FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Chapter 3: Oceans and Coastal Ecosystems and their Services 3 4 Coordinating Lead Authors: Sarah Cooley (USA) and David Schoeman (Australia) 5 6 Lead Authors: Laurent Bopp (France), Philip Boyd (Australia/United Kingdom), Simon Donner (Canada), 7 Shin-Ichi Ito (Japan), Wolfgang Kiessling (Germany), Paulina Martinetto (Argentina), Elena Ojea (Spain), 8 Marie-Fanny Racault (United Kingdom /France), Björn Rost (Germany), Mette Skern-Mauritzen (Norway), 9 Dawit Yemane Ghebrehiwet (South Africa/Eritrea) 10 11 Contributing Authors: Johann D. Bell (Australia), Julia Blanchard (Australia), Jessica Bolin (Australia), 12 William W. L. Cheung (Canada), Andrés Cisneros-Montemayor (Canada/Mexico), Sam Dupont 13 (Sweden/Belgium), Stephanie Dutkiewicz (USA), Thomas Frölicher (Switzerland), Juan Diego Gaitán- 14 Espitia (Hong Kong, Special Administrative Region, China/Colombia), Jorge García Molinos (Japan/Spain), 15 Helen Gurney-Smith (Canada), Stephanie Henson (United Kingdom), Manuel Hidalgo (Spain), Elisabeth 16 Holland (Fiji), Robert Kopp (USA), Rebecca Kordas (United Kingdom/USA), Lester Kwiatkowski 17 (France/United Kingdom), Nadine Le Bris (France), Salvador E. Lluch-Cota (Mexico), Cheryl Logan 18 (USA), Felix Mark (Germany), Yunus Mgaya (Tanzania), Coleen Moloney (South Africa), Norma Patricia 19 Muñoz Sevilla (Mexico), Gregoire Randin (Fiji/France/Switzerland), Nussaibah B. Raja 20 (Germany/Mauritius), Anusha Rajkaran (South Africa), Anthony Richardson (Australia), Stephanie Roe 21 (Philippines/USA), Raquel Ruiz Diaz (Spain), Diana Salili (Vanuatu), Jean-Baptiste Sallée (France), Kylie 22 Scales (Australia/United Kingdom), Michelle Scobie (Trinidad and Tobago), Craig T. Simmons (Australia), 23 Olivier Torres (France), Andrew Yool (United Kingdom) 24 25 Review Editors: Karim Hilmi (Morocco) and Lisa Levin (USA) 26 27 Chapter Scientist: Jessica Bolin (Australia) 28 29 Date of Draft: 1 October 2021 30 31 Notes: TSU Compiled Version 32 33 34 Table of Contents 35 36 Executive Summary..........................................................................................................................................3 37 3.1 Point of Departure ....................................................................................................................................8 38 FAQ3.1: How do we know which changes to marine ecosystems are specifically caused by climate 39 change? ....................................................................................................................................................10 40 3.2 Observed Trends and Projections of Climatic Impact-Drivers in the Global Ocean ......................16 41 3.2.1 Introduction...................................................................................................................................16 42 3.2.2 Physical Changes..........................................................................................................................18 43 3.2.3 Chemical Changes ........................................................................................................................22 44 3.2.4 Global Synthesis on Multiple Climatic Impact Drivers ................................................................25 45 3.3 Linking Biological Responses to Climatic-Impact Drivers.................................................................28 46 3.3.1 Introduction...................................................................................................................................28 47 3.3.2 Responses to Single Drivers..........................................................................................................29 48 3.3.3 Responses to Multiple Drivers ......................................................................................................32 49 3.3.4 Acclimation and Evolutionary Adaptation....................................................................................36 50 3.3.5 Ecological Response to Multiple Drivers .....................................................................................38 51 Box 3.1: Challenges for Multiple-Driver Research in Ecology and Evolution .........................................39 52 3.4 Observed and Projected Impacts of Climate Change on Marine Systems .......................................41 53 3.4.1 Introduction...................................................................................................................................41 54 3.4.2 Coastal Ecosystems and Seas .......................................................................................................42 55 3.4.3 Oceanic Systems and Cross Cutting Changes ..............................................................................68 56 Box 3.2: Marine Birds and Mammals...........................................................................................................82 57 Box 3.3: Deep Sea Ecosystems .......................................................................................................................96 Do Not Cite, Quote or Distribute 3-1 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 3.4.4 Reversibility and Impacts of Temporary Overshoot of 1.5°C or 2°C Warming ...........................98 2 3.5 Vulnerability, Resilience and Adaptive Capacity in Marine Social-Ecological Systems, including 3 Impacts to Ecosystem Services ..............................................................................................................99 4 3.5.1 Introduction...................................................................................................................................99 5 3.5.2 Biodiversity .................................................................................................................................104 6 3.5.3 Food Provision............................................................................................................................105 7 3.5.4 Other Provisioning Services .......................................................................................................107 8 3.5.5 Supporting and Regulating Services ...........................................................................................108 9 Box 3.4: Blue Carbon Ecosystems...............................................................................................................111 10 3.5.6 Cultural Services.........................................................................................................................114 11 3.6 Planned Adaptation and Governance to Achieve the Sustainable Development Goals (SDGs) ...116 12 3.6.1 Point of Departure ......................................................................................................................116 13 3.6.2 Adaptation Solutions ...................................................................................................................118 14 3.6.3 Implementation and Effectiveness of Adaptation and Mitigation Measures ..............................124 15 Cross-Chapter Box SLR: Sea Level Rise ...................................................................................................125 16 3.6.4 Contribution to the Sustainable Development Goals and Other Relevant Policy Frameworks.139 17 3.6.5 Emerging Best Practices for Ocean and Coastal Climate Adaptation.......................................143 18 FAQ3.2: Are we approaching so-called tipping points in the ocean and what can we do about it?.....144 19 FAQ3.3: How are marine heatwaves affecting marine life and human communities?..........................146 20 FAQ3.4: Which industries and jobs are most vulnerable to the impacts of climate change in the 21 oceans? ...................................................................................................................................................148 22 FAQ3.5: How can nature-based solutions, including marine protected areas, help us to adapt to 23 climate driven changes in the oceans? ................................................................................................150 24 References......................................................................................................................................................152 25 26 Do Not Cite, Quote or Distribute 3-2 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Executive Summary 2 3 Ocean and coastal ecosystems support life on Earth and many aspects of human well-being. Covering two- 4 thirds of the planet, the ocean hosts vast biodiversity and modulates the global climate system by regulating 5 cycles of heat, water, and elements including carbon. Marine systems are central to many cultures, and they 6 also provide food, minerals, energy and employment to people. Since previous assessments1, new laboratory 7 studies, field observations and process studies, a wider range of model simulations, Indigenous Knowledge, 8 and local knowledge provide increasing evidence on the impacts of climate change on ocean and coastal 9 systems, how human communities are experiencing these impacts, and the potential solutions for ecological 10 and human adaptation. 11 12 Observations: vulnerabilities and impacts 13 14 Anthropogenic climate change has exposed ocean and coastal ecosystems to conditions that are 15 unprecedented over millennia (high confidence2), and this has greatly impacted life in the ocean and 16 along its coasts (very high confidence). Fundamental changes in the physical and chemical characteristics 17 of the ocean acting individually and together are changing the timing of seasonal activities (very high 18 confidence), distribution (very high confidence), and abundance (very high confidence) of oceanic and 19 coastal organisms, from microbes to mammals and from individuals to ecosystems, in every region. 20 Evidence of these changes is apparent from multi-decadal observations, laboratory studies and mesocosms, 21 as well as meta-analyses of published data. Geographic range shifts of marine species generally follow the 22 pace and direction of climate warming (high confidence): surface warming since the 1950s has shifted 23 marine taxa and communities poleward at an average (mean ± very likely range) of 59.2 ± 15.5 km per 24 decade (high confidence), with substantial variation in responses among taxa and regions. Seasonal events 25 occur 4.3 ± 1.8 to 7.5 ± 1.5 days earlier per decade among planktonic organisms (very high confidence) and 26 on average 3 ± 2.1 days earlier per decade for fish (very high confidence). Warming, acidification and 27 deoxygenation are altering ecological communities by increasing the spread of physiologically sub-optimal 28 conditions for many marine fish and invertebrates (medium confidence). These and other responses have 29 subsequently driven habitat loss (very high confidence), population declines (high confidence), increased 30 risks of species extirpations and extinctions (medium confidence) and rearrangement of marine food webs 31 (medium to high confidence, depending on ecosystem). {3.2, 3.3, 3.3.2, 3.3.3, 3.3.3.2, 3.4.2.1, 3.4.2.3­ 32 3.4.2.8, 3.4.2.10, 3.4.3.1, 3.4.3.2, 3.4.3.3, Box 3.2} 33 34 Marine heatwaves lasting weeks to several months are exposing species and ecosystems to 35 environmental conditions beyond their tolerance and acclimation limits (very high confidence). WGI 36 AR6 concluded that marine heatwaves are more frequent (high confidence), more intense and longer 37 (medium confidence) since the 1980s, and since at least 2006, very likely3 attributable to anthropogenic 38 climate change. Open-ocean, coastal and shelf-sea ecosystems, including coral reefs, rocky shores, kelp 39 forests, seagrasses, mangroves, the Arctic Ocean and semi-enclosed seas, have recently undergone mass 40 mortalities from marine heatwaves (very high confidence). Marine heatwaves, including well-documented 41 events along the west coast of North America (2013­2016) and east coast of Australia (2015­2016, 2016­ 42 2017 and 2020), drive abrupt shifts in community composition that may persist for years (very high 43 confidence), with associated biodiversity loss (very high confidence), collapse of regional fisheries and 1 Previous IPCC assessments include the IPCC Fifth Assessment Report (AR5) (IPCC, 2013; IPCC, 2014c; IPCC, 2014b; IPCC, 2014d), the Special Report on Global Warming of 1.5°C (SR1.5) (IPCC, 2018), the Special Report on Ocean and Cryosphere in a Changing Climate (SROCC) (IPCC, 2019b) and the IPCC Sixth Assessment Report Working Group I (WGI AR6). 2 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. 3 In this Report, the following terms have been used to indicate the assessed likelihood of an outcome or a result: virtually certain 99­100% probability, very likely 90­100%, likely 66­100%, about as likely as not 33­ 66%, unlikely 0­33%, very unlikely 0­10%, exceptionally unlikely 0­1%. Additional terms (extremely likely: 95­ 100%, more likely than not >50­100%, and extremely unlikely 0­5%) may also be used when appropriate. Assessed likelihood is typeset in italics, e.g., very likely. Do Not Cite, Quote or Distribute 3-3 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 aquaculture (high confidence) and reduced capacity of habitat-forming species to protect shorelines (high 2 confidence). {WGI AR6 Chapter 9, 3.2.2.1, 3.4.2.1­3.4.2.5, 3.4.2.7, 3.4.2.10, 3.4.2.3, 3.4.3.3.3, 3.5.3} 3 4 At local to regional scales, climate change worsens the impacts on marine life of non-climate 5 anthropogenic drivers, such as habitat degradation, marine pollution, overfishing and overharvesting, 6 nutrient enrichment, and introduction of non-indigenous species (very high confidence). Although 7 impacts of multiple climate and non-climate drivers can be beneficial or neutral to marine life, most are 8 detrimental (high confidence). Warming exacerbates coastal eutrophication and associated hypoxia, causing 9 'dead zones' (very high confidence), which drive severe impacts on coastal and shelf-sea ecosystems (very 10 high confidence), including mass mortalities, habitat reduction and fisheries disruptions (medium 11 confidence). Overfishing exacerbates effects of multiple climate-impact drivers on predators at the top of the 12 marine food chain (medium confidence). Urbanization and associated changes in freshwater and sediment 13 dynamics increase the vulnerability of coastal ecosystems like sandy beaches, saltmarshes and mangrove 14 forests to sea-level rise and changes in wave energy (very high confidence). Although these non-climate 15 drivers confound attribution of impacts to climate change, adaptive, inclusive, and evidence-based 16 management reduces the cumulative pressure on ocean and coastal ecosystems, which will decrease their 17 vulnerability to climate change (high confidence). {3.3, 3.3.3, 3.4.2.4­3.4.2.8, 3.4.3.4, 3.5.3, 3.6.2, Cross- 18 Chapter Box SLR in Chapter 3} 19 20 Climate-driven impacts on ocean and coastal environments have caused measurable changes in 21 specific industries, economic losses, emotional harm, and altered cultural and recreational activities 22 around the world (high confidence). Climate-driven movement of fish stocks is causing commercial, 23 small-scale, artisanal, and recreational fishing activities to shift poleward and diversify harvests (high 24 confidence). Climate change is increasing the geographic spread and risk of marine-borne pathogens like 25 Vibrio sp. (very high confidence), which endanger human health and decrease provisioning and cultural 26 ecosystem services (high confidence). Interacting climatic impact-drivers and non-climate drivers are 27 enhancing movement and bioaccumulation of toxins and contaminants into marine food webs (medium 28 evidence, high agreement), and increasing salinity of coastal waters, aquifers, and soils (very high 29 confidence), which endangers human health (very high confidence). Combined climatic impact-drivers and 30 non-climate drivers also expose densely populated coastal zones to flooding (high confidence) and decrease 31 physical protection of people, property, and culturally important sites (very high confidence). {3.4.2.10, 32 3.5.3, 3.5.5, 3.5.5.3, 3.5.6, Cross-Chapter Box SLR in Chapter 3} 33 34 Projections: vulnerabilities, risks, and impacts 35 36 Ocean conditions are projected to continue diverging from a pre-industrial state (virtually certain), 37 with the magnitude of warming, acidification, deoxygenation, sea-level rise and other climatic impact- 38 drivers depending on the emission scenario (very high confidence), and to increase risk of regional 39 extirpations and global extinctions of marine species (medium confidence). Marine species richness near 40 the equator and in the Arctic is projected to continue declining, even with less than 2°C warming by the end 41 of the century (medium confidence). In the deep ocean, all global warming levels will cause faster 42 movements of temperature niches by 2100 than those that have driven extensive reorganisation of marine 43 biodiversity at the ocean surface over the past 50 years (medium confidence). "At warming levels beyond 44 2°C by 2100, risks of extirpation, extinction and ecosystem collapse escalate rapidly (high confidence)." 45 Paleorecords indicate that at extreme global warming levels (>5.2°C), mass extinction of marine species may 46 occur (medium confidence). {Box 3.2, 3.2.2.1, 3.4.2.5, 3.4.2.10, 3.4.3.3, Cross-Chapter Box PALEO in 47 Chapter 1} 48 49 Climate impacts on ocean and coastal ecosystems will be exacerbated by increases in intensity, 50 reoccurrence and duration of marine heatwaves (high confidence), in some cases, leading to species 51 extirpation, habitat collapse or surpassing ecological tipping points (very high confidence). Some 52 habitat-forming coastal ecosystems including many coral reefs, kelp forests and seagrass meadows, will 53 undergo irreversible phase shifts due to marine heatwaves with global warming levels >1.5°C and are at high 54 risk this century even in <1.5°C scenarios that include periods of temperature overshoot beyond 1.5°C (high 55 confidence). Under SSP1-2.6, coral reefs are at risk of widespread decline, loss of structural integrity and 56 transitioning to net erosion by mid-century due to increasing intensity and frequency of marine heatwaves 57 (very high confidence). Due to these impacts, the rate of sea-level rise is very likely to exceed that of reef Do Not Cite, Quote or Distribute 3-4 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 growth by 2050, absent adaptation. Other coastal ecosystems, including kelp forests, mangroves and 2 seagrasses, are vulnerable to phase shifts towards alternate states as marine heatwaves intensify (high 3 confidence). Loss of kelp forests are expected to be greatest at the low-latitude warm edge of species' ranges 4 (high confidence). {3.4.2.1, 3.4.2.3, 3.4.2.5, 3.4.4} 5 6 Escalating impacts of climate change on marine life will further alter biomass of marine animals 7 (medium confidence), the timing of seasonal ecological events (medium confidence) and the geographic 8 ranges of coastal and ocean taxa (medium confidence), disrupting life cycles (medium confidence), food 9 webs (medium confidence) and ecological connectivity throughout the water column (medium 10 confidence). Multiple lines of evidence suggest that climate-change responses are very likely to amplify up 11 marine food webs over large regions of the ocean. Modest projected declines in global phytoplankton 12 biomass translate into larger declines of total animal biomass (by 2080­2099 relative to 1995­2014) ranging 13 from (mean ± very likely range) ­5.7% ± 4.1% to ­15.5% ± 8.5% under SSP1-2.6 and SSP5-8.5, respectively 14 (medium confidence). Projected declines in upper-ocean nutrient concentrations, likely associated with 15 increases in stratification, will reduce carbon export flux to the mesopelagic and deep-sea ecosystems 16 (medium confidence). This will lead to a decline in the biomass of abyssal meio- and macrofauna (by 2081­ 17 2100 relative to 1995­2014) by ­9.8% and ­13.0% under SSP1-2.6 and SSP5-8.5, respectively (limited 18 evidence). By 2100, 18.8% ± 19.0% to 38.9% ± 9.4% of the ocean will very likely undergo a change of more 19 than 20 days (advances and delays) in the start of the phytoplankton growth period under SSP1-2.6 and 20 SSP5-8.5, respectively (low confidence). This altered timing increases the risk of temporal mismatches 21 between plankton blooms and fish spawning seasons (medium to high confidence) and increases the risk of 22 fish recruitment failure for species with restricted spawning locations, especially in mid-to-high latitudes of 23 the northern hemisphere (low confidence). Projected range shifts among marine species (medium confidence) 24 suggest extirpations and strongly decreasing tropical biodiversity. At higher latitudes, range expansions will 25 drive increased homogenisation of biodiversity. The projected loss of biodiversity ultimately threatens 26 marine ecosystem resilience (medium to high confidence), with subsequent effects on service provisioning 27 (medium to high confidence). {3.2.2.3, 3.4.2.10, 3.4.3.1­3.4.3.5, 3.5, WGI AR6 Section 2.3.4.2.3} 28 29 Risks from sea-level rise for coastal ecosystems and people are very likely to increase tenfold well 30 before 2100 without adaptation and mitigation action as agreed by Parties to the Paris Agreement 31 (very high confidence). Sea-level rise under emission scenarios that do not limit warming to 1.5°C will 32 increase the risk of coastal erosion and submergence of coastal land (high confidence), loss of coastal habitat 33 and ecosystems (high confidence) and worsen salinisation of groundwater (high confidence), compromising 34 coastal ecosystems and livelihoods (high confidence). Under SSP1-2.6, most coral reefs (very high 35 confidence), mangroves (likely, medium confidence) and saltmarshes (likely, medium confidence) will be 36 unable to keep up with sea-level rise by 2050, with ecological impacts escalating rapidly beyond 2050, 37 especially for scenarios coupling high emissions with aggressive coastal development (very high 38 confidence). Resultant decreases in natural shoreline protection will place increasing numbers of people at 39 risk (very high confidence). The ability to adapt to current coastal impacts, cope with future coastal risks, and 40 prevent further acceleration of sea-level rise beyond 2050 depends on immediate implementation of 41 mitigation and adaptation actions (very high confidence). {3.4.2.1, 3.4.2.4, 3.4.2.5, 3.4.2.6, 3.5.5.3, Cross- 42 Chapter Box SLR in Chapter 3} 43 44 Climate change will alter many ecosystem services provided by marine systems (high confidence), but 45 impacts to human communities will depend on people's overall vulnerability, which is strongly 46 influenced by local context and development pathways (very high confidence). Catch composition and 47 diversity of regional fisheries will change (high confidence), and fishers who are able to move, diversify, and 48 leverage technology to sustain harvests decrease their own vulnerability (medium confidence). Management 49 that eliminates overfishing facilitates successful future adaptation of fisheries to climate change (very high 50 confidence). Marine-dependent communities, including Indigenous Peoples and local peoples, will be at 51 increased risk of losing cultural heritage and traditional seafood-sourced nutrition (medium confidence). 52 Without adaptation, seafood-dependent people face increased risk of exposure to toxins, pathogens, and 53 contaminants (high confidence), and coastal communities face increasing risk from salinisation of 54 groundwater and soil (high confidence). Early-warning systems and public education about environmental 55 change, developed and implemented within the local and cultural context, can decrease those risks (high 56 confidence). Coastal development and management informed by sea-level rise projections will reduce the 57 number of people and amount of property at risk (high confidence), but historical coastal development and Do Not Cite, Quote or Distribute 3-5 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 policies impede change (high confidence). Current financial flows are globally uneven and overall 2 insufficient to meet the projected costs of climate impacts on coastal and marine socio-ecological systems 3 (very high confidence). Inclusive governance that accommodates geographically shifting marine life; 4 financially supports needed human transformations; provides effective public education; and incorporates 5 scientific evidence, Indigenous knowledge, and local knowledge to manage resources sustainably shows 6 greatest promise for decreasing human vulnerability to all of these projected changes in ocean and coastal 7 ecosystem services (very high confidence). {3.5.3, 3.5.5, 3.5.6, 3.6.3, Box 3.4, Cross-Chapter Box ILLNESS 8 in Chapter 2, Cross-Chapter Box SLR in Chapter 3} 9 10 Solutions, trade-offs, residual risk, decisions and governance 11 12 Humans are already adapting to climate-driven changes in marine systems, and while further 13 adaptations are required even under low-emission scenarios (high confidence), transformative 14 adaptation will be essential under high-emission scenarios (high confidence). Low-emission scenarios 15 permit a wider array of feasible, effective and low-risk nature-based adaptation options (e.g., restoration, 16 revegetation, conservation, early-warning systems for extreme events, and public education) (high 17 confidence). Under high-emission scenarios, adaptation options (e.g., hard infrastructure for coastal 18 protection, assisted migration or evolution, livelihood diversification, migration and relocation of people) are 19 more uncertain and require transformative governance changes (high confidence). Transformative climate 20 adaptation will reinvent institutions to overcome obstacles arising from historical precedents, reducing 21 current barriers to climate adaptation in cultural, financial, and governance sectors (high confidence). 22 Without transformation, global inequities will likely increase between regions (high confidence) and conflicts 23 between jurisdictions may emerge and escalate. {3.5, 3.5.2, 3.5.5.3, 3.6, 3.6.2.1, 3.6.3.1, 3.6.3.2, 3.6.3.3, 24 3.6.4.1, 3.6.4.2, 3.6.5, Cross-Chapter Box SLR in Chapter 3, Cross-Chapter Box ILLNESS in Chapter 2} 25 26 Available adaptation options are unable to offset climate-change impacts on marine ecosystems and 27 the services they provide (high confidence). Adaptation solutions implemented at appropriate scales, 28 when combined with ambitious and urgent mitigation measures, can meaningfully reduce impacts 29 (high confidence). Increasing evidence from implemented adaptations indicates that multi-level governance, 30 early-warning systems for climate-associated marine hazards, seasonal and dynamic forecasts, habitat 31 restoration, ecosystem-based management, climate-adaptive management, and sustainable harvesting tend to 32 be both feasible and effective (high confidence). Marine protected areas, as currently implemented, do not 33 confer resilience against warming and heatwaves (medium confidence) and are not expected to provide 34 substantial protection against climate impacts past 2050 (high confidence). However, marine protected areas 35 can contribute substantially to adaptation and mitigation if they are designed to address climate change, 36 strategically implemented, and well governed (high confidence). Habitat restoration limits climate-change 37 related loss of ecosystem services, including biodiversity, coastal protection, recreational use and tourism 38 (medium confidence), provides mitigation benefits on local to regional scales (e.g., via carbon-storing `blue 39 carbon' ecosystems) (high confidence), and may safeguard fish stock production in a warmer climate 40 (limited evidence). Ambitious and swift global mitigation offers more adaptation options and pathways to 41 sustain ecosystems and their services (high confidence). {3.4.2, 3.4.3.3, 3.5, 3.5.2, 3.5.3, 3.5.5.4, 3.5.5.5, 42 3.6.2.1, 3.6.2.2, 3.6.2.3, 3.6.3.1, 3.6.3.2, 3.6.3.3, 3.6.5, Figure 3.24, Figure 3.25} 43 44 Nature-based solutions for adaptation of ocean and coastal ecosystems can achieve multiple benefits 45 when well-designed and implemented (high confidence), but their effectiveness declines without 46 ambitious and urgent mitigation (high confidence). Nature-based solutions such as ecosystem-based 47 management, climate-smart conservation approaches (i.e., climate-adaptive fisheries and conservation) and 48 coastal habitat restoration can be cost-effective and generate social, economic and cultural co-benefits, while 49 contributing to the conservation of marine biodiversity and reducing cumulative anthropogenic drivers (high 50 confidence). The effectiveness of nature-based solutions declines with warming; conservation and restoration 51 will alone be insufficient to protect coral reefs beyond 2030 (high confidence) and to protect mangroves 52 beyond the 2040s (high confidence). The multi-dimensionality of climate change impacts and their 53 interactions with other anthropogenic stressors calls for integrated approaches that identify trade-offs and 54 synergies across sectors and scales in space and time to build resilience of ocean and coastal ecosystems and 55 the services they deliver (high confidence). {3.4.2, 3.5.2, 3.5.3, 3.5.5.3, 3.5.5.4, 3.5.5.5, 3.6.2.2, 3.6.3.2, 56 3.6.5, Figure 3.25, Table SM3.6} 57 Do Not Cite, Quote or Distribute 3-6 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Ocean-focused adaptations, especially those that employ nature-based solutions, address existing 2 inequalities, and incorporate just and inclusive decision-making and implementation processes, 3 support the UN Sustainable Development Goals (SDGs) (high confidence). There are predominantly 4 positive synergies between adaptation options for Life Below Water (SDG14), Climate Action (SDG13), and 5 social, economic and governance SDGs (SDG1­12, 16­17) (high confidence), but the ability of ocean 6 adaptation to contribute to the Sustainable Development Goals is constrained by the degree of mitigation 7 action (high confidence). Furthermore, existing inequalities and entrenched practices limit effective and just 8 responses to climate change in coastal communities (high confidence). Momentum is growing towards 9 transformative international and regional governance that will support comprehensive, equitable ocean and 10 coastal adaptation while also achieving SDG14 (robust evidence), without compromising achievement of 11 other SDGs. {3.6.4.0, 3.6.4.2, 3.6.4.3, Figure 3.26}. Do Not Cite, Quote or Distribute 3-7 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 3.1 Point of Departure 2 3 The ocean contains approximately 97% of Earth's water within a system of interconnected basins that cover 4 71% of its surface. Coastal systems mostly extend seaward from the high-water mark, or just beyond, to the 5 edge of the continental shelf and include shores of soft sediments, rocky shores and reefs, embayments, 6 estuaries, deltas and shelf systems. Oceanic systems comprise waters beyond the shelf edge, from ~200 m to 7 nearly 11,000 m deep (Stewart and Jamieson, 2019), with an average depth of approximately 3700 m. The 8 epipelagic zone, or upper 200 m of the ocean, is illuminated by sufficient sunlight to sustain photosynthesis 9 that supports the rich marine food web. Below the epipelagic zone lies the barely-lit mesopelagic zone (200­ 10 1000 m), the perpetually dark bathypelagic zone (depth >1000 m) and the deep seafloor (benthic ecosystems 11 at depths > 200 m), which spans rocky and sedimentary habitats on seamounts, mid-ocean ridges and 12 canyons, abyssal plains and sedimented margins. Semi-enclosed seas (SES) include both coastal and oceanic 13 systems. 14 15 The ocean sustains life on Earth by providing essential resources and modulating planetary flows of energy 16 and materials. Together, harvests from the ocean and inland waters provide more than 20% of dietary animal 17 protein for more than 3.3 billion people worldwide and livelihoods for about 60 million people (FAO, 18 2020b). The global ocean is centrally involved in sequestering anthropogenic atmospheric CO2 and recycling 19 many elements, and it regulates the global climate system by redistributing heat and water (WGI AR6 20 Chapter 9, Fox-Kemper et al., 2021). The ocean also provides a wealth of aesthetic and cultural resources 21 (Barbier et al., 2011), contains vast biodiversity (Appeltans et al., 2012), supports more animal biomass than 22 on land (Bar-On et al., 2018), and produces at least half the world's photosynthetic oxygen (Field et al., 23 1998). Ecosystem services (Annex II: Glossary) delivered by ocean and coastal ecosystems support 24 humanity by protecting coastlines, providing nutrition and economic opportunities (Figure 3.1, Selig et al., 25 2019), and providing many intangible benefits. Even though ecosystem services and biodiversity underpin 26 human well-being and support climate mitigation and adaptation (Pörtner et al., 2021b), there are also ethical 27 arguments for preserving biodiversity and ecosystem functions regardless of the beneficiary (e.g., Taylor et 28 al., 2020). This chapter assesses the impact of climate change on the full spectrum of ocean and coastal 29 ecosystems, on their services and on related human activities, and it assesses marine-related opportunities 30 within both ecological and social systems to adapt to climate change. 31 32 33 34 Figure 3.1: Estimated relative human dependence on marine ecosystems for coastal protection, nutrition, fisheries 35 economic benefits and overall. Each bar represents an index value that semi-quantitatively integrates the magnitude, 36 vulnerability to loss and substitutability of the benefit. Indices synthesize information on people's consumption of 37 marine protein and nutritional status, gross domestic product, fishing revenues, unemployment, education, governance 38 and coastal characteristics. Overall dependence is the mean of the three index values after standardization from 0­1 39 (Details are found in Table 1 and supplementary material of Selig et al. (2019)). This index does not include the Do Not Cite, Quote or Distribute 3-8 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 economic benefits from tourism or other ocean industries, and data limitations prevented including artisanal or 2 recreational fisheries or the protective impact of saltmarshes (Selig et al., 2019). Values for reference regions 3 established in the WGI AR6 Atlas (Gutiérrez et al., 2021) were computed as area-weighted means from original 4 country-level data. 5 6 7 Previous IPCC Assessment Reports (IPCC, 2014c; IPCC, 2014b; IPCC, 2018; IPCC, 2019b) have expressed 8 growing confidence in the detection of climate-change impacts in the ocean and their attribution to 9 anthropogenic greenhouse gas emissions. Heat and CO2 taken up by the ocean (high to very high confidence) 10 (IPCC, 2021b) directly affect marine systems, and the resultant "climatic impact-drivers" (CIDs, e.g., ocean 11 temperature and heatwaves, sea level, dissolved oxygen levels, acidification, Annex II: Glossary, WGI 12 Figure SPM.9, IPCC, 2021b) also influence ocean and coastal systems (Section 3.2, Cross-Chapter Box SLR 13 in Chapter 3, Cross-Chapter Box EXTREMES in Chapter 2, Figure SM3.1), from individual biophysical 14 processes to dependent human activities. Several marine outcomes of CIDs are themselves drivers of 15 ecological change (e.g., climate velocities, stratification, sea ice changes). This chapter updates and extends 16 the assessment of SROCC (IPCC, 2019b) and WGI AR6 by assessing the ecosystem effects of the CIDs in 17 WGI AR6 Figure SPM.9 (IPCC, 2021b) and their biologically relevant marine outcomes (detailed in Section 18 3.2), which are referred to collectively hereafter as "climate-impact drivers4". 4 We henceforth use the term "climate-impact drivers" in reference to all drivers of ecological change that are related directly to climate change (CIDs, IPCC, 2021a) as well as those that emerge in response to CIDs. Do Not Cite, Quote or Distribute 3-9 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Detrimental human impacts on ocean and coastal ecosystems are not only caused by climate. Other 3 anthropogenic activities are increasingly affecting the physical, chemical, and biological conditions of the 4 ocean (Doney, 2010; Halpern et al., 2019), and these "non-climate drivers5" also alter marine ecosystems and 5 their services. Fishing and other extractive activities are major non-climate drivers in many ocean and coastal 6 systems (Steneck and Pauly, 2019). Many activities, such as coastal development, shoreline hardening and 7 habitat destruction, physically alter marine spaces (Suchley and Alvarez-Filip, 2018; Ducrotoy et al., 2019; 8 Leo et al., 2019; Newton et al., 2020; Raw et al., 2020). Other human activities decrease water quality by 9 overloading coastal water with terrestrial nutrients (eutrophication) and by releasing runoff containing 10 chemical, biological and physical pollutants, toxins, and pathogens (Jambeck et al., 2015; Luek et al., 2017; 11 Breitburg et al., 2018; Froelich and Daines, 2020) Some human activities disturb marine organisms by 12 generating excess noise and light (Davies et al., 2014; Duarte et al., 2021), while others decrease natural light 13 penetration into the ocean (Wollschläger et al., 2021). Several anthropogenic activities alter processes that 14 span the land-sea interface by changing coastal hydrology or causing coastal subsidence (Michael et al., 15 2017; Phlips et al., 2020; Bagheri-Gavkosh et al., 2021). Atmospheric pollutants can harm marine systems or 16 unbalance natural marine processes (Doney et al., 2007; Hagens et al., 2014; Lamborg et al., 2014; Ito et al., 17 2016). Organisms frequently experience non-climate drivers simultaneously with climate-impact drivers 18 (Section 3.4), and feedbacks may exist between climate-impact drivers and non-climate drivers that enhance 19 the effects of climate change (Rocha et al., 2015; Ortiz et al., 2018; Wolff et al., 2018; Cabral et al., 2019; 20 Bowler et al., 2020; Gissi et al., 2021). SROCC assessed with high confidence that reduction of pollution and 21 other stressors, along with protection, restoration, and precautionary management, supports ocean and 22 coastal ecosystems and their services (IPCC, 2019b). This chapter examines the combined influence of 23 climate-impact drivers and primary non-climate drivers on many ecosystems assessed. 24 25 Detecting changes and attributing them to specific drivers have been especially difficult in ocean and coastal 26 ecosystems because drivers, responses and scales (temporal, spatial, organizational) often overlap and 27 interact (IPCC, 2014c; IPCC, 2014b; Abram et al., 2019; Gissi et al., 2021). In addition, some marine 28 systems have short, heterogeneous, or geographically biased observational records, which exacerbate the 29 interpretation challenge (Beaulieu et al., 2013; Christian, 2014; Huggel et al., 2016; Benway et al., 2019). It 30 is even more challenging to detect and attribute climate impacts on marine-dependent human systems, where 31 culture, governance and society also strongly influence observed outcomes. To assess climate-driven change 32 in natural and social systems robustly, IPCC reports rely on multiple lines of evidence, and the available 33 types of evidence differ depending on the system under study (Section 1.3.2.1, Cross-Working Group Box 34 ATTRIB). Lines of evidence used for ocean and coastal ecosystems for this and previous assessments 35 include observed phenomena, laboratory and field experiments, long-term monitoring, empirical and 36 dynamical model analyses, Indigenous knowledge (IK) and local knowledge (LK), and paleorecords (IPCC, 37 2014c; IPCC, 2014b; IPCC, 2019b). The growing body of climate research for ocean and coastal ecosystems 38 and their services increasingly provides multiple independent lines of evidence whose conclusions support 39 each other, raising the overall confidence in detection and attribution of impacts over time (Section 1.3.2.1, 40 Cross-Working Group Box ATTRIB in Chapter 3). 41 42 43 [START FAQ3.1 HERE] 44 45 FAQ3.1: How do we know which changes to marine ecosystems are specifically caused by climate 46 change? 47 48 To attribute changes in marine ecosystems to human-induced climate change, scientists use paleorecords 49 (reconstructing the links between climate, evolutionary and ecological changes in the geological past), 50 contemporary observations (assessing current climate and ecological responses in the field and through 51 experiments) and models. We refer to these as multiple lines of evidence, meaning that the evidence comes 52 from diverse approaches, as described below. 53 5 We henceforth use the term "non-climate drivers" in reference to drivers of ecological change that are not caused by climate change. Do Not Cite, Quote or Distribute 3-10 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Emissions of greenhouse gases like carbon dioxide from human activity cause ocean warming, acidification, 2 oxygen loss, and other physical and chemical changes that are affecting marine ecosystems around the 3 world. At the same time, natural climate variability and direct human impacts, such as overfishing and 4 pollution, also affect marine ecosystems locally, regionally and globally. These climate and non-climate 5 impact drivers counteract each other, add up or multiply to produce smaller or larger changes than expected 6 from individual drivers. Attribution of changes in marine ecosystems requires evaluating the often- 7 interacting roles of natural climate variability, non-climate drivers, and human-induced climate change. To 8 do this work, scientists use 9 - paleorecords: reconstructing the links between climate and evolutionary and ecological changes of the past 10 - contemporary observations: assessing current climate and ecological responses, 11 - manipulation experiments: measuring responses of organisms and ecosystems to different climate 12 conditions 13 - models: testing whether we understand how organisms and ecosystems are impacted by different stressors, 14 and quantifying the relative importance of different stressors. 15 16 Paleorecords can be used to trace the correlation between past changes in climate and marine life. 17 Paleoclimate is reconstructed from the chemical composition of shells and teeth or from sediments and ice 18 cores. Changes to sea life signalled by changing biodiversity, extinction or distributional shifts are 19 reconstructed from fossils. Using large datasets, we can infer the effects of climate change on sea life over 20 relatively long timescales ­ usually hundreds to millions of years. The advantage of paleorecords is that they 21 provide insights into how climate change affects life from organisms to ecosystems, without the 22 complicating influence of direct human impacts. A key drawback is that the paleo and modern worlds do not 23 have fully comparable paleoclimate regimes, dominant marine species, and rates of climate change. 24 Nevertheless, the paleorecord can be used to derive fundamental rules by which organisms, ecosystems, 25 environments and regions are typically most affected by climate change. For example, the paleorecord shows 26 that coral reefs repeatedly underwent declines during past warming events, supporting the inference that 27 corals may not be able to adapt to current climate warming. 28 29 Contemporary observations over recent decades allow scientists to relate the status of marine species and 30 ecosystems to changes in climate or other factors. For example, scientists compile large datasets to determine 31 whether species usually associated with warm water are appearing in traditionally cool-water areas that are 32 rapidly warming. A similar pattern observed in multiple regions and over several decades (i.e., longer than 33 timescales of natural variability) provides confidence that climate change is altering community structure. 34 This evidence is weighed against findings from other approaches, such as manipulation experiments, to 35 provide a robust picture of climate change impacts in the modern ocean. 36 37 In manipulation experiments, scientists expose organisms or communities of organisms to multiple stressors, 38 for example, elevated CO2, high temperature, or both, based on values drawn from future climate 39 projections. Such experiments will involve multiple treatments (i.e., different aquarium tanks) in which 40 organisms are exposed to different combinations of the stressors. This approach enables scientists to 41 understand the effects of individual stressors as well as their interactions to explore physiological thresholds 42 of marine organisms and communities. The scale of manipulation experiments can range from small tabletop 43 tanks to large installations or natural ocean experiments involving tens of thousands of litres of water. 44 45 Ecological effects of climate change are also explored within models developed from fundamental scientific 46 principles and observations. Using these numerical representations of marine ecosystems, scientists can 47 explore how different levels of climate change and non-climate stressors influence species and ecosystems at 48 scales not possible with experiments. Models are commonly used to simulate the ecological response to 49 climate change over recent decades and centuries. Convergence between the model results and the 50 observations suggests that our understanding of the key processes is sufficient to attribute the observed 51 ecological changes to climate change, and to use the models to project future ecological changes. Differences 52 between model results and observations indicate gaps in knowledge to be filled in order to better detect and 53 attribute the impacts of climate change on marine life. 54 55 Using peer-reviewed research spanning the full range of scientific approaches (paleorecords, observations, 56 experiments and models), we can assess the level of confidence in the impact of climate change on observed 57 modifications in marine ecosystems. We refer to this as multiple lines of evidence, meaning that the evidence Do Not Cite, Quote or Distribute 3-11 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 comes from the diverse approaches described above. This allows policy-makers and managers to address the 2 specific actions needed to reduce climate change and other impacts. 3 4 5 6 Figure FAQ3.1: Examples of well-known impacts of anthropogenic climate change and associated nature- 7 based adaptation. To attribute changes in marine ecosystems to anthropogenic climate change, scientists use 8 multiple lines of evidence including paleorecords, contemporary observations, manipulation experiments and 9 models. 10 11 [END FAQ3.1 HERE] 12 13 14 Natural adaptation to climate change in ocean and coastal systems includes an array of responses taking 15 place at scales from cells to ecosystems. Previous IPCC assessments have established that many marine 16 species "have shifted their geographic ranges, seasonal activities, migration patterns, abundances and species 17 interactions in response to climate change," (high confidence) (IPCC, 2014c; IPCC, 2014b), which has had 18 global impacts on species composition, abundance and biomass, and on ecosystem structure and function 19 (medium confidence) (IPCC, 2019b). Warming and acidification have affected coastal ecosystems in concert 20 with non-climate drivers (high confidence), which have affected habitat area, biodiversity, ecosystem 21 function and services (high confidence) (IPCC, 2019b). Confidence has grown in these assessments over 22 time as observational datasets have lengthened and other lines of evidence have corroborated observations. 23 AR5 and SROCC assessed how physiological sensitivity to climate-impact drivers is the underlying cause of Do Not Cite, Quote or Distribute 3-12 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 most marine organisms' vulnerability to climate (high confidence) (Pörtner et al., 2014; Bindoff et al., 2019). 2 Since those assessments, more evidence supports the empirical physiological models of tolerance and 3 plasticity (Sections 3.3.2, 3.3.4) and of interactions among multiple (climate and non-climate) drivers at 4 individual to ecosystem scales (Sections 3.3.3, 3.4.5). New experimental evidence about evolutionary 5 adaptation (Section 3.3.4) bolsters previous assessments that adaptation options to climate change are limited 6 for eukaryotic organisms. Tools such as ecosystem models can now constrain probable ecosystem states 7 (Sections 3.3.4­3.3.5 and 3.4). Observations have increased understanding of how extreme events affect 8 individuals, populations and ecosystems, helping refine understanding of both ecological tolerance to climate 9 impacts and ecological transformations (Section 3.4). 10 11 Human adaptation to climate impacts on ocean and coastal systems spans a variety of actions that change 12 human activity to maintain marine ecosystem services. After AR5 concluded that coastal adaptation could 13 reduce the effects of climate impacts on coastal human communities (high agreement, limited evidence) 14 (Wong et al., 2014), SROCC confirmed that mostly risk-reducing ocean and coastal adaptation responses 15 were underway (Bindoff et al., 2019). However, overlapping climate-impact and non-climate drivers 16 confound implementation and assessment of the success of marine adaptation, revealing the complexity of 17 attempting to maintain marine ecosystems and services through adaptation. SROCC assessed with high 18 confidence that while the benefits of many locally implemented adaptations exceed their disadvantages, 19 others are marginally effective and have large disadvantages, and overall, adaptation has a limited ability to 20 reduce the probable risks from climate change, being at best a temporary solution (Bindoff et al., 2019). 21 SROCC also concluded that a portfolio of many different types of adaptation actions, effective and inclusive 22 governance, and mitigation must be combined for successful adaptation (Bindoff et al., 2019). The portfolio 23 of adaptation measures has now been defined (Section 3.6.2), and individual and combined adaptation 24 solutions have been implemented in several marine sectors (Section 3.6.3). Delays in marine adaptation have 25 been partly attributed to the complexity of ocean governance (Section 3.6.4, Cross-Chapter Box 3 and Figure 26 CB3.1 in Abram et al., 2019) and to the low priority accorded the ocean in international development goals 27 (Nash et al., 2020), but in recent years the ocean is being increasingly incorporated in international climate 28 policy and multilateral environmental agreements (Section 3.6.4). 29 30 This chapter assesses the current understanding of climate-impact drivers, ecological vulnerability and 31 adaptability, risks to coastal and ocean ecosystems, and human vulnerability and adaptation to resulting 32 changes in ocean benefits, now and in the future (Figure 3.2). It starts by assessing the biologically relevant 33 outcomes of anthropogenic climate-impact drivers (Section 3.2). Next, it sets out the mechanisms that 34 determine the responses of ocean and coastal organisms to individual and combined drivers from the genetic 35 to the ecosystem level (Section 3.3). This supports a detailed assessment of the observed and projected 36 responses of coastal and ocean ecosystems to these hazards, placing them in context using the paleo-record 37 (Section 3.4). These observed and projected impacts are used to quantify consequent risks to delivery of 38 ecosystem services and the socioeconomic sectors that depend on them, with attention to the vulnerability, 39 resilience and adaptive capacity of social-ecological systems (Section 3.5). The chapter concludes by 40 assessing the state of adaptation and governance actions available to address these emerging threats while 41 also advancing human development (Section 3.6). Abbreviations used repeatedly in the chapter are defined 42 in Table 3.1. 43 44 Do Not Cite, Quote or Distribute 3-13 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure 3.2: WGII AR6 Chapter 3 concept map. This chapter assesses how climate changes both the properties (top of 3 wave, Sections 3.1­3.6) and the mechanisms (below wave, Sections 3.2­3.6) that influence the ocean and coastal 4 social-ecological system. The Sustainable Development Goals (top right) represent ideal outcomes and achievement of 5 equitable, healthy and sustainable ocean and coastal social-ecological systems. 6 7 8 Table 3.1: List of abbreviations frequently used in this chapter, with brief definitions for many of the abbreviations 9 used. Abbreviation Definition ABNJ Areas beyond national jurisdiction. The water column beyond the exclusive economic zone called the high seas and the seabed beyond the limits of the continental shelf established in conformity with United Nations Convention on the Law of the Sea. AMOC Atlantic meridional overturning circulation (WGI AR6 Glossary, IPCC, 2021a). AR5 The IPCC Fifth Assessment Report (IPCC, 2013; IPCC, 2014c; IPCC, 2014b; IPCC, 2014d). CBD Convention on Biological Diversity. An international legal instrument that has been ratified by 196 nations to conserve biological diversity, sustainably use its components and share its benefits fairly and equitably. CE Common era. CID Climatic Impact-Driver (WGI AR6 Glossary, IPCC, 2021a). CMIP5, CMIP6 The Coupled Model Intercomparison Project, Phase 5 or 6 (WGI AR6 Glossary, IPCC, 2021a). EbA Ecosystem-based adaptation. The use of ecosystem management activities to increase the resilience and reduce the vulnerability of people and ecosystems to climate change. EBUS Eastern boundary upwelling system (WGI AR6 Glossary, IPCC, 2021a). EBUE/EUS Eastern boundary upwelling systems/equatorial upwelling systems. EBUEs are marine ecosystems in EBUS. EUSs Do Not Cite, Quote or Distribute 3-14 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report EEZ ESM are located on the equator, mostly on the eastern side of major ocean basins, where trade winds drive upwelling. Fish-MIP Exclusive economic zone. The area from the coast to 200 GMSL/GMSLR nautical miles (370 km) off the coast, where a nation HAB exercises its sovereign rights and exclusive management ICZM authority. IK and LK MHW Earth-system model. A coupled atmosphere­ocean general MPA circulation model (AOGCM, WGI AR6 Glossary, IPCC, NbS 2021a) in which a representation of the carbon cycle is included, allowing for interactive calculation of NDC atmospheric CO2 or compatible emissions. NPP OECM The Fisheries and Marine Ecosystem Model OMZ Intercomparison Project. Fish-MIP is a component of the pCO2 Inter-Sectoral Impact Model Intercomparison pH Project (ISIMIP) that explores the long-term impacts of POC climate change on fisheries and marine ecosystems using scenarios from CMIP models. Do Not Cite, Quote or Distribute Global mean sea level/global mean sea-level rise (Sea- level change, WGI AR6 Glossary, IPCC, 2021a). Harmful algal bloom. A HAB is an algal bloom composed of phytoplankton known to naturally produce bio-toxins that are harmful to the resident population, as well as humans. Integrated coastal zone management. ICZM is a dynamic, multidisciplinary and iterative process to promote sustainable management of coastal zones (European Environmental Agency). Indigenous knowledge and Local knowledge (SROCC Glossary, IPCC, 2019a). Marine heatwaves (WGI AR6 Glossary, IPCC, 2021a). Marine protected area. MPA is an area-based management approach, commonly intended to conserve, preserve, or restore biodiversity and habitat, protect species, or manage resources (especially fisheries). Nature-based Solution. Actions to protect, sustainably manage and restore natural or modified ecosystems that address societal challenges effectively and adaptively, simultaneously providing human well-being and biodiversity benefits (IUCN, 2016). Nationally determined contribution by Parties to the Paris Agreement. Net primary production. The difference between how much CO2 vegetation takes in during photosynthesis (gross primary production) minus how much CO2 the plants release during respiration. Other effective area-based conservation measures. OECM is a conservation designation for areas that are achieving the effective in situ conservation of biodiversity outside of protected areas. Oxygen minimum zone (WGI AR6 Glossary, IPCC, 2021a). Partial pressure of carbon dioxide. For seawater, pCO2 is used to measure the amount of carbon dioxide dissolved in seawater. Potential of hydrogen (WGI AR6 Glossary, IPCC, 2021a). Particulate organic carbon. POC is a fraction of total organic carbon operationally defined as that which does 3-15 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report SDG not pass through a filter pore size that typically ranges in size from 0.053­2 mm. SES SIDS Sustainable Development Goals. The 17 global goals for SLR/RSLR/RSL development for all countries established by the United SR15 Nations through a participatory process and elaborated in the 2030 Agenda for Sustainable Development. Semi-enclosed sea. SES means a gulf, basin or sea surrounded by land and connected to another sea by a narrow outlet. Small Island Developing States (WGI AR6 Glossary, IPCC, 2021a). Sea-level rise/Relative rea-level rise/Relative Sea Level (Sea-level change, WGI AR6 Glossary, IPCC, 2021a). The IPCC Special Report on 1.5 Degrees C (IPCC, 2018). SROCC The IPCC Special Report on the Ocean and Cryosphere (IPCC, 2019b). SSP/RCP Shared socio-economic pathways/Representative concentration pathways (Pathways, IPCC, 2021a). SST Sea-surface temperature (WGI AR6 Glossary, IPCC, 2021a). aragonite Saturation state of seawater with respect to the calcium carbonate mineral aragonite, used as a proxy measurement for ocean acidification. 1 2 3 3.2 Observed Trends and Projections of Climatic Impact-Drivers in the Global Ocean 4 5 3.2.1 Introduction 6 7 Climate change exposes ocean and coastal ecosystems to changing environmental conditions, including 8 ocean warming, SLR, acidification, deoxygenation and other climatic-impact drivers, which have distinct 9 regional and temporal characteristics (Gruber, 2011; IPCC, 2018). This section aims to build on the WGI 10 AR6 assessment (Table 3.2) to provide an ecosystem-oriented framing of climatic impact-drivers. Updating 11 SROCC, projected trends assessed here are based on a new range of scenarios (Shared Socio-Economic 12 Pathways, SSPs), as used in the Coupled Model Intercomparison Project Phase 6 (CMIP6, Section 1.2.2). 13 14 15 Table 3.2: Overview of the main global ocean Climatic Impact-Drivers and their observed and projected trends from 16 WGI AR6, with corresponding confidence levels and links to WGI chapters, where these trends are assessed in detail. Climatic-Impact Observed trends over the WGI Section Projected trends over the 21st century WGI Drivers (Hazards) historical period Section Ocean Temperature Ocean Warming At the ocean surface, 2.3.3.1, 9.2.1 Ocean warming will continue over 9.2.1 (Fox- Kemper et temperature has on average (Fox-Kemper et the 21st century (virtually certain), al., 2021) increased by 0.88 [0.68­ al., 2021; Gulev and with the rate of global ocean 1.01] °C from 1850­1900 to et al., 2021) warming starting to be scenario- 2011­2020. dependent from about the mid-21st century (medium confidence). Marine Heatwaves MHW have become more Box 9.2 (Fox- MHW will become 4 [2­9, likely Box 9.2 (MHW) frequent (high confidence), Kemper et al., range] times more frequent in 2081­ (Fox- more intense, and longer 2021) 2100 compared to 1995­2014 under Kemper et (medium confidence) over SSP1-2.6, and 8 [3­15, likely range] al., 2021) the 20th century. times more frequent under SSP5-8.5. Climate Velocities Not assessed in WGI. Not assessed in WGI. Sea-Level Do Not Cite, Quote or Distribute 3-16 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report Global Mean Sea Since 1901, GMSL has risen 2.3.3, 9.6.1 There will be continued rise in GMSL 4.3.2.2, Level (GMSL) by 0.20 [0.15­0.25] m, and (Fox-Kemper et through the 21st century under all 9.6.3 (Fox- the rate of rise is al., 2021; Gulev assessed SSPs (virtually certain). Kemper et accelerating. et al., 2021) al., 2021; Lee et al., 2021) Extreme Sea Levels Relative sea-level rise is 9.6.4 (Fox- Rising mean relative sea level will 9.6.4 (Fox- driving a global increase in Kemper et al., continue to drive an increase in the Kemper et the frequency of extreme sea 2021) frequency of extreme sea levels (high al., 2021) levels (high confidence). confidence). Ocean circulation Ocean Stratification The upper ocean has 9.2.1.3 (Fox- Upper-ocean stratification will 9.2.1.3 continue to increase throughout the (Fox- become more stably Kemper et al., 21st century (virtually certain). Kemper et al., 2021) stratified since at least 1970 2021) (virtually certain). Eastern Boundary Only the California current 9.2.5 (Fox- Eastern boundary upwelling systems 9.2.5 (Fox- Upwelling Systems system has experienced Kemper et al., will change, with a dipole spatial Kemper et pattern within each system of al., 2021) some large-scale upwelling- 2021) reduction at low latitude and favourable wind enhancement at high latitude (high intensification since the confidence). 1980s (medium confidence). Atlantic Overturning For the 20th century, there is 2.3.3.4, 9.2.3 The AMOC will decline over the 21st 4.3.2.3, Circulation low confidence in (Fox-Kemper et century (high confidence, but low 9.2.3 (Fox- (AMOC) reconstructed and modelled al., 2021; Gulev confidence for quantitative Kemper et AMOC changes. et al., 2021) projections). al., 2021; Lee et al., 2021) Sea-Ice Arctic Sea-Ice Current Arctic sea-ice 2.3.2.1, 9.3.1 The Arctic will become practically 4.3.2.1, Changes coverage levels are the (Fox-Kemper et ice-free in September by the end of 9.3.1 (Fox- lowest since at least 1850 al., 2021; Gulev the 21st century under SSP2-4.5, Kemper et for both annual mean and et al., 2021) SSP3-7.0, and SSP5-8.5 (high al., 2021; late-summer values (high Lee et al., confidence). confidence) 2021) Antarctic Sea Ice There is no global 2.3.2.1, 9.3.2 There is low confidence in model 9.3.2 (Fox- Changes Kemper et significant trend in Antarctic (Fox-Kemper et simulations of future Antarctic sea al., 2021) sea-ice area from 1979 to al., 2021; Gulev ice. 2020 (high confidence). et al., 2021) Ocean Chemistry Changes in Salinity The large-scale, near-surface 2.3.3.2, 9.2.2.2 Fresh ocean regions will continue to 9.2.2.2 salinity contrasts have (Fox-Kemper et get fresher and salty ocean regions (Fox- intensified since at least al., 2021; Gulev will continue to get saltier in the 21st Kemper et 1950 (virtually certain). et al., 2021) century (medium confidence). al., 2021) Ocean Acidification Ocean surface pH has 2.3.3.5, 5.3.2.2 Ocean surface pH will continue to 4.3.2.5, declined globally over the 4.5.2.2, past four decades (virtually (Canadell et al., decrease through the 21st century, 5.3.4.1 certain). (Lee et al., 2021; Gulev et except for the lower-emission 2021) (Canadell al., 2021) scenarios SSP1-1.9 and SSP1-2.6, et al., 2021) (high confidence). Ocean Deoxygenation has occurred 2.3.3.6, 5.3.3.2 Subsurface oxygen content is 5.3.3.2 Deoxygenation in most open ocean regions (Canadell et al., projected to transition to historically (Canadell since the mid 20th (high 2021; Gulev et unprecedented condition with decline et al., confidence). al., 2021) over the 21st century (medium 2021) confidence). Changes in Nutrient Not assessed in WGI. Not assessed in WGI. Concentrations 1 Do Not Cite, Quote or Distribute 3-17 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 3.2.2 Physical Changes 3 4 3.2.2.1 Ocean Warming, Climate Velocities and Marine Heatwaves 5 6 Global mean SST has increased since the beginning of the 20th century by 0.88°C (very likely range: 0.68­ 7 1.01°C), and it is virtually certain that the global ocean has warmed since at least 1971 (WGI AR6 Section 8 9.2, Fox-Kemper et al., 2021). A key characteristic of ocean temperature change relevant for ecosystems is 9 climate velocity, a measure of the speed and direction at which isotherms move under climate change 10 (Burrows et al., 2011), which gives the rate at which species must migrate to maintain constant climate 11 conditions. It has been shown to be a useful and simple predictor of species distribution shifts in marine 12 ecosystems (Chen et al., 2011; Pinsky et al., 2013; Lenoir et al., 2020). Median climate velocity in the 13 surface ocean has been 21.7 km per decade since 1960, with higher values in the Arctic/sub-Arctic and 14 within 15° of the Equator (Figure 3.3, Burrows et al., 2011). While climate velocity has been slower in the 15 mesopelagic layer (200­1000 m) than in the epipelagic layer (0­200 m) over the last 50 years, it has been 16 shown to be faster in the bathypelagic (1000­4000 m) and abyssopelagic (>4000 m) layers (Figure 3.4, 17 Brito-Morales et al., 2020), suggesting that deep-ocean species could be as exposed to effects of warming as 18 species in the surface ocean (Brito-Morales et al., 2020). 19 20 21 22 Figure 3.3: Observed surface ocean warming, surface climate velocity and reconstructed changes in marine heatwaves 23 (MHWs) over the last 100 years. (a) Sea-surface temperature trend (°C per century) over 1925­2016 from Hadley 24 Centre Sea Ice and Sea Surface Temperature 1.1 (HadISST1.1), (b) surface climate velocity (km per decade) over Do Not Cite, Quote or Distribute 3-18 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 1925­2016 computed from HadISST1.1, and (c) change in total MHW days for the surface ocean from 1925­1954 to 2 1987­2016 based on monthly proxies, from Oliver et al. (2018). 3 4 5 Marine heatwaves (MHW) are periods of extreme seawater temperature relative to the long-term mean 6 seasonal cycle, that persist for days to months, and that may carry severe consequences for marine 7 ecosystems and their services (WGI AR6 Box 9.2, Hobday et al., 2016a; Smale et al., 2019; Fox-Kemper et 8 al., 2021). MHW have become more frequent over the 20th century (high confidence), approximately 9 doubling in frequency (high confidence) and becoming more intense and longer since the 1980s (medium 10 confidence) (WGI AR6 Box 9.2, Fox-Kemper et al., 2021). These trends in MHW are explained by an 11 increase in ocean mean temperatures (Oliver et al., 2018), and human influence has very likely contributed to 12 84­90% of them since at least 2006 (WGI AR6 Box 9.2, Fox-Kemper et al., 2021). The probability of 13 occurrence (as well as duration and intensity) of the largest and most impactful MHWs that have occurred in 14 the past 30 years has increased more than 20-fold due to anthropogenic climate change (Laufkötter et al., 15 2020). 16 17 18 19 Figure 3.4: Historical and projected climate velocity. Climate velocities (in km per decade) for the (a,d,g) historical 20 period (1965­2014), and for the last 50 years of the 21st century (2051­2100) under (b,e,h) SSP1-2.6 and (c,f,i) SSP5- 21 8.5. Shown are the epipelagic (0­200 m), mesopelagic (200­1000 m) and bathypelagic (1000­4000 m) domains. 22 Updated figure from Brito-Morales et al. (2020), with Coupled Model Intercomparison Project 6 models used in 23 Kwiatkowski et al. (2020). 24 25 26 Ocean warming will continue over the 21st century (virtually certain), and with the rate of global ocean 27 warming starting to be scenario-dependent from about the mid-21st century (medium confidence). At the 28 ocean surface, it is virtually certain that SST will continue to increase throughout the 21st century, with 29 increasing hazards to many marine ecosystems (WGI AR6 Box 9.2, Fox-Kemper et al., 2021). The future 30 global mean SST increase projected by CMIP6 models for the period 1995­2014 to 2081­2100 is 0.86°C 31 (very likely range: 0.43­1.47°C) under SSP1-2.6, 1.51 °C (1.02­2.19°C) under SSP2-4.5, 2.19°C (1.56­ 32 3.30°C) under SSP3-7.0, and 2.89°C (2.01­4.07°C) under SSP5-8.5 (WGI AR6 Section 9.2.1, Fox-Kemper 33 et al., 2021). Stronger surface warming occurs in parts of the tropics, in the North Pacific, and in the Arctic Do Not Cite, Quote or Distribute 3-19 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Ocean, where SST increases by >4°C in 2080­2099 under SSP5-8.5 (Kwiatkowski et al., 2020). The CMIP6 2 climate models also project ocean warming at the seafloor, with the magnitude of projected changes being 3 less than that of surface waters but having larger uncertainties (Kwiatkowski et al., 2020). The projected end- 4 of-the-century warming in CMIP6 as reported here is greater than assessed with Coupled Model 5 Intercomparison Project 5 (CMIP5) models in AR5 and in SROCC for similar radiative forcing scenarios 6 (Figure 3.5, Kwiatkowski et al., 2020), because of greater climate sensitivity in the CMIP6 model ensemble 7 than in CMIP5 (WGI AR6 Chapter 4, Forster et al., 2020; Lee et al., 2021). 8 9 MHWs will continue to increase in frequency, with a likely global increase of 2­9 times in 2081­2100 10 compared to 1995­2014 under SSP1-2.6, and 3­15 times under SSP5-8.5, with the largest increases in 11 tropical and Arctic oceans (WGI AR6 Box 9.2, Frölicher et al., 2018; Fox-Kemper et al., 2021). 12 13 3.2.2.2 Sea-Level Rise and Extreme Sea Levels 14 15 Global mean sea-level (GMSL, see also Cross-Chapter Box SLR in Chapter 3) has risen by about 0.20 m 16 since 1901 and continues to accelerate (WGI AR6 Section 2.3.3.3, Church and White, 2011; Jevrejeva et al., 17 2014; Hay et al., 2015; Kopp et al., 2016; Dangendorf et al., 2017; Cazenave et al., 2018; Kemp et al., 2018; 18 Ablain et al., 2019; Gulev et al., 2021). 19 20 Most coastal ecosystems (mangroves, sea grasses, saltmarshes, shallow coral reefs, rocky shores and sandy 21 beaches) are affected by changes in relative sea-level (RSL, the change in the mean sea level relative to the 22 land, Section 3.4.2). Regional rates of RSL rise differ from the global mean due to a range of factors, 23 including local subsidence driven by anthropogenic activities such as groundwater and hydrocarbon 24 extraction (WGI AR6 Box 9.1, Fox-Kemper et al., 2021). In many deltaic regions, anthropogenic subsidence 25 is currently the dominant driver of RSL rise (WGI AR6 Section 9.6.3.2, Tessler et al., 2018; Fox-Kemper et 26 al., 2021). RSL rise is driving a global increase in the frequency of extreme sea levels (high confidence) 27 (WGI AR6 Section 9.6.4.1, Fox-Kemper et al., 2021). 28 29 GMSL rise through the middle of the 21st century exhibits limited dependence on emissions scenario; 30 between 1995­2014 and 2050, GMSL is likely to rise by 0.15­0.23 m under SSP1-1.9 and 0.20­0.30 m 31 under SSP5-8.5 (WGI AR6 Section 9.6.3, Fox-Kemper et al., 2021). Beyond 2050, GMSL and RSL 32 projections are increasingly sensitive to the differences among emission scenarios. Considering only 33 processes in which there is at least medium confidence (thermal expansion, land water storage, land ice 34 surface mass balance, and some ice sheet dynamic processes), GMSL between 1995­2014 and 2100 is likely 35 to rise by 0.28­0.55 m under SSP1-1.9, 0.33­0.61 m under SSP1-2.6, 0.44­0.76 m under SSP2-4.5, 0.55­ 36 0.90 m under SSP3-7.0, and 0.63­1.02 m under SSP5-8.5 (Figure 3.5). Under high-emission scenarios, ice- 37 sheet processes in which there is low confidence and deep uncertainty might contribute more than one 38 additional metre to GMSL rise by 2100 (WGI AR6 Chapter 9, Fox-Kemper et al., 2021). 39 40 Rising mean RSL will continue to drive an increase in the frequency of extreme sea levels (high confidence). 41 The expected frequency of the current one-in-100-year extreme sea level is projected to increase by a median 42 of 20­30 times across tide-gauge sites by 2050, regardless of emission scenario (medium confidence). In 43 addition, extreme-sea-level frequency may be affected by changes in tropical cyclone climatology (low 44 confidence), wave climatology (low confidence), and tides (high confidence) associated with climate change 45 and sea-level change (WGI AR6 Section 9.6.4.2, Fox-Kemper et al., 2021). 46 47 Do Not Cite, Quote or Distribute 3-20 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure 3.5: Projected trends in climatic-impact drivers for ocean ecosystems. Panels (a,b,c,d) represent Coupled Model 3 Intercomparison Project 5 (CMIP5) Representative Concentration Pathway (RCP) and CMIP6 Shared Socioeconomic 4 Pathway (SSP) end-of-century changes in (a) global sea-level rise (SLR), (b) average surface pH, (c) subsurface (100­ 5 600 m) dissolved oxygen concentration and (d) euphotic-zone (0­100 m) nitrate (NO3) concentration against anomalies 6 in sea surface temperature. All anomalies are model-ensemble averages over 2080­2099 relative to the 1870­1899 7 baseline period (from Kwiatkowski et al., 2020), except for SLR, which shows model-ensemble median in 2100 relative 8 to 1901 (from AR6 WGI Chapter 9). Error bars represent very likely ranges, except for SLR where they represent likely 9 ranges. Very likely ranges for pH changes are too narrow to appear on the figure (see text). Panels (e,f,g,h) show regions 10 where end-of-century projected CMIP6 surface warming exceeds 2°C, where surface ocean pH decline exceeds 0.3, 11 where subsurface dissolved oxygen decline exceeds 30 mmol m-3 and where euphotic-zone (0­100 m) nitrate decline 12 exceeds 1 mmol m-3 in (e) SSP1-2.6, (f) SSP2-4.5, (g) SSP3-7.0 and (h) SSP5-8.5. All anomalies are 2080­2099 13 relative to the 1870­1899 baseline period (modified from Kwiatkowski et al., 2020). 14 15 16 3.2.2.3 Changes in Ocean Circulation, Stratification and Coastal Upwelling 17 18 Ocean circulation and its variations are key to the evolution of the physical, chemical and biological 19 properties of the ocean. Vertical mixing and upwelling are critical factors affecting the supply of nutrients to 20 the sunlit ocean and hence the magnitude of primary productivity. Ocean currents not only transport heat, 21 salt, carbon, and nutrients, but they also control the dispersion of many organisms and the connectivity 22 between distant populations. 23 24 Ocean stratification is an important factor controlling biogeochemical cycles and affecting marine 25 ecosystems. WGI AR6 Section 9.2.1.3 (Fox-Kemper et al., 2021) assessed that it is virtually certain that 26 stratification in the upper 200 m of the ocean has been increasing since 1970. Recent evidence has 27 strengthened estimates of the rate of change (Yamaguchi and Suga, 2019; Li et al., 2020a; Sallée et al., 28 2021), with an estimated increase of 1.0 ± 0.3% (very likely range) per decade over the period 1970­2018 29 (high confidence) (WGI AR6 Section 9.2.1.3, Fox-Kemper et al., 2021), higher than assessed in SROCC. It 30 is very likely that stratification in the upper few hundred metres of the ocean will increase substantially in the 31 21st century in all ocean basins, driven by intensified surface warming and near-surface freshening at high 32 latitudes (WGI AR6 Section 9.2.1.3, Capotondi et al., 2012; Fu et al., 2016; Bindoff et al., 2019; 33 Kwiatkowski et al., 2020; Fox-Kemper et al., 2021). 34 35 Contrasting changes among the major eastern boundary coastal upwelling systems (EBUS) were identified in 36 AR5 (Hoegh-Guldberg et al., 2014). While SROCC assessed with high confidence that three (Benguela, 37 Peru-Humboldt, California) out of the four major EBUS have experienced upwelling-favourable wind 38 intensification in the past 60 years (Sydeman et al., 2014; Bindoff et al., 2019), WGI AR6 revisited this Do Not Cite, Quote or Distribute 3-21 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 assessment based on evidence showing low agreement between studies that have investigated trends over 2 past decades (Varela et al., 2015). WGI AR6 assessed that only the California Current system has undergone 3 large-scale upwelling-favourable wind intensification since the 1980s (medium confidence) (WGI AR6 4 Section 9.2.1.5, García-Reyes and Largier, 2010; Seo et al., 2012; Fox-Kemper et al., 2021). 5 6 While no consistent pattern of contemporary changes in upwelling-favourable winds emerges from 7 observation-based studies, numerical and theoretical work projects that summertime winds near poleward 8 boundaries of upwelling zones will intensify, while winds near equatorward boundaries will weaken (high 9 confidence) (WGI AR6 Section 9.2.3.5, García-Reyes et al., 2015; Rykaczewski et al., 2015; Wang et al., 10 2015; Aguirre et al., 2019; Fox-Kemper et al., 2021). Nevertheless, projected future annual cumulative 11 upwelling wind changes at most locations and seasons remain within ±10­20% of present-day values 12 (medium confidence) (WGI AR6 Section 9.2.3.5, Fox-Kemper et al., 2021). 13 14 Continuous observation of the Atlantic meridional overturning circulation (AMOC) has improved the 15 understanding of its variability (Frajka-Williams et al., 2019), but there is low confidence in the 16 quantification of AMOC changes in the 20th century because of low agreement in quantitative reconstructed 17 and simulated trends (WGI AR6 Sections 2.3.3, 9.2.3.1, Fox-Kemper et al., 2021; Gulev et al., 2021). Direct 18 observational records since the mid-2000s remain too short to determine the relative contributions of internal 19 variability, natural forcing and anthropogenic forcing to AMOC change (high confidence) (WGI AR6 20 Sections 2.3.3, 9.2.3.1, Fox-Kemper et al., 2021; Gulev et al., 2021). Over the 21st century, AMOC will very 21 likely decline for all SSP scenarios, but will not involve an abrupt collapse before 2100 (WGI AR6 Sections 22 4.3.2, 9.2.3.1, Fox-Kemper et al., 2021; Lee et al., 2021). 23 24 3.2.2.4 Sea Ice Changes 25 26 Sea ice is a key driver of polar marine life, hosting unique ecosystems and affecting diverse marine 27 organisms and food webs through its impact on light penetration and supplies of nutrients and organic matter 28 (Arrigo, 2014). Since the late 1970s, Arctic sea-ice area has decreased for all months, with an estimated 29 decrease of two million km2 (or 25%) for summer sea-ice (averaged for August, September, October) in 30 2010­2019 as compared to 1979­1988 (WGI AR6 Section 9.3.1.1, Fox-Kemper et al., 2021). For Antarctic 31 sea-ice there is no significant global trend in satellite-observed sea-ice area from 1979 to 2020 in either 32 winter or summer, due to regionally opposing trends and large internal variability (WGI AR6 Section 33 9.3.2.1, Maksym, 2019; Fox-Kemper et al., 2021). 34 35 CMIP6 simulations project that the Arctic Ocean will likely become practically sea-ice free (area below 1 36 million km2) for the first time before 2050 and in the seasonal sea-ice minimum in each of the four emission 37 scenarios SSP1-1.9, SSP1-2.6, SSP2-4.5, and SSP5-8.5 (Figure 3.7, WGI AR6 Section 9.3.2.2, SIMIP 38 Community, 2020; Fox-Kemper et al., 2021). Antarctic sea-ice area is also projected to decrease during the 39 21st century, but due to mismatches between model simulations and observations, combined with a lack of 40 understanding of reasons for substantial inter-model spread, there is low confidence in model projections of 41 future Antarctic sea-ice changes, particularly at the regional level (WGI AR6 Section 9.3.2.2, Roach et al., 42 2020; Fox-Kemper et al., 2021). 43 44 3.2.3 Chemical Changes 45 46 3.2.3.1 Ocean Acidification 47 48 The ocean's uptake of anthropogenic carbon affects its chemistry in a process referred to as ocean 49 acidification, which increases the concentrations of aqueous CO2, bicarbonate and hydrogen ions, and 50 decreases pH, carbonate ion concentrations and calcium carbonate mineral saturation states (Doney et al., 51 2009). Ocean acidification affects a variety of biological processes with, for example, lower calcium 52 carbonate saturation states reducing net calcification rates for some shell-forming organisms and higher CO2 53 concentrations increasing photosynthesis for some phytoplankton and macroalgal species (Section 3.3.2). 54 55 Direct measurements of ocean acidity from ocean time series, as well as pH changes determined from other 56 shipboard studies, show consistent decreases in ocean surface pH over the past few decades (virtually Do Not Cite, Quote or Distribute 3-22 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 certain) (WGI AR6 Section 5.3.2.2, Takahashi et al., 2014; Bindoff et al., 2019; Sutton et al., 2019; Canadell 2 et al., 2021). 3 4 Since the 1980s, surface ocean pH has declined by a very likely rate of 0.016­0.020 per decade in the 5 subtropics and 0.002­0.026 per decade in the subpolar and polar zones (WGI AR6 Section 5.3.2.2, Canadell 6 et al., 2021). Typically, the pH of global surface waters has decreased from 8.2 to 8.1 since the pre-industrial 7 era (1750 CE), a trend attributable to rising atmospheric CO2 (virtually certain) (Orr et al., 2005; Jiang et al., 8 2019). 9 10 Ocean acidification is also developing in the ocean interior (very high confidence) due to the transport of 11 anthropogenic CO2 to depth by ocean currents and mixing (WGI AR6 Section 5.3.3.1, Canadell et al., 2021). 12 There, it leads to the shoaling of saturation horizons of aragonite and calcite (high confidence) (WGI AR6 13 Section 5.3.3.1, Canadell et al., 2021), below which dissolution of these calcium carbonate minerals is 14 thermodynamically favoured. The calcite or aragonite saturation horizons have migrated upwards in the 15 North Pacific (1­2 m yr­1 over 1991­2006, Feely et al., 2012) and in the Irminger Sea (10­15 m yr­1 for 16 the aragonite saturation horizon over 1991­2016, Perez et al., 2018). In some locations of the Western 17 Atlantic Ocean, calcite saturation depth has risen by ~300 m since the pre-industrial due to increasing 18 concentrations of deep-ocean dissolved inorganic carbon (Sulpis et al., 2018). In the Arctic, where some 19 coastal surface waters are already undersaturated with respect to aragonite due to the degradation of 20 terrestrial organic matter (Mathis et al., 2015; Semiletov et al., 2016), the deep aragonite saturation horizon 21 has shoaled on average by 270 ± 60 m during 1765­2005 (Terhaar et al., 2020). 22 23 Detection and attribution of ocean acidification in coastal environments are more difficult than in the open 24 ocean due to larger spatial and temporal variability of carbonate chemistry (Duarte et al., 2013; Laruelle et 25 al., 2017; Torres et al., 2021), and to the influence of other natural acidification drivers such as freshwater 26 and high-nutrient riverine inputs (Cai et al., 2011; Laurent et al., 2017; Fennel et al., 2019; Cai et al., 2020) 27 or anthropogenic acidification drivers (Section 3.1) like atmospherically deposited nitrogen and sulphur 28 (Doney et al., 2007; Hagens et al., 2014). Since AR5, the observing network in coastal oceans has expanded 29 substantially, improving understanding of both the drivers and amplitude of observed variability (Sutton et 30 al., 2016). Recent studies indicate that two more decades of observations may be required before 31 anthropogenic ocean acidification emerges over natural variability in some coastal sites and regions (WGI 32 AR6 Section 5.3.5.2, Sutton et al., 2019; Turk et al., 2019; Canadell et al., 2021). 33 34 Mean open-ocean surface pH is projected to decline by 0.08 ± 0.003 (very likely range), 0.17 ± 0.003, 0.27 ± 35 0.005 and 0.37 ± 0.007 pH units in 2081­2100 relative to 1995­2014, for SSP1-2.6, SSP2-4.5, SSP3-7.0 and 36 SSP5-8.5, respectively (Figure 3.5, WGI AR6 Section 4.3.2, Kwiatkowski et al., 2020; Lee et al., 2021). 37 Projected changes in surface pH are relatively uniform in contrast with those of other surface-ocean 38 variables, but they are largest in the Arctic Ocean (Figure 3.6, WGI AR6 Section 5.3.4.1, Canadell et al., 39 2021). Similar declines in the concentration of carbonate ions are projected by Earth System Models (ESMs, 40 Bopp et al., 2013; Gattuso et al., 2015; Kwiatkowski et al., 2020). The North Pacific, the Southern Ocean 41 and Arctic Ocean regions will become undersaturated for calcium carbonate minerals first (Orr et al., 2005; 42 Pörtner et al., 2014). Concurrent impacts on the seasonal amplitude of carbonate chemistry variables are 43 anticipated (i.e., increased amplitude for pCO2 and hydrogen ions, decreased amplitude for carbonate ions, 44 McNeil and Sasse, 2016; Kwiatkowski and Orr, 2018; Kwiatkowski et al., 2020). 45 46 Future declines in subsurface pH (Figure 3.6) will be modulated by changes in ocean overturning and water- 47 mass subduction (Resplandy et al., 2013), and in organic matter remineralisation (Chen et al., 2017). In 48 particular, decreases in pH will be less consistent at the seafloor than at the surface and will be linked to the 49 transport of surface anomalies to depth. For example, >20% of the North Atlantic seafloor deeper than 500 50 m, including canyons and seamounts designated as marine protected areas (MPAs), will experience pH 51 reductions >0.2 by 2100 under RCP8.5 (Gehlen et al., 2014). Changes in pH in the abyssal ocean (>3000 m 52 deep) are greatest in the Atlantic and Arctic Oceans, with lesser impacts in the Southern and Pacific Oceans 53 by 2100, mainly due to ventilation time scales (Sweetman et al., 2017). 54 55 3.2.3.2 Ocean Deoxygenation 56 Do Not Cite, Quote or Distribute 3-23 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Ocean deoxygenation, the loss of oxygen in the ocean, results from ocean warming, through a reduction in 2 oxygen saturation, increased oxygen consumption, increased ocean stratification and ventilation changes 3 (Keeling et al., 2010; IPCC, 2019a). In recent decades, anthropogenic inputs of nutrients and organic matter 4 (Section 3.1) have increased the extent, duration, and intensity of coastal hypoxia events worldwide (Diaz 5 and Rosenberg, 2008; Rabalais et al., 2010; Breitburg et al., 2018), while pollution-induced atmospheric 6 deposition of soluble iron over the ocean has accelerated open-ocean deoxygenation (Ito et al., 2016). 7 Dexoygenation and acidification often coincide because biological consumption of oxygen produces CO2. 8 Deoxygenation can have a range of detrimental effects on marine organisms and reduce the extent of marine 9 habitats (Sections 3.3.2, 3.4.3.1, Vaquer-Sunyer and Duarte, 2008; Chu and Tunnicliffe, 2015). 10 11 Changes in ocean oxygen concentrations have been analysed from compilations of in situ data dating back to 12 the 1960s (Helm et al., 2011; Ito et al., 2017; Schmidtko et al., 2017). SROCC concluded that a loss of 13 oxygen had occurred in the upper 1000 m of the ocean (medium confidence), with a global mean decrease of 14 0.5­3.3% (very likely range) over 1970­2010 (Bindoff et al., 2019). Based on new regional assessments 15 (Queste et al., 2018; Bronselaer et al., 2020; Cummins and Ross, 2020; Stramma et al., 2020). WGI AR6 16 assesses that ocean deoxygenation has occurred in most regions of the open ocean since the mid-20th century 17 (high confidence), but is modified by climate variability on interannual and decadal time-scales (medium 18 confidence) (WGI AR6 Sections 2.3.3.6, 5.3.3.2, Canadell et al., 2021; Gulev et al., 2021). New findings 19 since SROCC also confirm that the volume of oxygen minimum zones (OMZs) are expanding at many 20 locations (high confidence) (WGI AR6 Section 5.3.3.2, Canadell et al., 2021). 21 22 The most recent estimates of future oxygen loss in the subsurface ocean (100­600 m), using CMIP6 models, 23 amount to ­4.1 ± 4.2 (very likely range), ­6.6 ± 5.7, ­10.1 ± 6.7 and ­11.2 ± 7.7% in 2081­2100 relative to 24 1995­2014 for SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5, respectively (Figure 3.5, Kwiatkowski et al., 25 2020). Based on these CMIP6 projections, WGI AR6 concludes that the oxygen content of the subsurface 26 ocean is projected to decline to historically unprecedented conditions over the 21st century (medium 27 confidence) (WGI AR6 Section 5.3.3.2, Canadell et al., 2021). These declines are greater (by 31­72%) than 28 simulated by the CMIP5 models in their Representative Concentration Pathway (RCP) analogues, a likely 29 consequence of enhanced surface warming and stratification in CMIP6 models (Figure 3.5, Kwiatkowski et 30 al., 2020). At the regional scale and for subsurface waters, projected changes are not spatially uniform, and 31 there is lower agreement among models than they show for the global mean trend (Bopp et al., 2013; 32 Kwiatkowski et al., 2020). In particular, large uncertainties remain for these future projections of ocean 33 deoxygenation in the subsurface tropical oceans, where the major OMZs are located (Cabré et al., 2015; 34 Bopp et al., 2017). 35 36 3.2.3.3 Changes in Nutrient Availability 37 38 The availability of nutrients in the surface ocean often limits primary productivity, with implications for 39 marine food webs and the biological carbon pump. Nitrogen availability tends to limit phytoplankton 40 productivity throughout most of the low-latitude ocean, whereas dissolved iron availability limits 41 productivity in high-nutrient, low-chlorophyll regions, such as in the main upwelling region of the Southern 42 Ocean and the Eastern Equatorial Pacific (high confidence) (Moore et al., 2013; IPCC, 2019b). Phosphorus, 43 silicon, other micronutrients such as zinc, and vitamins can also co-limit marine phytoplankton productivity 44 in some ocean regions (Moore et al., 2013). Whereas some studies have shown coupling between climate 45 variability and nutrient trends in specific regions, such as in the North Atlantic (Hátún et al., 2016), North 46 Pacific (Di Lorenzo et al., 2009; Yasunaka et al., 2014) and tropical (Stramma and Schmidtko, 2021) 47 Oceans, very few studies have been able to detect long-term changes in ocean nutrient concentrations (but 48 see Yasunaka et al., 2016 ). 49 50 Future changes in nutrient concentrations have been estimated using ESMs, with future increases in 51 stratification generally leading to decreased nutrient levels in surface waters (IPCC, 2019b). CMIP6 models 52 project a decline in the nitrate concentration of the upper 100 m in 2080­2099 relative to 1995­2014 of ­ 53 0.46 ± 0.45 (very likely range), ­0.60 ± 0.58, ­0.80 ± 0.77 and ­1.00 ± 0.78 mmol m­3 under SSP1-2.6, 54 SSP2-4.5 and SSP5-8.5, respectively (Figure 3.5, Kwiatkowski et al., 2020). These declines in nitrate 55 concentration are greater than simulated by the CMIP5 models in their RCP analogues, a likely consequence 56 of enhanced surface warming and stratification in CMIP6 models (Figure 3.5, Kwiatkowski et al., 2020). It is Do Not Cite, Quote or Distribute 3-24 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 concluded that the surface ocean will encounter reduced nitrate concentrations in the 21st century (medium 2 confidence). 3 4 3.2.4 Global Synthesis on Multiple Climatic Impact Drivers 5 6 In the 21st century, ocean and coastal ecosystems are projected to face conditions unprecedented over past 7 centuries to millennia (high confidence) (Section 3.2, WGI AR6 Chapters 4, 9, Fox-Kemper et al., 2021; Lee 8 et al., 2021), with increased temperatures (virtually certain) and frequency and severity of MHWs (very high 9 confidence), stronger upper-ocean stratification (high confidence), continued rise in GMSL through the 21st 10 century (high confidence) and increased frequency of extreme sea levels (high confidence), further 11 acidification (virtually certain), oxygen decline (high confidence), and decreased surface nitrate inventories 12 (medium confidence). 13 14 The rates and magnitudes of these changes largely depend on the extent of future emissions (very high 15 confidence), with surface ocean warming and acidification (very likely range) at +3.47°C ± 1.28°C and ­0.44 16 pH units ± 0.008 in 2080­2099 (relative to 1870­1899) for SSP5-8.5 compared to +1.42°C ± 0.53°C and ­ 17 0.16 ± 0.003 for SSP1-2.6 (Figure 3.5, Kwiatkowski et al., 2020). 18 19 3.2.4.1 Compound Changes in the 21st Century 20 21 ESMs project distinct regional evolutions of the different climatic-impact drivers over the 21st century (very 22 high confidence) (Figures 3.5, 3.6, 3.7, Kwiatkowski et al., 2020). Tropical and subtropical oceans are 23 characterized by projected warming and acidification, accompanied by declining nitrate concentrations in 24 equatorial upwelling regions. The North Atlantic is characterized by a high exposure to acidification and 25 declining nitrate concentrations. The North Pacific is characterized by high sensitivity to compound changes, 26 with high rates of warming, acidification, deoxygenation and nutrient depletion. In contrast, the development 27 of compound hazards is limited in the Southern Ocean, where rates of warming and nutrient depletion are 28 lower. The Arctic Ocean is characterized by the highest rates of acidification and warming, strong nutrient 29 depletion, and it will likely become practically sea-ice free in the September mean for the first time before 30 the year 2050 in all SSP scenarios (high confidence) (Figures 3.5, 3.6, 3.7, Sections 3.2.2­3.2.3). 31 32 In general, the projected changes in climatic-impact drivers are less in absolute terms in the deep-sea 33 (mesopelagic and bathypelagic domains and deep-sea habitats) than in the surface ocean and in shallow- 34 waters habitats (kelp ecosystems, warm-water corals) (very high confidence) (Figures 3.6, 3.7, Mora et al., 35 2013; Sweetman et al., 2017). The mesopelagic domain will be nevertheless exposed to high rates of 36 deoxygenation (Figure 3.6) and high climate velocities (Figure 3.4, Section 3.2.2.1), as well as impacted by 37 the shoaling of aragonite or calcite saturation horizon (Section 3.2.3.2). Significant differences in projected 38 trends between the SSPs show that mitigation strategies will limit exposure of deep-sea ecosystems to 39 potential warming, acidification and deoxygenation during the 21st century (very high confidence) (Figure 40 3.6, Kwiatkowski et al., 2020). 41 42 Do Not Cite, Quote or Distribute 3-25 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure 3.6: Projected trends across open-ocean systems. Projected annual and global (a) average warming, (b) 3 acidification, (c) changes in dissolved oxygen concentrations and (d) changes in nitrate (NO3) concentrations for four 4 open-ocean systems, including the epipelagic (0­200 m depth), mesopelagic (200­1000 m), bathypelagic (>1000 m) 5 domains, and deep benthic waters (>200 m). All projections are based on Coupled Model Intercomparison Project 6 6 models and for three Shared Socioeconomic Pathways (SSPs), SSP1-2.6, SSP2-4.5 and SSP5-8.5 (Kwiatkowski et al., 7 2020). Anomalies in the near-term (2020­2041), mid-term (2041­2060) and long-term (2081­2100) are all relative to 8 1985­2014. Error bars represent very likely ranges. 9 10 Do Not Cite, Quote or Distribute 3-26 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure 3.7: Projected trends across coastal-ocean ecosystems. Projected (a) warming, (b) acidification, (c) changes in 3 dissolved oxygen concentrations, (d) changes in nitrate (NO3) concentrations and (e) changes in summer sea ice cover 4 fraction (September and north of 66°N for the Northern Polar Oceans, and March and south of 66°S for the Southern 5 Polar Ocean) for five coastal-ocean ecosystems. All projected trends are for the surface ocean, except oxygen 6 concentration changes that are computed for the subsurface ocean (100­600 m depth) for the upwelling ecosystems and 7 the polar seas. All projections are based on Coupled Model Intercomparison Project 6 (CMIP6) models and for three 8 Shared Socioeconomic Pathways (SSPs): SSP1-2.6, SSP2-4.5 and SSP5-8.5 (Kwiatkowski et al., 2020). Anomalies in 9 the near-term (2020­2041), mid-term (2041­2060) and long-term (2081­2100) are all relative to 1985­2014. Error bars 10 represent very likely ranges. Coastal seas are defined on a 1° × 1° grid when bathymetry is less than 200 m deep. 11 Distribution of warm-water corals is from UNEP-WCMC et al. (2018). Distribution of kelp ecosystems is from OBIS 12 (2020). Upwelling areas are defined according to Rykaczewski et al. (2015). 13 14 15 3.2.4.2 Time of Emergence 16 17 Anthropogenic changes in climatic impact-drivers assessed here exhibit vastly distinct times of emergence, 18 which is the time scale over which an anthropogenic signal related to climate change is statistically detected 19 to emerge from the background noise of natural climate for a specific region (Christensen et al., 2007; 20 Hawkins and Sutton, 2012). SROCC concluded that for ocean properties, the time of emergence ranges from 21 under a decade (e.g., surface ocean pH) to over a century (e.g., net primary production, see Section 3.4.3.3.4 22 for time of emergence of biological properties, Bindoff et al., 2019). 23 24 The literature assessed in SROCC mainly focused on surface ocean properties and gradual mean changes. 25 Since then, the time of emergence has also been investigated for subsurface properties, ocean extreme events 26 and particularly vulnerable regions, such as the Arctic Ocean (Hameau et al., 2019; Oliver et al., 2019; 27 Burger et al., 2020; Landrum and Holland, 2020; Schlunegger et al., 2020), but subsequent assessments are 28 low confidence due to limited evidence. Below the surface, changes in temperature typically emerge from 29 internal variability prior to changes in oxygen. However in about a third of the global thermocline, 30 deoxygenation emerges prior to warming (Hameau et al., 2019). Permanent MHW states, defined as when Do Not Cite, Quote or Distribute 3-27 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 SST exceeds the MHW threshold continuously over a full calendar year, will emerge during the 21st century 2 in many parts of the surface ocean (Oliver et al., 2019). Ocean acidification extremes have already emerged 3 from background natural internal variability during the 20th century in most of the surface ocean (Burger et 4 al., 2020). In the Arctic, anthropogenic sea-ice changes have already emerged from the background internal 5 variability, and anthropogenic alteration of air temperatures will emerge in the early- to mid-21st century 6 (Landrum and Holland, 2020). 7 8 3.2.4.4 Perspectives from Paleo Data 9 10 Paleo observations are useful to assess multiple hazards of environmental change while excluding direct 11 anthropogenic impacts (Section 3.4.3.3). Ancient intervals of rapid climate warming that occurred between 12 300 to 50 million years ago (Ma) were triggered by the release of greenhouse gases (high confidence). The 13 sources of greenhouse gases varied but include volcanic degassing from continental flood basalts and 14 methane hydrates stored in marine sediments and soils (Foster et al., 2018). Six extreme ancient 15 hyperthermal events are known from the last 300 Ma, when tropical SSTs reached 1.5°C­10°C warmer than 16 pre-industrial conditions, and with substantial impacts on ancient life (Cross-Chapter Box PALEO in 17 Chapter 1). Warming and deoxygenation in the oceans were closely associated in hyperthermal events (high 18 confidence), with anoxia reaching the photic zone and abyssal depths (Kaiho et al., 2014; Müller et al., 2017; 19 Penn et al., 2018; Weissert, 2019), whereas ocean acidification has not been demonstrated consistently 20 (medium confidence) (Hönisch et al., 2012; Penman et al., 2014; Clarkson et al., 2015; Harper et al., 2020a; 21 Jurikova et al., 2020; Müller et al., 2020). 22 23 Greenhouse gases also contributed substantially to shaping the longer-term climate trends over the last 50 24 million years, although changes in continental configuration and ocean circulation as well as planetary 25 orbital cycles were equally important (WGI AR6 Cross-Chapter Box 2.1, Westerhold et al., 2020; Gulev et 26 al., 2021). There is little evidence for ocean acidification in the last 2.6 Ma (low confidence) (Hönisch et al., 27 2012), but ocean ventilation was highly sensitive to even modest warming such as observed in the last 28 10,000 years (medium confidence) (Jaccard and Galbraith, 2012; Lembke-Jene et al., 2018). 29 30 31 3.3 Linking Biological Responses to Climatic-Impact Drivers 32 33 3.3.1 Introduction 34 35 This section assesses new evidence since AR5 (Pörtner et al., 2014) and SROCC (Bindoff et al., 2019) 36 regarding biotic responses to multiple environmental drivers. It assesses differential sensitivities among life 37 stages within individual organisms, changing responses across scales of biological organisation and the 38 potential for evolutionary adaptation to climate change (e.g., Przeslawski et al., 2015; Boyd et al., 2018; 39 Reddin et al., 2020), providing examples and identifying key gaps and uncertainties that limit our ability to 40 project the ecological impact of multiple climate-impact drivers (Figure 3.8a). The assessment includes 41 physiological responses to single environmental drivers and their underlying mechanisms (Section 3.3.2), the 42 characteristics of multiple drivers and organisms' responses to them (Section 3.3.3), short-term acclimation 43 and longer-term evolutionary adaptation of populations (Section 3.3.4), and it concludes with an assessment 44 of progress in upscaling laboratory findings to ecosystems within in situ settings (Figure 3.8b, Section 3.3.5). 45 46 Do Not Cite, Quote or Distribute 3-28 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure 3.8: The state of knowledge regarding ecological responses to environmental drivers in experimental settings. 3 (a) Schematic indicating where themes are discussed within Section 3.3, and how they jointly inform policy; adapted 4 from Riebesell and Gattuso (2014). (b) The hierarchy of accumulating physiological knowledge (grey layers), from 5 single (e.g., Pörtner et al., 2012) to multiple drivers, and from simple outcomes (e.g., Sciandra et al., 2003), interactions 6 among drivers (e.g., Crain et al., 2008) and identification of physiological roles of drivers (e.g., Bach et al., 2015) to 7 mechanistic understanding of drivers (e.g., Thomas et al., 2017). At present, the upper grey layer has been achieved, in 8 full, for two drivers e.g., temperature and nutrient concentrations, with validation of dual controls on phytoplankton 9 growth rate, Thomas et al. (2017). Hatched layers denote major advances since WGII AR5 Chapter 6 (Pörtner et al., 10 2014). The green layer indicates the level of understanding potentially needed to project the response of marine life 11 subjected to multiple drivers. Red horizontal arrows indicate the influence of confounding factors on our current 12 understanding, including population genetics, fluctuating oceanic conditions, or extreme events. 13 14 15 3.3.2 Responses to Single Drivers 16 17 Anthropogenic CO2 emissions trigger a suite of changes that alter ocean temperature, pH and CO2 18 concentration, oxygen concentration, and nutrient supply at global scales (Section 3.2). The response 19 pathways of these climate-impact drivers have been investigated primarily as single variables. 20 21 Temperature affects the movement and transport of molecules and, thereby, the rates of all biochemical 22 reactions. Thus, ongoing and projected warming (Section 3.2.2.1) that remains below an organism's 23 physiological optimum will generally raise metabolic rates (very high confidence) (Pörtner et al., 2014). 24 Beyond this optimum (Topt, Figure 3.9), metabolism typically decreases sharply, finally reaching a critical 25 threshold (Tcrit) beyond which enzymes become thermally inactivated and cells undergo oxidative stress. 26 Local and regional adaptation affect the heat tolerance thresholds of organisms. For example, organisms 27 adapted to thermally-stable environments (e.g., tropical, polar, deep-sea) are often more sensitive to warming 28 than those from thermally variable environments (e.g., estuaries) (very high confidence) (Section 3.4, Sunday 29 et al., 2019; Collins et al., 2020). Heat tolerance also decreases with increasing organisational complexity 30 (Storch et al., 2014; Pörtner and Gutt, 2016), and is lower in eggs, embryos and spawning fish than for their 31 larval stages or adults outside the spawning season (high confidence) (Dahlke et al., 2020b). By altering 32 physiological responses, projected changes in ocean warming (Section 3.2.2.1) will modify growth, 33 migration, distribution, competition, survival and reproduction (very high confidence) (Messmer et al., 2017; 34 Dahlke et al., 2018; Andrews et al., 2019; Pinsky et al., 2019; Anton et al., 2020). 35 36 Altered seawater carbonate chemistry (Section 3.2.3.1) affects specific processes to varying degrees. For 37 example, higher CO2 concentrations can increase photosynthesis and growth in some phytoplankton, 38 macroalgal and seagrass species (high confidence) (Pörtner et al., 2014; Seifert et al., 2020; Zimmerman, 39 2021), while lower pH levels decrease calcification (high confidence) (Pörtner et al., 2014; Falkenberg et al., 40 2018; Doney et al., 2020; Fox et al., 2020; Reddin et al., 2020) or silicification (low confidence) (Petrou et 41 al., 2019). Organisms' capacity to compensate for or resist acidification of internal fluids depends on their 42 capacity for acid-base regulation, which differs due to organisms' wide-ranging biological complexity and 43 adaptive abilities (low to medium confidence) (Vargas et al., 2017; Melzner et al., 2020). Detrimental 44 impacts of acidification include decreased growth and survival, and altered development, especially in early 45 life stages (high confidence) (Dahlke et al., 2018; Onitsuka et al., 2018; Hancock et al., 2020), along with Do Not Cite, Quote or Distribute 3-29 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 lowered recruitment and altered behaviour in animals (Kroeker et al., 2013a; Wittmann and Pörtner, 2013; 2 Clements and Hunt, 2015; Cattano et al., 2018; Esbaugh, 2018; Bednarsek et al., 2019; Reddin et al., 2020). 3 For finfish, laboratory studies of behavioural and sensory consequences of ocean acidification showed mixed 4 results (Rossi et al., 2018; Nagelkerken et al., 2019; Stiasny et al., 2019; Velez et al., 2019; Clark et al., 5 2020; Munday et al., 2020). Calcifiers are generally more sensitive to acidification (e.g., for growth and 6 survival) than non-calcifying groups (high confidence) (Kroeker et al., 2013a; Wittmann and Pörtner, 2013; 7 Clements and Hunt, 2015; Cattano et al., 2018; Bednarsek et al., 2019; Reddin et al., 2020; Seifert et al., 8 2020). For calcifying primary producers, including phytoplankton and coralline algae, ocean acidification 9 has different, often opposing effects, for example, decreasing calcification while photosynthetic rates 10 increase (high confidence) (Riebesell et al., 2000; Van de Waal et al., 2013; Bach et al., 2015; Cornwall et 11 al., 2017b; Gafar et al., 2019). 12 13 14 15 Do Not Cite, Quote or Distribute 3-30 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure 3.9: Organismal responses to single and multiple drivers. (a) The generic temperature-response curve describing 3 physiological process rates as a nonlinear function of a particular driver (e.g., temperature) with maximum rates (Rmax) 4 and temperature optima (Topt). The driver range that keeps physiological rates above a certain threshold represents the 5 organism's range of phenotypic plasticity, while below that threshold, the critical temperature (Tcrit), physiological 6 performance is so low as to constitute stressful conditions. (b) The response curve for one driver can depend on other 7 drivers, here exemplified for temperature and pH in the central panel. This interaction causes rates as well as optima to 8 change with (left) pH and (right) temperature, indicated by the coloured lines. (c) Impacts of multiple drivers on 9 processes can be (blue) additive, (red) synergistic or (green) antagonistic, that is, the cumulative effects of two (or 10 more) drivers are equal to, larger than or smaller than the sum of their individual effects, respectively. Potential 11 experimental outcomes affected by additive, synergistic, and antagonistic interactions are shown for scenarios where 12 drivers (left) increase rates, (centre) decrease rates, or (right) cause opposite responses, showing how experimental 13 outcomes can mask these mechanistic interactions. Adapted from Crain et al. (2008) and Piggott et al. (2015). For a 14 quantitative analysis of effects of driver pairs on animals, see Figure SM3.2. 15 16 17 Oxygen concentrations affect aerobic and anaerobic processes, including energy metabolism and 18 denitrification. Projected decreases in dissolved oxygen concentration (Section 3.2.3.2) will thus impact 19 organisms and their biogeography in ways dependent upon their oxygen requirements (Deutsch et al., 2020), 20 which are highest for large, multicellular organisms (Pörtner et al., 2014). The upper ocean generally 21 contains high dissolved-oxygen concentrations due to air-sea exchange and photosynthesis, but in subsurface 22 waters, deoxygenation may impair aerobic organisms in multiple ways (Oschlies et al., 2018; Galic et al., 23 2019; Thomas et al., 2019; Sampaio et al., 2021). Many processes contribute to lowered oxygen levels: 24 altered ventilation and stratification; microbial respiration enhanced by nearshore eutrophication; and less 25 oxygen solubility in warmer waters. For example, deoxygenation in highly eutrophic estuarine and coastal 26 marine ecosystems (Section 3.4.2) can result from accelerated microbial activity, leading to acute organismal 27 responses. Under hypoxia (oxygen concentrations 2 mg L­1, Limburg et al., 2020), physiological and 28 ecological processes are impaired and communities undergo species migration, replacement and loss, 29 transforming community composition (very high confidence) (Chu and Tunnicliffe, 2015; Gobler and 30 Baumann, 2016; Sampaio et al., 2021). Hypoxia can lead to expanding OMZs which will favour specialised 31 microbes and hypoxia-tolerant organisms (medium confidence) (Breitburg et al., 2018; Ramírez-Flandes et 32 al., 2019). As respiration consumes oxygen and produces CO2, lowered oxygen levels are often interlinked 33 with acidification in coastal and tropical habitats (Rosa et al., 2013; Gobler and Baumann, 2016; Feely et al., 34 2018) and is an example of a compound hazard (Sections 3.2.4.1, 3.4.2.4). 35 36 Increased density stratification and mixed-layer shallowing, caused by warming, freshening and sea-ice 37 decline, can alter light climate and nutrient availability within the surface mixed layer (high confidence) 38 (Section 3.2.2.3). As light and nutrient levels drive photosynthesis, changes in these drivers directly affect 39 primary producers, often in different directions (Matsumoto et al., 2014; Deppeler and Davidson, 2017). 40 Decreased upward nutrient supply is expected to decrease primary production in the low-latitude ocean 41 (medium confidence) (Section 3.4.4.2.1, Moore et al., 2018a; Kwiatkowski et al., 2019). Alternatively, higher 42 mean underwater light levels resulting from changes in sea ice and/or mixed layer shallowing can increase Do Not Cite, Quote or Distribute 3-31 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 primary production in high-latitude offshore regions, provided nutrient levels remain sufficiently high 2 (medium confidence) (Section 3.4.4.2.1, Cross-Chapter Paper 6, Vancoppenolle et al., 2013; Deppeler and 3 Davidson, 2017; Tedesco et al., 2019; Ardyna and Arrigo, 2020; Lannuzel et al., 2020). In some parts of the 4 open Southern Ocean, where iron limitation largely controls primary productivity (Tagliabue et al., 2017), 5 changes in wind fields will deepen the summer mixed-layer depth (Panassa et al., 2018), entrain more 6 nutrients, and raise primary productivity in the future (medium confidence) (Cross-Chapter Paper 6, Hauck et 7 al., 2015; Leung et al., 2015; Moore et al., 2018a; Kwiatkowski et al., 2020). 8 9 Climate-impact drivers fluctuate on time scales ranging from diurnal to annual, with potential consequences 10 for organismal responses (Figure 3.10), but these fluctuations are commonly not incorporated 11 experimentally. Experiments that simulate natural fluctuations in drivers, especially beyond tidal or diel 12 cycles, can result in more detrimental impacts than those based on quasi-constant conditions (Eriander et al., 13 2015; Sunday et al., 2019), but can also ameliorate effects (Comeau et al., 2014; Laubenstein et al., 2020; 14 Cabrerizo et al., 2021), confirming that the influence of environmental variability requires evaluation (Dowd 15 et al., 2015). MHWs exacerbate the impacts of rising mean temperatures, with major ecological 16 consequences (very high confidence) (Frölicher et al., 2018; IPCC, 2018; Arafeh-Dalmau et al., 2020; 17 Laufkötter et al., 2020). Higher temperature variability decreased phytoplankton growth and calcification in 18 Emiliania huxleyi relative to a stable warming regime (Wang et al., 2019b). Diel fluctuations (i.e., over 24 h) 19 in carbonate chemistry superimposed on current and future pCO2 levels influenced diatom species 20 differently, depending on their habitat (Li et al., 2016). CO2 fluctuations overlaid on changing mean values 21 also altered phenotypic evolutionary outcomes of picoeukaryotic algae (Schaum et al., 2016). In the bivalve 22 Mytilus edulis, fluctuating pH regimes exerted higher metabolic costs (Mangan et al., 2017), while salinity 23 fluctuations might be more influential than pH fluctuations in other bivalves (Velez et al., 2016). The 24 amplitude of diel and seasonal pH and CO2 changes are projected to increase in the future due to lowered 25 CO2 seawater buffering capacity (very high confidence) (Section 3.2.3.1, Burger et al., 2020), which can 26 impose additional stress on organisms. 27 28 3.3.3 Responses to Multiple Drivers 29 30 Each organism encounters a unique combination of local and climate-impact drivers, which vary in space 31 and time. The contribution of these drivers to an organism's overall biological response, and thereby also 32 potential risks for the organism, depends on the intensity and duration of its exposure to these drivers and 33 associated sensitivities. Both geographical location (e.g., polar, tropical) and marine habitat (e.g., benthic, 34 pelagic) strongly affect the combination of climate and non-climate drivers that organisms are exposed to. 35 Non-climate drivers (Section 3.1) can dominate outcomes or amplify vulnerability to climate-impact drivers, 36 with mostly detrimental effects such as extirpation (very high confidence) (Section 3.4, Boyd et al., 2018; 37 Gissi et al., 2021), and unique feedbacks may exist between climate change and drivers like habitat loss or 38 invasive species that further confound climate change effects (Ortiz et al., 2018; Wolff et al., 2018; Gissi et 39 al., 2021). Individual responses are further influenced by an organism's behaviour, trophic level and life- 40 history strategy (Figure 3.10, Przeslawski et al., 2015; Boyd et al., 2018). Evidence is increasing that some 41 life-history stages are more sensitive to specific drivers than others (Dahlke et al., 2020b). To identify the 42 most influential drivers for an organism requires targeting key traits (e.g., calcification, reproduction). The 43 trophic level of the organism must also be considered, because autotrophs directly depend on light and 44 nutrients while invertebrates are often more sensitive to changes in oxygen or altered prey, but temperature 45 plays a key role for both groups (Figure 3.10b). 46 Do Not Cite, Quote or Distribute 3-32 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure 3.10: The effect of environmental drivers differs depending upon organisms' life history, and trophic strategy or 3 habitat. (a) pH variability differs for benthic invertebrates, such as sea urchins (in blue), and their pelagic larvae (in 4 green); pH fluctuations over the annual cycle can be much larger in the water column (due to primary production) 5 relative to the seafloor. Variability associated with behaviour and life stage strongly defines organisms' niches and 6 sensitivities to present and future conditions. (b) Examples of organisms that are influenced by different suites of 7 drivers that are set jointly by their habitat (e.g., benthic versus epipelagic settings) and trophic strategy (e.g., nutrients 8 for phytoplankton, prey characteristics for grazers). 9 10 11 Co-occurring environmental drivers often cause complex organismal responses (high confidence) (Pörtner et 12 al., 2014). Individual drivers can have detrimental, neutral or beneficial effects, depending on the 13 relationship between driver and physiological process (Section 3.3.2, Figure 3.9a). Multiple drivers can have 14 interactive effects, where the response to one driver alters the sensitivity to another, and outcomes cannot be 15 deduced from individual drivers' effects (Figure 3.9b). Impacts of multiple drivers can be additive, 16 synergistic or antagonistic (Figure 3.9c, Crain et al., 2008; Piggott et al., 2015; Boyd et al., 2018; Bindoff et 17 al., 2019). Well-controlled laboratory studies on multiple-driver effects have revealed insights into the mode 18 of action of individual drivers and their interdependence (Kroeker et al., 2017; Gao et al., 2019; Reddin et 19 al., 2020; Seifert et al., 2020; Green et al., 2021b; Sampaio et al., 2021). Understanding the outcomes of 20 interactive drivers is important for robustly assessing risks to organisms under different climate-change 21 scenarios. 22 23 3.3.3.1 Effects of Multiple Drivers on Primary Producers 24 25 Warming and rising CO2 concentrations enhance growth and/or photosynthetic rates in many species of 26 cyanobacteria, picoeukaryotes, coccolithophores, dinoflagellates and diatoms (high confidence) (Fu et al., 27 2007; Sett et al., 2014; Hoppe et al., 2018a; Wolf et al., 2018; Brandenburg et al., 2019), and the optimum 28 pCO2 for growth and/or primary production shifts upward under warming (medium confidence) (Sett et al., 29 2014; Hoppe et al., 2018a). Warming and ocean acidification appear to jointly favour the proliferation and 30 toxicity of harmful algal bloom (HAB) species (limited evidence, high agreement) (Section 3.5.5.3, Bindoff 31 et al., 2019; Brandenburg et al., 2019; Griffith et al., 2019a; Wells et al., 2020), but a 2021 analysis found no 32 uniform global trend in HABs or their distribution over 1985­2018, once field data were adjusted for 33 regional variations in monitoring effort (Hallegraeff et al., 2021). The predominantly detrimental impacts of 34 ocean acidification on coccolithophores can partly be offset by warming (Seifert et al., 2020), but also be 35 exacerbated, depending on the magnitudes of drivers (D'Amario et al., 2020). For non-calcifying 36 macroalgae, responses are highly species-specific and often indicate synergistic interactions between 37 warming and acidification (Kram et al., 2016; Falkenberg et al., 2018). Ocean acidification poses a large risk 38 for coralline algae that is further amplified by warming (medium confidence) (Section 3.4.2.2, Cornwall et 39 al., 2019). However, temperatures up to 5°C above ambient do not decrease calcification (Cornwall et al., 40 2019), and there is limited evidence that some species have the physiological capacity to resist acidification 41 via pH upregulation at the calcification site (Cornwall et al., 2017a). For seagrass, warming beyond a 42 species' thermal tolerance will limit growth and impact germination, but ocean acidification appears to Do Not Cite, Quote or Distribute 3-33 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 increase thermal tolerance of some eelgrass species by increasing the photosynthesis-to-respiration ratio 2 (medium confidence) (Egea et al., 2018; Scalpone et al., 2020; Zimmerman, 2021). 3 4 Thermal sensitivity of pelagic primary producers changes with nutrient supply (high confidence) (Thomas et 5 al., 2017; Marañón et al., 2018; Fernández et al., 2020). Phosphorus limitation lowers the temperature 6 optimum for growth of phytoplankton, making these organisms more prone to heat stress (Thomas et al., 7 2017; Bestion et al., 2018). This trend may hold for open-ocean phytoplankton, which are often iron-limited 8 (medium confidence) (Boyd, 2019). Such temperature-nutrient interactions might be especially relevant 9 during summer MHWs (Section 3.2.2.1, Cross-Chapter Box EXTREMES in Chapter 2, IPCC, 2018; 10 Holbrook et al., 2019; DeCarlo et al., 2020; Hayashida et al., 2020), when primary producers are often 11 nutrient-limited and near their thermal limits. Increasingly frequent and intense MHWs along with projected 12 decreases in nutrient availability (Section 3.2.3.3) may push some primary producers beyond tolerance 13 thresholds. Temperature-nutrient interactions can also alter the photosynthesis-to-respiration ratio in 14 phytoplankton (Marañón et al., 2018). Overall, rising metabolic rates due to warming will be restricted to 15 primary producers in high-nutrient regions (medium confidence) (Thomas et al., 2017; Marañón et al., 2018). 16 For zooxanthellae-containing corals, nutrient supply from upwelling or from runoff can increase coral 17 susceptibility to bleaching during warm-season MHWs (DeCarlo et al., 2020; Wooldridge, 2020). 18 19 The effects of ocean acidification on growth, metabolic rates or elemental composition of primary producers 20 changes with nutrient availability and light conditions (high confidence) (Gao et al., 2019; Seifert et al., 21 2020). While interactions with nutrients are often additive in phytoplankton, diatoms revealed predominantly 22 synergistic interactions (Seifert et al., 2020). Growth or photosynthesis of some diatom and HAB species, for 23 instance, are stimulated by ocean acidification only if nutrients are replete (Hoppe et al., 2013; Boyd et al., 24 2015b; Eberlein et al., 2016; Griffith et al., 2019a). Interactions with light are more complex because relative 25 effects of ocean acidification are larger under limiting irradiances, while saturating light levels decrease 26 beneficial or detrimental effects on these processes (Kranz et al., 2010; Garcia et al., 2011; Rokitta and Rost, 27 2012; Heiden et al., 2016). For the coccolithophore Emiliania huxleyi, for example, the impacts of ocean 28 acidification are less detrimental under high light availability, which could partly explain why this species is 29 moving poleward (Winter et al., 2013; Kondrik et al., 2017; Neukermans et al., 2018), although acidification 30 is more pronounced in polar waters (Section 3.2.3.1, Cross-Chapter Paper 6). Under excess light, however, 31 the detrimental impacts of ocean acidification are amplified for many species (high confidence) (Gao et al., 32 2012; Li and Campbell, 2013; Zhang et al., 2015; Kottmeier et al., 2016; Gafar et al., 2019). Lowered photo- 33 physiological capacity to cope with high-light stress and avoid photodamage (Gao et al., 2012; Li and 34 Campbell, 2013; Hoppe et al., 2015; Kvernvik et al., 2020) is also consistent with observations that dynamic 35 light regimes can become more stressful under ocean acidification (Jin et al., 2013; Hoppe et al., 2015). 36 Given the expected mixed-layer shallowing in some regions (Section 3.2.2.3), the exposure to overall higher 37 mean irradiances could shift the effects of acidification from beneficial to detrimental for some primary 38 producers, depending on species and organismal traits (medium confidence) (Gao et al., 2019; Seifert et al., 39 2020). 40 41 Studies investigating two drivers provide most of the information on the wide range of interactive effects of 42 drivers on phytoplankton (Gao et al., 2019; Seifert et al., 2020), although climate change alters several 43 oceanic drivers concurrently (Section 3.2). The few experimental studies that have addressed three or more 44 drivers (Xu et al., 2014; Boyd et al., 2015b; Brennan and Collins, 2015; Brennan et al., 2017; Hoppe et al., 45 2018b; Moreno-Marín et al., 2018) indicate that one or two drivers generally dominate the cumulative 46 outcome, with others playing a subordinate role (medium confidence). In these studies, temperature had a 47 disproportionately large influence, while other drivers differed in importance, depending on the type of 48 primary producer, ecosystem characteristics and selected driver values. 49 50 3.3.3.2 Effects of Multiple Drivers on Animals 51 52 When changing CO2 concentrations affect marine ectotherms, they typically combine additively or 53 synergistically with warming (medium confidence) (e.g., Lefevre, 2016; Reddin et al., 2020; Sampaio et al., 54 2021), and their cumulative effects can lead to detrimental, neutral or beneficial effects (high confidence) 55 (Figure 3.9a, Bennett et al., 2017; Büscher et al., 2017; Dahlke et al., 2017; Foo and Byrne, 2017; Johnson et 56 al., 2017b; Cominassi et al., 2019). Higher ocean CO2 influences the thermal tolerance of species adapted to 57 extreme but stable habitats in tropical and polar regions, more than that of thermally tolerant generalists Do Not Cite, Quote or Distribute 3-34 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 (high confidence) (Byrne et al., 2013; Schiffer et al., 2014; Flynn et al., 2015; Kunz et al., 2016; Pörtner et 2 al., 2017; Kunz et al., 2018; Bindoff et al., 2019, but see) (Ern et al., 2017), especially in early life stages 3 (Dahlke et al., 2020a). In thermal generalists from temperate and subtropical species, warming and ocean 4 acidification generally have detrimental effects on growth and survival (e.g., Gao et al., 2020), but warming 5 can also alleviate the detrimental effects of ocean acidification by increasing metabolic rate and/or growth 6 (Garzke et al., 2020), provided that other conditions (e.g., thermal niche, food availability) are beneficial. For 7 example, larval growth and survival of Australasian snapper (Pagrus auratus) appear to benefit from 8 combined acidification and warming (but see Watson et al., 2018; McMahon et al., 2020), introducing major 9 uncertainties to population modelling (Section 3.3.4, Parsons et al., 2020). 10 11 As with ocean acidification, reduced oxygen availability further alters the influence of warming on metabolic 12 rates (high confidence). Acidification and hypoxia can contribute to a decrease or shift in thermal tolerance, 13 while the magnitude of this effect depends on the duration of exposure (Tripp-Valdez et al., 2017; Cattano et 14 al., 2018; Calderón-Liévanos et al., 2019; Schwieterman et al., 2019). Warming and hypoxia are mostly 15 positively correlated and tolerance to both phenomena are often linked after long-term acclimation (e.g., 16 Bouyoucos et al., 2020). Acute short-term heat shocks can impair hypoxia tolerance, for instance in intertidal 17 fish (McArley et al., 2020). This is relevant for shallow waters, specifically for MHWs (Section 3.2.2.1, 18 Hobday et al., 2016a; IPCC, 2018; Collins et al., 2019a). Ocean acidification can increase hypoxia tolerance 19 in some cases, possibly by downregulating activity (Faleiro et al., 2015) and/or changing blood oxygenation 20 (Montgomery et al., 2019). Other studies, however, reported additive negative effects of acidification and 21 warming on hypoxia tolerance (Schwieterman et al., 2019; Götze et al., 2020), in line with the oxygen- and 22 capacity-limited thermal tolerance (OCLTT) hypothesis presented in AR5 (Pörtner et al., 2014): warming 23 causes increased metabolic rates and oxygen demand in ectotherms, which at some point exceed supply 24 capacities (that also depend on environmental oxygen availability) and reduce aerobic scope. In 25 consequence, expansion of OMZs and other regions where warming, hypoxia and acidification combine will 26 further reduce habitat for many fish and invertebrates (high confidence) (Sections 3.4.3.2­3.4.3.3). 27 28 Food availability modulates, and may be more influential than, other driver responses by affecting the 29 energetic and nutritional status of animals (Cole et al., 2016; Stiasny et al., 2019; Cominassi et al., 2020). 30 Laboratory studies conducted under an excess of food risk underestimating the ecological effects of climate- 31 impact drivers, because increased feeding rates may help mitigate adverse effects (Nowicki et al., 2012; 32 Towle et al., 2015; Cominassi et al., 2020). Lowered food availability from reduced open-ocean primary 33 production (Sections 3.2.3.3, 3.4.4.2.1) will act as an additional driver, amplifying the detrimental effects of 34 other drivers. However, warming and higher CO2 availability may increase primary productivity in some 35 coastal areas (Section 3.4.4.1), ameliorating the adverse direct effects on animals (e.g., Sswat et al., 2018). 36 Due to the few studies addressing food availability under multiple-driver scenarios (Thomsen et al., 2013; 37 Pistevos et al., 2015; Towle et al., 2015; Ramajo et al., 2016; Brown et al., 2018a; Cominassi et al., 2020), 38 there is medium confidence in its modulating effect on climate-impact driver responses. 39 40 Animal behaviour can be affected by ocean acidification, warming and hypoxia. While warming and hypoxia 41 mostly induce avoidance behaviour, potentially leading to migration and habitat compression (Section 3.4, 42 McCormick and Levin, 2017; Limburg et al., 2020), the effects of acidification appear more complex. Some 43 studies reported that acidification dominates behavioural effects (Schmidt et al., 2017), although outcomes 44 vary with experimental design and duration of exposure (low confidence, low agreement) (Maximino and de 45 Brito, 2010; Munday et al., 2016; Laubenstein et al., 2018; Munday et al., 2019; Sundin et al., 2019; Clark et 46 al., 2020; Munday et al., 2020; Williamson et al., 2021). Behaviour represents an integrated phenomenon 47 that can be influenced both directly and indirectly by multiple drivers. For instance, increased pCO2 can 48 directly act on neuronal signalling pathways (e.g., Gamma-aminobutyric acid hypothesis, Nilsson et al., 49 2012; Thomas et al., 2020) and influence learning (Chivers et al., 2014), vision (Chung et al., 2014), and 50 choice and escape behaviour (Watson et al., 2014; Wang et al., 2017b). There is further evidence that 51 observed alterations in fish olfactory behaviour under ocean acidification may result from physiological and 52 molecular changes of the olfactory epithelium, influencing olfactory receptors (Roggatz et al., 2016; Porteus 53 et al., 2018; Velez et al., 2019; Mazurais et al., 2020). Temperature mainly drives metabolic processes and 54 thus energetic requirements, which can indirectly influence behaviour, including increased risk-taking during 55 feeding (Marangon et al., 2020). Ocean warming also accelerates the biochemical reactions and metabolic 56 processes that are primarily influenced by acidification. It is therefore difficult to generalise to what extent 57 co-occurring ocean warming ameliorates or exacerbates effects of acidification on behaviour (Laubenstein et Do Not Cite, Quote or Distribute 3-35 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 al., 2019); outcomes depend upon species and life stage (Faleiro et al., 2015; Chan et al., 2016; Tills et al., 2 2016; Wang et al., 2018b; Jarrold et al., 2020), interactions between species (e.g., Paula et al., 2019) along 3 with confounding factors including food availability and salinity (medium confidence) (Ferrari et al., 2015; 4 Pistevos et al., 2015; Pimentel et al., 2016; Pistevos et al., 2017; Horwitz et al., 2020). 5 6 While hypoxia can dominate multiple-driver responses locally (Sampaio et al., 2021), warming is the 7 fundamental physiological driver for most marine ectotherms, globally, as it directly affects their entire 8 biochemistry and energy metabolism. Other influential drivers include ocean acidification, salinity (high 9 confidence) (Lefevre, 2016; Whiteley et al., 2018; Reddin et al., 2020) or food availability/quality (medium 10 confidence) (Nagelkerken and Munday, 2016; Gao et al., 2020). Fluctuating and decreasing salinity may 11 aggravate the detrimental effects of warming and elevated CO2, because dilution with freshwater lowers 12 acid-base buffering capacity, resulting in lower pH and calcium carbonate saturation state (Dickinson et al., 13 2012; Shrivastava et al., 2019; Melzner et al., 2020). 14 15 3.3.4 Acclimation and Evolutionary Adaptation 16 17 Climate change is and will continue to be a major driver of natural selection, causing important changes in 18 fitness-related (e.g., growth, reproduction, survival) and functional (e.g., body/cell size, morphology, 19 physiology) traits, and in the genetic diversity of natural populations (medium confidence) (Pauls et al., 2013; 20 Merilä and Hendry, 2014). Climate-change impacts will continue to be exacerbated by interactions with non- 21 climate drivers such as habitat fragmentation or loss, pollution, or resource overexploitation, which limit the 22 adaptive potential of populations to future conditions (Trathan et al., 2015; Gaitán-Espitia and Hobday, 23 2021). However the ultimate responses to complex change are conditioned by the rate and magnitude of 24 environmental change, organisms' capacity for acclimation, the degree of local adaptation of natural 25 populations, and populations' potential for adaptive evolution (Figure 3.11, Pespeni et al., 2013; Calosi et al., 26 2017; Vargas et al., 2017). These controlling factors are mainly determined by local environmental 27 conditions encountered by populations across their geographical distribution (Boyd et al., 2016). In highly 28 fluctuating environments (e.g., upwelling regions, coastal zones), multiple drivers can change and interact 29 across temporal and spatial scales, generating geographical mosaics of tolerances and sensitivities to 30 environmental and climate change in marine organisms (medium confidence) (Pespeni et al., 2013; Boyd et 31 al., 2016; Vargas et al., 2017; Li et al., 2018a). A further challenge for marine life lies in its ability to cope 32 with extreme events such as MHWs (Cross-Chapter Box EXTREMES in Chapter 2). The interplay between 33 the abruptness, intensity, duration, magnitude and reoccurrence of extreme events may alter or prevent 34 evolutionary responses (e.g., adaptation) to climate change and the potential for acclimation to extreme 35 conditions such as MHWs (Cheung and Frölicher, 2020; Coleman et al., 2020a; Gurgel et al., 2020; Gruber 36 et al., 2021). 37 38 Some studies have documented higher phenotypic plasticity and tolerance to ocean warming and 39 acidification in marine invertebrates (Dam, 2013; Kelly et al., 2013; Pespeni et al., 2013; Gaitán-Espitia et 40 al., 2017a; Vargas et al., 2017; Li et al., 2018a), seaweeds (Noisette et al., 2013; Padilla-Gamiño et al., 2016; 41 Machado Monteiro et al., 2019), and fish (medium confidence) (Sandoval-Castillo et al., 2020; Enbody et al., 42 2021) living in coastal zones characterised by strong temporal fluctuations in temperature, pH, pCO2, light 43 and nutrients. For these populations, strong directional selection with intense and highly fluctuating 44 conditions may have favoured local adaptation and increased tolerance to environmental stress (low 45 confidence, low evidence) (Hong and Shurin, 2015; Gaitán-Espitia et al., 2017b; Li et al., 2018a). 46 47 Other mechanisms acting within and across generations can influence selection and inter-population 48 tolerances to environmental and climate-impact drivers. For instance, transgenerational effects and/or 49 developmental acclimation, both so-called "carry-over effects" (where the early-life environment affects the 50 expression of traits in later life stages or generations), can influence within- and cross-generational changes 51 in the tolerances of marine organisms (medium confidence) to ocean warming (Balogh and Byrne, 2020) and 52 acidification (Parker et al., 2012). Over longer time scales, increasing tolerance to these drivers may be 53 mediated by mechanisms such as transgenerational plasticity (Murray et al., 2014), leading to locally- 54 adapted genotypes, as seen in bivalves (Thomsen et al., 2017), annelids (Rodríguez-Romero et al., 2016; 55 Thibault et al., 2020), corals (Putnam et al., 2020), and coralline algae (Cornwall et al., 2020). However, 56 transgenerational plasticity is species-specific (Byrne et al., 2020; Thibault et al., 2020) and, depending on 57 the rate and magnitude of environmental change, it may either be insufficient for evolutionary rescue Do Not Cite, Quote or Distribute 3-36 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 (Morgan et al., 2020) or could induce maladaptive responses (i.e., reduced fitness) in marine organisms 2 exposed to multiple drivers (medium confidence, low evidence) (Figure 3.11, Griffith and Gobler, 2017; 3 Parker et al., 2017; Byrne et al., 2020). 4 5 Acclimation to environmental pressures and climate change via phenotypic plasticity (Section 3.3.3, Collins 6 et al., 2020) enables species to undergo niche shifts, such that their present-day climatic niche is altered to 7 incorporate new or shifted conditions (Fox et al., 2019). Although plasticity provides an adaptive 8 mechanism, it is unlikely to provide a long-term solution for species undergoing sustained directional 9 environmental change (e.g., global warming) (medium confidence) (Fox et al., 2019; Gaitán-Espitia and 10 Hobday, 2021). Beyond the limits for plastic responses (Figure 3.9, DeWitt et al., 1998; Valladares et al., 11 2007), genetic adjustments are required to persist in a changing world (Figure 3.11, Fox et al., 2019). The 12 ability of species and populations to undergo these adjustments (i.e., adaptive evolution) depends on 13 extrinsic factors including the rate and magnitude of environmental change (important determinants of the 14 strength and form of selection, Hoffmann and Sgrò, 2011; Munday et al., 2013), along with intrinsic factors 15 such as generation times and standing genetic variation (Mitchell-Olds et al., 2007; Lohbeck et al., 2012). 16 Accurately assessing the degree of acclimation and/or adaptation across space and time is difficult and 17 constrains studying adaptive evolution in natural populations. There is a major gap in climate-change 18 biology related to the study of evolutionary responses in complex and long-lived multicellular organisms. 19 Insights on organismal acclimation, adaptation, and evolution rely on studies of small, short-lived marine 20 organisms such as phytoplankton that divide rapidly and contain high genetic variation in large populations. 21 (Schaum et al., 2016; Cavicchioli et al., 2019; Collins et al., 2020). 22 23 Experimental evolution suggests that microbial populations can rapidly adapt (i.e., over 1­2 years) to 24 environmental changes mimicking projected effects of climate change (medium confidence). Phytoplankton 25 adaptive mechanisms include intraspecific strain sorting and genetic changes (Bach et al., 2018; Hoppe et al., 26 2018b; Wolf et al., 2019). The evolutionary responses of microbes are conditioned by the number and 27 characteristics of interacting drivers (low confidence) (Brennan et al., 2017). For example, in a high-salinity 28 adapted strain of the phytoplankton Chlamydomonas reinhardtii, the selection intensity and the adaptation 29 rate increased with the number of environmental drivers, accelerating the adaptive evolutionary response 30 (Brennan et al., 2017). For this and other phytoplankton species, a few dominant drivers explain most of the 31 phenotypic and evolutionary changes observed (Boyd et al., 2015a; Brennan and Collins, 2015; Brennan et 32 al., 2017). 33 34 Adaptation can be impeded, delayed or constrained in eukaryotic microbial populations as a result of reduced 35 genetic diversity, and/or the presence of functional and evolutionary trade-offs (Aranguren-Gassis et al., 36 2019; Lindberg and Collins, 2020; Walworth et al., 2020). In the marine diatom Chaetoceros simplex, a 37 functional trade-off between high-temperature tolerance and increased nitrogen requirements underlies 38 inhibited thermal adaptation under nitrogen-limited conditions (low confidence) (Aranguren-Gassis et al., 39 2019). When selection is strong due to unfavourable environmental conditions, microbial populations can 40 encounter functional and evolutionary trade-offs evidenced by reducing growth rates while increasing 41 tolerance and metabolism of reactive oxygen species (Lindberg and Collins, 2020). Other trade-offs can be 42 observed in offspring quality and number (Lindberg and Collins, 2020). These findings contribute towards a 43 mechanistic framework describing the range of evolutionary strategies in response to multiple drivers 44 (Collins et al., 2020), but other hazards such as extreme events (e.g., MHWs) still need to be included 45 because their characteristics may alter the potential for adaptation of species and populations to climate 46 change (Gruber et al., 2021). 47 48 Do Not Cite, Quote or Distribute 3-37 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure 3.11: Micro-evolutionary dynamics in response to environmental change. Simplified conceptual framework 3 shows two main eco-evolutionary trajectories for natural populations over time (vertical axis from top to bottom). If 4 environmental stress is low, rapid responses (within a generation) through plastic phenotypic adjustments and selection 5 (across generations) sustain fitness, enhancing maintenance of viable populations across generations. In contrast, if 6 environmental stress is high, ongoing phenotypic plasticity and acclimation may be insufficient to buffer the negative 7 effects, exacerbating the loss of fitness (change of colour to orange/yellow/red). Ultimately, very high stress conditions 8 accelerate population decline, enhancing the risk of species extinction. 9 10 11 3.3.5 Ecological Response to Multiple Drivers 12 13 Assessing ecological responses to multiple climate-impact drivers requires a combination of approaches, 14 including laboratory- and field-based experiments, field observations (e.g., natural gradients, climate 15 analogues), study of paleo-analogues and the development of mechanistic and empirical models (Clapham, 16 2019; Gissi et al., 2021). Experimental studies of food-web responses are often limited to an individual 17 driver, although recent manipulations have used a matrix of >1000-L mesocosms to explore ecological 18 responses to both warming and acidification (Box 3.1, Nagelkerken et al., 2020). Hence, complementary 19 approaches are needed to indirectly explore the mechanisms underlying ecosystem responses to global 20 climate change (Parmesan et al., 2013). Observations from time series longer than modes of natural 21 variability (i.e., decades) are essential for revealing and attributing ecological responses to climate change 22 (e.g., Section 3.4, Barton et al., 2015b; Brun et al., 2019). Also, paleo records provide insights into the 23 influence of multiple drivers on marine biota (Cross-Chapter Box PALEO in Chapter 1, Reddin et al., 2020). 24 Specifically, associations between vulnerabilities and traits of marine ectotherms in laboratory experiments 25 correspond with organismal responses to ancient hyperthermal events (medium confidence) (Reddin et al., 26 2020). This corroboration suggests that responses to multiple drivers inferred from the fossil record can help 27 provide insights into the future status of functional groups and hence food webs under rapid climate change. 28 Do Not Cite, Quote or Distribute 3-38 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Multi-species and integrated end-to-end ecosystem models are powerful tools to explore and project 2 outcomes to the often-interacting cumulative effects of climate change and other anthropogenic drivers 3 (Section 3.1, Kaplan and Marshall, 2016; Koenigstein et al., 2016; Peck and Pinnegar, 2018; Tittensor et al., 4 2018; Gissi et al., 2021). These models can integrate some aspects of the knowledge accrued from 5 manipulation experiments, paleo and contemporary observations, help test the relative importance of specific 6 drivers and driver combinations, and identify synergistic or antagonistic responses (Koenigstein et al., 2016; 7 Payne et al., 2016; Skogen et al., 2018; Tittensor et al., 2018). As these models are associated with wide- 8 ranging uncertainties (SM3.2.2, Payne et al., 2016; Trolle et al., 2019; Heneghan et al., 2021) they cannot be 9 expected to accurately project the trajectories of complex marine ecosystems under climate change. Hence, 10 they are most useful for assessing overall trends and in particular for providing a plausible envelope of 11 trajectories across a range of assumptions (Fulton et al., 2018; Peck et al., 2018; Tittensor et al., 2018). On a 12 global scale, ecosystem models project a ­5.7% ± 4.1% (very likely range) to ­15.5% ± 8.5% decline in 13 marine animal biomass with warming under SSP1-2.6 and SSP5-8.5, respectively, by 2080­2099 relative to 14 1995­2014, albeit with significant regional variation in both trends and uncertainties (medium confidence) 15 (Section 3.4.34, Tittensor et al., 2021). Biological interactions may exacerbate or buffer the projected 16 impacts. For instance, trophic amplification (strengthening of responses to climate-impact drivers at higher 17 trophic levels), may result from combined direct and indirect food-web mediated effects (medium 18 confidence) (Section 3.4.3.4, Lotze et al., 2019). Alternatively, compensatory species interactions can 19 dampen strong impacts on species from ocean acidification, resulting in weaker responses at functional- 20 group or community level than at species level (medium confidence) (Marshall et al., 2017; Hoppe et al., 21 2018b; Olsen et al., 2018; Gissi et al., 2021). Globally, the projected reduction of biomass due to climate- 22 impact drivers is relatively unaffected by fishing pressure, indicating additive responses of fisheries and 23 climate change (low confidence) (Lotze et al., 2019). Regionally, projected interactions of climate-impact 24 drivers, fisheries and other regional non-climate drivers can be both synergistic and antagonistic, varying 25 across regions, functional groups and species, and can cause non-linear dynamics with counterintuitive 26 outcomes, underlining the importance of adaptations and associated trade-offs (high confidence) (Sections 27 3.5.3, 3.6.3.1.2, 4.5, 4.6, Weijerman et al., 2015; Fulton et al., 2018; Hansen et al., 2019; Trolle et al., 2019; 28 Zeng et al., 2019; Holsman et al., 2020; Pethybridge et al., 2020; Gissi et al., 2021). 29 30 Given the limitations of individual ecological models discussed above, model intercomparisons, such as the 31 Fisheries and Marine Ecosystem Model Intercomparison Project (Fish-MIP, Tittensor et al., 2018) show 32 promise in increasing the robustness of projected ecological outcomes (Tittensor et al., 2018). Model 33 ensembles include a greater number of relevant processes and functional groups than any single model and 34 thus capture a wider range of plausible responses. Among the global Fish-MIP models, there is high 35 (temperate and tropical areas) to medium agreement (coastal and polar regions) on the direction of change, 36 but medium (temperate and tropical regions) to low agreement (coastal and polar regions) on magnitude of 37 change (Lotze et al., 2019; Heneghan et al., 2021). Although model outputs are validated relative to 38 observations to assess model skills (Payne et al., 2016; Tittensor et al., 2018), the Fish-MIP models under- 39 represent some sources of uncertainty, as they often do not include parameter uncertainties, and do not 40 usually include impacts of ocean acidification, oxygen loss, or evolutionary responses because there remains 41 high uncertainty regarding the influences of these processes across functional groups. Ensemble model 42 investigations like Fish-MIP have also identified gaps in our mechanistic understanding of ecosystems and 43 their responses to anthropogenic forcing, leading to model improvement and more rigorous benchmarking. 44 These investigations could inspire future targeted observational and experimental research to test the validity 45 of model assumptions (Payne et al., 2016; Lotze et al., 2019; Heneghan et al., 2021). The state-of-the-art in 46 such experimental research is presented in Box 3.1. 47 48 49 [START BOX 3.1 HERE] 50 51 Box 3.1: Challenges for Multiple-Driver Research in Ecology and Evolution 52 53 The majority of the examples in Section 3.3 are from studies mimicking projected conditions in the year 54 2100 that report the responses of an individual species or strain to multiple drivers. This powerful generic 55 experimental approach has largely been restricted to single species because it is logistically complex to 56 conduct experiments that straddle multiple trophic levels, and that also include more than two drivers (Figure 57 Box3.1.1b); the need for multiple replicates, drivers, and treatment levels greatly increase the work required Do Not Cite, Quote or Distribute 3-39 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 (Parmesan et al., 2013; Boyd et al., 2018). It is challenging to apply this experimental approach to 2 communities or ecosystems (Figure Box 3.1.1). To date, most research on community or ecosystem response 3 to climate-impact drivers has been in large-volume (>10,000 L) mesocosms (Riebesell and Gattuso, 2014), 4 or at natural analogues such as CO2 seeps, in which only one driver (ocean acidification) is altered (see (4) in 5 Figure Box 3.1.1). Only very recently have two drivers been incorporated into climate-change manipulation 6 studies examining responses of primary producers to secondary consumers ((5) in Figure Box3.1.1a, 7 Nagelkerken et al., 2020). Therefore, `natural experiments' from the geological past (Reddin et al., 2020) 8 provide insights into how food webs and their constituents respond to complex change involving multiple 9 drivers. Contemporary observations are occasionally long enough (>5 decades) to capture community 10 responses to complex climate change. For example, Brun et al. (2019) reported a shift in zooplankton 11 community structure in the North Atlantic (1960­2014), with major biogeochemical ramifications. 12 13 Conducting sufficiently long manipulation experiments to study the effect of adaptation on organisms is 14 equally difficult (Figure Box3.1.1b), with much research restricted to multi-year studies of the 15 microevolution of fast-growing (>one division day­1) phytoplankton species responding to single drivers 16 (Lohbeck et al., 2012; Schaum et al., 2016). In a few experimental evolution studies, ((7) in Figure 17 Box3.1.1a, Brennan et al., 2017), multiple drivers have been used, but none have used communities or 18 ecosystems (Figure Box3.1.1b). Nevertheless, the fossil record provides limited evidence of adaptations to 19 less rapid (relative to present-day) climate change (Jackson et al., 2018). Despite the need to explore 20 ecological or biogeochemical responses to projected future ocean conditions, logistical challenges require 21 that assessments of climate-change impacts at scales larger than mesocosms use large-scale, long-term in situ 22 observational studies (as documented within Section 3.4). 23 24 25 26 Figure Box 3.1.1: Knowledge gaps between current scientific understanding and that needed to inform policy. The 27 conceptual space relating driver number, (Driver axis), ecological organisation (Space axis) and evolutionary 28 acclimation state (Time axis), modified from Riebesell and Gattuso (2014). (a) Spheres indicate suites of studies that 29 illustrate the progress of research, including multiple drivers (Sphere (1) one species and one driver, Hutchins et al. 30 (2013) and (2) (one species and multiple drivers five, Boyd et al. (2015a)); ecology (Sphere (1) (one driver, one 31 species), (3) (one driver, planktonic community, Moustaka-Gouni et al., 2016), (4) one driver (high-CO2 seep) and 32 (benthic) ecosystem, Fabricius et al. (2014), and (5) two drivers and nearshore ecosystem, Nagelkerken et al. (2020)); 33 and evolution (Sphere (1) (acclimated organism and one driver), (6) adapted organisms and one driver, Listmann et al. 34 (2016) and (7) adapted organism and multiple drivers Brennan et al. (2017)). (b) Trends in research trajectories since 35 2000 from a survey of 171 studies, Boyd et al. (2018). Note the dominance of multiple-driver experiments at the species 36 level (lower left cluster); the focus on acclimation (red triangle) rather than adaptation (blue dot); the focus of 37 investigation on 3 drivers. Redrawn from Boyd et al. (2018). 38 39 40 [END BOX 3.1 HERE] 41 42 Do Not Cite, Quote or Distribute 3-40 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 3.4 Observed and Projected Impacts of Climate Change on Marine Systems 2 3 3.4.1 Introduction 4 5 Ocean and coastal ecosystems and their resident species are under increasing pressure from a multitude of 6 climate-impact drivers and non-climate drivers (Section 3.1, Figure 3.12, Bindoff et al., 2019). This section 7 builds from the assessment of biological responses to climate-impact drivers (Section 3.3) to examine the 8 new evidence about climate-change impacts at the level of marine ecosystems. It focuses on detection and 9 attribution of observed changes to marine ecosystems and the projected changes under different future 10 climate scenarios. This assessment considers emerging evidence on the effects of multiple non-climate 11 drivers and physiological acclimation and/or evolutionary adaptation on these observations and projections. 12 13 The section focuses first on coastal ecosystems and seas (Section 3.4.2), which have high spatial variability 14 in physical and chemical characteristics, are affected by many non-climate drivers (Section 3.1, Figure 3.12) 15 and support rich fisheries, high biodiversity and high levels of species endemism. The assessment begins 16 with warm-water coral reefs (Section 3.4.2.1) because these highly threatened systems are at the vanguard of 17 research on acclimation and evolutionary adaptation among coastal ecosystems. It follows with the other 18 shallow, nearshore ecosystems dominated by habitat-forming species (rocky shores, kelp systems) and then 19 nearshore sedimentary systems (estuaries, deltas, coastal wetlands, and sandy beaches), before moving on to 20 semi-enclosed seas, shelf seas, upwelling zones, and polar seas. 21 22 The section continues on to oceanic and cross-cutting changes (Section 3.4.3), which influence large areas of 23 the epipelagic zone (<200 m depth), while also affecting the mesopelagic (200­1000 m), the perpetually dark 24 bathypelagic (depth >1000 m) and the deep seafloor (benthic ecosystems at depths >200 m) zones. Assessed 25 in this section are species range shifts (Section 3.4.3.1), phenological shifts and trophic mismatches (Section 26 3.4.3.2), changes in communities and biodiversity (Section 3.4.3.3.2), time of emergence of climate-impact 27 signals in ecological systems from background natural variability (Section 3.4.3.3.4), and changes in 28 biomass, primary productivity, and carbon export (Sections 3.4.3.4­3.4.3.6). 29 30 31 32 Figure 3.12: Summary assessment of observed hazards to coastal ecosystems and seas as assessed in Section 3.4.2. Do Not Cite, Quote or Distribute 3-41 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 3 3.4.2 Coastal Ecosystems and Seas 4 5 3.4.2.1 Warm-Water Coral Reefs 6 7 Warm-water coral reef ecosystems house one-quarter of the marine biodiversity and provide services in the 8 form of food, income and shoreline protection to coastal communities around the world. These ecosystems 9 are threatened by climate and non-climate drivers, especially ocean warming, MHWs, ocean acidification, 10 SLR, tropical cyclones, fisheries/overharvesting, land-based pollution, disease spread and destructive 11 shoreline practices (Hoegh-Guldberg et al., 2018a; Bindoff et al., 2019; Hughes et al., 2020). Warm-water 12 coral reefs face near-term threats to their survival (Table 3.3), but research on observed and projected 13 impacts is very advanced. 14 15 16 Table 3.3: Summary of previous IPCC assessments of coral reefs. Observations Projections AR5 (Hoegh-Guldberg et al., 2014; Wong et al., 2014) Coral reefs are one of the most vulnerable marine Coral bleaching and mortality will increase in frequency ecosystems (high confidence), and more than half of the and magnitude over the next decades (very high world's reefs are under medium or high risk of confidence). Analysis of the Coupled Model degradation. Intercomparison Project 5 ensemble projects the loss of coral reefs from most sites globally by 2050 under mid to Mass coral bleaching and mortality, triggered by positive high rates of warming (very likely). temperature anomalies (high confidence), is the most widespread and conspicuous impact of climate change. Under the A1B scenario, 99% of the reef locations will Ocean acidification reduces biodiversity and the experience at least one severe bleaching event between calcification rate of corals (high confidence) while at the 2090 and 2099, with limited evidence and low agreement same time increasing the rate of dissolution of the reef that coral acclimation and/or adaptation will limit this framework (medium confidence). trend. In summary, ocean warming is the primary cause of mass The onset of global dissolution of coral reefs is at an coral bleaching and mortality (very high confidence), atmospheric CO2 of 560 ppm (medium confidence) and which, together with ocean acidification, deteriorates the dissolution will be widespread in 2100 (Representative balance between coral reef construction and erosion (high Concentration Pathway (RCP)8.5, medium confidence). confidence). A number of coral reefs could keep up with the maximum rate of sea-level rise (SLR) of 15.1 mm yr­1 projected for the end of the century, but lower net accretion and increased turbidity will weaken this capability (very high confidence). SR15 (Hoegh-Guldberg et al., 2018a; IPCC, 2019c) Climate change has emerged as the greatest threat to coral Multiple lines of evidence indicate that the majority (70­ reefs, with temperatures of just 1°C above the 1985­1993 90%) of warm water (tropical) coral reefs that exist today long-term summer maximum for an area over 4­6 weeks will disappear even if global warming is constrained to being enough to cause mass coral bleaching and mortality 1.5°C (very high confidence). (very high confidence). Coral reefs, for example, are projected to decline by a Predictions of back-to-back bleaching events have become further 70­90% at 1.5°C (high confidence) with larger reality over 2015-2017 as have projections of declining losses (>99%) at 2ºC (very high confidence). coral abundance (high confidence). SROCC (Bindoff et al., 2019) New evidence since AR5 and SR15 confirms the impacts Coral reefs will face very high risk at temperatures 1.5ºC of ocean warming and acidification on coral reefs (high of global sea surface warming (very high confidence). confidence), enhancing reef dissolution and bioerosion Almost all coral reefs will degrade from their current state, (high confidence), affecting coral species distribution, and even if global warming remains below 2ºC (very high leading to community changes (high confidence). The rate confidence), and the remaining shallow coral reef Do Not Cite, Quote or Distribute 3-42 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report of SLR (primarily noticed in small reef islands) may communities will differ in species composition and outpace the growth of reefs to keep up, although there is diversity from present reefs (very high confidence). This low agreement in the literature (low confidence). will greatly diminish the services they provide to society, such as food provision (high confidence), coastal Reefs are further exposed to other increased impacts, such protection (high confidence) and tourism (medium as enhanced storm intensity, turbidity and increased runoff confidence). from the land (high confidence). Recovery of coral reefs resulting from repeated disturbance events is slow (high The very high vulnerability of coral reefs to warming, confidence). Only few coral reef areas show some ocean acidification, increasing storm intensity and SLR resilience to global change drivers (low confidence). under climate, including enhanced bioerosion (high confidence), points to the importance of considering both mitigation and adaptation. 1 2 3 Global analyses published since AR5 show that mass coral bleaching events and disease outbreaks have 4 increased due to more frequent and severe heat stress associated with ocean warming (very high confidence, 5 virtually certain) (Donner et al., 2017; Hughes et al., 2018a; DeCarlo et al., 2019; Sully et al., 2019; Tracy et 6 al., 2019). The mass coral bleaching, which occurred continuously across different parts of the tropics from 7 2014­2016, is considered the longest and most severe global coral bleaching event on record (Section 10.4.3, 8 Box 15.2, Eakin et al., 2019). The Great Barrier Reef underwent mass bleaching three times between 2016­ 9 2020 (Box 11.2, Pratchett et al., 2021), validating past model projections that some warm-water coral reefs 10 would encounter bleaching-level heat stress multiple times per decade by the 2020s (Hoegh-Guldberg, 1999; 11 Donner, 2009). 12 13 Heat stress and mass bleaching events caused decreases in live coral cover (virtually certain) (Graham et al., 14 2014; Hughes et al., 2018b), loss of sensitive species (extremely likely) (Donner and Carilli, 2019; Lange and 15 Perry, 2019; Toth et al., 2019; Courtney et al., 2020), vulnerability to disease (extremely likely) (van Woesik 16 and Randall, 2017; Hadaidi et al., 2018; Brodnicke et al., 2019; Howells et al., 2020) and declines in coral 17 recruitment in the tropics (medium confidence) (Hughes et al., 2019; Price et al., 2019). Recent observations 18 also suggest that excess nutrients can increase the susceptibility of corals to heat stress (DeCarlo et al., 19 2020). Changes in coral community structure due to bleaching have caused declines in reef carbonate 20 production (high confidence) (Perry and Morgan, 2017; Lange and Perry, 2019; Perry and Alvarez-Filip, 21 2019; Courtney et al., 2020; van Woesik and Cacciapaglia, 2021) and in reef structural complexity (high 22 confidence, very likely) (Couch et al., 2017; Leggat et al., 2019; Magel et al., 2019), which increases water 23 depth, reduces wave attenuation and increases coastal flood risk (Yates et al., 2017; Beck et al., 2018). 24 Corals may also lose reproductive synchrony through climate change (Shlesinger and Loya, 2019), adding to 25 their vulnerability. Bleaching and other drivers promote phase shifts to ecosystems dominated by macroalgae 26 or other stress-tolerant species (very high confidence) (Graham et al., 2015; Stuart-Smith et al., 2018), 27 leading to changes in reef-fish species assemblages (high confidence) (Richardson et al., 2018; Robinson et 28 al., 2019a; Stuart-Smith et al., 2021). 29 30 Ocean acidification and associated declines in aragonite saturation state (aragonite) decrease rates of 31 calcification by corals and other calcifying reef organisms (very high confidence), reduce coral settlement 32 (medium confidence) and increase bioerosion and dissolution of reef substrates (high confidence) (Hoegh- 33 Guldberg et al., 2018a; Bindoff et al., 2019; Kline et al., 2019; Pitts et al., 2020). Warming can exacerbate 34 the coral response to ocean acidification (Kornder et al., 2018) and accelerate the decrease in coral skeletal 35 density (Guo et al., 2020). In addition, reefs with lower coral cover and a higher proportion of slow-growing 36 species, because of bleaching, are more sensitive to acidification (net dissolution occurs aragonite = 2.3 for 37 100% coral cover, and aragonite >3.5 for 30% coral cover, (Kline et al., 2019)). However, experimental 38 evidence suggests that coral responses to ocean acidification are species-specific (medium confidence) 39 (Fabricius et al., 2011; DeCarlo et al., 2018; Comeau et al., 2019). Evidence from experiments suggests that 40 crustose coralline algae, which contribute to reef structure and integrity and may be resistant to warming at 41 the RCP8.5 level by 2100 (Cornwall et al., 2019), are also sensitive to declines in aragonite (high confidence) 42 (Section 3.4.2.3, Fabricius et al., 2015; Smith et al., 2020). The integrated effect of acidification, bleaching, 43 storms and other non-climate drivers on corals, coralline algae and other calcifiers can further compromise 44 reef integrity and ecosystem services (Rivest et al., 2017; Cornwall et al., 2018; Perry and Alvarez-Filip, 45 2019). 46 Do Not Cite, Quote or Distribute 3-43 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Since SROCC, there have been advances in experimental, field and modelling research on the projected 2 response of coral cover and reef growth to bleaching and ocean acidification (Cziesielski et al., 2019; 3 Morikawa and Palumbi, 2019; Cornwall et al., 2021; Klein et al., 2021; Logan et al., 2021; McManus et al., 4 2021), and on the effect of possible human interventions like assisted evolution on coral resilience (Section 5 3.6.3.2.2, Condie et al., 2021; Hafezi et al., 2021; Kleypas et al., 2021). New model projections incorporating 6 physiological acclimation, larval dispersal, and evolutionary processes find limited ability to adapt this 7 century at rates of warming at or exceeding that in RCP4.5 (high confidence, very likely) (Bay et al., 2017; 8 Kubicek et al., 2019; Matz et al., 2020; McManus et al., 2020; Logan et al., 2021; McManus et al., 2021). 9 For example, a global analysis (Logan et al., 2021) finds that increased thermal tolerance via evolution or 10 switching to more stress-tolerant algal symbionts enable most (73­81%) coral to survive through 2100 under 11 RCP2.6, but coral-dominated communities with a historical mix of coral taxa still disappear (0­8% coral 12 survival) under RCP6.0 in simulations with adaptive mechanisms (Figure 3.13). Due to the impacts of 13 warming, and to a lesser extent ocean acidification, global reef carbonate production is estimated to decline 14 71% by 2050 in SSP1-2.6, and the rate of SLR is estimated to exceed that of reef growth for 97% of reefs 15 assessed, absent adaptation by corals and their symbionts (WGI AR6 Table 9.9, Cornwall et al., 2021; Fox- 16 Kemper et al., 2021). The increased water depth due to coral loss and reef erosion, as well as reduced 17 structural complexity, will limit wave attenuation and exacerbate the risk of flooding from SLR on reef- 18 fringed shorelines and reef islands (Yates et al., 2017; Beck et al., 2018; Harris et al., 2018). Local coral reef 19 fish species richness is projected to decline due to the impacts of warming on coral cover and diversity (high 20 confidence), with declines up to 40% by 2060 in SSP5-8.5 (Strona et al., 2021). 21 22 These observed and projected impacts are supported by geological and paleo-ecological evidence showing a 23 decline in coral reef extent and species richness under previous episodes of climate change and ocean 24 acidification (Kiessling and Simpson, 2011; Pandolfi et al., 2011; Kiessling et al., 2012; Pandolfi and 25 Kiessling, 2014; Kiessling and Kocsis, 2015). Major reef crises in the past 300 million years were governed 26 by hyperthermal events (medium confidence) (Section 3.2.4.4, Cross-Chapter Box PALEO in Chapter 1) 27 longer in timescale than anthropogenic climate change, during which net coral reef accretion was more 28 strongly affected than biodiversity (medium confidence). 29 30 In response to the global-scale decline in coral reefs and high future risk, recent literature focuses on finding 31 thermal refuges and identifying uniquely resilient species, populations or reefs for targeted restoration and 32 management (Hoegh-Guldberg et al., 2018b). Reefs exposed to internal waves (Storlazzi et al., 2020), 33 turbidity (Sully and van Woesik, 2020) or warm-season cloudiness (Gonzalez-Espinosa and Donner, 2021) 34 are expected to be less sensitive to thermal stress. Mesophotic reefs (30­150 m) have also been proposed as 35 thermal refugia (Bongaerts et al., 2010), although evidence from recent bleaching events, subsurface 36 temperature records, and species overlap is mixed (Frade et al., 2018; Rocha et al., 2018b; Eakin et al., 2019; 37 Venegas et al., 2019; Wyatt et al., 2020). A study of 2584 reef sites across the Indian and Pacific Oceans 38 estimated that 17% had sufficient cover of framework-building corals to warrant protection, 54% required 39 recovery efforts, and 28% were on a path to net erosion (Darling et al., 2019). There is medium evidence for 40 greater bleaching resistance among reefs subject to temperature variability or frequent heat stress (Barkley et 41 al., 2018; Gintert et al., 2018; Hughes et al., 2018a; Morikawa and Palumbi, 2019), but with trade-offs in 42 terms of diversity and structural complexity (Donner and Carilli, 2019; Magel et al., 2019). There is limited 43 agreement about the persistence of thermal tolerance in response to severe heat stress (Le Nohaïc et al., 44 2017; DeCarlo et al., 2019; Fordyce et al., 2019; Leggat et al., 2019; Schoepf et al., 2020). Recovery and 45 restoration efforts that target heat-resistant coral populations and culture heat-tolerant algal symbionts have 46 the greatest potential of effectiveness under future warming (high confidence) (Box 5.5 in SROCC Chapter 47 5, Bay et al., 2017; Darling and Côté, 2018; Baums et al., 2019; Bindoff et al., 2019; Howells et al., 2021); 48 however, there is low confidence that enhanced thermal tolerance can be sustained over time (Section 49 3.6.3.3.2, Buerger et al., 2020). The effectiveness of active restoration and other specific interventions (e.g., 50 reef shading) are further assessed in Section 3.6.3.3.2. 51 52 In sum, additional evidence since SROCC and SR15 (Table 3.3) finds that living coral and reef growth are 53 declining due to warming and MHWs (very high confidence). Coral reefs are under threat of transitioning to 54 net erosion with >1.5ºC of global warming (high confidence), with impacts expected to occur fastest in the 55 Atlantic Ocean. The effectiveness of conservation efforts to sustain living coral area, coral diversity, and reef 56 growth is limited for the majority of the world's reefs with >1.5ºC of global warming (high confidence) 57 (Section 3.6.3.3.2, Hoegh-Guldberg et al., 2018b; Bruno et al., 2019; Darling et al., 2019). Do Not Cite, Quote or Distribute 3-44 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 3 4 Figure 3.13: Coral reef futures, with and without adaptation. Graphs are based on a model of coral-symbiont 5 evolutionary dynamics from (Logan et al., 2021), which simulates two coral types and symbiont populations for 1925 6 reef cells worldwide, from 1950­2100 drawn from simulations with National Oceanic and Atmospheric 7 Administration-Geophysical Fluid Dynamics Laboratory Earth System Model (ESM2M) under four RCPs. Top panels 8 show the simulated fraction of cells with healthy reefs, when both coral types are not in a state of severe bleaching or 9 mortality, (i) without adaptive responses and (ii) with adaptive responses (symbiont evolution). Colours indicate 10 maximum monthly sea-surface temperature increase across all reef cells, versus a 1861­2010 baseline. Panels (a,b,c) 11 depict snapshots of coral reef conditions at time-points in the future each with different levels of warming, drawn from 12 the model-projected cover of the two coral types, and from a literature assessment (Section 3.4.2.1, Hughes et al., 13 2018b; Bindoff et al., 2019; Darling et al., 2019; Leggat et al., 2019; Cornwall et al., 2021). 14 15 16 3.4.2.2 Rocky Shores 17 18 Rocky shore ecosystems refer to a range of temperate intertidal and shallow coastal ecosystems that are 19 dominated by different foundational organisms, including mussels, oysters, fleshy macroalgae, hard and soft 20 corals, coralline algae, bryozoans and sponges, which create habitat for species-rich assemblages of 21 invertebrates, fish, marine mammals and other organisms. Rocky shores provide services including wave 22 attenuation, habitat provision and food resources, and these support commercial, recreational, and 23 Indigenous fisheries and shellfish aquaculture. 24 25 Observations since AR5 and SROCC (Table 3.4) find increased impacts of ocean warming on rocky shores. 26 This includes extirpation of species at the warm edge of their ranges (Yeruham et al., 2015; Martínez et al., 27 2018), extension of poleward range boundaries (Sanford et al., 2019), mortality from climate extremes 28 (Seuront et al., 2019), reduction in survival at shallower depths (Sorte et al., 2019; Wallingford and Sorte, 29 2019) and reorganisation of communities (Wilson et al., 2019; Mulders and Wernberg, 2020; Albano et al., 30 2021). Data collected after MHWs find ecological phase shifts (moderate evidence, high agreement) (e.g., 31 California (Rogers-Bennett and Catton, 2019; McPherson et al., 2021)) and homogenisation of communities 32 (limited evidence) (e.g., Alaska, (Weitzman et al., 2021)). For example, the collapse of sea star populations 33 in the Northeast Pacific due to a MHW-related disease outbreak (Hewson et al., 2014; Menge et al., 2016; 34 Miner et al., 2018; Schiebelhut et al., 2018), including 80­100% loss of the common predatory sunflower Do Not Cite, Quote or Distribute 3-45 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 star, Pycnopodia helianthoides (very high confidence) (Harvell et al., 2019), triggered shifts from kelp- to 2 urchin-dominated ecosystems (Schultz et al., 2016; Gravem and Morgan, 2017; McPherson et al., 2021). 3 4 5 Table 3.4: Summary of previous IPCC assessments of rocky shores. Observations Projections AR5: (Wong et al., 2014) The abundance and distribution of rocky shore species Rocky shores are among the better-understood coastal will continue to change in a warming world (high ecosystems in terms of potential impacts of climate confidence). For example, the long-term consequences of variability and change. The most prominent effects are ocean warming on mussel beds of the northeast Pacific range shifts of species in response to ocean warming are both positive (increased growth) and negative (high confidence) and changes in species distribution and (increased susceptibility to stress and of exposure to abundance (high confidence) mostly in relation to ocean predation) (medium confidence). warming and acidification. The dramatic decline of biodiversity in mussel beds of Observations performed near natural CO2 vents in the the Californian coast has been attributed to large-scale Mediterranean Sea show that diversity, biomass, and processes associated with climate-related drivers (high trophic complexity of rocky shore communities will confidence). decrease at future pH levels (high confidence). SR15 (Hoegh-Guldberg et al., 2018a) In the transition to 1.5°C, changes to water temperatures Changes in ocean circulation can have profound impacts will drive some species (e.g., plankton, fish) to relocate on temperate marine ecosystems by connecting regions to higher latitudes and for novel ecosystems to appear and facilitating the entry and establishment of species in (high confidence). Other ecosystems (e.g., kelp forests, areas where they were unknown before coral reefs) are relatively less able to move, however, (`tropicalization') as well as the arrival of novel disease and are projected to experience high rates of mortality agents (medium agreement, limited evidence). and loss (very high confidence). In the case of `less mobile' ecosystems (e.g., coral reefs, kelp forests, intertidal communities), shifts in biogeographical ranges may be limited, with mass mortalities and disease outbreaks increasing in frequency as the exposure to extreme temperatures have increased (high agreement, robust evidence) SROCC (Bindoff et al., 2019) Intertidal rocky shores ecosystems are highly sensitive to Intertidal rocky shores are also expected to be at very ocean warming, acidification and extreme heat exposure high risk (transition above 3ºC) under the RCP8.5 during low tide emersion (high confidence). scenario (medium confidence). These ecosystems have low to moderate adaptive capacity, as they are highly Sessile calcified organisms (e.g., barnacles and mussels) sensitive to ocean temperatures and acidification. in intertidal rocky shores are highly sensitive to extreme temperature events and acidification (high confidence), a Benthic species will continue to relocate in the intertidal reduction in their biodiversity and abundance have been zones and experience mass mortality events due to observed in naturally acidified rocky reef ecosystems warming (high confidence). Interactive effects between (medium confidence). acidification and warming will exacerbate the negative impacts on rocky shore communities, causing a shift towards a less diverse ecosystem in terms of species richness and complexity, increasingly dominated by macroalgae (high confidence). 6 7 8 Multiple lines of evidence find that foundational calcifying organisms such as mussels are at high risk of 9 decline due to both the individual and synergistic effects of warming, acidification and hypoxia (high 10 confidence) (Sunday et al., 2016; Sorte et al., 2017; Sorte et al., 2019; Newcomb et al., 2020). Warmer 11 temperatures reduce mussel and barnacle recruitment (e.g., NW Atlantic (Petraitis and Dudgeon, 2020)) and 12 the upper vertical limit of mussels (e.g., NE Pacific (Harley, 2011)) and (SW Pacific (Sorte et al., 2019)). 13 Experiments show that ocean acidification negatively impacts mussel physiology (very high confidence), 14 with evidence of reduced growth (Gazeau et al., 2010), attachment (Newcomb et al., 2020), 15 biomineralisation (Fitzer et al., 2014) and shell thickness (Pfister et al., 2016; McCoy et al., 2018). Net Do Not Cite, Quote or Distribute 3-46 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 calcification and abundance of mussels and other foundational species including oysters are expected to 2 decline due to ocean acidification (very high confidence) (Kwiatkowski et al., 2016; Sunday et al., 2016; 3 McCoy et al., 2018; Meng et al., 2018), causing the reorganisation of communities (high confidence) 4 (Kroeker et al., 2013b; Linares et al., 2015; Brown et al., 2016; Sunday et al., 2016; Agostini et al., 2018; 5 Teixidó et al., 2018). Experiments indicate that acidification can interact with warming and hypoxia to 6 increase the detrimental effects on mussels (Gu et al., 2019; Newcomb et al., 2020). In regions where food is 7 readily available to mussels, detrimental effects of ocean acidification may be dampened (Kroeker et al., 8 2016); however, recent findings are inconclusive (Brown et al., 2018a). 9 10 Coralline algae, foundational taxa that create habitat for sea urchins and abalone, form rhodolith beds in 11 temperate to Arctic habitats, and bind together substrates, are expected to be highly susceptible to ocean 12 acidification because they precipitate soluble magnesium calcite (Kuffner et al., 2008; Williams et al., 2021). 13 Damage from acidification varies among species and regions, and can be due to direct physiological stress 14 (Marchini et al., 2019) or interactions with non-calcifying competitors such as fleshy macroalgae (Smith et 15 al., 2020). Experiments indicate that warming reduces calcification by coralline algae (high confidence) 16 (Cornwall et al., 2019) and exacerbates the effect of acidification (Kim et al., 2020; Williams et al., 2021). 17 18 In contrast to warm-water coral reefs, there are no regional or global numerical models of rocky shore 19 ecosystem response to projected climate change and acidification. Experiments suggest that existing genetic 20 variation could be sufficient for some mussels (Bitter et al., 2019) and coralline algae (Cornwall et al., 2020) 21 to adapt over generations to ocean acidification. Populations exposed to variable environments often have a 22 greater capacity for phenotypic plasticity and resilience to environmental change (e.g., urchins (Gaitán- 23 Espitia et al., 2017b) and coralline algae (Section 3.3.2, Rivest et al., 2017; Cornwall et al., 2018)). Although 24 parental conditioning within and across generations is an acclimatisation mechanism to global change, there 25 is limited evidence from experimental studies that this is applicable for marine invertebrates on rocky shores 26 (Byrne et al., 2020). 27 28 This assessment concludes that MHWs, attributable to climate change (Section 3.2.2.1), can cause fatal 29 disease outbreaks or mass mortality among some key foundational species (high confidence) and contribute 30 to ecological phase shifts (medium confidence). The upper vertical limits of some species will also be 31 constrained by climate change (high confidence). Experimental evidence since previous assessments further 32 indicates that acidification decreases abundance and richness of calcifying species (high confidence), 33 although there is limited evidence for acclimation in some species. Synergistic effects of warming and 34 acidification will promote shifts toward macroalgal dominance in some ecosystems (medium confidence) and 35 lead to reorganisation of communities (medium confidence). 36 37 3.4.2.3 Kelp Ecosystems 38 39 Kelp are temperate, habitat-forming marine macroalgae or seaweeds, mostly of the order Laminariales, 40 which extend across one quarter of the world's coastlines (Assis et al., 2020; Jayathilake and Costello, 2020). 41 The perennial species form dense underwater forest canopies and three-dimensional habitat that provides 42 refuge for fish, crustaceans, invertebrates and marine mammals (Filbee-Dexter et al., 2016; Wernberg et al., 43 2019). Kelp ecosystems support fisheries, aquaculture, fertiliser, and food provision, including for local and 44 Indigenous Peoples, along with regulating services in the form of wave attenuation and habitat provision. 45 Kelp aquaculture can also buffer against local acidification (Xiao et al., 2021) and contribute to carbon 46 storage (Froehlich et al., 2019). 47 48 Recent research (Straub et al., 2019; Butler et al., 2020; Filbee-Dexter et al., 2020b; Tait et al., 2021) 49 supports the findings of previous assessments (Table 3.5) that kelp and other seaweeds in most regions are 50 undergoing mass mortalities from high temperature extremes and range shifts from warming (very high 51 confidence). Kelp are highly sensitive to the direct effect of high temperature on survival (Nepper-Davidsen 52 et al., 2019) and indirect impact of temperature on herbivorous species (Ling, 2008; Vergés et al., 2016), 53 upwelling and nutrient availability (Carr and Reed, 2015; Schiel and Foster, 2015). Synergies between 54 warming, storms, pollution and intensified herbivory (due to removal or loss of predators including sea stars 55 and otters that constrain herbivory by fish and urchin populations) can also cause physiological stress and 56 physical damage in kelp, reducing productivity and reproduction (Rogers-Bennett and Catton, 2019; Beas- 57 Luna et al., 2020; McPherson et al., 2021). Do Not Cite, Quote or Distribute 3-47 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 3 Table 3.5: Summary of previous IPCC assessments of kelp ecosystems. Observations Projections AR5: (Wong et al., 2014) Kelp forests have been reported to decline in temperate Kelp ecosystems will decline with the increased frequency areas in both hemispheres, a loss involving climate change of heatwaves and sea temperature extremes as well as (high confidence). Decline in kelp populations attributed to through the impact of invasive subtropical species (high ocean warming has been reported in southern Australia and confidence). the North Coast of Spain. Climate change will contribute to the continued decline in the extent of kelps in the temperate zone (medium confidence) and the range of kelp in the Northern Hemisphere will expand poleward (high confidence). SR15 (Hoegh-Guldberg et al., 2018a) In the transition to 1.5°C, changes to water temperatures Observed movement of kelp ecosystems not assessed. will drive some species (e.g., plankton, fish) to relocate to higher latitudes and for novel ecosystems to appear (high confidence). Other ecosystems (e.g., kelp forests, coral reefs) are relatively less able to move, however, and will experience high rates of mortality and loss (very high confidence). SROCC (Bindoff et al., 2019) Kelp forests have experienced large-scale habitat loss and Kelp forests will face moderate to high risk at temperatures degradation of ecosystem structure and functioning over above 1.5ºC global sea surface warming (high confidence). the past half century, implying a moderate to high level of risk at present conditions of global warming (high Due to their low capacity to relocate and high sensitivity to confidence). warming, kelp forests are projected to experience higher frequency of mass mortality events as the exposure to The abundance of kelp forests has decreased at a rate of extreme temperature rises (very high confidence). ~2% yr­1 over the past half century, mainly due to ocean warming and marine heatwaves, as well as from other Changes in ocean currents have facilitated the entry of human stressors (high confidence). tropical herbivorous fish into temperate kelp forests decreasing their distribution and abundance (medium Changes in ocean currents have facilitated the entry of confidence). tropical herbivorous fish into temperate kelp forests decreasing their distribution and abundance (medium Kelp forests at low latitudes will continue to retreat due to confidence). intensified extreme temperatures, and their low dispersal ability will elevate the risk of local extinction under The loss of kelp forests is followed by the colonisation of RCP8.5 (high confidence). turfs, which contributes to the reduction in habitat complexity, carbon storage and diversity (high confidence). 4 5 6 Trends in kelp abundance since 1950 are uneven globally (Krumhansl et al., 2016; Wernberg et al., 2019), 7 with population declines (e.g., giant kelp Macrocystis pyrifera in Tasmania, (Butler et al., 2020), sugar kelp 8 Saccharina latissima in the North Atlantic (Filbee-Dexter et al., 2020b)), more common than increases or no 9 change (e.g., giant kelp Macrocystis pyrifera in southern Chile (Friedlander et al., 2020). Warming is driving 10 range contraction and extirpation at the warm edge of species' ranges, and expansions at the cold range edge 11 (very high confidence) (Smale, 2019; Filbee-Dexter et al., 2020b). Local declines in populations of kelp and 12 other canopy-forming seaweeds driven by MHWs and other stressors have caused irreversible shifts to turf- 13 or urchin-dominated ecosystems, with lower productivity and biodiversity (high confidence) (Filbee-Dexter 14 and Scheibling, 2014; Filbee-Dexter and Wernberg, 2018; Rogers-Bennett and Catton, 2019; Beas-Luna et 15 al., 2020; Stuart-Smith et al., 2021), ecosystems dominated by warm-affinity seaweeds or coral (high 16 confidence) (Vergés et al., 2019), and loss of genetic diversity (Coleman et al., 2020a; Gurgel et al., 2020). 17 Do Not Cite, Quote or Distribute 3-48 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Species distribution models of kelp project range shifts and local extirpations with increasing levels of 2 warming (Japan (Takao et al., 2015; Sudo et al., 2020)), (Australia (Table 11.6, Assis et al., 2018; Martínez 3 et al., 2018; Castro et al., 2020)), (Europe (de la Hoz et al., 2019)), (North America (Wilson et al., 2019)), 4 (South America (Figure 12.3)). There is high agreement on the direction but not the magnitude of change 5 (Martínez et al., 2018; Castro et al., 2020), but effects of MHWs are not simulated. Where the length of 6 higher-latitude coastlines is limited, range contractions are projected to occur, even with 2°C of global 7 warming (i.e., SSP1-2.6) due to loss of habitat at the warm edge of species' ranges (Martínez et al., 2018). 8 Poleward expansion of warm-affinity herbivores including urchins could further reduce warm-edge kelp 9 populations (Castro et al., 2020; Mulders and Wernberg, 2020). Evidence from natural temperate CO2 seeps 10 suggests that ocean acidification at levels above those in RCP4.5 in 2100 could offset the increase in urchin 11 abundance (Coni et al., 2021). Genetic analyses suggest that kelp populations at the midpoint of species' 12 ranges will have lower tolerance of warming than implied by species distribution models, without assisted 13 gene flow from warm-edge populations (King et al., 2019; Wood et al., 2021). 14 15 While reducing non-climate drivers can help prevent kelp loss from warming and MHWs, there is limited 16 potential for restoration of kelp ecosystems after transition to urchin-dominant ecosystems (high confidence). 17 Current restoration efforts are generally small-scale (<0.1 km2) and less advanced than those in ecosystems 18 like coral reefs (Coleman et al., 2020b; Eger et al., 2020; Layton et al., 2020). Although abundance of 19 herbivores limits kelp populations, there is limited evidence that restoring predators of herbivores by creating 20 marine reserves, or directly removing grazing species, will increase kelp forest resilience to warming and 21 extremes (Vergés et al., 2019; Wernberg et al., 2019). Active reseeding of wild kelp populations through 22 transplantation and propagation of warm-tolerant genotypes (Coleman et al., 2020b; Alsuwaiyan et al., 2021) 23 can overcome low dispersal rates of many kelp species and facilitate effective restoration (medium 24 confidence) (Morris et al., 2020c). 25 26 Building on the conclusions of SROCC, this assessment finds that kelp ecosystems are expected to decline 27 and undergo changes in community structure in the future due to warming and increasing frequency and 28 intensity of MHWs (high confidence). Risk of loss of kelp ecosystems and shifts to turf- or urchin-dominated 29 ecosystems are highest at the warm edge of species' ranges (high confidence) and risks increase under 30 RCP6.0 and RCP8.5 by the end of the century (high confidence). 31 32 3.4.2.4 Estuaries, Deltas and Coastal Lagoons 33 34 Estuaries, deltas and lagoons encounter environmental gradients over small spatial scales, generating diverse 35 habitats that support myriad ecosystem services, including food provision, regulation of erosion, nutrient 36 recycling, carbon sequestration, recreation and tourism, and cultural significance (D'Alelio et al., 2021; 37 Keyes et al., 2021). Although these coastal ecosystems have historically been sensitive to erosion-accretion 38 cycles driven by sea level, drought and storms (high confidence) (Peteet et al., 2018; Wang et al., 2018c; 39 Jones et al., 2019b; Urrego et al., 2019; Hapsari et al., 2020; Zhao et al., 2020b), they were impacted for 40 much of the 20th century primarily by non-climate drivers (very high confidence) (Brown et al., 2018b; 41 Ducrotoy et al., 2019; Elliott et al., 2019; He and Silliman, 2019; Andersen et al., 2020; Newton et al., 2020; 42 Stein et al., 2020). Nevertheless, the influence of climate-impact drivers has become more apparent over 43 recent decades (medium confidence) (Table 3.6). 44 45 46 Table 3.6: Summary of previous IPCC assessments of estuaries, deltas and coastal lagoons. Observations Projections AR5: (Wong et al., 2014) Humans have impacted lagoons, estuaries and deltas (high Future changes in climate-impact drivers such as warming, to very high confidence), but non-climate drivers have acidification, waves, storms, sea-level rise (SLR), and been the primary agents of change (very high confidence). runoff will have consequences for ecosystem function and services in lagoons and estuaries (high confidence), but In estuaries and lagoons, nutrient inputs have driven with regional differences in magnitude of change in impact eutrophication, which has modified food-web structures drivers and ecosystem response. (high confidence) and caused more-intense and longer- lasting hypoxia, more-frequent occurrence of harmful algal Warming, changes in precipitation and changes in wind blooms, and enhanced emissions of nitrous oxide (high strength can interact to alter water-column salinity and confidence). Do Not Cite, Quote or Distribute 3-49 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report stratification (medium confidence), which could impact In deltas, land-use changes and associated disruption of water column oxygen content (medium confidence). sediment dynamics and land subsidence have driven changes that have been exacerbated by relative sea-level Land-use change, SLR and intensifying storms will alter rise and episodic events including river floods and oceanic deposition-erosion dynamics, impacting shoreline storm surges (very high confidence). vegetation and altering turbidity (medium confidence). Together with warming, these drivers will alter the Increased coastal flooding, erosion and saltwater intrusions seasonal pattern of primary production and the distribution have led to degradation of ecosystems (very high of biota throughout the ecosystems (medium to high confidence). confidence), impacting associated ecosystem services. The projected impacts of climate change on deltas are associated mainly with pluvial floods and SLR, which will amplify observed impacts of interacting climate and non- climate drivers (high confidence). SR15 (Hoegh-Guldberg et al., 2018a) Under both a 1.5°C and 2°C of warming, relative to pre- industrial, deltas are expected to be highly threatened by Estuaries, deltas, and lagoons were not assessed in this SLR and localised subsidence (high confidence). The report. slower rate of SLR associated with 1.5°C of warming poses smaller risks of flooding and salinisation (high confidence), and facilitates greater opportunities for adaptation, including managing and restoring natural coastal ecosystems and infrastructure reinforcement (medium confidence). Intact coastal ecosystems may be effective in reducing the adverse impacts of rising sea levels and intensifying storms by protecting coastal and deltaic regions (medium confidence). Natural sedimentation rates are expected to be able to offset the effect of rising sea levels, given the slower rates of SLR associated with 1.5°C of warming (medium confidence). Other feedbacks, such as landward migration and the adaptation of infrastructure, remain important (medium confidence). SROCC (Bindoff et al., 2019) Increased seawater intrusion caused by SLR has driven Salinisation and expansion of hypoxic conditions will upstream redistribution of marine biotic communities in intensify in eutrophic estuaries, especially in mid and high estuaries (medium confidence) where physical barriers latitudes with microtidal regimes (high confidence). such as the availability of benthic substrates do not limit availability of suitable habitats (medium confidence). The effects of warming will be more pronounced in high- latitude and temperate shallow estuaries characterised by Warming has driven poleward range shifts in species' limited exchange with the open ocean and strong distributions among estuaries (medium confidence). seasonality that already leads to development of dead zones (medium confidence). Interactions between warming, eutrophication and hypoxia have increased the incidence of harmful algal blooms (high Interaction between SLR and changes in precipitation will confidence), pathogenic bacteria such as Vibrio species have greater impacts on shallow than deep estuaries (low confidence), and mortalities of invertebrates and fish (medium confidence). communities (medium confidence). Estuaries characterised by large tidal exchanges and associated well-developed sediments will be more resilient to projected SLR and changes in river flow (medium confidence). Human activities that inhibit sediment dynamics in coastal deltas increase their vulnerability to SLR (medium confidence). 1 Do Not Cite, Quote or Distribute 3-50 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Estuarine biota are sensitive to warming (high confidence), with recent responses including changes in 3 abundance of some fish stocks (Erickson et al., 2021; Woodland et al., 2021), poleward shifts in distributions 4 of fish species, communities and associated biogeographic transition zones (Table 12.3, Franco et al., 2020; 5 Troast et al., 2020), recruits of warm-affinity species persisting into winter (Kimball et al., 2020), and 6 changes in seasonal timing of peaks in species abundance (Kimball et al., 2020). MHWs can be more severe 7 in estuaries than in adjacent coastal seas (Lonhart et al., 2019), causing conspicuous impacts (very high 8 confidence), including mass mortality of intertidal vegetation (Section 3.4.2.5), range shifts in algae and 9 animals (Lonhart et al., 2019), and reduced spawning success among invertebrates (Shanks et al., 2020). 10 11 RSLR extends the upstream limit of saline waters (high confidence) (Harvey et al., 2020; Jiang et al., 2020) 12 and alters tidal ranges (high confidence) (Idier et al., 2019; Talke et al., 2020). Elevated water levels also 13 alter submergence patterns for intertidal habitat (high confidence) (Andres et al., 2019), moving high-water 14 levels inland (high confidence) (Peteet et al., 2018; Appeaning Addo et al., 2020; Liu et al., 2020e) and 15 increasing the salinity of coastal water tables and soils (high confidence) (Eswar et al., 2021). These 16 processes favour inland and/or upstream migration of intertidal habitat, where it is unconstrained by 17 infrastructure, topography or other environmental features (high confidence) (Kirwan and Gedan, 2019; 18 Parker and Boyer, 2019; Langston et al., 2020; Magolan and Halls, 2020; Saintilan et al., 2020). The spread 19 of "ghost forests" along the North American east coast (Kirwan and Gedan, 2019) and elsewhere (Grieger et 20 al., 2020) illustrates this phenomenon. Along estuarine shorelines, changing submergence patterns and 21 upstream penetration of saline waters interact synergistically to stress intertidal plants, changing species 22 composition and reducing above-ground biomass, in some cases favouring invasive species (Xue et al., 23 2018; Buffington et al., 2020; Gallego-Tévar et al., 2020). Overall, changing salinity and submergence 24 patterns decrease the ability of shoreline vegetation to trap sediment (Xue et al., 2018), reducing accretion 25 rates and increasing the vulnerability of estuarine shorelines to submergence by SLR and erosion by wave 26 action (medium confidence) (Zhu et al., 2020b). 27 28 Drought and freshwater abstraction can reduce freshwater inflows to estuaries and lagoons, increasing 29 salinity, reducing water quality (Brooker and Scharler, 2020) and depleting resident macrophyte 30 communities (Scanes et al., 2020b). Changes in freshwater input and SLR, combined with land-use change, 31 can alter inputs of land-based sediments, causing expansion (Suyadi et al., 2019; Magolan and Halls, 2020) 32 or contraction (Andres et al., 2019; Appeaning Addo et al., 2020; Li et al., 2020b) of intertidal habitats. The 33 same phenomena alter salinity gradients, which are the primary drivers of estuarine species distributions 34 (high confidence) (Douglass et al., 2020; Lauchlan and Nagelkerken, 2020). Extreme reduction of freshwater 35 input can extend residence time of estuarine water, leading to persistent HABs (Lehman et al., 2020) and 36 converting estuaries to lagoons if the mouth clogs with sediment (Thom et al., 2020). 37 38 Acidification of estuarine water is a growing hazard (medium confidence) (Doney et al., 2020; Scanes et al., 39 2020a; Cai et al., 2021), and resident organisms display sensitivity to altered pH in laboratory settings 40 (medium confidence) (Young et al., 2019a; Morrell and Gobler, 2020; Pardo and Costa, 2021). However, 41 attribution of the biological effects of acidification is difficult because many biogeochemical processes affect 42 estuarine carbon chemistry (Sections 3.2.3.1, 3.3.2). Warming can exacerbate the impacts of both 43 acidification and hypoxia on estuarine organisms (Baumann and Smith, 2018; Collins et al., 2019b; Ni et al., 44 2020). These effects are further complicated by eutrophication, with high nitrogen loads associated with 45 lower pH (Rheuban et al., 2019). Warming (including MHWs) and eutrophication interact to decrease 46 estuarine oxygen content and pH, increasing the vulnerability of animals to MHWs (Brauko et al., 2020), and 47 exacerbating the incidence and impact of dead zones (medium confidence) (Altieri and Gedan, 2015). The 48 impacts of storms on estuaries are variable and are described in SM3.3.1. 49 50 All these impacts are projected to escalate under future climate change, but their magnitude depends on the 51 amount of warming, the socio-economic development pathway, and implementation of adaptation strategies 52 (medium confidence). Modelling studies (Lopes et al., 2019; Rodrigues et al., 2019; White et al., 2019; 53 Zhang and Li, 2019; Hong et al., 2020; Krvavica and Ruzi, 2020; Liu et al., 2020e; Shalby et al., 2020) 54 suggest that responses of estuaries to SLR will be complex and context dependent (Khojasteh et al., 2021), 55 but project that salinity, tidal range, storm-surge amplitude, depth and stratification will increase with SLR 56 (medium confidence), and that marine-dominated waters will penetrate farther upstream (high confidence). 57 Without careful management of freshwater inputs, sediment augmentation and/or the restoration of Do Not Cite, Quote or Distribute 3-51 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 shorelines to more natural states, transformation and loss of intertidal areas and wetland vegetation will 2 increase with SLR (high confidence) (Doughty et al., 2019; Leuven et al., 2019; Yu et al., 2019; Raw et al., 3 2020; Shih, 2020; Stein et al., 2020), with small, shallow microtidal estuaries being more vulnerable to 4 impacts than deeper estuaries with well-developed sediments (medium confidence) (Leuven et al., 2019; 5 Williamson and Guinder, 2021). Warming and MHWs will enhance stratification and deoxygenation in 6 shallow lagoons (medium confidence) (Derolez et al., 2020) and will continue to drive range shifts among 7 estuarine biota (medium confidence) (Veldkornet and Rajkaran, 2019; Zhang et al., 2020c), resulting in 8 extirpations where thermal habitat is lost and potentially generating new habitat for warm-affinity species 9 (limited evidence, medium agreement) (Veldkornet and Rajkaran, 2019). 10 11 3.4.2.5 Vegetated Blue Carbon Ecosystems 12 13 Mangroves, saltmarshes and seagrass beds (wetland ecosystems) are considered "blue carbon" ecosystems 14 due to their capacity to accumulate and store organic-carbon rich sediments (Box 3.4, Macreadie et al., 2019; 15 Rogers et al., 2019) and provide an extensive range of other ecosystem services (Box 3.4). Because these 16 ecosystems are often found within estuaries and along sheltered coastlines, they share vulnerabilities, 17 climate-impact (Table 3.7) and non-climate drivers with estuaries and coastal lagoons (Section 3.4.2.4). 18 19 20 Table 3.7: Summary of previous IPCC assessments of mangroves, saltmarshes and seagrass beds. Observations Projections AR5: (Wong et al., 2014) Seagrasses occurring close to their upper thermal limits are Climate change will drive ongoing declines in the extent of already stressed by climate change (high confidence). seagrasses in temperate waters (medium confidence) as well as poleward range expansions of seagrasses and Increased CO2 concentrations have increased seagrass mangroves, especially in the Northern Hemisphere (high photosynthetic rates by 20% (limited evidence, high confidence). agreement). Beneficial effects of elevated CO2 will increase seagrass productivity and carbon burial rates in salt marshes during the first half of the 21st century, but there is limited evidence that this will improve their survival or resistance to warming. As a result, interactions between climate change and non- climate drivers will continue to cause declines in estuarine vegetated systems (very high confidence). SR15 (Hoegh-Guldberg et al., 2018a) Intact wetland ecosystems can reduce the adverse impacts of rising sea levels and intensifying storms by protecting Vegetated blue carbon systems were not assessed in this shorelines (medium confidence) and their degradation report. could reduce remaining carbon budgets by up to 100 GtCO2. Under 1.5°C of warming, natural sedimentation rates are projected to outpace SLR (medium confidence), but other feedbacks, such as landward migration of wetlands and the adaptation of infrastructure, remain important (medium confidence). SROCC (Bindoff et al., 2019; Oppenheimer et al., 2019) Coastal ecosystems, including saltmarshes, mangroves, Seagrass meadows (high confidence) will face moderate to vegetated dunes and sandy beaches, can build vertically high risk at temperature above 1.5°C global sea surface and expand laterally in response to SLR, though this warming. capacity varies across sites (high confidence). These ecosystems provide important services that include coastal The transition from undetectable to moderate risk in salt protection and habitat for diverse biota. However, because marshes takes place between 0.7°C­1.2°C of global sea of human actions that fragment wetland habitats and surface warming (medium to high confidence), and Do Not Cite, Quote or Distribute 3-52 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report restrict landward migration, coastal ecosystems between 0.9°C­1.8°C (medium confidence) in sandy progressively lose their ability to adapt to climate-induced beaches, estuaries and mangrove forests. changes and provide ecosystem services, including acting as protective barriers (high confidence). The ecosystems at moderate to high risk under future emission scenarios are mangrove forests (transition from Warming and SLR-driven salinisation of wetlands are moderate to high risk at 2.5°C­2.7°C of global sea surface causing shifts in the distribution of plant species inland and warming), estuaries and sandy beaches (2.3°C­3.0°C) and poleward. Examples include mangrove encroachment into salt marshes (transition from moderate to high risk at subtropical salt marshes (high confidence) and contraction 1.8°C­2.7°C and from high to very high risk at 3.0°C­ in extent of low-latitude seagrass meadows (high 3.4°C) (medium confidence). confidence). Global coastal wetlands will lose between 20­90% of their Plants with low tolerance to flooding and extreme area depending on emissions scenario with impacts on temperatures are particularly vulnerable, increasing the their contributions to carbon sequestration and coastal risk of extirpation (medium confidence). protection (high confidence). Extreme weather events, including heatwaves, droughts Estuarine wetlands will remain resilient to modest rates of and storms, are causing mass mortalities and changes in SLR where their sediment dynamics are unconstrained. community composition in coastal wetlands (high But SLR and warming are projected to drive global loss of confidence). up to 90% of vegetated wetlands by the end of the century under the RCP8.5 (medium confidence), especially if Severe disturbance of wetlands or transitions among landward migration and sediment supply are limited by wetland community types can favour invasive species human modification of shorelines and river flows (medium (medium confidence). confidence). The degradation or loss of vegetated coastal ecosystems Moreover, pervasive coastal squeeze and human-driven reduces carbon storage, with positive feedbacks to the habitat deterioration will reduce the natural capacity of climate system (high confidence). these ecosystems to adapt to climate impacts (high confidence). 1 2 3 Since AR5 and SROCC, syntheses have emphasised that the vulnerability of rooted wetland ecosystems to 4 climate-impact drivers is exacerbated by non-climate drivers (high confidence) (Elliott et al., 2019; 5 Ostrowski et al., 2021; Williamson and Guinder, 2021) and climate variability (high confidence) (Day and 6 Rybczyk, 2019; Kendrick et al., 2019; Shields et al., 2019). Global rates of mangrove loss have been 7 extensive but are slowing (high confidence) at least partially due to management interventions (Friess et al., 8 2020b; Goldberg et al., 2020). From 2000 to 2010 mangrove loss averaged 0.16% yr­1, globally, but with 9 greatest loss in Southeast Asia (high confidence) (Hamilton and Casey, 2016; Friess et al., 2019; Goldberg et 10 al., 2020), and ubiquitous fragmentation leaving few mangroves intact (Bryan-Brown et al., 2020). Saltmarsh 11 ecosystems have also suffered extensive losses (up to 60% in places, since the 1980s), especially in 12 developed and rapidly developing countries (medium confidence) (Table 12.3, Gu et al., 2018; Stein et al., 13 2020). Similarly, 29% of seagrass meadows were lost from 1879­2006 due primarily to coastal development 14 and degradation of water quality, with climate-change impacts escalating since 1990 (medium confidence) 15 (Waycott et al., 2009; Sousa et al., 2019; Derolez et al., 2020; Green et al., 2021a). Local examples of habitat 16 stability or growth (e.g., de los Santos et al., 2019; Laengner et al., 2019; Sousa et al., 2019; Suyadi et al., 17 2019; Derolez et al., 2020; Goldberg et al., 2020; McKenzie and Yoshida, 2020) indicate some resilience to 18 climate change in the absence of non-climate drivers (high confidence). Nevertheless, previous declines have 19 left wetland ecosystems more vulnerable to impacts from climate-impact and non-climate drivers (high 20 confidence) (Friess et al., 2019; Williamson and Guinder, 2021). 21 22 Warming and MHWs have affected the range, species composition and survival of some wetland 23 ecosystems. Warming is allowing some, but not all (Rogers and Krauss, 2018; Saintilan et al., 2018), 24 mangrove species to expand their ranges poleward (high confidence) (Friess et al., 2019; Whitt et al., 2020). 25 This expansion can affect species interactions (Guo et al., 2017; Friess et al., 2019), and enhance sediment 26 accretion and carbon storage rates in some instances (medium confidence) (Guo et al., 2017; Kelleway et al., 27 2017; Chen et al., 2018b; Coldren et al., 2019; Raw et al., 2019). Drought, low sea levels and MHWs can 28 cause significant die-offs among mangroves (medium confidence) (Lovelock et al., 2017b; Duke et al., 29 2021). Seagrasses are similarly vulnerable to warming (high confidence) (Repolho et al., 2017; Duarte et al., 30 2018; Jayathilake and Costello, 2018; Savva et al., 2018), which has been attributed as one cause of observed Do Not Cite, Quote or Distribute 3-53 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 changes in distribution and community structure (medium confidence) (Hyndes et al., 2016; Nowicki et al., 2 2017). MHWs, together with storm-driven turbidity and structural damage, can cause seagrass die-offs (high 3 confidence) (Arias-Ortiz et al., 2018; Kendrick et al., 2019; Smale et al., 2019; Strydom et al., 2020), shifts 4 to small, fast-growing species (high confidence) (Kendrick et al., 2019; Shields et al., 2019; Strydom et al., 5 2020), and ecosystem collapse (Serrano et al., 2021). 6 7 The sensitivity of saltmarshes and mangroves to RSLR depends on whether they accrete inorganic sediment 8 and/or organic material at rates equivalent to rising water levels (very high confidence) (Peteet et al., 2018; 9 FitzGerald and Hughes, 2019; Friess et al., 2019; Gonneea et al., 2019; Leo et al., 2019; Marx et al., 2020; 10 Saintilan et al., 2020). Otherwise, wetland ecosystems must migrate either inland or upstream, or face 11 gradual submergence in deeper, increasingly saline water (very high confidence) (Section 3.4.2.4, Andres et 12 al., 2019; Jones et al., 2019b; Cohen et al., 2020; Mafi-Gholami et al., 2020; Magolan and Halls, 2020; Sklar 13 et al., 2021). Ability to migrate depends on local topography, the positioning of anthropogenic infrastructure, 14 and structures placed to defend such infrastructure (Schuerch et al., 2018; Fagherazzi et al., 2020; Cahoon et 15 al., 2021). Submergence drives changes in community structure (high confidence) (Jones et al., 2019b; Yu et 16 al., 2019; Douglass et al., 2020; Langston et al., 2020) and functioning (high confidence) (Charles et al., 17 2019; Buffington et al., 2020; Stein et al., 2020), and will eventually lead to extirpation of the most sensitive 18 vegetation (medium confidence) (Schepers et al., 2017; Scalpone et al., 2020) and associated animals (low 19 confidence) (Rosencranz et al., 2018). The impacts of storms on wetlands are variable and are described in 20 SM3.3.1. 21 22 As noted in SROCC, given the diversity of coastal wetlands as well as the dependence of their future 23 vulnerability to climate change on adaptation pathways (Krauss, 2021; Rogers, 2021), projections of future 24 impacts based on shoreline elevation estimated from satellite data and CMIP5 projections (Spencer et al., 25 2016; Schuerch et al., 2018) vary greatly. Although all approaches have individual strengths and weaknesses 26 (Törnqvist et al., 2021), paleo-records provide some clarity because they yield estimates of wetland 27 responses to changes in climate in the absence of other anthropogenic drivers and are therefore inherently 28 conservative. On the basis of paleorecords (Table 3.8), we assess that mangroves and saltmarshes are likely 29 at high risk from future SLR, even under SSP1-1.9, with impacts manifesting in the mid-term (medium 30 confidence). Under SSP5-8.5, wetlands are very likely at high risk from SLR, with larger impacts 31 manifesting before 2040 (medium confidence). By 2100, these ecosystems are at high risk of impacts under 32 all scenarios except SSP1-1.9 (high confidence), with impacts most severe along coastlines with gently- 33 sloping shorelines, limited sediment inputs, small tidal ranges and limited space for inland migration (very 34 high confidence) (Cross-Chapter Box SLR in Chapter 3, Schuerch et al., 2018; FitzGerald and Hughes, 2019; 35 Leo et al., 2019; Schuerch et al., 2019; Raw et al., 2020; Saintilan et al., 2020). 36 37 38 Table 3.8: Estimates of vulnerability of coastal wetlands to sea-level rise (SLR) on the basis of sediment cores. Region Habitat Reference Rates of SLR at which WGI AR6 Table 9.9 estimate of SLR (Fox- habitat loss is Kemper et al., 2021) Likely Very likely 2040­2060 2090­2100 Global Mangrove (Saintilan et al., 4.2* 6.1 SSP1-1.9: 2020) 4.2 (2.9­6.1) mm yr­1 4.3 (2.5­6.6) mm yr­1 Southeastern Saltmarsh (Törnqvist et al., 3.5# 4.2# SSP5-8.5: 7.3 (5.7­9.8) mm yr­1 12.2 (8.8­17.7) mm USA 2020) yr­1 UK Saltmarsh (Horton et al., 4.6* 7.1* 2018) * Estimate digitised from published figure # Published figure digitised and remodelled as binomial generalised linear model (number drowned vs. not). 39 40 41 For seagrasses, recent projections for climate-change impacts vary by species and region. Warming is 42 projected to increase the habitat available to Zostera marina on the east coast of the USA by 2100, but 43 contract its southern range edge by 150­650 km under RCP2.6 and RCP8.5, respectively (Wilson and Lotze, Do Not Cite, Quote or Distribute 3-54 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2019). Other species, such as Posidonia oceanica in the Mediterranean, might lose as much as 75% of their 2 habitat by 2050 under RCP8.5 and become functionally extinct (low confidence) by 2100 (Chefaoui et al., 3 2018). Observed impacts of MHWs (Kendrick et al., 2019; Strydom et al., 2020; Serrano et al., 2021) 4 indicate that increasing intensity and frequency of MHWs (Section 3.2.2.1) will have escalating impacts on 5 seagrass ecosystems (high confidence). Habitat suitability can also be reduced by moderate RSLR, due to its 6 impact on light attenuation (medium confidence) (Aoki et al., 2020; Ondiviela et al., 2020; Scalpone et al., 7 2020). 8 9 Overall, warming will drive range shifts in wetland species (medium to high confidence), but SLR poses 10 greatest risk for mangroves and saltmarshes, with significant losses projected under all future scenarios by 11 mid-century (medium confidence) and substantially greater losses by 2100 under all scenarios except SSP1- 12 1.9 (high confidence). MHWs pose the greatest risk to seagrasses (high confidence). In all cases, losses will 13 be greatest where accommodation space is constrained or where other non-climate drivers exacerbate risk 14 from climate-impact drivers (very high confidence). 15 16 3.4.2.6 Sandy Beaches 17 18 Sandy beaches comprise unvegetated, fine- to medium-grained sediments in the intertidal zones that line 19 roughly one-third of the length of the world's ice-free coastlines (Luijendijk et al., 2018). The amenity value 20 of beaches as cultural, recreational and residential destinations has driven extensive urbanisation of beach- 21 associated coastlines (Todd et al., 2019). Beaches also provide habitat for many resident species, nesting 22 habitat for marine vertebrates, filtration of coastal waters and protection of the coastline from erosion 23 (McLachlan and Defeo, 2018). These soft-sediment coastal ecosystems are particularly vulnerable to habitat 24 loss caused by erosion, especially where landward transgression is inhibited by infrastructure (Table 3.9). 25 26 27 Table 3.9: Summary of previous IPCC assessments of sandy beaches. Observations Projections AR5: (Wong et al., 2014) In the absence of adaptation, beaches, sand dunes, and Globally, beaches and dunes have in general undergone cliffs currently eroding will continue to do so under net erosion over the past century or longer. increasing sea level (high confidence). Attributing shoreline changes to climate change is still Coastal squeeze is expected to accelerate with sea-level difficult owing to the multiple natural and anthropogenic rise (SLR). In many locations, finding sufficient sand to drivers contributing to coastal erosion. rebuild beaches and dunes artificially will become increasingly difficult and expensive as present supplies near project sites are depleted (high confidence). In the absence of adaptation measures, beaches and sand dunes currently affected by erosion will continue to be affected under SLR (high confidence). SROCC (Bindoff et al., 2019) Sandy beach ecosystems will increasingly be at risk of Coastal ecosystems are already impacted by the eroding, reducing the habitable area for dependent combination of SLR, other climate-related ocean organisms (high confidence). changes, and adverse effects from human activities on ocean and land (high confidence). Attributing such Sandy shorelines are expected to continue to reduce their impacts to SLR, however, remains challenging due to the area and change their topography due to SLR and influence of other climate-related and non-climate increased extreme climatic erosive events. This will be drivers such as infrastructure development and human- especially important in low-lying coastal areas with high induced habitat degradation (high confidence). Coastal population and building densities (medium confidence). ecosystems, including saltmarshes, mangroves, vegetated dunes and sandy beaches, can build vertically and Assuming that the physiological underpinning of the expand laterally in response to SLR, though this capacity relationship between body size and temperature can be varies across sites (high confidence) as a consequence of applied to warming (medium confidence), the body size human actions that fragment wetland habitats and restrict of sandy beach crustaceans is expected to decrease under landward migration, coastal ecosystems progressively warming (low evidence, medium agreement). lose their ability to adapt to climate-induced changes and Do Not Cite, Quote or Distribute 3-55 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report provide ecosystem services, including acting as Sandy beaches transition from undetectable to moderate protective barriers (high confidence). risk between 0.9ºC­1.8ºC (medium confidence) of global sea surface warming and from moderate to high risk at Loss of breeding substrate, including mostly coastal 2.3ºC­3.0ºC of global sea surface warming (medium habitats such as sandy beaches, can reduce the available confidence). nesting or pupping habitat for land breeding marine turtles, lizards, seabirds and pinnipeds (high confidence). Projected changes in mean and extreme sea levels and warming under RCP8.5 are expected to result in high Overall, changes in sandy beach morphology have been risk of impacts on sandy beach ecosystems by the end of observed from climate related events, such as storm the 21st century (medium confidence), taking account of surges, intensified offshore winds, and from coastal the slow recovery rate of sandy beach vegetation, the degradation caused by humans (high confidence), with direct loss of habitats and the high climatic sensitivity of impacts on beach habitats (e.g., benthic megafauna) some fauna. (medium confidence). Under RCP2.6, the risk of impacts on sandy beaches is expected to be only slightly higher than the present-day level (low confidence). Coastal urbanisation lowers the buffering capacity and recovery potential of sandy beach ecosystems to impacts from SLR and warming and thus is expected to limit their resilience to climate change (high confidence). Coastal squeeze and human-driven habitat deterioration will reduce the natural capacity of these ecosystems to adapt to climate impacts (high confidence). 1 2 3 Since SROCC, observed trends in coastal erosion continue to be obscured by beach nourishment that 4 replaces eroded sediment or by coastal protection of areas at risk of erosion (Section 3.6.3.1.1, Cross-Chapter 5 Box SLR in Chapter 3). Nevertheless, RSLR, increases in wave energy and/or changes in wave direction, 6 disruptions to sediment supplies (including sand mining), and other anthropogenic modifications of the coast 7 have driven localised beach erosion (very high confidence) at rates up to 0.5­3 m yr­1 (Vitousek et al., 2017a; 8 Vitousek et al., 2017b; Cambers and Wynne, 2019; Enríquez-de-Salamanca, 2020; Sharples et al., 2020). 9 Corresponding analyses of coarse-scale (30-m resolution) global data estimate that 15% of tidal flats 10 (including beaches) have been lost since 1984 (medium confidence) (Mentaschi et al., 2018; Murray et al., 11 2019) but with a corresponding number of the world's beaches accreting (28%) as eroding (24%) (medium 12 confidence) (Luijendijk et al., 2018). 13 14 Progress is being made toward models that can project beach erosion under future scenarios despite inherent 15 uncertainties and the presence of multiple confounding drivers in the coastal zone (Vitousek et al., 2017b; Le 16 Cozannet et al., 2019; Cooper et al., 2020a; Vousdoukas et al., 2020b; Vousdoukas et al., 2020a). In the 17 interim, models with varying levels of complexity estimate local loss of beach area to SLR by 2100 under 18 RCP8.5-like scenarios, assuming minimal human intervention, ranging 30­70% (low confidence) (Vitousek 19 et al., 2017b; Mori et al., 2018; Ritphring et al., 2018; Hallin et al., 2019; Kasmi et al., 2020). Within 20 regions, projected impacts scale negatively with beach width and positively with the magnitude of projected 21 SLR. None of these local studies, however, considered high-energy storm events, which are known to also 22 impact sandy coasts (high confidence) (e.g., Burvingt et al., 2018; Garrote et al., 2018; Duvat et al., 2019; 23 Sharples et al., 2020), and model structure often had more influence on projected shoreline responses than 24 did physical drivers (Le Cozannet et al., 2019). Nevertheless, the most-advanced available models, which 25 incorporate multiple coastal processes, including SLR, project that without anthropogenic barriers to erosion, 26 13.6­15.2% and 35.7­49.5% of the world's beaches likely risk undergoing at least 100 m of shoreline retreat 27 (relative to 2010) by 2050 and 2100, respectively (low confidence) (Vousdoukas et al., 2020b). Aggregating 28 these trends regionally suggests that relative rates of shoreline change under RCP4.5 and RCP8.5 diverge 29 strongly after mid-century, with trends towards erosion escalating under RCP8.5 by 2100 (medium 30 confidence) (Figure 3.14, Vousdoukas et al., 2020b). This trend supports the WGI AR6 assessment that 31 projected SLR will contribute to erosion of sandy beaches, especially under high-emissions futures (high 32 confidence) (WGI AR6 Technical Summary, Arias et al., 2021). 33 34 Do Not Cite, Quote or Distribute 3-56 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure 3.14: Relative trends in projected regional shoreline change (advance/retreat relative to 2010). Frequency 3 distributions of median projected change by (a,c) 2050 and (b,d) 2100 under (a,b) RCP4.5 and (c,d) RCP8.5. 4 Projections account for both long-term shoreline dynamics and sea-level rise and assume no impediment to inland 5 transgression of sandy beaches. Data for small island states are aggregated and plotted in the Caribbean. Data from 6 Vousdoukas et al. (2020b). Values for reference regions established in the WGI AR6 Atlas (Gutiérrez et al., 2021) were 7 computed as area-weighted means from original country-level data. For model assumptions and associated debate, see 8 Vousdoukas et al. (2020a) and Cooper et al. (2020a). 9 10 11 For beach fauna, emerging evidence links range shifts, increasing representation by warm-affinity species 12 and mass mortalities to ocean warming (limited evidence, high agreement) (McLachlan and Defeo, 2018; 13 Martin et al., 2019). But even amongst the best-studied taxa, such as turtles, vulnerability to warming (high 14 confidence) and SLR (medium confidence) anticipated on the basis of theory (Poloczanska et al., 2009; Saba 15 et al., 2012; Pike, 2013; Laloë et al., 2017; Tilley et al., 2019) yields only a few detected impacts in the field 16 associated mainly with feminisation (female-skewed sex ratios driven by warmer nest temperatures) (Jensen 17 et al., 2018; Colman et al., 2019; Tilley et al., 2019), phenology (Monsinjon et al., 2019), reproductive 18 success (Bladow and Milton, 2019) and inter-nesting period (Valverde-Cantillo et al., 2019). Moreover, 19 although established vulnerabilities imply high projected future risk for turtles (high confidence) (e.g., 20 Almpanidou et al., 2019; Monsinjon et al., 2019; Patrício et al., 2019; Varela et al., 2019; Santidrián Tomillo 21 et al., 2020), many populations remain resilient to change (Fuentes et al., 2019; Valverde-Cantillo et al., 22 2019; Laloë et al., 2020; Lamont et al., 2020), perhaps because variation in sand temperatures at nesting 23 depth among beaches very likely exceeds the magnitude of warming anticipated by 2100, even under RCP8.5 24 (medium confidence) (Bentley et al., 2020a). As expected for a taxon with a long evolutionary history, turtles 25 display natural adaptation, not only by virtue of broad geographic distributions that include natural climate- 26 change refugia (Boissin et al., 2019; Jensen et al., 2019), but also because some initial responses to warming 27 might counteract anticipated impacts. For example, although feminisation poses a significant long-term risk 28 to turtle populations (high confidence), it might contribute to population growth in the near- to mid-term 29 (medium confidence) (Patrício et al., 2019). Resilience to climate change might be further enhanced by range 30 extensions, alterations in nesting phenology, and fine-scale nest-site selection (medium confidence) (Abella 31 Perez et al., 2016; Santos et al., 2017; Almpanidou et al., 2018; Rivas et al., 2019; Laloë et al., 2020). 32 33 New literature since SROCC on climate impacts and risks has been scarce for most beach taxa besides turtles 34 (the impacts of storms on beach fauna are variable and are described in SM3.3.1). Nevertheless, theoretical 35 sensitivity to warming (Section 3.3.2), together with the projected loss of habitat under future climate 36 scenarios, suggest substantial impacts for populations and communities of beach fauna into the future (high 37 confidence). These impacts will be exacerbated by coastal squeeze along urbanised coastlines (high Do Not Cite, Quote or Distribute 3-57 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 confidence), albeit with magnitudes that cannot yet be accurately projected (McLachlan and Defeo, 2018; Le 2 Cozannet et al., 2019; Leo et al., 2019). 3 4 3.4.2.7 Semi-Enclosed Seas 5 6 This section assesses impacts on five SES, or seas larger than 200,000 km2 with single entrances <120 km 7 wide, including the Persian Gulf, the Red Sea, the Black Sea, the Baltic Sea and the Mediterranean Sea. 8 These SES are largely landlocked and are thus heavily influenced by surrounding landscapes, local and 9 global climate-impact drivers, as well as non-climate drivers (Section 3.1), making them highly vulnerable to 10 cumulative threats. Key climate-impact drivers in SES are warming, increasing frequency and duration of 11 MHWs, acidification, and the increasing in size and number of OMZs (Figure 3.12, Hoegh-Guldberg et al., 12 2014). In AR5, SES were recognised as regionally significant for fisheries and tourism, but highly exposed 13 to both local and global stressors, offering limited options for organisms to migrate in response to climate 14 change (Table 3.10). 15 16 17 Table 3.10: Summary of past IPCC assessments of semi-enclosed seas (SES). Observations Projections AR5 (Hoegh-Guldberg et al., 2014) The surface waters of the SES exhibit significant warming Projected warming increases the risk of greater thermal from 1982 and most Coastal Boundary Systems show stratification in some regions, which can lead to reduced significant warming since 1950. Warming of the O2 ventilation of underlying waters and the formation of Mediterranean has led to the recent spread of tropical additional hypoxic zones, especially in the Baltic and species invading from the Atlantic and Indian Oceans. Black Seas (medium confidence). SES are highly vulnerable to changes in global Changing rainfall intensity can exert a strong influence on temperature on account of their small seawater volume and the physical and chemical conditions within SES, and in landlocked nature. Consequently, SES will respond faster some cases will combine with other climatic changes to than most other parts of the Ocean (high confidence). transform these areas. These changes are likely to increase the risk of reduced bottom-water O2 levels for Baltic and The impact of rising temperatures on SES is exacerbated Black Sea ecosystems (due to reduced solubility, increased by their vulnerability to other human influences such as stratification, and microbial respiration), which is very over-exploitation, pollution, and enhanced runoff from likely to affect fisheries. modified coastlines. Due to a combination of global and local human stressors, key fisheries have undergone Persian Gulf, Red Sea: Extreme temperature events such as fundamental changes in their abundance and distribution heatwaves are projected to increase (high confidence) and over the past 50 years (medium confidence). temperatures are very likely to increase above established thresholds for mass coral bleaching and mortality (very high confidence). SROCC (Bindoff et al., 2019) Semi-enclosed seas were not assessed in this report. Projections from distribution models for multiple fish species show hotspots of decreased species richness in the Indo-Pacific region, and semi-enclosed seas such as the Red Sea and Persian Gulf (medium evidence, high agreement). In addition, geographic barriers such as land boundaries or lower oxygen water in deeper waters are projected to limit species range shifts in semi-enclosed seas, resulting in larger relative decrease in species richness (medium confidence). 18 19 20 Since AR5, there is evidence for increasing frequency and duration of MHWs, extreme weather events and a 21 diversity of threats across depth strata causing mass mortality events, local extirpations and coral reef decline 22 (high confidence) (Section 3.4.2.1, SM3.3.2, Buchanan et al., 2016a; Shlesinger et al., 2018; Wabnitz et al., 23 2018b; Garrabou et al., 2019). In most SES, non-climate drivers, including pollution, habitat destruction, and 24 especially overfishing are decreasing the local adaptive capacity of organisms and the ability of ecosystems Do Not Cite, Quote or Distribute 3-58 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 to cope with climate change impacts (high confidence) (Cramer et al., 2018; Hidalgo et al., 2018; Ben-Hasan 2 and Christensen, 2019). SLR is accelerating faster than expected (high confidence) (Kulp and Strauss, 2019), 3 posing a key risk to SES' coastal ecosystems and the services they provide in urban areas, including drinking 4 water provision, housing, recreational activities, among others (Hérivaux et al., 2018; Reimann et al., 2018). 5 6 The size and number of OMZs are increasing worldwide and in most SES (high confidence) (Global Ocean 7 Oxygen Network, 2018), with growing impacts on fish species diversity and ecosystem functioning. In the 8 Persian Gulf and Red Sea, increasing nutrient loads associated with coastal activities and warming has 9 increased the size of OMZs (high confidence) (Al-Said et al., 2018; Lachkar et al., 2019). OMZs represent an 10 even greater problem in the Black and Baltic Seas, with broad implications for ecosystem function and 11 services (Levin et al., 2009), especially where actions to reduce nutrient loading from land have been unable 12 to reduce the OMZ coverage (high confidence) (Carstensen et al., 2014; Miladinova et al., 2017; Global 13 Ocean Oxygen Network, 2018). In the Baltic Sea, OMZs are affecting the extent of suitable spawning areas 14 of cod, Gadus morhua (high confidence) (Hinrichsen et al., 2016), while in the Black Sea, the combined 15 effect of OMZs and warming is influencing the distribution and physiology of fish species, and their 16 migration and schooling behavior in their overwintering grounds (medium confidence) (Güraslan et al., 17 2017). Cascading effects on food webs have been reported in the Baltic, where detrimental effects of 18 changing oxygen levels on zooplankton production, pelagic and piscivorous fish are influencing seasonal 19 succession and species composition of phytoplankton (high confidence) (Viitasalo et al., 2015). 20 21 In the Mediterranean Sea (Cross-Chapter Paper 4), the increase in climate extremes and mass mortality 22 events reported in AR5 has continued (very high confidence) (Gómez-Gras et al., 2021). Extreme weather 23 events (including deep convection, González-Alemán et al., 2019) and MHWs have become more frequent 24 (Darmaraki et al., 2019) and are associated with mass mortality of benthic sessile species across the basin 25 (high confidence) (Garrabou et al., 2019; Gómez-Gras et al., 2021). Since AR5, in the Persian Gulf and Red 26 Sea, extreme temperatures, together with disease and predation, have continued to cause bleaching-induced 27 mortality of corals, along with declines in the average coral colony size (high confidence) (Burt et al., 2019). 28 Poleward migration and tropicalisation of species (Section 3.4.2.3) has also continued in the Mediterranean, 29 and these phenomena have also become an issue in the Black Sea (high confidence) (Boltachev and Karpova, 30 2014; Hidalgo et al., 2018). Climate impacts on phytoplankton production and phenology show high spatial 31 heterogeneity across the Mediterranean Sea (medium evidence) (Marbà et al., 2015b; Salgado-Hernanz et al., 32 2019), with consequent effects on the diversity and abundance of zooplankton and fish species (medium 33 confidence) (Peristeraki et al., 2019). Changes in primary production and a decrease in river runoff have also 34 altered the optimum habitats for small pelagic fish in the Mediterranean, from local to the basin scale 35 (Piroddi et al., 2017). Evidence of impacts from ocean acidification is increasing, with the rates of coral 36 calcification showing major decline in the Red Sea (medium confidence) (Section 3.4.2.1, Steiner et al., 37 2018; Bindoff et al., 2019). In the Mediterranean Sea, evidence of acidification events have been reported at 38 local scale (Hassoun et al., 2015), with impacts on bivalves and coralligenous species (medium confidence) 39 (Lacoue-Labarthe et al., 2016). 40 41 Climate models project increasing frequency and intensity of MHWs (high confidence) (Section 3.2.2.1), 42 which will exacerbate warming-driven impacts in the Red Sea and Persian Gulf regions, and erode the 43 resilience of Red Sea coral reefs (high confidence) (Osman et al., 2018; Genevier et al., 2019; Kleinhaus et 44 al., 2020). In the Persian Gulf region, extreme temperatures, >35ºC (Pal and Eltahir, 2016), have been linked 45 with high rates of extirpation and a decrease in fisheries catch potential (medium confidence) (Wabnitz et al., 46 2018b). In the Mediterranean Sea, east-west gradients in rates of warming are projected to trigger spatially 47 different changes in primary production, which combined with the increasing arrival of non-indigenous 48 species, may trigger biogeographical changes in fish diversity, increasing in the eastern and decreasing in the 49 western Mediterranean (medium to high confidence) (Albouy et al., 2013; Macias et al., 2015). Projections 50 also show greater impacts from SLR than originally expected in the Mediterranean and Baltic (e.g., Dieterich 51 et al., 2019; Thiéblemont et al., 2019). In the Baltic Sea, under high nutrient load and warming climate 52 scenarios, eutrophication is projected to increase in the future (2069­2098) compared to historical (1976­ 53 2005) periods. In contrast, under continued nutrient load reductions following present management 54 regulations, environmental conditions and ecological state will continue to improve independently of the 55 climate warming scenarios (low to medium confidence) (Saraiva et al., 2019). 56 Do Not Cite, Quote or Distribute 3-59 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 3.4.2.8 Shelf Seas 2 3 Shelf seas overlie the continental margin, often with maximum depths of <200 m, and represent 7% of the 4 global ocean by area (Simpson and Sharples, 2012). These ecosystems are found offshore of every continent, 5 generate 10­30% (Mackenzie et al., 2000; Andersson and Mackenzie, 2004) of global marine net primary 6 production and play a key role in global biogeochemical cycling, including the export of land-borne carbon 7 and nutrients (Johnson et al., 1999; Nishioka et al., 2011; Li et al., 2019) to the deep ocean and recycling of 8 fixed nitrogen back to the atmosphere via denitrification (Devol, 2015). The shelf seas are home to several of 9 the world's major industrial capture fisheries, such as those of the mid-Atlantic Bight, Scotian Shelf, Eastern 10 Bering Sea Shelf and North Brazil Shelf (Barange et al., 2018), and support other marine industries, 11 including aquaculture, extractive industries (oil, gas, and mining), shipping, and renewable-energy 12 installations. 13 14 Similar to other coastal ecosystems, evidence since SROCC (Table 3.11) suggests shelf-sea ecosystems and 15 the fisheries and aquaculture they support are sensitive to the interactive effects of climate-impact drivers, as 16 well as non-climate drivers, including nutrient pollution, sedimentation, fishing pressure and resource 17 extraction (Table 3.12, Figure 3.12). Changes in freshwater, nutrient and sediment inputs from rivers due to 18 both climate and non-climate drivers can influence productivity and nutrient limitation, ecosystem structure, 19 carbon export and species diversity and abundance (Balch et al., 2012; Picado et al., 2014), and can result in 20 reduced water clarity and light penetration (Dupont and Aksnes, 2013; McGovern et al., 2019). Seasonal 21 bottom-water hypoxia occurs in some shelf seas (e.g., northern Gulf of Mexico, Bohai Sea, East China Sea) 22 due to riverine inputs of freshwater and nutrients, promoting stratification, enhanced primary production, and 23 organic carbon export to bottom waters (high confidence) (Zhao et al., 2017; Wei et al., 2019; Del Giudice et 24 al., 2020; Große et al., 2020; Jarvis et al., 2020; Rabalais and Baustian, 2020; Song et al., 2020a; Xiong et 25 al., 2020; Zhang et al., 2020a). 26 27 28 Table 3.11: Summary of past IPCC assessments of shelf seas. Observations Projections AR5 (Hoegh-Guldberg et al., 2014) Primary productivity, biomass yields, and fish capture Global warming will result in more frequent extreme rates have undergone large changes within the East China events and greater associated risks to ocean ecosystems Sea over the past decades (limited evidence, medium (high confidence). In some cases, projected increases will agreement, low confidence). eliminate ecosystems, and increase the risks and vulnerabilities to coastal livelihoods and vulnerabilities for Changing sea temperatures have influenced the abundance food security including Southeast Asia (medium to high of phytoplankton, benthic biomass, cephalopod fisheries, confidence). Reducing stressors not related to climate and the size of demersal trawl catches in the northern change represents an opportunity to strengthen the South China Sea observed over the period 1976­2004 ecological resilience within these regions, which may help (limited evidence, medium agreement). biota survive some projected changes in ocean temperature and chemistry. Concurrent with the retreat of the "cold pool" on the northern Bering Sea shelf, bottom trawl surveys of fish and Changes in eutrophication and hypoxia are likely to invertebrates show a significant community-wide influence shelf seas, but there is low confidence in the northward distributional shift and a colonisation of the understanding of the magnitude of potential changes and former cold pool areas by sub-Arctic fauna (high impacts on ecosystem functioning, fisheries and other confidence). industries. Observed changes in the phenology of plankton groups in the North Sea over the past 50 years are driven by climate forcing, in particular regional warming (high confidence). SROCC (Bindoff et al., 2019) Species composition of fisheries catches since the 1970s in Direct anthropogenic impacts include coastal land-use many shelf seas ecosystems of the world is increasingly change; indirect effects include increased nutrient delivery dominated by warm water species (medium confidence). and other changes in river catchments, and marine resource exploitation in shelf seas. There is high confidence that Estuaries, shelf seas and a wide range of other intertidal these human-driven changes will continue, reflecting and shallow-water habitats play an important role in the coastal settlement trends and global population growth. Do Not Cite, Quote or Distribute 3-60 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report global carbon cycle through their primary production by rooted plants, seaweeds (macroalgae) and phytoplankton, and also by processing riverine organic carbon. However, the natural carbon dynamics of these systems have been greatly changed by human activities (high confidence). 1 2 3 Table 3.12: Synthesis of interactive effects and their influence on shelf-sea ecosystems and the fisheries and 4 aquaculture they support. Factor Example of effect Example references Temperature Altered habitats for species, (Liang et al., 2018; Maharaj et al., change in plankton, fish and 2018; Ma et al., 2019; Meyer and macrofauna community structure, Kröncke, 2019; Yan et al., 2019; influence on species growth, Bargahi et al., 2020; Bedford et al., thermal stress, 2020; Denechaud et al., 2020; altered diversity, Friedland et al., 2020b; Mérillet et al., altered productivity, 2020; Nohe et al., 2020) altered phenology. pH Acidification with hypoxia. (Zhang and Wang, 2019) Salinity Change in species distribution due to (Liu et al., 2020c) altered salinity front distribution. Oxygen concentration Deoxygenation. (Wei et al., 2019; Del Giudice et al., 2020) River discharge Change in plankton community (Shi et al., 2020) structure. Nutrient pollution Enhanced primary production, (Kong et al., 2019; Nohe et al., 2020) change in plankton community structure. Sedimentation Modified ocean chemistry. (Hallett et al., 2018) Fishing pressure Increased vulnerability leading to (Maharaj et al., 2018; Wang et al., changes in community structure. 2019c; Hernvann and Gascuel, 2020) Resource extraction Contamination, (Hall, 2002) change in benthic community structure. 5 6 7 Key risks to shelf seas include shifts or declines in marine micro- and macro-organism abundance and 8 diversity driven by eutrophication, HABs and extreme events (storms and MHWs), and consequent effects 9 on fisheries, resource extraction, transportation, tourism and marine renewable energy (Figure 3.12). The 10 combined effects of deoxygenation and warming can affect the metabolism, growth, feeding behaviour and 11 mobility of fish species (Section 3.3.3). The increasing availability of observations mean that ecosystem 12 changes in shelf seas can be increasingly attributed to climate change (high confidence) (Liang et al., 2018; 13 Maharaj et al., 2018; Ma et al., 2019; Meyer and Kröncke, 2019; Bargahi et al., 2020; Bedford et al., 2020; 14 Friedland et al., 2020b; Mérillet et al., 2020). Eutrophication and seasonal bottom-water hypoxia in some 15 shelf seas have been linked to warming (high confidence) (Wei et al., 2019; Del Giudice et al., 2020) and 16 increased riverine nutrient loading (high confidence) (Wei et al., 2019; Del Giudice et al., 2020). Since 17 SROCC, some severe HABs have been attributed to extreme events, such as MHWs (Section 14.4.2, Roberts 18 et al., 2019; Trainer et al., 2019). However, a recent worldwide assessment of HABs attributed the increase 19 in observed HABs to intensified monitoring associated with increased aquaculture production (high 20 confidence) (Hallegraeff et al., 2021). 21 22 Since SROCC, changes in the community structure and diversity of plankton, macrofauna and infauna have 23 been detected in some shelf seas, although attribution has been regionally specific (e.g., bottom-water 24 warming or hypoxia) (Meyer and Kröncke, 2019; Rabalais and Baustian, 2020). Detection of the 25 picoplankton Synechococcus spp. in the North Sea is potentially linked to a summer decrease in copepod 26 stocks and declining food-web efficiency (low confidence) (Schmidt et al., 2020). The seasonally distinct 27 phytoplankton assemblages in the North Sea have begun to appear concurrently and homogenise (Nohe et Do Not Cite, Quote or Distribute 3-61 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 al., 2020). Changes in abundance, species composition, and size of zooplankton have been detected in some 2 shelf seas (Yellow Sea, North Sea, Celtic Sea, and Tasman Sea), including a decline in stocks of larger 3 copepods, increased abundances of gelatinous and meroplankton, and a shift to smaller species due to 4 warming, increased river discharge, circulation change, and/or extreme events (high confidence) (Wang et 5 al., 2018a; Bedford et al., 2020; Evans et al., 2020; Shi et al., 2020; Edwards et al., 2021). 6 7 Ocean warming has shifted distributions of fish (Free et al., 2019; Franco et al., 2020; Pinsky et al., 2020b; 8 Fredston et al., 2021) and marine mammal species (Salvadeo et al., 2010; García-Aguilar et al., 2018; Davis 9 et al., 2020) poleward (high confidence) or deeper (low to medium confidence) (Section 3.4.3.1, Nye et al., 10 2009; Pinsky et al., 2013; Pinsky et al., 2020b). Warming has also tropicalised the pelagic and demersal fish 11 assemblages of mid- and high-latitude shelves (high confidence) (Montero-Serra et al., 2015; Liang et al., 12 2018; Maharaj et al., 2018; Ma et al., 2019; Friedland et al., 2020a; Kakehi et al., 2021; Punzón et al., 2021). 13 Fisheries catch composition in many shelf-sea ecosystems has become increasingly dominated by warm- 14 water species since the 1970s (high confidence) (Cheung et al., 2013; Leitão et al., 2018; Maharaj et al., 15 2018; McLean et al., 2019). Warming has taxonomically diversified fish communities along a latitudinal 16 gradient in the North Sea, but homogenised functional diversity (McLean et al., 2019). However, in some 17 regions, changing predator or prey distributions, temperature-dependent hypoxia, population changes, 18 evolutionary adaptation, and other biotic or abiotic processes, including species' exploitation, confound 19 responses to climate-impact drivers, which must therefore be interpreted with caution (Frank et al., 2018). 20 For example, although, most species' range edges are tracking temperature change on the northeast shelf of 21 the USA (medium confidence) (Fredston-Hermann et al., 2020; Fredston et al., 2021), range edges of others 22 are not. 23 24 A wide range of responses by fish and invertebrate populations to warming have been observed. The 25 majority of responses have been detrimental, with the direction and magnitude of the response depending on 26 ecoregion, taxonomy, life history, and exploitation history (Free et al., 2019; Yati et al., 2020). For example, 27 fisheries productivity has strongly decreased in the North Sea (Free et al., 2019), and fisheries yields have 28 also decreased in the Celtic Sea, attributed primarily to warming and secondarily to long-term exploitation 29 (Hernvann and Gascuel, 2020; Mérillet et al., 2020). Conversely, fish species diversity and overall 30 productivity have increased in the Gulf of Maine, even with warming (Le Bris et al., 2018; Friedland et al., 31 2020a; Friedland et al., 2020b). Fisheries yields have decreased in the Yellow Sea, East China Sea and South 32 China Sea due primarily to overexploitation (Ma et al., 2019; Wang et al., 2019c), with warming exerting 33 more influence on the yield of cold-water species than on temperate- and warm-water groups (Ma et al., 34 2019). The combined effects of exploitation and multi-decadal climate fluctuations make it difficult to assess 35 global climate-change impacts on fisheries yields (Chapter 5, Ma et al., 2019; Bentley et al., 2020b; Johnson 36 et al., 2020). 37 38 Since AR5, increasing spatial and temporal extent of hypoxia has been projected due to enhanced benthic 39 respiration and reduced oxygen solubility from warming (Del Giudice et al., 2020). Similar to the open 40 ocean, large shifts in the phenology of phytoplankton blooms have been projected for shelf seas throughout 41 subpolar and polar waters (medium confidence) (Henson et al., 2018a; Asch et al., 2019). Zooplankton, 42 which are important prey for many fish species and sea birds, are expected to decrease in abundance on the 43 northeast shelf of the USA (Grieve et al., 2017). However, responses vary by shelf ecosystem (Chust et al., 44 2014b). Trends towards tropicalisation will continue in the future (high confidence) (Cheung et al., 2015; 45 Stortini et al., 2015; Allyn et al., 2020; Maltby et al., 2020; Costa et al., 2021), but uncertainty of future 46 projections of fisheries production increases substantially beyond 2040 (Maltby et al., 2020). Nevertheless, 47 shelf-sea fisheries at lower latitudes are most vulnerable to climate change (Monnereau et al., 2017). Under 48 future climate change marked by more frequent and intense extreme events and the influences of multiple 49 drivers, more flexible and adaptive management approaches could reduce climate impacts on species while 50 also supporting industry adaptation (high confidence) (Section 3.6.3.1.2, Shackell et al., 2014; Stortini et al., 51 2015; Hare et al., 2016; Stortini et al., 2017; Greenan et al., 2019; Ocaña et al., 2019; Maltby et al., 2020). 52 53 3.4.2.9 Upwelling Zones 54 55 EBUS comprise four important social-ecological systems in the Pacific (California and Peru-Humboldt) and 56 Atlantic (Canary and Benguela) ocean basins. Each is characterised by high primary production, sustained by 57 wind-driven upwelling that draws cold, nutrient-rich, generally low-pH and low-oxygen water to the surface Do Not Cite, Quote or Distribute 3-62 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 (Bindoff et al., 2019). Despite their small relative size, the primary productivity in EBUS supports a vast 2 biomass of marine consumers, including some of the world's most productive fisheries (Pauly and Zeller, 3 2016), along with many species of conservation significance (Bakun et al., 2015). 4 5 Although upwelling is important in many other oceanic regions, we focus here on the most documented 6 examples provided by the EBUS. Yet even here, observed changes in upwelling, temperature, acidification 7 and loss of oxygen (Seabra et al., 2019; Abrahams et al., 2021; Gallego et al., 2021; Varela et al., 2021) 8 cannot be robustly attributed to anthropogenic climate change, and projected future changes in upwelling are 9 expected to be relatively small and variable among and within EBUS (Section 3.2.2.3, WGI AR6 Chapter 9, 10 Fox-Kemper et al., 2021). We therefore have few updates to assessments provided by AR5 and SROCC 11 (Table 3.13) and restrict our brief assessment to the limited amount of new evidence (Figure 3.12). 12 13 14 Table 3.13: Summary of previous IPCC assessments of eastern boundary upwelling systems (EBUS). Observations Projections AR5 (Hoegh-Guldberg et al., 2014; Lluch-Cota et al., 2014) EBUS are vulnerable to changes that influence the Like other ocean sub-regions, EBUS are projected to warm intensity of currents, upwelling, and mixing (and hence under climate change, with increased stratification and changes in sea-surface temperature, wind strength and intensified winds as westerly winds shift poleward (likely). direction), as well as O2 content, carbonate chemistry, However, cooling has also been predicted for some EBUS, nutrient content, and the supply of organic carbon to deep resulting from the intensification of wind-driven offshore locations (high confidence). upwelling. Climate-change-induced intensification of ocean upwelling There is medium agreement, despite limited evidence, that in some EBUS, as observed in the last decades, may lead upwelling intensity and associated variables (e.g., to regional cooling rather than warming of surface waters temperature, nutrient, and O2 concentrations) from the and cause enhanced productivity (medium confidence), but Benguela system will change as a result of climate change. also enhanced hypoxia, acidification, and associated biomass reduction in fish and invertebrate stocks. Owing Any projected increase in upwelling intensity has potential to contradictory observations there is currently uncertainty disadvantages. Elevated primary productivity may lead to about the future trends of major upwelling systems and decreasing trophic transfer efficiency, thus increasing the how their drivers will shape ecosystem characteristics (low amount of organic carbon exported to the seabed, where it confidence). is virtually certain to increase microbial respiration and hence increase low O2 stress. Declining O2 and shoaling of the aragonite saturation horizon through ocean acidification increase the risk of upwelling water being low in pH and O2, with impacts on coastal ecosystems and fisheries. These risks and uncertainties are likely to involve significant challenges for fisheries and associated livelihoods along the west coasts of South America, Africa, and North America (low to medium confidence). There is robust evidence and medium agreement that the California Current has experienced an increase of the overall magnitude of upwelling events from 1967 to 2010 (high confidence). This is consistent with changes expected under climate change yet remains complicated by the influence of decadal-scale variability (low confidence). Declining oxygen concentrations and shoaling of the hypoxic boundary layer likely reduced the available habitat for key benthic communities as well as fish and other mobile species. Together with the shoaling of the saturation horizon, these changes have increased the incidence of low O2 and low pH water flowing onto the continental shelf (high confidence; 40 to 120 m), causing problems for industries such as the shellfish aquaculture industry. Do Not Cite, Quote or Distribute 3-63 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report Despite its apparent sensitivity to environmental variability, there is limited evidence of ecological changes in the Benguela Current EBUS due to climate change. SROCC (Bindoff et al., 2019; IPCC, 2019c; IPCC, 2019d) Increasing ocean acidification and oxygen loss are Anthropogenic changes in EBUS will emerge primarily in negatively impacting two of the four major upwelling the second half of the 21st century (medium confidence). systems: the California Current and Humboldt Current EBUS will be impacted by climate change in different (high confidence). Ocean acidification and decrease in ways, with strong regional variability with consequences oxygen level in the California Current upwelling system for fisheries, recreation, and climate regulation (medium have altered ecosystem structure, with direct negative confidence). The Pacific EBUS are projected to have impacts on biomass production and species composition calcium carbonate undersaturation in surface waters within (medium confidence). a few decades under Representative Concentration Pathway (RCP)8.5 (high confidence); combined with Three out of the four major Eastern Boundary Upwelling warming and decreasing oxygen levels, this will increase Systems (EBUS) have shown large-scale wind the impacts on shellfish larvae, benthic invertebrates, and intensification in the past 60 years (high confidence). demersal fishes (high confidence) and related fisheries and However, the interaction of coastal warming and local aquaculture (medium confidence). winds may have affected upwelling strength, with the direction of changes varies between and within EBUS (low The inherent natural variability of EBUS, together with confidence). Increasing trends in ocean acidification in the uncertainties in present and future trends in the intensity California Current EBUS and deoxygenation in California and seasonality of upwelling, coastal warming and Current and Humboldt Current EBUS are observed in the stratification, primary production and biogeochemistry of last few decades (high confidence), although there is low source waters poses large challenges in projecting the confidence to distinguish anthropogenic forcing from response of EBUS to climate change and to the adaptation internal climate variability. The expanding California of governance of biodiversity conservation and living EBUS oxygen minimum zone has altered ecosystem marine resources in EBUS (high confidence). structure and fisheries catches (medium confidence). Given the high sensitivity of the coupled human-natural Overall, EBUS have been changing with intensification of EBUS to oceanographic changes, the future sustainable winds that drives the upwelling, leading to changes in delivery of key ecosystem services from EBUS is at risk water temperature and other ocean biogeochemistry under climate change; those that are most at risk in the 21st (medium confidence). century include fisheries (high confidence), aquaculture (medium confidence), coastal tourism (low confidence) and The direction and magnitude of observed changes vary climate regulation (low confidence). among and within EBUS, with uncertainties regarding the driving mechanisms behind this variability. Moreover, the For vulnerable human communities with a strong high natural variability of EBUS and their insufficient dependence on EBUS services and low adaptive capacity, representation by global Earth System Models gives low such as those along the Canary Current system, confidence that these observed changes can be attributed to unmitigated climate change effects on EBUS (complicated anthropogenic causes. by other non-climatic stresses such as social unrest) have a high risk of altering their development pathways (high confidence). 1 2 3 The California EBUS is arguably the best-studied of the four ecosystems in terms of robust projections of 4 climate change, although even here, there is limited evidence and low agreement among projections. For 5 example, trends in outputs from high-resolution, downscaled models in the California EBUS generally 6 reflect those from underlying coarser-scale ESMs, but projections for physical variables are more convergent 7 among modelling approaches than are those for biogeochemical variables (high confidence) (Howard et al., 8 2020a; Pozo Buil et al., 2021). Models agree on general warming in the California EBUS, with concomitant 9 declines in oxygen content (medium confidence) (Howard et al., 2020b; Fiechter et al., 2021; Pozo Buil et 10 al., 2021). But implications for the future spatial distribution of species, including for some fisheries 11 resources (Howard et al., 2020b; Fiechter et al., 2021), are confounded by local-scale oceanographic process 12 (Siedlecki et al., 2021) and by lateral input of anthropogenic land-based nutrients (Kessouri et al., 2021), 13 suggesting that such projections should be accorded low confidence. 14 15 More generally, changes in upwelling intensity are observed to affect organismal metabolism, population 16 productivity and recruitment, and food-web structure (medium confidence) (van der Sleen et al., 2018; 17 Brodeur et al., 2019; Ramajo et al., 2020). But low confidence in projected trends in upwelling make it Do Not Cite, Quote or Distribute 3-64 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 difficult to extrapolate these results to understand potential changes in the ecology of EBUS. Projected 2 changes in fish biomass within EBUS (Carozza et al., 2019) are therefore accorded low confidence. Finally, 3 although MHWs are an important emerging hazard in the global ocean, with intensity, frequency and 4 duration increasing strongly (Section 3.2.2.1), the number of MHW days yr­1 within EBUS has been 5 increasing more slowly (or decreasing faster, in the case of the Peru-Humboldt system) than in surrounding 6 waters (Varela et al., 2021). Notwithstanding these trends, EBUS remain vulnerable both to MHWs (high 7 confidence) (Sen Gupta et al., 2020) and to their long-lasting impacts (high confidence) (Arafeh-Dalmau et 8 al., 2019; Harvell et al., 2019; McPherson et al., 2021). On this basis, the suggestion that EBUS may 9 represent refugia from MHWs is accorded low confidence. 10 11 Despite low confidence in detailed projections for ecological changes in EBUS, the WGI assessment (WGI 12 AR6 Chapter 9, Fox-Kemper et al., 2021) that upwelling-favourable winds will weaken (or be present for 13 shorter durations) at low latitude but intensify at high latitude (high confidence), albeit by no more than 20% 14 in either case (medium confidence), presents some key risks to associated EBUS ecosystems. These include 15 potential decreases in provisioning services, including fisheries and marine aquaculture (Bertrand et al., 16 2018; Kifani et al., 2018; Lluch-Cota et al., 2018; van der Lingen and Hampton, 2018), and cultural services 17 such as nature-based tourism (Section 3.5). 18 19 3.4.2.10 Polar Seas 20 21 The polar seas cover ~20% of the global ocean and include the deep Arctic Ocean and surrounding shelf seas 22 as well as the Southern Ocean south of the polar front. They play a significant role in ocean circulation and 23 absorption of anthropogenic CO2 (Meredith et al., 2019). The Arctic is characterised by polar seas 24 surrounded by land, while the Antarctic comprises continental Antarctica surrounded by the Southern Ocean. 25 These high-latitude ecosystems share key properties, including strong seasonality in solar radiation and sea- 26 ice coverage. Sea ice regulates water-column physics, chemistry and biology, air-sea exchange, and is a 27 critical habitat for many species. In spring, when solar radiation returns and sea ice melts, intense 28 phytoplankton blooms fuel food webs that include rich communities of both resident and summer-migrant 29 species, with typically high dependency on a few key species for trophic transfer (Meredith et al., 2019; 30 Rogers et al., 2020). Over the last two decades, Arctic Ocean surface temperature has increased in line with 31 the global average, while there has been no uniform warming across the Antarctic (high confidence) (WGI 32 AR6 Chapter 9, Fox-Kemper et al., 2021). Thus, the rate of change due to warming, and associated sea-ice 33 loss, is greater in the Arctic than in the Antarctic (high confidence) (Section 3.2, Table 3.14, WGI AR6 34 Chapter 9, Fox-Kemper et al., 2021). Both Arctic and Antarctic regions have a long history of living 35 resource extraction, including some of the largest fisheries on the globe in terms of catches. However, only 36 the Arctic hosts human populations, holding a rich Indigenous knowledge and Local knowledge on these 37 socio-ecological systems (Cross-Chapter Paper 6, Meredith et al., 2019). 38 39 Previous assessments of polar seas (Table 3.14) concluded that climate change has already profoundly 40 influenced polar ecosystems, through changing species distributions and abundances from primary producers 41 to top predators, including both ecologically and economically important species (high confidence) and that 42 it will continue to do so (Table 3.14). 43 44 45 Table 3.14: Summary of previous IPCC assessments for polar seas. Observations Projections AR5 (Wong et al., 2014) Poleward species distributional shifts are due to climate Some marine species will shift their ranges in response to warming (medium to high confidence). changing ocean and sea ice conditions in the polar regions (medium confidence). Impacts of shifts in ocean conditions affect fish and shellfish abundances in the Arctic (high confidence). Loss of sea ice in summer and increased ocean temperatures are expected to impact secondary pelagic Changes in sea ice and the physical environment to the production in some regions of the Arctic Ocean, with west of the Antarctic Peninsula are altering phytoplankton associated changes in the energy pathways within the stocks and productivity, and krill (high confidence). marine ecosystem (medium confidence). Do Not Cite, Quote or Distribute 3-65 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report Ocean acidification has the potential to inhibit embryo development and shell formation of some zooplankton and krill in the polar regions, with potentially far-reaching consequences to food webs in these regions (medium confidence). Shifts in the timing and magnitude of seasonal biomass production could disrupt coupled phenologies in the food webs, leading to decreased survival of dependent species (medium confidence). SR15 (Hoegh-Guldberg et al., 2018a) A fundamental transformation is occurring in polar The losses in sea ice at 1.5°C and 2°C of warming will organisms and ecosystems, driven by climate change (high result in habitat losses for organisms such as seals, polar confidence). bears, whales and seabirds. There is high agreement and robust evidence that phytoplankton species will change because of sea ice retreat and related changes in temperature and light penetration, and this is very likely to benefit fisheries productivity in the Arctic spring bloom system. `Unique and threatened systems' (RFC1), including Arctic and coral reefs, display a transition from high to very high risk of transition at temperatures between 1.5°C and 2°C of global warming, as opposed to at 2.6°C of global warming in AR5 (high confidence). SROCC (Bindoff et al., 2019) Climate-induced changes in seasonal sea ice extent and Future climate-induced changes in the polar oceans, sea thickness and ocean stratification are altering marine ice, snow and permafrost will drive habitat and biome primary production (high confidence), with impacts on shifts, with associated changes in the ranges and ecosystems (medium confidence). abundance of ecologically important species (medium confidence). Changes in the timing, duration and magnitude of primary production have occurred in both polar oceans, with Projected range expansion of subarctic marine species will marked regional or local variability (high confidence). increase pressure for high-Arctic species (medium confidence), with regionally variable impacts. In both polar regions, climate-induced changes in ocean Both polar oceans will be increasingly affected by CO2 and sea ice conditions have expanded the range of uptake, causing corrosive conditions for calcium carbonate temperate species and contracted the range of polar fish shell-producing organisms (high confidence), with and ice-associated species (high confidence). associated impacts on marine organisms and ecosystems (medium confidence). Ocean acidification will affect several key Arctic species (medium confidence). The projected effects of climate-induced stressors on polar marine ecosystems present risks for commercial and subsistence fisheries with implications for regional economies, cultures and the global supply of fish, shellfish, and Antarctic krill (high confidence). 1 2 3 Since SROCC, evidence demonstrates that warmer oceans, less sea ice and increased advection results in 4 increasing primary production in the Arctic, albeit with regional variation (high confidence), while trends 5 remain spatially heterogeneous and less clear in the Antarctic (medium confidence) (Cross-Chapter Paper 6, 6 Del Castillo et al., 2019; Lewis et al., 2020; Pinkerton et al., 2021; Song et al., 2021a). Furthermore, climate 7 warming influences key mechanisms determining energy transfer between trophic levels including: (1) 8 altered size spectra; (2) shifts in trophic pathways; (3) phenological mismatches; and (4) increased top-down 9 trophic regulation (Table 3.15). However, the scale of impacts from changes in these mechanisms on 10 ecosystem productivity in warming polar oceans remains unresolved and is hence assigned low confidence. 11 Do Not Cite, Quote or Distribute 3-66 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Table 3.15: Examples of mechanisms influencing transfer of energy between lower trophic levels in warmer polar 2 oceans. Mechanism Examples References Altered size spectra Shifts towards smaller algal cells and (Aarflot et al., 2018; Kimmel et al., zooplankton in warmer and more 2018; Weydmann et al., 2018; Hop et stratified oceans results in longer and al., 2019; Møller and Nielsen, 2020; less-efficient food chains, with lower Spear et al., 2020), but see Dong et al. lipid content. (2021) and Vernet et al. (2017) for opposite trends. Shifts in trophic pathways Changes in microbial food-web (Cross-Chapter Paper 6, Fujiwara et interactions, including strengthening of al., 2016; Onda et al., 2017; Vernet et microbial loop, may reduce overall al., 2017; Grebmeier et al., 2018; productivity. Transitions from sea-ice Moore et al., 2018b; Cavan et al., algae to open-water phytoplankton 2019; Vaqué et al., 2019; Yurkowski production may reduce benthic-pelagic et al., 2020)Braekcman et al 2021 coupling and benthic production; transition from autotroph to heterotroph benthic production with increased water turbidity; shifts from krill-dominated to salp-dominated ecosystems in the Antarctic may have negative impacts on higher trophic levels. Phenological mismatches Mismatches in timing arise between (Søreide et al., 2010; Renaud et al., spring phytoplankton blooms and 2018; Dezutter et al., 2019) zooplankton recruits. Increased top-down trophic regulation Increased predation efficiency and top- (Langbehn and Varpe, 2017; Kaartvedt down regulation of zooplankton by and Titelman, 2018; Hobbs et al., zooplanktivorous fish (due to more 2021) light with less sea ice) disconnects zooplankton and phytoplankton production. 3 4 5 Major community shifts, both gradual and abrupt, are observed in polar oceans in response to warming 6 trends and MHWs (Arctic only) (high confidence) (Cross-Chapter Paper 6, Figure 3.12, Beaugrand et al., 7 2019; Meredith et al., 2019; Huntington et al., 2020). In general, abundances and ranges of Arctic fish 8 species are declining and contracting, while ranges of boreal fish species are expanding, both geographically 9 and in terms of feeding interactions and ecological roles (high confidence) (Huserbråten et al., 2019; 10 Meredith et al., 2019; Huntington et al., 2020; Pecuchet et al., 2020a), with variable outcomes for large 11 commercial fish stocks (Cross-Chapter Paper 6, Kjesbu et al., 2014; Holsman et al., 2018; Free et al., 2019). 12 The extreme seasonal solar radiation cycles of these high latitudes may both act as a barrier for species 13 immigration and change predator-prey dynamics in previously ice-covered areas, factors not currently 14 considered in projections (limited evidence) (Kaartvedt and Titelman, 2018; Ljungström et al., 2021). 15 Responses by marine mammals and birds to the ongoing changes in polar ecosystems are both positive and 16 negative (Meredith et al., 2019; Bestley et al., 2020). Phenological, behavioural, physiological, and 17 distributional changes are observed in marine mammals and birds in response to altered ecological 18 interactions and habitat degradation, especially to loss of sea ice (high confidence) (Box 3.2, Cross-Chapter 19 Paper 6, Beltran et al., 2019; Cusset et al., 2019; Descamps et al., 2019; Meredith et al., 2019; Huntington et 20 al., 2020). Reproductive failures and declining abundances attributed to warmer polar oceans and less sea-ice 21 cover are observed in populations of polar bears, Ursus maritimus, seals, whales and marine birds (high 22 confidence) (Box 3.2, Duffy-Anderson et al., 2019; Ropert-Coudert et al., 2019; Bestley et al., 2020; 23 Chambault et al., 2020; Molnár et al., 2020; Stenson et al., 2020). The ongoing changes in polar marine 24 ecosystems can lead to temporary increases in biodiversity and functional diversity (e.g., due to immigration 25 of boreal species in the Arctic, high confidence), but reduced trophic-transfer efficiencies and functional 26 redundancy, with uncertain consequences for ecosystem resilience and vulnerability (limited evidence, low 27 agreement) (Griffith et al., 2019b; Alabia et al., 2020; du Pontavice et al., 2020; Alabia et al., 2021; Frainer 28 et al., 2021). 29 Do Not Cite, Quote or Distribute 3-67 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Calcareous polar organisms are among the groups most sensitive to ocean acidification (high confidence) 2 (Section 3.3.2). Niemi et al. (2021) reports that >80% of sampled sea snail, Limacina helicina, a key species 3 in pelagic food webs, displayed signs of shell dissolution in the Amundsen Gulf. However, bacteria, 4 phytoplankton, zooplankton and benthic communities are found to be detrimentally impacted, resilient, or 5 even positively influenced by ocean acidification in observational and experimental studies (Section 3.3, 6 Hildebrandt et al., 2016; Thor et al., 2018; Ericson et al., 2019; McLaskey et al., 2019; Meredith et al., 2019; 7 Petrou et al., 2019; Renaud et al., 2019; Brown et al., 2020; Hancock et al., 2020; Henley et al., 2020; 8 Johnson and Hofmann, 2020; Torstensson et al., 2021).While fish larval stages may be sensitive, adult fish 9 are expected to have low vulnerability to projected acidification levels (Section 3.3.3, Hancock et al., 2020), 10 although reduced swimming capacity in polar cod in an ocean acidification experiment has been observed 11 (Kunz et al., 2018). Polar organisms' sensitivity to ocean acidification may increase with increasing light 12 levels due to the loss of sea ice (algae, Donahue et al., 2019; Kvernvik et al., 2020), temperature stress 13 (pteropods, Johnson and Hofmann, 2020), or indirectly via increased heterotrophic bacterial productivity 14 (limited evidence) (Vaqué et al., 2019). Due to limited mechanistic understanding of observed effects, and 15 mixed responses among Arctic species, future impacts of ocean acidification are assigned medium 16 confidence for polar species, and low confidence for outcomes for polar ecosystems (Meredith et al., 2019; 17 Green et al., 2021b). 18 19 While levels of pollutants in biota (e.g., persistent organic pollutants, mercury) have generally declined over 20 the past decades, recent increasing levels are associated with release from reservoirs in ice, snow and 21 permafrost, and through changing food webs and pathways for trophic amplification (medium confidence) 22 (Box 3.2, Ma et al., 2016; Amélineau et al., 2019; Foster et al., 2019; Bourque et al., 2020; Kobusiska et al., 23 2020). Also, a warmer climate, altered ocean currents and increased human activities elevate the risk of 24 invasive species in the Arctic (medium confidence), potentially changing ecosystems in this region (high 25 confidence) (Chan et al., 2019; Goldsmit et al., 2020). In the remote Antarctic, there is a lower risk of 26 invasive species (limited evidence) (McCarthy et al., 2019; Holland et al., 2021). 27 28 Fisheries are largely sustainably managed, yet are expanding polewards following sea-ice melt in the Arctic 29 (high confidence) (Fauchald et al., 2021) and possibly in the Antarctic (limited evidence) (Santa Cruz et al., 30 2018). Tourism is increasing and expanding in both polar regions, while shipping and hydrocarbon 31 exploration are growing in the Arctic, increasing the risks of compound effects on vulnerable and already 32 stressed populations and ecosystems (high confidence) (Sections 3.6.3.1.3, 3.6.3.1.4, Cross-Chapter Paper 6, 33 Hauser et al., 2018; Meredith et al., 2019; Helle et al., 2020; Rogers et al., 2020; Cavanagh et al., 2021). 34 35 Ensemble global model projections indicate future increases in primary production and total animal biomass 36 towards 2100 under RCP2.6 (~ 5% and 50%, respectively) and RCP8.5 (~10% and 70%, respectively), in the 37 Arctic (Bryndum-Buchholz et al., 2019; Lotze et al., 2019; Nakamura and Oka, 2019), highlighting 38 opportunities for, and possibly conflicts over, new ecosystem services (Section 3.5). For the Southern Ocean, 39 no overall trends are apparent, but greater variability in both primary production and total animal biomass 40 are projected under RCP2.6, with a ~5% and 15% increase in primary production and total animal biomass 41 under RCP8.5, respectively (Bryndum-Buchholz et al., 2019; Lotze et al., 2019; Nakamura and Oka, 2019). 42 All projections presented exhibit high inter-model variability and hence uncertainty (Heneghan et al., 2021). 43 Furthermore, regional models project significant distributional shifts and wide-ranging trends (i.e., relatively 44 stable, increasing and declining) in productivity for key ecological and commercial species, and functional 45 groups, with weak to strong dependence on emission scenarios, indicating low confidence in future outcomes 46 for polar marine ecosystems and associated ecosystem services (Section 3.5, Piñones and Fedorov, 2016; 47 Griffiths et al., 2017; Klein et al., 2018; Hansen et al., 2019; Meredith et al., 2019; Steiner et al., 2019; Tai et 48 al., 2019; Alabia et al., 2020; Holsman et al., 2020; Reum et al., 2020; Veytia et al., 2020; Sandø et al., 49 2021). Potentially highly influential tipping points associated with Arctic sea-ice melt and Antarctic ocean 50 circulation change adds to this uncertainty (Cross-Chapter Paper 6, Heinze et al., 2021). Nevertheless, 51 increasing evidence supports that sustainable and adaptive ecosystem-based fisheries practices can reduce 52 detrimental impacts of climate change on harvested populations (medium confidence) (Section 3.6.3.1.2, 53 Klein et al., 2018; Free et al., 2019; Hansen et al., 2019; Holsman et al., 2020). 54 55 3.4.3 Oceanic Systems and Cross Cutting Changes 56 Do Not Cite, Quote or Distribute 3-68 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 The oceanic zone, comprising >99% of the ocean's volume, is highly exposed to climate-impact drivers 2 because of its proximity to the atmosphere (Section 3.2, Pörtner et al., 2014; Bindoff et al., 2019), while its 3 relative distance from human settlements and coastal ecosystems decreases variability and interactions, and 4 permits many phenomena to be detected clearly and attributed to climate change. This section assesses how 5 climate-driven changes influence oceanic biological systems over very large spatial scales and notes how 6 impacts on the epipelagic zone affect the mesopelagic, bathypelagic, and deep seafloor ecosystems. 7 8 3.4.3.1 Biogeography and Species Range Shifts 9 10 3.4.3.1.1 Observed species range shifts 11 Since previous assessments (Table 3.16), poleward range-shifts have remained a ubiquitous response to 12 climate change (high confidence), moving species from warmer regions into higher-latitude ecosystems 13 (Fossheim et al., 2015; Kumagai et al., 2018; Burrows et al., 2019; Lenoir et al., 2020). 14 15 16 Table 3.16: Summary of previous IPCC assessments of biogeography and species range shifts. Observations Projections AR5: (Hoegh-Guldberg et al., 2014; Pörtner et al., 2014) The distribution and abundance of many fishes and Spatial shifts of marine species due to projected warming invertebrates have shifted poleward and/or to deeper, will cause high-latitude invasions and high local-extinction cooler waters (high confidence). rates in the tropics and semi-enclosed seas (medium confidence). On average, species' distributions have shifted poleward by 72.0 ± 0.35 km per decade (high confidence). SROCC (Bindoff et al., 2019) Ocean warming has contributed to observed changes in Recent model projections since AR5 and SR15 continue to biogeography of organisms ranging from phytoplankton to support global-scale range shifts of marine fishes at rates marine mammals (high confidence). of tens to hundreds of km per decade in the 21st century, with rate of shifts being substantially higher under RCP8.5 The direction of the majority of the shifts of epipelagic than RCP2.6. organisms are consistent with a response to warming (high confidence) but are also shaped by oxygen concentrations and ocean currents across depth, latitudinal and longitudinal gradients (high confidence). Geographic ranges have shifted since the 1950s by 51.5 ± 33.3 km per decade (mean and very likely range) and 29.0 ± 15.5 km per decade for organisms in the epipelagic and seafloor ecosystems, respectively. 17 18 19 Thermal tolerances of epipelagic populations drive biogeographic change (Figures 3.10, 3.15), but the 20 strength and direction of range shifts tend to be modulated by both climate and non-climate drivers (Pinsky 21 et al., 2020b), including: interactive effects of hypoxia and ocean acidification (Sampaio et al., 2021); 22 oceanic dispersal barriers (Choo et al., 2021), food and critical habitat availability (Alabia et al., 2020; 23 Tanaka et al., 2021), geographic position (including depth, Mardones et al., 2021), and ocean currents 24 (Sunday et al., 2015; Chapman et al., 2020; Fuchs et al., 2020). The difference between physiological 25 thermal tolerances (Section 3.3.2) and local environmental conditions determines safety margins against 26 future climate warming in ectotherms (Pinsky et al., 2019). Acclimation and evolution (Section 3.3.4) and 27 life-history stage (Section 3.3.3) also alter species' thermal tolerances. Biogeographic responses are further 28 modulated by other interacting factors (Table 3.17). 29 30 A large global meta-analysis of range shifts across multiple levels of the marine food web (Lenoir et al., 31 2020) estimates that marine species are moving poleward at a rate of 59.2 km per decade (very likely range: 32 43.7­74.7 km per decade), closely matching the local climate velocity (high confidence). In some cases, 33 warming-related distribution shifts were followed by density-dependent use of these areas, influencing Do Not Cite, Quote or Distribute 3-69 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 associated fisheries (Baudron et al., 2020), and in others, warming influenced competitive interactions: in the 2 Arctic-Boreal Barents Sea, warming-induced increases in cod (Gadus morhua) abundance reduces haddock 3 (Melanogrammus aeglefinus) abundance (Durant et al., 2020). 4 5 6 7 Figure 3.15: Schematic of range-shift dynamics in marine ectotherms in response to climate warming. As the ocean 8 warms, conditions at the edge of the species' distribution may become warmer than the maximum thermal tolerance of 9 the species (such as with T2, see Figure 3.9), causing local populations to undergo a gradual decline in performance, a 10 decreasing population size and ultimately their extirpation, resulting in a range contraction. Conversely, at the cool 11 extreme of the distribution (such as with T1), habitats beyond the current range of the species will become thermally 12 suitable in the future (i.e., within the species' thermal tolerance range) and, providing the species can disperse to those 13 locations, allow for the colonisation and consolidation of new populations and subsequent range expansion. These are 14 processes conditioned by multiple drivers that interact with warming to ultimately define range shift responses; some of 15 which are described in Table 3.17. Note that physiological thermal tolerances relate to body temperatures of the 16 organism rather than ambient temperatures. 17 18 19 Table 3.17: Synthesis of selected processes conditioned by multiple environmental drivers that interact with warming 20 to ultimately define range-shift responses. Factor Effect Example references Evolution and acclimation Evolution of thermal tolerances and (Palumbi et al., 2014; Miller et al., acclimation under local climatic 2020a) conditions can increase resilience to future climate warming, slowing the loss of species at trailing (warm) range edges. Marine heatwaves (MHWs) Influence the evolution of thermal (Buckley and Huey, 2016; Sunday et tolerances by eliminating genotypes al., 2019) that are intolerant of elevated temperatures. MHWs can produce widespread die- (Smale and Wernberg, 2013) offs of shallow-water benthic organisms triggering extensive contractions of their ranges. MHWs can facilitate range expansions (Leriorato and Nakamura, 2019; by opening niches and/or enhancing Thomsen et al., 2019; Monaco et al., recruitment of warm-affiliated species. 2021) Cold-waves can halt or even reverse (Leriorato and Nakamura, 2019) range expansions at leading edges. Do Not Cite, Quote or Distribute 3-70 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report Ocean currents Influence range dynamics through (Hunt et al., 2016; Kumagai et al., Climatic refugia their effect on dispersal, depending on 2018; Fuchs et al., 2020) their magnitude, direction and seasonal Oxygen availability patterns. Habitat availability and quality Where currents align with spatial (García Molinos et al., 2017) Biotic interactions, including food availability gradients of warming, range Do Not Cite, Quote or Distribute expansions track thermal changes more closely. Conversely, directional mismatches result in consistently slower expansion rates and larger response lags; an effect more acute for benthic organisms relying on passive dispersion of larvae and propagules. Rates of range contraction across taxa (García Molinos et al., 2017) decreased (increased) under directional agreement (mismatch) with ocean currents, possibly associated with enhanced (reduced) flows of adaptive genes to warming in downstream (upstream) populations within the distributional range. Areas of locally stable climatic (Smith et al., 2014; Assis et al., 2016; conditions, such as deeper waters or Lourenço et al., 2016; Wyatt et al., regions with internal tides or localised 2020) upwelling, can buffer the effects of regional warming, facilitating species persistence and conserving genetic diversity at rear edge populations. Distributional shifts into deeper, cooler (Smith et al., 2014; Assis et al., 2016; habitats can offer an effective Lourenço et al., 2016) alternative response to latitudinal shifts, because sharper thermal gradients mean vertical displacements, needed to compensate for the same amount of warming, are several orders of magnitude smaller than planar displacements. Oxygen supersaturation may extend (Giomi et al., 2019) ectotherm survival to extreme temperatures and increase thermal tolerances by compensating for the increasing metabolic demand at high temperatures. Oxygen deprivation increases (Brown and Thatje, 2015; Roman et metabolic demand and respiration al., 2019; Hughes et al., 2020) rates. Shallowing of oxygen dead zones and subsequent hypoxic avoidance can render deep thermal refuges unsuitable for organisms. The availability and quality of habitat (Krause-Jensen et al., 2019; Tamir et (underwater light conditions, adequate al., 2019) substrate, nutrient and food supply) set limits to the distribution of organisms and range shift dynamics (e.g., resilience of populations to climate warming and the consolidation of range expansions). Species interactions can confer (Falkenberg et al., 2015; Giomi et al., resilience to warming by retarding 2019) habitat degradation and buffering the impacts of warming on organisms. 3-71 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report Changes in biotic interactions (e.g., (Selden et al., 2018; Westerbom et al., altered predation rates, food 2018; Figueira et al., 2019; Pinsky et availability, competition or trophic al., 2020b; Monaco et al., 2021) mismatches) induced by climate warming can modify range-shift dynamics. 1 2 3 Biogeographic shifts lead to novel communities and biotic interactions (high confidence) (Zarco-Perello et 4 al., 2017; Pecuchet et al., 2020b), with concomitant changes in ecosystem functioning and servicing (high 5 confidence) (Vergés et al., 2019; Nagelkerken et al., 2020; Peleg et al., 2020). For instance, temperature- 6 driven changes in distribution and abundance of copepods, the dominant zooplankton, were observed 7 between 1960­2014 in the North Atlantic. These changes subsequently affect biogenic carbon cycling 8 through alteration of microbial remineralisation and carbon sequestration in deep water (medium confidence) 9 (Section 3.4.3.6, Pitois and Fox, 2006; Brun et al., 2019). 10 11 3.4.3.1.2 Observed vertical redistributions 12 Epipelagic isotherms have recently (1980­2015) deepened at an average of 6.6 ± 18.8 m per decade (Pinsky 13 et al., 2019) but, there is low agreement on whether species move deeper in pursuit of thermal refuge. Prior 14 studies suggested range shifts to depth (Dulvy et al., 2008; Pinsky et al., 2013; Yemane et al., 2014), but 15 increasing evidence suggests that fish and planktonic communities across large parts of the North Atlantic, 16 sub-Arctic and northeast Pacific Ocean redistribute horizontally with horizontal climate velocity, except 17 where vertical temperature gradients are particularly steep. There is low confidence for temperature-driven 18 depth shifts in the epipelagic zone (Burrows et al., 2019; Campana et al., 2020; Caves and Johnsen, 2021). 19 At the same time, decreasing oxygen concentrations and the vertical expansion of OMZs have already 20 decreased suitable habitat of pelagic fishes, including tuna and billfishes, by ~15% primarily due to vertical 21 compression of environmental niches (Stramma et al., 2012; Deutsch et al., 2015). 22 23 3.4.3.1.3 Projected changes in species range shifts 24 Continued changes in the biogeography of marine predators and prey are anticipated under future climate 25 change, with climate velocity in the epipelagic zone during 2050­2100 under RCP8.5 projected to be 26 sevenfold faster than that during 1955­2005 (medium confidence) (Figure 3.4, Brito-Morales et al., 2020). 27 This have substantial ecological implications, as projections suggest near-elimination of overlaps between 28 the distributions of certain predator-prey pairs in the northeast Atlantic Ocean when their current joint 29 distributions (1989­2014) are compared with those projected (2037­2062) under RCP8.5 (Sadykova et al., 30 2020). 31 32 Deepening of epipelagic isotherms is projected to accelerate over 2006­2100 to rates of 8.5 m per decade 33 under RCP4.5 and 32 m per decade under RCP8.5 (Jorda et al., 2020). Although vertical redistribution of 34 thermal niches is three to four orders of magnitude slower than horizontal displacement, maximum depth 35 limits imposed by the seafloor and photic layer (both of which are projected to be reached in this century) 36 will likely vertically compress suitable habitat for most marine organisms (medium confidence) (Dueri et al., 37 2014; Jorda et al., 2020). 38 39 Projections from coupled biogeochemical and ecosystem models suggest a general decline in mesopelagic 40 biomass (Lefort et al., 2015), although this may vary among ocean basins. The volume of OMZs have been 41 expanding at many locations (high confidence), and the oxygen content of the subsurface ocean is projected 42 to decline to historically unprecedented conditions over the 21st century (medium confidence) (Section 43 3.2.3.2, WGI AR6 Section 5.3.3.2, Canadell et al., 2021) at a rate of 10­15 µM per decade in OMZs (Section 44 3.2.3.2, Breitburg et al., 2018). Oxygen availability and the effects of ocean acidification (Sections 3.3, 45 3.4.2) on zooplankton might become a dominant constraint in the upper ocean's metabolic index, which is 46 projected to decrease globally by 20% by 2100 (Deutsch et al., 2015; Steinberg and Landry, 2017). In 47 addition, extremely rapid acceleration of climate velocities projected in the mesopelagic under all emissions 48 scenarios suggest that species in this ocean stratum will be even more exposed to future warming than 49 species in the epipelagic (Figure 3.4, Brito-Morales et al., 2020). But projections also suggest that warming- 50 related increases in trophic efficiency lead to a 17% increase in the biomass of the deep scattering layer 51 (zooplankton and fish in the mesopelagic) by 2100 (low confidence) (Bindoff et al., 2019); observational Do Not Cite, Quote or Distribute 3-72 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 studies appears to show that mesopelagic fishes adapted to warm water increased in abundance and 2 distribution in the California Current associated with warming and the expansion of OMZ (Koslow et al., 3 2019), suggesting that some mesopelagic fish stocks might be resilient to a changing climate (medium 4 confidence). 5 6 3.4.3.2 Phenological shifts & trophic mismatches 7 8 3.4.3.2.1 Observed changes 9 SROCC reported high confidence in phenological shifts towards earlier onset of biological events (Table 10 3.18), with phenological shifts among epipelagic species attributed to ocean warming (high confidence). 11 12 13 Table 3.18: Summary of previous IPCC assessments of phenological shifts & trophic mismatches Observations Projections AR5 WGII (Hoegh-Guldberg et al., 2014; Larsen et al., Projections of phenological shifts and trophic mismatches 2014) were not assessed in this report. Changes to sea temperature have altered the phenology, or timing of key life-history events such as plankton blooms, and migratory patterns and spawning in fish and invertebrates, over recent decades (medium confidence). There is medium to high agreement that these changes pose significant uncertainties and risks to fisheries, aquaculture, and other coastal activities. The highly productive high-latitude spring bloom systems in the northeastern Atlantic are responding to warming (medium evidence, high agreement), with the greatest changes being observed since the late 1970s in the phenology, distribution, and abundance of plankton assemblages, and the reorganisation of fish assemblages, with a range of consequences for fisheries (high confidence). Observed changes in the phenology of plankton groups in the North Sea over the past 50 years are driven by climate forcing, in particular regional warming (high confidence). On average, spring events in the ocean have advanced by 4.4 ± 0.7 days per decade (mean ± SE). Shifts in the timing and magnitude of seasonal biomass production could disrupt matched phenologies in the food webs, leading to decreased survival of dependent species (medium confidence). If the timing of primary and secondary production is no longer matched to the timing of spawning or egg release, survival could be impacted, with cascading implications to higher trophic levels. This impact would be exacerbated if shifts in timing occur rapidly (medium confidence). There is medium to high confidence that climate- induced disruptions in the synchrony between timing of spawning and hatching of some fish and shellfish and the seasonal increases in prey availability can result in increased larval or juvenile mortality or changes in the condition factor of fish and shellfish species in the Arctic marine ecosystems. SROCC (Bindoff et al., 2019) Phenology of marine ectotherms in the epipelagic systems Projections of phenological shifts and trophic mismatches are related to ocean warming (high confidence) and that were not assessed in this report. Do Not Cite, Quote or Distribute 3-73 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report the timing of biological events has shifted earlier (high confidence). Timing of spring phenology of marine organisms is shifting to earlier in the year under warming, at an average rate of 4.4 ± 1.1 days per decade, although it is variable among taxonomic groups and among ocean regions. WGI AR6 Chapter 2 (Gulev et al., 2021) Phenological metrics for many species of marine Projections of phenological shifts and trophic mismatches organisms have changed in the last half-century (high were not assessed in this report. confidence), though many regions and many species of marine organisms remain under-sampled or even unsampled. The changes vary with location and with species (high confidence). There is a strong dependence of survival in higher trophic-level organisms (fish, exploited invertebrates, birds) on the availability of food at various stages in their life cycle, which in turn depends on phenologies of both (high confidence). There is a gap in our understanding of how the varied responses of marine organisms to climate change, from a phenological perspective, might threaten the stability and integrity of entire ecosystems. 1 2 3 Since SROCC, field data have continued to show that the phenology of biological events in the ocean is very 4 likely (high to very high confidence) advancing in response to climate change, with 71.9% of published 5 observations consistent with these anticipated effects (Figure 3.16a,b; Table 3.19), although most reports 6 (95.6%) were from the Northern Hemisphere (Figure 3.16a). Biological events that are shifting earlier in 7 response to climate change include phytoplankton blooms (Scharfe and Wiltshire, 2019; Chivers et al., 2020) 8 such as those of HAB species (Forsblom et al., 2019; Bucci et al., 2020); peaks in zooplankton abundance 9 (Chevillot et al., 2017; Forsblom et al., 2019); the migration (Otero et al., 2014; Kovach et al., 2015; Chust et 10 al., 2019) and spawning of commercial fish (McQueen and Marshall, 2017; Kanamori et al., 2019) including 11 crabs and squid (Henderson et al., 2017); and breeding of marine reptiles (Mazaris et al., 2008; Cherkiss et 12 al., 2020). Moreover, different trophic levels within epipelagic food webs are responding at different rates 13 (very high confidence) (Table 3.19, Figure 3.16b,c), with greater and more consistent responses by lower 14 trophic levels (phytoplankton, holozooplankton and meroplankton), but less consistent, weaker and more 15 varied responses by higher trophic levels. There were too few independent time series to make robust 16 estimates for benthic invertebrates, plants, marine reptiles and mammals. This differential response across 17 trophic levels could lead to trophic mismatches (Neuheimer et al., 2018), where predators and their prey 18 respond asynchronously to climate change (Edwards and Richardson, 2004; Rogers and Dougherty, 2019; 19 Rubenstein et al., 2019; Émond et al., 2020), with potential population-level consequences, including 20 declines in fish recruitment (Burthe et al., 2012; Chevillot et al., 2017; McQueen and Marshall, 2017; Asch 21 et al., 2019; Durant et al., 2019; Régnier et al., 2019). Available evidence also suggests that feeding 22 relationships could modulate species responses to climate change, as seen in breeding of surface-feeding and 23 deeper-diving seabirds (Descamps et al., 2019). These differential responses could determine `winners' and 24 `losers' under future climate change (Lindén, 2018). 25 26 Do Not Cite, Quote or Distribute 3-74 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure 3.16: Observed responses to climate change based on a systematic Web of Science review of marine phenology 3 studies exceeding 19 years in length to update the assessment in WGII AR5 Chapter 30 (Hoegh-Guldberg et al., 2014). 4 Error bars indicate 95% confidence limits (i.e., the extremely likely range). (a) Global data showing changes in seasonal 5 cycles of biota that are attributed (at least partly) to climate change (blue, n=297 observations), and changes that are 6 inconsistent with climate change (white, n=116 observations). Each circle represents the centre of a study area. (b) The 7 proportion of phenological observations (showing means and extremely likely ranges) that are attributed to climate 8 change (i.e., generally showing earlier timing) by taxonomic group. (c) Observed shift in timing (days per decade, 9 showing means and extremely likely ranges), by taxonomic group, that are attributed to climate change. Negative shifts 10 are earlier, positive shifts are later. Details and additional plots are presented in SM3.3.3, Figure SM3.3 and Table 11 SM3.1. 12 13 Do Not Cite, Quote or Distribute 3-75 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Table 3.19: Assessment of phenological shifts by taxon based on time series from field observations spanning at least 2 19 years published over the past 25 years. Taxon Rate of consistency of Estimated mean rate of Confidence Notes observations with change in seasonal climate change timing Phytoplankton 78.41% ­7.5 days per decade Very high confidence Evidence most robust (n=85) (n=83) for changes in timing of blooms in the North Atlantic (e.g., Chivers et al., 2020) and Baltic (e.g., Scharfe and Wiltshire, 2019; Wasmund et al., 2019), with limited evidence from the Southern Hemisphere. Holozooplankton 79.74% ­4.27 days per decade Very high confidence Evidence most robust (n=77) (n=58) in the northeast Atlantic (e.g., Chevillot et al., 2017), but sparse elsewhere. Meroplankton (taxa 81.06% ­4.34 days per decade Very high confidence Includes earlier peak that are only (n=72) (n=64) abundance of fish temporarily in the larvae in upwelling plankton) systems (e.g., Asch, 2015). Benthic invertebrates 72.34% ­8.5 days per decade Low confidence Evidence is limited, (n=5) uncertainty levels are (n=5) (limited evidence, high. Rate of consistency of medium agreement) responses with climate change is not significantly different from random chance. Plants 100% No estimate available Very low confidence Just a single study for (n=1) seagrasses, and only for consistency (Diaz- Almela et al., 2007). Fish 65.48% ­3.02 days per decade Very high confidence Includes earlier (n=109) (n=43) appearance of migratory fish in estuaries (e.g., Chevillot et al., 2017), earlier spawning migrations for anadromous fish such as salmon (e.g., Rubenstein et al., 2019), earlier migrations for sole (e.g., Fincham et al., 2013) and tuna (e.g., Dufour et al., 2010), and earlier spawning of key commercial demersal (bottom- dwelling) species such as cod (e.g., McQueen and Marshall, 2017). Do Not Cite, Quote or Distribute 3-76 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report Marine reptiles 100.0% ­2.89 days per decade Low confidence Evidence is limited, (n=4) (n=4) (limited evidence, low uncertainty levels are agreement) high. Mean phenological shift is not significantly different from zero. Seabirds 42.36% +0.77 days per decade Very low confidence Neither the rate of (n=56) (n=51) (limited evidence, low consistency with agreement) climate change nor the phenological shift differ significantly from null expectations (50% consistency and no shift). Many seabirds are breeding earlier (Byrd et al., 2008; Sydeman et al., 2009), while breeding among others in temperate and polar regions has been delayed, which has been linked to later sea-ice breakup or limited prey resources (Barbraud and Weimerskirch, 2006; Wanless et al., 2009; Chambers et al., 2014). Although the response of lifecycle events for many seabird species is variable in direction, there has usually been a more complex driver associated with climate that has been considered to be responsible (Sydeman et al., 2015). For many species, seasonal timing is moving earlier, especially in the Arctic (e.g., Byrd et al., 2008; Descamps et al., 2019), but for many species in the Southern Ocean, it is not (Barbraud and Weimerskirch, 2006; Chambers et al., 2014). This could be because of a much slower rate of warming in most of the Southern Ocean than in the Arctic. Marine mammals 100.0% ­0.34 days per decade Very low confidence All studies of (n=4) (n=4) (limited evidence, low phenological changes agreement) for marine mammals have focused on whales (e.g., Ramp et al., 2015; Hauser et al., Do Not Cite, Quote or Distribute 3-77 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 2017; Loseto et al., 2018) or polar bears (e.g., Cherry et al., 2013; Atwood et al., 2016; Escajeda et al., 2018) and have related timing to aspects of sea ice dynamics, highlighting the complexity of such processes. Mean phenological shift is not significantly different from zero at the global scale. 1 2 3 3.4.3.2.2 Projected changes 4 The CMIP6 ESM ensembles project that, by 2100, 18.8% ± 19.0% (mean ± very likely range) and 38.9% ± 5 9.4% of the ocean surface will very likely undergo a change of 20 days or more (advance or delay) in the 6 start of the phytoplankton growth period under SSP1-2.6 and SSP5-8.5, respectively (Figure 3.17a,b) (low 7 confidence due to the dependence with the projected changes in phytoplankton biomass which trends are 8 reported with low confidence) (Section 3.4.3.4 and SROCC Section 5.2.3, Bindoff et al., 2019). 9 Phytoplankton growth is projected to begin later in the Northern Hemisphere subtropics, and earlier at high 10 latitudes in some regions around the Antarctic Peninsula, and over large areas in the Northern Hemisphere 11 (low to medium confidence as there are improved constraints from historical variability in this region and 12 consistency with CMIP5-based studies results) (Henson et al., 2018b; Asch et al., 2019). There is high 13 agreement in model projections that the start of the phytoplankton growth period will very likely advance in 14 the Arctic Ocean under a high-emission scenario for CMIP5 and CMIP6 (Figure 3.17b, Henson et al., 2018b; 15 Asch et al., 2019; Tedesco et al., 2019; Lannuzel et al., 2020). The CMIP6 ensemble projections further 16 show limited changes in phenology across most of the Southern Ocean, but large regional variations in the 17 tropics (Figure 3.17). Overall, the regional patterns are qualitatively similar under SSP1-2.6 and SSP5-8.5, 18 but with greater magnitude and larger areas under SSP5-8.5 (low confidence). 19 20 At latitudes >40ºN, temperature-linked phenology of fish reproduction with high geographic fidelity to 21 spawning grounds (geographic spawners) is projected to change at double the rate of that for phytoplankton, 22 which will likely cause phenological mismatches resulting in increased risk of starvation for fish larvae 23 (medium to high confidence) (WGI AR6 Section 2.3.4.2.3, Asch et al., 2019; Durant et al., 2019; Régnier et 24 al., 2019; Gulev et al., 2021; Laurel et al., 2021). Furthermore, under RCP8.5, trophic mismatch events 25 exceeding ±30 days (Asch et al., 2019) leading to fish recruitment failure are expected to increase 10-fold for 26 geographic spawners across much of the North Atlantic, North Pacific and Arctic Ocean basins (low 27 confidence) (Neuheimer et al., 2018). In contrast, temporal mismatches between fish that relocate spawning 28 grounds in response to environmental variations (environmental spawners) and phytoplankton blooms are 29 projected to remain shorter and less varied, suggesting that across ocean basins, range shifts by 30 environmental spawners may increase their resilience. Nevertheless, this compensation mechanism might 31 fail at locations where phytoplankton bloom phenology is not controlled by temperature-driven water- 32 column stratification, leading to a possible six-fold local increase in extreme mismatches under climate 33 change (Asch et al., 2019). 34 Do Not Cite, Quote or Distribute 3-78 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure 3.17: Projected phytoplankton phenology. (a,c) Spatial patterns and (b,d) density distributions of projected 3 change in phytoplankton phenology by 2100 under Shared Socioeconomic Pathway (SSP)1-2.6 and SSP5-8.5, 4 respectively. Difference in the start of the phytoplankton growth period is calculated as 2090­2099 minus 1996­2013. 5 Negative (positive) values indicate earlier (later) start of the phytoplankton growth period by 2100. The ensemble 6 projections of global changes in phytoplankton phenology include, under SSP1-2.6 and SSP5-8.5, respectively, a total 7 of five Coupled Model Intercomparison Project 6 Earth System Models containing coupled ocean biogeochemical 8 models that cover a wide range of complexity (Kwiatkowski et al., 2019). The phenology calculations are based on 9 Racault et al. (2017) using updated data. 10 11 12 3.4.3.3 Changes in Community Composition and Biodiversity 13 14 3.4.3.3.1 Evidence of natural adaptive capacity based on species' responses to past climate variability 15 16 Responses to abrupt climate change in the geological past suggest that adaptive capacity is limited for marine 17 animals (Cross-Chapter Box PALEO in Chapter 1). Temperatures during the last Interglacial (~125 ka), 18 which were warmer than today, led to poleward range shifts of reef corals (medium confidence) (Kiessling et 19 al., 2012; Jones et al., 2019a). Temperature has also driven marine range shifts over multi-million-year 20 timescales (medium confidence) (Gibbs et al., 2016; Reddin et al., 2018). Warming climates, even with low 21 ocean warming rates, inevitably decreased tropical marine biodiversity compared with mid-latitudes (high 22 confidence) (Mannion et al., 2014; Crame, 2020; Yasuhara et al., 2020; Raja and Kiessling, 2021). 23 24 The paleo record confirms that marine biodiversity has been vulnerable to climate warming both globally 25 and regionally (very high confidence) (Cross-Chapter Box PALEO in Chapter 1, Stanley, 2016). In extreme 26 cases of warming (e.g., >5.2°C), marine mass extinctions occurred in the geological past, and there may be a 27 relationship between warming magnitude and extinction toll (medium confidence) (Song et al., 2021b). A 28 combination of warming and spreading anoxia caused marine extinctions in ancient episodes of rapid climate 29 warming (high confidence) (Bond and Grasby, 2017; Benton, 2018; Penn et al., 2018; Them et al., 2018; 30 Chen and Xu, 2019). The role of ocean acidification in ancient extinctions is yet to be resolved (low 31 confidence) (Clapham and Payne, 2011; Clarkson et al., 2015; Jurikova et al., 2020; Müller et al., 2020). 32 Do Not Cite, Quote or Distribute 3-79 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 3.4.3.3.2 Community Structure and Biodiversity 2 Observed contemporary changes 3 Ocean temperature is a major driver of species richness in the global ocean at evolutionary timescales 4 (Tittensor et al., 2010; Chaudhary et al., 2021). This, together with temperature-driven range and phenology 5 shifts evident across taxa and ocean ecosystems (Sections 3.4.3.1, 3.4.3.2), suggests that recent ocean 6 warming (Section 3.2.2.1) should alter biodiversity at regional to global scales. Since previous assessments 7 (Table 3.20), the most common evidence supporting these expected changes is replacement of cold-adapted 8 species by warm-adapted species within an ecosystem as waters warm (Worm and Lotze, 2021). Known as 9 tropicalisation (Section 3.4.2.3), this phenomenon has been attributed to ocean warming (medium to high 10 confidence) in communities as diverse as kelp, invertebrates, plankton and fish (Burrows et al., 2019; 11 Flanagan et al., 2019; Ajani et al., 2020; Villarino et al., 2020; Punzón et al., 2021; Smith et al., 2021). 12 13 14 Table 3.20: Summary of previous IPCC assessments of community composition and biodiversity. Observations Projections AR5: (Hoegh-Guldberg et al., 2014; Pörtner et al., Spatial shifts of marine species due to projected warming 2014) will cause high-latitude invasions and high local- The paleoecological record shows that global climate extinction rates in the tropics and semi-enclosed seas changes comparable in magnitudes to those projected for (medium confidence). the 21st century under all scenarios resulted in large- scale biome shifts and changes in community Species richness and fisheries catch potential are composition; and that for rates projected under RCP6 projected to increase, on average, at mid and high and 8.5 were associated with species extinctions in some latitudes (high confidence) and decrease at tropical groups (high confidence). latitudes (medium confidence). Loss of corals due to bleaching has a potentially critical Shifts in the geographical distributions of marine species influence on the maintenance of marine biodiversity in cause changes in community composition and the tropics (high confidence). interactions. Thereby, climate change will reassemble communities and affect biodiversity, with differences over time and between biomes and latitudes (high confidence). Models are currently useful for developing scenarios of directional changes in net primary productivity, species distributions, community structure, and trophic dynamics of marine ecosystems, as well as their implications for ecosystem goods and services under climate change. However, specific quantitative projections by these models remain imprecise (low confidence). SROCC (Bindoff et al., 2019) Poleward range shifts are projected to decrease species Ocean warming has contributed to observed changes in richness in tropical oceans, counterbalanced by increases biogeography of organisms ranging from phytoplankton in mid to high-latitude regions, leading to global-scale to marine mammals (high confidence), consequently species turnover (medium confidence on trends, low changing community composition (high confidence), and confidence on magnitude because of model uncertainties in some cases, altering interactions between organisms and limited number of published model simulations). and ecosystem function (medium confidence). The projected intensity of species turnover is lower under low-emission scenarios (high confidence). Projections from multiple fish species distribution models show hotspots of decrease in species richness in the Indo-Pacific region, and semi-enclosed seas such as the Red Sea and Persian Gulf (medium evidence, high agreement). In addition, geographic barriers, such as land, bounding the poleward species range edge in semi- enclosed seas or low-oxygen water in deeper waters are projected to limit range shifts, resulting in larger relative decrease in species richness (medium confidence). Do Not Cite, Quote or Distribute 3-80 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report The large variation in sensitivity of different zooplankton taxa to future conditions of warming and ocean acidification suggests elevated risk to community structure and inter-specific interactions of zooplankton in the 21st century (medium confidence). 1 2 3 At local to regional scales, tropicalisation often increases species richness where warm-water species extend 4 their ranges to overlap with existing communities, and decreases species richness where warming waters 5 extirpate species (medium to high confidence) (Friedland et al., 2020a; Chaudhary et al., 2021; Worm and 6 Lotze, 2021). Latitudinal estimates from catalogued observations show declining species richness in 7 equatorial waters over the past 50 years, with concomitant increases in species richness at mid-latitudes; the 8 pattern is especially prominent in free-swimming pelagic species (Figure 3.18, Chaudhary et al., 2021). 9 Similar patterns among marine animals have been described previously for historical warming events (Song 10 et al., 2020b). Tropicalisation is associated with increased representation of herbivorous species (Vergés et 11 al., 2016; Zarco-Perello et al., 2020; Smith et al., 2021), although observations and theory suggest that 12 dietary generalism can also favour range-shifting species (Monaco et al., 2020; Wallingford et al., 2020). 13 14 Projected changes 15 At the community level, the magnitude and shape of biodiversity changes differ, depending on what groups 16 are considered (medium confidence) (Chaudhary et al., 2021). Molecular-based richness measures indicate 17 that the most dramatic increases in diversity relative to current conditions are expected for photosynthetic 18 eukaryotes and copepods in the Arctic Ocean (Ibarbalz et al., 2019). However, component eukaryotic taxa, 19 for example diatoms Busseni et al. (2020), are projected to lose diversity by 2100 under RCP8.5. Ecosystem 20 models project a decline in nutrient supply that drives the disappearance of less-competitive and larger 21 phytoplankton types, leading to extinction of up to 30% of diatom types, particularly in the northern 22 hemisphere, by 2100 under RCP8.5 (Henson et al., 2021). Models further suggest that high latitudes are 23 likely to encounter entirely novel phytoplankton communities by 2100 under RCP8.5 (100% change in 24 community composition, Dutkiewicz et al., 2019; Reygondeau et al., 2020). At the polar edges, the increased 25 richness is projected to coincide with high species turnover and increasing dominance of smaller 26 phytoplankton types (Henson et al., 2021). These imply pronounced changes to both the oceans' ecological 27 and biogeochemical function, as regions dominated by small phytoplankton typically support less-productive 28 food webs (Section 3.4.3.4, Stock et al., 2017; Armengol et al., 2019) and sequester less particulate organic 29 carbon in the deep ocean (Section 3.4.3.5, Mouw et al., 2016; Cram et al., 2018) than areas dominated by 30 larger size classes (high confidence). 31 32 The profound climatic and environmental changes projected for the Arctic region by 2100 (Cross-Chapter 33 Paper 6) are also anticipated to alter the composition of apex assemblages like marine mammals (Albouy et 34 al., 2020, Box 3.2). Under both RCP2.6 and 8.5 scenarios the most vulnerable marine mammal species will 35 be the North Pacific right whale (Eubalaena japonica, listed as an endangered species (IUCN, 2020)) and the 36 gray whale (Eschrichtius robustus, which has critically endangered subpopulations (IUCN, 2020)). The 37 extinction of the most-vulnerable species will disproportionately eliminate unique and important 38 evolutionary lineages as well as functional diversity, with consequent impacts throughout the entire marine 39 ecosystem (section 3.3.4). More generally, future warming and acidification simulated in mesocosm 40 experiments support projections of a substantial increase in biomass and productivity of primary producers 41 and secondary consumers, but a decrease by >40% of primary consumers (Nagelkerken et al., 2020). On 42 longer time scales, alteration of energy flow through marine food webs may lead to ecological tipping points 43 (Wernberg et al., 2016; Harley et al., 2017) after which the food web collapses into shorter, bottom-heavy 44 trophic pyramids (medium confidence). 45 46 Global projections anticipate a likely future reorganisation of marine life of variable magnitude, contingent 47 on emission scenario (Beaugrand et al., 2015; Jones and Cheung, 2015; Barton et al., 2016; García Molinos 48 et al., 2016; Nagelkerken et al., 2020; Henson et al., 2021). Marine organism redistributions projected under 49 RCP4.5 and RCP8.5 include extirpations and range contractions in the tropics, strongly decreasing tropical 50 biodiversity, and range expansions at higher latitudes, associated with increased diversity and 51 homogenisation of marine communities (Figure 3.18b). Under continuing climate change, the projected loss 52 of biodiversity may ultimately threaten marine ecosystem stability (medium confidence) (Albouy et al., 2020; Do Not Cite, Quote or Distribute 3-81 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Nagelkerken et al., 2020; Henson et al., 2021), altering both the functioning and structure of marine 2 ecosystems and thus affecting service provisioning (medium confidence) (Section 3.5, Ibarbalz et al., 2019; 3 Righetti et al., 2019). 4 5 However, biodiversity observations remain sparse, and statistical and modelling tools can provide conflicting 6 diversity information (e.g., Righetti et al., 2019; Dutkiewicz et al., 2020) because correlative approaches 7 assume that the modern-day relationship between marine species distribution and environmental conditions 8 remains the same into the future, whereas mechanistic models permit marine species to respond dynamically 9 to changing environmental forcing. Moreover, existing global projections of future biodiversity 10 disproportionately focus on the effects sea surface temperature (Thomas et al., 2012), typically overlooking 11 other factors such as ocean acidification, deoxygenation and nutrient availability (Section 3.2.3), and often 12 failing to account for natural adaptation (e.g., Section 3.3.4, Box 3.1, Barton et al., 2016; Henson et al., 13 2021). 14 15 16 17 Figure 3.18: Changes in latitudinal marine species richness latitudinal distribution. (a) Observed species richness for 18 three historical periods. The observed latitudinal patterns in species richness, are for a suite of taxonomic groups based 19 on 48,661 marine species (Chaudhary et al., 2021). (b) Projected changes in species richness under RCP4.5 and RCP8.5 20 calculated as differences by grid cell by 2100 relative to 2006. Latitudinal global median (5° moving average) (based on 21 Figure 1b,c in García Molinos et al., 2016). The projected latitudinal patterns in changes in species richness under climate 22 change are based on a numerical model that includes species-specific information across a suite of taxonomic groups, 23 based on 12,796 marine species (García Molinos et al., 2016). 24 25 26 [START BOX 3.2 HERE] 27 28 Box 3.2: Marine Birds and Mammals 29 30 Marine birds (seabirds and shorebirds) and mammals include charismatic species and species that are 31 economically, culturally and ecologically important (Sydeman et al., 2015; Albouy et al., 2020; Pimiento et 32 al., 2020). Their long generation times and slow population growth suggests limited evolutionary resilience 33 to rapid climate change (Section 3.3.4, Sydeman et al., 2015; Miller et al., 2018). According to the Red List 34 Species Assessments of the International Union for Conservation of Nature (IUCN, 2020), the greatest 35 current hazards to these groups include human use of biological resources and areas, invasive species and 36 pollution (Figure Box3.2.1, Dias et al., 2019; Lusseau et al., 2021). Impacts of climate change and severe 37 weather are ranked among the five most-important hazards, influencing 131 and 45 bird and mammal 38 species, respectively (see Figure Box 3.2.1 for selection of species), including 24 bird and seven mammal 39 species that are currently listed as endangered, critically endangered or threatened. Furthermore, according to Do Not Cite, Quote or Distribute 3-82 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 these IUCN assessments, climate change and severe weather are expected to impact an additional 122 and 18 2 marine bird and mammal species over the next 50­100 years, respectively (Figure Box 3.2.1 Dias et al., 3 2019). 4 5 Marine birds and mammals are vulnerable to climate-induced loss of breeding and foraging habitats such as 6 sea ice (Section 3.4.2.12), sandy beaches (Section 3.4.2.6), salt marshes (Section 3.4.2.5) and seagrass beds 7 (high confidence) (Section 3.4.2.5, Sydeman et al., 2015; Bindoff et al., 2019; Ropert-Coudert et al., 2019; 8 Von Holle et al., 2019; Albouy et al., 2020; Amano et al., 2020; Bestley et al., 2020; Grose et al., 2020). 9 With warming, shorebird population abundances decline in the tropics, likely due to heat stress and habitat 10 loss, and increase at higher latitudes (Amano et al., 2020). Marine mammals dependent on sea-ice habitat are 11 particularly vulnerable to warming (medium confidence) (Albouy et al., 2020; Bestley et al., 2020; Lefort et 12 al., 2020), yet vulnerability can differ between populations. Ongoing sea-ice loss is decreasing some polar 13 bear populations while others remain stable, likely related to past harvesting history, regional differences in 14 sea-ice phenology and ecosystem productivity (Hamilton and Derocher, 2019; Molnár et al., 2020). 15 Nevertheless, even under an intermediate emission scenario RCP4.5, increasing ice-free periods will likely 16 reduce both recruitment and adult survival across most polar bear populations over the next four decades, 17 threatening their existence (medium confidence) (Figure Box3.2.2, Molnár et al., 2020). 18 19 Climate change is affecting marine food-web dynamics (high confidence) (Sections 3.4.2, 3.4.3), and the 20 vulnerability and adaptive capacity of marine birds and mammals to such changes is linked to the species' 21 breeding and feeding ecology. Higher-vulnerability species include central-place foragers (confined to, for 22 example, breeding colonies fixed in space), diet and habitat specialists, and species with restricted 23 distributions such as marine mammal populations in SES (medium confidence) (McMahon et al., 2019; 24 Ropert-Coudert et al., 2019; Albouy et al., 2020; Grose et al., 2020; Sydeman et al., 2021). Surface-feeding 25 and piscivorous marine birds appear to be more vulnerable to food-web changes than diving seabirds and 26 planktivorous seabirds (medium confidence) (Sydeman et al., 2021). During the 2014­2015 Pacific 27 heatwave, around one million piscivorous common murres died along a 1500 km coastal stretch in the 28 Pacific USA due to reduced prey availability (Jones et al., 2018b; Piatt et al., 2020). Marine birds are 29 vulnerable to phenological shifts in food-web dynamics, as they have limited phenotypic plasticity of 30 reproductive timing, with potentially little scope for evolutionary adaptation (medium confidence) (Keogan 31 et al., 2018), although changes in reproduction timing are observed in several species (Section 3.4.4.1, 32 Sydeman et al., 2015; Descamps et al., 2019; Sauve et al., 2019). There is limited evidence of marine 33 mammals' capacity to adapt to shifting phenologies, but observed responses include changes in the onset of 34 migrations, moulting and breeding (Section 3.4.4.1, Ramp et al., 2015; Hauser et al., 2017; Beltran et al., 35 2019; Bowen et al., 2020; Szesciorka et al., 2020). 36 37 Increased emergence of infectious disease in mammals and birds is expected with ocean warming, due to 38 new transmission pathways from changing species distributions, higher species densities caused by habitat 39 loss, and increased vulnerability due to environmental stress on individuals (limited evidence) (Sydeman et 40 al., 2015; VanWormer et al., 2019; Sanderson and Alexander, 2020). Marine birds and mammals are likely to 41 suffer from increased mortalities due to increasing frequencies of HABs, and of extreme weather, at sea, on 42 sea ice, and in terrestrial breeding habitats (Broadwater et al., 2018; Gibble and Hoover, 2018; Ropert- 43 Coudert et al., 2019; Grose et al., 2020). Also, climate-change driven distributional shifts have strengthened 44 interactions with other anthropogenic impacts, through, for example, increasing risks of ship strikes and 45 bycatch (medium confidence) (e.g., Hauser et al., 2018; Krüger et al., 2018; Record et al., 2019; Santora et 46 al., 2020). 47 48 Do Not Cite, Quote or Distribute 3-83 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure Box 3.2.1: Hazard assessment for marine birds and mammals. Number of (a) marine birds and (b) mammals 3 currently impacted by different hazards (blue), and numbers of additional species expected to be exposed to these 4 threats over the next 50­100 years (red), as assessed in the International Union for Conservation of Nature Red List 5 (IUCN, 2020). Seabird species include species in the key orders Sphenisciformes, Pelecaniformes, Suliformes, 6 Anseriformes, Procellariiformes and Charadriiformes categorised as inhabitants of marine ecosystems (n = 483 species, 7 assessed in the period 2016­2019). Marine mammal species include the species reviewed by Lusseau et al. (2021) (n = 8 136 species, assessed in the period 2008­2019). 9 10 Do Not Cite, Quote or Distribute 3-84 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure Box 3.2.2: Modelled risk timelines for demographic impacts on circumpolar polar bear subpopulations, and 3 associated confidence assessments, due to extended fasting periods with loss of sea ice. Years of first impact on cub 4 recruitment (yellow), adult male survival (blue) and adult female survival (red) are shown for the (a) RCP4.5 and (b) 5 RCP8.5. Data from Molnár et al. (2020). 6 7 8 [END BOX 3.2 HERE] 9 10 11 3.4.3.3.3 Abrupt ecosystem shifts and extreme events 12 13 Climate-change driven changes in ocean characteristics and the frequency and intensity of extreme events 14 (Section 3.2) increase the risk of persistent, rapid and abrupt ecosystem change (very high confidence), often 15 referred to as ecosystem collapses or regime shifts (AR6 WGI Chapter 9, Collins et al., 2019a; Canadell and 16 Jackson, 2021; Ma et al., 2021). Such abrupt changes include altering ecosystem structure, function and 17 biodiversity outside the range of natural fluctuations (Collins et al., 2019a; Canadell and Jackson, 2021). 18 They can involve mass mortality events and `tipping points' or `critical transitions,' where strong positive 19 feedbacks within an ecosystem lead to self-sustaining change (Figure 3.19a, Scheffer et al., 2012; Möllmann 20 et al., 2015; Biggs et al., 2018). Abrupt ecosystem shifts have been observed in both large open-ocean 21 ecosystems and coastal ecosystems (Section 3.4.2) with dramatic social consequences through significant 22 loss of diverse ecosystem services (high confidence) (Section 3.5, Biggs et al., 2018; Pinsky et al., 2018; 23 Beaugrand et al., 2019; Collins et al., 2019a; Filbee-Dexter et al., 2020b; Huntington et al., 2020; Trisos et 24 al., 2020; Turner et al., 2020b; Canadell and Jackson, 2021; Ma et al., 2021; Ruthrof et al., 2021). A 25 summary of previous assessments of abrupt ecosystem shifts and extreme events are provided in Table 3.21. 26 27 28 Table 3.21: Summary of previous IPCC assessments of observed and projected abrupt ecosystem shifts and extreme 29 events. Observations Projections AR5 (Wong et al., 2014) Observations of abrupt ecosystem shifts and extreme Warming and acidification will lead to coral bleaching, events were not assessed in this report. mortality, and decreased constructional ability (high confidence), making coral reefs the most vulnerable marine ecosystem with little scope for adaptation. Temperate seagrass and kelp ecosystems will decline with the increased frequency of heatwaves and sea temperature extremes as well as through the impact of invasive subtropical species (high confidence). Do Not Cite, Quote or Distribute 3-85 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report SROCC (Collins et al., 2019a) Marine heatwaves (MHWs), periods of extremely high Marine heatwaves are projected to further increase in ocean temperatures, have negatively impacted marine frequency, duration, spatial extent and intensity (maximum organisms and ecosystems in all ocean basins over the last temperature) (very high confidence). Climate models two decades, including critical foundation species such as project increases in the frequency of marine heatwaves by corals, seagrasses and kelps (very high confidence) 2081­2100, relative to 1850­1900, by approximately 50 times under RCP8.5 and 20 times under RCP2.6 (medium confidence). Extreme El Niño and La Niña events are projected to likely increase in frequency in the 21st century and to likely intensify existing hazards, with drier or wetter responses in several regions across the globe. Extreme El Niño events are projected to occur about as twice as often under both RCP2.6 and RCP8.5 in the 21st century when compared to the 20th century (medium confidence). Limiting global warming would reduce the risk of impacts of MHWs, but critical thresholds for some ecosystems (e.g., kelp forests, coral reefs) will be reached at relatively low levels of future global warming (high confidence) 1 2 3 Abrupt ecosystem shifts are associated with large-scale patterns of climate variability (Alheit et al., 2019; 4 Beaugrand et al., 2019; Lehodey et al., 2020), some of which are projected to intensify with climate change 5 (medium confidence) (WGI AR6 Chapter 1, Wang et al., 2017a; Collins et al., 2019a; Chen et al., 2021). 6 Over the past 60 years, abrupt ecosystem shifts have generally followed El Niño/Southern Oscillation events 7 of any strength, but some periods had geographically limited ecological shifts (~0.25% of the global ocean in 8 1984­1987) and others more extensive shifts (14% of the global ocean in 2012­2015) (medium confidence) 9 (Figure 3.19b, Beaugrand et al., 2019). Typically, interacting drivers, such as eutrophication and overharvest, 10 reduce ecosystem resilience to climate extremes (e.g., MHW, cyclones) or gradual warming, and hence 11 promote ecosystem shifts (high confidence) (Figure 3.19a, Rocha et al., 2015; Biggs et al., 2018; Babcock et 12 al., 2019; Turner et al., 2020b; Bergstrom et al., 2021; Canadell and Jackson, 2021; Tait et al., 2021). Also, 13 shifts in different ecosystems may be connected through common drivers or through cascading effects 14 (medium confidence) (Rocha et al., 2018a). 15 16 Recent MHWs (Section 3.2.2.1) have caused major ecosystem shifts and mass mortality in oceanic and 17 coastal ecosystems, including corals, kelp forests and seagrass meadows (Sections 3.4.2.1, 3.4.2.3, 3.4.2.5, 18 3.4.2.6, 3.4.2.10, Cross-Chapter Box MOVING SPECIES in Chapter 5 and Cross-Chapter Box EXTREMES 19 in Chapter 2), with dramatic declines in species foundational for habitat formation or trophic flow, 20 biodiversity declines, and biogeographic shifts in fish stocks (very high confidence) (Table 3.15, Cross- 21 Chapter Box MOVING SPECIES in Chapter 5, Canadell and Jackson, 2021). Three major bleaching 22 episodes on Australia's Great Barrier Reef in 5 years corresponded with extreme temperatures in 2016, 2017 23 and 2020 (Pratchett et al., 2021). Between 1981 and 2017, marine heatwaves have increased more than 20- 24 fold due to anthropogenic climate change (Section 3.2.2.1, WGI AR6 Chapter 9, Laufkötter et al., 2020; 25 Fox-Kemper et al., 2021), increasing the risk of abrupt ecosystem shifts (high confidence) (Figure 3.19a, 26 Cross-Chapter Box EXTREMES in Chapter 2, van der Bolt et al., 2018; Garrabou et al., 2021; Wernberg, 27 2021). 28 29 Ecosystems can recover from abrupt shifts (e.g., Babcock et al., 2019; Christie et al., 2019; Medrano et al., 30 2020). However, where climate change is a dominant driver, ecosystem collapses increasingly cause 31 permanent transitions (high confidence), although the extents of such transitions depend on emission 32 scenario (Trisos et al., 2020; Garrabou et al., 2021; Klein et al., 2021; Pratchett et al., 2021; Wernberg, 33 2021). Over the coming decades, MHW are projected to very likely become more frequent under all emission 34 scenarios (Section 3.2, WGI AR6 Chapter 9, Fox-Kemper et al., 2021), with intensities and rates too high for 35 recovery of degraded foundational species, habitats, or biodiversity (medium confidence) (Babcock et al., 36 2019; Garrabou et al., 2021; Klein et al., 2021; Serrano et al., 2021; Wernberg, 2021). Emission pathways 37 that result in temperature overshoot above 1.5oC will increase the risks of abrupt and irreversible shifts in 38 coral reefs and other vulnerable ecosystems (Section 3.4.4). Do Not Cite, Quote or Distribute 3-86 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 3 Figure 3.19: Observed ecological regime shifts and their drivers in the oceans. (a) A conceptual representation of 4 ecosystem resilience and regime shifts. Shift from Regime 1 to Regime 2 can be triggered by either a large shock (i.e., 5 an abrupt environmental transition) or gradual internal or external change that erodes the dominant balancing feedbacks, 6 reducing ecosystem resilience (indicated by the shallower dotted line, relative to the deeper `valley' reflecting higher 7 resilience). Figure based on Biggs et al. (2018). (b) The sum of the magnitude and extent of the abrupt community shifts 8 that has been estimated at each geographic cell in the global ocean during 1960­2014, calculated as the ratio of the 9 amplitude of the change in a particular year to the average magnitude of the change over the entire time series (thus is 10 dimensionless). Figure based on Beaugrand et al. (2019). 11 12 13 3.4.3.3.4 Time of emergence ­ species exposure to altered environments 14 Since SROCC, more studies have assessed the time of emergence for climate-impact drivers (Section 3.2.3), 15 and the ecosystem attributes through which the impacts manifest. However, as in previous assessments 16 (Table 3.22), the time of emergence for a given driver or ecosystem attribute depends on the reference 17 period, the definition of the signal emergence threshold and the spatial and temporal scales considered (Box 18 5.1 in SROCC, Kirtman et al., 2013; Bindoff et al., 2019). 19 20 Do Not Cite, Quote or Distribute 3-87 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Table 3.22: Summary of previous IPCC assessments of projected time of emergence. System Projections Coastal Multiple climate-impact drivers will emerge in the 21st Epipelagic century under RCP8.5, while the time of emergence will Open Ocean be later and with few climatic hazards under RCP2.6. Non- climate impacts such as eutrophication add to, and in some Deep sea cases, exacerbate these large-scale slow climate drivers beyond biological thresholds at local scale (e.g., deoxygenation). Observed range shifts in response to climate change in some regions such as the north Atlantic are strongly influenced by warming due to both multi-decadal climate change and variability, suggesting that there is a longer time-of-emergence of range shifts from natural variability and a need for longer biological time series for robust attribution. The timing for five primary drivers of marine ecosystem change (surface warming and acidification, oxygen loss, nitrate concentration and net primary production change) are all prior to 2100 for >60% of the ocean area under RCP8.5 and over 30% under RCP2.6 (very likely). Anthropogenic signals are expected to remain detectable over large parts of the ocean even under the RCP2.6 scenario for pH and SST but are likely to be less conspicuous for nutrients and NPP in the 21st century. For example, for the open ocean, the anthropogenic pH signal in Earth System Models (ESM) historical simulations is very likely to have emerged for three-quarters of the ocean prior to 1950 and it is very likely over 95% of the ocean has already been affected, with little discernible difference between scenarios. The climate signal of oxygen loss will very likely emerge by 2050 with a very likely range of 59­ 80% by 2031­2050 and increasing with a very likely range of 79­91% of the ocean area by 2081­2100 (RCP8.5 emissions scenario). The emergence of oxygen loss is smaller in area under RCP2.6 scenario in the 21st century and by 2090 the area where emergence is evident is declining. It has also been shown that signatures of altered oxygen solubility or utilisation may emerge earlier than for oxygen levels. Emergence of risk is expected to occur later at around the mid-21st century under RCP8.5 for abyssal plain and chemosynthetic ecosystems (vents and seeps). All deep seafloor ecosystems are expected to be subject to at least moderate risk under RCP8.5 by the end of the 21st century, with cold water corals undergoing a transition from moderate to high risk below 3ºC. 2 3 4 Anthropogenically driven changes in chlorophyll-a concentrations across an ensemble of 30 ESMs are 5 expected to exceed natural variability under RCP8.5 by 2100 in ~65­80% of the global oceans, when the 6 natural variability is calculated using the ensemble's standard deviation (Schlunegger et al., 2020). However, 7 if two standard deviations are used, then significant trends in chlorophyll-a concentration are expected under 8 RCP8.5 across ~31% of the global oceans by 2100 (Dutkiewicz et al., 2019). In contrast, the anthropogenic 9 signal in phytoplankton community structure, which has a lower natural variability, will emerge under 10 RCP8.5 across 63% of the ocean by 2100 when two standard deviations are used (limited evidence) 11 (Dutkiewicz et al., 2019). 12 13 The time of emergence of climate impacts on ecosystems will be modulated jointly by species-specific 14 adaptation potential (Section 3.3.4, Jones and Cheung, 2018; Collins et al., 2020; Gamliel et al., 2020; Miller Do Not Cite, Quote or Distribute 3-88 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 et al., 2020a), speed of range shifts and spatial reorganisation (high confidence) (Sections 3.3, 3.4.2,3.4.3). 2 These ecosystem responses complicate projections of the time of emergence of environmental properties that 3 impact biogeochemical cycling (Schlunegger et al., 2019; Schlunegger et al., 2020; Wrightson and 4 Tagliabue, 2020), ecosystem structure and biodiversity (Figure 3.20a,c, Dutkiewicz et al., 2019; Trisos et al., 5 2020), and higher trophic levels, including fisheries targets (Cheung and Frölicher, 2020). Better accounting 6 for multiple interacting factors in ESMs (Box 3.1), which will provide insight into how marine ecosystems 7 will respond to future climate (high confidence). 8 9 The time of emergence of ecosystem responses supports planning for specific time-bound actions to reduce 10 risks to ecosystems (Sections 3.6.3.2.1 Bruno et al., 2018; Trisos et al., 2020). Although under RCP 8.5, 11 climate refugia from SST after 2050 are primarily in the Southern Ocean in tropical waters, these refugia are 12 mainly from deoxygenation (Bruno et al., 2018). Marine assemblages in these places will be exposed to 13 unprecedented temperatures after 2050, peaking in 2075 (Figure 3.20a,b, Trisos et al., 2020)(). In contrast, 14 changes in phytoplankton community structure will emerge earlier, and primarily in the Pacific Ocean 15 subtropics, and through much of the North Atlantic Ocean (Figure 3.20c,d, Dutkiewicz et al., 2019). Under 16 RCP8.5, changes in phytoplankton community structure and, to a lesser extent, exposure of marine species to 17 unprecedented temperatures, will emerge earlier in MPAs, covering ~7.7% of the global oceans, (Section 18 3.6.2.3.2.1, UNEP-WCMC and IUCN, 2020; UNEP-WCMC and IUCN, 2021) compared to non-MPAs 19 (Figure 3.20b,d). Such assessment can support planning for future MPA placement and extent. Because 20 MPAs can serve as refugia from non-climate drivers (Sections 3.6.2.3, 3.6.3.2.1), they facilitate opportunities 21 for adaptation among marine species and communities in coastal oceans (Section 3.4.2). 22 23 24 25 Figure 3.20: Time of exposure to altered environments. (a) Simulated spatial variation in the time of exposure of 26 marine species to unprecedented temperatures under RCP8.5. Time of exposure is quantified as the median year after 27 which local species are projected to encounter temperatures warmer than the historical maximum within their full 28 geographic range for a period of at least five years. This estimate is based on 22 Coupled Model Intercomparison 29 Project 5 (CMIP5) models, and is drawn from data presented by Trisos et al. (2020). Only regions that have times of 30 emergence by 2100 are shown. (b) The distribution in the time of exposure to unprecedented temperatures within 31 marine assemblages (Trisos et al., 2020) under RCP8.5 in marine protected areas (in turquoise) and in non-marine 32 protected areas (in purple). Values were calculated after regridding to equal-area 0.5° hexagons. (c) Time of emergence 33 for phytoplankton community structure changes (based on a proxy ­ ecosystem-model reflectance at 500 nm) under 34 RCP8.5. Only regions with statistically significant (p <0.05) trends, that are presently largely ice-free and that have 35 times of emergence by 2100 are shown. Figure based on the results of one model numerical model from Dutkiewicz et 36 al. (2019). (d) The distribution in the time of emergence for changes in phytoplankton community structure (same proxy 37 as in Panel c) (Dutkiewicz et al., 2019) under RCP8.5 in marine protected areas (in turquoise) and in non-marine 38 protected areas (in purple). Values were calculated after regridding to equal-area 0.5° hexagons. 39 40 Do Not Cite, Quote or Distribute 3-89 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 3.4.3.4 Biomass 2 3 3.4.3.4.1 Observed changes 4 Observed changes in biomass in the global ocean, beyond those for phytoplankton (Table 3.23), have not 5 routinely been attributed to climate-impact drivers, but rather to the compound effects of multiple drivers, 6 especially fishing (Christensen et al., 2014; Palomares et al., 2020). We therefore do not assess observed 7 changes in ocean biomass here. 8 9 3.4.3.4.2 Projected changes 10 Zooplankton biomass 11 Based on an ensemble of CMIP5 ESMs, SROCC projected declines in global zooplankton biomass by 2100 12 dependent on emission scenario (low confidence) (Table 3.23). The new CMIP6 ESM ensemble projects a 13 decline in global zooplankton biomass by ­3.9% ± 8.2%, (very likely range) and ­9.0% ± 8.9% in the period 14 2081­2100 relative to 1995­2014 under SSP1-2.6 and SSP5-8.5, respectively (Figure 3.21d, Kwiatkowski et 15 al., 2020), thus reinforcing the SROCC assessment albeit with greater inter-model uncertainties. 16 17 18 Table 3.23: Summary of previous IPCC assessments of changes in open ocean and deep sea biomass. Measure Observations Projections AR5 WGII: (Hoegh-Guldberg et al., 2014; Pörtner et al., 2014) Chlorophyll-a/phytoplankton biomass Phytoplankton biomass: The Owing to contradictory observations approximately 15-year archived time there is currently uncertainty about the series of satellite-chlorophyll (as a future trends of major upwelling proxy of phytoplankton biomass) is too systems and how their drivers short to reveal trends over time and (enhanced productivity, acidification, their causes (WGII AR5 Section 6.1.2, and hypoxia) will shape ecosystem Pörtner et al., 2014). characteristics (low confidence) (WGII AR5 Chapter 6 Executive Summary, Chlorophyll concentrations measured Pörtner et al., 2014). by satellites have decreased in the subtropical gyres of the North Pacific, Indian, and North Atlantic Oceans by 9%, 12%, and 11%, respectively, over and above the inherent seasonal and interannual variability from 1998 to 2010 (high confidence; p-value 0.05). Significant warming over this period has resulted in increased water column stratification, reduced mixed layer depth, and possibly decreases in nutrient availability and ecosystem productivity (limited evidence, medium agreement). The short time frame of these studies against well-established patterns of long-term variability leads to the conclusion that these changes are about as likely as not due to climate change (WGII AR5 Chapter 30 Hoegh-Guldberg et al., 2014). Animal biomass The climate-change-induced intensification of ocean upwelling in some eastern boundary systems, as observed in the last decades, may lead to regional cooling rather than warming of surface waters and cause enhanced productivity (medium confidence), but also enhanced hypoxia, acidification, and associated Do Not Cite, Quote or Distribute 3-90 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report biomass reduction in fish and invertebrate stocks. SROCC (Bindoff et al., 2019) Chlorophyll-a/phytoplankton biomass Changes reported in overall open ocean chlorophyll levels (a proxy of phytoplankton biomass) of <±1% yr­1 for individual time periods. Regionally, trends of ±4% between 2002­2015 for different regions are found when different satellite products are merged, with increases at high latitudes and moderate decreases at low latitudes (SROCC Section 5.2.2.6, Bindoff et al., 2019). Animal biomass Observed changes in open ocean and There is high agreement in model WGI AR6 Chapter 2 (Gulev et al., deep sea biomass were not assessed in projections that global zooplankton 2021) this report. biomass will very likely reduce in the 21st century, with projected decline under RCP8.5 almost doubled that of RCP2.6 (very likely). However, the strong dependence of the projected declines on phytoplankton production (low confidence) and simplification in representation of the zooplankton communities and foodweb render their projections having low confidence. The global biomass of marine animals, including those that contribute to fisheries, is projected to decrease by 4.3 ± 2.0% (95% confidence intervals) and 15.0 ± 5.9% under RCP2.6 and RCP8.5, respectively, by 2080­2099 relative to 1986­2005, while the decrease is around 4.9% by 2031­2050 across all RCP2.6 and RCP8.5 (very likely). Regionally, total animal biomass decreases largely in tropical and mid-latitude oceans (very likely). Projected decrease in upper ocean export of organic carbon to the deep seafloor is expected to result in a loss of animal biomass on the deep seafloor by 5.2­17.6% by 2090­2100 compared to the present (2006­2015) under RCP8.5 with regional variations (medium confidence). Some increases are projected in the polar regions, due to enhanced stratification in the surface ocean, reduced primary production and shifts towards small phytoplankton (medium confidence). The projected impacts on biomass in the abyssal seafloor are larger under RCP8.5 than RCP4.5 (very likely). Do Not Cite, Quote or Distribute 3-91 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report Chlorophyll-a/phytoplankton biomass The multi-sensor time series of Projected changes in open ocean and chlorophyll-a concentration has been deep sea biomass were not assessed in updated to cover two decades (1998­ this report. 2018). Global trends in chlorophyll-a for the last two decades are insignificant over large areas of the global oceans, but some regions exhibit significant trends, with positive trends in parts of the Arctic and the Antarctic waters (>3% yr­1), and both negative and positive trends (within ±3% yr­1), in parts of the tropics, subtropics and temperate waters. In the last two decades, the concentration of phytoplankton at the base of the marine food web, as indexed by chlorophyll concentration, has shown weak and variable trends in low and mid-latitudes and an increase in high latitudes (medium confidence). 1 2 3 Marine animal biomass 4 Using an ensemble of global-scale marine ecosystem and fisheries models (Fish-MIP, Tittensor et al., 2018) 5 with the CMIP5 ESM ensemble, SROCC concludes that projected ocean warming and decreased 6 phytoplankton production and biomass will reduce global marine animal biomass during the 21st century 7 (medium confidence). The simulated declines (with very likely range) are ­3.5 ± 4.8% and ­14.0 ± 14.6% 8 under RCP2.6 and RCP8.5, respectively, by 2080­2099 relative to 1995­2014 (SROCC Section 5.2.3, 9 Bindoff et al., 2019; Lotze et al., 2019)1. Updated Fish-MIP simulations with CMIP6 (Figure 3.21g,h,i) 10 confirm the projected decline in total marine animal biomass in the 21st century (Tittensor et al., 2021). The 11 simulated declines (with very likely range) are ­5.7% ± 4.1% and ­15.5% ± 8.5% under SSP1-2.6 and SSP5- 12 8.5, respectively, by 2080­2099 relative to 1995­2014 (Figure 3.21g), showing greater declines and lower 13 inter-model uncertainties (Tittensor et al., 2021). These declines result from combined warming and 14 decreased primary production (with low confidence in future changes in primary production, Section 3.4.3.5) 15 and are amplified at each trophic level within all ESM and marine ecosystem model projections across all 16 scenarios (medium confidence) (Kwiatkowski et al., 2019; Lotze et al., 2019; Tittensor et al., 2021). 17 However, there is limited evidence about how underlying food-web mechanisms amplify the climate signal 18 from primary producers to higher trophic levels, and several putative mechanisms have been proposed 19 (Section 3.4.4.2.2, Chust et al., 2014a; Stock et al., 2014; Kwiatkowski et al., 2019; Lotze et al., 2019; 20 Heneghan et al., 2021). As assessed in SROCC, the biomass projections contain considerable regional 21 variation, with declines in tropical to temperate regions and strong increases in total animal biomass are 22 projected in polar regions under high-emission scenarios, with climate-change effects that are spatially 23 similar but less pronounced under lower-emission scenarios (Figure 3.21b,c,e,f,h,i; Tai et al., 2019; Tittensor 24 et al., 2021). 25 26 Benthic biomass 27 SROCC assessed that reduced food supply to the deep sea will drive a reduction in abyssal seafloor biota by 28 2100 for RCP8.5 (Table 3.23). Simulations from one size-resolved benthic biomass model coupled to an 29 ocean-biogeochemistry model forced with the CMIP5 ESM HadGEM2-ES (Yool et al., 2017) project a 30 decline in the globally integrated total seafloor biomass of ­1.1% and ­17.6% by 2100 under RCP2.6 and 31 RCP8.5, respectively (limited evidence, high agreement). In waters shallower than 100 m, total benthic 32 biomass is projected to increase by 3.2% on average by 2100 under RCP8.5, primarily driven warming- 1 SROCC reported declines in total marine animal biomass have been recomputed using 1995­2014 as the baseline period and the very likely ranges (5­95%) are now computed from the model ensemble ranges assuming a normal distribution. Do Not Cite, Quote or Distribute 3-92 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 increased growth rates (Yool et al., 2013), while at depths >2000 m (representing 83% of the ocean 2 seafloor), declines of ­32% arise from climate-driven decreases in surface primary production and 3 particulate organic carbon (POC) flux to the seafloor (Yool et al., 2013; Kelly-Gerreyn et al., 2014; Yool et 4 al., 2015; Yool et al., 2017). These patterns are qualitatively similar under RCP2.6, except in the Pacific and 5 Indian Ocean basins, where some increased total seafloor biomass is projected (Yool et al., 2013). Updated 6 simulations with the same benthic biomass model (Kelly-Gerreyn et al., 2014) forced with the CMIP6 ESM 7 UKESM-1 project declines in total seafloor biomass of ­9.8% and ­13.0% by 2081­2100 relative to 1995­ 8 2014 for SSP1-2.6 and SSP5-8.5, respectively (Figure 3.21j,k,l, updated from Yool et al., 2017). These 9 projected changes in benthic biomass are based on limited evidence. Development of ensemble projections 10 forced with a range of ESMs and a benthic model that considers the ecological roles of temperature (Hunt 11 and Roy, 2006; Reuman et al., 2014), oxygen (Mosch et al., 2012) and ocean acidification (Andersson et al., 12 2011) will provide opportunities to better quantify uncertainty in projected declines in total seafloor biomass 13 under climate change. 14 15 Conclusions 16 Overall, ocean warming and decreased phytoplankton production and biomass will drive a global decline in 17 biomass for zooplankton (low confidence), marine animals (medium confidence) and seafloor benthos (low 18 confidence), with regional differences in the direction and magnitude of changes (high confidence). There is 19 increasing evidence that responses will amplify throughout the food web and at ocean depths, with relatively 20 modest changes in surface primary producers translating into substantial changes at higher trophic levels and 21 for deep-water benthic communities (medium confidence). 22 23 24 25 Figure 3.21: Projected change in marine biomass. Simulated global biomass changes of (a,b,c) surface phytoplankton, 26 (d,e,f) zooplankton, (g,h,i) animals and (j,k,l) seafloor benthos. In (a,d,g,j), the multi-model mean (solid lines) and very 27 likely range (envelope) over 2000­2100 relative to 1995­2014, for SSP1-2.6 and SSP5-8.5. Spatial patterns of 28 simulated change by 2090­2099 are calculated relative to 1995­2014 for (b,e,h,k) SSP1-2.6 and (c,f,i,l) SSP5-8.5. 29 Confidence intervals can be affected by the number of models available for the Coupled Model Intercomparison Project 30 6 (CMIP6) scenarios and for different variables. Only one model was available for panel (j), so no confidence interval is Do Not Cite, Quote or Distribute 3-93 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 calculated. For panels (a­f), the ensemble projections of global changes in phytoplankton and zooplankton biomasses 2 updated based on Kwiatkowski et al. (2019) include, under SSP1-2.6 and SSP5-8.5, respectively, a total of nine and 10 3 CMIP6 Earth System Models (ESMs). For panels (g,h,i), the ensemble projections of global changes in total animal 4 biomass updated based on Tittensor et al. (2021) include 6­9 published global fisheries and marine ecosystem models 5 from the Fisheries and Marine Ecosystem Model Intercomparison Project (Fish-MIP, Tittensor et al., 2018; Tittensor et 6 al., 2021), forced with standardised outputs from two CMIP6 ESMs. For panels (j,k,l), globally integrated changes in 7 total seafloor biomass have been updated based on Yool et al. (2017) with one benthic model (Kelly-Gerreyn et al., 8 2014) forced with the CMIP6 ESM HadGEM2-ES. 9 10 11 3.4.3.5 Changes in Primary Production and Biological Carbon Export Flux 12 13 3.4.3.5.1 Observed changes in primary production 14 Analyses of satellite-derived primary production over the past two decades (1998­2018) showed generally 15 weak and negative trends (up to -3.0%) at low and mid latitudes (Kulk et al., 2020). In contrast, positive 16 trends occurred in large areas of the South Atlantic and South Pacific Oceans, as well as in polar and coastal 17 (upwelling) regions (up to +4.5%, Cross-Chapter Paper 6, Kulk et al., 2020). Data-assimilating ocean 18 biogeochemical models estimate a global decline in primary production of 2.1% per decade in the period 19 1998­2015, driven by the shoaling mixed layer and decreasing nitrate concentrations (Gregg and Rousseaux, 20 2019). This is consistent with previous assessments that identified ocean warming and increased 21 stratification as the main drivers (high confidence) affecting the regional variability in primary production 22 Bindoff et al. (2019). However, as noted in SROCC and WGI AR6 Chapter 2 (Table 3.24, Gulev et al., 23 2021), observed inter-annual changes in primary production on global and regional scales are nonlinear and 24 largely influenced by natural temporal variability, providing low confidence in the trends. 25 26 27 Table 3.24: Summary of previous IPCC assessments of ocean primary production and carbon export flux. Process Observed Impacts Projected Impacts SROCC (Bindoff et al., 2019) Open ocean Past open ocean productivity trends, Net primary productivity is very likely primary production including those determined by to decline by 4­11% by 2081­2100, satellites, are appraised with low relative to 2006­2015, across CMIP5 confidence, due to newly identified models for RCP8.5, but there is low region-specific drivers of microbial confidence for this estimate due to the growth and the lack of corroborating in medium agreement among models and situ time series datasets. the limited evidence from observations. The tropical ocean net primary productivity is very likely to decline by 7­16% for RCP8.5, with medium confidence as there are improved constraints from historical variability in this region. Open ocean carbon export Analyses of long-term trends in The projected changes in export primary production and particle export production can be larger than global production, as well as model primary production because they are simulations, reveal that increasing affected by both the net primary temperatures, leading to enhanced production changes, but also how stratification and nutrient limitation, shifts in food web structure modulates will have the greatest influence on the `transfer efficiency' of particulate decreasing the flux of particulate organic material, which then affects organic carbon (POC) to the deep the sinking speed and lability of ocean. However, different lines of exported particles through the ocean evidence (including observation, interior to the sea floor. modeling and experimental studies) provide low confidence on the As export production is a much better mechanistic understanding of how understood net integral of changing net climate-impact drivers affect different nutrient supply and can be constrained components of the biological pump in by interior ocean nutrient and oxygen the epipelagic ocean, as well as levels, there is medium confidence in Do Not Cite, Quote or Distribute 3-94 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report changes in the efficiency and projections for global export magnitude of carbon export in the deep production changes based on CMIP5 ocean. model runs. WGI AR6 Chapters 2 and 5 (Canadell et al., 2021; Gulev et al., 2021) Open ocean Global ocean marine primary In CMIP5 models run under RCP8.5, primary production production is estimated to be 47 ± 7.8 particulate organic carbon (POC) PgC yr­1 with low confidence because export flux is projected to decline by of the small number of recent studies 1­12% by 2100 (Taucher and and the insufficient length of the time Oschlies, 2011; Laufkötter et al., series analysed. A small decrease in 2015). Similar values are predicted in productivity is evident globally for the 18 CMIP6 models, with declines of period 1998­2015, but regional 2.5­21.5% (median ­14%) between changes are larger and of opposing 1900 and 2100 under the SSP5-8.5 signs (low confidence) (WGI AR6 scenario. The mechanisms driving Section 2.3.4.2.2, Gulev et al., 2021). these changes vary widely between models due to differences in parameterisation of particle formation, remineralisation and plankton community structure (WGI AR6 Section 5.4.4.2, Canadell et al., 2021). 1 2 3 3.4.3.5.2 Projected changes in primary production 4 Across 10 CMIP5 and 13 CMIP6 ESM ensembles, global mean net primary production is projected to 5 decline by 2080­2099 relative to 2006­2015, under all RCPs and SSPs (Kwiatkowski et al., 2020). 6 However, under comparable radiative forcing, the CMIP6 multi-model mean projections of primary 7 production declines (mean ± SD: ­0.56 ± 4.12% under SSP1-2.6, and ­3.00 ± 9.10% under SSP5-8.5) are 8 less than those of previous CMIP5 models (3.42 ± 2.47% under RCP2.6, and 8.54 ± 5.88% under RCP8.5) 9 (WGI AR6 Section 5.4.4.2, Kwiatkowski et al., 2020; Canadell et al., 2021). The inter-model uncertainty 10 associated with CMIP6 net primary production projections is larger than in CMIP5, and it is consistently 11 larger than the scenario uncertainty. For each SSP across the CMIP6 ensemble, individual models project 12 both increases and decreases in global primary production, reflecting a diverse suite of bottom-up and top- 13 down ecological processes, which are variously parameterised across models (Laufkötter et al., 2015; 14 Bindoff et al., 2019). Further, accurate simulation of many of the biogeochemical tracers upon which net 15 primary production depends (e.g., the distribution of iron, Tagliabue et al., 2016; Bindoff et al., 2019) 16 remains a significant and ongoing challenge to ESMs (high confidence) (Séférian et al., 2020). 17 18 Regionally, multi-model mean changes in primary production show generally similar patterns of large 19 declines in the North Atlantic and the western equatorial Pacific, while in the high latitudes, primary 20 production consistently increases in CMIP5 and CMIP6 by 2100 (Kwiatkowski et al., 2020, Cross-Chapter 21 Paper 6). In the Indian Ocean and sub-tropical North Pacific, which were regions of consistent net primary 22 production decline in CMIP5 projections (Bopp et al., 2013), the regional declines are reduced in magnitude, 23 less spatially extensive, and are typically less robust in CMIP6. Further assessment of simultaneous changes 24 in processes such as nutrient advection, nitrogen fixation, the microbial loop, and top-down grazing pressure 25 (WGI AR6 Section 5.4.4.2, Laufkötter et al., 2015; Bindoff et al., 2019; Canadell et al., 2021) are required to 26 fully understand the regional primary production response in CMIP6 (Kwiatkowski et al., 2020). Given the 27 regional variations in the estimates of primary production changes and the uncertainty in the representation 28 of the dominant drivers, there remains low confidence in the projected global decline in net primary 29 production. 30 31 3.4.3.5.3 Observed processes driving changes in global export flux 32 The SROCC medium confidence assessment that warming, stratification, declines in productivity and 33 changes in plankton community in the epipelagic zone result in reduced export of primary production to 34 deeper layers (Table 3.24) is supported by subsequent literature (Bach et al., 2019; Leung et al., 2021). POC 35 export efficiency is constrained by altered mixing and nutrient availability (Boyd et al., 2019; Lundgreen et 36 al., 2019), particle fragmentation (Briggs et al., 2020), as well as viral, microbial, and planktonic community 37 structure (Fu et al., 2016; Guidi et al., 2016; Flombaum et al., 2020; Kaneko et al., 2021) and metabolic rates Do Not Cite, Quote or Distribute 3-95 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 (Cavan et al., 2019). These processes are strongly interlinked and their net effect on primary production 2 export from the upper ocean remain difficult to quantify observationally (Boyd et al., 2019). Since SROCC, 3 there is increasing evidence that ocean deoxygenation can alter zooplankton community structure (Wishner 4 et al., 2018), zooplankton respiration rates (Cass and Daly, 2014; Cavan et al., 2017) and patterns of diel 5 vertical migration (Aumont et al., 2018), which may focus remineralisation of organic carbon at the upper 6 margins of OMZs (Section 3.4.3.4 on depth shifts due to OMZ, Bianchi et al., 2013; Archibald et al., 2019). 7 8 Data on export flux from the upper ocean are limited either in coverage and consistency (ship-board 9 sampling) or duration (sediment traps) and are subject to considerable spatial variability (as shown in 10 satellite observations, Boyd et al., 2019). As a result, trends are weak, inconsistent and often not statistically 11 significant (Lomas et al., 2010; Cael et al., 2017; Muller-Karger et al., 2019; Xie et al., 2019). Deep-ocean 12 fluxes are similarly equivocal (Smith et al., 2018; Fischer et al., 2019; Fischer et al., 2020). In coming years, 13 an increasing number of Argo floats equipped with bio-optical sensors should help improve estimates of 14 particle flux spatial and temporal variability (e.g., Dall'Olmo et al., 2016). 15 16 Projected changes 17 SROCC and WGI AR6 reported global declines in POC export flux, between ­8.9 to ­15.8% by 2100 18 relative to 2000 under RCP8.5 in CMIP5 models, and ­2.5 to ­21.5% (median value ­14%) between 1900 19 and 2100 under SSP5-8.5 in CMIP6 models (WGI AR6 5.4.4.2, Table 3.24, Bindoff et al., 2019; Canadell et 20 al., 2021). In CMIP5 model runs, the decrease in the sinking flux of organic matter from the upper ocean into 21 the ocean interior was strongly related to the changes in stratification that reduce net nutrient supply (Fu et 22 al., 2016; Bindoff et al., 2019), especially in tropical regions, and the projections for global export 23 production changes are reported with medium confidence. Increasing model complexity with more 24 widespread representation of ocean biogeochemical processes between CMIP5 and CMIP6, and inclusion of 25 more than one or two classes of phyto- and zooplankton will provide opportunities to improve assessments 26 of how climate-impact drivers affect different components of biological carbon pump in the epipelagic 27 ocean, as well as changes in the efficiency and magnitude of carbon export in the deep ocean (high 28 confidence) (Box 3.3, Le Quéré et al., 2016; Séférian et al., 2020; Wright et al., 2021). 29 30 31 [START BOX 3.3 HERE] 32 33 Box 3.3: Deep Sea Ecosystems 34 35 Deep-sea ecosystems include all waters below the 200 m isobath as well as the underlying benthos, and they 36 provide habitats for highly diversified and specialised biota, which play a key role in the cycling of carbon 37 and other nutrients (Figure Box3.3.1, Thurber et al., 2014; Middelburg, 2018; Snelgrove et al., 2018). The 38 deep sea covers >63% of Earth's surface (Costello and Cheung, 2010) and is exposed to climate-driven 39 changes in abyssal, intermediate, and surface waters that influence sinking fluxes of particulate organic 40 matter (high confidence) (Figure Box3.3.1, Sections 3.1, 3.2.1, 3.2.2, 3.4.3.4, WGII AR5 Section 30.5.7, 41 SROCC Sections 5.2.3, 5.2.4, Hoegh-Guldberg et al., 2014; Bindoff et al., 2019). These ecosystems are also 42 influenced by non-climate drivers, especially fisheries, oil and gas extraction (Thurber et al., 2014; Cordes et 43 al., 2016; Zhang et al., 2019a); cable laying (United Nations, 2021); and mineral resource exploration (Hein 44 et al., 2021); with proposed large-scale deep-sea mining a potential future source of impacts (Danovaro, 45 2018; Levin et al., 2020). 46 47 Ocean warming alters biological processes in deep-sea ecosystems in ways that affect deep-sea habitat, 48 biodiversity, and material processing. Enhancement of microbial respiration by warming attenuates sinking 49 POC, which has been associated with the globally projected declines in total seafloor biomass of ­9.8% and 50 ­13.0% by 2081­2100 relative to 1995­2014 under SSP1-2.6 and SSP5-8.5, respectively (limited evidence) 51 (Section 3.4.3.4). Additionally, climate-change-driven oxygen loss (Section 3.2.3.2, Luna et al., 2012; Belley 52 et al., 2016), and geographic shifts in predator distributions (Section 3.4.3.1) are anticipated to affect deep- 53 sea biodiversity (limited evidence, high agreement) (Smith et al., 2012; Morato et al., 2020). Complex 54 responses of some bathyal crustacean assemblages to environmental change suggest an increase in 55 phylogenetic diversity but limited decreases in abundances with temperature (Ashford et al., 2019). Acute 56 mortality of some reef-forming cold-water corals to laboratory-simulated warming (Lunden et al., 2014) 57 suggests that both long-term warming and the increase of MHWs in intermediate and deep waters (Elzahaby Do Not Cite, Quote or Distribute 3-96 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 and Schaeffer, 2019) could pose significant risk to associated ecosystems (high confidence). Thermal 2 tolerance thresholds (lethal and sub-lethal) of scleractinians in laboratory settings depend on their geographic 3 position and capacity for thermal adaptation, as well as other factors including food, oxygen and pH (medium 4 to high confidence) (Naumann et al., 2013; Hennige et al., 2014; Lunden et al., 2014; Naumann et al., 2014; 5 Georgian et al., 2016; Gori et al., 2016; Maier et al., 2016; Büscher et al., 2017). 6 7 The extension and intensification of deep-water acidification (Section 3.2.3.1) has been identified as a 8 further key risk to deep-water coral ecosystems (medium confidence) (Bindoff et al., 2019). Literature since 9 SROCC supports this assessment (Morato et al., 2020; Puerta et al., 2020), although scleractinians and 10 gorgonians are found in regions undersaturated with respect to aragonite (Thresher et al., 2011; Fillinger and 11 Richter, 2013; Baco et al., 2017). Laboratory experiments on reef-forming scleractinians show variable 12 results, with regional acclimation potential and population-genetic adaptations (Georgian et al., 2016; 13 Kurman et al., 2017). Desmophyllum pertusum6 and M. oculata maintain calcification in moderately low pH 14 (7.75) and near-saturation of aragonite (Hennige et al., 2014; Maier et al., 2016; Büscher et al., 2017), but 15 lower pH (7.6) and corrosive conditions lead to net dissolution of D. pertusum skeletons (high confidence) 16 (Lunden et al., 2014; Kurman et al., 2017; Gómez et al., 2018). Experiments suggest that D. dianthus is more 17 sensitive to warming than acidification and when both are high, as projected under climate change. Warming 18 appears to compensate for declines in calcification, with fitness also sensitive to food availability (Bramanti 19 et al., 2013; Movilla et al., 2014; Gori et al., 2016; Baussant et al., 2017; Büscher et al., 2017; Schönberg et 20 al., 2017; Höfer et al., 2018; Maier et al., 2019). 21 22 In OMZ regions (Section 3.2.3.2), benthic species distributions (Sperling et al., 2016; Levin, 2018; Gallo et 23 al., 2020), abundance and composition of demersal fishes in canyons (De Leo et al., 2012) and deep-pelagic 24 zooplankton (Wishner et al., 2018) follow oxygen gradients, indicating that deep-sea biodiversity and 25 ecosystem structure will be impacted by extension of hypoxic areas (medium confidence). Fossil records 26 show benthic population collapse and turnover when oxygen ranged from oxic to mildly or severely hypoxic 27 (Cross-Chapter Box PALEO in Chapter 1, Moffitt et al., 2015). Regional extirpations among cold-water 28 corals in the paleorecord were associated with substantial declines in oxygen, coincident with abrupt 29 warming and altered intermediate water masses properties (Wienberg et al., 2018; Hebbeln et al., 2019). 30 Despite mortality and functional impacts from low oxygen concentrations observed in aquaria (Lunden et al., 31 2014), recent observations of the deep-water coral D. pertusum suggest adaptive capacity to hypoxia among 32 specimens from OMZ regions that are highly productive (low confidence) (Hanz et al., 2019; Hebbeln et al., 33 2020). 34 35 Chemosynthetic ecosystems could be particularly prone to oxygen decline (low to medium confidence). 36 Projected OMZ expansion in the vicinity of seep communities could favour sulphide-tolerant species, as 37 suggested from fossil records (Moffitt et al., 2015), but this will exclude large symbiont-bearing foundation 38 species of methane seep ecosystems (Fischer et al., 2012; Sweetman et al., 2017). Projected warming, or 39 shifts in warm-current circulation along continental margins, could enhance dissociation of buried methane 40 hydrates (Phrampus and Hornbach, 2012; Phrampus et al., 2014), either increasing anaerobic methane 41 oxidation (Boetius and Wenzhöfer, 2013), which benefit seep communities, or increasing gas fluxes, which 42 would decrease anaerobic methane oxidation rates and exclude chemosynthetic fauna. 43 44 Environmental niche models (FAO, 2019; Morato et al., 2020; Puerta et al., 2020) project that under 45 RCP8.5, >50% of present-day scleractinian habitats in the North Atlantic Ocean, will become unsuitable by 46 2100, with greater impacts on D. pertusum than on D. dianthus or Madrepora oculata. For gorgonians, 47 corresponding habitat loss is likely >80%. Much less is known about the environmental niches of deep-sea 48 sponges, preventing a similar assessment (Kazanidis et al., 2019; Puerta et al., 2020). 49 50 Climate-driven impacts further limit the resilience of deep-sea ecosystems to impacts from human activities 51 (high confidence) (Levin and Le Bris, 2015; Rogers, 2015; Sweetman et al., 2017). However, assessing 52 cumulative climatic and non-climatic impacts is challenging for these data-poor environments (Ashford et 53 al., 2018; Levin, 2018; Armstrong et al., 2019; Heffernan, 2019; Kazanidis et al., 2020; Orejas et al., 2020), 54 where lack of knowledge increases the possibility of overlooking ecosystem vulnerabilities and risks (Levin, 55 2021). A paucity of information about the natural variability and historical trends of these habitats prevents 6 Previously named Lophelia pertusa Do Not Cite, Quote or Distribute 3-97 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 robust assessment of adaptive capacities and potential vulnerabilities to extreme events (Aguzzi et al., 2019; 2 Levin et al., 2019; Chapron et al., 2020; Danovaro et al., 2020; Le Bris and Levin, 2020; Levin, 2021). The 3 spatial resolution of CMIP5 models is too coarse to robustly project changes in mesoscale circulation at the 4 seafloor (Sulpis et al., 2019), on which deep-sea ecosystems depend for organic material supplies and 5 dispersal of planktonic and planktotrophic larvae (high confidence) (Fox et al., 2016; Mitarai et al., 2016; 6 Dunn et al., 2018). Higher-resolution modelling from CMIP6 (Orr et al., 2017), multiannual and high- 7 frequency records of ocean bottom-water properties (Meinen et al., 2020), and better understanding and 8 accounting of biogeochemical mechanisms of organic carbon transport to the ocean interior is expected to 9 improve this capacity (Boyd et al., 2019; Séférian et al., 2020). 10 11 12 13 Figure Box3.3.1: Schematic of the combination of climate-impact drivers in different deep-ocean ecosystems. Key 14 physical and biological drivers of change in the deep-sea and benthic habitats with specific vulnerabilities are discussed 15 in Section 3.4.3.3. 16 17 18 [END BOX 3.3 HERE] 19 20 21 3.4.4 Reversibility and Impacts of Temporary Overshoot of 1.5°C or 2°C Warming 22 23 Scenarios limiting warming to the 1.5°C and 2°C limits in the Paris Agreement can involve temporarily 24 exceeding those warming levels before declining again (WGI AR6 Section 4.6.2.1, Lee et al., 2021). The 25 effect of such "overshoot" on marine and coastal ecosystems depends on the reversibility of both the 26 response of climate-impact drivers, and the response of organisms and ecosystems to the climate-impact 27 drivers, during the overshoot period. WGI AR6 assessed that temporary overshoot of a 2°C warming 28 threshold has irreversible effects on global mean sea-level and also effects on ocean heat content that persist 29 beyond 2100 (WGI AR6 Section 4.6.2.1, Lee et al., 2021). Model results indicate that sea surface 30 temperatures (high confidence), Arctic sea ice (high confidence), surface ocean acidification (very high 31 confidence) and surface ocean deoxygenation (very high confidence) are reversible within years to decades if 32 net emissions reach zero or below (WGI AR6 Table 4.10, Lee et al., 2021). Although changes in these Do Not Cite, Quote or Distribute 3-98 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 surface ocean variables are reversible, habitat-forming ecosystems including coral reefs and kelp forests may 2 undergo irreversible phase shifts with >1.5°C warming (Section 3.4.2.1, 3.4.2.3), and are thus at high risk 3 this century in 1.5°C or 2°C scenarios involving overshoot (Tachiiri et al., 2019). In an overshoot scenario in 4 which CO2 returns to 2040 levels by 2100 (SSP5-3.4-OS, O'Neill et al., 2016), SST and Arctic sea ice do not 5 fully return by 2100 to levels prior to the CO2 peak (medium confidence) (WGI AR6 Section 4.6.2.1, Lee et 6 al., 2021), suggesting that reversal of marine ecological impacts from 21st century climate impacts would 7 extend into the 22nd century or beyond (McManus et al., 2021). Models also indicate that global sea level 8 rise, as well as warming, ocean acidification and deoxygenation at depth, are irreversible for centuries or 9 longer (very high confidence) (WGI AR6 Section 4.6.2.1 and Table 4.10, Palter et al., 2018; Li et al., 2020c; 10 Lee et al., 2021). 11 12 13 3.5 Vulnerability, Resilience and Adaptive Capacity in Marine Social-Ecological Systems, including 14 Impacts to Ecosystem Services 15 16 3.5.1 Introduction 17 18 This Section assesses the impacts of climate change on ecosystem services (Table 3.25, Chapter 1) and the 19 outcomes on social-ecological systems, building on previous assessments (Table 3.26). Section 3.5.2 20 assesses how changes in biodiversity influence ecosystem services. Then Sections 3.5.3 and 3.5.4 assess 21 provisioning services (food and non-food), Section 3.5.5 assesses supporting and regulating services, and 22 Section 3.5.6, cultural services. Where evidence exists, the section evaluates how the vulnerability and 23 adaptive capacity of social-ecological systems govern the manifestation of impacts on each ecosystem 24 service. 25 26 27 Table 3.25: Ocean and coastal ecosystem services. Adapted from IPBES (2017), with examples made specific to ocean 28 and coastal ecosystems by Chapter 3 authors. Ecosystem Service Category Components Ocean and Coastal Examples Provisioning Food and feed Status of harvested marine fish, invertebrates, mammals, and plants. Medicinal, biochemical and genetic Existence of and access to biological resources resources that could offer future prospects for development, including marine fish, invertebrates, mammals, plants, microbes, viruses. Materials and assistance Existence of and access to minerals, shells, stones, coral branches, dyes used to create other goods; availability of marine organisms to exhibit in zoos, aquariums, and as pets. Energy Existence of and access to sources of energy, including oil and gas reserves; solar, tidal, and thermal ocean energy; and biofuels from marine plants. Supporting and Regulating Habitat creation and maintenance Status of nesting, feeding, nursery, and mating sites for birds, mammals, and other marine life, and of resting and overwintering places for migratory marine life or insects. Connectivity of ocean habitats. Dispersal and other propagules Ability of marine life to spread gametes and larvae successfully by broadcast spawning reproduction, and ability of adults to disperse widely. Regulation of climate Status of carbon storage and sequestration, methane cycling in Do Not Cite, Quote or Distribute 3-99 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report Cultural wetlands, and dimethyl sulfide creation and destruction. 1 2 Regulation of air quality Status of aquatic processes that 3 maintain and balance CO2, oxygen, nitrogen oxides, sulfur oxides, volatile organic compounds, particulates, and aerosols. Regulation of ocean acidification Status of chemical and biological (Section 3.2.3.1) aquatic processes that maintain and balance CO2 and other acids/bases. Regulation of freshwater quantity, Status of water storage by coastal location and timing systems, including groundwater flow; aquifer recharge; and flooding responses of wetlands, coastal water bodies, and developed spaces. Regulation of freshwater and coastal Status of chemical and biological water quality aquatic processes that retain and filter coastal waters, capture pollutants and particles, and oxygenate water (e.g., natural filtration by sediments including adsorbent minerals and microbes). Regulation of organisms detrimental to Status of grazing that controls harmful humans and marine life algal blooms and algal overgrowth of key ecosystems. Environmental conditions that suppress marine pathogens. Formation, protection and Status of chemical and biological decontamination of soils and aquatic processes that capture sediments pollutants and particles (e.g., adsorption by minerals, microbial breakdown of pollutants). Regulation of hazards and extreme Ability of coastal environments to events serve as wave energy dissipators, barriers, and wave breaks. Regulation of key elements Status of aquatic processes that maintain and balance stocks of carbon, nitrogen, phosphorus, and other elements critical for life. Physical and psychological Existence of and access to recreational experiences opportunities including visiting beaches and coastal environments; and aquatic activities such as fishing, boating, swimming, and diving. Supporting identities Existence of and access to cultural, heritage, and religious activities, and opportunities for intergenerational knowledge transfer. Sense of place. Learning and inspiration Existence of educational opportunities and characteristics to be emulated, as in biomimicry. Maintenance of options Existence of opportunities to develop new medicines, materials, foods, and resources, or to adapt to a warmer climate and emergent diseases. Do Not Cite, Quote or Distribute 3-100 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Table 3.26: Conclusions from previous IPCC assessments about observed and projected climate impacts to ocean and 2 coastal biodiversity and ecosystem services. Ecosystem service and chapter Observed Impacts Projected Impacts subsection All (Section 3.5) Climate change has affected marine "Long-term loss and degradation of "ecosystem services with regionally marine ecosystems compromises the diverse outcomes, challenging their ocean's role in cultural, recreational, governance (high confidence). Both and intrinsic values important for positive and negative impacts result for human identity and well-being food security through fisheries (medium confidence) (3.2.4, 3.4.3, (medium confidence), local cultures 5.4.1, 5.4.2, 6.4)" (SROCC SPM B.8, and livelihoods (medium confidence), IPCC, 2019c). and tourism and recreation (medium confidence). The impacts on ecosystem services have negative consequences for health and well-being (medium confidence), and for Indigenous Peoples and local communities dependent on fisheries (high confidence) (1.1, 1.5, 3.2.1, 5.4.1, 5.4.2, Figure SPM.2)" (SROCC SPM A.8, IPCC, 2019c). Biodiversity (Section 3.5.2) "[Climate] Impacts are already "Risks of severe impacts on observed on [coastal ecosystem] biodiversity, structure and function of habitat area and biodiversity, as well as coastal ecosystems are projected to be ecosystem functioning and services higher for elevated temperatures under (high confidence) (4.3.2, 4.3.3, 5.3, high compared to low emissions 5.4.1, 6.4.2, Figure SPM.2)" (SROCC scenarios in the 21st century and SPM A.6, IPCC, 2019c). beyond." (SROCC SPM B.6, IPCC, 2019c). Food provision (Section 3.5.3) "Warming-induced changes in the "Future shifts in fish distribution and spatial distribution and abundance of decreases in their abundance and some fish and shellfish stocks have fisheries catch potential due to climate had positive and negative impacts on change are projected to affect income, catches, economic benefits, livelihoods, and food security of livelihoods, and local culture (high marine resource-dependent confidence). There are negative communities (medium confidence). consequences for Indigenous Peoples Long-term loss and degradation of and local communities that are marine ecosystems compromises the dependent on fisheries (high ocean's role in cultural, recreational, confidence). Shifts in species and intrinsic values important for distributions and abundance has human identity and well-being challenged international and national (medium confidence) (3.2.4, 3.4.3, ocean and fisheries governance, 5.4.1, 5.4.2, 6.4)" (SROCC SPM including in the Arctic, North Atlantic B.8,IPCC, 2019c). and Pacific, in terms of regulating fishing to secure ecosystem integrity and sharing of resources between fishing entities (high confidence) (3.2.4, 3.5.3, 5.4.2, 5.5.2, Figure SPM.2)". (SROCC SPM A.8.1 IPCC, 2019c). Non-food consumable provisioning Observed impacts on non-food "Reductions in marine biodiversity due services (Section 3.5.4.1) provisioning services not previously to climate change and other assessed. anthropogenic stressors (Tittensor et al., 2010), such as ocean acidification (CBD, 2009) and pollution, might reduce the discovery of genetic resources from marine species useful in pharmaceutical, aquaculture, agriculture, and other industries (Arrieta et al., 2010), leading to a loss Do Not Cite, Quote or Distribute 3-101 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report Renewable energy (Section 3.5.4.2) of option value from marine Habitat creation and maintenance ecosystems." (WGII AR5 Section (Section 3.5.5.1) 6.4.1.2, Pörtner et al., 2014) Climate regulation and air quality Observed impacts on ocean renewable "Ocean renewable energy can support (Section 3.5.5.2) energy not previously assessed. climate change mitigation, and can comprise energy extraction from offshore winds, tides, waves, thermal and salinity gradient and algal biofuels. The emerging demand for alternative energy sources is expected to generate economic opportunities for the ocean renewable energy sector (high confidence), although their potential may also be affected by climate change (low confidence) (5.4.2, 5.5.1, Figure 5.23)". (SROCC SPM C.2.5, IPCC, 2019c). "[Climate] Impacts are already "In the Southern Ocean, the habitat of observed on [coastal ecosystem] Antarctic krill, a key prey species for habitat area and biodiversity, as well as penguins, seals and whales, is ecosystem functioning and services projected to contract southwards under (high confidence) (4.3.2, 4.3.3, 5.3, both RCP2.6 and RCP8.5 (medium 5.4.1, 6.4.2, Figure SPM.2)" (SROCC confidence) (3.2.2, 3.2.3, 5.2.3)" SPM A.6, IPCC, 2019c). (SROCC SPM B5.3, IPCC, 2019c). "Ocean warming, oxygen loss, "In polar regions, ice associated acidification and a decrease in flux of marine mammals and seabirds have organic carbon from the surface to the experienced habitat contraction linked deep ocean are projected to harm to sea ice changes (high confidence)." habitat-forming cold-water corals, (SROCC SPM A.5.2, IPCC, 2019c). which support high biodiversity, partly through decreased calcification, increased dissolution of skeletons, and bioerosion (medium confidence)." (SROCC SPM B5.4, IPCC, 2019c). "Global ocean heat content continued "The increase in global ocean heat to increase throughout [the 1951- content (TS2.4) will likely continue present] period, indicating continuous until at least 2300 even for low- warming of the entire climate system emission scenarios." (WGI AR6 Box (very high confidence)" (WGI AR6 TS.9, Arias et al., 2021). TS1.2.3, Arias et al., 2021). "Land and ocean have taken up a near- "While natural land and ocean carbon constant proportion (globally about sinks are projected to take up, in 56% year­1) of CO2 emissions from absolute terms, a progressively larger human activities over the past six amount of CO2 under higher compared decades, with regional differences to lower CO2 emissions scenarios, they (high confidence)" (WGI AR6 SPM become less effective, that is, the A1.1, IPCC, 2021b). proportion of emissions taken up by land and ocean decrease with increasing cumulative CO2 emissions. This is projected to result in a higher proportion of emitted CO2 remaining in the atmosphere (high confidence)" (WGI AR6 SPM B4.1, IPCC, 2021b). Observed impacts on marine "The effect of climate change on organisms' contribution to climate marine biota will alter their regulation not previously assessed. contribution to climate regulation, that is, the maintenance of the chemical composition and physical processes in the atmosphere and oceans (high confidence) (Beaumont et al., 2007)" Do Not Cite, Quote or Distribute 3-102 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report (WGII AR5 Section 6.4.1.3, Pörtner et al., 2014). Provision of fresh water, maintenance Observed climate impacts on "In the absence of more ambitious of water quality, regulation of salinisation of coastal soil and adaptation efforts compared to today, pathogens (Section 3.5.5.3) groundwater not previously assessed. and under current trends of increasing exposure and vulnerability of coastal communities, risks, such as erosion and land loss, flooding, salinisation, and cascading impacts due to mean sea level rise and extreme events are projected to significantly increase throughout this century under all greenhouse gas emissions scenarios (very high confidence)." (SROCC SPM B9.1, IPCC, 2019c) "Global warming compromises "[Risks from marine-borne pollutants seafood safety (medium confidence) and pathogens] are projected to be through human exposure to elevated particularly large for human bioaccumulation of persistent organic communities with high consumption of pollutants and mercury in marine seafood, including coastal Indigenous plants and animals (medium communities (medium confidence), confidence), increasing prevalence of and for economic sectors such as waterborne Vibrio sp. pathogens fisheries, aquaculture, and tourism (medium confidence), and heightened (high confidence) (3.4.3, 5.4.2, Box likelihood of harmful algal blooms 5.3)" (SROCC SPM B.8.3, IPCC, (medium confidence)." (SROCC SPM 2019c). B.8.3, IPCC, 2019c). "Since the early 1980s, the occurrence "Overall, the occurrence of HABs, of harmful algal blooms (HABs) and their toxicity and risk on natural and pathogenic organisms (e.g., Vibrio) has human systems are projected to increased in coastal areas in response continue to increase with warming and to warming, deoxygenation and rising CO2 in the 21st century (Glibert eutrophication, with negative impacts et al., 2014; Martín-García et al., 2014; on food provisioning, tourism, the McCabe et al., 2016; Paerl et al., 2016; economy and human health (high Gobler et al., 2017; McKibben et al., confidence)." (SROCC Chapter 5 2017; Rodríguez et al., 2017; Paerl et Executive Summary, Bindoff et al., al., 2018; Riebesell et al., 2018) (high 2019). confidence)." (SROCC Box 5.4, Bindoff et al., 2019). Regulation of physical hazards "Coastal ecosystems are already "The decline in warm water coral reefs (Section 3.5.5.4) impacted by the combination of sea is projected to greatly compromise the level rise, other climate-related ocean services they provide to society, such changes, and adverse effects from as...coastal protection (high human activities on ocean and land confidence)..." (SROCC SPM B.8.2, (high confidence)... Coastal and near- IPCC, 2019c). shore ecosystems including saltmarshes, mangroves, and vegetated dunes in sandy beaches, ... provide important services including coastal protection... (high confidence)" (SROCC Chapter 4 Executive Summary, Oppenheimer et al., 2019). Ocean and coastal carbon storage "Recent observations show that ocean "Emission scenarios SSP4-6.0 and (Section 3.5.5.5) carbon processes are starting to change SSP5-8.5 lead to warming of the in response to the growing ocean sink, surface ocean and large reductions of and these changes are expected to the buffering capacity, which will slow contribute significantly to future the growth of the ocean sink after weakening of the ocean sink under 2050. Scenario SSP1-2.6 limits further medium- to high-emission scenarios. reductions in buffering capacity and However, the effects of these changes warming, and the ocean sink weakens is not yet reflected in a weakening in response to the declining rate of Do Not Cite, Quote or Distribute 3-103 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report trend of the contemporary (1960­ increasing atmospheric CO2. There is 2019) ocean sink (high confidence)" low confidence in how changes in the (WGI AR6 Chapter 5 Executive biological pump will influence the Summary, Canadell et al., 2021). magnitude and direction of the ocean carbon feedback" (WGI AR6 Chapter 5 Executive Summary, Canadell et al., 2021). "Mangrove, seagrass, and salt marsh "...under high emission scenarios, sea ecosystems offer important carbon level rise and warming are expected to storage and sequestration opportunities reduce carbon sequestration by (limited evidence, medium agreement), vegetated coastal ecosystems (medium in addition to ecosystem goods and confidence); however, under services such as protection against conditions of slow sea level rise, there coastal erosion and storm damage and may be net increase in carbon uptake maintenance of habitats for fisheries by some coastal wetlands (medium species." (WGII AR5 Technical confidence)" (SROCC Chapter 5, Summary). Bindoff et al., 2019). Cultural Services (Section 3.5.6) "Climate change impacts on marine "Future shifts in fish distribution and ecosystems and their services put key decreases in their abundance and cultural dimensions of lives and fisheries catch potential due to climate livelihoods at risk (medium change are projected to affect income, confidence), including through shifts in livelihoods, and food security of the distribution or abundance of marine resource-dependent harvested species and diminished communities (medium confidence). access to fishing or hunting areas. This Long-term loss and degradation of includes potentially rapid and marine ecosystems compromises the irreversible loss of culture and local ocean's role in cultural, recreational, knowledge and Indigenous knowledge, and intrinsic values important for and negative impacts on traditional human identity and well-being diets and food security, aesthetic (medium confidence)" (SROCC SPM aspects, and marine recreational B.8, IPCC, 2019c). activities (medium confidence)" (SROCC SPM B.8.4, IPCC, 2019c). 1 2 3 3.5.2 Biodiversity 4 5 Climate change is a key agent of biodiversity change in numerous ocean and coastal ecosystems (very high 6 confidence) (Table 3.26, Worm and Lotze, 2021), and climate change and biodiversity loss reinforce each 7 other (Pörtner et al., 2021b). Biodiversity has changed in association with ocean warming and loss of sea ice 8 (Sections 3.4.2.10, 3.4.3.3.3; Section CCP6 2.4.2), SLR (Section 3.4.2; Cross-Chapter Box SLR in Chapter 9 3), coral bleaching (Section 3.4.2.1), MHWs (Sections 3.4.2.1-3.4.2.5), and upwelling changes (high 10 confidence) (Section 3.4.2.9). Overlapping non-climate drivers (Section 3.1) also decrease ocean and coastal 11 ecosystem biodiversity (very high confidence) (O'Hara et al., 2021; Pörtner et al., 2021b). There is medium 12 confidence that local and regional marine biodiversity losses from climate disrupt ecosystem services 13 provided by specific ocean and coastal species or places (Sections 3.5.3­3.5.6, Figure 3.23, Table 3.26, Box 14 3.3, Dee et al., 2019a; Hossain, 2019; Smale et al., 2019; Teixeira et al., 2019; Martin et al., 2020; Pathak, 15 2020; Weiskopf et al., 2020; Zunino et al., 2020; Archer et al., 2021). However, adaptive capacity varies 16 greatly among ecosystems, and ecological functions sometimes remain, despite changes in species 17 assemblages, as in certain coral reef communities (Richardson et al., 2020). Projected changes in biodiversity 18 due to climate change (Section 3.4.3.3.3) are expected to alter the flow and array of ocean and coastal 19 ecosystem services (high confidence) (Smale et al., 2019; Cavanagh et al., 2021; Ruthrof et al., 2021; Worm 20 and Lotze, 2021), but data gaps hinder developing projections of ecosystem service changes detailed enough 21 to support decision making (Rosa et al., 2020). 22 23 Non-indigenous marine species are major agents of ocean and coastal biodiversity change, and climate and 24 non-climate drivers interact to support their movement and success (high confidence) (Iacarella et al., 2020). 25 At times, non-indigenous species act invasively and outcompete indigenous species, causing regional 26 biodiversity shifts and altering ecosystem function, as seen in the Mediterranean region (high confidence) Do Not Cite, Quote or Distribute 3-104 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 (e.g., Mannino et al., 2017; Bianchi et al., 2019; Hall-Spencer and Harvey, 2019; Verdura et al., 2019; 2 García-Gómez et al., 2020; Dimitriadis et al., 2021). Warming-related range expansions of non-indigenous 3 species have directly or indirectly decreased commercially important fishery species and nursery habitat 4 (Booth et al., 2018). Non-indigenous species outperform indigenous species in coastal zones experiencing 5 warming and freshening (McKnight et al., 2021). Non-climate drivers, especially marine shipping in newly 6 ice-free locations (Chan et al., 2019), fishing pressure (Last et al., 2011), aquaculture of non-indigenous 7 species (Mach et al., 2017; Ruby and Ahilan, 2018), and marine pollution and debris (Gall and Thompson, 8 2015; Carlton et al., 2018; Carlton and Fowler, 2018; Lasut et al., 2018; Miralles et al., 2018; Rech et al., 9 2018; Therriault et al., 2018), promote range shifts and movement of non-indigenous species (high 10 confidence). Non-climate drivers can also intensify the ecological effects of non-indigenous species (Geraldi 11 et al., 2020). Invasive marine species can alter species behaviour, reduce indigenous species abundance, 12 reduce water clarity, bioaccumulate more heavy metals than indigenous species, and inhibit ecosystem 13 resilience in the face of extreme events (medium confidence) (McDowell et al., 2017; Geburzi and 14 McCarthy, 2018; Anton et al., 2019; Ruthrof et al., 2021). Risks from invasive species to the sources of other 15 ecosystem services or aquatic goods, including valuable materials, mining activities, shipping, or ocean 16 energy installations, have not been evaluated. 17 18 Reducing risk to ecosystem functions and services that depend on biodiversity requires an integrated 19 approach that acknowledges the close linkages between the climate and biodiversity crises and common 20 governance challenges (Pörtner et al., 2021b). Climate-focused solutions that employ nature-based solutions 21 (NbS), technological interventions, and socio-institutional interventions (Section 3.6.2) can also safeguard 22 biodiversity (Pörtner et al., 2021b), which in turn will help ocean and coastal ecosystems adapt to climate 23 impacts as well as help sustain the services they provide to people (Sections 3.5.3­3.5.6). 24 25 3.5.3 Food Provision 26 27 Globally, about 17% of humans' average per capita intake of animal protein in 2017 came from marine and 28 freshwater wild-caught and aquacultured aquatic animals (Costello et al., 2020; FAO, 2020a). Per capita 29 intake of seafood is 50% or more in some Small Island Developing States (SIDS) (Vannuccini et al., 2018), 30 and consumption per capita is 15 times higher in Indigenous Peoples than non-Indigenous Peoples (Cisneros- 31 Montemayor et al., 2016). Fishery products also supply critical dietary micronutrients worldwide (Section 32 3.5.4.1, Hicks et al., 2019; Vianna et al., 2020). Marine and freshwater fisheries and aquaculture provide 33 livelihoods for an estimated 10­12% of the world's population (Barange et al., 2018). Fishing and 34 aquaculture provide women and their families with substantial amounts of food and income (Harper et al., 35 2020b), because at least 11% of small-scale fishers (Harper et al., 2020b) and up to half of all fishery and 36 aquaculture workers (FAO, 2018) are women. This section assesses how climate-driven alterations of the 37 abundance or nutritional quality of food from the sea could affect humans. Aquaculture, catch potential 38 changes, and human adaptations to changes in wild and cultured harvests, are assessed in Section 5.9. 39 40 Ocean and coastal fauna are moving towards higher latitudes globally due to warming (high confidence) 41 (Section 3.4.3.1, Table 3.26), challenging fishers and fisheries management (high confidence) as fishers also 42 move poleward and diversify harvests (medium evidence, high agreement) (Table 3.26, Sections 3.4.3.3.3, 43 5.8.4, Leitão et al., 2018; Liang et al., 2018; Ottosen et al., 2018; Peck and Pinnegar, 2018; Pinsky et al., 44 2018; Erauskin-Extramiana et al., 2019; Free et al., 2019; Gianelli et al., 2019; Scott et al., 2019; Smith et 45 al., 2019; Gervais et al., 2021). Model hindcasts have identified temperature-associated fisheries reductions 46 worldwide (Free et al., 2019), and they have implicated overfishing as the primary non-climate driver 47 increasing fishery vulnerability (Section 5.8.4, Peck and Pinnegar, 2018; Das et al., 2020). Catch 48 composition is changing in many locations fished by smaller-scale, less mobile commercial, artisanal, and 49 recreational fisheries (high confidence) (Booth et al., 2018; Townhill et al., 2019; Young et al., 2019b; 50 Robinson et al., 2020; Champion et al., 2021). Limited exceptions have been noted, with wild harvests in 51 some places remaining stable or increasing (e.g., Arreguín-Sánchez, 2019; Robinson et al., 2019b; Kainge et 52 al., 2020). Where possible, fishers are maintaining harvests by broadening catch diversity, traveling 53 poleward, and changing gear and strategies (high confidence) (Section 3.6.3.1.2, Barange et al., 2018; Dubik 54 et al., 2019; Townhill et al., 2019). Fisheries and aquaculture adaptations, including management, are 55 comprehensively assessed in Sections 3.6.3.1.2, 5.8.4, and 5.9.4. 56 Do Not Cite, Quote or Distribute 3-105 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Ocean acidification and deoxygenation caused by climate change are thought to influence fishing and 2 aquaculture harvests, but limited evidence prevents assessing their present global impact on harvests. 3 Substantial economic losses in the North American Pacific Coast shellfish aquaculture industry in the 2000s 4 assessed in SROCC (Bindoff et al., 2019) and WGII AR5 (Pörtner et al., 2014) remain the clearest example 5 of human harm from ocean acidification. Technology-based adaptations (Section 3.6.3) have minimised 6 aquaculture losses from ocean acidification, including early warning systems to guide hatchery operations 7 and culturing resilient shellfish strains (Section 5.9.4, Barton et al., 2015a). Laboratory studies show that 8 ocean acidification decreases the fitness, growth, or survival of many economically and culturally important 9 larval or juvenile shelled mollusks (high confidence) (Cao et al., 2018; Onitsuka et al., 2018; Stevens and 10 Gobler, 2018; Griffith et al., 2019a; Mellado et al., 2019), and of several valuable wild-harvest crab species 11 (Barton et al., 2015a; Punt et al., 2015; Miller et al., 2016; Swiney et al., 2017; Gravinese et al., 2018; 12 Tomasetti et al., 2018; Long et al., 2019; Trigg et al., 2019). Ocean acidification alters larval settlement and 13 metamorphosis of fish in laboratory studies (high confidence) (Cattano et al., 2018; Espinel-Velasco et al., 14 2018), suggesting possible changes in fish survival and thus fishery characteristics. Deoxygenation can 15 decrease size and abundance of marine species and suppress trophic interactions (Levin, 2003), decrease the 16 diversity within marine ecosystems (Sperling et al., 2016) while temporarily increasing catchability and 17 increasing the risk of overfishing (Breitburg et al., 2018), and decrease the ecosystem services provided by 18 specific fisheries (Orio et al., 2021). The chronic effects of deoxygenation on wild fisheries are complex and 19 highly interactive with co-occurring drivers and overall ecosystem responses (medium evidence, high 20 agreement) (Townhill et al., 2017; Rose et al., 2019). Detecting and attributing marine ecosystem responses 21 to ocean acidification and deoxygenation outside of laboratory studies remains challenging because of the 22 strong influence of co-occurring environmental changes on natural systems (Section 3.3.5, Rose et al., 2019; 23 Doo et al., 2020). 24 25 Ocean and coastal organisms will continue moving poleward under RCP8.5 (high confidence) (Section 26 3.4.3.1.3, Figure 3.18), and this is expected to decrease fisheries harvests in low latitudes and alter species 27 composition and abundance in higher latitudes (high confidence) (Table 3.26, Figure 3.23, Asch et al., 2018; 28 Morley et al., 2018; Tai et al., 2019; Erauskin-Extramiana et al., 2020; Shelton et al., 2021). Species that 29 succeed in new ranges or conditions may offer opportunities to diversify regional fisheries or aquaculture 30 (Sections 3.6.3.1.2, 5.8.4, 5.9.4, Bindoff et al., 2019), or they may outcompete indigenous species and act as 31 invasive species (Sections 3.4.2.10, 3.5.2). 32 33 Temperature will continue to be a major driver of fisheries changes globally, but other non-climate factors 34 like organism physiology and ecosystem response (Section 3.3) and fishing pressure (Chapter 5), as well as 35 other climate-impact drivers like acidification, deoxygenation, and sea-ice loss (Section 3.2), will play 36 critical roles in future global and local fisheries changes (high confidence). Warming, acidification, and 37 business-as-usual fishing policy under RCP8.5 are projected to place around 60% of global fisheries at very 38 high risk (medium confidence) (Cheung et al., 2018). Model intercomparison showed that ocean acidification 39 and protection affect ecosystems more than fishing pressure, and ecological adaptation greatly determines 40 impacts on fishery biomass, catch and value until approximately 2050 (medium confidence) (Olsen et al., 41 2018). Ecosystem responses to warming water, fishing pressure, food-web changes, MHWs, and sea ice 42 algal populations have been responsible for highly variable or collapsing populations of Northern 43 Hemisphere high-latitude forage fish species including sand lances (Ammodytes spp.), Arctic cod 44 (Boreogadus saida), capelin (Mallotus catervarius), and herring (Clupea spp.) (Lindegren et al., 2018; 45 Steiner et al., 2019; Arimitsu et al., 2021; Suca et al., 2021). Declining stocks of forage fish are expected to 46 have detrimental effects on seabirds, pelagic fish, and marine mammals (medium confidence) (Lindegren et 47 al., 2018; Steiner et al., 2019), which may harm dependent human communities, including Arctic Indigenous 48 Peoples (low confidence) (Arctic Monitoring and Assessment Programme, 2018; Steiner et al., 2019). 49 Modelled fishery futures and revenue depend on environmental scenario, fishing fleet composition and 50 management, and ocean acidification and temperature responses of harvested species (high confidence) (Punt 51 et al., 2014; Punt et al., 2015; Seung et al., 2015; Fernandes et al., 2017; Rheuban et al., 2018; Tai et al., 52 2019; Punt et al., 2020). Detrimental effects of ocean acidification are projected to begin emerging in 53 specific fisheries by 2030 (limited evidence, high agreement) [(southern Tanner crab (Chionoecetes bairdi) 54 (Punt et al., 2015); sea scallop (Placopecten magellanicus) (Rheuban et al., 2018); Northeast Arctic cod 55 (Gadus morhua) (Hänsel et al., 2020); Arctic fisheries (Lam et al., 2016)]. At the same time, projected 56 hypoxic conditions of ~2 mg l­1 of oxygen will be consistently detrimental across taxonomic groups, 57 developmental stages, and climate regions (high confidence) (Sampaio et al., 2021). Ecosystem-based Do Not Cite, Quote or Distribute 3-106 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 management (Section 3.6.3.1.2) shows promise for decreasing risk from interacting climate and non-climate 2 drivers to forage species and fished species. 3 4 3.5.4 Other Provisioning Services 5 6 3.5.4.1 Non-Food Consumable Products 7 8 The interaction of climate and non-climate drivers endangers the supply of non-food consumable products 9 developed from marine organisms (limited evidence, high agreement). This broad class includes 10 nutraceuticals (derived from fish, krill, shellfish, seaweeds, microbes), food preservatives or additives 11 (derived from crustaceans, fish, microalgae and seaweeds, cyanobacteria), pharmaceuticals (derived from 12 fish, shellfish, microbes, cyanobacteria, corals, sponges), or cosmetic products (derived from sponges, 13 phytoplankton and seaweeds, fish, etc.) (Freitas et al., 2012; Dewapriya and Kim, 2014; Leal and Calado, 14 2015; Stengel and Connan, 2015; Greene et al., 2016; Ciavatta et al., 2017; Gutiérrez-Rodríguez et al., 15 2018). But biodiversity changes, warming, acidification, and non-climate drivers (especially fishing 16 pressure) may decrease the availability of these organisms or the potency of the compounds they produce 17 (Section 5.7.5.1, Table 3.26, Figure 3.23, Webster and Taylor, 2012; Mehbub et al., 2014; Kotta et al., 2018; 18 Martins et al., 2018; Conrad et al., 2021). Observed and projected declines and movement of fish stocks due 19 to fishing pressure and climate change impacts (IPCC, 2019b) have generated concerns that the supply and 20 safety of fish and krill oil for human dietary supplements may decline (Section 5.7.5.1, Gribble et al., 2016; 21 Lloret et al., 2016). This risk can be lowered by technological adaptations (Section 3.6.2.2) such as 22 increasing the use of alternative sources, like marine phytoplankton, macroalgae, marine microbes 23 (Dewapriya and Kim, 2014; Greene et al., 2016; Dave and Routray, 2018; Nguyen et al., 2020) and 24 underutilised resources such as fish, seal, crab and shrimp byproducts (Dave and Routray, 2018), and by 25 improving extraction and processing efficiency(Cashion et al., 2017). Climate effects on non-food 26 consumable products could be widespread yet poorly detected, complicating assessment of impacts, risks, 27 and vulnerability reduction. 28 29 There is insufficient evidence to develop global projections of future climate impacts on humans through 30 changes in non-food consumable marine products, but specific local examples have been investigated, such 31 as the Arctic ooligan (eulachon) (Thaleichthys pacificus), a small smelt fish. Ooligan grease has been used 32 by Indigenous Peoples of the North Pacific coast (Phinney et al., 2009) for at least 5000 years to treat 33 stomach aches, colds and skin conditions and as a traditional food source high in omega-3 fatty acids (Byram 34 and Lewis, 2001; Cranmer, 2016; Patton et al., 2019). Analysis of remains have shown that ooligan could 35 comprise up to 67% of traditional historical fisheries catches (Patton et al., 2019). Because ooligan spawning 36 relies on the timing of the spring freshet, and because the species has declined in the last 25 years due to 37 fishing pressure and predation, the species may be at risk from combined climate and non-climate drivers 38 (medium confidence) (Talloni-Álvarez et al., 2019). Projections under RCP2.6 or RCP8.5 estimate 39 reductions by 21% or 31% by 2050 in essential nutrients from traditional seafood for Indigenous Peoples in 40 Canada, relative to 2000, with a modelled nutritional deficit that includes non-traditional dietary 41 substitutions (Marushka et al., 2019). 42 43 3.5.4.2 Non-Consumable Goods 44 45 Limited evidence about climate impacts exists for valuable non-food aquatic materials. Ocean warming and 46 acidification harm red coral (Corallium rubrum) (Bramanti et al., 2013) and communities hosting black coral 47 (Antipatharian spp.), both used for jewellery (Ross et al., 2020). While no-take MPAs (Section 3.6.3.2) 48 enhance red coral structural complexity, they only weakly compensate for warming effects (Cerrano et al., 49 2013; Montero-Serra et al., 2019). Antipatharian spp. are not well studied or monitored (Gress and Andradi- 50 Brown, 2018). Acidification and warming negatively impact pearl oysters (Welladsen et al., 2010; Liu and 51 He, 2012; Liu et al., 2012; Hoegh-Guldberg et al., 2014; Zhang et al., 2019b). Projected climate impacts for 52 2035 would decrease the average net present value of French Polynesia's pearl aquaculture industry by 53 29.1% compared to the present (Hilsenroth et al., 2021). Climate impacts on ornamental species sought by 54 aquarists have not been well studied (Dee et al., 2019b). 55 56 Decreasing the vulnerability of renewable energy installations, particularly wind turbines, to climate risks 57 (Table 3.26, Bindoff et al., 2019) could include technological adaptations (Section 3.6.2.2) such as storm Do Not Cite, Quote or Distribute 3-107 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 "survival mode" settings (Penalba et al., 2018); preparation for hazards such as icing, SLR, drifting sea ice, 2 and wave activity (Neill et al., 2018; Goodale and Milman, 2019; Solaun and Cerdá, 2019), and biofouling 3 (medium confidence) (Want and Porter, 2018; Joyce et al., 2019; Vinagre et al., 2020), which is expected to 4 increase in response to warming and acidification (medium confidence) (Dobretsov et al., 2019; Khosravi et 5 al., 2019; Liu et al., 2020d; Lamim and Procópio, 2021). Macroalgae and fish processing byproducts are 6 being tested for biofuel use (Greene et al., 2016; Alamsjah et al., 2017; Saifuddin and Boyce, 2017; 7 Sakthivel et al., 2018; Sudhakar et al., 2019; Nguyen et al., 2020; Ramachandra and Hebbale, 2020; Tan et 8 al., 2020), but weather variability could pose financial risk to this sector (Kleiman et al., 2021). 9 10 3.5.5 Supporting and Regulating Services 11 12 Ocean and coastal regulating services are detailed in Table 3.25. The economic value of all regulating 13 ecosystem services in 2015 was estimated at 29.1 trillion USD, with water- and climate-regulating services 14 contributing the most (Balasubramanian, 2019). 15 16 3.5.5.1 Habitat Creation and Maintenance and Larval Dispersal 17 18 Climate impacts have already altered ocean and coastal habitats (Section 3.4.2, Table 3.26, Gissi et al., 2021) 19 in ways that have led to species range shifts, biodiversity changes, phenology changes, and regime shifts 20 (Section 3.4.3) from the surface ocean to the seafloor (very high confidence) (Box 3.3, Figure 3.22). 21 Continued ocean and coastal habitat impacts are projected, and their severities will depend on emissions 22 scenario and co-occurring drivers (Section 3.4.3, Qiu et al., 2019) or extremes (e.g., Babcock et al., 2019). 23 Warming and physical circulation are projected to change larval dispersal, a habitat-related service 24 (Bashevkin et al., 2020), but identifying probable outcomes remains challenging owing to the high 25 variability among species, locations, and recruitment (Schilling et al., 2020; King et al., 2021; Le Corre et 26 al., 2021; Raventos et al., 2021). Climate risks to habitat can be decreased by reducing non-climate drivers, 27 preserving ecosystems, or restoring habitat (Sections 3.6.2, 3.6.3.2). Risk to larval dispersal cannot be 28 meaningfully addressed at scale by human-implemented adaptations; instead, declines in this service will 29 pressure natural systems to adapt via physiological plasticity or evolution (Section 3.3, Bashevkin et al., 30 2020). 31 32 3.5.5.2 Climate Regulation and Air Quality 33 34 Climate regulation by the ocean depends on physical and biogeochemical processes (Sections 3.2­3.4) that 35 create, move, and store heat, water vapor and other climate-active compounds including CO2, methane, and 36 dimethyl sulfide and methane (WGI AR6 Chapter 6, Naik et al., 2021). Over the 21st century, ocean heat 37 and CO2 uptake will continue (WGI AR6 SPMB4.1, B5.1, IPCC, 2021b) and sea ice loss from warming will 38 allow some additional CO2 uptake (Armstrong et al., 2019), but the ocean will take up a smaller fraction of 39 CO2 emissions as atmospheric CO2 concentrations rise (high confidence) (Table 3.26, WGI AR6 SPM B4.1, 40 IPCC, 2021b). 41 42 There is very limited evidence on climate-driven air-quality changes in the coastal zone. Increased humidity 43 decreases the lifetime of ozone and increases particulate matter and indoor mold levels (USGCRP, 2016), 44 potentially affecting near-shore air quality. However, coastal zone air pollution can enhance coastal climate 45 impacts by increasing risk of acid rain, which worsens ocean acidification (nitrogen oxides, sulphur oxides, 46 and mercury, Doney, 2010; Northcott et al., 2019). 47 48 3.5.5.3 Provision of Fresh Water, Maintenance of Water Quality, and Regulation of Pathogens 49 50 The salinities of many estuaries, deltas, coastal fresh water aquifers, and soils around the world are 51 increasing, and this decrease in water quality is endangering human health and agricultural yields (very high 52 confidence) (Table 3.26, Section 3.4.2.4, Bindoff et al., 2019; Bouderbala, 2019; Rahman et al., 2019; Naser 53 et al., 2020; Rakib et al., 2020; Mastrocicco and Colombani, 2021). Coastal salinisation is attributed to 54 regionally varying combinations of climate-impact drivers, like SLR and storm-related flooding by seawater, 55 and non-climate drivers, like water withdrawal and land use changes (very high confidence) (Islam et al., 56 2019; Rahman et al., 2019; Paldor and Michael, 2021). Monitoring-related adaptations (Section 3.6.2.2.2) 57 including advances in modeling and monitoring are providing decision-relevant, regional-scale information Do Not Cite, Quote or Distribute 3-108 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 (Colombani et al., 2016; Mukhopadhyay et al., 2019; Slama et al., 2020; Corwin, 2021). For example, new 2 projections indicate which drinking water intake stations on China's Pearl River Estuary will be unable to 3 meet demands by 2100 due to SLR and drought (Wang and Hong, 2021), while others show that SLR effects 4 on seawater intrusion into the coastal aquifer in Kerala, India under both RCP4.5 and RCP8.5 scenarios are 5 negligible (Sithara et al., 2020). Salinisation-associated changes may disproportionately burden women 6 responsible for securing drinking water and fuel, such as in the Indian Sundarbans (Mukhopadhyay et al., 7 2019). Salinisation will continue to endanger coastal water and soil quality in the future (high confidence) 8 (Islam et al., 2019; Paldor and Michael, 2021), but the evidence assessed above shows that subsequent 9 impacts to human health and agriculture will depend heavily on regional variations in environment and 10 human behaviour (medium confidence). 11 12 Together, climate and non-climate drivers can mobilise toxins and contaminants in ways that harm human 13 and marine species health (very high confidence) (Box 3.2), and climate change is altering these relationships 14 (high confidence) (Table 3.26, Bindoff et al., 2019). Under warming or ocean acidification, marine molluscs 15 exposed to pharmaceuticals via wastewater experience more detrimental biological consequences or greater 16 bioaccumulation (limited evidence, high agreement) (Costa et al., 2020a; Costa et al., 2020b; Dionísio et al., 17 2020; Freitas et al., 2020; Kibria et al., 2021). Physical circulation, temperature, and biogeochemical 18 characteristics (Bowman et al., 2020; Liu et al., 2020a; Liu et al., 2020b; Sun et al., 2020; Zhang et al., 19 2020b) control the ubiquitous oceanic distribution of methylmercury, and ocean acidification- and warming- 20 driven changes in planktonic speciation and interactions can promote additional food-web bioaccumulation 21 of methylmercury (Tada and Marumoto, 2020; Wu et al., 2020b; Zhang et al., 2020b; Zhang et al., 2021a). 22 Interactions among drivers also matter: temperature plus overfishing increased tissue methylmercury 23 concentrations in Atlantic bluefin tuna from the 1970s to the 2000s more than the decreases in the late 1990s 24 and 2000s from lower environmental mercury levels (Schartup et al., 2019). This appears true for persistent 25 organic pollutants as well, but their bioaccumulation is related more to temperature effects on animal 26 behaviour than on pollutant dynamics (Houde et al., 2019; Wagner et al., 2019; Kalia et al., 2021). By 2100 27 under RCP8.5, productivity changes and community structure shifts are expected to increase methylmercury 28 concentrations in polar oceans and high-latitude phytoplankton and decrease it in low latitudes (Zhang et al., 29 2021a). The estimated average global cost of mercury-related health effects by 2050, mainly from seafood 30 consumption during 2010­2050, will be 19 trillion USD (2020), assuming a 3% discount rate, if 31 methylmercury emissions are not reduced (Zhang et al., 2021b). 32 33 Since previous assessments, evidence has increased that climate impacts such as warming, extreme weather, 34 and SLR are increasing the geographic spread and risk of marine-borne human pathogen outbreaks, 35 including Vibrio spp. (very high confidence) (Table 3.26, Bindoff et al., 2019; Logar-Henderson et al., 2019; 36 Froelich and Daines, 2020; Montánchez and Kaberdin, 2020; Semenza, 2020; Ferchichi et al., 2021). 37 Climate change affects at least 30 human pathogens with aquatic-system infection routes (e.g., ingestion of 38 contaminated water or seafood, or contact with wounds, see Table SM3.2, Cross-Chapter Box ILLNESS in 39 Chapter 2, Nichols et al., 2018). Conditions favourable for Vibrio cholerae are increasing globally, which 40 raises risk to humans (Cross-Chapter Box ILLNESS in Chapter 2). Increased storm-related flooding and 41 SLR further increase human encounters with Vibrio spp. (Froelich and Daines, 2020). Aquatic diseases, 42 particularly Vibrio spp., have caused large economic losses in aquaculture by decreasing the quality or 43 survival of cultured species (Lafferty et al., 2015; Novriadi, 2016). Temperature-based model projections 44 show that all Canadian shellfish beds will experience conditions that promote high risk of Vibrio spp. growth 45 by 2100 for both RCP4.5 and RCP8.5 scenarios (Ferchichi et al., 2021). Climate-impact drivers may increase 46 Vibrio spp. loads in seafood species: laboratory-simulated heatwaves increase Vibrio spp. abundance in 47 Pacific oyster (Crassostrea gigas) (Green et al., 2019) and simulated ocean acidification increases hard clam 48 (Mercenaria mercenaria) susceptibility to Vibrio spp. infection (Schwaner et al., 2020). Projected increases 49 in temperature, extreme and variable rainfall conditions, coastal flooding, and SLR (Section 3.2, Cross- 50 Chapter Box SLR in Chapter 3) strongly increase the risk of frequent and severe aquatic human pathogen 51 outbreaks in ocean and coastal areas that will continue to harm human health and cause economic losses 52 (high confidence) (Cross-Chapter Box ILLNESS in Chapter 2, Froelich and Daines, 2020; Semenza, 2020; 53 Ferchichi et al., 2021). Section 3.6.3.1.5 assesses human adaptations to increasing risk of marine-borne 54 pathogens. 55 56 Climate-driven changes in temperature, salinity (from ice melt and precipitation changes), deoxygenation, 57 and ocean acidification can alter dynamics of infectious diseases that target ocean and coastal species by Do Not Cite, Quote or Distribute 3-109 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 increasing hosts' susceptibility or pathogens' abundance or virulence (high confidence) (Burge and 2 Hershberger, 2020; Byers, 2021). Coral and urchin diseases have increased over time driven by warming- 3 related declines in organism recovery and survival or immunity (medium confidence) (Cohen et al., 2018; 4 Tracy et al., 2019). Seagrass and sea star wasting disease outbreaks have occurred under combinations of 5 ocean warming or MHWs and non-climate drivers (e.g., eutrophication, bottom trawling), but attribution of 6 these outbreaks to specific drivers is still not resolved (Harvell et al., 2019; Jakobsson-Thor et al., 2020; 7 Krause-Jensen et al., 2021). Disease outbreaks threaten marine biodiversity, species that create habitat or 8 dampen wave action, and keystone species (Harvell and Lamb, 2020). Attributing observed changes in 9 marine disease patterns to climate remains extremely difficult owing to interacting climate and non-climate 10 drivers (Burge and Hershberger, 2020) and lack of baseline data (Tracy et al., 2019). Projected increases in 11 the frequency, duration, and intensity of warming events would reduce survival and recovery of some 12 species from hot events, reduce immunity of other species to pathogens, extend poleward ranges of some 13 pathogens, and increase infection risk when host species congregate in scarce habitat (Cohen et al., 2018). 14 Pathogens that target ocean and coastal organisms may themselves be sensitive to future climate conditions 15 or subsequent ecosystem changes, which challenges development of projections (Cohen et al., 2018; Burge 16 and Hershberger, 2020). 17 18 Following the high confidence assessment of SROCC Table 3.26, Bindoff et al. (2019) that risks associated 19 with HABs will continue to increase with warming and rising CO2 in the 21st century, new examples have 20 illustrated how toxic HABs interfere with regulating, provisioning (Section 3.5.3), and cultural ecosystem 21 services (Section 3.5.6) in interconnected ways (limited evidence, high agreement). A massive toxic Pseudo- 22 nitzschia spp. bloom in 2013­2016 along the United States (US) West Coast triggered Dungeness crab, rock 23 crab, and razor clam fishery closures to protect human consumers (Sections 3.6.2, 3.6.3.1.5, McCabe et al., 24 2016), and this disproportionately harmed fishers, especially small-vessel owners, and fishing-support 25 service industries, primarily through lost revenue (Ritzman et al., 2018; Moore et al., 2019; Trainer et al., 26 2019; Jardine et al., 2020; Moore et al., 2020a). Toxic Alexandrium spp. blooms promoted by climate-driven 27 coastal extremes (e.g., MHWs, stratification, runoff) in Tasmania, Australia, in 2012 and Chile in 2016 28 caused fish kills, shellfish product recalls, substantial economic losses, and human sickness and death 29 (Trainer et al., 2019). The Chile event caused an estimated loss of 800 million USD in the farmed salmon 30 industry (Díaz et al., 2019) and resulted in a series of large, long-lasting regional protests calling for national 31 aid (Delgado et al., 2019). New evidence, however, suggests that the perceived global increase in harmful 32 algal blooms results from better monitoring and more detrimental bloom impacts rather than a climate-linked 33 mechanism (Hallegraeff et al., 2021). 34 35 Natural and engineered systems have long been used effectively to manage precipitation and wastewater 36 safely (Chapter 4, Box 4.5), and maintaining and enhancing them is a key nature-based adaptation strategy 37 for coastal communities (Section 3.6.2.3, Cross-Chapter Paper 2). Estimated values of water purification and 38 stormwater management provided by coastal ecosystems are in the hundreds to thousands of USD per 39 hectare [e.g., 272 Euro per 0.01 km2 yr­1 from the Mediterranean's sandy coastline (Hérivaux et al., 2018); 40 1100­2800 USD per 0.01 km2 yr­1 from the state of Maryland, USA (Campbell et al., 2020b); 600 USD per 41 0.01 km2 yr­1 in Zhuzhou City, China (Zhan et al., 2020)]. Both wild and cultured organisms also provide 42 filtration services. Seagrasses' ability to purify water is well recognised by coastal residents and ocean 43 resource users in tropical and temperate locations (Ambo-Rappe et al., 2019; Quevedo et al., 2020; Heckwolf 44 et al., 2021; McKenzie et al., 2021a). Globally, aquacultured shellfish remove an estimated 49,000 tonnes of 45 nitrogen and 6000 tonnes of phosphorus from coastal waters, worth a potential 1.20 billion USD, and they 46 may help improve existing engineered wastewater treatment systems (van der Schatte Olivier et al., 2020). 47 Climate change, especially episodic extreme rains and RSLR (Romero-Lankao et al., 2014), is challenging 48 management and design of wastewater and stormwater systems (high confidence) (Flood and Cahoon, 2011; 49 Trtanj et al., 2016; Hummel et al., 2018; Kirshen et al., 2018; Nazarnia et al., 2020; Reznik et al., 2020; 50 McKenzie et al., 2021b) and integrity of coastal landfills (Beaven et al., 2020). Without substantial 51 adaptation that addresses projected wastewater management challenges and community needs (Section 52 4.2.6.1, Kirshen et al., 2018; Kirchhoff and Watson, 2019; Kool et al., 2020; Nazarnia et al., 2020; Hughes et 53 al., 2021), coastal water quality in many areas will decrease because of more frequent or severe releases of 54 untreated wastes (high confidence) (Flood and Cahoon, 2011; Hummel et al., 2018; Hughes et al., 2021; 55 McKenzie et al., 2021b), and this will have harmful consequences for human and coastal ecosystem health 56 (high confidence) (Cross-Chapter Box ILLNESS in Chapter 2, Section 4.2.6.1, Bindoff et al., 2019). 57 Do Not Cite, Quote or Distribute 3-110 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 3.5.5.4 Regulation of Physical Hazards 2 3 Coastal ecosystems physically protect people and property from storms and flooding, and climate change 4 threatens this protection function (Table 3.26, Figure 3.22). Increasingly detailed models show how warm- 5 water coral reefs (Reguero et al., 2019; Storlazzi et al., 2019; Reguero et al., 2021) mangroves (Blankespoor 6 et al., 2017; Menéndez et al., 2020; Trégarot et al., 2021) and wetlands (Sun and Carson, 2020) prevent 7 billions of USD of direct and indirect damage to private and public property and shield millions of people 8 from flooding each year. Protection by mangroves provides more economic benefits in higher-income 9 nations and shields more people in lower-income nations (Menéndez et al., 2020). Seagrasses (James et al., 10 2020; James et al., 2021), kelp (Morris et al., 2020b; Zhu, 2020), suspended shellfish aquaculture (Gentry et 11 al., 2020; Zhu et al., 2020a), oyster reefs (Chowdhury et al., 2019) coastal wetlands (Möller, 2019; Keimer et 12 al., 2021), and sandy coastlines (Section 3.4.2.6, Hérivaux et al., 2018) also measurably decrease wave 13 energy. Non-climate drivers (e.g., invasive species (James et al., 2020), sediment supply changes (Ganju, 14 2019; Ladd et al., 2019; Ilia, 2020), erosion and storm damage (Mehvar et al., 2019; Bacopoulos and Clark, 15 2021)], acting together with climate-impact drivers and impacts (e.g., sea level rise (Cross-Chapter Box SLR 16 in Chapter 3), changes in plant biodiversity (Section 3.5.2, Lee Smee, 2019; Silliman et al., 2019; Schoutens 17 et al., 2020), MHWs (Section 3.4.3.7), and acidification (Section 3.4.2.1)) compromise physical protection 18 by coastal ecosystems (very high confidence). See Cross-Chapter Box SLR in Chapter 3 and Sections 3.6.3.1 19 and 3.6.3.2.2 for assessment of adaptations that address this ecosystem service. 20 21 3.5.5.5 Regulation of Carbon Cycling in Ocean and Coastal Ecosystems 22 23 Current and future total carbon storage and cycling in the ocean are governed by past and future CO2 24 emissions trajectories (Table 3.26), but regional ocean and coastal carbon stocks and cycling vary over time 25 and space due to processes being altered by climate, including ocean circulation, sea-ice cover, coastal 26 upwelling, and thermal stratification (Section 3.2.2.3); ocean primary production and export (Sections 3.2.3, 27 3.4.4); and marine ecosystem biodiversity (high confidence) (Figure 3.22, Section 3.5.2). Quantifying 28 regional carbon fluxes and stocks is still challenging and relies on indirect measures (e.g., Fennel et al., 29 2019; Clay et al., 2020), especially in coastal ecosystems where drivers interact. Carbon cycling and storage 30 co-occurs with other regulating services such as habitat provision, water quality maintenance, and coastal 31 protection (Ouyang et al., 2018), particularly in vegetated coastal ecosystems (Box 3.4). Adaptations to 32 support regional carbon cycling and storage generally focus on area-based management and conservation 33 (Section 3.6.3.2), but interventions to enhance ocean carbon storage are being explored for mitigation 34 (WGIII AR6 Chapter 7). 35 36 37 [START BOX 3.4 HERE] 38 39 Box 3.4: Blue Carbon Ecosystems 40 41 Climate change and other anthropogenic drivers, including eutrophication, land-use changes, and 42 overexploitation, directly and indirectly threaten blue carbon ecosystems (Annex II: Glossary). Commonly 43 considered blue carbon ecosystems include vegetated coastal ecosystems (Sections 3.4.2.3­3.4.2.5), whose 44 mangroves, saltmarshes and seagrass beds host rooted, vascular plants known to store large amounts of 45 carbon for long periods and to be amenable to management (Lovelock and Duarte, 2019). Other ocean and 46 coastal taxa including rooted or floating macroalgae (e.g., non-vascular multicellular kelp or seaweed genera 47 such as Macrocystis spp., Sargassum spp., or Laminaria spp., Filbee-Dexter and Wernberg, 2020), 48 phytoplankton, and even pelagic fauna (e.g., finfish or whales, Chami et al., 2019) have also been proposed 49 as blue carbon ecosystems. Terrestrial vascular-plant-derived material can also carry and store significant 50 amounts of carbon in marine environments (Cragg et al., 2020). There is increasing evidence about the 51 coverage and carbon content of macroalgal, planktonic, and faunal taxa, but low agreement about their long- 52 term carbon storage potential and manageability (Alongi, 2018b; Wernberg and Filbee-Dexter, 2018; 53 Lovelock and Duarte, 2019; Ortega et al., 2019; Pfister et al., 2019; Queirós et al., 2019; Filbee-Dexter et al., 54 2020a; Gallagher, 2020; Mariani et al., 2020; Thorhaug et al., 2020; van Son et al., 2020; Bach et al., 2021; 55 Bayley et al., 2021; Cavanagh et al., 2021; Frontier et al., 2021; Martin et al., 2021; Pedersen et al., 2021; 56 Weigel and Pfister, 2021). This section focuses on the array of ecosystem services and adaptation 57 opportunities provided by vegetated coastal blue carbon ecosystems, where consensus and evidence are most Do Not Cite, Quote or Distribute 3-111 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 abundant. Mitigation potential of blue carbon ecosystems is assessed with land-based mitigation options in 2 WGIII AR6 Section 7.4. 3 4 Carbon storage and burial in mangroves, saltmarshes and seagrass meadows (Table Box3.4.1) help regulate 5 ocean and coastal carbon cycling and may contribute to nature-based mitigation, although regional estimates 6 vary widely based on climatic and edaphic conditions (WGIII AR6 Section 7.4). In addition, coastal 7 vegetated ecosystems provide substantial and interdependent regulating, provisioning and cultural ecosystem 8 services. These include disproportionately high biodiversity per unit area (Pörtner et al., 2021a); abundant 9 habitat (Section 3.5.5.1) and nurseries for aquatic, terrestrial, aerial, and microbial species; natural filtration 10 of waste and stormwater runoff into the coastal ocean (Sections 3.5.5.3, 4.2.7, Cross-Chapter Box ILLNESS 11 in Chapter 2); coastal protection (Section 3.5.5.4, Ouyang et al., 2018; Quevedo et al., 2020); food and 12 natural materials (Sections 3.5.3, 3.5.4); and support for tourism, livelihoods, and cultural activities (Section 13 3.5.6). Global estimates of services provided by coastal blue carbon ecosystems depend on the quality of 14 available mapping, which is currently best developed for mangroves (Macreadie et al., 2019), and improving 15 for saltmarshes and seagrasses (McOwen et al., 2017; McKenzie et al., 2020; Young et al., 2021). 16 17 18 Table Box3.4.1: Estimates of organic carbon storage and burial rates in mangroves, saltmarshes, and seagrass 19 meadows. Estimates are the mean ± 95% confidence interval, where available (indicating the extremely likely range) 20 and range. Carbon stocks for mangroves include above- and below-ground storage up to 3 m depth (sampling period 21 2007­2017). The estimates for saltmarsh and seagrass stocks are soil stocks up to 1 m depth (observations spanning 22 1983­2016 for saltmarshes and until 2016 for seagrass meadows). Date ranges for the burial rates are: 1989­2020, 23 1975­2020 and 1956­2016 for mangroves, saltmarshes and seagrass meadows, respectively. Mangroves Saltmarshes Seagrass meadows Carbon stocks (MgC ha­1) 856 ± 64.2 (79­2208) 317.2 ± 38.2 (27­1900) 139.7 (9.1­628) (Fourqurean (Kauffman et al., 2020) (Alongi, 2018c) et al., 2012; Alongi, 2018d) Carbon burial rate (g C m­2 194 ± 30 (6.2­1722) (Wang 168 ± 14 (1.2­1167.5) 220.7 ± 40.2 (-2094­2124) (Alongi, 2018d) yr­1) et al., 2020) (Wang et al., 2020) Global Carbon burial rate 41 (Wang et al., 2020) 12.63 (Wang et al., 2020) 35.31 (Alongi, 2020) (TgC yr­1) Global areal coverage (Mha) 13.7 (Richards et al., 2020) 5.5 (McOwen et al., 2017) 16 (McKenzie et al., 2020) 24 25 26 Coastal vegetated ecosystems are vulnerable to harm from multiple climate and non-climate drivers, and 27 together these have reduced wetland area globally (high confidence) (Section 3.4.2.5) and endangered the 28 services provided by these ecosystems (high confidence). Loss of coastal vegetated ecosystems changes 29 biodiversity (Sections 3.5.2, 3.4.2.3­3.4.2.5) (Numbere, 2019; Parreira et al., 2021), increases risk of damage 30 and erosion from SLR and storms (Sections 3.4.2.3­3.4.2.5, Cross-Chapter Box SLR in Chapter 3, Galeano 31 et al., 2017), and impacts provisioning (Sections 3.5.3­3.5.4, Li et al., 2018b; Maina et al., 2021). These 32 changes also strongly determine the quantity and longevity of blue carbon storage (high confidence) 33 (Macreadie et al., 2019; Lovelock and Reef, 2020). Specific site characteristics and ecosystem responses to 34 climate change will determine future local blue carbon storage or loss (high confidence) (Table Box3.4.2). 35 For instance, poleward migration of mangroves to areas dominated by salt marshes is expected to increase 36 carbon storage (Kelleway et al., 2016); however, this change in the dominant vegetation and associated 37 faunal changes can modify carbon stocks and sequestration, as well as other ecosystem services (Martinetto 38 et al., 2016; Kelleway et al., 2017; Smee et al., 2017; Macreadie et al., 2019; Macy et al., 2019). Landward 39 range expansion of mangroves, marshes, and seagrass in response to gradual RSLR can enhance carbon 40 sequestration (Cross-Chapter Box SLR in Chapter 3, Section 3.4.2.5, Macreadie et al., 2019), but coastal 41 squeeze can limit this (Phan et al., 2015; Schuerch et al., 2018) and RSLR can either submerge and bury or 42 erode and release stored blue carbon (Section 3.4.2.5, Macreadie et al., 2019; Lovelock and Reef, 2020). 43 Gains and losses of mangrove habitat area (and therefore carbon storage) projected for nations under RCP4.5 44 and RCP8.5 depend primarily on the combination of SLR rate, adaptation scenario (including coastal 45 development), and island or continental status (Lovelock and Reef, 2020). The influence of warming, 46 MHWs, and acidification on seagrass meadows (Kendrick et al., 2019; Strydom et al., 2020) and associated 47 coralligenous reefs (Zunino et al., 2019) suggests that future warming and especially MHWs will cause more 48 widespread loss of services from these ecosystems (Section 3.4.2.5). Loss of blue carbon ecosystems will not 49 only halt carbon storage, but also release stored carbon: emissions after 2000 due to global mangrove Do Not Cite, Quote or Distribute 3-112 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 deforestation have been estimated at 23.5­38.7 Tg Cyr­1 (Ouyang and Lee, 2020). Mitigation estimates for 2 avoided conversion and restoration of coastal wetlands and the implications of the impacts of climate change 3 are assessed in WGIII AR6 Section 7.4. 4 5 To date, initiatives aiming to restore coastal wetland ecosystems primarily address ecosystem characteristics 6 other than carbon storage (Herr et al., 2017; de los Santos et al., 2019; Lovelock and Duarte, 2019; Friess et 7 al., 2020a). But recovery of coastal vegetated ecosystems is expected to bring back the full suite of 8 ecosystem services they provide, not just carbon storage (medium confidence) (Marbà et al., 2015a; Burden 9 et al., 2019; Friess et al., 2020a), making coastal restoration a low-risk action that offers both adaptation and 10 mitigation benefits (Steven et al., 2020; Gattuso et al., 2021). Successful restoration requires using 11 appropriate plant species in suitable environmental settings (Wodehouse and Rayment, 2019; Friess et al., 12 2020a) with favourable geomorphology and biophysical conditions (Cameron et al., 2019; Ochoa-Gómez et 13 al., 2019) and considering social, economic, policy, and operational constraints (Section 3.6.3.2.2, Cross- 14 Chapter Box NATURAL in Chapter 2), now and in the future (high confidence) (Duarte et al., 2020; 15 Lovelock and Reef, 2020). Nevertheless, restored spaces may not store carbon at rates equal to those of 16 undisturbed spaces (Yang et al., 2020), and it may take decades to determine or achieve carbon storage 17 outcomes of restoration (Sasmito et al., 2019; Duarte et al., 2020; Oreska et al., 2020). Integration improves 18 efforts to restore or conserve coastal wetland ecosystems to accomplish both adaptation and mitigation 19 outcomes (Steven et al., 2020). Government-led conservation of blue carbon ecosystems as part of national 20 and subnational climate strategies (e.g., Friess et al., 2020a; Kelleway et al., 2020; Wedding et al., 2021) 21 benefits from coordination with private activities, such as incentivising conservation with payments for 22 ecosystem services (Muenzel and Martino, 2018; Friess et al., 2020a). Moreover, successful area-based 23 protection measures consider both environmental and social issues (Section 3.6.3.2). Continued integration 24 and alignment of policies at international to local levels (Section 3.6.5) will also support achieving the 25 adaptation and mitigation benefit of blue carbon spaces (Friess et al., 2020a; Steven et al., 2020; Wu et al., 26 2020a). 27 28 29 Table Box 3.4.2: Examples of vegetated blue carbon ecosystem carbon storage gains and losses in response to climate- 30 impact drivers, and key actions contributing to maintained or and increased carbon storage. "+C" indicates potential 31 positive effects on blue carbon stocks, "­C" indicates negative effects, "0" indicates no effects, and "C" indicates 32 positive or negative effects. Effects on carbon stocks from 3.4.2.5, Macreadie et al. (2019); Lovelock and Reef (2020); 33 Wang et al. (2020). Key actions to sustain blue carbon storage from Duarte et al. (2020); Wedding et al. (2021). Mangroves Saltmarshes Seagrasses Sea-level rise Landward expansion by +C +C +C vegetation Coastal squeeze ­C ­C ­C Loss of low-lying or ­C ­C ­C submerged land or vegetation Human adaptation to +C +C increase accommodation space Extreme storms Erosion/ loss of area/ ­C ­C 0 to ­C subsidence Enhanced sedimentation +C +C +C Vegetation damage and ­C to +C ­C mortality Warming Increased productivity +C +C Vegetation mortality ­C Increased decomposition of ­C ­C to +C soil Poleward expansion of +C ­C mangroves Do Not Cite, Quote or Distribute 3-113 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report Poleward expansion of +C seagrasses Poleward expansion of C bioturbators Change in dominant species C Rising concentrations of atmospheric CO2 Increased productivity of C C +C some species Biodiversity loss ­C Altered precipitation Vegetation mortality ­C Reduced productivity ­C ­C Increased productivity +C +C Increased remineralisation ­C ­C Low salinity events 0 to ­C Key actions to sustain blue carbon storage Protect ecosystems X X X Develop alternative X livelihoods Provide space for landward X X migration Restore hydrological X X connections Maintain/restore sediment X X supply Restore ecosystems X X Plant indigenous species X Reduce nutrient inputs X 1 2 3 [END BOX 3.4 HERE] 4 5 6 3.5.6 Cultural Services 7 8 Cultural services provided by ocean and coastal ecosystems help maintain psychological well-being, cultural 9 development, human identities, educational opportunities, and reserves that could support development of 10 future goods or activities (Table 3.25). Most recent studies of ocean and coastal cultural services simply 11 detail local benefits using replicable methods (e.g., Drakou et al., 2018; Folkersen, 2018; Förster et al., 2019; 12 Lau et al., 2019; Pouso et al., 2019; Weitzman, 2019; Yang et al., 2019), focusing on diverse ocean and 13 coastal environments and ecosystems (Jobstvogt et al., 2014; Balzan et al., 2018; Drakou et al., 2018; Ingram 14 et al., 2018; Pouso et al., 2018; Zapata et al., 2018; Ghermandi et al., 2019; Pouso et al., 2019; Tanner et al., 15 2019; Turner et al., 2019; Ortíz Liñán and Vázquez Solís, 2021). Cultural ecosystem services may directly 16 benefit from marine development activities, such as marine aquaculture (e.g., Alleway et al., 2018), and 17 indirectly benefit from marine activities that increase biodiversity (e.g., Causon and Gill, 2018). Cultural 18 services are generally quantified using interviews and revealed-preference or stated-preference valuation 19 (National Research Council, 2005; Sangha et al., 2019), but people often are especially reluctant to evaluate 20 cultural ecosystem services in monetary terms, given the spiritual and community linkages to these services 21 (Oleson et al., 2018). 22 23 Additional evidence since previous assessments (Table 3.26) confirms that climate-change impacts on ocean 24 and coastal cultural ecosystem services have already disrupted people's place-based emotional attachments 25 and cultural activities (limited evidence, high agreement) (Figure 3.22). Bleaching and mortality of corals in Do Not Cite, Quote or Distribute 3-114 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 the Great Barrier Reef have induced measurable "reef grief," a type of solastalgia, among reef visitors and 2 researchers (Conroy, 2019; Curnock et al., 2019; Marshall et al., 2019). The mental health of people in 3 Tuvalu (Gibson et al., 2020), Alaska (Allen, 2020), and Honduras (Kent and Brondo, 2020) have suffered 4 from both the experience of climate impacts on ocean and coastal ecosystems (e.g., SLR and changes in 5 fisheries and wildlife), and the anticipation of more in the future. The climate-associated MHWs and harmful 6 algal bloom events in 2014­2016 in the US Pacific Northwest (Moore et al., 2019) prevented seasonal razor 7 clam harvests culturally important to Indigenous Peoples and the local community (Section 3.5.5.3, Crosman 8 et al., 2019). SLR and storm-driven coastal erosion endanger coastal archaeological and heritage sites around 9 the world (very high confidence) (Hoque and Hoque, 2008; Carmichael et al., 2018; Reimann et al., 2018; 10 Elliott and Williams, 2019; Ravanelli et al., 2019; Anzidei et al., 2020; Chemeli et al., 2020; García Sánchez 11 et al., 2020; Harkin et al., 2020; Hil, 2020; Rivera-Collazo, 2020). 12 13 Disruptions in ocean and coastal ecosystem services partly attributable to climate change have also caused 14 economic losses (limited evidence, high agreement). Water quality deterioration over 24 years in a US 15 temperate bay due to nutrient enrichment and warming caused 0.08­0.67 million USD per decade in lost 16 recreational shellfish revenues (Luk et al., 2019). In southwestern Florida, USA, where nutrient enrichment, 17 lake hydrology, and rainfall conditions control cyanobacterial HAB formation (Havens et al., 2019), toxic 18 HAB events deterred visitors and recreation, leading to lodging and restaurant revenue losses (Bechard, 19 2020), decreased domestic and international arrivals and overall visitor spending (a 99 million USD loss 20 from August to October 2018, Scanlon, 2019), and lost recreational spending from loss of boat-ramp access 21 (a 3 million USD economic loss from June to September 2018, Alvarez et al., 2019). In Cornwall, England, 22 HABs from 2009­2016 disrupted residents' sense of place, identity, and well-being by interrupting 23 recreational and economic activities, and by creating feelings of uncertainty and unease around the safety or 24 dependability of future ocean-related activities (Willis et al., 2018). Increasingly abundant Sargassum spp. 25 floating macroalgae from the central Atlantic Ocean and Caribbean Sea, whose proliferation has been 26 attributed to high sea surface temperatures and nutrient enrichment (Wang et al., 2019a), has substantially 27 disrupted beach tourism in the Caribbean and Mexico and imposes millions of dollars of clean-up costs 28 annually on affected beaches (Milledge and Harvey, 2016). 29 30 Observed disruption of ocean and coastal cultural services by climate impacts, plus increasingly severe and 31 widespread projected climate change impacts on ocean and coastal ecosystems, imply that risk to cultural 32 ecosystem services will remain constant or grow (medium confidence) (Figure 3.22, Table 3.26). Recent 33 studies assert that cultural ecosystem services are at risk from climate change (high confidence) (Singh et al., 34 2019a; Koenigstein, 2020). However, limited evidence and complex social-ecological interactions (e.g., 35 Ingram et al., 2018) challenge development of specific projections. For instance, the little auk (Alle alle) in 36 the North Water Polynya is traditionally harvested by Indigenous Inughuit for food and community-wide 37 celebrations and seasonal activities, but harvests are threatened to an undetermined degree as the seabird 38 competes for food with recovering bowhead whale (Balaena mysticetus) populations and northward range 39 shifts of capelin (Mallotus villosus) due to warming (Mosbech et al., 2018). Section 3.6 assesses the cultural 40 implications of implemented human adaptations. 41 42 Do Not Cite, Quote or Distribute 3-115 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure 3.22: Observed global influence of climate-impact drivers on ecosystem services. (Coloured cells) The 3 "observed impact" indicates the total effect of all climate-impact drivers on a specific ecosystem service, using expert 4 judgement based on summary statements throughout Section 3.5. (Grey cells) Co-occurring non-climate drivers that 5 affect the service. Cell colour shows whether the observed impact of the climate-impact driver on a group of ecosystem 6 services is positive (beneficial), negative (detrimental) or mixed (usually resulting from location, the presence of 7 interacting drivers, or changing effects over time). No assessment indicates that not enough evidence is available to 8 assess the direction of impact. 9 10 11 3.6 Planned Adaptation and Governance to Achieve the Sustainable Development Goals (SDGs) 12 13 3.6.1 Point of Departure 14 15 Human adaptation comprises an array of measures (adaptation options, IPCC, 2014a) that modulate harm or 16 exploit opportunities from climate change (Section 1.2.1.3). Adaptation options that respond to key ocean 17 and coastal risks (Section 3.4) focus on individuals, livelihoods, and economic sectors that benefit from 18 ocean and coastal ecosystem services (Section 3.5). AR5 concluded that local adaptation measures would not 19 alone be enough to offset global effects of increased climate change on marine and coastal ecosystems, and 20 that mitigation of emissions would also be necessary (high confidence) (Pörtner et al., 2014; Oppenheimer et 21 al., 2019, Table 3.27). SROCC assessed that ecosystem-based adaptation, including MPAs (high confidence) 22 (Bindoff et al., 2019) and adaptive management are effective to reduce climate change impacts (IPCC, 2018; 23 IPCC, 2019b), but that existing marine governance is insufficient to provide an effective adaptation response 24 in the marine ecosystem (high confidence) (IPCC, 2019c). Do Not Cite, Quote or Distribute 3-116 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 3 Table 3.27: Conclusions from previous IPCC assessments about implemented adaptation, enablers and limits and 4 contribution to SDGs. AR5 SR15 SROCC Degree of Implementation The analysis and Adaptation (to SLR) is A diversity of adaptation (Section 3.6.3.1) implementation of coastal already happening (high responses to coastal impacts adaptation toward climate- confidence) and will remain and risks have been resilient and sustainable important over implemented around the coasts has progressed more multi-centennial time scales world, but mostly as a significantly in developed (Hoegh-Guldberg et al., reaction to current coastal countries than in developing 2018a). risk or experienced disasters countries (high confidence) (high confidence) (Wong et al., 2014). (Oppenheimer et al., 2019). Conservation and With continuing climate Existing and restored natural Ecosystem restoration may Restoration (Section 3.6.3.2) change, local adaptation coastal ecosystems may be be able to locally reduce measures (such as effective in reducing the climate risks (medium conservation) or a reduction adverse impacts of rising sea confidence) but at relatively in human activities (such as levels and intensifying high cost and effectiveness fishing) may not sufficiently storms by protecting coastal limited to low emissions offset global-scale effects on and deltaic regions (medium scenarios and to less marine ecosystems (high confidence) sensitive ecosystems (high confidence) (Hoegh-Guldberg et al., confidence) (Pörtner et al., 2014). 2018a). (Bindoff et al., 2019). Enablers, Barriers and Adaptation strategies for Lower rates of change There are a broad range of Limits of Adaptation ocean regions beyond (Section 3.6.3.3) coastal waters are generally [associated with a 1.5ºC identified barriers and limits poorly developed but will benefit from international temperature increase] for adaptation to climate legislation and expert networks, as well as marine enhance the ability of natural change in ecosystems and spatial planning (high agreement) and human systems to adapt, human systems (high (Hoegh-Guldberg et al., 2014). with substantial benefits for confidence). Limitations a wide range of terrestrial, include [...] availability of freshwater, wetland, coastal technology, knowledge and and ocean ecosystems financial support and (including coral reefs) (high existing governance confidence) structures (medium (Hoegh-Guldberg et al., confidence). 2018a). (Bindoff et al., 2019). Existing ocean governance structures are already facing multi-dimensional, scale- related challenges because of climate change [...] (high confidence) (Bindoff et al., 2019). SDGs and Other Policy Overall, there is a strong Adaptation strategies can Achieving [the SDGs] and Frameworks (Section 3.6.4) charting Climate Resilient need to develop ecosystem- result in trade-offs with and Development Pathways depends in part on ambitious based monitoring and among the SDGs (medium and sustained mitigation efforts to contain SLR adaptation strategies to evidence, high agreement) coupled with effective adaptation actions to reduce mitigate rapidly growing (Roy et al., 2018). SLR impacts and risk (medium evidence, high risks and uncertainties to the agreement) (Oppenheimer et al., 2019). coastal and oceanic industries, communities, and nations (high agreement) (Hoegh-Guldberg et al., 2014). 5 6 7 This section builds on the SROCC assessment of the portfolio of available solutions, their applicability, and 8 their effectiveness in reducing climate change-induced risks to ocean and coastal ecosystems. Section 3.6.2 9 assesses the set of planned adaptation measures. Section 3.6.3 assesses implementation of adaptation Do Not Cite, Quote or Distribute 3-117 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 solutions and the enablers, barriers, and limitations that affect their feasibility. Section 3.6.4 evaluates the 2 contribution of planned adaptation to the Sustainable Development Goals (SDGs) and other policy-relevant 3 frameworks, and Section 3.6.5 synthesises emerging evidence about best practices. 4 5 3.6.2 Adaptation Solutions 6 7 Adaptation in ocean and coastal ecosystems continues to be informed primarily by theory, as there is still 8 limited evidence about implemented solutions (high agreement) (Seddon et al., 2020) and their success 9 across regions, especially in low-income nations (Chausson et al., 2020). Adapting to climate change 10 depends on society's ability and willingness to anticipate the change, recognise its effects, plan to 11 accommodate its consequences (Ling and Hobday, 2019; Wilson et al., 2020b), and implement a coordinated 12 portfolio of informed solutions. Here, the complete portfolio of adaptation solutions is assessed using the 13 taxonomy of Abram et al. (2019): (1) socio-institutional adaptation, (2) built infrastructure and technology, 14 and (3) marine and coastal nature-based solutions (NbS) (Figure 3.23). 15 16 17 18 Figure 3.23: Adaptation solutions for ocean and coastal ecosystems that address climate-change risk in different ocean 19 ecosystems, communities and economic sectors. Box-outline weights indicate confidence in the solution's potential to 20 reduce mid-term risks (based on the amount of evidence and agreement supporting the solutions, see SM3.5.1 for full 21 assessment). The feasibility and effectiveness of each solution (low, medium or high) indicates its ability to support 22 ecosystems and societies as they adapt to climate change impacts, based on Table SM3.3. Do Not Cite, Quote or Distribute 3-118 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 3 3.6.2.1 Socio-Institutional Adaptation 4 5 Increasing evidence shows that an effective solution portfolio includes social and institutional adaptation 6 (Figure 3.23, top; Table 3.28). Social adaptation to climate change is already occurring, as people use 7 strategies ranging from accommodating change, to coping, adapting and transforming their livelihoods (Béné 8 and Doyen, 2018; Fedele et al., 2019; Galappaththi et al., 2019; Barnes et al., 2020; Ojea et al., 2020; Green 9 et al., 2021c). Although management and institutions have major roles in adaptation (Gaines et al., 2018; 10 Barange, 2019), marine governance is impeded by increasing numbers of often-competing users and uses 11 (Boyes and Elliott, 2014); sector-led, fragmented, efforts (Nunan et al., 2020); and a legal framework less 12 clear than those on land (Crespo et al., 2019; Guggisberg, 2019). Future social responses depend on warming 13 levels and on the institutional, socio-economic and cultural constructs that allow or limit livelihood changes 14 (medium confidence) (Chapter 18, Galappaththi et al., 2019; Ford et al., 2020; Green et al., 2021c). Both 15 social and institutional transformations are needed to change the structures of power, culture, politics and/or 16 identity associated with marine ecosystems (Section 1.5.2, Wilson et al., 2020b). Ideally, institutional and 17 social adaptation will work together to sustain knowledge systems and education, enhance participation and 18 social inclusion, facilitate livelihood support and transformational change of dependent coastal communities, 19 provide economic and financial instruments, and include polycentric and multi-level governance of 20 transboundary management (Fedele et al., 2019; Fulton et al., 2019). 21 22 23 Table 3.28: Assessment of socio-institutional adaptation solutions to reduce mid-term climate impacts in oceans and 24 coastal ecosystems. Confidence is assessed in SM3.5.1. Feasibility and effectiveness are assessed in Figure 3.24. Solution Confidence in solution Contribution to Selected references Examples of (mid-term potential) adaptation implementation Knowledge diversity High confidence Consideration of IK (Norström et al., 2020; Ecotourism (Section and LK systems is Petzold et al., 2020; 3.6.3.1.3), beneficial to Gianelli et al., 2021; conservation (Section communities, increases Schlingmann et al., 3.6.3.2.1) their resilience, and is 2021) relevant and transferable beyond the local scale. Socially inclusive High confidence Policies that promote (Brodie Rudolph et al., Finance (Section policies participation of a 2020; Ford et al., 3.6.3.4.2) diversity of groups are 2020; Friedman et al., able to address 2020) existing vulnerabilities in coastal communities, and promote adaptation and transformational change. Participation Medium confidence Participation in (Brodie Rudolph et al., Fisheries and decision making and 2020; Claudet et al., mariculture (Section adaptation processes is 2020a; Hügel and 3.6.3.1.2), Indigenous recommended across a Davies, 2020; Sumaila Peoples (Section range of different et al., 2021) 3.6.3.4.1). hazards and contexts and has the potential to improve adaptation outcomes. Livelihood Medium confidence Livelihood (Blanchard et al., Fisheries and diversification diversification in 2017; Cinner and mariculture (Section communities Barnes, 2019; 3.6.3.1.2), coastal dependent on marine Mohamed Shaffril et communities (Cross- and coastal ecosystems al., 2020; Owen, 2020; Chapter Box SLR in reduces climate risks Pinsky, 2021; Taylor Chapter 3), tourism and confers flexibility et al., 2021) (Section 3.6.3.1.3) Do Not Cite, Quote or Distribute 3-119 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report Mobility to individuals, which is Migration key to adaptive capacity. Finance and market mechanisms Medium confidence When individuals are (Barnett and Fisheries and Low confidence given the choice about McMichael, 2018) mariculture (Section Disaster response mobility, they may 3.6.3.1.2) programs High confidence elect to use this High confidence response to minimise Multi-level ocean High confidence climate risks and governance Medium confidence benefit their livelihoods. Institutional transboundary Migration often (Maharjan et al., 2020; Coastal livelihoods agreements involves different Biswas and Mallick, (Section 3.6.3.1.1) spatio-temporal scales 2021; Zickgraf, 2021) than mobility. Migration could be considered an adaptation solution for some coastal and island populations in the cases of extreme events, but also as a response to more gradual changes (e.g., coastal erosion from sea-level rise (SLR)). Financial mechanisms (Shaffril et al., 2017; Economic dimensions and credit provision Dunstan et al., 2018; (Section 3.6.3.4.2) for marine-dependent Hinkel et al., 2018; livelihoods are Moser et al., 2019; effective for Sainz et al., 2019; overcoming impacts Woodruff et al., 2020) from SLR, extreme events and other climate-impact drivers. Disaster response (Nurhidayah and Climate services (Section 3.6.3.4.3), programs confer McIlgorm, 2019) tourism cruise ship sector (Section resilience to 3.6.3.1.3) communities and contribute to adaptation, when designed to be inclusive, participatory and adaptive. The multi-scale nature (Miller et al., 2018; Policy frameworks of ocean and coastal Gilfillan, 2019; (Section 3.6.4.3) climate-change risk Holsman et al., 2019; demands adaptation Obura et al., 2021). solutions at multiple levels of governance that consider the objectives and perceptions of all stakeholders to support local implementation of broad strategies. Institutional (Engler, 2020; Mason Fisheries (Section agreements for the et al., 2020; Oremus et 3.6.3.1.2, Cross- management of al., 2020; Melbourne- Chapter Box transboundary marine Thomas et al., 2021) MOVING SPECIES in resources are key for a Chapter 5) Do Not Cite, Quote or Distribute 3-120 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report sustainable future given current impacts on marine species distribution due to climate change 1 2 3 3.6.2.2 Built Infrastructure and Technology 4 5 Engineering and technology support marine and coastal adaptation (Table 3.29). Built infrastructure includes 6 engineered solutions that protect, accommodate or relocate coastal assets using hard engineering, like 7 seawalls, and soft engineering, such as beach and shore nourishment (Cross-Chapter Box SLR in Chapter 3). 8 Technological tools include early-warning systems for extreme events (Bindoff et al., 2019; Collins et al., 9 2019a), improved forecast and hindcast models (Winter et al., 2020; Davidson et al., 2021; Spillman and 10 Smith, 2021) and environmental monitoring (Claudet et al., 2020a; Wilson et al., 2020a; Melbourne-Thomas 11 et al., 2021) that support informed decision making (Tommasi et al., 2017; Rilov et al., 2020; A. Maureaud 12 et al., 2021). Emerging adaptation technologies, such as habitat development, active restoration and assisted 13 evolution (Boström-Einarsson et al., 2020; Kleypas et al., 2021), intend to accelerate recovery of damaged 14 ecosystems and promote ecological adaptation to climate change (Jones et al., 2018a; Boström-Einarsson et 15 al., 2020; Kleypas et al., 2021). 16 17 18 Table 3.29: Assessment of built infrastructure and technology solutions to reduce mid-term climate impacts in oceans 19 and coastal ecosystems. Confidence is assessed in SM3.5.1. Feasibility and effectiveness are assessed in Figure 3.24. Solution Confidence in solution Contribution to Selected references Examples of (mid-term potential) adaptation implementation Accommodation and High confidence Asset modification and (Hanson and Nicholls, Cross-Chapter Box relocation relocation of 2020; Monios and SLR in Chapter 3, livelihoods to adapt to Wilmsmeier, 2020; coastal development sea-level rise (SLR), Zickgraf, 2021) (Section 3.6.3.1.1). extreme events and coastal erosion. Protection and beach Medium confidence Protection of coastal (Pinto et al., 2020; de Cross-Chapter Box and shore nourishment ecosystems with Schipper et al., 2021; SLR in Chapter 3, interventions such as Elko et al., 2021) coastal development beach and shore (Section 3.6.3.1.1). nourishment is a common response to beach erosion around the world, and an alternative to hard protection structures such as seawalls. Early-warning systems High confidence Early-warning systems (Bindoff et al., 2019; Coastal development (Section 3.6.3.1.1), can support decision- Collins et al., 2019a; human health (Section 3.6.3.1.5). making, limit Winter et al., 2020; economic losses from Neußner, 2021) extreme events, and aid in the enterprise and development of adaptive management systems. Seasonal and dynamic High confidence The proliferation of (Payne et al., 2017; Fisheries and forecasts real-time and seasonal Hazen et al., 2018; mariculture (Section forecasts of Fernández-Montblanc 3.6.3.1.2), MPAs temperature extremes, et al., 2019; Holbrook (Section 3.6.3.2.1), marine heatwaves et al., 2020; Winter et climate services (MHWs), storm al., 2020; Bever et al., (Section 3.6.3.2.4). surges, harmful algal 2021; Davidson et al., blooms (HABs), and Do Not Cite, Quote or Distribute 3-121 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report the distribution of 2021; Spillman and living marine Smith, 2021) resources greatly contribute to adaptation through monitoring, early- warning systems, adaptive management and ecosystem-based management. Monitoring systems Medium confidence Monitoring systems (Nichols et al., 2019; MPAs (Section 3.6.3.2.1), climate that address climate- Claudet et al., 2020a; services (Section 3.6.3.2.4), fisheries impact drivers, Wilson et al., 2020a) (Section 3.6.3.1.2). ecosystem impacts and social vulnerabilities in marine social- ecological systems are key for adaptation. Habitat development Low confidence Accelerates the (Jones et al., 2018a; Restoration (Section 3.6.3.2.2). recovery of damaged Boström-Einarsson et ecosystems and al., 2020; Kleypas et promotes ecological or al., 2021) biological adaptation to future climate change. Active restoration High confidence Reintroduces species (Boström-Einarsson et Restoration (3.6.3.2.2). or augments existing al., 2020; Rinkevich, populations. For 2021) example, propagating and transplanting heat- tolerant coral species. Assisted evolution High confidence Manipulates species' (Bulleri et al., 2018; Restoration (Section genes to accelerate National Academies of 3.6.3.2.2). natural selection. Sciences, 2019; Morris et al., 2020c) 1 2 3 3.6.2.3 Marine and Coastal Nature-Based Solutions (NbS) 4 5 The ocean and coastal adaptation portfolio (Figure 3.23) also includes marine and coastal NbS (Table 3.30). 6 NbS that contribute to climate adaptation, also known as ecosystem-based adaptations (EBA), are cross- 7 cutting actions that harness ecosystem functions to restore, protect, and sustainably manage marine 8 ecosystems facing climate change impacts, while also benefiting social systems and human security 9 (Abelson et al., 2015; Barkdull and Harris, 2019) and supporting biodiversity (high confidence) (Annex II: 10 Glossary, Cross-Chapter Box NATURAL in Chapter 2, Seddon et al., 2021). NbS are expected to contribute 11 to global adaptation and mitigation goals (high confidence) (Beck et al., 2018; Cooley et al., 2019; Hoegh- 12 Guldberg et al., 2019b; Menéndez et al., 2020; Morris et al., 2020a) by protecting coastal environments from 13 SLR and storms (Cross-Chapter Box SLR, Reguero et al., 2018), and by storing substantial quantities of 14 carbon (Sections 3.4.2.5, 3.6.3.1.5, WGIII AR6 Chapter 7, Howard et al., 2017; Chow, 2018; Smale et al., 15 2018; Singh et al., 2019b; Soper et al., 2019). Marine NbS are cost-effective, can generate social, economic 16 and cultural co-benefits and can contribute to the conservation of biodiversity in the near- to mid-term (high 17 confidence) (Secretariat of the Convention on Biological Diversity, 2009; Gattuso et al., 2018; Barkdull and 18 Harris, 2019; McLeod et al., 2019). 19 20 21 Table 3.30: Assessment of marine and coastal nature-based solutions to reduce mid-term climate impacts in oceans and 22 coastal ecosystems. Confidence is assessed in SM3.5.1. Feasibility and effectiveness are assessed in Figure 3.24. Solution Confidence in solution Contribution to Selected references Examples of (mid-term potential) adaptation implementation Do Not Cite, Quote or Distribute 3-122 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report Habitat restoration High confidence Marine habitat (Colls et al., 2009; Restoration (Section 3.6.3.2.2) restoration increases Arkema et al., 2017; biodiversity and Espeland and protects shorelines and Kettenring, 2018; coastal livelihoods McLeod et al., 2019) from climate oceanic hazards. Marine protected areas High confidence MPAs and MPA (Section 3.4.3.3.4, Conservation (Section (MPAs) and other networks that are Queirós et al., 2016; 3.6.3.2.1) effective area-based carefully designed to Roberts et al., 2017; conservation measures address climate Maxwell et al., 2020a; (OECMs) change, strategically Arafeh-Dalmau et al., placed, and well 2021; Gurney et al., enforced, hold great 2021; Sala et al., 2021) potential to deliver adaptation outcomes. OECMs can confer climate resilience to dependent communities outside of MPAs. Conservation of Medium confidence Protecting areas that (Section 3.4.3.3.4, Conservation (Section climate refugia retain climate and Cross-Chapter Box 3.6.3.2.1) biodiversity conditions MOVING SPECIES in for longer durations Chapter 5, Rilov et al., under climate-change 2020; Wilson et al., can increase the 2020a; Arafeh-Dalmau resilience of marine et al., 2021) ecosystems to warming and marine heatwaves (MHWs) and facilitate marine species range shifts. Transboundary marine Low confidence Transboundary MSP (Rosendo et al., 2018; Tourism (Section spatial planning (MSP) and integrated coastal and ICZM that Tittensor et al., 2019; 3.6.3.1.3), zone management (ICZM) incorporate climate- Frazão Santos et al., conservation, (Section Sustainable harvesting High confidence change impacts and 2020; Rilov et al., 3.6.3.2.1.) adaptation in their 2020; Pinsky et al., design can support 2021) climate adaptation and to foster international ocean cooperation. Sustainable harvesting (Gattuso et al., 2018; Fisheries and mariculture (Section is a nature-based Burden and Fujita, 3.6.3.1.2) solution (NbS) that 2019; Duarte et al., contributes to 2020) adaptation by safeguarding the provision of marine food and cultural services, while reducing the ecological vulnerability of marine ecosystems. Climate-adaptive High confidence Incorporating climate- (Cross-Chapter Box Fisheries and management adaptive management MOVING SPECIES in mariculture (Section allows climate Chapter 5, Rilov et al., 3.6.3.1.2), knowledge and 2019; Free et al., 2020; conservation, (Section information available Wilson et al., 2020a; 3.6.3.2.1), restoration for the system to be Melbourne-Thomas et (Section 3.6.3.2.2) iteratively updated in al., 2021) Do Not Cite, Quote or Distribute 3-123 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report the management plan. It also facilitates consideration of species distribution shifts and other climate-change responses. Ecosystem-based High confidence EbM focuses on (Fernandino et al., Fisheries and mariculture (Section management (EbM) ecosystems. By 2018; Lowerre- 3.6.3.1.2) incorporating many of Barbieri et al., 2019) the above tools, ecosystem-based adaptation (EbA) benefits adaptation of marine ecosystems and supports provision of ecosystem services to people. 1 2 3 3.6.3 Implementation and Effectiveness of Adaptation and Mitigation Measures 4 5 This section assesses implemented adaptations introduced in Section 3.6.2 for selected marine sectors 6 (Section 3.6.3.1) and ecosystems (Section 3.6.3.2), using case studies to emphasise characteristics that enable 7 or inhibit adaptation (Section 3.6.3.3). The feasibility and effectiveness of these adaptations is assessed in 8 Figure 3.24. 9 10 3.6.3.1 Degree of Implementation and Evidence of Effectiveness Across Sectors 11 12 3.6.3.1.1 Coastal community development and settlement 13 14 Coastal adaptation often addresses the risk of flooding and erosion from SLR, changes in storm activity and 15 degradation of coastal ecosystems and their services (high confidence) (Sections 3.4.2, 3.5, Oppenheimer et 16 al., 2019). Without coastal protection, people and property will be increasingly exposed to coastal flooding 17 after 2050, especially under RCP8.5 (Cross-Chapter Box SLR in Chapter 3, Bevacqua et al., 2020; Kirezci et 18 al., 2020). This section assesses adaptation responses for coastal ecosystems, addressing loss of natural 19 coastal protection (Sections 3.4.2.1, 3.4.2.4­3.4.2.6), and the need for relocation (Section 3.6.2.1.2). 20 Adaptation responses specific to SLR are assessed in detail in Cross-Chapter Box SLR in Chapter 3, while 21 adaptation in coastal cities and settlements is assessed in Chapter 6. 22 23 Coastal conservation tends to involve cost-effective, low-impact actions that aim to support both adaptation 24 and mitigation by conserving a wide array of ecosystem functions and services (Gattuso et al., 2018; Gattuso 25 et al., 2021), and that are achievable by nations with extensive coastlines or low-income status (Herr et al., 26 2017; Taillardat et al., 2018). Where coastlines are undeveloped, the lowest-risk option is to avoid new 27 development, but elsewhere, coastal conservation includes protection of key assets, accommodation of SLR, 28 advancing defences seawards or upwards, or planned retreat from the coast (Cross-Chapter Box SLR in 29 Chapter 3). 30 31 Hard engineered structures like seawalls are generally more costly than nature-based adaptations (high 32 confidence) (Hérivaux et al., 2018; Haasnoot et al., 2019; Nicholls et al., 2019; Oppenheimer et al., 2019) 33 and can lock communities into engineered responses in the future (Cross-Chapter Box SLR in Chapter 3), 34 creating trade-offs with mitigation goals, which constitutes maladaptation (Nunn et al., 2021) that carries 35 ecological and cultural costs (Sections 3.4.2.4, 3.4.2.6, 3.5.6). As a result there is high agreement on the 36 importance of shifting from hard infrastructure to soft infrastructure for coastal defence (Toimil et al., 2020; 37 Nunn et al., 2021). The common remedy for beach erosion is beach nourishment (Oppenheimer et al., 2019; 38 Pinto et al., 2020; Elko et al., 2021), which provides rapid results but poorly quantified trade-offs between 39 efficacy, long-term cost, utility to beach users and ecological damage (de Schipper et al., 2021). 40 Do Not Cite, Quote or Distribute 3-124 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Since SROCC, coastal adaptation using NbS, like restoration of coastal vegetation, has advanced 2 substantially (Cohen-Shacham et al., 2019; Kuhl et al., 2020; Kumar et al., 2020; Morris et al., 2020a). Field 3 and modelling studies confirm that wetland restoration and preservation are key actions to restore coastal 4 protection and reduce community vulnerability to flooding (very high confidence) (see also Section 3.6, 5 Chapter 15, Cross-Chapter Box SLR in Chapter 3, Jones et al., 2020; Menéndez et al., 2020; Van Coppenolle 6 and Temmerman, 2020), while maintaining coastal ecosystem services (Section 3.5). Restoring coral reefs, 7 oyster reefs and mangroves (Section 3.6.2.1), and protecting macrophyte meadows, dissipate wave energy 8 (Section 3.4.2.1, Yates et al., 2017; Beck et al., 2018; Wiberg et al., 2019; Menéndez et al., 2020), accrete 9 sediment, and elevate shorelines, which reduces exposure to waves and storm surges, and offsets erosional 10 losses (medium confidence) (Kench and Mann, 2017; Pomeroy et al., 2018; Dasgupta et al., 2019; James et 11 al., 2019; Morris et al., 2019; David and Schlurmann, 2020; Masselink et al., 2020). However, irreversible 12 regime shifts in ocean ecosystems due to SLR and extreme events such as MHW can limit or compromise 13 restoration in the long term (high confidence) (Section 3.4.3.3.3, Cross-Chapter Box SLR in Chapter 3, 14 Marzloff et al., 2016; Johnson et al., 2017a). Under all warming scenarios, coastal wetlands will be impacted 15 by warming and MHWs (Sections 3.2.2.1, 3.2.4.5, Cross-Chapter Box 9.1 in WGI Chapter 9, Fox-Kemper et 16 al., 2021), while also being pressed inland by RSLR (Section 3.4.2.5, Cross-Chapter Box SLR in Chapter 3). 17 Therefore, restoration and conservation are more successful when non-climate drivers are also minimised 18 (high confidence) (Brodie et al., 2020; Duarte et al., 2020; Liu et al., 2021). 19 20 For highly exposed human settlements, migration is an adaptation option (e.g., for some island populations 21 under extreme circumstances), but there are important uncertainties (Section 15.3.4.6), as international 22 regimes develop around human rights, migration (Scobie, 2019a), displacement (George Puthucherril, 2012) 23 and the implications for national sovereignty (Yamamoto and Esteban, 2014) of disappearing land spaces 24 caused by climate change. Colonial power dynamics can influence climate change responses (Chapter 18), 25 for example when external funders favour migration over local desires to adapt in place to preserve national 26 identity and sovereignty (Bordner et al., 2020). Examples of relocation within livelihoods' customary land 27 show some successes (Section 15.3.4.6). 28 29 Evidence since SROCC (Section, 5.5.2.3.3, Bindoff et al., 2019) continues to show that built infrastructure 30 cannot address all of the adaptation challenges that coastal communities face. Coastal squeeze creates 31 tensions between coastal development, armouring, and habitat management (Sections 3.4.2.4­3.4.2.6). 32 Managed realignment is the best option to reduce risks from SLR (high confidence) (Cross-Chapter Box 33 SLR in Chapter 3) but requires transformative changes in coastal development and settlement (Felipe Pérez 34 and Tomaselli, 2021; Fitton et al., 2021; Mach and Siders, 2021; Siders et al., 2021). Implementation of 35 protective measures varies among nations and lack of financial resources limits the options available (very 36 high confidence) (Cross-Chapter Box SLR in Chapter 3, Hinkel et al., 2018; Klöck and Nunn, 2019). 37 38 39 [START CROSS-CHAPTER BOX SLR HERE] 40 41 Cross-Chapter Box SLR: Sea Level Rise 42 43 Authors: Gonéri Le Cozannet (France, Chapter 13, CCP4), Judy Lawrence (New Zealand, Chapter 11), 44 David Schoeman (Australia, Chapter 3), Ibidun Adelekan (Nigeria, Chapter 9), Sarah Cooley (USA, Chapter 45 3), Bruce Glavovic (New Zealand/South Africa, Chapter 18, CCP2), Marjolijn Haasnoot (The Netherlands, 46 Chapter 13, CCP2), Rebecca Harris (Australia, Chapter 2), Wolfgang Kiessling (Germany, Chapter 3), 47 Robert E. Kopp (USA, WGI), Aditi Mukherji (Nepal, Chapter 4), Patrick Nunn (Australia, Chapter 15), 48 Dieter Piepenburg (Germany, Chapter 13), Daniela Schmidt (UK/Germany, Chapter 13), Craig T. Simmons 49 (Australia), Chandni Singh (India, Chapter 10, CCP2), Aimée Slangen (The Netherlands, WGI), Seree 50 Supratid (Thailand, Chapter 4). 51 52 53 Sea-level rise (SLR) is already impacting ecosystems, human livelihoods, infrastructure, food security and 54 climate mitigation at the coast and beyond. Ultimately, it threatens the existence of cities and settlements in 55 low lying areas, and some island nations and their cultural heritage (Chapters 9­ 15, Cross-Chapter Paper 2 56 and 4 Oppenheimer et al., 2019). The challenge can be addressed by mitigation of climate change and coastal 57 adaptation. Do Not Cite, Quote or Distribute 3-125 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Current impacts of sea level rise 3 The rate of global mean SLR was 1.35 mm yr­1 (0.78­1.92 mm yr­1, very likely range) during 1901­1990, 4 faster than during any century in at least 3000 years (high confidence) (WGI AR6 Chapter 9, Stanley and 5 Warne, 1994; Woodroffe et al., 2016; Fox-Kemper et al., 2021). Global mean SLR has accelerated to 3.25 6 mm yr­1 (2.88­3.61 mm yr­1, very likely range) during 1993­2018 (high confidence). Extreme sea levels 7 have increased consistently across most regions (WGI AR6 Chapter 9, Fox-Kemper et al., 2021). The largest 8 observed changes in coastal ecosystems are being caused by the concurrence of human activities, waves, 9 current-induced sediment transport, and extreme storm events (medium confidence) (Chapters 3, 15 and 16, 10 Takayabu et al., 2015; Mentaschi et al., 2018; Duvat, 2019; Murray et al., 2019; Oppenheimer et al., 2019). 11 Early impacts of accelerating SLR detected at sheltered or subsiding coasts include chronic flooding at high 12 tides, wetland salinisation and ecosystem transitions, increased erosion and coastal flood damage (Chapters 13 3, 11 and 13­16, WGI AR6 Chapter 9, Sweet and Park, 2014; Moftakhari et al., 2015; Nunn et al., 2017; 14 Oppenheimer et al., 2019; Sharples et al., 2020; Fox-Kemper et al., 2021; Strauss et al., 2021). The exposure 15 of many coastal populations and ecosystems to SLR is high: economic development is disproportionately 16 concentrated in and around coastal cities and settlements (virtually certain) (Chapters 3 and 9­15, Cross- 17 Chapter Papers 2 and 4). 18 19 Projected risks to coastal communities, infrastructure and ecosystems 20 Risks from SLR are very likely to increase by one order of magnitude well before 2100 without adaptation 21 and mitigation action as agreed by parties to the Paris Agreement (very high confidence). Global mean SLR 22 is likely to continue accelerating under SSP1-2.6 and more strongly forced scenarios (Figure SLR1, WGI 23 AR6 Chapter 9, Oppenheimer et al., 2019; Fox-Kemper et al., 2021), increasing the risk of chronic coastal 24 flooding at high-tide, extreme flooding during extreme events such as swells, storms and hurricanes, and 25 erosion, and coastal ecosystem losses across many low-lying and erodible coasts (very high confidence) 26 (Chapters 3 and 9­15, Cross-Chapter Paper 2, Hinkel et al., 2014; McLachlan and Defeo, 2018; Kulp and 27 Strauss, 2019; Vousdoukas et al., 2020b). The compounding of rainfall, river flooding, rising water tables, 28 coastal surges, and waves are projected to exacerbate SLR impacts on low-lying areas and rivers further 29 inland (Chapters 4 and 11­15) (Bevacqua et al., 2020). 30 31 There is high confidence that coastal risks will increase by at least one order of magnitude over the 21st 32 century due to committed SLR (Hinkel et al., 2013; Hinkel et al., 2014; IPCC, 2019b). Exposure of 33 population and economic assets to coastal hazards is projected to increase over the next decades, particularly 34 in coastal regions with fast-growing populations in Africa, Southeast Asia, and Small Islands (medium 35 evidence) (Chapters 9­15, Cross-Chapter Papers 2 and 4, Neumann et al., 2015; Jones and O'Neill, 2016; 36 Merkens et al., 2016; Merkens et al., 2018; Rasmussen et al., 2020). For RCP8.5, 2.5­9% of the global 37 population and 12­20% of the global gross domestic product (GDP) is projected to be exposed to coastal 38 flooding by 2100 (Kulp and Strauss, 2019; Kirezci et al., 2020; Rohmer et al., 2021). Above 3°C global 39 warming levels (GWL) and with low adaptation, SLR may cause disruptions to ports and coastal 40 infrastructure (Camus et al., 2019; Christodoulou et al., 2019; Verschuur et al., 2020; Yesudian and Dawson, 41 2021), which in turn may cascade and amplify across sectors and regions, generating impacts to financial 42 systems (Chapters 11, 13, Mandel et al., 2021). Depending on the hydrogeological context, SLR causes 43 salinisation of groundwater, estuaries, wetlands and soils, adding constraints to water management and 44 livelihoods in agriculture sectors, for example, in deltas (Chapters 9, 15, Cross-Chapter Paper 4, 45 Oppenheimer et al., 2019; Nicholls et al., 2020). 46 47 Coastal ecosystems can migrate landward or grow vertically in response to SLR, but their resilience and 48 capacity to keep up with SLR will be compromised by ocean warming and other drivers, depending on 49 regions and species, e.g., above 1.5°C for coral reefs (high confidence) (Chapter 3, 16, IPCC, 2018; 50 Melbourne et al., 2018; Perry et al., 2018; IPCC, 2019b; Cornwall et al., 2021). Sediments and space for 51 landward retreat are crucial for mangroves, saltmarshes and beach ecosystems (high confidence) (Chapter 3, 52 Peteet et al., 2018; Schuerch et al., 2018; FitzGerald and Hughes, 2019; Friess et al., 2019; Leo et al., 2019; 53 Schuerch et al., 2019). Loss of habitat is accompanied by loss of associated ecosystem services, including 54 wave-energy attenuation, habitat provision for biodiversity, food production and carbon storage (Chapter 3, 55 Cross-Chapter Box NATURAL in Chapter 2). 56 Do Not Cite, Quote or Distribute 3-126 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Under a high-emissions, low-likelihood/high-impact scenario, where low confidence ice-sheet mass loss 2 occurs, global mean SLR could exceed the likely range by more than one additional metre in 2100 (Figure 3 SLR1b, Cross-Chapter Box DEEP in Chapter 17, WGI AR6 Technical Summary and Chapter 9, Arias et al., 4 2021; Fox-Kemper et al., 2021). This is a reason for concern given that rapid SLR after the last glacial- 5 interglacial transition caused a drowning of coral reefs (high confidence) (Camoin and Webster, 2015; 6 Sanborn et al., 2017; Webster et al., 2018), extensive loss of coastal land and islands, habitats and associated 7 biodiversity (high confidence) (AR6 WGI Chapter 9, Fruergaard et al., 2015; Fernández-Palacios et al., 8 2016; Hamilton et al., 2019; Helfensdorfer et al., 2019; Kane and Fletcher, 2020; Fox-Kemper et al., 2021) 9 and triggered Neolithic migrations in Europe and Australia (medium confidence) (Cross-Chapter Box 10 PALEO in Chapter 1, Turney and Brown, 2007; Brisset et al., 2018; Williams et al., 2018). 11 12 At centennial timescales, projected SLR represents an existential threat for island nations, low-lying coastal 13 zones, and the communities, infrastructure, and cultural heritage therein (Chapters 9­15, Cross-Chapter 14 Paper 4). Even if climate warming is stabilised at 2°C to 2.5°C GWL, coastlines will continue to reshape 15 over millennia, affecting at least 25 megacities and drowning low-lying areas where 0.6­1.3 billion people 16 lived in 2010 (medium confidence) (WGI AR6 Chapter 9 Marzeion and Levermann, 2014; Clark et al., 2016; 17 Kulp and Strauss, 2019; Fox-Kemper et al., 2021; Strauss et al., 2021). 18 19 Solutions, opportunities and limits to adaptation 20 The ability to adapt to current coastal impacts, to cope with future coastal risks, and to prevent further 21 acceleration of SLR beyond 2050 depends on immediate mitigation and adaptation actions (very high 22 confidence). The most urgent adaptation challenge is chronic flooding at high tide (Chapters 10, 11, and 13­ 23 15). Reducing the acceleration of SLR beyond 2050 will only be achieved with fast and profound mitigation 24 of climate change (Nicholls et al., 2018; Oppenheimer et al., 2019). Until 2050, adaptation planning and 25 implementation needs are projected to increase significantly in most inhabited coastal regions (Figure SLR1, 26 WGI AR6 Chapter 9, IPCC, 2019b; Fox-Kemper et al., 2021). For SSP1-2.6 and more strongly forced 27 scenarios, SLR rates continue to increase (WGI AR6 Chapter 9, Fox-Kemper et al., 2021), and so do the 28 scale and the frequency of adaptation interventions needed in coastal zones (Figure SLR1, Haasnoot et al., 29 2020). 30 31 Risks can be anticipated, planned, and decided upon, and adaptation interventions can be implemented over 32 the coming decades, considering their often long lead- and life-times, irrespective of the large uncertainty 33 about SLR beyond 2050 (high confidence) (Figure SLR1, Cross-Chapter Box DEEP in Chapter 17, Cross- 34 Chapter Paper 2, Chapters 11, 13, Haasnoot et al., 2018; Stephens et al., 2018; Stammer et al., 2019). 35 Adaptation capacity and governance to manage risks from projected SLR typically require decades to 36 implement and institutionalise (high confidence) (Chapters 11,13, Haasnoot et al., 2021). Without 37 considering both short- and long-term adaptation needs, including beyond 2100, communities are 38 increasingly confronted with a shrinking solution space, and adverse consequences are disproportionately 39 borne by exposed and socially vulnerable people (Chapters 1, 8). SLR is likely to compound social conflict 40 in some settings (high confidence) (Oppenheimer et al., 2019). 41 42 Coastal impacts of SLR can be avoided by preventing new development in exposed coastal locations 43 (Chapters 3, 9­15, Cross-Chapter Paper 2, Doberstein et al., 2019; Oppenheimer et al., 2019). For existing 44 developments, a range of near-term adaptation options exists, including (1) engineered, sediment or 45 ecosystem-based protection; (2) accommodation and land use planning, to reduce the vulnerability of people 46 and infrastructure; (3) advance through, e.g., land reclamation; and (4) retreat through planned relocation or 47 displacements and migrations due to SLR (Chapters 9­15, Cross-Chapter Paper 2, Oppenheimer et al., 48 2019). Only avoidance and relocation can remove coastal risks for the coming decades, while other measures 49 only delay impacts for a time, have increasing residual risk or perpetuate risk and create ongoing legacy 50 effects and virtually certain property and ecosystem losses (high confidence) (Cross-Chapter Paper 2, Siders 51 et al., 2019). Large-scale relocation has immense cultural, political, social and economic costs, and equity 52 implications, which can be reduced by fast implementation of climate mitigation and adaptation policies 53 (Chapter 8, Cross-Chapter Paper 2, Gibbs, 2015; Haasnoot et al., 2021). While relocation may currently 54 appear socially unacceptable, economically inefficient, or technically infeasible today (Lincke and Hinkel, 55 2021), it becomes the only feasible option as protection costs become unaffordable and the limits to 56 accommodation become obvious (Chapters 11, 13, 15, Hino et al., 2017; Siders et al., 2019; Strauss et al., 57 2021). Effective responses to rising sea level involve locally applicable combinations of decision analysis, Do Not Cite, Quote or Distribute 3-127 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 land use planning, public participation and conflict resolution approaches; together these can anticipate 2 change and help to chart adaptation pathways, over time addressing the governance challenges due to rising 3 sea level (high confidence) (Oppenheimer et al., 2019). 4 5 Ecosystem-based adaptation can reduce impacts on human settlements and bring substantial co-benefits such 6 as ecosystem services restoration and carbon storage, but they require space for sediment and ecosystems 7 and have site-specific physical limits, at least above 1.5°C GWL (high confidence) (Cross-Chapter Box 8 NATURAL in Chapter 2, Chapters 3, 9, 11, 15, Herbert et al., 2015; Brown et al., 2019; Van Coppenolle and 9 Temmerman, 2019; Watanabe et al., 2019; Neijnens et al., 2021). For example, planting and conserving 10 vegetation helps sediment accumulation by dissipating wave energy and reducing impacts of storms, at least 11 at present-day sea levels (high confidence) (Temmerman et al., 2013; Narayan et al., 2016; Romañach et al., 12 2018; Laengner et al., 2019; Leo et al., 2019). Coastal wetlands and ecosystems can be preserved by 13 landward migration (Schuerch et al., 2018; Schuerch et al., 2019) or sediment supply (VanZomeren et al., 14 2018), but they can be seriously damaged by coastal defences designed to protect infrastructure (Chapters 3, 15 13, Cooper et al., 2020b). Sediment nourishment can prevent erosion, but it can also negatively impact beach 16 amenities and ecosystems through ongoing dredging, pumping and deposition of sand and silts (VanZomeren 17 et al., 2018; de Schipper et al., 2021; Harris et al., 2021). 18 19 There is increasing evidence that current governance and institutional arrangements are unable to address the 20 escalating risks in low-lying coastal areas worldwide (high confidence). Barriers to adaptation such as 21 decision-making driven by short-term thinking or vested interests, funding limitations, and inadequate 22 financial policies and insurance can be addressed equitably and sustainably through implementation of suites 23 of adaptation options and pathways, (Chapters 11, 13, 17­18, Cross-Chapter Paper 2). Improved coastal 24 adaptation governance can be supported by approaches that consider changing risks over time, such as 25 "dynamic adaptation pathways" planning (Chapters 11, 13, 18, Cross-Chapter Box DEEP in Chapter 17). 26 Integrated Coastal Zone Management and land-use and infrastructure planning are starting to consider SLR 27 by, for example, monitoring early signals (Haasnoot et al., 2018; Stephens et al., 2018; Kool et al., 2020), 28 updating sea-level projections (Stephens et al., 2017; Hinkel et al., 2019; Kopp et al., 2019; Stammer et al., 29 2019), considering uncertainties of sea-level projections and coastal impacts (e.g., Stephens et al., 2017; 30 Jevrejeva et al., 2019; Rohmer et al., 2019), as well as engaging with communities, practitioners and 31 scientists, recognising the values of current and future generations (e.g., Nicholls et al., 2014; Buchanan et 32 al., 2016b). While there is high agreement that the majority of adaptation needs are not met yet, there is 33 robust evidence of sea level rise increasingly being considered in coastal adaptation decision making and 34 being embedded in national and local guidance and regulations (Nicholls et al., 2014; Le Cozannet et al., 35 2017; Lawrence et al., 2018; Kopp et al., 2019; McEvoy et al., 2021). 36 Do Not Cite, Quote or Distribute 3-128 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure Cross-Chapter Box SLR.1: The challenge of coastal adaptation in the era of sea-level rise (SLR): (a): typical 3 timescales for the planning, implementation (grey triangles) and operational lifetime of current coastal risk-management 4 measures (blue bars); (b): global sea-level projections, which are representative of relative SLR projected for 60 to 70% 5 of global shorelines, within ±20% errors (WGI AR6 Chapter 9, Fox-Kemper et al., 2021); (c): Frequency of illustrative 6 adaptation decisions to +0.5 m of SLR under different SSP-RCP scenarios. In response to accelerated SLR, adaptation 7 either occurs earlier and faster, or accounts for higher amounts of SLR (e.g., to +1 m instead of to +0.5 m). Adaptation 8 to +0.5 m from today's sea-levels have a lifetime of 90 years for SSP1-2.6, but lifetime is reduced to 60 years for SSP5- 9 8.5 and 30 years for a high-end scenario involving low confidence processes. Adaptations to +0.5 m are comparable to 10 e.g., the Thames Barrier in the United Kingdom, or the Delta Works in the Netherlands, which primarily had an 11 intended lifetime of 100­200 years. Adaptation measures to +0.2 m may include nourishment or wetland or setback 12 zones. 13 14 15 [END CROSS-CHAPTER BOX SLR HERE] 16 17 18 3.6.3.1.2 Fisheries and mariculture 19 SROCC (Bindoff et al., 2019) assessed adaptation in fisheries and mariculture (marine aquaculture), and 20 socioeconomically focused updates are provided in Section 5.8.4 and Cross-Chapter Box MOVING 21 SPECIES in Chapter 5. Here, we present a brief synthesis of how fisheries and mariculture adaptations 22 interact with the natural environment, with further detail and supporting material in SM3.5.2. Do Not Cite, Quote or Distribute 3-129 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Mobility allows fishing fleets and fishers to adapt to shifts in marine species distributions (high agreement) 3 (Sections 3.4.3.1, 3.5.3, Peck and Pinnegar, 2018; Pinsky et al., 2018; Frazão Santos et al., 2020), but with 4 limits and unintended consequences (Pinsky and Fogarty, 2012; Bell et al., 2021). Diversification of target 5 species, harvest tactics and employment sectors, including transitions from fisheries to mariculture and 6 ecotourism, allows some fishers to accommodate some impacts on their livelihoods (Miller et al., 2018; 7 Robinson et al., 2020; Gonzalez-Mon et al., 2021). Technology and infrastructure adaptations can improve 8 marine harvest efficiency, reduce risk, and support resource management goals (Friedman et al., 2020; Bell 9 et al., 2021; Melbourne-Thomas et al., 2021), but their ability to overcome climate-change impacts remains 10 uncertain (Bell et al., 2020). Improving capacity to predict anomalous conditions in coastal and marine 11 ecosystems (Jacox et al., 2019; Holbrook et al., 2020; Jacox et al., 2020), storm-driven flooding in reef-lined 12 coasts (Scott et al., 2020; Winter et al., 2020) and fisheries stocks (Payne et al., 2017; Tommasi et al., 2017; 13 Muhling et al., 2018) can improve forecasts of coastal and marine resources. These can enhance 14 sustainability of wild-capture fisheries under climate change (high confidence) (Blanchard et al., 2017; 15 Tommasi et al., 2017; Pinsky et al., 2020a; Bell et al., 2021). Limiting overexploitation is the central goal of 16 fishery management, and it very likely benefits fisheries adaptation to climate change (Burden and Fujita, 17 2019; Free et al., 2019; Sumaila and Tai, 2020). Conventional tools include catch and size limits, spatial 18 management and adaptive management. Ecosystem-based fisheries management outperforms single-species 19 management (Fulton et al., 2019), is widely legislated (Bryndum-Buchholz et al., 2021), and can reduce 20 climate impacts in fisheries in the near-term, especially under low-emission scenarios (Karp et al., 2019; 21 Holsman et al., 2020). Transboundary agreements on shifting fisheries will reduce the risk of overharvesting 22 (medium confidence) (Gaines et al., 2018). Permits tradable across political boundaries could also address 23 this challenge, but limited evidence is available regarding their efficacy (Cross-Chapter Box MOVING 24 SPECIES in Chapter 5, Pinsky et al., 2018). Climate-smart conservation (Section 3.6.32.1) under the 25 negotiations on areas beyond national jurisdiction (ABNJ) (Pinsky et al., 2018; Tittensor et al., 2019; Frazão 26 Santos et al., 2020), and in the Convention on Biological Diversity (CBD) areas designed as other effective 27 area-based conservation measures (OECMs, Tittensor et al., 2019) provide further benefits. Despite the 28 potential for of adaptive management to achieve sustainable fisheries, outcomes will very likely be 29 inequitable (Gaines et al., 2018; Lam et al., 2020) with lower-income countries suffering the greater biomass 30 and economic losses, increasing inequalities especially under higher emission scenarios (high confidence) 31 (Boyce et al., 2020). Flexible and polycentric governance approaches have facilitated some short-term 32 successes in achieving equitable, sustainable fisheries practices, but these may be challenging to implement 33 where other governance systems, especially hierarchical systems, are well-established (Cvitanovic et al., 34 2018; Bell et al., 2020). 35 36 3.6.3.1.3 Tourism 37 Coastal areas, coastal infrastructure and beaches, sustaining tourism that contributes significantly to local 38 economies (James et al., 2019; Ruiz-Ramírez et al., 2019), are under threat from development, SLR and 39 increased wave energy during storms and (high confidence) (Sections 3.4.2.6, 3.4.4­3.4.6, 3.5.6, Lithgow et 40 al., 2019; Ruiz-Ramírez et al., 2019). Engineered solutions like seawalls and revetments have traditionally 41 been used to address coastal erosion (Section 3.6.3.1.1), but soft infrastructure approaches, including beach 42 nourishment, submerged breakwaters and groins, and NbS (Section 3.6.2.1), are becoming more common, 43 partly due to demand from the tourism industry (medium confidence) (Pranzini, 2018; Pranzini et al., 2018). 44 45 Elsewhere, interactions between tourism and climate impacts worsen outcomes for coastal and ocean 46 environments (Section 3.6.3.1.4). Climate change is opening up new cruise-ship routes in the Arctic (Sun et 47 al., 2018), increasing number of visitors and associated stressors, such as litter, to previously undisturbed 48 areas (Anfuso et al., 2020; Hovelsrud et al., 2020; Suaria et al., 2020). Risk reduction for cruise-ship tourism 49 includes disaster response management, improved mapping, and passenger codes of conduct ensuring social, 50 cultural and ecological sustainability (Stewart et al., 2015; Dawson et al., 2016). 51 52 Marine ecotourism, integrating conservation, education and provision of benefits to local communities 53 (Donohoe and Needham, 2006), can provide significant economic benefits (Wabnitz, 2019), and is among 54 the most common livelihood alternatives to support both marine conservation and climate change adaptation 55 (Kutzner, 2019; Pham, 2020; Prasetyo et al., 2020). Ecotourism can enhance social and political will for 56 marine conservation (Cisneros-Montemayor and Sumaila, 2014), and facilitates integration of local and 57 Indigenous Peoples in employment, ownership, and industry governance. The community of Cabo Pulmo, Do Not Cite, Quote or Distribute 3-130 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Mexico, self-imposed an MPA and replaced fishing with ecotourism, which now generates millions of USD 2 yr­1, sustains locally owned and operated tour companies, and has increased some fish populations ten-fold 3 (Knowlton, 2020). In Misool, Indonesia, local ecotourism incorporates IK by including local communities' 4 preferences and sustainable resource use (Prasetyo et al., 2020). 5 6 Unintended consequences of ecotourism, such as detrimental ecological impacts on reefs (Giglio et al., 7 2020), sharks, marine birds (Monti et al., 2018), and whales (Higham et al., 2016; Barra et al., 2020; Hoarau 8 et al., 2020), can be minimised by relying on evidence-based management of associated activities (Blumstein 9 et al., 2017). Public perception of climate change connections to tourism can create obstacles (Meynecke et 10 al., 2017; Atzori et al., 2018) such as deterring long-term investment in SIDS tourism initiatives (Santos- 11 Lacueva et al., 2017), or benefits like inclining tourists to participate in conservation projects (Curnock et al., 12 2019; Miller et al., 2020b; Ziegler et al., 2021). Social and cultural networks may decrease climate 13 vulnerability, as with Indigenous tourism operators in SIDS (Parsons et al., 2018). Tourism-based adaptation 14 can also be improved by equitable access to resources, and recognition and inclusion of all stakeholders 15 during policy planning and implementation. The principles of marine spatial planning (Papageorgiou, 2016) 16 provide for effectively incorporating stakeholders and could inform development of activities to assess 17 climate-associated risks (e.g., Tzoraki et al., 2018; Loehr, 2020). The recent decrease in global tourism due 18 to the COVID-19 pandemic may offer opportunities to transform existing practices to more sustainable 19 approaches (Cross-Chapter Box COVID in Chapter 7, Gössling et al., 2021). 20 21 3.6.3.1.4 Maritime transport 22 Increased maritime transport and cruise-ship tourism in the Arctic are already impacting local and 23 Indigenous Peoples, revealing conflicts over the uses of the ocean and the governance needed to support 24 local people and a sustainable blue economy (high confidence) (Debortoli et al., 2019; Palma et al., 2019; 25 Berman et al., 2020; Dundas et al., 2020). While shipping and its associated environmental impacts are 26 projected to grow (Palma et al., 2019; Dawson et al., 2020), adaptation efforts are only at the planning stage 27 (Debortoli et al., 2019). Increased Arctic traffic due to ice loss can benefit trade, transportation and tourism 28 (medium confidence), but will also affect Arctic marine ecosystems and livelihoods (high confidence) (Palma 29 et al., 2019; Dawson et al., 2020). Increasing search-and-rescue activities (Ford and Clark, 2019) reveal 30 capacity gaps to support future demands (Ford and Clark, 2019; Palma et al., 2019). The Low-Impact 31 Shipping Corridors initiative has been developed as an adaptation strategy in the Arctic, although with 32 limited inclusion of IK and LK (Dawson et al., 2020). 33 34 RSLR and the increased frequency and severity of storms are already affecting port activity, infrastructure, 35 and supply chains, sometimes disrupting trade and transport (Monios and Wilmsmeier, 2020), but these 36 hazards are not systematically incorporated into adaptation planning (medium evidence) (Monios and 37 Wilmsmeier, 2020; O'Keeffe et al., 2020). Climate-change impacts that increase food insecurity, income 38 loss, and poverty can exacerbate maritime criminal activity including illegal fishing, drug trafficking or 39 piracy (medium evidence) (Germond and Mazaris, 2019). A transformational adaptation approach to address 40 climate impacts on maritime activities and increase security (Germond and Mazaris, 2019) would relocate 41 ports, change centers of demand, reduce shipping distances, or shorten supply chains (medium agreement) 42 (Walsh et al., 2019; Monios and Wilmsmeier, 2020) as well as decrease marginalization of vulnerable 43 groups, develop polycentric governance systems and eliminate maladaptive environmental policies and 44 resource loss (Belhabib et al., 2020; O'Keeffe et al., 2020). 45 46 3.6.3.1.5 Human Health 47 Health-focused adaptations to climate-driven changes in ocean and coastal water quality (Section 3.5.5.3) 48 mainly leverage technology and infrastructure (Section 3.6.2.2) to improve water-quality monitoring and 49 forecasting to inform socio-institutional adaptation (Section 3.6.2.1) and NbS (Section 3.6.2.3). Seafood 50 quality and safety are decreasing due to climate-driven increases in marine-borne diseases (Cross-Chapter 51 Box ILLNESS in Chapter 2), toxic HABs, or toxin bioaccumulation (high agreement) (Karagas et al., 2012; 52 Krabbenhoft and Sunderland, 2013; Rafaj et al., 2013; Curtis et al., 2019; Schartup et al., 2019; Thackray 53 and Sunderland, 2019). Future exposure to seafood-borne contaminants also depends partly on consumers' 54 seafood preferences (Elsayed et al., 2020) and seafood supply (Sunderland et al., 2018). Reducing this risk 55 by decreasing seafood consumption increases risk of eating less nutritious foods, and loss of cultural 56 practices (Chapter 5, Cross-Chapter Box MOVING SPECIES in Chapter 5, Donatuto et al., 2011; Bindoff et 57 al., 2019). Models incorporating high-resolution satellite images, field survey data, meteorological Do Not Cite, Quote or Distribute 3-131 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 observations and historical records can provide early-warning forecasts of HABs or conditions that favour 2 microbial pathogen outbreaks (Cross-Chapter Box ILLNESS in Chapter 2, Semenza et al., 2017; Franks, 3 2018; Hattenrath-Lehmann et al., 2018; Borbor-Cordova et al., 2019; Davis et al., 2019; Campbell et al., 4 2020a; Davidson et al., 2021). Forecasts facilitate preventive public health measures (World Health 5 Organisation and United Nations Children's Fund, 2017), or seafood harvest guidance (Maguire et al., 2016; 6 Leadbetter et al., 2018; Anderson et al., 2019; Bolin et al., 2021), reducing risks of disease outbreaks, waste, 7 and contaminated seafood entering the market (medium confidence) (Cross-Chapter Box ILLNESS in 8 Chapter 2, Nichols et al., 2018). Monitoring of water quality and seafood safety (Cross-Chapter Box 9 ILLNESS in Chapter 2), paired with effective public communication and education (Ekstrom et al., 2020) 10 inform individual and local adaptations, including use of personal protective equipment, seafood selection 11 and preparation (Elsayed et al., 2020; Froelich and Daines, 2020; Fielding et al., 2021), income 12 diversification (Section 3.6.2.1, Moore et al., 2020b), public education (Borbor-Cordova et al., 2019), or 13 community-level actions to decrease risk from coastal aquifer and soil salinisation (Slama et al., 2020; 14 Mastrocicco and Colombani, 2021), HAB toxins (Ekstrom et al., 2020) and other contaminants (e.g., 15 methylmercury, metals, persistent organic pollutants) in seafood (Chan et al., 2021). A full assessment of 16 climate-change impacts on human health is found in Chapter 7 and Cross-Chapter Box ILLNESS in Chapter 17 2. 18 19 20 21 Figure 3.24: Assessment of feasibility and effectiveness of adaptation solutions for ocean and coastal ecosystems. 22 Feasibility dimensions assessed include: technical and economic capacity to deliver and implement the solution, the Do Not Cite, Quote or Distribute 3-132 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 institutional and geophysical capacity to implement a solution; and associated social and ecological implications that 2 make a solution more feasible. The general feasibility level is obtained from assessment of the three dimensions 3 together. Note that feasibility is assessed for marine and coastal ecosystems as a whole and not by ecosystem type or 4 region. Feasibility dimensions and assessment are updated and adapted from IPCC (2018) and Singh et al. (2020). 5 Effectiveness: ability of the adaptation solution to reduce climate change mid-term risks. Main solutions are assessed 6 per sector. Underlying data are available in Table SM3.3. 7 8 9 3.6.3.2 Cross-Cutting Solutions for Coastal and Ocean Ecosystems 10 11 SROCC concluded that protection, restoration and pollution reduction can support ocean and coastal 12 ecosystems (high confidence), and that EbA lowers climate risks locally and provides multiple societal 13 benefits (high confidence) (IPCC, 2019c). This section updates the assessment of the effectiveness of these 14 strategies for addressing climate impacts. 15 16 3.6.3.2.1 Area-based protection: MPAs for adapting to climate change 17 Marine protected areas (MPAs) are the most widely implemented area-based management approach (Section 18 3.6.2.3.2), commonly intended to conserve, preserve, or restore biodiversity and habitats, protect species, or 19 manage resources (especially fisheries) (National Research Council, 2001). By August 2021, 7.74% of the 20 ocean was protected (in both MPAs and other effective conservation measures, OECMs) (UNEP-WCMC 21 and IUCN, 2021), primarily within nations' Exclusive Economic Zones (EEZs). These MPAs support 22 adaptation by sustaining nearshore ecosystems that provide natural erosion barriers (Sections 3.4.2.­3.4.2.5, 23 Cross-Chapter Box SLR in Chapter 3), ecosystem function (Cheng et al., 2019), habitat, natural filtration, 24 carbon storage, livelihoods, and cultural opportunities (Sections 3.5.5, 3.5.6, Erskine et al., 2021) and help 25 ecosystems and livelihoods recover after extreme events (Roberts et al., 2017; Aalto et al., 2019; Wilson et 26 al., 2020a). However, in 2021 only 2.7% of the ocean was in fully or highly protected MPAs (Marine 27 Conservation Institute, 2021), the hard-to-achieve states that most effectively rebuild biomass and fish 28 community structure (Sala and Giakoumi, 2017; Bergseth, 2018; Zupan et al., 2018; Ohayon et al., 2021). 29 Only 1.18% of ABNJ is protected (UNEP-WCMC and IUCN, 2021), mostly due to governance limitations 30 (O'Leary and Roberts, 2017; Vijayaraghavan, 2021), but calls to protect more ABNJ emphasise the need to 31 protect habitat of long-range pelagic fish and marine mammals, maintain the ocean's regulating functions, 32 and minimise impacts from uses such as maritime shipping or deep-sea mining (Table 3.30). 33 34 MPAs are theorised to facilitate ecological climate adaptation and contribute to SDG14 ("Life below water", 35 Table 3.30, Figure 3.26, Bates et al., 2014; Lubchenco and Grorud-Colvert, 2015; Gattuso et al., 2018) 36 because they alleviate non-climate drivers and promote biodiversity (i.e., "managed resilience hypothesis", 37 Bruno et al., 2019; Maestro et al., 2019; Cinner et al., 2020). Current MPAs offer conservation benefits such 38 as increases in biomass and diversity of habitats, populations, and communities (high confidence) (Pendleton 39 et al., 2018; Bates et al., 2019; Stevenson et al., 2020; Lenihan et al., 2021; Ohayon et al., 2021), and these 40 benefits may last after some (possibly climate-enhanced) disturbances (e.g., tropical cyclones, McClure et 41 al., 2020). But current MPAs do not provide resilience against observed warming and heatwaves in tropical 42 to temperate ecosystems (medium confidence) (Bates et al., 2019; Bruno et al., 2019; Freedman et al., 2020; 43 Graham et al., 2020; Rilov et al., 2020). There is robust evidence that processes around MPA design and 44 implementation strongly influence whether outcomes are beneficial or harmful for adjacent human 45 communities (McNeill et al., 2018; Zupan et al., 2018; Ban et al., 2019). 46 47 Current placement and extent of MPAs will not provide substantial protections against projected climate 48 change past 2050 (high confidence), as the placement of MPAs has been driven more often by political 49 expediency (e.g., Leenhardt et al., 2013) than by managing key drivers of biodiversity loss (Cockerell et al., 50 2020; Stevenson et al., 2020) or climate-impact drivers (Bruno et al., 2018). Only 3.5% of the area currently 51 protected will provide refuges from both SST and deoxygenation by 2050 under both RCP4.5 and RCP8.5 52 (Bruno et al., 2018) and MPAs are more exposed to climate change under RCP8.5 than non-MPAs (Section 53 3.4.3.3.4, Figure 3.20d). Community thermal tolerances will be exceeded by 2050 in the tropics and by 2150 54 for many higher-latitude MPAs (Bruno et al., 2018). Most MPA design has focused on the surface ocean, but 55 MPAs are assumed to protect the entire water column and benthos. Climate-impact drivers (Section 3.2) 56 throughout the water column and rapidly accelerating climate velocities at depths below 200 m (Johnson et 57 al., 2018; Brito-Morales et al., 2020), are projected to affect virtually all North Atlantic deep-water and open 58 ocean area-based management zones in the next 20­50 years (Johnson et al., 2018) and the conservation Do Not Cite, Quote or Distribute 3-133 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 goals of benthic MPAs in the North Sea are not expected to be fulfilled (Weinert et al., 2021). Heightened 2 risk of non-indigenous species immigration from vessel traffic plus climate change further endangers MPA 3 success (Iacarella et al., 2020), a particular concern in the Mediterranean (D'Amen and Azzurro, 2020; 4 Mannino and Balistreri, 2021), where the current MPA network is already highly vulnerable to climate 5 change (Kyprioti et al., 2021). This new evidence supports SROCC's high confidence assessment that 6 present governance arrangements including MPAs are too fragmented to provide integrated responses to the 7 increasing and cascading risks from climate change in the ocean (SROCC SPMC1.2, IPCC, 2019c). 8 9 Strategic conservation planning can yield future MPA networks substantially more ready for climate change 10 (e.g., Section 3.6.3.1.5, SROCC SPM C2.1, IPCC, 2019c; Frazão Santos et al., 2020; Rassweiler et al., 11 2020). Global protection is increasing (Worm, 2017; Claudet et al., 2020b) as nations pursue international 12 targets (e.g., SDG14, "Life below water" aimed to conserve 10% of the ocean by 2020), and the UN CBD 13 proposes to protect 30% by 2030 (Section 3.6.4, SM3.5.3, CBD, 2020). A growing body of evidence 14 (Tittensor et al., 2019; Cabral et al., 2020; Zhao et al., 2020a; Pörtner et al., 2021b; Sala et al., 2021) 15 underscores the urgent need to pursue biodiversity, ecosystem-service provision, and climate-adaptation 16 goals simultaneously, while acknowledging inherent tradeoffs (Claudet et al., 2020a; Sala et al., 2021). 17 Frameworks to create "climate-smart" MPAs (Tittensor et al., 2019) generally include: defining conservation 18 goals that embrace resource vulnerabilities and co-occurring hazards; carefully selecting adaptation 19 strategies that include LK and IK while respecting Indigenous rights and accommodating human behaviour 20 (Kikiloi et al., 2017; Thomas, 2018; Yates et al., 2019; Failler et al., 2020; Wilson et al., 2020a; Croke, 2021; 21 Reimer et al., 2021; Vijayaraghavan, 2021); developing protection that is appropriate for all ocean depths 22 (Brito-Morales et al., 2018; Frazão Santos et al., 2020; Wilson et al., 2020a), especially considering climate 23 velocity (Arafeh-Dalmau et al., 2021); using dynamic national and international management tools to 24 accommodate extreme events or species distribution shifts (Gaines et al., 2018; Pinsky et al., 2018; Bindoff 25 et al., 2019; Scheffers and Pecl, 2019; Tittensor et al., 2019; Cashion et al., 2020; Crespo et al., 2020; Frazão 26 Santos et al., 2020; Maxwell et al., 2020b), which could build on dynamic regulations already in place for 27 fishing or ship strikes (Maxwell et al., 2020b); and seeking to increase connectivity (Wilson et al., 2020a), 28 using genomic or multispecies model insights (Xuereb et al., 2020; Friesen et al., 2021; Lima et al., 2021). 29 30 There is growing international support for a 30% conservation target for 2030 (Gurney et al., 2021), that will 31 need efforts beyond protected areas. For example, Other Effective area-based Conservation Measures 32 (OECMs) recognise management interventions that sustain biodiversity, irrespective of their main objective 33 (Maxwell et al., 2020b; Gurney et al., 2021). There is high agreement on the potential of OECMs to 34 contribute to conservation and equity, for example by recognising Indigenous territories as OECMs 35 (Maxwell et al., 2020b; Gurney et al., 2021). However, the capacity of these conservation tools to provide 36 adaptation outcomes remains unexplored. 37 38 In summary, MPAs and other marine spatial planning tools have great potential to address climate change 39 mitigation and adaptation in ocean and coastal ecosystems, if they are designed and implemented in a 40 coordinated way that takes into account ecosystem vulnerability and responses to projected climate 41 conditions, considers existing and future ecosystem uses and non-climate drivers, and supports effective 42 governance (high confidence). 43 44 3.6.3.2.2 Ecological restoration, interventions and their limitations 45 Restoration of degraded ecosystems is a common NbS increasingly deployed at local scales in response to 46 climate change (Cross-Chapter Box NATURAL in Chapter 2, Duarte et al., 2020; Bertolini and da Mosto, 47 2021; Braun de Torrez et al., 2021). Despite covering limited areas and having uncertain efficacy under 48 future climate change (Gordon et al., 2020), these actions have successfully restored marine populations and 49 ecosystems at regional to global scales (Duarte et al., 2020), and enhanced livelihoods and wellbeing of 50 coastal peoples as well as biodiversity and resilience of ecological communities (Silver et al., 2019; Gordon 51 et al., 2020; Braun de Torrez et al., 2021). Technology-based approaches like active restoration, assisted 52 evolution, and ecological forecasting can aid in moving beyond restoring ecosystems (Section 3.6.2.3) 53 towards enhancing resilience, reviving biodiversity and guarding against loss of foundational, ornamental or 54 iconic species (Bulleri et al., 2018; Collins et al., 2019a; da Silva et al., 2019; National Academies of 55 Sciences, 2019; Boström-Einarsson et al., 2020; Fredriksen et al., 2020; Morris et al., 2020c; Kleypas et al., 56 2021). 57 Do Not Cite, Quote or Distribute 3-134 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 Local restoration projects often target vegetated ecosystems like mangroves, seagrasses and saltmarshes that 2 are valued and used by coastal communities (Veettil et al., 2019; Duarte et al., 2020; Wu et al., 2020a; 3 Bertolini and da Mosto, 2021) Detail on mangroves and corals as EbA and protection/restoration hotspots is 4 provided in SuppMat 3.8. Common and effective actions (Sasmito et al., 2019; Duarte et al., 2020; Oreska et 5 al., 2020) include securing accommodation space (Sections 3.4.2.4­3.4.2.5), restoring hydrological (Kroeger 6 et al., 2017; Al-Haj and Fulweiler, 2020) and sediment dynamics; managing harvesting (particularly in 7 mangroves); reducing pollution (especially in seagrasses, de los Santos et al., 2019); and replanting 8 appropriate species in suitable environmental settings (Wodehouse and Rayment, 2019; Friess et al., 2020a). 9 Although efficacy is context dependent (Zeng et al., 2020; Krause-Jensen et al., 2021), and implementation 10 is most often local (Alongi, 2018a), such projects allow tangible community engagement in climate action. 11 Moreover, because these ecosystems sequester disproportionate amounts of carbon (blue carbon, Annex II: 12 Glossary, Box 3.4), restoration supports climate-change mitigation (Lovelock and Reef, 2020; Gattuso et al., 13 2021). Yet, constraints remain. For instance, Southeast Asia has 1.21 million km2 of terrestrial, freshwater 14 and mangrove area biophysically suitable for reforestation, which could mitigate 3.43 ± 1.29 Pg CO2e yr-1 15 through 2030; however, reforestation is only feasible in a small fraction of this area (0.3­18%) given 16 financial, land-use and operational constraints (Zeng et al., 2020). Nevertheless, the multiple benefits offered 17 by ecosystem restoration will likely outweigh competing costs, and increase its relevance as part of 18 adaptation strategy portfolios (Silver et al., 2019; Wedding et al., 2021), national carbon-accounting systems, 19 and NDCs (Friess et al., 2020a; Wu et al., 2020a). 20 21 Restoration efficacy of coral reefs, kelp forests and other habitat-forming coastal ecosystems (Section 22 3.4.2.2­3.4.2.6) are jeopardised by the near-term nature of climate-driven risks (McLeod et al., 2019; 23 National Academies of Sciences, 2019; Coleman et al., 2020b). Modelling studies indicate that available 24 practices will not prevent degradation of coral reefs from >1.5°C of global average surface warming (Figure 25 3.25, National Academies of Sciences Engineering and Medicine, 2019; Condie et al., 2021; Hafezi et al., 26 2021). Proposed interventions, not yet implemented, include assisted migration (Boström-Einarsson et al., 27 2020; Fredriksen et al., 2020; Morris et al., 2020c), assisted evolution (Bay et al., 2019; National Academies 28 of Sciences, 2019) and other engineering solutions like artificial shading and enhanced upwelling (Condie et 29 al., 2021; Kleypas et al., 2021). 30 31 Transplanting heat-tolerant coral colonies can increase reef resistance to bleaching (Morikawa and Palumbi, 32 2019; Howells et al., 2021), but potentially lowering species diversity and altering ecosystem function 33 (Section 3.4.2.1). Genetic manipulation or assisted evolution that propagates genes from heat-tolerant 34 populations could enhance restoration of corals (Anthony et al., 2017; Epstein et al., 2019) and kelp (medium 35 agreement, limited evidence) (Coleman and Goold, 2019; Coleman et al., 2020b; Fredriksen et al., 2020; 36 Wade et al., 2020). Managed breeding of corals has also had limited success in the laboratory and at small 37 local scales (National Academies of Sciences, 2019). There is also limited evidence that physiological 38 interventions like algal-symbiont or microbiome manipulation could increase coral thermal tolerance in the 39 field (National Academies of Sciences, 2019). Employing the natural adaptive capacity of species or 40 individuals in active restoration for corals and kelps with current technology involves fewer risks than 41 assisted evolution or long-distance relocation (high confidence) (Filbee-Dexter and Smajdor, 2019; National 42 Academies of Sciences, 2019). More ambitious engineered interventions like reef shading remain theoretical 43 and not scalable to the reef level (Condie et al., 2021). Debate continues on how to apply planned adaptation 44 in cost-effective ways that will accomplish the intended goals (National Academies of Sciences, 2019; 45 Duarte et al., 2020; Kleypas et al., 2021). 46 47 Models show that a combination of available management approaches (restoration, reducing non-climate 48 drivers) and speculative interventions (enhanced corals, reef shading) can contribute to sustaining some coral 49 reefs beyond 1.5°C of global warming with declining effectiveness beyond 2°C of global warming (medium 50 confidence) (Figure 3.25, WGII Chapter 17). These proposed interventions are also currently theoretical and 51 impractical over large scales; for example, engineered solutions like reef shading are untested and not 52 scalable at the reef level (Condie et al., 2021). Existing projects suggest that restoration and ecological 53 interventions to habitat-forming ecosystems have additional benefits of raising local awareness, promoting 54 tourism, and creating jobs and economic benefits (Fadli et al., 2012; Boström-Einarsson et al., 2020; Hafezi 55 et al., 2021), provided communities are involved in planning, operation and monitoring (Boström-Einarsson 56 et al., 2020). 57 Do Not Cite, Quote or Distribute 3-135 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 3 Figure 3.25: Implemented and potential future adaptations in ocean and coastal ecosystems. (a) Global implementation 4 since 1970 of (top) cumulative habitat-restoration projects (Duarte et al., 2020), (middle) cumulative area-based 5 conservation protected area (MPA total, Boonzaier and Pauly, 2016), no-take areas (UN Environment World 6 Conservation Monitoring Centre et al., 2018; UNEP-WCMC, 2019), and (bottom) percentage of total fish stocks rebuilt 7 (Kleisner et al., 2013). (b) Adaptation pathways for coral reefs to maintain healthy cover (line weight: solid lines, likely 8 effectiveness; dashed lines, more likely than not to likely; dotted lines = unlikely to more likely than not), with 9 confidence noted for each intervention (VH = Very High, H = High, M = Medium) (Section 3.4.2.1, 3.6.3.2, Anthony et 10 al., 2019; National Academies of Sciences Engineering and Medicine, 2019). Circles denote when other measures must 11 also be implemented. (c) As in (b), but for mangrove ecosystems. Underlying data are available in Tables SM3.4­3.6. 12 13 14 3.6.3.3 Enablers, Barriers and Limitations of Adaptation and Mitigation 15 16 Not only is mitigation necessary to support ocean and coastal adaptation (Pörtner et al., 2014; Oppenheimer 17 et al., 2019), but the global emission pathways also impose limits to ocean and coastal adaptation, with lower 18 warming levels enabling greater effectiveness of adaptations (high confidence) (Figure 3.25). Chapter 17 19 broadly assesses the limits to adaptation, while this section focuses on barriers and limits to adaptation 20 imposed by cultural (Section 3.6.3.3.1), economic (Section 3.6.3.3.2) and governance (Section 3.6.3.3.3) 21 dimensions (Hinkel et al., 2018). Globally, these factors more strongly influence ocean development than 22 does local natural resource availability (Cisneros-Montemayor et al., 2021), and are key to avoiding 23 maladaptation. This section also assesses enablers and limits to mitigation (Section 3.6.3.3.4). 24 25 3.6.3.3.1 Sociocultural dimensions (culture, ethics, identity, behaviour) 26 Every coastal community values marine ecosystems for more than the material and intangible resources they 27 deliver, or the physical protection they offer (Díaz et al., 2018). Cultural services that provide identity, 28 spiritual and cultural continuity, religious meaning, or options for the future (e.g., genetic or mineral 29 resources, Bindoff et al., 2019), are not substitutable. Furthermore, interactions between climate impacts and 30 existing inequalities can threaten the human rights of already-marginalised peoples by disrupting livelihoods 31 and food security, which further erodes people's social, economic, and cultural rights (Finkbeiner et al., Do Not Cite, Quote or Distribute 3-136 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2018). For instance, European colonisation and ongoing development blocked the Cucapá Indigenous 2 People's access and rights to resources in the Colorado River Delta, USA, over the 20th century. Recent 3 reallocation of water rights and fishing access is allowing the Cucapá people to reconstruct their cultural 4 identity (Sangha et al., 2019), but future climate change impacts could reverse the community's recovery of 5 their cultural heritage. Adaptations that consider local needs may help sustain cultural services (Ortíz Liñán 6 and Vázquez Solís, 2021). 7 8 Interactions with oceans are fundamental to the identities of many coastal Indigenous Peoples (Norman, 9 2017) and this influences Indigenous responses to climate hazards and adaptation. Around 30 million 10 Indigenous Peoples live along coasts (Cisneros-Montemayor et al., 2016). Seafood consumption among 11 Indigenous Peoples is much higher than for non-Indigenous populations, and marine species support many 12 cultural, medicinal and traditional activities contributing to public health (Section 3.5.3.1, Kenny et al., 13 2018). Perpetuation of Indigenous cultures depends on protecting marine ecosystems and on adapting to 14 changes in self-led ways (see Section 3.5.6, Sangha et al., 2019) that promote self-determination (von der 15 Porten et al., 2019). Indigenous resurgence, or reinvigorating Indigenous ways of life and traditional 16 management, can include marine resource protection and ocean-sector development founded on culturally 17 appropriate strategies and partnerships, that are consistent with traditional norms and beneficial to local 18 communities (von der Porten et al., 2019). Successful adaptation would simultaneously improve ecosystem 19 health and address current and historical inequities (Bennett, 2018). Examples include practicing traditional 20 resource management, protecting traditional territories, engaging with monitoring, collaborations with non- 21 Indigenous partners, and reinvesting benefits into capacity-building within communities (von der Porten et 22 al., 2019; Equator Initiative, 2020). The legitimacy of different adaptation strategies depends on local and 23 Indigenous Peoples' acceptance, which is based on cultural values (Adger et al., 2017); financial gain cannot 24 compensate for loss of IK or LK (Wilson et al., 2020b). Palau's recent goal of shifting seafood consumption 25 away from reef fishes (Remengesau Jr., 2019) and limiting and closely monitoring the expansion of 26 ecotourism was prompted by the cultural importance of protecting these reefs and associated traditional 27 fisheries for local consumption, a recognition of the importance of tourism, and the hazard of climate change 28 (Wabnitz et al., 2018a). 29 30 Adaptations implemented at the local level that consider IK and LK systems are beneficial (high confidence) 31 (Nalau et al., 2018; Sultana et al., 2019). Studies in SIDS and the Arctic have shown how IK and LK 32 facilitate the success of EbA (Nalau et al., 2018; Peñaherrera-Palma et al., 2018; Raymond-Yakoubian and 33 Daniel, 2018), reinforce and improve institutional approaches and enhance the provision of ecosystem 34 services (Ross et al., 2019; Terra Stori et al., 2019). Perspectives on adaptation also vary among groups of 35 age, race, (dis)ability, class, caste, and gender (Wilson et al., 2020b), so engaging different groups results in 36 more robust and equitable adaptation to climate change (Cross-Chapter Box GENDER in Chapter 18, 37 McLeod et al., 2018). Some coastal communities have developed substantial social capital and dense local 38 networks based on trust and reciprocity (Petzold and Ratter, 2015), with individual and community 39 flexibility to learn, adapt, and organise themselves to help local adaptation governance (Silva et al., 2020). 40 Recent evidence suggests that policies supporting local institutions can improve adaptation outcomes 41 (medium confidence) (Berman et al., 2020). Coastal communities can be engaged using novel approaches to 42 co-generate adaptation solutions (van der Voorn et al., 2017; Flood et al., 2018) that benefit education 43 (Koenigstein et al., 2020) and engagement in adaptation processes (Rumore et al., 2016). Successful 44 adaptation implementation in line with climate-resilient development pathways (WGII Chapter 18) depends 45 on bottom-up, participatory and inclusive processes (Section 3.6.1.2.1) that engage diverse stakeholders 46 (Basel et al., 2020; McNamara et al., 2020; Ogier et al., 2020; Williams et al., 2020), and that protect 47 Indigenous customary rights (Farbotko and McMichael, 2019; Ford et al., 2020), empower women, and give 48 rights to climate refugees (McLeod et al., 2018). 49 50 3.6.3.3.2 Economic dimensions (planning, finance, costs) 51 Finance is a key barrier globally for ocean health, governance and adaptation to climate change (high 52 agreement) (Annex II: Glossary, Cross-Chapter Box FINANCE in Chapter 17, Hinkel et al., 2018; Miller et 53 al., 2018; Wabnitz and Blasiak, 2019; Woodruff et al., 2020; Sumaila et al., 2021). Global adaptation finance 54 was estimated to total 30 billion USD yr­1 in 2017­2018, or 5% of all climate finance (CPI, 2019), with no 55 tracking specifically for coastal or marine adaptation in low- to middle-income countries. Marine-focused 56 adaptation finance is difficult to trace and label due to the cross-sectoral nature of many projects (Blasiak 57 and Wabnitz, 2018) and the lack of clear definitions about what qualifies as adaptation or as new and Do Not Cite, Quote or Distribute 3-137 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 additional finance (Donner et al., 2016; Weikmans and Roberts, 2019). Finance for marine conservation 2 from Overseas Development Assistance doubled between 2003 and 2016, reaching 634 million USD in 3 2016, similar to the level provided by philanthropic foundations (Berger et al., 2019). Yet coastal adaptation 4 to SLR alone is projected to cost hundreds of billions of USD yr­1, depending on the model and emission 5 scenario (e.g., Wong et al., 2014; Nicholls et al., 2019). Economic and financing barriers to marine 6 adaptation are often higher in low- to middle-income countries, where resources influence governance and 7 constrain options for implementation and maintenance (high confidence) (Hinkel et al., 2018; Klöck and 8 Nunn, 2019; Tompkins et al., 2020) and impacts on their coastal and marine ecosystems could total several 9 percentage points of their gross domestic product (Wong et al., 2014). Current financial flows are 10 insufficient to meet the costs of coastal and marine impacts of climate change (very high confidence) and 11 ocean-focused finance is unevenly distributed, with higher flows within and to developed countries (very 12 high confidence). 13 14 Development assistance can help resolve resource constraints, but additional governance and coordination 15 challenges can arise from short-term, project-based funding, shifting priorities of donor institutions, and 16 pressures placed on human resources in the receiving nation (Parsons and Nalau, 2019; Nunn et al., 2020). 17 Innovative policy instruments like concessional loans, tax-policy reforms, climate bonds and public-debt 18 forgiveness can supplement traditional financial instruments (Bisaro and Hinkel, 2018; McGowan et al., 19 2020). Mechanisms for solving the persistent problem of securing upfront investments for coastal protection 20 and other adaptation measures (Bisaro and Hinkel, 2018; Moser et al., 2019; Kok et al., 2021) include 21 integrating adaptation investments into insurance schemes (Reguero et al., 2020) and using debt financing to 22 bridge the time until benefits are realised (Ware and Banhalmi-Zakar, 2020). Insurance mechanisms that link 23 payments to losses from a trigger event (e.g., MHW) can confer resilience to marine-dependent communities 24 (Sumaila et al., 2021). All innovative financial instruments are most effective when they are inclusive and 25 reach vulnerable groups and marginalised communities (low evidence, high agreement) (Claudet et al., 26 2020a; Sumaila et al., 2021). 27 28 Countries with large ocean areas within their EEZs have opportunities to develop "blue-green economies" to 29 reduce emissions and finance adaptation pathways (Chen et al., 2018a; Lee et al., 2020). Shifting from grants 30 to results-based financing can help attract more private capital to ocean adaptation (Lubchenco et al., 2016; 31 Claudet et al., 2020a). Public-private partnerships can also increase ocean adaptation finance (Goldstein et 32 al., 2019; Sumaila et al., 2021). For example, the financial benefits that biodiversity conservation confers to 33 seafood harvest resilience could be used to leverage industry participation in adaptation and conservation 34 finance (Barbier et al., 2018). Connecting restoration of blue carbon ecosystems with offset markets (e.g., 35 Vanderklift et al., 2019) shows potential, but uncertainties remain about the international emissions trading 36 under the UN Framework Convention on Climate Change and climate impacts on blue-carbon ecosystems 37 (Section 3.6.3.1.6, Lovelock et al., 2017a; Macreadie et al., 2019). 38 39 Transparency, coherence between different actors and initiatives, and project monitoring and evaluation 40 enhance success in adapting and achieving SDG14 (Life below water) (Blasiak et al., 2019). Maladaptation 41 (WGII Chapter 16, Magnan et al., 2016), is a common risk of current project-based funding due to the 42 pressure to produce concrete results (medium confidence) (Parsons and Nalau, 2019; Nunn et al., 2020; Nunn 43 et al., 2021). Maladaptation can be avoided through a focus on building adaptive capacity, community-based 44 management, drivers of vulnerability and site-specific measures (low confidence) (Magnan and Duvat, 2018; 45 Piggott-McKellar et al., 2020; Schipper, 2020). More research is needed to identify ways that governance 46 and financing agreements can help overcome financial barriers and socio-cultural constraints to avoid 47 maladaptation in coastal ecosystems (high confidence) (Hinkel et al., 2018; Miller et al., 2018; Piggott- 48 McKellar et al., 2020; Schipper, 2020). 49 50 3.6.3.3.3 Governance dimension (institutional settings, decision making) 51 Ocean governance has become increasingly complex as new initiatives, new international agreements, 52 institutions, and scientific evidence arise at global, national, and sub-national scales (high agreement) 53 (Bindoff et al., 2019; Scobie, 2019b), limiting the present effectiveness of adaptation (IPCC, 2019c). Marine 54 climate governance is within the normatively contested marine governance space (Frazão Santos et al., 55 2020), which is influenced by geopolitics (Gray et al., 2020) and profit maximisation (Flannery et al., 2016; 56 Haas et al., 2021) in ways that can entrench exclusionary processes in decision making, science management 57 and funding (Levin et al., 2018). This limits just and inclusive ocean governance (Bennett, 2018), Do Not Cite, Quote or Distribute 3-138 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 perpetuates historical and cultural extractive practices and climate inaction, and leaves little space for 2 Indigenous-led adaptation frameworks and approaches (Nursey-Bray et al., 2019). At the national level, 3 ocean governance for climate-change adaptation is often transversal, requiring consideration of biophysical 4 and environmental conditions (Furlan et al., 2020), while fitting into existing economic (Kim, 2020) and 5 political processes. Adaptation governance that couples existing top-down structures with decentralised and 6 participatory approaches generates shared goals and unlocks required resources and monitoring (Gupta et al., 7 2016; Haas et al., 2021). 8 9 Communities and governments at all levels increasingly use decision-making frameworks (e.g., structured 10 decision making) or decision-analysis tools to evaluate trade-offs between different responses, rather than 11 applying generic best practices to different physical, technical or cultural contexts (high confidence) 12 (Watkiss et al., 2015; Haasnoot et al., 2019; Palutikof et al., 2019). Increased effort has also been devoted to 13 developing climate services (actionable information and data products) that bridge the gap between climate 14 prediction and decision-making (Hewitt et al., 2020). Climate services have the potential to inform decision 15 making related to disaster-risk reduction, adaptation responses, marine environmental management (e.g., 16 fisheries management and MPA management) and ocean-based climate mitigation (e.g., renewable energy 17 installations, Le Cozannet et al., 2017; Gattuso et al., 2019; Gattuso et al., 2021). Although improving 18 observational and modelling capacity is important to developing ocean-focused services, particularly in high- 19 risk regions like SIDS where regional climate projections are scarce (WGI AR6 Chapter 9, Morim et al., 20 2019; Fox-Kemper et al., 2021), data is not the only limiting factor in decision-making (Weichselgartner and 21 Arheimer, 2019). Focusing on user engagement, relationship-building and the decision-making context 22 ensures that climate services are useful to and used by different stakeholders (high confidence) (Soares et al., 23 2018; Mackenzie et al., 2019; Weichselgartner and Arheimer, 2019; Findlater et al., 2021; West et al., 2021). 24 25 3.6.3.3.4 Mitigation 26 Ocean and coastal NbS can contribute to global mitigation efforts, especially with ocean renewable energy 27 and restoration and preservation of carbon ecosystems (Box 3.4, Section 3.6.2.3). Technological, economic 28 and financing barriers presently hamper development of renewable ocean energy (AR6 WGIII Chapter 6). 29 Such development could help small nations reliant on imported fuel meet their climate-mitigation goals and 30 decrease risk from global fuel supply dynamics (Millar et al., 2017; Chen et al., 2018a), but progress is 31 limited by lack of investment (Millar et al., 2017; Lee et al., 2020) or equipment (Aderinto and Li, 2018; 32 Rusu and Onea, 2018). Wave-energy installations, possibly co-located with wind turbines(Perez-Collazo et 33 al., 2018), are promising for both low- to middle-income nations and areas with significant island or remote 34 coastal geographies (Lavidas and Venugopal, 2016; Bergillos et al., 2018; Jakimavicius et al., 2018; Kompor 35 et al., 2018; Penalba et al., 2018; Saprykina and Kuznetsov, 2018; Lavidas, 2019). Wave-energy capture may 36 also diminish storm-induced coastal erosion (Abanades et al., 2018; Bergillos et al., 2018). Tidal energy is a 37 relatively new technology (Haslett et al., 2018; Liu et al., 2018; Neill et al., 2018) with limiting siting 38 requirements (Mofor et al., 2013). Ocean renewable energy expansion faces other technological obstacles 39 including lack of implementable or scalable energy-capture devices, access to offshore sites, competing 40 coastal uses, potential environmental impacts, and lack of power-grid infrastructure at the coast (Aderinto 41 and Li, 2018; Neill et al., 2018). 42 43 3.6.4 Contribution to the Sustainable Development Goals and Other Relevant Policy Frameworks 44 45 The impacts of climate change on ocean and coastal ecosystems and their services threaten achievement of 46 the UN SDGs by 2030 (high confidence), particularly ocean targets (Table 3.31, Nilsson et al., 2016; Pecl et 47 al., 2017; IPCC, 2018; Singh et al., 2019a; Claudet et al., 2020a). Nevertheless, local to international 48 decision-making bodies have assigned the lowest priority to SDG14, Life Below Water (Nash et al., 2020). 49 50 51 Table 3.31: Sustainable Development Goals, grouped into broader categories as discussed in this section 52 (http://sdgs.un.org/goals). Category Goal Society SDG1: No Poverty SDG2: Zero Hunger SDG3: Good Health & Well-Being SDG4: Quality Education SDG5: Gender Equality Do Not Cite, Quote or Distribute 3-139 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report Economy SDG6: Clean Water & Sanitation SDG7: Affordable & Clean Energy Environment SDG8: Decent Work & Economic Growth Governance SDG9: Industry, Innovation & Infrastructure SDG10: Reduced Inequality SDG11: Sustainable Cities & Communities SDG12: Responsible Consumption & Production SDG13: Climate Action SDG14: Life Below Water SDG15: Life on Land SDG16: Peace and Justice Strong Institutions SDG17: Partnerships to achieve the Goals 1 2 3 3.6.4.1 Climate Mitigation Effects on Ocean-Related SDGs 4 5 SROCC underscored the need for ambitious mitigation to control climate hazards in the ocean to achieve 6 SDGs (medium evidence, high agreement) (Bindoff et al., 2019; Oppenheimer et al., 2019). Delays in 7 achieving ocean-dependent SDGs observed in SROCC and SR15 can be addressed with ambitious planned 8 adaptation and mitigation action (high agreement) (Hoegh-Guldberg et al., 2019b). Since the ocean can 9 contribute substantially to the attainment of mitigation targets aiming to limit warming to 1.5ºC above pre- 10 industrial (Hoegh-Guldberg et al., 2019b), and to adaptation solutions facilitating attainment of social and 11 economic SDGs, climate policy is treating the ocean less as a victim of climate change and more as a central 12 participant in solving the global climate challenge (Cooley et al., 2019; Hoegh-Guldberg et al., 2019a; 13 Dundas et al., 2020). 14 15 Relationships between Climate Action (SDG13) targets and SDG14 targets are mostly synergistic (Figure 16 3.26, Fuso Nerini et al., 2019). Responding to climate-change impacts requires transformative governance 17 (high confidence) (Chapters 1 and 18, Collins et al., 2019a; Brodie Rudolph et al., 2020; Claudet et al., 18 2020a), especially for extreme events and higher-impact scenarios (e.g., higher emissions) (Fedele et al., 19 2019), and for achieving SDGs through one of the global ecosystems transitions (Chapter 18, Sachs et al., 20 2019; Brodie Rudolph et al., 2020). Opportunities to transform ocean governance exist in developing new 21 international and local agreements, regulations and policies that reduce the risks of relocating ocean and 22 coastal activities (Section 3.6.3.1.1) or in reinventing established practices (Section 3.6.3.3.3). Policy 23 transformations improving ocean sustainability under SDG14 also help address SDG13 (Brodie Rudolph et 24 al., 2020; Dundas et al., 2020; Claudet, 2021; Sumaila et al., 2021). Emergent situations such as the COVID- 25 19 pandemic may provide opportunities to implement transformative "green recovery plans" that support 26 achievement of the SDGs and NDCs (Cross-Chapter Box COVID in Chapter 7). 27 28 3.6.4.2 Contribution of Ocean Adaptation to SDGs 29 30 Marine-focused adaptations show promise in helping achieve social SDGs, especially when they are 31 designed to achieve multiple benefits (medium confidence) (Figure 3.26, Ntona and Morgera, 2018; Claudet 32 et al., 2020a). Technology- and infrastructure-focused adaptations (Section 3.6.2.2) can help relieve coastal 33 communities from risks associated with poverty (SDG1), hunger (SDG2), health and water sanitation (SDG3 34 and SDG6), and inequality (SDG10) by supporting aquaculture (Sections 3.5.3, 3.6.3.1), alerting the public 35 about poor water quality (Sections 3.5.5.3, 3.6.3.1), and empowering marginalised groups, such as women 36 and Indigenous Peoples, with decision-relevant information (medium evidence, high agreement) (Sections 37 3.5.5.3, 3.6.3.1). Effectively implemented and managed marine NbS (Section 3.6.2.3) contribute to 38 attainment of social SDGs by preserving biodiversity (Carlton and Fowler, 2018; Warner, 2018; Scheffers 39 and Pecl, 2019), which benefits most ocean and coastal ecosystem services (Section 3.5.3, Figure 3.22), by 40 increasing marine fishery and aquaculture sustainability (Section 3.6.3), by including vulnerable people and 41 communities in management (Section 3.6.3.2.1), by lowering risk of flooding from storms and SLR (Cross- 42 Chapter Box SLR in Chapter 3, Sections 3.6.3.1.1), and by implementing spatial management tools that 43 make room for new uses like renewable energy development (Section 3.6.3.3.4). NbS can therefore help 44 support achievement of No Poverty (SDG1) (Ntona and Morgera, 2018), Zero Hunger (SDG2), Good Health 45 and Well-Being (SDG3) (Duarte et al., 2020), Affordable and Clean Energy (SDG7) (Fuso Nerini et al., 46 2019; Levin et al., 2020), and Reduced Inequality (SDG10). Socio-institutional marine adaptations (Section Do Not Cite, Quote or Distribute 3-140 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 3.6.2.2) that support current livelihoods and help develop alternatives can contribute to attainment of social 2 SDGs by enhancing social equity and supporting societal transformation (medium confidence) (Cisneros- 3 Montemayor et al., 2019; Pelling and Garschagen, 2019; Nash et al., 2021). Even societal changes that are 4 not directly marine-related can decrease human vulnerability to ocean and coastal climate risks by improving 5 overall human adaptive capacity (Section 1.2). 6 7 Marine adaptation also shows promise for helping support achievement of economic SDGs (medium 8 confidence) (Figure 3.26). Marine NbS could help blue economy frameworks achieve Decent Work and 9 Economic Growth (SDG8) (Lee et al., 2020), by sustainably and equitably incorporating ecosystem-based 10 fisheries management, restoration or conservation (Sections 3.6.3.1.2, 3.6.3.2.1 and 3.6.3.2.2) (Voyer et al., 11 2018; Cisneros-Montemayor et al., 2019; Cohen et al., 2019; Okafor-Yarwood et al., 2020). NbS that involve 12 active restoration or accommodation can contribute to Sustainable Cities and Communities (SDG11) and 13 Infrastructure (SDG9) (Section 3.6.3.1.1). Newly developed marine industries and livelihoods associated 14 with NbS might support attainment of Sustainable Communities (SDG11) (Cisneros-Montemayor et al., 15 2019). Finance and market mechanisms to support disaster relief or ocean ecosystem services, such as blue 16 carbon or food provisioning, and innovations (SDG9) including new technologies like vessel-monitoring 17 systems (Kroodsma et al., 2018), can contribute to Responsible Consumption and Production (SDG12) 18 (Sumaila and Tai, 2020). Blue economy growth that includes sustainable shipping, tourism, renewable ocean 19 energy, and transboundary fisheries management (Pinsky et al., 2018) have the potential to contribute to 20 Economic Development (SDG8), affordable and clean energy (SDG7) (as well as global mitigation efforts, 21 SDG13, (Hoegh-Guldberg et al., 2019b; Duarte et al., 2020)). Participatory approaches and co-management 22 systems (Section 3.6.2.1) in many maritime sectors can contribute to SDG11 and SDG12 while helping align 23 the blue economy and the SDGs (high agreement) (Lee et al., 2020; Okafor-Yarwood et al., 2020). 24 25 Developing marine adaptation pathways that offer multiple benefits requires transformational adaptation 26 (high confidence) (Claudet et al., 2020a; Friedman et al., 2020; Wilson et al., 2020b; Nash et al., 2021) that 27 avoids risky and maladaptive actions (Magnan and Duvat, 2018; Ojea et al., 2020). Ocean and coastal 28 extreme events and other hazards disproportionately harm the most vulnerable communities in SIDS, tropical 29 and Arctic regions, and Indigenous Peoples (Chapter 8.2.1.2). Presently implemented adaptation activity, at 30 the aggregate level, adversely affects multiple gender targets under SDG5 (high confidence) (Cross-Chapter 31 Box GENDER in Chapter 18). Although women make up over half of the global seafood production 32 workforce (fishing and processing sectors), provide more than half the artisanal landings in Pacific region 33 (Harper et al., 2013), dominate some seafood sectors such as seaweed (Howard and Pecl, 2019) and shellfish 34 harvesting (Turner et al., 2020a), and account for 11% of global artisanal fisheries participants (Harper et al., 35 2020b), they are often not specifically counted in datasets and excluded from decision-making and support 36 programs (Cross-Chapter Box GENDER in Chapter 18, Harper et al., 2020b; Michalena et al., 2020). 37 Targeted efforts to incorporate knowledge diversity, and include artisanal fishers, women and Indigenous 38 Peoples within international, regional, and local policy planning promote marine adaptation that supports 39 achievement of gender equality (SDG5) and reduces inequalities (SDG10) (limited evidence, high 40 agreement) (FAO, 2015). Integrated planning, financing, and implementation can help overcome these 41 limitations (Section 3.6.3.3.2, Cross-Chapter Box FINANCE in Chapter 17), ensuring that marine 42 adaptations do not compromise overall human equity or specific SDGs (Österblom et al., 2020; Nash et al., 43 2021), but are in fact fully synergistic with these goals (Bennett et al., 2021). 44 45 Do Not Cite, Quote or Distribute 3-141 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 Figure 3.26: Synergies and trade-offs between SDG13 Climate Action, SDG14 Life Below Water and social, economic 3 and governance SDGs. Achieving SDG13 provides positive outcomes and supports the achievement of all SDG14 4 targets. In turn, meeting SDG14 drives mostly positive interactions with social, economic and governance SDGs. The 5 interaction types, `Indivisible' (inextricably linked to the achievement of another goal), `Reinforcing' (aids the 6 achievement of another goal), `Enabling' (creates conditions that further another goal), `Consistent' (no significant 7 positive or negative interactions), `Constraining' (limits options on another goal), follows Nilsson et al. (2016)'s scoring 8 system based on authors' assessment, and agreement denotes consistency across author ratings. Full data available in 9 Table SM3.7. 10 11 12 3.6.4.3 Relevant Policy Frameworks for Ocean Adaptation 13 14 The intricacy, scope, timescales, and uncertainties associated with climate change challenge ocean 15 governance, which already is extremely complex because it encompasses a variety of overlapping spatial 16 scales, concerns, and governance structures (Figure CB3.1 in SROCC Chapter 1, Prakash et al., 2019). 17 Assessment of how established global agreements and regional, sectoral, or scientific bodies address climate 18 adaptation and resilience, and how current practices can be improved, is found in SM3.5.3. 19 20 There is growing momentum to include the ocean in international climate policy (robust evidence), paving 21 the way for a more integrated approach to both mitigation and adaptation. Following adoption of the Paris 22 Agreement in 2015, the UN SDGs (Table 3.31) came into force in 2016, including SDG14 specifically 23 dedicated to life below water (Table 3.31). In 2017, the first UN Ocean Conference was held (United 24 Nations, 2017), the UNFCCC adopted the Ocean Pathway to increase ocean-targeted multilateral climate 25 action (COP23, 2017), and the UN Assembly declared 2021­2030 the Decade for Ocean Science for 26 Sustainable Development (Visbeck, 2018; Lee et al., 2020). Next, 14 world leaders formed the High-Level 27 Panel for a Sustainable Ocean Economy to produce the New Ocean Action Agenda, founded on 100% 28 sustainable management of national ocean spaces by 2025 (Ocean Panel, 2020). All of these initiatives 29 position oceans centrally within the climate-policy and biodiversity-conservation landscapes and seek to 30 develop a coherent effort and common frameworks to achieve marine sustainability (Visbeck, 2018; Lee et 31 al., 2020), new economic opportunities (Konar and Ding, 2020; Lee et al., 2020), more equitable outcomes 32 (Österblom et al., 2020), and decisive climate mitigation and adaptation (Hoegh-Guldberg et al., 2019a), to 33 achieve truly transformative change (Claudet et al., 2020a). 34 35 There is high confidence in the literature that multilateral environmental agreements need better alignment 36 and integration to support achievement of ambitious international development, climate mitigation, and 37 adaptation goals (Swilling et al., 202; Duarte et al., 2020; Friedman et al., 2020; Conservation International Do Not Cite, Quote or Distribute 3-142 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 and IUCN, 2021; Pörtner et al., 2021b; Sumaila et al., 2021). The ocean targets of the CBD (e.g., the Post- 2 2020 Global Biodiversity Framework), the SDGs (Agenda 2030) and the Paris Agreement are already 3 inclusive and synergistic (Duarte et al., 2020). However, specific policy instruments and sectors within them 4 could be additionally integrated, especially to address such cross-cutting impacts as ocean acidification and 5 deoxygenation (Gallo et al., 2017; Bindoff et al., 2019), increasing plastic pollution (Ostle et al., 2019; 6 Duarte et al., 2020), high-seas governance (Johnson et al., 2019; Leary, 2019), or deep sea uses (Wright et 7 al., 2019; Levin et al., 2020; Orejas et al., 2020). National adaptation plans present opportunities to 8 synergistically build on mitigation to support equitable development (Morioka et al., 2020), economic 9 planning (Dundas et al., 2020; Lee et al., 2020), and ocean stewardship (von Schuckmann et al., 2020). 10 Alignment of multilateral agreements is expected to increase mitigation impact as well as increase adaptation 11 options (Section 3.6.3, Figure 3.25, Roberts et al., 2020). Opportunities to improve multilateral 12 environmental agreements and policies beyond UNFCCC and CBD processes are discussed in SM3.5.3, and 13 an assessment of commercial species management initiatives and needs is in Chapter 5. 14 15 3.6.5 Emerging Best Practices for Ocean and Coastal Climate Adaptation 16 17 There is robust evidence that a combination of global and local solutions offers the greatest benefit in 18 reducing climate risk (Gattuso et al., 2018; Hoegh-Guldberg et al., 2019a; Hoegh-Guldberg et al., 2019b). 19 Ambitious and swift global mitigation offers more adaptation options and pathways to sustain ecosystems 20 and their services (Figure 3.25). Some solutions target both mitigation and adaptation (e.g., blue carbon 21 conservation, Cross-Chapter Box NATURAL in Chapter 2, Box 3.4), and cross-cutting solutions 22 simultaneously support several ocean-related sectors (e.g., area-based measures support fishing, tourism; 23 Section 3.6.3.2.1) or ecosystem functions (e.g., NbS support coastal protection, biodiversity, habitat, etc., 24 Section 3.6.3.2.2, Sala et al., 2021). Combined solutions also leverage a variety of existing policies and 25 governance systems (Section 3.6.4.3, Duarte et al., 2020) to advance climate mitigation and adaptation. Even 26 communities that face the limits of adaptation, like those who must relocate to cope with rising seas 27 (McMichael et al., 2019; Bronen et al., 2020), urgently require solutions that combine scientific projections, 28 IK and LK, cultural and community values, and ways to preserve cultural identity to support planning and 29 implementation of relocation (McMichael and Katonivualiku, 2020). 30 31 NbS are showing promising results in achieving adaptation and mitigation outcomes across marine and 32 coastal ecosystems (Sections 3.6.3.2.1­3.6.3.2.2), but NbS have different degrees of readiness in marine 33 ecosystems (Duarte et al., 2020). Habitat restoration and recovery are highly effective in specific settings and 34 conditions (McLeod et al., 2019). Restoring and conserving vegetated coastal habitats (Sections 3.4.2.4­ 35 3.4.2.5) represent robust NbS, especially in the tropics, and particularly when paired with restoration and 36 conservation of terrestrial ecosystems (robust evidence) (e.g., peatlands and forests, WGIII AR6 Chapter 7, 37 Hoegh-Guldberg et al., 2019b; Duarte et al., 2020; Griscom et al., 2020). Although most of the focus on NbS 38 efficacy has been on coastal and shelf ecosystems (Section 3.6.3.2), recent advances point to an emerging 39 role of NbS beyond coastal waters in the form of area-based management tools in marine areas beyond 40 national jurisdiction (Section 3.6.2.3, Gaines et al., 2018; Pinsky et al., 2018; Crespo et al., 2020; O'Leary et 41 al., 2020; Visalli et al., 2020; Wagner et al., 2020), because sustainable fisheries and aquaculture and 42 climate-responsive MPAs have high potential to adapt (Tittensor et al., 2019). 43 44 Adaptation efforts (Sections 3.6.3.1­3.6.3.2) have three common characteristics that facilitate 45 implementation and success and contribute to climate-resilient development pathways (Chapter 18). First, 46 availability of multiple types of information (e.g., monitoring, models, climate services, Section 3.6.3.3) 47 exposes the magnitude and nature of the adaptation challenge. Well-developed observation and modelling 48 capabilities (Reusch et al., 2018) offer insights on climate-associated risks at different timescales (Cvitanovic 49 et al., 2018; Hobday et al., 2018), and this facilitates adaptation within multiple areas (e.g., industries over 50 shorter timescales, societies over longer scales) (Hobday et al., 2018). Environmental data has supported 51 building societal and political (socio-institutional) will to adopt national and subnational adaptive 52 management principles (Hobday et al., 2016b; Champion et al., 2018; McDonald et al., 2019). However, 53 incorporating IK and LK at the same time provides more diverse social-environmental insight (Section 54 3.6.3.4.1, Goeldner-Gianella et al., 2019; Petzold and Magnan, 2019; Wilson et al., 2020b). This can help 55 align adaptation solutions with cultural values and increase their legitimacy with Indigenous and local 56 communities (Chapter 1.3.2.3), achieving climate resilient development pathways (Chapter 18, Adger et al., 57 2017; Nalau et al., 2018; Peñaherrera-Palma et al., 2018; Raymond-Yakoubian and Daniel, 2018; Wamsler Do Not Cite, Quote or Distribute 3-143 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 and Brink, 2018). Second, implementation of multiple low-risk options (Hoegh-Guldberg et al., 2019a; 2 Gattuso et al., 2021) such as economic diversification (Section 3.6.2.1) can provide culturally acceptable 3 livelihood alternatives and food supplies (e.g., fishing to ecotourism and mariculture, (Froehlich et al., 2019) 4 while also providing environmental benefits (e.g., seaweed mariculture's potential carbon storage co-benefits 5 (WGIII AR6 Chapter 7, Hoegh-Guldberg et al., 2019a; Gattuso et al., 2021). Third, inclusive governance 6 that is well-aligned to the systems at risk from climate change is fundamental for effective adaptation 7 (Barange et al., 2018). Solutions implemented within polycentric governance systems (Section 3.6.3, 8 Bellanger et al., 2020) benefit from synergies between knowledge, action and socio-ecological contexts and 9 stimulate governance responses at appropriate spatial and temporal scales (Cvitanovic and Hobday, 2018). 10 Governance aligned with Indigenous structures and local structures supports successful outcomes that 11 prioritise the concerns and rights of involved communities (Section 3.6.3, Mawyer and Jacka, 2018) and 12 better leverages existing social organisation (i.e., network structures), learning processes and power 13 dynamics (Barnes et al., 2020). 14 15 There is an opportunity to improve current practices when developing new ocean and coastal adaptation 16 efforts so that they routinely contain these successful characteristics and resolve technical, economic, 17 institutional, geophysical, ecological and social constraints (Figure 3.25, Section 3.6.3.3, IPCC, 2018; Singh 18 et al., 2020). Enhancements are needed in human, technical and financial resources; regulatory frameworks 19 (Ojwang et al., 2017); political support (Rosendo et al., 2018); institutional conditions and resources for fair 20 governance (Gupta et al., 2016; Scobie, 2018); political leadership; stakeholder engagement; 21 multidisciplinary data availability (Gopalakrishnan et al., 2018); funding and public support for adaptation 22 (Cross-Chapter Box FINANCE in Chapter 17, Ford and King, 2015); and incorporating IK and LK in 23 decision making (Nalau et al., 2018; Jabali et al., 2020; Petzold et al., 2020). As climate change continues to 24 challenge ocean and coastal regions, there is high confidence associated with the benefits of developing 25 robust, equitable adaptation strategies that incorporate scientific projections, employ portfolios of low-risk 26 options, internalise IK and LK, and address social aspects of governance from international to local scales 27 (Finkbeiner et al., 2018; Gattuso et al., 2018; Miller et al., 2018; Raymond-Yakoubian and Daniel, 2018; 28 Cheung et al., 2019; Gattuso et al., 2021). 29 30 [START FAQ3.2 HERE] 31 32 FAQ3.2: Are we approaching so-called tipping points in the ocean and what can we do about it? 33 34 A tipping point is a threshold beyond which an abrupt or rapid change in a system occurs. Tipping points 35 that have already been reached in ocean systems include the melting of sea ice in the Arctic, thermal 36 bleaching of tropical coral reefs and the loss of kelp forests. Human-induced climate change will continue to 37 force ecosystems into abrupt and often irreversible change, absent strong mitigation and adaptation action. 38 39 A gradual change in water temperature or oxygen concentration can lead to a fundamental shift in the 40 structure and/or composition of an ecosystem when a tipping point is exceeded. For example, all species 41 have upper temperature limits below which they can thrive. In the tropics, prolonged warm temperatures can 42 cause fatal `bleaching' of tropical corals, leading reef ecosystems to degrade and become dominated by 43 algae. In temperate regions, marine heatwaves can kill or reduce the growth of kelp, threatening the other 44 species that depend on the tall canopy-forming marine plants for habitat. In the Arctic, rising temperatures 45 are melting sea ice, and reducing the available habitat for communities of ice-dependent species. 46 47 Once a tipping point is passed, the effects can be long-lasting and/or irreversible over timescales of decades 48 or longer. An ecosystem or a population can remain in the new state, even if the driver of the change returns 49 to previous levels. For example, once a coral reef has been affected by bleaching, it can take decades for 50 corals to grow back, even if temperatures remain below the bleaching threshold. Crossing a tipping point can 51 cause entire populations to collapse, causing local extinctions. 52 53 Tipping points are widespread across oceanic provinces and their ecosystems for climate variables like water 54 temperature, oxygen concentration and acidification. Evidence suggests that ocean tipping points are being 55 surpassed more frequently as the climate changes; scientists have estimated that abrupt shifts in communities 56 of marine species occurred over 14% of the ocean in 2015, up from 0.25% of the ocean in the 1980s. Other 57 human stresses to the ocean, including habitat destruction, overfishing, pollution and the spread of diseases, Do Not Cite, Quote or Distribute 3-144 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 combine with climate change to push marine systems beyond tipping points. As an example, nutrient 2 pollution from land together with climate change can lead to low-oxygen coastal areas referred to as "dead 3 zones". 4 5 Human communities can also experience tipping points that alter people's relationships with marine 6 ecosystem services. Indigenous Peoples and local communities may be forced to move from a particular 7 location due to sea-level rise, erosion, or loss of marine resources. Current activities that help sustain 8 Indigenous Peoples and their cultures may no longer be possible in the coming decades, and traditional diets 9 or territories may have to be abandoned. These tipping points have implications for physical and mental 10 health of marine-dependent human communities. 11 12 Adaptation solutions to the effects of ecological tipping points are rarely able to reverse their environmental 13 impacts, and instead often require human communities to transform their livelihoods in different ways. 14 Examples include diversifying income by shifting from fishing to tourism and relocating communities 15 threatened by flooding to other areas to continue their livelihoods. Tipping points are being passed already in 16 coral reefs and polar systems, and more will probably be reached in the near future, given climate-change 17 projections. Nevertheless, the chances of moving beyond additional tipping points in the future will be 18 minimised if we reduce greenhouse gas emissions, and we also act to limit other human impacts on the 19 ocean, such as overfishing and nutrient pollution. 20 21 22 23 Figure FAQ3.2: Global map with examples of tipping points that have been passed in ocean systems around the world. 24 Tipping points in ecological systems are linked to increasing impacts and vulnerability of dependent human 25 communities. SES: semi-enclosed sea, EBUS: eastern boundary upwelling system, CBC; coastal boundary current. 26 27 28 [END FAQ3.2 HERE] 29 30 31 [START FAQ3.3 HERE] Do Not Cite, Quote or Distribute 3-145 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 FAQ3.3: How are marine heatwaves affecting marine life and human communities? 3 4 Heatwaves happen in the ocean as well as in the atmosphere. Marine heatwaves (MHWs) are extended 5 periods of unusually warm ocean temperatures, relative to the typical temperatures for that location and 6 time of year. Due to climate change, the number of days with MHWs have increased by 54% over the past 7 century. MHWs cause mortalities in a wide variety of marine species, from corals to kelp to seagrasses to 8 fish to seabirds, and have consequent effects on ecosystems and industries like aquaculture and fisheries. 9 10 Extreme events in the ocean can have damaging effects on marine ecosystems and the human communities 11 that depend on them. The most common form of ocean extremes are marine heatwaves (MHWs), extended 12 periods of unusually warm ocean temperatures, which are becoming more frequent and intense due to global 13 warming. Because seawater absorbs and releases heat more slowly than air, temperature extremes in the 14 ocean are not as pronounced as over land, but they can persist for much longer, often for weeks to months 15 over areas covering hundreds of thousands of square kilometres. These MHWs can be more detrimental for 16 marine species, in comparison to land species, because marine species are usually adapted to relatively stable 17 temperatures. 18 19 A commonly used definition of MHWs is a period of at least five days whose temperatures are warmer than 20 90% of the historical records for that location and time of year. MHWs are described by their abruptness, 21 magnitude, duration, intensity, and other metrics. In addition, targeted methods are used to characterize 22 MHWs that threaten particular ecosystems; for example, the accumulated heat stress above typical summer 23 temperatures, described by "degree heating weeks", is used to estimate the likelihood of coral bleaching. 24 25 Over the past century, MHWs have doubled in frequency, become more intense, lasted for longer and 26 extended over larger areas. MHWs have occurred in every ocean region over the past few decades, most 27 markedly in association with regional climate phenomena such as the El Niño/Southern Oscillation. During 28 the 2015­2016 El Niño event, 70% of the world's ocean surface encountered MHWs. 29 30 MHWs cause mortality of a wide variety of marine species, from corals to kelp to seagrasses to fish to 31 seabirds, and they have consequent effects on ecosystems and industries such as mariculture and fisheries. 32 Warm-water coral reefs, estuarine seagrass meadows and cold-temperate kelp forests are among the 33 ecosystems most threatened by MHWs since they are attached to the seafloor (FAQ 3.2). Unusually warm 34 temperatures cause bleaching and associated death of warm-water corals, which can lead to shifts to low- 35 diversity or algae-dominated reefs, changes in fish communities, and deterioration of the physical reef 36 structure, which causes habitat loss and increases the vulnerability of nearby shorelines to large-wave events 37 and sea-level rise. Since the early 1980s, the frequency and severity of mass coral bleaching events have 38 increased sharply worldwide. For example, from 2016 through 2020, the Great Barrier Reef experienced 39 mass coral bleaching three times in five years. 40 41 Mass loss of kelp from MHWs effects on the canopy-forming species has occurred across ocean basins, 42 including the coasts of Japan, Canada, Mexico, Australia and New Zealand. In southern Norway and the 43 northeast U.S., mortality from MHWs contributed to the decline of sugar kelp over the last two decades, and 44 to the spread of turf algal ecosystems that prevent recolonisation by the original canopy-forming species. 45 46 One of the largest and longest-duration MHWs, nicknamed the 'Blob,' occurred in the Northeast Pacific 47 Ocean, extending from California north towards the Bering Sea, from 2013 through 2015. Warming from the 48 MHW persisted into 2016 off the U.S. West Coast and into 2018 in the deeper waters of a Canadian fjord. 49 The consequent effects of this expansive MHW included widespread shifts in abundance, distribution and 50 nutritional value of invertebrates and fish, a bloom of toxic algae off the US west coast that impacted 51 fisheries, the decline of California kelp forests that contributed to the collapse of the abalone fishery, and 52 mass mortality of seabirds. 53 54 The projected increase in the frequency, severity, duration, and areal extent of MHWs threaten many marine 55 species and ecosystems. MHWs may exceed the thermal limits of species, and they may occur too frequently 56 for the species to acclimate or for populations to recover. The majority of the world's coral reefs are 57 projected to decline and begin eroding due to more frequent bleaching-level MHWs if the world warms by Do Not Cite, Quote or Distribute 3-146 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 more than 1.5°C. Recent research suggests possible shifts to more heat-tolerant coral communities, but at the 2 expense of species and habitat diversity. Other systems, including kelp forests, are most threatened near the 3 edges of their ranges, although more research is needed into the effect of re-occurring MHWs on kelp forests 4 and other vulnerable systems. 5 6 The projected ecological impacts of MHWs threaten local communities' and Indigenous Peoples' cultures, 7 incomes, fisheries, tourism, and, in the case of coral reefs, shoreline protection from waves. High-resolution 8 forecasts and early-warning systems, currently most advanced for coral reefs, can help people and industries 9 prepare for MHWs and also collect data on their effects. Identifying and protecting locations and habitats 10 with reduced exposure to MHWs is a key scientific endeavour. For example, corals may be protected from 11 MHWs in tidally-stirred waters or in reefs where cooler water upwells from subsurface. Marine protected 12 areas and no-take zones, in addition to terrestrial protection surrounding vulnerable coastal ecosystems, 13 cannot prevent MHWs from occurring. But, depending on the location and adherence by people to 14 restrictions on certain activities, the cumulative effect of other stressors on vulnerable ecosystems can be 15 reduced, potentially helping to enhance the rate of recovery of marine life. 16 17 18 19 20 Figure FAQ 3.3: Impact pathway of a massive extreme marine heatwave, the NW Pacific "Blob," from causal 21 mechanisms, to initial effects, resulting non-linear effects, and the consequent impacts for humans. Lessons learnt from 22 the "Blob" include the need to advance seasonal forecasts, real-time predictions, monitoring responses, education, 23 possible fisheries impacts and adaptation. 24 25 26 [END FAQ3.3 HERE] 27 28 29 [START FAQ3.4 HERE] 30 Do Not Cite, Quote or Distribute 3-147 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 FAQ3.4: Which industries and jobs are most vulnerable to the impacts of climate change in the 2 oceans? 3 4 The global ocean underpins human well-being through the provision of resources that directly and indirectly 5 feed and employ many millions of people. In many regions, climate change is degrading ocean health and 6 altering stocks of marine resources. Together with over-harvesting, climate change is threatening the future 7 of the sustenance provided to Indigenous Peoples, the livelihoods of artisanal fisheries, and marine-based 8 industries including tourism, shipping and transportation. 9 10 The ocean is the lifeblood of the planet. In addition to regulating planetary cycles of carbon, water and heat, 11 the ocean and its vast resources support human livelihoods, cultural practices, jobs and industries. The 12 impacts of climate change on the ocean can influence human activities and employment by altering resource 13 availability, spreading pathogens, flooding shorelines and degrading ocean ecosystems. Fishing and 14 mariculture are highly exposed to change. The global ocean and inland waters together provide more than 15 3.3 billion people at least 20% of the protein they eat and provide livelihoods for 60 million people. Changes 16 in the nutritional quality or abundance of food from the oceans could influence billions of people. 17 18 Substantial economic losses for fisheries resulting from recent climate-driven harmful algal blooms and 19 marine pathogen outbreaks have been recorded in Asia, North America and South America. A 2016 event in 20 Chile caused an estimated loss of 800 million USD in the farmed salmon industry and led to regional 21 government protests. The recent closure of the U.S. Dungeness crab and razor clam fishery due to a climate- 22 driven algal bloom harmed 84% of surveyed residents from 16 California coastal communities. Fishers and 23 service industries that support commercial and recreational fishing experienced the most substantial 24 economic losses, and fishers were the least able to recover their losses. This same event also disrupted 25 subsistence and recreational fishing for razor clams, important activities for Indigenous Peoples and local 26 communities in the Pacific Northwest of the U.S.A. 27 28 Other goods from the ocean, including non-food products like dietary supplements, food preservatives, 29 pharmaceuticals, biofuels, sponges and cosmetic products, as well as luxury products like jewellery coral, 30 cultured pearls, and aquarium species, will change in abundance or quality due to climate change. For 31 instance, ocean warming is endangering the "candlefish" ooligan (Thaleichthys pacificus), whose oil is a 32 traditional food source and medicine of Indigenous Peoples of the Pacific Northwest of North America. 33 Declines in tourism and real estate values have also been recorded in the United States, France, and England 34 associated with climate-driven harmful algal blooms. 35 36 Small-scale fisheries livelihoods and jobs are the most vulnerable to climate-driven changes in marine 37 resources and ecosystem services. The abundance and composition of their harvest depend on suitable 38 environmental conditions and on Indigenous knowledge and local knowledge developed over generations. 39 Large-scale fisheries, though still vulnerable, are more able to adapt to climate change due to greater 40 mobility and greater resources for changing technologies. These fisheries are already adapting by broadening 41 catch diversity, increasing their mobility to follow shifting species, and changing gear, technology and 42 strategies. Adaptation in large-scale fisheries, however, is at times constrained by regulations and 43 governance challenges. 44 45 Jobs, industries and livelihoods which depend on particular species or are tied to the coast can also be at risk 46 to climate change. Species-dependent livelihoods (e.g., a lobster fishery or oyster farm) are vulnerable due to 47 a lack of substitutes if the fished species are declining, biodiversity is reduced, or mariculture is threatened 48 by climate change or ocean acidification. Coastal activities and industries ranging from fishing (e.g., 49 gleaning on a tidal flat) to tourism to shipping and transportation are also vulnerable to sea-level rise and 50 other climate-change impacts on the coastal environment. The ability of coastal systems to protect the 51 shoreline will decline due to sea-level rise and simultaneous degradation of nearshore systems including 52 coral reefs, kelp forests and coastal wetlands. 53 54 The vulnerability of communities to losses in marine ecosystem services varies within and among 55 communities. Tourists seeking to replace lost cultural services can adapt by engaging in the activity 56 elsewhere. But communities who depend on tourism for income or who have strong cultural identity linked 57 to the ocean have a more difficult time. Furthermore, climate-change impacts exacerbate existing inequalities Do Not Cite, Quote or Distribute 3-148 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 already experienced by some communities, including Indigenous Peoples, Pacific Island countries and 2 territories, and marginalized peoples, like migrants and women in fisheries and mariculture. These inequities 3 increase the risk to their fundamental human rights by disrupting livelihoods and food security, while leading 4 to loss of social, economic, and cultural rights. These maladaptive outcomes can be avoided by securing 5 tenure and access rights to resources and territories for all people depending on the ocean, and by supporting 6 decision-making processes that are just, participatory and equitable. 7 8 A key adaptation solution is improving access to credit and insurance in order to buffer against variability in 9 resource access and abundance. Further actions that decrease social and institutional vulnerability are also 10 important, such as inclusive decision-making processes, access to resources and land to Indigenous Peoples, 11 and participatory approaches in management. For the fishing industry, international fisheries agreements and 12 investing in sustainable mariculture and fisheries reforms is often recommended. Immediate adaptations to 13 other challenges, such as harmful algal blooms, frequently include fishing-area closures. These can be 14 informed by early-warning forecasts, public communications, and education. These types of adaptations are 15 more effective when built on trusted relationships and effective coordination among involved parties, and are 16 inclusive of the diversity of actors in a coastal community. 17 18 19 20 Figure FAQ3.4: Illustration that identifies vulnerable groups and stresses the hazards or impacts over coastal and ocean 21 systems. 22 Do Not Cite, Quote or Distribute 3-149 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 [END FAQ3.4 HERE] 3 4 5 [START FAQ3.5 HERE] 6 7 8 FAQ3.5: How can nature-based solutions, including marine protected areas, help us to adapt to 9 climate driven changes in the oceans? 10 11 Coastal habitats like mangroves or vegetated dunes protect coastal communities from sea-level rise and 12 storm surges, while supporting fisheries, sequestering carbon and providing other ecosystem services as 13 well. Efforts to restore, conserve and/or recover these natural habitats help people confront the impacts of 14 climate change. These marine nature-based solutions like marine protected areas, habitat restoration and 15 sustainable fisheries are cost-effective and provide myriad benefits to society. 16 17 In the oceans, nature-based solutions comprise attempts to recover, restore or conserve coastal and marine 18 habitats to reduce the impacts of climate change on nature and society. Marine habitats such as seagrasses 19 and coral reefs provide services like food and flood regulation, in the same way as forests do so on land. 20 Coastal habitats like mangroves or vegetated dunes protect coastal communities from sea-level rise and 21 storm surges, while supporting fisheries, recreational and aesthetic services as well. Seagrasses, coral reefs 22 and kelp forests also provide important benefits that help humans adapt to climate change, including 23 sustainable fishing, recreation and shoreline protection services. By recognizing these services and benefits 24 of the ocean, nature-based solutions can improve the quality and integrity of the marine ecosystems. 25 26 Nature-based solutions offer a wide range of potential benefits, including protecting ecosystem services, 27 supporting biodiversity and mitigating climate change. Coastal and marine examples include marine 28 protected areas, habitat restoration, habitat development and maintaining sustainable fisheries. While local 29 communities with limited resources might find nature-based solutions challenging to implement, they are 30 generally "no-regret" options, which bring societal and ecological benefits regardless of the level of climate 31 change. 32 33 Carefully designed and placed marine protected areas, especially when they exclude fishing, can increase 34 resilience to climate change by removing additional stressors on ecosystems. While marine protected areas 35 do not prevent extreme events like marine heatwaves (FAQ3.3), they can provide marine plants and animals 36 with a better chance to adapt to a changing climate. Current marine protected areas, however, are often too 37 small, too poorly connected and too static to account for climate-induced shifts in the range of marine 38 species. Marine protected area networks that are large, are connected, have adaptable boundaries, and are 39 designed following systematic analysis of future climate projections can better support climate resilience. 40 41 Habitat restoration and development in coastal systems can support biodiversity, protect communities from 42 flooding and erosion, support the local economy, and enhance the livelihoods and wellbeing of coastal 43 peoples. Restoration of mangroves, saltmarshes and seagrass meadows provide effective ways to remove 44 carbon dioxide from the atmosphere and at the same time to protect coasts from the impacts of storms and 45 sea-level rise. Active restoration techniques that target heat-resistant individuals or species are increasingly 46 recommended for coral reefs and kelp forests, which are highly vulnerable to marine heatwaves and climate 47 change. 48 49 Sustainable fishing is also seen as a nature-based solution because managing marine commercial species 50 within sustainable limits maximizes the catch and food production, thus contributing to the UN's Sustainable 51 Development Goal 2 ­ Zero Hunger. Currently, the oceans provide 17% of the animal protein eaten by the 52 global population, but the contribution could be larger if fisheries were managed sustainably. Aquaculture, 53 such as oyster farming, can be efficient and sustainable means of food production and also provide additional 54 benefits like shoreline protection. Through nature-based solutions that conserve and restore marine habitats 55 and species, we can sustain marine biodiversity, respond to climate change, and provide benefits to society. Do Not Cite, Quote or Distribute 3-150 Total pages: 236 FINAL DRAFT Chapter 3 IPCC WGII Sixth Assessment Report 1 2 3 Figure FAQ3.5: Contributions of nature-based solutions in the oceans to the Sustainable Development Goals. The 4 icons in the bottom show the Sustainable Development Goals to which nature-based solutions in the ocean possibly 5 contribute. [Placeholder figure -- authoritative version on FMS] 6 7 8 [END FAQ3.5 HERE] 9 10 11 Acknowledgements 12 We acknowledge the kind contributions of Rita Erven (GEOMAR Helmholtz Centre for Ocean Research 13 Kiel, Germany), Miriam Seifert (Alfred Wegener Institute for Polar and Marine Research, Germany), 14 Sebastian Rokitta (Alfred Wegener Institute for Polar and Marine Research, Germany), Amy Marie 15 Campbell (National Oceanography Centre, Southampton/Centre for Environment, Fisheries and Aquaculture 16 Science, United Kingdom), Mariana Castaneda-Guzman (Virginia Polytechnic Institute and State University, 17 USA), Stephen Goult (Plymouth Marine Laboratory/National Centre for Earth Observation, United 18 Kingdom), Josh Douglas (Plymouth Marine Laboratory, United Kingdom), Carl Reddin (Museum für 19 Naturkunde, Berlin, Germany) and the PML Communications and Graphics Team (Plymouth Marine 20 Laboratory, United Kingdom) who assisted in drafting figures and tables. 21 22 23 Do Not Cite, Quote or Distribute 3-151 Total pages: 236