
13   area, much of which has been affected to at least some extent by human pressures and restoring the total
14   areas will cost considerably more. De Groot et al. (2013) analysed 94 restoration projects globally and found
15   costs varied by several orders of magnitude but terrestrial and freshwater ecosystems mostly in the range of
16   100–10,000 USD per ha. They did however estimate that the majority of these projects provided net benefits
17   and should be considered as high yield investments. Some methods can however be much cheaper than
18   others even in the same type of ecosystems in the same Country: estimated cost of restoring forest cover in
19   Brazil varied between a mean of 49 USD using natural regeneration compared to a mean of 2041 USD per
20   hectare using planting (Brancalion et al., 2019). In assessing costs it is also important to take account of the
21   benefits delivered by different options, both in economic terms and other wider benefits.
23   The ‘technological’ dimension of feasibility in the context of ecosystems can be regarded as the range of
24   techniques available and the capacity to implement them. As described in Sections 2.6.2 and 2.6.3, above, a
25   wide range of techniques have been developed and are starting to be implemented. There is good evidence to
26   support adaptation for biodiversity and EbA in general terms and in many cases adaptation draws on
27   techniques for habitat creation and restoration which have been develop to mee other objectives. However,
28   feasibility needs to be assessed alongside likely effectiveness: a feasible but ineffective scheme is of no
29   value and the evaluation of success for specific interventions remains poorly developed (Morecroft et al.,
30   2019). It is often therefore important to proceed with the use of pilot studies and good monitoring and
31   evaluation of outcomes to build confidence before wider deployment of approaches. A linked technical area
32   is the availability of specialist skills and knowledge to implement adaptation which can vary considerably
33   according to the type of adaptation measure.
35   Institutional dimensions are dealt with more fully in other chapters, but in the specific context of the natural
36   environment, it is notable that EbA is relevant to a wide range of organisations and policy objectives, in
37   addition to environmental departments, NGOs and agencies, which conservation has traditionally been
38   delivered by. Upscaling implementation is likely to be dependent on this wider range of interests. There can
39   however be problems in that appropriate geographies for decision making on ecosystems (such as a
40   catchment) may not directly map onto governance arrangements
42   Socio-cultural factors are important in adaptation of the natural environment, in reviewing constraints on
43   EbA Nalau et al. (2018) found that risk perceptions and cultural preferences for particular types of
44   management approaches were frequently identified in studies.
46   In the IPCC feasibility assessment framework, one integral dimension is ‘Environmental / ecological’. In this
47   respect adaptation by and for ecosystems should perform well and this may be a reason to prefer EbA to

     Do Not Cite, Quote or Distribute                      2-144                                         Total pages: 237
     FINAL DRAFT                                    Chapter 2                  IPCC WGII Sixth Assessment Report

 1   other approaches when there is an alternative. It should however be noted that sometimes apparently
 2   environmentally positive approaches such as forest creation can be done in ways which are damaging
 3   (Section 2.6.7 and Box 2.2) and impacts need to be critically assessed for local circumstances.
 5   Geophysical dimensions are important for ecosystems as they have typically shaped which ecosystems can
 6   occur where and feasibility will depend on implementing adaptation options in places where they are
 7   appropriate. Paleoecological studies can help inform potential options (Wingard et al., 2017)
10   [START BOX 2.2 HERE]
12   Box 2.2: Risks of Maladaptive Mitigation
14   To hold global temperature rise to well below 2°C and pursue efforts to limit it to 1.5°C as required by the
15   Paris Agreement requires major changes in land use and management. There are many opportunities for
16   Nature-based Solutions, which can provide climate change mitigation and adaptation in ways, which protect
17   and restore biodiversity and provide a wide range of benefits to people (Cross Chapter Box NATURAL this
18   Chapter). There are also new technologies and approaches to develop the bioeconomy in ways, which
19   provide many benefits (Cross-Working Group Box BIOECONOMY in Chapter 5). Nevertheless, renewable
20   energy is a large and essential element of the climate change mitigation and there are adverse impacts on
21   biodiversity associated with some renewable energy, including wind and solar technologies (Rehbein et al.,
22   2020) However, of the most serious emerging conflicts are between land-based approaches to mitigation
23   and the protection of biodiversity, particularly as a result of afforestation strategies and potentially large
24   areas devoted to bioenergy, including bioenergy with carbon capture and storage (BECCS). It is important to
25   recognise the impacts of climate change mitigation at the same time as assessing the direct impacts of
26   climate change and ensure that adaptation and mitigation are joined up.
28   BECCS is an integral part of all widely accepted pathways to holding global temperature rise to 1.5°C
29   (IPCC, 2018b). This requires large areas of land which can conflict with the need to produce food and
30   protect biodiversity (Smith et al., 2018). One study that examined the combined impacts of climate change
31   and land use change for bioenergy and found severe impacts on species were likely if bioenergy were a
32   major component of climate change mitigation strategies (Hof et al., 2018). A study on the potential impacts
33   of bioenergy production and climate change on European birds found that land conversion for biodiversity to
34   meet a 2°C target would have greater impacts on species range loss than a global temperature increase of
35   4°C, if bioenergy were the only mitigation option (Meller et al., 2015). To avoid the worst impacts of
36   BECCS, it will need to be carefully targeted, according to context and local conditions (and other mitigation
37   strategies prioritised so its use can be minimised IPCC, 2019, Special Report on Land; Ohashi, 2019,
38   Biodiversity can benefit).
40   Reforestation of formerly forested areas can bring multiple benefits, but planting trees in places where they
41   do not naturally grow can have serious environmental impacts, including potentially exacerbating the effects
42   of climate change. Savannas, are at amongst those that are at risk from afforestation programmes. Savannas
43   are grass dominated, high diversity ecosystems with endemic species adapted to high light environments,
44   herbivory and fire (Staver et al., 2011; Murphy et al., 2016). Interactions between climate change, elevated
45   CO2 and the disruption of natural disturbance regimes have led to widespread woody plant encroachment
46   (Stevens et al., 2016) causing a fundamental shift in ecosystem structure and function with loss of grass,
47   reduced fire frequency (Archibald et al., 2009) and streamflow (Honda and Durigan, 2016). Afforestation
48   exacerbates this degradation (Bremer and Farley, 2010; Veldman et al., 2015; Abreu et al., 2017). Global-
49   scale analyses aimed at identifying degraded forest areas suitable for afforestation (Veldman et al., 2019)
50   cannot reliably separate grassy ecosystems with sparse tree cover from degraded forests and local knowledge
51   is essential to ensure tree planting is targeted where it can most benefit and avoid harm. Figure Box 2.2.1
52   indicates where these issues are most likely to arise.

     Do Not Cite, Quote or Distribute                    2-145                                      Total pages: 237
     FINAL DRAFT                                       Chapter 2                     IPCC WGII Sixth Assessment Report

 2   Figure Box 2.2.1: Regions where savannas at potential risk from afforestation. Based on (Veldman et al., 2015)
 5   A similar issue can occur in naturally treeless peatlands which can be afforested if they are drained, but this
 6   leads to the loss of distinctive peatland species and communities as well as high greenhouse gas emissions
 7   (Wilson et al., 2014). Mitigation benefits of growing timber are reduced or become negative in these
 8   conditions by CO2 emissions from the oxidation of the drained peat - they can be a net sources rather than a
 9   sinks (Simola et al., 2012; Crump, 2017; Goldstein et al., 2020).
11   [END BOX 2.2 HERE]
14   [START FAQ2.6 HERE]
16   FAQ2.6: Can tree planting tackle climate change?
18   Restoring and preventing further loss of native forests is essential for combating climate change. Planting
19   trees in historically unforested areas (grasslands, shrublands, savannas, some peatlands) can reduce
20   biodiversity and increase the risks of damage from climate change. It is therefore essential to target tree
21   planting to the appropriate locations and use appropriate species. Restoring and protecting forests reduces
22   human vulnerability to climate change, reduces air pollution, stores carbon and builds natural systems
23   resilience.
25   Like all living plants, trees remove carbon dioxide from the atmosphere through the process of
26   photosynthesis. In trees, this carbon uptake is relatively long-term, since much of it is stored in the trees’
27   woody stems and roots. Therefore, tree planting can be a valuable contribution to reducing climate change.
28   Besides capturing carbon, planting trees can reduce some negative impacts of climate change by providing
29   shade and cooling. It can also help prevent erosion and reduce flood risk by slowing water flow. Restoring
30   forest in degraded areas supports biodiversity and can provide benefits to people, ranging from timber to
31   food and recreation.
33   There are some areas where replacing lost trees is useful. These include forest that has been recently cut
34   down, and where reforestation is usually practical. However, it is very important to identify correctly areas
35   of forest that are degraded or have definitely been deforested. Reforesting places, especially where existing
36   native forest patches occur, brings benefits both in sucking up carbon from the atmosphere and helping us
37   adapt to climate change. Plantations of a non-native species, although offering economic benefits, do not
38   usually provide the same range of positive impacts and generally have lower biodiversity and carbon uptake
39   and storage.

     Do Not Cite, Quote or Distribute                        2-146                                        Total pages: 237
     FINAL DRAFT                                      Chapter 2                   IPCC WGII Sixth Assessment Report

 1   Reforestation options include the natural regeneration of the forest, assisted restoration, enrichment planting,
 2   native tree plantations, commercial plantations and directed tree planting can occur in agroforestry systems
 3   and urban areas. Reforestation with native species, usually contributes to a wide range of sustainability
 4   goals, including biodiversity recovery, improved water filtration and groundwater recharge. It can reduce the
 5   risks of soil erosion and flood risk. In cities, planting trees can support climate change adaptation by
 6   reducing the heat of the area and can promote a wide range of social benefits such as providing shade and
 7   benefiting outdoor recreation. Urban trees can also lower energy costs by reducing demand for conventional
 8   sources of cooling like air-conditioning, especially during peak-demand periods. It is therefore important to
 9   recognise that there are a wide range of different planting and forest management strategies. The choice will
10   depend on the objectives and the location.
12   Not everywhere is suitable for tree planting. It is particularly problematic in native non-forested ecosystems.
13   These natural ecosystems are not deforested and degraded but are instead naturally occurring non-forested
14   ecosystems. These areas vary from being open grasslands to densely wooded savannas and shrublands. Here
15   restoring the natural ecosystems instead of afforesting them will better contribute to increasing carbon
16   storage and increasing the areas resilience. It is important to remember, just because a tree can grow
17   somewhere, it does not mean that it should. These systems are very important in their own right, storing
18   carbon in soils, supporting a rich biodiversity and providing people with important ecosystem services such
19   as grazing. Planting trees in these areas destroys the ecosystem and threatens the biodiversity, which is
20   adapted to these environments. They can also impact on ecosystem services such as forage for livestock, on
21   which many people rely.
23   Many of these open areas also occur in low rainfall areas. Planting trees there uses a lot of water, can cause
24   reductions in stream flow and groundwater. Many of these locations also burn regularly and planting trees
25   threatens both the establishing trees but can also increase the intensity of the fires from that of a grass-fuelled
26   fire to that of a woody-fuelled fire. Swapping grassy ecosystems for forests may contribute to warming, as
27   forests absorb more incoming radiation (warmth) than grasslands. Aside from the negative impacts to
28   adaptation, it is also questionable just how much carbon can be sequestered in these landscapes asplanting
29   trees in grassy ecosystems can reduce carbon gains. Furthermore, a high belowground carbon store prevents
30   carbon loss to fire in these fire-prone environments.
32   Another example is with peatlands. Peats store an incredible amount of carbon within them and are therefore
33   important in maintaining and restoring to reduce atmospheric carbon. However, the restoration actions
34   depend on what type of peatland it is and where it is located. Many temperate and boreal peatlands are
35   naturally treeless. Here planting trees is often only possible following drainage, but draining and planting
36   (especially with non-native species) destroys native biodiversity and releases greenhouse gases. Yet many
37   peatlands, especially in the tropics, are naturally forested and restoring these peatlands requires re-wetting
38   and restoring natural tree cover (see Figure FAQ2.2.1) which will increase carbon storage.
40   There are actions we can do instead of planting trees in non-forested ecosystems, and these include:
41      • Address the causes of deforestation, forest degradation and widespread ecosystem loss;
42      • Reduce carbon emissions from fossil fuels;
43      • Focus on ecosystem restoration over tree planting. For example, in restoring tropical grassy
44           ecosystems, we can look at actions that cut down trees, enhance grass regrowth, and restore natural
45           fire regimes. We have then a much better chance of both enhancing carbon capture and reducing
46           some of the harmful effects of climate change.
48   In between the two extremes of where planting trees is highly suitable and areas where it is not, it is
49   important to remember that the context matters and decisions to (re)forest should look beyond simply the act
50   of planting trees. We can consider what the ecological, social and economic goals are of tree planting. It
51   is then important to verify the local context and then decide what restoration action will be most effective.
52   It is also very important to conserve forests before worrying about reforesting.

     Do Not Cite, Quote or Distribute                      2-147                                        Total pages: 237
     FINAL DRAFT                                       Chapter 2                     IPCC WGII Sixth Assessment Report

 2   Figure FAQ2.6.1: Some places are more appropriate for tree planting than others and caution needs to be applied when
 3   planting in different biomes with some biomes being more suitable than others. This figure highlights some basic biome
 4   specific guidelines when planting in natural and semi-natural vegetation.
 6   [END FAQ2.6 HERE]
 9   2.7     Reducing Scientific Uncertainties to Inform Policy and Management Decisions
11   Research since the IPCC Fifth Assessment Report (Settele et al., 2014) has increased understanding of
12   climate change impacts and vulnerability in ecosystems. Evidence gaps remain and geographic coverage of
13   research is uneven. This section assesses gaps in ecosystem science where research is necessary for
14   environmental policies and natural resource management, including under the UN Framework Convention
15   on Climate Change and the Convention on Biological Diversity.
17   2.7.1    Observed impacts
19   Detection and attribution efforts have yet to give robust assessments of the roles of climate change in
20   wildfires, tree mortality and human infectious diseases. Only one fire impact – the increase of the area
21   burned by wildfire in western North America in the period 1984-2017 (Section and just a few
22   cases of tree mortality (Section have been formally attributed to anthropogenic climate change.
23   Global changes in soil and freshwater ecosystem carbon over time remain uncertainties in global carbon
24   stocks and changes (Section; due to physical inabilities to conduct repeat monitoring and the lack of
25   remote sensing to scale up point measurements, no global methods can yet produce repeating spatial
26   estimates of soil carbon stock changes.

     Do Not Cite, Quote or Distribute                        2-148                                        Total pages: 237
     FINAL DRAFT                                     Chapter 2                   IPCC WGII Sixth Assessment Report

 1   Despite the growing understanding of the importance of ecosystem services this assessment found limited
 2   research on observed impacts of climate change on ecosystem services for 14 of 18 ecosystem services
 3   (Table 2.1)
 5   2.7.2   Projected risks
 7   A challenge for future projections that continues from previous IPCC reports is accurately characterising and
 8   quantifying interactions among climate change and other factors causing ecological change, including
 9   deforestation, agricultural expansion, urbanisation, and air and water pollution. Interactions can be
10   particularly complex for invasive species, pests and pathogens, and human infectious diseases. Modelling of
11   risks at the species level requires comprehensive databases of physiological, life-history, and functional traits
12   relevant to ecosystem resilience to climate change. Taxa that particularly lack this basis for model
13   projections include fungi and bacteria. For numerous plant and animal species, research on genotypic and
14   phenotypic diversity as a source of ecosystem resilience would inform projections of risk.
16   Soil plays a vital role in ecosystem function, is the habitat of a large number of species and is a large carbon
17   store which is currently a major source of greenhouse gas emissions, it is therefore a priority for climate
18   change research (Hashimoto et al., 2015). Major uncertainties remain in our understanding of soil functions.
19   Earth System Models (ESMs) predict soil respiration to increase with rising temperature (Friedlingstein et
20   al., 2014). However, there is evidence of acclimation post-increase (Carey et al., 2016) as the opposite
21   response of decrease in respiration with warming (Li et al., 2013; Reynolds et al., 2015). Long-term, large
22   scale field observations combined with a better conceptual understanding of factors governing soil process
23   responses to climate change is needed. A better understanding of plant water relations is also necessary,
24   including the response of plant transpiration to increased CO2, climate warming and changes in soil moisture
25   and groundwater elevation.
27   2.7.3   Adaptation and Climate Resilient Development
29   There are significant evidence gaps in developing adaptation, both for biodiversity conservation and EbA. In
30   particular, whilst many adaptation measures have been proposed and implementation is starting, there are
31   very few evaluations of success in the scientific literature (Morecroft et al., 2019; Prober et al., 2019). As
32   detailed in Section 2.6.2, there is a strong literature on conceptual approaches to climate change adaptation
33   for biodiversity, but very little empirical testing of which approaches actual work best. Going forward it is
34   important put in place good monitoring and evaluation of adaptation strategies. For EbA, there are good
35   examples of measuring changes in response to new adaptation measures, but these remain relatively rare
36   globally.
38   Human factors which promote or hinder adaptation are important as well as the technical issues. There are
39   few studies incorporating climate change and ecosystem services in integrated decision making, and even
40   fewer aimed to identify solutions robust to uncertainty (Runting et al., 2017).
41   Over the last decades, losses from natural disasters including those from events related to extreme weather
42   have strongly increased (Mechler and Bouwer, 2015). There is a need for better assessment of global
43   adaptation costs, funding and investment (Micale et al., 2018). Potential synergies between international
44   finance for disaster risk management and adaptation have not yet been fully realised. Research has almost
45   exclusively focused on normalising losses for changes in exposure, yet not for vulnerability, a major gap
46   given the dynamic nature of vulnerability (Mechler and Bouwer, 2015).

     Do Not Cite, Quote or Distribute                     2-149                                       Total pages: 237