Editing IPCC:AR6/WGII/Chapter-15 (section)

=== 15.3.3 Observed Impacts and Projected Risks on Natural Systems ===

<div id="h2-5-siblings" class="h2-siblings"></div>

<div id="15.3.3.1" class="h3-container"></div>

<span id="impacts-on-marine-and-coastal-systems"></span>
==== 15.3.3.1 Impacts on Marine and Coastal Systems ====

<div id="h3-1-siblings" class="h3-siblings"></div>

<div id="15.3.3.1.1" class="h4-container"></div>

<span id="submergence-and-flooding-of-islands-and-coastal-areas"></span>
===== 15.3.3.1.1 Submergence and flooding of islands and coastal areas =====

<div id="h4-5-siblings" class="h4-siblings"></div>

Recent studies confirmed that observed ESL events causing extensive flooding generally resulted from compound effects, including the combination of SLR ( [[IPCC:Wg2:Chapter:Chapter-3#3.2.2.2|Section 3.2.2.2]] and Cross-Chapter Box SLR in Chapter 3) with ETCs, TCs and tropical depressions (WGI AR6 Sections 11.7.1 and 11.7.2, Seneviratne, 2021), ENSO-related high-water levels associated with high or spring tide and/or local human disturbances amplifying impacts ( ''high confidence'' ). For example, the major floods that occurred in 1987 and 2007 in the Maldives involved the combination of distant-source swells and high spring tides and the settlement of reclaimed low-lying areas (Box 15.1; [[#Wadey--2017|Wadey et al., 2017]] ). In the Tuamotu atolls, French Polynesia, the 1996 and 2011 floods were due to the combination of distant-source swells causing lagoon filling and the obstruction of inter-islet channels by human-built structures ( [[#Canavesio--2019|Canavesio, 2019]] ). In 2011, the flooding of the lagoon-facing coast of Majuro Atoll, Marshall Islands, resulted from the combination of high sea levels occurring during La Niña conditions and seasonally high tides ( [[#Ford--2018|Ford et al., 2018]] ). Another example is the widespread flooding caused by distant TC Pam (2015) in Kiribati and Tuvalu, which was attributed to the strong swell generated, the long duration of the event and exceptionally high regional sea levels ( [[#Hoeke--2021|Hoeke et al., 2021]] ). On high tropical islands, major floods often occurred during TC events, due to the cumulative effects of storm surge and river flooding, the impacts of which were exacerbated by human-induced changes to natural processes in urban areas. This, for example, occurred in 2014 (TC Bejisa) in Reunion Island, France, in a harbour area favourable to water accumulation ( [[#Duvat--2016|Duvat et al., 2016]] ); in 2015 (TC Pam) in Port Vila, Vanuatu, where urbanisation and human-induced changes to the river exacerbated flooding (Rey et al., 2017); and in 2017 (TC Irma) in Saint-Martin, Caribbean, where urbanisation had the same effect ( [[#Rey--2019|Rey et al., 2019]] ). Successive tropical depressions generating heavy rains were also involved in extensive flooding, for example, in 2012 in Fiji ( [[#Kuleshov--2014|Kuleshov et al., 2014]] ) and in 2014 in the Solomon Islands ( [[#Ha’apio--2019|Ha’apio et al., 2019]] ).

Reconstructions of past storm surges and modelling studies assessing storm surge risk similarly highlighted high variations of risk along island coasts, due to variations in exposure, topography and bathymetry ( ''high confidence'' ). For example, the storm surge caused by TC Oli (2010) on the high volcanic island of Tubuai, French Polynesia, ranged from a few centimetres to 2.5 m, depending on coast exposure ( [[#Barriot--2016|Barriot et al., 2016]] ). Investigating the contribution of reef characteristics to variations in wave-driven flooding on Roi-Namur Island, Kwajalein Atoll, Marshall Islands, [[#Quataert--2015|Quataert et al. (2015)]] found that the coasts fronted by narrow reefs with steep fore reef slopes and smoother reef flats are the most flood-prone. Modelling studies assessing storm surge risk in Fiji ( [[#McInnes--2014|McInnes et al., 2014]] ) and Samoa ( [[#McInnes--2016|McInnes et al., 2016]] ) confirmed the influence of coast exposure and water depth on risk distribution. In Apia, Samoa, Hoeke et al. (2015, p. 1117) found ‘differences in extreme sea levels in the order of 1 m at spatial scales of less than 1 km’ and estimated (p. 1131) that a ‘1 m SLR relative to constant topography increases wave energy reaching the shore by up to 200% during storm surges.’ These studies reaffirmed the main control exerted by SLR on ESL events and associated storm surges compared to ENSO ( ''high confidence'' ). In Hawaii and the Caribbean, SLR is projected to exponentially increase flooding, with nearly every centimetre of SLR causing a doubling of the probability of flooding ( [[#Taherkhani--2020|Taherkhani et al., 2020]] ). Simulations of SLR-induced flooding resulting from the combination of (a) direct marine flooding, (b) flow reversal in drainage networks caused by extreme tide levels and (c) the elevation of groundwater levels, at Honolulu, Hawaii, highlighted the major influence of this latter component (which is the most difficult to manage), as well as the increase of the proportion of triple-mechanism flooding as sea level rises ( [[#Habel--2020|Habel et al., 2020]] ). Where coral reefs buffer flooding through wave attenuation, flooding will be further aggravated by reef decline over time ( [[#15.3.3.1.3|Section 15.3.3.1.3]] ).

Larger-scale studies confirmed that projected changes in the wave climate superimposed on SLR will rapidly increase flooding in small islands, despite highly contrasting exposure profiles between ocean sub-regions ( ''high confidence'' ) ( [[#Shope--2016|Shope et al., 2016]] ; [[#Mentaschi--2017|Mentaschi et al., 2017]] ; [[#Shope--2017|Shope et al., 2017]] ; [[#Vitousek--2017|Vitousek et al., 2017]] ; [[#Morim--2019|Morim et al., 2019]] ). In particular, [[#Vitousek--2017|Vitousek et al. (2017)]] showed that even a 5–10-cm additional SLR (expected for ~2030–2050) will double flooding frequency in much of the Indian Ocean and Tropical Pacific, while TCs will remain the main driver of (rarer) flooding in the Caribbean Sea and Southern Tropical Pacific (Figure 15.3). Some Pacific atoll islands, which already experience major floods, will ''likely'' undergo annual wave-driven flooding over their entire surface from the 2060s–2070s ( [[#Storlazzi--2018|Storlazzi et al., 2018]] ) to 2090s ( [[#Beetham--2017|Beetham et al., 2017]] ) under RCP8.5, although future reef growth may delay the onset of flooding ( ''limited evidence, low agreement'' ) (key risk KR2 in Figure 15.5).

<div id="15.3.3.1.2" class="h4-container"></div>

<span id="reef-island-destabilisation-and-coastal-erosion"></span>
===== 15.3.3.1.2 Reef island destabilisation and coastal erosion =====

<div id="h4-6-siblings" class="h4-siblings"></div>

Over the past three to five decades, shoreline changes were dominated by stability on reef islands and erosion on high islands; attribution of observed erosion to SLR and other climate change-related drivers is challenged by the complex interplay of multiple climatic, ecological and human drivers ( ''high confidence'' ). Since the 1950s–1970s, and even in regions exhibiting higher than global-averaged SLR rates, atoll islands maintained their land area ( ''high confidence).'' A literature review including 709 Indian Ocean and Pacific Ocean atoll islands showed that 73.1% of these islands were stable in area, while, respectively, 15.5% and 11.4% increased and decreased in area ( [[#Duvat--2018|Duvat, 2018]] ). The rates of change did not correlate with SLR rates, suggesting that the impact of SLR on island land area was obscured by other climate drivers and human disturbances on some islands ( ''high confidence'' ) ( [[#Kench--2015|Kench et al., 2015]] ; [[#McLean--2015|McLean and Kench, 2015]] ; [[#Duvat--2018|Duvat, 2018]] ). However, reef island disappearance and reduction in land area was clearly observed in New Caledonia and the Solomon Islands, and was attributed to the synergistic interactions of gradual SLR with stronger trade winds causing higher sea levels and local tectonics in the Solomon Islands ( [[#Albert--2016|Albert et al., 2016]] ; [[#Garcin--2016|Garcin et al., 2016]] ). Despite important knowledge gaps on coastal erosion in high tropical islands, recent studies confirmed increasing shoreline retreat and beach loss over the past decades, mainly due to TC and ETC waves and human disturbances ( ''high confidence'' ) (e.g., in the Caribbean region: Anguilla, Saint-Kitts, Nevis, Montserrat, Dominica and Grenada ( [[#Cambers--2009|Cambers, 2009]] ; [[#Reguero--2018|Reguero et al., 2018]] )), and Pacific (Hawaii ( [[#Romine--2013|Romine and Fletcher, 2013]] ); Tubuai, French Polynesia ( [[#Salmon--2019|Salmon et al., 2019]] )) and Indian Oceans (Anjouan, Comoros ( [[#Ratter--2016|Ratter et al., 2016]] ).

Despite storm-induced erosion prevailing along some shoreline sections, recent studies reaffirmed the contribution of TC and ETC waves to coastal and reef island vertical building through massive reef-to-island sediment transfer ''(high confidence'' ). For example, TC Ophelia (1958) and Category 5 TC Fantala (2016), which eroded the islands of Jaluit Atoll, Marshall Islands ( [[#Ford--2016|Ford and Kench, 2016]] ), and Farquhar Atoll, Seychelles ( [[#Duvat--2017c|Duvat et al., 2017c]] ), respectively, also contributed to island and beach expansion. Likewise, tropical depressions can have constructional effects, as reported on Fakarava Atoll, French Polynesia ( [[#Duvat--2020b|Duvat et al., 2020b]] ). On Saint-Martin/Sint Maarten and Saint-Barthélemy, the 2017 hurricanes, which caused marked shoreline retreat at most beach sites, also enabled beach formation and beach ridge development along some natural coasts ( [[#Duvat--2019a|Duvat et al., 2019a]] ; [[#Pillet--2019|Pillet et al., 2019]] ). Similarly, El Niño and La Niña were involved in rapid and highly contrasting shoreline changes ( ''high confidence'' ), including reef island accretion in the Ryukyu Islands, Japan ( [[#Kayanne--2016|Kayanne et al., 2016]] ), beach shifts on Maiana and Aranuka atolls, Kiribati (Rankey, 2011), and beach erosion on Hawaii, USA ( [[#Barnard--2015|Barnard et al., 2015]] ). These contrasting shoreline responses were, respectively, due to coral reef degradation from past bleaching events providing material to islands, wave directional shifts, and increased wave energy. The role of bleaching events in increasing short-term sediment generation in atoll contexts was confirmed by a study conducted on Gaafu Dhaalu Atoll, Maldives, which reported an increase of sediment production from ~0.5 kg CaCO 3 m –2 yr -1 to ~3.7 kg CaCO 3 m –2 yr -1 between 2016 (pre-bleaching) and 2019 (bleaching + 3 years) ( [[#Perry--2020|Perry et al., 2020]] ).

There is ''high confidence'' that accelerating SLR and increased wave height will affect the geomorphology of reef islands ( [[#Baldock--2015|Baldock et al., 2015]] ; [[#Costa--2019|Costa et al., 2019]] ; [[#Tuck--2019|Tuck et al., 2019]] ) and coastal systems on high islands ( [[#Grady--2013|Grady et al., 2013]] ; [[#Barnard--2015|Barnard et al., 2015]] ; [[#Bindoff--2019|Bindoff et al., 2019]] ), and that the responses of these systems will highly depend on changes in boundary conditions (wave regime and direction, exposure to extreme events, impacts of ocean warming and acidification on supporting ecosystems, bathymetry and reef flat roughness) and the degree of disturbance of their natural dynamics by human activities ( [[#Smithers--2014|Smithers and Hoeke, 2014]] ; [[#McLean--2015|McLean and Kench, 2015]] ; [[#Bheeroo--2016|Bheeroo et al., 2016]] ; [[#Ratter--2016|Ratter et al., 2016]] ; [[#Shope--2016|Shope et al., 2016]] ; [[#Duvat--2017a|Duvat et al., 2017a]] ; [[#Kench--2017|Kench and Mann, 2017]] ; [[#Kench--2018|Kench et al., 2018]] ; [[#Duvat--2019a|Duvat et al., 2019a]] ). Reef islands and beach and beach-dune systems that are not disturbed by human activities are, respectively, expected to migrate lagoonward ( [[#Webb--2010|Webb and Kench, 2010]] ; [[#Albert--2016|Albert et al., 2016]] ; [[#Beetham--2017|Beetham et al., 2017]] ; [[#Costa--2019|Costa et al., 2019]] ; [[#Tuck--2019|Tuck et al., 2019]] ) and landward ( [[#Bindoff--2019|Bindoff et al., 2019]] ), and to also experience increased erosion as well as changes in configuration, volume and elevation ( [[#Kench--2017|Kench and Mann, 2017]] ; [[#Tuck--2019|Tuck et al., 2019]] ) ( [[#Bramante--2020|Bramante et al., 2020]] ; [[#Kane--2020|Kane and Fletcher, 2020]] ). Small reef islands and narrow coastal systems affected by human disturbances will increasingly be at risk of disappearance due to SLR (KR2 in Figure 15.5), enhanced sediment loss caused by extreme events ( [[#Duvat--2019a|Duvat et al., 2019a]] ) and/or human activities ( ''high confidence'' ), as reported in Hawaii ( [[#Romine--2013|Romine and Fletcher, 2013]] ), Puerto Rico ( [[#Jackson--2012|Jackson et al., 2012]] ), Sicily ( [[#Anfuso--2012|Anfuso et al., 2012]] ), and Takuu, Papua New Guinea ( [[#Mann--2014|Mann and Westphal, 2014]] ). SLR will also increase coastal erosion in the Mediterranean Sea, (e.g., in the Aegean Archipelago, Greece ( [[#Monioudi--2017|Monioudi et al., 2017]] ), and Mallorca, Spain ( [[#Enríquez--2017|Enríquez et al., 2017]] ).

<div id="_idContainer021" class="Figure"></div>

[[File:a1cb2010756818ae2f65cc852f12fad6 IPCC_AR6_WGII_Figure_15_005.png]]

'''Figure 15.5 |''' '''Key risks in small islands'''. KR1 to KR8 are interconnected as shown by ''arrows'' , which causes risk accumulation leading to reduced island habitability. The main interconnections are shown in this figure: for example, loss of marine and coastal and terrestrial biodiversity and ecosystem services (KR1 and KR3, respectively) are projected to cause the submergence of reef islands (KR2), water insecurity (KR4), destruction of settlements and infrastructure (KR5), degradation of human health and well-being (KR6), economic decline and livelihood failure (KR7), and loss of cultural resources and heritage (KR8). Importantly, KRs result from both direct effects (e.g., decrease in rainfall will increase water insecurity) and indirect effects (e.g., loss of terrestrial biodiversity and ecosystem services will increase water insecurity, which will in turn cause the degradation of human health and well-being).

<div id="15.3.3.1.3" class="h4-container"></div>

<span id="impacts-on-marine-and-coastal-ecosystems"></span>
===== 15.3.3.1.3 Impacts on marine and coastal ecosystems =====

<div id="h4-7-siblings" class="h4-siblings"></div>

Loss of marine and coastal biodiversity and ecosystem services is a key risk in small islands (see KR1 in Figure 15.5). Coral bleaching caused by elevated water temperatures is the most visible and widespread manifestation of a climate change impact on coastal ecosystems in most small islands but is far from being the only one (Sections 3.4.2.1 and [[IPCC:Wg2:Chapter:Chapter-5#5.3|Section 5.3.4]] ; [[#Spalding--2015|Spalding and Brown, 2015]] ; [[#Hoegh-Guldberg--2017|Hoegh-Guldberg et al., 2017]] ; [[#IPCC--2018|IPCC, 2018]] ; [[#Bindoff--2019|Bindoff et al., 2019]] ; [[#Sully--2019|Sully et al., 2019]] ). Severe coral bleaching, together with declines in coral abundance have been documented in many small islands, especially those in the Pacific Ocean and Indian Ocean (e.g., Guam, Fiji, Palau, Vanuatu, Chagos, Comoros, Mauritius, Seychelles, and the Maldives ( ''high confidence'' ) (Box 15.1; [[#Golbuu--2007|Golbuu et al., 2007]] ; [[#Woesik--2012|Woesik et al., 2012]] ; [[#Perry--2017|Perry and Morgan, 2017]] ; [[#Hughes--2018|Hughes et al., 2018]] ). During severe bleaching events, not only do reefs lose a significant amount of live coral cover, but they also experience a decrease in growth potential, and thus reef erosion surpasses reef accretion ( [[#Perry--2017|Perry and Morgan, 2017]] ). Median return time between two severe bleaching events has diminished steadily since 1980 and is now only 6 years (e.g., [[#Hughes--2017b|Hughes et al., 2017b]] ; [[#Hughes--2018|Hughes et al., 2018]] ) and is often associated with warm phase of ENSO events ( ''high confidence'' ) ( [[#Lix--2016|Lix et al., 2016]] ). Modelling of both bleaching and ocean acidification effects under future climate scenarios suggested that some Pacific small islands (e.g., Nauru, Guam, Northern Marianas Islands) will experience conditions that cause severe bleaching on an annual basis before 2040 and that 90% of the world reefs are projected to experience conditions that result in severe bleaching annually by 2055 ( ''medium confidence'' ) ( [[#van%20Hooidonk--2016|van Hooidonk et al., 2016]] ). Models are currently predicting the large-scale loss of coral reefs by mid-century under even low-emission scenarios. Even achieving emission reduction targets consistent with the ambitious goal of 1.5°C of global warming under the Paris Agreement will result in the further loss of 70–90% of reef-building corals compared to today, with 99% of corals being lost under warming of 2°C or more above the pre-industrial period ( ''high confidence'' ) ( [[#Hoegh-Guldberg--2018|Hoegh-Guldberg et al., 2018]] ).

Satellite data and local field studies at 3351 sites in 81 countries including small islands show that not all coral reefs are equally exposed to severe temperature stress events, and even similar coral reefs exposed to similar conditions show local and regional variation and species-specific responses ( [[#Sully--2019|Sully et al., 2019]] ). There is great variability in terms of sensitivity of corals to climate change, as also demonstrated in the Comoros Archipelago ( [[#Cowburn--2018|Cowburn et al., 2018]] ), in the Pacific ( [[#Fox--2019|Fox et al., 2019]] ; [[#Mollica--2019|Mollica et al., 2019]] ; [[#Romero-Torres--2020|Romero-Torres et al., 2020]] ) and globally ( [[#Sully--2019|Sully et al., 2019]] ; [[#McClanahan--2020|McClanahan et al., 2020]] ). It has been hypothesised that low-latitude tropical reefs bleached less than those in higher latitudes because: (a) of the geographical differences in species composition, (b) of the higher genotypic diversity at low latitudes, and (c) some corals were pre-adapted to thermal stress because of consistently warmer temperatures at low latitude prior to thermal stress events ( [[#Sully--2019|Sully et al., 2019]] ). However, latitudinal variation was not reported in other global surveys of coral bleaching occurrence ( [[#Donner--2017|Donner et al., 2017]] ; [[#Hughes--2017a|Hughes et al., 2017a]] ; [[#Hughes--2017b|Hughes et al., 2017b]] ; [[#McClanahan--2019|McClanahan et al., 2019]] ). [[#Ainsworth--2016|Ainsworth et al. (2016)]] and [[#Ateweberhan--2013|Ateweberhan et al. (2013)]] showed that coral bleaching can be mitigated by pre-exposure to elevated temperatures. Regionally, recovery is also highly variable. While some reefs in the Seychelles and Maldives were shown to recover to pre-disturbance levels of coral cover after previous bleaching events (Box 15.1; [[#Pisapia--2016|Pisapia et al., 2016]] ; [[#Koester--2020|Koester et al., 2020]] ), other reefs underwent seemingly permanent regime shifts toward domination by fleshy macro algae ( [[#Graham--2015|Graham et al., 2015]] ), or major declines in carbonate budgets, and thus the capacity of reefs to sustain vertical growth under rising sea levels ( [[#Perry--2017|Perry and Morgan, 2017]] ).

Despite their vital social and ecological value, substantial declines in seagrass communities have been documented in many small islands ( [[IPCC:Wg2:Chapter:Chapter-3#3.4.2.5|Section 3.4.2.5]] ; [[#Arias-Ortiz--2018|Arias-Ortiz et al., 2018]] ; [[#Kendrick--2019|Kendrick et al., 2019]] ; [[#Brodie--2020|Brodie et al., 2020]] ), including Fiji ( [[#Joseph--2019|Joseph et al., 2019]] ), Reunion Island ( [[#Cuvillier--2017|Cuvillier et al., 2017]] ), Bermuda, Cayman Islands, US Virgin Islands ( [[#Waycott--2009|Waycott et al., 2009]] ), Kiribati ( [[#Brodie--2020|Brodie et al., 2020]] ), Federated States of Micronesia, and Palau ( [[#Short--2016|Short et al., 2016]] ), but attribution of such declines to climatic influences remains weak ( ''low confidence'' ). The impact of climate change on seagrasses goes beyond the loss of seagrass but includes acceleration of seagrass decomposition ( [[#Kelaher--2018|Kelaher et al., 2018]] ), palatability ( [[#Jimenez-Ramos--2017|Jimenez-Ramos et al., 2017]] ) and the cumulative effect of warming and eutrophication ( [[#Ontoria--2019|Ontoria et al., 2019]] ). Seagrasses face a multitude of threats including physical disturbance and direct damage caused by rapidly growing human populations, declines in water quality, and coastal erosion ( [[#Short--2016|Short et al., 2016]] ). Experimental studies have shown increased mortality, leaf necrosis, and respiration when seagrasses are exposed to higher-than-normal temperatures ( [[#Hernan--2017|Hernan et al., 2017]] ). As such, seagrass meadows growing near the edge of their thermal tolerance are at risk from rising temperatures ( [[#Pedersen--2016|Pedersen et al., 2016]] ). In the Mediterranean, seagrass meadows are already showing signs of regression, which may have been aggravated by climate change ( ''high confidence'' ). Some studies suggest seagrasses have potential for acclimation and adaptation ( [[#Duarte--2018|Duarte et al., 2018]] ; [[#Ruiz--2018|Ruiz et al., 2018]] ; [[#Beca-Carretero--2020|Beca-Carretero et al., 2020]] ). Chefaoui et al. (2018) attempted to forecast the distribution of two seagrasses in the future, including around the islands of Cyprus, Malta, Sicily and the Balearic Islands. Under the worst-case scenario, ''Posidonia oceanica'' was projected to lose 75% of suitable habitat by 2050. Conversely, it has been suggested that seagrasses could actually benefit from an increase in anthropogenic CO 2 because of increased growth and photosynthesis ( [[#Hopley--2007|Hopley et al., 2007]] ; [[#Waycott--2011|Waycott et al., 2011]] ; [[#Sunday--2016|Sunday et al., 2016]] ; [[#Repolho--2017|Repolho et al., 2017]] ). However, [[#Collier--2017|Collier et al. (2017)]] argued that when faced with increased heat waves, thermal stress will rarely be offset by the benefit of elevated CO 2 and therefore that the widespread belief that seagrasses will be a ‘winner’ under future climate change conditions seems unlikely ''(low confidence'' ).

Since 2011, the Caribbean region has been experiencing unprecedented influxes of the pelagic seaweed ''Sargassum'' . These extraordinary sargassum ‘blooms’ have resulted in mass strandings of sargassum throughout the Lesser Antilles, with significant damage to coastal habitats, mortality of seagrass beds and associated corals ( [[#van%20Tussenbroek--2017|van Tussenbroek et al., 2017]] ), as well as consequences for fisheries and tourism. Whether or not such events are related to long-term climate change remains unclear; however, it has been suggested that the influx may be related to strong Amazon discharge, enhanced West African upwelling, together with rising seawater temperatures in the Atlantic ( ''low confidence'' ) ( [[#Oviatt--2019|Oviatt et al., 2019]] ; [[#Wang--2019|Wang et al., 2019]] ). Since 2011, the Pacific atoll nation of Tuvalu has also been affected by algal blooms, the most recent being a large growth of ''Sargassum'' on the main atoll of Funafuti, and this phenomenon has been related to anthropogenic eutrophication and high seawater temperatures ( [[#De%20Ramon%20N’Yeurt--2014|De Ramon N’Yeurt and Iese, 2014]] ).

Mangroves face serious risks from deforestation and unsustainable coastal development ( [[IPCC:Wg2:Chapter:Chapter-3#3.4.2.5|Section 3.4.2.5]] ; [[#Gattuso--2015|Gattuso et al., 2015]] ). Large-scale die-offs around many small islands suggest that mangroves face increased risks from climate change ( [[#Sippo--2018|Sippo et al., 2018]] ). Mangrove seaward edge retreat has been demonstrated in American Samoa and at Tikina Wai in Fiji, in Bermuda, West Papua, Grand Cayman and attributed to long-term SLR or tectonic subsidence ( [[#Ellison--1993|Ellison, 1993]] ; [[#Ellison--2005|Ellison, 2005]] ; [[#Gilman--2007|Gilman et al., 2007]] ; [[#Ellison--2015|Ellison and Strickland, 2015]] ). Inundation-related mortality of mangroves could, in theory, be mitigated if mangrove substrates can ‘keep up’ with rising sea level by accretion. Pacific Island studies using radionuclides (e.g., 210Pb, 137Cs) have suggested that most mangroves are keeping up with current rates of SLR ( [[#Alongi--2008|Alongi, 2008]] ; [[#MacKenzie--2016|MacKenzie et al., 2016]] ), while surface elevation tables (SETs) suggest otherwise. [[#Lovelock--2015|Lovelock et al. (2015)]] reported that nearly 70% of the mangroves monitored with SETs are not keeping up with current SLR rates. If SLR exceeds 6 mm yr –1 , mangroves may be unable to maintain their elevation relative to sea level, a threshold likely to be surpassed in the next 30 years under high emission scenarios ( [[#Ellison--1993|Ellison, 1993]] ; [[#Saintilan--2020|Saintilan et al., 2020]] ). In these worst-case scenarios, flooding would result in tree, root and rhizome death and an abrupt change in elevation through peat collapse ( [[#Krauss--2010|Krauss et al., 2010]] ; [[#Lang’at--2014|Lang’at et al., 2014]] ), creating a positive feedback loop between SLR and elevation loss. Geomorphology, hydrology, tidal range and suspended sediments are important factors that will determine if mangroves will survive increased rates of SLR ( [[#Lovelock--2015|Lovelock et al., 2015]] ; [[#Sasmito--2015|Sasmito et al., 2015]] ; [[#Rogers--2019|Rogers et al., 2019]] ). TCs can cause extensive damage to mangroves ( [[#Short--2016|Short et al., 2016]] ). While immediate physical damage is often considerable, trees can sometimes recover by re-foliating, re-sprouting or regenerating ( [[#Kauffman--2010|Kauffman and Cole, 2010]] ). Examples of substantive mangrove recovery include the regrowth of trees in the Bay Islands of Honduras following Hurricane Mitch (October 1998) ( [[#Fickert--2018|Fickert, 2018]] ) and in the Nicobar Islands, India, following the December 2004 Indian Ocean Tsunami ( [[#Nehru--2018|Nehru and Balasubramanian, 2018]] ).

Sandy beaches are an important ecosystem in small islands, with high socioeconomic as well as ecosystem services value ( [[#Ellison--2018|Ellison, 2018]] ). Turtles and many seabirds nest just above the high-water mark on sandy beaches or among sand dunes, but TCs, rising seas, storm surges and heavy rainfall as well as inappropriate coastal development can erode beaches ( [[#15.3.1|Section 15.3.1.2]] ) resulting in damage to nests and eggs ( [[#Fuentes--2011|Fuentes et al., 2011]] ). Beach-nesting turtle populations are projected to become threatened around many small islands as a result of future climate change (e.g., Bonaire – Netherlands Antilles ( [[#Fish--2005|Fish et al., 2005]] ), Bioko Island – Equatorial Guinea ( [[#Veelenturf--2020|Veelenturf et al., 2020]] ), Cyprus ( [[#Varela--2019|Varela et al., 2019]] ), Raine Island – Australia ( [[#Pike--2015|Pike et al., 2015]] )), although other populations such as those around the Cape Verde Islands are projected to remain relatively robust ( [[#Abella%20Perez--2016|Abella Perez et al., 2016]] ). Turtles are also threatened by temperature rise around some small islands as warmer temperatures on nesting beaches can lead to an unbalanced sex ratio in the population (e.g., St. Eustatius island, ( [[#Laloë--2016|Laloë et al., 2016]] )).

<div id="15.3.3.1.4" class="h4-container"></div>

<span id="marine-and-coastal-ecosystem-services"></span>
===== 15.3.3.1.4 Marine and coastal ecosystem services =====

<div id="h4-8-siblings" class="h4-siblings"></div>

Intact coral reefs ( [[#Woodhead--2019|Woodhead et al., 2019]] ), seagrass meadows ( [[#Hejnowicz--2015|Hejnowicz et al., 2015]] ) and mangroves ( [[#UNEP--2014b|UNEP, 2014b]] ) ( [[#Friess--2016|Friess, 2016]] ) provide a variety of ecosystem services that are key to island communities, including provisioning services (e.g., timber, fisheries, aquaculture), regulating services (e.g., coastal protection, carbon storage, filtering of pollutants), cultural services ( [[#Pascua--2017|Pascua et al., 2017]] ) as well as supporting community resilience ( [[#Förster--2019|Förster et al., 2019]] ). If coastal ecosystems are degraded and lost, then the benefits they provide are also lost ( [[#Oleson--2018|Oleson et al., 2018]] ; [[#Förster--2019|Förster et al., 2019]] ; [[#Brodie--2020|Brodie et al., 2020]] ). In small islands where the risk of loss to ecosystem services is high (Cross-Chapter Box DEEP in Chapter 17), many of these ecosystem services cannot be easily replaced ( ''medium confidence'' ). The beneficial role that coral reefs play in coastal protection through wave attenuation, and therefore enhancing climate resilience in small islands, has been extensively studied (e.g., [[#Elliff--2017|Elliff and Silva, 2017]] ; [[#Harris--2018|Harris et al., 2018]] ; [[#Reguero--2018|Reguero et al., 2018]] ). Indeed, it has been demonstrated that in small islands (such as the Cayman Islands, Grenada, Bahamas) averted damages as a result of protecting intact coral reefs can be considerable when expressed as a percentage of GDP ( [[#Beck--2018|Beck et al., 2018]] ). [[#Ferrario--2014|Ferrario et al. (2014)]] conducted a global meta-analysis including many small islands across the Atlantic, Pacific and Indian oceans and found that coral reefs reduce wave height by an average of 84% (and wave energy by 97%) and that reef crests alone dissipate most of this energy. Based on another meta-analysis of 69 case studies worldwide (wave heights measured before and after the habitat), [[#Narayan--2016|Narayan et al. (2016)]] observed that coral reefs, mangroves and seagrass reduced wave height by 70%, 31% and 36%, respectively (Figure 15.4) and thus perform an essential role in protecting human lives and livelihoods ( ''high confidence'' ). Post-TC studies have provided additional evidence for the protection services offered by coastal ecosystems. On some Caribbean islands (e.g., Saint-Martin/Sint Maarten) where the dense indigenous vegetation belt was preserved, the vegetative structure buffered the waves of TCs Irma and José (2017), reducing the extent of marine inundation and shoreline retreat to a 30-m-wide coastal strip against values &gt;160 m in deforested areas ( [[#Duvat--2019a|Duvat et al., 2019a]] ; [[#Pillet--2019|Pillet et al., 2019]] ). By contrast, the destruction of mangrove ecosystems, even a few trees around the fringes, can accelerate coastal erosion, as exemplified by observations in Micronesia ( [[#Krauss--2010|Krauss et al., 2010]] ; [[#Nunn--2017a|Nunn et al., 2017a]] ).

<div id="_idContainer012" class="Figure"></div>

[[File:70b5381bbe7e40eceb637fbdf4241eba IPCC_AR6_WGII_Figure_15_004.png]]

'''Figure 15.4 |''' '''Ridge-to-reef interrelated protection services delivered by ecosystems on small islands.''' On small islands, terrestrial, coastal and marine ecosystems are interconnected and interdependent, with each ecosystem contributing towards maintaining the health of the others. Together, these ecosystems provide protection services against natural hazards (including flooding, erosion, landslides, mudflows, glacial melting and sedimentation) to human populations living on islands. As a consequence, the degradation of one or more of these ecosystems significantly reduces the protection services provided by this continuum of ecosystems. Conversely, the protection or restoration of one or more of these ecosystems also provides benefits to the other ecosystems and enhances the protection services provided to island inhabitants. See Box [https://www.ipcc.ch/chapter/15#CCP1.1 CCP1.1] for more details.

As corals, mangroves and seagrasses disappear, so do fish and other dependent organisms that directly benefit industries such as ecotourism and fisheries ( ''high confidence'' ) ( [[#Graham--2015|Graham et al., 2015]] ; [[#Cinner--2016|Cinner et al., 2016]] ). These impacts are sometimes exacerbated by catastrophic events such as tropical storms and marine heatwaves that destroy habitats and hence the resources upon which coastal fisheries depend ( [[#Sainsbury--2018|Sainsbury et al., 2018]] ). There is ''high confidence'' that climate change impacts, together with local human disturbances, will continue to denude coastal and marine ecosystem services in many small islands with serious consequences for vulnerable communities ( [[#Elliff--2017|Elliff and Silva, 2017]] ; [[#Bindoff--2019|Bindoff et al., 2019]] ).

<div id="15.3.3.2" class="h3-container"></div>

<span id="impacts-on-freshwater-systems"></span>
==== 15.3.3.2 Impacts on Freshwater Systems ====

<div id="h3-2-siblings" class="h3-siblings"></div>

Freshwater systems on small islands are exposed to dynamic climate impacts and are considered to be among the most threatened on the planet (key risk 3 in Box 15.1; [[#Settele--2014|Settele et al., 2014]] ; [[#IPCC--2018|IPCC, 2018]] ; [[#Butchart--2019|Butchart et al., 2019]] ). [[#Hoegh-Guldberg--2019|Hoegh-Guldberg et al. (2019)]] estimated that freshwater stress on small islands would be 25% less with a warming of 1.5°C or less as compared to 2.0°C. While some island regions are projected to experience substantial freshwater decline, an opposite trend is observed for some western Pacific and northern Indian Ocean islands ( [[#Holding--2016|Holding et al., 2016]] ; [[#Karnauskas--2016|Karnauskas et al., 2016]] ). Island topography and ecophysiology influence water storage capacity and rainfall response potential ( [[#Dunn--2018|Dunn et al., 2018]] ). On high volcanic and granitic islands, freshwater ecosystems are often closely connected with coastal spaces, and changes in freshwater supply from river systems have direct implications for salinity and sediment loads ( ''high confidence'' ) ( [[#Yang--2015|Yang et al., 2015]] ; [[#Zahid--2018|Zahid et al., 2018]] ). Climate impacts on streamflow patterns in tropical islands also create shifts in water supply for downstream users and habitat conditions for organisms supporting a wide range of ecosystem services ( ''high confidence'' ) ( [[#Strauch--2015|Strauch et al., 2015]] ; [[#Frazier--2019|Frazier and Brewington, 2019]] ; [[#Frauendorf--2020|Frauendorf et al., 2020]] ).

Projected changes in aridity are expected to impose freshwater stress on many small islands, especially SIDS ''(high confidence'' ).These changes are congruent with drought risk projections for Caribbean SIDS ( [[#Lehner--2017|Lehner et al., 2017]] ; [[#Taylor--2018|Taylor et al., 2018]] ) and aligned with observations from the Shared Socioeconomic Pathway (SSP) 2 scenario, where a 1°C increase in temperature (from 1.7°C to 2.7°C) could result in a 60% increase in the number of people projected to experience severe water resources stress from 2043 to 2071 ( [[#Schewe--2014|Schewe et al., 2014]] ; [[#Karnauskas--2018|Karnauskas et al., 2018]] ). In the Mediterranean region, freshwater resources will decline by 10–30% ''(medium confidence)'' ( [[#Koutroulis--2016|Koutroulis et al., 2016]] ; [[#Kumar--2020|Kumar et al., 2020]] ). For example, analysis of annual and seasonal streamflow data on the island of Mallorca shows a decreasing trend during spring and summer, with a reduction of up to 17% in some basins (Garcia, 2017).

The influence of climate change spans several variables for atoll islands with multiple, interacting forces that exacerbate impacts on freshwater ecosystems ( [[#Connell--2016|Connell, 2016]] ), including groundwater and freshwater resources ( [[#Warix--2017|Warix et al., 2017]] ). Analysis of groundwater resources on Roi-Namur, in the Marshall Islands, reveals that the extent of salinisation of fresh groundwater lenses varies with the scale of the overwash ( [[#Gingerich--2017|Gingerich et al., 2017]] ). [[#Alsumaiei--2018|Alsumaiei and Bailey (2018)]] estimated an 11–36% reduction in the fresh groundwater lens volume of the small atoll islands (area &lt; 0.6 km²) of the Maldives due to SLR. Small overwash events lead to saline conditions that last for up to 3 months ( [[#Oberle--2017|Oberle et al., 2017]] ).

SLR undermines the long-term persistence of freshwater-dependent ecosystems on islands ( [[#Goodman--2012|Goodman et al., 2012]] ) and is one of the greatest threats to the goods and services these environments provide (Box 16.1; [[#Mitsch--2013|Mitsch and Hernandez, 2013]] ). [[#Hoegh-Guldberg--2019|Hoegh-Guldberg et al. (2019)]] posit that as sea level rises, managing the risk of salinisation of freshwater resources will become increasingly important. On Roi-Namur, Marshall Islands, [[#Storlazzi--2018|Storlazzi et al. (2018)]] found that the availability of freshwater is impacted by the compounding effect of SLR and coastal flooding. In other Pacific atolls, [[#Terry--2012|Terry and Chui (2012)]] showed that freshwater resources could be significantly affected by a 0.40-m SLR. Similar impacts are anticipated for some Caribbean countries (Stennett- [[#Brown--2017|Brown et al., 2017]] ). Such changes in SLR could increase salinity in estuarine and aquifer water, affecting ground and surface water resources for drinking and irrigation water ( [[#Mycoo--2018a|Mycoo, 2018a]] ) across the region ( ''high confidence'' ). SLR also affects groundwater quality ( [[#Bailey--2016|Bailey et al., 2016]] ), salinity ( [[#Gingerich--2017|Gingerich et al., 2017]] ) and water-table height ( [[#Masterson--2014|Masterson et al., 2014]] ).

<div id="15.3.3.3" class="h3-container"></div>

<span id="impacts-on-terrestrial-biodiversity-systems"></span>
==== 15.3.3.3 Impacts on Terrestrial Biodiversity Systems ====

<div id="h3-3-siblings" class="h3-siblings"></div>

Despite encompassing approximately 2% of the Earth’s terrestrial surface, oceanic and other high-endemicity islands are estimated to harbour substantial proportions of existing species including ~25% extant global flora, ~12% birds and ~10% mammals ( [[#Alcover--1998|Alcover et al., 1998]] ; [[#Wetzel--2013|Wetzel et al., 2013]] ; [[#Kumar--2017|Kumar and Tehrany, 2017]] ). Islands also have higher densities of critically endangered species, hosting just under half of all species currently considered to be at risk of extinction ( [[#Spatz--2017a|Spatz et al., 2017a]] ; 2017b), hence making the loss of terrestrial biodiversity and related ecosystem services a KR (KR3) for small islands (Figure 15.5). Impacts from developing synergies between changing climate, natural and anthropogenic stressors on islands (Cross-Chapter Box DEEP in Chapter 17) could lead to disproportionate changes in global biodiversity. The most prominent drivers include: SLR, increasing intensities of extreme events (human activities—especially continuing/accelerating habitat destruction/degradation) and the introduction of invasive alien species (IAS) ( [[#Tershy--2015|Tershy et al., 2015]] ). When coupled with characteristic small island traits such as spatial and other resource limitations, these synergies play a critical role towards increasing the vulnerability of these insular ecosystems (Box [https://www.ipcc.ch/chapter/15#CCP1.1 CCP1.1] ). This is likely to hinder the adaptation response of terrestrial biota–increasing the risk of biodiversity loss and, in turn, impairing the resilience capacity of ecosystem functioning and services ( ''high confidence'' ) ( [[#Heller--2009|Heller and Zavaleta, 2009]] ; [[#Ferreira--2016|Ferreira et al., 2016]] ; [[#Vogiatzakis--2016|Vogiatzakis et al., 2016]] ).

Current observations of insular species response to climate change generally report geographic range shifts/reductions for species and vegetation associations in addition to resulting impacts on local ecology ( [[#Virah-Sawmy--2016|Virah-Sawmy et al., 2016]] ; [[#Koide--2017|Koide et al., 2017]] ; [[#Maharaj--2019|Maharaj et al., 2019]] ). These include changes in plant/animal phenology and resulting community alterations such as for the common Mediterranean island species ''Quercus ilex'' (holly oak) and ''Ficus carica'' (common fig). Species have been shifting greater distances to access not only suitable climate conditions but also, by association, suitable breeding conditions and seasonal food. Examples include: migratory birds such as ''Coturnix coturnix'' now having earlier spring arrival dates in the Mediterranean compared to six decades ago and the increased mortality of the iconic ''Argyroxiphium sandwicense'' (Hinahina) as result of warmer drier trends at Hawaiian high altitudes ( [[#Krushelnycky--2012|Krushelnycky et al., 2012]] ; [[#Taylor--2016a|Taylor and Kumar, 2016a]] ; [[#Vogiatzakis--2016|Vogiatzakis et al., 2016]] ). There have also been die-offs of some species from temperature extremes (e.g., flying fox species: ''Pteropus'' species) within the Pacific islands ( [[#Taylor--2016a|Taylor and Kumar, 2016a]] ).

Recorded alterations of ecological interactions include increased competition, changes to migratory routes ( [[#Harter--2015|Harter et al., 2015]] ) and mismatches between species, such as increased pathogen attacks on Mediterranean forest species ( [[#Vogiatzakis--2016|Vogiatzakis et al., 2016]] ). Also, in some areas of Madagascar there has been increased vulnerability to fire, due to the replacement of succulents by less fire-resilient species ( [[#Virah-Sawmy--2016|Virah-Sawmy et al., 2016]] ). Further, the low functional redundancy of island ecosystems implies a comparatively higher proportion of keystone species than continents, many of them being endemic ( [[#Harter--2015|Harter et al., 2015]] ), with potentially unpredictable system consequences due to climate-induced ecological changes. For example, Caribbean land crabs have been observed to alter their food intake as a response to drying conditions ( [[#McGaw--2019|McGaw et al., 2019]] ) and Aldabra giant land tortoises have reduced their activity in response to increasing temperature and decreasing precipitation ( [[#Falcon--2018|Falcon and Hansen, 2018]] ); such changes in both these ecosystem engineers are of potential consequence for seed dispersal, among other ecological functions.

The majority of studies modelling geographical range changes of small island species, to even the most optimistic 21st century climate change scenarios, imply a reduction in climate refugia (Table 15.3, Box [https://www.ipcc.ch/chapter/15#CCP1.1 CCP1.1] ). This is due to projected strong shifts, reductions or even complete losses of climatic niches resulting from inadequate geographic space for species to track suitable climate envelopes ( ''high confidence'' ) (e.g., [[#Maharaj--2013|Maharaj and New, 2013]] ; [[#Fortini--2015|Fortini et al., 2015]] ; [[#Struebig--2015b|Struebig et al., 2015b]] ). Because of the high proportion of global endemics hosted within small and especially isolated islands, the resulting increased extinction risk of such species (up to 100%) could lead to disproportionate losses in global biodiversity ( ''medium'' to ''high confidence'' ) ( [[#Harter--2015|Harter et al., 2015]] ; [[#Manes--2021|Manes et al., 2021]] ).

SLR has been projected to impact the terrestrial biodiversity of low-lying islands and coastal regions via large habitat losses both directly (e.g., submergence) and indirectly (e.g., salinity intrusion, salinisation of coastal wetlands and soil erosion) at even the 1-m scenario ( ''medium'' to ''high confidence'' ). However, these impacts vary depending on the islands’ topographical differences. In a study of SLR impacts on insular biodiversity hotspots, Bellard et al. (2013a) reported that the Caribbean islands, Sundaland and the Philippines were projected to suffer the most habitat loss while the East Melanesian islands were projected to be less (but not minimally) affected. The most threatened of these, the Caribbean, was projected to have between 8.7% and 49.2% of its islands entirely submerged, respectively, from 1-m to 6-m SLR ( [[#Bellard--2013a|Bellard et al., 2013a]] ). However, many current projection studies consider marine flooding directly and seldom incorporate other indirect impacts such as increased habitat losses from horizontal erosion loss, increased salinity levels, tidal ranges and extreme events. These projections are considered to be conservative, underestimating the extent of habitat loss to terrestrial biodiversity ( [[#Bellard--2013b|Bellard et al., 2013b]] ).

Marine flooding is expected to destroy habitats of coastal species, particularly range-restricted coastal and/or single-island endemics (many already listed as ''at least'' ‘threatened’ by the International Union for Conservation of Nature) within the limited terrain on atoll islands. These species have limited opportunities to accommodate such direct impacts of climate change apart from shifting further inland or to other neighbouring atolls which might have favourable habitat. However, fragmentation of habitat due to anthropogenic activity may hinder migration further inland, while shifting to neighbouring islands is not viable due to the water barrier between islands ( ''high confidence'' ) ( [[#Bellard--2013b|Bellard et al., 2013b]] ; [[#Wetzel--2013|Wetzel et al., 2013]] ; [[#Kumar--2017|Kumar and Tehrany, 2017]] ). Additionally, migratory birds, which use small islands (e.g., atolls) for stopovers or breeding/nesting sites, are projected to become impacted. Within the Mediterranean and Caribbean, significant losses to coastal wetlands—critical habitat for migratory birds—has already been observed, with further significant habitat losses, redistribution and changes in quality being projected across island systems such as the Bahamas (Caribbean) and Sardinia (Mediterranean) ( [[#Vogiatzakis--2016|Vogiatzakis et al., 2016]] ; [[#Wolcott--2018|Wolcott et al., 2018]] ).

Indirect impacts of SLR may potentially result in equal or more biodiversity loss than direct impacts ( ''medium confidence'' ). Relocation of displaced coastal human populations and associated intensive agriculture and urban areas inland to natural habitat may result in greater biodiversity loss than direct impacts—especially on islands with large coastal populations and urban centres ( [[#Wetzel--2012|Wetzel et al., 2012]] ; [[#Bellard--2013b|Bellard et al., 2013b]] ). Given the dense population of insular hotspots (~31.8% of existing humans within ~15.9% of inhabited global land area) and the fact that on many islands, large proportions of human populations live within coastal regions, it has been suggested that immense impacts from such relocations should be factored into projection and adaptation studies ( [[#Wetzel--2012|Wetzel et al., 2012]] ).

Tropical island natural habitats/systems are highly vulnerable to extreme weather events such as TCs, due to their small size, unique ecological systems and often low socioeconomic capacity ( ''high confidence'' ) (Box 15.2; [[#Goulding--2016|Goulding et al., 2016]] ; [[#Schütte--2018|Schütte et al., 2018]] ). Growing evidence suggests high resilience of forest habitats ( [[#Keppel--2014|Keppel et al., 2014]] ; [[#Luke--2017|Luke et al., 2017]] ), especially within intact forest ecosystems to hurricanes and cyclones ( [[#Goulding--2016|Goulding et al., 2016]] ). While initial damage can be high, relatively fast recovery rates have been reported for both floral and faunal components of these ecosystems ( [[#Cantrell--2014|Cantrell et al., 2014]] ; [[#Shiels--2014|Shiels et al., 2014]] ; [[#Monoy--2016|Monoy et al., 2016]] ; [[#Richardson--2018|Richardson et al., 2018]] ). Within the Caribbean in particular, high resilience of forest types has been associated with the ''current'' intensity and return rate of hurricanes over the past 150 years.

It should, however, be underscored that these relatively fast recovery rates are associated with the ''present'' intensity and return rate of TCs. They do not reflect the impacts of increasingly intense events such as Hurricane Dorian (2019), which resulted in almost complete inundation of several low-lying islands of the Bahamas from storm surges. Severe weather events also have indirect effects on the biodiversity of islands—interacting synergistically with other stressors, such as increased invasion by non-native species and land use change. For example, TCs within Papua New Guinea resulted in the destruction of subsistence gardens, which led inhabitants to clear forest areas for new farming areas and for harvesting of timber resources to rebuild ( [[#Goulding--2016|Goulding et al., 2016]] ).

The most recent projections suggest that TC intensity is predicted to increase as climate continues to change ( [[#Walsh--2016|Walsh et al., 2016]] ; [[#Kossin--2017|Kossin et al., 2017]] ). There are too few studies available to suggest potential future response trends of these ecosystems to this increased intensity; however, it seems plausible that present resilience capacities may be adversely impacted ( ''medium confidence'' ) ( [[#Marler--2014|Marler, 2014]] ). Further, the potential for stressors such as forest fragmentation/degradation or IAS combining with these increasingly intense events to cause precipitating ecosystem cascades is a real concern ( [[#Goulding--2016|Goulding et al., 2016]] ).

Continued high rates of habitat loss and degradation have been reported for many small islands as natural habitats continue to be cleared to meet increasing demands upon natural resources from rising human populations, agriculture, urbanisation, unsustainable tourism, overgrazing and fires. This increases the vulnerability of ecosystems within especially oceanic islands—where isolation has given rise to high levels of endemism but simple biotic communities, with low functional redundancy (Box [https://www.ipcc.ch/chapter/15#CCP1.1 CCP1.1] ). There is ''high confidence'' that climate change may exacerbate the effects of this habitat loss upon the biodiversity of these islands as the climate refugia (Table 15.3) and the upslope shifts of range-restricted, dispersal-limited and poorly competitive species, confined within narrow latitudinal (and decreasing altitudinal) gradients, are increasingly challenged by fragmented and degraded landscapes (e.g., [[#Struebig--2015a|Struebig et al., 2015a]] ; [[#IPBES--2019|IPBES, 2019]] ). Additionally, high-altitude ecosystems such as cloud forests which harbour high levels of endemism are projected to shrink due to increasing atmospheric temperature and competition from upward-shifting lowland species ( [[#Taylor--2016a|Taylor and Kumar, 2016a]] ). These may ultimately increase the risk of multiple extinctions, negatively impacting upon global biodiversity levels ( ''high confidence'' ) ( [[#Taylor--2016a|Taylor and Kumar, 2016a]] ; [[#Portner--2021|Portner et al., 2021]] ).

Analyses of historical and current threats indicate that IAS and disease have been the primary drivers of insular extinctions in modern history ( [[#Bellard--2016|Bellard et al., 2016]] ). Impacts of IAS on islands are projected to increase with time due to synergies between climate change and other traditional drivers such as increasing global trade, tourism, agricultural intensification, overexploitation and urbanisation ( [[#Bellard--2014|Bellard et al., 2014]] ; [[#Russell--2017|Russell et al., 2017]] ). Changing climate conditions may not necessarily increase the rate of IAS introductions but is expected to improve chances of IAS establishment via (a) altering IAS transport and introduction mechanisms, (b) increasing the impacts and distributions of existing IAS and (ci) altering the effectiveness of existing control strategies ( [[#Hellmann--2008|Hellmann et al., 2008]] ; [[#Russell--2017|Russell et al., 2017]] ). These are likely to enhance IAS impacts on islands including: restructuring of ecological communities leading to declines and extinctions/extirpations in flora and fauna, habitat degradation, declining ecosystem functioning, services and resilience and, in extreme cases, potential community homogenisation ( ''high confidence'' ) ( [[#Russell--2017|Russell and Blackburn, 2017]] ; [[#IPBES--2019|IPBES, 2019]] ). Given the high degree of endemicity within oceanic islands and their associated vulnerabilities, such exacerbation by changing climate poses a serious threat to decreasing global biodiversity ( ''medium'' to ''high confidence'' ) ( [[#van%20Kleunen--2015|van Kleunen et al., 2015]] ).

Compared to continents, terrestrial IAS are disproportionately prevalent on islands (almost three quarters of global species currently threatened by IAS and disease are found on islands) and also generate stronger impacts (e.g., within alpine ecosystems of high islands) than on continents ( ''high confidence)'' ( [[#Bellard--2014|Bellard et al., 2014]] ; [[#Bellard--2016|Bellard et al., 2016]] ; [[#Frazier--2019|Frazier and Brewington, 2019]] ). [[#Russell--2017|Russell and Blackburn (2017)]] suggested a correlation between small island size and increased numbers of IAS. SIDS within the Indian Ocean and in particular the Pacific SIDS region were reported to have significantly more IAS ( ''medium confidence'' ), while the Caribbean and Atlantic SIDS have fewer numbers but faster accumulation of IAS. Finally, while there have been developments in the eradication of IAS on islands ( [[#Jones--2016|Jones et al., 2016]] ), there is sparse evidence and hence assessment of the degree to which measures designed to prevent introduction and to manage invasion pathways and establishment have been successful.

<div id="15.3.4" class="h2-container"></div>

<span id="observed-impacts-and-projected-risks-on-human-systems"></span>