Editing IPCC:AR6/SROCC/Chapter-5 (section)

== Box 5.2 Cold Water Corals and Sponges ==

<div id="section-5-2-4-4chemosynthetic-ecosystems-block-1"></div>

Cold water corals and sponges form large reefs at the deep seafloor mostly between 200–1500 m, creating complex 3D habitat that supports high biodiversity; they are found at the highest densities on hard substrates of continental slopes, canyons, and seamounts (Buhl-Mortensen et al., 2010). The meta-analysis reported in AR5 Chapter 6 Table 6-3 (Pörtner et al., 2014), identifies 10 studies involving 6 species of cold water corals that suggest low vulnerability to CO 2 changes at RCP6.0 and medium vulnerability at RCP8.5, with negative effects starting at pCO 2 of 445 µatm.

Scleractinian corals have the capacity to acclimate to high CO 2 conditions due to their capacity to upregulate the pH at the calcification site (Form and Riebesell, 2011; Rodolfo-Metalpa et al., 2015; Gori et al., 2016). The most widely distributed, habitat-forming species in deep water (e.g., ''Lophelia pertusa ['' renamed ''Desmophyllum pertusum'' ) (Addamo et al., 2016)can continue to calcify at aragonite undersaturation and high CO 2 levels projected for 2100 (750–1100 uatm) based on experiments (Georgian et al., 2016; Kurman et al., 2017) and observations along the natural gradient of carbon chemistry in their distributions (Fillinger and Richter, 2013; Movilla et al., 2014; Baco et al., 2017) (Appendix 1) ( ''robust evidence'' , ''medium agreement'' , ''medium confidence)'' and thus appear to be able to acclimate to rising CO 2 levels (Hennige et al., 2015). However, net calcification rates (difference between calcification and dissolution) of ''L. pertusa'' exposed to aragonite-undersaturated conditions (Ω ''arag'' &lt; 1, where Ω ''arag'' =aragonite saturation state) often decreases to close to zero or even becomes negative (Lunden et al., 2014; Hennige et al., 2015; Büscher et al., 2017), with genetic variability underpinning ability to calcify at low aragonite saturation states (Kurman et al., 2017). Additionally, skeletons become longer, thinner and weaker (Hennige et al., 2015), and bioerosion is enhanced (e.g., by bacteria, fungi, annelids and sponges) (Schönberg et al., 2017), exacerbating effects of dissolution of the skeleton. ''L. pertusa'' can calcify when exposed to multiple environmental stresses in the laboratory (Hennige et al., 2015; Büscher et al., 2017), but cannot survive with warming above water temperatures of 14°C–15 o C or oxygen concentrations below 1.6 ml l -1 in the Gulf of Mexico, 3.3 ml l -1 in the north Atlantic, 2 ml l -1 in the Mediterranean, and 0.5–1. 5 ml l -1 in the SE Atlantic (Brooke et al., 2013; Lunden et al., 2014; Hanz et al., 2019), highlighting the existence of critical thresholds for cold water coral populations living at the edge of their tolerance. The role of temporal dynamics, species-specific thermal tolerances, and food availability in mediating the response to combinations of stressors is recognised but is still poorly studied under ''in situ'' conditions (Lartaud et al., 2014; Naumann et al., 2014; Baco et al., 2017).

Sponges also form critical habitat in the deep ocean but are much less well studied than cold-water corals with respect to climate change. The geologic record, modern distributions and evolutionary and metabolic pathways suggest that sponges are more tolerant to warm temperatures, high CO 2 and low oxygen than are cold-water corals (Schulz et al., 2013). One habitat forming, deep sea sponge along with its microbiome (microbial inhabitants) has been shown in laboratory experiments to tolerate a 5°C increase in temperature, albeit with evidence of stress (Strand et al., 2017), while ocean acidification (pH 7.5) reduces the feeding of two deep sea demosponge taxa (Robertson et al., 2017).

Generally, the deep sea areas where cold water corals may be found are projected to be exposed to multiple climate hazards in the 21st century because of the projected ocean warming, oxygen loss, and decrease in POC flux (Table 5.5) under scenarios of greenhouse gas emissions. The average changes in these climate hazards for coral-water corals are projected to be almost halved under RCP2.6 relative to RCP8.5 (Table 5.5). Under RCP8.5, 95 ± 2% (95% CI) of cold-water coral habitats are projected to experience animal biomass decline (–8.6 ± 2.0%) globally by 2091–2100 relative to 2006–2015, driven by a projected 21 ± 9% drop in POC flux ( ''medium confidence'' ) (Jones et al., 2014). However, nutritional co-reliance of cold-water corals on zooplankton (Höfer et al., 2018) and carbon fixation by symbiotic microbes (Middelburg et al., 2015), is not incorporated into the models, adding uncertainty to these estimates. Regionally, suitable habitat for coral-water corals in the NE Atlantic is projected to decrease with multiple climatic hazards (warming, acidification, decreases in oxygen and POC flux) under RCP8.5 for 2081–2100 (FAO, 2019), with up to 98% loss of suitable habitat by 2099 due to shoaling aragonite saturation horizons. In the Southern hemisphere, a tolerance threshold of 7°C and decline of aragonite saturation below that required for survival ( ''Ω'' arag &lt;0.84) can cause large loss of cold water corals habitat ( ''Solenosmilia variabilis'' ) on seamounts off Australia and New Zealand under future projections of warming and acidification to 2099 at RCP4.5 and nearly complete loss under RCP8.5 (Thresher et al., 2015).

Overall, cold water corals can survive conditions of aragonite-undersaturation associated with ocean acidification but sensitivity varies among species and skeletons will be weakened ( ''medium confidence'' ). The largest impacts on calcification and growth will occur when aragonite saturation is accompanied by warming and/or decrease in oxygen concentration beyond the tolerance limits of these corals ( ''medium confidence)'' . Given present day occurrence of 95% of cold water corals above the aragonite saturation horizon (Guinotte et al., 2006) and that no adaptation has been detected with regard to increased dissolution of exposed aragonite (Eyre et al., 2014), there is limited scope for the non-living components of cold water corals and for the large, non-living reef framework that comprises deep water reefs to avoid dissolution under RCP8.5 in the 21st century ( ''high confidence'' ). Multiple climatic hazards of warming, deoxygenation, aragonite under-saturation and decrease in POC flux are projected to negatively affect cold water corals worldwide from the present day by 2100 ( ''high confidence'' ). Uncertainty remains in the adaptive capacity of living cold water corals to cope with these changes and in the influence of altered regional current patterns on connectivity (Fox et al. 2016; Roberts et al., 2017). Sponges and the habitat they form may be less vulnerable than cold water corals to warming, acidification and deoxygenation that will occur under RCP8.5 in 2100 ( ''low confidence'' ).

<span id="risk-assessment-of-open-ocean-ecosystems"></span>
=== 5.2.5 Risk Assessment of Open Ocean Ecosystems ===

<div id="section-5-2-5risk-assessment-of-open-ocean-ecosystems-block-1"></div>

This section synthesises the assessment of climate impacts on open ocean and deep seafloor ecosystem structure and functioning and the levels of risk under future conditions of global warming (see SM5.2). The format for Figure 5.16 matches that of Figure 19.4 of AR5 (Pörtner et al., 2014) and Figure 3.20 of SR15 (Hoegh-Guldberg et al., 2018), indicating the levels of additional risk as colours (white, yellow, red and purple). Each column in Figure 5.16 indicates how risks increase with ocean warming, acidification (OA), deoxygenation, and POC flux with a focus on present day conditions (2000s) and future conditions by the year 2100 under low (RCP2.6) and high (RCP8.5) CO 2 emission scenarios. The transition between the levels of risk to each type of ecosystem is estimated from key evidence assessed in earlier parts of this chapter (Sections 5.2.2, 5.2.3, 5.2.4). SST is chosen to provide an indication of the changes in all these variables because it is closely related to cumulative carbon emission (Gattuso et al., 2015) which is the main climatic driver of the hazards. SST scales with Global Mean Surface Temperature (GMST) by a factor of 1.44 according to changes in an ensemble of RCP8.5 simulations; with an uncertainty of about 4 % in this scaling factor based on differences between the RCP2.6 and RCP8.5 scenarios. The transition values may have an error of ±0.3°C depending on the consensus of expert judgment. The deep seafloor embers are generated based on earth system model projection of climate variables to the seafloor under RCP2.6 and RCP8.5 scenarios, and then translated to RCP associated change in SST. The assessed confidence in assigning the levels of risk at present day and future scenarios are ''low, medium, high'' and ''very high'' levels of confidence. A detailed account of the procedures involved in the ember for each type of ecosystem, such as their exposure to climate hazards, sensitivity of key biotic and abiotic components, natural adaptive capacity, observed impacts and projected risks, and regional hotspots of vulnerability is provided in the SM5.2 and Table 5.5. The risk assessment for cold water corals is in agreement with the conclusions in AR5 Chapter 6.3.1.4.1, although more recent literature is assessed in Box 5.2 and Table SM5.5.

Overall, the upper ocean (0−700 m) and 700−2000 m layers have both warmed from 2004 to 2016 ( ''virtually certain'' ) and the abyssal ocean continues to warm in the Southern Hemisphere ( ''high confidence'' ). The ocean is stratifying; observed warming and high-latitude freshening are both surface intensified trends making the surface ocean lighter at a faster rate than deeper in the ocean ( ''high confidence'' ) (Section 5.2.2.2). It is  ''very likely''  that stratification in the upper few hundred meters of the ocean will increase significantly in the 21st century. It is ''virtually certain'' that ocean pH is declining by ~0.02 pH units per decade where time series observations exist (Section 5.2.2.3). The anthropogenic pH signal has already emerged over the entire surface ocean ( ''high confidence'' ) and emission scenarios are the most important control of surface ocean pH relative to internal variability for most of the 21st century at both global and local scale ( ''virtually certain'' ). The oxygen content of the global ocean has declined by about 0.5−3.3% in 0−1000 m layer (Section 5.2.2.4). Over the next century oxygen declines of 3.5% by 2100 are predicted by CMIP5 models globally ( ''medium confidence'' ), with ''low confidence'' at regional scales, especially in the tropics. The largest changes in the deep sea will occur after 2100 (Section 5.2.2.3). CMIP5 models project a decrease in global NPP ( ''medium confidence'' ) with increases in high-latitude ( ''low confidence'' ) and decreases in low latitude ( ''medium confidence'' ) (Section 5.2.2.6) in response to changes in ocean nutrient supply (Section 5.2.2.5). These models also project reductions by 8.9−15.8% in the globally integrated POC flux for RCP8.5, with decreases in tropical regions and increases at higher latitudes ''(medium confidence)'' , affecting the organic carbon supply to the deep sea floor ecosystems ( ''high confidence'' ) (Section 5.2.2.6). However, there is ''low confidence'' on the mechanistic understanding of how climatic drivers will affect the different components of the biological pump in the epipelagic ocean (Table 5.4). Therefore, the exposure to hazard for epipelagic ecosystems ranges from moderate (RCP2.6) to high (RCP8.5), with uncertain effects and tolerance of planktonic organisms, fishes and large vertebrates to interactive climate stressors. Major risks are predicted for declining productivity and fish biomass in tropical and subtropical waters (RCP8.5) (SM5.2).

The climatic hazards for pelagic organisms from plankton to mammals are driving changes in eco-physiology, biogeography and ecology and biodiversity ( ''high confidence'' ) (Section 5.2.3.1). Observed and projected population declines in the equator-ward range boundary ( ''medium confidence'' ), expansion in the poleward boundary ( ''high confidence'' ), earlier timing of biological events ( ''high confidence'' ), overall shift species composition ( ''high confidence'' ) and decreases in animal biomass ( ''medium confidence'' ), are consistent with expected responses to climate change (Section 5.2.3; Figure 5.13). It is ''likely'' that increased OA has not yet caused sufficient reduction in fitness to decrease abundances of calcifying phytoplankton and zooplankton, but ''is very'' ''likely'' ( ''high confidence'' ) that calcifying planktonic organisms will experience great decreases in abundance and diversity under high emission scenarios by the end of the century. Therefore, impacts to the epipelagic ecosystems are already observed in the present day (Figure 5.16). Based on simulation modelling and experimental findings, the combined effects of warming, ocean deoxygenation, OA and changes in NPP in the 21st century are projected to exacerbate the impacts on the growth, reproduction and mortality of fishes, and consequently increase the risk of population decline ( ''high confidence'' ) (Section 5.2.3.1). There may be some capacity for adjustment and evolutionary adaptation that lowers their sensitivity to warming and decrease in oxygen ( ''low confidence'' ). However, historical responses in abundance and ranges of marine fishes to ocean warming and decrease in oxygen in the past suggest that adaptation is not always sufficient to mitigate the observed impacts ( ''medium confidence'' ) (Section 5.2.3) (SM5.2).

Despite its remoteness, most of the deep seafloor ecosystems already have or are projected to experience rising temperatures and declining oxygen, pH and POC flux beyond natural variability within the next half century (See Section 5.2.4). On slopes, seamounts and canyons these changes are projected to be much larger under RCP8.5 than under RCP2.6 ( ''high confidence'' ), with greatest effects on seafloor community diversity and function from expansion of low oxygen zones and aragonite undersaturation ( ''medium confidence'' ). As critical thresholds of temperature, oxygen and CO 2 are exceeded, coral species will alter their depth distributions, non-living carbonate will experience dissolution and bioerosion, and stress will be exacerbated by lower food supply. These changes are projected to cause loss of cold water coral habitat with highest climate hazard in the Arctic and north Atlantic Ocean ( ''medium confidence'' ), while sponges may be more tolerant (Box 5.2) ( ''low confidence'' ). Projected changes in food supply to the seafloor at abyssal depths combined with warmer temperatures are anticipated to cause reductions in biomass and body size ( ''medium confidence'' ) that could affect the carbon cycle in this century under RCP8.5 ( ''low confidence'' ). Even at hydrothermal vents and methane seeps, some dominant species such as mussels may be vulnerable to reduced photosynthetically-based food supply or have planktonic larvae or oxidising symbionts that are negatively affected by warming, acidification and oxygen loss ( ''low confidence'' ).

Widespread attributes of deep seafloor fauna (e.g., great longevity '','' high levels of habitat specialisation including well-defined physiological tolerances and thresholds, dependence on environmental triggers for reproduction, and highly developed mutualistic interactions) can increase the vulnerability of selected taxa to changing conditions (FAO, 2019) ( ''medium confidence'' ). However, some deep sea taxa (e.g., foraminifera and nematodes) may be more resilient to environmental change than their shallow-water counterparts ( ''low confidence'' ). Observations, experiments and model projections indicate that impacts of climate change have or are expected to take place in this century, indicating a transition from undetectable risk to moderate risk at &lt;1.5 o C warming of sea surface temperature for continental slope, canyon and seamount habitats, and for cold water corals (Figure 5.16). 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) (Figure 5.16). 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 experiencing a transition from moderate to high risk below 3 o C (SM5.2).

<span id="changing-coastal-ecosystems-and-biodiversity"></span>