Reviews and Syntheses: Ocean acidification and its potential impacts on marine ecosystems

Abstract. Ocean acidification, a complex phenomenon that lowers seawater pH, is the net outcome of several contributions. They include the dissolution of increasing atmospheric CO2 that adds up with dissolved inorganic carbon (dissolved CO2, H2CO3, HCO3−, and CO32−) generated upon mineralization of primary producers (PP) and dissolved organic matter (DOM). The aquatic processes leading to inorganic carbon are substantially affected by increased DOM and nutrients via terrestrial runoff, acidic rainfall, increased PP and algal blooms, nitrification, denitrification, sulfate reduction, global warming (GW), and by atmospheric CO2 itself through enhanced photosynthesis. They are consecutively associated with enhanced ocean acidification, hypoxia in acidified deeper seawater, pathogens, algal toxins, oxidative stress by reactive oxygen species, and thermal stress caused by longer stratification periods as an effect of GW. We discuss the mechanistic insights into the aforementioned processes and pH changes, with particular focus on processes taking place with different timescales (including the diurnal one) in surface and subsurface seawater. This review also discusses these collective influences to assess their potential detrimental effects to marine organisms, and of ecosystem processes and services. Our review of the effects operating in synergy with ocean acidification will provide a broad insight into the potential impact of acidification itself on biological processes. The foreseen danger to marine organisms by acidification is in fact expected to be amplified by several concurrent and interacting phenomena.


Introduction
Ocean acidification is typically defined as a process of increasing seawater acidity or lowering seawater pH, as a consequence of the dissolution of elevated atmospheric CO 2 . Carbon dioxide from the atmosphere (Orr et al., 2005;Feely et al., 2008) adds to the dissolved inorganic carbon (DIC; dissolved CO 2 , H 2 CO 3 , HCO − 3 , and CO 2− 3 ) originated from the degradation of dissolved organic matter (DOM) (Mostofa et al., 2013a), primary producers (PP) Mostofa et al., 2013a), CO 2 seeps from sub-seabed storage (Taylor et al., 2014) and volcanic vents (Lidbury et al., 2012;Hall-Spencer et al., 2008) in shallow submarine zones, anaerobic oxidation of methane (Haroon et al., 2013), and sulfide oxidation coupled to carbonate dissolution (Torres et al., 2014) in seawater. The sources of elevated atmospheric CO 2 first of all include anthropogenic activities such as fossil fuels combustion, i.e., coal, petroleum, and natural gas (Le Quéré et al., 2009), enhanced land-use practices (Le Quéré et al., 2009), as well as deforestation (van der Werf et al., 2009;Lapola et al., 2014). Additionally, there could be significant contributions from natural sources such as plant litter decomposition (King et al., 2012), volcanic eruptions (Hall-Spencer et al., 2008), emission of CO 2 from freshwater including the Amazon River basin (Sobek et al., 2005;Abril et al., 2014), and enhanced respiration of soil organic matter (OM) under global warming (GW) conditions (Knorr et al., 2005).
The emissions of CO 2 by fossil fuels combustion increased by 29 % in 2000-2008(Le Quéré et al., 2009) and, as far as natural-water sources are concerned, the contribution from European estuaries is for instance equivalent to approximately 5-10 % of the anthropogenic CO 2 emissions in western Europe (Frankignoulle et al., 1998).
Recent studies demonstrate that ocean acidification under elevated CO 2 and temperature levels could increase primary productivity of specific species (Holding et al., 2015;Coello-Camba et al., 2014;Li et al., 2012). Additionally, such specific species-based primary productivity is also found to increase either by an increasing seawater CO 2 level (Kim et al., 2006;Olischläger et al., 2013) or elevated temperature alone because of the effects of global warming (Yvon-Durocher et al., 2015;Lewandowska et al., 2012). The primary production in the oceans contributes approximately 48.5 petagrams (1 Pg = 10 15 g) of C yr −1 (46.2 % of the total), as estimated using the integrated the Carnegie-Ames-Stanford Approach -the Vertically Generalized Production Model (CASA-VGPM) biosphere model (Field et al., 1998). As a consequence, approximately one-third to 50 % of the atmospheric CO 2 is fixed annually worldwide by marine phytoplankton (Sabine et al., 2004;Toseland et al., 2013). However, one should also consider that the photoinduced and biological mineralization of organic matter (OM), including DOM and dead organisms, is an important source of DIC in seawater and liberates again an important fraction of the CO 2 fixed by photosynthesis (Bates and Mathis, 2009;Mostofa et al., 2013a).
Ocean acidification is responsible for changes in the oceanic carbonate system, with effects on partial pressure of CO 2 (pCO 2 ), DIC, pH, alkalinity, and calcium carbonate saturation state Beaufort et al., 2011). In the case of calcifying organisms one observes a marked pattern of decreasing calcification with increasing pCO 2 , which follows the corresponding decreasing concentrations of CO 2− 3 as a consequence of decreasing pH (Beaufort et al., 2011). Such effects finally cause a decline in calcification and growth rates of shellfish (Talmage and Gobler, 2010;Wittmann and Pörtner, 2013), of shell-forming marine plankton and of benthic organisms including corals (Kleypas et al., 1999;Doney et al., 2009;Beaufort et al., 2011;Pandolfi et al., 2011;McCulloch et al., 2012). The latter have already been lost or are highly damaged in coastal areas near many countries including Indonesia, Hawaii, the Caribbean, Fiji, Maldives, and Australia (Erez et al., 2011). A 30 % decline or damage of coral reef ecosystems has been estimated worldwide, and it is predicted that as much as 60 % of the world's coral reefs might be lost by 2030 (Hughes et al., 2003).
The extent and effects of ocean acidification can be exacerbated by several complex processes, some of which act as stimulating factors, such as local environmental impacts including terrestrial or riverine runoff (Sunda and Cai, 2012;Bauer et al., 2013), modified land use practices (Lapola et al., 2014), and atmospheric acid rain (Baker et al., 2007). An additional effect could be represented by the enhanced mineralization of DOM and PP (e.g., phytoplankton) as a consequence of global warming (Mostofa et al., 2013a). Such mineralization could be biological (respiration) or abiotic via different (mainly) photochemical processes. Most of the cited effects are expected to cause eutrophication or algal blooms in coastal seawater, which would in turn affect the carbon cycling and the carbonate chemistry and influence the overall acidification process (Beaufort et al., 2011;Sunda and Cai, 2012;Bauer et al., 2013). Such acidification is responsible for changes in the oceanic carbonate system Beaufort et al., 2011), which subsequently impacts on marine living organisms and the related ecosystem processes or services (Cooley et al., 2009;Mora et al., 2013;Mostofa et al., 2013a). Considering the possible devastating consequences on the marine ecosystems, their organisms and the related ecosystem services (Cooley et al., 2009;Doney et al., , 2012Cai, 2011), it is important to ascertain all the possible causes of ocean acidification and their interlinks. showing either dissolution of atmospheric CO 2 or emission of aquatic CO 2 plus DIC originated from the photoinduced and/or biological respiration of primary producers (PP). The latter includes both dissolved organic matter (DOM) and PP (1). Uptake of such CO 2 is primarily responsible for the occurrence of photosynthesis and PP (2) that can generate algal toxins or pathogens in the euphotic zone, along with generation of CO 2 , DIC, and other products; PP can also be enhanced by autochthonous DOM (2), by DOM or sinking cells in subsurface or deeper seawater (2), and by riverine DOM (2). Atmospheric acid rain (mostly HNO 3 and H 2 SO 4 ) can contribute directly to the acidification (3). Global warming can lengthen the stratification period with a subsequent decline in vertical mixing, which reduces the exchange with surface oxygenated water (4). This review will provide a general overview of the ocean acidification such as chemistry and ecological consequences, including the interactions between acidification by CO 2 and other processes that could in turn modify the seawater pH. We shall discuss changes in the pH values in both sea surface and subsurface/deeper water extensively with different timescales, from diurnal to multi-annual. We shall also address potential impacts of ocean acidification on marine organisms, along with possible indirect impact processes from a series of stimulating factors (oxidative stress in surface seawater, hypoxia in subsurface/deeper seawater, stress caused by algal, or red-tide toxins and pathogens) for both sea surface and subsurface/deeper water. Our review from the point of synergistic effects of ocean acidification with such stimulating factors will broaden the understanding of the potential impact of acidification on biological processes. Such an impact is based on the conceptual model provided for both surface and deeper seawaters.

Potential mechanisms behind ocean acidification
Ocean acidification includes several potential phenomena that may be operational at the global and/or local scales ( Fig. 1): (i) increasing dissolution of atmospheric CO 2 to seawater: anthropogenic ocean acidification; (ii) input of CO 2 plus DIC upon mineralization of PP influenced by elevated atmospheric CO 2 : natural ocean acidification; (iii) enhanced PP and respiration due to the effects of global warming and other processes: natural ocean acidification; and (iv) direct acidification and stimulation of PP by atmospheric acid rain: natural and anthropogenic ocean acidification. A pictorial scheme of the main operational processes affecting the ocean acidification is depicted in Fig. 1.

Increasing dissolution of atmospheric CO 2 to seawater: anthropogenic ocean acidification
Enhanced dissolution of atmospheric CO 2 to seawater lowers pH and modifies the carbonate chemistry, affecting both biogenic and sedimentary CaCO 3 . This process has extensively been discussed in earlier reviews (Pearson and Palmer, 2000;Feely et al., 2008;Beaufort et al., 2011). For the given seawater, net CO 2 fluxes (either from atmosphere to water or the reverse) may significantly vary depending mostly on time (day or night) and season. Based on a series of studies, six scenarios can be formulated for the net sea-air fluxes of CO 2 . They are (i) sinking or balance of atmospheric CO 2 to seawater under sunlight, and emission or balance of CO 2 to the atmosphere during the night; (ii) emission or balance of CO 2 to the atmosphere during daytime, and sinking or balance of atmospheric CO 2 to surface water during the night; (iii) emission or balance of seawater CO 2 to the atmosphere during both day and night; (iv) sinking or balance of atmospheric CO 2 to surface water during both day and night; (v) sinking or source or balance of atmospheric CO 2 to surface water during the warm period; and (vi) emission or sinking or balance of seawater CO 2 to the atmosphere during the cold period. These scenarios are described in the Supplement.

Input of CO 2 plus DIC upon mineralization of PP influenced by elevated atmospheric CO 2 : natural ocean acidification
The formation and seawater dissolution of CO 2 and DIC produced from photoinduced and biological mineralization of PP or DOM also lowers pH and modifies the carbonate chemistry ( Fig. 2) (Cai et al., 2006Feely et al., 2010;Sunda and Cai, 2012;Bates et al., 2013;Mostofa et al., 2013a). Anticorrelation between pH and CO 2 levels during the diurnal cycle has been observed in surface and sub-surface waters (Fig. 2), where CO 2 is mainly originated from the biological respiration of PP or DOM. Such an issue is further complicated by the fact that enhanced levels of CO 2 are partially responsible for the increase of photosynthesis (Behrenfeld et al., 2006;Kranz et al., 2009), and they may have a deep impact on the net primary production (PP) (Hein and Sand-Jensen, 1997;Behrenfeld et al., 2006;Jiao et al., 2010). The upper ocean organisms, mostly the autotrophs, are a massive carbon-processing machine that can uptake atmospheric CO 2 (Hein and Sand-Jensen, 1997;Falkowski et al., 1998;Sarmento et al., 2010) or CO 2 plus DIC regenerated from DOM or PP, particularly during the daytime ( Fig. 2a; see also Supplement) (Takahashi et al., 2002(Takahashi et al., , 2009Yates et al., 2007;Chen and Borges, 2009;Mostofa et al., 2013a). In contrast, during the night seawater can become a source of CO 2 , as shown in Fig. 2 in three different contexts. The ability of water to act as a CO 2 source is shown by the higher values of pCO 2 in seawater compared to that in the atmosphere (Zhai et al., 2005(Zhai et al., , 2014Yates et al., 2007;Chen and Borges, 2009). The daytime uptake of CO 2 is the consequence of primary production through photosynthesis, which mostly uses dissolved CO 2 via the enzyme ribulosebiphosphate carboxylase (RUBISCO), which governs the carbon-concentrating mechanisms (CCMs) (Yoshioka, 1997;Behrenfeld et al., 2006;Kranz et al., 2009). Mesocosm experiments using 14 C-bottle incubations indicate that elevated CO 2 can increase 14 C primary production or bacterial biomass production, also leading to the formation of dissolved organic carbon (DOC) and to its rapid utilization .
ported to the ocean depths . Furthermore, all primary producers including the large and small cells can contribute to the carbon export from the surface layer of the ocean, at rates proportional to their production rates (Richardson and Jackson, 2007). The reprocessing of this organic material can cause a decrease in the pH of seawater via the CO 2 produced by respiration (Jiao et al., 2010). If, in addition, organic N and P are biologically transformed into NO − 3 and phosphate (Mostofa et al., 2013a) and if there is also transformation of NH + 4 to N 2 (Doney et al., 2007), there can be a further decrease of seawater alkalinity. Such processes also decrease the buffering capacity of seawater (Thomas et al., 2009), which would become more susceptible to acidification caused by the dissolution of atmospheric CO 2 (Thomas et al., 2009;Cai et al., 2011). A decrease in alkalinity and accompanying acidification may have negative impacts on shellfish production (Hu et al., 2015).
Heterotrophic bacteria are the main organisms that are responsible for respiration in the ocean (> 95 %) (Del Giorgio and Duarte, 2002), and half of the respiration (approximately 37 Gt of C per year) takes place in the euphotic layer (del Giorgio and Williams, 2005). An interesting issue is that such bacteria are also important sources of the superoxide radical anion (O 2 − • ) , the dismutation of which (2O 2 − • + 2H + → H 2 O 2 + O 2 ) consumes H + and could partially buffer at local scale the acidification that is connected to the degradation of OM (Mostofa et al., 2013b).
The biological transformation of DOM and PP is active constantly at the sea surface as well as in the subsurface/deeper water, whilst photoinduced degradation is merely active during daytime in the sea surface layer. Of course, such processes show variations associated with seasonal and annual changes in deep-sea geochemistry and biology, along with phenomena associated with ocean circulation (Asper et al., 1992;Thomas et al., 2004). The entire phytoplankton biomass of the global oceans is consumed every 2 to 6 days (Behrenfeld and Falkowski, 1997) and part of the carbon fixed by the autotrophs is actually respired in situ (Sarmento et al., 2010), also providing nutrients for the microbial food web (Behrenfeld et al., 2006;Sarmento et al., 2010). In some cases, the reprocessing of nutrients is involved in harmful algal blooms or eutrophication by enhanced photosynthesis in surface seawater (Sunda and Cai, 2012;Mostofa et al., 2013a).

Enhanced PP and respiration due to the effects of global warming and other processes: natural ocean acidification
Anthropogenic global warming could also enhance the natural acidification process. The dissolution of CO 2(g) and DIC released from PP and its subsequent respiration/degradation can be enhanced by the effects of GW (Behrenfeld et al., 2006;Cai et al., 2006;Kranz et al., 2009;Cai et al., 2011;Sunda and Cai, 2012;Mostofa et al., 2013a;Holding et al., 2015). GW is a key factor to increase water temperature (WT), which can affect the extent and the duration of the vertical stratification during the summer season. Furthermore, the prolonged exposure of the surface water layer to sunlight may cause photoinduced bleaching of sunlight-absorbing DOM, the so-called color dissolved organic matter (CDOM), thereby enhancing the water column transparency and modifying the depth of the mixing layer or euphotic zone (Behren-feld et al., 2006;Huisman et al., 2006). The increased stability of the water column may also enhance the photoinduced and biological mineralization of OM, due to the combination of higher temperature and of the longer exposure of the water surface layer to sunlight (Huisman et al., 2006;Vázquez-Domínguez et al., 2007). A further effect is the reduction of subsurface dissolved O 2 because of the decline of vertical winter mixing, which subsequently reduces the exchange of surface oxygenated water to the deeper layers ( Fig. 1). Increasing temperature increases the respiration rates in natural waters (Vázquez-Domínguez et al., 2007), and it affects phytoplankton metabolism nearly as significantly as nutrients and light do (Toseland et al., 2013). Various photoinduced and microbial products/compounds formed from DOM or PP (e.g., CO 2 , DIC, H 2 O 2 , NH + 4 , NO − 3 , PO 3− 4 , CH 4 , autochthonous DOM), the generation of which can be higher in stratified surface water as a consequence of GW, may enhance photosynthesis and, consequently, primary production as schematized in Fig. S1 in Supplement (Bates and Mathis, 2009;Cai et al., 2011;Mostofa et al., 2013a). Further details are reported in the Supplement.

Diurnal, abrupt and homogeneous pH changes in seawater
In some locations, the pH of the sea surface water gradually increases during the period before sunrise to noon and then decreases after sunset as a function of the solar irradiation flux (Fig. 3a, b) (Fransson et al., 2004b;Arakaki et al., 2005;Akhand et al., 2013). Furthermore, substantial fluctuations of the pH values during daytime are also observed ( Fig. 3a, b) (Fransson et al., 2004a;Arakaki et al., 2005;Clark et al., 2010). The magnitude of the diurnal pH variation can be substantial, ranging from ∼ 0.01 in waters with low biological activity to 1.60 in waters with high biological activity that are influenced by riverine inputs, particularly in coastal areas (Table S1 in Supplement). More specifically, pH has been observed to increase by 0.03-0.81 units in surface coastal seawater, from 0.26 to 1.60 in macroalgae, 0.01 to 0.75 in coral reefs, from 0.17 to > 1.00 in the seagrass community, from 0.03 to 1.59 in CO 2 venting sites, and from 0.04 to 0.10 in polar oceans (Table S1) (Semesi et al., 2009;Taguchi and Fujiwara, 2010;Hofmann et al., 2011). Diurnal pH changes in sea surface waters are apparently triggered by two phenomena. The first and key issue is the consumption or dissolution in seawater of CO 2 that is involved in primary production ( Fig. 2a, b) (Akhand et al., 2013;Zhai et al., 2014). Depending on the ratio between photosynthesis and respiration, diurnal fluctuations of pCO 2 are observed in seawater and the pCO 2 maxima correspond to pH minima and vice versa. In the case of Fig. 2a and b, the pH maxima are observed at noon or soon after noon; in other locations they may occur in different times of the day, but the anticorrelation between pH and pCO 2 is always observed. At the sea surface one may observe a diurnal decrease in pCO 2 with an increase in pH during the daytime or in the presence of sunlight (due to the prevalence of photosynthesis), along with an increase in pCO 2 with a decrease in pH at night when respiration prevails (Yates et al., 2007;Semesi et al., 2009). A second issue that might affect pH is the photoinduced generation of H 2 O 2 , primarily by dismutation of superoxide radical anion (2O 2 (Arakaki et al., 2005;Clark et al., 2010) and the subsequent production of the strong oxidant, hydroxyl radicals (HO • ) via photolysis or Fenton and photo-Fenton processes, which are responsible for the degradation of DOM and POM (Vione et al., 2006;Minakata et al., 2009). The linear correlation between pH / [H 2 O 2 ] and the UV (ultraviolet) intensity (Fig. 3c, d) can be elucidated by considering that both variables are directly influenced by solar irradiation.
Seawater pH is predominantly determined by the balance between consumption (photosynthesis) and release (respiration) of CO 2 as a consequence of the PP activity. In the reported cases the maximum consumption of dissolved CO 2 takes place at the same time as the maximum activity of the photo-stimulated biota. In addition, the positive correlation between [H 2 O 2 ] and UV intensity (Fig. 3c, d) is linked to the fact that the O 2 − • production rate overlaps with the maxi-mum of solar irradiation, because the biological and photochemical production of O 2 − • is activated by light absorption. The concentration of H 2 O 2 in sea surface water gradually increases during the period before sunrise to noon and then decreases after sunset as a function of solar irradiation (Fig. 3a, b). The amplitude of the H 2 O 2 diurnal cycle (the highest concentration at noontime minus concentration during the period before sunrise) ranged from 20 to 365 nM in coastal seas to marine bathing waters (Table S1). Both the O 2 − • production and its dismutation with formation of H 2 O 2 involve H + exchange and can consequently affect the ocean pH. O 2 − • is largely produced by the enzyme nicotinamide adenine dinucleotide phosphate (NADPH) oxidase through the synthesis of HO 2 • that is a weak acid (pKa = 4.88) (Bielski et al., 1985), which dissociates at the oceanic pH releasing H + ions according to the following reactions: The production and dismutation of O 2 − • is a H + -neutral process, but the fate of the superoxide anion is also a consequence of the redox state of the environment.  (Fig. 3a, b) would give an additional contribution to the daytime pH maxima and, as a consequence, could be a further actor in the definition of the daytime pH fluctuation.
Apart from the diurnal cycle, abrupt pH changes caused by both photoinduced and biological processes (overlapping to diurnal changes) have been observed in surface seawater and among the branches of Pocillopora colonies in the Great Barrier Reef (Gagliano et al., 2010), in the surface seawater of Okinawa Island (Fig. 3a, b) (Arakaki et al., 2005), in marine bathing waters (southern California) (Clark et al., 2010), in the North Sea (Blackford and Gilbert, 2007), in the North Pacific Ocean (Byrne et al., 2010), in the Chwaka Bay (Semesi et al., 2009), and in the northeast Atlantic (Findlay et al., 2014). Such rapid changes in pH are supposed to be a consequence of the primary production as well, although the details of the pH-modifying pathway(s) are still poorly understood. Proposals include several processes in which an intracellular microenvironment is produced, with very different pH values compared to the surrounding seawater, with possible release of intracellular material as a consequence of, e.g., cell lysis. Among these processes the main ones are (i) pH variation connected with aggregates present in photosynthetically active cells or inside colonies (Lubbers et al., 1990); (ii) polyanion-mediated formation of mineralpolymer composites inside alginate microgels or in the Golgi of coccolithophorid algae (Chin et al., 1998); (iii) processes occurring at the site of calcification such as conventional H + -channeling, Ca 2+ -H + exchanging ATPase, transcellular symporter, and co-transporter H + -solute shuttling (Ries, 2011); (iv) cellular extrusion of hydroxyl ions (OH − ) into the calcifying medium (Ries, 2011); and (v) CO 2 consumption via photosynthesis (Ries, 2011). Furthermore, the ability to up-regulate pH at the site of calcification can provide corals with enhanced resilience to the effects of ocean acidification (McCulloch et al., 2012). Increased pH during high primary productivity can be justified by the observation of a parallel increase in the δ 13 C values of POM, which may reflect a shift by phytoplankton from using CO 2 to using HCO − 3 for photosynthesis (Doi et al., 2006;Akhand et al., 2013). Therefore, uptake of HCO − 3 for phytoplankton photosynthesis at high pH might be the effect of its enhanced occurrence in seawater.
Homogeneous (longer-term and constant-rate) acidification in subsurface/deeper seawater is characteristically observed in oceans (Fig. 2c, Table S1; Feely et al., 2008;Byrne et al., 2010;Taguchi and Fujiwara, 2010;Cai et al., 2011;Zhai et al., 2012;Bates et al., 2013), estuaries , and experimentally in dark incubation (Lubbers et al., 1990). Such a homogeneous pH behavior is also followed in the subsurface water of a large freshwater lake ( Fig. S2a). At the beginning of the summer stratification period, pH in subsurface water (at depths of 40 and 80 m) gradually decreases whilst pH in the surface lake water (at depths of 2.5 and 10 m) increases, while DOC (Fig. S2b) and PP (chlorophyll a, Fig. S2c) also increase. Similar results, particularly monthly pH variations in surface and deeper seawater, are observed in the Seto Inland Sea during the summer stratification period and during convective mixing periods (Taguchi and Fujiwara, 2010). Homogeneous acidification can vary on a timescale of days to weeks or even months in a wide range of subsurface water at a specific depth ( Fig. S3; Byrne et al., 2010;Taguchi and Fujiwara, 2010). For example, pH was 7.4 at ∼ 2000-2500 m depth and 7.5 at ∼ 2500-3400 m depth along 25-55 • N in the North Pacific Ocean (Byrne et al., 2010), or pH was 7.0 at 80 m depth during the August-November period (Fig. S3). In the dark, pH decreases gradually inside colonies and also "nightly" decreases of pH occur (Lubbers et al., 1990). Such homogeneous acidification is primarily linked to the dissolution of CO 2 plus DIC originated from the biological degradation of sinking microorganisms (Bates and Mathis, 2009;Cai et al., 2011) and of the DOM originally produced by such organisms (Mostofa et al., 2013a). Enhanced acidification due to the biological degradation of OM can cause undersaturation of aragonite and calcite during the summer period in subsurface/deeper seawater in the Yellow Sea (Figs. 2b, 4) (Zhai et al., 2013), Gulf of Mexico Sunda and Cai, 2012), North Pacific Ocean (Byrne et al., 2010), Arctic Ocean (Bates et al., 2013), and Arctic shelves (Bates and Mathis, 2009).
The biological degradation processes are constantly occurring in subsurface/deeper seawater after the onset of early summer, and they continue during the summer stratification period for several months, until the start of winter vertical mixing (Fig. 1). The occurrence and importance of these processes is shown by the increasing trend in subsurface CO 2 followed by a similar decreasing trend of pH. Significant anticorrelation between the two parameters (r 2 = 0.5) has been observed in subsurface seawater (13-75 m depth) along 37 • 25 -39 • 67 N to 121 • 16 -124 • 10 E in the Yellow Sea (Fig. 4a). Furthermore, the same evidence was observed in the Seto Inland Sea (Taguchi and Fujiwara, 2010) and in the diurnal samples of the Luhuitou fringing reef (Sanya Bay) of the South China Sea . Strong anticorrelation between pCO 2[seawater] and dissolved O 2 (r 2 = 0.8; Fig. 4b) supports the production of CO 2 plus DIC from the biological respiration/degradation of DOM and PP by heterotrophic bacteria as discussed earlier. Such bacteria also produce the superoxide radical anion (O 2 • − )  that might be further involved in the processing/oxidation of DOM or PP by producing H 2 O 2 and consequently • OH via photolysis, photo-Fenton, or Fenton-like processes. Such trends of CO 2 (or DIC) vs. dissolved O 2 are also observed in California coastal waters (DeGrandpre et al., 1998), in the East China Sea ), in the South China Sea , and in Seto Inland Sea (Taguchi and Fujiwara, 2010). Biological respiration can be evidenced from an experiment conducted using subsurface water (37 m depth) collected from the East China Sea, where the decline in dissolved O 2 is significantly coupled with an increase of DIC production during a 60hour study period (Fig. 4c). The heterotrophic bacteria carry out the largest fraction of respiration (> 95 %) in the ocean (Del Giorgio and Duarte, 2002). This means that the heterotrophic community catabolizes an important percentage of the OM produced by the autotrophs (e.g., plants, algae, or bacteria) (Laws et al., 2000). Therefore, enhanced primary production or algal blooms in surface seawater and the subsequent sinking are the key processes for homogeneous acidification of the subsurface layer during the summer stratification period, through the degradation of sinking organic material. Finally, different regions or ecosystems are expected to give different responses to ocean acidification (Gattuso et al., 2015). Unfortunately, little has been documented on geographical comparisons of this aspect.

Possible forthcoming impacts on ocean acidification
An increase in world population (9 billion estimated in 2050) with increasing needs of energy, food, medicines, and habitats is one of the key issues (Mostofa et al., 2013a) that will probably contribute not only to the increase of atmospheric CO 2 , but also to the exacerbation of other factors that may also be related to ocean acidification. Such factors include enhanced photosynthesis (because of the release of terrestrial OM and nutrients from increased land use), the increment of OM and nutrients in wastewater, acid rain, and so on. The following issues can be foreseen in the next decades, unless remedial actions of some sort are taken: i. Long-term homogenous acidification in the deeper waters of both coastal and oligotrophic oceans, apparently caused by biological respiration of DOM and PP and their subsequent release of CO 2 or DIC, could have key impacts on marine organisms Bates et al., 2013;Zhai et al., 2013Zhai et al., , 2014Byrne et al., 2010;Mostofa et al., 2013a). Such homogenous effects of acidification are directly linked to the effects of GW that can enhance the surface water temperature. The consequence is an extension of the summer stratification period, which would determine acidification in deeper oceans.
ii. Coastal seawater, particularly in locations that are highly influenced by terrestrial river freshwater inputs, is at risk of substantial acidification, to a higher extent compared to the open oceans (Zhai et al., 2014;Thomas et al., 2009;Bate et al., 2013;Barton et al., 2014;Cai et al., 2011;Cai, 2011;Bauer et al., 2013;Hu et al., 2015). In fact, in addition to the dissolution of atmospheric CO 2 , coastal seawater would be subjected to acidification processes connected with eutrophication, acid rain, and pollution-affected respiration (Doney et al., 2007;Cai et al., 2011;Sunda and Cai, 2012;Zeng et al., 2015). Indeed, OM is substantially increasing in coastal oceans (Bauer et al., 2013). Furthermore, transport phenomena (e.g., oceanic pump) will gradually increase the level of nutrients, DOM and PP from coastal areas in the direction of the oligotrophic open ocean (Fig. 1) (Thomas et al., 2004). Therefore, additional acidification processes in the oligotrophic open ocean could be operational and more significant in the coming decades.
iii. Enhanced PP and respiration could increase pCO 2 in open-ocean water and decrease the ability of seawater itself to act as a sink of atmospheric CO 2 . The consequence will be an extension of the zones where seawater acts as a source of CO 2 , which has increased at an average rate of 1.5 µatm yr −1 in 1970-2007 (Takahashi et al., 2002(Takahashi et al., , 2009). In addition to the contribution to ocean acidification, the decreasing ability of seawater to act as CO 2 sink will also exacerbate the problems related to GW.
iv. The present sea-air fluxes of CO 2 (Takahashi et al., 2009) suggest that the equatorial oceans are prevailingly a CO 2 source to the atmosphere while the temperate ones are mainly a sink. Figure 5 reports the predicted pH changes by 2100 (Mora et al., 2013), showing that acidification is expected to affect all the world's oceans but that the most important effects are predicted for the elevated northern and southern latitudes. Such locations are presently the sites that mostly act as CO 2 sinks, because seawater pCO 2 is lower than the atmospheric one, and they will experience the most important pH-associated increase of seawater pCO 2 . It is thus likely that the global map of sea-air CO 2 fluxes will undergo important changes during the 21st century.

Impacts of acidification on marine organisms
Marine organisms at low and high latitudes do not respond uniformly to ocean acidification (Hendriks et al., 2010;Toseland et al., 2013), and the expected effects can thus be stimulative, inhibitive, or neutral (Anthony et al., 2008;Gao et al., 2012a;Hutchins et al., 2013). Considering the overall processes that are involved in ocean acidification (see Fig. 1), it can be assumed that marine organisms would face detrimental impacts under the following conditions: (i) they are peculiarly susceptible to pH changes with different timescales and particularly to acidification, which applies for instance to the majority of marine calcifiers; (ii) they live under hypoxia in long-term homogeneous acidified subsurface/deeper seawater, where they cannot carry out respiration and metabolism properly (this would happen during a stratification period of increasing duration due to GW, which can damage their natural growth and development); and (iii) they are subjected to death/damage in surface seawater by the action of algal toxins and pathogens (e.g., viruses, coliform bacteria, fungi), and/or to oxidative stress caused by reactive oxygen species (ROS) and increased water temperature. In many cases it is extremely difficult (or even next to impossible) to disentangle acidification from other processes that are taking place at the same time. Actually, the impacts of increasing acidification on marine organisms may derive from several processes that are closely interlinked: (i) acidification, (ii) synergistic effects of acidification and oxidative stress in surface seawater, (iii) low dissolved O 2 (hypoxia) and acidification in subsurface/deeper seawater, and (iv) stress by algal or red-tide toxins and pathogens.

Acidification
Impacts induced by seawater acidification or reduced seawater pH are recognized phenomena and they are discussed in many early reviews. However, seawater acidification or reduced seawater pH may produce undersaturation of aragonite and calcite, with the following effects in a variety of seawaters: (i) dissolution of biogenic shells or skeletons, mostly composed of CaCO 3 in the form of calcite or aragonite, of adult marine calcifiers such as corals (Kleypas et al., 1999;Erez et al., 2011;Pandolfi et al., 2011;Wittmann and Pörtner, 2013), crustose coralline algae (Anthony et al., 2008;Hall-Spencer et al., 2008), shellfish (Talmage and Gobler, 2010;Barton et al., 2012;Wittmann and Pörtner, 2013), marine plankton including foraminifera (de Moel et al., 2009;Moy et al., 2009) and coccolithophores (Riebesell et al., 2000;Beaufort et al., 2011), mollusks Wittmann and Pörtner, 2013), and echinoderms Wittmann and Pörtner, 2013); sedimentary CaCO 3 would be affected as well (Kleypas et al., 1999;Bates et al., 2013); (ii) inability to form new shells or skeletons of framework builders by larvae or juvenile calcifiers (e.g., the larval and juvenile stages or smaller individuals), particularly at the early development stages. The effect would be operational through the decline of calcification rates, which substantially decreases the growth and development of the organisms including corals (Kleypas et al., 1999;Anthony et al., 2008;Kroeker et al., 2013); and (iii) ocean acidification could indirectly enhance heterotrophic bacterial activities with increasing bacterial protein production and growth rate at elevated pCO 2 levels (Grossart et al., 2006;Endres et al., 2014;Baragi et al., 2015); higher bacterial abundance has been reported under high pCO 2 treatments (Endres et al., 2014;Tait et al., 2013), which could consequently accelerate respiration processes and increase the respiratory CO 2 production in the future ocean (Piontek et al., 2010). As discussed in Sect. 3, seawater pH varies in different timescales and shows short-term variations (e.g., minutes to hours: diurnal and abrupt) in upper surface seawater and long-term variations (e.g., weeks to several months: homogeneous) in subsurface and deeper seawater. Longterm homogenous acidification is apparently responsible for the majority of impacts on marine organisms. However, the impact on marine calcifiers of pH variations in different timescales, and most notably the diurnal ones, is presently poorly known and should be the focus of future research.

Synergistic effects of acidification and oxidative stress in surface seawater
The rapidly rising levels of atmospheric CO 2 will result in ocean warming in addition to lowering the seawater pH (Solomon et al., 2009;McCulloch et al., 2012). Marine calcifiers are for instance more sensitive to increased temperature under low pH conditions, because of the combination of two stressors (Wood et al., 2010;Pandolfi et al., 2011;Hiebenthal et al., 2013;Kroeker et al., 2013). The synergistic effects of ocean acidification and oxidative stress, elevated water temperature, or high irradiance, all connected with in-creasing CO 2 and GW, can affect marine ecosystems to a variable degree. In some cases the marine primary productivity is decreased (Boyce et al., 2010;Gao et al., 2012a), while in other cases the decrease is not so obvious, as tolerance to elevated CO 2 levels may be developed (Feng et al., 2009;Gao et al., 2009;Connell and Russell, 2010). However, even in the latter instances one may observe deep changes in species composition (Meron et al., 2011;Witt et al., 2011), and sometimes even an increase in coral productivity in experimental studies (Anthony et al., 2008). However, a drop in biodiversity is generally observed that is always to the detriment of calcifying organisms (Hall-Spencer et al., 2008;Connell and Russell, 2010). The observed negative effects include bleaching and productivity loss in coral reef builders (Hoegh-Guldberg et al., 2007;Anthony et al., 2008), high mortality and reduction of shell growth and shell breaking force (Hobbs and McDonald, 2010;Lischka et al., 2011;Hiebenthal et al., 2013), declining calcification and enhanced dissolution (Rodolfo-Metalpa et al., 2010), decline in abundance of the juveniles population (Lischka et al., 2011), and increased N : P ratios of eukaryotic phytoplankton (Toseland et al., 2013). The mechanism behind the oxidative stress at elevated WT or high irradiance is caused by a substantial generation of ROS, such as O 2 • − , H 2 O 2 , HO • , or 1 O 2 , in the surface water layer. The hydroxyl radical (HO • ), a strong oxidizing agent, is produced from either endogenic or exogenic H 2 O 2 through Fenton and photo-Fenton reactions in the presence of metal ions, and upon photolysis of NO − 2 or NO − 3 (Zepp et al., 1992;Mostofa et al., 2013c;Gligorovski et al., 2015). Inside organisms, HO • can damage the photosystem II activities and finally cause cell death (Blokhina et al., 2003;Mostofa et al., 2013c). H 2 O 2 concentration levels of approximately 100 nM (compared to up to 1700 nM values that have been detected in coastal waters) (Mostofa et al., 2013c) can cause oxidative stress to bacteria, as determined on the basis of increasing catalase enzyme concentration (Angel et al., 1999). H 2 O 2 can also reduce bacterial abundances by inducing elevated mortality in seawater (Clark et al., 2008). The oxidative stress that is related to the Fenton processes would even increase in acidified water, where the HO • yield is higher (Zepp et al., 1992). Interestingly and coherently with the expected HO • yield, the degree of oxidative stress in mollusks has been found to increase with decreasing pH (Tomanek et al., 2011), and the pH effect is further exacerbated by an increase in temperature (Matozzo et al., 2013). Furthermore, the synergistic effect of high H 2 O 2 combined with high seawater temperature resulted in a 134 % increase in coral metabolism/respiration rates (Higuchi et al., 2009).
Moreover, one should be sure not to only focus on the direct detrimental effects at the organism or single-species level: the negative impacts on the dynamics, structure, composition, and biodiversity of the coral reefs (Findlay et al., 2010;Wittmann and Pörtner, 2013), of other marine calcifiers (Feng et al., 2009;Wittmann and Pörtner, 2013), and of marine ecosystem processes would be linked to changes in species abundance, distribution, predator vulnerability, and competitive fitness (Hiscock et al., 2004;Feng et al., 2009;Gao et al., 2012b).

Synergistic effects of low dissolved O 2 (hypoxia) and acidification in subsurface/deeper seawater
Declining dissolved O 2 in deeper seawater would mostly be caused by reduced vertical mixing as a consequence of GW (Huisman et al., 2006;Keeling et al., 2010), which inhibits re-oxygenation while O 2 in deep water is consumed by biological respiration/degradation of sinking organisms and DOM ( Fig. 1) (Stramma et al., 2008;Cai et al., 2011;Sunda and Cai, 2012;Zhai et al., 2012;Mostofa et al., 2013a). The key reason for hypoxia is the long-term biological respiration/degradation of sinking OM in the absence of mixing, which is also a key pathway for acidification in sea subsurface water during the summer stratification period, as is discussed in earlier sections. The net decrease of dissolved O 2 in subsurface seawater in the Bohai Sea (China) between June and August 2011 was 34-62 % (see Fig. S3a), which would be the result of OM respiration during the summer stratification period. The hypoxia in subsurface water (40 and 70-80 m depths) (Fig. S3b) along with changes in pH, DOC, and PP or Chl a (Fig. S2) is linked with enhanced sinking of PP at the end of the summer stratification period. The connection between hypoxia (through respiration of OM) and acidification can be assessed by the positive correlation between pH and dissolved O 2 (Fig. S4), which shows that declining O 2 is directly associated with reduced pH in subsurface/deeper seawaters ( Fig. S4; Cai et al., 2011;Zhai et al., 2012Zhai et al., , 2013.The connection between hypoxia and acidification could be exacerbated, and long-term hypoxia could be induced, by two important factors, namely (i) the increase in algal blooms and the subsequently enhanced sinking of dead algae in subsurface/deeper seawater, and (ii) the effects of GW that would induce longer stratification periods as a consequence of a longer summer season, as previously discussed.
Recent study reveals that hypoxia and acidification have synergistic detrimental effects on living organisms, because they can separately affect growth and mortality and their combination can cause damage to organisms that are resistant to the separate stresses (Gobler et al., 2014). Moreover, acidification can cause an additional worsening of survival conditions in oxygen-poor waters, which are already made more acidic by the degradation of OM (Melzner et al., 2013). The overall consequences of hypoxia and acidification affect the natural growth and development of organisms (Boyce et al., 2010) and have implications for habitat loss (Keeling et al., 2010;Stramma et al., 2010), fish mortality (Hobbs and McDonald, 2010), nutrient cycling (Keeling et al., 2010;Toseland et al., 2013), carbon cycling (Keeling et al., 2010), ecosystem functioning (Diaz and Rosen-berg, 2008), and diversity, with possible changes of species composition in the bentho-pelagic communities (Diaz and Rosenberg, 2008;Stramma et al., 2010).

Stress caused by algal or red-tide toxins and pathogens
Ocean acidification or elevated CO 2 could increase the toxic algal blooms, involving for instance the diazotrophic cyanobacterium Nodularia spumigena Olli et al., 2015). They could also increase the accumulation of toxic phenolic compounds across trophic levels in phytoplankton grown under elevated CO 2 concentrations (Jin et al., 2015). Ocean acidification combined with nutrient limitation or temperature changes could considerably enhance the toxicity of some harmful groups (Fu et al., 2012). Correspondingly, harmful algal blooms are expected to increase in coastal waters because of increasing WT and eutrophication (Anderson et al., 2008;Glibert et al., 2010;Mostofa et al., 2013a), which would enhance net primary productivity that is the essential backdrop for the development of such blooms. The same phenomena are also involved in acidification; thus, it can be expected that more frequent algal blooms will take place along with ongoing acidification as an additional stress to marine organisms. Algal blooms and acidification could also be more closely linked Sunda and Cai, 2012), because the decline of marine algae with a calcareous skeleton could produce a selective advantage for harmful species (Irigoien et al., 2005;Mostofa et al., 2013a). Harmful algal blooms can produce algal toxins (e.g., microcystins) or red-tide toxins (e.g., brevetoxins) (Flewelling et al., 2005;Anderson et al., 2008), and the occurrence of pathogens (e.g., potentially hazardous fecal-oral viruses, coliform bacteria, parasites, or fungi) (Littler and Littler, 1995;Suttle, 2005) is also more likely in the presence of large phytoplankton cells and during algal blooms (Fuhrman, 1999;Suttle, 2005). Toxins and pathogens are a major cause of morbidity and mortality for marine organisms and they can affect humans as well (Harvell et al., 1999;Flewelling et al., 2005;Anderson et al., 2008). The most common toxins are microcystins, cyanotoxins (blue green algal toxins), okadaic acid (OA), dinophysis toxins (DTXs) and pectenotoxins (PTXs) produced by dinoflagellates (Takahashi et al., 2007), domoic acid (DA) produced by diatoms (Takahashi et al., 2007), and brevetoxins produced by the red-tide dinoflagellate Kareniabrevis (Flewelling et al., 2005;Anderson et al., 2008). Brevetoxins are potent neurotoxins that kill vast numbers of fish and even large marine mammals: for instance, 34 endangered Florida manatees (Trichechus manatuslatirostris) died in southwest Florida in the spring of 2002, and 107 bottlenose dolphins (Tursiops truncatus) died in waters off the Florida panhandle in the spring of 2004 as a consequence of exposure to brevetoxins (Flewelling et al., 2005). Furthermore, brevetoxins cause illness in humans who ingest contaminated filter-feeding shellfish or inhale toxic aerosols (Flewelling et al., 2005).
Ocean acidification/elevated CO 2 could indirectly affect bacterial activity and abundance (see Sect. 5.1;Grossart et al., 2006;Allgaier et al., 2008;Endres et al., 2014;Baragi et al., 2015;Witt et al., 2011;Tait et al., 2013). However, the abundance of different bacterial communities could respond differently (increase, remain unchanged or even decrease) under the effect of global warming (Allgaier et al., 2008;Witt et al., 2011;Baragi et al., 2015). However, acidification is also connected to an increase of pathogenic microbiota in corals (Meron et al., 2011). The latter effect is particularly alarming, because coral reefs are already directly endangered by acidification (inhibition of the calcification process, as already discussed) and GW. The reduction in reef-building coral species would be exacerbated by 18 coral diseases identified so far, with increasing prevalence and virulence in most marine taxa (Sutherland et al., 2004). The most concerning diseases are the black band disease (BBD), probably caused by several species of cyanobacteria including most notably Phormidium corallyticum (Rudnick and Ferrari, 1999); the coralline lethal orange disease (CLOD; a bacterial disease affecting coralline algae), which greatly impacts coral reefs and reef-building processes (Rudnick and Ferrari, 1999); a virulent disease known as white plague type II, which caused widespread mortality in most Caribbean coral species through physical contact with the macroalga Halimeda opuntia (Nugues et al., 2004); and, finally, corals bleaching or disease caused by the temperature-dependent bacteria Vibrio shiloi (Vidal-Dupiol et al., 2011). Further proposed pathogens for BBD, in addition to Phormidium corallyticum, include different genera of cyanobacteria, sulfate-reducing bacteria including Desulfovibrio spp., sulfide-oxidizing bacteria presumed to be Beggiatoa spp., several other heterotrophs, and marine fungi (Sekar et al., 2006). Any bacterial community shifted by elevated CO 2 could thus impact on other marine organisms. Finally, experimental research is warranted to find out links and mechanisms between harmful algal blooms and ocean acidification/elevated CO 2 .

Potential ecological and biogeochemical consequences arising from future ocean acidification
An overview of the potential upcoming ecological and biogeochemical consequences, linking different environmental drivers, processes, and cycles related to acidification in the future ocean is provided in Fig. 6. A recent study demonstrated that different types of tropical cyclones (hurricanes and typhoons) could increase significantly in oceans and on land over the 21st century (Lin and Emanuel, 2016). Extreme daily rainfall is thought to increase with temperature in some regions (Chan et al., 2016 and reference therein). Watersheds with high precipitation induce higher riverine discharge rates (Bauer et al., 2013) and, for instance, a single tropical storm can export approximately 43 % of the average annual riverine DOC (Yoon and Raymond, 2012). Similarly, on decadal timescales, single large, cyclone-induced floods can transport 77-92% of particulate organic carbon from mountainous regions (Hilton et al., 2008). Correspondingly, enhanced human activities due to increasing population will unquestionably jeopardize Earth's natural systems. Soil erosion is gradually intensified in regions where forests are converted into croplands (Ito, 2007), and humans have increased the sediment transport of global rivers through soil erosion by 2.3 ± 0.6 billion metric tons per year (Syvitski et al., 2005).
Potential changes in erosion rates in the midwestern United States under climate change is predicted and runoff could increase from +10 to +310 % (along with soil loss increase from +33 to +274%) in 2040-2059 relative to 1990-1999(O'Neal et al., 2005. The transfer of OM or organic carbon from the terrestrial soil to the oceans via erosion and riverine transport could significantly affect the coastal oceans (Hilton et al., 2008;Bauer et al., 2013;Galy et al., 2015). Particulate organic carbon (POC) export from the terrestrial biosphere into the oceans is mostly controlled by physical erosion, which is thus predicted to become the dominant long-term atmospheric CO 2 sink under a fourfold increase in global physical erosion rate at constant temperature (Galy et al., 2015). Such enhanced input of OM with rising temperatures under future global warming conditions will have a drastic impact on ocean acidification, which is concomitantly linked with other biogeochemical processes (Jin et al., 2015;Mora et al., 2013). Moreover, temperature regulates important abiotic and biotic processes that can alter water throughput, flow paths, dissolution rates, and watershed carbon stocks (Bauer et al., 2013) as well as the stratification period or euphotic zone ( Fig. 1; Mora et al., 2013;Huisman et al., 2006;Jöhnk et al., 2008). In addition, elevated temperature under global warming conditions could potentially enhance the proliferation of harmful Cyanobacteria in surface water (Paerl and Huisman, 2008;Jöhnk et al., 2008). The overall ecological and biogeochemical consequences of future ocean acidification under forthcoming global warming conditions in oceans could severely impact coastal seas, with a spreading of anoxic dead zones and a frequent occurrence of toxic dinoflagellate blooms (Jackson, 2008). Possible evolutions could involve expanding hypoxia in the deeper water layers Stramma et al., 2008); changes in food-web dynamics (Fabry et al., 2008;Wannicke et al., 2013); changes in the biogeochemical cycling dynamics of C, N, and P (Keeling et al., 2010;Wannicke et al., 2013;Toseland et al., 2013;Unger et al., 2013;Olli et al., 2015;Baragi et al., 2015); changes in metabolic pathways (Jin et al., 2015); increases in coral susceptibility to disease, pathogen abundance, and pathogen virulence (Maynard et al., 2015); negative consequences to mortality for various marine organisms, particularly for the shell-forming ones (Haigh et al., 2015;Doney et al., 2009); structural changes in Figure 6. An overview of the potential upcoming ecological and biogeochemical consequences, linking different environmental drivers, processes, and cycles related to acidification in the future ocean. phytoplankton communities (Dutkiewicz et al., 2015) and in some marine keystone species (Waldbusser et al., 2015;Barton et al., 2012); setting up of the Lilliput effect that causes organisms to evolve towards becoming smaller and exploit related physiological advantages (Garilli et al., 2015); increasing appearance of harmful marine species (e.g., Nodularia spumigena sp.; Olli et al., 2015;Jackson, 2008;Paerl and Huisman, 2008) and of toxic compounds (e.g., of the phenolic type; Jin et al., 2015); alteration of fish populations through habitat modification (Nagelkerken et al., 2016), as well as increasing global redistribution of marine biodiversity (Molinos et al., 2016). Finally, such ecological and biogeochemical changes in the oceans could have profound consequences for marine biodiversity, ecosystem services or processes, and seafood quality with deep implications for fishery industries in the upcoming decades Mora et al., 2013;Jin et al., 2015).

Perspectives
Ocean acidification is the outcome of a series of anthropic and natural processes that take place at the same time and are often interlinked. The dissolution of increasing atmospheric CO 2 into seawater obviously plays an important role (Pearson and Palmer, 2000;Feely et al., 2008;Beaufort et al., 2011), but there are also important contributions from the degradation of primary producers and DOM Sunda and Cai, 2012;Mostofa et al., 2013a). The latter process could be enhanced by an increased oceanic primary productivity (Feng et al., 2009;Sunda and Cai, 2012;Mostofa et al., 2013a), which is one of the possible consequences of global warming (see also Fig. S1) (Feng et al., 2009;Mostofa et al., 2013c). In coastal areas, acid rains and eutrophication caused by the runoff of terrestrial organic matter including DOM and nutrients (Sunda and Cai, 2012;Bauer et al., 2013), combined with microbial and photochemical degradation (Mostofa et al., 2013a), may be important or even the major causes of acidification. All the described processes would increase the supersaturation of the seawater CO 2 that correspondingly reduces the ability of seawater to take up atmospheric CO 2 , thereby extending the oceanic areas that constitute a source instead of a sink or carbon dioxide (presently, such areas are mostly concentrated in the equatorial zone) (see Fig. 1). An important issue is that acidification takes place at varying degrees, with different roles of the factors involved and with different impacts depending on the latitude, on the water temperature range as modified by the effects of GW, and on the distance from the coast (Vitousek et al., 1997;Copin-Montégut et al., 2004;Feely et al., 2008;Yamamoto-Kawai et al., 2009;Beaufort et al., 2011;Bates et al., 2013;Kroeker et al., 2013).
Acidification of seawater would be detrimental to marine organisms, and particularly to marine calcifiers for the longterm (e.g., homogeneous) acidification of subsurface/deeper seawater and possibly also the short-term (e.g., diurnal and abrupt) acidification of upper surface seawater. Therefore, living organisms will have to face multiple stresses at the same time, such as increasing occurrence of reactive oxygen species in the sea surface water, hypoxia in subsurface water, toxic algal blooms, and pathogens. Some of these additional stressors and/or their effects could be enhanced by acidification: the oxidative stress tends to be more severe at lower pH values and in the presence of diurnal and abrupt pH variations in surface water; the effects of hypoxia are exacerbated in long-term homogeneously acidified subsurface/deeper seawater, and a decline in marine calcifiers could provide a competitive advantage for toxic algae. Therefore, ocean acidification is expected to introduce deep changes in marine habitats, and food web processes.
Based on the discussed mechanisms, some of the possible actions that could be taken to limit the future impacts of acidification can be listed here (i) a reduction of anthropic CO 2 emissions to the atmosphere, which should be carried out in the wider context of fighting global warming and will face the same difficulties; (ii) the implementation of measures aimed at CO 2 capture, such as a worldwide increase in green plantation. In coastal areas, to limit the effects of acidification, some measures could be taken that are probably of somewhat easier implementation: (a) reduction of the inputs to seawater of OM from soil runoff, which implies the control and limitation of land use practices, of soil erosion and of wastewater discharges; (b) limitation of the primary productivity by controlling eutrophication, including the release of nutrients from agricultural activities; (c) removal of algae (e.g., by means of nets) during bloom periods, to avoid fertilization of seawater by the associated nutrients; (d) limitation of the emission of pollutants such as nitrogen and sulfur oxides to the atmosphere, as they are precursors of HNO 3 and H 2 SO 4 that are involved in acid rains. Finally, marine oceanographers should focus on how marine organisms are affected by short-term pH variations (e.g., diurnal and abrupt) in surface waters and by long-term (e.g., homo-geneous) ones in response to the effects of GW, which may further influence such pH variations.
The Supplement related to this article is available online at doi:10.5194/bg-13-1767-2016-supplement.